H8.5 top features

Top feature 0 in H8.5: (feature 13977)

TOP ACTIVATIONS
MAX = 3.533

responsible
Token responsible
Feature activation+1.050
journalists
Token journalists
Feature activation+2.571
who
Token who
Feature activation+0.195
covered
Token covered
Feature activation+1.175
the
Token the
Feature activation+0.577
story
Token story
Feature activation+3.533
âĢĶ
Token âĢĶ
Feature activation+0.000
that
Token that
Feature activation+0.000
"
Token "
Feature activation+0.000
we
Tokenwe
Feature activation+0.000
get
Token get
Feature activation+0.000
the
Token the
Feature activation+0.000
flashlight
Token flashlight
Feature activation+2.775
attachments
Token attachments
Feature activation+3.377
from
Token from
Feature activation+0.000
Blue
Token Blue
Feature activation+0.366
Team
Token Team
Feature activation+3.512
's
Token's
Feature activation+0.795
weapon
Token weapon
Feature activation+0.919
--
Token --
Feature activation+0.263
a
Token a
Feature activation+0.000
Gr
Token Gr
Feature activation+0.000
lit
Token lit
Feature activation+0.453
only
Token only
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
flashlight
Token flashlight
Feature activation+2.775
attachments
Token attachments
Feature activation+3.377
from
Token from
Feature activation+0.000
Blue
Token Blue
Feature activation+0.366
Team
Token Team
Feature activation+3.512
's
Token's
Feature activation+0.795
weapon
Token weapon
Feature activation+0.919
characters
Token characters
Feature activation+1.643
,
Token,
Feature activation+0.153
and
Token and
Feature activation+0.000
people
Token people
Feature activation+0.402
and
Token and
Feature activation+0.000
aliens
Token aliens
Feature activation+3.331
of
Token of
Feature activation+0.000
every
Token every
Feature activation+0.000
color
Token color
Feature activation+3.310
âĢĶ
TokenâĢĶ
Feature activation+0.000
had
Tokenhad
Feature activation+0.000
people
Token people
Feature activation+0.402
and
Token and
Feature activation+0.000
aliens
Token aliens
Feature activation+3.331
of
Token of
Feature activation+0.000
every
Token every
Feature activation+0.000
color
Token color
Feature activation+3.310
âĢĶ
TokenâĢĶ
Feature activation+0.000
had
Tokenhad
Feature activation+0.000
g
Token g
Feature activation+0.000
amed
Tokenamed
Feature activation+0.000
the
Token the
Feature activation+0.000
ana
Tokenana
Feature activation+0.817
Mend
Token Mend
Feature activation+0.291
o
Tokeno
Feature activation+0.000
za
Tokenza
Feature activation+1.882
from
Token from
Feature activation+0.128
Venezuela
Token Venezuela
Feature activation+3.177
-
Token -
Feature activation+0.000
visited
Token visited
Feature activation+0.000
the
Token the
Feature activation+0.000
facility
Token facility
Feature activation+0.000
last
Token last
Feature activation+0.000
two
Token two
Feature activation+0.000
minutes
Token minutes
Feature activation+2.523
of
Token of
Feature activation+0.000
every
Token every
Feature activation+0.305
NBA
Token NBA
Feature activation+0.358
game
Token game
Feature activation+3.055
--
Token --
Feature activation+0.000
the
Token the
Feature activation+0.000
league
Token league
Feature activation+0.000
says
Token says
Feature activation+0.000
that
Token that
Feature activation+0.000
to
Token to
Feature activation+0.000
substitute
Token substitute
Feature activation+0.408
N
Token N
Feature activation+0.000
aby
Tokenaby
Feature activation+0.000
Ke
Token Ke
Feature activation+0.000
ita
Tokenita
Feature activation+2.997
âĢ
TokenâĢ
Feature activation+0.255
Ļ
TokenĻ
Feature activation+2.796
s
Tokens
Feature activation+0.689
89
Token 89
Feature activation+0.277
-
Token-
Feature activation+0.000
and
Token and
Feature activation+0.000
no
Token no
Feature activation+0.000
buildings
Token buildings
Feature activation+1.676
were
Token were
Feature activation+0.000
present
Token present
Feature activation+2.941
there
Token there
Feature activation+2.942
at
Token at
Feature activation+0.377
that
Token that
Feature activation+0.920
time
Token time
Feature activation+2.569
âĢĵ
Token âĢĵ
Feature activation+0.000
many
Token many
Feature activation+0.000
shops
Token shops
Feature activation+0.252
and
Token and
Feature activation+0.000
no
Token no
Feature activation+0.000
buildings
Token buildings
Feature activation+1.676
were
Token were
Feature activation+0.000
present
Token present
Feature activation+2.941
there
Token there
Feature activation+2.942
at
Token at
Feature activation+0.377
that
Token that
Feature activation+0.920
time
Token time
Feature activation+2.569
âĢĵ
Token âĢĵ
Feature activation+0.000
same
Token same
Feature activation+0.110
day
Token day
Feature activation+0.000
the
Token the
Feature activation+0.000
report
Token report
Feature activation+0.000
was
Token was
Feature activation+0.196
launched
Token launched
Feature activation+2.932
in
Token in
Feature activation+0.083
Moscow
Token Moscow
Feature activation+2.716
âĢĵ
Token âĢĵ
Feature activation+0.000
a
Token a
Feature activation+0.000
high
Token high
Feature activation+0.000
Go
TokenGo
Feature activation+0.054
in
Token in
Feature activation+0.000
trouble
Token trouble
Feature activation+1.766
earlier
Token earlier
Feature activation+1.805
this
Token this
Feature activation+0.000
year
Token year
Feature activation+2.883
âĢĶ
Token âĢĶ
Feature activation+0.000
it
Token it
Feature activation+0.000
often
Token often
Feature activation+0.000
lays
Token lays
Feature activation+0.000
und
Token und
Feature activation+0.000
policy
Token policy
Feature activation+0.000
as
Token as
Feature activation+0.360
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.239
common
Tokencommon
Feature activation+0.338
sense
Token sense
Feature activation+2.874
,
Token,
Feature activation+0.584
âĢ
TokenâĢ
Feature activation+0.502
Ŀ
TokenĿ
Feature activation+1.930
despite
Token despite
Feature activation+0.000
the
Token the
Feature activation+0.000
formally
Token formally
Feature activation+0.000
sworn
Token sworn
Feature activation+0.000
in
Token in
Feature activation+0.292
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
role
Token role
Feature activation+2.867
this
Token this
Feature activation+0.000
morning
Token morning
Feature activation+2.721
âĢĶ
Token âĢĶ
Feature activation+0.000
has
Token has
Feature activation+0.000
rejected
Token rejected
Feature activation+0.000
âĢĵ
TokenâĢĵ
Feature activation+0.000
our
Tokenour
Feature activation+0.000
local
Token local
Feature activation+0.464
version
Token version
Feature activation+1.561
of
Token of
Feature activation+0.403
Obamacare
Token Obamacare
Feature activation+2.865
âĢĵ
TokenâĢĵ
Feature activation+0.000
cr
Tokencr
Feature activation+0.000
ashed
Tokenashed
Feature activation+0.000
yesterday
Token yesterday
Feature activation+0.000
.
Token.
Feature activation+0.000
the
Token the
Feature activation+0.000
school
Token school
Feature activation+0.000
gets
Token gets
Feature activation+0.950
its
Token its
Feature activation+0.546
public
Token public
Feature activation+0.000
funding
Token funding
Feature activation+2.864
-
Token -
Feature activation+0.000
is
Token is
Feature activation+0.000
investigating
Token investigating
Feature activation+0.000
alleged
Token alleged
Feature activation+0.000
financial
Token financial
Feature activation+0.000
it
Token it
Feature activation+0.000
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
foster
Token foster
Feature activation+0.000
investment
Token investment
Feature activation+1.719
"
Token"
Feature activation+2.855
âĢĶ
Token âĢĶ
Feature activation+0.000
was
Token was
Feature activation+0.000
struggling
Token struggling
Feature activation+0.000
to
Token to
Feature activation+0.000
find
Token find
Feature activation+0.000
points
Token points
Feature activation+0.000
âĢĵ
Token âĢĵ
Feature activation+0.000
allowing
Token allowing
Feature activation+0.075
just
Token just
Feature activation+0.749
seven
Token seven
Feature activation+1.091
goals
Token goals
Feature activation+2.813
âĢĵ
Token âĢĵ
Feature activation+0.000
the
Token the
Feature activation+0.000
offense
Token offense
Feature activation+0.000
has
Token has
Feature activation+0.000
been
Token been
Feature activation+0.000
like
Token like
Feature activation+0.000
Reading
Token Reading
Feature activation+1.195
,
Token,
Feature activation+0.000
Brighton
Token Brighton
Feature activation+0.000
and
Token and
Feature activation+0.000
Bristol
Token Bristol
Feature activation+2.785
âĢĶ
TokenâĢĶ
Feature activation+0.000
to
Tokento
Feature activation+0.000
commute
Token commute
Feature activation+0.000
to
Token to
Feature activation+0.000
high
Token high
Feature activation+0.000
who
Tokenwho
Feature activation+0.000
is
Token is
Feature activation+0.178
rumored
Token rumored
Feature activation+0.206
to
Token to
Feature activation+0.085
be
Token be
Feature activation+0.118
vegan
Token vegan
Feature activation+2.777
âĢĶ
TokenâĢĶ
Feature activation+0.000
pre
Tokenpre
Feature activation+0.000
viously
Tokenviously
Feature activation+0.000
lent
Token lent
Feature activation+0.000
his
Token his
Feature activation+0.000

Top DFA by src position
MAX = 4.112

's
Token's
Feature activation+0.000
Top resid features:
Phil
Token Phil
Feature activation+0.000
Top resid features:
An
Token An
Feature activation+0.000
Top resid features:
ast
Tokenast
Feature activation-0.000
Top resid features:
asia
Tokenasia
Feature activation-0.001
Top resid features:
âĢĶ
Token âĢĶ
Feature activation+4.112
Top resid features:
one
Token one
Feature activation+0.030
Top resid features:
of
Token of
Feature activation+0.006
Top resid features:
the
Token the
Feature activation+0.014
Top resid features:
few
Token few
Feature activation-0.002
Top resid features:
responsible
Token responsible
Feature activation+0.007
Top resid features:
I
TokenI
Feature activation-0.010
Top resid features:
research
Token research
Feature activation+0.004
Top resid features:
facility
Token facility
Feature activation-0.000
Top resid features:
Argent
Token Argent
Feature activation-0.006
Top resid features:
Moon
Token Moon
Feature activation-0.003
Top resid features:
--
Token --
Feature activation+3.905
Top resid features:
lit
Token lit
Feature activation+0.000
Top resid features:
only
Token only
Feature activation-0.007
Top resid features:
by
Token by
Feature activation-0.003
Top resid features:
the
Token the
Feature activation-0.003
Top resid features:
flashlight
Token flashlight
Feature activation+0.004
Top resid features:
I
TokenI
Feature activation-0.001
Top resid features:
research
Token research
Feature activation+0.001
Top resid features:
facility
Token facility
Feature activation-0.002
Top resid features:
Argent
Token Argent
Feature activation-0.000
Top resid features:
Moon
Token Moon
Feature activation-0.002
Top resid features:
--
Token --
Feature activation+3.830
Top resid features:
lit
Token lit
Feature activation+0.001
Top resid features:
only
Token only
Feature activation-0.023
Top resid features:
by
Token by
Feature activation+0.036
Top resid features:
the
Token the
Feature activation-0.002
Top resid features:
flashlight
Token flashlight
Feature activation+0.009
Top resid features:
multicultural
Token multicultural
Feature activation+0.006
Top resid features:
makeup
Token makeup
Feature activation+0.003
Top resid features:
of
Token of
Feature activation-0.001
Top resid features:
Hugo
Token Hugo
Feature activation-0.003
Top resid features:
winners
Token winners
Feature activation+0.004
Top resid features:
âĢĶ
TokenâĢĶ
Feature activation+3.581
Top resid features:
books
Tokenbooks
Feature activation+0.017
Top resid features:
featuring
Token featuring
Feature activation+0.041
Top resid features:
women
Token women
Feature activation-0.005
Top resid features:
,
Token,
Feature activation+0.004
Top resid features:
gay
Token gay
Feature activation-0.006
Top resid features:
multicultural
Token multicultural
Feature activation+0.001
Top resid features:
makeup
Token makeup
Feature activation+0.002
Top resid features:
of
Token of
Feature activation+0.001
Top resid features:
Hugo
Token Hugo
Feature activation-0.000
Top resid features:
winners
Token winners
Feature activation+0.004
Top resid features:
âĢĶ
TokenâĢĶ
Feature activation+3.623
Top resid features:
books
Tokenbooks
Feature activation+0.014
Top resid features:
featuring
Token featuring
Feature activation+0.026
Top resid features:
women
Token women
Feature activation-0.002
Top resid features:
,
Token,
Feature activation+0.002
Top resid features:
gay
Token gay
Feature activation-0.006
Top resid features:
Washington
Token Washington
Feature activation+0.001
Top resid features:
,
Token,
Feature activation+0.002
Top resid features:
the
Token the
Feature activation-0.001
Top resid features:
beauty
Token beauty
Feature activation-0.001
Top resid features:
queen
Token queen
Feature activation+0.012
Top resid features:
-
Token -
Feature activation+3.535
Top resid features:
who
Token who
Feature activation+0.051
Top resid features:
is
Token is
Feature activation+0.011
Top resid features:
also
Token also
Feature activation+0.005
Top resid features:
known
Token known
Feature activation+0.008
Top resid features:
as
Token as
Feature activation+0.023
Top resid features:
Last
TokenLast
Feature activation-0.012
Top resid features:
Two
Token Two
Feature activation-0.009
Top resid features:
Minute
Token Minute
Feature activation+0.001
Top resid features:
Report
Token Report
Feature activation+0.002
Top resid features:
"
Token"
Feature activation+0.010
Top resid features:
--
Token --
Feature activation+3.434
Top resid features:
its
Token its
Feature activation+0.039
Top resid features:
public
Token public
Feature activation+0.008
Top resid features:
review
Token review
Feature activation+0.032
Top resid features:
of
Token of
Feature activation+0.011
Top resid features:
offic
Token offic
Feature activation+0.006
Top resid features:
win
Token win
Feature activation-0.005
Top resid features:
over
Token over
Feature activation-0.005
Top resid features:
Bor
Token Bor
Feature activation-0.000
Top resid features:
ussia
Tokenussia
Feature activation-0.006
Top resid features:
Dortmund
Token Dortmund
Feature activation-0.001
Top resid features:
âĢĵ
Token âĢĵ
Feature activation+3.567
Top resid features:
thanks
Token thanks
Feature activation+0.019
Top resid features:
to
Token to
Feature activation+0.065
Top resid features:
substitute
Token substitute
Feature activation-0.013
Top resid features:
N
Token N
Feature activation-0.010
Top resid features:
aby
Tokenaby
Feature activation-0.000
Top resid features:
in
Token in
Feature activation-0.001
Top resid features:
Kat
Token Kat
Feature activation-0.005
Top resid features:
ima
Tokenima
Feature activation-0.004
Top resid features:
Mul
Token Mul
Feature activation-0.004
Top resid features:
ilo
Tokenilo
Feature activation-0.009
Top resid features:
âĢĵ
Token âĢĵ
Feature activation+3.793
Top resid features:
no
Token no
Feature activation-0.008
Top resid features:
shops
Token shops
Feature activation+0.013
Top resid features:
and
Token and
Feature activation-0.024
Top resid features:
no
Token no
Feature activation-0.026
Top resid features:
buildings
Token buildings
Feature activation+0.030
Top resid features:
in
Token in
Feature activation-0.005
Top resid features:
Kat
Token Kat
Feature activation-0.009
Top resid features:
ima
Tokenima
Feature activation-0.004
Top resid features:
Mul
Token Mul
Feature activation-0.003
Top resid features:
ilo
Tokenilo
Feature activation-0.013
Top resid features:
âĢĵ
Token âĢĵ
Feature activation+3.774
Top resid features:
no
Token no
Feature activation-0.004
Top resid features:
shops
Token shops
Feature activation+0.013
Top resid features:
and
Token and
Feature activation-0.029
Top resid features:
no
Token no
Feature activation-0.031
Top resid features:
buildings
Token buildings
Feature activation+0.012
Top resid features:
.
Token.
Feature activation-0.027
Top resid features:
Ċ
TokenĊ
Feature activation+0.005
Top resid features:
Ċ
TokenĊ
Feature activation-0.001
Top resid features:
On
TokenOn
Feature activation-0.010
Top resid features:
Wednesday
Token Wednesday
Feature activation-0.005
Top resid features:
âĢĵ
Token âĢĵ
Feature activation+3.283
Top resid features:
the
Token the
Feature activation+0.017
Top resid features:
same
Token same
Feature activation+0.035
Top resid features:
day
Token day
Feature activation+0.019
Top resid features:
the
Token the
Feature activation+0.015
Top resid features:
report
Token report
Feature activation+0.047
Top resid features:
there
Token there
Feature activation-0.003
Top resid features:
is
Token is
Feature activation-0.002
Top resid features:
a
Token a
Feature activation-0.003
Top resid features:
certificate
Token certificate
Feature activation+0.002
Top resid features:
hack
Token hack
Feature activation+0.000
Top resid features:
âĢĶ
Token âĢĶ
Feature activation+3.480
Top resid features:
like
Token like
Feature activation+0.016
Top resid features:
the
Token the
Feature activation-0.011
Top resid features:
one
Token one
Feature activation-0.012
Top resid features:
that
Token that
Feature activation+0.038
Top resid features:
got
Token got
Feature activation+0.016
Top resid features:
new
Token new
Feature activation+0.000
Top resid features:
voter
Token voter
Feature activation+0.015
Top resid features:
-
Token-
Feature activation+0.003
Top resid features:
ID
TokenID
Feature activation+0.030
Top resid features:
law
Token law
Feature activation+0.008
Top resid features:
âĢĵ
Token âĢĵ
Feature activation+3.145
Top resid features:
he
Token he
Feature activation+0.000
Top resid features:
described
Token described
Feature activation+0.006
Top resid features:
the
Token the
Feature activation+0.001
Top resid features:
policy
Token policy
Feature activation+0.013
Top resid features:
as
Token as
Feature activation+0.130
Top resid features:
,
Token,
Feature activation-0.001
Top resid features:
Communications
Token Communications
Feature activation-0.003
Top resid features:
Minister
Token Minister
Feature activation+0.010
Top resid features:
Malcolm
Token Malcolm
Feature activation-0.007
Top resid features:
Turnbull
Token Turnbull
Feature activation-0.020
Top resid features:
âĢĶ
Token âĢĶ
Feature activation+3.308
Top resid features:
formally
Token formally
Feature activation+0.052
Top resid features:
sworn
Token sworn
Feature activation-0.015
Top resid features:
in
Token in
Feature activation+0.010
Top resid features:
to
Token to
Feature activation+0.009
Top resid features:
the
Token the
Feature activation+0.022
Top resid features:
M
Token M
Feature activation-0.000
Top resid features:
NS
TokenNS
Feature activation+0.001
Top resid features:
ure
Tokenure
Feature activation+0.010
Top resid features:
web
Token web
Feature activation-0.010
Top resid features:
site
Token site
Feature activation-0.012
Top resid features:
âĢĵ
TokenâĢĵ
Feature activation+3.204
Top resid features:
our
Tokenour
Feature activation+0.047
Top resid features:
local
Token local
Feature activation+0.007
Top resid features:
version
Token version
Feature activation+0.028
Top resid features:
of
Token of
Feature activation+0.061
Top resid features:
Obamacare
Token Obamacare
Feature activation+0.054
Top resid features:
Ċ
TokenĊ
Feature activation-0.008
Top resid features:
The
TokenThe
Feature activation-0.003
Top resid features:
Education
Token Education
Feature activation+0.005
Top resid features:
Funding
Token Funding
Feature activation-0.009
Top resid features:
Agency
Token Agency
Feature activation+0.000
Top resid features:
-
Token -
Feature activation+3.463
Top resid features:
from
Token from
Feature activation-0.041
Top resid features:
which
Token which
Feature activation+0.051
Top resid features:
the
Token the
Feature activation-0.003
Top resid features:
school
Token school
Feature activation+0.006
Top resid features:
gets
Token gets
Feature activation+0.027
Top resid features:
February
Token February
Feature activation+0.001
Top resid features:
2016
Token 2016
Feature activation-0.005
Top resid features:
,
Token,
Feature activation-0.006
Top resid features:
the
Token the
Feature activation-0.007
Top resid features:
bank
Token bank
Feature activation+0.012
Top resid features:
âĢĶ
Token âĢĶ
Feature activation+3.512
Top resid features:
whose
Token whose
Feature activation+0.061
Top resid features:
stated
Token stated
Feature activation+0.020
Top resid features:
mission
Token mission
Feature activation+0.007
Top resid features:
is
Token is
Feature activation-0.024
Top resid features:
to
Token to
Feature activation-0.013
Top resid features:
their
Token their
Feature activation-0.002
Top resid features:
last
Token last
Feature activation-0.011
Top resid features:
30
Token 30
Feature activation+0.006
Top resid features:
possible
Token possible
Feature activation+0.005
Top resid features:
points
Token points
Feature activation+0.010
Top resid features:
âĢĵ
Token âĢĵ
Feature activation+3.430
Top resid features:
allowing
Token allowing
Feature activation-0.061
Top resid features:
just
Token just
Feature activation-0.013
Top resid features:
seven
Token seven
Feature activation-0.027
Top resid features:
goals
Token goals
Feature activation+0.020
Top resid features:
âĢĵ
Token âĢĵ
Feature activation+0.000
Top resid features:
many
Token many
Feature activation-0.002
Top resid features:
cities
Token cities
Feature activation+0.002
Top resid features:
in
Token in
Feature activation-0.001
Top resid features:
Southern
Token Southern
Feature activation-0.002
Top resid features:
England
Token England
Feature activation-0.018
Top resid features:
âĢĶ
TokenâĢĶ
Feature activation+3.561
Top resid features:
including
Tokenincluding
Feature activation+0.006
Top resid features:
far
Token far
Feature activation+0.002
Top resid features:
away
Token away
Feature activation+0.014
Top resid features:
cities
Token cities
Feature activation+0.021
Top resid features:
like
Token like
Feature activation-0.003
Top resid features:
ethical
Token ethical
Feature activation-0.002
Top resid features:
issues
Token issues
Feature activation+0.001
Top resid features:
.
Token.
Feature activation-0.023
Top resid features:
The
Token The
Feature activation-0.005
Top resid features:
actor
Token actor
Feature activation+0.021
Top resid features:
âĢĶ
TokenâĢĶ
Feature activation+3.120
Top resid features:
who
Tokenwho
Feature activation+0.067
Top resid features:
is
Token is
Feature activation+0.037
Top resid features:
rumored
Token rumored
Feature activation+0.020
Top resid features:
to
Token to
Feature activation+0.034
Top resid features:
be
Token be
Feature activation+0.070
Top resid features:

Decoder Weights Distribution

Head 0: 0.07

Head 1: 0.02

Head 2: 0.05

Head 3: 0.03

Head 4: 0.03

Head 5: 0.58

Head 6: 0.02

Head 7: 0.06

Head 8: 0.03

Head 9: 0.03

Head 10: 0.05

Head 11: 0.02

Positive logits

)—3.20

(@3.02

)--2.57

2.54

-2.48

2.46

*)2.45

2.41

"—2.20

--2.19

—"2.14

)-2.06

)*2.00

2.00

—-1.99

)1.98

,—1.98

1.95

--1.91

)1.89

Negative logits

.).-4.12

.''.-3.79

%.-3.50

>.-3.27

.]-3.19

].-3.16

.}-3.06

.'-3.05

.�-2.99

'.-2.94

$.-2.94

.''-2.94

().-2.91

.'"-2.91

`.-2.90

'.-2.84

''.-2.80

+.-2.80

.」-2.79

.",-2.77

INTERVAL 3.180 - 3.533
CONTAINS 0.001%

responsible
Token responsible
Feature activation+1.050
journalists
Token journalists
Feature activation+2.571
who
Token who
Feature activation+0.195
covered
Token covered
Feature activation+1.175
the
Token the
Feature activation+0.577
story
Token story
Feature activation+3.533
âĢĶ
Token âĢĶ
Feature activation+0.000
that
Token that
Feature activation+0.000
"
Token "
Feature activation+0.000
we
Tokenwe
Feature activation+0.000
get
Token get
Feature activation+0.000
characters
Token characters
Feature activation+1.643
,
Token,
Feature activation+0.153
and
Token and
Feature activation+0.000
people
Token people
Feature activation+0.402
and
Token and
Feature activation+0.000
aliens
Token aliens
Feature activation+3.331
of
Token of
Feature activation+0.000
every
Token every
Feature activation+0.000
color
Token color
Feature activation+3.310
âĢĶ
TokenâĢĶ
Feature activation+0.000
had
Tokenhad
Feature activation+0.000
people
Token people
Feature activation+0.402
and
Token and
Feature activation+0.000
aliens
Token aliens
Feature activation+3.331
of
Token of
Feature activation+0.000
every
Token every
Feature activation+0.000
color
Token color
Feature activation+3.310
âĢĶ
TokenâĢĶ
Feature activation+0.000
had
Tokenhad
Feature activation+0.000
g
Token g
Feature activation+0.000
amed
Tokenamed
Feature activation+0.000
the
Token the
Feature activation+0.000
lit
Token lit
Feature activation+0.453
only
Token only
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
flashlight
Token flashlight
Feature activation+2.775
attachments
Token attachments
Feature activation+3.377
from
Token from
Feature activation+0.000
Blue
Token Blue
Feature activation+0.366
Team
Token Team
Feature activation+3.512
's
Token's
Feature activation+0.795
weapon
Token weapon
Feature activation+0.919
the
Token the
Feature activation+0.000
flashlight
Token flashlight
Feature activation+2.775
attachments
Token attachments
Feature activation+3.377
from
Token from
Feature activation+0.000
Blue
Token Blue
Feature activation+0.366
Team
Token Team
Feature activation+3.512
's
Token's
Feature activation+0.795
weapon
Token weapon
Feature activation+0.919
--
Token --
Feature activation+0.263
a
Token a
Feature activation+0.000
Gr
Token Gr
Feature activation+0.000

INTERVAL 2.827 - 3.180
CONTAINS 0.001%

the
Token the
Feature activation+0.000
school
Token school
Feature activation+0.000
gets
Token gets
Feature activation+0.950
its
Token its
Feature activation+0.546
public
Token public
Feature activation+0.000
funding
Token funding
Feature activation+2.864
-
Token -
Feature activation+0.000
is
Token is
Feature activation+0.000
investigating
Token investigating
Feature activation+0.000
alleged
Token alleged
Feature activation+0.000
financial
Token financial
Feature activation+0.000
and
Token and
Feature activation+0.000
no
Token no
Feature activation+0.000
buildings
Token buildings
Feature activation+1.676
were
Token were
Feature activation+0.000
present
Token present
Feature activation+2.941
there
Token there
Feature activation+2.942
at
Token at
Feature activation+0.377
that
Token that
Feature activation+0.920
time
Token time
Feature activation+2.569
âĢĵ
Token âĢĵ
Feature activation+0.000
many
Token many
Feature activation+0.000
two
Token two
Feature activation+0.000
minutes
Token minutes
Feature activation+2.523
of
Token of
Feature activation+0.000
every
Token every
Feature activation+0.305
NBA
Token NBA
Feature activation+0.358
game
Token game
Feature activation+3.055
--
Token --
Feature activation+0.000
the
Token the
Feature activation+0.000
league
Token league
Feature activation+0.000
says
Token says
Feature activation+0.000
that
Token that
Feature activation+0.000
formally
Token formally
Feature activation+0.000
sworn
Token sworn
Feature activation+0.000
in
Token in
Feature activation+0.292
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
role
Token role
Feature activation+2.867
this
Token this
Feature activation+0.000
morning
Token morning
Feature activation+2.721
âĢĶ
Token âĢĶ
Feature activation+0.000
has
Token has
Feature activation+0.000
rejected
Token rejected
Feature activation+0.000
policy
Token policy
Feature activation+0.000
as
Token as
Feature activation+0.360
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.239
common
Tokencommon
Feature activation+0.338
sense
Token sense
Feature activation+2.874
,
Token,
Feature activation+0.584
âĢ
TokenâĢ
Feature activation+0.502
Ŀ
TokenĿ
Feature activation+1.930
despite
Token despite
Feature activation+0.000
the
Token the
Feature activation+0.000

INTERVAL 2.473 - 2.827
CONTAINS 0.003%

Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.376
ed
Token ed
Feature activation+0.000
gy
Tokengy
Feature activation+2.552
or
Token or
Feature activation+0.000
controversial
Token controversial
Feature activation+2.728
âĢĵ
TokenâĢĵ
Feature activation+0.000
min
Tokenmin
Feature activation+0.000
ors
Tokenors
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
180
Token 180
Feature activation+0.470
votes
Token votes
Feature activation+0.754
as
Token as
Feature activation+0.000
of
Token of
Feature activation+0.000
Saturday
Token Saturday
Feature activation+1.591
night
Token night
Feature activation+2.489
--
Token --
Feature activation+0.000
the
Token the
Feature activation+0.000
issue
Token issue
Feature activation+0.000
of
Token of
Feature activation+0.000
how
Token how
Feature activation+0.000
in
Token in
Feature activation+0.292
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
role
Token role
Feature activation+2.867
this
Token this
Feature activation+0.000
morning
Token morning
Feature activation+2.721
âĢĶ
Token âĢĶ
Feature activation+0.000
has
Token has
Feature activation+0.000
rejected
Token rejected
Feature activation+0.000
the
Token the
Feature activation+0.000
online
Token online
Feature activation+0.000
points
Token points
Feature activation+0.000
âĢĵ
Token âĢĵ
Feature activation+0.000
allowing
Token allowing
Feature activation+0.075
just
Token just
Feature activation+0.749
seven
Token seven
Feature activation+1.091
goals
Token goals
Feature activation+2.813
âĢĵ
Token âĢĵ
Feature activation+0.000
the
Token the
Feature activation+0.000
offense
Token offense
Feature activation+0.000
has
Token has
Feature activation+0.000
been
Token been
Feature activation+0.000
,
Token,
Feature activation+0.000
Wright
Token Wright
Feature activation+0.000
âĢĵ
Token âĢĵ
Feature activation+0.000
still
Token still
Feature activation+0.260
a
Token a
Feature activation+0.493
rookie
Token rookie
Feature activation+2.763
âĢĵ
Token âĢĵ
Feature activation+0.467
shouldn
Token shouldn
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
t
Tokent
Feature activation+0.000

INTERVAL 2.120 - 2.473
CONTAINS 0.004%

tens
Token tens
Feature activation+0.000
of
Token of
Feature activation+0.000
thousands
Token thousands
Feature activation+0.190
of
Token of
Feature activation+0.000
different
Token different
Feature activation+0.000
managers
Token managers
Feature activation+2.149
and
Token and
Feature activation+0.000
employed
Token employed
Feature activation+0.000
in
Token in
Feature activation+0.000
thousands
Token thousands
Feature activation+0.575
of
Token of
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+0.000
superst
Token superst
Feature activation+0.000
itious
Tokenitious
Feature activation+0.000
as
Token as
Feature activation+0.000
always
Token always
Feature activation+2.475
,
Token,
Feature activation+2.138
and
Token and
Feature activation+0.000
with
Token with
Feature activation+0.161
enough
Token enough
Feature activation+0.085
in
Token in
Feature activation+0.000
poll
Token poll
Feature activation+0.000
of
Token of
Feature activation+0.128
dropping
Token dropping
Feature activation+0.794
out
Token out
Feature activation+1.973
of
Token of
Feature activation+0.000
school
Token school
Feature activation+1.980
entirely
Token entirely
Feature activation+2.431
âĢĵ
Token âĢĵ
Feature activation+0.000
Morrow
Token Morrow
Feature activation+0.000
transferred
Token transferred
Feature activation+0.000
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
no
Token no
Feature activation+0.000
place
Token place
Feature activation+0.000
among
Token among
Feature activation+0.000
us
Token us
Feature activation+0.981
âĢ
TokenâĢ
Feature activation+1.414
Ŀ
TokenĿ
Feature activation+2.287
--
Token --
Feature activation+0.000
were
Token were
Feature activation+0.000
government
Token government
Feature activation+0.000
and
Token and
Feature activation+0.000
military
Token military
Feature activation+0.000
rare
Token rare
Feature activation+0.141
amendments
Token amendments
Feature activation+1.786
sadd
Token sadd
Feature activation+0.387
led
Tokenled
Feature activation+0.047
by
Token by
Feature activation+0.062
Congress
Token Congress
Feature activation+2.459
with
Token with
Feature activation+0.370
a
Token a
Feature activation+0.000
ratification
Token ratification
Feature activation+0.258
deadline
Token deadline
Feature activation+1.922
--
Token --
Feature activation+0.000

INTERVAL 1.767 - 2.120
CONTAINS 0.007%

use
Token use
Feature activation+0.401
-
Token-
Feature activation+0.000
of
Tokenof
Feature activation+0.000
-
Token-
Feature activation+0.000
force
Tokenforce
Feature activation+0.141
scenarios
Token scenarios
Feature activation+2.001
âĢĵ
Token âĢĵ
Feature activation+0.000
and
Token and
Feature activation+0.000
his
Token his
Feature activation+0.000
patrol
Token patrol
Feature activation+0.000
car
Token car
Feature activation+0.000
ball
Tokenball
Feature activation+0.000
,
Token,
Feature activation+0.000
football
Token football
Feature activation+0.000
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
wrestling
Token wrestling
Feature activation+1.938
âĢĶ
TokenâĢĶ
Feature activation+0.000
Z
TokenZ
Feature activation+0.000
immer
Tokenimmer
Feature activation+0.000
was
Token was
Feature activation+0.000
a
Token a
Feature activation+0.000
was
Token was
Feature activation+0.434
projected
Token projected
Feature activation+0.216
prior
Token prior
Feature activation+0.203
to
Token to
Feature activation+0.198
its
Token its
Feature activation+0.301
release
Token release
Feature activation+1.846
to
Token to
Feature activation+0.155
gross
Token gross
Feature activation+0.041
about
Token about
Feature activation+0.000
$
Token $
Feature activation+0.000
70
Token70
Feature activation+0.263
process
Token process
Feature activation+0.000
ions
Tokenions
Feature activation+0.000
to
Token to
Feature activation+0.000
riot
Token riot
Feature activation+0.094
ous
Tokenous
Feature activation+0.000
arson
Token arson
Feature activation+1.836
âĢĵ
Token âĢĵ
Feature activation+0.000
have
Token have
Feature activation+0.000
taken
Token taken
Feature activation+0.000
place
Token place
Feature activation+0.000
across
Token across
Feature activation+0.000
got
Token got
Feature activation+0.000
Go
Token Go
Feature activation+0.066
Go
TokenGo
Feature activation+0.054
in
Token in
Feature activation+0.000
trouble
Token trouble
Feature activation+1.766
earlier
Token earlier
Feature activation+1.805
this
Token this
Feature activation+0.000
year
Token year
Feature activation+2.883
âĢĶ
Token âĢĶ
Feature activation+0.000
it
Token it
Feature activation+0.000
often
Token often
Feature activation+0.000

INTERVAL 1.413 - 1.767
CONTAINS 0.011%

âĢĶ
TokenâĢĶ
Feature activation+0.000
and
Tokenand
Feature activation+0.000
Filip
Token Filip
Feature activation+0.000
inos
Tokeninos
Feature activation+0.042
by
Token by
Feature activation+0.119
extension
Token extension
Feature activation+1.731
âĢĶ
TokenâĢĶ
Feature activation+0.000
sn
Tokensn
Feature activation+0.000
ip
Tokenip
Feature activation+0.000
that
Token that
Feature activation+0.000
bit
Token bit
Feature activation+0.000
the
Token the
Feature activation+0.000
most
Token most
Feature activation+0.000
sought
Token sought
Feature activation+0.000
-
Token-
Feature activation+0.000
after
Tokenafter
Feature activation+0.000
part
Token part
Feature activation+1.503
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
product
Token product
Feature activation+1.975
âĢĶ
Token âĢĶ
Feature activation+0.000
stay
Token stay
Feature activation+0.000
the
Token the
Feature activation+0.000
elderly
Token elderly
Feature activation+0.487
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
the
Token the
Feature activation+0.000
disabled
Token disabled
Feature activation+1.713
âĢĶ
Token âĢĶ
Feature activation+0.000
who
Token who
Feature activation+0.000
rely
Token rely
Feature activation+0.000
on
Token on
Feature activation+0.000
Medicaid
Token Medicaid
Feature activation+0.000
but
Token but
Feature activation+0.100
are
Token are
Feature activation+0.198
not
Token not
Feature activation+0.069
as
Token as
Feature activation+0.000
regularly
Token regularly
Feature activation+0.253
available
Token available
Feature activation+1.628
a
Token a
Feature activation+0.000
source
Token source
Feature activation+1.001
of
Token of
Feature activation+0.000
data
Token data
Feature activation+1.293
)
Token)
Feature activation+0.000
A
TokenA
Feature activation+0.000
911
Token911
Feature activation+0.000
.
Token.
Feature activation+0.000
24
Token24
Feature activation+0.000
billion
Token billion
Feature activation+0.000
)
Token)
Feature activation+1.508
-
Token -
Feature activation+0.000
would
Token would
Feature activation+0.000
take
Token take
Feature activation+0.000
place
Token place
Feature activation+0.000
immediately
Token immediately
Feature activation+0.000

INTERVAL 1.060 - 1.413
CONTAINS 0.012%

Ļ
TokenĻ
Feature activation+0.000
m
Tokenm
Feature activation+0.000
not
Token not
Feature activation+0.139
passing
Token passing
Feature activation+0.719
judgment
Token judgment
Feature activation+1.717
on
Token on
Feature activation+1.162
âĢĵ
Token âĢĵ
Feature activation+0.000
is
Token is
Feature activation+0.000
that
Token that
Feature activation+0.000
they
Token they
Feature activation+0.000
need
Token need
Feature activation+0.000
âĢĶ
TokenâĢĶ
Feature activation+0.000
according
Tokenaccording
Feature activation+0.000
to
Token to
Feature activation+0.205
that
Token that
Feature activation+0.654
Movie
Token Movie
Feature activation+0.349
Web
TokenWeb
Feature activation+1.287
interview
Token interview
Feature activation+2.193
âĢĶ
TokenâĢĶ
Feature activation+0.000
suggest
Tokensuggest
Feature activation+0.000
ed
Tokened
Feature activation+0.000
buying
Token buying
Feature activation+0.000
ce
Tokence
Feature activation+0.000
g
Tokeng
Feature activation+0.000
after
Token after
Feature activation+0.000
her
Token her
Feature activation+0.143
red
Token red
Feature activation+0.000
card
Token card
Feature activation+1.222
was
Token was
Feature activation+0.000
overturned
Token overturned
Feature activation+2.017
âĢĶ
TokenâĢĶ
Feature activation+0.000
it
Tokenit
Feature activation+0.000
will
Token will
Feature activation+0.000
âĢĶ
TokenâĢĶ
Feature activation+0.000
where
Tokenwhere
Feature activation+0.000
soldiers
Token soldiers
Feature activation+0.000
died
Token died
Feature activation+0.260
for
Token for
Feature activation+0.000
nothing
Token nothing
Feature activation+1.266
more
Token more
Feature activation+0.184
than
Token than
Feature activation+0.000
violence
Token violence
Feature activation+2.129
for
Token for
Feature activation+0.000
the
Token the
Feature activation+0.000
âĢĶ
TokenâĢĶ
Feature activation+0.000
a
Tokena
Feature activation+0.040
self
Token self
Feature activation+0.068
-
Token-
Feature activation+0.000
described
Tokendescribed
Feature activation+0.081
socialist
Token socialist
Feature activation+1.220
âĢĶ
TokenâĢĶ
Feature activation+0.000
has
Tokenhas
Feature activation+0.000
a
Token a
Feature activation+0.000
respectable
Token respectable
Feature activation+0.000
amount
Token amount
Feature activation+0.000

INTERVAL 0.707 - 1.060
CONTAINS 0.021%

telling
Token telling
Feature activation+0.000
me
Token me
Feature activation+0.000
since
Token since
Feature activation+0.000
the
Token the
Feature activation+0.000
90
Token 90
Feature activation+0.000
s
Tokens
Feature activation+0.867
!
Token!
Feature activation+1.355
âĢĵ
Token âĢĵ
Feature activation+0.000
but
Token but
Feature activation+0.000
to
Token to
Feature activation+0.000
dismiss
Token dismiss
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+0.000
men
Token men
Feature activation+0.000
from
Token from
Feature activation+0.000
an
Token an
Feature activation+0.000
elite
Token elite
Feature activation+0.000
unit
Token unit
Feature activation+0.870
who
Token who
Feature activation+0.000
were
Token were
Feature activation+0.000
sent
Token sent
Feature activation+0.000
into
Token into
Feature activation+0.000
position
Token position
Feature activation+0.010
ek
Tokenek
Feature activation+0.000
stra
Tokenstra
Feature activation+0.000
(
Token (
Feature activation+0.000
R
TokenR
Feature activation+0.000
-
Token-
Feature activation+0.000
MI
TokenMI
Feature activation+0.933
),
Token),
Feature activation+0.000
then
Token then
Feature activation+0.000
ranking
Token ranking
Feature activation+0.000
minority
Token minority
Feature activation+0.000
member
Token member
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
world
Token world
Feature activation+0.000
and
Token and
Feature activation+0.000
âĢĵ
Token âĢĵ
Feature activation+0.000
perhaps
Token perhaps
Feature activation+0.881
as
Token as
Feature activation+0.000
importantly
Token importantly
Feature activation+1.202
âĢĵ
Token âĢĵ
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
as
Tokenas
Feature activation+0.000
it
Token it
Feature activation+0.000
has
Token has
Feature activation+0.000
lately
Token lately
Feature activation+0.131
been
Token been
Feature activation+0.417
labeled
Token labeled
Feature activation+0.730
by
Token by
Feature activation+0.000
virtue
Token virtue
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
dele
Token dele
Feature activation+0.000

INTERVAL 0.353 - 0.707
CONTAINS 0.046%

here
Token here
Feature activation+0.000
call
Token call
Feature activation+0.000
j
Token j
Feature activation+0.000
uh
Tokenuh
Feature activation+0.000
yo
Tokenyo
Feature activation+0.581
(
Token (
Feature activation+0.388
âĢ
TokenâĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
ice
Tokenice
Feature activation+0.000
trees
Token trees
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.378
s
Tokens
Feature activation+0.129
first
Token first
Feature activation+0.000
professional
Token professional
Feature activation+0.000
rugby
Token rugby
Feature activation+0.104
league
Token league
Feature activation+0.670
(
Token (
Feature activation+0.732
PRO
TokenPRO
Feature activation+0.000
)
Token)
Feature activation+0.445
locked
Token locked
Feature activation+0.000
in
Token in
Feature activation+0.004
methods
Token methods
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
âĢ¢
TokenâĢ¢
Feature activation+0.000
Experience
Token Experience
Feature activation+0.000
with
Token with
Feature activation+0.587
C
Token C
Feature activation+0.000
++
Token++
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
âĢ¢
TokenâĢ¢
Feature activation+0.000
son
Token son
Feature activation+0.560
John
Token John
Feature activation+0.203
who
Token who
Feature activation+0.000
fought
Token fought
Feature activation+0.000
in
Token in
Feature activation+0.000
Afghanistan
Token Afghanistan
Feature activation+0.411
and
Token and
Feature activation+0.000
Iraq
Token Iraq
Feature activation+0.480
and
Token and
Feature activation+0.000
who
Token who
Feature activation+0.000
thank
Token thank
Feature activation+0.000
the
Token the
Feature activation+0.000
how
Token how
Feature activation+0.000
.
Token.
Feature activation+0.000
Then
Token Then
Feature activation+0.000
âĢĵ
Token âĢĵ
Feature activation+0.000
maybe
Token maybe
Feature activation+0.517
âĢĵ
Token âĢĵ
Feature activation+0.000
it
Token it
Feature activation+0.000
will
Token will
Feature activation+0.000
tackle
Token tackle
Feature activation+0.000
the
Token the
Feature activation+0.000

INTERVAL 0.000 - 0.353
CONTAINS 99.894%

.
Token.
Feature activation+0.000
For
Token For
Feature activation+0.000
her
Token her
Feature activation+0.000
it
Token it
Feature activation+0.000
was
Token was
Feature activation+0.000
even
Token even
Feature activation+0.000
worse
Token worse
Feature activation+0.000
because
Token because
Feature activation+0.000
she
Token she
Feature activation+0.000
felt
Token felt
Feature activation+0.000
betrayed
Token betrayed
Feature activation+0.000
Ad
Token Ad
Feature activation+0.000
elson
Tokenelson
Feature activation+0.000
and
Token and
Feature activation+0.000
Goldman
Token Goldman
Feature activation+0.000
Sachs
Token Sachs
Feature activation+0.000
,
Token,
Feature activation+0.000
fell
Token fell
Feature activation+0.000
apart
Token apart
Feature activation+0.000
toward
Token toward
Feature activation+0.000
the
Token the
Feature activation+0.000
end
Token end
Feature activation+0.000
on
Token on
Feature activation+0.000
rubber
Token rubber
Feature activation+0.000
gloves
Token gloves
Feature activation+0.000
like
Token like
Feature activation+0.000
a
Token a
Feature activation+0.000
pro
Token pro
Feature activation+0.000
ct
Tokenct
Feature activation+0.000
ologist
Tokenologist
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
s
Tokens
Feature activation+0.000
disappearance
Token disappearance
Feature activation+0.000
was
Token was
Feature activation+0.000
launched
Token launched
Feature activation+0.000
last
Token last
Feature activation+0.000
June
Token June
Feature activation+0.000
,
Token,
Feature activation+0.000
with
Token with
Feature activation+0.000
Maj
Token Maj
Feature activation+0.000
.
Token.
Feature activation+0.000
Gen
Token Gen
Feature activation+0.000
well
Token well
Feature activation+0.000
liked
Token liked
Feature activation+0.000
within
Token within
Feature activation+0.000
the
Token the
Feature activation+0.000
organization
Token organization
Feature activation+0.000
and
Token and
Feature activation+0.000
has
Token has
Feature activation+0.000
forged
Token forged
Feature activation+0.000
close
Token close
Feature activation+0.000
relationships
Token relationships
Feature activation+0.000
with
Token with
Feature activation+0.000

BOTTOM ACTIVATIONS
MIN = 0.000

atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
&
Token &
Feature activation+0.000
Development
Token Development
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
&
Token &
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
rik
Tokenrik
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
rik
Tokenrik
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000

Top feature 1 in H8.5: (feature 22344)

TOP ACTIVATIONS
MAX = 8.046

Total
Token Total
Feature activation+4.181
losses
Token losses
Feature activation+4.642
of
Token of
Feature activation+5.594
the
Token the
Feature activation+7.468
National
Token National
Feature activation+3.380
ist
Tokenist
Feature activation+8.046
Air
Token Air
Feature activation+3.431
Force
Token Force
Feature activation+6.312
were
Token were
Feature activation+5.663
2
Token 2
Feature activation+2.970
,
Token,
Feature activation+7.051
the
Token the
Feature activation+0.000
center
Token center
Feature activation+0.000
,
Token,
Feature activation+0.423
whether
Token whether
Feature activation+2.007
in
Token in
Feature activation+2.654
Havana
Token Havana
Feature activation+7.578
or
Token or
Feature activation+1.684
Hiroshima
Token Hiroshima
Feature activation+1.080
,
Token,
Feature activation+0.782
too
Token too
Feature activation+1.610
often
Token often
Feature activation+3.686
shoot
Token shoot
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
âĢĵ
TokenâĢĵ
Feature activation+0.000
When
Token When
Feature activation+7.503
something
Token something
Feature activation+2.555
is
Token is
Feature activation+1.109
said
Token said
Feature activation+5.597
on
Token on
Feature activation+5.980
channel
Token channel
Feature activation+6.751
Ċ
TokenĊ
Feature activation+0.000
:
Token:
Feature activation+0.000
Total
Token Total
Feature activation+4.181
losses
Token losses
Feature activation+4.642
of
Token of
Feature activation+5.594
the
Token the
Feature activation+7.468
National
Token National
Feature activation+3.380
ist
Tokenist
Feature activation+8.046
Air
Token Air
Feature activation+3.431
Force
Token Force
Feature activation+6.312
were
Token were
Feature activation+5.663
POS
Token POS
Feature activation+6.435
RE
Token RE
Feature activation+0.860
P
TokenP
Feature activation+6.275
was
Token was
Feature activation+3.753
instrumental
Token instrumental
Feature activation+0.633
in
Token in
Feature activation+7.059
a
Token a
Feature activation+5.294
suicide
Token suicide
Feature activation+6.509
intervention
Token intervention
Feature activation+0.919
.
Token.
Feature activation+0.000
Knowing
Token Knowing
Feature activation+0.000
ist
Tokenist
Feature activation+8.046
Air
Token Air
Feature activation+3.431
Force
Token Force
Feature activation+6.312
were
Token were
Feature activation+5.663
2
Token 2
Feature activation+2.970
,
Token,
Feature activation+7.051
468
Token468
Feature activation+4.136
aircraft
Token aircraft
Feature activation+1.607
(
Token (
Feature activation+5.421
According
TokenAccording
Feature activation+3.179
to
Token to
Feature activation+5.840
is
Token is
Feature activation+1.109
said
Token said
Feature activation+5.597
on
Token on
Feature activation+5.980
channel
Token channel
Feature activation+6.751
1
Token 1
Feature activation+3.989
that
Token that
Feature activation+6.875
needs
Token needs
Feature activation+2.955
to
Token to
Feature activation+3.949
be
Token be
Feature activation+3.158
passed
Token passed
Feature activation+0.509
along
Token along
Feature activation+3.877
When
Token When
Feature activation+7.503
something
Token something
Feature activation+2.555
is
Token is
Feature activation+1.109
said
Token said
Feature activation+5.597
on
Token on
Feature activation+5.980
channel
Token channel
Feature activation+6.751
1
Token 1
Feature activation+3.989
that
Token that
Feature activation+6.875
needs
Token needs
Feature activation+2.955
to
Token to
Feature activation+3.949
be
Token be
Feature activation+3.158
to
Token to
Feature activation+2.866
the
Token the
Feature activation+2.186
Fed
Token Fed
Feature activation+0.000
,
Token,
Feature activation+0.512
saying
Token saying
Feature activation+2.403
that
Token that
Feature activation+6.730
for
Token for
Feature activation+6.658
now
Token now
Feature activation+2.530
,
Token,
Feature activation+4.771
competition
Token competition
Feature activation+3.043
of
Token of
Feature activation+3.719
.
Token.
Feature activation+2.024
Most
Token Most
Feature activation+2.153
people
Token people
Feature activation+1.631
are
Token are
Feature activation+2.123
used
Token used
Feature activation+0.235
to
Token to
Feature activation+6.666
practicing
Token practicing
Feature activation+3.963
the
Token the
Feature activation+4.361
art
Token art
Feature activation+3.361
of
Token of
Feature activation+5.239
sc
Token sc
Feature activation+0.847
the
Token the
Feature activation+2.186
Fed
Token Fed
Feature activation+0.000
,
Token,
Feature activation+0.512
saying
Token saying
Feature activation+2.403
that
Token that
Feature activation+6.730
for
Token for
Feature activation+6.658
now
Token now
Feature activation+2.530
,
Token,
Feature activation+4.771
competition
Token competition
Feature activation+3.043
of
Token of
Feature activation+3.719
currencies
Token currencies
Feature activation+0.973
Ċ
TokenĊ
Feature activation+0.000
This
TokenThis
Feature activation+0.119
is
Token is
Feature activation+2.560
the
Token the
Feature activation+0.742
ability
Token ability
Feature activation+1.438
of
Token of
Feature activation+6.607
Te
Token Te
Feature activation+0.000
emo
Tokenemo
Feature activation+3.048
in
Token in
Feature activation+0.741
which
Token which
Feature activation+0.109
Te
Token Te
Feature activation+0.000
Led
Token Led
Feature activation+2.099
ward
Tokenward
Feature activation+4.558
Spencer
Token Spencer
Feature activation+3.321
,
Token,
Feature activation+3.853
accused
Token accused
Feature activation+2.818
of
Token of
Feature activation+6.595
burg
Token burg
Feature activation+1.034
lar
Tokenlar
Feature activation+0.573
izing
Tokenizing
Feature activation+4.794
Texas
Token Texas
Feature activation+4.051
State
Token State
Feature activation+2.879
guess
Token guess
Feature activation+2.515
what
Token what
Feature activation+2.989
the
Token the
Feature activation+3.726
child
Token child
Feature activation+2.507
has
Token has
Feature activation+2.848
in
Token in
Feature activation+6.557
his
Token his
Feature activation+4.444
hand
Token hand
Feature activation+0.000
?
Token?
Feature activation+0.000
A
Token A
Feature activation+0.015
knife
Token knife
Feature activation+0.000
P
TokenP
Feature activation+6.275
was
Token was
Feature activation+3.753
instrumental
Token instrumental
Feature activation+0.633
in
Token in
Feature activation+7.059
a
Token a
Feature activation+5.294
suicide
Token suicide
Feature activation+6.509
intervention
Token intervention
Feature activation+0.919
.
Token.
Feature activation+0.000
Knowing
Token Knowing
Feature activation+0.000
that
Token that
Feature activation+1.087
we
Token we
Feature activation+0.000
email
Token email
Feature activation+2.319
from
Token from
Feature activation+3.345
someone
Token someone
Feature activation+0.490
detailing
Token detailing
Feature activation+0.529
how
Token how
Feature activation+3.256
POS
Token POS
Feature activation+6.435
RE
Token RE
Feature activation+0.860
P
TokenP
Feature activation+6.275
was
Token was
Feature activation+3.753
instrumental
Token instrumental
Feature activation+0.633
in
Token in
Feature activation+7.059
to
Token to
Feature activation+0.983
wear
Token wear
Feature activation+1.972
a
Token a
Feature activation+1.617
lot
Token lot
Feature activation+1.106
of
Token of
Feature activation+1.984
padding
Token padding
Feature activation+6.408
underneath
Token underneath
Feature activation+5.626
your
Token your
Feature activation+5.969
colorful
Token colorful
Feature activation+4.625
suit
Token suit
Feature activation+2.283
.
Token.
Feature activation+2.712
the
Token the
Feature activation+3.166
case
Token case
Feature activation+6.008
because
Token because
Feature activation+5.771
they
Token they
Feature activation+4.010
claimed
Token claimed
Feature activation+4.254
they
Token they
Feature activation+6.369
were
Token were
Feature activation+4.551
unable
Token unable
Feature activation+2.618
to
Token to
Feature activation+5.143
identify
Token identify
Feature activation+5.036
the
Token the
Feature activation+5.104
of
Token of
Feature activation+5.594
the
Token the
Feature activation+7.468
National
Token National
Feature activation+3.380
ist
Tokenist
Feature activation+8.046
Air
Token Air
Feature activation+3.431
Force
Token Force
Feature activation+6.312
were
Token were
Feature activation+5.663
2
Token 2
Feature activation+2.970
,
Token,
Feature activation+7.051
468
Token468
Feature activation+4.136
aircraft
Token aircraft
Feature activation+1.607
someone
Token someone
Feature activation+0.490
detailing
Token detailing
Feature activation+0.529
how
Token how
Feature activation+3.256
POS
Token POS
Feature activation+6.435
RE
Token RE
Feature activation+0.860
P
TokenP
Feature activation+6.275
was
Token was
Feature activation+3.753
instrumental
Token instrumental
Feature activation+0.633
in
Token in
Feature activation+7.059
a
Token a
Feature activation+5.294
suicide
Token suicide
Feature activation+6.509

