H5.3 top features

Top feature 0 in H5.3: (feature 7207)

TOP ACTIVATIONS
MAX = 3.466

's
Token's
Feature activation+0.000
what
Token what
Feature activation+0.000
was
Token was
Feature activation+0.000
going
Token going
Feature activation+0.000
on
Token on
Feature activation+0.000
."
Token."
Feature activation+3.466
Ċ
TokenĊ
Feature activation+0.500
Ċ
TokenĊ
Feature activation+0.303
What
TokenWhat
Feature activation+0.000
was
Token was
Feature activation+0.000
going
Token going
Feature activation+0.000
gay
Token gay
Feature activation+0.000
men
Token men
Feature activation+0.000
in
Token in
Feature activation+0.000
this
Token this
Feature activation+0.000
country
Token country
Feature activation+0.000
."
Token."
Feature activation+3.303
Ċ
TokenĊ
Feature activation+1.146
Ċ
TokenĊ
Feature activation+1.129
C
TokenC
Feature activation+0.000
atherine
Tokenatherine
Feature activation+0.000
Frazier
Token Frazier
Feature activation+0.000
basically
Token basically
Feature activation+1.007
any
Token any
Feature activation+0.678
shape
Token shape
Feature activation+0.479
you
Token you
Feature activation+0.000
want
Token want
Feature activation+0.000
."
Token."
Feature activation+3.265
Ċ
TokenĊ
Feature activation+0.669
Ċ
TokenĊ
Feature activation+0.827
That
TokenThat
Feature activation+0.000
's
Token's
Feature activation+0.000
the
Token the
Feature activation+0.000
one
Token one
Feature activation+0.070
day
Token day
Feature activation+0.000
will
Token will
Feature activation+0.000
also
Token also
Feature activation+0.224
die
Token die
Feature activation+0.368
.
Token.
Feature activation+3.186
The
Token The
Feature activation+1.669
new
Token new
Feature activation+0.436
horror
Token horror
Feature activation+0.043
that
Token that
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
it
Token it
Feature activation+0.000
is
Token is
Feature activation+0.257
provided
Token provided
Feature activation+0.345
to
Token to
Feature activation+0.000
all
Token all
Feature activation+0.000
.
Token.
Feature activation+3.147
Ċ
TokenĊ
Feature activation+0.871
Ċ
TokenĊ
Feature activation+1.358
The
TokenThe
Feature activation+0.579
same
Token same
Feature activation+0.000
policy
Token policy
Feature activation+0.203
decides
Token decides
Feature activation+0.000
to
Token to
Feature activation+0.958
remain
Token remain
Feature activation+0.606
in
Token in
Feature activation+0.026
India
Token India
Feature activation+0.000
.
Token.
Feature activation+3.120
Ċ
TokenĊ
Feature activation+0.575
Ċ
TokenĊ
Feature activation+0.960
Dream
TokenDream
Feature activation+0.019
sequences
Token sequences
Feature activation+0.000
[
Token [
Feature activation+0.064
we
Token we
Feature activation+0.000
know
Token know
Feature activation+0.000
what
Token what
Feature activation+0.000
actually
Token actually
Feature activation+0.045
happened
Token happened
Feature activation+0.178
!
Token!
Feature activation+3.044
Ċ
TokenĊ
Feature activation+0.108
Ċ
TokenĊ
Feature activation+0.654
Ex
TokenEx
Feature activation+0.158
isting
Tokenisting
Feature activation+0.000
photographic
Token photographic
Feature activation+0.000
and
Token and
Feature activation+0.000
shots
Token shots
Feature activation+0.000
of
Token of
Feature activation+0.000
rubber
Token rubber
Feature activation+0.000
bullets
Token bullets
Feature activation+0.000
.
Token.
Feature activation+3.009
At
Token At
Feature activation+1.232
least
Token least
Feature activation+0.148
38
Token 38
Feature activation+0.290
people
Token people
Feature activation+0.000
have
Token have
Feature activation+0.000
can
Token can
Feature activation+0.000
you
Token you
Feature activation+0.000
do
Token do
Feature activation+0.000
to
Token to
Feature activation+0.000
help
Token help
Feature activation+0.230
?
Token?
Feature activation+2.961
The
Token The
Feature activation+0.465
United
Token United
Feature activation+0.000
States
Token States
Feature activation+0.000
is
Token is
Feature activation+0.000
among
Token among
Feature activation+0.000
into
Token into
Feature activation+0.000
something
Token something
Feature activation+0.000
it
Token it
Feature activation+0.000
's
Token's
Feature activation+0.000
not
Token not
Feature activation+0.000
."
Token."
Feature activation+2.948
Ċ
TokenĊ
Feature activation+0.333
Ċ
TokenĊ
Feature activation+0.330
Indeed
TokenIndeed
Feature activation+0.000
,
Token,
Feature activation+0.000
race
Token race
Feature activation+0.000
think
Token think
Feature activation+0.202
they
Token they
Feature activation+0.000
kind
Token kind
Feature activation+0.000
of
Token of
Feature activation+0.000
forgot
Token forgot
Feature activation+0.000
."
Token."
Feature activation+2.926
Ċ
TokenĊ
Feature activation+0.199
Ċ
TokenĊ
Feature activation+0.204
K
TokenK
Feature activation+0.089
raft
Tokenraft
Feature activation+0.000
started
Token started
Feature activation+0.000
students
Token students
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
other
Token other
Feature activation+0.000
sex
Token sex
Feature activation+0.015
.
Token.
Feature activation+2.849
âĢ
TokenâĢ
Feature activation+0.526
Ŀ
TokenĿ
Feature activation+0.089
Ċ
TokenĊ
Feature activation+0.115
Ċ
TokenĊ
Feature activation+0.000
Yet
TokenYet
Feature activation+0.000
will
Token will
Feature activation+0.000
find
Token find
Feature activation+0.129
hard
Token hard
Feature activation+0.000
to
Token to
Feature activation+0.000
resist
Token resist
Feature activation+0.321
.
Token.
Feature activation+2.837
Last
Token Last
Feature activation+0.926
year
Token year
Feature activation+0.174
they
Token they
Feature activation+0.294
gave
Token gave
Feature activation+0.067
the
Token the
Feature activation+0.000
in
Token in
Feature activation+0.000
use
Token use
Feature activation+0.000
by
Token by
Feature activation+0.000
foreign
Token foreign
Feature activation+0.000
governments
Token governments
Feature activation+0.306
.[
Token.[
Feature activation+2.827
11
Token11
Feature activation+0.100
]
Token]
Feature activation+1.185
Ċ
TokenĊ
Feature activation+0.269
Ċ
TokenĊ
Feature activation+0.860
R
TokenR
Feature activation+0.000
usually
Token usually
Feature activation+0.191
at
Token at
Feature activation+0.000
their
Token their
Feature activation+0.000
lowest
Token lowest
Feature activation+0.000
level
Token level
Feature activation+0.435
.
Token.
Feature activation+2.796
Ċ
TokenĊ
Feature activation+0.271
Ċ
TokenĊ
Feature activation+1.208
A
TokenA
Feature activation+0.321
Nation
Token Nation
Feature activation+0.000
of
Token of
Feature activation+0.000
get
Token get
Feature activation+0.000
the
Token the
Feature activation+0.000
treatment
Token treatment
Feature activation+0.000
you
Token you
Feature activation+0.000
need
Token need
Feature activation+0.000
.
Token.
Feature activation+2.738
âĢ
TokenâĢ
Feature activation+0.430
Ŀ
TokenĿ
Feature activation+0.144
Ċ
TokenĊ
Feature activation+0.002
Ċ
TokenĊ
Feature activation+0.574
âĢ
TokenâĢ
Feature activation+0.000
may
Token may
Feature activation+0.000
deem
Token deem
Feature activation+0.000
to
Token to
Feature activation+0.000
be
Token be
Feature activation+0.000
appropriate
Token appropriate
Feature activation+0.000
.
Token.
Feature activation+2.717
Ċ
TokenĊ
Feature activation+0.538
Ċ
TokenĊ
Feature activation+0.963
Advertisement
TokenAdvertisement
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
I
Token I
Feature activation+0.003
should
Token should
Feature activation+0.013
specify
Token specify
Feature activation+0.678
some
Token some
Feature activation+0.000
caveats
Token caveats
Feature activation+0.795
:
Token:
Feature activation+2.710
Ċ
TokenĊ
Feature activation+0.471
Ċ
TokenĊ
Feature activation+0.520
Ex
TokenEx
Feature activation+0.000
cessive
Tokencessive
Feature activation+0.000
webs
Token webs
Feature activation+0.000
words
Token words
Feature activation+0.245
to
Token to
Feature activation+0.289
communicate
Token communicate
Feature activation+0.332
his
Token his
Feature activation+0.017
message
Token message
Feature activation+0.039
.
Token.
Feature activation+2.704
He
Token He
Feature activation+0.549
doesn
Token doesn
Feature activation+0.354
't
Token't
Feature activation+0.000
have
Token have
Feature activation+0.000
the
Token the
Feature activation+0.000
the
Token the
Feature activation+0.000
eye
Token eye
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
audience
Token audience
Feature activation+0.000
.
Token.
Feature activation+2.633
âĢ
TokenâĢ
Feature activation+0.461
Ŀ
TokenĿ
Feature activation+0.122
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
M
TokenM
Feature activation+0.000

Top DFA by src position
MAX = 4.109

<|endoftext|>
Token<|endoftext|>
Feature activation+1.766
Top resid features:
,
Token,
Feature activation+4.109
Top resid features:
this
Token this
Feature activation+0.055
Top resid features:
statute
Token statute
Feature activation+0.013
Top resid features:
that
Token that
Feature activation+0.023
Top resid features:
does
Token does
Feature activation+0.006
Top resid features:
something
Token something
Feature activation+0.005
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.454
Top resid features:
,
Token,
Feature activation+3.778
Top resid features:
scientifically
Token scientifically
Feature activation+0.009
Top resid features:
false
Token false
Feature activation+0.013
Top resid features:
,"
Token,"
Feature activation+0.038
Top resid features:
he
Token he
Feature activation+0.034
Top resid features:
said
Token said
Feature activation+0.019
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.857
Top resid features:
,
Token,
Feature activation+3.910
Top resid features:
you
Token you
Feature activation+0.062
Top resid features:
can
Token can
Feature activation+0.069
Top resid features:
prop
Token prop
Feature activation+0.034
Top resid features:
it
Token it
Feature activation+0.002
Top resid features:
up
Token up
Feature activation+0.016
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.866
Top resid features:
,
Token,
Feature activation+3.645
Top resid features:
which
Token which
Feature activation+0.051
Top resid features:
by
Token by
Feature activation+0.015
Top resid features:
implication
Token implication
Feature activation+0.010
Top resid features:
means
Token means
Feature activation-0.001
Top resid features:
that
Token that
Feature activation+0.037
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.836
Top resid features:
,
Token,
Feature activation+3.682
Top resid features:
that
Token that
Feature activation+0.025
Top resid features:
it
Token it
Feature activation+0.001
Top resid features:
is
Token is
Feature activation+0.009
Top resid features:
a
Token a
Feature activation+0.005
Top resid features:
core
Token core
Feature activation+0.009
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.715
Top resid features:
,
Token,
Feature activation+3.535
Top resid features:
who
Token who
Feature activation+0.029
Top resid features:
has
Token has
Feature activation+0.019
Top resid features:
found
Token found
Feature activation+0.030
Top resid features:
not
Token not
Feature activation+0.007
Top resid features:
only
Token only
Feature activation+0.004
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.916
Top resid features:
,
Token,
Feature activation+3.623
Top resid features:
motive
Token motive
Feature activation+0.033
Top resid features:
and
Token and
Feature activation+0.038
Top resid features:
opportunity
Token opportunity
Feature activation+0.017
Top resid features:
can
Token can
Feature activation+0.021
Top resid features:
only
Token only
Feature activation+0.016
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.627
Top resid features:
,
Token,
Feature activation+3.307
Top resid features:
high
Token high
Feature activation-0.001
Top resid features:
-
Token-
Feature activation+0.002
Top resid features:
ranking
Tokenranking
Feature activation+0.005
Top resid features:
retired
Token retired
Feature activation+0.015
Top resid features:
Venezuelan
Token Venezuelan
Feature activation+0.014
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.861
Top resid features:
,
Token,
Feature activation+3.327
Top resid features:
arms
Token arms
Feature activation+0.008
Top resid features:
,
Token,
Feature activation+0.024
Top resid features:
and
Token and
Feature activation+0.022
Top resid features:
even
Token even
Feature activation+0.007
Top resid features:
people
Token people
Feature activation+0.010
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.304
Top resid features:
,
Token,
Feature activation+3.544
Top resid features:
had
Token had
Feature activation+0.023
Top resid features:
dressed
Token dressed
Feature activation+0.009
Top resid features:
up
Token up
Feature activation+0.000
Top resid features:
as
Token as
Feature activation+0.003
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.775
Top resid features:
,
Token,
Feature activation+3.404
Top resid features:
but
Token but
Feature activation+0.091
Top resid features:
that
Token that
Feature activation+0.010
Top resid features:
's
Token's
Feature activation+0.018
Top resid features:
the
Token the
Feature activation+0.005
Top resid features:
story
Token story
Feature activation+0.007
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.682
Top resid features:
,
Token,
Feature activation+3.263
Top resid features:
interpre
Token interpre
Feature activation+0.024
Top resid features:
ts
Tokents
Feature activation+0.110
Top resid features:
the
Token the
Feature activation+0.021
Top resid features:
statute
Token statute
Feature activation+0.021
Top resid features:
to
Token to
Feature activation+0.027
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.899
Top resid features:
,
Token,
Feature activation+3.234
Top resid features:
especially
Token especially
Feature activation+0.013
Top resid features:
the
Token the
Feature activation+0.002
Top resid features:
older
Token older
Feature activation+0.016
Top resid features:
members
Token members
Feature activation+0.053
Top resid features:
,
Token,
Feature activation+0.028
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.966
Top resid features:
,
Token,
Feature activation+3.393
Top resid features:
as
Token as
Feature activation+0.024
Top resid features:
they
Token they
Feature activation+0.022
Top resid features:
were
Token were
Feature activation+0.030
Top resid features:
told
Token told
Feature activation+0.009
Top resid features:
that
Token that
Feature activation+0.031
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.738
Top resid features:
,
Token,
Feature activation+3.186
Top resid features:
a
Token a
Feature activation+0.054
Top resid features:
time
Token time
Feature activation+0.021
Top resid features:
when
Token when
Feature activation+0.021
Top resid features:
the
Token the
Feature activation+0.012
Top resid features:
political
Token political
Feature activation+0.017
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.711
Top resid features:
,
Token,
Feature activation+3.540
Top resid features:
in
Token in
Feature activation+0.036
Top resid features:
case
Token case
Feature activation+0.035
Top resid features:
you
Token you
Feature activation+0.008
Top resid features:
are
Token are
Feature activation+0.018
Top resid features:
going
Token going
Feature activation+0.012
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.750
Top resid features:
,
Token,
Feature activation+3.090
Top resid features:
he
Token he
Feature activation+0.064
Top resid features:
may
Token may
Feature activation+0.081
Top resid features:
by
Token by
Feature activation+0.017
Top resid features:
proclamation
Token proclamation
Feature activation+0.026
Top resid features:
,
Token,
Feature activation+0.027
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.066
Top resid features:
,
Token,
Feature activation+3.219
Top resid features:
I
Token I
Feature activation+0.006
Top resid features:
should
Token should
Feature activation+0.010
Top resid features:
specify
Token specify
Feature activation+0.024
Top resid features:
some
Token some
Feature activation-0.026
Top resid features:
caveats
Token caveats
Feature activation+0.068
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.687
Top resid features:
,
Token,
Feature activation+2.906
Top resid features:
less
Token less
Feature activation+0.314
Top resid features:
skilled
Token skilled
Feature activation+0.025
Top resid features:
lower
Token lower
Feature activation+0.014
Top resid features:
-
Token-
Feature activation+0.002
Top resid features:
status
Tokenstatus
Feature activation+0.003
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.695
Top resid features:
,
Token,
Feature activation+2.986
Top resid features:
opening
Token opening
Feature activation+0.013
Top resid features:
it
Token it
Feature activation+0.012
Top resid features:
in
Token in
Feature activation+0.014
Top resid features:
the
Token the
Feature activation+0.003
Top resid features:
concentration
Token concentration
Feature activation+0.016
Top resid features:

Decoder Weights Distribution

Head 0: 0.02

Head 1: 0.03

Head 2: 0.07

Head 3: 0.53

Head 4: 0.04

Head 5: 0.05

Head 6: 0.02

Head 7: 0.02

Head 8: 0.05

Head 9: 0.07

Head 10: 0.06

Head 11: 0.04

Positive logits

ui2.30

david1.95

irez1.85

"]=>1.84

gency1.81

1.79

taboola1.77

artment1.74

irlf1.74

']1.71

warr1.70

cv1.70

ead1.68

olean1.68

captcha1.66

inese1.66

[+1.63

rimp1.63

yond1.63

onent1.61

Negative logits

----------1.63

—"-1.61

Hath-1.56

Cassidy-1.56

Finish-1.51

vous-1.49

Standing-1.48

!?"-1.45

Papa-1.45

!'"-1.44

beware-1.41

Cla-1.41

Rule-1.40

Inquiry-1.40

Breach-1.38

Madness-1.38

brut-1.37

McCull-1.36

impunity-1.36

Sonny-1.36

INTERVAL 3.119 - 3.466
CONTAINS 0.001%

basically
Token basically
Feature activation+1.007
any
Token any
Feature activation+0.678
shape
Token shape
Feature activation+0.479
you
Token you
Feature activation+0.000
want
Token want
Feature activation+0.000
."
Token."
Feature activation+3.265
Ċ
TokenĊ
Feature activation+0.669
Ċ
TokenĊ
Feature activation+0.827
That
TokenThat
Feature activation+0.000
's
Token's
Feature activation+0.000
the
Token the
Feature activation+0.000
's
Token's
Feature activation+0.000
what
Token what
Feature activation+0.000
was
Token was
Feature activation+0.000
going
Token going
Feature activation+0.000
on
Token on
Feature activation+0.000
."
Token."
Feature activation+3.466
Ċ
TokenĊ
Feature activation+0.500
Ċ
TokenĊ
Feature activation+0.303
What
TokenWhat
Feature activation+0.000
was
Token was
Feature activation+0.000
going
Token going
Feature activation+0.000
one
Token one
Feature activation+0.070
day
Token day
Feature activation+0.000
will
Token will
Feature activation+0.000
also
Token also
Feature activation+0.224
die
Token die
Feature activation+0.368
.
Token.
Feature activation+3.186
The
Token The
Feature activation+1.669
new
Token new
Feature activation+0.436
horror
Token horror
Feature activation+0.043
that
Token that
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
it
Token it
Feature activation+0.000
is
Token is
Feature activation+0.257
provided
Token provided
Feature activation+0.345
to
Token to
Feature activation+0.000
all
Token all
Feature activation+0.000
.
Token.
Feature activation+3.147
Ċ
TokenĊ
Feature activation+0.871
Ċ
TokenĊ
Feature activation+1.358
The
TokenThe
Feature activation+0.579
same
Token same
Feature activation+0.000
policy
Token policy
Feature activation+0.203
gay
Token gay
Feature activation+0.000
men
Token men
Feature activation+0.000
in
Token in
Feature activation+0.000
this
Token this
Feature activation+0.000
country
Token country
Feature activation+0.000
."
Token."
Feature activation+3.303
Ċ
TokenĊ
Feature activation+1.146
Ċ
TokenĊ
Feature activation+1.129
C
TokenC
Feature activation+0.000
atherine
Tokenatherine
Feature activation+0.000
Frazier
Token Frazier
Feature activation+0.000

INTERVAL 2.773 - 3.119
CONTAINS 0.001%

will
Token will
Feature activation+0.000
find
Token find
Feature activation+0.129
hard
Token hard
Feature activation+0.000
to
Token to
Feature activation+0.000
resist
Token resist
Feature activation+0.321
.
Token.
Feature activation+2.837
Last
Token Last
Feature activation+0.926
year
Token year
Feature activation+0.174
they
Token they
Feature activation+0.294
gave
Token gave
Feature activation+0.067
the
Token the
Feature activation+0.000
students
Token students
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
other
Token other
Feature activation+0.000
sex
Token sex
Feature activation+0.015
.
Token.
Feature activation+2.849
âĢ
TokenâĢ
Feature activation+0.526
Ŀ
TokenĿ
Feature activation+0.089
Ċ
TokenĊ
Feature activation+0.115
Ċ
TokenĊ
Feature activation+0.000
Yet
TokenYet
Feature activation+0.000
usually
Token usually
Feature activation+0.191
at
Token at
Feature activation+0.000
their
Token their
Feature activation+0.000
lowest
Token lowest
Feature activation+0.000
level
Token level
Feature activation+0.435
.
Token.
Feature activation+2.796
Ċ
TokenĊ
Feature activation+0.271
Ċ
TokenĊ
Feature activation+1.208
A
TokenA
Feature activation+0.321
Nation
Token Nation
Feature activation+0.000
of
Token of
Feature activation+0.000
we
Token we
Feature activation+0.000
know
Token know
Feature activation+0.000
what
Token what
Feature activation+0.000
actually
Token actually
Feature activation+0.045
happened
Token happened
Feature activation+0.178
!
Token!
Feature activation+3.044
Ċ
TokenĊ
Feature activation+0.108
Ċ
TokenĊ
Feature activation+0.654
Ex
TokenEx
Feature activation+0.158
isting
Tokenisting
Feature activation+0.000
photographic
Token photographic
Feature activation+0.000
in
Token in
Feature activation+0.000
use
Token use
Feature activation+0.000
by
Token by
Feature activation+0.000
foreign
Token foreign
Feature activation+0.000
governments
Token governments
Feature activation+0.306
.[
Token.[
Feature activation+2.827
11
Token11
Feature activation+0.100
]
Token]
Feature activation+1.185
Ċ
TokenĊ
Feature activation+0.269
Ċ
TokenĊ
Feature activation+0.860
R
TokenR
Feature activation+0.000

INTERVAL 2.426 - 2.773
CONTAINS 0.002%

of
Token of
Feature activation+0.147
the
Token the
Feature activation+0.180
Church
Token Church
Feature activation+0.073
of
Token of
Feature activation+0.000
England
Token England
Feature activation+0.241
.[
Token.[
Feature activation+2.427
2
Token2
Feature activation+0.319
]
Token]
Feature activation+1.582
Ċ
TokenĊ
Feature activation+0.348
Ċ
TokenĊ
Feature activation+0.922
There
TokenThere
Feature activation+0.000
players
Token players
Feature activation+0.000
and
Token and
Feature activation+0.000
the
Token the
Feature activation+0.000
TO
Token TO
Feature activation+0.000
s
Tokens
Feature activation+0.000
.
Token.
Feature activation+2.500
This
TokenThis
Feature activation+0.429
is
Token is
Feature activation+0.270
the
Token the
Feature activation+0.111
perfect
Token perfect
Feature activation+0.000
time
Token time
Feature activation+0.000
ass
Token ass
Feature activation+0.165
u
Tokenu
Feature activation+0.000
age
Tokenage
Feature activation+0.000
liberal
Token liberal
Feature activation+0.000
guilt
Token guilt
Feature activation+0.267
.
Token.
Feature activation+2.452
Thus
Token Thus
Feature activation+0.504
,
Token,
Feature activation+0.228
former
Token former
Feature activation+0.000
Hillary
Token Hillary
Feature activation+0.000
Clinton
Token Clinton
Feature activation+0.000
words
Token words
Feature activation+0.245
to
Token to
Feature activation+0.289
communicate
Token communicate
Feature activation+0.332
his
Token his
Feature activation+0.017
message
Token message
Feature activation+0.039
.
Token.
Feature activation+2.704
He
Token He
Feature activation+0.549
doesn
Token doesn
Feature activation+0.354
't
Token't
Feature activation+0.000
have
Token have
Feature activation+0.000
the
Token the
Feature activation+0.000
may
Token may
Feature activation+0.000
deem
Token deem
Feature activation+0.000
to
Token to
Feature activation+0.000
be
Token be
Feature activation+0.000
appropriate
Token appropriate
Feature activation+0.000
.
Token.
Feature activation+2.717
Ċ
TokenĊ
Feature activation+0.538
Ċ
TokenĊ
Feature activation+0.963
Advertisement
TokenAdvertisement
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000

INTERVAL 2.079 - 2.426
CONTAINS 0.004%

you
Token you
Feature activation+2.902
to
Token to
Feature activation+2.273
vote
Token vote
Feature activation+0.912
for
Token for
Feature activation+0.841
him
Token him
Feature activation+1.248
.
Token.
Feature activation+2.351
âĢ
TokenâĢ
Feature activation+0.307
Ŀ
TokenĿ
Feature activation+0.211
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.251
In
TokenIn
Feature activation+0.000
Indo
Token Indo
Feature activation+0.000
-
Token-
Feature activation+0.000
Crime
TokenCrime
Feature activation+0.000
an
Tokenan
Feature activation+0.000
Partnership
Token Partnership
Feature activation+0.000
.
Token.
Feature activation+2.386
Ċ
TokenĊ
Feature activation+0.986
Ċ
TokenĊ
Feature activation+1.562
The
TokenThe
Feature activation+0.164
symbolic
Token symbolic
Feature activation+0.000
show
Token show
Feature activation+0.000
kissed
Token kissed
Feature activation+0.521
her
Token her
Feature activation+0.012
on
Token on
Feature activation+0.315
the
Token the
Feature activation+0.102
cheek
Token cheek
Feature activation+0.258
.
Token.
Feature activation+2.315
Ċ
TokenĊ
Feature activation+0.086
Ċ
TokenĊ
Feature activation+0.813
A
TokenA
Feature activation+0.544
do
Token do
Feature activation+0.000
orman
Tokenorman
Feature activation+0.004
North
Token North
Feature activation+0.000
Dakota
Token Dakota
Feature activation+0.000
and
Token and
Feature activation+0.000
eastern
Token eastern
Feature activation+0.000
Montana
Token Montana
Feature activation+0.000
.
Token.
Feature activation+2.234
Ċ
TokenĊ
Feature activation+1.020
Ċ
TokenĊ
Feature activation+1.876
Un
TokenUn
Feature activation+0.000
employment
Tokenemployment
Feature activation+0.000
has
Token has
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
,
Token,
Feature activation+0.000
you
Token you
Feature activation+0.000
'll
Token'll
Feature activation+0.255
need
Token need
Feature activation+0.817
to
Token to
Feature activation+2.143
pick
Token pick
Feature activation+1.512
up
Token up
Feature activation+0.444
some
Token some
Feature activation+0.667
NFC
Token NFC
Feature activation+0.432
tags
Token tags
Feature activation+0.722

INTERVAL 1.733 - 2.079
CONTAINS 0.008%

to
Token to
Feature activation+0.000
enhance
Token enhance
Feature activation+1.039
the
Token the
Feature activation+0.938
public
Token public
Feature activation+1.009
's
Token's
Feature activation+1.361
understanding
Token understanding
Feature activation+1.830
,
Token,
Feature activation+1.858
your
Token your
Feature activation+0.900
editorial
Token editorial
Feature activation+0.265
might
Token might
Feature activation+0.576
have
Token have
Feature activation+0.144
soils
Token soils
Feature activation+0.585
,
Token,
Feature activation+0.899
among
Token among
Feature activation+0.000
other
Token other
Feature activation+0.395
sources
Token sources
Feature activation+0.000
.
Token.
Feature activation+1.800
The
Token The
Feature activation+1.481
study
Token study
Feature activation+1.666
says
Token says
Feature activation+0.726
scientists
Token scientists
Feature activation+0.020
have
Token have
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
,
Token,
Feature activation+0.000
Lady
Token Lady
Feature activation+0.300
of
Token of
Feature activation+0.356
Del
Token Del
Feature activation+0.771
ight
Tokenight
Feature activation+1.898
,
Token,
Feature activation+0.883
in
Token in
Feature activation+0.263
thy
Token thy
Feature activation+0.595
name
Token name
Feature activation+0.530
and
Token and
Feature activation+0.000
and
Token and
Feature activation+0.000
a
Token a
Feature activation+0.000
strong
Token strong
Feature activation+0.000
romantic
Token romantic
Feature activation+0.000
storyline
Token storyline
Feature activation+0.000
.
Token.
Feature activation+2.054
As
Token As
Feature activation+0.929
for
Token for
Feature activation+0.000
the
Token the
Feature activation+0.000
US
Token US
Feature activation+0.000
,
Token,
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
,
Token,
Feature activation+0.000
this
Token this
Feature activation+0.000
sort
Token sort
Feature activation+1.051
of
Token of
Feature activation+1.335
thing
Token thing
Feature activation+1.827
has
Token has
Feature activation+0.959
,
Token,
Feature activation+0.177
in
Token in
Feature activation+0.000
fact
Token fact
Feature activation+0.000
,
Token,
Feature activation+0.000

INTERVAL 1.386 - 1.733
CONTAINS 0.018%

ľ
Tokenľ
Feature activation+0.000
Miss
TokenMiss
Feature activation+0.000
Car
Token Car
Feature activation+0.000
us
Tokenus
Feature activation+0.000
o
Tokeno
Feature activation+0.000
).
Token).
Feature activation+1.412
We
Token We
Feature activation+0.661
talk
Token talk
Feature activation+0.000
Bond
Token Bond
Feature activation+0.000
and
Token and
Feature activation+0.000
Live
Token Live
Feature activation+0.000
'
Token '
Feature activation+0.000
What
TokenWhat
Feature activation+0.074
instrument
Token instrument
Feature activation+0.000
is
Token is
Feature activation+0.000
that
Token that
Feature activation+0.000
?'
Token?'
Feature activation+1.724
I
Token I
Feature activation+0.266
loved
Token loved
Feature activation+0.295
jazz
Token jazz
Feature activation+0.000
fusion
Token fusion
Feature activation+0.000
and
Token and
Feature activation+0.000
the
Token the
Feature activation+0.000
political
Token political
Feature activation+0.733
influence
Token influence
Feature activation+0.584
of
Token of
Feature activation+0.400
big
Token big
Feature activation+1.200
business
Token business
Feature activation+1.511
,
Token,
Feature activation+1.243
and
Token and
Feature activation+0.211
declining
Token declining
Feature activation+0.165
living
Token living
Feature activation+0.000
standards
Token standards
Feature activation+0.256
land
Token land
Feature activation+0.344
-
Token-
Feature activation+0.000
use
Tokenuse
Feature activation+0.000
scenarios
Token scenarios
Feature activation+0.493
and
Token and
Feature activation+0.181
found
Token found
Feature activation+1.486
that
Token that
Feature activation+1.641
increasing
Token increasing
Feature activation+0.000
yields
Token yields
Feature activation+0.000
could
Token could
Feature activation+0.000
reduce
Token reduce
Feature activation+0.000
Thunder
Token Thunder
Feature activation+0.379
roster
Token roster
Feature activation+0.009
to
Token to
Feature activation+0.000
a
Token a
Feature activation+0.000
championship
Token championship
Feature activation+0.000
.
Token.
Feature activation+1.458
Brooks
Token Brooks
Feature activation+1.079
is
Token is
Feature activation+0.422
well
Token well
Feature activation+0.000
liked
Token liked
Feature activation+0.000
within
Token within
Feature activation+0.000

INTERVAL 1.040 - 1.386
CONTAINS 0.028%

un
Token un
Feature activation+1.105
pe
Token pe
Feature activation+0.310
u
Tokenu
Feature activation+0.000
part
Token part
Feature activation+0.000
out
Tokenout
Feature activation+0.000
sur
Token sur
Feature activation+1.379
le
Token le
Feature activation+0.727
terrain
Token terrain
Feature activation+0.000
.
Token.
Feature activation+1.168
Jet
Token Jet
Feature activation+0.000
ez
Tokenez
Feature activation+0.000
of
Token of
Feature activation+0.000
any
Token any
Feature activation+0.000
future
Token future
Feature activation+0.000
council
Token council
Feature activation+0.000
meeting
Token meeting
Feature activation+0.000
.
Token.
Feature activation+1.145
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
S
TokenS
Feature activation+0.000
olving
Tokenolving
Feature activation+0.000
the
Token the
Feature activation+0.000
Student
Token Student
Feature activation+0.000
encourage
Token encourage
Feature activation+0.490
you
Token you
Feature activation+2.902
to
Token to
Feature activation+2.273
vote
Token vote
Feature activation+0.912
for
Token for
Feature activation+0.841
him
Token him
Feature activation+1.248
.
Token.
Feature activation+2.351
âĢ
TokenâĢ
Feature activation+0.307
Ŀ
TokenĿ
Feature activation+0.211
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.251
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
,
Token,
Feature activation+0.000
one
Token one
Feature activation+0.000
would
Token would
Feature activation+1.106
have
Token have
Feature activation+0.732
to
Token to
Feature activation+1.253
wonder
Token wonder
Feature activation+0.837
if
Token if
Feature activation+0.905
the
Token the
Feature activation+0.755
system
Token system
Feature activation+0.946
had
Token had
Feature activation+0.027
National
Token National
Feature activation+0.398
Marine
Token Marine
Feature activation+0.156
Fisheries
Token Fisheries
Feature activation+0.369
Service
Token Service
Feature activation+0.795
,
Token,
Feature activation+1.015
called
Token called
Feature activation+1.216
the
Token the
Feature activation+0.910
legal
Token legal
Feature activation+0.211
challenge
Token challenge
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.641
ľ
Tokenľ
Feature activation+0.000

INTERVAL 0.693 - 1.040
CONTAINS 0.062%

the
Token the
Feature activation+0.296
National
Token National
Feature activation+0.398
Marine
Token Marine
Feature activation+0.156
Fisheries
Token Fisheries
Feature activation+0.369
Service
Token Service
Feature activation+0.795
,
Token,
Feature activation+1.015
called
Token called
Feature activation+1.216
the
Token the
Feature activation+0.910
legal
Token legal
Feature activation+0.211
challenge
Token challenge
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.641
rers
Tokenrers
Feature activation+0.000
,
Token,
Feature activation+0.000
etc
Token etc
Feature activation+0.000
.)
Token.)
Feature activation+0.000
."
Token."
Feature activation+0.841
Ċ
TokenĊ
Feature activation+0.787
Ċ
TokenĊ
Feature activation+1.022
Major
TokenMajor
Feature activation+0.000
Viol
Token Viol
Feature activation+0.000
ations
Tokenations
Feature activation+0.000
vs
Token vs
Feature activation+0.000
his
Token his
Feature activation+0.322
heart
Token heart
Feature activation+0.514
on
Token on
Feature activation+1.302
his
Token his
Feature activation+0.300
sleeve
Token sleeve
Feature activation+0.834
,
Token,
Feature activation+0.852
and
Token and
Feature activation+0.645
that
Token that
Feature activation+0.691
desire
Token desire
Feature activation+0.137
helped
Token helped
Feature activation+0.000
the
Token the
Feature activation+0.000
,
Token,
Feature activation+0.000
bringing
Token bringing
Feature activation+0.362
an
Token an
Feature activation+0.600
end
Token end
Feature activation+1.943
to
Token to
Feature activation+1.475
more
Token more
Feature activation+0.915
than
Token than
Feature activation+0.614
a
Token a
Feature activation+0.645
decade
Token decade
Feature activation+0.479
of
Token of
Feature activation+0.000
Won
Token Won
Feature activation+0.826
-
Token-
Feature activation+0.000
year
Tokenyear
Feature activation+0.000
budget
Token budget
Feature activation+0.000
.
Token.
Feature activation+0.261
Ċ
TokenĊ
Feature activation+0.391
Ċ
TokenĊ
Feature activation+0.797
"
Token"
Feature activation+0.000
I
TokenI
Feature activation+0.000
don
Token don
Feature activation+0.000
't
Token't
Feature activation+0.000
have
Token have
Feature activation+0.000

INTERVAL 0.347 - 0.693
CONTAINS 0.116%

,
Token,
Feature activation+0.000
as
Token as
Feature activation+0.000
DHS
Token DHS
Feature activation+0.295
officials
Token officials
Feature activation+1.073
,
Token,
Feature activation+0.155
including
Token including
Feature activation+0.382
former
Token former
Feature activation+0.235
Secretary
Token Secretary
Feature activation+0.406
Je
Token Je
Feature activation+0.018
h
Tokenh
Feature activation+0.000
Johnson
Token Johnson
Feature activation+0.787
any
Token any
Feature activation+0.678
shape
Token shape
Feature activation+0.479
you
Token you
Feature activation+0.000
want
Token want
Feature activation+0.000
."
Token."
Feature activation+3.265
Ċ
TokenĊ
Feature activation+0.669
Ċ
TokenĊ
Feature activation+0.827
That
TokenThat
Feature activation+0.000
's
Token's
Feature activation+0.000
the
Token the
Feature activation+0.000
T
Token T
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
,
Token,
Feature activation+0.000
but
Token but
Feature activation+0.000
that
Token that
Feature activation+1.107
's
Token's
Feature activation+0.790
the
Token the
Feature activation+0.505
story
Token story
Feature activation+0.704
there
Token there
Feature activation+0.000
.
Token.
Feature activation+0.617
I
Token I
Feature activation+0.286
think
Token think
Feature activation+0.202
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
security
Token security
Feature activation+1.078
consultant
Token consultant
Feature activation+0.477
and
Token and
Feature activation+0.606
one
Token one
Feature activation+0.495
of
Token of
Feature activation+0.220
P
Token P
Feature activation+0.662
ORT
TokenORT
Feature activation+0.241
AL
TokenAL
Feature activation+0.000
's
Token's
Feature activation+0.443
fing
Token fing
Feature activation+0.000
ern
Tokenern
Feature activation+0.060
ail
Tokenail
Feature activation+0.796
cl
Token cl
Feature activation+0.639
ipp
Tokenipp
Feature activation+0.031
ings
Tokenings
Feature activation+0.622
.
Token.
Feature activation+2.466
Not
Token Not
Feature activation+0.601
only
Token only
Feature activation+0.538
are
Token are
Feature activation+0.000
a
Token a
Feature activation+0.000

INTERVAL 0.000 - 0.347
CONTAINS 99.761%

might
Token might
Feature activation+0.000
start
Token start
Feature activation+0.000
a
Token a
Feature activation+0.000
fire
Token fire
Feature activation+0.000
.
Token.
Feature activation+1.130
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.244
Ky
TokenKy
Feature activation+0.000
ew
Tokenew
Feature activation+0.000
aly
Tokenaly
Feature activation+0.000
anga
Tokenanga
Feature activation+0.000
be
Token be
Feature activation+0.000
sold
Token sold
Feature activation+0.000
to
Token to
Feature activation+0.000
raise
Token raise
Feature activation+0.000
money
Token money
Feature activation+0.000
for
Token for
Feature activation+0.000
other
Token other
Feature activation+0.000
productions
Token productions
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
on
Token on
Feature activation+0.000
the
Token the
Feature activation+0.000
run
Token run
Feature activation+0.000
!
Token!
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Master
TokenMaster
Feature activation+0.000
All
Token All
Feature activation+0.000
Modes
Token Modes
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Minister
Token Minister
Feature activation+0.000
Theresa
Token Theresa
Feature activation+0.000
May
Token May
Feature activation+0.000
attends
Token attends
Feature activation+0.000
the
Token the
Feature activation+0.000
European
Token European
Feature activation+0.000
Union
Token Union
Feature activation+0.000
leaders
Token leaders
Feature activation+0.000
summit
Token summit
Feature activation+0.000
in
Token in
Feature activation+0.000
Val
Token Val
Feature activation+0.000
Palestine
Token Palestine
Feature activation+0.000
?
Token?
Feature activation+0.000
Share
Token Share
Feature activation+0.000
your
Token your
Feature activation+0.000
thoughts
Token thoughts
Feature activation+0.000
below
Token below
Feature activation+0.000
!
Token!
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Are
TokenAre
Feature activation+0.000
you
Token you
Feature activation+0.000
trying
Token trying
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 H5.3: (feature 961)

TOP ACTIVATIONS
MAX = 4.886

is
Token is
Feature activation+0.304
the
Token the
Feature activation+1.106
setting
Token setting
Feature activation+0.775
different
Token different
Feature activation+3.474
and
Token and
Feature activation+2.867
interesting
Token interesting
Feature activation+4.886
,
Token,
Feature activation+2.905
but
Token but
Feature activation+0.000
the
Token the
Feature activation+0.000
game
Token game
Feature activation+0.000
forces
Token forces
Feature activation+0.000
that
Token that
Feature activation+0.000
is
Token is
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
more
Token more
Feature activation+2.782
powerful
Token powerful
Feature activation+4.580
,
Token,
Feature activation+2.014
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
has
Token has
Feature activation+0.000
a
Token a
Feature activation+0.000
beer
Token beer
Feature activation+0.000
:
Token:
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
it
Token it
Feature activation+4.393
possible
Token possible
Feature activation+3.275
that
Token that
Feature activation+2.641
RP
Token RP
Feature activation+0.000
As
TokenAs
Feature activation+0.000
could
Token could
Feature activation+1.363
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
an
Token an
Feature activation+0.626
impe
Token impe
Feature activation+0.689
achable
Tokenachable
Feature activation+0.707
offense
Token offense
Feature activation+3.996
.
Token.
Feature activation+0.000
It
Token It
Feature activation+0.000
's
Token's
Feature activation+0.000
also
Token also
Feature activation+0.000
illegal
Token illegal
Feature activation+0.000
education
Token education
Feature activation+0.000
,
Token,
Feature activation+0.000
is
Token is
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
ineffective
Token ineffective
Feature activation+3.818
at
Token at
Feature activation+1.741
preventing
Token preventing
Feature activation+2.727
teenage
Token teenage
Feature activation+0.908
pregnancy
Token pregnancy
Feature activation+2.365
,
Token,
Feature activation+2.309
shaming
Token shaming
Feature activation+0.023
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+1.456
accepted
Token accepted
Feature activation+3.790
,
Token,
Feature activation+1.241
but
Token but
Feature activation+0.000
welcomed
Token welcomed
Feature activation+0.758
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
inspect
Token inspect
Feature activation+1.317
the
Token the
Feature activation+1.624
photographs
Token photographs
Feature activation+1.330
and
Token and
Feature activation+3.358
videos
Token videos
Feature activation+2.001
,
Token,
Feature activation+3.790
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+0.246
the
Token the
Feature activation+1.664
material
Token material
Feature activation+1.252
soft
Token soft
Feature activation+3.771
,
Token,
Feature activation+2.206
comfortable
Token comfortable
Feature activation+2.165
,
Token,
Feature activation+0.887
and
Token and
Feature activation+2.768
warm
Token warm
Feature activation+0.961
the
Token the
Feature activation+0.375
next
Token next
Feature activation+0.368
six
Token six
Feature activation+0.536
months
Token months
Feature activation+1.678
drastically
Token drastically
Feature activation+3.639
impact
Token impact
Feature activation+3.708
Toronto
Token Toronto
Feature activation+1.278
âĢ
TokenâĢ
Feature activation+1.768
Ļ
TokenĻ
Feature activation+1.720
s
Tokens
Feature activation+1.219
team
Token team
Feature activation+0.249
over
Token over
Feature activation+0.636
the
Token the
Feature activation+0.375
next
Token next
Feature activation+0.368
six
Token six
Feature activation+0.536
months
Token months
Feature activation+1.678
drastically
Token drastically
Feature activation+3.639
impact
Token impact
Feature activation+3.708
Toronto
Token Toronto
Feature activation+1.278
âĢ
TokenâĢ
Feature activation+1.768
Ļ
TokenĻ
Feature activation+1.720
s
Tokens
Feature activation+1.219
Not
Token Not
Feature activation+0.000
only
Token only
Feature activation+0.000
is
Token is
Feature activation+0.304
the
Token the
Feature activation+1.106
setting
Token setting
Feature activation+0.775
different
Token different
Feature activation+3.474
and
Token and
Feature activation+2.867
interesting
Token interesting
Feature activation+4.886
,
Token,
Feature activation+2.905
but
Token but
Feature activation+0.000
the
Token the
Feature activation+0.000
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+3.444
for
Token for
Feature activation+0.837
women
Token women
Feature activation+1.479
âĢ
TokenâĢ
Feature activation+0.018
Ļ
TokenĻ
Feature activation+0.388
s
Tokens
Feature activation+0.800
for
Token for
Feature activation+2.151
12
Token 12
Feature activation+1.447
to
Token to
Feature activation+0.356
15
Token 15
Feature activation+0.286
hours
Token hours
Feature activation+2.122
,
Token,
Feature activation+3.415
that
Token that
Feature activation+2.146
temperature
Token temperature
Feature activation+0.000
has
Token has
Feature activation+0.005
to
Token to
Feature activation+0.000
rise
Token rise
Feature activation+0.000
entitled
Token entitled
Feature activation+1.531
to
Token to
Feature activation+0.679
inspect
Token inspect
Feature activation+1.317
the
Token the
Feature activation+1.624
photographs
Token photographs
Feature activation+1.330
and
Token and
Feature activation+3.358
videos
Token videos
Feature activation+2.001
,
Token,
Feature activation+3.790
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
to
Token to
Feature activation+0.000
only
Token only
Feature activation+0.000
by
Token by
Feature activation+0.000
single
Token single
Feature activation+0.491
-
Token-
Feature activation+0.000
letter
Tokenletter
Feature activation+0.000
changes
Token changes
Feature activation+3.328
in
Token in
Feature activation+1.736
their
Token their
Feature activation+1.932
genetic
Token genetic
Feature activation+1.122
codes
Token codes
Feature activation+1.702
,
Token,
Feature activation+2.752
:
Token:
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
it
Token it
Feature activation+4.393
possible
Token possible
Feature activation+3.275
that
Token that
Feature activation+2.641
RP
Token RP
Feature activation+0.000
As
TokenAs
Feature activation+0.000
could
Token could
Feature activation+1.363
become
Token become
Feature activation+1.001
wars
Token wars
Feature activation+0.828
in
Token in
Feature activation+0.413
Afghanistan
Token Afghanistan
Feature activation+1.258
and
Token and
Feature activation+2.368
Iraq
Token Iraq
Feature activation+0.452
,
Token,
Feature activation+3.113
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
á
Tokená
Feature activation+0.000
seriously
Token seriously
Feature activation+0.000
and
Token and
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
is
Token is
Feature activation+1.291
he
Token he
Feature activation+3.061
a
Token a
Feature activation+2.182
good
Token good
Feature activation+2.698
friend
Token friend
Feature activation+1.943
of
Token of
Feature activation+1.824
Gi
Token Gi
Feature activation+0.629
/
Token/
Feature activation+0.000
or
Tokenor
Feature activation+0.000
over
Token over
Feature activation+0.798
in
Tokenin
Feature activation+0.000
fl
Tokenfl
Feature activation+0.000
ated
Tokenated
Feature activation+3.013
,
Token,
Feature activation+1.965
some
Token some
Feature activation+0.000
have
Token have
Feature activation+0.000
become
Token become
Feature activation+0.000
downright
Token downright
Feature activation+0.000
only
Token only
Feature activation+0.000
was
Token was
Feature activation+0.379
the
Token the
Feature activation+0.726
Range
Token Range
Feature activation+0.000
Rover
Token Rover
Feature activation+0.884
a
Token a
Feature activation+2.978
new
Token new
Feature activation+1.183
vehicle
Token vehicle
Feature activation+1.327
and
Token and
Feature activation+2.932
brand
Token brand
Feature activation+0.937
in
Token in
Feature activation+1.698