Top DFA by src position
MAX = 8.477

Ċ
TokenĊ
Feature activation+0.013
Top resid features:
Ċ
TokenĊ
Feature activation+0.072
Top resid features:
:
Token:
Feature activation+1.170
Top resid features:
Total
Token Total
Feature activation+0.607
Top resid features:
losses
Token losses
Feature activation+1.405
Top resid features:
of
Token of
Feature activation+3.141
Top resid features:
the
Token the
Feature activation+2.021
Top resid features:
National
Token National
Feature activation+0.405
Top resid features:
ist
Tokenist
Feature activation+0.062
Top resid features:
Air
Token Air
Feature activation+0.000
Top resid features:
Force
Token Force
Feature activation+0.000
Top resid features:
him
Token him
Feature activation+0.008
Top resid features:
at
Token at
Feature activation+0.023
Top resid features:
the
Token the
Feature activation+0.024
Top resid features:
center
Token center
Feature activation+0.014
Top resid features:
,
Token,
Feature activation+0.123
Top resid features:
whether
Token whether
Feature activation+4.598
Top resid features:
in
Token in
Feature activation+2.983
Top resid features:
Havana
Token Havana
Feature activation+0.460
Top resid features:
or
Token or
Feature activation+0.000
Top resid features:
Hiroshima
Token Hiroshima
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.008
Top resid features:
shoot
Token shoot
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.003
Top resid features:
Ċ
TokenĊ
Feature activation+0.028
Top resid features:
Ċ
TokenĊ
Feature activation+0.042
Top resid features:
âĢĵ
TokenâĢĵ
Feature activation+8.477
Top resid features:
When
Token When
Feature activation+0.107
Top resid features:
something
Token something
Feature activation+0.000
Top resid features:
is
Token is
Feature activation+0.000
Top resid features:
said
Token said
Feature activation+0.000
Top resid features:
on
Token on
Feature activation+0.000
Top resid features:
Taiwanese
Token Taiwanese
Feature activation+0.008
Top resid features:
Sources
Token Sources
Feature activation+0.001
Top resid features:
).
Token).
Feature activation+0.016
Top resid features:
Ċ
TokenĊ
Feature activation+0.028
Top resid features:
Ċ
TokenĊ
Feature activation+0.040
Top resid features:
:
Token:
Feature activation+2.597
Top resid features:
Total
Token Total
Feature activation+2.022
Top resid features:
losses
Token losses
Feature activation+0.998
Top resid features:
of
Token of
Feature activation+1.877
Top resid features:
the
Token the
Feature activation+0.783
Top resid features:
National
Token National
Feature activation+0.000
Top resid features:
received
Token received
Feature activation+1.036
Top resid features:
an
Token an
Feature activation+0.582
Top resid features:
email
Token email
Feature activation+1.003
Top resid features:
from
Token from
Feature activation+0.182
Top resid features:
someone
Token someone
Feature activation+0.471
Top resid features:
detailing
Token detailing
Feature activation+2.309
Top resid features:
how
Token how
Feature activation+0.257
Top resid features:
POS
Token POS
Feature activation+0.012
Top resid features:
RE
Token RE
Feature activation+0.014
Top resid features:
P
TokenP
Feature activation+0.140
Top resid features:
was
Token was
Feature activation+0.381
Top resid features:
the
Token the
Feature activation+0.771
Top resid features:
National
Token National
Feature activation+0.052
Top resid features:
ist
Tokenist
Feature activation+0.083
Top resid features:
Air
Token Air
Feature activation+0.023
Top resid features:
Force
Token Force
Feature activation+0.091
Top resid features:
were
Token were
Feature activation+3.649
Top resid features:
2
Token 2
Feature activation+0.495
Top resid features:
,
Token,
Feature activation+0.146
Top resid features:
468
Token468
Feature activation+0.000
Top resid features:
aircraft
Token aircraft
Feature activation+0.000
Top resid features:
(
Token (
Feature activation+0.000
Top resid features:
something
Token something
Feature activation+0.575
Top resid features:
is
Token is
Feature activation+1.110
Top resid features:
said
Token said
Feature activation+0.384
Top resid features:
on
Token on
Feature activation+0.301
Top resid features:
channel
Token channel
Feature activation+0.346
Top resid features:
1
Token 1
Feature activation+3.572
Top resid features:
that
Token that
Feature activation+0.222
Top resid features:
needs
Token needs
Feature activation+0.000
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
be
Token be
Feature activation+0.000
Top resid features:
passed
Token passed
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.010
Top resid features:
shoot
Token shoot
Feature activation+0.001
Top resid features:
.
Token.
Feature activation+0.009
Top resid features:
Ċ
TokenĊ
Feature activation+0.017
Top resid features:
Ċ
TokenĊ
Feature activation+0.069
Top resid features:
âĢĵ
TokenâĢĵ
Feature activation+2.231
Top resid features:
When
Token When
Feature activation+1.431
Top resid features:
something
Token something
Feature activation+0.244
Top resid features:
is
Token is
Feature activation+0.935
Top resid features:
said
Token said
Feature activation+1.195
Top resid features:
on
Token on
Feature activation+1.457
Top resid features:
end
Token end
Feature activation+0.047
Top resid features:
to
Token to
Feature activation+0.003
Top resid features:
the
Token the
Feature activation+0.016
Top resid features:
Fed
Token Fed
Feature activation+0.341
Top resid features:
,
Token,
Feature activation+1.964
Top resid features:
saying
Token saying
Feature activation+4.227
Top resid features:
that
Token that
Feature activation+0.376
Top resid features:
for
Token for
Feature activation+0.000
Top resid features:
now
Token now
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
competition
Token competition
Feature activation+0.000
Top resid features:
seeing
Token seeing
Feature activation-0.024
Top resid features:
.
Token.
Feature activation+0.724
Top resid features:
Most
Token Most
Feature activation+0.796
Top resid features:
people
Token people
Feature activation+0.643
Top resid features:
are
Token are
Feature activation+2.410
Top resid features:
used
Token used
Feature activation+2.651
Top resid features:
to
Token to
Feature activation-0.058
Top resid features:
practicing
Token practicing
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
art
Token art
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
to
Token to
Feature activation+0.054
Top resid features:
the
Token the
Feature activation+0.031
Top resid features:
Fed
Token Fed
Feature activation+0.092
Top resid features:
,
Token,
Feature activation+0.125
Top resid features:
saying
Token saying
Feature activation+2.292
Top resid features:
that
Token that
Feature activation+4.614
Top resid features:
for
Token for
Feature activation-0.030
Top resid features:
now
Token now
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
competition
Token competition
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
damage
Token damage
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.001
Top resid features:
Ċ
TokenĊ
Feature activation+0.044
Top resid features:
Ċ
TokenĊ
Feature activation+0.237
Top resid features:
This
TokenThis
Feature activation+2.010
Top resid features:
is
Token is
Feature activation+2.744
Top resid features:
the
Token the
Feature activation+1.141
Top resid features:
ability
Token ability
Feature activation+1.426
Top resid features:
of
Token of
Feature activation-0.020
Top resid features:
Te
Token Te
Feature activation+0.000
Top resid features:
emo
Tokenemo
Feature activation+0.000
Top resid features:
rights
Token rights
Feature activation+0.037
Top resid features:
reserved
Token reserved
Feature activation+0.765
Top resid features:
Led
Token Led
Feature activation+0.096
Top resid features:
ward
Tokenward
Feature activation+0.037
Top resid features:
Spencer
Token Spencer
Feature activation+0.115
Top resid features:
,
Token,
Feature activation+2.060
Top resid features:
accused
Token accused
Feature activation+1.268
Top resid features:
of
Token of
Feature activation+0.050
Top resid features:
burg
Token burg
Feature activation+0.000
Top resid features:
lar
Tokenlar
Feature activation+0.000
Top resid features:
izing
Tokenizing
Feature activation+0.000
Top resid features:
to
Token to
Feature activation+0.297
Top resid features:
guess
Token guess
Feature activation+1.831
Top resid features:
what
Token what
Feature activation+0.912
Top resid features:
the
Token the
Feature activation+0.275
Top resid features:
child
Token child
Feature activation+1.173
Top resid features:
has
Token has
Feature activation+2.280
Top resid features:
in
Token in
Feature activation+0.038
Top resid features:
his
Token his
Feature activation+0.000
Top resid features:
hand
Token hand
Feature activation+0.000
Top resid features:
?
Token?
Feature activation+0.000
Top resid features:
A
Token A
Feature activation+0.000
Top resid features:
received
Token received
Feature activation+0.494
Top resid features:
an
Token an
Feature activation+0.383
Top resid features:
email
Token email
Feature activation+0.413
Top resid features:
from
Token from
Feature activation+0.205
Top resid features:
someone
Token someone
Feature activation+0.307
Top resid features:
detailing
Token detailing
Feature activation+1.774
Top resid features:
how
Token how
Feature activation+0.624
Top resid features:
POS
Token POS
Feature activation+0.039
Top resid features:
RE
Token RE
Feature activation+0.028
Top resid features:
P
TokenP
Feature activation+0.057
Top resid features:
was
Token was
Feature activation+0.584
Top resid features:
an
Token an
Feature activation+0.178
Top resid features:
email
Token email
Feature activation+0.099
Top resid features:
from
Token from
Feature activation+0.298
Top resid features:
someone
Token someone
Feature activation+0.129
Top resid features:
detailing
Token detailing
Feature activation+1.060
Top resid features:
how
Token how
Feature activation+4.814
Top resid features:
POS
Token POS
Feature activation+0.086
Top resid features:
RE
Token RE
Feature activation+0.000
Top resid features:
P
TokenP
Feature activation+0.000
Top resid features:
was
Token was
Feature activation+0.000
Top resid features:
instrumental
Token instrumental
Feature activation+0.000
Top resid features:
advice
Token advice
Feature activation+0.014
Top resid features:
is
Token is
Feature activation+0.156
Top resid features:
to
Token to
Feature activation+0.316
Top resid features:
wear
Token wear
Feature activation+0.543
Top resid features:
a
Token a
Feature activation+0.152
Top resid features:
lot
Token lot
Feature activation+3.042
Top resid features:
of
Token of
Feature activation+1.512
Top resid features:
padding
Token padding
Feature activation+1.506
Top resid features:
underneath
Token underneath
Feature activation+0.000
Top resid features:
your
Token your
Feature activation+0.000
Top resid features:
colorful
Token colorful
Feature activation+0.000
Top resid features:
closed
Token closed
Feature activation+1.556
Top resid features:
the
Token the
Feature activation+0.122
Top resid features:
case
Token case
Feature activation+0.357
Top resid features:
because
Token because
Feature activation+1.889
Top resid features:
they
Token they
Feature activation+0.219
Top resid features:
claimed
Token claimed
Feature activation+1.994
Top resid features:
they
Token they
Feature activation+0.190
Top resid features:
were
Token were
Feature activation+0.000
Top resid features:
unable
Token unable
Feature activation+0.000
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
identify
Token identify
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.010
Top resid features:
Ċ
TokenĊ
Feature activation+0.067
Top resid features:
:
Token:
Feature activation+1.085
Top resid features:
Total
Token Total
Feature activation+1.264
Top resid features:
losses
Token losses
Feature activation+1.603
Top resid features:
of
Token of
Feature activation+2.237
Top resid features:
the
Token the
Feature activation+0.714
Top resid features:
National
Token National
Feature activation+0.080
Top resid features:
ist
Tokenist
Feature activation+0.191
Top resid features:
Air
Token Air
Feature activation+0.104
Top resid features:
Force
Token Force
Feature activation+0.018
Top resid features:
an
Token an
Feature activation+0.288
Top resid features:
email
Token email
Feature activation+0.099
Top resid features:
from
Token from
Feature activation+0.129
Top resid features:
someone
Token someone
Feature activation+0.055
Top resid features:
detailing
Token detailing
Feature activation+2.460
Top resid features:
how
Token how
Feature activation+3.463
Top resid features:
POS
Token POS
Feature activation+0.078
Top resid features:
RE
Token RE
Feature activation+0.125
Top resid features:
P
TokenP
Feature activation+0.038
Top resid features:
was
Token was
Feature activation+0.000
Top resid features:
instrumental
Token instrumental
Feature activation+0.000
Top resid features:

Decoder Weights Distribution

Head 0: 0.05

Head 1: 0.05

Head 2: 0.06

Head 3: 0.02

Head 4: 0.04

Head 5: 0.57

Head 6: 0.03

Head 7: 0.05

Head 8: 0.03

Head 9: 0.03

Head 10: 0.05

Head 11: 0.02

Positive logits

aga2.20

Synopsis2.17

plet2.01

rex2.01

onica1.94

iona1.93

apon1.88

athing1.86

dyl1.85

athe1.85

ratch1.85

okemon1.84

̶1.83

edIn1.80

rawdownloadcloneembedreportprint1.80

athed1.80

ryn1.78

ocyte1.78

irit1.77

anch1.76

Negative logits

��-2.09

lihood-2.05

Canaver-2.00

wiser-1.99

lashes-1.89

brill-1.87

サーティワン-1.78

Andromeda-1.77

embr-1.77

Haku-1.73

trickle-1.69

praise-1.68

seams-1.66

Rockefeller-1.64

hairs-1.62

seiz-1.62

.).-1.62

Conc-1.61

misdem-1.61

Triumph-1.61

INTERVAL 7.242 - 8.046
CONTAINS 0.000%

shoot
Token shoot
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
âĢĵ
TokenâĢĵ
Feature activation+0.000
When
Token When
Feature activation+7.503
something
Token something
Feature activation+2.555
is
Token is
Feature activation+1.109
said
Token said
Feature activation+5.597
on
Token on
Feature activation+5.980
channel
Token channel
Feature activation+6.751
Ċ
TokenĊ
Feature activation+0.000
:
Token:
Feature activation+0.000
Total
Token Total
Feature activation+4.181
losses
Token losses
Feature activation+4.642
of
Token of
Feature activation+5.594
the
Token the
Feature activation+7.468
National
Token National
Feature activation+3.380
ist
Tokenist
Feature activation+8.046
Air
Token Air
Feature activation+3.431
Force
Token Force
Feature activation+6.312
were
Token were
Feature activation+5.663
Total
Token Total
Feature activation+4.181
losses
Token losses
Feature activation+4.642
of
Token of
Feature activation+5.594
the
Token the
Feature activation+7.468
National
Token National
Feature activation+3.380
ist
Tokenist
Feature activation+8.046
Air
Token Air
Feature activation+3.431
Force
Token Force
Feature activation+6.312
were
Token were
Feature activation+5.663
2
Token 2
Feature activation+2.970
,
Token,
Feature activation+7.051
the
Token the
Feature activation+0.000
center
Token center
Feature activation+0.000
,
Token,
Feature activation+0.423
whether
Token whether
Feature activation+2.007
in
Token in
Feature activation+2.654
Havana
Token Havana
Feature activation+7.578
or
Token or
Feature activation+1.684
Hiroshima
Token Hiroshima
Feature activation+1.080
,
Token,
Feature activation+0.782
too
Token too
Feature activation+1.610
often
Token often
Feature activation+3.686

INTERVAL 6.437 - 7.242
CONTAINS 0.001%

the
Token the
Feature activation+2.186
Fed
Token Fed
Feature activation+0.000
,
Token,
Feature activation+0.512
saying
Token saying
Feature activation+2.403
that
Token that
Feature activation+6.730
for
Token for
Feature activation+6.658
now
Token now
Feature activation+2.530
,
Token,
Feature activation+4.771
competition
Token competition
Feature activation+3.043
of
Token of
Feature activation+3.719
currencies
Token currencies
Feature activation+0.973
to
Token to
Feature activation+2.866
the
Token the
Feature activation+2.186
Fed
Token Fed
Feature activation+0.000
,
Token,
Feature activation+0.512
saying
Token saying
Feature activation+2.403
that
Token that
Feature activation+6.730
for
Token for
Feature activation+6.658
now
Token now
Feature activation+2.530
,
Token,
Feature activation+4.771
competition
Token competition
Feature activation+3.043
of
Token of
Feature activation+3.719
Ċ
TokenĊ
Feature activation+0.000
This
TokenThis
Feature activation+0.119
is
Token is
Feature activation+2.560
the
Token the
Feature activation+0.742
ability
Token ability
Feature activation+1.438
of
Token of
Feature activation+6.607
Te
Token Te
Feature activation+0.000
emo
Tokenemo
Feature activation+3.048
in
Token in
Feature activation+0.741
which
Token which
Feature activation+0.109
Te
Token Te
Feature activation+0.000
Led
Token Led
Feature activation+2.099
ward
Tokenward
Feature activation+4.558
Spencer
Token Spencer
Feature activation+3.321
,
Token,
Feature activation+3.853
accused
Token accused
Feature activation+2.818
of
Token of
Feature activation+6.595
burg
Token burg
Feature activation+1.034
lar
Tokenlar
Feature activation+0.573
izing
Tokenizing
Feature activation+4.794
Texas
Token Texas
Feature activation+4.051
State
Token State
Feature activation+2.879
When
Token When
Feature activation+7.503
something
Token something
Feature activation+2.555
is
Token is
Feature activation+1.109
said
Token said
Feature activation+5.597
on
Token on
Feature activation+5.980
channel
Token channel
Feature activation+6.751
1
Token 1
Feature activation+3.989
that
Token that
Feature activation+6.875
needs
Token needs
Feature activation+2.955
to
Token to
Feature activation+3.949
be
Token be
Feature activation+3.158

INTERVAL 5.632 - 6.437
CONTAINS 0.002%

and
Token and
Feature activation+1.269
author
Token author
Feature activation+2.605
of
Token of
Feature activation+3.232
historical
Token historical
Feature activation+3.331
fiction
Token fiction
Feature activation+2.735
that
Token that
Feature activation+5.750
compl
Token compl
Feature activation+1.745
icates
Tokenicates
Feature activation+4.230
the
Token the
Feature activation+3.269
past
Token past
Feature activation+2.226
by
Token by
Feature activation+4.577
you
Token you
Feature activation+0.484
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+1.148
re
Tokenre
Feature activation+5.017
taking
Token taking
Feature activation+1.999
off
Token off
Feature activation+5.784
your
Token your
Feature activation+4.485
walk
Token walk
Feature activation+0.530
ie
Tokenie
Feature activation+4.185
talk
Token talk
Feature activation+0.015
ie
Tokenie
Feature activation+3.173
t
Tokent
Feature activation+0.000
hed
Tokenhed
Feature activation+1.894
.
Token .
Feature activation+0.512
But
Token But
Feature activation+5.075
,
Token,
Feature activation+2.988
after
Token after
Feature activation+5.649
that
Token that
Feature activation+1.663
period
Token period
Feature activation+1.149
he
Token he
Feature activation+1.691
incredibly
Token incredibly
Feature activation+1.257
increases
Token increases
Feature activation+2.579
ulence
Tokenulence
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
red
Tokenred
Feature activation+0.000
uces
Tokenuces
Feature activation+5.649
digestive
Token digestive
Feature activation+3.191
irritation
Token irritation
Feature activation+3.628
and
Token and
Feature activation+3.360
the
Token the
Feature activation+1.016
risk
Token risk
Feature activation+2.169
art
Token art
Feature activation+3.361
of
Token of
Feature activation+5.239
sc
Token sc
Feature activation+0.847
rying
Tokenrying
Feature activation+4.870
with
Token with
Feature activation+5.186
a
Token a
Feature activation+5.802
mirror
Token mirror
Feature activation+4.800
or
Token or
Feature activation+4.631
a
Token a
Feature activation+4.166
crystal
Token crystal
Feature activation+2.407
ball
Token ball
Feature activation+0.000

INTERVAL 4.828 - 5.632
CONTAINS 0.006%

the
Token the
Feature activation+0.529
water
Token water
Feature activation+0.310
in
Token in
Feature activation+0.295
the
Token the
Feature activation+0.210
microwave
Token microwave
Feature activation+3.230
for
Token for
Feature activation+5.159
45
Token 45
Feature activation+2.700
seconds
Token seconds
Feature activation+0.000
.
Token.
Feature activation+1.458
Alternatively
Token Alternatively
Feature activation+1.976
use
Token use
Feature activation+2.310
,
Token,
Feature activation+4.771
competition
Token competition
Feature activation+3.043
of
Token of
Feature activation+3.719
currencies
Token currencies
Feature activation+0.973
âĢĵ
Token âĢĵ
Feature activation+2.129
and
Token and
Feature activation+5.389
banking
Token banking
Feature activation+3.914
structures
Token structures
Feature activation+0.000
âĢĵ
Token âĢĵ
Feature activation+3.394
should
Token should
Feature activation+4.043
be
Token be
Feature activation+2.435
little
Token little
Feature activation+2.513
steal
Token steal
Feature activation+1.662
t
Tokent
Feature activation+0.000
hed
Tokenhed
Feature activation+1.894
.
Token .
Feature activation+0.512
But
Token But
Feature activation+5.075
,
Token,
Feature activation+2.988
after
Token after
Feature activation+5.649
that
Token that
Feature activation+1.663
period
Token period
Feature activation+1.149
he
Token he
Feature activation+1.691
Ċ
TokenĊ
Feature activation+0.000
âĢĵ
TokenâĢĵ
Feature activation+0.000
When
Token When
Feature activation+7.503
something
Token something
Feature activation+2.555
is
Token is
Feature activation+1.109
said
Token said
Feature activation+5.597
on
Token on
Feature activation+5.980
channel
Token channel
Feature activation+6.751
1
Token 1
Feature activation+3.989
that
Token that
Feature activation+6.875
needs
Token needs
Feature activation+2.955
Ċ
TokenĊ
Feature activation+0.000
:
Token:
Feature activation+0.043
3
Token 3
Feature activation+3.592
R
TokenR
Feature activation+3.082
V
TokenV
Feature activation+2.021
X
TokenX
Feature activation+5.333
2
Token 2
Feature activation+1.436
.
Token.
Feature activation+3.624
9
Token9
Feature activation+2.483
.
Token.
Feature activation+3.925
2
Token2
Feature activation+1.749

INTERVAL 4.023 - 4.828
CONTAINS 0.013%

a
Token a
Feature activation+3.010
draft
Token draft
Feature activation+3.069
letter
Token letter
Feature activation+3.479
to
Token to
Feature activation+4.390
debate
Token debate
Feature activation+5.494
media
Token media
Feature activation+4.430
sponsors
Token sponsors
Feature activation+2.304
.
Token.
Feature activation+1.649
Leading
Token Leading
Feature activation+3.766
Republican
Token Republican
Feature activation+5.426
presidential
Token presidential
Feature activation+3.398
lot
Token lot
Feature activation+1.106
of
Token of
Feature activation+1.984
padding
Token padding
Feature activation+6.408
underneath
Token underneath
Feature activation+5.626
your
Token your
Feature activation+5.969
colorful
Token colorful
Feature activation+4.625
suit
Token suit
Feature activation+2.283
.
Token.
Feature activation+2.712
Then
Token Then
Feature activation+3.436
,
Token,
Feature activation+3.727
buy
Token buy
Feature activation+3.806
and
Token and
Feature activation+2.053
children
Token children
Feature activation+3.037
cut
Token cut
Feature activation+3.523
the
Token the
Feature activation+3.011
ribbon
Token ribbon
Feature activation+3.722
to
Token to
Feature activation+4.096
officially
Token officially
Feature activation+5.186
open
Token open
Feature activation+4.371
the
Token the
Feature activation+2.060
new
Token new
Feature activation+2.095
Ronald
Token Ronald
Feature activation+0.745
passed
Token passed
Feature activation+0.509
along
Token along
Feature activation+3.877
to
Token to
Feature activation+4.501
other
Token other
Feature activation+2.549
channels
Token channels
Feature activation+2.611
.
Token.
Feature activation+4.128
This
Token This
Feature activation+2.069
is
Token is
Feature activation+2.545
usually
Token usually
Feature activation+1.457
assigned
Token assigned
Feature activation+0.830
to
Token to
Feature activation+3.143
,
Token,
Feature activation+3.853
accused
Token accused
Feature activation+2.818
of
Token of
Feature activation+6.595
burg
Token burg
Feature activation+1.034
lar
Tokenlar
Feature activation+0.573
izing
Tokenizing
Feature activation+4.794
Texas
Token Texas
Feature activation+4.051
State
Token State
Feature activation+2.879
dorm
Token dorm
Feature activation+2.466
(
Token (
Feature activation+2.796
San
TokenSan
Feature activation+2.399

INTERVAL 3.219 - 4.023
CONTAINS 0.021%

historian
Token historian
Feature activation+3.009
,
Token,
Feature activation+3.228
and
Token and
Feature activation+1.269
author
Token author
Feature activation+2.605
of
Token of
Feature activation+3.232
historical
Token historical
Feature activation+3.331
fiction
Token fiction
Feature activation+2.735
that
Token that
Feature activation+5.750
compl
Token compl
Feature activation+1.745
icates
Tokenicates
Feature activation+4.230
the
Token the
Feature activation+3.269
also
Token also
Feature activation+2.861
tells
Token tells
Feature activation+4.042
stories
Token stories
Feature activation+2.912
about
Token about
Feature activation+5.186
many
Token many
Feature activation+4.344
of
Token of
Feature activation+4.007
the
Token the
Feature activation+3.342
saints
Token saints
Feature activation+0.000
Appears
Token Appears
Feature activation+0.166
in
Token in
Feature activation+1.677
the
Token the
Feature activation+1.257
neighborhood
Token neighborhood
Feature activation+3.611
on
Token on
Feature activation+1.058
suspicion
Token suspicion
Feature activation+3.757
of
Token of
Feature activation+2.325
resisting
Token resisting
Feature activation+2.723
arrest
Token arrest
Feature activation+3.774
and
Token and
Feature activation+2.591
weapons
Token weapons
Feature activation+2.533
violations
Token violations
Feature activation+0.000
charges
Token charges
Feature activation+0.000
.
Token.
Feature activation+0.000
turrets
Token turrets
Feature activation+2.919
or
Token or
Feature activation+1.066
target
Token target
Feature activation+3.076
weak
Token weak
Feature activation+0.000
-
Token-
Feature activation+0.000
points
Tokenpoints
Feature activation+3.284
to
Token to
Feature activation+3.884
blow
Token blow
Feature activation+1.317
out
Token out
Feature activation+3.871
entire
Token entire
Feature activation+1.996
sections
Token sections
Feature activation+2.049
allegedly
Token allegedly
Feature activation+2.676
masturb
Token masturb
Feature activation+0.000
ating
Tokenating
Feature activation+2.424
while
Token while
Feature activation+3.094
ferry
Token ferry
Feature activation+0.000
ing
Tokening
Feature activation+3.244
a
Token a
Feature activation+3.175
25
Token 25
Feature activation+1.621
-
Token-
Feature activation+0.773
year
Tokenyear
Feature activation+1.479
-
Token-
Feature activation+0.000

INTERVAL 2.414 - 3.219
CONTAINS 0.028%

a
Token a
Feature activation+0.809
new
Token new
Feature activation+0.957
plan
Token plan
Feature activation+1.698
that
Token that
Feature activation+3.476
will
Token will
Feature activation+3.736
actually
Token actually
Feature activation+3.071
create
Token create
Feature activation+4.293
jobs
Token jobs
Feature activation+1.693
?"
Token?"
Feature activation+0.000
Ambrose
Token Ambrose
Feature activation+0.000
asked
Token asked
Feature activation+0.000
that
Token that
Feature activation+2.995
the
Token the
Feature activation+3.197
prime
Token prime
Feature activation+1.077
minister
Token minister
Feature activation+3.678
of
Token of
Feature activation+4.126
a
Token a
Feature activation+2.647
country
Token country
Feature activation+3.230
would
Token would
Feature activation+3.500
be
Token be
Feature activation+3.233
owing
Token owing
Feature activation+2.070
something
Token something
Feature activation+2.163
,
Token,
Feature activation+1.081
George
Token George
Feature activation+0.635
Le
Token Le
Feature activation+0.555
pp
Tokenpp
Feature activation+1.663
,
Token,
Feature activation+1.895
Mark
Token Mark
Feature activation+2.494
Bram
Token Bram
Feature activation+0.000
ley
Tokenley
Feature activation+1.899
,
Token,
Feature activation+1.982
Marc
Token Marc
Feature activation+0.978
Urban
Token Urban
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
|
Token|
Feature activation+0.000
If
Token If
Feature activation+0.305
you
Token you
Feature activation+1.489
are
Token are
Feature activation+2.985
using
Token using
Feature activation+2.906
color
Token color
Feature activation+2.277
,
Token,
Feature activation+2.983
add
Token add
Feature activation+3.879
a
Token a
Feature activation+4.361
dash
Token dash
Feature activation+2.992
6
Token 6
Feature activation+0.000
,
Token,
Feature activation+0.000
2010
Token 2010
Feature activation+0.221
at
Token at
Feature activation+1.189
10
Token 10
Feature activation+2.137
:
Token:
Feature activation+2.574
38
Token38
Feature activation+2.720
AM
TokenAM
Feature activation+1.135
by
Token by
Feature activation+0.203
Tyler
Token Tyler
Feature activation+0.059
Thompson
Token Thompson
Feature activation+0.572

INTERVAL 1.609 - 2.414
CONTAINS 0.040%

port
Token port
Feature activation+0.000
to
Token to
Feature activation+0.746
make
Token make
Feature activation+0.654
this
Token this
Feature activation+0.664
,
Token,
Feature activation+0.349
this
Token this
Feature activation+2.196
is
Token is
Feature activation+3.424
really
Token really
Feature activation+3.723
awesome
Token awesome
Feature activation+1.108
that
Token that
Feature activation+0.000
Cle
Token Cle
Feature activation+0.000
for
Token for
Feature activation+5.159
45
Token 45
Feature activation+2.700
seconds
Token seconds
Feature activation+0.000
.
Token.
Feature activation+1.458
Alternatively
Token Alternatively
Feature activation+1.976
use
Token use
Feature activation+2.310
very
Token very
Feature activation+3.109
hot
Token hot
Feature activation+4.115
tap
Token tap
Feature activation+0.506
water
Token water
Feature activation+0.363
and
Token and
Feature activation+1.301
ignored
Token ignored
Feature activation+4.324
by
Token by
Feature activation+4.493
the
Token the
Feature activation+2.773
server
Token server
Feature activation+1.626
configuration
Token configuration
Feature activation+2.821
because
Token because
Feature activation+2.019
it
Token it
Feature activation+2.379
was
Token was
Feature activation+2.442
being
Token being
Feature activation+0.803
looked
Token looked
Feature activation+0.163
at
Token at
Feature activation+3.241
ating
Tokenating
Feature activation+2.424
while
Token while
Feature activation+3.094
ferry
Token ferry
Feature activation+0.000
ing
Tokening
Feature activation+3.244
a
Token a
Feature activation+3.175
25
Token 25
Feature activation+1.621
-
Token-
Feature activation+0.773
year
Tokenyear
Feature activation+1.479
-
Token-
Feature activation+0.000
old
Tokenold
Feature activation+2.482
woman
Token woman
Feature activation+0.000
to
Token to
Feature activation+0.729
help
Token help
Feature activation+0.000
us
Token us
Feature activation+0.549
continue
Token continue
Feature activation+0.000
to
Token to
Feature activation+0.593
write
Token write
Feature activation+1.769
about
Token about
Feature activation+0.871
the
Token the
Feature activation+0.000
issues
Token issues
Feature activation+0.702
that
Token that
Feature activation+0.701
matter
Token matter
Feature activation+0.634

INTERVAL 0.805 - 1.609
CONTAINS 0.073%

violence
Token violence
Feature activation+0.070
complaint
Token complaint
Feature activation+0.000
by
Token by
Feature activation+0.000
his
Token his
Feature activation+0.000
ex
Token ex
Feature activation+0.045
-
Token-
Feature activation+0.988
girlfriend
Tokengirlfriend
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
this
Token this
Feature activation+0.664
,
Token,
Feature activation+0.349
this
Token this
Feature activation+2.196
is
Token is
Feature activation+3.424
really
Token really
Feature activation+3.723
awesome
Token awesome
Feature activation+1.108
that
Token that
Feature activation+0.000
Cle
Token Cle
Feature activation+0.000
my
Tokenmy
Feature activation+2.936
is
Token is
Feature activation+5.007
now
Token now
Feature activation+5.231
good
Token good
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+0.000
Tre
Token Tre
Feature activation+0.000
on
Tokenon
Feature activation+0.000
Harris
Token Harris
Feature activation+0.000
(@
Token (@
Feature activation+1.491
t
Tokent
Feature activation+2.240
5
Token5
Feature activation+2.756
_
Token_
Feature activation+2.227
har
Tokenhar
Feature activation+0.602
ris
Tokenris
Feature activation+2.422
-
Token-
Feature activation+0.109
Dead
TokenDead
Feature activation+0.326
-
Token-
Feature activation+0.570
Red
TokenRed
Feature activation+0.000
emption
Tokenemption
Feature activation+0.677
-
Token-
Feature activation+0.965
Step
TokenStep
Feature activation+0.516
-
Token-
Feature activation+1.246
1
Token1
Feature activation+0.676
-
Token-
Feature activation+0.971
Version
TokenVersion
Feature activation+0.174
,
Token,
Feature activation+3.470
improved
Token improved
Feature activation+3.272
security
Token security
Feature activation+3.454
,
Token,
Feature activation+2.578
convenient
Token convenient
Feature activation+1.846
updates
Token updates
Feature activation+1.187
,
Token,
Feature activation+1.056
and
Token and
Feature activation+0.012
longer
Token longer
Feature activation+0.873
battery
Token battery
Feature activation+0.417
life
Token life
Feature activation+0.565

INTERVAL 0.000 - 0.805
CONTAINS 99.815%

than
Token than
Feature activation+0.000
anyone
Token anyone
Feature activation+0.000
,
Token,
Feature activation+0.000
told
Token told
Feature activation+0.000
me
Token me
Feature activation+0.000
today
Token today
Feature activation+0.000
.
Token.
Feature activation+0.000
In
Token In
Feature activation+0.000
short
Token short
Feature activation+0.000
,
Token,
Feature activation+0.000
many
Token many
Feature activation+0.000
ion
Tokenion
Feature activation+0.000
ario
Tokenario
Feature activation+0.000
Instit
Token Instit
Feature activation+0.000
uc
Tokenuc
Feature activation+0.000
ional
Tokenional
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
from
Token from
Feature activation+0.000
government
Token government
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
're
Token're
Feature activation+0.000
driving
Token driving
Feature activation+0.000
like
Token like
Feature activation+0.000
an
Token an
Feature activation+0.000
absolute
Token absolute
Feature activation+0.000
c
Token c
Feature activation+0.000
***
Token***
Feature activation+0.000
mate
Token mate
Feature activation+0.000
,"
Token,"
Feature activation+0.000
he
Token he
Feature activation+0.000
said
Token said
Feature activation+0.000
play
Token play
Feature activation+0.000
a
Token a
Feature activation+0.000
part
Token part
Feature activation+0.000
at
Token at
Feature activation+0.000
Chelsea
Token Chelsea
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Def
TokenDef
Feature activation+0.000
ender
Tokenender
Feature activation+0.000
De
Token De
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
Brien
TokenBrien
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
desk
Token desk
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Search
TokenSearch
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000

BOTTOM ACTIVATIONS
MIN = 0.000

atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
&
Token &
Feature activation+0.000
Development
Token Development
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
&
Token &
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
rik
Tokenrik
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
rik
Tokenrik
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000

Top feature 2 in H8.5: (feature 20863)

TOP ACTIVATIONS
MAX = 6.534

shaming
Token shaming
Feature activation+0.000
to
Token to
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
be
Token be
Feature activation+2.800
accepted
Token accepted
Feature activation+6.534
,
Token,
Feature activation+5.268
but
Token but
Feature activation+0.000
welcomed
Token welcomed
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
intact
Token intact
Feature activation+0.000
not
Token not
Feature activation+0.014
just
Token just
Feature activation+0.441
for
Token for
Feature activation+1.890
us
Token us
Feature activation+5.623
today
Token today
Feature activation+6.454
,
Token,
Feature activation+5.912
but
Token but
Feature activation+0.000
for
Token for
Feature activation+0.581
the
Token the
Feature activation+0.194
generations
Token generations
Feature activation+0.000
and
Token and
Feature activation+0.408
U
Token U
Feature activation+0.000
H
TokenH
Feature activation+0.000
H
Token H
Feature activation+0.000
ilo
Tokenilo
Feature activation+2.722
,
Token,
Feature activation+6.202
but
Token but
Feature activation+0.000
several
Token several
Feature activation+0.000
community
Token community
Feature activation+0.000
colleges
Token colleges
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+4.224
only
Token only
Feature activation+0.000
throughout
Token throughout
Feature activation+1.390
the
Token the
Feature activation+0.060
U
Token U
Feature activation+0.217
H
TokenH
Feature activation+2.107
system
Token system
Feature activation+6.150
âĢĶ
Token âĢĶ
Feature activation+3.041
a
Token a
Feature activation+0.385
system
Token system
Feature activation+0.000
that
Token that
Feature activation+0.000
includes
Token includes
Feature activation+0.000
about
Token about
Feature activation+0.000
owning
Token owning
Feature activation+0.000
not
Token not
Feature activation+0.000
just
Token just
Feature activation+0.000
search
Token search
Feature activation+2.726
capacities
Token capacities
Feature activation+6.139
,
Token,
Feature activation+5.352
video
Token video
Feature activation+0.546
and
Token and
Feature activation+0.232
maps
Token maps
Feature activation+1.926
,
Token,
Feature activation+6.095
capacities
Token capacities
Feature activation+6.139
,
Token,
Feature activation+5.352
video
Token video
Feature activation+0.546
and
Token and
Feature activation+0.232
maps
Token maps
Feature activation+1.926
,
Token,
Feature activation+6.095
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
the
Token the
Feature activation+0.000
shape
Token shape
Feature activation+0.000
of
Token of
Feature activation+0.000
offensive
Token offensive
Feature activation+1.004
to
Token to
Feature activation+0.134
people
Token people
Feature activation+1.538
with
Token with
Feature activation+0.000
bipolar
Token bipolar
Feature activation+0.002
disorder
Token disorder
Feature activation+6.091
,
Token,
Feature activation+5.854
but
Token but
Feature activation+0.000
then
Token then
Feature activation+0.000
also
Token also
Feature activation+0.000
managing
Token managing
Feature activation+0.000
not
Token not
Feature activation+0.000
just
Token just
Feature activation+0.000
the
Token the
Feature activation+0.589
standard
Token standard
Feature activation+1.368
Medicare
Token Medicare
Feature activation+1.921
,
Token,
Feature activation+6.066
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
the
Token the
Feature activation+0.000
[
Token [
Feature activation+0.000
private
Tokenprivate
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
for
Token for
Feature activation+0.998
violence
Token violence
Feature activation+5.446
and
Token and
Feature activation+0.930
nudity
Token nudity
Feature activation+6.060
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
political
Token political
Feature activation+0.000
sensitivity
Token sensitivity
Feature activation+0.000
.
Token.
Feature activation+0.000
not
Token not
Feature activation+0.014
just
Token just
Feature activation+0.441
for
Token for
Feature activation+1.890
us
Token us
Feature activation+5.623
today
Token today
Feature activation+6.454
,
Token,
Feature activation+5.912
but
Token but
Feature activation+0.000
for
Token for
Feature activation+0.581
the
Token the
Feature activation+0.194
generations
Token generations
Feature activation+0.000
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.465
wars
Token wars
Feature activation+2.874
in
Token in
Feature activation+0.592
Afghanistan
Token Afghanistan
Feature activation+2.256
and
Token and
Feature activation+0.000
Iraq
Token Iraq
Feature activation+5.886
,
Token,
Feature activation+4.754
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
to
Token to
Feature activation+0.000
Guant
Token Guant
Feature activation+0.000
just
Token just
Feature activation+0.000
how
Token how
Feature activation+1.791
loud
Token loud
Feature activation+2.370
the
Token the
Feature activation+0.435
noise
Token noise
Feature activation+0.945
is
Token is
Feature activation+5.874
,
Token,
Feature activation+5.407
but
Token but
Feature activation+0.000
how
Token how
Feature activation+0.000
far
Token far
Feature activation+0.000
away
Token away
Feature activation+0.000
to
Token to
Feature activation+0.134
people
Token people
Feature activation+1.538
with
Token with
Feature activation+0.000
bipolar
Token bipolar
Feature activation+0.002
disorder
Token disorder
Feature activation+6.091
,
Token,
Feature activation+5.854
but
Token but
Feature activation+0.000
then
Token then
Feature activation+0.000
also
Token also
Feature activation+0.000
managing
Token managing
Feature activation+0.000
to
Token to
Feature activation+0.000
embraces
Token embraces
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
indict
Token indict
Feature activation+1.075
able
Tokenable
Feature activation+1.531
crimes
Token crimes
Feature activation+5.849
but
Token but
Feature activation+0.000
"
Token "
Feature activation+0.000
conduct
Tokenconduct
Feature activation+0.000
...
Token ...
Feature activation+0.000
grossly
Token grossly
Feature activation+0.000
with
Token with
Feature activation+1.233
the
Token the
Feature activation+0.227
money
Token money
Feature activation+3.299
in
Token in
Feature activation+1.189
the
Token the
Feature activation+0.171
suit
Token suit
Feature activation+5.770
cases
Tokencases
Feature activation+5.715
,
Token,
Feature activation+4.911
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
having
Token having
Feature activation+0.000
only
Token only
Feature activation+0.000
a
Token a
Feature activation+1.688
sequel
Token sequel
Feature activation+4.224
to
Token to
Feature activation+1.165
Out
Token Out
Feature activation+1.212
Run
Token Run
Feature activation+5.756
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
a
Token a
Feature activation+0.000
new
Token new
Feature activation+0.000
entry
Token entry
Feature activation+0.000
for
Token for
Feature activation+1.382
women
Token women
Feature activation+5.285
âĢ
TokenâĢ
Feature activation+2.030
Ļ
TokenĻ
Feature activation+3.171
s
Tokens
Feature activation+2.280
health
Token health
Feature activation+5.750
,
Token,
Feature activation+4.460
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
moral
Token moral
Feature activation+0.000
and
Token and
Feature activation+0.000
the
Token the
Feature activation+0.227
money
Token money
Feature activation+3.299
in
Token in
Feature activation+1.189
the
Token the
Feature activation+0.171
suit
Token suit
Feature activation+5.770
cases
Tokencases
Feature activation+5.715
,
Token,
Feature activation+4.911
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
having
Token having
Feature activation+0.000
attempted
Token attempted
Feature activation+0.000
remain
Token remain
Feature activation+0.000
intact
Token intact
Feature activation+0.000
not
Token not
Feature activation+0.014
just
Token just
Feature activation+0.441
for
Token for
Feature activation+1.890
us
Token us
Feature activation+5.623
today
Token today
Feature activation+6.454
,
Token,
Feature activation+5.912
but
Token but
Feature activation+0.000
for
Token for
Feature activation+0.581
the
Token the
Feature activation+0.194
only
Token only
Feature activation+0.000
ruined
Token ruined
Feature activation+1.630
its
Token its
Feature activation+0.668
own
Token own
Feature activation+0.804
balance
Token balance
Feature activation+0.709
sheet
Token sheet
Feature activation+5.607
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
undermined
Token undermined
Feature activation+0.000
the
Token the
Feature activation+0.000
Puerto
Token Puerto
Feature activation+0.000

Top DFA by src position
MAX = 7.331

kind
Token kind
Feature activation+0.001
Top resid features:
of
Token of
Feature activation-0.005
Top resid features:
shaming
Token shaming
Feature activation+0.003
Top resid features:
to
Token to
Feature activation-0.030
Top resid features:
not
Token not
Feature activation+0.039
Top resid features:
only
Token only
Feature activation+7.331
Top resid features:
be
Token be
Feature activation+0.060
Top resid features:
accepted
Token accepted
Feature activation+0.098
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
but
Token but
Feature activation+0.000
Top resid features:
welcomed
Token welcomed
Feature activation+0.000
Top resid features:
rights
Token rights
Feature activation+0.020
Top resid features:
will
Token will
Feature activation-0.003
Top resid features:
remain
Token remain
Feature activation-0.006
Top resid features:
intact
Token intact
Feature activation+0.013
Top resid features:
not
Token not
Feature activation+0.074
Top resid features:
just
Token just
Feature activation+6.480
Top resid features:
for
Token for
Feature activation+0.258
Top resid features:
us
Token us
Feature activation+0.355
Top resid features:
today
Token today
Feature activation+0.179
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
but
Token but
Feature activation+0.000
Top resid features:
a
Token a
Feature activation+0.004
Top resid features:
system
Token system
Feature activation+0.008
Top resid features:
that
Token that
Feature activation+0.017
Top resid features:
includes
Token includes
Feature activation+0.040
Top resid features:
not
Token not
Feature activation-0.052
Top resid features:
only
Token only
Feature activation+5.550
Top resid features:
such
Token such
Feature activation+0.024
Top resid features:
colleges
Token colleges
Feature activation+0.064
Top resid features:
such
Token such
Feature activation+0.075
Top resid features:
as
Token as
Feature activation-0.005
Top resid features:
U
Token U
Feature activation-0.010
Top resid features:
much
Token much
Feature activation-0.010
Top resid features:
tuition
Token tuition
Feature activation-0.013
Top resid features:
costs
Token costs
Feature activation-0.022
Top resid features:
now
Token now
Feature activation-0.045
Top resid features:
not
Token not
Feature activation+0.045
Top resid features:
only
Token only
Feature activation+6.893
Top resid features:
throughout
Token throughout
Feature activation+0.184
Top resid features:
the
Token the
Feature activation+0.046
Top resid features:
U
Token U
Feature activation-0.009
Top resid features:
H
TokenH
Feature activation+0.016
Top resid features:
system
Token system
Feature activation+0.054
Top resid features:
become
Token become
Feature activation+0.003
Top resid features:
greedy
Token greedy
Feature activation+0.011
Top resid features:
about
Token about
Feature activation+0.027
Top resid features:
owning
Token owning
Feature activation-0.080
Top resid features:
not
Token not
Feature activation+0.178
Top resid features:
just
Token just
Feature activation+6.617
Top resid features:
search
Token search
Feature activation+0.136
Top resid features:
capacities
Token capacities
Feature activation+0.119
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
video
Token video
Feature activation+0.000
Top resid features:
and
Token and
Feature activation+0.000
Top resid features:
become
Token become
Feature activation+0.005
Top resid features:
greedy
Token greedy
Feature activation+0.011
Top resid features:
about
Token about
Feature activation+0.015
Top resid features:
owning
Token owning
Feature activation-0.025
Top resid features:
not
Token not
Feature activation+0.057
Top resid features:
just
Token just
Feature activation+6.879
Top resid features:
search
Token search
Feature activation+0.040
Top resid features:
capacities
Token capacities
Feature activation+0.155
Top resid features:
,
Token,
Feature activation-0.128
Top resid features:
video
Token video
Feature activation-0.019
Top resid features:
and
Token and
Feature activation+0.013
Top resid features:
make
Token make
Feature activation-0.004
Top resid features:
an
Token an
Feature activation-0.000
Top resid features:
apology
Token apology
Feature activation-0.000
Top resid features:
after
Token after
Feature activation-0.052
Top resid features:
not
Token not
Feature activation+0.012
Top resid features:
only
Token only
Feature activation+6.892
Top resid features:
stocking
Token stocking
Feature activation+0.089
Top resid features:
a
Token a
Feature activation+0.044
Top resid features:
product
Token product
Feature activation+0.020
Top resid features:
offensive
Token offensive
Feature activation+0.012
Top resid features:
to
Token to
Feature activation+0.036
Top resid features:
to
Token to
Feature activation-0.004
Top resid features:
take
Token take
Feature activation-0.007
Top resid features:
advantage
Token advantage
Feature activation-0.005
Top resid features:
of
Token of
Feature activation-0.010
Top resid features:
not
Token not
Feature activation-0.032
Top resid features:
just
Token just
Feature activation+7.106
Top resid features:
the
Token the
Feature activation+0.186
Top resid features:
standard
Token standard
Feature activation+0.031
Top resid features:
Medicare
Token Medicare
Feature activation-0.020
Top resid features:
,
Token,
Feature activation-0.101
Top resid features:
but
Token but
Feature activation+0.000
Top resid features:
who
Token who
Feature activation+0.007
Top resid features:
often
Token often
Feature activation+0.011
Top resid features:
flag
Token flag
Feature activation-0.026
Top resid features:
scenes
Token scenes
Feature activation+0.039
Top resid features:
not
Token not
Feature activation+0.094
Top resid features:
only
Token only
Feature activation+6.766
Top resid features:
for
Token for
Feature activation+0.196
Top resid features:
violence
Token violence
Feature activation+0.063
Top resid features:
and
Token and
Feature activation+0.031
Top resid features:
nudity
Token nudity
Feature activation+0.053
Top resid features:
but
Token but
Feature activation+0.000
Top resid features:
rights
Token rights
Feature activation+0.026
Top resid features:
will
Token will
Feature activation-0.001
Top resid features:
remain
Token remain
Feature activation-0.007
Top resid features:
intact
Token intact
Feature activation+0.020
Top resid features:
not
Token not
Feature activation-0.001
Top resid features:
just
Token just
Feature activation+6.571
Top resid features:
for
Token for
Feature activation+0.130
Top resid features:
us
Token us
Feature activation+0.137
Top resid features:
today
Token today
Feature activation+0.350
Top resid features:
,
Token,
Feature activation-0.249
Top resid features:
but
Token but
Feature activation+0.000
Top resid features:
ann
Tokenann
Feature activation-0.002
Top resid features:
elled
Tokenelled
Feature activation+0.011
Top resid features:
the
Token the
Feature activation+0.018
Top resid features:
opposition
Token opposition
Feature activation+0.001
Top resid features:
not
Token not
Feature activation+0.155
Top resid features:
only
Token only
Feature activation+6.388
Top resid features:
to
Token to
Feature activation+0.155
Top resid features:
the
Token the
Feature activation+0.086
Top resid features:
wars
Token wars
Feature activation+0.057
Top resid features:
in
Token in
Feature activation+0.016
Top resid features:
Afghanistan
Token Afghanistan
Feature activation-0.055
Top resid features:
to
Token to
Feature activation-0.000
Top resid features:
take
Token take
Feature activation+0.002
Top resid features:
into
Token into
Feature activation+0.001
Top resid features:
account
Token account
Feature activation+0.007
Top resid features:
not
Token not
Feature activation+0.043
Top resid features:
just
Token just
Feature activation+6.318
Top resid features:
how
Token how
Feature activation+0.370
Top resid features:
loud
Token loud
Feature activation+0.140
Top resid features:
the
Token the
Feature activation-0.017
Top resid features:
noise
Token noise
Feature activation+0.042
Top resid features:
is
Token is
Feature activation-0.043
Top resid features:
make
Token make
Feature activation-0.006
Top resid features:
an
Token an
Feature activation+0.005
Top resid features:
apology
Token apology
Feature activation-0.004
Top resid features:
after
Token after
Feature activation-0.053
Top resid features:
not
Token not
Feature activation-0.043
Top resid features:
only
Token only
Feature activation+7.236
Top resid features:
stocking
Token stocking
Feature activation+0.068
Top resid features:
a
Token a
Feature activation+0.040
Top resid features:
product
Token product
Feature activation+0.005
Top resid features:
offensive
Token offensive
Feature activation+0.015
Top resid features:
to
Token to
Feature activation+0.026
Top resid features:
good
Tokengood
Feature activation-0.011
Top resid features:
behavior
Token behavior
Feature activation-0.001
Top resid features:
"
Token"
Feature activation-0.003
Top resid features:
embraces
Token embraces
Feature activation-0.021
Top resid features:
not
Token not
Feature activation+0.072
Top resid features:
only
Token only
Feature activation+6.614
Top resid features:
indict
Token indict
Feature activation+0.056
Top resid features:
able
Tokenable
Feature activation+0.010
Top resid features:
crimes
Token crimes
Feature activation+0.073
Top resid features:
but
Token but
Feature activation+0.000
Top resid features:
"
Token "
Feature activation+0.000
Top resid features:
You
TokenYou
Feature activation+0.006
Top resid features:
have
Token have
Feature activation-0.003
Top resid features:
been
Token been
Feature activation+0.006
Top resid features:
caught
Token caught
Feature activation+0.003
Top resid features:
not
Token not
Feature activation+0.020
Top resid features:
only
Token only
Feature activation+6.561
Top resid features:
red
Token red
Feature activation+0.005
Top resid features:
-
Token-
Feature activation+0.000
Top resid features:
handed
Tokenhanded
Feature activation+0.068
Top resid features:
with
Token with
Feature activation+0.031
Top resid features:
the
Token the
Feature activation+0.033
Top resid features:
had
Token had
Feature activation-0.000
Top resid features:
it
Token it
Feature activation-0.020
Top resid features:
by
Token by
Feature activation-0.006
Top resid features:
releasing
Token releasing
Feature activation+0.014
Top resid features:
not
Token not
Feature activation+0.040
Top resid features:
only
Token only
Feature activation+6.260
Top resid features:
a
Token a
Feature activation+0.163
Top resid features:
sequel
Token sequel
Feature activation+0.151
Top resid features:
to
Token to
Feature activation+0.121
Top resid features:
Out
Token Out
Feature activation+0.035
Top resid features:
Run
Token Run
Feature activation+0.002
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
abortions
Token abortions
Feature activation+0.037
Top resid features:
might
Token might
Feature activation+0.019
Top resid features:
be
Token be
Feature activation-0.009
Top resid features:
not
Token not
Feature activation+0.030
Top resid features:
only
Token only
Feature activation+6.308
Top resid features:
good
Token good
Feature activation+0.050
Top resid features:
for
Token for
Feature activation+0.056
Top resid features:
women
Token women
Feature activation+0.126
Top resid features:
âĢ
TokenâĢ
Feature activation+0.019
Top resid features:
Ļ
TokenĻ
Feature activation+0.080
Top resid features:
You
TokenYou
Feature activation-0.000
Top resid features:
have
Token have
Feature activation+0.004
Top resid features:
been
Token been
Feature activation+0.008
Top resid features:
caught
Token caught
Feature activation-0.006
Top resid features:
not
Token not
Feature activation+0.018
Top resid features:
only
Token only
Feature activation+6.460
Top resid features:
red
Token red
Feature activation+0.003
Top resid features:
-
Token-
Feature activation+0.003
Top resid features:
handed
Tokenhanded
Feature activation+0.063
Top resid features:
with
Token with
Feature activation+0.032
Top resid features:
the
Token the
Feature activation+0.038
Top resid features:
rights
Token rights
Feature activation+0.040
Top resid features:
will
Token will
Feature activation-0.002
Top resid features:
remain
Token remain
Feature activation-0.005
Top resid features:
intact
Token intact
Feature activation+0.012
Top resid features:
not
Token not
Feature activation+0.096
Top resid features:
just
Token just
Feature activation+5.900
Top resid features:
for
Token for
Feature activation+0.318
Top resid features:
us
Token us
Feature activation+0.075
Top resid features:
today
Token today
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
but
Token but
Feature activation+0.000
Top resid features:
Ļ
TokenĻ
Feature activation+0.012
Top resid features:
s
Tokens
Feature activation+0.005
Top resid features:
poor
Token poor
Feature activation+0.028
Top resid features:
performance
Token performance
Feature activation+0.005
Top resid features:
not
Token not
Feature activation+0.019
Top resid features:
only
Token only
Feature activation+6.572
Top resid features:
ruined
Token ruined
Feature activation+0.096
Top resid features:
its
Token its
Feature activation+0.021
Top resid features:
own
Token own
Feature activation+0.005
Top resid features:
balance
Token balance
Feature activation-0.002
Top resid features:
sheet
Token sheet
Feature activation+0.001
Top resid features:

Decoder Weights Distribution

Head 0: 0.09

Head 1: 0.02

Head 2: 0.04

Head 3: 0.03

Head 4: 0.06

Head 5: 0.57

Head 6: 0.02

Head 7: 0.05

Head 8: 0.03

Head 9: 0.03

Head 10: 0.06

Head 11: 0.02

Positive logits

but2.99

but2.67

;2.35

);2.24

.;2.07

But2.06

But2.05

;;;;2.04

BUT2.01

";1.88

%;1.85

BUT1.80

';1.74

)}1.71

];1.67

;}1.65

;;1.65

fine1.62

;"1.61

frail1.61

Negative logits

".-2.48

>.-2.42

].-2.40

`.-2.31

":{"-2.24

ccording-2.18

=================================================================-2.16

."-2.16

'.-2.12

.''.-2.11

[/-2.07

''.-2.06

[/-1.97

.-1.85

!.-1.85

"!-1.84

+++-1.81

nonetheless-1.79

_.-1.79

etheless-1.77

INTERVAL 5.881 - 6.534
CONTAINS 0.001%

shaming
Token shaming
Feature activation+0.000
to
Token to
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
be
Token be
Feature activation+2.800
accepted
Token accepted
Feature activation+6.534
,
Token,
Feature activation+5.268
but
Token but
Feature activation+0.000
welcomed
Token welcomed
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
and
Token and
Feature activation+0.408
U
Token U
Feature activation+0.000
H
TokenH
Feature activation+0.000
H
Token H
Feature activation+0.000
ilo
Tokenilo
Feature activation+2.722
,
Token,
Feature activation+6.202
but
Token but
Feature activation+0.000
several
Token several
Feature activation+0.000
community
Token community
Feature activation+0.000
colleges
Token colleges
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+4.224
about
Token about
Feature activation+0.000
owning
Token owning
Feature activation+0.000
not
Token not
Feature activation+0.000
just
Token just
Feature activation+0.000
search
Token search
Feature activation+2.726
capacities
Token capacities
Feature activation+6.139
,
Token,
Feature activation+5.352
video
Token video
Feature activation+0.546
and
Token and
Feature activation+0.232
maps
Token maps
Feature activation+1.926
,
Token,
Feature activation+6.095
the
Token the
Feature activation+0.465
wars
Token wars
Feature activation+2.874
in
Token in
Feature activation+0.592
Afghanistan
Token Afghanistan
Feature activation+2.256
and
Token and
Feature activation+0.000
Iraq
Token Iraq
Feature activation+5.886
,
Token,
Feature activation+4.754
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
to
Token to
Feature activation+0.000
Guant
Token Guant
Feature activation+0.000
only
Token only
Feature activation+0.000
throughout
Token throughout
Feature activation+1.390
the
Token the
Feature activation+0.060
U
Token U
Feature activation+0.217
H
TokenH
Feature activation+2.107
system
Token system
Feature activation+6.150
âĢĶ
Token âĢĶ
Feature activation+3.041
a
Token a
Feature activation+0.385
system
Token system
Feature activation+0.000
that
Token that
Feature activation+0.000
includes
Token includes
Feature activation+0.000

INTERVAL 5.227 - 5.881
CONTAINS 0.002%

flag
Token flag
Feature activation+0.000
scenes
Token scenes
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
for
Token for
Feature activation+0.998
violence
Token violence
Feature activation+5.446
and
Token and
Feature activation+0.930
nudity
Token nudity
Feature activation+6.060
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
political
Token political
Feature activation+0.000
just
Token just
Feature activation+0.000
how
Token how
Feature activation+1.791
loud
Token loud
Feature activation+2.370
the
Token the
Feature activation+0.435
noise
Token noise
Feature activation+0.945
is
Token is
Feature activation+5.874
,
Token,
Feature activation+5.407
but
Token but
Feature activation+0.000
how
Token how
Feature activation+0.000
far
Token far
Feature activation+0.000
away
Token away
Feature activation+0.000
for
Token for
Feature activation+1.382
women
Token women
Feature activation+5.285
âĢ
TokenâĢ
Feature activation+2.030
Ļ
TokenĻ
Feature activation+3.171
s
Tokens
Feature activation+2.280
health
Token health
Feature activation+5.750
,
Token,
Feature activation+4.460
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
moral
Token moral
Feature activation+0.000
and
Token and
Feature activation+0.000
only
Token only
Feature activation+0.000
ruined
Token ruined
Feature activation+1.630
its
Token its
Feature activation+0.668
own
Token own
Feature activation+0.804
balance
Token balance
Feature activation+0.709
sheet
Token sheet
Feature activation+5.607
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
undermined
Token undermined
Feature activation+0.000
the
Token the
Feature activation+0.000
Puerto
Token Puerto
Feature activation+0.000
the
Token the
Feature activation+0.791
walls
Token walls
Feature activation+4.509
of
Token of
Feature activation+0.991
the
Token the
Feature activation+0.664
university
Token university
Feature activation+5.556
,
Token,
Feature activation+5.351
but
Token but
Feature activation+0.000
to
Token to
Feature activation+0.000
focus
Token focus
Feature activation+0.000
externally
Token externally
Feature activation+0.000
around
Token around
Feature activation+0.000

INTERVAL 4.574 - 5.227
CONTAINS 0.002%

that
Token that
Feature activation+1.552
an
Token an
Feature activation+0.408
attack
Token attack
Feature activation+0.074
is
Token is
Feature activation+0.742
highly
Token highly
Feature activation+0.733
likely
Token likely
Feature activation+4.925
,
Token,
Feature activation+3.404
but
Token but
Feature activation+0.000
that
Token that
Feature activation+0.000
another
Token another
Feature activation+0.000
attack
Token attack
Feature activation+0.000
knows
Token knows
Feature activation+0.000
not
Token not
Feature activation+0.212
only
Token only
Feature activation+0.000
what
Token what
Feature activation+0.837
we
Token we
Feature activation+1.435
buy
Token buy
Feature activation+4.854
,
Token,
Feature activation+2.822
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
what
Token what
Feature activation+0.000
time
Token time
Feature activation+0.000
that
Token that
Feature activation+0.000
includes
Token includes
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
such
Token such
Feature activation+1.599
colleges
Token colleges
Feature activation+4.946
such
Token such
Feature activation+1.772
as
Token as
Feature activation+1.323
U
Token U
Feature activation+0.036
H
TokenH
Feature activation+1.457
West
Token West
Feature activation+1.138
only
Token only
Feature activation+0.000
be
Token be
Feature activation+2.017
very
Token very
Feature activation+1.392
progressive
Token progressive
Feature activation+3.934
for
Token for
Feature activation+2.246
women
Token women
Feature activation+5.149
in
Token in
Feature activation+1.786
the
Token the
Feature activation+1.060
genre
Token genre
Feature activation+5.333
,
Token,
Feature activation+3.082
but
Token but
Feature activation+0.000
the
Token the
Feature activation+0.000
people
Token people
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
of
Token of
Feature activation+0.458
Panama
Token Panama
Feature activation+4.752
,
Token,
Feature activation+4.300
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000

INTERVAL 3.920 - 4.574
CONTAINS 0.005%

from
Token from
Feature activation+0.204
team
Token team
Feature activation+0.413
Santos
Token Santos
Feature activation+1.970
-
Token-
Feature activation+0.000
Glass
TokenGlass
Feature activation+3.138
,
Token,
Feature activation+4.539
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
serves
Token serves
Feature activation+0.000
Victor
Token Victor
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
changes
Token changes
Feature activation+4.221
in
Token in
Feature activation+0.758
their
Token their
Feature activation+0.354
genetic
Token genetic
Feature activation+0.307
codes
Token codes
Feature activation+5.407
,
Token,
Feature activation+3.998
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
by
Token by
Feature activation+0.000
larger
Token larger
Feature activation+0.000
mutations
Token mutations
Feature activation+0.000
only
Token only
Feature activation+0.000
by
Token by
Feature activation+0.000
single
Token single
Feature activation+0.000
-
Token-
Feature activation+0.241
letter
Tokenletter
Feature activation+0.338
changes
Token changes
Feature activation+4.221
in
Token in
Feature activation+0.758
their
Token their
Feature activation+0.354
genetic
Token genetic
Feature activation+0.307
codes
Token codes
Feature activation+5.407
,
Token,
Feature activation+3.998
abortions
Token abortions
Feature activation+0.000
might
Token might
Feature activation+0.000
be
Token be
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
good
Token good
Feature activation+4.126
for
Token for
Feature activation+1.382
women
Token women
Feature activation+5.285
âĢ
TokenâĢ
Feature activation+2.030
Ļ
TokenĻ
Feature activation+3.171
s
Tokens
Feature activation+2.280
to
Token to
Feature activation+0.000
satisfy
Token satisfy
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
the
Token the
Feature activation+0.643
story
Token story
Feature activation+4.381
buffs
Token buffs
Feature activation+4.302
,
Token,
Feature activation+4.334
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
the
Token the
Feature activation+0.000

INTERVAL 3.267 - 3.920
CONTAINS 0.003%

red
Token red
Feature activation+0.893
-
Token-
Feature activation+0.000
handed
Tokenhanded
Feature activation+3.288
with
Token with
Feature activation+1.233
the
Token the
Feature activation+0.227
money
Token money
Feature activation+3.299
in
Token in
Feature activation+1.189
the
Token the
Feature activation+0.171
suit
Token suit
Feature activation+5.770
cases
Tokencases
Feature activation+5.715
,
Token,
Feature activation+4.911
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
to
Token to
Feature activation+2.002
sell
Token sell
Feature activation+1.864
Apple
Token Apple
Feature activation+1.145
devices
Token devices
Feature activation+3.873
,
Token,
Feature activation+4.412
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
to
Token to
Feature activation+0.000
entirely
Token entirely
Feature activation+0.000
inspect
Token inspect
Feature activation+1.505
the
Token the
Feature activation+0.726
photographs
Token photographs
Feature activation+3.195
and
Token and
Feature activation+0.624
videos
Token videos
Feature activation+3.705
,
Token,
Feature activation+3.401
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
to
Token to
Feature activation+0.000
copy
Token copy
Feature activation+0.000
them
Token them
Feature activation+0.000
Not
Token Not
Feature activation+0.000
only
Token only
Feature activation+0.000
is
Token is
Feature activation+0.000
the
Token the
Feature activation+0.000
material
Token material
Feature activation+0.598
soft
Token soft
Feature activation+3.534
,
Token,
Feature activation+2.986
comfortable
Token comfortable
Feature activation+1.234
,
Token,
Feature activation+0.906
and
Token and
Feature activation+0.000
warm
Token warm
Feature activation+2.777
include
Token include
Feature activation+2.024
hundreds
Token hundreds
Feature activation+0.388
of
Token of
Feature activation+0.202
software
Token software
Feature activation+0.206
enhancements
Token enhancements
Feature activation+2.933
,
Token,
Feature activation+3.663
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
add
Token add
Feature activation+0.000
content
Token content
Feature activation+0.000
of
Token of
Feature activation+0.000

INTERVAL 2.614 - 3.267
CONTAINS 0.007%

by
Token by
Feature activation+2.501
the
Token the
Feature activation+1.792
U
Token U
Feature activation+0.473
.
Token.
Feature activation+0.000
S
TokenS
Feature activation+2.494
.,
Token.,
Feature activation+2.898
but
Token but
Feature activation+0.000
widely
Token widely
Feature activation+0.000
supported
Token supported
Feature activation+0.000
by
Token by
Feature activation+0.000
governments
Token governments
Feature activation+0.000
growing
Token growing
Feature activation+0.154
number
Token number
Feature activation+1.061
of
Token of
Feature activation+0.390
published
Token published
Feature activation+0.000
specs
Token specs
Feature activation+0.000
misleading
Token misleading
Feature activation+2.729
and
Token and
Feature activation+0.000
/
Token/
Feature activation+0.000
or
Tokenor
Feature activation+0.000
over
Token over
Feature activation+0.000
in
Tokenin
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
ve
Tokenve
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
hired
Token hired
Feature activation+2.712
talent
Token talent
Feature activation+3.719
in
Token in
Feature activation+2.438
South
Token South
Feature activation+0.660
East
Token East
Feature activation+0.474
Asia
Token Asia
Feature activation+4.007
them
Token them
Feature activation+0.000
not
Token not
Feature activation+0.000
simply
Token simply
Feature activation+0.360
to
Token to
Feature activation+1.487
one
Token one
Feature activation+0.380
another
Token another
Feature activation+3.074
but
Token but
Feature activation+0.000
to
Token to
Feature activation+0.000
one
Token one
Feature activation+0.000
another
Token another
Feature activation+0.000
through
Token through
Feature activation+0.000
good
Token good
Feature activation+0.353
friend
Token friend
Feature activation+2.152
of
Token of
Feature activation+0.000
Gi
Token Gi
Feature activation+0.000
Review
TokenReview
Feature activation+0.000
s
Tokens
Feature activation+2.976
,
Token,
Feature activation+3.100
but
Token but
Feature activation+0.000
a
Token a
Feature activation+0.000
fantastic
Token fantastic
Feature activation+0.000
instructor
Token instructor
Feature activation+0.000

INTERVAL 1.960 - 2.614
CONTAINS 0.009%

Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
aimed
Token aimed
Feature activation+0.000
not
Token not
Feature activation+0.000
at
Token at
Feature activation+0.010
Latinos
Token Latinos
Feature activation+2.482
,
Token,
Feature activation+1.664
but
Token but
Feature activation+0.000
at
Token at
Feature activation+0.000
stupid
Token stupid
Feature activation+0.000
Republicans
Token Republicans
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
illegal
Token illegal
Feature activation+2.461
aliens
Token aliens
Feature activation+0.641
who
Token who
Feature activation+0.012
are
Token are
Feature activation+0.537
escaping
Token escaping
Feature activation+0.710
enforcement
Token enforcement
Feature activation+0.420
,
Token,
Feature activation+0.000
not
Token not
Feature activation+0.000
because
Token because
Feature activation+0.000
they
Token they
Feature activation+0.063
like
Token like
Feature activation+1.419
Yanukovych
Token Yanukovych
Feature activation+2.009
(
Token (
Feature activation+0.064
except
Tokenexcept
Feature activation+0.000
perhaps
Token perhaps
Feature activation+0.000
in
Token in
Feature activation+0.000
his
Token his
Feature activation+0.000
reductions
Token reductions
Feature activation+0.000
not
Token not
Feature activation+0.000
because
Token because
Feature activation+0.000
their
Token their
Feature activation+0.179
product
Token product
Feature activation+0.958
is
Token is
Feature activation+2.044
effective
Token effective
Feature activation+3.680
but
Token but
Feature activation+0.000
because
Token because
Feature activation+0.000
they
Token they
Feature activation+0.000
want
Token want
Feature activation+0.000
not
Token not
Feature activation+0.000
just
Token just
Feature activation+0.000
because
Token because
Feature activation+0.000
the
Token the
Feature activation+0.357
state
Token state
Feature activation+0.698
has
Token has
Feature activation+1.996
resources
Token resources
Feature activation+1.998
that
Token that
Feature activation+0.426
small
Token small
Feature activation+0.226
towns
Token towns
Feature activation+0.000
like
Token like
Feature activation+0.000

INTERVAL 1.307 - 1.960
CONTAINS 0.016%

re
Tokenre
Feature activation+0.000
about
Token about
Feature activation+0.000
to
Token to
Feature activation+0.000
blow
Token blow
Feature activation+0.000
up
Token up
Feature activation+0.000
,
Token,
Feature activation+1.458
but
Token but
Feature activation+0.000
nothing
Token nothing
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
next
Token next
Feature activation+0.637
big
Token big
Feature activation+0.000
thing
Token thing
Feature activation+3.468
in
Token in
Feature activation+0.000
Ontario
Token Ontario
Feature activation+2.349
brewing
Token brewing
Feature activation+1.619
,
Token,
Feature activation+2.882
but
Token but
Feature activation+0.000
this
Token this
Feature activation+0.000
beer
Token beer
Feature activation+0.000
is
Token is
Feature activation+0.000
is
Token is
Feature activation+0.000
good
Token good
Feature activation+0.000
news
Token news
Feature activation+0.000
not
Token not
Feature activation+0.000
for
Token for
Feature activation+0.000
China
Token China
Feature activation+1.481
,
Token,
Feature activation+0.954
but
Token but
Feature activation+0.000
for
Token for
Feature activation+0.000
the
Token the
Feature activation+0.000
world
Token world
Feature activation+0.000
usually
Token usually
Feature activation+0.000
meant
Token meant
Feature activation+0.000
adopting
Token adopting
Feature activation+0.000
not
Token not
Feature activation+0.000
merely
Token merely
Feature activation+0.000
pro
Token pro
Feature activation+1.644
-
Token-
Feature activation+0.543
Israel
TokenIsrael
Feature activation+2.293
talking
Token talking
Feature activation+0.367
points
Token points
Feature activation+4.117
but
Token but
Feature activation+0.000
for
Token for
Feature activation+0.470
the
Token the
Feature activation+0.304
weight
Token weight
Feature activation+0.005
-
Token-
Feature activation+0.000
loss
Tokenloss
Feature activation+0.241
company
Token company
Feature activation+1.819
,
Token,
Feature activation+3.092
she
Token she
Feature activation+0.000
also
Token also
Feature activation+0.000
owns
Token owns
Feature activation+0.000
10
Token 10
Feature activation+0.000

INTERVAL 0.653 - 1.307
CONTAINS 0.036%

hospital
Token hospital
Feature activation+0.000
doctors
Token doctors
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
sat
Token sat
Feature activation+1.121
over
Token over
Feature activation+0.833
oxygen
Token oxygen
Feature activation+2.507
shortage
Token shortage
Feature activation+3.376
for
Token for
Feature activation+0.667
almost
Token almost
Feature activation+0.245
a
Token a
Feature activation+0.000
disrespectful
Token disrespectful
Feature activation+0.000
.
Token.
Feature activation+0.000
And
Token And
Feature activation+0.000
not
Token not
Feature activation+0.000
just
Token just
Feature activation+0.000
Mormon
Token Mormon
Feature activation+0.834
doctrine
Token doctrine
Feature activation+1.067
:
Token:
Feature activation+0.000
any
Token any
Feature activation+0.000
non
Token non
Feature activation+0.000
-
Token-
Feature activation+0.000
.
Token.
Feature activation+0.000
El
Token El
Feature activation+0.000
sie
Tokensie
Feature activation+0.000
tried
Token tried
Feature activation+0.000
to
Token to
Feature activation+0.111
walk
Token walk
Feature activation+0.952
with
Token with
Feature activation+0.829
him
Token him
Feature activation+1.159
and
Token and
Feature activation+0.000
hold
Token hold
Feature activation+0.283
hands
Token hands
Feature activation+0.508
its
Token its
Feature activation+0.000
report
Token report
Feature activation+0.000
on
Token on
Feature activation+0.381
information
Token information
Feature activation+0.595
supplied
Token supplied
Feature activation+0.460
by
Token by
Feature activation+0.739
the
Token the
Feature activation+0.393
Saudi
Token Saudi
Feature activation+0.012
-
Token-
Feature activation+0.000
backed
Tokenbacked
Feature activation+0.000
Yemeni
Token Yemeni
Feature activation+0.000
is
Token is
Feature activation+0.000
not
Token not
Feature activation+0.000
to
Token to
Feature activation+0.000
add
Token add
Feature activation+0.907
new
Token new
Feature activation+0.580
devices
Token devices
Feature activation+1.083
as
Token as
Feature activation+0.118
they
Token they
Feature activation+0.406
appear
Token appear
Feature activation+1.112
on
Token on
Feature activation+0.478
the
Token the
Feature activation+0.208

INTERVAL 0.000 - 0.653
CONTAINS 99.919%

time
Token time
Feature activation+0.000
to
Token to
Feature activation+0.000
create
Token create
Feature activation+0.000
the
Token the
Feature activation+0.000
app
Token app
Feature activation+0.000
.
Token.
Feature activation+0.000
I
Token I
Feature activation+0.000
will
Token will
Feature activation+0.000
name
Token name
Feature activation+0.000
this
Token this
Feature activation+0.000
app
Token app
Feature activation+0.000
order
Token order
Feature activation+0.000
to
Token to
Feature activation+0.000
be
Token be
Feature activation+0.000
an
Token an
Feature activation+0.000
informative
Token informative
Feature activation+0.000
speaker
Token speaker
Feature activation+0.000
and
Token and
Feature activation+0.000
in
Token in
Feature activation+0.000
order
Token order
Feature activation+0.000
to
Token to
Feature activation+0.000
have
Token have
Feature activation+0.000
.
Token.
Feature activation+0.000
They
Token They
Feature activation+0.000
were
Token were
Feature activation+0.000
looking
Token looking
Feature activation+0.000
for
Token for
Feature activation+0.000
ser
Token ser
Feature activation+0.000
o
Tokeno
Feature activation+0.000
-
Token-
Feature activation+0.000
different
Tokendifferent
Feature activation+0.000
couples
Token couples
Feature activation+0.000
where
Token where
Feature activation+0.000
along
Token along
Feature activation+0.000
the
Token the
Feature activation+0.000
border
Token border
Feature activation+0.000
,
Token,
Feature activation+0.000
though
Token though
Feature activation+0.000
it
Token it
Feature activation+0.000
transports
Token transports
Feature activation+0.000
more
Token more
Feature activation+0.000
serious
Token serious
Feature activation+0.000
cases
Token cases
Feature activation+0.000
to
Token to
Feature activation+0.000
it
Token it
Feature activation+0.000
adds
Token adds
Feature activation+0.000
up
Token up
Feature activation+0.000
to
Token to
Feature activation+0.000
make
Token make
Feature activation+0.000
a
Token a
Feature activation+0.000
huge
Token huge
Feature activation+0.000
difference
Token difference
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000

BOTTOM ACTIVATIONS
MIN = 0.000

atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
&
Token &
Feature activation+0.000
Development
Token Development
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
&
Token &
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
rik
Tokenrik
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
rik
Tokenrik
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000

Top feature 3 in H8.5: (feature 15566)

TOP ACTIVATIONS
MAX = 8.163

serves
Token serves
Feature activation+0.000
low
Token low
Feature activation+0.000
-
Token-
Feature activation+0.000
income
Tokenincome
Feature activation+0.000
Americans
Token Americans
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+8.163
or
Token or
Feature activation+0.405
for
Token for
Feature activation+0.237
tax
Token tax
Feature activation+0.000
subsidies
Token subsidies
Feature activation+0.000
to
Token to
Feature activation+0.000
up
Token up
Feature activation+4.048
her
Token her
Feature activation+0.637
rights
Token rights
Feature activation+6.643
to
Token to
Feature activation+0.823
the
Token the
Feature activation+0.233
child
Token child
Feature activation+7.301
or
Token or
Feature activation+0.000
leave
Token leave
Feature activation+0.000
Canada
Token Canada
Feature activation+0.000
.
Token.
Feature activation+0.000
Neither
Token Neither
Feature activation+0.199
Web
Token Web
Feature activation+0.140
GL
TokenGL
Feature activation+3.672
on
Token on
Feature activation+1.156
their
Token their
Feature activation+0.070
hardware
Token hardware
Feature activation+4.103
,
Token,
Feature activation+7.299
or
Token or
Feature activation+0.000
remain
Token remain
Feature activation+0.000
insecure
Token insecure
Feature activation+0.000
if
Token if
Feature activation+0.000
a
Token a
Feature activation+0.000
the
Token the
Feature activation+0.000
sewer
Token sewer
Feature activation+0.000
shows
Token shows
Feature activation+0.000
either
Token either
Feature activation+0.000
laz
Token laz
Feature activation+1.976
iness
Tokeniness
Feature activation+6.840
or
Token or
Feature activation+0.000
corruption
Token corruption
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
police
Token police
Feature activation+0.000
he
Token he
Feature activation+0.000
would
Token would
Feature activation+0.000
either
Token either
Feature activation+0.000
be
Token be
Feature activation+0.466
still
Token still
Feature activation+2.884
born
Tokenborn
Feature activation+6.837
or
Token or
Feature activation+0.000
die
Token die
Feature activation+0.000
very
Token very
Feature activation+0.000
shortly
Token shortly
Feature activation+0.000
afterwards
Token afterwards
Feature activation+0.000
inous
Tokeninous
Feature activation+0.000
.
Token.
Feature activation+0.000
Muslims
Token Muslims
Feature activation+0.000
are
Token are
Feature activation+0.000
either
Token either
Feature activation+0.000
silent
Token silent
Feature activation+6.652
or
Token or
Feature activation+0.000
putting
Token putting
Feature activation+0.000
up
Token up
Feature activation+0.000
their
Token their
Feature activation+0.000
own
Token own
Feature activation+0.000
to
Token to
Feature activation+0.000
either
Token either
Feature activation+0.000
give
Token give
Feature activation+1.509
up
Token up
Feature activation+4.048
her
Token her
Feature activation+0.637
rights
Token rights
Feature activation+6.643
to
Token to
Feature activation+0.823
the
Token the
Feature activation+0.233
child
Token child
Feature activation+7.301
or
Token or
Feature activation+0.000
leave
Token leave
Feature activation+0.000
energy
Token energy
Feature activation+1.774
,
Token,
Feature activation+2.382
finance
Token finance
Feature activation+0.855
and
Token and
Feature activation+0.000
telecom
Token telecom
Feature activation+0.252
,
Token,
Feature activation+6.368
or
Token or
Feature activation+1.051
controlling
Token controlling
Feature activation+0.000
capital
Token capital
Feature activation+0.000
and
Token and
Feature activation+0.000
foreign
Token foreign
Feature activation+0.000
produced
Token produced
Feature activation+0.000
art
Token art
Feature activation+0.583
of
Token of
Feature activation+0.000
this
Token this
Feature activation+0.000
type
Token type
Feature activation+3.563
,
Token,
Feature activation+6.041
and
Token and
Feature activation+4.275
did
Token did
Feature activation+0.000
so
Token so
Feature activation+1.262
within
Token within
Feature activation+0.000
the
Token the
Feature activation+0.000
to
Token to
Feature activation+3.082
travel
Token travel
Feature activation+3.277
abroad
Token abroad
Feature activation+4.947
during
Token during
Feature activation+2.478
holiday
Token holiday
Feature activation+1.365
season
Token season
Feature activation+5.978
or
Token or
Feature activation+0.000
spend
Token spend
Feature activation+0.000
a
Token a
Feature activation+0.000
few
Token few
Feature activation+0.000
days
Token days
Feature activation+0.000
it
Token it
Feature activation+0.000
,
Token,
Feature activation+0.000
either
Token either
Feature activation+0.000
by
Token by
Feature activation+0.000
debating
Token debating
Feature activation+3.052
it
Token it
Feature activation+5.914
or
Token or
Feature activation+0.000
voting
Token voting
Feature activation+0.000
it
Token it
Feature activation+0.000
down
Token down
Feature activation+0.000
.
Token.
Feature activation+0.000
lifestyle
Token lifestyle
Feature activation+0.000
and
Token and
Feature activation+0.000
either
Token either
Feature activation+0.000
start
Token start
Feature activation+1.138
a
Token a
Feature activation+0.759
business
Token business
Feature activation+5.908
or
Token or
Feature activation+0.000
become
Token become
Feature activation+0.000
that
Token that
Feature activation+0.000
free
Token free
Feature activation+0.000
agent
Token agent
Feature activation+0.000
other
Token other
Feature activation+0.000
projects
Token projects
Feature activation+0.000
were
Token were
Feature activation+0.000
either
Token either
Feature activation+0.000
disappoint
Token disappoint
Feature activation+0.815
ments
Tokenments
Feature activation+5.698
or
Token or
Feature activation+0.000
were
Token were
Feature activation+0.000
frustrated
Token frustrated
Feature activation+0.000
due
Token due
Feature activation+0.000
to
Token to
Feature activation+0.000
of
Token of
Feature activation+1.773
the
Token the
Feature activation+0.128
scene
Token scene
Feature activation+4.968
dedicated
Token dedicated
Feature activation+0.888
to
Token to
Feature activation+0.791
him
Token him
Feature activation+5.689
or
Token or
Feature activation+0.000
becomes
Token becomes
Feature activation+0.000
an
Token an
Feature activation+0.000
inconvenience
Token inconvenience
Feature activation+0.000
.
Token.
Feature activation+0.000
ized
Tokenized
Feature activation+0.000
films
Token films
Feature activation+0.000
that
Token that
Feature activation+0.000
were
Token were
Feature activation+0.000
either
Token either
Feature activation+0.000
killed
Token killed
Feature activation+5.666
at
Token at
Feature activation+2.375
the
Token the
Feature activation+1.211
treatment
Token treatment
Feature activation+2.485
or
Token or
Feature activation+0.000
script
Token script
Feature activation+0.000
Family
Token Family
Feature activation+0.000
Association
Token Association
Feature activation+0.000
is
Token is
Feature activation+0.000
either
Token either
Feature activation+0.000
consciously
Token consciously
Feature activation+4.618
lying
Token lying
Feature activation+5.588
or
Token or
Feature activation+0.000
has
Token has
Feature activation+0.000
absolutely
Token absolutely
Feature activation+0.000
failed
Token failed
Feature activation+0.000
to
Token to
Feature activation+0.000
it
Token it
Feature activation+0.000
will
Token will
Feature activation+0.000
either
Token either
Feature activation+0.000
be
Token be
Feature activation+1.565
your
Token your
Feature activation+0.000
friend
Token friend
Feature activation+5.547
lies
Tokenlies
Feature activation+5.041
or
Token or
Feature activation+0.000
the
Token the
Feature activation+0.000
enemy
Token enemy
Feature activation+0.000
who
Token who
Feature activation+0.000
issue
Token issue
Feature activation+0.000
that
Token that
Feature activation+0.000
was
Token was
Feature activation+0.000
either
Token either
Feature activation+0.000
left
Token left
Feature activation+4.411
out
Token out
Feature activation+5.318
of
Token of
Feature activation+3.323
the
Token the
Feature activation+1.020
presidential
Token presidential
Feature activation+0.064
debates
Token debates
Feature activation+4.714
and
Token and
Feature activation+2.480
the
Token the
Feature activation+0.000
US
Token US
Feature activation+0.000
either
Token either
Feature activation+0.000
still
Token still
Feature activation+3.534
getting
Token getting
Feature activation+3.045
now
Token now
Feature activation+5.256
or
Token or
Feature activation+0.142
digging
Token digging
Feature activation+0.455
out
Token out
Feature activation+0.016
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
O
TokenO
Feature activation+0.000
.
Token.
Feature activation+0.000
s
Tokens
Feature activation+0.000
were
Token were
Feature activation+0.000
either
Token either
Feature activation+0.000
confused
Token confused
Feature activation+5.119
or
Token or
Feature activation+0.000
lying
Token lying
Feature activation+0.000
:
Token:
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000

Top DFA by src position
MAX = 8.473

Instead
Token Instead
Feature activation-0.023
Top resid features:
,
Token,
Feature activation+0.002
Top resid features:
most
Token most
Feature activation-0.002
Top resid features:
will
Token will
Feature activation-0.001
Top resid features:
qualify
Token qualify
Feature activation+0.048
Top resid features:
either
Token either
Feature activation+8.473
Top resid features:
for
Token for
Feature activation+0.183
Top resid features:
Medicaid
Token Medicaid
Feature activation+0.208
Top resid features:
âĢĶ
Token âĢĶ
Feature activation+0.007
Top resid features:
the
Token the
Feature activation+0.010
Top resid features:
public
Token public
Feature activation+0.015
Top resid features:
now
Token now
Feature activation+0.036
Top resid features:
the
Token the
Feature activation+0.004
Top resid features:
mother
Token mother
Feature activation+0.004
Top resid features:
has
Token has
Feature activation+0.009
Top resid features:
to
Token to
Feature activation+0.031
Top resid features:
either
Token either
Feature activation+7.496
Top resid features:
give
Token give
Feature activation+0.071
Top resid features:
up
Token up
Feature activation-0.014
Top resid features:
her
Token her
Feature activation+0.020
Top resid features:
rights
Token rights
Feature activation+0.055
Top resid features:
to
Token to
Feature activation+0.088
Top resid features:
Windows
Token Windows
Feature activation+0.005
Top resid features:
Update
Token Update
Feature activation-0.001
Top resid features:
),
Token),
Feature activation-0.014
Top resid features:
users
Token users
Feature activation+0.014
Top resid features:
may
Token may
Feature activation-0.018
Top resid features:
either
Token either
Feature activation+8.133
Top resid features:
choose
Token choose
Feature activation-0.036
Top resid features:
to
Token to
Feature activation+0.038
Top resid features:
override
Token override
Feature activation-0.028
Top resid features:
the
Token the
Feature activation-0.015
Top resid features:
protection
Token protection
Feature activation+0.004
Top resid features:
army
Token army
Feature activation+0.008
Top resid features:
into
Token into
Feature activation+0.001
Top resid features:
the
Token the
Feature activation-0.002
Top resid features:
sewer
Token sewer
Feature activation+0.007
Top resid features:
shows
Token shows
Feature activation+0.028
Top resid features:
either
Token either
Feature activation+7.546
Top resid features:
laz
Token laz
Feature activation+0.016
Top resid features:
iness
Tokeniness
Feature activation-0.020
Top resid features:
or
Token or
Feature activation+0.000
Top resid features:
corruption
Token corruption
Feature activation+0.000
Top resid features:
in
Token in
Feature activation+0.000
Top resid features:
parents
Token parents
Feature activation+0.010
Top resid features:
were
Token were
Feature activation-0.002
Top resid features:
told
Token told
Feature activation+0.003
Top resid features:
he
Token he
Feature activation-0.019
Top resid features:
would
Token would
Feature activation-0.017
Top resid features:
either
Token either
Feature activation+7.603
Top resid features:
be
Token be
Feature activation+0.175
Top resid features:
still
Token still
Feature activation+0.004
Top resid features:
born
Tokenborn
Feature activation+0.028
Top resid features:
or
Token or
Feature activation+0.000
Top resid features:
die
Token die
Feature activation+0.000
Top resid features:
mut
Token mut
Feature activation+0.002
Top resid features:
inous
Tokeninous
Feature activation-0.001
Top resid features:
.
Token.
Feature activation-0.006
Top resid features:
Muslims
Token Muslims
Feature activation+0.004
Top resid features:
are
Token are
Feature activation-0.017
Top resid features:
either
Token either
Feature activation+7.269
Top resid features:
silent
Token silent
Feature activation+0.052
Top resid features:
or
Token or
Feature activation+0.000
Top resid features:
putting
Token putting
Feature activation+0.000
Top resid features:
up
Token up
Feature activation+0.000
Top resid features:
their
Token their
Feature activation+0.000
Top resid features:
now
Token now
Feature activation+0.017
Top resid features:
the
Token the
Feature activation+0.003
Top resid features:
mother
Token mother
Feature activation+0.011
Top resid features:
has
Token has
Feature activation-0.000
Top resid features:
to
Token to
Feature activation+0.034
Top resid features:
either
Token either
Feature activation+6.906
Top resid features:
give
Token give
Feature activation+0.101
Top resid features:
up
Token up
Feature activation-0.022
Top resid features:
her
Token her
Feature activation-0.021
Top resid features:
rights
Token rights
Feature activation+0.109
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
sectors
Token sectors
Feature activation-0.016
Top resid features:
of
Token of
Feature activation+0.002
Top resid features:
the
Token the
Feature activation-0.001
Top resid features:
economy
Token economy
Feature activation+0.007
Top resid features:
by
Token by
Feature activation-0.037
Top resid features:
either
Token either
Feature activation+7.061
Top resid features:
monopol
Token monopol
Feature activation+0.023
Top resid features:
izing
Tokenizing
Feature activation-0.107
Top resid features:
pillar
Token pillar
Feature activation+0.004
Top resid features:
industries
Token industries
Feature activation+0.028
Top resid features:
such
Token such
Feature activation+0.009
Top resid features:
âĢ
TokenâĢ
Feature activation+0.015
Top resid features:
Ļ
TokenĻ
Feature activation+0.015
Top resid features:
s
Tokens
Feature activation+0.030
Top resid features:
followers
Token followers
Feature activation-0.004
Top resid features:
are
Token are
Feature activation-0.007
Top resid features:
either
Token either
Feature activation+6.942
Top resid features:
unaware
Token unaware
Feature activation-0.054
Top resid features:
that
Token that
Feature activation-0.017
Top resid features:
some
Token some
Feature activation-0.022
Top resid features:
of
Token of
Feature activation-0.008
Top resid features:
the
Token the
Feature activation-0.011
Top resid features:
the
Token the
Feature activation-0.003
Top resid features:
year
Token year
Feature activation-0.002
Top resid features:
when
Token when
Feature activation-0.024
Top resid features:
Australian
Token Australian
Feature activation+0.013
Top resid features:
families
Token families
Feature activation+0.041
Top resid features:
either
Token either
Feature activation+6.653
Top resid features:
choose
Token choose
Feature activation+0.113
Top resid features:
to
Token to
Feature activation+0.126
Top resid features:
travel
Token travel
Feature activation+0.019
Top resid features:
abroad
Token abroad
Feature activation+0.030
Top resid features:
during
Token during
Feature activation-0.123
Top resid features:
days
Token days
Feature activation-0.007
Top resid features:
to
Token to
Feature activation-0.004
Top resid features:
address
Token address
Feature activation+0.001
Top resid features:
it
Token it
Feature activation+0.007
Top resid features:
,
Token,
Feature activation-0.021
Top resid features:
either
Token either
Feature activation+5.672
Top resid features:
by
Token by
Feature activation+0.550
Top resid features:
debating
Token debating
Feature activation+0.296
Top resid features:
it
Token it
Feature activation+0.025
Top resid features:
or
Token or
Feature activation+0.000
Top resid features:
voting
Token voting
Feature activation+0.000
Top resid features:
people
Token people
Feature activation-0.008
Top resid features:
choose
Token choose
Feature activation-0.023
Top resid features:
the
Token the
Feature activation-0.034
Top resid features:
lifestyle
Token lifestyle
Feature activation+0.039
Top resid features:
and
Token and
Feature activation-0.025
Top resid features:
either
Token either
Feature activation+6.699
Top resid features:
start
Token start
Feature activation+0.085
Top resid features:
a
Token a
Feature activation-0.031
Top resid features:
business
Token business
Feature activation+0.058
Top resid features:
or
Token or
Feature activation+0.000
Top resid features:
become
Token become
Feature activation+0.000
Top resid features:
),
Token),
Feature activation-0.014
Top resid features:
many
Token many
Feature activation-0.003
Top resid features:
other
Token other
Feature activation-0.003
Top resid features:
projects
Token projects
Feature activation+0.001
Top resid features:
were
Token were
Feature activation+0.010
Top resid features:
either
Token either
Feature activation+6.527
Top resid features:
disappoint
Token disappoint
Feature activation+0.032
Top resid features:
ments
Tokenments
Feature activation+0.028
Top resid features:
or
Token or
Feature activation+0.000
Top resid features:
were
Token were
Feature activation+0.000
Top resid features:
frustrated
Token frustrated
Feature activation+0.000
Top resid features:
affects
Token affects
Feature activation+0.003
Top resid features:
one
Token one
Feature activation-0.021
Top resid features:
character
Token character
Feature activation-0.037
Top resid features:
,
Token,
Feature activation-0.022
Top resid features:
who
Token who
Feature activation-0.035
Top resid features:
either
Token either
Feature activation+6.304
Top resid features:
gets
Token gets
Feature activation+0.076
Top resid features:
part
Token part
Feature activation+0.047
Top resid features:
of
Token of
Feature activation+0.005
Top resid features:
the
Token the
Feature activation+0.002
Top resid features:
scene
Token scene
Feature activation-0.002
Top resid features:
unreal
Token unreal
Feature activation+0.014
Top resid features:
ized
Tokenized
Feature activation+0.020
Top resid features:
films
Token films
Feature activation-0.011
Top resid features:
that
Token that
Feature activation-0.027
Top resid features:
were
Token were
Feature activation+0.031
Top resid features:
either
Token either
Feature activation+6.368
Top resid features:
killed
Token killed
Feature activation+0.101
Top resid features:
at
Token at
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
treatment
Token treatment
Feature activation+0.000
Top resid features:
or
Token or
Feature activation+0.000
Top resid features:
the
Token the
Feature activation-0.001
Top resid features:
American
Token American
Feature activation-0.003
Top resid features:
Family
Token Family
Feature activation-0.001
Top resid features:
Association
Token Association
Feature activation+0.013
Top resid features:
is
Token is
Feature activation-0.025
Top resid features:
either
Token either
Feature activation+6.323
Top resid features:
consciously
Token consciously
Feature activation+0.098
Top resid features:
lying
Token lying
Feature activation+0.035
Top resid features:
or
Token or
Feature activation+0.000
Top resid features:
has
Token has
Feature activation+0.000
Top resid features:
absolutely
Token absolutely
Feature activation+0.000
Top resid features:
the
Token the
Feature activation-0.004
Top resid features:
match
Token match
Feature activation-0.014
Top resid features:
,
Token,
Feature activation-0.018
Top resid features:
it
Token it
Feature activation-0.002
Top resid features:
will
Token will
Feature activation+0.010
Top resid features:
either
Token either
Feature activation+6.363
Top resid features:
be
Token be
Feature activation+0.185
Top resid features:
your
Token your
Feature activation+0.042
Top resid features:
friend
Token friend
Feature activation-0.009
Top resid features:
lies
Tokenlies
Feature activation+0.000
Top resid features:
or
Token or
Feature activation+0.000
Top resid features:
hugely
Token hugely
Feature activation+0.016
Top resid features:
important
Token important
Feature activation+0.008
Top resid features:
issue
Token issue
Feature activation+0.011
Top resid features:
that
Token that
Feature activation+0.004
Top resid features:
was
Token was
Feature activation-0.010
Top resid features:
either
Token either
Feature activation+5.901
Top resid features:
left
Token left
Feature activation+0.106
Top resid features:
out
Token out
Feature activation+0.027
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
presidential
Token presidential
Feature activation+0.000
Top resid features:
swat
Token swat
Feature activation+0.009
Top resid features:
hes
Tokenhes
Feature activation+0.002
Top resid features:
of
Token of
Feature activation+0.004
Top resid features:
the
Token the
Feature activation-0.004
Top resid features:
US
Token US
Feature activation-0.002
Top resid features:
either
Token either
Feature activation+6.009
Top resid features:
still
Token still
Feature activation+0.002
Top resid features:
getting
Token getting
Feature activation+0.054
Top resid features:
now
Token now
Feature activation+0.007
Top resid features:
or
Token or
Feature activation+0.000
Top resid features:
digging
Token digging
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.001
Top resid features:
O
TokenO
Feature activation-0.009
Top resid features:
.
Token.
Feature activation+0.010
Top resid features:
s
Tokens
Feature activation-0.017
Top resid features:
were
Token were
Feature activation-0.036
Top resid features:
either
Token either
Feature activation+5.896
Top resid features:
confused
Token confused
Feature activation+0.004
Top resid features:
or
Token or
Feature activation+0.000
Top resid features:
lying
Token lying
Feature activation+0.000
Top resid features:
:
Token:
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:

Decoder Weights Distribution

Head 0: 0.16

Head 1: 0.01

Head 2: 0.04

Head 3: 0.02

Head 4: 0.04

Head 5: 0.55

Head 6: 0.02

Head 7: 0.04

Head 8: 0.03

Head 9: 0.02

Head 10: 0.04

Head 11: 0.02

Positive logits

or3.74

nor2.48

OR2.28

Or2.20

whatsoever2.01

either2.00

Either1.92

Or1.85

1.71

pree1.69

themselves1.69

commits1.65

gobl1.58

(>1.57

/"1.56

Alternatively1.55

Nope1.53

outright1.52

withholding1.50

ourselves1.47

Negative logits

.).-2.65

.}-2.30

.]-2.18

}.-2.14

%.-2.12

].-2.12

!.-2.02

`.-1.99

.)-1.97

].-1.94

"]=>-1.91

>.-1.91

.�-1.85

$.-1.80

.''.-1.78

.")-1.78

+.-1.75

Owl-1.72

]."-1.72

veyard-1.68

INTERVAL 7.347 - 8.163
CONTAINS 0.000%

serves
Token serves
Feature activation+0.000
low
Token low
Feature activation+0.000
-
Token-
Feature activation+0.000
income
Tokenincome
Feature activation+0.000
Americans
Token Americans
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+8.163
or
Token or
Feature activation+0.405
for
Token for
Feature activation+0.237
tax
Token tax
Feature activation+0.000
subsidies
Token subsidies
Feature activation+0.000
to
Token to
Feature activation+0.000

INTERVAL 6.530 - 7.347
CONTAINS 0.001%

Web
Token Web
Feature activation+0.140
GL
TokenGL
Feature activation+3.672
on
Token on
Feature activation+1.156
their
Token their
Feature activation+0.070
hardware
Token hardware
Feature activation+4.103
,
Token,
Feature activation+7.299
or
Token or
Feature activation+0.000
remain
Token remain
Feature activation+0.000
insecure
Token insecure
Feature activation+0.000
if
Token if
Feature activation+0.000
a
Token a
Feature activation+0.000
to
Token to
Feature activation+0.000
either
Token either
Feature activation+0.000
give
Token give
Feature activation+1.509
up
Token up
Feature activation+4.048
her
Token her
Feature activation+0.637
rights
Token rights
Feature activation+6.643
to
Token to
Feature activation+0.823
the
Token the
Feature activation+0.233
child
Token child
Feature activation+7.301
or
Token or
Feature activation+0.000
leave
Token leave
Feature activation+0.000
the
Token the
Feature activation+0.000
sewer
Token sewer
Feature activation+0.000
shows
Token shows
Feature activation+0.000
either
Token either
Feature activation+0.000
laz
Token laz
Feature activation+1.976
iness
Tokeniness
Feature activation+6.840
or
Token or
Feature activation+0.000
corruption
Token corruption
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
police
Token police
Feature activation+0.000
up
Token up
Feature activation+4.048
her
Token her
Feature activation+0.637
rights
Token rights
Feature activation+6.643
to
Token to
Feature activation+0.823
the
Token the
Feature activation+0.233
child
Token child
Feature activation+7.301
or
Token or
Feature activation+0.000
leave
Token leave
Feature activation+0.000
Canada
Token Canada
Feature activation+0.000
.
Token.
Feature activation+0.000
Neither
Token Neither
Feature activation+0.199
he
Token he
Feature activation+0.000
would
Token would
Feature activation+0.000
either
Token either
Feature activation+0.000
be
Token be
Feature activation+0.466
still
Token still
Feature activation+2.884
born
Tokenborn
Feature activation+6.837
or
Token or
Feature activation+0.000
die
Token die
Feature activation+0.000
very
Token very
Feature activation+0.000
shortly
Token shortly
Feature activation+0.000
afterwards
Token afterwards
Feature activation+0.000

INTERVAL 5.714 - 6.530
CONTAINS 0.001%

it
Token it
Feature activation+0.000
,
Token,
Feature activation+0.000
either
Token either
Feature activation+0.000
by
Token by
Feature activation+0.000
debating
Token debating
Feature activation+3.052
it
Token it
Feature activation+5.914
or
Token or
Feature activation+0.000
voting
Token voting
Feature activation+0.000
it
Token it
Feature activation+0.000
down
Token down
Feature activation+0.000
.
Token.
Feature activation+0.000
produced
Token produced
Feature activation+0.000
art
Token art
Feature activation+0.583
of
Token of
Feature activation+0.000
this
Token this
Feature activation+0.000
type
Token type
Feature activation+3.563
,
Token,
Feature activation+6.041
and
Token and
Feature activation+4.275
did
Token did
Feature activation+0.000
so
Token so
Feature activation+1.262
within
Token within
Feature activation+0.000
the
Token the
Feature activation+0.000
to
Token to
Feature activation+3.082
travel
Token travel
Feature activation+3.277
abroad
Token abroad
Feature activation+4.947
during
Token during
Feature activation+2.478
holiday
Token holiday
Feature activation+1.365
season
Token season
Feature activation+5.978
or
Token or
Feature activation+0.000
spend
Token spend
Feature activation+0.000
a
Token a
Feature activation+0.000
few
Token few
Feature activation+0.000
days
Token days
Feature activation+0.000
energy
Token energy
Feature activation+1.774
,
Token,
Feature activation+2.382
finance
Token finance
Feature activation+0.855
and
Token and
Feature activation+0.000
telecom
Token telecom
Feature activation+0.252
,
Token,
Feature activation+6.368
or
Token or
Feature activation+1.051
controlling
Token controlling
Feature activation+0.000
capital
Token capital
Feature activation+0.000
and
Token and
Feature activation+0.000
foreign
Token foreign
Feature activation+0.000
lifestyle
Token lifestyle
Feature activation+0.000
and
Token and
Feature activation+0.000
either
Token either
Feature activation+0.000
start
Token start
Feature activation+1.138
a
Token a
Feature activation+0.759
business
Token business
Feature activation+5.908
or
Token or
Feature activation+0.000
become
Token become
Feature activation+0.000
that
Token that
Feature activation+0.000
free
Token free
Feature activation+0.000
agent
Token agent
Feature activation+0.000

INTERVAL 4.898 - 5.714
CONTAINS 0.002%

life
Token life
Feature activation+1.667
jeopard
Token jeopard
Feature activation+0.000
ized
Tokenized
Feature activation+4.470
by
Token by
Feature activation+0.948
the
Token the
Feature activation+0.358
police
Token police
Feature activation+4.928
or
Token or
Feature activation+0.000
by
Token by
Feature activation+0.000
catastrophic
Token catastrophic
Feature activation+0.000
climate
Token climate
Feature activation+0.000
change
Token change
Feature activation+0.000
of
Token of
Feature activation+0.000
either
Token either
Feature activation+0.000
cra
Token cra
Feature activation+0.042
ven
Tokenven
Feature activation+2.134
capit
Token capit
Feature activation+0.000
ulation
Tokenulation
Feature activation+5.059
to
Token to
Feature activation+0.875
Beijing
Token Beijing
Feature activation+1.740
or
Token or
Feature activation+0.015
inex
Token inex
Feature activation+0.000
cus
Tokencus
Feature activation+0.000
of
Token of
Feature activation+1.773
the
Token the
Feature activation+0.128
scene
Token scene
Feature activation+4.968
dedicated
Token dedicated
Feature activation+0.888
to
Token to
Feature activation+0.791
him
Token him
Feature activation+5.689
or
Token or
Feature activation+0.000
becomes
Token becomes
Feature activation+0.000
an
Token an
Feature activation+0.000
inconvenience
Token inconvenience
Feature activation+0.000
.
Token.
Feature activation+0.000
piece
Token piece
Feature activation+0.000
of
Token of
Feature activation+0.000
evidence
Token evidence
Feature activation+0.000
establishing
Token establishing
Feature activation+0.000
either
Token either
Feature activation+0.000
guilt
Token guilt
Feature activation+4.936
or
Token or
Feature activation+0.000
innocence
Token innocence
Feature activation+0.000
in
Token in
Feature activation+0.000
DUI
Token DUI
Feature activation+0.000
cases
Token cases
Feature activation+0.000
choice
Token choice
Feature activation+0.953
between
Token between
Feature activation+0.330
starvation
Token starvation
Feature activation+1.400
and
Token and
Feature activation+0.097
surrender
Token surrender
Feature activation+2.502
,
Token,
Feature activation+5.009
or
Token or
Feature activation+0.036
a
Token a
Feature activation+0.000
halt
Token halt
Feature activation+0.000
to
Token to
Feature activation+0.000
fighting
Token fighting
Feature activation+0.000

INTERVAL 4.081 - 4.898
CONTAINS 0.003%

they
Token they
Feature activation+0.000
need
Token need
Feature activation+0.000
either
Token either
Feature activation+0.000
a
Token a
Feature activation+0.142
wholesale
Token wholesale
Feature activation+1.144
rethink
Token rethink
Feature activation+4.533
of
Token of
Feature activation+0.000
approach
Token approach
Feature activation+1.638
or
Token or
Feature activation+0.000
a
Token a
Feature activation+0.000
root
Token root
Feature activation+0.000
of
Token of
Feature activation+0.316
me
Token me
Feature activation+1.684
on
Token on
Feature activation+0.000
the
Token the
Feature activation+0.000
BBC
Token BBC
Feature activation+0.976
'
Token'
Feature activation+4.243
or
Token or
Feature activation+0.000
'
Token '
Feature activation+0.000
my
Tokenmy
Feature activation+0.000
side
Token side
Feature activation+0.000
of
Token of
Feature activation+0.000
those
Token those
Feature activation+0.000
things
Token things
Feature activation+0.000
are
Token are
Feature activation+0.000
either
Token either
Feature activation+0.000
never
Token never
Feature activation+4.003
stored
Token stored
Feature activation+4.420
,
Token,
Feature activation+4.352
or
Token or
Feature activation+0.000
deleted
Token deleted
Feature activation+0.000
when
Token when
Feature activation+0.000
you
Token you
Feature activation+0.000
of
Token of
Feature activation+0.000
poor
Token poor
Feature activation+0.000
people
Token people
Feature activation+0.000
either
Token either
Feature activation+0.000
be
Token be
Feature activation+2.546
slashed
Token slashed
Feature activation+4.251
or
Token or
Feature activation+0.000
eliminated
Token eliminated
Feature activation+0.000
.
Token.
Feature activation+0.000
The
Token The
Feature activation+0.000
Obama
Token Obama
Feature activation+0.000
encouraging
Token encouraging
Feature activation+1.544
mil
Token mil
Feature activation+0.000
let
Tokenlet
Feature activation+1.268
domest
Token domest
Feature activation+0.000
ication
Tokenication
Feature activation+3.329
,
Token,
Feature activation+4.807
or
Token or
Feature activation+0.033
early
Token early
Feature activation+0.000
mil
Token mil
Feature activation+0.000
let
Tokenlet
Feature activation+0.000
farming
Token farming
Feature activation+0.000

INTERVAL 3.265 - 4.081
CONTAINS 0.003%

people
Token people
Feature activation+0.000
have
Token have
Feature activation+0.000
to
Token to
Feature activation+0.000
be
Token be
Feature activation+0.000
either
Token either
Feature activation+0.000
straight
Token straight
Feature activation+3.673
or
Token or
Feature activation+0.000
gay
Token gay
Feature activation+0.000
is
Token is
Feature activation+0.000
pervasive
Token pervasive
Feature activation+0.000
and
Token and
Feature activation+0.000
el
Tokenel
Feature activation+0.000
and
Token and
Feature activation+0.000
adding
Token adding
Feature activation+0.000
either
Token either
Feature activation+0.000
coconut
Token coconut
Feature activation+2.594
milk
Token milk
Feature activation+3.621
or
Token or
Feature activation+0.000
almond
Token almond
Feature activation+0.000
jelly
Token jelly
Feature activation+0.000
,
Token,
Feature activation+0.000
called
Token called
Feature activation+0.000
that
Token that
Feature activation+0.000
was
Token was
Feature activation+0.000
either
Token either
Feature activation+0.000
left
Token left
Feature activation+4.411
out
Token out
Feature activation+5.318
of
Token of
Feature activation+3.323
the
Token the
Feature activation+1.020
presidential
Token presidential
Feature activation+0.064
debates
Token debates
Feature activation+4.714
and
Token and
Feature activation+2.480
campaigns
Token campaigns
Feature activation+2.768
done
Token done
Feature activation+0.000
.
Token.
Feature activation+0.000
Either
Token Either
Feature activation+0.000
through
Token through
Feature activation+0.595
medical
Token medical
Feature activation+0.827
treatment
Token treatment
Feature activation+3.759
or
Token or
Feature activation+0.000
divine
Token divine
Feature activation+0.000
intervention
Token intervention
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
overall
Token overall
Feature activation+0.000
-
Token -
Feature activation+0.000
either
Token either
Feature activation+0.000
for
Token for
Feature activation+2.442
security
Token security
Feature activation+2.083
forces
Token forces
Feature activation+4.077
,
Token,
Feature activation+1.758
civilians
Token civilians
Feature activation+0.969
or
Token or
Feature activation+0.000
Islamic
Token Islamic
Feature activation+0.000
State
Token State
Feature activation+0.000

INTERVAL 2.449 - 3.265
CONTAINS 0.004%

commit
Token commit
Feature activation+0.000
to
Token to
Feature activation+0.000
either
Token either
Feature activation+0.000
a
Token a
Feature activation+0.064
strong
Token strong
Feature activation+1.159
message
Token message
Feature activation+2.692
or
Token or
Feature activation+0.000
solutions
Token solutions
Feature activation+0.000
.
Token.
Feature activation+0.000
Again
Token Again
Feature activation+0.000
,
Token,
Feature activation+0.000
be
Token be
Feature activation+0.584
free
Token free
Feature activation+1.563
for
Token for
Feature activation+0.813
a
Token a
Feature activation+0.955
short
Token short
Feature activation+0.639
period
Token period
Feature activation+2.576
,
Token,
Feature activation+3.391
or
Token or
Feature activation+0.000
potentially
Token potentially
Feature activation+0.000
cost
Token cost
Feature activation+0.000
a
Token a
Feature activation+0.000
It
Token It
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
either
Token either
Feature activation+0.000
honor
Token honor
Feature activation+2.984
or
Token or
Feature activation+0.000
shame
Token shame
Feature activation+0.000
,
Token,
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
year
Token year
Feature activation+0.000
when
Token when
Feature activation+0.000
Australian
Token Australian
Feature activation+0.000
families
Token families
Feature activation+0.000
either
Token either
Feature activation+0.000
choose
Token choose
Feature activation+3.204
to
Token to
Feature activation+3.082
travel
Token travel
Feature activation+3.277
abroad
Token abroad
Feature activation+4.947
during
Token during
Feature activation+2.478
holiday
Token holiday
Feature activation+1.365
single
Token single
Feature activation+0.000
origin
Token origin
Feature activation+0.000
,
Token,
Feature activation+0.000
with
Token with
Feature activation+0.000
either
Token either
Feature activation+0.000
rice
Token rice
Feature activation+3.019
agriculture
Token agriculture
Feature activation+3.429
spreading
Token spreading
Feature activation+2.867
north
Token north
Feature activation+2.671
and
Token and
Feature activation+0.810
encouraging
Token encouraging
Feature activation+1.544

INTERVAL 1.633 - 2.449
CONTAINS 0.007%

Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.358
pumping
Token pumping
Feature activation+0.000
up
Token up
Feature activation+0.273
our
Token our
Feature activation+0.000
guys
Token guys
Feature activation+1.677
or
Token or
Feature activation+0.000
ch
Token ch
Feature activation+0.000
ir
Tokenir
Feature activation+0.000
ping
Tokenping
Feature activation+0.000
the
Token the
Feature activation+0.000
It
Token It
Feature activation+0.000
will
Token will
Feature activation+0.000
either
Token either
Feature activation+0.000
be
Token be
Feature activation+2.284
sold
Token sold
Feature activation+3.339
delivered
Token delivered
Feature activation+1.813
,
Token,
Feature activation+2.937
or
Token or
Feature activation+0.000
not
Token not
Feature activation+0.000
..
Token..
Feature activation+0.000
in
Tokenin
Feature activation+0.000
a
Token a
Feature activation+0.000
major
Token major
Feature activation+0.000
threat
Token threat
Feature activation+0.000
in
Token in
Feature activation+0.000
either
Token either
Feature activation+0.000
Pakistan
Token Pakistan
Feature activation+2.238
or
Token or
Feature activation+0.000
Afghanistan
Token Afghanistan
Feature activation+0.000
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
group
Token group
Feature activation+0.000
a
Token a
Feature activation+0.000
gesture
Token gesture
Feature activation+0.000
of
Token of
Feature activation+0.000
either
Token either
Feature activation+0.000
cra
Token cra
Feature activation+0.042
ven
Tokenven
Feature activation+2.134
capit
Token capit
Feature activation+0.000
ulation
Tokenulation
Feature activation+5.059
to
Token to
Feature activation+0.875
Beijing
Token Beijing
Feature activation+1.740
or
Token or
Feature activation+0.015
izing
Tokenizing
Feature activation+4.835
pillar
Token pillar
Feature activation+0.371
industries
Token industries
Feature activation+3.681
such
Token such
Feature activation+0.099
as
Token as
Feature activation+0.288
energy
Token energy
Feature activation+1.774
,
Token,
Feature activation+2.382
finance
Token finance
Feature activation+0.855
and
Token and
Feature activation+0.000
telecom
Token telecom
Feature activation+0.252
,
Token,
Feature activation+6.368

INTERVAL 0.816 - 1.633
CONTAINS 0.013%

via
Token via
Feature activation+0.000
USPS
Token USPS
Feature activation+0.000
either
Token either
Feature activation+0.000
via
Token via
Feature activation+0.675
First
Token First
Feature activation+0.142
Class
Token Class
Feature activation+1.539
or
Token or
Feature activation+0.000
Priority
Token Priority
Feature activation+0.000
Mail
Token Mail
Feature activation+0.000
,
Token,
Feature activation+0.000
depending
Token depending
Feature activation+0.000
eligible
Token eligible
Feature activation+0.000
for
Token for
Feature activation+0.000
either
Token either
Feature activation+0.000
Mr
Token Mr
Feature activation+0.000
.
Token.
Feature activation+0.000
Obama
Token Obama
Feature activation+1.383
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
nond
Token nond
Feature activation+0.000
eport
Tokeneport
Feature activation+0.000
-
Token-
Feature activation+0.000
No
TokenNo
Feature activation+0.000
entries
Token entries
Feature activation+0.000
on
Token on
Feature activation+0.000
either
Token either
Feature activation+0.000
mobile
Token mobile
Feature activation+1.082
or
Token or
Feature activation+0.000
handheld
Token handheld
Feature activation+0.000
platforms
Token platforms
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
the
Token the
Feature activation+0.000
ple
Token ple
Feature activation+0.000
b
Tokenb
Feature activation+0.000
isc
Tokenisc
Feature activation+0.000
ite
Tokenite
Feature activation+1.097
,
Token,
Feature activation+1.056
"
Token "
Feature activation+1.119
altern
Tokenaltern
Feature activation+0.000
ate
Tokenate
Feature activation+0.486
sources
Token sources
Feature activation+0.822
of
Token of
Feature activation+0.190
either
Token either
Feature activation+0.000
cra
Token cra
Feature activation+0.042
ven
Tokenven
Feature activation+2.134
capit
Token capit
Feature activation+0.000
ulation
Tokenulation
Feature activation+5.059
to
Token to
Feature activation+0.875
Beijing
Token Beijing
Feature activation+1.740
or
Token or
Feature activation+0.015
inex
Token inex
Feature activation+0.000
cus
Tokencus
Feature activation+0.000
able
Tokenable
Feature activation+0.000

INTERVAL 0.000 - 0.816
CONTAINS 99.966%

than
Token than
Feature activation+0.000
the
Token the
Feature activation+0.000
previous
Token previous
Feature activation+0.000
week
Token week
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
had
Token had
Feature activation+0.000
fewer
Token fewer
Feature activation+0.000
top
Token top
Feature activation+0.000
three
Token three
Feature activation+0.000
teams
Token teams
Feature activation+0.000
on
Token on
Feature activation+0.000
that
Token that
Feature activation+0.000
list
Token list
Feature activation+0.000
.
Token.
Feature activation+0.000
All
Token All
Feature activation+0.000
of
Token of
Feature activation+0.000
these
Token these
Feature activation+0.000
teams
Token teams
Feature activation+0.000
gonna
Token gonna
Feature activation+0.000
like
Token like
Feature activation+0.000
this
Token this
Feature activation+0.000
one
Token one
Feature activation+0.000
.
Token.
Feature activation+0.000
It
Token It
Feature activation+0.000
has
Token has
Feature activation+0.000
all
Token all
Feature activation+0.000
the
Token the
Feature activation+0.000
things
Token things
Feature activation+0.000
I
Token I
Feature activation+0.000
City
Token City
Feature activation+0.000
,
Token,
Feature activation+0.000
were
Token were
Feature activation+0.000
impacted
Token impacted
Feature activation+0.000
by
Token by
Feature activation+0.000
this
Token this
Feature activation+0.000
decision
Token decision
Feature activation+0.000
.
Token.
Feature activation+0.000
IBM
Token IBM
Feature activation+0.000
is
Token is
Feature activation+0.000
hiring
Token hiring
Feature activation+0.000
severe
Token severe
Feature activation+0.000
pain
Token pain
Feature activation+0.000
and
Token and
Feature activation+0.000
to
Token to
Feature activation+0.000
vomit
Token vomit
Feature activation+0.000
repeatedly
Token repeatedly
Feature activation+0.000
.
Token.
Feature activation+0.000
The
Token The
Feature activation+0.000
sheriff
Token sheriff
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000

BOTTOM ACTIVATIONS
MIN = 0.000

atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
&
Token &
Feature activation+0.000
Development
Token Development
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
&
Token &
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
rik
Tokenrik
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
rik
Tokenrik
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000

Top feature 4 in H8.5: (feature 10548)

TOP ACTIVATIONS
MAX = 5.265

âĢĶ
Token âĢĶ
Feature activation+0.000
including
Token including
Feature activation+0.138
Jewish
Token Jewish
Feature activation+0.871
schools
Token schools
Feature activation+4.760
and
Token and
Feature activation+0.503
neighbourhoods
Token neighbourhoods
Feature activation+5.265
âĢĶ
Token âĢĶ
Feature activation+0.094
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
wake
Token wake
Feature activation+0.000
of
Token of
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.934
ed
Token ed
Feature activation+0.272
gy
Tokengy
Feature activation+4.882
or
Token or
Feature activation+0.000
controversial
Token controversial
Feature activation+5.186
âĢĵ
TokenâĢĵ
Feature activation+0.000
min
Tokenmin
Feature activation+0.000
ors
Tokenors
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
and
Token and
Feature activation+0.000
no
Token no
Feature activation+0.000
buildings
Token buildings
Feature activation+3.336
were
Token were
Feature activation+0.414
present
Token present
Feature activation+4.985
there
Token there
Feature activation+5.071
at
Token at
Feature activation+1.098
that
Token that
Feature activation+1.722
time
Token time
Feature activation+4.612
âĢĵ
Token âĢĵ
Feature activation+0.000
many
Token many
Feature activation+0.000
shops
Token shops
Feature activation+0.371
and
Token and
Feature activation+0.000
no
Token no
Feature activation+0.000
buildings
Token buildings
Feature activation+3.336
were
Token were
Feature activation+0.414
present
Token present
Feature activation+4.985
there
Token there
Feature activation+5.071
at
Token at
Feature activation+1.098
that
Token that
Feature activation+1.722
time
Token time
Feature activation+4.612
âĢĵ
Token âĢĵ
Feature activation+0.000
accidentally
Token accidentally
Feature activation+1.147
dropped
Token dropped
Feature activation+3.022
by
Token by
Feature activation+0.898
someone
Token someone
Feature activation+2.765
living
Token living
Feature activation+0.144
nearby
Token nearby
Feature activation+4.935
âĢĶ
Token âĢĶ
Feature activation+0.000
could
Token could
Feature activation+0.000
breed
Token breed
Feature activation+0.000
a
Token a
Feature activation+0.000
sense
Token sense
Feature activation+0.000
âĢĶ
TokenâĢĶ
Feature activation+0.000
both
Tokenboth
Feature activation+0.370
with
Token with
Feature activation+0.000
Europe
Token Europe
Feature activation+0.616
and
Token and
Feature activation+0.004
globally
Token globally
Feature activation+4.901
âĢĶ
TokenâĢĶ
Feature activation+0.360
will
Tokenwill
Feature activation+0.000
be
Token be
Feature activation+0.000
essential
Token essential
Feature activation+0.000
.
Token.
Feature activation+0.000
it
Token it
Feature activation+0.662
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.934
ed
Token ed
Feature activation+0.272
gy
Tokengy
Feature activation+4.882
or
Token or
Feature activation+0.000
controversial
Token controversial
Feature activation+5.186
âĢĵ
TokenâĢĵ
Feature activation+0.000
min
Tokenmin
Feature activation+0.000
ors
Tokenors
Feature activation+0.000
use
Token use
Feature activation+1.722
-
Token-
Feature activation+0.000
of
Tokenof
Feature activation+0.000
-
Token-
Feature activation+0.000
force
Tokenforce
Feature activation+1.942
scenarios
Token scenarios
Feature activation+4.880
âĢĵ
Token âĢĵ
Feature activation+0.000
and
Token and
Feature activation+0.000
his
Token his
Feature activation+0.000
patrol
Token patrol
Feature activation+0.000
car
Token car
Feature activation+0.000
âĢĶ
TokenâĢĶ
Feature activation+0.000
including
Tokenincluding
Feature activation+0.000
the
Token the
Feature activation+0.464
Mil
Token Mil
Feature activation+0.000
itor
Tokenitor
Feature activation+0.000
itself
Token itself
Feature activation+4.770
âĢĶ
TokenâĢĶ
Feature activation+0.000
sl
Tokensl
Feature activation+0.000
id
Tokenid
Feature activation+0.000
into
Token into
Feature activation+0.000
a
Token a
Feature activation+0.000
sensitive
Token sensitive
Feature activation+0.000
sites
Token sites
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+0.000
including
Token including
Feature activation+0.138
Jewish
Token Jewish
Feature activation+0.871
schools
Token schools
Feature activation+4.760
and
Token and
Feature activation+0.503
neighbourhoods
Token neighbourhoods
Feature activation+5.265
âĢĶ
Token âĢĶ
Feature activation+0.094
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
Greens
Token Greens
Feature activation+0.787
pan
Tokenpan
Feature activation+1.716
,
Token,
Feature activation+0.524
and
Token and
Feature activation+0.000
Bern
Token Bern
Feature activation+1.344
anke
Tokenanke
Feature activation+4.739
âĢĶ
TokenâĢĶ
Feature activation+0.200
have
Tokenhave
Feature activation+0.012
been
Token been
Feature activation+0.000
Jewish
Token Jewish
Feature activation+0.000
.
Token.
Feature activation+0.000
hated
Token hated
Feature activation+1.049
it
Token it
Feature activation+2.761
and
Token and
Feature activation+0.000
some
Token some
Feature activation+0.000
loved
Token loved
Feature activation+0.490
it
Token it
Feature activation+4.721
âĢĵ
TokenâĢĵ
Feature activation+0.000
were
Tokenwere
Feature activation+0.000
often
Token often
Feature activation+0.000
the
Token the
Feature activation+0.000
ones
Token ones
Feature activation+0.000
Go
TokenGo
Feature activation+0.000
in
Token in
Feature activation+0.000
trouble
Token trouble
Feature activation+2.901
earlier
Token earlier
Feature activation+3.079
this
Token this
Feature activation+0.000
year
Token year
Feature activation+4.641
âĢĶ
Token âĢĶ
Feature activation+0.000
it
Token it
Feature activation+0.000
often
Token often
Feature activation+0.000
lays
Token lays
Feature activation+0.000
und
Token und
Feature activation+0.000
violence
Token violence
Feature activation+4.348
for
Token for
Feature activation+0.173
the
Token the
Feature activation+0.000
sake
Token sake
Feature activation+0.533
of
Token of
Feature activation+0.000
violence
Token violence
Feature activation+4.626
âĢĶ
TokenâĢĶ
Feature activation+0.000
the
Tokenthe
Feature activation+0.000
Civil
Token Civil
Feature activation+0.000
War
Token War
Feature activation+0.000
,
Token,
Feature activation+0.000
were
Token were
Feature activation+0.414
present
Token present
Feature activation+4.985
there
Token there
Feature activation+5.071
at
Token at
Feature activation+1.098
that
Token that
Feature activation+1.722
time
Token time
Feature activation+4.612
âĢĵ
Token âĢĵ
Feature activation+0.000
many
Token many
Feature activation+0.000
buildings
Token buildings
Feature activation+0.000
were
Token were
Feature activation+0.000
taken
Token taken
Feature activation+0.000
filters
Token filters
Feature activation+1.892
in
Token in
Feature activation+0.326
gardens
Token gardens
Feature activation+1.292
,
Token,
Feature activation+0.123
for
Token for
Feature activation+0.000
example
Token example
Feature activation+4.573
,
Token,
Feature activation+2.025
to
Token to
Feature activation+0.000
keep
Token keep
Feature activation+0.000
it
Token it
Feature activation+0.000
from
Token from
Feature activation+0.000
area
Token area
Feature activation+4.387
,
Token,
Feature activation+3.127
virtually
Token virtually
Feature activation+0.693
impossible
Token impossible
Feature activation+1.375
to
Token to
Feature activation+0.029
reach
Token reach
Feature activation+4.568
âĢĶ
TokenâĢĶ
Feature activation+0.000
their
Tokentheir
Feature activation+0.000
bewild
Token bewild
Feature activation+0.000
er
Tokener
Feature activation+0.000
ment
Tokenment
Feature activation+0.000
-
Token-
Feature activation+0.000
int
Tokenint
Feature activation+0.000
ense
Tokenense
Feature activation+0.000
exercise
Token exercise
Feature activation+2.940
a
Token a
Feature activation+0.021
week
Token week
Feature activation+4.534
âĢĶ
TokenâĢĶ
Feature activation+0.000
sh
Tokensh
Feature activation+0.000
owed
Tokenowed
Feature activation+0.000
similar
Token similar
Feature activation+0.000
patterns
Token patterns
Feature activation+0.000
ball
Tokenball
Feature activation+0.000
,
Token,
Feature activation+0.000
football
Token football
Feature activation+0.000
,
Token,
Feature activation+0.284
and
Token and
Feature activation+0.086
wrestling
Token wrestling
Feature activation+4.474
âĢĶ
TokenâĢĶ
Feature activation+0.000
Z
TokenZ
Feature activation+0.000
immer
Tokenimmer
Feature activation+0.000
was
Token was
Feature activation+0.000
a
Token a
Feature activation+0.000
points
Token points
Feature activation+0.000
âĢĵ
Token âĢĵ
Feature activation+0.000
allowing
Token allowing
Feature activation+0.109
just
Token just
Feature activation+0.996
seven
Token seven
Feature activation+1.705
goals
Token goals
Feature activation+4.451
âĢĵ
Token âĢĵ
Feature activation+0.000
the
Token the
Feature activation+0.000
offense
Token offense
Feature activation+0.000
has
Token has
Feature activation+0.000
been
Token been
Feature activation+0.000

Top DFA by src position
MAX = 6.177

troops
Token troops
Feature activation-0.001
Top resid features:
to
Token to
Feature activation+0.003
Top resid features:
protect
Token protect
Feature activation+0.001
Top resid features:
sensitive
Token sensitive
Feature activation+0.007
Top resid features:
sites
Token sites
Feature activation-0.004
Top resid features:
âĢĶ
Token âĢĶ
Feature activation+5.939
Top resid features:
including
Token including
Feature activation+0.046
Top resid features:
Jewish
Token Jewish
Feature activation+0.013
Top resid features:
schools
Token schools
Feature activation+0.006
Top resid features:
and
Token and
Feature activation-0.001
Top resid features:
neighbourhoods
Token neighbourhoods
Feature activation+0.025
Top resid features:
copied
Token copied
Feature activation-0.004
Top resid features:
elsewhere
Token elsewhere
Feature activation-0.009
Top resid features:
on
Token on
Feature activation-0.001
Top resid features:
the
Token the
Feature activation-0.003
Top resid features:
Internet
Token Internet
Feature activation+0.001
Top resid features:
âĢĵ
TokenâĢĵ
Feature activation+6.177
Top resid features:
especially
Tokenespecially
Feature activation+0.030
Top resid features:
when
Token when
Feature activation+0.008
Top resid features:
it
Token it
Feature activation+0.006
Top resid features:
âĢ
TokenâĢ
Feature activation+0.004
Top resid features:
Ļ
TokenĻ
Feature activation+0.000
Top resid features:
in
Token in
Feature activation+0.000
Top resid features:
Kat
Token Kat
Feature activation+0.005
Top resid features:
ima
Tokenima
Feature activation+0.001
Top resid features:
Mul
Token Mul
Feature activation-0.001
Top resid features:
ilo
Tokenilo
Feature activation-0.004
Top resid features:
âĢĵ
Token âĢĵ
Feature activation+5.804
Top resid features:
no
Token no
Feature activation-0.001
Top resid features:
shops
Token shops
Feature activation+0.015
Top resid features:
and
Token and
Feature activation-0.003
Top resid features:
no
Token no
Feature activation-0.000
Top resid features:
buildings
Token buildings
Feature activation+0.040
Top resid features:
in
Token in
Feature activation+0.001
Top resid features:
Kat
Token Kat
Feature activation+0.013
Top resid features:
ima
Tokenima
Feature activation+0.006
Top resid features:
Mul
Token Mul
Feature activation-0.000
Top resid features:
ilo
Tokenilo
Feature activation-0.002
Top resid features:
âĢĵ
Token âĢĵ
Feature activation+5.623
Top resid features:
no
Token no
Feature activation+0.005
Top resid features:
shops
Token shops
Feature activation+0.012
Top resid features:
and
Token and
Feature activation+0.001
Top resid features:
no
Token no
Feature activation+0.007
Top resid features:
buildings
Token buildings
Feature activation+0.048
Top resid features:
the
Token the
Feature activation-0.004
Top resid features:
sidewalk
Token sidewalk
Feature activation+0.003
Top resid features:
of
Token of
Feature activation-0.000
Top resid features:
a
Token a
Feature activation-0.004
Top resid features:
neighborhood
Token neighborhood
Feature activation+0.008
Top resid features:
âĢĶ
Token âĢĶ
Feature activation+5.732
Top resid features:
as
Token as
Feature activation+0.085
Top resid features:
if
Token if
Feature activation+0.006
Top resid features:
accidentally
Token accidentally
Feature activation-0.005
Top resid features:
dropped
Token dropped
Feature activation-0.002
Top resid features:
by
Token by
Feature activation-0.017
Top resid features:
international
Token international
Feature activation-0.004
Top resid features:
research
Token research
Feature activation+0.003
Top resid features:
and
Token and
Feature activation+0.001
Top resid features:
trading
Token trading
Feature activation+0.015
Top resid features:
relationships
Token relationships
Feature activation-0.007
Top resid features:
âĢĶ
TokenâĢĶ
Feature activation+5.652
Top resid features:
both
Tokenboth
Feature activation+0.022
Top resid features:
with
Token with
Feature activation+0.038
Top resid features:
Europe
Token Europe
Feature activation-0.007
Top resid features:
and
Token and
Feature activation+0.047
Top resid features:
globally
Token globally
Feature activation+0.001
Top resid features:
copied
Token copied
Feature activation-0.003
Top resid features:
elsewhere
Token elsewhere
Feature activation-0.005
Top resid features:
on
Token on
Feature activation-0.001
Top resid features:
the
Token the
Feature activation-0.006
Top resid features:
Internet
Token Internet
Feature activation-0.000
Top resid features:
âĢĵ
TokenâĢĵ
Feature activation+5.805
Top resid features:
especially
Tokenespecially
Feature activation+0.031
Top resid features:
when
Token when
Feature activation+0.032
Top resid features:
it
Token it
Feature activation+0.015
Top resid features:
âĢ
TokenâĢ
Feature activation+0.003
Top resid features:
Ļ
TokenĻ
Feature activation-0.002
Top resid features:
not
Token not
Feature activation-0.002
Top resid features:
activate
Token activate
Feature activation-0.000
Top resid features:
his
Token his
Feature activation+0.001
Top resid features:
body
Token body
Feature activation-0.001
Top resid features:
camera
Token camera
Feature activation-0.002
Top resid features:
âĢĵ
Token âĢĵ
Feature activation+5.655
Top resid features:
as
Token as
Feature activation+0.010
Top resid features:
required
Token required
Feature activation+0.008
Top resid features:
by
Token by
Feature activation-0.016
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
department
Token department
Feature activation+0.000
Top resid features:
Plat
Token Plat
Feature activation-0.000
Top resid features:
te
Tokente
Feature activation+0.001
Top resid features:
,
Token,
Feature activation-0.001
Top resid features:
25
Token 25
Feature activation+0.005
Top resid features:
trucks
Token trucks
Feature activation+0.017
Top resid features:
âĢĶ
TokenâĢĶ
Feature activation+5.335
Top resid features:
including
Tokenincluding
Feature activation+0.110
Top resid features:
the
Token the
Feature activation+0.055
Top resid features:
Mil
Token Mil
Feature activation+0.013
Top resid features:
itor
Tokenitor
Feature activation+0.005
Top resid features:
itself
Token itself
Feature activation+0.076
Top resid features:
troops
Token troops
Feature activation-0.002
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
protect
Token protect
Feature activation-0.003
Top resid features:
sensitive
Token sensitive
Feature activation+0.004
Top resid features:
sites
Token sites
Feature activation-0.004
Top resid features:
âĢĶ
Token âĢĶ
Feature activation+5.382
Top resid features:
including
Token including
Feature activation+0.094
Top resid features:
Jewish
Token Jewish
Feature activation+0.043
Top resid features:
schools
Token schools
Feature activation+0.002
Top resid features:
and
Token and
Feature activation+0.000
Top resid features:
neighbourhoods
Token neighbourhoods
Feature activation+0.000
Top resid features:
on
Token on
Feature activation-0.000
Top resid features:
this
Token this
Feature activation-0.008
Top resid features:
list
Token list
Feature activation-0.001
Top resid features:
,
Token,
Feature activation+0.007
Top resid features:
five
Token five
Feature activation-0.003
Top resid features:
âĢĶ
TokenâĢĶ
Feature activation+5.573
Top resid features:
M
TokenM
Feature activation-0.004
Top resid features:
ey
Tokeney
Feature activation-0.001
Top resid features:
er
Tokener
Feature activation-0.001
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
Burns
Token Burns
Feature activation-0.000
Top resid features:
class
Token class
Feature activation+0.001
Top resid features:
argued
Token argued
Feature activation+0.001
Top resid features:
the
Token the
Feature activation-0.006
Top resid features:
most
Token most
Feature activation-0.010
Top resid features:
about
Token about
Feature activation-0.004
Top resid features:
âĢĵ
TokenâĢĵ
Feature activation+5.717
Top resid features:
meaning
Tokenmeaning
Feature activation-0.087
Top resid features:
some
Token some
Feature activation-0.014
Top resid features:
of
Token of
Feature activation-0.003
Top resid features:
the
Token the
Feature activation-0.003
Top resid features:
students
Token students
Feature activation+0.003
Top resid features:
there
Token there
Feature activation+0.000
Top resid features:
is
Token is
Feature activation+0.002
Top resid features:
a
Token a
Feature activation-0.005
Top resid features:
certificate
Token certificate
Feature activation+0.002
Top resid features:
hack
Token hack
Feature activation-0.004
Top resid features:
âĢĶ
Token âĢĶ
Feature activation+5.752
Top resid features:
like
Token like
Feature activation+0.008
Top resid features:
the
Token the
Feature activation-0.021
Top resid features:
one
Token one
Feature activation-0.052
Top resid features:
that
Token that
Feature activation-0.061
Top resid features:
got
Token got
Feature activation-0.008
Top resid features:
unnecessary
Token unnecessary
Feature activation-0.001
Top resid features:
loss
Token loss
Feature activation+0.001
Top resid features:
of
Token of
Feature activation-0.001
Top resid features:
human
Token human
Feature activation-0.001
Top resid features:
life
Token life
Feature activation-0.001
Top resid features:
âĢĶ
TokenâĢĶ
Feature activation+5.299
Top resid features:
where
Tokenwhere
Feature activation+0.013
Top resid features:
soldiers
Token soldiers
Feature activation-0.003
Top resid features:
died
Token died
Feature activation-0.003
Top resid features:
for
Token for
Feature activation+0.004
Top resid features:
nothing
Token nothing
Feature activation-0.001
Top resid features:
in
Token in
Feature activation+0.001
Top resid features:
Kat
Token Kat
Feature activation+0.008
Top resid features:
ima
Tokenima
Feature activation+0.003
Top resid features:
Mul
Token Mul
Feature activation-0.001
Top resid features:
ilo
Tokenilo
Feature activation-0.000
Top resid features:
âĢĵ
Token âĢĵ
Feature activation+5.236
Top resid features:
no
Token no
Feature activation+0.002
Top resid features:
shops
Token shops
Feature activation+0.003
Top resid features:
and
Token and
Feature activation-0.001
Top resid features:
no
Token no
Feature activation+0.002
Top resid features:
buildings
Token buildings
Feature activation+0.022
Top resid features:
storm
Token storm
Feature activation-0.001
Top resid features:
water
Tokenwater
Feature activation+0.000
Top resid features:
from
Token from
Feature activation+0.001
Top resid features:
their
Token their
Feature activation-0.002
Top resid features:
properties
Token properties
Feature activation+0.007
Top resid features:
âĢĶ
Token âĢĶ
Feature activation+5.178
Top resid features:
channel
Token channel
Feature activation+0.001
Top resid features:
ing
Tokening
Feature activation+0.010
Top resid features:
water
Token water
Feature activation+0.001
Top resid features:
from
Token from
Feature activation+0.002
Top resid features:
gut
Token gut
Feature activation-0.001
Top resid features:
just
Token just
Feature activation-0.001
Top resid features:
begun
Token begun
Feature activation+0.001
Top resid features:
in
Token in
Feature activation+0.001
Top resid features:
south
Token south
Feature activation-0.008
Top resid features:
Sudan
Token Sudan
Feature activation+0.004
Top resid features:
âĢĶ
TokenâĢĶ
Feature activation+5.272
Top resid features:
this
Tokenthis
Feature activation+0.018
Top resid features:
one
Token one
Feature activation+0.003
Top resid features:
in
Token in
Feature activation-0.002
Top resid features:
an
Token an
Feature activation-0.002
Top resid features:
isolated
Token isolated
Feature activation-0.002
Top resid features:
âĢ
TokenâĢ
Feature activation-0.038
Top resid features:
Ļ
TokenĻ
Feature activation+0.003
Top resid features:
t
Tokent
Feature activation-0.015
Top resid features:
physically
Token physically
Feature activation-0.000
Top resid features:
active
Token active
Feature activation+0.019
Top resid features:
âĢĶ
TokenâĢĶ
Feature activation+5.290
Top resid features:
meaning
Tokenmeaning
Feature activation+0.049
Top resid features:
they
Token they
Feature activation+0.010
Top resid features:
did
Token did
Feature activation+0.014
Top resid features:
not
Token not
Feature activation+0.001
Top resid features:
meet
Token meet
Feature activation+0.003
Top resid features:
port
Tokenport
Feature activation+0.007
Top resid features:
athlete
Token athlete
Feature activation+0.009
Top resid features:
in
Token in
Feature activation-0.010
Top resid features:
high
Token high
Feature activation+0.002
Top resid features:
school
Token school
Feature activation+0.023
Top resid features:
âĢĶ
TokenâĢĶ
Feature activation+4.814
Top resid features:
base
Tokenbase
Feature activation+0.000
Top resid features:
ball
Tokenball
Feature activation+0.040
Top resid features:
,
Token,
Feature activation+0.025
Top resid features:
football
Token football
Feature activation+0.021
Top resid features:
,
Token,
Feature activation-0.004
Top resid features:
their
Token their
Feature activation-0.006
Top resid features:
last
Token last
Feature activation-0.003
Top resid features:
30
Token 30
Feature activation-0.003
Top resid features:
possible
Token possible
Feature activation+0.005
Top resid features:
points
Token points
Feature activation+0.011
Top resid features:
âĢĵ
Token âĢĵ
Feature activation+5.245
Top resid features:
allowing
Token allowing
Feature activation-0.016
Top resid features:
just
Token just
Feature activation-0.002
Top resid features:
seven
Token seven
Feature activation+0.003
Top resid features:
goals
Token goals
Feature activation+0.041
Top resid features:
âĢĵ
Token âĢĵ
Feature activation+0.000
Top resid features:

Decoder Weights Distribution

Head 0: 0.07

Head 1: 0.02

Head 2: 0.08

Head 3: 0.03

Head 4: 0.04

Head 5: 0.54

Head 6: 0.03

Head 7: 0.06

Head 8: 0.02

Head 9: 0.02

Head 10: 0.08

Head 11: 0.02

Positive logits

)—4.25

4.19

4.08

3.82

--3.40

—"3.29

3.05

"—3.05

-3.00

––2.84

--2.83

—-2.81

(@2.79

2.73

,—2.67

.—2.65

)--2.24

——2.21

!--2.16

---2.06

Negative logits

});-2.21

},-2.03

"]-1.99

.",-1.98

});-1.98

};-1.93

"],-1.87

];-1.87

`.-1.84

]);-1.84

]).-1.83

');-1.81

};-1.79

].-1.78

''.-1.76

.]-1.76

']-1.76

].-1.76

Skydragon-1.75

.}-1.72

INTERVAL 4.739 - 5.265
CONTAINS 0.001%

Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.934
ed
Token ed
Feature activation+0.272
gy
Tokengy
Feature activation+4.882
or
Token or
Feature activation+0.000
controversial
Token controversial
Feature activation+5.186
âĢĵ
TokenâĢĵ
Feature activation+0.000
min
Tokenmin
Feature activation+0.000
ors
Tokenors
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
use
Token use
Feature activation+1.722
-
Token-
Feature activation+0.000
of
Tokenof
Feature activation+0.000
-
Token-
Feature activation+0.000
force
Tokenforce
Feature activation+1.942
scenarios
Token scenarios
Feature activation+4.880
âĢĵ
Token âĢĵ
Feature activation+0.000
and
Token and
Feature activation+0.000
his
Token his
Feature activation+0.000
patrol
Token patrol
Feature activation+0.000
car
Token car
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+0.000
including
Token including
Feature activation+0.138
Jewish
Token Jewish
Feature activation+0.871
schools
Token schools
Feature activation+4.760
and
Token and
Feature activation+0.503
neighbourhoods
Token neighbourhoods
Feature activation+5.265
âĢĶ
Token âĢĶ
Feature activation+0.094
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
wake
Token wake
Feature activation+0.000
of
Token of
Feature activation+0.000
âĢĶ
TokenâĢĶ
Feature activation+0.000
both
Tokenboth
Feature activation+0.370
with
Token with
Feature activation+0.000
Europe
Token Europe
Feature activation+0.616
and
Token and
Feature activation+0.004
globally
Token globally
Feature activation+4.901
âĢĶ
TokenâĢĶ
Feature activation+0.360
will
Tokenwill
Feature activation+0.000
be
Token be
Feature activation+0.000
essential
Token essential
Feature activation+0.000
.
Token.
Feature activation+0.000
sensitive
Token sensitive
Feature activation+0.000
sites
Token sites
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+0.000
including
Token including
Feature activation+0.138
Jewish
Token Jewish
Feature activation+0.871
schools
Token schools
Feature activation+4.760
and
Token and
Feature activation+0.503
neighbourhoods
Token neighbourhoods
Feature activation+5.265
âĢĶ
Token âĢĶ
Feature activation+0.094
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000

INTERVAL 4.212 - 4.739
CONTAINS 0.002%

the
Token the
Feature activation+0.000
flashlight
Token flashlight
Feature activation+3.497
attachments
Token attachments
Feature activation+4.277
from
Token from
Feature activation+0.181
Blue
Token Blue
Feature activation+0.427
Team
Token Team
Feature activation+4.257
's
Token's
Feature activation+0.748
weapon
Token weapon
Feature activation+0.928
--
Token --
Feature activation+0.000
a
Token a
Feature activation+0.000
Gr
Token Gr
Feature activation+0.000
blood
Token blood
Feature activation+0.360
flow
Token flow
Feature activation+3.045
into
Token into
Feature activation+0.192
people
Token people
Feature activation+1.969
âĢ
TokenâĢ
Feature activation+0.763
Ļ
TokenĻ
Feature activation+4.280
s
Tokens
Feature activation+0.831
limbs
Token limbs
Feature activation+3.301
.
Token.
Feature activation+0.000
Per
Token Per
Feature activation+0.000
ipher
Tokenipher
Feature activation+0.000
doesn
Token doesn
Feature activation+0.545
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
t
Tokent
Feature activation+0.000
require
Token require
Feature activation+0.251
movement
Token movement
Feature activation+4.425
âĢĶ
TokenâĢĶ
Feature activation+0.000
was
Tokenwas
Feature activation+0.000
the
Token the
Feature activation+0.000
better
Token better
Feature activation+0.000
construction
Token construction
Feature activation+0.000
people
Token people
Feature activation+1.112
and
Token and
Feature activation+0.025
aliens
Token aliens
Feature activation+4.212
of
Token of
Feature activation+0.035
every
Token every
Feature activation+0.000
color
Token color
Feature activation+4.376
âĢĶ
TokenâĢĶ
Feature activation+0.000
had
Tokenhad
Feature activation+0.000
g
Token g
Feature activation+0.000
amed
Tokenamed
Feature activation+0.000
the
Token the
Feature activation+0.000
of
Token of
Feature activation+0.242
different
Token different
Feature activation+0.216
stores
Token stores
Feature activation+4.280
throughout
Token throughout
Feature activation+0.452
the
Token the
Feature activation+0.025
country
Token country
Feature activation+4.318
--
Token--
Feature activation+0.000
assert
Token assert
Feature activation+0.000
highly
Token highly
Feature activation+0.000
individual
Token individual
Feature activation+0.000
ized
Tokenized
Feature activation+0.000

INTERVAL 3.686 - 4.212
CONTAINS 0.004%

âĢĶ
TokenâĢĶ
Feature activation+0.000
orig
Tokenorig
Feature activation+0.000
inally
Tokeninally
Feature activation+0.000
enacted
Token enacted
Feature activation+0.207
in
Token in
Feature activation+0.000
2001
Token 2001
Feature activation+4.004
(
Token (
Feature activation+1.878
in
Tokenin
Feature activation+0.000
the
Token the
Feature activation+0.000
Economic
Token Economic
Feature activation+0.000
Growth
Token Growth
Feature activation+0.000
policy
Token policy
Feature activation+0.000
as
Token as
Feature activation+0.234
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.153
common
Tokencommon
Feature activation+0.201
sense
Token sense
Feature activation+4.096
,
Token,
Feature activation+0.426
âĢ
TokenâĢ
Feature activation+0.734
Ŀ
TokenĿ
Feature activation+3.156
despite
Token despite
Feature activation+0.135
the
Token the
Feature activation+0.000
is
Token is
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+0.000
to
Token to
Feature activation+0.421
the
Token the
Feature activation+0.165
most
Token most
Feature activation+1.029
extent
Token extent
Feature activation+3.767
âĢĶ
Token âĢĶ
Feature activation+0.000
recognized
Token recognized
Feature activation+0.000
as
Token as
Feature activation+0.000
a
Token a
Feature activation+0.000
beautiful
Token beautiful
Feature activation+0.000
scientists
Token scientists
Feature activation+1.036
but
Token but
Feature activation+0.327
many
Token many
Feature activation+0.312
conservatives
Token conservatives
Feature activation+2.458
as
Token as
Feature activation+0.231
well
Token well
Feature activation+4.112
âĢĶ
Token âĢĶ
Feature activation+0.000
have
Token have
Feature activation+0.000
been
Token been
Feature activation+0.000
quick
Token quick
Feature activation+0.000
to
Token to
Feature activation+0.000
doctor
Token doctor
Feature activation+0.000
wants
Token wants
Feature activation+0.000
a
Token a
Feature activation+0.000
selfie
Token selfie
Feature activation+3.151
with
Token with
Feature activation+0.000
him
Token him
Feature activation+3.858
âĢĶ
TokenâĢĶ
Feature activation+0.000
H
TokenH
Feature activation+0.000
ark
Tokenark
Feature activation+0.000
ens
Tokenens
Feature activation+0.000
pre
Token pre
Feature activation+0.000

INTERVAL 3.159 - 3.686
CONTAINS 0.009%

't
Token't
Feature activation+0.266
shipping
Token shipping
Feature activation+1.505
until
Token until
Feature activation+0.058
later
Token later
Feature activation+1.636
this
Token this
Feature activation+0.000
year
Token year
Feature activation+3.240
,
Token,
Feature activation+1.616
and
Token and
Feature activation+0.000
if
Token if
Feature activation+0.000
Microsoft
Token Microsoft
Feature activation+0.000
follows
Token follows
Feature activation+0.000
âĢĵ
TokenâĢĵ
Feature activation+0.000
our
Tokenour
Feature activation+0.000
local
Token local
Feature activation+0.245
version
Token version
Feature activation+1.779
of
Token of
Feature activation+0.030
Obamacare
Token Obamacare
Feature activation+3.185
âĢĵ
TokenâĢĵ
Feature activation+0.000
cr
Tokencr
Feature activation+0.000
ashed
Tokenashed
Feature activation+0.000
yesterday
Token yesterday
Feature activation+0.000
.
Token.
Feature activation+0.000
âĢĶ
TokenâĢĶ
Feature activation+0.000
the
Tokenthe
Feature activation+0.000
elephant
Token elephant
Feature activation+0.526
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
room
Token room
Feature activation+3.175
here
Token here
Feature activation+3.190
in
Token in
Feature activation+0.349
Buffalo
Token Buffalo
Feature activation+2.271
âĢĶ
TokenâĢĶ
Feature activation+0.000
would
Tokenwould
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.145
ľ
Tokenľ
Feature activation+0.462
ice
Tokenice
Feature activation+0.596
trees
Token trees
Feature activation+1.019
âĢ
TokenâĢ
Feature activation+0.170
Ŀ
TokenĿ
Feature activation+3.524
)
Token)
Feature activation+4.034
or
Token or
Feature activation+0.000
snow
Token snow
Feature activation+0.028
monsters
Token monsters
Feature activation+2.076
,
Token,
Feature activation+1.630
in
Token in
Feature activation+0.183
Korea
Token Korea
Feature activation+3.815
or
Token or
Feature activation+0.000
the
Token the
Feature activation+0.000
Middle
Token Middle
Feature activation+0.082
East
Token East
Feature activation+3.682
âĢĶ
TokenâĢĶ
Feature activation+0.000
the
Tokenthe
Feature activation+0.000
active
Token active
Feature activation+0.000
duty
Token duty
Feature activation+0.000
force
Token force
Feature activation+0.000

INTERVAL 2.633 - 3.159
CONTAINS 0.014%

âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.841
band
Tokenband
Feature activation+0.000
its
Tokenits
Feature activation+2.434
âĢ
TokenâĢ
Feature activation+0.005
Ŀ
TokenĿ
Feature activation+3.014
âĢĶ
TokenâĢĶ
Feature activation+0.000
had
Tokenhad
Feature activation+0.000
just
Token just
Feature activation+0.000
committed
Token committed
Feature activation+0.000
the
Token the
Feature activation+0.000
even
Token even
Feature activation+0.099
his
Token his
Feature activation+0.000
doctor
Token doctor
Feature activation+0.000
wants
Token wants
Feature activation+0.000
a
Token a
Feature activation+0.000
selfie
Token selfie
Feature activation+3.151
with
Token with
Feature activation+0.000
him
Token him
Feature activation+3.858
âĢĶ
TokenâĢĶ
Feature activation+0.000
H
TokenH
Feature activation+0.000
ark
Tokenark
Feature activation+0.000
consultation
Token consultation
Feature activation+0.929
with
Token with
Feature activation+0.292
GOP
Token GOP
Feature activation+0.000
lawyer
Token lawyer
Feature activation+0.068
Trevor
Token Trevor
Feature activation+0.000
Stanley
Token Stanley
Feature activation+2.662
âĢĶ
TokenâĢĶ
Feature activation+0.000
he
Tokenhe
Feature activation+0.000
decided
Token decided
Feature activation+0.000
that
Token that
Feature activation+0.000
the
Token the
Feature activation+0.000
over
Token over
Feature activation+0.000
policy
Token policy
Feature activation+1.373
briefs
Token briefs
Feature activation+1.950
and
Token and
Feature activation+0.000
press
Token press
Feature activation+0.000
conferences
Token conferences
Feature activation+2.929
.
Token.
Feature activation+0.000
His
Token His
Feature activation+0.000
pitch
Token pitch
Feature activation+0.000
:
Token:
Feature activation+0.000
Create
Token Create
Feature activation+0.000
leaders
Tokenleaders
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+0.000
including
Token including
Feature activation+0.000
some
Token some
Feature activation+0.299
Clinton
Token Clinton
Feature activation+0.796
supporters
Token supporters
Feature activation+2.860
âĢĶ
Token âĢĶ
Feature activation+0.000
prefer
Token prefer
Feature activation+0.000
to
Token to
Feature activation+0.000
excuse
Token excuse
Feature activation+0.000
it
Token it
Feature activation+0.000

INTERVAL 2.106 - 2.633
CONTAINS 0.019%

third
Token third
Feature activation+0.000
of
Token of
Feature activation+0.000
Americans
Token Americans
Feature activation+0.000
âĢĵ
Token âĢĵ
Feature activation+0.000
36
Token 36
Feature activation+2.104
percent
Token percent
Feature activation+2.561
âĢĵ
Token âĢĵ
Feature activation+0.000
would
Token would
Feature activation+0.000
support
Token support
Feature activation+0.000
Obama
Token Obama
Feature activation+0.000
green
Token green
Feature activation+0.000
democracy
Token democracy
Feature activation+1.234
,
Token,
Feature activation+0.000
not
Token not
Feature activation+0.000
the
Token the
Feature activation+0.000
real
Token real
Feature activation+0.000
thing
Token thing
Feature activation+2.181
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Lead
TokenLead
Feature activation+0.000
ers
Tokeners
Feature activation+0.000
threats
Token threats
Feature activation+1.678
that
Token that
Feature activation+0.000
previously
Token previously
Feature activation+0.000
had
Token had
Feature activation+0.000
not
Token not
Feature activation+0.016
existed
Token existed
Feature activation+2.573
.
Token.
Feature activation+0.000
The
Token The
Feature activation+0.000
city
Token city
Feature activation+0.000
was
Token was
Feature activation+0.000
slowly
Token slowly
Feature activation+0.000
37
Token 37
Feature activation+0.636
miles
Token miles
Feature activation+1.319
outside
Token outside
Feature activation+0.168
the
Token the
Feature activation+0.000
capital
Token capital
Feature activation+1.278
city
Token city
Feature activation+2.161
of
Token of
Feature activation+0.000
T
Token T
Feature activation+0.000
bil
Tokenbil
Feature activation+0.000
isi
Tokenisi
Feature activation+2.851
âĢĵ
Token âĢĵ
Feature activation+0.000
super
Token super
Feature activation+0.000
delegates
Token delegates
Feature activation+0.000
to
Token to
Feature activation+0.049
front
Token front
Feature activation+0.000
-
Token-
Feature activation+0.000
loading
Tokenloading
Feature activation+2.152
the
Token the
Feature activation+0.341
more
Token more
Feature activation+0.000
conservative
Token conservative
Feature activation+0.095
southern
Token southern
Feature activation+0.027
states
Token states
Feature activation+2.839

INTERVAL 1.580 - 2.106
CONTAINS 0.030%

of
Token of
Feature activation+0.000
death
Token death
Feature activation+0.664
or
Token or
Feature activation+0.000
serious
Token serious
Feature activation+0.000
physical
Token physical
Feature activation+0.000
injury
Token injury
Feature activation+1.584
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
officer
Token officer
Feature activation+0.514
or
Token or
Feature activation+0.000
to
Token to
Feature activation+0.000
ahu
Tokenahu
Feature activation+0.000
and
Token and
Feature activation+0.000
U
Token U
Feature activation+0.000
H
TokenH
Feature activation+0.000
H
Token H
Feature activation+0.000
ilo
Tokenilo
Feature activation+1.683
,
Token,
Feature activation+0.000
but
Token but
Feature activation+0.000
several
Token several
Feature activation+0.000
community
Token community
Feature activation+0.000
colleges
Token colleges
Feature activation+0.385
much
Token much
Feature activation+0.000
gold
Token gold
Feature activation+0.000
âĢĶ
TokenâĢĶ
Feature activation+0.000
a
Tokena
Feature activation+0.000
great
Token great
Feature activation+0.111
target
Token target
Feature activation+2.067
for
Token for
Feature activation+0.727
thieves
Token thieves
Feature activation+2.236
âĢĶ
TokenâĢĶ
Feature activation+0.000
con
Tokencon
Feature activation+0.000
vin
Tokenvin
Feature activation+0.000
similar
Token similar
Feature activation+0.000
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
2012
Token 2012
Feature activation+0.000
consent
Token consent
Feature activation+0.000
order
Token order
Feature activation+1.834
with
Token with
Feature activation+0.000
Facebook
Token Facebook
Feature activation+0.344
.
Token.
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
sued
Token sued
Feature activation+0.000
the
Token the
Feature activation+0.000
City
Token City
Feature activation+0.000
claiming
Token claiming
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+0.000
rightfully
Token rightfully
Feature activation+1.643
so
Token so
Feature activation+2.166
âĢĶ
Token âĢĶ
Feature activation+0.488
that
Token that
Feature activation+0.000
if
Token if
Feature activation+0.000
a
Token a
Feature activation+0.000

INTERVAL 1.053 - 1.580
CONTAINS 0.048%

âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
sexuality
Token sexuality
Feature activation+0.000
âĢĵ
Token âĢĵ
Feature activation+0.000
the
Token the
Feature activation+0.000
practice
Token practice
Feature activation+1.341
of
Token of
Feature activation+0.414
dressing
Token dressing
Feature activation+0.019
up
Token up
Feature activation+0.259
in
Token in
Feature activation+0.000
fuzzy
Token fuzzy
Feature activation+0.172
whims
Token whims
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
--
Token--
Feature activation+0.000
the
Token the
Feature activation+0.253
same
Token same
Feature activation+1.324
industry
Token industry
Feature activation+0.230
that
Token that
Feature activation+0.000
uses
Token uses
Feature activation+0.000
children
Token children
Feature activation+0.000
as
Token as
Feature activation+0.000
deep
Token deep
Feature activation+0.000
sound
Token sound
Feature activation+0.174
beds
Token beds
Feature activation+0.677
and
Token and
Feature activation+0.000
drifting
Token drifting
Feature activation+0.000
melodies
Token melodies
Feature activation+1.332
.
Token.
Feature activation+0.000
But
Token But
Feature activation+0.000
this
Token this
Feature activation+0.000
track
Token track
Feature activation+0.000
had
Token had
Feature activation+0.000
if
Token if
Feature activation+0.000
any
Token any
Feature activation+0.000
ceasefire
Token ceasefire
Feature activation+0.000
were
Token were
Feature activation+0.000
to
Token to
Feature activation+0.000
endure
Token endure
Feature activation+1.322
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
UN
Token UN
Feature activation+0.000
of
Token of
Feature activation+0.211
bad
Token bad
Feature activation+1.720
teams
Token teams
Feature activation+2.657
in
Token in
Feature activation+0.034
the
Token the
Feature activation+0.000
NHL
Token NHL
Feature activation+1.376
âĢĵ
Token âĢĵ
Feature activation+0.000
would
Token would
Feature activation+0.000
sink
Token sink
Feature activation+0.000
like
Token like
Feature activation+0.000
a
Token a
Feature activation+0.000

INTERVAL 0.527 - 1.053
CONTAINS 0.101%

gold
Token gold
Feature activation+0.000
âĢĶ
TokenâĢĶ
Feature activation+0.000
a
Tokena
Feature activation+0.000
great
Token great
Feature activation+0.111
target
Token target
Feature activation+2.067
for
Token for
Feature activation+0.727
thieves
Token thieves
Feature activation+2.236
âĢĶ
TokenâĢĶ
Feature activation+0.000
con
Tokencon
Feature activation+0.000
vin
Tokenvin
Feature activation+0.000
ced
Tokenced
Feature activation+0.000
vs
Token vs
Feature activation+0.119
.
Token.
Feature activation+0.000
Qatar
Token Qatar
Feature activation+0.085
âĢĵ
Token âĢĵ
Feature activation+0.000
the
Token the
Feature activation+0.000
country
Token country
Feature activation+0.895
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.132
first
Token first
Feature activation+0.588
match
Token match
Feature activation+0.401
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.274
right
Token right
Feature activation+0.126
âĢĵ
Token âĢĵ
Feature activation+0.022
men
Token men
Feature activation+0.732
who
Token who
Feature activation+0.000
have
Token have
Feature activation+0.000
been
Token been
Feature activation+0.000
isolated
Token isolated
Feature activation+0.000
for
Token for
Feature activation+0.000
say
Token say
Feature activation+0.000
yet
Token yet
Feature activation+0.000
âĢĵ
TokenâĢĵ
Feature activation+0.000
probably
Tokenprobably
Feature activation+0.105
the
Token the
Feature activation+0.005
same
Token same
Feature activation+0.712
thing
Token thing
Feature activation+0.877
I
Token I
Feature activation+0.000
tell
Token tell
Feature activation+0.000
anyone
Token anyone
Feature activation+0.264
who
Token who
Feature activation+0.000
began
Token began
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+0.000
five
Token five
Feature activation+0.000
years
Token years
Feature activation+0.445
of
Token of
Feature activation+0.000
it
Token it
Feature activation+0.697
.
Token.
Feature activation+0.000
All
Token All
Feature activation+0.000
that
Token that
Feature activation+0.000
ended
Token ended
Feature activation+0.000
this
Token this
Feature activation+0.000