Top DFA by src position
MAX = 5.381

back
Token back
Feature activation-0.001
Top resid features:
to
Token to
Feature activation-0.002
Top resid features:
it
Token it
Feature activation+0.002
Top resid features:
.
Token.
Feature activation-0.014
Top resid features:
Not
Token Not
Feature activation+0.022
Top resid features:
only
Token only
Feature activation+4.786
Top resid features:
is
Token is
Feature activation+0.713
Top resid features:
the
Token the
Feature activation+0.157
Top resid features:
setting
Token setting
Feature activation+0.071
Top resid features:
different
Token different
Feature activation+0.058
Top resid features:
and
Token and
Feature activation+0.094
Top resid features:
called
Token called
Feature activation-0.002
Top resid features:
Swift
Token Swift
Feature activation+0.051
Top resid features:
that
Token that
Feature activation+0.026
Top resid features:
is
Token is
Feature activation+0.021
Top resid features:
not
Token not
Feature activation-0.028
Top resid features:
only
Token only
Feature activation+4.914
Top resid features:
more
Token more
Feature activation+0.239
Top resid features:
powerful
Token powerful
Feature activation+0.100
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
but
Token but
Feature activation+0.000
Top resid features:
also
Token also
Feature activation+0.000
Top resid features:
on
Token on
Feature activation-0.000
Top resid features:
this
Token this
Feature activation+0.014
Top resid features:
beer
Token beer
Feature activation-0.036
Top resid features:
:
Token:
Feature activation-0.072
Top resid features:
not
Token not
Feature activation+0.021
Top resid features:
only
Token only
Feature activation+5.381
Top resid features:
is
Token is
Feature activation+0.373
Top resid features:
it
Token it
Feature activation-0.045
Top resid features:
possible
Token possible
Feature activation+0.000
Top resid features:
that
Token that
Feature activation+0.000
Top resid features:
RP
Token RP
Feature activation+0.000
Top resid features:
lying
Token lying
Feature activation-0.017
Top resid features:
to
Token to
Feature activation-0.011
Top resid features:
Congress
Token Congress
Feature activation+0.020
Top resid features:
is
Token is
Feature activation-0.045
Top resid features:
not
Token not
Feature activation+0.038
Top resid features:
only
Token only
Feature activation+4.609
Top resid features:
an
Token an
Feature activation+0.218
Top resid features:
impe
Token impe
Feature activation+0.108
Top resid features:
achable
Tokenachable
Feature activation+0.049
Top resid features:
offense
Token offense
Feature activation+0.041
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
sex
Token sex
Feature activation+0.017
Top resid features:
education
Token education
Feature activation+0.026
Top resid features:
,
Token,
Feature activation+0.006
Top resid features:
is
Token is
Feature activation+0.068
Top resid features:
not
Token not
Feature activation+0.047
Top resid features:
only
Token only
Feature activation+4.479
Top resid features:
ineffective
Token ineffective
Feature activation+0.116
Top resid features:
at
Token at
Feature activation+0.000
Top resid features:
preventing
Token preventing
Feature activation+0.000
Top resid features:
teenage
Token teenage
Feature activation+0.000
Top resid features:
pregnancy
Token pregnancy
Feature activation+0.000
Top resid features:
kind
Token kind
Feature activation-0.003
Top resid features:
of
Token of
Feature activation-0.003
Top resid features:
shaming
Token shaming
Feature activation+0.030
Top resid features:
to
Token to
Feature activation-0.007
Top resid features:
not
Token not
Feature activation+0.063
Top resid features:
only
Token only
Feature activation+3.824
Top resid features:
be
Token be
Feature activation+0.481
Top resid features:
accepted
Token accepted
Feature activation+0.094
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:
"
Token"
Feature activation+0.019
Top resid features:
The
TokenThe
Feature activation+0.010
Top resid features:
defense
Token defense
Feature activation+0.006
Top resid features:
is
Token is
Feature activation+0.010
Top resid features:
not
Token not
Feature activation+0.007
Top resid features:
only
Token only
Feature activation+4.043
Top resid features:
entitled
Token entitled
Feature activation+0.199
Top resid features:
to
Token to
Feature activation+0.077
Top resid features:
inspect
Token inspect
Feature activation+0.307
Top resid features:
the
Token the
Feature activation+0.076
Top resid features:
photographs
Token photographs
Feature activation+0.053
Top resid features:
for
Token for
Feature activation+0.003
Top resid features:
my
Token my
Feature activation+0.003
Top resid features:
brother
Token brother
Feature activation+0.001
Top resid features:
.
Token.
Feature activation-0.006
Top resid features:
Not
Token Not
Feature activation+0.006
Top resid features:
only
Token only
Feature activation+2.617
Top resid features:
is
Token is
Feature activation+1.353
Top resid features:
the
Token the
Feature activation+0.368
Top resid features:
material
Token material
Feature activation+0.204
Top resid features:
soft
Token soft
Feature activation+0.129
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
with
Token with
Feature activation+0.010
Top resid features:
.
Token.
Feature activation+0.028
Top resid features:
Ċ
TokenĊ
Feature activation-0.008
Top resid features:
Ċ
TokenĊ
Feature activation+0.001
Top resid features:
Not
TokenNot
Feature activation-0.012
Top resid features:
only
Token only
Feature activation+2.776
Top resid features:
will
Token will
Feature activation+1.365
Top resid features:
the
Token the
Feature activation+0.304
Top resid features:
growth
Token growth
Feature activation+0.069
Top resid features:
of
Token of
Feature activation+0.014
Top resid features:
these
Token these
Feature activation+0.045
Top resid features:
with
Token with
Feature activation-0.002
Top resid features:
.
Token.
Feature activation+0.017
Top resid features:
Ċ
TokenĊ
Feature activation-0.011
Top resid features:
Ċ
TokenĊ
Feature activation-0.003
Top resid features:
Not
TokenNot
Feature activation-0.002
Top resid features:
only
Token only
Feature activation+2.968
Top resid features:
will
Token will
Feature activation+1.246
Top resid features:
the
Token the
Feature activation+0.308
Top resid features:
growth
Token growth
Feature activation+0.090
Top resid features:
of
Token of
Feature activation+0.022
Top resid features:
these
Token these
Feature activation+0.022
Top resid features:
back
Token back
Feature activation-0.008
Top resid features:
to
Token to
Feature activation-0.002
Top resid features:
it
Token it
Feature activation+0.002
Top resid features:
.
Token.
Feature activation+0.011
Top resid features:
Not
Token Not
Feature activation+0.005
Top resid features:
only
Token only
Feature activation+2.842
Top resid features:
is
Token is
Feature activation+0.901
Top resid features:
the
Token the
Feature activation+0.316
Top resid features:
setting
Token setting
Feature activation+0.170
Top resid features:
different
Token different
Feature activation+0.152
Top resid features:
and
Token and
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.013
Top resid features:
abortions
Token abortions
Feature activation+0.026
Top resid features:
might
Token might
Feature activation-0.032
Top resid features:
be
Token be
Feature activation+0.018
Top resid features:
not
Token not
Feature activation+0.037
Top resid features:
only
Token only
Feature activation+4.377
Top resid features:
good
Token good
Feature activation+0.043
Top resid features:
for
Token for
Feature activation+0.000
Top resid features:
women
Token women
Feature activation+0.000
Top resid features:
âĢ
TokenâĢ
Feature activation+0.000
Top resid features:
Ļ
TokenĻ
Feature activation+0.000
Top resid features:
.
Token.
Feature activation-0.013
Top resid features:
That
Token That
Feature activation+0.032
Top resid features:
means
Token means
Feature activation+0.084
Top resid features:
that
Token that
Feature activation+0.061
Top resid features:
not
Token not
Feature activation+0.009
Top resid features:
only
Token only
Feature activation+2.173
Top resid features:
does
Token does
Feature activation+1.860
Top resid features:
the
Token the
Feature activation+0.096
Top resid features:
brisk
Token brisk
Feature activation+0.013
Top resid features:
et
Tokenet
Feature activation+0.007
Top resid features:
have
Token have
Feature activation+0.065
Top resid features:
"
Token"
Feature activation+0.004
Top resid features:
The
TokenThe
Feature activation+0.005
Top resid features:
defense
Token defense
Feature activation+0.008
Top resid features:
is
Token is
Feature activation+0.007
Top resid features:
not
Token not
Feature activation+0.002
Top resid features:
only
Token only
Feature activation+3.601
Top resid features:
entitled
Token entitled
Feature activation+0.188
Top resid features:
to
Token to
Feature activation+0.087
Top resid features:
inspect
Token inspect
Feature activation+0.356
Top resid features:
the
Token the
Feature activation+0.143
Top resid features:
photographs
Token photographs
Feature activation+0.216
Top resid features:
these
Token these
Feature activation+0.015
Top resid features:
organisms
Token organisms
Feature activation+0.024
Top resid features:
is
Token is
Feature activation+0.017
Top resid features:
driven
Token driven
Feature activation+0.002
Top resid features:
not
Token not
Feature activation+0.024
Top resid features:
only
Token only
Feature activation+3.496
Top resid features:
by
Token by
Feature activation+0.491
Top resid features:
single
Token single
Feature activation+0.154
Top resid features:
-
Token-
Feature activation+0.007
Top resid features:
letter
Tokenletter
Feature activation+0.028
Top resid features:
changes
Token changes
Feature activation+0.089
Top resid features:
on
Token on
Feature activation+0.001
Top resid features:
this
Token this
Feature activation+0.014
Top resid features:
beer
Token beer
Feature activation+0.022
Top resid features:
:
Token:
Feature activation-0.095
Top resid features:
not
Token not
Feature activation+0.023
Top resid features:
only
Token only
Feature activation+3.466
Top resid features:
is
Token is
Feature activation+0.729
Top resid features:
it
Token it
Feature activation+0.156
Top resid features:
possible
Token possible
Feature activation+0.081
Top resid features:
that
Token that
Feature activation+0.000
Top resid features:
RP
Token RP
Feature activation+0.000
Top resid features:
ann
Tokenann
Feature activation+0.001
Top resid features:
elled
Tokenelled
Feature activation+0.133
Top resid features:
the
Token the
Feature activation-0.001
Top resid features:
opposition
Token opposition
Feature activation-0.008
Top resid features:
not
Token not
Feature activation-0.037
Top resid features:
only
Token only
Feature activation+3.403
Top resid features:
to
Token to
Feature activation+0.263
Top resid features:
the
Token the
Feature activation+0.046
Top resid features:
wars
Token wars
Feature activation+0.068
Top resid features:
in
Token in
Feature activation+0.143
Top resid features:
Afghanistan
Token Afghanistan
Feature activation-0.003
Top resid features:
very
Token very
Feature activation-0.014
Top resid features:
very
Token very
Feature activation-0.008
Top resid features:
seriously
Token seriously
Feature activation+0.019
Top resid features:
and
Token and
Feature activation+0.022
Top resid features:
not
Token not
Feature activation+0.021
Top resid features:
only
Token only
Feature activation+3.531
Top resid features:
is
Token is
Feature activation+0.721
Top resid features:
he
Token he
Feature activation-0.008
Top resid features:
a
Token a
Feature activation+0.000
Top resid features:
good
Token good
Feature activation+0.000
Top resid features:
friend
Token friend
Feature activation+0.000
Top resid features:
ipp
Tokenipp
Feature activation-0.004
Top resid features:
ings
Tokenings
Feature activation-0.001
Top resid features:
.
Token.
Feature activation-0.013
Top resid features:
Not
Token Not
Feature activation+0.034
Top resid features:
only
Token only
Feature activation+1.581
Top resid features:
are
Token are
Feature activation+1.976
Top resid features:
a
Token a
Feature activation+0.171
Top resid features:
growing
Token growing
Feature activation+0.033
Top resid features:
number
Token number
Feature activation-0.044
Top resid features:
of
Token of
Feature activation+0.036
Top resid features:
published
Token published
Feature activation+0.027
Top resid features:
Rover
Token Rover
Feature activation+0.009
Top resid features:
.
Token.
Feature activation-0.008
Top resid features:
Ċ
TokenĊ
Feature activation-0.019
Top resid features:
Ċ
TokenĊ
Feature activation-0.015
Top resid features:
Not
TokenNot
Feature activation-0.039
Top resid features:
only
Token only
Feature activation+2.806
Top resid features:
was
Token was
Feature activation+0.907
Top resid features:
the
Token the
Feature activation+0.144
Top resid features:
Range
Token Range
Feature activation+0.049
Top resid features:
Rover
Token Rover
Feature activation+0.058
Top resid features:
a
Token a
Feature activation+0.086
Top resid features:

Decoder Weights Distribution

Head 0: 0.02

Head 1: 0.01

Head 2: 0.09

Head 3: 0.52

Head 4: 0.09

Head 5: 0.02

Head 6: 0.03

Head 7: 0.04

Head 8: 0.06

Head 9: 0.03

Head 10: 0.04

Head 11: 0.04

Positive logits

undergrad1.58

(),1.57

disparate1.53

but1.50

binaries1.46

fro1.43

+,1.43

accol1.42

?,1.40

bounty1.39

peg1.37

overdue1.37

})1.36

?",1.36

ostensibly1.35

quot1.33

bene1.33

locally1.33

AFTA1.31

fascists1.31

Negative logits

etheless-2.46

elin-1.93

ktop-1.87

laughs-1.83

laugh-1.74

UTERS-1.70

.''.-1.68

interstitial-1.67

cellaneous-1.66

rint-1.64

ctuary-1.64

tics-1.63

��-1.63

skirts-1.61

thia-1.60

ocobo-1.60

attRot-1.59

ntil-1.58

tical-1.58

ebin-1.58

INTERVAL 4.398 - 4.886
CONTAINS 0.000%

is
Token is
Feature activation+0.304
the
Token the
Feature activation+1.106
setting
Token setting
Feature activation+0.775
different
Token different
Feature activation+3.474
and
Token and
Feature activation+2.867
interesting
Token interesting
Feature activation+4.886
,
Token,
Feature activation+2.905
but
Token but
Feature activation+0.000
the
Token the
Feature activation+0.000
game
Token game
Feature activation+0.000
forces
Token forces
Feature activation+0.000
that
Token that
Feature activation+0.000
is
Token is
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
more
Token more
Feature activation+2.782
powerful
Token powerful
Feature activation+4.580
,
Token,
Feature activation+2.014
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
has
Token has
Feature activation+0.000
a
Token a
Feature activation+0.000

INTERVAL 3.909 - 4.398
CONTAINS 0.000%

not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
an
Token an
Feature activation+0.626
impe
Token impe
Feature activation+0.689
achable
Tokenachable
Feature activation+0.707
offense
Token offense
Feature activation+3.996
.
Token.
Feature activation+0.000
It
Token It
Feature activation+0.000
's
Token's
Feature activation+0.000
also
Token also
Feature activation+0.000
illegal
Token illegal
Feature activation+0.000
beer
Token beer
Feature activation+0.000
:
Token:
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
it
Token it
Feature activation+4.393
possible
Token possible
Feature activation+3.275
that
Token that
Feature activation+2.641
RP
Token RP
Feature activation+0.000
As
TokenAs
Feature activation+0.000
could
Token could
Feature activation+1.363

INTERVAL 3.420 - 3.909
CONTAINS 0.001%

inspect
Token inspect
Feature activation+1.317
the
Token the
Feature activation+1.624
photographs
Token photographs
Feature activation+1.330
and
Token and
Feature activation+3.358
videos
Token videos
Feature activation+2.001
,
Token,
Feature activation+3.790
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
education
Token education
Feature activation+0.000
,
Token,
Feature activation+0.000
is
Token is
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
ineffective
Token ineffective
Feature activation+3.818
at
Token at
Feature activation+1.741
preventing
Token preventing
Feature activation+2.727
teenage
Token teenage
Feature activation+0.908
pregnancy
Token pregnancy
Feature activation+2.365
,
Token,
Feature activation+2.309
the
Token the
Feature activation+0.375
next
Token next
Feature activation+0.368
six
Token six
Feature activation+0.536
months
Token months
Feature activation+1.678
drastically
Token drastically
Feature activation+3.639
impact
Token impact
Feature activation+3.708
Toronto
Token Toronto
Feature activation+1.278
âĢ
TokenâĢ
Feature activation+1.768
Ļ
TokenĻ
Feature activation+1.720
s
Tokens
Feature activation+1.219
team
Token team
Feature activation+0.249
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+3.444
for
Token for
Feature activation+0.837
women
Token women
Feature activation+1.479
âĢ
TokenâĢ
Feature activation+0.018
Ļ
TokenĻ
Feature activation+0.388
s
Tokens
Feature activation+0.800
shaming
Token shaming
Feature activation+0.023
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+1.456
accepted
Token accepted
Feature activation+3.790
,
Token,
Feature activation+1.241
but
Token but
Feature activation+0.000
welcomed
Token welcomed
Feature activation+0.758
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000

INTERVAL 2.932 - 3.420
CONTAINS 0.001%

for
Token for
Feature activation+2.151
12
Token 12
Feature activation+1.447
to
Token to
Feature activation+0.356
15
Token 15
Feature activation+0.286
hours
Token hours
Feature activation+2.122
,
Token,
Feature activation+3.415
that
Token that
Feature activation+2.146
temperature
Token temperature
Feature activation+0.000
has
Token has
Feature activation+0.005
to
Token to
Feature activation+0.000
rise
Token rise
Feature activation+0.000
treatment
Token treatment
Feature activation+1.018
for
Token for
Feature activation+0.755
anxiety
Token anxiety
Feature activation+0.541
and
Token and
Feature activation+2.230
depression
Token depression
Feature activation+1.593
,
Token,
Feature activation+2.966
it
Token it
Feature activation+0.000
is
Token is
Feature activation+0.000
also
Token also
Feature activation+0.000
the
Token the
Feature activation+0.000
best
Token best
Feature activation+0.000
only
Token only
Feature activation+0.000
are
Token are
Feature activation+0.286
city
Token city
Feature activation+0.000
police
Token police
Feature activation+0.300
more
Token more
Feature activation+2.027
involved
Token involved
Feature activation+2.960
,
Token,
Feature activation+1.738
she
Token she
Feature activation+0.000
observes
Token observes
Feature activation+0.000
,
Token,
Feature activation+0.000
but
Token but
Feature activation+0.000
Range
Token Range
Feature activation+0.000
Rover
Token Rover
Feature activation+0.884
a
Token a
Feature activation+2.978
new
Token new
Feature activation+1.183
vehicle
Token vehicle
Feature activation+1.327
and
Token and
Feature activation+2.932
brand
Token brand
Feature activation+0.937
in
Token in
Feature activation+1.698
the
Token the
Feature activation+0.631
U
Token U
Feature activation+0.375
.
Token.
Feature activation+0.370
a
Token a
Feature activation+0.150
vast
Token vast
Feature activation+0.503
region
Token region
Feature activation+0.475
impossible
Token impossible
Feature activation+1.734
,
Token,
Feature activation+1.236
and
Token and
Feature activation+2.966
disrupt
Token disrupt
Feature activation+2.350
the
Token the
Feature activation+0.742
climate
Token climate
Feature activation+0.278
,
Token,
Feature activation+0.373
but
Token but
Feature activation+0.000

INTERVAL 2.443 - 2.932
CONTAINS 0.003%

ful
Tokenful
Feature activation+0.000
.
Token.
Feature activation+0.000
Not
Token Not
Feature activation+0.000
only
Token only
Feature activation+0.000
is
Token is
Feature activation+0.297
that
Token that
Feature activation+2.466
negativity
Token negativity
Feature activation+0.890
unw
Token unw
Feature activation+1.689
arranted
Tokenarranted
Feature activation+2.420
,
Token,
Feature activation+2.621
the
Token the
Feature activation+0.765
the
Token the
Feature activation+1.106
setting
Token setting
Feature activation+0.775
different
Token different
Feature activation+3.474
and
Token and
Feature activation+2.867
interesting
Token interesting
Feature activation+4.886
,
Token,
Feature activation+2.905
but
Token but
Feature activation+0.000
the
Token the
Feature activation+0.000
game
Token game
Feature activation+0.000
forces
Token forces
Feature activation+0.000
players
Token players
Feature activation+0.000
is
Token is
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
ineffective
Token ineffective
Feature activation+3.818
at
Token at
Feature activation+1.741
preventing
Token preventing
Feature activation+2.727
teenage
Token teenage
Feature activation+0.908
pregnancy
Token pregnancy
Feature activation+2.365
,
Token,
Feature activation+2.309
it
Token it
Feature activation+0.000
contributes
Token contributes
Feature activation+0.311
that
Token that
Feature activation+0.164
an
Token an
Feature activation+0.671
attack
Token attack
Feature activation+0.751
is
Token is
Feature activation+1.717
highly
Token highly
Feature activation+1.320
likely
Token likely
Feature activation+2.872
,
Token,
Feature activation+1.930
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
material
Token material
Feature activation+1.252
soft
Token soft
Feature activation+3.771
,
Token,
Feature activation+2.206
comfortable
Token comfortable
Feature activation+2.165
,
Token,
Feature activation+0.887
and
Token and
Feature activation+2.768
warm
Token warm
Feature activation+0.961
,
Token,
Feature activation+1.909
but
Token but
Feature activation+0.000
it
Token it
Feature activation+0.000
seems
Token seems
Feature activation+0.000

INTERVAL 1.954 - 2.443
CONTAINS 0.005%

Not
Token Not
Feature activation+0.000
only
Token only
Feature activation+0.000
are
Token are
Feature activation+0.286
city
Token city
Feature activation+0.000
police
Token police
Feature activation+0.300
more
Token more
Feature activation+2.027
involved
Token involved
Feature activation+2.960
,
Token,
Feature activation+1.738
she
Token she
Feature activation+0.000
observes
Token observes
Feature activation+0.000
,
Token,
Feature activation+0.000
from
Token from
Feature activation+0.887
team
Token team
Feature activation+0.664
Santos
Token Santos
Feature activation+0.185
-
Token-
Feature activation+0.000
Glass
TokenGlass
Feature activation+0.000
,
Token,
Feature activation+2.035
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
him
Token him
Feature activation+1.408
up
Token up
Feature activation+1.150
on
Token on
Feature activation+0.253
federal
Token federal
Feature activation+0.893
charges
Token charges
Feature activation+1.665
,
Token,
Feature activation+2.242
they
Token they
Feature activation+0.183
granted
Token granted
Feature activation+0.573
immunity
Token immunity
Feature activation+0.619
to
Token to
Feature activation+0.273
his
Token his
Feature activation+0.000
vast
Token vast
Feature activation+0.503
region
Token region
Feature activation+0.475
impossible
Token impossible
Feature activation+1.734
,
Token,
Feature activation+1.236
and
Token and
Feature activation+2.966
disrupt
Token disrupt
Feature activation+2.350
the
Token the
Feature activation+0.742
climate
Token climate
Feature activation+0.278
,
Token,
Feature activation+0.373
but
Token but
Feature activation+0.000
it
Token it
Feature activation+0.000
been
Token been
Feature activation+0.000
not
Token not
Feature activation+0.000
just
Token just
Feature activation+0.000
a
Token a
Feature activation+0.000
very
Token very
Feature activation+0.346
good
Token good
Feature activation+2.379
hitter
Token hitter
Feature activation+1.528
,
Token,
Feature activation+0.209
but
Token but
Feature activation+0.000
one
Token one
Feature activation+0.000
of
Token of
Feature activation+0.000

INTERVAL 1.466 - 1.954
CONTAINS 0.009%

the
Token the
Feature activation+0.751
U
Token U
Feature activation+0.391
H
TokenH
Feature activation+0.000
system
Token system
Feature activation+1.669
âĢĶ
Token âĢĶ
Feature activation+1.869
a
Token a
Feature activation+1.847
system
Token system
Feature activation+0.743
that
Token that
Feature activation+1.065
includes
Token includes
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
feds
Token feds
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
promised
Token promised
Feature activation+0.834
not
Token not
Feature activation+1.695
to
Token to
Feature activation+1.606
ring
Token ring
Feature activation+0.589
him
Token him
Feature activation+1.408
up
Token up
Feature activation+1.150
on
Token on
Feature activation+0.253
federal
Token federal
Feature activation+0.893
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
peaked
Token peaked
Feature activation+0.000
early
Token early
Feature activation+1.808
in
Token in
Feature activation+1.939
the
Token the
Feature activation+1.859
Obama
Token Obama
Feature activation+0.537
years
Token years
Feature activation+1.178
,
Token,
Feature activation+1.314
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
does
Token does
Feature activation+0.284
the
Token the
Feature activation+0.845
brisk
Token brisk
Feature activation+0.615
et
Tokenet
Feature activation+1.652
have
Token have
Feature activation+1.619
to
Token to
Feature activation+1.898
be
Token be
Feature activation+2.104
cooked
Token cooked
Feature activation+1.864
at
Token at
Feature activation+1.024
low
Token low
Feature activation+0.989
temperatures
Token temperatures
Feature activation+1.893
by
Token by
Feature activation+0.000
single
Token single
Feature activation+0.491
-
Token-
Feature activation+0.000
letter
Tokenletter
Feature activation+0.000
changes
Token changes
Feature activation+3.328
in
Token in
Feature activation+1.736
their
Token their
Feature activation+1.932
genetic
Token genetic
Feature activation+1.122
codes
Token codes
Feature activation+1.702
,
Token,
Feature activation+2.752
but
Token but
Feature activation+0.000

INTERVAL 0.977 - 1.466
CONTAINS 0.017%

not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
the
Token the
Feature activation+0.000
AV
Token AV
Feature activation+0.188
research
Token research
Feature activation+1.190
community
Token community
Feature activation+1.455
,
Token,
Feature activation+0.490
but
Token but
Feature activation+0.000
anyone
Token anyone
Feature activation+0.000
who
Token who
Feature activation+0.000
cares
Token cares
Feature activation+0.000
only
Token only
Feature activation+0.000
is
Token is
Feature activation+1.035
this
Token this
Feature activation+2.614
book
Token book
Feature activation+1.007
pure
Token pure
Feature activation+2.569
science
Token science
Feature activation+1.166
fiction
Token fiction
Feature activation+1.413
,
Token,
Feature activation+1.812
it
Token it
Feature activation+0.000
is
Token is
Feature activation+0.000
also
Token also
Feature activation+0.000
only
Token only
Feature activation+0.000
a
Token a
Feature activation+0.000
way
Token way
Feature activation+0.161
to
Token to
Feature activation+1.074
get
Token get
Feature activation+0.424
more
Token more
Feature activation+0.985
readers
Token readers
Feature activation+1.660
but
Token but
Feature activation+0.000
in
Token in
Feature activation+0.000
general
Token general
Feature activation+0.000
,
Token,
Feature activation+0.000
for
Token for
Feature activation+1.201
the
Token the
Feature activation+0.926
weight
Token weight
Feature activation+0.476
-
Token-
Feature activation+0.122
loss
Tokenloss
Feature activation+0.305
company
Token company
Feature activation+1.262
,
Token,
Feature activation+2.166
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
is
Token is
Feature activation+1.035
this
Token this
Feature activation+2.614
book
Token book
Feature activation+1.007
pure
Token pure
Feature activation+2.569
science
Token science
Feature activation+1.166
fiction
Token fiction
Feature activation+1.413
,
Token,
Feature activation+1.812
it
Token it
Feature activation+0.000
is
Token is
Feature activation+0.000
also
Token also
Feature activation+0.000
a
Token a
Feature activation+0.000

INTERVAL 0.489 - 0.977
CONTAINS 0.025%

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+0.000
talent
Token talent
Feature activation+0.613
in
Token in
Feature activation+1.220
South
Token South
Feature activation+0.000
East
Token East
Feature activation+0.000
Asia
Token Asia
Feature activation+1.304
myself
Token myself
Feature activation+2.010
that
Token that
Feature activation+0.000
is
Token is
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
funny
Token funny
Feature activation+2.602
,
Token,
Feature activation+0.829
but
Token but
Feature activation+0.000
allows
Token allows
Feature activation+0.000
your
Token your
Feature activation+0.000
friends
Token friends
Feature activation+0.000
to
Token to
Feature activation+0.000
8
Token 8
Feature activation+0.000
seed
Token seed
Feature activation+0.000
hasn
Token hasn
Feature activation+0.000
't
Token't
Feature activation+0.000
just
Token just
Feature activation+0.000
tied
Token tied
Feature activation+0.711
their
Token their
Feature activation+1.202
series
Token series
Feature activation+1.116
with
Token with
Feature activation+1.353
the
Token the
Feature activation+1.148
Atlanta
Token Atlanta
Feature activation+0.000
.
Token.
Feature activation+0.000
You
Token You
Feature activation+0.000
are
Token are
Feature activation+0.000
not
Token not
Feature activation+0.000
just
Token just
Feature activation+0.000
helping
Token helping
Feature activation+0.816
EN
Token EN
Feature activation+0.213
KK
TokenKK
Feature activation+0.000
O
TokenO
Feature activation+0.072
but
Token but
Feature activation+0.000
the
Token the
Feature activation+0.000
federal
Token federal
Feature activation+0.893
charges
Token charges
Feature activation+1.665
,
Token,
Feature activation+2.242
they
Token they
Feature activation+0.183
granted
Token granted
Feature activation+0.573
immunity
Token immunity
Feature activation+0.619
to
Token to
Feature activation+0.273
his
Token his
Feature activation+0.000
various
Token various
Feature activation+0.000
co
Token co
Feature activation+0.000
-
Token-
Feature activation+0.000

INTERVAL 0.000 - 0.489
CONTAINS 99.939%

the
Token the
Feature activation+0.000
material
Token material
Feature activation+0.000
,"
Token,"
Feature activation+0.000
said
Token said
Feature activation+0.000
Rep
Token Rep
Feature activation+0.000
.
Token.
Feature activation+0.000
Jer
Token Jer
Feature activation+0.000
ro
Tokenro
Feature activation+0.000
ld
Tokenld
Feature activation+0.000
Nad
Token Nad
Feature activation+0.000
ler
Tokenler
Feature activation+0.000
,
Token,
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
squeeze
Token squeeze
Feature activation+0.000
in
Token in
Feature activation+0.000
your
Token your
Feature activation+0.000
30
Token 30
Feature activation+0.000
-
Token-
Feature activation+0.000
minute
Tokenminute
Feature activation+0.000
lunch
Token lunch
Feature activation+0.000
can
Token can
Feature activation+0.000
get
Token get
Feature activation+0.000
stored
Token stored
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
database
Token database
Feature activation+0.000
.
Token.
Feature activation+0.000
At
Token At
Feature activation+0.000
this
Token this
Feature activation+0.000
point
Token point
Feature activation+0.000
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
arguing
Token arguing
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
crew
Token crew
Feature activation+0.000
played
Token played
Feature activation+0.000
games
Token games
Feature activation+0.000
designed
Token designed
Feature activation+0.000
to
Token to
Feature activation+0.000
a
Token a
Feature activation+0.000
new
Token new
Feature activation+0.000
genus
Token genus
Feature activation+0.000
and
Token and
Feature activation+0.000
species
Token species
Feature activation+0.000
of
Token of
Feature activation+0.000
frog
Token frog
Feature activation+0.000
fish
Tokenfish
Feature activation+0.000
,
Token,
Feature activation+0.000
which
Token which
Feature activation+0.000
are
Token are
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 H5.3: (feature 2938)

TOP ACTIVATIONS
MAX = 3.356

Te
Token Te
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.208
Ļ
TokenĻ
Feature activation+0.414
o
Tokeno
Feature activation+0.478
,
Token,
Feature activation+2.882
in
Token in
Feature activation+3.356
action
Token action
Feature activation+1.242
against
Token against
Feature activation+0.798
Bund
Token Bund
Feature activation+0.000
ee
Tokenee
Feature activation+0.000
Aki
Token Aki
Feature activation+0.000
the
Token the
Feature activation+0.561
best
Token best
Feature activation+0.719
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
world
Token world
Feature activation+1.060
.
Token.
Feature activation+3.193
Photograph
Token Photograph
Feature activation+0.997
:
Token:
Feature activation+0.000
Jonathan
Token Jonathan
Feature activation+0.000
Franklin
Token Franklin
Feature activation+0.000
for
Token for
Feature activation+0.000
Ben
Token Ben
Feature activation+0.674
Te
Token Te
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.208
Ļ
TokenĻ
Feature activation+0.414
o
Tokeno
Feature activation+0.478
,
Token,
Feature activation+2.882
in
Token in
Feature activation+3.356
action
Token action
Feature activation+1.242
against
Token against
Feature activation+0.798
Bund
Token Bund
Feature activation+0.000
ee
Tokenee
Feature activation+0.000
ila
Tokenila
Feature activation+0.146
-
Token-
Feature activation+0.000
Mat
TokenMat
Feature activation+0.000
as
Tokenas
Feature activation+0.000
,
Token,
Feature activation+2.093
The
Token The
Feature activation+2.719
Ill
Token Ill
Feature activation+0.000
og
Tokenog
Feature activation+0.000
ic
Tokenic
Feature activation+0.141
of
Token of
Feature activation+0.000
Kas
Token Kas
Feature activation+0.000
Facebook
TokenFacebook
Feature activation+0.000
Twitter
Token Twitter
Feature activation+0.000
Pinterest
Token Pinterest
Feature activation+0.000
Calais
Token Calais
Feature activation+0.613
migrants
Token migrants
Feature activation+1.044
:
Token:
Feature activation+2.702
âĢ
Token âĢ
Feature activation+0.288
ĺ
Tokenĺ
Feature activation+0.009
Get
TokenGet
Feature activation+0.000
to
Token to
Feature activation+0.000
England
Token England
Feature activation+0.000
Aki
Token Aki
Feature activation+0.000
of
Token of
Feature activation+0.230
Conn
Token Conn
Feature activation+0.000
acht
Tokenacht
Feature activation+0.013
,
Token,
Feature activation+1.621
joins
Token joins
Feature activation+2.633
Worcester
Token Worcester
Feature activation+0.560
this
Token this
Feature activation+0.718
summer
Token summer
Feature activation+1.210
and
Token and
Feature activation+0.806
is
Token is
Feature activation+1.274
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
media
Tokenmedia
Feature activation+0.000
_
Token_
Feature activation+0.000
camera
Tokencamera
Feature activation+0.000
Patrick
Token Patrick
Feature activation+2.467
Danger
Token Danger
Feature activation+1.651
field
Tokenfield
Feature activation+2.050
pre
Token pre
Feature activation+1.590
pping
Tokenpping
Feature activation+1.556
the
Token the
Feature activation+1.266
Ċ
TokenĊ
Feature activation+0.000
Facebook
TokenFacebook
Feature activation+0.000
Twitter
Token Twitter
Feature activation+0.000
Pinterest
Token Pinterest
Feature activation+0.000
Pop
Token Pop
Feature activation+1.140
:
Token:
Feature activation+2.458
30
Token 30
Feature activation+0.195
,
Token,
Feature activation+0.000
000
Token000
Feature activation+0.511
.
Token.
Feature activation+0.810
Tr
Token Tr
Feature activation+0.130
Le
Token Le
Feature activation+1.339
in
Tokenin
Feature activation+0.013
ster
Tokenster
Feature activation+1.771
âĢ
TokenâĢ
Feature activation+1.642
Ļ
TokenĻ
Feature activation+1.491
s
Tokens
Feature activation+2.333
Ben
Token Ben
Feature activation+0.674
Te
Token Te
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.208
Ļ
TokenĻ
Feature activation+0.414
o
Tokeno
Feature activation+0.478
is
Token is
Feature activation+1.274
eligible
Token eligible
Feature activation+0.000
for
Token for
Feature activation+0.000
England
Token England
Feature activation+0.000
selection
Token selection
Feature activation+0.000
.
Token.
Feature activation+2.231
Photograph
Token Photograph
Feature activation+0.000
:
Token:
Feature activation+0.000
IN
Token IN
Feature activation+0.000
PH
TokenPH
Feature activation+0.000
O
TokenO
Feature activation+0.000
V
Token V
Feature activation+0.000
ila
Tokenila
Feature activation+0.146
-
Token-
Feature activation+0.000
Mat
TokenMat
Feature activation+0.000
as
Tokenas
Feature activation+0.000
,
Token,
Feature activation+2.093
The
Token The
Feature activation+2.719
Ill
Token Ill
Feature activation+0.000
og
Tokenog
Feature activation+0.000
ic
Tokenic
Feature activation+0.141
of
Token of
Feature activation+0.000
og
Tokenog
Feature activation+0.000
ic
Tokenic
Feature activation+0.141
of
Token of
Feature activation+0.000
Kas
Token Kas
Feature activation+0.000
sel
Tokensel
Feature activation+1.585
.
Token.
Feature activation+1.958
Photograph
Token Photograph
Feature activation+0.069
:
Token:
Feature activation+0.000
Public
Token Public
Feature activation+0.000
ity
Tokenity
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
NZ
Token NZ
Feature activation+0.000
The
Token The
Feature activation+1.031
Anglic
Token Anglic
Feature activation+0.000
an
Tokenan
Feature activation+0.000
Church
Token Church
Feature activation+0.274
has
Token has
Feature activation+1.920
agreed
Token agreed
Feature activation+0.818
to
Token to
Feature activation+0.228
consider
Token consider
Feature activation+0.142
rein
Token rein
Feature activation+0.633
st
Tokenst
Feature activation+0.000
Facebook
TokenFacebook
Feature activation+0.000
Twitter
Token Twitter
Feature activation+0.000
Pinterest
Token Pinterest
Feature activation+0.196
Le
Token Le
Feature activation+1.339
in
Tokenin
Feature activation+0.013
ster
Tokenster
Feature activation+1.771
âĢ
TokenâĢ
Feature activation+1.642
Ļ
TokenĻ
Feature activation+1.491
s
Tokens
Feature activation+2.333
Ben
Token Ben
Feature activation+0.674
Te
Token Te
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Facebook
TokenFacebook
Feature activation+0.000
Twitter
Token Twitter
Feature activation+0.000
Pinterest
Token Pinterest
Feature activation+0.536
Extreme
Token Extreme
Feature activation+1.062
attention
Token attention
Feature activation+1.761
to
Token to
Feature activation+1.412
the
Token the
Feature activation+1.240
finest
Token finest
Feature activation+1.183
details
Token details
Feature activation+1.726
makes
Token makes
Feature activation+0.927
Extreme
Token Extreme
Feature activation+1.062
attention
Token attention
Feature activation+1.761
to
Token to
Feature activation+1.412
the
Token the
Feature activation+1.240
finest
Token finest
Feature activation+1.183
details
Token details
Feature activation+1.726
makes
Token makes
Feature activation+0.927
the
Token the
Feature activation+1.017
Per
Token Per
Feature activation+0.196
uvian
Tokenuvian
Feature activation+0.857
counterfe
Token counterfe
Feature activation+0.000
Twitter
Token Twitter
Feature activation+0.000
Pinterest
Token Pinterest
Feature activation+0.196
Le
Token Le
Feature activation+1.339
in
Tokenin
Feature activation+0.013
ster
Tokenster
Feature activation+1.771
âĢ
TokenâĢ
Feature activation+1.642
Ļ
TokenĻ
Feature activation+1.491
s
Tokens
Feature activation+2.333
Ben
Token Ben
Feature activation+0.674
Te
Token Te
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.208
ee
Tokenee
Feature activation+0.000
Aki
Token Aki
Feature activation+0.000
of
Token of
Feature activation+0.230
Conn
Token Conn
Feature activation+0.000
acht
Tokenacht
Feature activation+0.013
,
Token,
Feature activation+1.621
joins
Token joins
Feature activation+2.633
Worcester
Token Worcester
Feature activation+0.560
this
Token this
Feature activation+0.718
summer
Token summer
Feature activation+1.210
and
Token and
Feature activation+0.806
Ill
Token Ill
Feature activation+0.000
og
Tokenog
Feature activation+0.000
ic
Tokenic
Feature activation+0.141
of
Token of
Feature activation+0.000
Kas
Token Kas
Feature activation+0.000
sel
Tokensel
Feature activation+1.585
.
Token.
Feature activation+1.958
Photograph
Token Photograph
Feature activation+0.069
:
Token:
Feature activation+0.000
Public
Token Public
Feature activation+0.000
ity
Tokenity
Feature activation+0.000
Pinterest
Token Pinterest
Feature activation+0.196
Le
Token Le
Feature activation+1.339
in
Tokenin
Feature activation+0.013
ster
Tokenster
Feature activation+1.771
âĢ
TokenâĢ
Feature activation+1.642
Ļ
TokenĻ
Feature activation+1.491
s
Tokens
Feature activation+2.333
Ben
Token Ben
Feature activation+0.674
Te
Token Te
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.208
Ļ
TokenĻ
Feature activation+0.414

Top DFA by src position
MAX = 3.893

.
Token.
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation-0.012
Top resid features:
Ċ
TokenĊ
Feature activation-0.001
Top resid features:
Facebook
TokenFacebook
Feature activation+0.108
Top resid features:
Twitter
Token Twitter
Feature activation+0.075
Top resid features:
Pinterest
Token Pinterest
Feature activation+3.893
Top resid features:
Le
Token Le
Feature activation+0.029
Top resid features:
in
Tokenin
Feature activation-0.002
Top resid features:
ster
Tokenster
Feature activation-0.004
Top resid features:
âĢ
TokenâĢ
Feature activation+0.065
Top resid features:
Ļ
TokenĻ
Feature activation+0.122
Top resid features:
.
Token.
Feature activation+0.015
Top resid features:
Ċ
TokenĊ
Feature activation+0.030
Top resid features:
Ċ
TokenĊ
Feature activation+0.040
Top resid features:
Facebook
TokenFacebook
Feature activation+0.195
Top resid features:
Twitter
Token Twitter
Feature activation+0.280
Top resid features:
Pinterest
Token Pinterest
Feature activation+3.561
Top resid features:
Extreme
Token Extreme
Feature activation+0.192
Top resid features:
attention
Token attention
Feature activation+0.018
Top resid features:
to
Token to
Feature activation+0.032
Top resid features:
the
Token the
Feature activation+0.018
Top resid features:
finest
Token finest
Feature activation+0.016
Top resid features:
.
Token.
Feature activation-0.000
Top resid features:
Ċ
TokenĊ
Feature activation-0.015
Top resid features:
Ċ
TokenĊ
Feature activation-0.000
Top resid features:
Facebook
TokenFacebook
Feature activation+0.076
Top resid features:
Twitter
Token Twitter
Feature activation+0.051
Top resid features:
Pinterest
Token Pinterest
Feature activation+3.220
Top resid features:
Le
Token Le
Feature activation+0.063
Top resid features:
in
Tokenin
Feature activation-0.011
Top resid features:
ster
Tokenster
Feature activation-0.003
Top resid features:
âĢ
TokenâĢ
Feature activation+0.107
Top resid features:
Ļ
TokenĻ
Feature activation+0.177
Top resid features:
sel
Tokensel
Feature activation+0.006
Top resid features:
Ċ
TokenĊ
Feature activation-0.021
Top resid features:
Ċ
TokenĊ
Feature activation-0.012
Top resid features:
Facebook
TokenFacebook
Feature activation+0.236
Top resid features:
Twitter
Token Twitter
Feature activation+0.116
Top resid features:
Pinterest
Token Pinterest
Feature activation+3.460
Top resid features:
Enrique
Token Enrique
Feature activation+0.193
Top resid features:
V
Token V
Feature activation-0.015
Top resid features:
ila
Tokenila
Feature activation+0.014
Top resid features:
-
Token-
Feature activation-0.000
Top resid features:
Mat
TokenMat
Feature activation-0.014
Top resid features:
port
Token port
Feature activation-0.006
Top resid features:
Ċ
TokenĊ
Feature activation-0.035
Top resid features:
Ċ
TokenĊ
Feature activation-0.052
Top resid features:
Facebook
TokenFacebook
Feature activation+0.148
Top resid features:
Twitter
Token Twitter
Feature activation+0.157
Top resid features:
Pinterest
Token Pinterest
Feature activation+3.549
Top resid features:
Calais
Token Calais
Feature activation+0.127
Top resid features:
migrants
Token migrants
Feature activation+0.348
Top resid features:
:
Token:
Feature activation+0.018
Top resid features:
âĢ
Token âĢ
Feature activation+0.000
Top resid features:
ĺ
Tokenĺ
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.001
Top resid features:
Ċ
TokenĊ
Feature activation-0.014
Top resid features:
Ċ
TokenĊ
Feature activation+0.001
Top resid features:
Facebook
TokenFacebook
Feature activation+0.114
Top resid features:
Twitter
Token Twitter
Feature activation+0.069
Top resid features:
Pinterest
Token Pinterest
Feature activation+3.681
Top resid features:
Le
Token Le
Feature activation+0.044
Top resid features:
in
Tokenin
Feature activation-0.000
Top resid features:
ster
Tokenster
Feature activation-0.034
Top resid features:
âĢ
TokenâĢ
Feature activation+0.023
Top resid features:
Ļ
TokenĻ
Feature activation+0.023
Top resid features:
.
Token.
Feature activation+0.020
Top resid features:
Ċ
TokenĊ
Feature activation+0.018
Top resid features:
Ċ
TokenĊ
Feature activation+0.017
Top resid features:
media
Tokenmedia
Feature activation-0.012
Top resid features:
_
Token_
Feature activation-0.014
Top resid features:
camera
Tokencamera
Feature activation+3.588
Top resid features:
Patrick
Token Patrick
Feature activation+0.095
Top resid features:
Danger
Token Danger
Feature activation+0.000
Top resid features:
field
Tokenfield
Feature activation+0.000
Top resid features:
pre
Token pre
Feature activation+0.000
Top resid features:
pping
Tokenpping
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.009
Top resid features:
Ċ
TokenĊ
Feature activation-0.012
Top resid features:
Ċ
TokenĊ
Feature activation+0.050
Top resid features:
Facebook
TokenFacebook
Feature activation+0.159
Top resid features:
Twitter
Token Twitter
Feature activation+0.144
Top resid features:
Pinterest
Token Pinterest
Feature activation+3.024
Top resid features:
Pop
Token Pop
Feature activation+0.440
Top resid features:
:
Token:
Feature activation-0.030
Top resid features:
30
Token 30
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
000
Token000
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.013
Top resid features:
Ċ
TokenĊ
Feature activation-0.011
Top resid features:
Ċ
TokenĊ
Feature activation-0.001
Top resid features:
Facebook
TokenFacebook
Feature activation+0.102
Top resid features:
Twitter
Token Twitter
Feature activation+0.077
Top resid features:
Pinterest
Token Pinterest
Feature activation+2.970
Top resid features:
Le
Token Le
Feature activation+0.124
Top resid features:
in
Tokenin
Feature activation-0.004
Top resid features:
ster
Tokenster
Feature activation+0.068
Top resid features:
âĢ
TokenâĢ
Feature activation+0.283
Top resid features:
Ļ
TokenĻ
Feature activation+0.102
Top resid features:
.
Token.
Feature activation-0.006
Top resid features:
Ċ
TokenĊ
Feature activation+0.019
Top resid features:
Ċ
TokenĊ
Feature activation+0.026
Top resid features:
Facebook
TokenFacebook
Feature activation+0.101
Top resid features:
Twitter
Token Twitter
Feature activation+0.072
Top resid features:
Pinterest
Token Pinterest
Feature activation+3.566
Top resid features:
Le
Token Le
Feature activation+0.048
Top resid features:
in
Tokenin
Feature activation+0.001
Top resid features:
ster
Tokenster
Feature activation-0.012
Top resid features:
âĢ
TokenâĢ
Feature activation+0.023
Top resid features:
Ļ
TokenĻ
Feature activation+0.026
Top resid features:
sel
Tokensel
Feature activation+0.005
Top resid features:
Ċ
TokenĊ
Feature activation-0.077
Top resid features:
Ċ
TokenĊ
Feature activation-0.129
Top resid features:
Facebook
TokenFacebook
Feature activation+0.258
Top resid features:
Twitter
Token Twitter
Feature activation+0.106
Top resid features:
Pinterest
Token Pinterest
Feature activation+3.204
Top resid features:
Enrique
Token Enrique
Feature activation+0.113
Top resid features:
V
Token V
Feature activation-0.010
Top resid features:
ila
Tokenila
Feature activation+0.016
Top resid features:
-
Token-
Feature activation+0.003
Top resid features:
Mat
TokenMat
Feature activation-0.032
Top resid features:
sel
Tokensel
Feature activation+0.004
Top resid features:
Ċ
TokenĊ
Feature activation-0.088
Top resid features:
Ċ
TokenĊ
Feature activation-0.071
Top resid features:
Facebook
TokenFacebook
Feature activation+0.322
Top resid features:
Twitter
Token Twitter
Feature activation+0.107
Top resid features:
Pinterest
Token Pinterest
Feature activation+2.930
Top resid features:
Enrique
Token Enrique
Feature activation+0.220
Top resid features:
V
Token V
Feature activation-0.019
Top resid features:
ila
Tokenila
Feature activation+0.011
Top resid features:
-
Token-
Feature activation+0.003
Top resid features:
Mat
TokenMat
Feature activation-0.005
Top resid features:
/
Token/
Feature activation-0.042
Top resid features:
FA
TokenFA
Feature activation-0.005
Top resid features:
IR
TokenIR
Feature activation-0.017
Top resid features:
FA
TokenFA
Feature activation-0.005
Top resid features:
X
TokenX
Feature activation+0.021
Top resid features:
NZ
Token NZ
Feature activation+2.842
Top resid features:
The
Token The
Feature activation+0.411
Top resid features:
Anglic
Token Anglic
Feature activation+0.022
Top resid features:
an
Tokenan
Feature activation+0.010
Top resid features:
Church
Token Church
Feature activation+0.015
Top resid features:
has
Token has
Feature activation+0.065
Top resid features:
.
Token.
Feature activation+0.003
Top resid features:
Ċ
TokenĊ
Feature activation-0.015
Top resid features:
Ċ
TokenĊ
Feature activation-0.008
Top resid features:
Facebook
TokenFacebook
Feature activation+0.156
Top resid features:
Twitter
Token Twitter
Feature activation+0.094
Top resid features:
Pinterest
Token Pinterest
Feature activation+2.520
Top resid features:
Le
Token Le
Feature activation+0.071
Top resid features:
in
Tokenin
Feature activation+0.017
Top resid features:
ster
Tokenster
Feature activation+0.030
Top resid features:
âĢ
TokenâĢ
Feature activation+0.000
Top resid features:
Ļ
TokenĻ
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.008
Top resid features:
Ċ
TokenĊ
Feature activation-0.003
Top resid features:
Ċ
TokenĊ
Feature activation+0.023
Top resid features:
Facebook
TokenFacebook
Feature activation+0.090
Top resid features:
Twitter
Token Twitter
Feature activation+0.127
Top resid features:
Pinterest
Token Pinterest
Feature activation+2.263
Top resid features:
Extreme
Token Extreme
Feature activation+0.386
Top resid features:
attention
Token attention
Feature activation+0.007
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
finest
Token finest
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.016
Top resid features:
Ċ
TokenĊ
Feature activation+0.005
Top resid features:
Ċ
TokenĊ
Feature activation+0.022
Top resid features:
Facebook
TokenFacebook
Feature activation+0.094
Top resid features:
Twitter
Token Twitter
Feature activation+0.104
Top resid features:
Pinterest
Token Pinterest
Feature activation+1.636
Top resid features:
Extreme
Token Extreme
Feature activation+0.574
Top resid features:
attention
Token attention
Feature activation+0.052
Top resid features:
to
Token to
Feature activation+0.456
Top resid features:
the
Token the
Feature activation+0.336
Top resid features:
finest
Token finest
Feature activation+0.178
Top resid features:
.
Token.
Feature activation+0.002
Top resid features:
Ċ
TokenĊ
Feature activation-0.007
Top resid features:
Ċ
TokenĊ
Feature activation+0.001
Top resid features:
Facebook
TokenFacebook
Feature activation+0.141
Top resid features:
Twitter
Token Twitter
Feature activation+0.075
Top resid features:
Pinterest
Token Pinterest
Feature activation+2.583
Top resid features:
Le
Token Le
Feature activation+0.104
Top resid features:
in
Tokenin
Feature activation+0.025
Top resid features:
ster
Tokenster
Feature activation+0.108
Top resid features:
âĢ
TokenâĢ
Feature activation+0.073
Top resid features:
Ļ
TokenĻ
Feature activation+0.000
Top resid features:
.
Token.
Feature activation-0.020
Top resid features:
Ċ
TokenĊ
Feature activation+0.003
Top resid features:
Ċ
TokenĊ
Feature activation+0.007
Top resid features:
Facebook
TokenFacebook
Feature activation+0.044
Top resid features:
Twitter
Token Twitter
Feature activation+0.026
Top resid features:
Pinterest
Token Pinterest
Feature activation+2.867
Top resid features:
Le
Token Le
Feature activation+0.037
Top resid features:
in
Tokenin
Feature activation-0.001
Top resid features:
ster
Tokenster
Feature activation+0.010
Top resid features:
âĢ
TokenâĢ
Feature activation+0.052
Top resid features:
Ļ
TokenĻ
Feature activation+0.068
Top resid features:
sel
Tokensel
Feature activation+0.011
Top resid features:
Ċ
TokenĊ
Feature activation-0.231
Top resid features:
Ċ
TokenĊ
Feature activation+0.018
Top resid features:
Facebook
TokenFacebook
Feature activation+0.130
Top resid features:
Twitter
Token Twitter
Feature activation+0.042
Top resid features:
Pinterest
Token Pinterest
Feature activation+2.891
Top resid features:
Enrique
Token Enrique
Feature activation+0.152
Top resid features:
V
Token V
Feature activation-0.016
Top resid features:
ila
Tokenila
Feature activation+0.017
Top resid features:
-
Token-
Feature activation-0.009
Top resid features:
Mat
TokenMat
Feature activation-0.034
Top resid features:
.
Token.
Feature activation+0.001
Top resid features:
Ċ
TokenĊ
Feature activation-0.025
Top resid features:
Ċ
TokenĊ
Feature activation+0.006
Top resid features:
Facebook
TokenFacebook
Feature activation+0.134
Top resid features:
Twitter
Token Twitter
Feature activation+0.102
Top resid features:
Pinterest
Token Pinterest
Feature activation+1.783
Top resid features:
Le
Token Le
Feature activation+0.125
Top resid features:
in
Tokenin
Feature activation+0.015
Top resid features:
ster
Tokenster
Feature activation+0.003
Top resid features:
âĢ
TokenâĢ
Feature activation+0.261
Top resid features:
Ļ
TokenĻ
Feature activation+0.122
Top resid features:

Decoder Weights Distribution

Head 0: 0.02

Head 1: 0.02

Head 2: 0.06

Head 3: 0.50

Head 4: 0.06

Head 5: 0.05

Head 6: 0.03

Head 7: 0.03

Head 8: 0.05

Head 9: 0.05

Head 10: 0.05

Head 11: 0.08

Positive logits

?'1.64

?:1.62

ldon1.54

weeney1.43

?」1.41

Paraly1.39

rador1.38

ucha1.36

reacts1.36

?'"1.36

veland1.33

sunglasses1.30

detained1.30

ierra1.29

tein1.29

inferred1.29

Osama1.29

]'1.28

amas1.27

irin1.27

Negative logits

-2.08

luaj-1.99

��極-1.91

��-1.88

cffffcc-1.75

��-1.70

enthusi-1.68

-1.67

Secondly-1.66

SourceFile-1.62

��-1.61

anyways-1.56

��-1.56

fortunately-1.54

details-1.54

��-1.53

Story-1.52

however-1.51

veter-1.51

thous-1.50

INTERVAL 3.021 - 3.356
CONTAINS 0.000%

the
Token the
Feature activation+0.561
best
Token best
Feature activation+0.719
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
world
Token world
Feature activation+1.060
.
Token.
Feature activation+3.193
Photograph
Token Photograph
Feature activation+0.997
:
Token:
Feature activation+0.000
Jonathan
Token Jonathan
Feature activation+0.000
Franklin
Token Franklin
Feature activation+0.000
for
Token for
Feature activation+0.000
Te
Token Te
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.208
Ļ
TokenĻ
Feature activation+0.414
o
Tokeno
Feature activation+0.478
,
Token,
Feature activation+2.882
in
Token in
Feature activation+3.356
action
Token action
Feature activation+1.242
against
Token against
Feature activation+0.798
Bund
Token Bund
Feature activation+0.000
ee
Tokenee
Feature activation+0.000
Aki
Token Aki
Feature activation+0.000

INTERVAL 2.685 - 3.021
CONTAINS 0.000%

ila
Tokenila
Feature activation+0.146
-
Token-
Feature activation+0.000
Mat
TokenMat
Feature activation+0.000
as
Tokenas
Feature activation+0.000
,
Token,
Feature activation+2.093
The
Token The
Feature activation+2.719
Ill
Token Ill
Feature activation+0.000
og
Tokenog
Feature activation+0.000
ic
Tokenic
Feature activation+0.141
of
Token of
Feature activation+0.000
Kas
Token Kas
Feature activation+0.000
Facebook
TokenFacebook
Feature activation+0.000
Twitter
Token Twitter
Feature activation+0.000
Pinterest
Token Pinterest
Feature activation+0.000
Calais
Token Calais
Feature activation+0.613
migrants
Token migrants
Feature activation+1.044
:
Token:
Feature activation+2.702
âĢ
Token âĢ
Feature activation+0.288
ĺ
Tokenĺ
Feature activation+0.009
Get
TokenGet
Feature activation+0.000
to
Token to
Feature activation+0.000
England
Token England
Feature activation+0.000
Ben
Token Ben
Feature activation+0.674
Te
Token Te
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.208
Ļ
TokenĻ
Feature activation+0.414
o
Tokeno
Feature activation+0.478
,
Token,
Feature activation+2.882
in
Token in
Feature activation+3.356
action
Token action
Feature activation+1.242
against
Token against
Feature activation+0.798
Bund
Token Bund
Feature activation+0.000
ee
Tokenee
Feature activation+0.000

INTERVAL 2.350 - 2.685
CONTAINS 0.000%

Ċ
TokenĊ
Feature activation+0.000
Facebook
TokenFacebook
Feature activation+0.000
Twitter
Token Twitter
Feature activation+0.000
Pinterest
Token Pinterest
Feature activation+0.000
Pop
Token Pop
Feature activation+1.140
:
Token:
Feature activation+2.458
30
Token 30
Feature activation+0.195
,
Token,
Feature activation+0.000
000
Token000
Feature activation+0.511
.
Token.
Feature activation+0.810
Tr
Token Tr
Feature activation+0.130
Aki
Token Aki
Feature activation+0.000
of
Token of
Feature activation+0.230
Conn
Token Conn
Feature activation+0.000
acht
Tokenacht
Feature activation+0.013
,
Token,
Feature activation+1.621
joins
Token joins
Feature activation+2.633
Worcester
Token Worcester
Feature activation+0.560
this
Token this
Feature activation+0.718
summer
Token summer
Feature activation+1.210
and
Token and
Feature activation+0.806
is
Token is
Feature activation+1.274
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
media
Tokenmedia
Feature activation+0.000
_
Token_
Feature activation+0.000
camera
Tokencamera
Feature activation+0.000
Patrick
Token Patrick
Feature activation+2.467
Danger
Token Danger
Feature activation+1.651
field
Tokenfield
Feature activation+2.050
pre
Token pre
Feature activation+1.590
pping
Tokenpping
Feature activation+1.556
the
Token the
Feature activation+1.266

INTERVAL 2.014 - 2.350
CONTAINS 0.000%

V
Token V
Feature activation+0.000
ila
Tokenila
Feature activation+0.146
-
Token-
Feature activation+0.000
Mat
TokenMat
Feature activation+0.000
as
Tokenas
Feature activation+0.000
,
Token,
Feature activation+2.093
The
Token The
Feature activation+2.719
Ill
Token Ill
Feature activation+0.000
og
Tokenog
Feature activation+0.000
ic
Tokenic
Feature activation+0.141
of
Token of
Feature activation+0.000
Le
Token Le
Feature activation+1.339
in
Tokenin
Feature activation+0.013
ster
Tokenster
Feature activation+1.771
âĢ
TokenâĢ
Feature activation+1.642
Ļ
TokenĻ
Feature activation+1.491
s
Tokens
Feature activation+2.333
Ben
Token Ben
Feature activation+0.674
Te
Token Te
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.208
Ļ
TokenĻ
Feature activation+0.414
o
Tokeno
Feature activation+0.478
is
Token is
Feature activation+1.274
eligible
Token eligible
Feature activation+0.000
for
Token for
Feature activation+0.000
England
Token England
Feature activation+0.000
selection
Token selection
Feature activation+0.000
.
Token.
Feature activation+2.231
Photograph
Token Photograph
Feature activation+0.000
:
Token:
Feature activation+0.000
IN
Token IN
Feature activation+0.000
PH
TokenPH
Feature activation+0.000
O
TokenO
Feature activation+0.000

INTERVAL 1.678 - 2.014
CONTAINS 0.001%

Ċ
TokenĊ
Feature activation+0.000
Facebook
TokenFacebook
Feature activation+0.000
Twitter
Token Twitter
Feature activation+0.000
Pinterest
Token Pinterest
Feature activation+0.536
Extreme
Token Extreme
Feature activation+1.062
attention
Token attention
Feature activation+1.761
to
Token to
Feature activation+1.412
the
Token the
Feature activation+1.240
finest
Token finest
Feature activation+1.183
details
Token details
Feature activation+1.726
makes
Token makes
Feature activation+0.927
Extreme
Token Extreme
Feature activation+1.062
attention
Token attention
Feature activation+1.761
to
Token to
Feature activation+1.412
the
Token the
Feature activation+1.240
finest
Token finest
Feature activation+1.183
details
Token details
Feature activation+1.726
makes
Token makes
Feature activation+0.927
the
Token the
Feature activation+1.017
Per
Token Per
Feature activation+0.196
uvian
Tokenuvian
Feature activation+0.857
counterfe
Token counterfe
Feature activation+0.000
og
Tokenog
Feature activation+0.000
ic
Tokenic
Feature activation+0.141
of
Token of
Feature activation+0.000
Kas
Token Kas
Feature activation+0.000
sel
Tokensel
Feature activation+1.585
.
Token.
Feature activation+1.958
Photograph
Token Photograph
Feature activation+0.069
:
Token:
Feature activation+0.000
Public
Token Public
Feature activation+0.000
ity
Tokenity
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Facebook
TokenFacebook
Feature activation+0.000
Twitter
Token Twitter
Feature activation+0.000
Pinterest
Token Pinterest
Feature activation+0.196
Le
Token Le
Feature activation+1.339
in
Tokenin
Feature activation+0.013
ster
Tokenster
Feature activation+1.771
âĢ
TokenâĢ
Feature activation+1.642
Ļ
TokenĻ
Feature activation+1.491
s
Tokens
Feature activation+2.333
Ben
Token Ben
Feature activation+0.674
Te
Token Te
Feature activation+0.000
NZ
Token NZ
Feature activation+0.000
The
Token The
Feature activation+1.031
Anglic
Token Anglic
Feature activation+0.000
an
Tokenan
Feature activation+0.000
Church
Token Church
Feature activation+0.274
has
Token has
Feature activation+1.920
agreed
Token agreed
Feature activation+0.818
to
Token to
Feature activation+0.228
consider
Token consider
Feature activation+0.142
rein
Token rein
Feature activation+0.633
st
Tokenst
Feature activation+0.000

INTERVAL 1.343 - 1.678
CONTAINS 0.001%

Facebook
TokenFacebook
Feature activation+0.000
Twitter
Token Twitter
Feature activation+0.000
Pinterest
Token Pinterest
Feature activation+0.536
Extreme
Token Extreme
Feature activation+1.062
attention
Token attention
Feature activation+1.761
to
Token to
Feature activation+1.412
the
Token the
Feature activation+1.240
finest
Token finest
Feature activation+1.183
details
Token details
Feature activation+1.726
makes
Token makes
Feature activation+0.927
the
Token the
Feature activation+1.017
Ill
Token Ill
Feature activation+0.000
og
Tokenog
Feature activation+0.000
ic
Tokenic
Feature activation+0.141
of
Token of
Feature activation+0.000
Kas
Token Kas
Feature activation+0.000
sel
Tokensel
Feature activation+1.585
.
Token.
Feature activation+1.958
Photograph
Token Photograph
Feature activation+0.069
:
Token:
Feature activation+0.000
Public
Token Public
Feature activation+0.000
ity
Tokenity
Feature activation+0.000
Twitter
Token Twitter
Feature activation+0.000
Pinterest
Token Pinterest
Feature activation+0.196
Le
Token Le
Feature activation+1.339
in
Tokenin
Feature activation+0.013
ster
Tokenster
Feature activation+1.771
âĢ
TokenâĢ
Feature activation+1.642
Ļ
TokenĻ
Feature activation+1.491
s
Tokens
Feature activation+2.333
Ben
Token Ben
Feature activation+0.674
Te
Token Te
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.208
Pinterest
Token Pinterest
Feature activation+0.196
Le
Token Le
Feature activation+1.339
in
Tokenin
Feature activation+0.013
ster
Tokenster
Feature activation+1.771
âĢ
TokenâĢ
Feature activation+1.642
Ļ
TokenĻ
Feature activation+1.491
s
Tokens
Feature activation+2.333
Ben
Token Ben
Feature activation+0.674
Te
Token Te
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.208
Ļ
TokenĻ
Feature activation+0.414
ee
Tokenee
Feature activation+0.000
Aki
Token Aki
Feature activation+0.000
of
Token of
Feature activation+0.230
Conn
Token Conn
Feature activation+0.000
acht
Tokenacht
Feature activation+0.013
,
Token,
Feature activation+1.621
joins
Token joins
Feature activation+2.633
Worcester
Token Worcester
Feature activation+0.560
this
Token this
Feature activation+0.718
summer
Token summer
Feature activation+1.210
and
Token and
Feature activation+0.806

INTERVAL 1.007 - 1.343
CONTAINS 0.002%

Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Facebook
TokenFacebook
Feature activation+0.000
Twitter
Token Twitter
Feature activation+0.000
Pinterest
Token Pinterest
Feature activation+0.536
Extreme
Token Extreme
Feature activation+1.062
attention
Token attention
Feature activation+1.761
to
Token to
Feature activation+1.412
the
Token the
Feature activation+1.240
finest
Token finest
Feature activation+1.183
details
Token details
Feature activation+1.726
goods
Token goods
Feature activation+0.000
to
Token to
Feature activation+0.000
and
Token and
Feature activation+0.000
from
Token from
Feature activation+0.000
surrounding
Token surrounding
Feature activation+0.000
cities
Token cities
Feature activation+1.244
,
Token,
Feature activation+0.215
cutting
Token cutting
Feature activation+0.000
down
Token down
Feature activation+0.000
traffic
Token traffic
Feature activation+0.000
and
Token and
Feature activation+0.000
acht
Tokenacht
Feature activation+0.013
,
Token,
Feature activation+1.621
joins
Token joins
Feature activation+2.633
Worcester
Token Worcester
Feature activation+0.560
this
Token this
Feature activation+0.718
summer
Token summer
Feature activation+1.210
and
Token and
Feature activation+0.806
is
Token is
Feature activation+1.274
eligible
Token eligible
Feature activation+0.000
for
Token for
Feature activation+0.000
England
Token England
Feature activation+0.000
to
Token to
Feature activation+1.412
the
Token the
Feature activation+1.240
finest
Token finest
Feature activation+1.183
details
Token details
Feature activation+1.726
makes
Token makes
Feature activation+0.927
the
Token the
Feature activation+1.017
Per
Token Per
Feature activation+0.196
uvian
Tokenuvian
Feature activation+0.857
counterfe
Token counterfe
Feature activation+0.000
its
Tokenits
Feature activation+0.679
among
Token among
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.208
Ļ
TokenĻ
Feature activation+0.414
o
Tokeno
Feature activation+0.478
,
Token,
Feature activation+2.882
in
Token in
Feature activation+3.356
action
Token action
Feature activation+1.242
against
Token against
Feature activation+0.798
Bund
Token Bund
Feature activation+0.000
ee
Tokenee
Feature activation+0.000
Aki
Token Aki
Feature activation+0.000
of
Token of
Feature activation+0.230

INTERVAL 0.671 - 1.007
CONTAINS 0.002%

The
Token The
Feature activation+1.031
Anglic
Token Anglic
Feature activation+0.000
an
Tokenan
Feature activation+0.000
Church
Token Church
Feature activation+0.274
has
Token has
Feature activation+1.920
agreed
Token agreed
Feature activation+0.818
to
Token to
Feature activation+0.228
consider
Token consider
Feature activation+0.142
rein
Token rein
Feature activation+0.633
st
Tokenst
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
and
Token and
Feature activation+0.000
pin
Token pin
Feature activation+0.000
ball
Tokenball
Feature activation+0.000
machines
Token machines
Feature activation+0.000
.
Token.
Feature activation+0.677
It
Token It
Feature activation+0.674
even
Token even
Feature activation+0.091
has
Token has
Feature activation+0.000
an
Token an
Feature activation+0.000
old
Token old
Feature activation+0.000
-
Token-
Feature activation+0.000
attention
Token attention
Feature activation+1.761
to
Token to
Feature activation+1.412
the
Token the
Feature activation+1.240
finest
Token finest
Feature activation+1.183
details
Token details
Feature activation+1.726
makes
Token makes
Feature activation+0.927
the
Token the
Feature activation+1.017
Per
Token Per
Feature activation+0.196
uvian
Tokenuvian
Feature activation+0.857
counterfe
Token counterfe
Feature activation+0.000
its
Tokenits
Feature activation+0.679
or
Token or
Feature activation+0.000
die
Token die
Feature activation+0.000
trying
Token trying
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+1.107
Ļ
TokenĻ
Feature activation+0.541
Ċ
TokenĊ
Feature activation+0.734
Ċ
TokenĊ
Feature activation+0.000
At
TokenAt
Feature activation+0.000
least
Token least
Feature activation+0.000
15
Token 15
Feature activation+0.000
migrants
Token migrants
Feature activation+0.000
,
Token,
Feature activation+0.000
000
Token000
Feature activation+0.511
.
Token.
Feature activation+0.810
Tr
Token Tr
Feature activation+0.130
ains
Tokenains
Feature activation+0.039
bring
Token bring
Feature activation+0.797
people
Token people
Feature activation+0.054
and
Token and
Feature activation+0.000
goods
Token goods
Feature activation+0.000
to
Token to
Feature activation+0.000
and
Token and
Feature activation+0.000

INTERVAL 0.336 - 0.671
CONTAINS 0.001%

Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Image
TokenImage
Feature activation+0.000
:
Token:
Feature activation+0.000
Vladimir
Token Vladimir
Feature activation+0.000
Putin
Token Putin
Feature activation+0.440
says
Token says
Feature activation+0.327
Ukraine
Token Ukraine
Feature activation+0.240
is
Token is
Feature activation+0.052
playing
Token playing
Feature activation+0.000
a
Token a
Feature activation+0.000
Ļ
TokenĻ
Feature activation+1.491
s
Tokens
Feature activation+2.333
Ben
Token Ben
Feature activation+0.674
Te
Token Te
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.208
Ļ
TokenĻ
Feature activation+0.414
o
Tokeno
Feature activation+0.478
,
Token,
Feature activation+2.882
in
Token in
Feature activation+3.356
action
Token action
Feature activation+1.242
against
Token against
Feature activation+0.798
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Facebook
TokenFacebook
Feature activation+0.000
Twitter
Token Twitter
Feature activation+0.000
Pinterest
Token Pinterest
Feature activation+0.000
Calais
Token Calais
Feature activation+0.613
migrants
Token migrants
Feature activation+1.044
:
Token:
Feature activation+2.702
âĢ
Token âĢ
Feature activation+0.288
ĺ
Tokenĺ
Feature activation+0.009
Get
TokenGet
Feature activation+0.000
en
Tokenen
Feature activation+0.000
heim
Tokenheim
Feature activation+0.000
,
Token,
Feature activation+0.000
is
Token is
Feature activation+0.078
full
Token full
Feature activation+0.000
of
Token of
Feature activation+0.543
racing
Token racing
Feature activation+0.000
games
Token games
Feature activation+0.000
,
Token,
Feature activation+0.059
shoot
Token shoot
Feature activation+0.000
'
Token '
Feature activation+0.000
depicting
Token depicting
Feature activation+0.485
Democratic
Token Democratic
Feature activation+0.432
pres
Token pres
Feature activation+0.000
iden
Tokeniden
Feature activation+0.000
ital
Tokenital
Feature activation+0.000
nominee
Token nominee
Feature activation+0.355
Hillary
Token Hillary
Feature activation+0.055
Clinton
Token Clinton
Feature activation+0.000
appeared
Token appeared
Feature activation+0.000
near
Token near
Feature activation+0.000
the
Token the
Feature activation+0.000

INTERVAL 0.000 - 0.336
CONTAINS 99.993%

that
Token that
Feature activation+0.000
the
Token the
Feature activation+0.000
striker
Token striker
Feature activation+0.000
does
Token does
Feature activation+0.000
have
Token have
Feature activation+0.000
a
Token a
Feature activation+0.000
release
Token release
Feature activation+0.000
clause
Token clause
Feature activation+0.000
in
Token in
Feature activation+0.000
his
Token his
Feature activation+0.000
contract
Token contract
Feature activation+0.000
le
Tokenle
Feature activation+0.000
ent
Tokenent
Feature activation+0.000
ation
Tokenation
Feature activation+0.000
is
Token is
Feature activation+0.000
trivial
Token trivial
Feature activation+0.000
.
Token.
Feature activation+0.000
I
Token I
Feature activation+0.000
just
Token just
Feature activation+0.000
write
Token write
Feature activation+0.000
!
Token !
Feature activation+0.000
!
Token !
Feature activation+0.000
IP
Token IP
Feature activation+0.000
Act
Token Act
Feature activation+0.000
,
Token,
Feature activation+0.000
first
Token first
Feature activation+0.000
hinted
Token hinted
Feature activation+0.000
at
Token at
Feature activation+0.000
the
Token the
Feature activation+0.000
possibility
Token possibility
Feature activation+0.000
of
Token of
Feature activation+0.000
an
Token an
Feature activation+0.000
alternative
Token alternative
Feature activation+0.000
the
Token the
Feature activation+0.000
reader
Token reader
Feature activation+0.000
of
Token of
Feature activation+0.000
that
Token that
Feature activation+0.000
these
Token these
Feature activation+0.000
are
Token are
Feature activation+0.000
internal
Token internal
Feature activation+0.000
projections
Token projections
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Senator
Token Senator
Feature activation+0.000
Le
Token Le
Feature activation+0.000
ila
Tokenila
Feature activation+0.000
de
Token de
Feature activation+0.000
Lima
Token Lima
Feature activation+0.000
insisted
Token insisted
Feature activation+0.000
she
Token she
Feature activation+0.000
was
Token was
Feature activation+0.000
innocent
Token innocent
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
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 H5.3: (feature 33785)

TOP ACTIVATIONS
MAX = 3.387

Ċ
TokenĊ
Feature activation+0.000
Of
TokenOf
Feature activation+0.000
course
Token course
Feature activation+0.000
the
Token the
Feature activation+0.695
elite
Token elite
Feature activation+0.537
are
Token are
Feature activation+3.387
scared
Token scared
Feature activation+2.412
of
Token of
Feature activation+1.672
the
Token the
Feature activation+0.476
free
Token free
Feature activation+0.000
exchange
Token exchange
Feature activation+0.000
free
Token free
Feature activation+0.000
exchange
Token exchange
Feature activation+0.000
of
Token of
Feature activation+0.000
ideas
Token ideas
Feature activation+0.027
,
Token,
Feature activation+1.567
because
Token because
Feature activation+3.302
that
Token that
Feature activation+1.190
gives
Token gives
Feature activation+0.818
the
Token the
Feature activation+0.018
people
Token people
Feature activation+0.000
way
Token way
Feature activation+0.000
because
Token because
Feature activation+0.000
,
Token,
Feature activation+0.000
of
Token of
Feature activation+0.000
course
Token course
Feature activation+0.088
,
Token,
Feature activation+1.044
he
Token he
Feature activation+2.819
does
Token does
Feature activation+2.106
not
Token not
Feature activation+2.333
back
Token back
Feature activation+0.955
Mitt
Token Mitt
Feature activation+0.579
Romney
Token Romney
Feature activation+0.614
from
Token from
Feature activation+0.038
their
Token their
Feature activation+0.068
established
Token established
Feature activation+0.000
media
Token media
Feature activation+0.000
commentators
Token commentators
Feature activation+0.000
.
Token.
Feature activation+2.585
Ċ
TokenĊ
Feature activation+0.022
Ċ
TokenĊ
Feature activation+0.000
And
TokenAnd
Feature activation+0.080
Labour
Token Labour
Feature activation+0.000
?
Token?
Feature activation+0.000
Of
TokenOf
Feature activation+0.000
course
Token course
Feature activation+0.000
the
Token the
Feature activation+0.695
elite
Token elite
Feature activation+0.537
are
Token are
Feature activation+3.387
scared
Token scared
Feature activation+2.412
of
Token of
Feature activation+1.672
the
Token the
Feature activation+0.476
free
Token free
Feature activation+0.000
exchange
Token exchange
Feature activation+0.000
of
Token of
Feature activation+0.000
of
Token of
Feature activation+0.000
course
Token course
Feature activation+0.088
,
Token,
Feature activation+1.044
he
Token he
Feature activation+2.819
does
Token does
Feature activation+2.106
not
Token not
Feature activation+2.333
back
Token back
Feature activation+0.955
Mitt
Token Mitt
Feature activation+0.579
Romney
Token Romney
Feature activation+0.614
.
Token.
Feature activation+1.119
And
Token And
Feature activation+0.276
of
Token of
Feature activation+0.000
course
Token course
Feature activation+0.000
,
Token,
Feature activation+0.903
already
Token already
Feature activation+1.490
exists
Token exists
Feature activation+1.177
;
Token;
Feature activation+2.257
this
Token this
Feature activation+0.339
bill
Token bill
Feature activation+0.000
simply
Token simply
Feature activation+0.270
removes
Token removes
Feature activation+0.113
any
Token any
Feature activation+0.000
first
Token first
Feature activation+0.000
priority
Token priority
Feature activation+0.000
is
Token is
Feature activation+0.000
of
Token of
Feature activation+0.000
course
Token course
Feature activation+0.393
to
Token to
Feature activation+2.202
get
Token get
Feature activation+1.707
drafted
Token drafted
Feature activation+0.934
anywhere
Token anywhere
Feature activation+0.457
.
Token.
Feature activation+1.283
And
Token And
Feature activation+0.439
are
Token are
Feature activation+1.746
due
Token due
Feature activation+1.511
to
Token to
Feature activation+0.753
misconduct
Token misconduct
Feature activation+0.054
,
Token,
Feature activation+1.023
and
Token and
Feature activation+2.126
it
Token it
Feature activation+1.068
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
"
Token "
Feature activation+0.000
Of
TokenOf
Feature activation+0.000
course
Token course
Feature activation+0.000
I
Token I
Feature activation+0.705
'd
Token'd
Feature activation+1.745
like
Token like
Feature activation+2.116
to
Token to
Feature activation+1.965
move
Token move
Feature activation+1.006
to
Token to
Feature activation+0.976
one
Token one
Feature activation+0.691
of
Token of
Feature activation+1.066
,
Token,
Feature activation+0.000
of
Token of
Feature activation+0.000
course
Token course
Feature activation+0.088
,
Token,
Feature activation+1.044
he
Token he
Feature activation+2.819
does
Token does
Feature activation+2.106
not
Token not
Feature activation+2.333
back
Token back
Feature activation+0.955
Mitt
Token Mitt
Feature activation+0.579
Romney
Token Romney
Feature activation+0.614
.
Token.
Feature activation+1.119
this
Token this
Feature activation+0.761
ecological
Token ecological
Feature activation+0.308
DO
Token DO
Feature activation+0.000
OM
TokenOM
Feature activation+0.119
is
Token is
Feature activation+1.947
just
Token just
Feature activation+2.095
an
Token an
Feature activation+1.493
excuse
Token excuse
Feature activation+1.075
,
Token,
Feature activation+1.278
as
Token as
Feature activation+1.615
urban
Token urban
Feature activation+0.000
the
Token the
Feature activation+0.000
biggest
Token biggest
Feature activation+0.000
ones
Token ones
Feature activation+0.000
,
Token,
Feature activation+0.110
and
Token and
Feature activation+1.417
has
Token has
Feature activation+2.092
since
Token since
Feature activation+0.071
January
Token January
Feature activation+0.000
,
Token,
Feature activation+0.149
2008
Token 2008
Feature activation+0.000
.
Token.
Feature activation+0.988
low
Token low
Feature activation+0.000
-
Token-
Feature activation+0.000
income
Tokenincome
Feature activation+0.000
status
Token status
Feature activation+0.250
does
Token does
Feature activation+1.091
not
Token not
Feature activation+2.089
itself
Token itself
Feature activation+0.613
depress
Token depress
Feature activation+0.739
achievement
Token achievement
Feature activation+0.274
,
Token,
Feature activation+1.826
but
Token but
Feature activation+0.863
crypto
Token crypto
Feature activation+0.020
/
Token/
Feature activation+0.000
block
Tokenblock
Feature activation+0.000
chain
Tokenchain
Feature activation+0.000
movement
Token movement
Feature activation+0.265
,
Token,
Feature activation+2.063
but
Token but
Feature activation+0.428
the
Token the
Feature activation+0.000
noise
Token noise
Feature activation+0.000
it
Token it
Feature activation+0.000
produces
Token produces
Feature activation+0.000
Of
Token Of
Feature activation+0.000
course
Token course
Feature activation+0.000
,
Token,
Feature activation+0.568
that
Token that
Feature activation+0.837
reality
Token reality
Feature activation+0.421
is
Token is
Feature activation+2.030
likely
Token likely
Feature activation+1.288
still
Token still
Feature activation+1.110
off
Token off
Feature activation+0.441
in
Token in
Feature activation+0.307
the
Token the
Feature activation+0.275
claim
Token claim
Feature activation+0.000
,
Token,
Feature activation+0.000
of
Token of
Feature activation+0.000
course
Token course
Feature activation+0.000
,
Token,
Feature activation+0.877
was
Token was
Feature activation+1.996
absurd
Token absurd
Feature activation+0.941
.
Token.
Feature activation+0.995
The
Token The
Feature activation+0.000
threat
Token threat
Feature activation+0.000
those
Token those
Feature activation+0.000
economic
Token economic
Feature activation+0.000
powerhouse
Token powerhouse
Feature activation+0.030
of
Token of
Feature activation+0.000
Europe
Token Europe
Feature activation+0.202
,
Token,
Feature activation+1.448
nobody
Token nobody
Feature activation+1.986
sees
Token sees
Feature activation+1.388
the
Token the
Feature activation+0.488
trouble
Token trouble
Feature activation+0.238
brewing
Token brewing
Feature activation+0.000
.
Token.
Feature activation+1.414
Of
TokenOf
Feature activation+0.000
course
Token course
Feature activation+0.000
I
Token I
Feature activation+0.705
'd
Token'd
Feature activation+1.745
like
Token like
Feature activation+2.116
to
Token to
Feature activation+1.965
move
Token move
Feature activation+1.006
to
Token to
Feature activation+0.976
one
Token one
Feature activation+0.691
of
Token of
Feature activation+1.066
those
Token those
Feature activation+0.963
,
Token,
Feature activation+0.737
this
Token this
Feature activation+0.761
ecological
Token ecological
Feature activation+0.308
DO
Token DO
Feature activation+0.000
OM
TokenOM
Feature activation+0.119
is
Token is
Feature activation+1.947
just
Token just
Feature activation+2.095
an
Token an
Feature activation+1.493
excuse
Token excuse
Feature activation+1.075
,
Token,
Feature activation+1.278
as
Token as
Feature activation+1.615

Top DFA by src position
MAX = 3.536

end
Token end
Feature activation-0.000
Top resid features:
.
Token.
Feature activation-0.003
Top resid features:
Ċ
TokenĊ
Feature activation+0.017
Top resid features:
Ċ
TokenĊ
Feature activation+0.017
Top resid features:
Of
TokenOf
Feature activation+0.021
Top resid features:
course
Token course
Feature activation+3.490
Top resid features:
the
Token the
Feature activation+0.376
Top resid features:
elite
Token elite
Feature activation+0.110
Top resid features:
are
Token are
Feature activation+0.088
Top resid features:
scared
Token scared
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
end
Token end
Feature activation-0.000
Top resid features:
.
Token.
Feature activation+0.003
Top resid features:
Ċ
TokenĊ
Feature activation+0.005
Top resid features:
Ċ
TokenĊ
Feature activation+0.007
Top resid features:
Of
TokenOf
Feature activation+0.015
Top resid features:
course
Token course
Feature activation+3.536
Top resid features:
the
Token the
Feature activation+0.243
Top resid features:
elite
Token elite
Feature activation+0.016
Top resid features:
are
Token are
Feature activation+0.044
Top resid features:
scared
Token scared
Feature activation+0.002
Top resid features:
of
Token of
Feature activation-0.026
Top resid features:
RNC
Token RNC
Feature activation+0.005
Top resid features:
stage
Token stage
Feature activation-0.004
Top resid features:
because
Token because
Feature activation-0.061
Top resid features:
,
Token,
Feature activation+0.028
Top resid features:
of
Token of
Feature activation+0.057
Top resid features:
course
Token course
Feature activation+1.940
Top resid features:
,
Token,
Feature activation+1.085
Top resid features:
he
Token he
Feature activation+0.170
Top resid features:
does
Token does
Feature activation+0.000
Top resid features:
not
Token not
Feature activation+0.000
Top resid features:
back
Token back
Feature activation+0.000
Top resid features:
.
Token.
Feature activation-0.009
Top resid features:
Tory
Token Tory
Feature activation+0.011
Top resid features:
voters
Token voters
Feature activation+0.015
Top resid features:
were
Token were
Feature activation+0.012
Top resid features:
of
Token of
Feature activation+0.023
Top resid features:
course
Token course
Feature activation+2.960
Top resid features:
largely
Token largely
Feature activation+0.028
Top resid features:
conspicuous
Token conspicuous
Feature activation+0.078
Top resid features:
by
Token by
Feature activation+0.016
Top resid features:
their
Token their
Feature activation-0.011
Top resid features:
absence
Token absence
Feature activation+0.005
Top resid features:
end
Token end
Feature activation-0.001
Top resid features:
.
Token.
Feature activation+0.001
Top resid features:
Ċ
TokenĊ
Feature activation+0.013
Top resid features:
Ċ
TokenĊ
Feature activation+0.009
Top resid features:
Of
TokenOf
Feature activation+0.022
Top resid features:
course
Token course
Feature activation+2.699
Top resid features:
the
Token the
Feature activation+0.240
Top resid features:
elite
Token elite
Feature activation+0.035
Top resid features:
are
Token are
Feature activation+0.134
Top resid features:
scared
Token scared
Feature activation-0.003
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
RNC
Token RNC
Feature activation+0.003
Top resid features:
stage
Token stage
Feature activation-0.008
Top resid features:
because
Token because
Feature activation-0.053
Top resid features:
,
Token,
Feature activation+0.028
Top resid features:
of
Token of
Feature activation+0.042
Top resid features:
course
Token course
Feature activation+2.192
Top resid features:
,
Token,
Feature activation+0.436
Top resid features:
he
Token he
Feature activation+0.188
Top resid features:
does
Token does
Feature activation+0.046
Top resid features:
not
Token not
Feature activation+0.036
Top resid features:
back
Token back
Feature activation+0.000
Top resid features:
.
Token.
Feature activation-0.014
Top resid features:
That
Token That
Feature activation+0.069
Top resid features:
right
Token right
Feature activation+0.014
Top resid features:
,
Token,
Feature activation-0.024
Top resid features:
of
Token of
Feature activation+0.007
Top resid features:
course
Token course
Feature activation+2.427
Top resid features:
,
Token,
Feature activation+0.354
Top resid features:
already
Token already
Feature activation+0.050
Top resid features:
exists
Token exists
Feature activation+0.150
Top resid features:
;
Token;
Feature activation+0.025
Top resid features:
this
Token this
Feature activation+0.000
Top resid features:
's
Token's
Feature activation+0.017
Top resid features:
first
Token first
Feature activation+0.034
Top resid features:
priority
Token priority
Feature activation-0.017
Top resid features:
is
Token is
Feature activation+0.035
Top resid features:
of
Token of
Feature activation+0.050
Top resid features:
course
Token course
Feature activation+2.545
Top resid features:
to
Token to
Feature activation+0.093
Top resid features:
get
Token get
Feature activation+0.000
Top resid features:
drafted
Token drafted
Feature activation+0.000
Top resid features:
anywhere
Token anywhere
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
have
Token have
Feature activation-0.002
Top resid features:
been
Token been
Feature activation-0.002
Top resid features:
tried
Token tried
Feature activation+0.001
Top resid features:
.
Token.
Feature activation-0.013
Top resid features:
Of
Token Of
Feature activation+0.012
Top resid features:
course
Token course
Feature activation+2.101
Top resid features:
,
Token,
Feature activation+0.482
Top resid features:
a
Token a
Feature activation+0.065
Top resid features:
good
Token good
Feature activation+0.011
Top resid features:
number
Token number
Feature activation+0.003
Top resid features:
of
Token of
Feature activation+0.007
Top resid features:
Sport
Token Sport
Feature activation-0.020
Top resid features:
B
TokenB
Feature activation-0.019
Top resid features:
ild
Tokenild
Feature activation+0.002
Top resid features:
"
Token "
Feature activation-0.021
Top resid features:
Of
TokenOf
Feature activation+0.016
Top resid features:
course
Token course
Feature activation+2.411
Top resid features:
I
Token I
Feature activation+0.444
Top resid features:
'd
Token'd
Feature activation+0.017
Top resid features:
like
Token like
Feature activation-0.028
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
move
Token move
Feature activation+0.000
Top resid features:
RNC
Token RNC
Feature activation+0.003
Top resid features:
stage
Token stage
Feature activation-0.000
Top resid features:
because
Token because
Feature activation-0.060
Top resid features:
,
Token,
Feature activation+0.030
Top resid features:
of
Token of
Feature activation+0.038
Top resid features:
course
Token course
Feature activation+1.801
Top resid features:
,
Token,
Feature activation+0.737
Top resid features:
he
Token he
Feature activation+0.160
Top resid features:
does
Token does
Feature activation+0.002
Top resid features:
not
Token not
Feature activation+0.000
Top resid features:
back
Token back
Feature activation+0.000
Top resid features:
familiar
Token familiar
Feature activation+0.020
Top resid features:
âĢ¦
TokenâĢ¦
Feature activation+0.004
Top resid features:
Ċ
TokenĊ
Feature activation+0.016
Top resid features:
Ċ
TokenĊ
Feature activation-0.001
Top resid features:
Of
TokenOf
Feature activation+0.039
Top resid features:
course
Token course
Feature activation+1.658
Top resid features:
,
Token,
Feature activation+0.673
Top resid features:
this
Token this
Feature activation+0.269
Top resid features:
ecological
Token ecological
Feature activation+0.040
Top resid features:
DO
Token DO
Feature activation-0.001
Top resid features:
OM
TokenOM
Feature activation+0.001
Top resid features:
made
Token made
Feature activation-0.008
Top resid features:
?
Token?
Feature activation-0.006
Top resid features:
Ċ
TokenĊ
Feature activation-0.011
Top resid features:
Ċ
TokenĊ
Feature activation-0.007
Top resid features:
Of
TokenOf
Feature activation+0.014
Top resid features:
course
Token course
Feature activation+2.204
Top resid features:
,
Token,
Feature activation+0.290
Top resid features:
Greens
Token Greens
Feature activation+0.011
Top resid features:
pan
Tokenpan
Feature activation-0.001
Top resid features:
loves
Token loves
Feature activation+0.254
Top resid features:
hedge
Token hedge
Feature activation-0.021
Top resid features:
teachers
Token teachers
Feature activation+0.002
Top resid features:
?
Token?
Feature activation-0.018
Top resid features:
Ċ
TokenĊ
Feature activation-0.021
Top resid features:
Ċ
TokenĊ
Feature activation-0.017
Top resid features:
Of
TokenOf
Feature activation+0.012
Top resid features:
course
Token course
Feature activation+1.934
Top resid features:
,
Token,
Feature activation+0.669
Top resid features:
minority
Token minority
Feature activation+0.051
Top resid features:
or
Token or
Feature activation+0.013
Top resid features:
low
Token low
Feature activation+0.001
Top resid features:
-
Token-
Feature activation+0.002
Top resid features:
inspire
Token inspire
Feature activation+0.001
Top resid features:
unconditional
Token unconditional
Feature activation+0.001
Top resid features:
respect
Token respect
Feature activation+0.006
Top resid features:
.
Token.
Feature activation-0.037
Top resid features:
Of
Token Of
Feature activation+0.019
Top resid features:
course
Token course
Feature activation+1.694
Top resid features:
,
Token,
Feature activation+0.894
Top resid features:
ICO
Token ICO
Feature activation+0.065
Top resid features:
s
Tokens
Feature activation+0.002
Top resid features:
are
Token are
Feature activation+0.051
Top resid features:
only
Token only
Feature activation+0.076
Top resid features:
of
Token of
Feature activation-0.001
Top resid features:
interactive
Token interactive
Feature activation-0.002
Top resid features:
possibilities
Token possibilities
Feature activation+0.006
Top resid features:
.
Token.
Feature activation-0.030
Top resid features:
Of
Token Of
Feature activation+0.005
Top resid features:
course
Token course
Feature activation+1.903
Top resid features:
,
Token,
Feature activation+0.573
Top resid features:
that
Token that
Feature activation+0.221
Top resid features:
reality
Token reality
Feature activation+0.095
Top resid features:
is
Token is
Feature activation+0.019
Top resid features:
likely
Token likely
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.002
Top resid features:
That
TokenThat
Feature activation+0.008
Top resid features:
claim
Token claim
Feature activation+0.004
Top resid features:
,
Token,
Feature activation-0.019
Top resid features:
of
Token of
Feature activation+0.150
Top resid features:
course
Token course
Feature activation+1.879
Top resid features:
,
Token,
Feature activation+0.590
Top resid features:
was
Token was
Feature activation+0.056
Top resid features:
absurd
Token absurd
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
The
Token The
Feature activation+0.000
Top resid features:
bubble
Token bubble
Feature activation+0.002
Top resid features:
.
Token.
Feature activation-0.014
Top resid features:
Ċ
TokenĊ
Feature activation-0.005
Top resid features:
Ċ
TokenĊ
Feature activation-0.009
Top resid features:
Of
TokenOf
Feature activation+0.011
Top resid features:
course
Token course
Feature activation+1.644
Top resid features:
,
Token,
Feature activation+0.661
Top resid features:
given
Token given
Feature activation+0.158
Top resid features:
that
Token that
Feature activation+0.205
Top resid features:
Germany
Token Germany
Feature activation+0.003
Top resid features:
is
Token is
Feature activation+0.007
Top resid features:
Sport
Token Sport
Feature activation-0.020
Top resid features:
B
TokenB
Feature activation-0.006
Top resid features:
ild
Tokenild
Feature activation+0.000
Top resid features:
"
Token "
Feature activation-0.023
Top resid features:
Of
TokenOf
Feature activation+0.013
Top resid features:
course
Token course
Feature activation+2.361
Top resid features:
I
Token I
Feature activation+0.266
Top resid features:
'd
Token'd
Feature activation+0.073
Top resid features:
like
Token like
Feature activation+0.027
Top resid features:
to
Token to
Feature activation-0.030
Top resid features:
move
Token move
Feature activation+0.000
Top resid features:
familiar
Token familiar
Feature activation+0.063
Top resid features:
âĢ¦
TokenâĢ¦
Feature activation+0.006
Top resid features:
Ċ
TokenĊ
Feature activation+0.027
Top resid features:
Ċ
TokenĊ
Feature activation+0.008
Top resid features:
Of
TokenOf
Feature activation+0.023
Top resid features:
course
Token course
Feature activation+1.670
Top resid features:
,
Token,
Feature activation+0.646
Top resid features:
this
Token this
Feature activation+0.190
Top resid features:
ecological
Token ecological
Feature activation+0.077
Top resid features:
DO
Token DO
Feature activation-0.009
Top resid features:
OM
TokenOM
Feature activation-0.061
Top resid features:

Decoder Weights Distribution

Head 0: 0.02

Head 1: 0.03

Head 2: 0.09

Head 3: 0.50

Head 4: 0.06

Head 5: 0.03

Head 6: 0.03

Head 7: 0.04

Head 8: 0.04

Head 9: 0.04

Head 10: 0.06

Head 11: 0.06

Positive logits

🙂1.93

)).1.87

));1.82

),"1.70

)=1.68

)—1.68

spoilers1.68

=>1.66

*)1.66

);1.63

1.61

))1.59

>)1.58

caveats1.57

)),1.57

)}1.55

;)1.54

obligatory1.52

!).1.50

admittedly1.48

Negative logits

-1.65

Chattanooga-1.58

WARN-1.53

hens-1.48

]-1.48

Trem-1.41

pson-1.41

Hes-1.40

Chic-1.40

guiName-1.39

..."-1.38

panic-1.37

hab-1.36

isively-1.36

[/-1.35

sic-1.34

Quote-1.34

Harlem-1.33

-1.33

Leth-1.33

INTERVAL 3.048 - 3.387
CONTAINS 0.000%

Ċ
TokenĊ
Feature activation+0.000
Of
TokenOf
Feature activation+0.000
course
Token course
Feature activation+0.000
the
Token the
Feature activation+0.695
elite
Token elite
Feature activation+0.537
are
Token are
Feature activation+3.387
scared
Token scared
Feature activation+2.412
of
Token of
Feature activation+1.672
the
Token the
Feature activation+0.476
free
Token free
Feature activation+0.000
exchange
Token exchange
Feature activation+0.000
free
Token free
Feature activation+0.000
exchange
Token exchange
Feature activation+0.000
of
Token of
Feature activation+0.000
ideas
Token ideas
Feature activation+0.027
,
Token,
Feature activation+1.567
because
Token because
Feature activation+3.302
that
Token that
Feature activation+1.190
gives
Token gives
Feature activation+0.818
the
Token the
Feature activation+0.018
people
Token people
Feature activation+0.000
way
Token way
Feature activation+0.000

INTERVAL 2.710 - 3.048
CONTAINS 0.000%

because
Token because
Feature activation+0.000
,
Token,
Feature activation+0.000
of
Token of
Feature activation+0.000
course
Token course
Feature activation+0.088
,
Token,
Feature activation+1.044
he
Token he
Feature activation+2.819
does
Token does
Feature activation+2.106
not
Token not
Feature activation+2.333
back
Token back
Feature activation+0.955
Mitt
Token Mitt
Feature activation+0.579
Romney
Token Romney
Feature activation+0.614

INTERVAL 2.371 - 2.710
CONTAINS 0.000%

Of
TokenOf
Feature activation+0.000
course
Token course
Feature activation+0.000
the
Token the
Feature activation+0.695
elite
Token elite
Feature activation+0.537
are
Token are
Feature activation+3.387
scared
Token scared
Feature activation+2.412
of
Token of
Feature activation+1.672
the
Token the
Feature activation+0.476
free
Token free
Feature activation+0.000
exchange
Token exchange
Feature activation+0.000
of
Token of
Feature activation+0.000
from
Token from
Feature activation+0.038
their
Token their
Feature activation+0.068
established
Token established
Feature activation+0.000
media
Token media
Feature activation+0.000
commentators
Token commentators
Feature activation+0.000
.
Token.
Feature activation+2.585
Ċ
TokenĊ
Feature activation+0.022
Ċ
TokenĊ
Feature activation+0.000
And
TokenAnd
Feature activation+0.080
Labour
Token Labour
Feature activation+0.000
?
Token?
Feature activation+0.000

INTERVAL 2.032 - 2.371
CONTAINS 0.001%

first
Token first
Feature activation+0.000
priority
Token priority
Feature activation+0.000
is
Token is
Feature activation+0.000
of
Token of
Feature activation+0.000
course
Token course
Feature activation+0.393
to
Token to
Feature activation+2.202
get
Token get
Feature activation+1.707
drafted
Token drafted
Feature activation+0.934
anywhere
Token anywhere
Feature activation+0.457
.
Token.
Feature activation+1.283
And
Token And
Feature activation+0.439
,
Token,
Feature activation+0.000
of
Token of
Feature activation+0.000
course
Token course
Feature activation+0.088
,
Token,
Feature activation+1.044
he
Token he
Feature activation+2.819
does
Token does
Feature activation+2.106
not
Token not
Feature activation+2.333
back
Token back
Feature activation+0.955
Mitt
Token Mitt
Feature activation+0.579
Romney
Token Romney
Feature activation+0.614
.
Token.
Feature activation+1.119
this
Token this
Feature activation+0.761
ecological
Token ecological
Feature activation+0.308
DO
Token DO
Feature activation+0.000
OM
TokenOM
Feature activation+0.119
is
Token is
Feature activation+1.947
just
Token just
Feature activation+2.095
an
Token an
Feature activation+1.493
excuse
Token excuse
Feature activation+1.075
,
Token,
Feature activation+1.278
as
Token as
Feature activation+1.615
urban
Token urban
Feature activation+0.000
"
Token "
Feature activation+0.000
Of
TokenOf
Feature activation+0.000
course
Token course
Feature activation+0.000
I
Token I
Feature activation+0.705
'd
Token'd
Feature activation+1.745
like
Token like
Feature activation+2.116
to
Token to
Feature activation+1.965
move
Token move
Feature activation+1.006
to
Token to
Feature activation+0.976
one
Token one
Feature activation+0.691
of
Token of
Feature activation+1.066
low
Token low
Feature activation+0.000
-
Token-
Feature activation+0.000
income
Tokenincome
Feature activation+0.000
status
Token status
Feature activation+0.250
does
Token does
Feature activation+1.091
not
Token not
Feature activation+2.089
itself
Token itself
Feature activation+0.613
depress
Token depress
Feature activation+0.739
achievement
Token achievement
Feature activation+0.274
,
Token,
Feature activation+1.826
but
Token but
Feature activation+0.863

INTERVAL 1.693 - 2.032
CONTAINS 0.003%

economic
Token economic
Feature activation+0.000
powerhouse
Token powerhouse
Feature activation+0.030
of
Token of
Feature activation+0.000
Europe
Token Europe
Feature activation+0.202
,
Token,
Feature activation+1.448
nobody
Token nobody
Feature activation+1.986
sees
Token sees
Feature activation+1.388
the
Token the
Feature activation+0.488
trouble
Token trouble
Feature activation+0.238
brewing
Token brewing
Feature activation+0.000
.
Token.
Feature activation+1.414
This
Token This
Feature activation+0.000
,
Token,
Feature activation+0.000
of
Token of
Feature activation+0.000
course
Token course
Feature activation+0.000
,
Token,
Feature activation+0.696
means
Token means
Feature activation+1.815
that
Token that
Feature activation+1.575
your
Token your
Feature activation+0.000
doctor
Token doctor
Feature activation+0.000
may
Token may
Feature activation+0.013
no
Token no
Feature activation+0.000
inflation
Token inflation
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.062
years
Token years
Feature activation+0.003
since
Token since
Feature activation+0.288
:
Token:
Feature activation+1.851
Resident
Token Resident
Feature activation+0.000
tuition
Token tuition
Feature activation+0.000
at
Token at
Feature activation+0.000
Michigan
Token Michigan
Feature activation+0.000
this
Token this
Feature activation+0.000
of
Token of
Feature activation+0.000
course
Token course
Feature activation+0.000
largely
Token largely
Feature activation+1.105
conspicuous
Token conspicuous
Feature activation+1.696
by
Token by
Feature activation+0.728
their
Token their
Feature activation+1.759
absence
Token absence
Feature activation+0.488
on
Token on
Feature activation+0.097
Twitter
Token Twitter
Feature activation+0.299
away
Token away
Feature activation+0.141
from
Token from
Feature activation+0.038
want
Token want
Feature activation+0.000
,
Token,
Feature activation+0.000
of
Token of
Feature activation+0.000
course
Token course
Feature activation+0.000
,
Token,
Feature activation+0.000
is
Token is
Feature activation+1.849
access
Token access
Feature activation+0.364
,
Token,
Feature activation+0.331
should
Token should
Feature activation+0.187
she
Token she
Feature activation+0.000
become
Token become
Feature activation+0.000