INTERVAL 0.000 - 0.527
CONTAINS 99.772%

now
Token now
Feature activation+0.000
forced
Token forced
Feature activation+0.000
to
Token to
Feature activation+0.000
breast
Token breast
Feature activation+0.000
feed
Tokenfeed
Feature activation+0.000
unless
Token unless
Feature activation+0.000
she
Token she
Feature activation+0.000
has
Token has
Feature activation+0.000
a
Token a
Feature activation+0.000
medical
Token medical
Feature activation+0.000
reason
Token reason
Feature activation+0.000
is
Token is
Feature activation+0.000
more
Token more
Feature activation+0.000
straightforward
Token straightforward
Feature activation+0.000
to
Token to
Feature activation+0.000
encode
Token encode
Feature activation+0.000
and
Token and
Feature activation+0.000
decode
Token decode
Feature activation+0.000
with
Token with
Feature activation+0.000
,
Token,
Feature activation+0.000
but
Token but
Feature activation+0.000
has
Token has
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
be
Token be
Feature activation+0.000
clear
Token clear
Feature activation+0.000
:
Token:
Feature activation+0.000
Canadian
Token Canadian
Feature activation+0.000
law
Token law
Feature activation+0.000
enforcement
Token enforcement
Feature activation+0.000
is
Token is
Feature activation+0.000
not
Token not
Feature activation+0.000
way
Token way
Feature activation+0.000
,
Token,
Feature activation+0.000
with
Token with
Feature activation+0.000
all
Token all
Feature activation+0.000
this
Token this
Feature activation+0.000
am
Token am
Feature activation+0.000
ending
Tokenending
Feature activation+0.000
of
Token of
Feature activation+0.000
rights
Token rights
Feature activation+0.000
don
Token don
Feature activation+0.000
't
Token't
Feature activation+0.000
For
Token For
Feature activation+0.000
my
Token my
Feature activation+0.000
father
Token father
Feature activation+0.000
,
Token,
Feature activation+0.000
it
Token it
Feature activation+0.000
is
Token is
Feature activation+0.000
the
Token the
Feature activation+0.000
art
Token art
Feature activation+0.000
of
Token of
Feature activation+0.000
negotiating
Token negotiating
Feature activation+0.000
and
Token and
Feature activation+0.000

BOTTOM ACTIVATIONS
MIN = 0.000

atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
&
Token &
Feature activation+0.000
Development
Token Development
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
&
Token &
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
rik
Tokenrik
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
rik
Tokenrik
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000

Top feature 5 in H8.5: (feature 10754)

TOP ACTIVATIONS
MAX = 5.928

our
Token our
Feature activation+1.866
safety
Token safety
Feature activation+4.773
than
Token than
Feature activation+2.805
we
Token we
Feature activation+3.724
are
Token are
Feature activation+4.679
today
Token today
Feature activation+5.928
?
Token?
Feature activation+0.000
The
Token The
Feature activation+0.000
answer
Token answer
Feature activation+0.000
clearly
Token clearly
Feature activation+0.000
is
Token is
Feature activation+0.000
friendly
Tokenfriendly
Feature activation+2.568
public
Token public
Feature activation+2.579
transport
Token transport
Feature activation+4.293
where
Token where
Feature activation+0.896
you
Token you
Feature activation+3.953
live
Token live
Feature activation+5.836
?
Token?
Feature activation+0.424
Tell
Token Tell
Feature activation+0.113
us
Token us
Feature activation+0.000
about
Token about
Feature activation+0.000
it
Token it
Feature activation+0.000
spawned
Token spawned
Feature activation+2.616
of
Token of
Feature activation+2.777
unusually
Token unusually
Feature activation+0.578
warm
Token warm
Feature activation+1.541
ocean
Token ocean
Feature activation+2.450
waters
Token waters
Feature activation+5.784
?
Token?
Feature activation+0.000
In
Token In
Feature activation+0.262
one
Token one
Feature activation+0.034
sense
Token sense
Feature activation+0.000
,
Token,
Feature activation+0.000
,
Token,
Feature activation+0.000
do
Token do
Feature activation+0.000
they
Token they
Feature activation+0.000
take
Token take
Feature activation+1.482
more
Token more
Feature activation+2.912
risks
Token risks
Feature activation+5.499
like
Token like
Feature activation+2.940
starting
Token starting
Feature activation+2.139
a
Token a
Feature activation+1.422
business
Token business
Feature activation+2.897
or
Token or
Feature activation+1.378
or
Tokenor
Feature activation+2.287
most
Token most
Feature activation+0.016
all
Token all
Feature activation+0.025
)
Token)
Feature activation+3.531
human
Token human
Feature activation+1.937
labor
Token labor
Feature activation+5.401
?
Token?
Feature activation+0.000
First
Token First
Feature activation+0.000
we
Token we
Feature activation+0.000
raised
Token raised
Feature activation+0.000
living
Token living
Feature activation+0.000
it
Token it
Feature activation+0.000
supposed
Token supposed
Feature activation+0.000
to
Token to
Feature activation+1.455
be
Token be
Feature activation+1.861
200
Token 200
Feature activation+4.137
G
TokenG
Feature activation+5.369
?
Token?
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
we
Token we
Feature activation+0.000
as
Token as
Feature activation+1.479
individuals
Token individuals
Feature activation+0.694
be
Token be
Feature activation+0.058
more
Token more
Feature activation+3.363
casual
Token casual
Feature activation+5.366
about
Token about
Feature activation+3.747
our
Token our
Feature activation+1.866
safety
Token safety
Feature activation+4.773
than
Token than
Feature activation+2.805
we
Token we
Feature activation+3.724
business
Token business
Feature activation+2.897
or
Token or
Feature activation+1.378
going
Token going
Feature activation+2.094
back
Token back
Feature activation+2.413
to
Token to
Feature activation+1.312
school
Token school
Feature activation+5.309
?
Token?
Feature activation+0.225
And
Token And
Feature activation+0.472
does
Token does
Feature activation+0.000
that
Token that
Feature activation+0.558
security
Token security
Feature activation+0.408
TPP
Token TPP
Feature activation+2.587
,
Token,
Feature activation+1.827
opposed
Token opposed
Feature activation+1.312
by
Token by
Feature activation+1.313
many
Token many
Feature activation+1.845
voters
Token voters
Feature activation+5.266
?
Token?
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
According
TokenAccording
Feature activation+0.000
to
Token to
Feature activation+0.000
something
Token something
Feature activation+1.568
hits
Token hits
Feature activation+1.271
your
Token your
Feature activation+1.113
craft
Token craft
Feature activation+2.671
and
Token and
Feature activation+1.312
you
Token you
Feature activation+5.247
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.074
re
Tokenre
Feature activation+1.100
stranded
Token stranded
Feature activation+1.900
millions
Token millions
Feature activation+3.853
.
Token.
Feature activation+0.000
Would
Token Would
Feature activation+0.000
more
Token more
Feature activation+0.168
easing
Token easing
Feature activation+0.000
convince
Token convince
Feature activation+0.080
them
Token them
Feature activation+5.163
to
Token to
Feature activation+3.665
do
Token do
Feature activation+2.711
that
Token that
Feature activation+4.303
?
Token?
Feature activation+0.000
Probably
Token Probably
Feature activation+0.000
checked
Tokenchecked
Feature activation+1.024
delivery
Token delivery
Feature activation+2.179
of
Token of
Feature activation+0.802
data
Token data
Feature activation+4.959
âĢ
TokenâĢ
Feature activation+4.708
Ŀ
TokenĿ
Feature activation+5.144
?
Token?
Feature activation+0.000
Well
Token Well
Feature activation+0.000
,
Token,
Feature activation+0.000
an
Token an
Feature activation+0.000
attacker
Token attacker
Feature activation+0.000
like
Token like
Feature activation+0.465
Blast
Token Blast
Feature activation+0.000
Processing
Token Processing
Feature activation+0.000
sounded
Token sounded
Feature activation+3.546
when
Token when
Feature activation+0.665
you
Token you
Feature activation+5.098
were
Token were
Feature activation+2.056
15
Token 15
Feature activation+3.170
?
Token?
Feature activation+0.000
Made
Token Made
Feature activation+0.000
the
Token the
Feature activation+0.000
and
Token and
Feature activation+2.038
under
Token under
Feature activation+0.595
study
Tokenstudy
Feature activation+4.374
to
Token to
Feature activation+0.883
C
Token C
Feature activation+0.162
ech
Tokenech
Feature activation+5.072
?
Token?
Feature activation+0.000
Of
Token Of
Feature activation+1.175
course
Token course
Feature activation+0.000
.
Token.
Feature activation+0.000
It
Token It
Feature activation+0.000
after
Token after
Feature activation+2.331
a
Token a
Feature activation+1.460
couple
Token couple
Feature activation+2.230
of
Token of
Feature activation+1.762
loan
Token loan
Feature activation+0.913
spells
Token spells
Feature activation+5.017
and
Token and
Feature activation+3.591
have
Token have
Feature activation+2.057
him
Token him
Feature activation+1.844
as
Token as
Feature activation+0.806
backup
Token backup
Feature activation+3.797
was
Token was
Feature activation+0.496
a
Token a
Feature activation+0.720
mod
Token mod
Feature activation+0.000
icum
Tokenicum
Feature activation+1.628
of
Token of
Feature activation+0.389
security
Token security
Feature activation+4.974
?
Token?
Feature activation+0.000
If
Token If
Feature activation+0.000
Syrian
Token Syrian
Feature activation+0.000
authorities
Token authorities
Feature activation+0.000
organized
Token organized
Feature activation+0.106
error
Token error
Feature activation+0.342
-
Token-
Feature activation+0.000
checked
Tokenchecked
Feature activation+1.024
delivery
Token delivery
Feature activation+2.179
of
Token of
Feature activation+0.802
data
Token data
Feature activation+4.959
âĢ
TokenâĢ
Feature activation+4.708
Ŀ
TokenĿ
Feature activation+5.144
?
Token?
Feature activation+0.000
Well
Token Well
Feature activation+0.000
,
Token,
Feature activation+0.000
,
Token,
Feature activation+0.869
or
Token or
Feature activation+0.348
an
Token an
Feature activation+2.712
ang
Token ang
Feature activation+0.552
rier
Tokenrier
Feature activation+2.361
child
Token child
Feature activation+4.945
spawned
Token spawned
Feature activation+2.616
of
Token of
Feature activation+2.777
unusually
Token unusually
Feature activation+0.578
warm
Token warm
Feature activation+1.541
ocean
Token ocean
Feature activation+2.450
my
Token my
Feature activation+1.258
height
Token height
Feature activation+1.328
,
Token,
Feature activation+1.326
or
Token or
Feature activation+0.196
even
Token even
Feature activation+2.895
someone
Token someone
Feature activation+4.923
taller
Token taller
Feature activation+4.039
?
Token?
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Advertisement
TokenAdvertisement
Feature activation+0.000
strongly
Token strongly
Feature activation+1.279
correlated
Token correlated
Feature activation+3.035
with
Token with
Feature activation+2.116
the
Token the
Feature activation+1.874
divorce
Token divorce
Feature activation+2.868
rate
Token rate
Feature activation+4.907
in
Token in
Feature activation+2.463
Maine
Token Maine
Feature activation+3.917
?
Token?
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000

Top DFA by src position
MAX = 5.616

,
Token,
Feature activation+0.011
Top resid features:
could
Token could
Feature activation+0.475
Top resid features:
we
Token we
Feature activation+2.665
Top resid features:
as
Token as
Feature activation+0.024
Top resid features:
individuals
Token individuals
Feature activation+0.054
Top resid features:
be
Token be
Feature activation+2.800
Top resid features:
more
Token more
Feature activation+0.007
Top resid features:
casual
Token casual
Feature activation+0.005
Top resid features:
about
Token about
Feature activation+0.019
Top resid features:
our
Token our
Feature activation+0.008
Top resid features:
safety
Token safety
Feature activation+0.014
Top resid features:
inside
Token inside
Feature activation-0.001
Top resid features:
).
Token).
Feature activation+0.000
Top resid features:
Do
Token Do
Feature activation-0.027
Top resid features:
any
Token any
Feature activation+0.062
Top resid features:
of
Token of
Feature activation+0.009
Top resid features:
you
Token you
Feature activation+2.238
Top resid features:
readers
Token readers
Feature activation+0.953
Top resid features:
from
Token from
Feature activation+0.197
Top resid features:
around
Token around
Feature activation+0.009
Top resid features:
the
Token the
Feature activation+0.002
Top resid features:
world
Token world
Feature activation+0.159
Top resid features:
Sandy
Token Sandy
Feature activation+0.108
Top resid features:
just
Token just
Feature activation+1.429
Top resid features:
a
Token a
Feature activation+0.048
Top resid features:
hurricane
Token hurricane
Feature activation+0.007
Top resid features:
,
Token,
Feature activation+2.005
Top resid features:
or
Token or
Feature activation+3.105
Top resid features:
an
Token an
Feature activation+0.009
Top resid features:
ang
Token ang
Feature activation+0.001
Top resid features:
rier
Tokenrier
Feature activation+0.010
Top resid features:
child
Token child
Feature activation+0.006
Top resid features:
spawned
Token spawned
Feature activation+0.006
Top resid features:
they
Token they
Feature activation+0.003
Top resid features:
cannot
Token cannot
Feature activation+0.005
Top resid features:
fall
Token fall
Feature activation+0.006
Top resid features:
,
Token,
Feature activation+0.006
Top resid features:
do
Token do
Feature activation+0.010
Top resid features:
they
Token they
Feature activation+5.054
Top resid features:
take
Token take
Feature activation+0.732
Top resid features:
more
Token more
Feature activation+0.166
Top resid features:
risks
Token risks
Feature activation+0.035
Top resid features:
like
Token like
Feature activation+0.000
Top resid features:
starting
Token starting
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.005
Top resid features:
time
Token time
Feature activation+0.012
Top resid features:
of
Token of
Feature activation+0.005
Top resid features:
Smith
Token Smith
Feature activation+0.001
Top resid features:
forward
Token forward
Feature activation+0.007
Top resid features:
lead
Token lead
Feature activation+4.115
Top resid features:
to
Token to
Feature activation+0.624
Top resid features:
the
Token the
Feature activation+0.154
Top resid features:
dis
Token dis
Feature activation-0.002
Top resid features:
-
Token-
Feature activation+0.002
Top resid features:
employment
Tokenemployment
Feature activation+0.030
Top resid features:
n
Tokenn
Feature activation-0.004
Top resid features:
âĢ
TokenâĢ
Feature activation+0.009
Top resid features:
Ļ
TokenĻ
Feature activation-0.001
Top resid features:
t
Tokent
Feature activation+0.006
Top resid features:
it
Token it
Feature activation+2.578
Top resid features:
supposed
Token supposed
Feature activation+2.786
Top resid features:
to
Token to
Feature activation+0.233
Top resid features:
be
Token be
Feature activation+0.390
Top resid features:
200
Token 200
Feature activation-0.005
Top resid features:
G
TokenG
Feature activation+0.029
Top resid features:
?
Token?
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.012
Top resid features:
could
Token could
Feature activation+0.313
Top resid features:
we
Token we
Feature activation+2.358
Top resid features:
as
Token as
Feature activation+0.019
Top resid features:
individuals
Token individuals
Feature activation+0.102
Top resid features:
be
Token be
Feature activation+2.811
Top resid features:
more
Token more
Feature activation+0.204
Top resid features:
casual
Token casual
Feature activation+0.052
Top resid features:
about
Token about
Feature activation+0.000
Top resid features:
our
Token our
Feature activation+0.000
Top resid features:
safety
Token safety
Feature activation+0.000
Top resid features:
they
Token they
Feature activation+0.002
Top resid features:
cannot
Token cannot
Feature activation+0.005
Top resid features:
fall
Token fall
Feature activation+0.004
Top resid features:
,
Token,
Feature activation+0.007
Top resid features:
do
Token do
Feature activation-0.004
Top resid features:
they
Token they
Feature activation+4.441
Top resid features:
take
Token take
Feature activation+0.838
Top resid features:
more
Token more
Feature activation-0.019
Top resid features:
risks
Token risks
Feature activation+0.040
Top resid features:
like
Token like
Feature activation+0.303
Top resid features:
starting
Token starting
Feature activation-0.008
Top resid features:
with
Token with
Feature activation-0.000
Top resid features:
them
Token them
Feature activation+0.007
Top resid features:
on
Token on
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.001
Top resid features:
TPP
Token TPP
Feature activation+0.023
Top resid features:
,
Token,
Feature activation+4.675
Top resid features:
opposed
Token opposed
Feature activation+0.351
Top resid features:
by
Token by
Feature activation+0.212
Top resid features:
many
Token many
Feature activation+0.012
Top resid features:
voters
Token voters
Feature activation+0.018
Top resid features:
?
Token?
Feature activation+0.000
Top resid features:
Commun
TokenCommun
Feature activation+0.001
Top resid features:
ication
Tokenication
Feature activation-0.006
Top resid features:
Ċ
TokenĊ
Feature activation-0.022
Top resid features:
Ċ
TokenĊ
Feature activation-0.029
Top resid features:
What
TokenWhat
Feature activation+0.036
Top resid features:
if
Token if
Feature activation+5.616
Top resid features:
something
Token something
Feature activation-0.016
Top resid features:
hits
Token hits
Feature activation+0.079
Top resid features:
your
Token your
Feature activation+0.009
Top resid features:
craft
Token craft
Feature activation+0.041
Top resid features:
and
Token and
Feature activation+0.116
Top resid features:
doing
Token doing
Feature activation+0.016
Top resid features:
.
Token.
Feature activation+0.009
Top resid features:
Would
Token Would
Feature activation+0.484
Top resid features:
more
Token more
Feature activation+0.021
Top resid features:
easing
Token easing
Feature activation+0.016
Top resid features:
convince
Token convince
Feature activation+4.993
Top resid features:
them
Token them
Feature activation+0.078
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
do
Token do
Feature activation+0.000
Top resid features:
that
Token that
Feature activation+0.000
Top resid features:
?
Token?
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation-0.005
Top resid features:
Frag
TokenFrag
Feature activation+0.002
Top resid features:
mentation
Tokenmentation
Feature activation+0.003
Top resid features:
attacks
Token attacks
Feature activation-0.001
Top resid features:
Remember
Token Remember
Feature activation+0.058
Top resid features:
how
Token how
Feature activation+4.641
Top resid features:
I
Token I
Feature activation+0.417
Top resid features:
said
Token said
Feature activation+0.520
Top resid features:
TCP
Token TCP
Feature activation+0.021
Top resid features:
âĢ
Token âĢ
Feature activation+0.016
Top resid features:
ľ
Tokenľ
Feature activation+0.001
Top resid features:
worse
Token worse
Feature activation+0.002
Top resid features:
.
Token.
Feature activation-0.006
Top resid features:
Ċ
TokenĊ
Feature activation+0.003
Top resid features:
Ċ
TokenĊ
Feature activation-0.011
Top resid features:
Remember
TokenRemember
Feature activation+0.108
Top resid features:
how
Token how
Feature activation+5.092
Top resid features:
impressive
Token impressive
Feature activation+0.046
Top resid features:
something
Token something
Feature activation+0.006
Top resid features:
like
Token like
Feature activation+0.005
Top resid features:
Blast
Token Blast
Feature activation+0.010
Top resid features:
Processing
Token Processing
Feature activation+0.012
Top resid features:
rapid
Token rapid
Feature activation+0.001
Top resid features:
rise
Token rise
Feature activation+0.009
Top resid features:
?
Token?
Feature activation+0.002
Top resid features:
Is
Token Is
Feature activation+0.862
Top resid features:
it
Token it
Feature activation+0.931
Top resid features:
possible
Token possible
Feature activation+1.389
Top resid features:
that
Token that
Feature activation+1.378
Top resid features:
the
Token the
Feature activation+0.006
Top resid features:
intention
Token intention
Feature activation+0.013
Top resid features:
was
Token was
Feature activation+0.097
Top resid features:
always
Token always
Feature activation+0.014
Top resid features:
rapid
Token rapid
Feature activation+0.002
Top resid features:
rise
Token rise
Feature activation+0.010
Top resid features:
?
Token?
Feature activation+0.070
Top resid features:
Is
Token Is
Feature activation+0.927
Top resid features:
it
Token it
Feature activation+1.199
Top resid features:
possible
Token possible
Feature activation+1.515
Top resid features:
that
Token that
Feature activation+0.861
Top resid features:
the
Token the
Feature activation+0.016
Top resid features:
intention
Token intention
Feature activation+0.016
Top resid features:
was
Token was
Feature activation+0.177
Top resid features:
always
Token always
Feature activation+0.026
Top resid features:
confirmed
Token confirmed
Feature activation+0.050
Top resid features:
by
Token by
Feature activation+0.454
Top resid features:
the
Token the
Feature activation+0.012
Top resid features:
low
Token low
Feature activation+0.006
Top resid features:
turnout
Token turnout
Feature activation+0.027
Top resid features:
,
Token,
Feature activation+2.997
Top resid features:
even
Token even
Feature activation+0.029
Top resid features:
in
Token in
Feature activation+0.031
Top resid features:
areas
Token areas
Feature activation+0.022
Top resid features:
where
Token where
Feature activation+0.047
Top resid features:
there
Token there
Feature activation+0.017
Top resid features:
Ċ
TokenĊ
Feature activation+0.001
Top resid features:
Frag
TokenFrag
Feature activation+0.011
Top resid features:
mentation
Tokenmentation
Feature activation+0.006
Top resid features:
attacks
Token attacks
Feature activation-0.009
Top resid features:
Remember
Token Remember
Feature activation+0.041
Top resid features:
how
Token how
Feature activation+4.423
Top resid features:
I
Token I
Feature activation+0.365
Top resid features:
said
Token said
Feature activation+0.451
Top resid features:
TCP
Token TCP
Feature activation+0.046
Top resid features:
âĢ
Token âĢ
Feature activation+0.009
Top resid features:
ľ
Tokenľ
Feature activation+0.024
Top resid features:
Sandy
Token Sandy
Feature activation+0.097
Top resid features:
just
Token just
Feature activation+0.522
Top resid features:
a
Token a
Feature activation+0.074
Top resid features:
hurricane
Token hurricane
Feature activation+0.015
Top resid features:
,
Token,
Feature activation+1.048
Top resid features:
or
Token or
Feature activation+3.621
Top resid features:
an
Token an
Feature activation+0.025
Top resid features:
ang
Token ang
Feature activation+0.005
Top resid features:
rier
Tokenrier
Feature activation-0.047
Top resid features:
child
Token child
Feature activation+0.024
Top resid features:
spawned
Token spawned
Feature activation+0.000
Top resid features:
taller
Token taller
Feature activation+0.098
Top resid features:
than
Token than
Feature activation+0.370
Top resid features:
someone
Token someone
Feature activation+0.110
Top resid features:
my
Token my
Feature activation+0.167
Top resid features:
height
Token height
Feature activation+0.078
Top resid features:
,
Token,
Feature activation+2.117
Top resid features:
or
Token or
Feature activation+1.037
Top resid features:
even
Token even
Feature activation+0.642
Top resid features:
someone
Token someone
Feature activation+0.035
Top resid features:
taller
Token taller
Feature activation+0.000
Top resid features:
?
Token?
Feature activation+0.000
Top resid features:
warning
Token warning
Feature activation-0.002
Top resid features:
.
Token.
Feature activation-0.008
Top resid features:
Did
Token Did
Feature activation+0.908
Top resid features:
you
Token you
Feature activation+0.666
Top resid features:
realise
Token realise
Feature activation+1.583
Top resid features:
that
Token that
Feature activation+2.242
Top resid features:
consumption
Token consumption
Feature activation+0.001
Top resid features:
of
Token of
Feature activation-0.002
Top resid features:
marg
Token marg
Feature activation+0.002
Top resid features:
arine
Tokenarine
Feature activation+0.002
Top resid features:
is
Token is
Feature activation+0.017
Top resid features:

Decoder Weights Distribution

Head 0: 0.08

Head 1: 0.01

Head 2: 0.05

Head 3: 0.03

Head 4: 0.04

Head 5: 0.54

Head 6: 0.02

Head 7: 0.04

Head 8: 0.03

Head 9: 0.06

Head 10: 0.07

Head 11: 0.01

Positive logits

)?6.81

?6.52

"?6.16

?"5.96

'?5.89

?).5.75

?!5.74

?)5.66

?:5.63

??5.59

?".5.57

?",5.45

?),5.44

?5.42

!?5.34

?!"5.27

?'5.25

???5.22

?」5.19

?'"4.93

Negative logits

renheit-2.23

ctuary-2.13

jong-2.01

DERR-1.99

��極-1.98

lication-1.94

anmar-1.94

perties-1.90

inguishable-1.85

inery-1.85

agraph-1.84

adelphia-1.82

xon-1.81

trak-1.77

igmatic-1.76

mast-1.76

achine-1.75

odox-1.75

qqa-1.73

upiter-1.72

INTERVAL 5.335 - 5.928
CONTAINS 0.001%

,
Token,
Feature activation+0.000
do
Token do
Feature activation+0.000
they
Token they
Feature activation+0.000
take
Token take
Feature activation+1.482
more
Token more
Feature activation+2.912
risks
Token risks
Feature activation+5.499
like
Token like
Feature activation+2.940
starting
Token starting
Feature activation+2.139
a
Token a
Feature activation+1.422
business
Token business
Feature activation+2.897
or
Token or
Feature activation+1.378
it
Token it
Feature activation+0.000
supposed
Token supposed
Feature activation+0.000
to
Token to
Feature activation+1.455
be
Token be
Feature activation+1.861
200
Token 200
Feature activation+4.137
G
TokenG
Feature activation+5.369
?
Token?
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
friendly
Tokenfriendly
Feature activation+2.568
public
Token public
Feature activation+2.579
transport
Token transport
Feature activation+4.293
where
Token where
Feature activation+0.896
you
Token you
Feature activation+3.953
live
Token live
Feature activation+5.836
?
Token?
Feature activation+0.424
Tell
Token Tell
Feature activation+0.113
us
Token us
Feature activation+0.000
about
Token about
Feature activation+0.000
it
Token it
Feature activation+0.000
spawned
Token spawned
Feature activation+2.616
of
Token of
Feature activation+2.777
unusually
Token unusually
Feature activation+0.578
warm
Token warm
Feature activation+1.541
ocean
Token ocean
Feature activation+2.450
waters
Token waters
Feature activation+5.784
?
Token?
Feature activation+0.000
In
Token In
Feature activation+0.262
one
Token one
Feature activation+0.034
sense
Token sense
Feature activation+0.000
,
Token,
Feature activation+0.000
we
Token we
Feature activation+0.000
as
Token as
Feature activation+1.479
individuals
Token individuals
Feature activation+0.694
be
Token be
Feature activation+0.058
more
Token more
Feature activation+3.363
casual
Token casual
Feature activation+5.366
about
Token about
Feature activation+3.747
our
Token our
Feature activation+1.866
safety
Token safety
Feature activation+4.773
than
Token than
Feature activation+2.805
we
Token we
Feature activation+3.724

INTERVAL 4.742 - 5.335
CONTAINS 0.002%

after
Token after
Feature activation+2.331
a
Token a
Feature activation+1.460
couple
Token couple
Feature activation+2.230
of
Token of
Feature activation+1.762
loan
Token loan
Feature activation+0.913
spells
Token spells
Feature activation+5.017
and
Token and
Feature activation+3.591
have
Token have
Feature activation+2.057
him
Token him
Feature activation+1.844
as
Token as
Feature activation+0.806
backup
Token backup
Feature activation+3.797
business
Token business
Feature activation+2.897
or
Token or
Feature activation+1.378
going
Token going
Feature activation+2.094
back
Token back
Feature activation+2.413
to
Token to
Feature activation+1.312
school
Token school
Feature activation+5.309
?
Token?
Feature activation+0.225
And
Token And
Feature activation+0.472
does
Token does
Feature activation+0.000
that
Token that
Feature activation+0.558
security
Token security
Feature activation+0.408
checked
Tokenchecked
Feature activation+1.024
delivery
Token delivery
Feature activation+2.179
of
Token of
Feature activation+0.802
data
Token data
Feature activation+4.959
âĢ
TokenâĢ
Feature activation+4.708
Ŀ
TokenĿ
Feature activation+5.144
?
Token?
Feature activation+0.000
Well
Token Well
Feature activation+0.000
,
Token,
Feature activation+0.000
an
Token an
Feature activation+0.000
attacker
Token attacker
Feature activation+0.000
ropes
Token ropes
Feature activation+1.687
and
Token and
Feature activation+1.718
hit
Token hit
Feature activation+1.428
the
Token the
Feature activation+0.879
road
Token road
Feature activation+2.759
,
Token,
Feature activation+4.900
in
Token in
Feature activation+1.633
hot
Token hot
Feature activation+0.442
pursuit
Token pursuit
Feature activation+2.660
of
Token of
Feature activation+1.660
this
Token this
Feature activation+1.833
TPP
Token TPP
Feature activation+2.587
,
Token,
Feature activation+1.827
opposed
Token opposed
Feature activation+1.312
by
Token by
Feature activation+1.313
many
Token many
Feature activation+1.845
voters
Token voters
Feature activation+5.266
?
Token?
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
According
TokenAccording
Feature activation+0.000
to
Token to
Feature activation+0.000

INTERVAL 4.149 - 4.742
CONTAINS 0.003%

horrible
Token horrible
Feature activation+1.906
inc
Token inc
Feature activation+0.031
lem
Tokenlem
Feature activation+0.000
ency
Tokenency
Feature activation+3.567
of
Token of
Feature activation+2.695
life
Token life
Feature activation+4.515
?
Token?
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
âĢķ
Token âĢķ
Feature activation+0.000
Ald
Token Ald
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
t
Tokent
Feature activation+0.069
we
Token we
Feature activation+0.000
all
Token all
Feature activation+0.183
know
Token know
Feature activation+4.241
it
Token it
Feature activation+2.080
âĢ
TokenâĢ
Feature activation+0.125
Ļ
TokenĻ
Feature activation+1.742
s
Tokens
Feature activation+2.558
never
Token never
Feature activation+2.631
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
t
Tokent
Feature activation+2.989
even
Token even
Feature activation+2.324
voted
Token voted
Feature activation+3.326
yet
Token yet
Feature activation+4.725
?
Token?
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
In
TokenIn
Feature activation+0.000
answer
Token answer
Feature activation+0.000
of
Token of
Feature activation+2.816
Avengers
Token Avengers
Feature activation+1.041
:
Token:
Feature activation+0.801
Age
Token Age
Feature activation+0.264
of
Token of
Feature activation+0.000
Ultron
Token Ultron
Feature activation+4.243
?
Token?
Feature activation+0.217
Email
Token Email
Feature activation+0.000
arts
Token arts
Feature activation+0.000
correspondent
Token correspondent
Feature activation+0.000
Emma
Token Emma
Feature activation+0.000
the
Token the
Feature activation+1.878
faith
Token faith
Feature activation+4.061
of
Token of
Feature activation+1.959
some
Token some
Feature activation+2.595
weaker
Token weaker
Feature activation+2.300
Christians
Token Christians
Feature activation+4.524
and
Token and
Feature activation+2.194
stun
Token stun
Feature activation+1.580
ting
Tokenting
Feature activation+2.631
their
Token their
Feature activation+1.415
growth
Token growth
Feature activation+4.419

INTERVAL 3.557 - 4.149
CONTAINS 0.010%

llo
Tokenllo
Feature activation+0.603
would
Token would
Feature activation+2.689
be
Token be
Feature activation+1.148
the
Token the
Feature activation+1.108
biggest
Token biggest
Feature activation+0.871
beneficiaries
Token beneficiaries
Feature activation+3.596
of
Token of
Feature activation+0.902
the
Token the
Feature activation+0.852
Rangers
Token Rangers
Feature activation+0.749
âĢ
TokenâĢ
Feature activation+0.504
Ļ
TokenĻ
Feature activation+1.710
is
Token is
Feature activation+0.000
that
Token that
Feature activation+0.000
the
Token the
Feature activation+0.062
connection
Token connection
Feature activation+2.649
that
Token that
Feature activation+1.982
inspired
Token inspired
Feature activation+3.838
the
Token the
Feature activation+1.766
Eagles
Token Eagles
Feature activation+2.068
to
Token to
Feature activation+1.914
sign
Token sign
Feature activation+1.754
the
Token the
Feature activation+1.234
same
Token same
Feature activation+1.351
old
Token old
Feature activation+3.549
,
Token,
Feature activation+1.633
billions
Token billions
Feature activation+2.979
of
Token of
Feature activation+0.831
dollars
Token dollars
Feature activation+3.595
in
Token in
Feature activation+1.035
spending
Token spending
Feature activation+2.955
and
Token and
Feature activation+0.849
higher
Token higher
Feature activation+1.141
taxes
Token taxes
Feature activation+3.408
it
Token it
Feature activation+0.276
the
Token the
Feature activation+1.137
right
Token right
Feature activation+1.618
kind
Token kind
Feature activation+1.627
of
Token of
Feature activation+1.390
action
Token action
Feature activation+3.793
?
Token?
Feature activation+0.000
Extrem
Token Extrem
Feature activation+0.000
ism
Tokenism
Feature activation+0.454
researchers
Token researchers
Feature activation+0.035
said
Token said
Feature activation+0.000
How
Token How
Feature activation+0.000
will
Token will
Feature activation+0.000
you
Token you
Feature activation+0.265
proceed
Token proceed
Feature activation+2.773
to
Token to
Feature activation+2.419
build
Token build
Feature activation+3.558
and
Token and
Feature activation+2.186
publish
Token publish
Feature activation+3.362
an
Token an
Feature activation+2.161
app
Token app
Feature activation+3.584
like
Token like
Feature activation+2.480

INTERVAL 2.964 - 3.557
CONTAINS 0.026%

"
Token "
Feature activation+1.675
order
Tokenorder
Feature activation+0.000
"
Token"
Feature activation+0.004
come
Token come
Feature activation+0.213
from
Token from
Feature activation+1.617
"
Token "
Feature activation+3.548
dis
Tokendis
Feature activation+0.986
order
Tokenorder
Feature activation+1.999
"...
Token"...
Feature activation+2.081
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
coronary
Token coronary
Feature activation+0.658
arteries
Token arteries
Feature activation+3.436
that
Token that
Feature activation+1.120
feed
Token feed
Feature activation+1.788
our
Token our
Feature activation+1.351
heart
Token heart
Feature activation+3.416
?
Token?
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
There
TokenThere
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
"
Token "
Feature activation+0.000
How
TokenHow
Feature activation+0.000
can
Token can
Feature activation+0.000
you
Token you
Feature activation+0.924
charge
Token charge
Feature activation+1.150
developers
Token developers
Feature activation+3.037
60
Token 60
Feature activation+0.468
or
Token or
Feature activation+0.648
70
Token 70
Feature activation+0.760
percent
Token percent
Feature activation+1.260
?
Token?
Feature activation+0.000
!
Token!
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Ready
TokenReady
Feature activation+0.000
to
Token to
Feature activation+0.000
fight
Token fight
Feature activation+2.065
back
Token back
Feature activation+3.081
?
Token?
Feature activation+0.000
Sign
Token Sign
Feature activation+0.000
up
Token up
Feature activation+0.000
for
Token for
Feature activation+0.000
Take
Token Take
Feature activation+0.000
the
Token the
Feature activation+1.083
new
Token new
Feature activation+1.733
line
Token line
Feature activation+1.289
of
Token of
Feature activation+1.056
wear
Token wear
Feature activation+0.345
ables
Tokenables
Feature activation+2.968
?
Token?
Feature activation+0.000
Are
Token Are
Feature activation+0.000
you
Token you
Feature activation+0.716
interested
Token interested
Feature activation+0.796
in
Token in
Feature activation+1.616

INTERVAL 2.371 - 2.964
CONTAINS 0.066%

issue
Token issue
Feature activation+0.000
who
Token who
Feature activation+0.000
was
Token was
Feature activation+0.199
it
Token it
Feature activation+0.000
who
Token who
Feature activation+1.436
stood
Token stood
Feature activation+2.580
against
Token against
Feature activation+3.019
so
Token so
Feature activation+1.693
much
Token much
Feature activation+2.461
world
Token world
Feature activation+0.961
opinion
Token opinion
Feature activation+2.541
.
Token.
Feature activation+0.000
How
Token How
Feature activation+0.000
do
Token do
Feature activation+0.000
you
Token you
Feature activation+0.000
sell
Token sell
Feature activation+2.124
hope
Token hope
Feature activation+2.391
to
Token to
Feature activation+2.727
a
Token a
Feature activation+2.729
generation
Token generation
Feature activation+2.915
that
Token that
Feature activation+0.775
can
Token can
Feature activation+1.089
Stanley
Token Stanley
Feature activation+2.828
's
Token's
Feature activation+1.611
officials
Token officials
Feature activation+2.799
when
Token when
Feature activation+0.602
she
Token she
Feature activation+2.094
was
Token was
Feature activation+2.395
with
Token with
Feature activation+1.821
them
Token them
Feature activation+2.812
on
Token on
Feature activation+1.951
the
Token the
Feature activation+1.089
TPP
Token TPP
Feature activation+2.587
Volcano
Token Volcano
Feature activation+1.087
with
Token with
Feature activation+1.376
all
Token all
Feature activation+0.799
of
Token of
Feature activation+1.778
its
Token its
Feature activation+0.808
amenities
Token amenities
Feature activation+2.432
?
Token?
Feature activation+0.000
The
Token The
Feature activation+0.000
secret
Token secret
Feature activation+0.000
headquarters
Token headquarters
Feature activation+0.000
of
Token of
Feature activation+0.000
an
Token an
Feature activation+2.712
ang
Token ang
Feature activation+0.552
rier
Tokenrier
Feature activation+2.361
child
Token child
Feature activation+4.945
spawned
Token spawned
Feature activation+2.616
of
Token of
Feature activation+2.777
unusually
Token unusually
Feature activation+0.578
warm
Token warm
Feature activation+1.541
ocean
Token ocean
Feature activation+2.450
waters
Token waters
Feature activation+5.784
?
Token?
Feature activation+0.000

INTERVAL 1.778 - 2.371
CONTAINS 0.121%

t
Tokent
Feature activation+0.000
the
Token the
Feature activation+0.000
casting
Token casting
Feature activation+0.121
of
Token of
Feature activation+0.815
votes
Token votes
Feature activation+1.494
online
Token online
Feature activation+1.843
pretty
Token pretty
Feature activation+0.683
much
Token much
Feature activation+1.021
the
Token the
Feature activation+1.500
same
Token same
Feature activation+2.621
as
Token as
Feature activation+2.672
incumbent
Token incumbent
Feature activation+0.000
land
Token land
Feature activation+0.137
owners
Token owners
Feature activation+0.541
gain
Token gain
Feature activation+1.152
or
Token or
Feature activation+1.379
lose
Token lose
Feature activation+2.340
?
Token?
Feature activation+0.000
3
Token3
Feature activation+0.000
.
Token.
Feature activation+0.000
For
Token For
Feature activation+0.000
urban
Token urban
Feature activation+0.000
asured
Tokenasured
Feature activation+0.614
factor
Token factor
Feature activation+1.936
that
Token that
Feature activation+0.828
is
Token is
Feature activation+1.785
affecting
Token affecting
Feature activation+1.871
all
Token all
Feature activation+2.016
three
Token three
Feature activation+2.121
variables
Token variables
Feature activation+2.459
?
Token?
Feature activation+0.000
The
Token The
Feature activation+0.000
special
Token special
Feature activation+0.000
he
Token he
Feature activation+0.090
focus
Token focus
Feature activation+0.000
on
Token on
Feature activation+1.545
Ukraine
Token Ukraine
Feature activation+2.860
again
Token again
Feature activation+2.803
,
Token,
Feature activation+1.821
or
Token or
Feature activation+0.768
will
Token will
Feature activation+0.000
he
Token he
Feature activation+1.325
get
Token get
Feature activation+2.108
involved
Token involved
Feature activation+2.275
âĤ¬
TokenâĤ¬
Feature activation+0.317
Ëľ
TokenËľ
Feature activation+0.876
for
Tokenfor
Feature activation+1.185
loops
Token loops
Feature activation+1.421
âĢ
TokenâĢ
Feature activation+0.468
Ļ
TokenĻ
Feature activation+1.978
,
Token,
Feature activation+2.096
Ã
Token Ã
Feature activation+0.191
¢
Token¢
Feature activation+0.211
âĤ¬
TokenâĤ¬
Feature activation+0.460
Ëľ
TokenËľ
Feature activation+2.003

INTERVAL 1.186 - 1.778
CONTAINS 0.171%

on
Token on
Feature activation+0.000
them
Token them
Feature activation+0.276
.
Token.
Feature activation+0.000
Can
Token Can
Feature activation+0.000
you
Token you
Feature activation+0.134
imagine
Token imagine
Feature activation+1.568
the
Token the
Feature activation+0.184
backlash
Token backlash
Feature activation+1.832
?
Token?
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
as
Token as
Feature activation+1.231
we
Token we
Feature activation+3.164
allowed
Token allowed
Feature activation+2.679
dom
Token dom
Feature activation+0.233
ino
Tokenino
Feature activation+0.348
es
Tokenes
Feature activation+1.567
to
Token to
Feature activation+1.132
dominant
Token dominant
Feature activation+1.469
strategy
Token strategy
Feature activation+3.328
in
Token in
Feature activation+1.747
the
Token the
Feature activation+1.005
to
Token to
Feature activation+2.215
find
Token find
Feature activation+1.877
Founders
Token Founders
Feature activation+0.726
beer
Token beer
Feature activation+2.086
somewhere
Token somewhere
Feature activation+1.838
you
Token you
Feature activation+1.600
wish
Token wish
Feature activation+1.900
you
Token you
Feature activation+0.479
could
Token could
Feature activation+0.655
?
Token?
Feature activation+0.000
Are
Token Are
Feature activation+0.000
does
Token does
Feature activation+0.000
the
Token the
Feature activation+0.000
NBA
Token NBA
Feature activation+0.000
really
Token really
Feature activation+0.662
have
Token have
Feature activation+1.115
to
Token to
Feature activation+1.252
make
Token make
Feature activation+2.056
such
Token such
Feature activation+0.942
threats
Token threats
Feature activation+2.207
?
Token?
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
?
Token?
Feature activation+0.000
Would
Token Would
Feature activation+0.000
he
Token he
Feature activation+0.000
ever
Token ever
Feature activation+0.740
want
Token want
Feature activation+1.687
to
Token to
Feature activation+1.631
do
Token do
Feature activation+1.834
anything
Token anything
Feature activation+1.698
?
Token?
Feature activation+0.000
What
Token What
Feature activation+0.000
is
Token is
Feature activation+0.195

INTERVAL 0.593 - 1.186
CONTAINS 0.231%

going
Token going
Feature activation+0.000
on
Token on
Feature activation+0.829
,
Token,
Feature activation+0.000
precisely
Token precisely
Feature activation+0.236
,
Token,
Feature activation+0.625
in
Token in
Feature activation+0.947
1996
Token 1996
Feature activation+1.303
?
Token?
Feature activation+0.000
Let
Token Let
Feature activation+0.000
's
Token's
Feature activation+0.000
take
Token take
Feature activation+0.000
might
Token might
Feature activation+0.000
Rick
Token Rick
Feature activation+0.000
's
Token's
Feature activation+0.000
reaction
Token reaction
Feature activation+0.000
to
Token to
Feature activation+1.290
that
Token that
Feature activation+0.871
look
Token look
Feature activation+0.000
like
Token like
Feature activation+2.808
?
Token?
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
abuse
Token abuse
Feature activation+0.000
,
Token,
Feature activation+0.000
but
Token but
Feature activation+0.000
what
Token what
Feature activation+0.000
about
Token about
Feature activation+0.000
the
Token the
Feature activation+0.969
film
Token film
Feature activation+2.303
industry
Token industry
Feature activation+1.727
?
Token?
Feature activation+0.000
The
Token The
Feature activation+0.115
world
Token world
Feature activation+0.000
,
Token,
Feature activation+0.356
or
Token or
Feature activation+0.000
just
Token just
Feature activation+0.736
want
Token want
Feature activation+0.837
to
Token to
Feature activation+1.004
tell
Token tell
Feature activation+0.861
me
Token me
Feature activation+1.093
something
Token something
Feature activation+1.267
funny
Token funny
Feature activation+1.691
?
Token?
Feature activation+0.398
Email
Token Email
Feature activation+0.000
on
Token on
Feature activation+1.956
your
Token your
Feature activation+1.104
way
Token way
Feature activation+1.191
home
Token home
Feature activation+3.376
after
Token after
Feature activation+0.862
you
Token you
Feature activation+0.879
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
The
TokenThe
Feature activation+0.000
extremely
Token extremely
Feature activation+0.000
small
Token small
Feature activation+0.000
end
Token end
Feature activation+0.000

INTERVAL 0.000 - 0.593
CONTAINS 99.371%

j
Tokenj
Feature activation+0.000
s
Token s
Feature activation+0.000
k
Token k
Feature activation+0.000
oj
Tokenoj
Feature activation+0.000
eg
Tokeneg
Feature activation+0.000
se
Token se
Feature activation+0.000
gra
Token gra
Feature activation+0.000
Ä
TokenÄ
Feature activation+0.000
ij
Tokenij
Feature activation+0.000
an
Tokenan
Feature activation+0.000
ima
Tokenima
Feature activation+0.000
do
Token do
Feature activation+0.000
not
Token not
Feature activation+0.000
have
Token have
Feature activation+0.000
the
Token the
Feature activation+0.000
luxury
Token luxury
Feature activation+0.000
to
Token to
Feature activation+0.000
stop
Token stop
Feature activation+0.000
,
Token,
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
Ãĸ
Token Ãĸ
Feature activation+0.000
distortions
Token distortions
Feature activation+0.000
they
Token they
Feature activation+0.000
are
Token are
Feature activation+0.000
."
Token."
Feature activation+0.000
"
Token "
Feature activation+0.000
The
TokenThe
Feature activation+0.000
burning
Token burning
Feature activation+0.000
flag
Token flag
Feature activation+0.000
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
nationalist
Token nationalist
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
at
Token at
Feature activation+0.000
home
Token home
Feature activation+0.000
,
Token,
Feature activation+0.000
otherwise
Token otherwise
Feature activation+0.000
you
Token you
Feature activation+0.000
might
Token might
Feature activation+0.000
leave
Token leave
Feature activation+0.000
the
Token the
Feature activation+0.000
A
Token A
Feature activation+0.000
Western
Token Western
Feature activation+0.000
Canada
Token Canada
Feature activation+0.000
have
Token have
Feature activation+0.000
a
Token a
Feature activation+0.000
semi
Token semi
Feature activation+0.000
-
Token-
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
id
Tokenid
Feature activation+0.000
climate
Token climate
Feature activation+0.000
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000

BOTTOM ACTIVATIONS
MIN = 0.000

atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
&
Token &
Feature activation+0.000
Development
Token Development
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
&
Token &
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
rik
Tokenrik
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
rik
Tokenrik
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000

Top feature 6 in H8.5: (feature 16981)

TOP ACTIVATIONS
MAX = 4.926

as
Token as
Feature activation+0.000
the
Token the
Feature activation+0.814
result
Token result
Feature activation+4.917
of
Token of
Feature activation+2.715
so
Token so
Feature activation+0.945
doing
Token doing
Feature activation+4.926
,
Token,
Feature activation+0.000
civilization
Token civilization
Feature activation+0.000
may
Token may
Feature activation+0.000
be
Token be
Feature activation+0.000
checked
Token checked
Feature activation+0.000
and
Token and
Feature activation+0.000
how
Token how
Feature activation+0.000
,
Token,
Feature activation+0.000
as
Token as
Feature activation+0.000
the
Token the
Feature activation+0.814
result
Token result
Feature activation+4.917
of
Token of
Feature activation+2.715
so
Token so
Feature activation+0.945
doing
Token doing
Feature activation+4.926
,
Token,
Feature activation+0.000
civilization
Token civilization
Feature activation+0.000
--
Token--
Feature activation+0.000
a
Tokena
Feature activation+0.000
sign
Token sign
Feature activation+0.000
,
Token,
Feature activation+0.000
he
Token he
Feature activation+0.442
says
Token says
Feature activation+4.273
,
Token,
Feature activation+0.000
of
Token of
Feature activation+0.000
my
Token my
Feature activation+0.000
"
Token "
Feature activation+0.000
cultural
Tokencultural
Feature activation+0.000
egregious
Token egregious
Feature activation+0.000
a
Token a
Feature activation+0.000
failing
Token failing
Feature activation+0.000
than
Token than
Feature activation+0.000
,
Token,
Feature activation+0.000
say
Token say
Feature activation+4.109
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
Jays
Token Jays
Feature activation+0.000
on
Tokenon
Feature activation+0.000
Blair
Token Blair
Feature activation+0.000
differences
Token differences
Feature activation+0.000
are
Token are
Feature activation+0.000
,
Token,
Feature activation+0.000
in
Token in
Feature activation+0.366
many
Token many
Feature activation+0.631
cases
Token cases
Feature activation+3.940
,
Token,
Feature activation+0.000
central
Token central
Feature activation+0.000
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
very
Token very
Feature activation+0.000
solar
Token solar
Feature activation+0.000
is
Token is
Feature activation+0.000
,
Token,
Feature activation+0.000
at
Token at
Feature activation+0.000
long
Token long
Feature activation+0.204
last
Token last
Feature activation+3.920
,
Token,
Feature activation+0.000
ready
Token ready
Feature activation+0.000
to
Token to
Feature activation+0.000
move
Token move
Feature activation+0.000
from
Token from
Feature activation+0.000
the
Token the
Feature activation+0.000
result
Token result
Feature activation+0.000
,
Token,
Feature activation+0.000
it
Token it
Feature activation+0.374
turned
Token turned
Feature activation+2.601
out
Token out
Feature activation+3.890
,
Token,
Feature activation+0.000
of
Token of
Feature activation+0.000
an
Token an
Feature activation+0.000
influx
Token influx
Feature activation+0.000
of
Token of
Feature activation+0.000
helpful
Token helpful
Feature activation+0.000
to
Token to
Feature activation+0.000
know
Token know
Feature activation+0.000
,
Token,
Feature activation+0.000
for
Token for
Feature activation+0.014
example
Token example
Feature activation+3.876
,
Token,
Feature activation+0.000
that
Token that
Feature activation+0.000
sometime
Token sometime
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
is
Token is
Feature activation+0.000
,
Token,
Feature activation+0.000
for
Token for
Feature activation+0.000
the
Token the
Feature activation+0.524
most
Token most
Feature activation+1.168
part
Token part
Feature activation+3.841
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
order
Token order
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
exceed
Token exceed
Feature activation+0.000
,
Token,
Feature activation+0.000
by
Token by
Feature activation+2.302
orders
Token orders
Feature activation+2.194
of
Token of
Feature activation+1.982
magnitude
Token magnitude
Feature activation+3.807
,
Token,
Feature activation+0.000
those
Token those
Feature activation+0.000
found
Token found
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
way
Token way
Feature activation+0.000
that
Token that
Feature activation+0.000
,
Token,
Feature activation+0.000
despite
Token despite
Feature activation+0.075
those
Token those
Feature activation+0.659
changes
Token changes
Feature activation+3.757
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
economy
Token economy
Feature activation+0.000
does
Token does
Feature activation+0.000
not
Token not
Feature activation+0.000
happens
Token happens
Feature activation+0.000
when
Token when
Feature activation+0.000
,
Token,
Feature activation+0.000
for
Token for
Feature activation+0.236
whatever
Token whatever
Feature activation+2.050
reason
Token reason
Feature activation+3.752
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
Kim
Token Kim
Feature activation+0.000
regime
Token regime
Feature activation+0.000
can
Token can
Feature activation+0.000
for
Token for
Feature activation+0.087
more
Token more
Feature activation+0.516
than
Token than
Feature activation+1.062
a
Token a
Feature activation+0.707
year
Token year
Feature activation+3.669
now
Token now
Feature activation+3.695
,
Token,
Feature activation+0.000
been
Token been
Feature activation+0.000
pushing
Token pushing
Feature activation+0.000
for
Token for
Feature activation+0.000
professional
Token professional
Feature activation+0.000
,
Token,
Feature activation+0.000
for
Token for
Feature activation+0.087
more
Token more
Feature activation+0.516
than
Token than
Feature activation+1.062
a
Token a
Feature activation+0.707
year
Token year
Feature activation+3.669
now
Token now
Feature activation+3.695
,
Token,
Feature activation+0.000
been
Token been
Feature activation+0.000
pushing
Token pushing
Feature activation+0.000
for
Token for
Feature activation+0.000
a
Token a
Feature activation+0.000
career
Token career
Feature activation+0.000
and
Token and
Feature activation+0.000
,
Token,
Feature activation+0.000
one
Token one
Feature activation+0.000
day
Token day
Feature activation+3.658
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
family
Token family
Feature activation+0.000
.
Token.
Feature activation+0.000
But
Token But
Feature activation+0.000
to
Token to
Feature activation+0.000
describe
Token describe
Feature activation+0.000
,
Token,
Feature activation+0.000
in
Token in
Feature activation+0.266
great
Token great
Feature activation+0.898
detail
Token detail
Feature activation+3.527
,
Token,
Feature activation+0.000
other
Token other
Feature activation+0.000
incidents
Token incidents
Feature activation+0.000
of
Token of
Feature activation+0.000
alleged
Token alleged
Feature activation+0.000
Independent
Token Independent
Feature activation+0.000
who
Token who
Feature activation+0.000
,
Token,
Feature activation+0.000
until
Token until
Feature activation+0.593
this
Token this
Feature activation+1.018
summer
Token summer
Feature activation+3.509
,
Token,
Feature activation+0.000
had
Token had
Feature activation+0.000
never
Token never
Feature activation+0.000
attended
Token attended
Feature activation+0.000
a
Token a
Feature activation+0.000
,
Token,
Feature activation+0.000
one
Token one
Feature activation+0.000
-
Token-
Feature activation+0.000
by
Tokenby
Feature activation+0.804
-
Token-
Feature activation+0.977
one
Tokenone
Feature activation+3.458
,
Token,
Feature activation+0.000
our
Token our
Feature activation+0.000
musical
Token musical
Feature activation+0.000
tastes
Token tastes
Feature activation+0.000
,
Token,
Feature activation+0.000
,
Token,
Feature activation+0.000
imperfect
Token imperfect
Feature activation+0.368
as
Token as
Feature activation+1.336
it
Token it
Feature activation+1.753
may
Token may
Feature activation+2.136
be
Token be
Feature activation+3.457
,
Token,
Feature activation+0.000
is
Token is
Feature activation+0.000
a
Token a
Feature activation+0.000
democracy
Token democracy
Feature activation+0.000
and
Token and
Feature activation+0.000
as
Token as
Feature activation+0.000
such
Token such
Feature activation+0.000
was
Token was
Feature activation+0.000
,
Token,
Feature activation+0.000
in
Token in
Feature activation+0.092
effect
Token effect
Feature activation+3.445
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
local
Token local
Feature activation+0.000
stamp
Token stamp
Feature activation+0.000
.
Token.
Feature activation+0.000