INTERVAL 1.355 - 1.693
CONTAINS 0.006%

Of
TokenOf
Feature activation+0.000
course
Token course
Feature activation+0.000
,
Token,
Feature activation+0.000
Greens
Token Greens
Feature activation+0.000
pan
Tokenpan
Feature activation+0.035
loves
Token loves
Feature activation+1.562
hedge
Token hedge
Feature activation+0.166
funds
Token funds
Feature activation+0.422
now
Token now
Feature activation+0.598
,
Token,
Feature activation+1.050
especially
Token especially
Feature activation+0.644
the
Token the
Feature activation+0.331
various
Token various
Feature activation+0.082
agencies
Token agencies
Feature activation+0.057
report
Token report
Feature activation+0.000
on
Token on
Feature activation+0.000
include
Token include
Feature activation+1.465
things
Token things
Feature activation+0.785
that
Token that
Feature activation+0.587
aren
Token aren
Feature activation+0.000
't
Token't
Feature activation+0.000
really
Token really
Feature activation+0.000
,
Token,
Feature activation+0.261
your
Token your
Feature activation+0.249
privacy
Token privacy
Feature activation+0.647
is
Token is
Feature activation+1.887
worth
Token worth
Feature activation+1.366
it
Token it
Feature activation+1.365
,
Token,
Feature activation+1.903
but
Token but
Feature activation+0.655
they
Token they
Feature activation+0.000
're
Token're
Feature activation+0.000
all
Token all
Feature activation+0.000
powerhouse
Token powerhouse
Feature activation+0.030
of
Token of
Feature activation+0.000
Europe
Token Europe
Feature activation+0.202
,
Token,
Feature activation+1.448
nobody
Token nobody
Feature activation+1.986
sees
Token sees
Feature activation+1.388
the
Token the
Feature activation+0.488
trouble
Token trouble
Feature activation+0.238
brewing
Token brewing
Feature activation+0.000
.
Token.
Feature activation+1.414
They
Token They
Feature activation+0.032
building
Token building
Feature activation+0.228
block
Token block
Feature activation+0.762
of
Token of
Feature activation+0.836
all
Token all
Feature activation+0.306
life
Token life
Feature activation+0.173
.
Token.
Feature activation+1.625
DNA
Token DNA
Feature activation+0.096
stores
Token stores
Feature activation+0.007
our
Token our
Feature activation+0.051
genetic
Token genetic
Feature activation+0.000
information
Token information
Feature activation+0.000

INTERVAL 1.016 - 1.355
CONTAINS 0.011%

course
Token course
Feature activation+0.000
,
Token,
Feature activation+0.010
ICO
Token ICO
Feature activation+0.000
s
Tokens
Feature activation+0.505
are
Token are
Feature activation+1.473
only
Token only
Feature activation+1.295
a
Token a
Feature activation+0.539
subset
Token subset
Feature activation+0.706
of
Token of
Feature activation+0.489
the
Token the
Feature activation+0.633
broader
Token broader
Feature activation+0.741
exchange
Token exchange
Feature activation+0.000
of
Token of
Feature activation+0.000
ideas
Token ideas
Feature activation+0.027
,
Token,
Feature activation+1.567
because
Token because
Feature activation+3.302
that
Token that
Feature activation+1.190
gives
Token gives
Feature activation+0.818
the
Token the
Feature activation+0.018
people
Token people
Feature activation+0.000
way
Token way
Feature activation+0.000
too
Token too
Feature activation+0.000
,
Token,
Feature activation+0.568
that
Token that
Feature activation+0.837
reality
Token reality
Feature activation+0.421
is
Token is
Feature activation+2.030
likely
Token likely
Feature activation+1.288
still
Token still
Feature activation+1.110
off
Token off
Feature activation+0.441
in
Token in
Feature activation+0.307
the
Token the
Feature activation+0.275
distance
Token distance
Feature activation+0.488
âĢĶ
Token âĢĶ
Feature activation+1.363
this
Token this
Feature activation+0.515
and
Token and
Feature activation+0.749
other
Token other
Feature activation+0.605
matters
Token matters
Feature activation+0.769
will
Token will
Feature activation+1.711
be
Token be
Feature activation+1.326
part
Token part
Feature activation+0.674
of
Token of
Feature activation+0.734
the
Token the
Feature activation+0.596
agenda
Token agenda
Feature activation+0.948
that
Token that
Feature activation+0.952
course
Token course
Feature activation+0.000
at
Token at
Feature activation+0.154
some
Token some
Feature activation+0.257
point
Token point
Feature activation+0.517
we
Token we
Feature activation+0.697
would
Token would
Feature activation+1.305
like
Token like
Feature activation+0.879
to
Token to
Feature activation+1.293
have
Token have
Feature activation+0.697
one
Token one
Feature activation+0.412
big
Token big
Feature activation+0.169

INTERVAL 0.677 - 1.016
CONTAINS 0.023%

the
Token the
Feature activation+0.471
mask
Token mask
Feature activation+0.865
,
Token,
Feature activation+0.577
he
Token he
Feature activation+1.120
was
Token was
Feature activation+1.787
raised
Token raised
Feature activation+0.738
to
Token to
Feature activation+0.460
love
Token love
Feature activation+1.221
.
Token.
Feature activation+1.065
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
but
Token but
Feature activation+0.000
of
Token of
Feature activation+0.000
course
Token course
Feature activation+0.000
some
Token some
Feature activation+1.265
silly
Token silly
Feature activation+0.766
sequences
Token sequences
Feature activation+0.680
as
Token as
Feature activation+0.559
well
Token well
Feature activation+0.278
.
Token.
Feature activation+1.012
At
Token At
Feature activation+0.000
one
Token one
Feature activation+0.000
,
Token,
Feature activation+0.000
of
Token of
Feature activation+0.000
course
Token course
Feature activation+0.000
,
Token,
Feature activation+0.735
have
Token have
Feature activation+0.899
some
Token some
Feature activation+0.991
responsibility
Token responsibility
Feature activation+1.070
.
Token.
Feature activation+1.180
If
Token If
Feature activation+0.000
you
Token you
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
,
Token,
Feature activation+0.384
the
Token the
Feature activation+0.862
times
Token times
Feature activation+0.460
,
Token,
Feature activation+0.330
they
Token they
Feature activation+0.559
are
Token are
Feature activation+0.834
a
Token a
Feature activation+0.801
-
Token-
Feature activation+0.018
ch
Tokench
Feature activation+0.000
ang
Tokenang
Feature activation+0.000
in
Tokenin
Feature activation+0.174
.
Token.
Feature activation+0.000
But
Token But
Feature activation+0.000
of
Token of
Feature activation+0.000
course
Token course
Feature activation+0.016
,
Token,
Feature activation+1.102
for
Token for
Feature activation+0.753
far
Token far
Feature activation+0.613
too
Token too
Feature activation+0.465
many
Token many
Feature activation+0.415
at
Token at
Feature activation+0.000
Git
Token Git
Feature activation+0.000

INTERVAL 0.339 - 0.677
CONTAINS 0.038%

be
Token be
Feature activation+0.885
able
Token able
Feature activation+1.443
to
Token to
Feature activation+1.092
do
Token do
Feature activation+0.915
much
Token much
Feature activation+0.781
more
Token more
Feature activation+0.625
.
Token.
Feature activation+1.504
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
This
TokenThis
Feature activation+0.000
vertical
Token vertical
Feature activation+0.000
interactive
Token interactive
Feature activation+0.000
possibilities
Token possibilities
Feature activation+0.000
.
Token.
Feature activation+0.000
Of
Token Of
Feature activation+0.000
course
Token course
Feature activation+0.000
,
Token,
Feature activation+0.568
that
Token that
Feature activation+0.837
reality
Token reality
Feature activation+0.421
is
Token is
Feature activation+2.030
likely
Token likely
Feature activation+1.288
still
Token still
Feature activation+1.110
,
Token,
Feature activation+0.967
it
Token it
Feature activation+1.132
was
Token was
Feature activation+1.588
an
Token an
Feature activation+1.391
uphill
Token uphill
Feature activation+0.343
struggle
Token struggle
Feature activation+0.675
:
Token:
Feature activation+1.787
life
Token life
Feature activation+0.000
gets
Token gets
Feature activation+0.000
bus
Token bus
Feature activation+0.000
ier
Tokenier
Feature activation+0.000
have
Token have
Feature activation+0.361
any
Token any
Feature activation+0.000
questions
Token questions
Feature activation+0.000
,
Token,
Feature activation+0.000
I
Token I
Feature activation+0.344
am
Token am
Feature activation+0.543
waiting
Token waiting
Feature activation+0.551
for
Token for
Feature activation+0.000
comments
Token comments
Feature activation+0.000
,
Token,
Feature activation+0.000
I
Token I
Feature activation+0.000
advertisements
Token advertisements
Feature activation+0.760
have
Token have
Feature activation+0.147
taken
Token taken
Feature activation+0.045
place
Token place
Feature activation+0.019
and
Token and
Feature activation+0.906
candidates
Token candidates
Feature activation+0.355
are
Token are
Feature activation+0.993
still
Token still
Feature activation+0.348
getting
Token getting
Feature activation+0.000
into
Token into
Feature activation+0.000
the
Token the
Feature activation+0.000

INTERVAL 0.000 - 0.339
CONTAINS 99.918%

S
TokenS
Feature activation+0.000
V
TokenV
Feature activation+0.000
N
TokenN
Feature activation+0.000
9
Token 9
Feature activation+0.000
36
Token36
Feature activation+0.000
M
TokenM
Feature activation+0.000
)
Token)
Feature activation+0.000
Tue
Token Tue
Feature activation+0.000
Sep
Token Sep
Feature activation+0.000
24
Token 24
Feature activation+0.000
03
Token 03
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
when
Token when
Feature activation+0.000
j
Token j
Feature activation+0.000
oc
Tokenoc
Feature activation+0.000
keys
Tokenkeys
Feature activation+0.000
Dare
Token Dare
Feature activation+0.000
k
Tokenk
Feature activation+0.000
M
Token M
Feature activation+0.000
azz
Tokenazz
Feature activation+0.000
one
Tokenone
Feature activation+0.000
of
Token of
Feature activation+0.000
and
Token and
Feature activation+0.000
fund
Token fund
Feature activation+0.000
a
Token a
Feature activation+0.000
recount
Token recount
Feature activation+0.000
attempt
Token attempt
Feature activation+0.000
that
Token that
Feature activation+0.000
only
Token only
Feature activation+0.000
serves
Token serves
Feature activation+0.000
the
Token the
Feature activation+0.000
interest
Token interest
Feature activation+0.000
of
Token of
Feature activation+0.000
who
Token who
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
the
Token the
Feature activation+0.000
real
Token real
Feature activation+0.000
racist
Token racist
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
2011
Token2011
Feature activation+0.000
:
Token:
Feature activation+0.000
17
Token 17
Feature activation+0.000
.
Token.
Feature activation+0.000
9
Token9
Feature activation+0.000
%
Token%
Feature activation+0.000
(
Token (
Feature activation+0.000
11
Token11
Feature activation+0.000
th
Tokenth
Feature activation+0.000
)
Token)
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 4 in H5.3: (feature 35753)

TOP ACTIVATIONS
MAX = 4.660

Talk
Token Talk
Feature activation+0.855
show
Token show
Feature activation+2.621
host
Token host
Feature activation+4.485
Stephen
Token Stephen
Feature activation+0.673
Colbert
Token Colbert
Feature activation+3.991
isn
Token isn
Feature activation+4.660
âĢ
TokenâĢ
Feature activation+1.559
Ļ
TokenĻ
Feature activation+0.205
t
Tokent
Feature activation+0.803
backing
Token backing
Feature activation+0.771
down
Token down
Feature activation+0.282
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
CLOSE
TokenCLOSE
Feature activation+0.000
Talk
Token Talk
Feature activation+0.855
show
Token show
Feature activation+2.621
host
Token host
Feature activation+4.485
Stephen
Token Stephen
Feature activation+0.673
Colbert
Token Colbert
Feature activation+3.991
isn
Token isn
Feature activation+4.660
âĢ
TokenâĢ
Feature activation+1.559
Ļ
TokenĻ
Feature activation+0.205
"
Token "
Feature activation+0.000
would
Tokenwould
Feature activation+0.000
do
Token do
Feature activation+0.000
it
Token it
Feature activation+0.000
again
Token again
Feature activation+0.000
."
Token."
Feature activation+4.288
USA
Token USA
Feature activation+0.108
TODAY
Token TODAY
Feature activation+0.677
Ċ
TokenĊ
Feature activation+0.479
Ċ
TokenĊ
Feature activation+0.043
F
TokenF
Feature activation+0.000
speech
Token speech
Feature activation+0.000
transcripts
Token transcripts
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
CLOSE
TokenCLOSE
Feature activation+0.000
For
Token For
Feature activation+4.081
the
Token the
Feature activation+3.001
first
Token first
Feature activation+1.255
time
Token time
Feature activation+0.358
,
Token,
Feature activation+0.865
presidential
Token presidential
Feature activation+2.082
CLOSE
TokenCLOSE
Feature activation+0.000
Talk
Token Talk
Feature activation+0.855
show
Token show
Feature activation+2.621
host
Token host
Feature activation+4.485
Stephen
Token Stephen
Feature activation+0.673
Colbert
Token Colbert
Feature activation+3.991
isn
Token isn
Feature activation+4.660
âĢ
TokenâĢ
Feature activation+1.559
Ļ
TokenĻ
Feature activation+0.205
t
Tokent
Feature activation+0.803
backing
Token backing
Feature activation+0.771
are
Token are
Feature activation+1.798
apologizing
Token apologizing
Feature activation+1.191
for
Token for
Feature activation+0.676
the
Token the
Feature activation+0.629
error
Token error
Feature activation+0.635
.
Token.
Feature activation+3.804
Sus
Token Sus
Feature activation+0.300
ana
Tokenana
Feature activation+0.303
Victoria
Token Victoria
Feature activation+0.144
Perez
Token Perez
Feature activation+1.022
(@
Token (@
Feature activation+0.000
's
Token's
Feature activation+1.537
Betty
Token Betty
Feature activation+0.494
Cock
Token Cock
Feature activation+0.290
rum
Tokenrum
Feature activation+1.483
talks
Token talks
Feature activation+2.392
about
Token about
Feature activation+3.444
the
Token the
Feature activation+2.912
changes
Token changes
Feature activation+2.132
she
Token she
Feature activation+1.215
's
Token's
Feature activation+1.374
seen
Token seen
Feature activation+0.760
video
Token video
Feature activation+1.052
of
Token of
Feature activation+1.124
him
Token him
Feature activation+0.199
from
Token from
Feature activation+0.523
2005
Token 2005
Feature activation+0.904
.
Token.
Feature activation+3.079
In
Token In
Feature activation+2.173
the
Token the
Feature activation+1.336
video
Token video
Feature activation+0.911
,
Token,
Feature activation+0.122
Trump
Token Trump
Feature activation+0.517
time
Token time
Feature activation+0.358
,
Token,
Feature activation+0.865
presidential
Token presidential
Feature activation+2.082
candidates
Token candidates
Feature activation+2.938
Hillary
Token Hillary
Feature activation+1.732
Clinton
Token Clinton
Feature activation+3.048
and
Token and
Feature activation+2.033
Bernie
Token Bernie
Feature activation+0.847
Sanders
Token Sanders
Feature activation+1.953
faced
Token faced
Feature activation+0.914
off
Token off
Feature activation+0.304
other
Token other
Feature activation+0.323
during
Token during
Feature activation+0.940
a
Token a
Feature activation+1.097
Democratic
Token Democratic
Feature activation+0.960
debate
Token debate
Feature activation+1.000
.
Token.
Feature activation+3.013
USA
Token USA
Feature activation+0.286
TODAY
Token TODAY
Feature activation+0.563
Ċ
TokenĊ
Feature activation+1.329
Ċ
TokenĊ
Feature activation+0.347
Wall
TokenWall
Feature activation+0.000
transcripts
Token transcripts
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
CLOSE
TokenCLOSE
Feature activation+0.000
For
Token For
Feature activation+4.081
the
Token the
Feature activation+3.001
first
Token first
Feature activation+1.255
time
Token time
Feature activation+0.358
,
Token,
Feature activation+0.865
presidential
Token presidential
Feature activation+2.082
candidates
Token candidates
Feature activation+2.938
the
Token the
Feature activation+3.001
first
Token first
Feature activation+1.255
time
Token time
Feature activation+0.358
,
Token,
Feature activation+0.865
presidential
Token presidential
Feature activation+2.082
candidates
Token candidates
Feature activation+2.938
Hillary
Token Hillary
Feature activation+1.732
Clinton
Token Clinton
Feature activation+3.048
and
Token and
Feature activation+2.033
Bernie
Token Bernie
Feature activation+0.847
Sanders
Token Sanders
Feature activation+1.953
Betty
Token Betty
Feature activation+0.494
Cock
Token Cock
Feature activation+0.290
rum
Tokenrum
Feature activation+1.483
talks
Token talks
Feature activation+2.392
about
Token about
Feature activation+3.444
the
Token the
Feature activation+2.912
changes
Token changes
Feature activation+2.132
she
Token she
Feature activation+1.215
's
Token's
Feature activation+1.374
seen
Token seen
Feature activation+0.760
over
Token over
Feature activation+0.235
tree
Token tree
Feature activation+0.335
looking
Token looking
Feature activation+0.000
fresh
Token fresh
Feature activation+0.000
throughout
Token throughout
Feature activation+0.208
December
Token December
Feature activation+1.342
.
Token.
Feature activation+2.847
Alicia
Token Alicia
Feature activation+1.504
St
Token St
Feature activation+0.734
ice
Tokenice
Feature activation+0.410
Ċ
TokenĊ
Feature activation+0.201
Ċ
TokenĊ
Feature activation+0.000
her
Token her
Feature activation+0.322
plans
Token plans
Feature activation+1.146
for
Token for
Feature activation+0.000
the
Token the
Feature activation+0.000
future
Token future
Feature activation+0.225
.
Token.
Feature activation+2.832
Kelly
Token Kelly
Feature activation+0.296
Wilkinson
Token Wilkinson
Feature activation+1.217
/
Token/
Feature activation+1.077
Ind
TokenInd
Feature activation+0.819
y
Tokeny
Feature activation+0.000
member
Token member
Feature activation+0.000
states
Token states
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
CLOSE
TokenCLOSE
Feature activation+0.000
Out
Token Out
Feature activation+2.831
going
Tokengoing
Feature activation+0.961
Indiana
Token Indiana
Feature activation+0.000
and
Token and
Feature activation+0.154
Kentucky
Token Kentucky
Feature activation+0.589
Planned
Token Planned
Feature activation+0.886
/
Token/
Feature activation+0.000
Getty
TokenGetty
Feature activation+0.000
Images
Token Images
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
CLOSE
TokenCLOSE
Feature activation+0.000
This
Token This
Feature activation+2.646
million
Token million
Feature activation+0.082
dollar
Token dollar
Feature activation+1.632
t
Token t
Feature activation+0.000
-
Token-
Feature activation+0.000
shirt
Tokenshirt
Feature activation+1.257
turn
Token turn
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
CLOSE
TokenCLOSE
Feature activation+0.000
Talk
Token Talk
Feature activation+0.855
show
Token show
Feature activation+2.621
host
Token host
Feature activation+4.485
Stephen
Token Stephen
Feature activation+0.673
Colbert
Token Colbert
Feature activation+3.991
isn
Token isn
Feature activation+4.660
âĢ
TokenâĢ
Feature activation+1.559
us
Token us
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
CLOSE
TokenCLOSE
Feature activation+0.000
Donald
Token Donald
Feature activation+1.983
Trump
Token Trump
Feature activation+2.583
is
Token is
Feature activation+2.168
facing
Token facing
Feature activation+0.000
severe
Token severe
Feature activation+0.157
backlash
Token backlash
Feature activation+0.221
after
Token after
Feature activation+2.083
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
CLOSE
TokenCLOSE
Feature activation+0.000
Fox
Token Fox
Feature activation+0.856
News
Token News
Feature activation+0.728
's
Token's
Feature activation+2.472
Brit
Token Brit
Feature activation+0.011
Hume
Token Hume
Feature activation+1.099
generated
Token generated
Feature activation+1.810
controversy
Token controversy
Feature activation+1.102
with
Token with
Feature activation+1.674

Top DFA by src position
MAX = 6.049

units
Token units
Feature activation+0.000
Top resid features:
per
Token per
Feature activation-0.000
Top resid features:
turn
Token turn
Feature activation-0.001
Top resid features:
.
Token.
Feature activation-0.001
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.447
Top resid features:
CLOSE
TokenCLOSE
Feature activation+6.049
Top resid features:
Talk
Token Talk
Feature activation+0.191
Top resid features:
show
Token show
Feature activation-0.010
Top resid features:
host
Token host
Feature activation-0.106
Top resid features:
Stephen
Token Stephen
Feature activation+0.038
Top resid features:
Colbert
Token Colbert
Feature activation+0.073
Top resid features:
units
Token units
Feature activation+0.000
Top resid features:
per
Token per
Feature activation-0.001
Top resid features:
turn
Token turn
Feature activation-0.001
Top resid features:
.
Token.
Feature activation-0.003
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.896
Top resid features:
CLOSE
TokenCLOSE
Feature activation+5.605
Top resid features:
Talk
Token Talk
Feature activation+0.151
Top resid features:
show
Token show
Feature activation+0.096
Top resid features:
host
Token host
Feature activation+0.109
Top resid features:
Stephen
Token Stephen
Feature activation+0.000
Top resid features:
Colbert
Token Colbert
Feature activation+0.000
Top resid features:
units
Token units
Feature activation-0.000
Top resid features:
per
Token per
Feature activation+0.000
Top resid features:
turn
Token turn
Feature activation+0.000
Top resid features:
.
Token.
Feature activation-0.004
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.304
Top resid features:
CLOSE
TokenCLOSE
Feature activation+5.553
Top resid features:
Talk
Token Talk
Feature activation+0.092
Top resid features:
show
Token show
Feature activation-0.010
Top resid features:
host
Token host
Feature activation-0.004
Top resid features:
Stephen
Token Stephen
Feature activation+0.019
Top resid features:
Colbert
Token Colbert
Feature activation+0.142
Top resid features:
paid
Token paid
Feature activation+0.083
Top resid features:
speech
Token speech
Feature activation+0.095
Top resid features:
transcripts
Token transcripts
Feature activation+0.044
Top resid features:
Ċ
TokenĊ
Feature activation+0.023
Top resid features:
Ċ
TokenĊ
Feature activation+0.104
Top resid features:
CLOSE
TokenCLOSE
Feature activation+5.491
Top resid features:
For
Token For
Feature activation+0.139
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
first
Token first
Feature activation+0.000
Top resid features:
time
Token time
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
units
Token units
Feature activation+0.003
Top resid features:
per
Token per
Feature activation-0.001
Top resid features:
turn
Token turn
Feature activation-0.000
Top resid features:
.
Token.
Feature activation-0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.726
Top resid features:
CLOSE
TokenCLOSE
Feature activation+5.262
Top resid features:
Talk
Token Talk
Feature activation+0.250
Top resid features:
show
Token show
Feature activation+0.016
Top resid features:
host
Token host
Feature activation-0.169
Top resid features:
Stephen
Token Stephen
Feature activation+0.006
Top resid features:
Colbert
Token Colbert
Feature activation+0.096
Top resid features:
Bennett
Token Bennett
Feature activation+0.005
Top resid features:
/
Token/
Feature activation-0.002
Top resid features:
Getty
TokenGetty
Feature activation-0.012
Top resid features:
Images
Token Images
Feature activation+0.011
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.257
Top resid features:
CLOSE
TokenCLOSE
Feature activation+4.854
Top resid features:
This
Token This
Feature activation+0.051
Top resid features:
million
Token million
Feature activation+0.003
Top resid features:
dollar
Token dollar
Feature activation+0.002
Top resid features:
t
Token t
Feature activation+0.020
Top resid features:
-
Token-
Feature activation+0.010
Top resid features:
its
Token its
Feature activation-0.000
Top resid features:
member
Token member
Feature activation-0.000
Top resid features:
states
Token states
Feature activation-0.001
Top resid features:
.
Token.
Feature activation-0.001
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.352
Top resid features:
CLOSE
TokenCLOSE
Feature activation+4.387
Top resid features:
Out
Token Out
Feature activation+0.046
Top resid features:
going
Tokengoing
Feature activation+0.010
Top resid features:
Indiana
Token Indiana
Feature activation+0.123
Top resid features:
and
Token and
Feature activation+0.027
Top resid features:
Kentucky
Token Kentucky
Feature activation+0.045
Top resid features:
thinks
Token thinks
Feature activation+0.002
Top resid features:
of
Token of
Feature activation+0.001
Top resid features:
us
Token us
Feature activation-0.001
Top resid features:
Ċ
TokenĊ
Feature activation-0.005
Top resid features:
Ċ
TokenĊ
Feature activation+0.042
Top resid features:
CLOSE
TokenCLOSE
Feature activation+4.040
Top resid features:
Donald
Token Donald
Feature activation+0.124
Top resid features:
Trump
Token Trump
Feature activation+0.087
Top resid features:
is
Token is
Feature activation+0.029
Top resid features:
facing
Token facing
Feature activation+0.020
Top resid features:
severe
Token severe
Feature activation+0.007
Top resid features:
paid
Token paid
Feature activation-0.004
Top resid features:
speech
Token speech
Feature activation+0.037
Top resid features:
transcripts
Token transcripts
Feature activation+0.030
Top resid features:
Ċ
TokenĊ
Feature activation+0.025
Top resid features:
Ċ
TokenĊ
Feature activation+0.096
Top resid features:
CLOSE
TokenCLOSE
Feature activation+3.971
Top resid features:
For
Token For
Feature activation+0.176
Top resid features:
the
Token the
Feature activation+0.049
Top resid features:
first
Token first
Feature activation+0.039
Top resid features:
time
Token time
Feature activation+0.051
Top resid features:
,
Token,
Feature activation+0.076
Top resid features:
paid
Token paid
Feature activation+0.001
Top resid features:
speech
Token speech
Feature activation+0.007
Top resid features:
transcripts
Token transcripts
Feature activation+0.002
Top resid features:
Ċ
TokenĊ
Feature activation+0.019
Top resid features:
Ċ
TokenĊ
Feature activation+0.055
Top resid features:
CLOSE
TokenCLOSE
Feature activation+3.835
Top resid features:
For
Token For
Feature activation+0.167
Top resid features:
the
Token the
Feature activation+0.025
Top resid features:
first
Token first
Feature activation+0.013
Top resid features:
time
Token time
Feature activation+0.028
Top resid features:
,
Token,
Feature activation+0.028
Top resid features:
paid
Token paid
Feature activation+0.027
Top resid features:
speech
Token speech
Feature activation+0.065
Top resid features:
transcripts
Token transcripts
Feature activation+0.025
Top resid features:
Ċ
TokenĊ
Feature activation+0.024
Top resid features:
Ċ
TokenĊ
Feature activation+0.079
Top resid features:
CLOSE
TokenCLOSE
Feature activation+3.652
Top resid features:
For
Token For
Feature activation+0.709
Top resid features:
the
Token the
Feature activation+0.123
Top resid features:
first
Token first
Feature activation+0.000
Top resid features:
time
Token time
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
paid
Token paid
Feature activation-0.004
Top resid features:
speech
Token speech
Feature activation+0.061
Top resid features:
transcripts
Token transcripts
Feature activation+0.026
Top resid features:
Ċ
TokenĊ
Feature activation+0.035
Top resid features:
Ċ
TokenĊ
Feature activation+0.097
Top resid features:
CLOSE
TokenCLOSE
Feature activation+4.063
Top resid features:
For
Token For
Feature activation+0.217
Top resid features:
the
Token the
Feature activation+0.061
Top resid features:
first
Token first
Feature activation+0.041
Top resid features:
time
Token time
Feature activation+0.139
Top resid features:
,
Token,
Feature activation+0.257
Top resid features:
its
Token its
Feature activation+0.000
Top resid features:
member
Token member
Feature activation-0.000
Top resid features:
states
Token states
Feature activation-0.002
Top resid features:
.
Token.
Feature activation+0.001
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.348
Top resid features:
CLOSE
TokenCLOSE
Feature activation+3.783
Top resid features:
Out
Token Out
Feature activation+0.041
Top resid features:
going
Tokengoing
Feature activation+0.005
Top resid features:
Indiana
Token Indiana
Feature activation+0.062
Top resid features:
and
Token and
Feature activation+0.016
Top resid features:
Kentucky
Token Kentucky
Feature activation+0.020
Top resid features:
ready
Token ready
Feature activation+0.001
Top resid features:
.
Token.
Feature activation-0.002
Top resid features:
âĢ
TokenâĢ
Feature activation-0.001
Top resid features:
Ŀ
TokenĿ
Feature activation+0.004
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.289
Top resid features:
CLOSE
TokenCLOSE
Feature activation+3.448
Top resid features:
We
Token We
Feature activation+0.168
Top resid features:
talked
Token talked
Feature activation+0.165
Top resid features:
to
Token to
Feature activation+0.131
Top resid features:
experts
Token experts
Feature activation+0.042
Top resid features:
and
Token and
Feature activation+0.048
Top resid features:
its
Token its
Feature activation+0.000
Top resid features:
member
Token member
Feature activation-0.000
Top resid features:
states
Token states
Feature activation-0.001
Top resid features:
.
Token.
Feature activation-0.002
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.231
Top resid features:
CLOSE
TokenCLOSE
Feature activation+3.805
Top resid features:
Out
Token Out
Feature activation+0.058
Top resid features:
going
Tokengoing
Feature activation-0.029
Top resid features:
Indiana
Token Indiana
Feature activation+0.072
Top resid features:
and
Token and
Feature activation+0.030
Top resid features:
Kentucky
Token Kentucky
Feature activation+0.015
Top resid features:
its
Token its
Feature activation+0.002
Top resid features:
member
Token member
Feature activation+0.000
Top resid features:
states
Token states
Feature activation-0.002
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.536
Top resid features:
CLOSE
TokenCLOSE
Feature activation+4.063
Top resid features:
Out
Token Out
Feature activation+0.081
Top resid features:
going
Tokengoing
Feature activation+0.000
Top resid features:
Indiana
Token Indiana
Feature activation+0.000
Top resid features:
and
Token and
Feature activation+0.000
Top resid features:
Kentucky
Token Kentucky
Feature activation+0.000
Top resid features:
Bennett
Token Bennett
Feature activation+0.004
Top resid features:
/
Token/
Feature activation+0.000
Top resid features:
Getty
TokenGetty
Feature activation+0.007
Top resid features:
Images
Token Images
Feature activation+0.027
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.451
Top resid features:
CLOSE
TokenCLOSE
Feature activation+4.033
Top resid features:
This
Token This
Feature activation+0.042
Top resid features:
million
Token million
Feature activation+0.000
Top resid features:
dollar
Token dollar
Feature activation+0.000
Top resid features:
t
Token t
Feature activation+0.000
Top resid features:
-
Token-
Feature activation+0.000
Top resid features:
units
Token units
Feature activation+0.001
Top resid features:
per
Token per
Feature activation+0.000
Top resid features:
turn
Token turn
Feature activation+0.000
Top resid features:
.
Token.
Feature activation-0.002
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.704
Top resid features:
CLOSE
TokenCLOSE
Feature activation+3.570
Top resid features:
Talk
Token Talk
Feature activation+0.158
Top resid features:
show
Token show
Feature activation+0.057
Top resid features:
host
Token host
Feature activation+0.000
Top resid features:
Stephen
Token Stephen
Feature activation+0.000
Top resid features:
Colbert
Token Colbert
Feature activation+0.000
Top resid features:
thinks
Token thinks
Feature activation+0.395
Top resid features:
of
Token of
Feature activation+0.018
Top resid features:
us
Token us
Feature activation-0.023
Top resid features:
Ċ
TokenĊ
Feature activation-0.074
Top resid features:
Ċ
TokenĊ
Feature activation+0.015
Top resid features:
CLOSE
TokenCLOSE
Feature activation+4.405
Top resid features:
Donald
Token Donald
Feature activation+0.287
Top resid features:
Trump
Token Trump
Feature activation+0.046
Top resid features:
is
Token is
Feature activation+0.000
Top resid features:
facing
Token facing
Feature activation+0.000
Top resid features:
severe
Token severe
Feature activation+0.000
Top resid features:
necessarily
Token necessarily
Feature activation+0.001
Top resid features:
attractive
Token attractive
Feature activation-0.000
Top resid features:
'
Token'
Feature activation+0.008
Top resid features:
Ċ
TokenĊ
Feature activation-0.017
Top resid features:
Ċ
TokenĊ
Feature activation+0.004
Top resid features:
CLOSE
TokenCLOSE
Feature activation+4.010
Top resid features:
Fox
Token Fox
Feature activation+0.191
Top resid features:
News
Token News
Feature activation+0.274
Top resid features:
's
Token's
Feature activation+0.054
Top resid features:
Brit
Token Brit
Feature activation+0.000
Top resid features:
Hume
Token Hume
Feature activation+0.000
Top resid features:

Decoder Weights Distribution

Head 0: 0.02

Head 1: 0.02

Head 2: 0.06

Head 3: 0.50

Head 4: 0.05

Head 5: 0.04

Head 6: 0.03

Head 7: 0.03

Head 8: 0.05

Head 9: 0.07

Head 10: 0.07

Head 11: 0.07

Positive logits

Replay2.01

avanaugh1.72

?'1.62

OTUS1.54

awaits1.54

?:1.51

reacts1.48

TIM1.47

waits1.45

gamer1.43

?'"1.40

watchdog1.40

tweets1.39

NYT1.37

DEM1.37

unve1.37

?]1.36

guiName1.35

]'1.35

MUST1.35

Negative logits

"),-1.98

laughs-1.98

��極-1.86

ascript-1.83

Newsletter-1.82

equivalents-1.81

withstanding-1.81

SourceFile-1.81

ItemThumbnailImage-1.79

etheless-1.78

Laughs-1.73

Anyway-1.71

orically-1.66

laugh-1.65

EStream-1.65

luaj-1.64

viation-1.63

prisingly-1.63

cffffcc-1.60

").-1.60

INTERVAL 4.194 - 4.660
CONTAINS 0.000%

.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
CLOSE
TokenCLOSE
Feature activation+0.000
Talk
Token Talk
Feature activation+0.855
show
Token show
Feature activation+2.621
host
Token host
Feature activation+4.485
Stephen
Token Stephen
Feature activation+0.673
Colbert
Token Colbert
Feature activation+3.991
isn
Token isn
Feature activation+4.660
âĢ
TokenâĢ
Feature activation+1.559
Ļ
TokenĻ
Feature activation+0.205
Talk
Token Talk
Feature activation+0.855
show
Token show
Feature activation+2.621
host
Token host
Feature activation+4.485
Stephen
Token Stephen
Feature activation+0.673
Colbert
Token Colbert
Feature activation+3.991
isn
Token isn
Feature activation+4.660
âĢ
TokenâĢ
Feature activation+1.559
Ļ
TokenĻ
Feature activation+0.205
t
Tokent
Feature activation+0.803
backing
Token backing
Feature activation+0.771
down
Token down
Feature activation+0.282
"
Token "
Feature activation+0.000
would
Tokenwould
Feature activation+0.000
do
Token do
Feature activation+0.000
it
Token it
Feature activation+0.000
again
Token again
Feature activation+0.000
."
Token."
Feature activation+4.288
USA
Token USA
Feature activation+0.108
TODAY
Token TODAY
Feature activation+0.677
Ċ
TokenĊ
Feature activation+0.479
Ċ
TokenĊ
Feature activation+0.043
F
TokenF
Feature activation+0.000

INTERVAL 3.728 - 4.194
CONTAINS 0.000%

are
Token are
Feature activation+1.798
apologizing
Token apologizing
Feature activation+1.191
for
Token for
Feature activation+0.676
the
Token the
Feature activation+0.629
error
Token error
Feature activation+0.635
.
Token.
Feature activation+3.804
Sus
Token Sus
Feature activation+0.300
ana
Tokenana
Feature activation+0.303
Victoria
Token Victoria
Feature activation+0.144
Perez
Token Perez
Feature activation+1.022
(@
Token (@
Feature activation+0.000
speech
Token speech
Feature activation+0.000
transcripts
Token transcripts
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
CLOSE
TokenCLOSE
Feature activation+0.000
For
Token For
Feature activation+4.081
the
Token the
Feature activation+3.001
first
Token first
Feature activation+1.255
time
Token time
Feature activation+0.358
,
Token,
Feature activation+0.865
presidential
Token presidential
Feature activation+2.082
CLOSE
TokenCLOSE
Feature activation+0.000
Talk
Token Talk
Feature activation+0.855
show
Token show
Feature activation+2.621
host
Token host
Feature activation+4.485
Stephen
Token Stephen
Feature activation+0.673
Colbert
Token Colbert
Feature activation+3.991
isn
Token isn
Feature activation+4.660
âĢ
TokenâĢ
Feature activation+1.559
Ļ
TokenĻ
Feature activation+0.205
t
Tokent
Feature activation+0.803
backing
Token backing
Feature activation+0.771

INTERVAL 3.262 - 3.728
CONTAINS 0.000%

's
Token's
Feature activation+1.537
Betty
Token Betty
Feature activation+0.494
Cock
Token Cock
Feature activation+0.290
rum
Tokenrum
Feature activation+1.483
talks
Token talks
Feature activation+2.392
about
Token about
Feature activation+3.444
the
Token the
Feature activation+2.912
changes
Token changes
Feature activation+2.132
she
Token she
Feature activation+1.215
's
Token's
Feature activation+1.374
seen
Token seen
Feature activation+0.760

INTERVAL 2.796 - 3.262
CONTAINS 0.001%

other
Token other
Feature activation+0.323
during
Token during
Feature activation+0.940
a
Token a
Feature activation+1.097
Democratic
Token Democratic
Feature activation+0.960
debate
Token debate
Feature activation+1.000
.
Token.
Feature activation+3.013
USA
Token USA
Feature activation+0.286
TODAY
Token TODAY
Feature activation+0.563
Ċ
TokenĊ
Feature activation+1.329
Ċ
TokenĊ
Feature activation+0.347
Wall
TokenWall
Feature activation+0.000
transcripts
Token transcripts
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
CLOSE
TokenCLOSE
Feature activation+0.000
For
Token For
Feature activation+4.081
the
Token the
Feature activation+3.001
first
Token first
Feature activation+1.255
time
Token time
Feature activation+0.358
,
Token,
Feature activation+0.865
presidential
Token presidential
Feature activation+2.082
candidates
Token candidates
Feature activation+2.938
member
Token member
Feature activation+0.000
states
Token states
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
CLOSE
TokenCLOSE
Feature activation+0.000
Out
Token Out
Feature activation+2.831
going
Tokengoing
Feature activation+0.961
Indiana
Token Indiana
Feature activation+0.000
and
Token and
Feature activation+0.154
Kentucky
Token Kentucky
Feature activation+0.589
Planned
Token Planned
Feature activation+0.886
the
Token the
Feature activation+3.001
first
Token first
Feature activation+1.255
time
Token time
Feature activation+0.358
,
Token,
Feature activation+0.865
presidential
Token presidential
Feature activation+2.082
candidates
Token candidates
Feature activation+2.938
Hillary
Token Hillary
Feature activation+1.732
Clinton
Token Clinton
Feature activation+3.048
and
Token and
Feature activation+2.033
Bernie
Token Bernie
Feature activation+0.847
Sanders
Token Sanders
Feature activation+1.953
video
Token video
Feature activation+1.052
of
Token of
Feature activation+1.124
him
Token him
Feature activation+0.199
from
Token from
Feature activation+0.523
2005
Token 2005
Feature activation+0.904
.
Token.
Feature activation+3.079
In
Token In
Feature activation+2.173
the
Token the
Feature activation+1.336
video
Token video
Feature activation+0.911
,
Token,
Feature activation+0.122
Trump
Token Trump
Feature activation+0.517

INTERVAL 2.330 - 2.796
CONTAINS 0.001%

Parenthood
Token Parenthood
Feature activation+1.079
's
Token's
Feature activation+1.537
Betty
Token Betty
Feature activation+0.494
Cock
Token Cock
Feature activation+0.290
rum
Tokenrum
Feature activation+1.483
talks
Token talks
Feature activation+2.392
about
Token about
Feature activation+3.444
the
Token the
Feature activation+2.912
changes
Token changes
Feature activation+2.132
she
Token she
Feature activation+1.215
's
Token's
Feature activation+1.374
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
CLOSE
TokenCLOSE
Feature activation+0.000
Fox
Token Fox
Feature activation+0.856
News
Token News
Feature activation+0.728
's
Token's
Feature activation+2.472
Brit
Token Brit
Feature activation+0.011
Hume
Token Hume
Feature activation+1.099
generated
Token generated
Feature activation+1.810
controversy
Token controversy
Feature activation+1.102
with
Token with
Feature activation+1.674
us
Token us
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
CLOSE
TokenCLOSE
Feature activation+0.000
Donald
Token Donald
Feature activation+1.983
Trump
Token Trump
Feature activation+2.583
is
Token is
Feature activation+2.168
facing
Token facing
Feature activation+0.000
severe
Token severe
Feature activation+0.157
backlash
Token backlash
Feature activation+0.221
after
Token after
Feature activation+2.083
/
Token/
Feature activation+0.000
Getty
TokenGetty
Feature activation+0.000
Images
Token Images
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
CLOSE
TokenCLOSE
Feature activation+0.000
This
Token This
Feature activation+2.646
million
Token million
Feature activation+0.082
dollar
Token dollar
Feature activation+1.632
t
Token t
Feature activation+0.000
-
Token-
Feature activation+0.000
shirt
Tokenshirt
Feature activation+1.257
turn
Token turn
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
CLOSE
TokenCLOSE
Feature activation+0.000
Talk
Token Talk
Feature activation+0.855
show
Token show
Feature activation+2.621
host
Token host
Feature activation+4.485
Stephen
Token Stephen
Feature activation+0.673
Colbert
Token Colbert
Feature activation+3.991
isn
Token isn
Feature activation+4.660
âĢ
TokenâĢ
Feature activation+1.559

INTERVAL 1.864 - 2.330
CONTAINS 0.001%

,
Token,
Feature activation+0.865
presidential
Token presidential
Feature activation+2.082
candidates
Token candidates
Feature activation+2.938
Hillary
Token Hillary
Feature activation+1.732
Clinton
Token Clinton
Feature activation+3.048
and
Token and
Feature activation+2.033
Bernie
Token Bernie
Feature activation+0.847
Sanders
Token Sanders
Feature activation+1.953
faced
Token faced
Feature activation+0.914
off
Token off
Feature activation+0.304
with
Token with
Feature activation+1.401
.
Token.
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
CLOSE
TokenCLOSE
Feature activation+0.000
We
Token We
Feature activation+2.100
talked
Token talked
Feature activation+1.383
to
Token to
Feature activation+1.755
experts
Token experts
Feature activation+1.817
and
Token and
Feature activation+0.601
got
Token got
Feature activation+0.628
Cock
Token Cock
Feature activation+0.290
rum
Tokenrum
Feature activation+1.483
talks
Token talks
Feature activation+2.392
about
Token about
Feature activation+3.444
the
Token the
Feature activation+2.912
changes
Token changes
Feature activation+2.132
she
Token she
Feature activation+1.215
's
Token's
Feature activation+1.374
seen
Token seen
Feature activation+0.760
over
Token over
Feature activation+0.235
15
Token 15
Feature activation+0.720
candidates
Token candidates
Feature activation+2.938
Hillary
Token Hillary
Feature activation+1.732
Clinton
Token Clinton
Feature activation+3.048
and
Token and
Feature activation+2.033
Bernie
Token Bernie
Feature activation+0.847
Sanders
Token Sanders
Feature activation+1.953
faced
Token faced
Feature activation+0.914
off
Token off
Feature activation+0.304
with
Token with
Feature activation+1.401
just
Token just
Feature activation+0.542
each
Token each
Feature activation+0.908
For
Token For
Feature activation+4.081
the
Token the
Feature activation+3.001
first
Token first
Feature activation+1.255
time
Token time
Feature activation+0.358
,
Token,
Feature activation+0.865
presidential
Token presidential
Feature activation+2.082
candidates
Token candidates
Feature activation+2.938
Hillary
Token Hillary
Feature activation+1.732
Clinton
Token Clinton
Feature activation+3.048
and
Token and
Feature activation+2.033
Bernie
Token Bernie
Feature activation+0.847

INTERVAL 1.398 - 1.864
CONTAINS 0.002%

from
Token from
Feature activation+0.087
all
Token all
Feature activation+0.000
of
Token of
Feature activation+0.895
the
Token the
Feature activation+0.995
criticism
Token criticism
Feature activation+0.631
regarding
Token regarding
Feature activation+1.524
his
Token his
Feature activation+1.088
rant
Token rant
Feature activation+1.220
on
Token on
Feature activation+0.199
President
Token President
Feature activation+0.900
Donald
Token Donald
Feature activation+0.289
Ŀ
TokenĿ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
CLOSE
TokenCLOSE
Feature activation+0.000
We
Token We
Feature activation+2.100
talked
Token talked
Feature activation+1.383
to
Token to
Feature activation+1.755
experts
Token experts
Feature activation+1.817
and
Token and
Feature activation+0.601
got
Token got
Feature activation+0.628
their
Token their
Feature activation+1.054
advice
Token advice
Feature activation+2.271
severe
Token severe
Feature activation+0.157
backlash
Token backlash
Feature activation+0.221
after
Token after
Feature activation+2.083
The
Token The
Feature activation+1.244
Washington
Token Washington
Feature activation+0.426
Post
Token Post
Feature activation+1.509
released
Token released
Feature activation+1.596
a
Token a
Feature activation+1.501
video
Token video
Feature activation+1.052
of
Token of
Feature activation+1.124
him
Token him
Feature activation+0.199
after
Token after
Feature activation+2.083
The
Token The
Feature activation+1.244
Washington
Token Washington
Feature activation+0.426
Post
Token Post
Feature activation+1.509
released
Token released
Feature activation+1.596
a
Token a
Feature activation+1.501
video
Token video
Feature activation+1.052
of
Token of
Feature activation+1.124
him
Token him
Feature activation+0.199
from
Token from
Feature activation+0.523
2005
Token 2005
Feature activation+0.904
Images
Token Images
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
CLOSE
TokenCLOSE
Feature activation+0.000
This
Token This
Feature activation+2.646
million
Token million
Feature activation+0.082
dollar
Token dollar
Feature activation+1.632
t
Token t
Feature activation+0.000
-
Token-
Feature activation+0.000
shirt
Tokenshirt
Feature activation+1.257
making
Token making
Feature activation+0.153
company
Token company
Feature activation+1.835

INTERVAL 0.932 - 1.398
CONTAINS 0.003%

backing
Token backing
Feature activation+0.771
down
Token down
Feature activation+0.282
from
Token from
Feature activation+0.087
all
Token all
Feature activation+0.000
of
Token of
Feature activation+0.895
the
Token the
Feature activation+0.995
criticism
Token criticism
Feature activation+0.631
regarding
Token regarding
Feature activation+1.524
his
Token his
Feature activation+1.088
rant
Token rant
Feature activation+1.220
on
Token on
Feature activation+0.199
each
Token each
Feature activation+0.908
other
Token other
Feature activation+0.323
during
Token during
Feature activation+0.940
a
Token a
Feature activation+1.097
Democratic
Token Democratic
Feature activation+0.960
debate
Token debate
Feature activation+1.000
.
Token.
Feature activation+3.013
USA
Token USA
Feature activation+0.286
TODAY
Token TODAY
Feature activation+0.563
Ċ
TokenĊ
Feature activation+1.329
Ċ
TokenĊ
Feature activation+0.347
This
Token This
Feature activation+2.646
million
Token million
Feature activation+0.082
dollar
Token dollar
Feature activation+1.632
t
Token t
Feature activation+0.000
-
Token-
Feature activation+0.000
shirt
Tokenshirt
Feature activation+1.257
making
Token making
Feature activation+0.153
company
Token company
Feature activation+1.835
failed
Token failed
Feature activation+0.938
to
Token to
Feature activation+0.038
catch
Token catch
Feature activation+0.000
all
Token all
Feature activation+0.000
of
Token of
Feature activation+0.895
the
Token the
Feature activation+0.995
criticism
Token criticism
Feature activation+0.631
regarding
Token regarding
Feature activation+1.524
his
Token his
Feature activation+1.088
rant
Token rant
Feature activation+1.220
on
Token on
Feature activation+0.199
President
Token President
Feature activation+0.900
Donald
Token Donald
Feature activation+0.289
Trump
Token Trump
Feature activation+1.464
of
Token of
Feature activation+0.895
the
Token the
Feature activation+0.995
criticism
Token criticism
Feature activation+0.631
regarding
Token regarding
Feature activation+1.524
his
Token his
Feature activation+1.088
rant
Token rant
Feature activation+1.220
on
Token on
Feature activation+0.199
President
Token President
Feature activation+0.900
Donald
Token Donald
Feature activation+0.289
Trump
Token Trump
Feature activation+1.464
the
Token the
Feature activation+0.185

INTERVAL 0.466 - 0.932
CONTAINS 0.004%

now
Token now
Feature activation+1.024
they
Token they
Feature activation+1.527
are
Token are
Feature activation+1.798
apologizing
Token apologizing
Feature activation+1.191
for
Token for
Feature activation+0.676
the
Token the
Feature activation+0.629
error
Token error
Feature activation+0.635
.
Token.
Feature activation+3.804
Sus
Token Sus
Feature activation+0.300
ana
Tokenana
Feature activation+0.303
Victoria
Token Victoria
Feature activation+0.144
and
Token and
Feature activation+0.601
got
Token got
Feature activation+0.628
their
Token their
Feature activation+1.054
advice
Token advice
Feature activation+2.271
for
Token for
Feature activation+0.589
keeping
Token keeping
Feature activation+0.468
your
Token your
Feature activation+0.119
Christmas
Token Christmas
Feature activation+0.275
tree
Token tree
Feature activation+0.335
looking
Token looking
Feature activation+0.000
fresh
Token fresh
Feature activation+0.000
a
Token a
Feature activation+1.097
Democratic
Token Democratic
Feature activation+0.960
debate
Token debate
Feature activation+1.000
.
Token.
Feature activation+3.013
USA
Token USA
Feature activation+0.286
TODAY
Token TODAY
Feature activation+0.563
Ċ
TokenĊ
Feature activation+1.329
Ċ
TokenĊ
Feature activation+0.347
Wall
TokenWall
Feature activation+0.000
Street
Token Street
Feature activation+0.000
,
Token,
Feature activation+0.000
ana
Tokenana
Feature activation+0.000
_
Token_
Feature activation+0.000
vp
Tokenvp
Feature activation+0.000
)
Token)
Feature activation+0.000
has
Token has
Feature activation+0.619
more
Token more
Feature activation+0.523
.
Token.
Feature activation+0.000
Buzz
Token Buzz
Feature activation+0.065
60
Token60
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
and
Token and
Feature activation+0.000
now
Token now
Feature activation+1.024
they
Token they
Feature activation+1.527
are
Token are
Feature activation+1.798
apologizing
Token apologizing
Feature activation+1.191
for
Token for
Feature activation+0.676
the
Token the
Feature activation+0.629
error
Token error
Feature activation+0.635
.
Token.
Feature activation+3.804
Sus
Token Sus
Feature activation+0.300
ana
Tokenana
Feature activation+0.303