Top DFA by src position
MAX = 5.463

with
Token with
Feature activation-0.066
Top resid features:
them
Token them
Feature activation-0.046
Top resid features:
,
Token,
Feature activation-0.013
Top resid features:
and
Token and
Feature activation+0.000
Top resid features:
how
Token how
Feature activation-0.029
Top resid features:
,
Token,
Feature activation+5.463
Top resid features:
as
Token as
Feature activation+0.023
Top resid features:
the
Token the
Feature activation+0.001
Top resid features:
result
Token result
Feature activation+0.008
Top resid features:
of
Token of
Feature activation+0.004
Top resid features:
so
Token so
Feature activation+0.009
Top resid features:
with
Token with
Feature activation-0.016
Top resid features:
them
Token them
Feature activation-0.027
Top resid features:
,
Token,
Feature activation-0.012
Top resid features:
and
Token and
Feature activation+0.001
Top resid features:
how
Token how
Feature activation-0.012
Top resid features:
,
Token,
Feature activation+5.248
Top resid features:
as
Token as
Feature activation+0.091
Top resid features:
the
Token the
Feature activation+0.008
Top resid features:
result
Token result
Feature activation+0.003
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
so
Token so
Feature activation+0.000
Top resid features:
National
Token National
Feature activation-0.001
Top resid features:
Anthem
Token Anthem
Feature activation+0.003
Top resid features:
--
Token--
Feature activation-0.002
Top resid features:
a
Tokena
Feature activation+0.005
Top resid features:
sign
Token sign
Feature activation-0.002
Top resid features:
,
Token,
Feature activation+4.549
Top resid features:
he
Token he
Feature activation+0.063
Top resid features:
says
Token says
Feature activation+0.019
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
my
Token my
Feature activation+0.000
Top resid features:
more
Token more
Feature activation+0.004
Top resid features:
egregious
Token egregious
Feature activation+0.004
Top resid features:
a
Token a
Feature activation+0.001
Top resid features:
failing
Token failing
Feature activation+0.001
Top resid features:
than
Token than
Feature activation+0.012
Top resid features:
,
Token,
Feature activation+4.335
Top resid features:
say
Token say
Feature activation+0.020
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
Jays
Token Jays
Feature activation+0.000
Top resid features:
on
Tokenon
Feature activation+0.000
Top resid features:
existence
Token existence
Feature activation+0.005
Top resid features:
.
Token.
Feature activation+0.002
Top resid features:
Their
Token Their
Feature activation+0.028
Top resid features:
differences
Token differences
Feature activation+0.019
Top resid features:
are
Token are
Feature activation+0.003
Top resid features:
,
Token,
Feature activation+4.061
Top resid features:
in
Token in
Feature activation+0.114
Top resid features:
many
Token many
Feature activation+0.017
Top resid features:
cases
Token cases
Feature activation+0.059
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
central
Token central
Feature activation+0.000
Top resid features:
advocates
Token advocates
Feature activation+0.002
Top resid features:
say
Token say
Feature activation-0.001
Top resid features:
that
Token that
Feature activation+0.003
Top resid features:
solar
Token solar
Feature activation+0.014
Top resid features:
is
Token is
Feature activation+0.006
Top resid features:
,
Token,
Feature activation+4.183
Top resid features:
at
Token at
Feature activation+0.062
Top resid features:
long
Token long
Feature activation+0.006
Top resid features:
last
Token last
Feature activation+0.029
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
ready
Token ready
Feature activation+0.000
Top resid features:
in
Token in
Feature activation-0.000
Top resid features:
attendance
Token attendance
Feature activation-0.002
Top resid features:
âĢĶ
Token âĢĶ
Feature activation-0.012
Top resid features:
the
Token the
Feature activation+0.007
Top resid features:
result
Token result
Feature activation+0.020
Top resid features:
,
Token,
Feature activation+3.876
Top resid features:
it
Token it
Feature activation+0.095
Top resid features:
turned
Token turned
Feature activation+0.260
Top resid features:
out
Token out
Feature activation+0.065
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
will
Token will
Feature activation+0.000
Top resid features:
be
Token be
Feature activation-0.001
Top resid features:
helpful
Token helpful
Feature activation-0.003
Top resid features:
to
Token to
Feature activation-0.008
Top resid features:
know
Token know
Feature activation-0.031
Top resid features:
,
Token,
Feature activation+4.264
Top resid features:
for
Token for
Feature activation+0.074
Top resid features:
example
Token example
Feature activation+0.013
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
that
Token that
Feature activation+0.000
Top resid features:
sometime
Token sometime
Feature activation+0.000
Top resid features:
Europe
Token Europe
Feature activation+0.003
Top resid features:
,
Token,
Feature activation-0.006
Top resid features:
civic
Token civic
Feature activation+0.004
Top resid features:
nationalism
Token nationalism
Feature activation-0.006
Top resid features:
is
Token is
Feature activation-0.010
Top resid features:
,
Token,
Feature activation+4.202
Top resid features:
for
Token for
Feature activation+0.012
Top resid features:
the
Token the
Feature activation+0.008
Top resid features:
most
Token most
Feature activation+0.021
Top resid features:
part
Token part
Feature activation+0.001
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
radically
Token radically
Feature activation-0.003
Top resid features:
and
Token and
Feature activation+0.013
Top resid features:
unjust
Token unjust
Feature activation+0.006
Top resid features:
ly
Tokenly
Feature activation-0.006
Top resid features:
exceed
Token exceed
Feature activation-0.043
Top resid features:
,
Token,
Feature activation+3.997
Top resid features:
by
Token by
Feature activation+0.160
Top resid features:
orders
Token orders
Feature activation+0.030
Top resid features:
of
Token of
Feature activation+0.009
Top resid features:
magnitude
Token magnitude
Feature activation+0.075
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
in
Token in
Feature activation-0.001
Top resid features:
such
Token such
Feature activation-0.000
Top resid features:
a
Token a
Feature activation-0.004
Top resid features:
way
Token way
Feature activation-0.001
Top resid features:
that
Token that
Feature activation-0.002
Top resid features:
,
Token,
Feature activation+3.929
Top resid features:
despite
Token despite
Feature activation+0.096
Top resid features:
those
Token those
Feature activation+0.158
Top resid features:
changes
Token changes
Feature activation+0.003
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
Chinese
Token Chinese
Feature activation-0.051
Top resid features:
counterparts
Token counterparts
Feature activation+0.007
Top resid features:
what
Token what
Feature activation+0.006
Top resid features:
happens
Token happens
Feature activation+0.007
Top resid features:
when
Token when
Feature activation-0.016
Top resid features:
,
Token,
Feature activation+3.887
Top resid features:
for
Token for
Feature activation+0.037
Top resid features:
whatever
Token whatever
Feature activation+0.048
Top resid features:
reason
Token reason
Feature activation+0.065
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.021
Top resid features:
because
Token because
Feature activation+0.015
Top resid features:
Adobe
Token Adobe
Feature activation-0.029
Top resid features:
has
Token has
Feature activation-0.004
Top resid features:
,
Token,
Feature activation+3.847
Top resid features:
for
Token for
Feature activation+0.052
Top resid features:
more
Token more
Feature activation+0.013
Top resid features:
than
Token than
Feature activation+0.017
Top resid features:
a
Token a
Feature activation+0.007
Top resid features:
year
Token year
Feature activation+0.092
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.019
Top resid features:
because
Token because
Feature activation+0.010
Top resid features:
Adobe
Token Adobe
Feature activation-0.015
Top resid features:
has
Token has
Feature activation-0.010
Top resid features:
,
Token,
Feature activation+3.884
Top resid features:
for
Token for
Feature activation+0.046
Top resid features:
more
Token more
Feature activation+0.016
Top resid features:
than
Token than
Feature activation+0.026
Top resid features:
a
Token a
Feature activation+0.008
Top resid features:
year
Token year
Feature activation+0.067
Top resid features:
to
Token to
Feature activation-0.022
Top resid features:
pursue
Token pursue
Feature activation-0.012
Top resid features:
a
Token a
Feature activation-0.010
Top resid features:
career
Token career
Feature activation-0.006
Top resid features:
and
Token and
Feature activation-0.004
Top resid features:
,
Token,
Feature activation+3.899
Top resid features:
one
Token one
Feature activation+0.158
Top resid features:
day
Token day
Feature activation+0.029
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
a
Token a
Feature activation+0.000
Top resid features:
family
Token family
Feature activation+0.000
Top resid features:
and
Token and
Feature activation+0.001
Top resid features:
went
Token went
Feature activation+0.006
Top resid features:
on
Token on
Feature activation+0.004
Top resid features:
to
Token to
Feature activation-0.002
Top resid features:
describe
Token describe
Feature activation-0.062
Top resid features:
,
Token,
Feature activation+3.998
Top resid features:
in
Token in
Feature activation+0.082
Top resid features:
great
Token great
Feature activation+0.016
Top resid features:
detail
Token detail
Feature activation+0.007
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
other
Token other
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.010
Top resid features:
the
Token the
Feature activation-0.005
Top resid features:
registered
Token registered
Feature activation-0.005
Top resid features:
Independent
Token Independent
Feature activation+0.003
Top resid features:
who
Token who
Feature activation-0.078
Top resid features:
,
Token,
Feature activation+3.794
Top resid features:
until
Token until
Feature activation+0.063
Top resid features:
this
Token this
Feature activation+0.000
Top resid features:
summer
Token summer
Feature activation+0.013
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
had
Token had
Feature activation+0.000
Top resid features:
ideas
Token ideas
Feature activation+0.001
Top resid features:
âĢĵ
Token âĢĵ
Feature activation-0.004
Top resid features:
which
Token which
Feature activation+0.007
Top resid features:
is
Token is
Feature activation-0.007
Top resid features:
why
Token why
Feature activation-0.008
Top resid features:
,
Token,
Feature activation+3.836
Top resid features:
one
Token one
Feature activation+0.026
Top resid features:
-
Token-
Feature activation+0.010
Top resid features:
by
Tokenby
Feature activation-0.015
Top resid features:
-
Token-
Feature activation+0.001
Top resid features:
one
Tokenone
Feature activation-0.014
Top resid features:
time
Token time
Feature activation-0.003
Top resid features:
in
Token in
Feature activation-0.003
Top resid features:
a
Token a
Feature activation-0.002
Top resid features:
country
Token country
Feature activation-0.019
Top resid features:
that
Token that
Feature activation+0.013
Top resid features:
,
Token,
Feature activation+3.807
Top resid features:
imperfect
Token imperfect
Feature activation+0.048
Top resid features:
as
Token as
Feature activation+0.012
Top resid features:
it
Token it
Feature activation-0.014
Top resid features:
may
Token may
Feature activation+0.014
Top resid features:
be
Token be
Feature activation-0.035
Top resid features:
Ireland
Token Ireland
Feature activation-0.009
Top resid features:
and
Token and
Feature activation+0.003
Top resid features:
as
Token as
Feature activation-0.007
Top resid features:
such
Token such
Feature activation-0.001
Top resid features:
was
Token was
Feature activation-0.013
Top resid features:
,
Token,
Feature activation+3.735
Top resid features:
in
Token in
Feature activation+0.112
Top resid features:
effect
Token effect
Feature activation+0.058
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
a
Token a
Feature activation+0.000
Top resid features:
local
Token local
Feature activation+0.000
Top resid features:

Decoder Weights Distribution

Head 0: 0.06

Head 1: 0.01

Head 2: 0.05

Head 3: 0.03

Head 4: 0.04

Head 5: 0.51

Head 6: 0.02

Head 7: 0.10

Head 8: 0.04

Head 9: 0.04

Head 10: 0.07

Head 11: 0.02

Positive logits

?,3.84

%,3.48

+,3.31

*,3.26

,[3.17

,,3.16

$,3.11

,...3.10

/,3.09

(),3.08

,3.03

!,3.02

",2.97

(),2.96

?",2.86

,2.83

,,,,,,,,2.79

',2.78

.,2.57

,,,,2.51

Negative logits

(@-2.20

anwhile-2.03

ngth-1.88

etheless-1.83

pione-1.79

namely-1.74

prus-1.73

��-1.73

ディ-1.71

razil-1.68

ウス-1.66

abwe-1.65

srf-1.64

perty-1.63

landish-1.61

(%)-1.57

conservancy-1.56

earthqu-1.51

cellaneous-1.49

odox-1.48

INTERVAL 4.433 - 4.926
CONTAINS 0.000%

and
Token and
Feature activation+0.000
how
Token how
Feature activation+0.000
,
Token,
Feature activation+0.000
as
Token as
Feature activation+0.000
the
Token the
Feature activation+0.814
result
Token result
Feature activation+4.917
of
Token of
Feature activation+2.715
so
Token so
Feature activation+0.945
doing
Token doing
Feature activation+4.926
,
Token,
Feature activation+0.000
civilization
Token civilization
Feature activation+0.000
as
Token as
Feature activation+0.000
the
Token the
Feature activation+0.814
result
Token result
Feature activation+4.917
of
Token of
Feature activation+2.715
so
Token so
Feature activation+0.945
doing
Token doing
Feature activation+4.926
,
Token,
Feature activation+0.000
civilization
Token civilization
Feature activation+0.000
may
Token may
Feature activation+0.000
be
Token be
Feature activation+0.000
checked
Token checked
Feature activation+0.000

INTERVAL 3.941 - 4.433
CONTAINS 0.000%

--
Token--
Feature activation+0.000
a
Tokena
Feature activation+0.000
sign
Token sign
Feature activation+0.000
,
Token,
Feature activation+0.000
he
Token he
Feature activation+0.442
says
Token says
Feature activation+4.273
,
Token,
Feature activation+0.000
of
Token of
Feature activation+0.000
my
Token my
Feature activation+0.000
"
Token "
Feature activation+0.000
cultural
Tokencultural
Feature activation+0.000
egregious
Token egregious
Feature activation+0.000
a
Token a
Feature activation+0.000
failing
Token failing
Feature activation+0.000
than
Token than
Feature activation+0.000
,
Token,
Feature activation+0.000
say
Token say
Feature activation+4.109
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
Jays
Token Jays
Feature activation+0.000
on
Tokenon
Feature activation+0.000
Blair
Token Blair
Feature activation+0.000

INTERVAL 3.448 - 3.941
CONTAINS 0.002%

differences
Token differences
Feature activation+0.000
are
Token are
Feature activation+0.000
,
Token,
Feature activation+0.000
in
Token in
Feature activation+0.366
many
Token many
Feature activation+0.631
cases
Token cases
Feature activation+3.940
,
Token,
Feature activation+0.000
central
Token central
Feature activation+0.000
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
very
Token very
Feature activation+0.000
Independent
Token Independent
Feature activation+0.000
who
Token who
Feature activation+0.000
,
Token,
Feature activation+0.000
until
Token until
Feature activation+0.593
this
Token this
Feature activation+1.018
summer
Token summer
Feature activation+3.509
,
Token,
Feature activation+0.000
had
Token had
Feature activation+0.000
never
Token never
Feature activation+0.000
attended
Token attended
Feature activation+0.000
a
Token a
Feature activation+0.000
,
Token,
Feature activation+0.000
one
Token one
Feature activation+0.000
-
Token-
Feature activation+0.000
by
Tokenby
Feature activation+0.804
-
Token-
Feature activation+0.977
one
Tokenone
Feature activation+3.458
,
Token,
Feature activation+0.000
our
Token our
Feature activation+0.000
musical
Token musical
Feature activation+0.000
tastes
Token tastes
Feature activation+0.000
,
Token,
Feature activation+0.000
is
Token is
Feature activation+0.000
,
Token,
Feature activation+0.000
for
Token for
Feature activation+0.000
the
Token the
Feature activation+0.524
most
Token most
Feature activation+1.168
part
Token part
Feature activation+3.841
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
order
Token order
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
way
Token way
Feature activation+0.000
that
Token that
Feature activation+0.000
,
Token,
Feature activation+0.000
despite
Token despite
Feature activation+0.075
those
Token those
Feature activation+0.659
changes
Token changes
Feature activation+3.757
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
economy
Token economy
Feature activation+0.000
does
Token does
Feature activation+0.000
not
Token not
Feature activation+0.000

INTERVAL 2.955 - 3.448
CONTAINS 0.003%

Ċ
TokenĊ
Feature activation+0.000
This
TokenThis
Feature activation+0.000
is
Token is
Feature activation+0.000
,
Token,
Feature activation+0.000
yet
Token yet
Feature activation+2.147
again
Token again
Feature activation+2.986
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
pivotal
Token pivotal
Feature activation+0.000
point
Token point
Feature activation+0.000
for
Token for
Feature activation+0.000
and
Token and
Feature activation+0.000
,
Token,
Feature activation+0.000
in
Token in
Feature activation+0.586
the
Token the
Feature activation+0.819
long
Token long
Feature activation+0.599
run
Token run
Feature activation+2.972
,
Token,
Feature activation+0.000
saves
Token saves
Feature activation+0.000
money
Token money
Feature activation+0.000
as
Token as
Feature activation+0.000
well
Token well
Feature activation+0.000
so
Token so
Feature activation+0.000
good
Token good
Feature activation+0.000
and
Token and
Feature activation+0.000
,
Token,
Feature activation+0.000
until
Token until
Feature activation+0.590
recently
Token recently
Feature activation+3.018
,
Token,
Feature activation+0.000
it
Token it
Feature activation+0.000
was
Token was
Feature activation+0.000
assumed
Token assumed
Feature activation+0.000
they
Token they
Feature activation+0.000
goes
Token goes
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+0.000
barring
Token barring
Feature activation+0.000
,
Token,
Feature activation+0.000
of
Token of
Feature activation+0.609
course
Token course
Feature activation+3.041
,
Token,
Feature activation+0.000
space
Token space
Feature activation+0.000
travel
Token travel
Feature activation+0.000
and
Token and
Feature activation+0.000
repl
Token repl
Feature activation+0.000
interactions
Token interactions
Feature activation+0.000
that
Token that
Feature activation+0.000
,
Token,
Feature activation+0.000
without
Token without
Feature activation+0.837
the
Token the
Feature activation+0.710
device
Token device
Feature activation+3.336
,
Token,
Feature activation+0.052
would
Token would
Feature activation+0.000
likely
Token likely
Feature activation+0.000
include
Token include
Feature activation+0.000
frustrating
Token frustrating
Feature activation+0.000

INTERVAL 2.463 - 2.955
CONTAINS 0.005%

classical
Token classical
Feature activation+0.000
texts
Token texts
Feature activation+0.000
and
Token and
Feature activation+0.000
,
Token,
Feature activation+0.000
in
Token in
Feature activation+1.024
general
Token general
Feature activation+2.690
,
Token,
Feature activation+0.000
of
Token of
Feature activation+0.000
Catholic
Token Catholic
Feature activation+0.000
literature
Token literature
Feature activation+0.000
,
Token,
Feature activation+0.000
,
Token,
Feature activation+0.000
in
Token in
Feature activation+0.000
one
Token one
Feature activation+0.152
way
Token way
Feature activation+2.225
or
Token or
Feature activation+0.726
another
Token another
Feature activation+2.856
and
Token and
Feature activation+0.416
in
Token in
Feature activation+0.651
their
Token their
Feature activation+0.294
own
Token own
Feature activation+0.174
best
Token best
Feature activation+0.148
and
Token and
Feature activation+0.000
diplomats
Token diplomats
Feature activation+0.000
have
Token have
Feature activation+0.000
,
Token,
Feature activation+0.000
for
Token for
Feature activation+0.098
decades
Token decades
Feature activation+2.654
,
Token,
Feature activation+0.000
tried
Token tried
Feature activation+0.000
to
Token to
Feature activation+0.000
discuss
Token discuss
Feature activation+0.000
with
Token with
Feature activation+0.000
,
Token,
Feature activation+0.000
according
Token according
Feature activation+0.306
to
Token to
Feature activation+1.164
accepted
Token accepted
Feature activation+0.134
Islamic
Token Islamic
Feature activation+0.161
sources
Token sources
Feature activation+2.827
,
Token,
Feature activation+0.000
Mohammed
Token Mohammed
Feature activation+0.000
,
Token,
Feature activation+0.000
at
Token at
Feature activation+0.000
age
Token age
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+0.000
the
Token the
Feature activation+0.000
result
Token result
Feature activation+0.000
,
Token,
Feature activation+0.000
it
Token it
Feature activation+0.374
turned
Token turned
Feature activation+2.601
out
Token out
Feature activation+3.890
,
Token,
Feature activation+0.000
of
Token of
Feature activation+0.000
an
Token an
Feature activation+0.000
influx
Token influx
Feature activation+0.000

INTERVAL 1.970 - 2.463
CONTAINS 0.009%

Friday
Token Friday
Feature activation+0.000
that
Token that
Feature activation+0.000
read
Token read
Feature activation+0.000
,
Token,
Feature activation+0.000
in
Token in
Feature activation+0.092
part
Token part
Feature activation+2.316
:
Token:
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
R
TokenR
Feature activation+0.000
ansom
Tokenansom
Feature activation+0.000
determine
Token determine
Feature activation+0.000
who
Token who
Feature activation+0.000
was
Token was
Feature activation+0.000
,
Token,
Feature activation+0.000
in
Token in
Feature activation+0.583
fact
Token fact
Feature activation+2.107
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
best
Token best
Feature activation+0.000
within
Token within
Feature activation+0.000
a
Token a
Feature activation+0.000
This
Token This
Feature activation+0.000
should
Token should
Feature activation+0.000
be
Token be
Feature activation+0.000
regarded
Token regarded
Feature activation+0.000
,
Token,
Feature activation+0.000
psychologically
Token psychologically
Feature activation+2.019
,
Token,
Feature activation+0.192
as
Token as
Feature activation+0.000
a
Token a
Feature activation+0.000
compens
Token compens
Feature activation+0.000
atory
Tokenatory
Feature activation+0.000
violence
Token violence
Feature activation+0.000
.
Token.
Feature activation+0.000
That
Token That
Feature activation+0.000
,
Token,
Feature activation+0.000
it
Token it
Feature activation+0.212
seems
Token seems
Feature activation+2.177
,
Token,
Feature activation+0.148
is
Token is
Feature activation+0.015
insufficient
Token insufficient
Feature activation+0.000
ly
Tokenly
Feature activation+0.000
ideological
Token ideological
Feature activation+0.000
announcement
Token announcement
Feature activation+0.000
is
Token is
Feature activation+0.000
,
Token,
Feature activation+0.000
in
Token in
Feature activation+0.522
some
Token some
Feature activation+0.169
ways
Token ways
Feature activation+2.230
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
big
Token big
Feature activation+0.000
change
Token change
Feature activation+0.000
that
Token that
Feature activation+0.000

INTERVAL 1.478 - 1.970
CONTAINS 0.015%

that
Token that
Feature activation+0.000
have
Token have
Feature activation+0.000
,
Token,
Feature activation+0.000
as
Token as
Feature activation+0.000
a
Token a
Feature activation+0.734
central
Token central
Feature activation+1.685
defining
Token defining
Feature activation+0.311
ten
Token ten
Feature activation+0.000
et
Tokenet
Feature activation+2.269
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
the
Token the
Feature activation+0.000
Army
Token Army
Feature activation+0.000
when
Token when
Feature activation+0.000
,
Token,
Feature activation+0.000
in
Token in
Feature activation+0.000
1969
Token 1969
Feature activation+1.849
,
Token,
Feature activation+0.000
he
Token he
Feature activation+0.000
found
Token found
Feature activation+0.000
himself
Token himself
Feature activation+0.000
in
Token in
Feature activation+0.000
"
Token "
Feature activation+0.000
which
Tokenwhich
Feature activation+0.000
,
Token,
Feature activation+0.000
six
Token six
Feature activation+0.400
months
Token months
Feature activation+1.734
ago
Token ago
Feature activation+1.798
,
Token,
Feature activation+0.000
was
Token was
Feature activation+0.000
very
Token very
Feature activation+0.000
much
Token much
Feature activation+0.000
in
Token in
Feature activation+0.000
has
Token has
Feature activation+0.000
no
Token no
Feature activation+0.000
business
Token business
Feature activation+0.000
,
Token,
Feature activation+0.000
in
Token in
Feature activation+0.396
general
Token general
Feature activation+1.892
,
Token,
Feature activation+0.134
decree
Token decree
Feature activation+0.657
ing
Tokening
Feature activation+0.000
who
Token who
Feature activation+0.000
may
Token may
Feature activation+0.000
am
Tokenam
Feature activation+0.155
oured
Tokenoured
Feature activation+1.231
of
Token of
Feature activation+0.744
their
Token their
Feature activation+0.237
tough
Token tough
Feature activation+0.000
guys
Token guys
Feature activation+1.482
,
Token,
Feature activation+0.138
appreciate
Token appreciate
Feature activation+0.000
.
Token.
Feature activation+0.000
In
Token In
Feature activation+0.000
the
Token the
Feature activation+0.000

INTERVAL 0.985 - 1.478
CONTAINS 0.039%

if
Token if
Feature activation+0.100
they
Token they
Feature activation+0.461
had
Token had
Feature activation+0.422
accepted
Token accepted
Feature activation+0.443
the
Token the
Feature activation+0.096
parts
Token parts
Feature activation+1.111
,
Token,
Feature activation+0.000
may
Token may
Feature activation+0.000
have
Token have
Feature activation+0.000
completely
Token completely
Feature activation+0.000
altered
Token altered
Feature activation+0.000
The
Token The
Feature activation+0.000
fundamental
Token fundamental
Feature activation+0.000
problem
Token problem
Feature activation+0.000
,
Token,
Feature activation+0.000
they
Token they
Feature activation+0.000
said
Token said
Feature activation+1.302
,
Token,
Feature activation+0.000
is
Token is
Feature activation+0.000
temperature
Token temperature
Feature activation+0.000
control
Token control
Feature activation+0.000
.
Token.
Feature activation+0.000
and
Token and
Feature activation+0.000
,
Token,
Feature activation+0.000
in
Token in
Feature activation+0.498
the
Token the
Feature activation+0.872
years
Token years
Feature activation+0.865
just
Token just
Feature activation+1.180
before
Token before
Feature activation+1.020
we
Token we
Feature activation+0.000
meet
Token meet
Feature activation+1.200
them
Token them
Feature activation+2.739
,
Token,
Feature activation+0.000
some
Token some
Feature activation+0.000
research
Token research
Feature activation+0.000
,
Token,
Feature activation+0.000
especially
Token especially
Feature activation+0.924
in
Token in
Feature activation+1.027
the
Token the
Feature activation+1.034
area
Token area
Feature activation+1.008
of
Token of
Feature activation+0.444
heart
Token heart
Feature activation+0.000
disease
Token disease
Feature activation+0.870
,
Token,
Feature activation+0.000
at
Token at
Feature activation+0.173
the
Token the
Feature activation+0.004
public
Token public
Feature activation+0.000
meeting
Token meeting
Feature activation+0.570
last
Token last
Feature activation+0.000
night
Token night
Feature activation+0.995
,
Token,
Feature activation+0.000
said
Token said
Feature activation+0.000
both
Token both
Feature activation+0.000
Dublin
Token Dublin
Feature activation+0.000
City
Token City
Feature activation+0.000

INTERVAL 0.493 - 0.985
CONTAINS 0.151%

,
Token,
Feature activation+0.000
expecting
Token expecting
Feature activation+0.580
fan
Token fan
Feature activation+0.190
atical
Tokenatical
Feature activation+0.221
Japanese
Token Japanese
Feature activation+0.323
resistance
Token resistance
Feature activation+0.826
,
Token,
Feature activation+0.106
American
Token American
Feature activation+0.000
military
Token military
Feature activation+0.000
planners
Token planners
Feature activation+0.000
repeatedly
Token repeatedly
Feature activation+0.000
of
Token of
Feature activation+0.000
,
Token,
Feature activation+0.000
if
Token if
Feature activation+0.000
not
Token not
Feature activation+0.000
doing
Token doing
Feature activation+0.669
something
Token something
Feature activation+0.778
exactly
Token exactly
Feature activation+0.482
,
Token,
Feature activation+0.165
at
Token at
Feature activation+0.000
least
Token least
Feature activation+0.000
establishing
Token establishing
Feature activation+0.000
metabolic
Token metabolic
Feature activation+0.000
disorders
Token disorders
Feature activation+0.276
related
Token related
Feature activation+0.000
to
Token to
Feature activation+0.000
circadian
Token circadian
Feature activation+0.000
disruption
Token disruption
Feature activation+0.748
,
Token,
Feature activation+0.000
will
Token will
Feature activation+0.000
be
Token be
Feature activation+0.000
published
Token published
Feature activation+0.000
by
Token by
Feature activation+0.000
ings
Tokenings
Feature activation+0.000
which
Token which
Feature activation+0.000
,
Token,
Feature activation+0.000
if
Token if
Feature activation+0.000
not
Token not
Feature activation+0.593
by
Token by
Feature activation+0.686
design
Token design
Feature activation+1.473
,
Token,
Feature activation+0.000
it
Token it
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
said
Token said
Feature activation+0.000
,
Token,
Feature activation+0.000
with
Token with
Feature activation+0.090
a
Token a
Feature activation+0.796
sm
Token sm
Feature activation+0.207
irk
Tokenirk
Feature activation+0.843
on
Token on
Feature activation+0.307
his
Token his
Feature activation+0.000
face
Token face
Feature activation+1.015
,
Token,
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.000

INTERVAL 0.000 - 0.493
CONTAINS 99.778%

ľ
Tokenľ
Feature activation+0.000
...
Token...
Feature activation+0.000
It
TokenIt
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
very
Token very
Feature activation+0.000
easy
Token easy
Feature activation+0.000
to
Token to
Feature activation+0.000
arm
Token arm
Feature activation+0.000
chair
Tokenchair
Feature activation+0.000
remains
Token remains
Feature activation+0.000
imprisoned
Token imprisoned
Feature activation+0.000
in
Token in
Feature activation+0.000
Iran
Token Iran
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
The
TokenThe
Feature activation+0.000
Government
Token Government
Feature activation+0.000
Accountability
Token Accountability
Feature activation+0.000
Office
Token Office
Feature activation+0.000
,
Token,
Feature activation+0.000
Comics
Token Comics
Feature activation+0.000
#
Token #
Feature activation+0.000
260
Token260
Feature activation+0.000
J
Token J
Feature activation+0.000
AN
TokenAN
Feature activation+0.000
Super
TokenSuper
Feature activation+0.000
girl
Tokengirl
Feature activation+0.000
becomes
Token becomes
Feature activation+0.000
Super
Token Super
Feature activation+0.000
-
Token-
Feature activation+0.000
baby
Tokenbaby
Feature activation+0.000
to
Token to
Feature activation+0.000
clean
Token clean
Feature activation+0.000
up
Token up
Feature activation+0.000
articles
Token articles
Feature activation+0.000
.
Token.
Feature activation+0.000
Many
Token Many
Feature activation+0.000
Wik
Token Wik
Feature activation+0.000
iped
Tokeniped
Feature activation+0.000
ians
Tokenians
Feature activation+0.000
--
Token --
Feature activation+0.000
frequent
Token frequent
Feature activation+0.000
[
Token [
Feature activation+0.000
21
Token21
Feature activation+0.000
:
Token:
Feature activation+0.000
53
Token53
Feature activation+0.000
mark
Token mark
Feature activation+0.000
]
Token]
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
[
Token[
Feature activation+0.000
21
Token21
Feature activation+0.000
:
Token:
Feature activation+0.000

BOTTOM ACTIVATIONS
MIN = 0.000

,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.089
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
&
Token &
Feature activation+0.000
Development
Token Development
Feature activation+0.000
Organisation
Token Organisation
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.089
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
&
Token &
Feature activation+0.000
Development
Token Development
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.089
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.089
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
rik
Tokenrik
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
rik
Tokenrik
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.089
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.089
of
Token of
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.089
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000

Top feature 7 in H8.5: (feature 8285)

TOP ACTIVATIONS
MAX = 2.713

gender
Token gender
Feature activation+0.000
transition
Token transition
Feature activation+1.779
treatment
Token treatment
Feature activation+2.409
in
Token in
Feature activation+0.286
the
Token the
Feature activation+0.242
military
Token military
Feature activation+2.713
would
Token would
Feature activation+0.000
be
Token be
Feature activation+0.000
some
Token some
Feature activation+0.000
14
Token 14
Feature activation+0.000
times
Token times
Feature activation+0.000
in
Token in
Feature activation+0.358
the
Token the
Feature activation+0.488
name
Token name
Feature activation+0.273
of
Token of
Feature activation+0.666
religious
Token religious
Feature activation+0.215
liberty
Token liberty
Feature activation+2.677
are
Token are
Feature activation+0.344
very
Token very
Feature activation+0.000
much
Token much
Feature activation+0.000
alive
Token alive
Feature activation+0.000
in
Token in
Feature activation+0.000
in
Token in
Feature activation+0.219
the
Token the
Feature activation+0.341
U
Token U
Feature activation+0.086
.
Token.
Feature activation+0.000
S
TokenS
Feature activation+2.618
.
Token.
Feature activation+2.641
are
Token are
Feature activation+0.167
"
Token "
Feature activation+0.365
extremely
Tokenextremely
Feature activation+0.000
low
Token low
Feature activation+0.000
.
Token.
Feature activation+0.000
outbreak
Token outbreak
Feature activation+2.125
in
Token in
Feature activation+0.219
the
Token the
Feature activation+0.341
U
Token U
Feature activation+0.086
.
Token.
Feature activation+0.000
S
TokenS
Feature activation+2.618
.
Token.
Feature activation+2.641
are
Token are
Feature activation+0.167
"
Token "
Feature activation+0.365
extremely
Tokenextremely
Feature activation+0.000
low
Token low
Feature activation+0.000
fallen
Token fallen
Feature activation+0.000
U
Token U
Feature activation+0.000
.
Token.
Feature activation+0.000
S
TokenS
Feature activation+0.000
.
Token.
Feature activation+0.000
soldier
Token soldier
Feature activation+2.596
whose
Token whose
Feature activation+0.000
father
Token father
Feature activation+0.000
spoke
Token spoke
Feature activation+0.000
out
Token out
Feature activation+0.000
against
Token against
Feature activation+0.000
last
Token last
Feature activation+0.000
week
Token week
Feature activation+0.041
's
Token's
Feature activation+0.000
Democratic
Token Democratic
Feature activation+0.000
National
Token National
Feature activation+0.000
Convention
Token Convention
Feature activation+2.568
were
Token were
Feature activation+0.722
"
Token "
Feature activation+0.598
blown
Tokenblown
Feature activation+0.000
hugely
Token hugely
Feature activation+0.000
out
Token out
Feature activation+0.000
last
Token last
Feature activation+0.000
year
Token year
Feature activation+0.000
or
Token or
Feature activation+0.000
two
Token two
Feature activation+0.416
have
Token have
Feature activation+1.333
brought
Token brought
Feature activation+2.500
into
Token into
Feature activation+0.395
the
Token the
Feature activation+0.464
arena
Token arena
Feature activation+1.550
are
Token are
Feature activation+0.155
central
Token central
Feature activation+0.292
energy
Token energy
Feature activation+0.019
,
Token,
Feature activation+0.000
transport
Token transport
Feature activation+0.000
and
Token and
Feature activation+0.000
digital
Token digital
Feature activation+0.000
networks
Token networks
Feature activation+2.462
was
Token was
Feature activation+0.000
slashed
Token slashed
Feature activation+0.000
by
Token by
Feature activation+0.000
a
Token a
Feature activation+0.000
quarter
Token quarter
Feature activation+0.000
odds
Token odds
Feature activation+0.000
of
Token of
Feature activation+0.504
the
Token the
Feature activation+0.000
asteroid
Token asteroid
Feature activation+0.165
hitting
Token hitting
Feature activation+0.352
Earth
Token Earth
Feature activation+2.449
were
Token were
Feature activation+0.062
first
Token first
Feature activation+0.000
estimated
Token estimated
Feature activation+0.000
at
Token at
Feature activation+0.000
one
Token one
Feature activation+0.000
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
accusations
Token accusations
Feature activation+0.000
of
Token of
Feature activation+0.542
fake
Token fake
Feature activation+0.656
religion
Token religion
Feature activation+2.438
have
Token have
Feature activation+0.598
long
Token long
Feature activation+0.564
been
Token been
Feature activation+0.049
a
Token a
Feature activation+0.000
strategy
Token strategy
Feature activation+0.000
choose
Token choose
Feature activation+0.008
to
Token to
Feature activation+0.000
undergo
Token undergo
Feature activation+0.115
gender
Token gender
Feature activation+0.000
transition
Token transition
Feature activation+1.779
treatment
Token treatment
Feature activation+2.409
in
Token in
Feature activation+0.286
the
Token the
Feature activation+0.242
military
Token military
Feature activation+2.713
would
Token would
Feature activation+0.000
be
Token be
Feature activation+0.000
U
Token U
Feature activation+0.000
.
Token.
Feature activation+0.000
N
TokenN
Feature activation+0.020
.
Token.
Feature activation+0.000
ausp
Token ausp
Feature activation+0.000
ices
Tokenices
Feature activation+2.302
in
Token in
Feature activation+0.620
Geneva
Token Geneva
Feature activation+2.176
were
Token were
Feature activation+0.811
not
Token not
Feature activation+0.406
progressing
Token progressing
Feature activation+0.012
regarding
Token regarding
Feature activation+0.000
contributions
Token contributions
Feature activation+0.940
to
Token to
Feature activation+0.757
a
Token a
Feature activation+0.694
political
Token political
Feature activation+0.115
entity
Token entity
Feature activation+2.255
are
Token are
Feature activation+0.000
contained
Token contained
Feature activation+0.000
in
Token in
Feature activation+0.000
sections
Token sections
Feature activation+0.000
404
Token 404
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.000
ĺ
Tokenĺ
Feature activation+0.273
Mad
TokenMad
Feature activation+0.000
Men
Token Men
Feature activation+1.690
âĢ
TokenâĢ
Feature activation+0.271
Ļ
TokenĻ
Feature activation+2.245
with
Token with
Feature activation+0.860
these
Token these
Feature activation+0.000
downtown
Token downtown
Feature activation+0.000
campuses
Token campuses
Feature activation+1.877
with
Token with
Feature activation+0.361
between
Token between
Feature activation+0.595
flare
Token flare
Feature activation+0.000
and
Token and
Feature activation+0.000
non
Token non
Feature activation+0.000
fl
Tokenfl
Feature activation+0.000
are
Tokenare
Feature activation+2.235
groups
Token groups
Feature activation+1.736
for
Token for
Feature activation+0.543
presence
Token presence
Feature activation+0.582
or
Token or
Feature activation+0.000
number
Token number
Feature activation+0.703
stemming
Token stemming
Feature activation+0.171
from
Token from
Feature activation+0.436
melting
Token melting
Feature activation+0.777
perm
Token perm
Feature activation+0.000
af
Tokenaf
Feature activation+0.000
rost
Tokenrost
Feature activation+2.232
could
Token could
Feature activation+0.301
be
Token be
Feature activation+0.000
critical
Token critical
Feature activation+0.000
to
Token to
Feature activation+0.000
determining
Token determining
Feature activation+0.000
comments
Token comments
Feature activation+0.000
about
Token about
Feature activation+0.514
the
Token the
Feature activation+0.274
Gold
Token Gold
Feature activation+0.000
Star
Token Star
Feature activation+0.000
family
Token family
Feature activation+2.205
of
Token of
Feature activation+0.000
a
Token a
Feature activation+0.000
fallen
Token fallen
Feature activation+0.000
U
Token U
Feature activation+0.000
.
Token.
Feature activation+0.000
N
TokenN
Feature activation+0.020
.
Token.
Feature activation+0.000
ausp
Token ausp
Feature activation+0.000
ices
Tokenices
Feature activation+2.302
in
Token in
Feature activation+0.620
Geneva
Token Geneva
Feature activation+2.176
were
Token were
Feature activation+0.811
not
Token not
Feature activation+0.406
progressing
Token progressing
Feature activation+0.012
.
Token.
Feature activation+0.000
Local
Token Local
Feature activation+0.000
her
Token her
Feature activation+0.000
attempts
Token attempts
Feature activation+0.000
to
Token to
Feature activation+0.028
up
Token up
Feature activation+0.055
the
Token the
Feature activation+0.336
stakes
Token stakes
Feature activation+2.168
in
Token in
Feature activation+0.566
talks
Token talks
Feature activation+1.749
with
Token with
Feature activation+0.874
the
Token the
Feature activation+0.617
EU
Token EU
Feature activation+2.017
to
Token to
Feature activation+0.396
interrog
Token interrog
Feature activation+0.269
ate
Tokenate
Feature activation+1.196
him
Token him
Feature activation+0.853
without
Token without
Feature activation+0.312
counsel
Token counsel
Feature activation+2.165
and
Token and
Feature activation+0.000
the
Token the
Feature activation+0.000
fact
Token fact
Feature activation+0.000
that
Token that
Feature activation+0.226
the
Token the
Feature activation+0.123

Top DFA by src position
MAX = 3.120

reveals
Token reveals
Feature activation-0.017
Top resid features:
that
Token that
Feature activation-0.051
Top resid features:
the
Token the
Feature activation-0.012
Top resid features:
health
Token health
Feature activation+0.016
Top resid features:
care
Token care
Feature activation+0.029
Top resid features:
costs
Token costs
Feature activation+3.120
Top resid features:
of
Token of
Feature activation+0.143
Top resid features:
trans
Token trans
Feature activation+0.002
Top resid features:
g
Tokeng
Feature activation+0.000
Top resid features:
enders
Tokenenders
Feature activation+0.022
Top resid features:
who
Token who
Feature activation+0.017
Top resid features:
people
Token people
Feature activation+0.007
Top resid features:
.
Token.
Feature activation+0.041
Top resid features:
Ċ
TokenĊ
Feature activation+0.018
Top resid features:
Ċ
TokenĊ
Feature activation+0.044
Top resid features:
But
TokenBut
Feature activation-0.032
Top resid features:
claims
Token claims
Feature activation+2.828
Top resid features:
in
Token in
Feature activation+0.263
Top resid features:
the
Token the
Feature activation+0.032
Top resid features:
name
Token name
Feature activation+0.196
Top resid features:
of
Token of
Feature activation+0.084
Top resid features:
religious
Token religious
Feature activation-0.001
Top resid features:
are
Token are
Feature activation-0.009
Top resid features:
safe
Token safe
Feature activation-0.006
Top resid features:
and
Token and
Feature activation-0.004
Top resid features:
that
Token that
Feature activation-0.008
Top resid features:
the
Token the
Feature activation+0.012
Top resid features:
chances
Token chances
Feature activation+2.815
Top resid features:
of
Token of
Feature activation+0.140
Top resid features:
an
Token an
Feature activation+0.013
Top resid features:
outbreak
Token outbreak
Feature activation+0.129
Top resid features:
in
Token in
Feature activation+0.168
Top resid features:
the
Token the
Feature activation+0.081
Top resid features:
are
Token are
Feature activation-0.005
Top resid features:
safe
Token safe
Feature activation-0.004
Top resid features:
and
Token and
Feature activation-0.002
Top resid features:
that
Token that
Feature activation-0.010
Top resid features:
the
Token the
Feature activation+0.010
Top resid features:
chances
Token chances
Feature activation+2.748
Top resid features:
of
Token of
Feature activation+0.224
Top resid features:
an
Token an
Feature activation+0.015
Top resid features:
outbreak
Token outbreak
Feature activation+0.089
Top resid features:
in
Token in
Feature activation+0.139
Top resid features:
the
Token the
Feature activation+0.121
Top resid features:
Ċ
TokenĊ
Feature activation+0.055
Top resid features:
Ċ
TokenĊ
Feature activation+0.016
Top resid features:
Donald
TokenDonald
Feature activation-0.003
Top resid features:
Trump
Token Trump
Feature activation-0.004
Top resid features:
's
Token's
Feature activation+0.013
Top resid features:
comments
Token comments
Feature activation+2.887
Top resid features:
about
Token about
Feature activation+0.085
Top resid features:
the
Token the
Feature activation+0.015
Top resid features:
Gold
Token Gold
Feature activation+0.002
Top resid features:
Star
Token Star
Feature activation+0.001
Top resid features:
family
Token family
Feature activation+0.020
Top resid features:
Ċ
TokenĊ
Feature activation+0.048
Top resid features:
Ċ
TokenĊ
Feature activation+0.015
Top resid features:
Donald
TokenDonald
Feature activation-0.003
Top resid features:
Trump
Token Trump
Feature activation-0.004
Top resid features:
's
Token's
Feature activation+0.001
Top resid features:
comments
Token comments
Feature activation+2.878
Top resid features:
about
Token about
Feature activation+0.122
Top resid features:
the
Token the
Feature activation+0.010
Top resid features:
Gold
Token Gold
Feature activation-0.001
Top resid features:
Star
Token Star
Feature activation-0.001
Top resid features:
family
Token family
Feature activation+0.003
Top resid features:
covering
Token covering
Feature activation-0.006
Top resid features:
it
Token it
Feature activation-0.006
Top resid features:
âĢĶ
TokenâĢĶ
Feature activation+0.007
Top resid features:
but
Tokenbut
Feature activation-0.010
Top resid features:
the
Token the
Feature activation+0.045
Top resid features:
matters
Token matters
Feature activation+2.217
Top resid features:
that
Token that
Feature activation+0.928
Top resid features:
the
Token the
Feature activation+0.008
Top resid features:
last
Token last
Feature activation+0.005
Top resid features:
year
Token year
Feature activation+0.017
Top resid features:
or
Token or
Feature activation+0.006
Top resid features:
figures
Token figures
Feature activation+0.001
Top resid features:
.
Token.
Feature activation+0.026
Top resid features:
Ċ
TokenĊ
Feature activation+0.026
Top resid features:
Ċ
TokenĊ
Feature activation+0.022
Top resid features:
But
TokenBut
Feature activation-0.085
Top resid features:
plans
Token plans
Feature activation+2.885
Top resid features:
to
Token to
Feature activation+0.204
Top resid features:
use
Token use
Feature activation+0.026
Top resid features:
40
Token 40
Feature activation-0.006
Top resid features:
billion
Token billion
Feature activation-0.007
Top resid features:
euros
Token euros
Feature activation+0.009
Top resid features:
Nature
Token Nature
Feature activation+0.003
Top resid features:
.
Token.
Feature activation+0.018
Top resid features:
Ċ
TokenĊ
Feature activation+0.014
Top resid features:
Ċ
TokenĊ
Feature activation+0.013
Top resid features:
The
TokenThe
Feature activation+0.000
Top resid features:
odds
Token odds
Feature activation+2.877
Top resid features:
of
Token of
Feature activation+0.298
Top resid features:
the
Token the
Feature activation+0.008
Top resid features:
asteroid
Token asteroid
Feature activation+0.018
Top resid features:
hitting
Token hitting
Feature activation-0.011
Top resid features:
Earth
Token Earth
Feature activation+0.005
Top resid features:
and
Token and
Feature activation-0.012
Top resid features:
superst
Token superst
Feature activation+0.001
Top resid features:
ition
Tokenition
Feature activation+0.005
Top resid features:
,
Token,
Feature activation+0.012
Top resid features:
and
Token and
Feature activation-0.053
Top resid features:
accusations
Token accusations
Feature activation+2.969
Top resid features:
of
Token of
Feature activation+0.111
Top resid features:
fake
Token fake
Feature activation+0.077
Top resid features:
religion
Token religion
Feature activation+0.015
Top resid features:
have
Token have
Feature activation+0.000
Top resid features:
long
Token long
Feature activation+0.000
Top resid features:
reveals
Token reveals
Feature activation-0.008
Top resid features:
that
Token that
Feature activation-0.050
Top resid features:
the
Token the
Feature activation-0.011
Top resid features:
health
Token health
Feature activation+0.027
Top resid features:
care
Token care
Feature activation+0.046
Top resid features:
costs
Token costs
Feature activation+2.809
Top resid features:
of
Token of
Feature activation+0.172
Top resid features:
trans
Token trans
Feature activation+0.005
Top resid features:
g
Tokeng
Feature activation+0.001
Top resid features:
enders
Tokenenders
Feature activation+0.033
Top resid features:
who
Token who
Feature activation+0.016
Top resid features:
al
Tokenal
Feature activation-0.000
Top resid features:
em
Tokenem
Feature activation-0.000
Top resid features:
said
Token said
Feature activation+0.010
Top resid features:
separate
Token separate
Feature activation+0.083
Top resid features:
peace
Token peace
Feature activation+0.087
Top resid features:
talks
Token talks
Feature activation+2.540
Top resid features:
under
Token under
Feature activation+0.023
Top resid features:
U
Token U
Feature activation-0.000
Top resid features:
.
Token.
Feature activation-0.000
Top resid features:
N
TokenN
Feature activation+0.003
Top resid features:
.
Token.
Feature activation-0.000
Top resid features:
Political
Token Political
Feature activation+0.001
Top resid features:
Contributions
Token Contributions
Feature activation+0.015
Top resid features:
Ċ
TokenĊ
Feature activation+0.027
Top resid features:
Ċ
TokenĊ
Feature activation+0.021
Top resid features:
The
TokenThe
Feature activation-0.004
Top resid features:
provisions
Token provisions
Feature activation+2.229
Top resid features:
regarding
Token regarding
Feature activation+0.303
Top resid features:
contributions
Token contributions
Feature activation+0.175
Top resid features:
to
Token to
Feature activation+0.071
Top resid features:
a
Token a
Feature activation+0.091
Top resid features:
political
Token political
Feature activation+0.022
Top resid features:
Ċ
TokenĊ
Feature activation+0.013
Top resid features:
Ċ
TokenĊ
Feature activation+0.018
Top resid features:
âĢ
TokenâĢ
Feature activation+0.003
Top resid features:
ľ
Tokenľ
Feature activation-0.007
Top resid features:
The
TokenThe
Feature activation-0.011
Top resid features:
days
Token days
Feature activation+2.534
Top resid features:
of
Token of
Feature activation+0.364
Top resid features:
âĢ
Token âĢ
Feature activation+0.015
Top resid features:
ĺ
Tokenĺ
Feature activation-0.004
Top resid features:
Mad
TokenMad
Feature activation+0.002
Top resid features:
Men
Token Men
Feature activation+0.068
Top resid features:
micro
Token micro
Feature activation+0.007
Top resid features:
organ
Tokenorgan
Feature activation+0.001
Top resid features:
ism
Tokenism
Feature activation+0.005
Top resid features:
species
Token species
Feature activation+0.010
Top resid features:
.
Token.
Feature activation+0.003
Top resid features:
Differences
Token Differences
Feature activation+2.595
Top resid features:
between
Token between
Feature activation+0.219
Top resid features:
flare
Token flare
Feature activation+0.082
Top resid features:
and
Token and
Feature activation+0.006
Top resid features:
non
Token non
Feature activation+0.002
Top resid features:
fl
Tokenfl
Feature activation+0.004
Top resid features:
.
Token.
Feature activation+0.002
Top resid features:
Ċ
TokenĊ
Feature activation+0.003
Top resid features:
Ċ
TokenĊ
Feature activation+0.008
Top resid features:
The
TokenThe
Feature activation-0.011
Top resid features:
methane
Token methane
Feature activation+0.087
Top resid features:
emissions
Token emissions
Feature activation+2.056
Top resid features:
stemming
Token stemming
Feature activation+0.327
Top resid features:
from
Token from
Feature activation+0.108
Top resid features:
melting
Token melting
Feature activation+0.202
Top resid features:
perm
Token perm
Feature activation+0.031
Top resid features:
af
Tokenaf
Feature activation+0.005
Top resid features:
Ċ
TokenĊ
Feature activation+0.045
Top resid features:
Ċ
TokenĊ
Feature activation+0.013
Top resid features:
Donald
TokenDonald
Feature activation-0.009
Top resid features:
Trump
Token Trump
Feature activation-0.006
Top resid features:
's
Token's
Feature activation+0.016
Top resid features:
comments
Token comments
Feature activation+2.449
Top resid features:
about
Token about
Feature activation+0.156
Top resid features:
the
Token the
Feature activation+0.047
Top resid features:
Gold
Token Gold
Feature activation+0.004
Top resid features:
Star
Token Star
Feature activation-0.011
Top resid features:
family
Token family
Feature activation+0.065
Top resid features:
al
Tokenal
Feature activation+0.000
Top resid features:
em
Tokenem
Feature activation+0.001
Top resid features:
said
Token said
Feature activation+0.002
Top resid features:
separate
Token separate
Feature activation+0.033
Top resid features:
peace
Token peace
Feature activation+0.041
Top resid features:
talks
Token talks
Feature activation+2.420
Top resid features:
under
Token under
Feature activation+0.003
Top resid features:
U
Token U
Feature activation+0.001
Top resid features:
.
Token.
Feature activation+0.001
Top resid features:
N
TokenN
Feature activation+0.010
Top resid features:
.
Token.
Feature activation+0.008
Top resid features:
Herman
Token Herman
Feature activation-0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.003
Top resid features:
Ċ
TokenĊ
Feature activation+0.035
Top resid features:
But
TokenBut
Feature activation+0.040
Top resid features:
her
Token her
Feature activation-0.016
Top resid features:
attempts
Token attempts
Feature activation+2.644
Top resid features:
to
Token to
Feature activation+0.132
Top resid features:
up
Token up
Feature activation-0.016
Top resid features:
the
Token the
Feature activation-0.026
Top resid features:
stakes
Token stakes
Feature activation+0.002
Top resid features:
in
Token in
Feature activation+0.000
Top resid features:
with
Token with
Feature activation-0.010
Top resid features:
the
Token the
Feature activation+0.004
Top resid features:
order
Token order
Feature activation-0.007
Top resid features:
,
Token,
Feature activation+0.024
Top resid features:
my
Token my
Feature activation-0.008
Top resid features:
emails
Token emails
Feature activation+2.585
Top resid features:
with
Token with
Feature activation+0.343
Top resid features:
the
Token the
Feature activation+0.014
Top resid features:
advice
Token advice
Feature activation+0.045
Top resid features:
not
Token not
Feature activation+0.028
Top resid features:
to
Token to
Feature activation-0.023
Top resid features:

Decoder Weights Distribution

Head 0: 0.04

Head 1: 0.01

Head 2: 0.03

Head 3: 0.03

Head 4: 0.08

Head 5: 0.51

Head 6: 0.02

Head 7: 0.09

Head 8: 0.02

Head 9: 0.06

Head 10: 0.07

Head 11: 0.02

Positive logits

abound4.22

are4.14

were3.99

aren3.74

weren3.65

vary3.47

differ3.44

remain3.36

seem3.35

include3.34

were3.31

appear3.29

converge3.18

comprise3.15

differed3.02

illustrate3.01

resonate2.96

have2.96

constitute2.92

indicate2.82

Negative logits

Learns-2.79

laughs-2.41

Kills-2.37

wants-2.07

hopes-2.05

loves-2.04

expects-2.02

Gets-2.02

Moves-2.00

Own-1.99

`.-1.98

thinks-1.97

wishes-1.95

Uses-1.95

ˈ-1.93

\",-1.92

buys-1.90

regrets-1.90

hires-1.90

alloc-1.90

INTERVAL 2.442 - 2.713
CONTAINS 0.001%

outbreak
Token outbreak
Feature activation+2.125
in
Token in
Feature activation+0.219
the
Token the
Feature activation+0.341
U
Token U
Feature activation+0.086
.
Token.
Feature activation+0.000
S
TokenS
Feature activation+2.618
.
Token.
Feature activation+2.641
are
Token are
Feature activation+0.167
"
Token "
Feature activation+0.365
extremely
Tokenextremely
Feature activation+0.000
low
Token low
Feature activation+0.000
odds
Token odds
Feature activation+0.000
of
Token of
Feature activation+0.504
the
Token the
Feature activation+0.000
asteroid
Token asteroid
Feature activation+0.165
hitting
Token hitting
Feature activation+0.352
Earth
Token Earth
Feature activation+2.449
were
Token were
Feature activation+0.062
first
Token first
Feature activation+0.000
estimated
Token estimated
Feature activation+0.000
at
Token at
Feature activation+0.000
one
Token one
Feature activation+0.000
gender
Token gender
Feature activation+0.000
transition
Token transition
Feature activation+1.779
treatment
Token treatment
Feature activation+2.409
in
Token in
Feature activation+0.286
the
Token the
Feature activation+0.242
military
Token military
Feature activation+2.713
would
Token would
Feature activation+0.000
be
Token be
Feature activation+0.000
some
Token some
Feature activation+0.000
14
Token 14
Feature activation+0.000
times
Token times
Feature activation+0.000
in
Token in
Feature activation+0.219
the
Token the
Feature activation+0.341
U
Token U
Feature activation+0.086
.
Token.
Feature activation+0.000
S
TokenS
Feature activation+2.618
.
Token.
Feature activation+2.641
are
Token are
Feature activation+0.167
"
Token "
Feature activation+0.365
extremely
Tokenextremely
Feature activation+0.000
low
Token low
Feature activation+0.000
.
Token.
Feature activation+0.000
fallen
Token fallen
Feature activation+0.000
U
Token U
Feature activation+0.000
.
Token.
Feature activation+0.000
S
TokenS
Feature activation+0.000
.
Token.
Feature activation+0.000
soldier
Token soldier
Feature activation+2.596
whose
Token whose
Feature activation+0.000
father
Token father
Feature activation+0.000
spoke
Token spoke
Feature activation+0.000
out
Token out
Feature activation+0.000
against
Token against
Feature activation+0.000

INTERVAL 2.170 - 2.442
CONTAINS 0.001%

comments
Token comments
Feature activation+0.000
about
Token about
Feature activation+0.514
the
Token the
Feature activation+0.274
Gold
Token Gold
Feature activation+0.000
Star
Token Star
Feature activation+0.000
family
Token family
Feature activation+2.205
of
Token of
Feature activation+0.000
a
Token a
Feature activation+0.000
fallen
Token fallen
Feature activation+0.000
U
Token U
Feature activation+0.000
.
Token.
Feature activation+0.000
between
Token between
Feature activation+0.595
flare
Token flare
Feature activation+0.000
and
Token and
Feature activation+0.000
non
Token non
Feature activation+0.000
fl
Tokenfl
Feature activation+0.000
are
Tokenare
Feature activation+2.235
groups
Token groups
Feature activation+1.736
for
Token for
Feature activation+0.543
presence
Token presence
Feature activation+0.582
or
Token or
Feature activation+0.000
number
Token number
Feature activation+0.703
N
TokenN
Feature activation+0.020
.
Token.
Feature activation+0.000
ausp
Token ausp
Feature activation+0.000
ices
Tokenices
Feature activation+2.302
in
Token in
Feature activation+0.620
Geneva
Token Geneva
Feature activation+2.176
were
Token were
Feature activation+0.811
not
Token not
Feature activation+0.406
progressing
Token progressing
Feature activation+0.012
.
Token.
Feature activation+0.000
Local
Token Local
Feature activation+0.000
U
Token U
Feature activation+0.000
.
Token.
Feature activation+0.000
N
TokenN
Feature activation+0.020
.
Token.
Feature activation+0.000
ausp
Token ausp
Feature activation+0.000
ices
Tokenices
Feature activation+2.302
in
Token in
Feature activation+0.620
Geneva
Token Geneva
Feature activation+2.176
were
Token were
Feature activation+0.811
not
Token not
Feature activation+0.406
progressing
Token progressing
Feature activation+0.012
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
accusations
Token accusations
Feature activation+0.000
of
Token of
Feature activation+0.542
fake
Token fake
Feature activation+0.656
religion
Token religion
Feature activation+2.438
have
Token have
Feature activation+0.598
long
Token long
Feature activation+0.564
been
Token been
Feature activation+0.049
a
Token a
Feature activation+0.000
strategy
Token strategy
Feature activation+0.000

INTERVAL 1.899 - 2.170
CONTAINS 0.004%

to
Token to
Feature activation+0.077
the
Token the
Feature activation+0.006
remainder
Token remainder
Feature activation+0.672
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
world
Token world
Feature activation+1.989
is
Token is
Feature activation+0.000
$
Token $
Feature activation+0.000
18
Token18
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
at
Token at
Feature activation+0.467
the
Token the
Feature activation+0.364
start
Token start
Feature activation+1.893
of
Token of
Feature activation+0.846
the
Token the
Feature activation+0.478
song
Token song
Feature activation+2.122
are
Token are
Feature activation+0.273
perfect
Token perfect
Feature activation+0.094
.
Token.
Feature activation+0.000
I
Token I
Feature activation+0.000
wish
Token wish
Feature activation+0.000
rape
Token rape
Feature activation+0.628
at
Token at
Feature activation+0.000
Phi
Token Phi
Feature activation+0.000
Kappa
Token Kappa
Feature activation+0.000
P
Token P
Feature activation+0.000
si
Tokensi
Feature activation+2.083
were
Token were
Feature activation+0.000
later
Token later
Feature activation+0.000
cast
Token cast
Feature activation+0.000
into
Token into
Feature activation+0.000
doubt
Token doubt
Feature activation+0.000
received
Tokenreceived
Feature activation+0.000
silver
Token silver
Feature activation+0.000
screen
Token screen
Feature activation+1.275
so
Token so
Feature activation+0.464
ire
Tokenire
Feature activation+0.000
es
Tokenes
Feature activation+2.144
don
Token don
Feature activation+0.000
't
Token't
Feature activation+0.057
apply
Token apply
Feature activation+0.000
.
Token.
Feature activation+0.000
Peter
Token Peter
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Police
TokenPolice
Feature activation+0.000
investigations
Token investigations
Feature activation+0.000
into
Token into
Feature activation+0.028
the
Token the
Feature activation+0.330
matter
Token matter
Feature activation+1.900
are
Token are
Feature activation+0.707
continuing
Token continuing
Feature activation+0.000
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
officers
Token officers
Feature activation+0.000

INTERVAL 1.628 - 1.899
CONTAINS 0.007%

As
TokenAs
Feature activation+0.000
the
Token the
Feature activation+0.000
ranks
Token ranks
Feature activation+0.000
of
Token of
Feature activation+0.000
adjunct
Token adjunct
Feature activation+0.278
s
Tokens
Feature activation+1.680
have
Token have
Feature activation+0.805
sw
Token sw
Feature activation+0.000
elled
Tokenelled
Feature activation+0.000
and
Token and
Feature activation+0.000
pay
Token pay
Feature activation+0.000
discussions
Token discussions
Feature activation+0.000
I
Token I
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
ve
Tokenve
Feature activation+0.531
seen
Token seen
Feature activation+1.796
have
Token have
Feature activation+1.165
been
Token been
Feature activation+0.376
about
Token about
Feature activation+0.098
1
Token 1
Feature activation+0.000
)
Token)
Feature activation+0.258
of
Tokenof
Feature activation+0.000
ana
Tokenana
Feature activation+1.697
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.459
s
Tokens
Feature activation+0.678
case
Token case
Feature activation+1.699
take
Token take
Feature activation+0.046
advantage
Token advantage
Feature activation+0.000
of
Token of
Feature activation+0.122
such
Token such
Feature activation+0.000
injust
Token injust
Feature activation+0.000
billion
Token billion
Feature activation+0.000
euros
Token euros
Feature activation+0.722
to
Token to
Feature activation+0.105
leverage
Token leverage
Feature activation+0.433
private
Token private
Feature activation+0.056
investment
Token investment
Feature activation+1.824
in
Token in
Feature activation+0.339
cross
Token cross
Feature activation+0.000
-
Token-
Feature activation+0.000
border
Tokenborder
Feature activation+0.326
energy
Token energy
Feature activation+0.019
and
Token and
Feature activation+0.093
other
Token other
Feature activation+0.000
details
Token details
Feature activation+0.411
of
Token of
Feature activation+0.984
the
Token the
Feature activation+0.612
campaign
Token campaign
Feature activation+1.645
were
Token were
Feature activation+0.383
revealed
Token revealed
Feature activation+0.000
in
Token in
Feature activation+0.000
a
Token a
Feature activation+0.000
briefing
Token briefing
Feature activation+0.000

INTERVAL 1.357 - 1.628
CONTAINS 0.013%

,
Token,
Feature activation+0.000
relations
Token relations
Feature activation+0.000
between
Token between
Feature activation+0.000
Tokyo
Token Tokyo
Feature activation+0.000
and
Token and
Feature activation+0.000
Seoul
Token Seoul
Feature activation+1.413
have
Token have
Feature activation+0.386
been
Token been
Feature activation+0.043
m
Token m
Feature activation+0.000
ired
Tokenired
Feature activation+0.000
in
Token in
Feature activation+0.002
physical
Token physical
Feature activation+0.000
features
Token features
Feature activation+0.000
of
Token of
Feature activation+0.124
the
Token the
Feature activation+0.293
eastern
Token eastern
Feature activation+0.261
coast
Token coast
Feature activation+1.536
were
Token were
Feature activation+0.372
advantageous
Token advantageous
Feature activation+0.000
for
Token for
Feature activation+0.000
intr
Token intr
Feature activation+0.000
aco
Tokenaco
Feature activation+0.000
on
Token on
Feature activation+0.156
the
Token the
Feature activation+0.118
swearing
Token swearing
Feature activation+0.229
-
Token-
Feature activation+0.000
in
Tokenin
Feature activation+1.108
ceremony
Token ceremony
Feature activation+1.502
will
Token will
Feature activation+0.000
be
Token be
Feature activation+0.000
announced
Token announced
Feature activation+0.000
early
Token early
Feature activation+0.000
next
Token next
Feature activation+0.000
technology
Token technology
Feature activation+0.632
during
Token during
Feature activation+0.510
and
Token and
Feature activation+0.130
after
Token after
Feature activation+0.738
the
Token the
Feature activation+0.557
war
Token war
Feature activation+1.572
convinced
Token convinced
Feature activation+0.000
many
Token many
Feature activation+0.000
that
Token that
Feature activation+0.000
"
Token "
Feature activation+0.000
the
Tokenthe
Feature activation+0.000
between
Token between
Feature activation+0.409
Ki
Token Ki
Feature activation+0.000
ir
Tokenir
Feature activation+0.000
and
Token and
Feature activation+0.305
Mach
Token Mach
Feature activation+0.000
ar
Tokenar
Feature activation+1.517
first
Token first
Feature activation+1.276
erupted
Token erupted
Feature activation+0.000
into
Token into
Feature activation+0.000
conflict
Token conflict
Feature activation+0.000
in
Token in
Feature activation+0.000

INTERVAL 1.085 - 1.357
CONTAINS 0.019%

was
Token was
Feature activation+0.000
written
Token written
Feature activation+0.000
,
Token,
Feature activation+0.000
sketches
Token sketches
Feature activation+0.000
of
Token of
Feature activation+0.000
it
Token it
Feature activation+1.337
were
Token were
Feature activation+0.000
shown
Token shown
Feature activation+0.000
to
Token to
Feature activation+0.000
some
Token some
Feature activation+0.000
outside
Token outside
Feature activation+0.000
assumed
Token assumed
Feature activation+0.000
that
Token that
Feature activation+0.000
methane
Token methane
Feature activation+0.000
emissions
Token emissions
Feature activation+0.000
from
Token from
Feature activation+0.000
wetlands
Token wetlands
Feature activation+1.147
are
Token are
Feature activation+0.000
high
Token high
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
trop
Token trop
Feature activation+0.000
costs
Token costs
Feature activation+0.000
of
Token of
Feature activation+0.016
the
Token the
Feature activation+0.108
military
Token military
Feature activation+0.874
air
Token air
Feature activation+0.000
operation
Token operation
Feature activation+1.197
would
Token would
Feature activation+0.292
be
Token be
Feature activation+0.000
met
Token met
Feature activation+0.000
from
Token from
Feature activation+0.000
the
Token the
Feature activation+0.000
opened
Token opened
Feature activation+0.000
my
Token my
Feature activation+0.000
eyes
Token eyes
Feature activation+0.000
the
Token the
Feature activation+0.000
lights
Token lights
Feature activation+0.000
are
Token are
Feature activation+1.188
out
Token out
Feature activation+0.361
and
Token and
Feature activation+0.000
here
Token here
Feature activation+0.000
comes
Token comes
Feature activation+0.000
a
Token a
Feature activation+0.000
threats
Token threats
Feature activation+0.000
to
Token to
Feature activation+0.237
human
Token human
Feature activation+0.146
health
Token health
Feature activation+1.259
and
Token and
Feature activation+0.001
wildlife
Token wildlife
Feature activation+1.323
from
Token from
Feature activation+0.271
rising
Token rising
Feature activation+0.493
waters
Token waters
Feature activation+1.299
that
Token that
Feature activation+0.466
inund
Token inund
Feature activation+0.274

INTERVAL 0.814 - 1.085
CONTAINS 0.039%

targets
Token targets
Feature activation+0.000
for
Token for
Feature activation+0.144
in
Token in
Feature activation+0.278
-
Token-
Feature activation+0.015
depth
Tokendepth
Feature activation+0.877
analysis
Token analysis
Feature activation+0.883
,
Token,
Feature activation+0.201
including
Token including
Feature activation+0.088
drilling
Token drilling
Feature activation+0.217
for
Token for
Feature activation+0.000
samples
Token samples
Feature activation+0.021
and
Token and
Feature activation+0.000
safe
Token safe
Feature activation+0.000
havens
Token havens
Feature activation+0.604
of
Token of
Feature activation+0.491
the
Token the
Feature activation+0.578
Atlantic
Token Atlantic
Feature activation+1.061
coast
Token coast
Feature activation+1.633
have
Token have
Feature activation+0.553
been
Token been
Feature activation+0.079
well
Token well
Feature activation+0.000
known
Token known
Feature activation+0.000
S
TokenS
Feature activation+0.000
ect
Tokenect
Feature activation+0.000
ors
Tokenors
Feature activation+0.000
targeted
Token targeted
Feature activation+0.000
for
Token for
Feature activation+0.000
cuts
Token cuts
Feature activation+0.930
in
Token in
Feature activation+0.798
the
Token the
Feature activation+0.365
bill
Token bill
Feature activation+0.716
will
Token will
Feature activation+0.000
immediately
Token immediately
Feature activation+0.031
s
Tokens
Feature activation+0.000
for
Token for
Feature activation+0.000
almost
Token almost
Feature activation+0.103
400
Token 400
Feature activation+0.584
NHS
Token NHS
Feature activation+0.000
services
Token services
Feature activation+0.861
,
Token,
Feature activation+0.162
worth
Token worth
Feature activation+0.178
a
Token a
Feature activation+0.000
quarter
Token quarter
Feature activation+0.000
of
Token of
Feature activation+0.000
that
Token that
Feature activation+0.000
the
Token the
Feature activation+0.000
benefits
Token benefits
Feature activation+0.000
to
Token to
Feature activation+0.000
our
Token our
Feature activation+0.000
community
Token community
Feature activation+1.021
outwe
Token outwe
Feature activation+0.000
ighed
Tokenighed
Feature activation+0.000
the
Token the
Feature activation+0.000
potential
Token potential
Feature activation+0.000
risks
Token risks
Feature activation+0.000

INTERVAL 0.543 - 0.814
CONTAINS 0.087%

dun
Token dun
Feature activation+0.000
ams
Tokenams
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.008
cemetery
Token cemetery
Feature activation+0.810
ground
Token ground
Feature activation+0.666
were
Token were
Feature activation+0.000
seized
Token seized
Feature activation+0.000
including
Token including
Feature activation+0.000
70
Token 70
Feature activation+0.000
percent
Token percent
Feature activation+0.000
angle
Token angle
Feature activation+0.000
pictures
Token pictures
Feature activation+0.000
taken
Token taken
Feature activation+0.318
of
Token of
Feature activation+0.213
your
Token your
Feature activation+0.122
house
Token house
Feature activation+0.655
.
Token.
Feature activation+0.000
Thankfully
Token Thankfully
Feature activation+0.000
I
Token I
Feature activation+0.000
'm
Token'm
Feature activation+0.000
able
Token able
Feature activation+0.000
to
Token to
Feature activation+0.000
reach
Token reach
Feature activation+0.746
Ali
Token Ali
Feature activation+0.446
Oz
Token Oz
Feature activation+0.000
tur
Tokentur
Feature activation+0.000
k
Tokenk
Feature activation+0.752
,
Token,
Feature activation+0.149
a
Token a
Feature activation+0.000
spokesman
Token spokesman
Feature activation+0.000
for
Token for
Feature activation+0.000
the
Token the
Feature activation+0.000
available
Token available
Feature activation+0.136
for
Token for
Feature activation+0.070
confirmation
Token confirmation
Feature activation+0.648
of
Token of
Feature activation+0.542
human
Token human
Feature activation+0.116
LD
Token LD
Feature activation+0.714
have
Token have
Feature activation+0.259
variable
Token variable
Feature activation+0.000
sensitivity
Token sensitivity
Feature activation+0.000
and
Token and
Feature activation+0.103
specificity
Token specificity
Feature activation+0.000
packages
Token packages
Feature activation+0.000
from
Token from
Feature activation+0.143
the
Token the
Feature activation+0.141
@
Token @
Feature activation+0.000
angular
Tokenangular
Feature activation+0.067
scope
Token scope
Feature activation+0.576
need
Token need
Feature activation+0.000
to
Token to
Feature activation+0.000
be
Token be
Feature activation+0.000
updated
Token updated
Feature activation+0.000
too
Token too
Feature activation+0.000

INTERVAL 0.271 - 0.543
CONTAINS 0.234%

itself
Token itself
Feature activation+0.000
-
Token -
Feature activation+0.000
and
Token and
Feature activation+0.000
other
Token other
Feature activation+0.000
instances
Token instances
Feature activation+0.000
of
Token of
Feature activation+0.430
aggression
Token aggression
Feature activation+0.261
in
Token in
Feature activation+0.265
Europe
Token Europe
Feature activation+0.266
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
the
Token the
Feature activation+0.000
league
Token league
Feature activation+0.000
as
Token as
Feature activation+0.000
quickly
Token quickly
Feature activation+0.000
as
Token as
Feature activation+0.077
possible
Token possible
Feature activation+0.332
adds
Token adds
Feature activation+0.000
an
Token an
Feature activation+0.000
element
Token element
Feature activation+0.000
of
Token of
Feature activation+0.000
risk
Token risk
Feature activation+0.000
legal
Token legal
Feature activation+0.000
problems
Token problems
Feature activation+0.122
that
Token that
Feature activation+0.672
proved
Token proved
Feature activation+0.743
impossible
Token impossible
Feature activation+0.529
for
Token for
Feature activation+0.493
him
Token him
Feature activation+0.000
to
Token to
Feature activation+0.000
overcome
Token overcome
Feature activation+0.000
.
Token.
Feature activation+0.000
Under
Token Under
Feature activation+0.000
experience
Token experience
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Last
Token Last
Feature activation+0.000
2
Token 2
Feature activation+0.000
years
Token years
Feature activation+0.365
has
Token has
Feature activation+0.151
shown
Token shown
Feature activation+0.000
me
Token me
Feature activation+0.000
that
Token that
Feature activation+0.000
it
Token it
Feature activation+0.000
of
Token of
Feature activation+0.000
Mil
Token Mil
Feature activation+0.000
gram
Tokengram
Feature activation+0.352
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.075
s
Tokens
Feature activation+0.344
study
Token study
Feature activation+0.341
have
Token have
Feature activation+0.000
been
Token been
Feature activation+0.000
conducted
Token conducted
Feature activation+0.000
in
Token in
Feature activation+0.000

INTERVAL 0.000 - 0.271
CONTAINS 99.596%

never
Tokennever
Feature activation+0.000
possible
Token possible
Feature activation+0.000
,
Token,
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
the
Token the
Feature activation+0.000
code
Token code
Feature activation+0.000
of
Token of
Feature activation+0.000
ethics
Token ethics
Feature activation+0.000
and
Token and
Feature activation+0.000
a
Token a
Feature activation+0.000
traditional
Token traditional
Feature activation+0.000
marriage
Token marriage
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
But
TokenBut
Feature activation+0.000
marriage
Token marriage
Feature activation+0.000
is
Token is
Feature activation+0.000
no
Token no
Feature activation+0.000
Nights
Token Nights
Feature activation+0.000
,
Token,
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
and
Token and
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
Step
TokenStep
Feature activation+0.000
Brothers
Token Brothers
Feature activation+0.000
,
Token,
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
lists
Token lists
Feature activation+0.000
of
Token of
Feature activation+0.000
their
Token their
Feature activation+0.000
preferred
Token preferred
Feature activation+0.000
pronouns
Token pronouns
Feature activation+0.000
on
Token on
Feature activation+0.000
their
Token their
Feature activation+0.000
self
Token self
Feature activation+0.000
-
Token-
Feature activation+0.000
absor
Tokenabsor
Feature activation+0.000
bed
Tokenbed
Feature activation+0.000
.
Token.
Feature activation+0.000
air
Tokenair
Feature activation+0.000
soft
Tokensoft
Feature activation+0.000
gi
Tokengi
Feature activation+0.000
.
Token.
Feature activation+0.000
com
Tokencom
Feature activation+0.000
/
Token/
Feature activation+0.000
information
Tokeninformation
Feature activation+0.000
.
Token.
Feature activation+0.000
php
Tokenphp
Feature activation+0.000
?
Token?
Feature activation+0.000

BOTTOM ACTIVATIONS
MIN = 0.000

atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
&
Token &
Feature activation+0.000
Development
Token Development
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
&
Token &
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
rik
Tokenrik
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
rik
Tokenrik
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000

Top feature 8 in H8.5: (feature 5442)

TOP ACTIVATIONS
MAX = 5.096

by
Token by
Feature activation+0.000
a
Token a
Feature activation+0.000
process
Token process
Feature activation+0.000
which
Token which
Feature activation+0.000
neither
Token neither
Feature activation+0.000
they
Token they
Feature activation+5.096
nor
Token nor
Feature activation+0.354
we
Token we
Feature activation+0.000
could
Token could
Feature activation+0.009
successfully
Token successfully
Feature activation+0.000
control
Token control
Feature activation+0.000
No
Token No
Feature activation+0.000
ida
Tokenida
Feature activation+0.000
itself
Token itself
Feature activation+0.000
which
Token which
Feature activation+0.000
neither
Token neither
Feature activation+0.000
maintain
Token maintain
Feature activation+5.010
the
Token the
Feature activation+3.042
we
Token we
Feature activation+0.386
blog
Tokenblog
Feature activation+2.227
of
Token of
Feature activation+1.931
their
Token their
Feature activation+1.240
Republicans
Token Republicans
Feature activation+0.000
will
Token will
Feature activation+0.000
become
Token become
Feature activation+0.000
neither
Token neither
Feature activation+0.000
the
Token the
Feature activation+1.324
party
Token party
Feature activation+4.969
of
Token of
Feature activation+1.669
Trump
Token Trump
Feature activation+3.617
nor
Token nor
Feature activation+0.000
the
Token the
Feature activation+0.000
party
Token party
Feature activation+0.000
,
Token,
Feature activation+0.000
but
Token but
Feature activation+0.000
that
Token that
Feature activation+0.000
's
Token's
Feature activation+0.000
neither
Token neither
Feature activation+0.000
here
Token here
Feature activation+4.664
nor
Token nor
Feature activation+0.000
there
Token there
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
CIS
Token CIS
Feature activation+0.000
,
Token,
Feature activation+0.000
Ukraine
Token Ukraine
Feature activation+0.000
has
Token has
Feature activation+0.000
neither
Token neither
Feature activation+0.000
signed
Token signed
Feature activation+4.532
nor
Token nor
Feature activation+0.000
ratified
Token ratified
Feature activation+0.000
the
Token the
Feature activation+0.000
charter
Token charter
Feature activation+0.000
of
Token of
Feature activation+0.000
for
Token for
Feature activation+0.000
whom
Token whom
Feature activation+0.000
there
Token there
Feature activation+0.000
's
Token's
Feature activation+0.000
neither
Token neither
Feature activation+0.000
dear
Token dear
Feature activation+4.514
nor
Token nor
Feature activation+0.000
und
Token und
Feature activation+0.000
ear
Tokenear
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
says
Token says
Feature activation+0.000
he
Token he
Feature activation+0.000
is
Token is
Feature activation+0.000
neither
Token neither
Feature activation+0.000
a
Token a
Feature activation+1.723
Republican
Token Republican
Feature activation+4.285
nor
Token nor
Feature activation+0.000
a
Token a
Feature activation+0.000
Democrat
Token Democrat
Feature activation+0.000
but
Token but
Feature activation+0.000
an
Token an
Feature activation+0.000
the
Token the
Feature activation+3.042
we
Token we
Feature activation+0.386
blog
Tokenblog
Feature activation+2.227
of
Token of
Feature activation+1.931
their
Token their
Feature activation+1.240
customers
Token customers
Feature activation+4.283
nor
Token nor
Feature activation+0.000
did
Token did
Feature activation+0.000
they
Token they
Feature activation+0.000
have
Token have
Feature activation+0.000
any
Token any
Feature activation+0.000
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
Vermont
Token Vermont
Feature activation+0.000
senator
Token senator
Feature activation+0.000
neither
Token neither
Feature activation+0.000
conceded
Token conceded
Feature activation+4.245
defeat
Token defeat
Feature activation+3.877
nor
Token nor
Feature activation+0.000
endorsed
Token endorsed
Feature activation+0.000
his
Token his
Feature activation+0.000
primary
Token primary
Feature activation+0.000
including
Token including
Feature activation+0.000
those
Token those
Feature activation+0.000
who
Token who
Feature activation+0.000
were
Token were
Feature activation+0.000
neither
Token neither
Feature activation+0.000
Jewish
Token Jewish
Feature activation+4.174
nor
Token nor
Feature activation+0.000
Black
Token Black
Feature activation+0.000
.
Token.
Feature activation+0.000
In
Token In
Feature activation+0.000
a
Token a
Feature activation+0.000
be
Token be
Feature activation+0.000
difficult
Token difficult
Feature activation+0.000
and
Token and
Feature activation+0.000
neither
Token neither
Feature activation+0.000
the
Token the
Feature activation+0.525
vehicle
Token vehicle
Feature activation+4.144
,
Token,
Feature activation+2.136
or
Token or
Feature activation+0.265
a
Token a
Feature activation+0.000
suspect
Token suspect
Feature activation+0.000
,
Token,
Feature activation+0.000
manager
Token manager
Feature activation+0.000
,
Token,
Feature activation+0.000
argues
Token argues
Feature activation+0.000
that
Token that
Feature activation+0.000
neither
Token neither
Feature activation+0.000
smartphones
Token smartphones
Feature activation+3.886
based
Token based
Feature activation+2.544
on
Token on
Feature activation+1.791
Google
Token Google
Feature activation+2.549
's
Token's
Feature activation+0.976
Android
Token Android
Feature activation+2.129
the
Token the
Feature activation+0.000
Vermont
Token Vermont
Feature activation+0.000
senator
Token senator
Feature activation+0.000
neither
Token neither
Feature activation+0.000
conceded
Token conceded
Feature activation+4.245
defeat
Token defeat
Feature activation+3.877
nor
Token nor
Feature activation+0.000
endorsed
Token endorsed
Feature activation+0.000
his
Token his
Feature activation+0.000
primary
Token primary
Feature activation+0.000
rival
Token rival
Feature activation+0.000
become
Token become
Feature activation+0.000
neither
Token neither
Feature activation+0.000
the
Token the
Feature activation+1.324
party
Token party
Feature activation+4.969
of
Token of
Feature activation+1.669
Trump
Token Trump
Feature activation+3.617
nor
Token nor
Feature activation+0.000
the
Token the
Feature activation+0.000
party
Token party
Feature activation+0.000
of
Token of
Feature activation+0.000
Bannon
Token Bannon
Feature activation+0.000
States
Token States
Feature activation+0.000
wanted
Token wanted
Feature activation+0.000
,
Token,
Feature activation+0.000
when
Token when
Feature activation+0.000
neither
Token neither
Feature activation+0.000
Baghdad
Token Baghdad
Feature activation+3.456
nor
Token nor
Feature activation+0.000
Washington
Token Washington
Feature activation+1.399
wanted
Token wanted
Feature activation+0.000
to
Token to
Feature activation+0.000
leave
Token leave
Feature activation+0.000
Asian
Token Asian
Feature activation+0.000
Americans
Token Americans
Feature activation+0.000
revealed
Token revealed
Feature activation+0.000
that
Token that
Feature activation+0.000
neither
Token neither
Feature activation+0.000
Democrats
Token Democrats
Feature activation+3.417
nor
Token nor
Feature activation+0.000
Republicans
Token Republicans
Feature activation+0.000
are
Token are
Feature activation+0.000
doing
Token doing
Feature activation+0.000
much
Token much
Feature activation+0.000
.
Token.
Feature activation+0.000
Neither
Token Neither
Feature activation+0.000
the
Token the
Feature activation+0.962
Catalan
Token Catalan
Feature activation+1.711
regional
Token regional
Feature activation+1.426
government
Token government
Feature activation+3.386
nor
Token nor
Feature activation+0.408
the
Token the
Feature activation+0.137
national
Token national
Feature activation+0.000
government
Token government
Feature activation+0.000
in
Token in
Feature activation+0.048
OH
TokenOH
Feature activation+0.000
R
TokenR
Feature activation+0.000
claims
Token claims
Feature activation+0.000
,
Token,
Feature activation+0.000
neither
Token neither
Feature activation+0.000
Reuters
Token Reuters
Feature activation+3.343
nor
Token nor
Feature activation+0.000
the
Token the
Feature activation+0.000
UK
Token UK
Feature activation+0.000
-
Token-
Feature activation+0.000
based
Tokenbased
Feature activation+0.000
ida
Tokenida
Feature activation+0.000
itself
Token itself
Feature activation+0.000
which
Token which
Feature activation+0.000
neither
Token neither
Feature activation+0.000
maintain
Token maintain
Feature activation+5.010
the
Token the
Feature activation+3.042
we
Token we
Feature activation+0.386
blog
Tokenblog
Feature activation+2.227
of
Token of
Feature activation+1.931
their
Token their
Feature activation+1.240
customers
Token customers
Feature activation+4.283
âĢ
TokenâĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
Neither
TokenNeither
Feature activation+0.000
the
Token the
Feature activation+0.348
Treasury
Token Treasury
Feature activation+1.451
Department
Token Department
Feature activation+3.026
nor
Token nor
Feature activation+0.000
the
Token the
Feature activation+0.000
Federal
Token Federal
Feature activation+0.000
Reserve
Token Reserve
Feature activation+0.000
believes
Token believes
Feature activation+0.000

Top DFA by src position
MAX = 5.884

and
Token and
Feature activation+0.002
Top resid features:
by
Token by
Feature activation-0.015
Top resid features:
a
Token a
Feature activation-0.015
Top resid features:
process
Token process
Feature activation+0.070
Top resid features:
which
Token which
Feature activation+0.087
Top resid features:
neither
Token neither
Feature activation+5.672
Top resid features:
they
Token they
Feature activation+0.266
Top resid features:
nor
Token nor
Feature activation+0.000
Top resid features:
we
Token we
Feature activation+0.000
Top resid features:
could
Token could
Feature activation+0.000
Top resid features:
successfully
Token successfully
Feature activation+0.000
Top resid features:
in
Token in
Feature activation+0.019
Top resid features:
No
Token No
Feature activation-0.006
Top resid features:
ida
Tokenida
Feature activation+0.007
Top resid features:
itself
Token itself
Feature activation+0.011
Top resid features:
which
Token which
Feature activation+0.047
Top resid features:
neither
Token neither
Feature activation+5.265
Top resid features:
maintain
Token maintain
Feature activation+0.598
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
we
Token we
Feature activation+0.000
Top resid features:
blog
Tokenblog
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
,
Token,
Feature activation-0.042
Top resid features:
the
Token the
Feature activation-0.004
Top resid features:
Republicans
Token Republicans
Feature activation-0.005
Top resid features:
will
Token will
Feature activation+0.009
Top resid features:
become
Token become
Feature activation+0.022
Top resid features:
neither
Token neither
Feature activation+5.612
Top resid features:
the
Token the
Feature activation+0.426
Top resid features:
party
Token party
Feature activation+0.164
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
Trump
Token Trump
Feature activation+0.000
Top resid features:
nor
Token nor
Feature activation+0.000
Top resid features:
reject
Token reject
Feature activation-0.042
Top resid features:
,
Token,
Feature activation+0.010
Top resid features:
but
Token but
Feature activation-0.002
Top resid features:
that
Token that
Feature activation-0.024
Top resid features:
's
Token's
Feature activation+0.008
Top resid features:
neither
Token neither
Feature activation+5.884
Top resid features:
here
Token here
Feature activation+0.086
Top resid features:
nor
Token nor
Feature activation+0.000
Top resid features:
there
Token there
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.011
Top resid features:
CIS
Token CIS
Feature activation+0.046
Top resid features:
,
Token,
Feature activation+0.053
Top resid features:
Ukraine
Token Ukraine
Feature activation+0.007
Top resid features:
has
Token has
Feature activation+0.013
Top resid features:
neither
Token neither
Feature activation+5.195
Top resid features:
signed
Token signed
Feature activation+0.090
Top resid features:
nor
Token nor
Feature activation+0.000
Top resid features:
ratified
Token ratified
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
charter
Token charter
Feature activation+0.000
Top resid features:
those
Token those
Feature activation+0.022
Top resid features:
for
Token for
Feature activation+0.008
Top resid features:
whom
Token whom
Feature activation+0.047
Top resid features:
there
Token there
Feature activation-0.003
Top resid features:
's
Token's
Feature activation+0.164
Top resid features:
neither
Token neither
Feature activation+5.162
Top resid features:
dear
Token dear
Feature activation+0.087
Top resid features:
nor
Token nor
Feature activation+0.000
Top resid features:
und
Token und
Feature activation+0.000
Top resid features:
ear
Tokenear
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.019
Top resid features:
who
Token who
Feature activation+0.012
Top resid features:
says
Token says
Feature activation-0.006
Top resid features:
he
Token he
Feature activation+0.007
Top resid features:
is
Token is
Feature activation+0.046
Top resid features:
neither
Token neither
Feature activation+5.197
Top resid features:
a
Token a
Feature activation+0.175
Top resid features:
Republican
Token Republican
Feature activation+0.085
Top resid features:
nor
Token nor
Feature activation+0.000
Top resid features:
a
Token a
Feature activation+0.000
Top resid features:
Democrat
Token Democrat
Feature activation+0.000
Top resid features:
in
Token in
Feature activation+0.004
Top resid features:
No
Token No
Feature activation-0.003
Top resid features:
ida
Tokenida
Feature activation+0.015
Top resid features:
itself
Token itself
Feature activation+0.012
Top resid features:
which
Token which
Feature activation+0.041
Top resid features:
neither
Token neither
Feature activation+4.409
Top resid features:
maintain
Token maintain
Feature activation+0.439
Top resid features:
the
Token the
Feature activation+0.096
Top resid features:
we
Token we
Feature activation+0.008
Top resid features:
blog
Tokenblog
Feature activation+0.169
Top resid features:
of
Token of
Feature activation+0.289
Top resid features:
video
Token video
Feature activation+0.004
Top resid features:
,
Token,
Feature activation+0.010
Top resid features:
the
Token the
Feature activation-0.004
Top resid features:
Vermont
Token Vermont
Feature activation-0.002
Top resid features:
senator
Token senator
Feature activation-0.002
Top resid features:
neither
Token neither
Feature activation+4.989
Top resid features:
conceded
Token conceded
Feature activation+0.130
Top resid features:
defeat
Token defeat
Feature activation+0.000
Top resid features:
nor
Token nor
Feature activation+0.000
Top resid features:
endorsed
Token endorsed
Feature activation+0.000
Top resid features:
his
Token his
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.016
Top resid features:
including
Token including
Feature activation+0.009
Top resid features:
those
Token those
Feature activation-0.002
Top resid features:
who
Token who
Feature activation-0.006
Top resid features:
were
Token were
Feature activation-0.011
Top resid features:
neither
Token neither
Feature activation+5.024
Top resid features:
Jewish
Token Jewish
Feature activation+0.144
Top resid features:
nor
Token nor
Feature activation+0.000
Top resid features:
Black
Token Black
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
In
Token In
Feature activation+0.000
Top resid features:
proving
Token proving
Feature activation-0.017
Top resid features:
to
Token to
Feature activation+0.019
Top resid features:
be
Token be
Feature activation-0.012
Top resid features:
difficult
Token difficult
Feature activation+0.009
Top resid features:
and
Token and
Feature activation+0.028
Top resid features:
neither
Token neither
Feature activation+4.764
Top resid features:
the
Token the
Feature activation+0.381
Top resid features:
vehicle
Token vehicle
Feature activation+0.102
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
or
Token or
Feature activation+0.000
Top resid features:
a
Token a
Feature activation+0.000
Top resid features:
programs
Token programs
Feature activation-0.007
Top resid features:
manager
Token manager
Feature activation+0.001
Top resid features:
,
Token,
Feature activation-0.010
Top resid features:
argues
Token argues
Feature activation+0.005
Top resid features:
that
Token that
Feature activation+0.007
Top resid features:
neither
Token neither
Feature activation+4.557
Top resid features:
smartphones
Token smartphones
Feature activation+0.417
Top resid features:
based
Token based
Feature activation+0.000
Top resid features:
on
Token on
Feature activation+0.000
Top resid features:
Google
Token Google
Feature activation+0.000
Top resid features:
's
Token's
Feature activation+0.000
Top resid features:
video
Token video
Feature activation+0.001
Top resid features:
,
Token,
Feature activation-0.001
Top resid features:
the
Token the
Feature activation-0.022
Top resid features:
Vermont
Token Vermont
Feature activation-0.001
Top resid features:
senator
Token senator
Feature activation+0.005
Top resid features:
neither
Token neither
Feature activation+4.557
Top resid features:
conceded
Token conceded
Feature activation+0.253
Top resid features:
defeat
Token defeat
Feature activation+0.091
Top resid features:
nor
Token nor
Feature activation+0.000
Top resid features:
endorsed
Token endorsed
Feature activation+0.000
Top resid features:
his
Token his
Feature activation+0.000
Top resid features:
,
Token,
Feature activation-0.084
Top resid features:
the
Token the
Feature activation-0.012
Top resid features:
Republicans
Token Republicans
Feature activation-0.001
Top resid features:
will
Token will
Feature activation+0.008
Top resid features:
become
Token become
Feature activation-0.019
Top resid features:
neither
Token neither
Feature activation+4.056
Top resid features:
the
Token the
Feature activation+0.294
Top resid features:
party
Token party
Feature activation+0.144
Top resid features:
of
Token of
Feature activation+0.412
Top resid features:
Trump
Token Trump
Feature activation+0.064
Top resid features:
nor
Token nor
Feature activation+0.000
Top resid features:
United
Token United
Feature activation+0.003
Top resid features:
States
Token States
Feature activation-0.004
Top resid features:
wanted
Token wanted
Feature activation+0.007
Top resid features:
,
Token,
Feature activation-0.029
Top resid features:
when
Token when
Feature activation-0.059
Top resid features:
neither
Token neither
Feature activation+4.475
Top resid features:
Baghdad
Token Baghdad
Feature activation+0.210
Top resid features:
nor
Token nor
Feature activation+0.000
Top resid features:
Washington
Token Washington
Feature activation+0.000
Top resid features:
wanted
Token wanted
Feature activation+0.000
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
of
Token of
Feature activation-0.003
Top resid features:
Asian
Token Asian
Feature activation-0.000
Top resid features:
Americans
Token Americans
Feature activation-0.029
Top resid features:
revealed
Token revealed
Feature activation-0.010
Top resid features:
that
Token that
Feature activation-0.012
Top resid features:
neither
Token neither
Feature activation+4.293
Top resid features:
Democrats
Token Democrats
Feature activation+0.261
Top resid features:
nor
Token nor
Feature activation+0.000
Top resid features:
Republicans
Token Republicans
Feature activation+0.000
Top resid features:
are
Token are
Feature activation+0.000
Top resid features:
doing
Token doing
Feature activation+0.000
Top resid features:
October
Token October
Feature activation+0.005
Top resid features:
7
Token 7
Feature activation+0.011
Top resid features:
in
Token in
Feature activation-0.005
Top resid features:
Madrid
Token Madrid
Feature activation-0.007
Top resid features:
.
Token.
Feature activation+0.009
Top resid features:
Neither
Token Neither
Feature activation+4.084
Top resid features:
the
Token the
Feature activation+0.228
Top resid features:
Catalan
Token Catalan
Feature activation+0.200
Top resid features:
regional
Token regional
Feature activation+0.038
Top resid features:
government
Token government
Feature activation+0.183
Top resid features:
nor
Token nor
Feature activation+0.000
Top resid features:
S
Token S
Feature activation-0.000
Top resid features:
OH
TokenOH
Feature activation-0.003
Top resid features:
R
TokenR
Feature activation+0.008
Top resid features:
claims
Token claims
Feature activation+0.024
Top resid features:
,
Token,
Feature activation-0.031
Top resid features:
neither
Token neither
Feature activation+4.318
Top resid features:
Reuters
Token Reuters
Feature activation+0.056
Top resid features:
nor
Token nor
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
UK
Token UK
Feature activation+0.000
Top resid features:
-
Token-
Feature activation+0.000
Top resid features:
in
Token in
Feature activation+0.011
Top resid features:
No
Token No
Feature activation-0.008
Top resid features:
ida
Tokenida
Feature activation-0.004
Top resid features:
itself
Token itself
Feature activation-0.002
Top resid features:
which
Token which
Feature activation-0.014
Top resid features:
neither
Token neither
Feature activation+2.596
Top resid features:
maintain
Token maintain
Feature activation+1.568
Top resid features:
the
Token the
Feature activation+0.153
Top resid features:
we
Token we
Feature activation+0.000
Top resid features:
blog
Tokenblog
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
.
Token.
Feature activation-0.011
Top resid features:
Ċ
TokenĊ
Feature activation+0.014
Top resid features:
Ċ
TokenĊ
Feature activation-0.001
Top resid features:
âĢ
TokenâĢ
Feature activation-0.017
Top resid features:
ľ
Tokenľ
Feature activation+0.030
Top resid features:
Neither
TokenNeither
Feature activation+3.979
Top resid features:
the
Token the
Feature activation+0.188
Top resid features:
Treasury
Token Treasury
Feature activation+0.019
Top resid features:
Department
Token Department
Feature activation+0.013
Top resid features:
nor
Token nor
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:

Decoder Weights Distribution

Head 0: 0.09

Head 1: 0.01

Head 2: 0.07

Head 3: 0.04

Head 4: 0.04

Head 5: 0.51

Head 6: 0.02

Head 7: 0.07

Head 8: 0.04

Head 9: 0.03

Head 10: 0.05

Head 11: 0.03

Positive logits

nor5.29

nor3.40

]"2.53

];2.30

Nor2.28

']2.26

';2.19

;}2.16

or2.14

}"2.14

]2.13

"/>2.11

}2.08

";1.99

;"1.93

],1.91

soever1.91

"}1.91

ught1.90

OR1.90

Negative logits

sometime-2.27

ometime-2.10

eventually-1.98

sorts-1.93

pretty-1.90

ospons-1.88

fortunately-1.87

soDeliveryDate-1.85

somewhat-1.77

somew-1.76

kinda-1.76

beware-1.74

coinc-1.74

awfully-1.73

culmin-1.71

somewhere-1.71

basically-1.71

interesting-1.70

entially-1.70

Somewhere-1.69

INTERVAL 4.586 - 5.096
CONTAINS 0.000%

No
Token No
Feature activation+0.000
ida
Tokenida
Feature activation+0.000
itself
Token itself
Feature activation+0.000
which
Token which
Feature activation+0.000
neither
Token neither
Feature activation+0.000
maintain
Token maintain
Feature activation+5.010
the
Token the
Feature activation+3.042
we
Token we
Feature activation+0.386
blog
Tokenblog
Feature activation+2.227
of
Token of
Feature activation+1.931
their
Token their
Feature activation+1.240
,
Token,
Feature activation+0.000
but
Token but
Feature activation+0.000
that
Token that
Feature activation+0.000
's
Token's
Feature activation+0.000
neither
Token neither
Feature activation+0.000
here
Token here
Feature activation+4.664
nor
Token nor
Feature activation+0.000
there
Token there
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Republicans
Token Republicans
Feature activation+0.000
will
Token will
Feature activation+0.000
become
Token become
Feature activation+0.000
neither
Token neither
Feature activation+0.000
the
Token the
Feature activation+1.324
party
Token party
Feature activation+4.969
of
Token of
Feature activation+1.669
Trump
Token Trump
Feature activation+3.617
nor
Token nor
Feature activation+0.000
the
Token the
Feature activation+0.000
party
Token party
Feature activation+0.000
by
Token by
Feature activation+0.000
a
Token a
Feature activation+0.000
process
Token process
Feature activation+0.000
which
Token which
Feature activation+0.000
neither
Token neither
Feature activation+0.000
they
Token they
Feature activation+5.096
nor
Token nor
Feature activation+0.354
we
Token we
Feature activation+0.000
could
Token could
Feature activation+0.009
successfully
Token successfully
Feature activation+0.000
control
Token control
Feature activation+0.000

INTERVAL 4.077 - 4.586
CONTAINS 0.001%

including
Token including
Feature activation+0.000
those
Token those
Feature activation+0.000
who
Token who
Feature activation+0.000
were
Token were
Feature activation+0.000
neither
Token neither
Feature activation+0.000
Jewish
Token Jewish
Feature activation+4.174
nor
Token nor
Feature activation+0.000
Black
Token Black
Feature activation+0.000
.
Token.
Feature activation+0.000
In
Token In
Feature activation+0.000
a
Token a
Feature activation+0.000
CIS
Token CIS
Feature activation+0.000
,
Token,
Feature activation+0.000
Ukraine
Token Ukraine
Feature activation+0.000
has
Token has
Feature activation+0.000
neither
Token neither
Feature activation+0.000
signed
Token signed
Feature activation+4.532
nor
Token nor
Feature activation+0.000
ratified
Token ratified
Feature activation+0.000
the
Token the
Feature activation+0.000
charter
Token charter
Feature activation+0.000
of
Token of
Feature activation+0.000
says
Token says
Feature activation+0.000
he
Token he
Feature activation+0.000
is
Token is
Feature activation+0.000
neither
Token neither
Feature activation+0.000
a
Token a
Feature activation+1.723
Republican
Token Republican
Feature activation+4.285
nor
Token nor
Feature activation+0.000
a
Token a
Feature activation+0.000
Democrat
Token Democrat
Feature activation+0.000
but
Token but
Feature activation+0.000
an
Token an
Feature activation+0.000
for
Token for
Feature activation+0.000
whom
Token whom
Feature activation+0.000
there
Token there
Feature activation+0.000
's
Token's
Feature activation+0.000
neither
Token neither
Feature activation+0.000
dear
Token dear
Feature activation+4.514
nor
Token nor
Feature activation+0.000
und
Token und
Feature activation+0.000
ear
Tokenear
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
be
Token be
Feature activation+0.000
difficult
Token difficult
Feature activation+0.000
and
Token and
Feature activation+0.000
neither
Token neither
Feature activation+0.000
the
Token the
Feature activation+0.525
vehicle
Token vehicle
Feature activation+4.144
,
Token,
Feature activation+2.136
or
Token or
Feature activation+0.265
a
Token a
Feature activation+0.000
suspect
Token suspect
Feature activation+0.000
,
Token,
Feature activation+0.000

INTERVAL 3.567 - 4.077
CONTAINS 0.000%

the
Token the
Feature activation+0.000
Vermont
Token Vermont
Feature activation+0.000
senator
Token senator
Feature activation+0.000
neither
Token neither
Feature activation+0.000
conceded
Token conceded
Feature activation+4.245
defeat
Token defeat
Feature activation+3.877
nor
Token nor
Feature activation+0.000
endorsed
Token endorsed
Feature activation+0.000
his
Token his
Feature activation+0.000
primary
Token primary
Feature activation+0.000
rival
Token rival
Feature activation+0.000
manager
Token manager
Feature activation+0.000
,
Token,
Feature activation+0.000
argues
Token argues
Feature activation+0.000
that
Token that
Feature activation+0.000
neither
Token neither
Feature activation+0.000
smartphones
Token smartphones
Feature activation+3.886
based
Token based
Feature activation+2.544
on
Token on
Feature activation+1.791
Google
Token Google
Feature activation+2.549
's
Token's
Feature activation+0.976
Android
Token Android
Feature activation+2.129
become
Token become
Feature activation+0.000
neither
Token neither
Feature activation+0.000
the
Token the
Feature activation+1.324
party
Token party
Feature activation+4.969
of
Token of
Feature activation+1.669
Trump
Token Trump
Feature activation+3.617
nor
Token nor
Feature activation+0.000
the
Token the
Feature activation+0.000
party
Token party
Feature activation+0.000
of
Token of
Feature activation+0.000
Bannon
Token Bannon
Feature activation+0.000

INTERVAL 3.057 - 3.567
CONTAINS 0.000%

States
Token States
Feature activation+0.000
wanted
Token wanted
Feature activation+0.000
,
Token,
Feature activation+0.000
when
Token when
Feature activation+0.000
neither
Token neither
Feature activation+0.000
Baghdad
Token Baghdad
Feature activation+3.456
nor
Token nor
Feature activation+0.000
Washington
Token Washington
Feature activation+1.399
wanted
Token wanted
Feature activation+0.000
to
Token to
Feature activation+0.000
leave
Token leave
Feature activation+0.000
Asian
Token Asian
Feature activation+0.000
Americans
Token Americans
Feature activation+0.000
revealed
Token revealed
Feature activation+0.000
that
Token that
Feature activation+0.000
neither
Token neither
Feature activation+0.000
Democrats
Token Democrats
Feature activation+3.417
nor
Token nor
Feature activation+0.000
Republicans
Token Republicans
Feature activation+0.000
are
Token are
Feature activation+0.000
doing
Token doing
Feature activation+0.000
much
Token much
Feature activation+0.000
OH
TokenOH
Feature activation+0.000
R
TokenR
Feature activation+0.000
claims
Token claims
Feature activation+0.000
,
Token,
Feature activation+0.000
neither
Token neither
Feature activation+0.000
Reuters
Token Reuters
Feature activation+3.343
nor
Token nor
Feature activation+0.000
the
Token the
Feature activation+0.000
UK
Token UK
Feature activation+0.000
-
Token-
Feature activation+0.000
based
Tokenbased
Feature activation+0.000
.
Token.
Feature activation+0.000
Neither
Token Neither
Feature activation+0.000
the
Token the
Feature activation+0.962
Catalan
Token Catalan
Feature activation+1.711
regional
Token regional
Feature activation+1.426
government
Token government
Feature activation+3.386
nor
Token nor
Feature activation+0.408
the
Token the
Feature activation+0.137
national
Token national
Feature activation+0.000
government
Token government
Feature activation+0.000
in
Token in
Feature activation+0.048

INTERVAL 2.548 - 3.057
CONTAINS 0.001%

ida
Tokenida
Feature activation+0.000
itself
Token itself
Feature activation+0.000
which
Token which
Feature activation+0.000
neither
Token neither
Feature activation+0.000
maintain
Token maintain
Feature activation+5.010
the
Token the
Feature activation+3.042
we
Token we
Feature activation+0.386
blog
Tokenblog
Feature activation+2.227
of
Token of
Feature activation+1.931
their
Token their
Feature activation+1.240
customers
Token customers
Feature activation+4.283
kids
Token kids
Feature activation+0.000
in
Token in
Feature activation+0.000
tow
Token tow
Feature activation+0.000
.
Token.
Feature activation+0.000
Neither
Token Neither
Feature activation+0.000
she
Token she
Feature activation+2.734
nor
Token nor
Feature activation+0.073
her
Token her
Feature activation+0.293
daughter
Token daughter
Feature activation+0.000
were
Token were
Feature activation+0.000
home
Token home
Feature activation+0.150
related
Token related
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Neither
TokenNeither
Feature activation+0.000
Amtrak
Token Amtrak
Feature activation+2.913
nor
Token nor
Feature activation+0.000
NJ
Token NJ
Feature activation+0.000
Transit
Token Transit
Feature activation+0.000
have
Token have
Feature activation+0.000
offered
Token offered
Feature activation+0.000
that
Token that
Feature activation+0.000
neither
Token neither
Feature activation+0.000
smartphones
Token smartphones
Feature activation+3.886
based
Token based
Feature activation+2.544
on
Token on
Feature activation+1.791
Google
Token Google
Feature activation+2.549
's
Token's
Feature activation+0.976
Android
Token Android
Feature activation+2.129
nor
Token nor
Feature activation+1.327
Apple
Token Apple
Feature activation+0.000
's
Token's
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
writer
Token writer
Feature activation+0.000
said
Token said
Feature activation+0.000
neither
Token neither
Feature activation+0.000
he
Token he
Feature activation+2.775
nor
Token nor
Feature activation+0.000
his
Token his
Feature activation+0.000
family
Token family
Feature activation+0.000
had
Token had
Feature activation+0.149
heard
Token heard
Feature activation+0.000