INTERVAL 0.000 - 0.466
CONTAINS 99.988%

),
Token),
Feature activation+0.000
cog
Token cog
Feature activation+0.000
ent
Tokenent
Feature activation+0.000
comment
Tokencomment
Feature activation+0.000
.
Token.
Feature activation+0.000
com
Tokencom
Feature activation+0.000
,
Token,
Feature activation+0.000
heavy
Token heavy
Feature activation+0.000
.
Token.
Feature activation+0.000
com
Tokencom
Feature activation+0.000
,
Token,
Feature activation+0.000
John
Token John
Feature activation+0.000
F
Token F
Feature activation+0.000
.
Token.
Feature activation+0.000
Kennedy
Token Kennedy
Feature activation+0.000
were
Token were
Feature activation+0.000
both
Token both
Feature activation+0.000
sons
Token sons
Feature activation+0.000
of
Token of
Feature activation+0.000
privileged
Token privileged
Feature activation+0.000
,
Token,
Feature activation+0.000
politically
Token politically
Feature activation+0.000
President
Token President
Feature activation+0.000
Trump
Token Trump
Feature activation+0.000
's
Token's
Feature activation+0.000
White
Token White
Feature activation+0.000
House
Token House
Feature activation+0.000
on
Token on
Feature activation+0.000
the
Token the
Feature activation+0.000
future
Token future
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
payments
Token payments
Feature activation+0.000
met
Token met
Feature activation+0.000
in
Token in
Feature activation+0.000
an
Token an
Feature activation+0.000
organised
Token organised
Feature activation+0.000
fashion
Token fashion
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Despite
TokenDespite
Feature activation+0.000
their
Token their
Feature activation+0.000
proximity
Token proximity
Feature activation+0.000
the
Token the
Feature activation+0.000
Falcons
Token Falcons
Feature activation+0.000
by
Token by
Feature activation+0.000
about
Token about
Feature activation+0.000
6
Token 6
Feature activation+0.000
inches
Token inches
Feature activation+0.000
--
Token --
Feature activation+0.000
on
Token on
Feature activation+0.000
a
Token a
Feature activation+0.000
day
Token day
Feature activation+0.000
when
Token when
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 H5.3: (feature 36245)

TOP ACTIVATIONS
MAX = 3.604

home
Token home
Feature activation+0.000
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
with
Token with
Feature activation+0.000
MLS
Token MLS
Feature activation+0.547
not
Token not
Feature activation+3.604
looking
Token looking
Feature activation+2.578
like
Token like
Feature activation+1.698
a
Token a
Feature activation+0.000
viable
Token viable
Feature activation+0.145
option
Token option
Feature activation+0.503
millions
Token millions
Feature activation+0.049
of
Token of
Feature activation+0.000
dollars
Token dollars
Feature activation+0.076
of
Token of
Feature activation+0.000
payments
Token payments
Feature activation+0.981
looming
Token looming
Feature activation+3.265
and
Token and
Feature activation+1.887
little
Token little
Feature activation+1.300
cash
Token cash
Feature activation+0.855
on
Token on
Feature activation+1.423
hand
Token hand
Feature activation+2.678
that
Token that
Feature activation+0.180
surround
Token surround
Feature activation+0.000
her
Token her
Feature activation+0.000
property
Token property
Feature activation+0.000
,
Token,
Feature activation+2.228
it
Token it
Feature activation+3.257
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.083
s
Tokens
Feature activation+0.302
about
Token about
Feature activation+0.073
as
Token as
Feature activation+0.000
the
Token the
Feature activation+0.000
next
Token next
Feature activation+0.000
month
Token month
Feature activation+0.367
or
Token or
Feature activation+0.305
two
Token two
Feature activation+0.000
,
Token,
Feature activation+3.191
there
Token there
Feature activation+2.941
could
Token could
Feature activation+1.440
be
Token be
Feature activation+1.400
bills
Token bills
Feature activation+0.000
introduced
Token introduced
Feature activation+0.000
problem
Token problem
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
With
TokenWith
Feature activation+0.000
so
Token so
Feature activation+3.114
many
Token many
Feature activation+1.365
benefits
Token benefits
Feature activation+0.209
to
Token to
Feature activation+0.000
be
Token be
Feature activation+0.111
gained
Token gained
Feature activation+0.316
time
Token time
Feature activation+0.000
.
Token.
Feature activation+0.000
And
Token And
Feature activation+0.000
with
Token with
Feature activation+0.810
Google
Token Google
Feature activation+1.280
helping
Token helping
Feature activation+3.090
fund
Token fund
Feature activation+0.875
its
Token its
Feature activation+0.065
development
Token development
Feature activation+0.369
,
Token,
Feature activation+1.685
we
Token we
Feature activation+0.000
and
Token and
Feature activation+0.000
Alt
Token Alt
Feature activation+0.000
-
Token-
Feature activation+0.000
J
TokenJ
Feature activation+0.000
,
Token,
Feature activation+0.000
it
Token it
Feature activation+3.086
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.179
wouldn
Token wouldn
Feature activation+0.912
âĢ
TokenâĢ
Feature activation+0.000
games
Token games
Feature activation+0.000
.
Token.
Feature activation+0.000
With
Token With
Feature activation+0.000
preparation
Token preparation
Feature activation+1.258
time
Token time
Feature activation+0.831
so
Token so
Feature activation+3.009
short
Token short
Feature activation+1.281
and
Token and
Feature activation+0.152
with
Token with
Feature activation+0.397
no
Token no
Feature activation+1.552
mid
Token mid
Feature activation+0.136
Audi
Token Audi
Feature activation+0.189
,
Token,
Feature activation+0.000
BMW
Token BMW
Feature activation+0.000
and
Token and
Feature activation+0.000
Mercedes
Token Mercedes
Feature activation+0.000
stepping
Token stepping
Feature activation+3.008
up
Token up
Feature activation+0.697
,
Token,
Feature activation+2.523
it
Token it
Feature activation+1.492
will
Token will
Feature activation+0.822
only
Token only
Feature activation+0.744
next
Token next
Feature activation+0.000
month
Token month
Feature activation+0.367
or
Token or
Feature activation+0.305
two
Token two
Feature activation+0.000
,
Token,
Feature activation+3.191
there
Token there
Feature activation+2.941
could
Token could
Feature activation+1.440
be
Token be
Feature activation+1.400
bills
Token bills
Feature activation+0.000
introduced
Token introduced
Feature activation+0.000
in
Token in
Feature activation+0.000
next
Token next
Feature activation+0.000
year
Token year
Feature activation+0.174
's
Token's
Feature activation+0.000
European
Token European
Feature activation+0.000
elections
Token elections
Feature activation+0.000
,
Token,
Feature activation+2.850
the
Token the
Feature activation+1.991
"
Token "
Feature activation+0.300
modern
Tokenmodern
Feature activation+0.000
isers
Tokenisers
Feature activation+0.000
"
Token"
Feature activation+0.000
With
Token With
Feature activation+0.000
the
Token the
Feature activation+0.000
economy
Token economy
Feature activation+0.765
on
Token on
Feature activation+0.734
the
Token the
Feature activation+0.065
rise
Token rise
Feature activation+2.818
,
Token,
Feature activation+1.245
the
Token the
Feature activation+0.037
neighborhood
Token neighborhood
Feature activation+0.000
is
Token is
Feature activation+0.000
reinvent
Token reinvent
Feature activation+0.000
yet
Token yet
Feature activation+2.443
to
Token to
Feature activation+2.384
see
Token see
Feature activation+1.648
a
Token a
Feature activation+0.863
release
Token release
Feature activation+1.631
,
Token,
Feature activation+2.790
it
Token it
Feature activation+1.346
âĢ
TokenâĢ
Feature activation+0.012
Ļ
TokenĻ
Feature activation+0.241
s
Tokens
Feature activation+0.239
hard
Token hard
Feature activation+0.465
eventual
Token eventual
Feature activation+0.824
vaccine
Token vaccine
Feature activation+0.000
against
Token against
Feature activation+0.005
the
Token the
Feature activation+0.000
virus
Token virus
Feature activation+0.237
likely
Token likely
Feature activation+2.781
years
Token years
Feature activation+1.125
away
Token away
Feature activation+2.457
,
Token,
Feature activation+1.984
a
Token a
Feature activation+0.719
widely
Token widely
Feature activation+0.030
votes
Token votes
Feature activation+0.000
as
Token as
Feature activation+0.000
of
Token of
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
night
Token night
Feature activation+0.000
--
Token --
Feature activation+2.737
the
Token the
Feature activation+1.759
issue
Token issue
Feature activation+0.323
of
Token of
Feature activation+0.570
how
Token how
Feature activation+0.000
and
Token and
Feature activation+0.000
el
Tokenel
Feature activation+0.192
retiring
Token retiring
Feature activation+2.240
from
Token from
Feature activation+0.000
the
Token the
Feature activation+0.000
Senate
Token Senate
Feature activation+0.000
,
Token,
Feature activation+2.709
there
Token there
Feature activation+0.678
are
Token are
Feature activation+0.000
few
Token few
Feature activation+0.000
prominent
Token prominent
Feature activation+0.000
Republican
Token Republican
Feature activation+0.000
looming
Token looming
Feature activation+3.265
and
Token and
Feature activation+1.887
little
Token little
Feature activation+1.300
cash
Token cash
Feature activation+0.855
on
Token on
Feature activation+1.423
hand
Token hand
Feature activation+2.678
,
Token,
Feature activation+2.321
PRE
Token PRE
Feature activation+0.000
PA
TokenPA
Feature activation+0.000
has
Token has
Feature activation+0.106
been
Token been
Feature activation+0.000
national
Token national
Feature activation+0.000
targets
Token targets
Feature activation+1.491
for
Token for
Feature activation+0.130
renewable
Token renewable
Feature activation+0.000
energy
Token energy
Feature activation+0.000
,
Token,
Feature activation+2.672
China
Token China
Feature activation+0.315
is
Token is
Feature activation+0.000
poised
Token poised
Feature activation+0.344
to
Token to
Feature activation+0.522
be
Token be
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.752
plan
Token plan
Feature activation+0.897
for
Token for
Feature activation+0.170
Syria
Token Syria
Feature activation+0.000
,
Token,
Feature activation+2.658
the
Token the
Feature activation+1.060
odds
Token odds
Feature activation+0.382
of
Token of
Feature activation+0.000
a
Token a
Feature activation+0.000
rational
Token rational
Feature activation+0.000
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
with
Token with
Feature activation+0.000
MLS
Token MLS
Feature activation+0.547
not
Token not
Feature activation+3.604
looking
Token looking
Feature activation+2.578
like
Token like
Feature activation+1.698
a
Token a
Feature activation+0.000
viable
Token viable
Feature activation+0.145
option
Token option
Feature activation+0.503
,
Token,
Feature activation+1.284

Top DFA by src position
MAX = 4.135

a
Token a
Feature activation+0.006
Top resid features:
return
Token return
Feature activation+0.015
Top resid features:
home
Token home
Feature activation+0.002
Top resid features:
,
Token,
Feature activation+0.003
Top resid features:
and
Token and
Feature activation+0.014
Top resid features:
with
Token with
Feature activation+4.135
Top resid features:
MLS
Token MLS
Feature activation+0.089
Top resid features:
not
Token not
Feature activation+0.056
Top resid features:
looking
Token looking
Feature activation+0.000
Top resid features:
like
Token like
Feature activation+0.000
Top resid features:
a
Token a
Feature activation+0.000
Top resid features:
pay
Tokenpay
Feature activation+0.001
Top resid features:
able
Tokenable
Feature activation-0.001
Top resid features:
.
Token.
Feature activation-0.018
Top resid features:
âĢ
TokenâĢ
Feature activation-0.001
Top resid features:
Ŀ
TokenĿ
Feature activation+0.014
Top resid features:
With
Token With
Feature activation+3.865
Top resid features:
hundreds
Token hundreds
Feature activation-0.013
Top resid features:
of
Token of
Feature activation-0.002
Top resid features:
millions
Token millions
Feature activation+0.002
Top resid features:
of
Token of
Feature activation+0.005
Top resid features:
dollars
Token dollars
Feature activation+0.017
Top resid features:
a
Token a
Feature activation+0.001
Top resid features:
map
Token map
Feature activation-0.004
Top resid features:
and
Token and
Feature activation+0.001
Top resid features:
directions
Token directions
Feature activation-0.003
Top resid features:
.
Token.
Feature activation+0.016
Top resid features:
With
Token With
Feature activation+4.027
Top resid features:
mature
Token mature
Feature activation+0.008
Top resid features:
apple
Token apple
Feature activation-0.000
Top resid features:
,
Token,
Feature activation-0.012
Top resid features:
plum
Token plum
Feature activation-0.001
Top resid features:
,
Token,
Feature activation-0.005
Top resid features:
sales
Token sales
Feature activation-0.001
Top resid features:
price
Token price
Feature activation+0.000
Top resid features:
.
Token.
Feature activation-0.005
Top resid features:
Ċ
TokenĊ
Feature activation-0.008
Top resid features:
Ċ
TokenĊ
Feature activation-0.007
Top resid features:
With
TokenWith
Feature activation+3.478
Top resid features:
many
Token many
Feature activation+0.022
Top resid features:
state
Token state
Feature activation+0.002
Top resid features:
legislative
Token legislative
Feature activation+0.001
Top resid features:
sessions
Token sessions
Feature activation+0.008
Top resid features:
just
Token just
Feature activation+0.012
Top resid features:
this
Token this
Feature activation+0.002
Top resid features:
problem
Token problem
Feature activation+0.002
Top resid features:
.
Token.
Feature activation-0.005
Top resid features:
Ċ
TokenĊ
Feature activation-0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.004
Top resid features:
With
TokenWith
Feature activation+3.703
Top resid features:
so
Token so
Feature activation+0.027
Top resid features:
many
Token many
Feature activation+0.000
Top resid features:
benefits
Token benefits
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:
amount
Token amount
Feature activation-0.001
Top resid features:
of
Token of
Feature activation-0.000
Top resid features:
time
Token time
Feature activation+0.001
Top resid features:
.
Token.
Feature activation+0.045
Top resid features:
And
Token And
Feature activation+0.014
Top resid features:
with
Token with
Feature activation+3.636
Top resid features:
Google
Token Google
Feature activation-0.013
Top resid features:
helping
Token helping
Feature activation+0.046
Top resid features:
fund
Token fund
Feature activation+0.000
Top resid features:
its
Token its
Feature activation+0.000
Top resid features:
development
Token development
Feature activation+0.000
Top resid features:
film
Token film
Feature activation-0.009
Top resid features:
personality
Token personality
Feature activation+0.003
Top resid features:
and
Token and
Feature activation-0.001
Top resid features:
style
Token style
Feature activation+0.002
Top resid features:
.
Token.
Feature activation+0.001
Top resid features:
With
Token With
Feature activation+3.529
Top resid features:
our
Token our
Feature activation+0.023
Top resid features:
main
Token main
Feature activation-0.001
Top resid features:
characters
Token characters
Feature activation+0.001
Top resid features:
being
Token being
Feature activation+0.014
Top resid features:
two
Token two
Feature activation+0.010
Top resid features:
s
Tokens
Feature activation+0.002
Top resid features:
Six
Token Six
Feature activation+0.003
Top resid features:
Nations
Token Nations
Feature activation+0.000
Top resid features:
games
Token games
Feature activation+0.003
Top resid features:
.
Token.
Feature activation+0.013
Top resid features:
With
Token With
Feature activation+3.763
Top resid features:
preparation
Token preparation
Feature activation-0.074
Top resid features:
time
Token time
Feature activation-0.028
Top resid features:
so
Token so
Feature activation-0.009
Top resid features:
short
Token short
Feature activation+0.000
Top resid features:
and
Token and
Feature activation+0.000
Top resid features:
already
Token already
Feature activation-0.011
Top resid features:
is
Token is
Feature activation+0.009
Top resid features:
now
Token now
Feature activation+0.016
Top resid features:
âĢĵ
Token âĢĵ
Feature activation+0.009
Top resid features:
but
Token but
Feature activation+0.030
Top resid features:
with
Token with
Feature activation+3.625
Top resid features:
Audi
Token Audi
Feature activation-0.076
Top resid features:
,
Token,
Feature activation-0.050
Top resid features:
BMW
Token BMW
Feature activation+0.004
Top resid features:
and
Token and
Feature activation+0.002
Top resid features:
Mercedes
Token Mercedes
Feature activation+0.001
Top resid features:
sales
Token sales
Feature activation-0.000
Top resid features:
price
Token price
Feature activation-0.000
Top resid features:
.
Token.
Feature activation-0.006
Top resid features:
Ċ
TokenĊ
Feature activation-0.008
Top resid features:
Ċ
TokenĊ
Feature activation-0.014
Top resid features:
With
TokenWith
Feature activation+3.416
Top resid features:
many
Token many
Feature activation+0.014
Top resid features:
state
Token state
Feature activation+0.008
Top resid features:
legislative
Token legislative
Feature activation-0.002
Top resid features:
sessions
Token sessions
Feature activation+0.005
Top resid features:
just
Token just
Feature activation+0.012
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.414
Top resid features:
exhibition
Token exhibition
Feature activation-0.038
Top resid features:
hall
Token hall
Feature activation-0.002
Top resid features:
.
Token.
Feature activation-0.000
Top resid features:
But
Token But
Feature activation+0.055
Top resid features:
with
Token with
Feature activation+3.273
Top resid features:
Farage
Token Farage
Feature activation+0.005
Top resid features:
's
Token's
Feature activation+0.002
Top resid features:
nationalists
Token nationalists
Feature activation-0.004
Top resid features:
("
Token ("
Feature activation+0.006
Top resid features:
Give
TokenGive
Feature activation-0.002
Top resid features:
the
Token the
Feature activation-0.005
Top resid features:
bust
Token bust
Feature activation-0.000
Top resid features:
in
Token in
Feature activation-0.035
Top resid features:
2008
Token 2008
Feature activation-0.003
Top resid features:
.
Token.
Feature activation+0.026
Top resid features:
With
Token With
Feature activation+4.006
Top resid features:
the
Token the
Feature activation+0.142
Top resid features:
economy
Token economy
Feature activation-0.218
Top resid features:
on
Token on
Feature activation-0.274
Top resid features:
the
Token the
Feature activation-0.017
Top resid features:
rise
Token rise
Feature activation-0.055
Top resid features:
.
Token.
Feature activation+0.020
Top resid features:
âĢ
TokenâĢ
Feature activation+0.002
Top resid features:
Ŀ
TokenĿ
Feature activation-0.001
Top resid features:
Ċ
TokenĊ
Feature activation-0.005
Top resid features:
Ċ
TokenĊ
Feature activation-0.028
Top resid features:
With
TokenWith
Feature activation+3.507
Top resid features:
GTA
Token GTA
Feature activation+0.005
Top resid features:
V
Token V
Feature activation-0.003
Top resid features:
yet
Token yet
Feature activation-0.081
Top resid features:
to
Token to
Feature activation-0.021
Top resid features:
see
Token see
Feature activation-0.036
Top resid features:
critical
Token critical
Feature activation-0.007
Top resid features:
to
Token to
Feature activation-0.018
Top resid features:
public
Token public
Feature activation-0.001
Top resid features:
health
Token health
Feature activation+0.003
Top resid features:
.
Token.
Feature activation+0.028
Top resid features:
With
Token With
Feature activation+3.383
Top resid features:
an
Token an
Feature activation+0.064
Top resid features:
eventual
Token eventual
Feature activation-0.056
Top resid features:
vaccine
Token vaccine
Feature activation+0.022
Top resid features:
against
Token against
Feature activation+0.003
Top resid features:
the
Token the
Feature activation+0.013
Top resid features:
of
Token of
Feature activation+0.002
Top resid features:
rejected
Token rejected
Feature activation-0.003
Top resid features:
absentee
Token absentee
Feature activation+0.001
Top resid features:
ballots
Token ballots
Feature activation-0.003
Top resid features:
.
Token.
Feature activation-0.006
Top resid features:
With
Token With
Feature activation+3.002
Top resid features:
Republican
Token Republican
Feature activation+0.015
Top resid features:
U
Token U
Feature activation-0.000
Top resid features:
.
Token.
Feature activation-0.000
Top resid features:
S
TokenS
Feature activation+0.000
Top resid features:
.
Token.
Feature activation-0.000
Top resid features:
veterans
Token veterans
Feature activation+0.002
Top resid features:
and
Token and
Feature activation+0.007
Top resid features:
military
Token military
Feature activation-0.001
Top resid features:
voters
Token voters
Feature activation+0.006
Top resid features:
.
Token.
Feature activation-0.044
Top resid features:
With
Token With
Feature activation+3.280
Top resid features:
McCain
Token McCain
Feature activation-0.006
Top resid features:
facing
Token facing
Feature activation+0.023
Top resid features:
a
Token a
Feature activation+0.012
Top resid features:
diminished
Token diminished
Feature activation+0.007
Top resid features:
role
Token role
Feature activation+0.010
Top resid features:
pay
Tokenpay
Feature activation+0.001
Top resid features:
able
Tokenable
Feature activation-0.000
Top resid features:
.
Token.
Feature activation-0.006
Top resid features:
âĢ
TokenâĢ
Feature activation-0.002
Top resid features:
Ŀ
TokenĿ
Feature activation+0.018
Top resid features:
With
Token With
Feature activation+2.966
Top resid features:
hundreds
Token hundreds
Feature activation-0.038
Top resid features:
of
Token of
Feature activation-0.002
Top resid features:
millions
Token millions
Feature activation-0.004
Top resid features:
of
Token of
Feature activation+0.002
Top resid features:
dollars
Token dollars
Feature activation+0.009
Top resid features:
.
Token.
Feature activation+0.001
Top resid features:
Ċ
TokenĊ
Feature activation+0.003
Top resid features:
Ċ
TokenĊ
Feature activation+0.007
Top resid features:
âĢ
TokenâĢ
Feature activation+0.064
Top resid features:
ľ
Tokenľ
Feature activation+0.032
Top resid features:
With
TokenWith
Feature activation+3.068
Top resid features:
extensive
Token extensive
Feature activation+0.014
Top resid features:
manufacturing
Token manufacturing
Feature activation+0.001
Top resid features:
capacity
Token capacity
Feature activation+0.003
Top resid features:
in
Token in
Feature activation+0.005
Top resid features:
the
Token the
Feature activation+0.001
Top resid features:
Obama
Token Obama
Feature activation-0.002
Top resid features:
administration
Token administration
Feature activation-0.001
Top resid features:
every
Token every
Feature activation+0.024
Top resid features:
day
Token day
Feature activation+0.003
Top resid features:
.
Token.
Feature activation+0.036
Top resid features:
With
Token With
Feature activation+2.945
Top resid features:
the
Token the
Feature activation+0.070
Top resid features:
Trump
Token Trump
Feature activation-0.003
Top resid features:
regime
Token regime
Feature activation-0.008
Top resid features:
refusing
Token refusing
Feature activation+0.008
Top resid features:
to
Token to
Feature activation-0.001
Top resid features:
a
Token a
Feature activation-0.004
Top resid features:
return
Token return
Feature activation+0.013
Top resid features:
home
Token home
Feature activation+0.009
Top resid features:
,
Token,
Feature activation+0.003
Top resid features:
and
Token and
Feature activation+0.018
Top resid features:
with
Token with
Feature activation+3.210
Top resid features:
MLS
Token MLS
Feature activation+0.059
Top resid features:
not
Token not
Feature activation+0.021
Top resid features:
looking
Token looking
Feature activation+0.008
Top resid features:
like
Token like
Feature activation+0.000
Top resid features:
a
Token a
Feature activation+0.000
Top resid features:

Decoder Weights Distribution

Head 0: 0.02

Head 1: 0.03

Head 2: 0.12

Head 3: 0.50

Head 4: 0.06

Head 5: 0.02

Head 6: 0.03

Head 7: 0.03

Head 8: 0.04

Head 9: 0.03

Head 10: 0.06

Head 11: 0.06

Positive logits

looming1.75

resurg1.71

cele1.60

scarce1.59

tempting1.59

unwilling1.58

vez1.58

trending1.52

scram1.51

uncertainties1.50

pressures1.49

dwindling1.49

eager1.49

scar1.48

keen1.47

busy1.47

poised1.46

odds1.46

ticking1.46

churn1.44

Negative logits

>"-1.80

!".-1.60

Recommend-1.55

aut-1.52

".[-1.51

ntil-1.50

bestos-1.50

ONSORED-1.49

IRT-1.48

Dialogue-1.47

rotein-1.45

¯-1.45

ARDS-1.45

.(-1.45

.�-1.44

ortment-1.43

ENE-1.43

IMAGES-1.43

UTERS-1.42

essional-1.42

INTERVAL 3.243 - 3.604
CONTAINS 0.000%

millions
Token millions
Feature activation+0.049
of
Token of
Feature activation+0.000
dollars
Token dollars
Feature activation+0.076
of
Token of
Feature activation+0.000
payments
Token payments
Feature activation+0.981
looming
Token looming
Feature activation+3.265
and
Token and
Feature activation+1.887
little
Token little
Feature activation+1.300
cash
Token cash
Feature activation+0.855
on
Token on
Feature activation+1.423
hand
Token hand
Feature activation+2.678
that
Token that
Feature activation+0.180
surround
Token surround
Feature activation+0.000
her
Token her
Feature activation+0.000
property
Token property
Feature activation+0.000
,
Token,
Feature activation+2.228
it
Token it
Feature activation+3.257
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.083
s
Tokens
Feature activation+0.302
about
Token about
Feature activation+0.073
as
Token as
Feature activation+0.000
home
Token home
Feature activation+0.000
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
with
Token with
Feature activation+0.000
MLS
Token MLS
Feature activation+0.547
not
Token not
Feature activation+3.604
looking
Token looking
Feature activation+2.578
like
Token like
Feature activation+1.698
a
Token a
Feature activation+0.000
viable
Token viable
Feature activation+0.145
option
Token option
Feature activation+0.503

INTERVAL 2.883 - 3.243
CONTAINS 0.001%

Audi
Token Audi
Feature activation+0.189
,
Token,
Feature activation+0.000
BMW
Token BMW
Feature activation+0.000
and
Token and
Feature activation+0.000
Mercedes
Token Mercedes
Feature activation+0.000
stepping
Token stepping
Feature activation+3.008
up
Token up
Feature activation+0.697
,
Token,
Feature activation+2.523
it
Token it
Feature activation+1.492
will
Token will
Feature activation+0.822
only
Token only
Feature activation+0.744
the
Token the
Feature activation+0.000
next
Token next
Feature activation+0.000
month
Token month
Feature activation+0.367
or
Token or
Feature activation+0.305
two
Token two
Feature activation+0.000
,
Token,
Feature activation+3.191
there
Token there
Feature activation+2.941
could
Token could
Feature activation+1.440
be
Token be
Feature activation+1.400
bills
Token bills
Feature activation+0.000
introduced
Token introduced
Feature activation+0.000
games
Token games
Feature activation+0.000
.
Token.
Feature activation+0.000
With
Token With
Feature activation+0.000
preparation
Token preparation
Feature activation+1.258
time
Token time
Feature activation+0.831
so
Token so
Feature activation+3.009
short
Token short
Feature activation+1.281
and
Token and
Feature activation+0.152
with
Token with
Feature activation+0.397
no
Token no
Feature activation+1.552
mid
Token mid
Feature activation+0.136
time
Token time
Feature activation+0.000
.
Token.
Feature activation+0.000
And
Token And
Feature activation+0.000
with
Token with
Feature activation+0.810
Google
Token Google
Feature activation+1.280
helping
Token helping
Feature activation+3.090
fund
Token fund
Feature activation+0.875
its
Token its
Feature activation+0.065
development
Token development
Feature activation+0.369
,
Token,
Feature activation+1.685
we
Token we
Feature activation+0.000
and
Token and
Feature activation+0.000
Alt
Token Alt
Feature activation+0.000
-
Token-
Feature activation+0.000
J
TokenJ
Feature activation+0.000
,
Token,
Feature activation+0.000
it
Token it
Feature activation+3.086
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.179
wouldn
Token wouldn
Feature activation+0.912
âĢ
TokenâĢ
Feature activation+0.000

INTERVAL 2.523 - 2.883
CONTAINS 0.001%

votes
Token votes
Feature activation+0.000
as
Token as
Feature activation+0.000
of
Token of
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
night
Token night
Feature activation+0.000
--
Token --
Feature activation+2.737
the
Token the
Feature activation+1.759
issue
Token issue
Feature activation+0.323
of
Token of
Feature activation+0.570
how
Token how
Feature activation+0.000
and
Token and
Feature activation+0.000
next
Token next
Feature activation+0.000
year
Token year
Feature activation+0.174
's
Token's
Feature activation+0.000
European
Token European
Feature activation+0.000
elections
Token elections
Feature activation+0.000
,
Token,
Feature activation+2.850
the
Token the
Feature activation+1.991
"
Token "
Feature activation+0.300
modern
Tokenmodern
Feature activation+0.000
isers
Tokenisers
Feature activation+0.000
"
Token"
Feature activation+0.000
eventual
Token eventual
Feature activation+0.824
vaccine
Token vaccine
Feature activation+0.000
against
Token against
Feature activation+0.005
the
Token the
Feature activation+0.000
virus
Token virus
Feature activation+0.237
likely
Token likely
Feature activation+2.781
years
Token years
Feature activation+1.125
away
Token away
Feature activation+2.457
,
Token,
Feature activation+1.984
a
Token a
Feature activation+0.719
widely
Token widely
Feature activation+0.030
national
Token national
Feature activation+0.000
targets
Token targets
Feature activation+1.491
for
Token for
Feature activation+0.130
renewable
Token renewable
Feature activation+0.000
energy
Token energy
Feature activation+0.000
,
Token,
Feature activation+2.672
China
Token China
Feature activation+0.315
is
Token is
Feature activation+0.000
poised
Token poised
Feature activation+0.344
to
Token to
Feature activation+0.522
be
Token be
Feature activation+0.000
With
Token With
Feature activation+0.000
the
Token the
Feature activation+0.000
economy
Token economy
Feature activation+0.765
on
Token on
Feature activation+0.734
the
Token the
Feature activation+0.065
rise
Token rise
Feature activation+2.818
,
Token,
Feature activation+1.245
the
Token the
Feature activation+0.037
neighborhood
Token neighborhood
Feature activation+0.000
is
Token is
Feature activation+0.000
reinvent
Token reinvent
Feature activation+0.000

INTERVAL 2.162 - 2.523
CONTAINS 0.002%

Committee
Token Committee
Feature activation+0.000
and
Token and
Feature activation+0.000
the
Token the
Feature activation+0.074
provincial
Token provincial
Feature activation+0.000
leaders
Token leaders
Feature activation+1.059
,
Token,
Feature activation+2.272
his
Token his
Feature activation+0.582
political
Token political
Feature activation+0.000
power
Token power
Feature activation+0.000
is
Token is
Feature activation+0.000
heavily
Token heavily
Feature activation+0.000
town
Token town
Feature activation+0.000
hall
Token hall
Feature activation+0.000
on
Token on
Feature activation+0.000
Monday
Token Monday
Feature activation+0.000
)
Token)
Feature activation+0.987
threatening
Token threatening
Feature activation+2.458
to
Token to
Feature activation+0.856
overrun
Token overrun
Feature activation+0.000
the
Token the
Feature activation+0.000
Conservative
Token Conservative
Feature activation+0.000
heart
Token heart
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
With
TokenWith
Feature activation+0.184
GTA
Token GTA
Feature activation+0.194
V
Token V
Feature activation+0.767
yet
Token yet
Feature activation+2.443
to
Token to
Feature activation+2.384
see
Token see
Feature activation+1.648
a
Token a
Feature activation+0.863
release
Token release
Feature activation+1.631
,
Token,
Feature activation+2.790
it
Token it
Feature activation+1.346
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
With
TokenWith
Feature activation+0.184
GTA
Token GTA
Feature activation+0.194
V
Token V
Feature activation+0.767
yet
Token yet
Feature activation+2.443
to
Token to
Feature activation+2.384
see
Token see
Feature activation+1.648
a
Token a
Feature activation+0.863
release
Token release
Feature activation+1.631
,
Token,
Feature activation+2.790
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
now
Token now
Feature activation+0.000
with
Token with
Feature activation+0.009
Caesar
Token Caesar
Feature activation+0.001
seemingly
Token seemingly
Feature activation+2.215
clumsy
Token clumsy
Feature activation+0.043
-
Token-
Feature activation+0.000
footed
Tokenfooted
Feature activation+0.000
on
Token on
Feature activation+0.000
another
Token another
Feature activation+0.508

INTERVAL 1.802 - 2.162
CONTAINS 0.003%

.
Token.
Feature activation+0.000
Indeed
TokenIndeed
Feature activation+0.000
,
Token,
Feature activation+0.000
with
Token with
Feature activation+0.000
Cook
Token Cook
Feature activation+0.000
at
Token at
Feature activation+1.924
the
Token the
Feature activation+0.869
helm
Token helm
Feature activation+0.371
,
Token,
Feature activation+0.434
Apple
Token Apple
Feature activation+0.000
has
Token has
Feature activation+0.000
and
Token and
Feature activation+0.215
the
Token the
Feature activation+0.292
economy
Token economy
Feature activation+0.771
shrinking
Token shrinking
Feature activation+2.272
,
Token,
Feature activation+2.284
not
Token not
Feature activation+1.863
growing
Token growing
Feature activation+1.188
,
Token,
Feature activation+1.550
since
Token since
Feature activation+0.692
the
Token the
Feature activation+0.000
spending
Token spending
Feature activation+0.000
role
Token role
Feature activation+0.333
in
Token in
Feature activation+0.247
the
Token the
Feature activation+0.000
GOP
Token GOP
Feature activation+0.250
,
Token,
Feature activation+1.192
and
Token and
Feature activation+1.919
Chuck
Token Chuck
Feature activation+0.000
Hag
Token Hag
Feature activation+0.000
el
Tokenel
Feature activation+0.192
retiring
Token retiring
Feature activation+2.240
from
Token from
Feature activation+0.000
of
Token of
Feature activation+0.000
dollars
Token dollars
Feature activation+0.076
of
Token of
Feature activation+0.000
payments
Token payments
Feature activation+0.981
looming
Token looming
Feature activation+3.265
and
Token and
Feature activation+1.887
little
Token little
Feature activation+1.300
cash
Token cash
Feature activation+0.855
on
Token on
Feature activation+1.423
hand
Token hand
Feature activation+2.678
,
Token,
Feature activation+2.321
as
Tokenas
Feature activation+0.000
now
Token now
Feature activation+0.997
banned
Token banned
Feature activation+1.160
in
Token in
Feature activation+0.487
Canada
Token Canada
Feature activation+0.620
,
Token,
Feature activation+2.006
the
Token the
Feature activation+0.226
United
Token United
Feature activation+0.000
States
Token States
Feature activation+0.163
and
Token and
Feature activation+0.000
Europe
Token Europe
Feature activation+0.000

INTERVAL 1.442 - 1.802
CONTAINS 0.007%

Ċ
TokenĊ
Feature activation+0.000
With
TokenWith
Feature activation+0.000
the
Token the
Feature activation+0.000
NRL
Token NRL
Feature activation+0.000
finals
Token finals
Feature activation+0.324
starting
Token starting
Feature activation+1.480
this
Token this
Feature activation+0.927
week
Token week
Feature activation+0.599
,
Token,
Feature activation+0.937
clubs
Token clubs
Feature activation+0.000
across
Token across
Feature activation+0.000
little
Token little
Feature activation+0.000
sea
Token sea
Feature activation+0.000
-
Token-
Feature activation+0.000
side
Tokenside
Feature activation+0.000
town
Token town
Feature activation+0.000
,
Token,
Feature activation+1.584
and
Token and
Feature activation+2.577
a
Token a
Feature activation+1.598
fitting
Token fitting
Feature activation+0.000
soundtrack
Token soundtrack
Feature activation+0.160
that
Token that
Feature activation+2.045
time
Token time
Feature activation+0.831
so
Token so
Feature activation+3.009
short
Token short
Feature activation+1.281
and
Token and
Feature activation+0.152
with
Token with
Feature activation+0.397
no
Token no
Feature activation+1.552
mid
Token mid
Feature activation+0.136
week
Tokenweek
Feature activation+0.678
games
Token games
Feature activation+1.405
in
Token in
Feature activation+0.893
which
Token which
Feature activation+1.457
companies
Token companies
Feature activation+0.121
and
Token and
Feature activation+0.629
by
Token by
Feature activation+0.000
going
Token going
Feature activation+0.000
public
Token public
Feature activation+0.093
,
Token,
Feature activation+1.612
Turkey
Token Turkey
Feature activation+0.088
will
Token will
Feature activation+0.038
attract
Token attract
Feature activation+0.000
more
Token more
Feature activation+0.000
international
Token international
Feature activation+0.000
mayor
Token mayor
Feature activation+0.000
Giovanni
Token Giovanni
Feature activation+0.000
Cal
Token Cal
Feature activation+0.000
abre
Tokenabre
Feature activation+0.000
se
Tokense
Feature activation+0.000
,
Token,
Feature activation+1.662
Sporting
Token Sporting
Feature activation+0.000
Loc
Token Loc
Feature activation+0.000
ri
Tokenri
Feature activation+0.000
will
Token will
Feature activation+0.000
square
Token square
Feature activation+0.000

INTERVAL 1.081 - 1.442
CONTAINS 0.010%

bed
Token bed
Feature activation+0.000
.
Token.
Feature activation+0.000
With
Token With
Feature activation+0.000
currents
Token currents
Feature activation+1.552
being
Token being
Feature activation+1.965
slow
Token slow
Feature activation+1.148
to
Token to
Feature activation+0.790
nearly
Token nearly
Feature activation+0.483
non
Token non
Feature activation+0.118
-
Token-
Feature activation+0.000
existent
Tokenexistent
Feature activation+1.016
pro
Token pro
Feature activation+0.000
-
Token-
Feature activation+0.000
Israel
TokenIsrael
Feature activation+0.000
positioning
Token positioning
Feature activation+1.263
,
Token,
Feature activation+2.086
and
Token and
Feature activation+1.300
support
Token support
Feature activation+0.239
for
Token for
Feature activation+1.091
a
Token a
Feature activation+0.173
very
Token very
Feature activation+0.207
questionable
Token questionable
Feature activation+0.013
of
Token of
Feature activation+0.000
War
Token War
Feature activation+0.000
,
Token,
Feature activation+0.357
âĢ
TokenâĢ
Feature activation+0.166
Ŀ
TokenĿ
Feature activation+0.158
everything
Token everything
Feature activation+1.237
is
Token is
Feature activation+0.000
starting
Token starting
Feature activation+0.000
to
Token to
Feature activation+0.088
fall
Token fall
Feature activation+0.000
into
Token into
Feature activation+0.000
than
Token than
Feature activation+1.630
three
Token three
Feature activation+0.801
weeks
Token weeks
Feature activation+1.426
after
Token after
Feature activation+0.884
that
Token that
Feature activation+0.070
,
Token,
Feature activation+1.151
it
Token it
Feature activation+0.356
is
Token is
Feature activation+0.007
possible
Token possible
Feature activation+0.035
to
Token to
Feature activation+0.000
discern
Token discern
Feature activation+0.000
it
Token it
Feature activation+1.346
âĢ
TokenâĢ
Feature activation+0.012
Ļ
TokenĻ
Feature activation+0.241
s
Tokens
Feature activation+0.239
hard
Token hard
Feature activation+0.465
to
Token to
Feature activation+1.370
tell
Token tell
Feature activation+0.000
if
Token if
Feature activation+0.000
this
Token this
Feature activation+0.000
idea
Token idea
Feature activation+0.000
has
Token has
Feature activation+0.000

INTERVAL 0.721 - 1.081
CONTAINS 0.018%

an
Token an
Feature activation+2.138
all
Token all
Feature activation+0.229
but
Token but
Feature activation+0.201
certain
Token certain
Feature activation+0.147
birth
Token birth
Feature activation+0.287
in
Token in
Feature activation+0.769
the
Token the
Feature activation+0.280
NCAA
Token NCAA
Feature activation+0.000
postseason
Token postseason
Feature activation+0.666
,
Token,
Feature activation+1.487
you
Token you
Feature activation+0.000
U
Token U
Feature activation+0.028
.
Token.
Feature activation+0.000
S
TokenS
Feature activation+0.233
.
Token.
Feature activation+0.000
statement
Token statement
Feature activation+1.469
read
Token read
Feature activation+0.990
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
TH
TokenTH
Feature activation+0.000
A
TokenA
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
With
TokenWith
Feature activation+0.000
all
Token all
Feature activation+0.000
this
Token this
Feature activation+0.920
beer
Token beer
Feature activation+0.000
drinking
Token drinking
Feature activation+0.219
experience
Token experience
Feature activation+0.000
under
Token under
Feature activation+0.241
my
Token my
Feature activation+0.162
code
Tokencode
Feature activation+0.149
6
Token 6
Feature activation+0.264
and
Token and
Feature activation+0.353
iOS
Token iOS
Feature activation+0.042
8
Token 8
Feature activation+0.520
,
Token,
Feature activation+0.890
Apple
Token Apple
Feature activation+0.181
have
Token have
Feature activation+0.267
created
Token created
Feature activation+0.006
a
Token a
Feature activation+0.000
new
Token new
Feature activation+0.000
the
Token the
Feature activation+0.392
hottest
Token hottest
Feature activation+0.109
year
Token year
Feature activation+0.193
on
Token on
Feature activation+0.000
record
Token record
Feature activation+0.752
,
Token,
Feature activation+1.041
India
Token India
Feature activation+0.120
and
Token and
Feature activation+0.000
China
Token China
Feature activation+0.000
will
Token will
Feature activation+0.118
negotiate
Token negotiate
Feature activation+0.000

INTERVAL 0.360 - 0.721
CONTAINS 0.038%

London
Token London
Feature activation+0.000
.
Token.
Feature activation+0.000
With
Token With
Feature activation+0.000
a
Token a
Feature activation+0.000
Kickstarter
Token Kickstarter
Feature activation+0.000
launch
Token launch
Feature activation+0.582
party
Token party
Feature activation+0.000
and
Token and
Feature activation+0.575
private
Token private
Feature activation+0.143
view
Token view
Feature activation+0.209
for
Token for
Feature activation+0.504
fights
Token fights
Feature activation+0.008
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
NFL
Token NFL
Feature activation+0.000
PA
TokenPA
Feature activation+0.000
,
Token,
Feature activation+0.460
the
Token the
Feature activation+0.000
average
Token average
Feature activation+0.000
NFL
Token NFL
Feature activation+0.000
player
Token player
Feature activation+0.000
salary
Token salary
Feature activation+0.000
that
Token that
Feature activation+0.000
un
Token un
Feature activation+0.000
spoken
Tokenspoken
Feature activation+0.300
threat
Token threat
Feature activation+0.209
-
Token -
Feature activation+0.766
that
Token that
Feature activation+0.677
this
Token this
Feature activation+0.000
is
Token is
Feature activation+0.000
what
Token what
Feature activation+0.000
awaits
Token awaits
Feature activation+0.000
you
Token you
Feature activation+0.000
as
Token as
Feature activation+0.000
well
Token well
Feature activation+0.000
?
Token?
Feature activation+0.000
With
Token With
Feature activation+0.000
housing
Token housing
Feature activation+0.375
benefit
Token benefit
Feature activation+0.396
being
Token being
Feature activation+2.406
cut
Token cut
Feature activation+1.261
as
Token as
Feature activation+0.869
well
Token well
Feature activation+0.802
what
Token what
Feature activation+1.247
months
Token months
Feature activation+0.588
of
Token of
Feature activation+0.000
his
Token his
Feature activation+0.000
son
Token son
Feature activation+0.000
's
Token's
Feature activation+0.000
contract
Token contract
Feature activation+0.608
,
Token,
Feature activation+1.248
all
Token all
Feature activation+0.407
options
Token options
Feature activation+0.000
are
Token are
Feature activation+0.000
being
Token being
Feature activation+0.000

INTERVAL 0.000 - 0.360
CONTAINS 99.920%

the
Token the
Feature activation+0.000
existing
Token existing
Feature activation+0.000
number
Token number
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
organisation
Token organisation
Feature activation+0.000
has
Token has
Feature activation+0.000
had
Token had
Feature activation+0.000
the
Token the
Feature activation+0.000
item
Token item
Feature activation+0.000
--
Token --
Feature activation+0.000
and
Token and
Feature activation+0.000
then
Token then
Feature activation+0.000
the
Token the
Feature activation+0.000
people
Token people
Feature activation+0.000
who
Token who
Feature activation+0.000
associated
Token associated
Feature activation+0.000
with
Token with
Feature activation+0.000
them
Token them
Feature activation+0.000
,
Token,
Feature activation+0.000
as
Token as
Feature activation+0.000
an
Token an
Feature activation+0.000
evil
Token evil
Feature activation+0.000
gold
Token gold
Feature activation+0.000
dig
Token dig
Feature activation+0.000
ger
Tokenger
Feature activation+0.000
who
Token who
Feature activation+0.000
was
Token was
Feature activation+0.000
cheating
Token cheating
Feature activation+0.000
on
Token on
Feature activation+0.000
Johnny
Token Johnny
Feature activation+0.000
provided
Token provided
Feature activation+0.000
a
Token a
Feature activation+0.000
nice
Token nice
Feature activation+0.000
contrast
Token contrast
Feature activation+0.000
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
rough
Token rough
Feature activation+0.000
and
Token and
Feature activation+0.000
powerful
Token powerful
Feature activation+0.000
sight
Token sight
Feature activation+0.000
of
Token of
Feature activation+0.000
-
Token-
Feature activation+0.000
minded
Tokenminded
Feature activation+0.000
donors
Token donors
Feature activation+0.000
in
Token in
Feature activation+0.000
2003
Token 2003
Feature activation+0.000
.
Token.
Feature activation+0.000
Since
Token Since
Feature activation+0.000
then
Token then
Feature activation+0.000
,
Token,
Feature activation+0.000
he
Token he
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 H5.3: (feature 14184)