INTERVAL 2.038 - 2.548
CONTAINS 0.001%

Still
Token Still
Feature activation+0.000
,
Token,
Feature activation+0.000
neither
Token neither
Feature activation+0.000
world
Token world
Feature activation+0.536
of
Token of
Feature activation+0.292
performances
Token performances
Feature activation+2.133
hinted
Token hinted
Feature activation+0.234
at
Token at
Feature activation+0.209
a
Token a
Feature activation+0.095
blockbuster
Token blockbuster
Feature activation+0.000
-
Token-
Feature activation+0.000
difficult
Token difficult
Feature activation+0.000
and
Token and
Feature activation+0.000
neither
Token neither
Feature activation+0.000
the
Token the
Feature activation+0.525
vehicle
Token vehicle
Feature activation+4.144
,
Token,
Feature activation+2.136
or
Token or
Feature activation+0.265
a
Token a
Feature activation+0.000
suspect
Token suspect
Feature activation+0.000
,
Token,
Feature activation+0.000
has
Token has
Feature activation+0.000
was
Token was
Feature activation+0.000
so
Token so
Feature activation+0.000
great
Token great
Feature activation+0.000
that
Token that
Feature activation+0.000
neither
Token neither
Feature activation+0.000
man
Token man
Feature activation+2.529
nor
Token nor
Feature activation+0.000
beast
Token beast
Feature activation+0.000
could
Token could
Feature activation+0.097
endure
Token endure
Feature activation+0.000
it
Token it
Feature activation+0.000
,
Token,
Feature activation+0.000
having
Token having
Feature activation+0.000
learned
Token learned
Feature activation+0.000
neither
Token neither
Feature activation+0.000
how
Token how
Feature activation+1.493
to
Token to
Feature activation+2.515
read
Token read
Feature activation+1.987
nor
Token nor
Feature activation+0.000
write
Token write
Feature activation+0.000
,
Token,
Feature activation+0.000
Kem
Token Kem
Feature activation+0.000
,
Token,
Feature activation+0.000
argues
Token argues
Feature activation+0.000
that
Token that
Feature activation+0.000
neither
Token neither
Feature activation+0.000
smartphones
Token smartphones
Feature activation+3.886
based
Token based
Feature activation+2.544
on
Token on
Feature activation+1.791
Google
Token Google
Feature activation+2.549
's
Token's
Feature activation+0.976
Android
Token Android
Feature activation+2.129
nor
Token nor
Feature activation+1.327

INTERVAL 1.529 - 2.038
CONTAINS 0.001%

who
Token who
Feature activation+0.000
says
Token says
Feature activation+0.000
he
Token he
Feature activation+0.000
is
Token is
Feature activation+0.000
neither
Token neither
Feature activation+0.000
a
Token a
Feature activation+1.723
Republican
Token Republican
Feature activation+4.285
nor
Token nor
Feature activation+0.000
a
Token a
Feature activation+0.000
Democrat
Token Democrat
Feature activation+0.000
but
Token but
Feature activation+0.000
in
Token in
Feature activation+0.000
Madrid
Token Madrid
Feature activation+0.000
.
Token.
Feature activation+0.000
Neither
Token Neither
Feature activation+0.000
the
Token the
Feature activation+0.962
Catalan
Token Catalan
Feature activation+1.711
regional
Token regional
Feature activation+1.426
government
Token government
Feature activation+3.386
nor
Token nor
Feature activation+0.408
the
Token the
Feature activation+0.137
national
Token national
Feature activation+0.000
neither
Token neither
Feature activation+0.000
maintain
Token maintain
Feature activation+5.010
the
Token the
Feature activation+3.042
we
Token we
Feature activation+0.386
blog
Tokenblog
Feature activation+2.227
of
Token of
Feature activation+1.931
their
Token their
Feature activation+1.240
customers
Token customers
Feature activation+4.283
nor
Token nor
Feature activation+0.000
did
Token did
Feature activation+0.000
they
Token they
Feature activation+0.000
will
Token will
Feature activation+0.000
become
Token become
Feature activation+0.000
neither
Token neither
Feature activation+0.000
the
Token the
Feature activation+1.324
party
Token party
Feature activation+4.969
of
Token of
Feature activation+1.669
Trump
Token Trump
Feature activation+3.617
nor
Token nor
Feature activation+0.000
the
Token the
Feature activation+0.000
party
Token party
Feature activation+0.000
of
Token of
Feature activation+0.000
having
Token having
Feature activation+0.000
learned
Token learned
Feature activation+0.000
neither
Token neither
Feature activation+0.000
how
Token how
Feature activation+1.493
to
Token to
Feature activation+2.515
read
Token read
Feature activation+1.987
nor
Token nor
Feature activation+0.000
write
Token write
Feature activation+0.000
,
Token,
Feature activation+0.000
Kem
Token Kem
Feature activation+0.000
m
Tokenm
Feature activation+0.000

INTERVAL 1.019 - 1.529
CONTAINS 0.001%

Madrid
Token Madrid
Feature activation+0.000
.
Token.
Feature activation+0.000
Neither
Token Neither
Feature activation+0.000
the
Token the
Feature activation+0.962
Catalan
Token Catalan
Feature activation+1.711
regional
Token regional
Feature activation+1.426
government
Token government
Feature activation+3.386
nor
Token nor
Feature activation+0.408
the
Token the
Feature activation+0.137
national
Token national
Feature activation+0.000
government
Token government
Feature activation+0.000
10
Token 10
Feature activation+0.000
,
Token,
Feature activation+0.000
having
Token having
Feature activation+0.000
learned
Token learned
Feature activation+0.000
neither
Token neither
Feature activation+0.000
how
Token how
Feature activation+1.493
to
Token to
Feature activation+2.515
read
Token read
Feature activation+1.987
nor
Token nor
Feature activation+0.000
write
Token write
Feature activation+0.000
,
Token,
Feature activation+0.000
,
Token,
Feature activation+0.000
when
Token when
Feature activation+0.000
neither
Token neither
Feature activation+0.000
Baghdad
Token Baghdad
Feature activation+3.456
nor
Token nor
Feature activation+0.000
Washington
Token Washington
Feature activation+1.399
wanted
Token wanted
Feature activation+0.000
to
Token to
Feature activation+0.000
leave
Token leave
Feature activation+0.000
them
Token them
Feature activation+0.000
there
Token there
Feature activation+0.000
the
Token the
Feature activation+0.000
Republicans
Token Republicans
Feature activation+0.000
will
Token will
Feature activation+0.000
become
Token become
Feature activation+0.000
neither
Token neither
Feature activation+0.000
the
Token the
Feature activation+1.324
party
Token party
Feature activation+4.969
of
Token of
Feature activation+1.669
Trump
Token Trump
Feature activation+3.617
nor
Token nor
Feature activation+0.000
the
Token the
Feature activation+0.000
in
Tokenin
Feature activation+0.000
English
Token English
Feature activation+0.000
,
Token,
Feature activation+0.000
for
Token for
Feature activation+0.000
neither
Token neither
Feature activation+0.000
spoke
Token spoke
Feature activation+1.164
the
Token the
Feature activation+0.000
other
Token other
Feature activation+0.000
's
Token's
Feature activation+0.000
mother
Token mother
Feature activation+0.000
tongue
Token tongue
Feature activation+0.000

INTERVAL 0.510 - 1.019
CONTAINS 0.001%

.
Token.
Feature activation+0.000
Huawei
Token Huawei
Feature activation+0.000
refuses
Token refuses
Feature activation+0.000
to
Token to
Feature activation+0.000
either
Token either
Feature activation+0.000
confirm
Token confirm
Feature activation+0.754
or
Token or
Feature activation+0.000
deny
Token deny
Feature activation+0.000
any
Token any
Feature activation+0.000
launch
Token launch
Feature activation+0.000
plans
Token plans
Feature activation+0.000
realized
Token realized
Feature activation+0.000
that
Token that
Feature activation+0.000
neither
Token neither
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
boys
Token boys
Feature activation+0.742
had
Token had
Feature activation+0.886
returned
Token returned
Feature activation+0.000
home
Token home
Feature activation+0.000
by
Token by
Feature activation+0.000
5
Token 5
Feature activation+0.000
said
Token said
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Neither
TokenNeither
Feature activation+0.000
Lockheed
Token Lockheed
Feature activation+0.862
nor
Token nor
Feature activation+0.000
the
Token the
Feature activation+0.000
U
Token U
Feature activation+0.000
.
Token.
Feature activation+0.000
S
TokenS
Feature activation+0.000
to
Token to
Feature activation+0.000
be
Token be
Feature activation+0.000
difficult
Token difficult
Feature activation+0.000
and
Token and
Feature activation+0.000
neither
Token neither
Feature activation+0.000
the
Token the
Feature activation+0.525
vehicle
Token vehicle
Feature activation+4.144
,
Token,
Feature activation+2.136
or
Token or
Feature activation+0.265
a
Token a
Feature activation+0.000
suspect
Token suspect
Feature activation+0.000
7
Token 7
Feature activation+0.000
in
Token in
Feature activation+0.000
Madrid
Token Madrid
Feature activation+0.000
.
Token.
Feature activation+0.000
Neither
Token Neither
Feature activation+0.000
the
Token the
Feature activation+0.962
Catalan
Token Catalan
Feature activation+1.711
regional
Token regional
Feature activation+1.426
government
Token government
Feature activation+3.386
nor
Token nor
Feature activation+0.408
the
Token the
Feature activation+0.137

INTERVAL 0.000 - 0.510
CONTAINS 99.993%

Core
Token Core
Feature activation+0.000
Design
Token Design
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
1996
Token 1996
Feature activation+0.000
original
Token original
Feature activation+0.000
.
Token.
Feature activation+0.000
In
Token In
Feature activation+0.000
the
Token the
Feature activation+0.000
intervening
Token intervening
Feature activation+0.000
fight
Token fight
Feature activation+0.000
can
Token can
Feature activation+0.000
be
Token be
Feature activation+0.000
made
Token made
Feature activation+0.000
I
Token I
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
d
Tokend
Feature activation+0.000
love
Token love
Feature activation+0.000
it
Token it
Feature activation+0.000
.
Token.
Feature activation+0.000
said
Token said
Feature activation+0.000
that
Token that
Feature activation+0.000
there
Token there
Feature activation+0.000
is
Token is
Feature activation+0.000
now
Token now
Feature activation+0.000
more
Token more
Feature activation+0.000
racism
Token racism
Feature activation+0.000
against
Token against
Feature activation+0.000
whites
Token whites
Feature activation+0.000
than
Token than
Feature activation+0.000
against
Token against
Feature activation+0.000
IL
TokenIL
Feature activation+0.000
O
TokenO
Feature activation+0.000
had
Token had
Feature activation+0.000
a
Token a
Feature activation+0.000
number
Token number
Feature activation+0.000
of
Token of
Feature activation+0.000
high
Token high
Feature activation+0.000
-
Token-
Feature activation+0.000
impact
Tokenimpact
Feature activation+0.000
plays
Token plays
Feature activation+0.000
on
Token on
Feature activation+0.000
s
Tokens
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
and
Token and
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
O
TokenO
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000

BOTTOM ACTIVATIONS
MIN = 0.000

atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
&
Token &
Feature activation+0.000
Development
Token Development
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
&
Token &
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
rik
Tokenrik
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
rik
Tokenrik
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000

Top feature 9 in H8.5: (feature 20460)

TOP ACTIVATIONS
MAX = 5.580

as
Token as
Feature activation+0.000
Plan
Token Plan
Feature activation+0.000
ck
Tokenck
Feature activation+2.089
and
Token and
Feature activation+0.380
WM
Token WM
Feature activation+0.000
AP
TokenAP
Feature activation+5.580
)
Token)
Feature activation+0.000
have
Token have
Feature activation+0.000
mapped
Token mapped
Feature activation+0.000
temperature
Token temperature
Feature activation+0.000
fluctuations
Token fluctuations
Feature activation+0.000
and
Token and
Feature activation+0.000
the
Token the
Feature activation+0.131
B
Token B
Feature activation+0.000
aj
Tokenaj
Feature activation+0.000
aj
Tokenaj
Feature activation+1.145
s
Tokens
Feature activation+5.035
)
Token)
Feature activation+0.000
have
Token have
Feature activation+0.000
been
Token been
Feature activation+0.000
merged
Token merged
Feature activation+0.000
to
Token to
Feature activation+0.000
,
Token,
Feature activation+0.154
Dylan
Token Dylan
Feature activation+2.202
,
Token,
Feature activation+0.328
and
Token and
Feature activation+0.247
Orb
Token Orb
Feature activation+0.514
ison
Tokenison
Feature activation+4.944
)
Token)
Feature activation+0.000
decided
Token decided
Feature activation+0.000
not
Token not
Feature activation+0.000
to
Token to
Feature activation+0.000
use
Token use
Feature activation+0.000
which
Tokenwhich
Feature activation+0.608
permanently
Token permanently
Feature activation+0.000
ended
Token ended
Feature activation+0.363
his
Token his
Feature activation+0.049
flying
Token flying
Feature activation+0.380
career
Token career
Feature activation+4.607
)
Token)
Feature activation+0.000
from
Token from
Feature activation+0.000
the
Token the
Feature activation+0.000
Command
Token Command
Feature activation+0.000
ant
Tokenant
Feature activation+0.000
,
Token,
Feature activation+0.093
c
Token c
Feature activation+0.000
ate
Tokenate
Feature activation+0.000
rers
Tokenrers
Feature activation+0.427
,
Token,
Feature activation+0.229
etc
Token etc
Feature activation+4.592
.)
Token.)
Feature activation+0.000
."
Token."
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Major
TokenMajor
Feature activation+0.000
paid
Token paid
Feature activation+1.528
for
Token for
Feature activation+1.937
or
Token or
Feature activation+0.000
are
Token are
Feature activation+0.254
being
Token being
Feature activation+0.000
sold
Token sold
Feature activation+4.550
on
Token on
Feature activation+0.000
cons
Token cons
Feature activation+0.000
ignment
Tokenignment
Feature activation+3.639
),
Token),
Feature activation+0.000
we
Token we
Feature activation+0.000
the
Token the
Feature activation+0.000
78
Token 78
Feature activation+0.000
(
Token (
Feature activation+0.000
53
Token53
Feature activation+3.356
.
Token.
Feature activation+0.912
8
Token8
Feature activation+4.522
percent
Token percent
Feature activation+2.041
)
Token)
Feature activation+0.000
special
Token special
Feature activation+0.000
teams
Token teams
Feature activation+0.000
tackles
Token tackles
Feature activation+0.000
(
Token (
Feature activation+0.000
including
Tokenincluding
Feature activation+0.063
the
Token the
Feature activation+0.549
Am
Token Am
Feature activation+0.000
ban
Tokenban
Feature activation+0.000
is
Tokenis
Feature activation+4.460
,
Token,
Feature activation+2.625
the
Token the
Feature activation+0.498
J
Token J
Feature activation+0.000
ind
Tokenind
Feature activation+0.000
als
Tokenals
Feature activation+0.000
of
Token of
Feature activation+0.073
encryption
Token encryption
Feature activation+2.380
now
Token now
Feature activation+0.220
widely
Token widely
Feature activation+0.000
considered
Token considered
Feature activation+0.750
insecure
Token insecure
Feature activation+4.420
)
Token)
Feature activation+0.000
and
Token and
Feature activation+0.000
security
Token security
Feature activation+0.000
question
Token question
Feature activation+0.000
answers
Token answers
Feature activation+0.000
in
Tokenin
Feature activation+1.254
a
Token a
Feature activation+1.248
good
Token good
Feature activation+1.391
,
Token,
Feature activation+0.000
big
Token big
Feature activation+0.948
way
Token way
Feature activation+4.350
)
Token)
Feature activation+0.000
on
Token on
Feature activation+0.000
the
Token the
Feature activation+0.000
whole
Token whole
Feature activation+0.000
global
Token global
Feature activation+0.000
online
Token online
Feature activation+2.461
or
Token or
Feature activation+0.370
in
Token in
Feature activation+0.000
boutique
Token boutique
Feature activation+1.745
baby
Token baby
Feature activation+0.000
stores
Token stores
Feature activation+4.323
)
Token)
Feature activation+0.000
and
Token and
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
Un
Token Un
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
ver
Tokenver
Feature activation+1.847
and
Token and
Feature activation+0.225
D
Token D
Feature activation+0.000
HL
TokenHL
Feature activation+4.270
),
Token),
Feature activation+0.000
and
Token and
Feature activation+0.000
academics
Token academics
Feature activation+0.000
and
Token and
Feature activation+0.000
researchers
Token researchers
Feature activation+0.000
fs
Tokenfs
Feature activation+0.000
and
Token and
Feature activation+0.000
old
Token old
Feature activation+0.000
or
Token or
Feature activation+0.108
dusty
Token dusty
Feature activation+0.000
galaxies
Token galaxies
Feature activation+4.263
)
Token)
Feature activation+0.000
can
Token can
Feature activation+0.000
mimic
Token mimic
Feature activation+0.000
the
Token the
Feature activation+0.000
appearance
Token appearance
Feature activation+0.000
creature
Token creature
Feature activation+0.000
hunters
Token hunters
Feature activation+0.000
(
Token (
Feature activation+0.000
aka
Tokenaka
Feature activation+0.000
Pokemon
Token Pokemon
Feature activation+1.295
trainers
Token trainers
Feature activation+4.239
)
Token)
Feature activation+0.000
was
Token was
Feature activation+0.000
a
Token a
Feature activation+0.000
big
Token big
Feature activation+0.000
part
Token part
Feature activation+0.000
bike
Token bike
Feature activation+0.000
spaces
Token spaces
Feature activation+0.428
than
Token than
Feature activation+0.000
the
Token the
Feature activation+0.000
old
Token old
Feature activation+1.055
ones
Token ones
Feature activation+4.221
),
Token),
Feature activation+0.000
neither
Token neither
Feature activation+0.000
is
Token is
Feature activation+0.000
as
Token as
Feature activation+0.000
impressive
Token impressive
Feature activation+0.000
care
Token care
Feature activation+0.000
(
Token (
Feature activation+0.000
for
Tokenfor
Feature activation+0.733
children
Token children
Feature activation+2.705
under
Token under
Feature activation+0.061
18
Token 18
Feature activation+4.177
months
Token months
Feature activation+2.974
)
Token)
Feature activation+0.000
much
Token much
Feature activation+0.000
more
Token more
Feature activation+0.000
expensive
Token expensive
Feature activation+0.000
to
Token to
Feature activation+0.000
other
Token other
Feature activation+0.000
organizations
Token organizations
Feature activation+0.000
(
Token (
Feature activation+0.000
and
Tokenand
Feature activation+0.000
churches
Token churches
Feature activation+4.128
)
Token)
Feature activation+0.012
for
Token for
Feature activation+0.000
whom
Token whom
Feature activation+0.000
this
Token this
Feature activation+0.000
was
Token was
Feature activation+0.000
mil
Tokenmil
Feature activation+0.000
k
Tokenk
Feature activation+0.000
,
Token,
Feature activation+0.454
eggs
Token eggs
Feature activation+0.500
,
Token,
Feature activation+0.471
etc
Token etc
Feature activation+4.104
.),
Token.),
Feature activation+0.000
as
Token as
Feature activation+0.000
well
Token well
Feature activation+0.000
as
Token as
Feature activation+0.000
hot
Token hot
Feature activation+0.000
washing
Token washing
Feature activation+0.000
machine
Token machine
Feature activation+0.954
and
Token and
Feature activation+0.000
air
Token air
Feature activation+0.505
condition
Token condition
Feature activation+0.000
er
Tokener
Feature activation+4.096
)
Token)
Feature activation+0.000
âĢĵ
Token âĢĵ
Feature activation+0.000
popular
Token popular
Feature activation+0.000
ly
Tokenly
Feature activation+0.000
known
Token known
Feature activation+0.000
spoke
Token spoke
Feature activation+0.000
the
Token the
Feature activation+0.000
other
Token other
Feature activation+3.192
's
Token's
Feature activation+0.023
mother
Token mother
Feature activation+0.665
tongue
Token tongue
Feature activation+4.051
confidently
Token confidently
Feature activation+1.815
)
Token)
Feature activation+0.000
about
Token about
Feature activation+0.000
everything
Token everything
Feature activation+0.000
under
Token under
Feature activation+0.000

Top DFA by src position
MAX = 6.163

moments
Token moments
Feature activation+0.003
Top resid features:
.
Token.
Feature activation+0.007
Top resid features:
Most
Token Most
Feature activation-0.001
Top resid features:
observ
Token observ
Feature activation+0.005
Top resid features:
atories
Tokenatories
Feature activation+0.013
Top resid features:
(
Token (
Feature activation+6.163
Top resid features:
such
Tokensuch
Feature activation+0.083
Top resid features:
as
Token as
Feature activation+0.148
Top resid features:
Plan
Token Plan
Feature activation+0.004
Top resid features:
ck
Tokenck
Feature activation-0.015
Top resid features:
and
Token and
Feature activation+0.115
Top resid features:
in
Token in
Feature activation+0.005
Top resid features:
the
Token the
Feature activation+0.001
Top resid features:
period
Token period
Feature activation+0.000
Top resid features:
under
Token under
Feature activation+0.004
Top resid features:
consideration
Token consideration
Feature activation-0.007
Top resid features:
(
Token (
Feature activation+5.515
Top resid features:
including
Tokenincluding
Feature activation+0.053
Top resid features:
the
Token the
Feature activation+0.019
Top resid features:
Am
Token Am
Feature activation+0.002
Top resid features:
ban
Tokenban
Feature activation+0.004
Top resid features:
is
Tokenis
Feature activation+0.015
Top resid features:
.
Token.
Feature activation-0.018
Top resid features:
The
Token The
Feature activation-0.006
Top resid features:
five
Token five
Feature activation+0.027
Top resid features:
front
Token front
Feature activation-0.001
Top resid features:
men
Tokenmen
Feature activation+0.013
Top resid features:
(
Token (
Feature activation+5.701
Top resid features:
H
TokenH
Feature activation-0.001
Top resid features:
arrison
Tokenarrison
Feature activation-0.001
Top resid features:
,
Token,
Feature activation+0.011
Top resid features:
Lyn
Token Lyn
Feature activation+0.002
Top resid features:
ne
Tokenne
Feature activation-0.001
Top resid features:
still
Token still
Feature activation-0.005
Top resid features:
suffered
Token suffered
Feature activation+0.060
Top resid features:
from
Token from
Feature activation+0.004
Top resid features:
double
Token double
Feature activation-0.011
Top resid features:
vision
Token vision
Feature activation+0.010
Top resid features:
(
Token (
Feature activation+5.379
Top resid features:
which
Tokenwhich
Feature activation+0.077
Top resid features:
permanently
Token permanently
Feature activation-0.014
Top resid features:
ended
Token ended
Feature activation+0.066
Top resid features:
his
Token his
Feature activation-0.023
Top resid features:
flying
Token flying
Feature activation-0.010
Top resid features:
inspectors
Token inspectors
Feature activation-0.000
Top resid features:
inspect
Token inspect
Feature activation+0.023
Top resid features:
retail
Token retail
Feature activation+0.006
Top resid features:
food
Token food
Feature activation+0.004
Top resid features:
establishments
Token establishments
Feature activation+0.010
Top resid features:
(
Token (
Feature activation+5.411
Top resid features:
rest
Tokenrest
Feature activation+0.005
Top resid features:
aur
Tokenaur
Feature activation-0.002
Top resid features:
ants
Tokenants
Feature activation+0.007
Top resid features:
,
Token,
Feature activation+0.032
Top resid features:
markets
Token markets
Feature activation+0.008
Top resid features:
nikov
Tokennikov
Feature activation+0.001
Top resid features:
Concern
Token Concern
Feature activation-0.000
Top resid features:
has
Token has
Feature activation-0.000
Top resid features:
an
Token an
Feature activation+0.000
Top resid features:
interest
Token interest
Feature activation-0.001
Top resid features:
(
Token (
Feature activation+5.067
Top resid features:
for
Tokenfor
Feature activation+0.040
Top resid features:
example
Token example
Feature activation+0.014
Top resid features:
,
Token,
Feature activation+0.037
Top resid features:
the
Token the
Feature activation+0.020
Top resid features:
products
Token products
Feature activation+0.020
Top resid features:
for
Token for
Feature activation+0.011
Top resid features:
42
Token 42
Feature activation+0.005
Top resid features:
of
Token of
Feature activation-0.027
Top resid features:
the
Token the
Feature activation-0.004
Top resid features:
78
Token 78
Feature activation+0.013
Top resid features:
(
Token (
Feature activation+5.565
Top resid features:
53
Token53
Feature activation+0.025
Top resid features:
.
Token.
Feature activation+0.105
Top resid features:
8
Token8
Feature activation-0.042
Top resid features:
percent
Token percent
Feature activation+0.000
Top resid features:
)
Token)
Feature activation+0.000
Top resid features:
in
Token in
Feature activation+0.006
Top resid features:
the
Token the
Feature activation+0.002
Top resid features:
period
Token period
Feature activation-0.075
Top resid features:
under
Token under
Feature activation-0.000
Top resid features:
consideration
Token consideration
Feature activation-0.012
Top resid features:
(
Token (
Feature activation+5.145
Top resid features:
including
Tokenincluding
Feature activation+0.161
Top resid features:
the
Token the
Feature activation+0.069
Top resid features:
Am
Token Am
Feature activation+0.010
Top resid features:
ban
Tokenban
Feature activation+0.007
Top resid features:
is
Tokenis
Feature activation+0.059
Top resid features:
5
Token5
Feature activation+0.011
Top resid features:
-
Token-
Feature activation+0.011
Top resid features:
h
Tokenh
Feature activation+0.000
Top resid features:
ashed
Tokenashed
Feature activation+0.006
Top resid features:
passwords
Token passwords
Feature activation+0.027
Top resid features:
(
Token (
Feature activation+4.796
Top resid features:
a
Tokena
Feature activation+0.052
Top resid features:
form
Token form
Feature activation+0.025
Top resid features:
of
Token of
Feature activation+0.031
Top resid features:
encryption
Token encryption
Feature activation+0.026
Top resid features:
now
Token now
Feature activation+0.133
Top resid features:
The
Token The
Feature activation+0.005
Top resid features:
game
Token game
Feature activation+0.015
Top resid features:
also
Token also
Feature activation+0.006
Top resid features:
scales
Token scales
Feature activation-0.016
Top resid features:
back
Token back
Feature activation+0.004
Top resid features:
(
Token (
Feature activation+5.109
Top resid features:
in
Tokenin
Feature activation+0.086
Top resid features:
a
Token a
Feature activation+0.005
Top resid features:
good
Token good
Feature activation+0.040
Top resid features:
,
Token,
Feature activation+0.060
Top resid features:
big
Token big
Feature activation+0.013
Top resid features:
Ŀ
TokenĿ
Feature activation+0.002
Top resid features:
z
Token z
Feature activation+0.004
Top resid features:
ipp
Tokenipp
Feature activation-0.002
Top resid features:
ered
Tokenered
Feature activation-0.000
Top resid features:
sheets
Token sheets
Feature activation+0.012
Top resid features:
(
Token (
Feature activation+5.110
Top resid features:
available
Tokenavailable
Feature activation+0.035
Top resid features:
online
Token online
Feature activation+0.082
Top resid features:
or
Token or
Feature activation+0.022
Top resid features:
in
Token in
Feature activation+0.020
Top resid features:
boutique
Token boutique
Feature activation+0.023
Top resid features:
companies
Token companies
Feature activation+0.021
Top resid features:
with
Token with
Feature activation+0.011
Top resid features:
significant
Token significant
Feature activation-0.006
Top resid features:
logistics
Token logistics
Feature activation-0.114
Top resid features:
needs
Token needs
Feature activation-0.001
Top resid features:
(
Token (
Feature activation+5.098
Top resid features:
including
Tokenincluding
Feature activation+0.180
Top resid features:
Un
Token Un
Feature activation+0.007
Top resid features:
ile
Tokenile
Feature activation+0.004
Top resid features:
ver
Tokenver
Feature activation+0.011
Top resid features:
and
Token and
Feature activation+0.039
Top resid features:
important
Token important
Feature activation+0.001
Top resid features:
,
Token,
Feature activation+0.004
Top resid features:
as
Token as
Feature activation-0.031
Top resid features:
nearby
Token nearby
Feature activation+0.027
Top resid features:
objects
Token objects
Feature activation+0.013
Top resid features:
(
Token (
Feature activation+4.835
Top resid features:
such
Tokensuch
Feature activation+0.038
Top resid features:
as
Token as
Feature activation+0.106
Top resid features:
red
Token red
Feature activation+0.004
Top resid features:
stars
Token stars
Feature activation+0.017
Top resid features:
,
Token,
Feature activation+0.015
Top resid features:
ize
Tokenize
Feature activation+0.005
Top resid features:
with
Token with
Feature activation-0.010
Top resid features:
other
Token other
Feature activation+0.010
Top resid features:
creature
Token creature
Feature activation+0.013
Top resid features:
hunters
Token hunters
Feature activation+0.010
Top resid features:
(
Token (
Feature activation+4.955
Top resid features:
aka
Tokenaka
Feature activation+0.239
Top resid features:
Pokemon
Token Pokemon
Feature activation+0.026
Top resid features:
trainers
Token trainers
Feature activation+0.009
Top resid features:
)
Token)
Feature activation+0.000
Top resid features:
was
Token was
Feature activation+0.000
Top resid features:
and
Token and
Feature activation+0.006
Top resid features:
Berlin
Token Berlin
Feature activation+0.004
Top resid features:
both
Token both
Feature activation+0.006
Top resid features:
allow
Token allow
Feature activation-0.021
Top resid features:
bikes
Token bikes
Feature activation+0.005
Top resid features:
(
Token (
Feature activation+4.661
Top resid features:
although
Tokenalthough
Feature activation+0.126
Top resid features:
I
Token I
Feature activation+0.021
Top resid features:
hear
Token hear
Feature activation-0.009
Top resid features:
the
Token the
Feature activation+0.014
Top resid features:
new
Token new
Feature activation+0.039
Top resid features:
older
Token older
Feature activation+0.008
Top resid features:
,
Token,
Feature activation-0.004
Top resid features:
with
Token with
Feature activation+0.023
Top resid features:
infant
Token infant
Feature activation+0.011
Top resid features:
care
Token care
Feature activation+0.029
Top resid features:
(
Token (
Feature activation+4.751
Top resid features:
for
Tokenfor
Feature activation+0.167
Top resid features:
children
Token children
Feature activation+0.134
Top resid features:
under
Token under
Feature activation+0.017
Top resid features:
18
Token 18
Feature activation+0.063
Top resid features:
months
Token months
Feature activation+0.000
Top resid features:
barriers
Token barriers
Feature activation+0.009
Top resid features:
and
Token and
Feature activation+0.011
Top resid features:
to
Token to
Feature activation+0.015
Top resid features:
other
Token other
Feature activation-0.015
Top resid features:
organizations
Token organizations
Feature activation+0.016
Top resid features:
(
Token (
Feature activation+4.539
Top resid features:
and
Tokenand
Feature activation+0.449
Top resid features:
churches
Token churches
Feature activation+0.041
Top resid features:
)
Token)
Feature activation+0.000
Top resid features:
for
Token for
Feature activation+0.000
Top resid features:
whom
Token whom
Feature activation+0.000
Top resid features:
goods
Token goods
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.005
Top resid features:
and
Token and
Feature activation+0.024
Top resid features:
grocery
Token grocery
Feature activation-0.004
Top resid features:
staples
Token staples
Feature activation-0.002
Top resid features:
(
Token (
Feature activation+4.766
Top resid features:
mil
Tokenmil
Feature activation+0.009
Top resid features:
k
Tokenk
Feature activation+0.010
Top resid features:
,
Token,
Feature activation+0.010
Top resid features:
eggs
Token eggs
Feature activation+0.012
Top resid features:
,
Token,
Feature activation+0.027
Top resid features:
turn
Token turn
Feature activation+0.038
Top resid features:
Internet
Token Internet
Feature activation+0.022
Top resid features:
-
Token-
Feature activation+0.008
Top resid features:
connected
Tokenconnected
Feature activation+0.014
Top resid features:
devices
Token devices
Feature activation+0.019
Top resid features:
(
Token (
Feature activation+4.838
Top resid features:
pr
Tokenpr
Feature activation+0.011
Top resid features:
inter
Tokeninter
Feature activation-0.002
Top resid features:
,
Token,
Feature activation-0.009
Top resid features:
washing
Token washing
Feature activation-0.004
Top resid features:
machine
Token machine
Feature activation+0.010
Top resid features:
heated
Token heated
Feature activation-0.002
Top resid features:
discussion
Token discussion
Feature activation+0.004
Top resid features:
with
Token with
Feature activation+0.014
Top resid features:
one
Token one
Feature activation+0.002
Top resid features:
another
Token another
Feature activation+0.004
Top resid features:
(
Token (
Feature activation+4.945
Top resid features:
in
Tokenin
Feature activation+0.076
Top resid features:
English
Token English
Feature activation+0.015
Top resid features:
,
Token,
Feature activation+0.025
Top resid features:
for
Token for
Feature activation+0.042
Top resid features:
neither
Token neither
Feature activation+0.023
Top resid features:

Decoder Weights Distribution

Head 0: 0.07

Head 1: 0.02

Head 2: 0.07

Head 3: 0.04

Head 4: 0.03

Head 5: 0.51

Head 6: 0.02

Head 7: 0.07

Head 8: 0.03

Head 9: 0.03

Head 10: 0.08

Head 11: 0.02

Positive logits

)5.27

')4.93

),4.71

*)4.61

)?4.51

")4.43

?)4.41

+)4.41

)"4.36

)4.35

)",4.26

),"4.26

'),4.24

!)4.23

%)4.22

),4.12

/)4.11

?),4.07

)-4.06

)—4.04

Negative logits

guiName-2.59

.''.-2.19

%.-1.95

.」-1.94

":{"-1.84

proverb-1.78

."-1.70

-1.66

.-1.63

.''-1.63

Poké-1.62

!.-1.61

'.-1.56

-1.56

Reviewed-1.55

luaj-1.54

.:-1.53

orem-1.49

Edited-1.49

."[-1.48

INTERVAL 5.022 - 5.580
CONTAINS 0.000%

as
Token as
Feature activation+0.000
Plan
Token Plan
Feature activation+0.000
ck
Tokenck
Feature activation+2.089
and
Token and
Feature activation+0.380
WM
Token WM
Feature activation+0.000
AP
TokenAP
Feature activation+5.580
)
Token)
Feature activation+0.000
have
Token have
Feature activation+0.000
mapped
Token mapped
Feature activation+0.000
temperature
Token temperature
Feature activation+0.000
fluctuations
Token fluctuations
Feature activation+0.000
and
Token and
Feature activation+0.000
the
Token the
Feature activation+0.131
B
Token B
Feature activation+0.000
aj
Tokenaj
Feature activation+0.000
aj
Tokenaj
Feature activation+1.145
s
Tokens
Feature activation+5.035
)
Token)
Feature activation+0.000
have
Token have
Feature activation+0.000
been
Token been
Feature activation+0.000
merged
Token merged
Feature activation+0.000
to
Token to
Feature activation+0.000

INTERVAL 4.464 - 5.022
CONTAINS 0.001%

,
Token,
Feature activation+0.154
Dylan
Token Dylan
Feature activation+2.202
,
Token,
Feature activation+0.328
and
Token and
Feature activation+0.247
Orb
Token Orb
Feature activation+0.514
ison
Tokenison
Feature activation+4.944
)
Token)
Feature activation+0.000
decided
Token decided
Feature activation+0.000
not
Token not
Feature activation+0.000
to
Token to
Feature activation+0.000
use
Token use
Feature activation+0.000
the
Token the
Feature activation+0.000
78
Token 78
Feature activation+0.000
(
Token (
Feature activation+0.000
53
Token53
Feature activation+3.356
.
Token.
Feature activation+0.912
8
Token8
Feature activation+4.522
percent
Token percent
Feature activation+2.041
)
Token)
Feature activation+0.000
special
Token special
Feature activation+0.000
teams
Token teams
Feature activation+0.000
tackles
Token tackles
Feature activation+0.000
,
Token,
Feature activation+0.093
c
Token c
Feature activation+0.000
ate
Tokenate
Feature activation+0.000
rers
Tokenrers
Feature activation+0.427
,
Token,
Feature activation+0.229
etc
Token etc
Feature activation+4.592
.)
Token.)
Feature activation+0.000
."
Token."
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Major
TokenMajor
Feature activation+0.000
which
Tokenwhich
Feature activation+0.608
permanently
Token permanently
Feature activation+0.000
ended
Token ended
Feature activation+0.363
his
Token his
Feature activation+0.049
flying
Token flying
Feature activation+0.380
career
Token career
Feature activation+4.607
)
Token)
Feature activation+0.000
from
Token from
Feature activation+0.000
the
Token the
Feature activation+0.000
Command
Token Command
Feature activation+0.000
ant
Tokenant
Feature activation+0.000
paid
Token paid
Feature activation+1.528
for
Token for
Feature activation+1.937
or
Token or
Feature activation+0.000
are
Token are
Feature activation+0.254
being
Token being
Feature activation+0.000
sold
Token sold
Feature activation+4.550
on
Token on
Feature activation+0.000
cons
Token cons
Feature activation+0.000
ignment
Tokenignment
Feature activation+3.639
),
Token),
Feature activation+0.000
we
Token we
Feature activation+0.000

INTERVAL 3.906 - 4.464
CONTAINS 0.002%

creature
Token creature
Feature activation+0.000
hunters
Token hunters
Feature activation+0.000
(
Token (
Feature activation+0.000
aka
Tokenaka
Feature activation+0.000
Pokemon
Token Pokemon
Feature activation+1.295
trainers
Token trainers
Feature activation+4.239
)
Token)
Feature activation+0.000
was
Token was
Feature activation+0.000
a
Token a
Feature activation+0.000
big
Token big
Feature activation+0.000
part
Token part
Feature activation+0.000
to
Token to
Feature activation+0.383
55
Token 55
Feature activation+1.888
.
Token.
Feature activation+0.000
1
Token1
Feature activation+2.374
per
Token per
Feature activation+0.000
cent
Token cent
Feature activation+3.964
)
Token)
Feature activation+0.000
and
Token and
Feature activation+0.000
in
Token in
Feature activation+0.000
Ontario
Token Ontario
Feature activation+0.000
and
Token and
Feature activation+0.000
(
Token (
Feature activation+0.000
including
Tokenincluding
Feature activation+0.063
the
Token the
Feature activation+0.549
Am
Token Am
Feature activation+0.000
ban
Tokenban
Feature activation+0.000
is
Tokenis
Feature activation+4.460
,
Token,
Feature activation+2.625
the
Token the
Feature activation+0.498
J
Token J
Feature activation+0.000
ind
Tokenind
Feature activation+0.000
als
Tokenals
Feature activation+0.000
paper
Token paper
Feature activation+2.098
and
Token and
Feature activation+0.791
drawing
Token drawing
Feature activation+1.610
lines
Token lines
Feature activation+3.499
and
Token and
Feature activation+0.802
shapes
Token shapes
Feature activation+3.967
),
Token),
Feature activation+0.000
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
fared
Token fared
Feature activation+0.000
better
Token better
Feature activation+0.000
proportion
Token proportion
Feature activation+0.140
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
speed
Token speed
Feature activation+0.076
of
Token of
Feature activation+0.000
light
Token light
Feature activation+3.944
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
laboratory
Token laboratory
Feature activation+0.000
.
Token.
Feature activation+0.000

INTERVAL 3.348 - 3.906
CONTAINS 0.004%

production
Token production
Feature activation+0.283
in
Token in
Feature activation+0.000
-
Token-
Feature activation+0.000
car
Tokencar
Feature activation+1.610
navigation
Token navigation
Feature activation+1.516
system
Token system
Feature activation+3.421
),
Token),
Feature activation+0.000
and
Token and
Feature activation+0.000
the
Token the
Feature activation+0.000
later
Token later
Feature activation+0.000
models
Token models
Feature activation+0.000
04
Token04
Feature activation+1.813
which
Token which
Feature activation+0.381
is
Token is
Feature activation+0.000
no
Token no
Feature activation+0.000
longer
Token longer
Feature activation+0.000
supported
Token supported
Feature activation+3.770
)
Token)
Feature activation+0.077
from
Token from
Feature activation+0.000
later
Token later
Feature activation+0.000
today
Token today
Feature activation+0.000
.
Token.
Feature activation+0.000
be
Token be
Feature activation+0.000
interpreted
Token interpreted
Feature activation+0.000
as
Token as
Feature activation+0.000
running
Token running
Feature activation+0.000
(
Token (
Feature activation+0.000
mostly
Tokenmostly
Feature activation+3.855
)
Token)
Feature activation+0.000
from
Token from
Feature activation+0.000
money
Token money
Feature activation+0.000
to
Token to
Feature activation+0.000
output
Token output
Feature activation+0.000
are
Token are
Feature activation+0.254
being
Token being
Feature activation+0.000
sold
Token sold
Feature activation+4.550
on
Token on
Feature activation+0.000
cons
Token cons
Feature activation+0.000
ignment
Tokenignment
Feature activation+3.639
),
Token),
Feature activation+0.000
we
Token we
Feature activation+0.000
advise
Token advise
Feature activation+0.000
that
Token that
Feature activation+0.000
U
Token U
Feature activation+0.000
the
Token the
Feature activation+0.000
bolt
Token bolt
Feature activation+0.000
will
Token will
Feature activation+0.000
cock
Token cock
Feature activation+0.000
(
Token (
Feature activation+0.000
catch
Tokencatch
Feature activation+3.439
)
Token)
Feature activation+0.000
on
Token on
Feature activation+0.000
the
Token the
Feature activation+0.000
sear
Token sear
Feature activation+0.000
mechanism
Token mechanism
Feature activation+0.000

INTERVAL 2.790 - 3.348
CONTAINS 0.006%

,
Token,
Feature activation+0.000
also
Token also
Feature activation+0.000
(
Token (
Feature activation+0.000
ch
Tokench
Feature activation+0.000
ances
Tokenances
Feature activation+0.000
are
Token are
Feature activation+3.034
,
Token,
Feature activation+0.000
you
Token you
Feature activation+0.518
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
ve
Tokenve
Feature activation+0.058
so
Tokenso
Feature activation+0.000
only
Token only
Feature activation+0.000
an
Token an
Feature activation+0.000
icon
Token icon
Feature activation+0.000
is
Token is
Feature activation+0.000
displayed
Token displayed
Feature activation+3.320
),
Token),
Feature activation+0.000
toggle
Token toggle
Feature activation+0.000
Unity
Token Unity
Feature activation+0.000
launcher
Token launcher
Feature activation+0.000
visibility
Token visibility
Feature activation+0.000
caused
Token caused
Feature activation+0.000
actual
Token actual
Feature activation+0.000
(
Token (
Feature activation+0.000
per
Tokenper
Feature activation+0.524
cept
Tokencept
Feature activation+0.545
ible
Tokenible
Feature activation+2.889
)
Token)
Feature activation+0.000
harm
Token harm
Feature activation+0.000
by
Token by
Feature activation+0.000
stealing
Token stealing
Feature activation+0.000
only
Token only
Feature activation+0.000
any
Token any
Feature activation+0.000
ecosystem
Token ecosystem
Feature activation+0.000
(
Token (
Feature activation+0.000
present
Tokenpresent
Feature activation+0.178
or
Token or
Feature activation+0.000
past
Token past
Feature activation+2.791
[
Token[
Feature activation+0.562
2
Token2
Feature activation+0.065
])
Token])
Feature activation+0.000
that
Token that
Feature activation+0.000
is
Token is
Feature activation+0.000
it
Tokenit
Feature activation+0.066
os
Tokenos
Feature activation+1.236
and
Token and
Feature activation+0.218
Tw
Token Tw
Feature activation+0.000
ink
Tokenink
Feature activation+0.000
ies
Tokenies
Feature activation+3.059
,
Token,
Feature activation+0.377
in
Token in
Feature activation+0.000
case
Token case
Feature activation+0.050
you
Token you
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000

INTERVAL 2.232 - 2.790
CONTAINS 0.014%

Swift
Token Swift
Feature activation+0.000
(
Token (
Feature activation+0.000
Jason
TokenJason
Feature activation+0.000
Mer
Token Mer
Feature activation+0.000
r
Tokenr
Feature activation+0.000
itt
Tokenitt
Feature activation+2.437
/
Token/
Feature activation+0.000
Getty
TokenGetty
Feature activation+0.365
Images
Token Images
Feature activation+0.000
)
Token)
Feature activation+0.000
and
Token and
Feature activation+0.000
is
Token is
Feature activation+0.000
now
Token now
Feature activation+0.000
(
Token (
Feature activation+0.000
supp
Tokensupp
Feature activation+0.000
osed
Tokenosed
Feature activation+0.804
ly
Tokenly
Feature activation+2.452
)
Token)
Feature activation+0.000
the
Token the
Feature activation+0.000
most
Token most
Feature activation+0.000
paid
Token paid
Feature activation+0.000
Pac
Token Pac
Feature activation+0.000
with
Tokenwith
Feature activation+0.098
paddle
Token paddle
Feature activation+0.194
-
Token-
Feature activation+0.000
sh
Tokensh
Feature activation+0.000
if
Tokenif
Feature activation+0.000
ters
Tokenters
Feature activation+2.355
)
Token)
Feature activation+0.000
will
Token will
Feature activation+0.000
be
Token be
Feature activation+0.000
available
Token available
Feature activation+0.000
and
Token and
Feature activation+0.000
using
Tokenusing
Feature activation+0.200
the
Token the
Feature activation+0.000
same
Token same
Feature activation+0.000
maths
Token maths
Feature activation+0.000
as
Token as
Feature activation+0.094
before
Token before
Feature activation+2.636
)
Token)
Feature activation+0.000
and
Token and
Feature activation+0.000
gives
Token gives
Feature activation+0.000
you
Token you
Feature activation+0.000
an
Token an
Feature activation+0.000
menu
Token menu
Feature activation+0.000
,
Token,
Feature activation+0.000
(
Token (
Feature activation+0.000
Che
TokenChe
Feature activation+0.000
f
Tokenf
Feature activation+1.126
Jorge
Token Jorge
Feature activation+2.407
)
Token)
Feature activation+0.000
Val
Token Val
Feature activation+0.000
le
Tokenle
Feature activation+0.000
jo
Tokenjo
Feature activation+0.000
aims
Token aims
Feature activation+0.000

INTERVAL 1.674 - 2.232
CONTAINS 0.026%

(
Token (
Feature activation+0.000
and
Tokenand
Feature activation+0.000
Marvin
Token Marvin
Feature activation+0.000
Gay
Token Gay
Feature activation+0.000
e
Tokene
Feature activation+0.461
songs
Token songs
Feature activation+2.130
)
Token)
Feature activation+0.000
never
Token never
Feature activation+0.000
go
Token go
Feature activation+0.000
out
Token out
Feature activation+0.000
of
Token of
Feature activation+0.000
idespread
Tokenidespread
Feature activation+0.000
Panic
Token Panic
Feature activation+1.151
lead
Token lead
Feature activation+0.777
guitarist
Token guitarist
Feature activation+2.881
who
Token who
Feature activation+1.701
died
Token died
Feature activation+2.181
of
Token of
Feature activation+0.000
pancreat
Token pancreat
Feature activation+0.000
ic
Tokenic
Feature activation+0.000
cancer
Token cancer
Feature activation+3.738
in
Token in
Feature activation+0.000
(
Token (
Feature activation+0.093
B
TokenB
Feature activation+0.000
-
Token-
Feature activation+0.000
gap
Tokengap
Feature activation+2.319
defensive
Token defensive
Feature activation+0.000
tackle
Token tackle
Feature activation+1.845
),
Token),
Feature activation+0.000
and
Token and
Feature activation+0.000
then
Token then
Feature activation+0.000
run
Token run
Feature activation+0.000
inside
Token inside
Feature activation+0.000
directors
Token directors
Feature activation+0.000
both
Token both
Feature activation+0.000
great
Token great
Feature activation+0.000
(
Token (
Feature activation+0.000
John
TokenJohn
Feature activation+0.000
Woo
Token Woo
Feature activation+1.746
)
Token)
Feature activation+0.000
and
Token and
Feature activation+0.000
awful
Token awful
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.288
needs
Token needs
Feature activation+0.000
(
Token (
Feature activation+0.000
including
Tokenincluding
Feature activation+0.000
Un
Token Un
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
ver
Tokenver
Feature activation+1.847
and
Token and
Feature activation+0.225
D
Token D
Feature activation+0.000
HL
TokenHL
Feature activation+4.270
),
Token),
Feature activation+0.000
and
Token and
Feature activation+0.000

INTERVAL 1.116 - 1.674
CONTAINS 0.055%

Mc
Token Mc
Feature activation+0.000
Queen
TokenQueen
Feature activation+1.951
and
Token and
Feature activation+0.000
D
Token D
Feature activation+0.000
ior
Tokenior
Feature activation+1.281
have
Token have
Feature activation+1.611
been
Token been
Feature activation+0.532
heavily
Token heavily
Feature activation+0.263
featured
Token featured
Feature activation+1.553
)
Token)
Feature activation+0.000
and
Token and
Feature activation+0.000
tem
Token tem
Feature activation+0.000
ps
Tokenps
Feature activation+0.428
afe
Tokenafe
Feature activation+0.073
,
Token,
Feature activation+0.740
grouping
Token grouping
Feature activation+0.838
,
Token,
Feature activation+1.124
aut
Token aut
Feature activation+0.000
ol
Tokenol
Feature activation+0.000
oot
Tokenoot
Feature activation+0.000
,
Token,
Feature activation+1.247
aut
Token aut
Feature activation+0.000
and
Token and
Feature activation+0.000
higher
Token higher
Feature activation+0.000
incomes
Token incomes
Feature activation+0.000
permit
Token permit
Feature activation+0.000
(
Token (
Feature activation+0.000
through
Tokenthrough
Feature activation+1.662
Engel
Token Engel
Feature activation+1.163
curve
Token curve
Feature activation+2.218
consumption
Token consumption
Feature activation+2.329
paths
Token paths
Feature activation+2.107
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
GHC
Token GHC
Feature activation+0.000
(
Token (
Feature activation+0.000
cost
Tokencost
Feature activation+0.000
-
Token-
Feature activation+0.000
cent
Tokencent
Feature activation+0.000
res
Tokenres
Feature activation+1.241
,
Token,
Feature activation+0.000
S
Token S
Feature activation+0.000
CC
TokenCC
Feature activation+0.000
annotations
Token annotations
Feature activation+0.549
,
Token,
Feature activation+0.615
ignorant
Token ignorant
Feature activation+0.436
of
Token of
Feature activation+0.000
your
Token your
Feature activation+0.000
interpret
Token interpret
Feature activation+0.000
ive
Tokenive
Feature activation+0.000
framework
Token framework
Feature activation+1.129
).
Token).
Feature activation+0.000
They
Token They
Feature activation+0.000
are
Token are
Feature activation+0.000
faithful
Token faithful
Feature activation+0.000
and
Token and
Feature activation+0.000

INTERVAL 0.558 - 1.116
CONTAINS 0.111%

e
Tokene
Feature activation+0.000
.
Token.
Feature activation+0.000
g
Tokeng
Feature activation+0.000
.
Token.
Feature activation+0.089
Bristol
Token Bristol
Feature activation+0.000
Ridge
Token Ridge
Feature activation+0.911
)
Token)
Feature activation+0.000
will
Token will
Feature activation+0.000
only
Token only
Feature activation+0.000
be
Token be
Feature activation+0.000
supported
Token supported
Feature activation+0.000
row
Token row
Feature activation+0.000
(
Token (
Feature activation+0.000
and
Tokenand
Feature activation+0.000
received
Token received
Feature activation+0.325
a
Token a
Feature activation+0.000
fourth
Token fourth
Feature activation+0.944
straight
Token straight
Feature activation+0.808
the
Token the
Feature activation+0.331
following
Token following
Feature activation+0.102
year
Token year
Feature activation+1.510
),
Token),
Feature activation+0.000
(
Token (
Feature activation+0.000
along
Tokenalong
Feature activation+0.000
with
Token with
Feature activation+0.002
Z
Token Z
Feature activation+0.000
eb
Tokeneb
Feature activation+0.000
oyd
Tokenoyd
Feature activation+0.937
's
Token's
Feature activation+0.000
first
Token first
Feature activation+0.000
game
Token game
Feature activation+1.209
,
Token,
Feature activation+1.397
the
Token the
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.403
s
Tokens
Feature activation+0.000
ability
Token ability
Feature activation+0.171
to
Token to
Feature activation+0.000
remain
Token remain
Feature activation+0.000
credible
Token credible
Feature activation+1.007
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
member
Token member
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
str
Token str
Feature activation+0.000
(
Token(
Feature activation+0.000
p
Tokenp
Feature activation+0.000
okemon
Tokenokemon
Feature activation+0.405
_
Token_
Feature activation+0.551
number
Tokennumber
Feature activation+0.756
)
Token)
Feature activation+0.000
+
Token +
Feature activation+0.000
".
Token ".
Feature activation+0.000
png
Tokenpng
Feature activation+0.000
")
Token")
Feature activation+0.000

INTERVAL 0.000 - 0.558
CONTAINS 99.781%

than
Token than
Feature activation+0.000
half
Token half
Feature activation+0.000
a
Token a
Feature activation+0.000
century
Token century
Feature activation+0.000
of
Token of
Feature activation+0.000
feminism
Token feminism
Feature activation+0.000
and
Token and
Feature activation+0.000
the
Token the
Feature activation+0.000
welfare
Token welfare
Feature activation+0.000
state
Token state
Feature activation+0.000
,
Token,
Feature activation+0.000
all
Token all
Feature activation+0.000
ure
Tokenure
Feature activation+0.000
of
Token of
Feature activation+0.000
adding
Token adding
Feature activation+0.000
to
Token to
Feature activation+0.000
my
Token my
Feature activation+0.000
small
Token small
Feature activation+0.000
but
Token but
Feature activation+0.000
growing
Token growing
Feature activation+0.000
pint
Token pint
Feature activation+0.000
glass
Token glass
Feature activation+0.000
,
Token,
Feature activation+0.000
would
Token would
Feature activation+0.000
you
Token you
Feature activation+0.000
care
Token care
Feature activation+0.000
where
Token where
Feature activation+0.000
that
Token that
Feature activation+0.000
doctor
Token doctor
Feature activation+0.000
's
Token's
Feature activation+0.000
education
Token education
Feature activation+0.000
came
Token came
Feature activation+0.000
from
Token from
Feature activation+0.000
during
Token during
Feature activation+0.000
the
Token the
Feature activation+0.093
early
Token early
Feature activation+0.000
hours
Token hours
Feature activation+0.000
of
Token of
Feature activation+0.000
April
Token April
Feature activation+0.000
16
Token 16
Feature activation+0.000
,
Token,
Feature activation+0.000
someone
Token someone
Feature activation+0.000
pulled
Token pulled
Feature activation+0.000
a
Token a
Feature activation+0.000
than
Token than
Feature activation+0.000
exploded
Token exploded
Feature activation+0.000
by
Token by
Feature activation+0.000
a
Token a
Feature activation+0.000
mun
Token mun
Feature activation+0.000
ition
Tokenition
Feature activation+0.000
within
Token within
Feature activation+0.000
it
Token it
Feature activation+0.000
.
Token.
Feature activation+0.000
All
Token All
Feature activation+0.000
of
Token of
Feature activation+0.000

BOTTOM ACTIVATIONS
MIN = 0.000

atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
&
Token &
Feature activation+0.000
Development
Token Development
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
&
Token &
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Defence
Token Defence
Feature activation+0.000
Research
Token Research
Feature activation+0.000
rik
Tokenrik
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
rik
Tokenrik
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
said
Token said
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
A
Token A
Feature activation+0.000
atre
Tokenatre
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boss
Token boss
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
K
TokenK
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
defence
Token defence
Feature activation+0.000
ministry
Token ministry
Feature activation+0.000
committee
Token committee
Feature activation+0.000
headed
Token headed
Feature activation+0.000
by
Token by
Feature activation+0.000
V
Token V
Feature activation+0.000
.
Token.
Feature activation+0.000