TOP ACTIVATIONS
MAX = 3.745

Ċ
TokenĊ
Feature activation+0.000
Sure
TokenSure
Feature activation+0.000
,
Token,
Feature activation+0.055
sure
Token sure
Feature activation+0.171
she
Token she
Feature activation+0.655
was
Token was
Feature activation+3.745
screaming
Token screaming
Feature activation+2.305
and
Token and
Feature activation+2.038
running
Token running
Feature activation+1.606
there
Token there
Feature activation+0.460
for
Token for
Feature activation+0.000
season
Token season
Feature activation+0.000
.
Token.
Feature activation+0.000
Sure
Token Sure
Feature activation+0.000
,
Token,
Feature activation+0.676
we
Token we
Feature activation+2.540
knew
Token knew
Feature activation+3.063
they
Token they
Feature activation+0.986
might
Token might
Feature activation+0.911
exist
Token exist
Feature activation+0.531
,
Token,
Feature activation+2.518
but
Token but
Feature activation+0.237
conditioning
Token conditioning
Feature activation+0.000
.
Token.
Feature activation+0.000
Sure
Token Sure
Feature activation+0.000
,
Token,
Feature activation+0.463
everyone
Token everyone
Feature activation+2.340
had
Token had
Feature activation+2.975
to
Token to
Feature activation+1.819
deal
Token deal
Feature activation+1.030
with
Token with
Feature activation+1.143
that
Token that
Feature activation+0.965
heat
Token heat
Feature activation+0.000
you
Token you
Feature activation+0.823
received
Token received
Feature activation+0.014
from
Token from
Feature activation+0.000
your
Token your
Feature activation+0.000
designer
Token designer
Feature activation+0.235
look
Token look
Feature activation+2.925
great
Token great
Feature activation+2.471
with
Token with
Feature activation+1.745
those
Token those
Feature activation+1.922
mocked
Token mocked
Feature activation+0.922
out
Token out
Feature activation+0.250
Sure
TokenSure
Feature activation+0.000
,
Token,
Feature activation+0.000
the
Token the
Feature activation+2.409
mixed
Token mixed
Feature activation+0.135
division
Token division
Feature activation+0.000
has
Token has
Feature activation+2.603
always
Token always
Feature activation+2.572
been
Token been
Feature activation+1.957
looked
Token looked
Feature activation+0.626
at
Token at
Feature activation+0.649
as
Token as
Feature activation+0.517
,
Token,
Feature activation+0.000
the
Token the
Feature activation+2.409
mixed
Token mixed
Feature activation+0.135
division
Token division
Feature activation+0.000
has
Token has
Feature activation+2.603
always
Token always
Feature activation+2.572
been
Token been
Feature activation+1.957
looked
Token looked
Feature activation+0.626
at
Token at
Feature activation+0.649
as
Token as
Feature activation+0.517
something
Token something
Feature activation+0.000
the
Token the
Feature activation+0.000
season
Token season
Feature activation+0.000
.
Token.
Feature activation+0.000
Sure
Token Sure
Feature activation+0.000
,
Token,
Feature activation+0.676
we
Token we
Feature activation+2.540
knew
Token knew
Feature activation+3.063
they
Token they
Feature activation+0.986
might
Token might
Feature activation+0.911
exist
Token exist
Feature activation+0.531
,
Token,
Feature activation+2.518
we
Token we
Feature activation+2.540
knew
Token knew
Feature activation+3.063
they
Token they
Feature activation+0.986
might
Token might
Feature activation+0.911
exist
Token exist
Feature activation+0.531
,
Token,
Feature activation+2.518
but
Token but
Feature activation+0.237
our
Token our
Feature activation+0.137
defense
Token defense
Feature activation+0.000
thought
Token thought
Feature activation+0.015
they
Token they
Feature activation+0.000
received
Token received
Feature activation+0.014
from
Token from
Feature activation+0.000
your
Token your
Feature activation+0.000
designer
Token designer
Feature activation+0.235
look
Token look
Feature activation+2.925
great
Token great
Feature activation+2.471
with
Token with
Feature activation+1.745
those
Token those
Feature activation+1.922
mocked
Token mocked
Feature activation+0.922
out
Token out
Feature activation+0.250
items
Token items
Feature activation+0.757
to
Token to
Feature activation+0.000
display
Token display
Feature activation+0.000
.
Token.
Feature activation+0.000
Sure
Token Sure
Feature activation+0.000
,
Token,
Feature activation+0.000
the
Token the
Feature activation+2.466
files
Token files
Feature activation+0.979
you
Token you
Feature activation+0.823
received
Token received
Feature activation+0.014
from
Token from
Feature activation+0.000
your
Token your
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Sure
TokenSure
Feature activation+0.000
,
Token,
Feature activation+0.000
the
Token the
Feature activation+2.409
mixed
Token mixed
Feature activation+0.135
division
Token division
Feature activation+0.000
has
Token has
Feature activation+2.603
always
Token always
Feature activation+2.572
been
Token been
Feature activation+1.957
air
Token air
Feature activation+0.000
conditioning
Token conditioning
Feature activation+0.000
.
Token.
Feature activation+0.000
Sure
Token Sure
Feature activation+0.000
,
Token,
Feature activation+0.463
everyone
Token everyone
Feature activation+2.340
had
Token had
Feature activation+2.975
to
Token to
Feature activation+1.819
deal
Token deal
Feature activation+1.030
with
Token with
Feature activation+1.143
that
Token that
Feature activation+0.965
running
Token running
Feature activation+1.606
there
Token there
Feature activation+0.460
for
Token for
Feature activation+0.000
a
Token a
Feature activation+0.691
second
Token second
Feature activation+0.601
.
Token.
Feature activation+2.327
It
Token It
Feature activation+1.103
took
Token took
Feature activation+0.221
me
Token me
Feature activation+0.098
a
Token a
Feature activation+0.000
minute
Token minute
Feature activation+0.000
Sure
Token Sure
Feature activation+0.000
,
Token,
Feature activation+0.223
it
Token it
Feature activation+1.474
âĢ
TokenâĢ
Feature activation+0.011
Ļ
TokenĻ
Feature activation+0.619
s
Tokens
Feature activation+2.305
not
Token not
Feature activation+1.817
technically
Token technically
Feature activation+0.762
a
Token a
Feature activation+0.000
novel
Token novel
Feature activation+0.000
,
Token,
Feature activation+0.269
Sure
TokenSure
Feature activation+0.000
,
Token,
Feature activation+0.055
sure
Token sure
Feature activation+0.171
she
Token she
Feature activation+0.655
was
Token was
Feature activation+3.745
screaming
Token screaming
Feature activation+2.305
and
Token and
Feature activation+2.038
running
Token running
Feature activation+1.606
there
Token there
Feature activation+0.460
for
Token for
Feature activation+0.000
a
Token a
Feature activation+0.691
anymore
Token anymore
Feature activation+0.000
.
Token.
Feature activation+0.000
Sure
Token Sure
Feature activation+0.000
a
Token a
Feature activation+0.000
woman
Token woman
Feature activation+0.721
looks
Token looks
Feature activation+2.293
hot
Token hot
Feature activation+0.707
in
Token in
Feature activation+0.754
a
Token a
Feature activation+0.815
tight
Token tight
Feature activation+0.484
dress
Token dress
Feature activation+0.740
with
Token with
Feature activation+1.745
those
Token those
Feature activation+1.922
mocked
Token mocked
Feature activation+0.922
out
Token out
Feature activation+0.250
items
Token items
Feature activation+0.757
.
Token.
Feature activation+2.186
But
Token But
Feature activation+0.000
,
Token,
Feature activation+0.000
how
Token how
Feature activation+0.000
does
Token does
Feature activation+0.000
the
Token the
Feature activation+0.000
,
Token,
Feature activation+0.055
sure
Token sure
Feature activation+0.171
she
Token she
Feature activation+0.655
was
Token was
Feature activation+3.745
screaming
Token screaming
Feature activation+2.305
and
Token and
Feature activation+2.038
running
Token running
Feature activation+1.606
there
Token there
Feature activation+0.460
for
Token for
Feature activation+0.000
a
Token a
Feature activation+0.691
second
Token second
Feature activation+0.601
the
Token the
Feature activation+2.409
mixed
Token mixed
Feature activation+0.135
division
Token division
Feature activation+0.000
has
Token has
Feature activation+2.603
always
Token always
Feature activation+2.572
been
Token been
Feature activation+1.957
looked
Token looked
Feature activation+0.626
at
Token at
Feature activation+0.649
as
Token as
Feature activation+0.517
something
Token something
Feature activation+0.000
of
Token of
Feature activation+0.000
your
Token your
Feature activation+0.000
designer
Token designer
Feature activation+0.235
look
Token look
Feature activation+2.925
great
Token great
Feature activation+2.471
with
Token with
Feature activation+1.745
those
Token those
Feature activation+1.922
mocked
Token mocked
Feature activation+0.922
out
Token out
Feature activation+0.250
items
Token items
Feature activation+0.757
.
Token.
Feature activation+2.186
But
Token But
Feature activation+0.000

Top DFA by src position
MAX = 2.871

<|endoftext|>
Token<|endoftext|>
Feature activation+1.458
Top resid features:
Ċ
TokenĊ
Feature activation-0.013
Top resid features:
Ċ
TokenĊ
Feature activation+0.004
Top resid features:
Sure
TokenSure
Feature activation+1.421
Top resid features:
,
Token,
Feature activation+1.046
Top resid features:
sure
Token sure
Feature activation+2.195
Top resid features:
she
Token she
Feature activation+0.455
Top resid features:
was
Token was
Feature activation+0.038
Top resid features:
screaming
Token screaming
Feature activation+0.000
Top resid features:
and
Token and
Feature activation+0.000
Top resid features:
running
Token running
Feature activation+0.000
Top resid features:
most
Token most
Feature activation+0.009
Top resid features:
of
Token of
Feature activation+0.002
Top resid features:
the
Token the
Feature activation+0.008
Top resid features:
season
Token season
Feature activation+0.024
Top resid features:
.
Token.
Feature activation+0.070
Top resid features:
Sure
Token Sure
Feature activation+2.403
Top resid features:
,
Token,
Feature activation+1.504
Top resid features:
we
Token we
Feature activation+0.241
Top resid features:
knew
Token knew
Feature activation+0.060
Top resid features:
they
Token they
Feature activation+0.000
Top resid features:
might
Token might
Feature activation+0.000
Top resid features:
's
Token's
Feature activation+0.018
Top resid features:
faulty
Token faulty
Feature activation+0.039
Top resid features:
air
Token air
Feature activation+0.011
Top resid features:
conditioning
Token conditioning
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.072
Top resid features:
Sure
Token Sure
Feature activation+2.374
Top resid features:
,
Token,
Feature activation+1.846
Top resid features:
everyone
Token everyone
Feature activation-0.003
Top resid features:
had
Token had
Feature activation+0.002
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
deal
Token deal
Feature activation+0.000
Top resid features:
any
Token any
Feature activation+0.003
Top resid features:
items
Token items
Feature activation+0.018
Top resid features:
to
Token to
Feature activation+0.013
Top resid features:
display
Token display
Feature activation-0.014
Top resid features:
.
Token.
Feature activation+0.034
Top resid features:
Sure
Token Sure
Feature activation+2.246
Top resid features:
,
Token,
Feature activation+1.751
Top resid features:
the
Token the
Feature activation+0.702
Top resid features:
files
Token files
Feature activation+0.083
Top resid features:
you
Token you
Feature activation-0.007
Top resid features:
received
Token received
Feature activation+0.035
Top resid features:
seriously
Token seriously
Feature activation-0.001
Top resid features:
.
Token.
Feature activation+0.054
Top resid features:
Ċ
TokenĊ
Feature activation+0.007
Top resid features:
Ċ
TokenĊ
Feature activation+0.011
Top resid features:
Sure
TokenSure
Feature activation+1.526
Top resid features:
,
Token,
Feature activation+1.977
Top resid features:
the
Token the
Feature activation+0.247
Top resid features:
mixed
Token mixed
Feature activation+0.026
Top resid features:
division
Token division
Feature activation-0.028
Top resid features:
has
Token has
Feature activation+0.115
Top resid features:
always
Token always
Feature activation+0.000
Top resid features:
seriously
Token seriously
Feature activation+0.004
Top resid features:
.
Token.
Feature activation+0.037
Top resid features:
Ċ
TokenĊ
Feature activation+0.006
Top resid features:
Ċ
TokenĊ
Feature activation+0.008
Top resid features:
Sure
TokenSure
Feature activation+0.867
Top resid features:
,
Token,
Feature activation+2.703
Top resid features:
the
Token the
Feature activation+0.301
Top resid features:
mixed
Token mixed
Feature activation+0.012
Top resid features:
division
Token division
Feature activation-0.008
Top resid features:
has
Token has
Feature activation+0.097
Top resid features:
always
Token always
Feature activation-0.076
Top resid features:
most
Token most
Feature activation+0.012
Top resid features:
of
Token of
Feature activation+0.001
Top resid features:
the
Token the
Feature activation+0.006
Top resid features:
season
Token season
Feature activation+0.073
Top resid features:
.
Token.
Feature activation+0.062
Top resid features:
Sure
Token Sure
Feature activation+2.194
Top resid features:
,
Token,
Feature activation+1.271
Top resid features:
we
Token we
Feature activation+0.165
Top resid features:
knew
Token knew
Feature activation+0.000
Top resid features:
they
Token they
Feature activation+0.000
Top resid features:
might
Token might
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.002
Top resid features:
the
Token the
Feature activation+0.008
Top resid features:
season
Token season
Feature activation+0.033
Top resid features:
.
Token.
Feature activation+0.072
Top resid features:
Sure
Token Sure
Feature activation+1.173
Top resid features:
,
Token,
Feature activation+1.577
Top resid features:
we
Token we
Feature activation+0.261
Top resid features:
knew
Token knew
Feature activation+0.141
Top resid features:
they
Token they
Feature activation+0.070
Top resid features:
might
Token might
Feature activation+0.076
Top resid features:
exist
Token exist
Feature activation+0.365
Top resid features:
any
Token any
Feature activation+0.006
Top resid features:
items
Token items
Feature activation+0.065
Top resid features:
to
Token to
Feature activation+0.018
Top resid features:
display
Token display
Feature activation+0.006
Top resid features:
.
Token.
Feature activation+0.072
Top resid features:
Sure
Token Sure
Feature activation+1.789
Top resid features:
,
Token,
Feature activation+1.406
Top resid features:
the
Token the
Feature activation+0.391
Top resid features:
files
Token files
Feature activation+0.052
Top resid features:
you
Token you
Feature activation+0.008
Top resid features:
received
Token received
Feature activation-0.022
Top resid features:
any
Token any
Feature activation+0.014
Top resid features:
items
Token items
Feature activation+0.016
Top resid features:
to
Token to
Feature activation+0.011
Top resid features:
display
Token display
Feature activation-0.001
Top resid features:
.
Token.
Feature activation+0.096
Top resid features:
Sure
Token Sure
Feature activation+1.900
Top resid features:
,
Token,
Feature activation+1.794
Top resid features:
the
Token the
Feature activation+0.234
Top resid features:
files
Token files
Feature activation+0.000
Top resid features:
you
Token you
Feature activation+0.000
Top resid features:
received
Token received
Feature activation+0.000
Top resid features:
seriously
Token seriously
Feature activation-0.011
Top resid features:
.
Token.
Feature activation+0.058
Top resid features:
Ċ
TokenĊ
Feature activation+0.014
Top resid features:
Ċ
TokenĊ
Feature activation+0.020
Top resid features:
Sure
TokenSure
Feature activation+1.159
Top resid features:
,
Token,
Feature activation+2.211
Top resid features:
the
Token the
Feature activation+0.171
Top resid features:
mixed
Token mixed
Feature activation+0.000
Top resid features:
division
Token division
Feature activation+0.000
Top resid features:
has
Token has
Feature activation+0.000
Top resid features:
always
Token always
Feature activation+0.000
Top resid features:
faulty
Token faulty
Feature activation+0.060
Top resid features:
air
Token air
Feature activation+0.035
Top resid features:
conditioning
Token conditioning
Feature activation+0.007
Top resid features:
.
Token.
Feature activation+0.063
Top resid features:
Sure
Token Sure
Feature activation+1.683
Top resid features:
,
Token,
Feature activation+1.755
Top resid features:
everyone
Token everyone
Feature activation+0.049
Top resid features:
had
Token had
Feature activation+0.000
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
deal
Token deal
Feature activation+0.000
Top resid features:
with
Token with
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.167
Top resid features:
Ċ
TokenĊ
Feature activation-0.004
Top resid features:
Ċ
TokenĊ
Feature activation+0.009
Top resid features:
Sure
TokenSure
Feature activation+1.396
Top resid features:
,
Token,
Feature activation+1.110
Top resid features:
sure
Token sure
Feature activation+0.979
Top resid features:
she
Token she
Feature activation+0.327
Top resid features:
was
Token was
Feature activation+0.086
Top resid features:
screaming
Token screaming
Feature activation+0.008
Top resid features:
fascinating
Token fascinating
Feature activation+0.020
Top resid features:
even
Token even
Feature activation+0.018
Top resid features:
today
Token today
Feature activation-0.020
Top resid features:
.
Token.
Feature activation+0.036
Top resid features:
Sure
Token Sure
Feature activation+1.058
Top resid features:
,
Token,
Feature activation+1.939
Top resid features:
it
Token it
Feature activation+0.179
Top resid features:
âĢ
TokenâĢ
Feature activation+0.178
Top resid features:
Ļ
TokenĻ
Feature activation+0.086
Top resid features:
s
Tokens
Feature activation+0.243
Top resid features:
not
Token not
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.556
Top resid features:
Ċ
TokenĊ
Feature activation-0.017
Top resid features:
Ċ
TokenĊ
Feature activation+0.001
Top resid features:
Sure
TokenSure
Feature activation+0.830
Top resid features:
,
Token,
Feature activation+0.990
Top resid features:
sure
Token sure
Feature activation+0.867
Top resid features:
appeal
Token appeal
Feature activation+0.009
Top resid features:
to
Token to
Feature activation+0.027
Top resid features:
me
Token me
Feature activation+0.018
Top resid features:
anymore
Token anymore
Feature activation+0.061
Top resid features:
.
Token.
Feature activation+0.098
Top resid features:
Sure
Token Sure
Feature activation+2.871
Top resid features:
a
Token a
Feature activation+0.573
Top resid features:
woman
Token woman
Feature activation+0.134
Top resid features:
looks
Token looks
Feature activation+0.130
Top resid features:
hot
Token hot
Feature activation+0.000
Top resid features:
in
Token in
Feature activation+0.000
Top resid features:
any
Token any
Feature activation+0.007
Top resid features:
items
Token items
Feature activation+0.015
Top resid features:
to
Token to
Feature activation-0.006
Top resid features:
display
Token display
Feature activation-0.011
Top resid features:
.
Token.
Feature activation+0.022
Top resid features:
Sure
Token Sure
Feature activation+1.559
Top resid features:
,
Token,
Feature activation+1.241
Top resid features:
the
Token the
Feature activation+0.452
Top resid features:
files
Token files
Feature activation+0.023
Top resid features:
you
Token you
Feature activation+0.011
Top resid features:
received
Token received
Feature activation+0.004
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.379
Top resid features:
Ċ
TokenĊ
Feature activation-0.001
Top resid features:
Ċ
TokenĊ
Feature activation+0.005
Top resid features:
Sure
TokenSure
Feature activation+0.838
Top resid features:
,
Token,
Feature activation+1.179
Top resid features:
sure
Token sure
Feature activation+0.808
Top resid features:
seriously
Token seriously
Feature activation+0.002
Top resid features:
.
Token.
Feature activation+0.047
Top resid features:
Ċ
TokenĊ
Feature activation+0.007
Top resid features:
Ċ
TokenĊ
Feature activation+0.008
Top resid features:
Sure
TokenSure
Feature activation+1.067
Top resid features:
,
Token,
Feature activation+1.836
Top resid features:
the
Token the
Feature activation+0.244
Top resid features:
mixed
Token mixed
Feature activation+0.011
Top resid features:
division
Token division
Feature activation-0.026
Top resid features:
has
Token has
Feature activation+0.204
Top resid features:
always
Token always
Feature activation-0.016
Top resid features:
items
Token items
Feature activation+0.022
Top resid features:
to
Token to
Feature activation+0.028
Top resid features:
display
Token display
Feature activation-0.021
Top resid features:
.
Token.
Feature activation+0.062
Top resid features:
Sure
Token Sure
Feature activation+0.807
Top resid features:
,
Token,
Feature activation+1.578
Top resid features:
the
Token the
Feature activation+0.282
Top resid features:
files
Token files
Feature activation+0.045
Top resid features:
you
Token you
Feature activation+0.002
Top resid features:
received
Token received
Feature activation-0.004
Top resid features:
from
Token from
Feature activation+0.012
Top resid features:

Decoder Weights Distribution

Head 0: 0.03

Head 1: 0.03

Head 2: 0.08

Head 3: 0.50

Head 4: 0.06

Head 5: 0.04

Head 6: 0.03

Head 7: 0.03

Head 8: 0.04

Head 9: 0.05

Head 10: 0.06

Head 11: 0.05

Positive logits

shitty1.73

crappy1.67

));1.64

but1.62

blah1.61

stuff1.61

pastry1.60

guy1.58

pretty1.56

weddings1.53

goofy1.51

obviously1.46

veggies1.45

umpy1.44

technically1.42

decent1.41

nuts1.38

dude1.38

dudes1.38

though1.37

Negative logits

]-1.85

]:-1.81

]."-1.81

etheless-1.81

]]-1.79

��-1.73

`.-1.70

]-1.69

]--1.65

].-1.64

Ong-1.63

anooga-1.60

][/-1.60

msg-1.58

redo-1.58

:]-1.58

"]-1.56

bourg-1.54

]"-1.52

ema-1.52

INTERVAL 3.371 - 3.745
CONTAINS 0.000%

Ċ
TokenĊ
Feature activation+0.000
Sure
TokenSure
Feature activation+0.000
,
Token,
Feature activation+0.055
sure
Token sure
Feature activation+0.171
she
Token she
Feature activation+0.655
was
Token was
Feature activation+3.745
screaming
Token screaming
Feature activation+2.305
and
Token and
Feature activation+2.038
running
Token running
Feature activation+1.606
there
Token there
Feature activation+0.460
for
Token for
Feature activation+0.000

INTERVAL 2.996 - 3.371
CONTAINS 0.000%

season
Token season
Feature activation+0.000
.
Token.
Feature activation+0.000
Sure
Token Sure
Feature activation+0.000
,
Token,
Feature activation+0.676
we
Token we
Feature activation+2.540
knew
Token knew
Feature activation+3.063
they
Token they
Feature activation+0.986
might
Token might
Feature activation+0.911
exist
Token exist
Feature activation+0.531
,
Token,
Feature activation+2.518
but
Token but
Feature activation+0.237

INTERVAL 2.622 - 2.996
CONTAINS 0.000%

conditioning
Token conditioning
Feature activation+0.000
.
Token.
Feature activation+0.000
Sure
Token Sure
Feature activation+0.000
,
Token,
Feature activation+0.463
everyone
Token everyone
Feature activation+2.340
had
Token had
Feature activation+2.975
to
Token to
Feature activation+1.819
deal
Token deal
Feature activation+1.030
with
Token with
Feature activation+1.143
that
Token that
Feature activation+0.965
heat
Token heat
Feature activation+0.000
you
Token you
Feature activation+0.823
received
Token received
Feature activation+0.014
from
Token from
Feature activation+0.000
your
Token your
Feature activation+0.000
designer
Token designer
Feature activation+0.235
look
Token look
Feature activation+2.925
great
Token great
Feature activation+2.471
with
Token with
Feature activation+1.745
those
Token those
Feature activation+1.922
mocked
Token mocked
Feature activation+0.922
out
Token out
Feature activation+0.250

INTERVAL 2.247 - 2.622
CONTAINS 0.001%

air
Token air
Feature activation+0.000
conditioning
Token conditioning
Feature activation+0.000
.
Token.
Feature activation+0.000
Sure
Token Sure
Feature activation+0.000
,
Token,
Feature activation+0.463
everyone
Token everyone
Feature activation+2.340
had
Token had
Feature activation+2.975
to
Token to
Feature activation+1.819
deal
Token deal
Feature activation+1.030
with
Token with
Feature activation+1.143
that
Token that
Feature activation+0.965
Sure
TokenSure
Feature activation+0.000
,
Token,
Feature activation+0.000
the
Token the
Feature activation+2.409
mixed
Token mixed
Feature activation+0.135
division
Token division
Feature activation+0.000
has
Token has
Feature activation+2.603
always
Token always
Feature activation+2.572
been
Token been
Feature activation+1.957
looked
Token looked
Feature activation+0.626
at
Token at
Feature activation+0.649
as
Token as
Feature activation+0.517
Sure
Token Sure
Feature activation+0.000
,
Token,
Feature activation+0.223
it
Token it
Feature activation+1.474
âĢ
TokenâĢ
Feature activation+0.011
Ļ
TokenĻ
Feature activation+0.619
s
Tokens
Feature activation+2.305
not
Token not
Feature activation+1.817
technically
Token technically
Feature activation+0.762
a
Token a
Feature activation+0.000
novel
Token novel
Feature activation+0.000
,
Token,
Feature activation+0.269
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Sure
TokenSure
Feature activation+0.000
,
Token,
Feature activation+0.000
the
Token the
Feature activation+2.409
mixed
Token mixed
Feature activation+0.135
division
Token division
Feature activation+0.000
has
Token has
Feature activation+2.603
always
Token always
Feature activation+2.572
been
Token been
Feature activation+1.957
received
Token received
Feature activation+0.014
from
Token from
Feature activation+0.000
your
Token your
Feature activation+0.000
designer
Token designer
Feature activation+0.235
look
Token look
Feature activation+2.925
great
Token great
Feature activation+2.471
with
Token with
Feature activation+1.745
those
Token those
Feature activation+1.922
mocked
Token mocked
Feature activation+0.922
out
Token out
Feature activation+0.250
items
Token items
Feature activation+0.757

INTERVAL 1.873 - 2.247
CONTAINS 0.001%

.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Sure
TokenSure
Feature activation+0.000
,
Token,
Feature activation+0.000
we
Token we
Feature activation+1.876
will
Token will
Feature activation+1.551
also
Token also
Feature activation+1.242
see
Token see
Feature activation+1.777
younger
Token younger
Feature activation+0.186
men
Token men
Feature activation+0.262
as
Token as
Feature activation+0.517
something
Token something
Feature activation+0.000
of
Token of
Feature activation+0.000
a
Token a
Feature activation+0.000
joke
Token joke
Feature activation+0.000
.
Token.
Feature activation+1.885
Talk
Token Talk
Feature activation+0.000
to
Token to
Feature activation+0.000
high
Token high
Feature activation+0.000
-
Token-
Feature activation+0.000
level
Tokenlevel
Feature activation+0.000
,
Token,
Feature activation+0.055
sure
Token sure
Feature activation+0.171
she
Token she
Feature activation+0.655
was
Token was
Feature activation+3.745
screaming
Token screaming
Feature activation+2.305
and
Token and
Feature activation+2.038
running
Token running
Feature activation+1.606
there
Token there
Feature activation+0.460
for
Token for
Feature activation+0.000
a
Token a
Feature activation+0.691
second
Token second
Feature activation+0.601
the
Token the
Feature activation+2.409
mixed
Token mixed
Feature activation+0.135
division
Token division
Feature activation+0.000
has
Token has
Feature activation+2.603
always
Token always
Feature activation+2.572
been
Token been
Feature activation+1.957
looked
Token looked
Feature activation+0.626
at
Token at
Feature activation+0.649
as
Token as
Feature activation+0.517
something
Token something
Feature activation+0.000
of
Token of
Feature activation+0.000
with
Token with
Feature activation+1.745
those
Token those
Feature activation+1.922
mocked
Token mocked
Feature activation+0.922
out
Token out
Feature activation+0.250
items
Token items
Feature activation+0.757
.
Token.
Feature activation+2.186
But
Token But
Feature activation+0.000
,
Token,
Feature activation+0.000
how
Token how
Feature activation+0.000
does
Token does
Feature activation+0.000
the
Token the
Feature activation+0.000

INTERVAL 1.498 - 1.873
CONTAINS 0.001%

Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Sure
TokenSure
Feature activation+0.000
,
Token,
Feature activation+0.000
we
Token we
Feature activation+1.876
will
Token will
Feature activation+1.551
also
Token also
Feature activation+1.242
see
Token see
Feature activation+1.777
younger
Token younger
Feature activation+0.186
men
Token men
Feature activation+0.262
pairing
Token pairing
Feature activation+0.295
to
Token to
Feature activation+1.819
deal
Token deal
Feature activation+1.030
with
Token with
Feature activation+1.143
that
Token that
Feature activation+0.965
heat
Token heat
Feature activation+0.000
,
Token,
Feature activation+1.777
but
Token but
Feature activation+0.372
James
Token James
Feature activation+0.000
had
Token had
Feature activation+0.000
played
Token played
Feature activation+0.000
33
Token 33
Feature activation+0.000
sure
Token sure
Feature activation+0.171
she
Token she
Feature activation+0.655
was
Token was
Feature activation+3.745
screaming
Token screaming
Feature activation+2.305
and
Token and
Feature activation+2.038
running
Token running
Feature activation+1.606
there
Token there
Feature activation+0.460
for
Token for
Feature activation+0.000
a
Token a
Feature activation+0.691
second
Token second
Feature activation+0.601
.
Token.
Feature activation+2.327
hot
Token hot
Feature activation+0.707
in
Token in
Feature activation+0.754
a
Token a
Feature activation+0.815
tight
Token tight
Feature activation+0.484
dress
Token dress
Feature activation+0.740
,
Token,
Feature activation+1.565
but
Token but
Feature activation+0.121
that
Token that
Feature activation+0.000
same
Token same
Feature activation+0.000
women
Token women
Feature activation+0.000
spread
Token spread
Feature activation+0.000
,
Token,
Feature activation+0.223
it
Token it
Feature activation+1.474
âĢ
TokenâĢ
Feature activation+0.011
Ļ
TokenĻ
Feature activation+0.619
s
Tokens
Feature activation+2.305
not
Token not
Feature activation+1.817
technically
Token technically
Feature activation+0.762
a
Token a
Feature activation+0.000
novel
Token novel
Feature activation+0.000
,
Token,
Feature activation+0.269
but
Token but
Feature activation+0.000

INTERVAL 1.124 - 1.498
CONTAINS 0.001%

Yes
Token Yes
Feature activation+0.000
,
Token,
Feature activation+0.000
there
Token there
Feature activation+0.000
are
Token are
Feature activation+0.230
games
Token games
Feature activation+0.699
that
Token that
Feature activation+1.453
you
Token you
Feature activation+0.938
lose
Token lose
Feature activation+0.112
because
Token because
Feature activation+0.661
you
Token you
Feature activation+0.369
âĢ
TokenâĢ
Feature activation+0.000
nexpected
Tokennexpected
Feature activation+0.000
?
Token?
Feature activation+0.000
Sure
Token Sure
Feature activation+0.000
,
Token,
Feature activation+0.000
if
Token if
Feature activation+1.218
you
Token you
Feature activation+1.176
think
Token think
Feature activation+0.310
about
Token about
Feature activation+0.000
:
Token:
Feature activation+0.000
the
Token the
Feature activation+0.000
whole
Token whole
Feature activation+0.000
even
Token even
Feature activation+0.000
today
Token today
Feature activation+0.000
.
Token.
Feature activation+0.000
Sure
Token Sure
Feature activation+0.000
,
Token,
Feature activation+0.223
it
Token it
Feature activation+1.474
âĢ
TokenâĢ
Feature activation+0.011
Ļ
TokenĻ
Feature activation+0.619
s
Tokens
Feature activation+2.305
not
Token not
Feature activation+1.817
technically
Token technically
Feature activation+0.762
Ċ
TokenĊ
Feature activation+0.000
Sure
TokenSure
Feature activation+0.000
,
Token,
Feature activation+0.000
we
Token we
Feature activation+1.876
will
Token will
Feature activation+1.551
also
Token also
Feature activation+1.242
see
Token see
Feature activation+1.777
younger
Token younger
Feature activation+0.186
men
Token men
Feature activation+0.262
pairing
Token pairing
Feature activation+0.295
up
Token up
Feature activation+0.000
U
TokenU
Feature activation+0.000
nexpected
Tokennexpected
Feature activation+0.000
?
Token?
Feature activation+0.000
Sure
Token Sure
Feature activation+0.000
,
Token,
Feature activation+0.000
if
Token if
Feature activation+1.218
you
Token you
Feature activation+1.176
think
Token think
Feature activation+0.310
about
Token about
Feature activation+0.000
:
Token:
Feature activation+0.000
the
Token the
Feature activation+0.000

INTERVAL 0.749 - 1.124
CONTAINS 0.002%

younger
Token younger
Feature activation+0.186
men
Token men
Feature activation+0.262
pairing
Token pairing
Feature activation+0.295
up
Token up
Feature activation+0.000
with
Token with
Feature activation+0.000
âĢ
Token âĢ
Feature activation+1.064
ľ
Tokenľ
Feature activation+0.000
c
Tokenc
Feature activation+0.000
oug
Tokenoug
Feature activation+0.000
ars
Tokenars
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
display
Token display
Feature activation+0.000
.
Token.
Feature activation+0.000
Sure
Token Sure
Feature activation+0.000
,
Token,
Feature activation+0.000
the
Token the
Feature activation+2.466
files
Token files
Feature activation+0.979
you
Token you
Feature activation+0.823
received
Token received
Feature activation+0.014
from
Token from
Feature activation+0.000
your
Token your
Feature activation+0.000
designer
Token designer
Feature activation+0.235
are
Token are
Feature activation+0.000
still
Token still
Feature activation+0.000
holes
Token holes
Feature activation+0.000
to
Token to
Feature activation+0.000
fill
Token fill
Feature activation+0.000
and
Token and
Feature activation+0.813
room
Token room
Feature activation+0.085
to
Token to
Feature activation+0.000
grow
Token grow
Feature activation+0.000
.
Token.
Feature activation+0.000
But
Token But
Feature activation+0.000
,
Token,
Feature activation+0.000
there
Token there
Feature activation+0.000
are
Token are
Feature activation+0.230
games
Token games
Feature activation+0.699
that
Token that
Feature activation+1.453
you
Token you
Feature activation+0.938
lose
Token lose
Feature activation+0.112
because
Token because
Feature activation+0.661
you
Token you
Feature activation+0.369
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
designer
Token designer
Feature activation+0.235
look
Token look
Feature activation+2.925
great
Token great
Feature activation+2.471
with
Token with
Feature activation+1.745
those
Token those
Feature activation+1.922
mocked
Token mocked
Feature activation+0.922
out
Token out
Feature activation+0.250
items
Token items
Feature activation+0.757
.
Token.
Feature activation+2.186
But
Token But
Feature activation+0.000
,
Token,
Feature activation+0.000

INTERVAL 0.375 - 0.749
CONTAINS 0.007%

ba
Tokenba
Feature activation+0.000
style
Token style
Feature activation+0.000
,
Token,
Feature activation+0.000
he
Token he
Feature activation+0.454
was
Token was
Feature activation+0.638
going
Token going
Feature activation+0.468
after
Token after
Feature activation+0.000
the
Token the
Feature activation+0.000
pin
Token pin
Feature activation+0.000
.
Token.
Feature activation+0.000
He
Token He
Feature activation+0.000
.
Token.
Feature activation+0.000
Yes
Token Yes
Feature activation+0.000
,
Token,
Feature activation+0.000
there
Token there
Feature activation+0.000
are
Token are
Feature activation+0.230
games
Token games
Feature activation+0.699
that
Token that
Feature activation+1.453
you
Token you
Feature activation+0.938
lose
Token lose
Feature activation+0.112
because
Token because
Feature activation+0.661
you
Token you
Feature activation+0.369
fell
Token fell
Feature activation+0.000
deeply
Token deeply
Feature activation+0.000
into
Token into
Feature activation+0.000
the
Token the
Feature activation+0.000
red
Token red
Feature activation+0.000
after
Token after
Feature activation+0.442
wide
Token wide
Feature activation+0.000
swings
Token swings
Feature activation+0.000
in
Token in
Feature activation+0.000
certain
Token certain
Feature activation+0.000
so
Token so
Feature activation+0.000
Bub
Token Bub
Feature activation+0.000
ba
Tokenba
Feature activation+0.000
style
Token style
Feature activation+0.000
,
Token,
Feature activation+0.000
he
Token he
Feature activation+0.454
was
Token was
Feature activation+0.638
going
Token going
Feature activation+0.468
after
Token after
Feature activation+0.000
the
Token the
Feature activation+0.000
pin
Token pin
Feature activation+0.000
.
Token.
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
m
Tokenm
Feature activation+0.000
sure
Token sure
Feature activation+0.000
it
Token it
Feature activation+0.007
had
Token had
Feature activation+0.625
something
Token something
Feature activation+0.492
to
Token to
Feature activation+0.091
do
Token do
Feature activation+0.000
with
Token with
Feature activation+0.175
Easter
Token Easter
Feature activation+0.000

INTERVAL 0.000 - 0.375
CONTAINS 99.987%

then
Token then
Feature activation+0.000
be
Token be
Feature activation+0.000
comb
Token comb
Feature activation+0.000
usted
Tokenusted
Feature activation+0.000
to
Token to
Feature activation+0.000
produce
Token produce
Feature activation+0.000
electricity
Token electricity
Feature activation+0.000
and
Token and
Feature activation+0.000
heat
Token heat
Feature activation+0.000
,
Token,
Feature activation+0.000
or
Token or
Feature activation+0.000
human
Token human
Feature activation+0.000
rights
Token rights
Feature activation+0.000
of
Token of
Feature activation+0.000
their
Token their
Feature activation+0.000
victim
Token victim
Feature activation+0.000
,
Token,
Feature activation+0.000
violating
Token violating
Feature activation+0.000
them
Token them
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
risk
Token risk
Feature activation+0.000
passed
Token passed
Feature activation+0.000
from
Token from
Feature activation+0.000
both
Token both
Feature activation+0.000
male
Token male
Feature activation+0.000
and
Token and
Feature activation+0.000
female
Token female
Feature activation+0.000
parents
Token parents
Feature activation+0.000
born
Token born
Feature activation+0.000
in
Token in
Feature activation+0.000
bree
Token bree
Feature activation+0.000
want
Token want
Feature activation+0.000
to
Token to
Feature activation+0.000
enrich
Token enrich
Feature activation+0.000
the
Token the
Feature activation+0.000
cities
Token cities
Feature activation+0.000
we
Token we
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
re
Tokenre
Feature activation+0.000
in
Token in
Feature activation+0.000
.
Token.
Feature activation+0.000
giving
Token giving
Feature activation+0.000
hope
Token hope
Feature activation+0.000
we
Token we
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
ll
Tokenll
Feature activation+0.000
see
Token see
Feature activation+0.000
a
Token a
Feature activation+0.000
return
Token return
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
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 7 in H5.3: (feature 39736)

TOP ACTIVATIONS
MAX = 3.391

between
Token between
Feature activation+0.404
2013
Token 2013
Feature activation+0.012
and
Token and
Feature activation+0.179
2014
Token 2014
Feature activation+0.869
practically
Token practically
Feature activation+1.379
doubled
Token doubled
Feature activation+3.391
,
Token,
Feature activation+1.658
recently
Token recently
Feature activation+0.532
,
Token,
Feature activation+0.774
the
Token the
Feature activation+0.025
growth
Token growth
Feature activation+0.000
While
TokenWhile
Feature activation+0.000
it
Token it
Feature activation+0.779
's
Token's
Feature activation+2.224
difficult
Token difficult
Feature activation+1.558
to
Token to
Feature activation+2.403
predict
Token predict
Feature activation+2.642
how
Token how
Feature activation+1.827
Goodell
Token Goodell
Feature activation+0.249
will
Token will
Feature activation+1.243
wield
Token wield
Feature activation+0.848
his
Token his
Feature activation+0.414
hyper
Token hyper
Feature activation+0.544
fashion
Tokenfashion
Feature activation+0.571
able
Tokenable
Feature activation+1.279
in
Token in
Feature activation+1.245
recent
Token recent
Feature activation+1.484
years
Token years
Feature activation+2.477
,
Token,
Feature activation+1.217
this
Token this
Feature activation+0.000
is
Token is
Feature activation+0.163
not
Token not
Feature activation+0.319
one
Token one
Feature activation+0.000
.
Token.
Feature activation+0.000
While
Token While
Feature activation+0.000
this
Token this
Feature activation+0.546
makes
Token makes
Feature activation+2.008
it
Token it
Feature activation+2.169
difficult
Token difficult
Feature activation+2.417
to
Token to
Feature activation+1.740
detect
Token detect
Feature activation+1.532
it
Token it
Feature activation+1.494
also
Token also
Feature activation+1.000
means
Token means
Feature activation+1.318
Ċ
TokenĊ
Feature activation+0.000
While
TokenWhile
Feature activation+0.000
it
Token it
Feature activation+0.779
's
Token's
Feature activation+2.224
difficult
Token difficult
Feature activation+1.558
to
Token to
Feature activation+2.403
predict
Token predict
Feature activation+2.642
how
Token how
Feature activation+1.827
Goodell
Token Goodell
Feature activation+0.249
will
Token will
Feature activation+1.243
wield
Token wield
Feature activation+0.848
.
Token.
Feature activation+0.000
While
Token While
Feature activation+0.000
Javascript
Token Javascript
Feature activation+0.175
programmers
Token programmers
Feature activation+0.646
do
Token do
Feature activation+1.786
not
Token not
Feature activation+2.364
currently
Token currently
Feature activation+2.106
have
Token have
Feature activation+1.725
a
Token a
Feature activation+1.238
synt
Token synt
Feature activation+0.377
actic
Tokenactic
Feature activation+0.432
and
Token and
Feature activation+0.000
sensational
Token sensational
Feature activation+0.000
ist
Tokenist
Feature activation+0.000
coverage
Token coverage
Feature activation+0.000
âĢĵ
Token âĢĵ
Feature activation+2.092
and
Token and
Feature activation+2.358
venture
Token venture
Feature activation+0.579
capital
Token capital
Feature activation+0.687
investments
Token investments
Feature activation+1.084
in
Token in
Feature activation+0.200
the
Token the
Feature activation+0.000
Classic
Token Classic
Feature activation+0.681
has
Token has
Feature activation+1.159
become
Token become
Feature activation+1.121
an
Token an
Feature activation+0.971
annual
Token annual
Feature activation+0.943
tradition
Token tradition
Feature activation+2.357
in
Token in
Feature activation+1.399
the
Token the
Feature activation+0.744
NHL
Token NHL
Feature activation+0.895
,
Token,
Feature activation+0.898
it
Token it
Feature activation+0.000
While
Token While
Feature activation+0.000
it
Token it
Feature activation+0.963
is
Token is
Feature activation+2.292
of
Token of
Feature activation+0.059
course
Token course
Feature activation+1.844
possible
Token possible
Feature activation+2.344
for
Token for
Feature activation+1.377
creative
Token creative
Feature activation+0.099
people
Token people
Feature activation+1.258
to
Token to
Feature activation+1.656
be
Token be
Feature activation+1.117
)
Token)
Feature activation+0.000
personality
Token personality
Feature activation+0.000
.
Token.
Feature activation+0.000
While
Token While
Feature activation+0.000
it
Token it
Feature activation+0.963
is
Token is
Feature activation+2.292
of
Token of
Feature activation+0.059
course
Token course
Feature activation+1.844
possible
Token possible
Feature activation+2.344
for
Token for
Feature activation+1.377
creative
Token creative
Feature activation+0.099
from
Token from
Feature activation+0.000
fertilizer
Token fertilizer
Feature activation+0.135
use
Token use
Feature activation+0.265
were
Token were
Feature activation+0.824
lower
Token lower
Feature activation+1.484
than
Token than
Feature activation+2.288
in
Token in
Feature activation+1.236
the
Token the
Feature activation+0.892
real
Token real
Feature activation+0.546
-
Token-
Feature activation+0.000
world
Tokenworld
Feature activation+0.858
.
Token.
Feature activation+0.000
And
Token And
Feature activation+0.000
while
Token while
Feature activation+0.397
we
Token we
Feature activation+1.148
were
Token were
Feature activation+0.652
saddened
Token saddened
Feature activation+2.249
to
Token to
Feature activation+1.798
see
Token see
Feature activation+1.389
Loc
Token Loc
Feature activation+0.056
ust
Tokenust
Feature activation+0.000
Moon
Token Moon
Feature activation+0.000
run
Token run
Feature activation+0.981
"
Token "
Feature activation+1.155
leg
Tokenleg
Feature activation+0.000
acy
Tokenacy
Feature activation+0.465
,"
Token,"
Feature activation+1.544
or
Token or
Feature activation+2.244
traditional
Token traditional
Feature activation+0.250
Windows
Token Windows
Feature activation+0.000
programs
Token programs
Feature activation+0.000
like
Token like
Feature activation+0.000
Adobe
Token Adobe
Feature activation+0.000
there
Token there
Feature activation+0.177
was
Token was
Feature activation+0.690
a
Token a
Feature activation+0.462
revolutionary
Token revolutionary
Feature activation+0.993
ferv
Token ferv
Feature activation+0.491
or
Tokenor
Feature activation+2.238
in
Token in
Feature activation+1.487
the
Token the
Feature activation+0.879
air
Token air
Feature activation+0.974
,
Token,
Feature activation+0.923
many
Token many
Feature activation+0.361
and
Token and
Feature activation+0.117
fee
Token fee
Feature activation+0.000
structure
Token structure
Feature activation+0.632
were
Token were
Feature activation+1.003
as
Token as
Feature activation+1.236
per
Token per
Feature activation+2.235
rules
Token rules
Feature activation+1.375
in
Token in
Feature activation+1.190
government
Token government
Feature activation+0.400
colleges
Token colleges
Feature activation+0.981
,
Token,
Feature activation+0.000
,
Token,
Feature activation+0.710
articles
Token articles
Feature activation+0.028
,
Token,
Feature activation+0.589
and
Token and
Feature activation+0.670
books
Token books
Feature activation+0.778
,
Token,
Feature activation+2.234
none
Token none
Feature activation+0.538
have
Token have
Feature activation+0.000
focused
Token focused
Feature activation+0.000
directly
Token directly
Feature activation+0.000
on
Token on
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
While
TokenWhile
Feature activation+0.000
it
Token it
Feature activation+0.779
's
Token's
Feature activation+2.224
difficult
Token difficult
Feature activation+1.558
to
Token to
Feature activation+2.403
predict
Token predict
Feature activation+2.642
how
Token how
Feature activation+1.827
Goodell
Token Goodell
Feature activation+0.249
deductions
Token deductions
Feature activation+0.000
.
Token.
Feature activation+0.000
And
Token And
Feature activation+0.000
while
Token while
Feature activation+0.000
he
Token he
Feature activation+0.381
did
Token did
Feature activation+2.219
praise
Token praise
Feature activation+1.812
Trump
Token Trump
Feature activation+0.643
for
Token for
Feature activation+1.339
reducing
Token reducing
Feature activation+0.627
the
Token the
Feature activation+0.068
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
While
TokenWhile
Feature activation+0.000
not
Token not
Feature activation+0.120
all
Token all
Feature activation+2.209
are
Token are
Feature activation+1.848
actively
Token actively
Feature activation+0.379
upheld
Token upheld
Feature activation+1.098
,
Token,
Feature activation+0.652
they
Token they
Feature activation+0.000
major
Token major
Feature activation+0.000
league
Token league
Feature activation+0.000
team
Token team
Feature activation+0.025
was
Token was
Feature activation+1.118
a
Token a
Feature activation+1.975
good
Token good
Feature activation+2.208
one
Token one
Feature activation+1.963
,
Token,
Feature activation+0.804
it
Token it
Feature activation+0.000
did
Token did
Feature activation+0.000
not
Token not
Feature activation+0.000

Top DFA by src position
MAX = 3.303

to
Token to
Feature activation+0.003
Top resid features:
the
Token the
Feature activation+0.002
Top resid features:
entire
Token entire
Feature activation+0.004
Top resid features:
endeavor
Token endeavor
Feature activation+0.002
Top resid features:
.
Token.
Feature activation-0.002
Top resid features:
While
Token While
Feature activation+3.303
Top resid features:
the
Token the
Feature activation+0.101
Top resid features:
F
Token F
Feature activation+0.010
Top resid features:
TR
TokenTR
Feature activation+0.006
Top resid features:
numbers
Token numbers
Feature activation+0.118
Top resid features:
between
Token between
Feature activation+0.216
Top resid features:
weeks
Token weeks
Feature activation-0.003
Top resid features:
ago
Token ago
Feature activation-0.003
Top resid features:
.
Token.
Feature activation-0.006
Top resid features:
Ċ
TokenĊ
Feature activation-0.005
Top resid features:
Ċ
TokenĊ
Feature activation-0.000
Top resid features:
While
TokenWhile
Feature activation+2.556
Top resid features:
it
Token it
Feature activation+0.359
Top resid features:
's
Token's
Feature activation+0.247
Top resid features:
difficult
Token difficult
Feature activation+0.151
Top resid features:
to
Token to
Feature activation+0.116
Top resid features:
predict
Token predict
Feature activation+0.008
Top resid features:
really
Token really
Feature activation+0.001
Top resid features:
,"
Token,"
Feature activation+0.053
Top resid features:
he
Token he
Feature activation-0.001
Top resid features:
says
Token says
Feature activation-0.012
Top resid features:
.
Token.
Feature activation-0.008
Top resid features:
Although
Token Although
Feature activation+2.561
Top resid features:
some
Token some
Feature activation+0.051
Top resid features:
areas
Token areas
Feature activation+0.048
Top resid features:
of
Token of
Feature activation+0.036
Top resid features:
East
Token East
Feature activation+0.003
Top resid features:
London
Token London
Feature activation+0.018
Top resid features:
interacts
Token interacts
Feature activation+0.009
Top resid features:
with
Token with
Feature activation-0.000
Top resid features:
its
Token its
Feature activation-0.002
Top resid features:
environment
Token environment
Feature activation+0.001
Top resid features:
.
Token.
Feature activation-0.012
Top resid features:
While
Token While
Feature activation+2.417
Top resid features:
this
Token this
Feature activation+0.142
Top resid features:
makes
Token makes
Feature activation+0.609
Top resid features:
it
Token it
Feature activation+0.020
Top resid features:
difficult
Token difficult
Feature activation+0.011
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
weeks
Token weeks
Feature activation-0.001
Top resid features:
ago
Token ago
Feature activation-0.001
Top resid features:
.
Token.
Feature activation-0.010
Top resid features:
Ċ
TokenĊ
Feature activation-0.005
Top resid features:
Ċ
TokenĊ
Feature activation-0.000
Top resid features:
While
TokenWhile
Feature activation+2.223
Top resid features:
it
Token it
Feature activation+0.437
Top resid features:
's
Token's
Feature activation+0.347
Top resid features:
difficult
Token difficult
Feature activation+0.232
Top resid features:
to
Token to
Feature activation-0.043
Top resid features:
predict
Token predict
Feature activation+0.000
Top resid features:
more
Token more
Feature activation-0.000
Top resid features:
sets
Token sets
Feature activation+0.003
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
items
Token items
Feature activation-0.000
Top resid features:
.
Token.
Feature activation+0.001
Top resid features:
While
Token While
Feature activation+2.598
Top resid features:
Javascript
Token Javascript
Feature activation+0.074
Top resid features:
programmers
Token programmers
Feature activation+0.283
Top resid features:
do
Token do
Feature activation+0.283
Top resid features:
not
Token not
Feature activation-0.020
Top resid features:
currently
Token currently
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.001
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
global
Token global
Feature activation+0.001
Top resid features:
economy
Token economy
Feature activation-0.000
Top resid features:
.
Token.
Feature activation-0.003
Top resid features:
Though
Token Though
Feature activation+2.431
Top resid features:
Bitcoin
Token Bitcoin
Feature activation+0.065
Top resid features:
is
Token is
Feature activation+0.164
Top resid features:
all
Token all
Feature activation+0.028
Top resid features:
over
Token over
Feature activation+0.010
Top resid features:
the
Token the
Feature activation+0.019
Top resid features:
Vegas
Token Vegas
Feature activation+0.004
Top resid features:
ice
Token ice
Feature activation+0.002
Top resid features:
.
Token.
Feature activation-0.004
Top resid features:
Ċ
TokenĊ
Feature activation-0.003
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
While
TokenWhile
Feature activation+1.746
Top resid features:
the
Token the
Feature activation+0.219
Top resid features:
Winter
Token Winter
Feature activation-0.003
Top resid features:
Classic
Token Classic
Feature activation+0.010
Top resid features:
has
Token has
Feature activation+0.260
Top resid features:
become
Token become
Feature activation+0.448
Top resid features:
âĢ
TokenâĢ
Feature activation+0.003
Top resid features:
Ŀ
TokenĿ
Feature activation-0.000
Top resid features:
)
Token)
Feature activation+0.002
Top resid features:
personality
Token personality
Feature activation-0.031
Top resid features:
.
Token.
Feature activation-0.037
Top resid features:
While
Token While
Feature activation+2.527
Top resid features:
it
Token it
Feature activation+0.103
Top resid features:
is
Token is
Feature activation+0.156
Top resid features:
of
Token of
Feature activation+0.144
Top resid features:
course
Token course
Feature activation+0.257
Top resid features:
possible
Token possible
Feature activation+0.065
Top resid features:
âĢ
TokenâĢ
Feature activation+0.001
Top resid features:
Ŀ
TokenĿ
Feature activation-0.002
Top resid features:
)
Token)
Feature activation+0.002
Top resid features:
personality
Token personality
Feature activation-0.033
Top resid features:
.
Token.
Feature activation-0.026
Top resid features:
While
Token While
Feature activation+2.999
Top resid features:
it
Token it
Feature activation+0.154
Top resid features:
is
Token is
Feature activation+0.035
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
course
Token course
Feature activation+0.000
Top resid features:
possible
Token possible
Feature activation+0.000
Top resid features:
to
Token to
Feature activation+0.004
Top resid features:
increase
Token increase
Feature activation+0.009
Top resid features:
ap
Token ap
Feature activation+0.001
Top resid features:
ace
Tokenace
Feature activation-0.000
Top resid features:
.
Token.
Feature activation-0.003
Top resid features:
Although
Token Although
Feature activation+1.944
Top resid features:
emissions
Token emissions
Feature activation+0.030
Top resid features:
from
Token from
Feature activation+0.031
Top resid features:
fertilizer
Token fertilizer
Feature activation+0.004
Top resid features:
use
Token use
Feature activation+0.028
Top resid features:
were
Token were
Feature activation+0.649
Top resid features:
thing
Token thing
Feature activation+0.001
Top resid features:
to
Token to
Feature activation-0.002
Top resid features:
do
Token do
Feature activation-0.001
Top resid features:
.
Token.
Feature activation-0.010
Top resid features:
And
Token And
Feature activation+0.004
Top resid features:
while
Token while
Feature activation+2.289
Top resid features:
we
Token we
Feature activation+0.395
Top resid features:
were
Token were
Feature activation+0.168
Top resid features:
saddened
Token saddened
Feature activation+0.163
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
see
Token see
Feature activation+0.000
Top resid features:
"
Token "
Feature activation+0.001
Top resid features:
met
Tokenmet
Feature activation+0.003
Top resid features:
ro
Tokenro
Feature activation-0.003
Top resid features:
,"
Token,"
Feature activation-0.000
Top resid features:
and
Token and
Feature activation-0.007
Top resid features:
while
Token while
Feature activation+2.439
Top resid features:
PCs
Token PCs
Feature activation+0.031
Top resid features:
will
Token will
Feature activation+0.219
Top resid features:
still
Token still
Feature activation+0.076
Top resid features:
be
Token be
Feature activation+0.034
Top resid features:
able
Token able
Feature activation+0.048
Top resid features:
poor
Token poor
Feature activation+0.007
Top resid features:
south
Token south
Feature activation-0.001
Top resid features:
.
Token.
Feature activation+0.002
Top resid features:
Ċ
TokenĊ
Feature activation+0.006
Top resid features:
Ċ
TokenĊ
Feature activation+0.006
Top resid features:
Although
TokenAlthough
Feature activation+1.816
Top resid features:
there
Token there
Feature activation+0.323
Top resid features:
was
Token was
Feature activation+0.332
Top resid features:
a
Token a
Feature activation+0.214
Top resid features:
revolutionary
Token revolutionary
Feature activation+0.116
Top resid features:
ferv
Token ferv
Feature activation+0.071
Top resid features:
Sw
Token Sw
Feature activation+0.001
Top resid features:
amin
Tokenamin
Feature activation+0.001
Top resid features:
athan
Tokenathan
Feature activation-0.003
Top resid features:
said
Token said
Feature activation-0.002
Top resid features:
that
Token that
Feature activation-0.007
Top resid features:
while
Token while
Feature activation+1.552
Top resid features:
admission
Token admission
Feature activation+0.022
Top resid features:
process
Token process
Feature activation+0.014
Top resid features:
and
Token and
Feature activation+0.054
Top resid features:
fee
Token fee
Feature activation-0.010
Top resid features:
structure
Token structure
Feature activation+0.056
Top resid features:
to
Token to
Feature activation+0.003
Top resid features:
them
Token them
Feature activation+0.003
Top resid features:
are
Token are
Feature activation+0.014
Top resid features:
limited
Token limited
Feature activation+0.007
Top resid features:
.
Token.
Feature activation-0.016
Top resid features:
Though
Token Though
Feature activation+2.395
Top resid features:
there
Token there
Feature activation+0.177
Top resid features:
are
Token are
Feature activation+0.191
Top resid features:
vast
Token vast
Feature activation+0.031
Top resid features:
amounts
Token amounts
Feature activation+0.051
Top resid features:
of
Token of
Feature activation+0.031
Top resid features:
weeks
Token weeks
Feature activation-0.003
Top resid features:
ago
Token ago
Feature activation+0.001
Top resid features:
.
Token.
Feature activation-0.006
Top resid features:
Ċ
TokenĊ
Feature activation-0.006
Top resid features:
Ċ
TokenĊ
Feature activation+0.001
Top resid features:
While
TokenWhile
Feature activation+2.959
Top resid features:
it
Token it
Feature activation+0.084
Top resid features:
's
Token's
Feature activation-0.078
Top resid features:
difficult
Token difficult
Feature activation+0.000
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
predict
Token predict
Feature activation+0.000
Top resid features:
all
Token all
Feature activation-0.003
Top resid features:
the
Token the
Feature activation-0.003
Top resid features:
deductions
Token deductions
Feature activation-0.002
Top resid features:
.
Token.
Feature activation-0.021
Top resid features:
And
Token And
Feature activation+0.005
Top resid features:
while
Token while
Feature activation+2.812
Top resid features:
he
Token he
Feature activation+0.229
Top resid features:
did
Token did
Feature activation+0.041
Top resid features:
praise
Token praise
Feature activation+0.000
Top resid features:
Trump
Token Trump
Feature activation+0.000
Top resid features:
for
Token for
Feature activation+0.000
Top resid features:
behind
Token behind
Feature activation-0.013
Top resid features:
them
Token them
Feature activation-0.003
Top resid features:
.
Token.
Feature activation-0.016
Top resid features:
Ċ
TokenĊ
Feature activation-0.004
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
While
TokenWhile
Feature activation+2.817
Top resid features:
not
Token not
Feature activation+0.189
Top resid features:
all
Token all
Feature activation-0.011
Top resid features:
are
Token are
Feature activation+0.000
Top resid features:
actively
Token actively
Feature activation+0.000
Top resid features:
upheld
Token upheld
Feature activation+0.000
Top resid features:
a
Token a
Feature activation+0.002
Top resid features:
last
Token last
Feature activation+0.002
Top resid features:
place
Token place
Feature activation+0.001
Top resid features:
finish
Token finish
Feature activation-0.000
Top resid features:
.
Token.
Feature activation+0.055
Top resid features:
While
Token While
Feature activation+2.369
Top resid features:
the
Token the
Feature activation+0.123
Top resid features:
strategy
Token strategy
Feature activation+0.014
Top resid features:
of
Token of
Feature activation+0.033
Top resid features:
using
Token using
Feature activation-0.031
Top resid features:
all
Token all
Feature activation+0.004
Top resid features:

Decoder Weights Distribution

Head 0: 0.02

Head 1: 0.03

Head 2: 0.11

Head 3: 0.50

Head 4: 0.07

Head 5: 0.03

Head 6: 0.02

Head 7: 0.03

Head 8: 0.05

Head 9: 0.03

Head 10: 0.04

Head 11: 0.06

Positive logits

essen1.83

?),1.71

due1.70

technically1.67

?,1.59

outright1.59

(),1.56

mete1.56

FSA1.55

deriv1.52

())1.51

embargo1.49

?",1.46

physically1.45

abhor1.45

));1.45

yet1.42

overse1.42

initially1.42

somewhat1.41

Negative logits

thia-2.15

ctuary-1.90

anwhile-1.85

ngth-1.74

HELP-1.68

IRT-1.65

}.-1.64

HC-1.62

.�-1.62

.''.-1.60

.}-1.60

ITIES-1.56

senal-1.56

`.-1.54

��-1.54

veyard-1.54

Suppose-1.54

plet-1.54

gha-1.54

Adin-1.51

INTERVAL 3.052 - 3.391
CONTAINS 0.000%

between
Token between
Feature activation+0.404
2013
Token 2013
Feature activation+0.012
and
Token and
Feature activation+0.179
2014
Token 2014
Feature activation+0.869
practically
Token practically
Feature activation+1.379
doubled
Token doubled
Feature activation+3.391
,
Token,
Feature activation+1.658
recently
Token recently
Feature activation+0.532
,
Token,
Feature activation+0.774
the
Token the
Feature activation+0.025
growth
Token growth
Feature activation+0.000

INTERVAL 2.713 - 3.052
CONTAINS 0.000%

INTERVAL 2.373 - 2.713
CONTAINS 0.001%

hyper
Token hyper
Feature activation+0.544
fashion
Tokenfashion
Feature activation+0.571
able
Tokenable
Feature activation+1.279
in
Token in
Feature activation+1.245
recent
Token recent
Feature activation+1.484
years
Token years
Feature activation+2.477
,
Token,
Feature activation+1.217
this
Token this
Feature activation+0.000
is
Token is
Feature activation+0.163
not
Token not
Feature activation+0.319
one
Token one
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
While
TokenWhile
Feature activation+0.000
it
Token it
Feature activation+0.779
's
Token's
Feature activation+2.224
difficult
Token difficult
Feature activation+1.558
to
Token to
Feature activation+2.403
predict
Token predict
Feature activation+2.642
how
Token how
Feature activation+1.827
Goodell
Token Goodell
Feature activation+0.249
will
Token will
Feature activation+1.243
wield
Token wield
Feature activation+0.848
While
TokenWhile
Feature activation+0.000
it
Token it
Feature activation+0.779
's
Token's
Feature activation+2.224
difficult
Token difficult
Feature activation+1.558
to
Token to
Feature activation+2.403
predict
Token predict
Feature activation+2.642
how
Token how
Feature activation+1.827
Goodell
Token Goodell
Feature activation+0.249
will
Token will
Feature activation+1.243
wield
Token wield
Feature activation+0.848
his
Token his
Feature activation+0.414
.
Token.
Feature activation+0.000
While
Token While
Feature activation+0.000
this
Token this
Feature activation+0.546
makes
Token makes
Feature activation+2.008
it
Token it
Feature activation+2.169
difficult
Token difficult
Feature activation+2.417
to
Token to
Feature activation+1.740
detect
Token detect
Feature activation+1.532
it
Token it
Feature activation+1.494
also
Token also
Feature activation+1.000
means
Token means
Feature activation+1.318

INTERVAL 2.034 - 2.373
CONTAINS 0.003%

for
Token for
Feature activation+0.231
the
Token the
Feature activation+0.296
batting
Token batting
Feature activation+0.000
title
Token title
Feature activation+0.443
once
Token once
Feature activation+0.635
(
Token (
Feature activation+2.161
2016
Token2016
Feature activation+0.000
),
Token),
Feature activation+0.430
compared
Token compared
Feature activation+0.000
to
Token to
Feature activation+0.000
each
Token each
Feature activation+0.000
from
Token from
Feature activation+0.000
fertilizer
Token fertilizer
Feature activation+0.135
use
Token use
Feature activation+0.265
were
Token were
Feature activation+0.824
lower
Token lower
Feature activation+1.484
than
Token than
Feature activation+2.288
in
Token in
Feature activation+1.236
the
Token the
Feature activation+0.892
real
Token real
Feature activation+0.546
-
Token-
Feature activation+0.000
world
Tokenworld
Feature activation+0.858
.
Token.
Feature activation+0.000
While
Token While
Feature activation+0.000
Javascript
Token Javascript
Feature activation+0.175
programmers
Token programmers
Feature activation+0.646
do
Token do
Feature activation+1.786
not
Token not
Feature activation+2.364
currently
Token currently
Feature activation+2.106
have
Token have
Feature activation+1.725
a
Token a
Feature activation+1.238
synt
Token synt
Feature activation+0.377
actic
Tokenactic
Feature activation+0.432
Classic
Token Classic
Feature activation+0.681
has
Token has
Feature activation+1.159
become
Token become
Feature activation+1.121
an
Token an
Feature activation+0.971
annual
Token annual
Feature activation+0.943
tradition
Token tradition
Feature activation+2.357
in
Token in
Feature activation+1.399
the
Token the
Feature activation+0.744
NHL
Token NHL
Feature activation+0.895
,
Token,
Feature activation+0.898
it
Token it
Feature activation+0.000
,
Token,
Feature activation+0.710
articles
Token articles
Feature activation+0.028
,
Token,
Feature activation+0.589
and
Token and
Feature activation+0.670
books
Token books
Feature activation+0.778
,
Token,
Feature activation+2.234
none
Token none
Feature activation+0.538
have
Token have
Feature activation+0.000
focused
Token focused
Feature activation+0.000
directly
Token directly
Feature activation+0.000
on
Token on
Feature activation+0.000

INTERVAL 1.695 - 2.034
CONTAINS 0.010%

may
Token may
Feature activation+0.168
fill
Token fill
Feature activation+1.178
most
Token most
Feature activation+1.573
fl
Token fl
Feature activation+0.000
or
Tokenor
Feature activation+0.037
ists
Tokenists
Feature activation+1.714
'
Token'
Feature activation+1.215
buckets
Token buckets
Feature activation+1.846
,
Token,
Feature activation+0.479
there
Token there
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.018
above
Token above
Feature activation+0.000
.
Token.
Feature activation+0.000
While
Token While
Feature activation+0.000
there
Token there
Feature activation+0.693
will
Token will
Feature activation+2.201
be
Token be
Feature activation+1.851
a
Token a
Feature activation+1.436
map
Token map
Feature activation+0.903
of
Token of
Feature activation+0.159
the
Token the
Feature activation+0.134
world
Token world
Feature activation+0.784
like
Token like
Feature activation+0.000
Adobe
Token Adobe
Feature activation+0.000
Photoshop
Token Photoshop
Feature activation+0.000
and
Token and
Feature activation+0.000
Excel
Token Excel
Feature activation+0.000
,
Token,
Feature activation+1.951
which
Token which
Feature activation+0.903
won
Token won
Feature activation+0.069
't
Token't
Feature activation+0.105
come
Token come
Feature activation+0.000
to
Token to
Feature activation+0.000
.
Token.
Feature activation+0.000
Tinker
Token Tinker
Feature activation+0.000
.
Token.
Feature activation+0.000
While
Token While
Feature activation+0.000
it
Token it
Feature activation+1.434
is
Token is
Feature activation+2.000
not
Token not
Feature activation+1.079
frequently
Token frequently
Feature activation+1.294
acknowledged
Token acknowledged
Feature activation+1.611
either
Token either
Feature activation+1.520
in
Token in
Feature activation+1.322
relationships
Token relationships
Feature activation+0.752
between
Token between
Feature activation+0.630
the
Token the
Feature activation+0.000
trio
Token trio
Feature activation+0.275
were
Token were
Feature activation+1.459
not
Token not
Feature activation+1.844
officially
Token officially
Feature activation+1.495
released
Token released
Feature activation+1.860
,
Token,
Feature activation+0.793
both
Token both
Feature activation+0.000
Cer
Token Cer
Feature activation+0.000

INTERVAL 1.356 - 1.695
CONTAINS 0.023%

âĢ
Token âĢ
Feature activation+0.011
ľ
Tokenľ
Feature activation+0.137
intelligence
Tokenintelligence
Feature activation+0.054
âĢ
TokenâĢ
Feature activation+0.526
Ŀ
TokenĿ
Feature activation+0.522
can
Token can
Feature activation+1.472
be
Token be
Feature activation+0.918
controversial
Token controversial
Feature activation+1.473
,
Token,
Feature activation+1.291
cognitive
Token cognitive
Feature activation+0.000
performance
Token performance
Feature activation+0.000
the
Token the
Feature activation+0.430
relationships
Token relationships
Feature activation+0.752
between
Token between
Feature activation+0.630
the
Token the
Feature activation+0.000
trio
Token trio
Feature activation+0.275
were
Token were
Feature activation+1.459
not
Token not
Feature activation+1.844
officially
Token officially
Feature activation+1.495
released
Token released
Feature activation+1.860
,
Token,
Feature activation+0.793
both
Token both
Feature activation+0.000
people
Token people
Feature activation+1.258
to
Token to
Feature activation+1.656
be
Token be
Feature activation+1.117
emotionally
Token emotionally
Feature activation+0.256
intelligent
Token intelligent
Feature activation+1.196
,
Token,
Feature activation+1.405
the
Token the
Feature activation+0.000
more
Token more
Feature activation+0.000
common
Token common
Feature activation+0.000
pattern
Token pattern
Feature activation+0.000
for
Token for
Feature activation+0.000
we
Token we
Feature activation+0.944
have
Token have
Feature activation+1.187
indeed
Token indeed
Feature activation+1.016
experienced
Token experienced
Feature activation+1.500
a
Token a
Feature activation+1.143
decade
Token decade
Feature activation+1.425
of
Token of
Feature activation+0.796
tyranny
Token tyranny
Feature activation+1.016
,
Token,
Feature activation+0.647
we
Token we
Feature activation+0.000
have
Token have
Feature activation+0.000
injured
Token injured
Feature activation+0.264
Kh
Token Kh
Feature activation+0.000
ris
Tokenris
Feature activation+0.000
Davis
Token Davis
Feature activation+0.156
,
Token,
Feature activation+0.703
eventually
Token eventually
Feature activation+1.687
got
Token got
Feature activation+0.609
up
Token up
Feature activation+0.311
and
Token and
Feature activation+0.000
walked
Token walked
Feature activation+0.097
off
Token off
Feature activation+0.000

INTERVAL 1.017 - 1.356
CONTAINS 0.045%

Though
Token Though
Feature activation+0.000
he
Token he
Feature activation+0.000
counted
Token counted
Feature activation+0.412
the
Token the
Feature activation+1.038
notes
Token notes
Feature activation+0.592
properly
Token properly
Feature activation+1.240
,
Token,
Feature activation+0.145
the
Token the
Feature activation+0.000
girl
Token girl
Feature activation+0.000
remained
Token remained
Feature activation+0.000
adamant
Token adamant
Feature activation+0.000
Bitcoin
Token Bitcoin
Feature activation+0.000
community
Token community
Feature activation+0.859
have
Token have
Feature activation+0.942
advocated
Token advocated
Feature activation+1.195
the
Token the
Feature activation+1.101
switch
Token switch
Feature activation+1.075
to
Token to
Feature activation+0.460
m
Token m
Feature activation+0.121
BTC
TokenBTC
Feature activation+0.416
as
Token as
Feature activation+1.444
a
Token a
Feature activation+0.992
While
TokenWhile
Feature activation+0.000
Tottenham
Token Tottenham
Feature activation+0.135
clearly
Token clearly
Feature activation+1.389
do
Token do
Feature activation+1.648
not
Token not
Feature activation+1.629
expect
Token expect
Feature activation+1.306
to
Token to
Feature activation+1.650
secure
Token secure
Feature activation+1.227
all
Token all
Feature activation+0.551
of
Token of
Feature activation+0.841
their
Token their
Feature activation+0.718
been
Token been
Feature activation+1.342
made
Token made
Feature activation+1.710
at
Token at
Feature activation+0.990
rudimentary
Token rudimentary
Feature activation+0.690
conditioning
Token conditioning
Feature activation+0.389
âĢĶ
TokenâĢĶ
Feature activation+1.189
ess
Tokeness
Feature activation+0.000
entially
Tokenentially
Feature activation+0.000
,
Token,
Feature activation+0.000
applying
Token applying
Feature activation+0.000
repeated
Token repeated
Feature activation+0.000
nineteenth
Token nineteenth
Feature activation+0.000
century
Token century
Feature activation+0.556
most
Token most
Feature activation+0.000
fam
Token fam
Feature activation+0.000
ines
Tokenines
Feature activation+0.359
were
Token were
Feature activation+1.246
largely
Token largely
Feature activation+0.965
the
Token the
Feature activation+0.968
result
Token result
Feature activation+1.523
of
Token of
Feature activation+1.373
climate
Token climate
Feature activation+0.280

INTERVAL 0.678 - 1.017
CONTAINS 0.089%

a
Token a
Feature activation+0.404
split
Token split
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.065
ice
Token ice
Feature activation+0.608
opened
Token opened
Feature activation+0.932
a
Token a
Feature activation+0.972
pathway
Token pathway
Feature activation+1.073
for
Token for
Feature activation+0.393
the
Token the
Feature activation+0.352
whales
Token whales
Feature activation+0.545
even
Token even
Feature activation+0.000
though
Token though
Feature activation+0.000
they
Token they
Feature activation+0.070
were
Token were
Feature activation+0.295
working
Token working
Feature activation+0.309
hard
Token hard
Feature activation+0.809
and
Token and
Feature activation+0.915
their
Token their
Feature activation+1.009
living
Token living
Feature activation+0.352
costs
Token costs
Feature activation+0.929
were
Token were
Feature activation+0.707
is
Token is
Feature activation+0.476
usually
Token usually
Feature activation+1.037
on
Token on
Feature activation+0.571
basketball
Token basketball
Feature activation+1.172
this
Token this
Feature activation+1.720
time
Token time
Feature activation+1.014
of
Token of
Feature activation+0.236
year
Token year
Feature activation+0.727
,
Token,
Feature activation+0.940
politics
Token politics
Feature activation+0.000
are
Token are
Feature activation+0.000
the
Token the
Feature activation+0.526
infamous
Token infamous
Feature activation+0.655
Bit
Token Bit
Feature activation+0.000
Torrent
TokenTorrent
Feature activation+0.141
site
Token site
Feature activation+0.974
certainly
Token certainly
Feature activation+0.880
gained
Token gained
Feature activation+1.335
some
Token some
Feature activation+1.026
new
Token new
Feature activation+0.749
visitors
Token visitors
Feature activation+1.366
in
Token in
Feature activation+1.111
the
Token the
Feature activation+0.560
U
Token U
Feature activation+0.000
.
Token.
Feature activation+0.039
S
TokenS
Feature activation+0.067
.
Token.
Feature activation+0.416
industry
Token industry
Feature activation+0.794
gu
Token gu
Feature activation+0.000
es
Tokenes
Feature activation+0.000
st
Tokenst
Feature activation+0.000
imates
Tokenimates
Feature activation+0.626
costs
Token costs
Feature activation+1.056

INTERVAL 0.339 - 0.678
CONTAINS 0.163%

athan
Tokenathan
Feature activation+0.000
said
Token said
Feature activation+0.000
that
Token that
Feature activation+0.000
while
Token while
Feature activation+0.000
admission
Token admission
Feature activation+0.000
process
Token process
Feature activation+0.406
and
Token and
Feature activation+0.117
fee
Token fee
Feature activation+0.000
structure
Token structure
Feature activation+0.632
were
Token were
Feature activation+1.003
as
Token as
Feature activation+1.236
me
Token me
Feature activation+0.000
wrong
Token wrong
Feature activation+0.000
,
Token,
Feature activation+0.350
I
Token I
Feature activation+0.388
'm
Token'm
Feature activation+0.150
not
Token not
Feature activation+0.339
comparing
Token comparing
Feature activation+0.233
him
Token him
Feature activation+0.000
to
Token to
Feature activation+0.000
these
Token these
Feature activation+0.000
two
Token two
Feature activation+0.000
be
Token be
Feature activation+0.506
seen
Token seen
Feature activation+1.065
on
Token on
Feature activation+0.751
the
Token the
Feature activation+0.667
paper
Token paper
Feature activation+0.416
held
Token held
Feature activation+0.539
in
Token in
Feature activation+0.000
Mar
Token Mar
Feature activation+0.351
at
Tokenat
Feature activation+0.304
's
Token's
Feature activation+0.000
left
Token left
Feature activation+0.219
of
Token of
Feature activation+0.000
2004
Token 2004
Feature activation+0.000
.
Token.
Feature activation+0.000
Despite
Token Despite
Feature activation+0.000
this
Token this
Feature activation+0.000
decrease
Token decrease
Feature activation+0.672
,
Token,
Feature activation+0.000
there
Token there
Feature activation+0.000
are
Token are
Feature activation+0.000
a
Token a
Feature activation+0.000
number
Token number
Feature activation+0.000
has
Token has
Feature activation+0.426
never
Token never
Feature activation+0.275
held
Token held
Feature activation+0.000
elected
Token elected
Feature activation+0.000
office
Token office
Feature activation+0.000
,
Token,
Feature activation+0.588
opinion
Token opinion
Feature activation+0.000
polls
Token polls
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
run
Token run
Feature activation+0.000

INTERVAL 0.000 - 0.339
CONTAINS 99.666%

a
Token a
Feature activation+0.000
final
Token final
Feature activation+0.000
opportunity
Token opportunity
Feature activation+0.000
to
Token to
Feature activation+0.000
another
Token another
Feature activation+0.000
app
Token app
Feature activation+0.000
-
Token-
Feature activation+0.000
based
Tokenbased
Feature activation+0.000
cab
Token cab
Feature activation+0.000
service
Token service
Feature activation+0.000
provider
Token provider
Feature activation+0.000
es
Tokenes
Feature activation+0.000
Ã
Token Ã
Feature activation+0.000
®
Token®
Feature activation+0.000
nc
Tokennc
Feature activation+0.000
Ä
TokenÄ
Feature activation+0.000
ĥ
Tokenĥ
Feature activation+0.000
c
Token c
Feature activation+0.000
ine
Tokenine
Feature activation+0.000
s
Token s
Feature activation+0.000
Ä
TokenÄ
Feature activation+0.000
ĥ
Tokenĥ
Feature activation+0.000
let
Token let
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
take
Token take
Feature activation+0.000
the
Token the
Feature activation+0.000
IP
Token IP
Feature activation+0.000
address
Token address
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
frontal
Token frontal
Feature activation+0.000
supported
Token supported
Feature activation+0.000
the
Token the
Feature activation+0.000
resolutions
Token resolutions
Feature activation+0.000
,
Token,
Feature activation+0.000
including
Token including
Feature activation+0.000
the
Token the
Feature activation+0.000
lowest
Token lowest
Feature activation+0.000
priority
Token priority
Feature activation+0.000
language
Token language
Feature activation+0.000
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
do
Token do
Feature activation+0.000
think
Token think
Feature activation+0.000
that
Token that
Feature activation+0.000
the
Token the
Feature activation+0.000
conversation
Token conversation
Feature activation+0.000
has
Token has
Feature activation+0.000
been
Token been
Feature activation+0.000
overs
Token overs
Feature activation+0.000
impl
Tokenimpl
Feature activation+0.000
ified
Tokenified
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 H5.3: (feature 15631)

TOP ACTIVATIONS
MAX = 3.880

of
Token of
Feature activation+0.759
rising
Token rising
Feature activation+0.923
numbers
Token numbers
Feature activation+0.733
real
Token real
Feature activation+1.840
,
Token,
Feature activation+2.392
or
Token or
Feature activation+3.880
are
Token are
Feature activation+2.169
humans
Token humans
Feature activation+0.492
just
Token just
Feature activation+2.288
spending
Token spending
Feature activation+1.408
more
Token more
Feature activation+0.645
play
Token play
Feature activation+0.319
shutdown
Token shutdown
Feature activation+0.280
soccer
Token soccer
Feature activation+0.490
away
Token away
Feature activation+0.285
,
Token,
Feature activation+1.708
or
Token or
Feature activation+3.822
aim
Token aim
Feature activation+1.634
to
Token to
Feature activation+1.120
secure
Token secure
Feature activation+0.953
an
Token an
Feature activation+0.061
advantage
Token advantage
Feature activation+0.142
organizers
Token organizers
Feature activation+0.000
to
Token to
Feature activation+0.225
explore
Token explore
Feature activation+0.122
their
Token their
Feature activation+0.254
work
Token work
Feature activation+0.133
than
Token than
Feature activation+3.819
through
Token through
Feature activation+0.945
writing
Token writing
Feature activation+0.358
science
Token science
Feature activation+0.031
fiction
Token fiction
Feature activation+0.000
stories
Token stories
Feature activation+0.062
focus
Token focus
Feature activation+1.055
on
Token on
Feature activation+1.937
Ukraine
Token Ukraine
Feature activation+0.856
again
Token again
Feature activation+1.948
,
Token,
Feature activation+3.040
or
Token or
Feature activation+3.695
will
Token will
Feature activation+3.248
he
Token he
Feature activation+1.959
get
Token get
Feature activation+1.454
involved
Token involved
Feature activation+1.152
somewhere
Token somewhere
Feature activation+1.783
an
Token an
Feature activation+1.260
explicit
Token explicit
Feature activation+1.287
for
Token for
Feature activation+0.497
-
Token-
Feature activation+0.000
loop
Tokenloop
Feature activation+1.278
or
Token or
Feature activation+3.537
do
Token do
Feature activation+2.574
you
Token you
Feature activation+0.423
instead
Token instead
Feature activation+2.518
use
Token use
Feature activation+1.652
a
Token a
Feature activation+2.556
school
Token school
Feature activation+0.000
no
Token no
Feature activation+0.000
less
Token less
Feature activation+0.000
--
Token --
Feature activation+1.564
be
Token be
Feature activation+1.317
unable
Token unable
Feature activation+3.406
to
Token to
Feature activation+2.230
read
Token read
Feature activation+0.245
her
Token her
Feature activation+0.392
own
Token own
Feature activation+0.000
words
Token words
Feature activation+0.346
Sandy
Token Sandy
Feature activation+0.096
just
Token just
Feature activation+1.930
a
Token a
Feature activation+1.677
hurricane
Token hurricane
Feature activation+0.615
,
Token,
Feature activation+1.833
or
Token or
Feature activation+3.371
an
Token an
Feature activation+1.380
ang
Token ang
Feature activation+0.000
rier
Tokenrier
Feature activation+0.147
child
Token child
Feature activation+0.000
spawned
Token spawned
Feature activation+0.252
Who
TokenWho
Feature activation+0.000
says
Token says
Feature activation+0.000
this
Token this
Feature activation+0.397
administration
Token administration
Feature activation+0.666
doesn
Token doesn
Feature activation+2.698
't
Token't
Feature activation+3.292
get
Token get
Feature activation+1.594
things
Token things
Feature activation+0.313
done
Token done
Feature activation+0.897
?
Token?
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
coron
Token coron
Feature activation+0.351
ating
Tokenating
Feature activation+0.496
Clinton
Token Clinton
Feature activation+0.896
already
Token already
Feature activation+1.590
,
Token,
Feature activation+1.742
when
Token when
Feature activation+3.260
national
Token national
Feature activation+0.223
polls
Token polls
Feature activation+0.662
have
Token have
Feature activation+1.147
Bernie
Token Bernie
Feature activation+0.000
neck
Token neck
Feature activation+0.022
on
Token on
Feature activation+1.937
Ukraine
Token Ukraine
Feature activation+0.856
again
Token again
Feature activation+1.948
,
Token,
Feature activation+3.040
or
Token or
Feature activation+3.695
will
Token will
Feature activation+3.248
he
Token he
Feature activation+1.959
get
Token get
Feature activation+1.454
involved
Token involved
Feature activation+1.152
somewhere
Token somewhere
Feature activation+1.783
else
Token else
Feature activation+1.308
a
Token a
Feature activation+0.782
fiery
Token fiery
Feature activation+0.419
mess
Token mess
Feature activation+0.526
of
Token of
Feature activation+0.000
bone
Token bone
Feature activation+0.000
,
Token,
Feature activation+3.245
cart
Token cart
Feature activation+0.508
ilage
Tokenilage
Feature activation+0.380
,
Token,
Feature activation+0.351
and
Token and
Feature activation+0.087
bl
Token bl
Feature activation+0.000
circulation
Token circulation
Feature activation+1.900
in
Token in
Feature activation+1.787
the
Token the
Feature activation+0.871
coronary
Token coronary
Feature activation+0.211
arteries
Token arteries
Feature activation+0.943
that
Token that
Feature activation+3.219
feed
Token feed
Feature activation+0.211
our
Token our
Feature activation+0.451
heart
Token heart
Feature activation+0.131
?
Token?
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
a
Token a
Feature activation+0.474
Justice
Token Justice
Feature activation+0.610
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Peace
Token Peace
Feature activation+1.087
or
Token or
Feature activation+3.188
a
Token a
Feature activation+2.000
traffic
Token traffic
Feature activation+0.000
officer
Token officer
Feature activation+0.657
?
Token?
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
is
Token is
Feature activation+0.000
it
Token it
Feature activation+0.260
something
Token something
Feature activation+2.671
to
Token to
Feature activation+2.457
do
Token do
Feature activation+2.724
with
Token with
Feature activation+3.158
women
Token women
Feature activation+1.430
and
Token and
Feature activation+2.511
single
Token single
Feature activation+0.159
-
Token-
Feature activation+0.000
sex
Tokensex
Feature activation+0.323
numbers
Token numbers
Feature activation+1.396
of
Token of
Feature activation+0.160
people
Token people
Feature activation+0.553
being
Token being
Feature activation+0.911
evacuated
Token evacuated
Feature activation+0.577
or
Token or
Feature activation+3.134
isolated
Token isolated
Feature activation+0.879
?
Token?
Feature activation+0.000
Is
Token Is
Feature activation+0.000
a
Token a
Feature activation+0.000
plane
Token plane
Feature activation+0.000
a
Token a
Feature activation+1.418
financial
Token financial
Feature activation+0.173
point
Token point
Feature activation+2.196
of
Token of
Feature activation+0.401
view
Token view
Feature activation+2.796
,
Token,
Feature activation+3.091
also
Token also
Feature activation+2.149
taking
Token taking
Feature activation+0.929
into
Token into
Feature activation+0.200
account
Token account
Feature activation+1.939
if
Token if
Feature activation+1.236
lead
Token lead
Feature activation+0.252
character
Token character
Feature activation+1.136
,
Token,
Feature activation+1.402
to
Token to
Feature activation+0.947
someone
Token someone
Feature activation+2.267
who
Token who
Feature activation+3.068
hasn
Token hasn
Feature activation+1.196
âĢ
TokenâĢ
Feature activation+0.395
Ļ
TokenĻ
Feature activation+0.111
t
Tokent
Feature activation+0.000
encountered
Token encountered
Feature activation+0.077
your
Token your
Feature activation+0.341
doctor
Token doctor
Feature activation+0.312
,
Token,
Feature activation+1.308
though
Token though
Feature activation+1.838
,
Token,
Feature activation+1.758
if
Token if
Feature activation+3.055
your
Token your
Feature activation+0.798
doctor
Token doctor
Feature activation+0.000
cannot
Token cannot
Feature activation+0.216
keep
Token keep
Feature activation+0.000
his
Token his
Feature activation+0.000
he
Token he
Feature activation+0.422
focus
Token focus
Feature activation+1.055
on
Token on
Feature activation+1.937
Ukraine
Token Ukraine
Feature activation+0.856
again
Token again
Feature activation+1.948
,
Token,
Feature activation+3.040
or
Token or
Feature activation+3.695
will
Token will
Feature activation+3.248
he
Token he
Feature activation+1.959
get
Token get
Feature activation+1.454
involved
Token involved
Feature activation+1.152
how
Token how
Feature activation+0.000
does
Token does
Feature activation+0.254
the
Token the
Feature activation+0.573
implementation
Token implementation
Feature activation+0.965
look
Token look
Feature activation+1.561
when
Token when
Feature activation+3.024
there
Token there
Feature activation+1.392
are
Token are
Feature activation+1.464
zero
Token zero
Feature activation+0.498
results
Token results
Feature activation+0.337
?
Token?
Feature activation+0.000

Top DFA by src position
MAX = 4.416

sh
Token sh
Feature activation+0.000
Top resid features:
rim
Tokenrim
Feature activation-0.000
Top resid features:
pers
Tokenpers
Feature activation-0.000
Top resid features:
.
Token.
Feature activation-0.006
Top resid features:
But
Token But
Feature activation-0.003
Top resid features:
are
Token are
Feature activation+4.120
Top resid features:
these
Token these
Feature activation+0.165
Top resid features:
accounts
Token accounts
Feature activation+0.028
Top resid features:
of
Token of
Feature activation+0.007
Top resid features:
rising
Token rising
Feature activation+0.011
Top resid features:
numbers
Token numbers
Feature activation+0.004
Top resid features:
at
Token at
Feature activation+0.000
Top resid features:
B
Token B
Feature activation+0.003
Top resid features:
MO
TokenMO
Feature activation-0.000
Top resid features:
Field
Token Field
Feature activation-0.001
Top resid features:
,
Token,
Feature activation+0.005
Top resid features:
will
Token will
Feature activation+4.210
Top resid features:
T
Token T
Feature activation-0.002
Top resid features:
FC
TokenFC
Feature activation-0.000
Top resid features:
look
Token look
Feature activation-0.006
Top resid features:
to
Token to
Feature activation-0.000
Top resid features:
play
Token play
Feature activation+0.006
Top resid features:
other
Token other
Feature activation+0.002
Top resid features:
worlds
Token worlds
Feature activation+0.007
Top resid features:
,
Token,
Feature activation+0.003
Top resid features:
so
Token so
Feature activation+0.012
Top resid features:
what
Token what
Feature activation+0.002
Top resid features:
better
Token better
Feature activation+4.416
Top resid features:
venue
Token venue
Feature activation-0.031
Top resid features:
for
Token for
Feature activation-0.126
Top resid features:
organizers
Token organizers
Feature activation-0.004
Top resid features:
to
Token to
Feature activation-0.001
Top resid features:
explore
Token explore
Feature activation+0.009
Top resid features:
next
Token next
Feature activation+0.003
Top resid features:
move
Token move
Feature activation+0.007
Top resid features:
might
Token might
Feature activation-0.004
Top resid features:
be
Token be
Feature activation-0.004
Top resid features:
.
Token.
Feature activation-0.009
Top resid features:
Will
Token Will
Feature activation+3.644
Top resid features:
he
Token he
Feature activation+0.639
Top resid features:
focus
Token focus
Feature activation-0.004
Top resid features:
on
Token on
Feature activation-0.013
Top resid features:
Ukraine
Token Ukraine
Feature activation+0.007
Top resid features:
again
Token again
Feature activation-0.016
Top resid features:
compute
Token compute
Feature activation+0.004
Top resid features:
some
Token some
Feature activation+0.004
Top resid features:
result
Token result
Feature activation+0.007
Top resid features:
âĢĶ
TokenâĢĶ
Feature activation+0.032
Top resid features:
do
Tokendo
Feature activation+0.282
Top resid features:
you
Token you
Feature activation+3.514
Top resid features:
write
Token write
Feature activation+0.111
Top resid features:
an
Token an
Feature activation-0.010
Top resid features:
explicit
Token explicit
Feature activation-0.000
Top resid features:
for
Token for
Feature activation+0.086
Top resid features:
-
Token-
Feature activation+0.018
Top resid features:
deeper
Token deeper
Feature activation+0.019
Top resid features:
.
Token.
Feature activation+0.005
Top resid features:
Ċ
TokenĊ
Feature activation-0.000
Top resid features:
Ċ
TokenĊ
Feature activation-0.000
Top resid features:
How
TokenHow
Feature activation+0.138
Top resid features:
could
Token could
Feature activation+3.681
Top resid features:
a
Token a
Feature activation-0.021
Top resid features:
19
Token 19
Feature activation-0.002
Top resid features:
-
Token-
Feature activation+0.000
Top resid features:
year
Tokenyear
Feature activation+0.000
Top resid features:
-
Token-
Feature activation+0.000
Top resid features:
years
Token years
Feature activation+0.001
Top resid features:
."
Token."
Feature activation-0.001
Top resid features:
Ċ
TokenĊ
Feature activation-0.006
Top resid features:
Ċ
TokenĊ
Feature activation-0.006
Top resid features:
But
TokenBut
Feature activation-0.002
Top resid features:
was
Token was
Feature activation+4.019
Top resid features:
Sandy
Token Sandy
Feature activation-0.014
Top resid features:
just
Token just
Feature activation-0.041
Top resid features:
a
Token a
Feature activation-0.036
Top resid features:
hurricane
Token hurricane
Feature activation-0.063
Top resid features:
,
Token,
Feature activation+0.022
Top resid features:
Europe
Token Europe
Feature activation+0.006
Top resid features:
."
Token."
Feature activation-0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.001
Top resid features:
Ċ
TokenĊ
Feature activation+0.002
Top resid features:
Who
TokenWho
Feature activation+0.151
Top resid features:
says
Token says
Feature activation+3.708
Top resid features:
this
Token this
Feature activation-0.081
Top resid features:
administration
Token administration
Feature activation+0.022
Top resid features:
doesn
Token doesn
Feature activation-0.018
Top resid features:
't
Token't
Feature activation-0.050
Top resid features:
get
Token get
Feature activation+0.000
Top resid features:
are
Token are
Feature activation-0.002
Top resid features:
naturally
Token naturally
Feature activation-0.004
Top resid features:
asking
Token asking
Feature activation+0.005
Top resid features:
:
Token:
Feature activation+0.042
Top resid features:
How
Token How
Feature activation+0.474
Top resid features:
can
Token can
Feature activation+2.881
Top resid features:
they
Token they
Feature activation+0.238
Top resid features:
be
Token be
Feature activation+0.024
Top resid features:
coron
Token coron
Feature activation+0.007
Top resid features:
ating
Tokenating
Feature activation-0.000
Top resid features:
Clinton
Token Clinton
Feature activation+0.030
Top resid features:
next
Token next
Feature activation+0.002
Top resid features:
move
Token move
Feature activation+0.006
Top resid features:
might
Token might
Feature activation-0.000
Top resid features:
be
Token be
Feature activation-0.002
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
Will
Token Will
Feature activation+3.509
Top resid features:
he
Token he
Feature activation+0.067
Top resid features:
focus
Token focus
Feature activation+0.027
Top resid features:
on
Token on
Feature activation-0.010
Top resid features:
Ukraine
Token Ukraine
Feature activation+0.004
Top resid features:
again
Token again
Feature activation+0.005
Top resid features:
end
Token end
Feature activation-0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
it
Token it
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.005
Top resid features:
When
Token When
Feature activation-0.071
Top resid features:
is
Token is
Feature activation+3.238
Top resid features:
he
Token he
Feature activation+0.320
Top resid features:
supposed
Token supposed
Feature activation+0.177
Top resid features:
to
Token to
Feature activation+0.002
Top resid features:
turn
Token turn
Feature activation+0.000
Top resid features:
into
Token into
Feature activation-0.012
Top resid features:
is
Token is
Feature activation+0.000
Top resid features:
great
Token great
Feature activation-0.011
Top resid features:
,
Token,
Feature activation-0.000
Top resid features:
but
Token but
Feature activation+0.000
Top resid features:
what
Token what
Feature activation+0.106
Top resid features:
about
Token about
Feature activation+3.595
Top resid features:
circulation
Token circulation
Feature activation+0.006
Top resid features:
in
Token in
Feature activation+0.030
Top resid features:
the
Token the
Feature activation-0.004
Top resid features:
coronary
Token coronary
Feature activation+0.000
Top resid features:
arteries
Token arteries
Feature activation+0.003
Top resid features:
TV
Token TV
Feature activation-0.000
Top resid features:
asked
Token asked
Feature activation-0.004
Top resid features:
the
Token the
Feature activation-0.000
Top resid features:
judge
Token judge
Feature activation-0.003
Top resid features:
,
Token,
Feature activation+0.033
Top resid features:
was
Token was
Feature activation+2.974
Top resid features:
he
Token he
Feature activation+0.653
Top resid features:
elected
Token elected
Feature activation-0.023
Top resid features:
to
Token to
Feature activation-0.003
Top resid features:
be
Token be
Feature activation-0.001
Top resid features:
a
Token a
Feature activation-0.011
Top resid features:
sharing
Tokensharing
Feature activation-0.006
Top resid features:
in
Token in
Feature activation+0.004
Top resid features:
books
Token books
Feature activation+0.009
Top resid features:
âĢĵ
Token âĢĵ
Feature activation+0.020
Top resid features:
is
Token is
Feature activation+0.596
Top resid features:
it
Token it
Feature activation+3.003
Top resid features:
something
Token something
Feature activation-0.001
Top resid features:
to
Token to
Feature activation-0.013
Top resid features:
do
Token do
Feature activation+0.038
Top resid features:
with
Token with
Feature activation+0.103
Top resid features:
women
Token women
Feature activation+0.000
Top resid features:
serious
Token serious
Feature activation+0.000
Top resid features:
disruption
Token disruption
Feature activation+0.001
Top resid features:
or
Token or
Feature activation-0.001
Top resid features:
reaction
Token reaction
Feature activation+0.001
Top resid features:
.
Token.
Feature activation-0.005
Top resid features:
Are
Token Are
Feature activation+3.677
Top resid features:
buildings
Token buildings
Feature activation+0.001
Top resid features:
being
Token being
Feature activation-0.081
Top resid features:
closed
Token closed
Feature activation-0.012
Top resid features:
and
Token and
Feature activation+0.000
Top resid features:
substantial
Token substantial
Feature activation+0.008
Top resid features:
P
Token P
Feature activation+0.002
Top resid features:
ire
Tokenire
Feature activation-0.000
Top resid features:
lli
Tokenlli
Feature activation+0.001
Top resid features:
.
Token.
Feature activation+0.001
Top resid features:
Can
Token Can
Feature activation+0.341
Top resid features:
you
Token you
Feature activation+2.468
Top resid features:
explain
Token explain
Feature activation+0.575
Top resid features:
the
Token the
Feature activation+0.034
Top resid features:
plans
Token plans
Feature activation+0.016
Top resid features:
to
Token to
Feature activation+0.009
Top resid features:
make
Token make
Feature activation+0.004
Top resid features:
entirely
Token entirely
Feature activation+0.005
Top resid features:
.
Token.
Feature activation+0.008
Top resid features:
Ċ
TokenĊ
Feature activation-0.000
Top resid features:
Ċ
TokenĊ
Feature activation-0.000
Top resid features:
How
TokenHow
Feature activation+0.075
Top resid features:
would
Token would
Feature activation+1.939
Top resid features:
you
Token you
Feature activation+0.866
Top resid features:
describe
Token describe
Feature activation+0.128
Top resid features:
Furious
Token Furious
Feature activation-0.003
Top resid features:
,
Token,
Feature activation+0.001
Top resid features:
be
Token be
Feature activation+0.077
Top resid features:
June
Token June
Feature activation+0.000
Top resid features:
2010
Token 2010
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.002
Top resid features:
How
TokenHow
Feature activation+0.090
Top resid features:
can
Token can
Feature activation+3.210
Top resid features:
you
Token you
Feature activation+0.220
Top resid features:
keep
Token keep
Feature activation+0.010
Top resid features:
your
Token your
Feature activation+0.002
Top resid features:
doctor
Token doctor
Feature activation+0.032
Top resid features:
,
Token,
Feature activation+0.046
Top resid features:
next
Token next
Feature activation+0.004
Top resid features:
move
Token move
Feature activation+0.007
Top resid features:
might
Token might
Feature activation-0.003
Top resid features:
be
Token be
Feature activation-0.003
Top resid features:
.
Token.
Feature activation-0.017
Top resid features:
Will
Token Will
Feature activation+2.038
Top resid features:
he
Token he
Feature activation+1.373
Top resid features:
focus
Token focus
Feature activation+0.103
Top resid features:
on
Token on
Feature activation-0.000
Top resid features:
Ukraine
Token Ukraine
Feature activation+0.009
Top resid features:
again
Token again
Feature activation-0.036
Top resid features:
items
Token items
Feature activation+0.001
Top resid features:
.
Token.
Feature activation+0.001
Top resid features:
But
Token But
Feature activation+0.033
Top resid features:
,
Token,
Feature activation+0.010
Top resid features:
how
Token how
Feature activation+0.578
Top resid features:
does
Token does
Feature activation+2.745
Top resid features:
the
Token the
Feature activation+0.105
Top resid features:
implementation
Token implementation
Feature activation+0.027
Top resid features:
look
Token look
Feature activation+0.055
Top resid features:
when
Token when
Feature activation+0.037
Top resid features:
there
Token there
Feature activation+0.000
Top resid features:

Decoder Weights Distribution

Head 0: 0.02

Head 1: 0.02

Head 2: 0.10

Head 3: 0.49

Head 4: 0.06

Head 5: 0.04

Head 6: 0.02

Head 7: 0.03

Head 8: 0.04

Head 9: 0.05

Head 10: 0.06

Head 11: 0.05

Positive logits

?4.03

?:3.96

?)3.91

)?3.89

?),3.87

?).3.85

?'3.69

?"3.51

?",3.39

?'"3.38

?".3.30

'?3.30

?3.30

.?3.29

?]3.28

??3.20

?,3.17

?!3.07

?????3.04

!?3.03

Negative logits

iHUD-1.80

Pwr-1.77

}}}-1.58

emale-1.53

igi-1.52

wcsstore-1.47

trak-1.46

anmar-1.45

camp-1.44

Ocean-1.40

iverpool-1.39

coord-1.38

APD-1.37

tip-1.36

research-1.36

},"-1.36

drivers-1.35

CTV-1.35

DERR-1.33

ombat-1.33

INTERVAL 3.492 - 3.880
CONTAINS 0.001%

focus
Token focus
Feature activation+1.055
on
Token on
Feature activation+1.937
Ukraine
Token Ukraine
Feature activation+0.856
again
Token again
Feature activation+1.948
,
Token,
Feature activation+3.040
or
Token or
Feature activation+3.695
will
Token will
Feature activation+3.248
he
Token he
Feature activation+1.959
get
Token get
Feature activation+1.454
involved
Token involved
Feature activation+1.152
somewhere
Token somewhere
Feature activation+1.783
of
Token of
Feature activation+0.759
rising
Token rising
Feature activation+0.923
numbers
Token numbers
Feature activation+0.733
real
Token real
Feature activation+1.840
,
Token,
Feature activation+2.392
or
Token or
Feature activation+3.880
are
Token are
Feature activation+2.169
humans
Token humans
Feature activation+0.492
just
Token just
Feature activation+2.288
spending
Token spending
Feature activation+1.408
more
Token more
Feature activation+0.645
an
Token an
Feature activation+1.260
explicit
Token explicit
Feature activation+1.287
for
Token for
Feature activation+0.497
-
Token-
Feature activation+0.000
loop
Tokenloop
Feature activation+1.278
or
Token or
Feature activation+3.537
do
Token do
Feature activation+2.574
you
Token you
Feature activation+0.423
instead
Token instead
Feature activation+2.518
use
Token use
Feature activation+1.652
a
Token a
Feature activation+2.556
organizers
Token organizers
Feature activation+0.000
to
Token to
Feature activation+0.225
explore
Token explore
Feature activation+0.122
their
Token their
Feature activation+0.254
work
Token work
Feature activation+0.133
than
Token than
Feature activation+3.819
through
Token through
Feature activation+0.945
writing
Token writing
Feature activation+0.358
science
Token science
Feature activation+0.031
fiction
Token fiction
Feature activation+0.000
stories
Token stories
Feature activation+0.062
play
Token play
Feature activation+0.319
shutdown
Token shutdown
Feature activation+0.280
soccer
Token soccer
Feature activation+0.490
away
Token away
Feature activation+0.285
,
Token,
Feature activation+1.708
or
Token or
Feature activation+3.822
aim
Token aim
Feature activation+1.634
to
Token to
Feature activation+1.120
secure
Token secure
Feature activation+0.953
an
Token an
Feature activation+0.061
advantage
Token advantage
Feature activation+0.142

INTERVAL 3.104 - 3.492
CONTAINS 0.001%

Who
TokenWho
Feature activation+0.000
says
Token says
Feature activation+0.000
this
Token this
Feature activation+0.397
administration
Token administration
Feature activation+0.666
doesn
Token doesn
Feature activation+2.698
't
Token't
Feature activation+3.292
get
Token get
Feature activation+1.594
things
Token things
Feature activation+0.313
done
Token done
Feature activation+0.897
?
Token?
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
numbers
Token numbers
Feature activation+1.396
of
Token of
Feature activation+0.160
people
Token people
Feature activation+0.553
being
Token being
Feature activation+0.911
evacuated
Token evacuated
Feature activation+0.577
or
Token or
Feature activation+3.134
isolated
Token isolated
Feature activation+0.879
?
Token?
Feature activation+0.000
Is
Token Is
Feature activation+0.000
a
Token a
Feature activation+0.000
plane
Token plane
Feature activation+0.000
school
Token school
Feature activation+0.000
no
Token no
Feature activation+0.000
less
Token less
Feature activation+0.000
--
Token --
Feature activation+1.564
be
Token be
Feature activation+1.317
unable
Token unable
Feature activation+3.406
to
Token to
Feature activation+2.230
read
Token read
Feature activation+0.245
her
Token her
Feature activation+0.392
own
Token own
Feature activation+0.000
words
Token words
Feature activation+0.346
is
Token is
Feature activation+0.000
it
Token it
Feature activation+0.260
something
Token something
Feature activation+2.671
to
Token to
Feature activation+2.457
do
Token do
Feature activation+2.724
with
Token with
Feature activation+3.158
women
Token women
Feature activation+1.430
and
Token and
Feature activation+2.511
single
Token single
Feature activation+0.159
-
Token-
Feature activation+0.000
sex
Tokensex
Feature activation+0.323
Sandy
Token Sandy
Feature activation+0.096
just
Token just
Feature activation+1.930
a
Token a
Feature activation+1.677
hurricane
Token hurricane
Feature activation+0.615
,
Token,
Feature activation+1.833
or
Token or
Feature activation+3.371
an
Token an
Feature activation+1.380
ang
Token ang
Feature activation+0.000
rier
Tokenrier
Feature activation+0.147
child
Token child
Feature activation+0.000
spawned
Token spawned
Feature activation+0.252

INTERVAL 2.716 - 3.104
CONTAINS 0.003%

it
Token it
Feature activation+1.047
the
Token the
Feature activation+2.908
right
Token right
Feature activation+2.898
kind
Token kind
Feature activation+2.065
of
Token of
Feature activation+1.675
action
Token action
Feature activation+2.799
?
Token?
Feature activation+0.184
Extrem
Token Extrem
Feature activation+0.000
ism
Tokenism
Feature activation+0.072
researchers
Token researchers
Feature activation+0.000
said
Token said
Feature activation+0.000
arer
Tokenarer
Feature activation+0.550
inherently
Token inherently
Feature activation+0.538
has
Token has
Feature activation+0.820
,
Token,
Feature activation+2.724
which
Token which
Feature activation+2.929
you
Token you
Feature activation+2.752
support
Token support
Feature activation+0.405
?
Token?
Feature activation+0.305
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
We
TokenWe
Feature activation+0.000
.
Token.
Feature activation+0.000
When
Token When
Feature activation+0.000
is
Token is
Feature activation+0.000
he
Token he
Feature activation+0.581
supposed
Token supposed
Feature activation+1.175
to
Token to
Feature activation+2.862
turn
Token turn
Feature activation+0.984
into
Token into
Feature activation+1.011
a
Token a
Feature activation+0.782
fiery
Token fiery
Feature activation+0.419
mess
Token mess
Feature activation+0.526
World
Token World
Feature activation+0.000
Series
Token Series
Feature activation+0.000
,
Token,
Feature activation+0.000
are
Token are
Feature activation+0.000
you
Token you
Feature activation+0.421
really
Token really
Feature activation+2.858
going
Token going
Feature activation+2.143
to
Token to
Feature activation+2.914
push
Token push
Feature activation+1.356
your
Token your
Feature activation+1.552
limits
Token limits
Feature activation+1.372
related
Token related
Feature activation+0.000
to
Token to
Feature activation+0.000
.
Token.
Feature activation+0.000
Is
Token Is
Feature activation+0.000
it
Token it
Feature activation+0.381
acceptable
Token acceptable
Feature activation+2.897
to
Token to
Feature activation+2.703
whip
Token whip
Feature activation+0.786
rape
Token rape
Feature activation+0.522
victims
Token victims
Feature activation+0.104
if
Token if
Feature activation+2.716

INTERVAL 2.328 - 2.716
CONTAINS 0.006%

acceptable
Token acceptable
Feature activation+2.091
to
Token to
Feature activation+2.245
kill
Token kill
Feature activation+1.291
a
Token a
Feature activation+1.125
doctor
Token doctor
Feature activation+0.967
if
Token if
Feature activation+2.487
you
Token you
Feature activation+2.313
believe
Token believe
Feature activation+2.031
God
Token God
Feature activation+1.158
wants
Token wants
Feature activation+0.750
you
Token you
Feature activation+0.543
sc
Token sc
Feature activation+0.878
ari
Tokenari
Feature activation+0.577
est
Tokenest
Feature activation+1.514
hive
Token hive
Feature activation+0.488
mind
Tokenmind
Feature activation+0.593
like
Token like
Feature activation+2.706
in
Token in
Feature activation+2.137
real
Token real
Feature activation+0.960
life
Token life
Feature activation+1.313
?
Token?
Feature activation+0.000
A
Token A
Feature activation+0.000
a
Tokena
Feature activation+0.000
need
Token need
Feature activation+1.668
US
Token US
Feature activation+0.060
taxpayer
Token taxpayer
Feature activation+0.002
help
Token help
Feature activation+0.593
when
Token when
Feature activation+2.462
his
Token his
Feature activation+1.301
firm
Token firm
Feature activation+0.583
(
Token (
Feature activation+0.477
and
Tokenand
Feature activation+0.000
his
Token his
Feature activation+0.000
specific
Token specific
Feature activation+0.821
episodes
Token episodes
Feature activation+0.306
of
Token of
Feature activation+0.117
care
Token care
Feature activation+0.006
,
Token,
Feature activation+1.274
or
Token or
Feature activation+2.379
toward
Token toward
Feature activation+0.863
accountable
Token accountable
Feature activation+0.277
-
Token-
Feature activation+0.000
care
Tokencare
Feature activation+0.000
organizations
Token organizations
Feature activation+0.185
books
Token books
Feature activation+0.000
âĢĵ
Token âĢĵ
Feature activation+0.000
is
Token is
Feature activation+0.000
it
Token it
Feature activation+0.260
something
Token something
Feature activation+2.671
to
Token to
Feature activation+2.457
do
Token do
Feature activation+2.724
with
Token with
Feature activation+3.158
women
Token women
Feature activation+1.430
and
Token and
Feature activation+2.511
single
Token single
Feature activation+0.159

INTERVAL 1.940 - 2.328
CONTAINS 0.014%

But
Token But
Feature activation+0.000
is
Token is
Feature activation+0.000
it
Token it
Feature activation+0.000
primarily
Token primarily
Feature activation+1.509
caused
Token caused
Feature activation+1.295
by
Token by
Feature activation+2.008
humanity
Token humanity
Feature activation+1.245
,
Token,
Feature activation+2.141
or
Token or
Feature activation+2.953
are
Token are
Feature activation+1.970
other
Token other
Feature activation+0.737
Ukraine
Token Ukraine
Feature activation+0.856
again
Token again
Feature activation+1.948
,
Token,
Feature activation+3.040
or
Token or
Feature activation+3.695
will
Token will
Feature activation+3.248
he
Token he
Feature activation+1.959
get
Token get
Feature activation+1.454
involved
Token involved
Feature activation+1.152
somewhere
Token somewhere
Feature activation+1.783
else
Token else
Feature activation+1.308
in
Token in
Feature activation+0.648
ins
Token ins
Feature activation+0.283
ip
Tokenip
Feature activation+0.000
id
Tokenid
Feature activation+0.789
bl
Token bl
Feature activation+0.083
uster
Tokenuster
Feature activation+0.817
,
Token,
Feature activation+2.204
the
Token the
Feature activation+1.928
ED
Token ED
Feature activation+0.000
M
TokenM
Feature activation+0.000
equivalent
Token equivalent
Feature activation+0.443
of
Token of
Feature activation+0.279
Ċ
TokenĊ
Feature activation+0.000
How
TokenHow
Feature activation+0.000
is
Token is
Feature activation+0.000
this
Token this
Feature activation+0.000
different
Token different
Feature activation+0.754
than
Token than
Feature activation+2.318
before
Token before
Feature activation+0.324
?
Token?
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
People
TokenPeople
Feature activation+0.000
"
Token "
Feature activation+0.000
What
TokenWhat
Feature activation+0.000
is
Token is
Feature activation+0.000
this
Token this
Feature activation+0.029
,
Token,
Feature activation+1.499
a
Token a
Feature activation+1.990
first
Token first
Feature activation+0.977
-
Token-
Feature activation+0.288
person
Tokenperson
Feature activation+0.477
shooter
Token shooter
Feature activation+0.832
?"
Token?"
Feature activation+0.000

INTERVAL 1.552 - 1.940
CONTAINS 0.028%

need
Token need
Feature activation+1.052
to
Token to
Feature activation+1.008
defend
Token defend
Feature activation+1.097
themselves
Token themselves
Feature activation+0.398
against
Token against
Feature activation+0.945
such
Token such
Feature activation+1.558
a
Token a
Feature activation+0.916
ludicrous
Token ludicrous
Feature activation+0.540
charge
Token charge
Feature activation+0.754
?
Token?
Feature activation+0.000
Who
Token Who
Feature activation+0.000
you
Token you
Feature activation+0.075
keep
Token keep
Feature activation+0.707
your
Token your
Feature activation+0.341
doctor
Token doctor
Feature activation+0.312
,
Token,
Feature activation+1.308
though
Token though
Feature activation+1.838
,
Token,
Feature activation+1.758
if
Token if
Feature activation+3.055
your
Token your
Feature activation+0.798
doctor
Token doctor
Feature activation+0.000
cannot
Token cannot
Feature activation+0.216
system
Token system
Feature activation+0.816
of
Token of
Feature activation+0.268
selling
Token selling
Feature activation+0.431
stocks
Token stocks
Feature activation+0.391
a
Token a
Feature activation+1.555
complex
Token complex
Feature activation+1.555
one
Token one
Feature activation+1.741
?
Token?
Feature activation+0.000
What
Token What
Feature activation+0.000
about
Token about
Feature activation+0.568
the
Token the
Feature activation+0.511
,
Token,
Feature activation+0.000
what
Token what
Feature activation+0.000
does
Token does
Feature activation+0.273
this
Token this
Feature activation+0.043
mean
Token mean
Feature activation+1.165
for
Token for
Feature activation+1.635
science
Token science
Feature activation+0.662
?
Token?
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
For
TokenFor
Feature activation+0.000
warning
Token warning
Feature activation+0.000
.
Token.
Feature activation+0.000
Did
Token Did
Feature activation+0.000
you
Token you
Feature activation+0.000
realise
Token realise
Feature activation+0.306
that
Token that
Feature activation+1.581
consumption
Token consumption
Feature activation+0.151
of
Token of
Feature activation+0.000
marg
Token marg
Feature activation+0.000
arine
Tokenarine
Feature activation+0.074
is
Token is
Feature activation+0.910

INTERVAL 1.164 - 1.552
CONTAINS 0.061%

can
Token can
Feature activation+0.000
you
Token you
Feature activation+0.000
take
Token take
Feature activation+0.000
advantage
Token advantage
Feature activation+1.011
of
Token of
Feature activation+0.914
this
Token this
Feature activation+1.256
opportunity
Token opportunity
Feature activation+0.889
?
Token?
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Simply
TokenSimply
Feature activation+0.000
the
Token the
Feature activation+0.573
implementation
Token implementation
Feature activation+0.965
look
Token look
Feature activation+1.561
when
Token when
Feature activation+3.024
there
Token there
Feature activation+1.392
are
Token are
Feature activation+1.464
zero
Token zero
Feature activation+0.498
results
Token results
Feature activation+0.337
?
Token?
Feature activation+0.000
Should
Token Should
Feature activation+0.000
there
Token there
Feature activation+0.558
ag
Token ag
Feature activation+0.000
ro
Tokenro
Feature activation+0.000
-
Token-
Feature activation+0.000
industrial
Tokenindustrial
Feature activation+0.000
complex
Token complex
Feature activation+0.294
,
Token,
Feature activation+1.468
and
Token and
Feature activation+0.945
what
Token what
Feature activation+0.000
the
Token the
Feature activation+0.442
relative
Token relative
Feature activation+0.000
impact
Token impact
Feature activation+0.266
that
Token that
Feature activation+1.954
Washington
Token Washington
Feature activation+1.086
state
Token state
Feature activation+0.134
could
Token could
Feature activation+1.513
be
Token be
Feature activation+0.925
ignored
Token ignored
Feature activation+1.480
when
Token when
Feature activation+1.976
the
Token the
Feature activation+1.057
quake
Token quake
Feature activation+0.210
hits
Token hits
Feature activation+0.370
?
Token?
Feature activation+0.000
our
Token our
Feature activation+0.365
reliance
Token reliance
Feature activation+0.232
on
Token on
Feature activation+0.156
search
Token search
Feature activation+0.294
engines
Token engines
Feature activation+0.429
making
Token making
Feature activation+1.416
it
Token it
Feature activation+0.825
harder
Token harder
Feature activation+1.727
for
Token for
Feature activation+0.641
us
Token us
Feature activation+0.586
to
Token to
Feature activation+0.712

INTERVAL 0.776 - 1.164
CONTAINS 0.127%

pray
Token pray
Feature activation+0.000
tell
Token tell
Feature activation+0.000
,
Token,
Feature activation+0.000
did
Token did
Feature activation+0.000
they
Token they
Feature activation+0.128
feel
Token feel
Feature activation+0.799
the
Token the
Feature activation+0.664
need
Token need
Feature activation+1.052
to
Token to
Feature activation+1.008
defend
Token defend
Feature activation+1.097
themselves
Token themselves
Feature activation+0.398
ARY
TokenARY
Feature activation+0.336
CL
Token CL
Feature activation+0.273
INTON
TokenINTON
Feature activation+0.387
ALWAYS
Token ALWAYS
Feature activation+1.859
C
Token C
Feature activation+0.661
ACK
TokenACK
Feature activation+0.829
LE
TokenLE
Feature activation+0.071
WITH
Token WITH
Feature activation+0.906
G
Token G
Feature activation+0.350
LE
TokenLE
Feature activation+0.000
E
TokenE
Feature activation+0.000
respondents
Token respondents
Feature activation+0.766
consider
Token consider
Feature activation+1.135
when
Token when
Feature activation+1.875
they
Token they
Feature activation+1.397
assess
Token assess
Feature activation+1.177
whether
Token whether
Feature activation+0.841
they
Token they
Feature activation+0.544
âĢ
TokenâĢ
Feature activation+0.473
Ļ
TokenĻ
Feature activation+0.176
re
Tokenre
Feature activation+0.000
regarded
Token regarded
Feature activation+0.356
?
Token?
Feature activation+0.000
Why
Token Why
Feature activation+0.000
should
Token should
Feature activation+0.000
it
Token it
Feature activation+0.924
take
Token take
Feature activation+1.253
more
Token more
Feature activation+0.934
than
Token than
Feature activation+0.888
nine
Token nine
Feature activation+0.463
months
Token months
Feature activation+0.460
to
Token to
Feature activation+0.969
resolve
Token resolve
Feature activation+0.847
Or
Token Or
Feature activation+0.231
is
Token is
Feature activation+0.700
there
Token there
Feature activation+0.812
yet
Token yet
Feature activation+0.271
some
Token some
Feature activation+1.595
other
Token other
Feature activation+1.012
,
Token,
Feature activation+0.516
un
Token un
Feature activation+0.000
me
Tokenme
Feature activation+0.000
asured
Tokenasured
Feature activation+0.408
factor
Token factor
Feature activation+0.964

INTERVAL 0.388 - 0.776
CONTAINS 0.223%

them
Token them
Feature activation+0.000
.
Token.
Feature activation+0.000
Why
Token Why
Feature activation+0.000
will
Token will
Feature activation+0.000
they
Token they
Feature activation+0.270
associate
Token associate
Feature activation+0.626
with
Token with
Feature activation+0.577
tra
Token tra
Feature activation+0.222
itors
Tokenitors
Feature activation+0.400
?
Token?
Feature activation+0.000
The
Token The
Feature activation+0.000
something
Token something
Feature activation+0.703
?
Token?
Feature activation+0.000
What
Token What
Feature activation+0.000
separates
Token separates
Feature activation+0.670
knowledge
Token knowledge
Feature activation+0.681
from
Token from
Feature activation+0.644
mere
Token mere
Feature activation+0.384
belief
Token belief
Feature activation+0.193
?
Token?
Feature activation+0.000
Within
Token Within
Feature activation+0.000
Plato
Token Plato
Feature activation+0.000
perfect
Token perfect
Feature activation+1.193
conditions
Token conditions
Feature activation+1.342
?
Token?
Feature activation+0.000
Do
Token Do
Feature activation+0.000
you
Token you
Feature activation+0.139
want
Token want
Feature activation+0.583
the
Token the
Feature activation+0.865
elements
Token elements
Feature activation+0.980
to
Token to
Feature activation+0.576
impact
Token impact
Feature activation+0.920
it
Token it
Feature activation+0.991
even
Token even
Feature activation+0.000
exist
Token exist
Feature activation+0.000
.
Token.
Feature activation+0.000
Do
Token Do
Feature activation+0.000
you
Token you
Feature activation+0.000
want
Token want
Feature activation+0.417
real
Token real
Feature activation+0.904
cuts
Token cuts
Feature activation+0.861
?
Token?
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.496
Ċ
TokenĊ
Feature activation+0.000
should
Token should
Feature activation+0.778
be
Token be
Feature activation+0.971
cut
Token cut
Feature activation+0.641
off
Token off
Feature activation+0.607
?,
Token?,
Feature activation+0.265
âĢ
TokenâĢ
Feature activation+0.432
Ŀ
TokenĿ
Feature activation+0.070
the
Token the
Feature activation+0.459
RM
Token RM
Feature activation+0.000
U
TokenU
Feature activation+0.000
poll
Token poll
Feature activation+0.000

INTERVAL 0.000 - 0.388
CONTAINS 99.536%

25
Token25
Feature activation+0.000
/
Token/
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Here
TokenHere
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
how
Token how
Feature activation+0.000
to
Token to
Feature activation+0.000
judge
Token judge
Feature activation+0.000
the
Token the
Feature activation+0.000
In
TokenIn
Feature activation+0.000
any
Token any
Feature activation+0.000
event
Token event
Feature activation+0.000
,
Token,
Feature activation+0.000
it
Token it
Feature activation+0.000
is
Token is
Feature activation+0.000
unlikely
Token unlikely
Feature activation+0.000
that
Token that
Feature activation+0.000
North
Token North
Feature activation+0.000
Koreans
Token Koreans
Feature activation+0.000
want
Token want
Feature activation+0.000
bulky
Token bulky
Feature activation+0.000
organic
Token organic
Feature activation+0.000
oxid
Token oxid
Feature activation+0.000
izing
Tokenizing
Feature activation+0.000
agent
Token agent
Feature activation+0.000
and
Token and
Feature activation+0.000
an
Token an
Feature activation+0.000
optimized
Token optimized
Feature activation+0.000
solvent
Token solvent
Feature activation+0.000
system
Token system
Feature activation+0.000
to
Token to
Feature activation+0.000
her
Tokenher
Feature activation+0.000
in
Token in
Feature activation+0.000
Jason
Token Jason
Feature activation+0.000
Hat
Token Hat
Feature activation+0.000
cher
Tokencher
Feature activation+0.000
and
Token and
Feature activation+0.000
three
Token three
Feature activation+0.000
inside
Token inside
Feature activation+0.000
linebackers
Token linebackers
Feature activation+0.000
(
Token (
Feature activation+0.000
two
Tokentwo
Feature activation+0.000
value
Token value
Feature activation+0.000
of
Token of
Feature activation+0.000
dropping
Token dropping
Feature activation+0.000
high
Token high
Feature activation+0.000
tier
Token tier
Feature activation+0.000
maps
Token maps
Feature activation+0.000
that
Token that
Feature activation+0.000
was
Token was
Feature activation+0.000
a
Token a
Feature activation+0.000
part
Token part
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

Top feature 9 in H5.3: (feature 36407)

TOP ACTIVATIONS
MAX = 3.703

READ
TokenREAD
Feature activation+0.000
MORE
Token MORE
Feature activation+0.000
:
Token:
Feature activation+0.000
North
Token North
Feature activation+1.344
Van
Token Van
Feature activation+1.045
developer
Token developer
Feature activation+3.703
says
Token says
Feature activation+0.906
the
Token the
Feature activation+2.042
industry
Token industry
Feature activation+1.479
needs
Token needs
Feature activation+1.601
to
Token to
Feature activation+1.267
Ċ
TokenĊ
Feature activation+0.000
SEE
TokenSEE
Feature activation+0.000
ALSO
Token ALSO
Feature activation+0.000
:
Token:
Feature activation+0.000
I
Token I
Feature activation+1.136
tried
Token tried
Feature activation+3.585
to
Token to
Feature activation+2.376
find
Token find
Feature activation+2.958
a
Token a
Feature activation+1.878
new
Token new
Feature activation+2.562
B
Token B
Feature activation+2.439
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
MORE
TokenMORE
Feature activation+0.000
:
Token:
Feature activation+0.000
Cookies
Token Cookies
Feature activation+3.571
n
Token n
Feature activation+2.393
âĢ
TokenâĢ
Feature activation+0.340
Ļ
TokenĻ
Feature activation+1.067
C
Token C
Feature activation+1.522
reme
Tokenreme
Feature activation+0.566
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Read
TokenRead
Feature activation+0.000
more
Token more
Feature activation+0.000
:
Token:
Feature activation+0.000
Student
Token Student
Feature activation+3.372
comm
Token comm
Feature activation+2.441
utes
Tokenutes
Feature activation+2.216
1
Token 1
Feature activation+0.943
,
Token,
Feature activation+0.650
000
Token000
Feature activation+0.752
Ċ
TokenĊ
Feature activation+0.000
Previous
TokenPrevious
Feature activation+0.000
Coverage
Token Coverage
Feature activation+0.000
:
Token:
Feature activation+0.233
Tampa
Token Tampa
Feature activation+1.958
's
Token's
Feature activation+3.367
University
Token University
Feature activation+1.369
Mall
Token Mall
Feature activation+1.265
to
Token to
Feature activation+1.201
get
Token get
Feature activation+0.302
major
Token major
Feature activation+0.635
MORE
TokenMORE
Feature activation+0.000
:
Token:
Feature activation+0.000
7
Token 7
Feature activation+1.399
Weird
Token Weird
Feature activation+2.984
Head
Token Head
Feature activation+1.429
ache
Tokenache
Feature activation+3.289
Tr
Token Tr
Feature activation+1.235
ig
Tokenig
Feature activation+0.194
gers
Tokengers
Feature activation+1.539
Ċ
TokenĊ
Feature activation+1.862
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
*
Token*
Feature activation+1.097
Victory
Token Victory
Feature activation+2.988
at
Token at
Feature activation+2.362
hand
Token hand
Feature activation+3.156
in
Token in
Feature activation+1.424
multinational
Token multinational
Feature activation+1.993
tax
Token tax
Feature activation+0.853
battle
Token battle
Feature activation+1.212
but
Token but
Feature activation+1.200
:
Token:
Feature activation+0.145
B
Token B
Feature activation+0.771
.
Token.
Feature activation+0.414
C
TokenC
Feature activation+1.270
.
Token.
Feature activation+0.877
photographer
Token photographer
Feature activation+3.126
captures
Token captures
Feature activation+1.300
rescue
Token rescue
Feature activation+1.570
of
Token of
Feature activation+0.876
bald
Token bald
Feature activation+0.652
eagle
Token eagle
Feature activation+1.811
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
[
Token[
Feature activation+0.000
RELATED
TokenRELATED
Feature activation+0.000
:
Token:
Feature activation+0.672
Students
Token Students
Feature activation+3.102
fear
Token fear
Feature activation+2.863
âĢ
Token âĢ
Feature activation+1.165
ĺ
Tokenĺ
Feature activation+1.544
m
Tokenm
Feature activation+0.492
ilit
Tokenilit
Feature activation+0.461
MORE
Token MORE
Feature activation+0.000
:
Token:
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
*
Token*
Feature activation+1.467
Games
Token Games
Feature activation+3.063
arcade
Token arcade
Feature activation+2.957
proves
Token proves
Feature activation+2.356
'
Token '
Feature activation+0.906
old
Tokenold
Feature activation+0.285
school
Token school
Feature activation+0.000
BMW
Token BMW
Feature activation+1.966
Concept
Token Concept
Feature activation+2.431
X
Token X
Feature activation+2.315
7
Token7
Feature activation+2.688
i
Token i
Feature activation+0.939
Performance
TokenPerformance
Feature activation+3.037
Preview
Token Preview
Feature activation+0.890
ed
Tokened
Feature activation+1.027
Ahead
Token Ahead
Feature activation+0.471
Of
Token Of
Feature activation+0.462
De
Token De
Feature activation+0.356
Ċ
TokenĊ
Feature activation+0.000
ST
TokenST
Feature activation+0.000
ORY
TokenORY
Feature activation+0.000
:
Token:
Feature activation+0.028
Kevin
Token Kevin
Feature activation+1.167
Smith
Token Smith
Feature activation+2.991
Slam
Token Slam
Feature activation+0.749
s
Tokens
Feature activation+1.557
Indie
Token Indie
Feature activation+1.185
Spirits
Token Spirits
Feature activation+0.981
on
Token on
Feature activation+0.721
MORE
Token MORE
Feature activation+0.000
:
Token:
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
*
Token*
Feature activation+1.097
Victory
Token Victory
Feature activation+2.988
at
Token at
Feature activation+2.362
hand
Token hand
Feature activation+3.156
in
Token in
Feature activation+1.424
multinational
Token multinational
Feature activation+1.993
tax
Token tax
Feature activation+0.853
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
MORE
TokenMORE
Feature activation+0.000
:
Token:
Feature activation+0.000
7
Token 7
Feature activation+1.399
Weird
Token Weird
Feature activation+2.984
Head
Token Head
Feature activation+1.429
ache
Tokenache
Feature activation+3.289
Tr
Token Tr
Feature activation+1.235
ig
Tokenig
Feature activation+0.194
gers
Tokengers
Feature activation+1.539
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
RELATED
TokenRELATED
Feature activation+0.000
:
Token:
Feature activation+0.615
John
Token John
Feature activation+0.974
Boehner
Token Boehner
Feature activation+2.963
Claims
Token Claims
Feature activation+1.142
He
Token He
Feature activation+0.000
Was
Token Was
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.729
ĺ
Tokenĺ
Feature activation+0.778
âĢ
TokenâĢ
Feature activation+0.340
Ļ
TokenĻ
Feature activation+1.067
C
Token C
Feature activation+1.522
reme
Tokenreme
Feature activation+0.566
Ore
Token Ore
Feature activation+0.424
os
Tokenos
Feature activation+2.958
:
Token:
Feature activation+1.451
The
Token The
Feature activation+1.041
Most
Token Most
Feature activation+0.694
Meta
Token Meta
Feature activation+0.635
Sn
Token Sn
Feature activation+0.288
ALSO
Token ALSO
Feature activation+0.000
:
Token:
Feature activation+0.000
I
Token I
Feature activation+1.136
tried
Token tried
Feature activation+3.585
to
Token to
Feature activation+2.376
find
Token find
Feature activation+2.958
a
Token a
Feature activation+1.878
new
Token new
Feature activation+2.562
B
Token B
Feature activation+2.439
FF
TokenFF
Feature activation+2.334
on
Token on
Feature activation+2.547
:
Token:
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
*
Token*
Feature activation+1.467
Games
Token Games
Feature activation+3.063
arcade
Token arcade
Feature activation+2.957
proves
Token proves
Feature activation+2.356
'
Token '
Feature activation+0.906
old
Tokenold
Feature activation+0.285
school
Token school
Feature activation+0.000
'
Token'
Feature activation+1.604
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
RELATED
TokenRELATED
Feature activation+0.000
:
Token:
Feature activation+0.755
Col
Token Col
Feature activation+2.955
qu
Tokenqu
Feature activation+1.602
itt
Tokenitt
Feature activation+2.754
County
Token County
Feature activation+2.623
kicker
Token kicker
Feature activation+2.869
leads
Token leads
Feature activation+2.065
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Related
TokenRelated
Feature activation+0.000
:
Token:
Feature activation+0.937
Diablo
Token Diablo
Feature activation+2.624
4
Token 4
Feature activation+2.866
âĢĵ
Token âĢĵ
Feature activation+2.401
everything
Token everything
Feature activation+1.110
you
Token you
Feature activation+1.511
need
Token need
Feature activation+0.969
to
Token to
Feature activation+0.608

Top DFA by src position
MAX = 4.212

.
Token.
Feature activation-0.004
Top resid features:
Ċ
TokenĊ
Feature activation+0.034
Top resid features:
Ċ
TokenĊ
Feature activation+0.062
Top resid features:
READ
TokenREAD
Feature activation+0.218
Top resid features:
MORE
Token MORE
Feature activation+0.342
Top resid features:
:
Token:
Feature activation+3.384
Top resid features:
North
Token North
Feature activation+0.280
Top resid features:
Van
Token Van
Feature activation+0.151
Top resid features:
developer
Token developer
Feature activation+0.052
Top resid features:
says
Token says
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation-0.002
Top resid features:
Ċ
TokenĊ
Feature activation+0.045
Top resid features:
SEE
TokenSEE
Feature activation+0.222
Top resid features:
ALSO
Token ALSO
Feature activation+0.886
Top resid features:
:
Token:
Feature activation+2.517
Top resid features:
I
Token I
Feature activation+0.701
Top resid features:
tried
Token tried
Feature activation+0.047
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
find
Token find
Feature activation+0.000
Top resid features:
a
Token a
Feature activation+0.000
Top resid features:
said
Token said
Feature activation-0.030
Top resid features:
.
Token.
Feature activation-0.013
Top resid features:
Ċ
TokenĊ
Feature activation-0.014
Top resid features:
Ċ
TokenĊ
Feature activation-0.005
Top resid features:
MORE
TokenMORE
Feature activation+0.221
Top resid features:
:
Token:
Feature activation+4.212
Top resid features:
Cookies
Token Cookies
Feature activation+0.096
Top resid features:
n
Token n
Feature activation+0.000
Top resid features:
âĢ
TokenâĢ
Feature activation+0.000
Top resid features:
Ļ
TokenĻ
Feature activation+0.000
Top resid features:
C
Token C
Feature activation+0.000
Top resid features:
.
Token.
Feature activation-0.022
Top resid features:
Ċ
TokenĊ
Feature activation+0.032
Top resid features:
Ċ
TokenĊ
Feature activation+0.038
Top resid features:
Read
TokenRead
Feature activation+0.146
Top resid features:
more
Token more
Feature activation+0.267
Top resid features:
:
Token:
Feature activation+3.402
Top resid features:
Student
Token Student
Feature activation+0.199
Top resid features:
comm
Token comm
Feature activation+0.000
Top resid features:
utes
Tokenutes
Feature activation+0.000
Top resid features:
1
Token 1
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
.
Token.
Feature activation-0.001
Top resid features:
Ċ
TokenĊ
Feature activation+0.004
Top resid features:
Ċ
TokenĊ
Feature activation+0.011
Top resid features:
Previous
TokenPrevious
Feature activation-0.010
Top resid features:
Coverage
Token Coverage
Feature activation+1.644
Top resid features:
:
Token:
Feature activation+2.117
Top resid features:
Tampa
Token Tampa
Feature activation+0.210
Top resid features:
's
Token's
Feature activation+0.077
Top resid features:
University
Token University
Feature activation+0.000
Top resid features:
Mall
Token Mall
Feature activation+0.000
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
doctor
Token doctor
Feature activation-0.001
Top resid features:
.)
Token .)
Feature activation-0.018
Top resid features:
Ċ
TokenĊ
Feature activation+0.013
Top resid features:
Ċ
TokenĊ
Feature activation+0.037
Top resid features:
MORE
TokenMORE
Feature activation+0.165
Top resid features:
:
Token:
Feature activation+2.984
Top resid features:
7
Token 7
Feature activation+0.465
Top resid features:
Weird
Token Weird
Feature activation+0.352
Top resid features:
Head
Token Head
Feature activation+0.112
Top resid features:
ache
Tokenache
Feature activation-0.006
Top resid features:
Tr
Token Tr
Feature activation+0.000
Top resid features:
READ
TokenREAD
Feature activation+0.093
Top resid features:
MORE
Token MORE
Feature activation+0.117
Top resid features:
:
Token:
Feature activation+0.407
Top resid features:
Ċ
TokenĊ
Feature activation+0.249
Top resid features:
Ċ
TokenĊ
Feature activation+0.315
Top resid features:
*
Token*
Feature activation+2.556
Top resid features:
Victory
Token Victory
Feature activation+0.100
Top resid features:
at
Token at
Feature activation+0.041
Top resid features:
hand
Token hand
Feature activation+0.092
Top resid features:
in
Token in
Feature activation+0.000
Top resid features:
multinational
Token multinational
Feature activation+0.000
Top resid features:
.
Token.
Feature activation-0.007
Top resid features:
Ċ
TokenĊ
Feature activation+0.003
Top resid features:
Ċ
TokenĊ
Feature activation+0.017
Top resid features:
READ
TokenREAD
Feature activation+0.130
Top resid features:
MORE
Token MORE
Feature activation+0.409
Top resid features:
:
Token:
Feature activation+3.035
Top resid features:
B
Token B
Feature activation+0.196
Top resid features:
.
Token.
Feature activation+0.009
Top resid features:
C
TokenC
Feature activation+0.063
Top resid features:
.
Token.
Feature activation+0.030
Top resid features:
photographer
Token photographer
Feature activation+0.067
Top resid features:
opinion
Token opinion
Feature activation+0.006
Top resid features:
."
Token."
Feature activation-0.001
Top resid features:
Ċ
TokenĊ
Feature activation-0.005
Top resid features:
Ċ
TokenĊ
Feature activation+0.007
Top resid features:
[
Token[
Feature activation+0.988
Top resid features:
RELATED
TokenRELATED
Feature activation+1.301
Top resid features:
:
Token:
Feature activation+1.180
Top resid features:
Students
Token Students
Feature activation+0.162
Top resid features:
fear
Token fear
Feature activation+0.000
Top resid features:
âĢ
Token âĢ
Feature activation+0.000
Top resid features:
ĺ
Tokenĺ
Feature activation+0.000
Top resid features:
READ
TokenREAD
Feature activation+0.016
Top resid features:
MORE
Token MORE
Feature activation+0.169
Top resid features:
:
Token:
Feature activation+0.869
Top resid features:
Ċ
TokenĊ
Feature activation+0.766
Top resid features:
Ċ
TokenĊ
Feature activation+0.840
Top resid features:
*
Token*
Feature activation+1.128
Top resid features:
Games
Token Games
Feature activation+0.079
Top resid features:
arcade
Token arcade
Feature activation+0.000
Top resid features:
proves
Token proves
Feature activation+0.000
Top resid features:
'
Token '
Feature activation+0.000
Top resid features:
old
Tokenold
Feature activation+0.000
Top resid features:
.
Token.
Feature activation-0.011
Top resid features:
Ċ
TokenĊ
Feature activation+0.011
Top resid features:
Ċ
TokenĊ
Feature activation+0.016
Top resid features:
Also
TokenAlso
Feature activation+0.022
Top resid features:
Read
Token Read
Feature activation+1.189
Top resid features:
:
Token:
Feature activation+2.238
Top resid features:
BMW
Token BMW
Feature activation+0.205
Top resid features:
Concept
Token Concept
Feature activation+0.096
Top resid features:
X
Token X
Feature activation+0.078
Top resid features:
7
Token7
Feature activation-0.004
Top resid features:
i
Token i
Feature activation+0.079
Top resid features:
Ŀ
TokenĿ
Feature activation+0.013
Top resid features:
Ċ
TokenĊ
Feature activation+0.006
Top resid features:
Ċ
TokenĊ
Feature activation+0.019
Top resid features:
ST
TokenST
Feature activation+0.072
Top resid features:
ORY
TokenORY
Feature activation+0.691
Top resid features:
:
Token:
Feature activation+2.460
Top resid features:
Kevin
Token Kevin
Feature activation+0.362
Top resid features:
Smith
Token Smith
Feature activation+0.183
Top resid features:
Slam
Token Slam
Feature activation+0.000
Top resid features:
s
Tokens
Feature activation+0.000
Top resid features:
Indie
Token Indie
Feature activation+0.000
Top resid features:
READ
TokenREAD
Feature activation+0.076
Top resid features:
MORE
Token MORE
Feature activation+0.106
Top resid features:
:
Token:
Feature activation+0.493
Top resid features:
Ċ
TokenĊ
Feature activation+0.325
Top resid features:
Ċ
TokenĊ
Feature activation+0.365
Top resid features:
*
Token*
Feature activation+2.144
Top resid features:
Victory
Token Victory
Feature activation+0.043
Top resid features:
at
Token at
Feature activation+0.000
Top resid features:
hand
Token hand
Feature activation+0.000
Top resid features:
in
Token in
Feature activation+0.000
Top resid features:
multinational
Token multinational
Feature activation+0.000
Top resid features:
doctor
Token doctor
Feature activation+0.008
Top resid features:
.)
Token .)
Feature activation-0.024
Top resid features:
Ċ
TokenĊ
Feature activation+0.001
Top resid features:
Ċ
TokenĊ
Feature activation+0.011
Top resid features:
MORE
TokenMORE
Feature activation+0.173
Top resid features:
:
Token:
Feature activation+2.922
Top resid features:
7
Token 7
Feature activation+0.481
Top resid features:
Weird
Token Weird
Feature activation+0.077
Top resid features:
Head
Token Head
Feature activation+0.000
Top resid features:
ache
Tokenache
Feature activation+0.000
Top resid features:
Tr
Token Tr
Feature activation+0.000
Top resid features:
like
Token like
Feature activation-0.017
Top resid features:
children
Token children
Feature activation-0.005
Top resid features:
.
Token.
Feature activation-0.017
Top resid features:
Ċ
TokenĊ
Feature activation-0.033
Top resid features:
Ċ
TokenĊ
Feature activation-0.001
Top resid features:
RELATED
TokenRELATED
Feature activation+2.051
Top resid features:
:
Token:
Feature activation+1.611
Top resid features:
John
Token John
Feature activation+0.147
Top resid features:
Boehner
Token Boehner
Feature activation+0.050
Top resid features:
Claims
Token Claims
Feature activation+0.000
Top resid features:
He
Token He
Feature activation+0.000
Top resid features:
said
Token said
Feature activation-0.019
Top resid features:
.
Token.
Feature activation-0.008
Top resid features:
Ċ
TokenĊ
Feature activation-0.006
Top resid features:
Ċ
TokenĊ
Feature activation+0.005
Top resid features:
MORE
TokenMORE
Feature activation+0.189
Top resid features:
:
Token:
Feature activation+3.210
Top resid features:
Cookies
Token Cookies
Feature activation+0.095
Top resid features:
n
Token n
Feature activation+0.076
Top resid features:
âĢ
TokenâĢ
Feature activation+0.053
Top resid features:
Ļ
TokenĻ
Feature activation+0.084
Top resid features:
C
Token C
Feature activation+0.240
Top resid features:
.
Token.
Feature activation-0.008
Top resid features:
Ċ
TokenĊ
Feature activation-0.014
Top resid features:
Ċ
TokenĊ
Feature activation+0.038
Top resid features:
SEE
TokenSEE
Feature activation+0.065
Top resid features:
ALSO
Token ALSO
Feature activation+0.946
Top resid features:
:
Token:
Feature activation+1.978
Top resid features:
I
Token I
Feature activation+0.651
Top resid features:
tried
Token tried
Feature activation+0.252
Top resid features:
to
Token to
Feature activation+0.041
Top resid features:
find
Token find
Feature activation-0.075
Top resid features:
a
Token a
Feature activation+0.000
Top resid features:
READ
TokenREAD
Feature activation+0.058
Top resid features:
MORE
Token MORE
Feature activation+0.221
Top resid features:
:
Token:
Feature activation+0.915
Top resid features:
Ċ
TokenĊ
Feature activation+0.609
Top resid features:
Ċ
TokenĊ
Feature activation+0.545
Top resid features:
*
Token*
Feature activation+1.249
Top resid features:
Games
Token Games
Feature activation+0.241
Top resid features:
arcade
Token arcade
Feature activation+0.010
Top resid features:
proves
Token proves
Feature activation+0.000
Top resid features:
'
Token '
Feature activation+0.000
Top resid features:
old
Tokenold
Feature activation+0.000
Top resid features:
stand
Token stand
Feature activation+0.004
Top resid features:
up
Token up
Feature activation-0.005
Top resid features:
.
Token.
Feature activation-0.007
Top resid features:
Ċ
TokenĊ
Feature activation+0.041
Top resid features:
Ċ
TokenĊ
Feature activation+0.048
Top resid features:
RELATED
TokenRELATED
Feature activation+2.015
Top resid features:
:
Token:
Feature activation+1.445
Top resid features:
Col
Token Col
Feature activation+0.051
Top resid features:
qu
Tokenqu
Feature activation+0.000
Top resid features:
itt
Tokenitt
Feature activation+0.000
Top resid features:
County
Token County
Feature activation+0.000
Top resid features:
âĢ
TokenâĢ
Feature activation+0.019
Top resid features:
Ŀ
TokenĿ
Feature activation+0.020
Top resid features:
Ċ
TokenĊ
Feature activation+0.008
Top resid features:
Ċ
TokenĊ
Feature activation+0.045
Top resid features:
Related
TokenRelated
Feature activation+0.960
Top resid features:
:
Token:
Feature activation+2.359
Top resid features:
Diablo
Token Diablo
Feature activation+0.137
Top resid features:
4
Token 4
Feature activation+0.072
Top resid features:
âĢĵ
Token âĢĵ
Feature activation+0.000
Top resid features:
everything
Token everything
Feature activation+0.000
Top resid features:
you
Token you
Feature activation+0.000
Top resid features:

Decoder Weights Distribution

Head 0: 0.03

Head 1: 0.03

Head 2: 0.09

Head 3: 0.49

Head 4: 0.08

Head 5: 0.02

Head 6: 0.03

Head 7: 0.03

Head 8: 0.07

Head 9: 0.03

Head 10: 0.06

Head 11: 0.05

Positive logits

?]3.17

Replay3.08

!]2.86

').2.82

!'2.67

']2.54

]'2.50

!'"2.45

>]2.44

?'2.40

]}2.28

}}2.25

.)2.22

});2.22

?).2.18

.]2.14

…]2.14

})2.12

)'2.09

');2.08

Negative logits

cffffcc-1.93

nown-1.86

entimes-1.84

esm-1.79

POSE-1.75

ODUCT-1.73

xual-1.73

inarily-1.72

omever-1.71

龍喚士-1.70

norm-1.69

theless-1.66

Whereas-1.65

initions-1.62

ircraft-1.62

oldown-1.59

whatever-1.57

asionally-1.56

76561-1.55

FTWARE-1.54

INTERVAL 3.333 - 3.703
CONTAINS 0.001%

READ
TokenREAD
Feature activation+0.000
MORE
Token MORE
Feature activation+0.000
:
Token:
Feature activation+0.000
North
Token North
Feature activation+1.344
Van
Token Van
Feature activation+1.045
developer
Token developer
Feature activation+3.703
says
Token says
Feature activation+0.906
the
Token the
Feature activation+2.042
industry
Token industry
Feature activation+1.479
needs
Token needs
Feature activation+1.601
to
Token to
Feature activation+1.267
Ċ
TokenĊ
Feature activation+0.000
Previous
TokenPrevious
Feature activation+0.000
Coverage
Token Coverage
Feature activation+0.000
:
Token:
Feature activation+0.233
Tampa
Token Tampa
Feature activation+1.958
's
Token's
Feature activation+3.367
University
Token University
Feature activation+1.369
Mall
Token Mall
Feature activation+1.265
to
Token to
Feature activation+1.201
get
Token get
Feature activation+0.302
major
Token major
Feature activation+0.635
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
MORE
TokenMORE
Feature activation+0.000
:
Token:
Feature activation+0.000
Cookies
Token Cookies
Feature activation+3.571
n
Token n
Feature activation+2.393
âĢ
TokenâĢ
Feature activation+0.340
Ļ
TokenĻ
Feature activation+1.067
C
Token C
Feature activation+1.522
reme
Tokenreme
Feature activation+0.566
Ċ
TokenĊ
Feature activation+0.000
SEE
TokenSEE
Feature activation+0.000
ALSO
Token ALSO
Feature activation+0.000
:
Token:
Feature activation+0.000
I
Token I
Feature activation+1.136
tried
Token tried
Feature activation+3.585
to
Token to
Feature activation+2.376
find
Token find
Feature activation+2.958
a
Token a
Feature activation+1.878
new
Token new
Feature activation+2.562
B
Token B
Feature activation+2.439
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Read
TokenRead
Feature activation+0.000
more
Token more
Feature activation+0.000
:
Token:
Feature activation+0.000
Student
Token Student
Feature activation+3.372
comm
Token comm
Feature activation+2.441
utes
Tokenutes
Feature activation+2.216
1
Token 1
Feature activation+0.943
,
Token,
Feature activation+0.650
000
Token000
Feature activation+0.752

INTERVAL 2.962 - 3.333
CONTAINS 0.001%

Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
RELATED
TokenRELATED
Feature activation+0.000
:
Token:
Feature activation+0.615
John
Token John
Feature activation+0.974
Boehner
Token Boehner
Feature activation+2.963
Claims
Token Claims
Feature activation+1.142
He
Token He
Feature activation+0.000
Was
Token Was
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.729
ĺ
Tokenĺ
Feature activation+0.778
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
*
Token*
Feature activation+1.097
Victory
Token Victory
Feature activation+2.988
at
Token at
Feature activation+2.362
hand
Token hand
Feature activation+3.156
in
Token in
Feature activation+1.424
multinational
Token multinational
Feature activation+1.993
tax
Token tax
Feature activation+0.853
battle
Token battle
Feature activation+1.212
but
Token but
Feature activation+1.200
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
MORE
TokenMORE
Feature activation+0.000
:
Token:
Feature activation+0.000
7
Token 7
Feature activation+1.399
Weird
Token Weird
Feature activation+2.984
Head
Token Head
Feature activation+1.429
ache
Tokenache
Feature activation+3.289
Tr
Token Tr
Feature activation+1.235
ig
Tokenig
Feature activation+0.194
gers
Tokengers
Feature activation+1.539
MORE
TokenMORE
Feature activation+0.000
:
Token:
Feature activation+0.000
7
Token 7
Feature activation+1.399
Weird
Token Weird
Feature activation+2.984
Head
Token Head
Feature activation+1.429
ache
Tokenache
Feature activation+3.289
Tr
Token Tr
Feature activation+1.235
ig
Tokenig
Feature activation+0.194
gers
Tokengers
Feature activation+1.539
Ċ
TokenĊ
Feature activation+1.862
Ċ
TokenĊ
Feature activation+0.000
:
Token:
Feature activation+0.145
B
Token B
Feature activation+0.771
.
Token.
Feature activation+0.414
C
TokenC
Feature activation+1.270
.
Token.
Feature activation+0.877
photographer
Token photographer
Feature activation+3.126
captures
Token captures
Feature activation+1.300
rescue
Token rescue
Feature activation+1.570
of
Token of
Feature activation+0.876
bald
Token bald
Feature activation+0.652
eagle
Token eagle
Feature activation+1.811

INTERVAL 2.592 - 2.962
CONTAINS 0.002%

Ċ
TokenĊ
Feature activation+0.000
MORE
TokenMORE
Feature activation+0.000
:
Token:
Feature activation+0.000
Storm
Token Storm
Feature activation+2.470
ch
Token ch
Feature activation+2.109
aser
Tokenaser
Feature activation+2.604
video
Token video
Feature activation+1.359
of
Token of
Feature activation+0.547
Tup
Token Tup
Feature activation+0.675
elo
Tokenelo
Feature activation+1.241
tornado
Token tornado
Feature activation+1.609
READ
Token READ
Feature activation+0.000
ING
TokenING
Feature activation+0.000
:
Token:
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.564
Ċ
TokenĊ
Feature activation+0.000
One
TokenOne
Feature activation+2.855
man
Token man
Feature activation+2.205
âĢ
TokenâĢ
Feature activation+1.014
Ļ
TokenĻ
Feature activation+0.850
s
Tokens
Feature activation+0.346
hard
Token hard
Feature activation+0.539
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Also
TokenAlso
Feature activation+0.000
Read
Token Read
Feature activation+0.000
:
Token:
Feature activation+0.888
Donald
Token Donald
Feature activation+2.614
Trump
Token Trump
Feature activation+2.441
Makes
Token Makes
Feature activation+1.383
History
Token History
Feature activation+0.995
With
Token With
Feature activation+1.526
Zero
Token Zero
Feature activation+0.714
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Related
TokenRelated
Feature activation+0.000
:
Token:
Feature activation+0.903
White
Token White
Feature activation+1.550
House
Token House
Feature activation+2.691
,
Token,
Feature activation+1.493
GOP
Token GOP
Feature activation+1.601
Intel
Token Intel
Feature activation+0.921
Committee
Token Committee
Feature activation+1.911
Heads
Token Heads
Feature activation+0.658
:
Token:
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
*
Token*
Feature activation+1.467
Games
Token Games
Feature activation+3.063
arcade
Token arcade
Feature activation+2.957
proves
Token proves
Feature activation+2.356
'
Token '
Feature activation+0.906
old
Tokenold
Feature activation+0.285
school
Token school
Feature activation+0.000
'
Token'
Feature activation+1.604

INTERVAL 2.222 - 2.592
CONTAINS 0.004%

Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
MORE
TokenMORE
Feature activation+0.000
:
Token:
Feature activation+0.000
North
Token North
Feature activation+1.339
Korea
Token Korea
Feature activation+2.297
just
Token just
Feature activation+2.120
built
Token built
Feature activation+1.962
a
Token a
Feature activation+1.898
really
Token really
Feature activation+1.730
stunning
Token stunning
Feature activation+1.205
Also
TokenAlso
Feature activation+0.000
Read
Token Read
Feature activation+0.000
:
Token:
Feature activation+0.040
BMW
Token BMW
Feature activation+1.966
Concept
Token Concept
Feature activation+2.431
X
Token X
Feature activation+2.315
7
Token7
Feature activation+2.688
i
Token i
Feature activation+0.939
Performance
TokenPerformance
Feature activation+3.037
Preview
Token Preview
Feature activation+0.890
ed
Tokened
Feature activation+1.027
'
Token'
Feature activation+1.403
Mad
TokenMad
Feature activation+0.637
Men
Token Men
Feature activation+1.410
'
Token'
Feature activation+2.458
recap
Token recap
Feature activation+1.604
:
Token:
Feature activation+2.540
Out
Token Out
Feature activation+2.356
with
Token with
Feature activation+1.064
a
Token a
Feature activation+1.110
whim
Token whim
Feature activation+0.553
per
Tokenper
Feature activation+0.358
more
Token more
Feature activation+0.000
:
Token:
Feature activation+0.077
Net
Token Net
Feature activation+0.547
an
Tokenan
Feature activation+0.197
jah
Tokenjah
Feature activation+1.620
i
Tokeni
Feature activation+2.537
says
Token says
Feature activation+1.204
Israel
Token Israel
Feature activation+1.113
will
Token will
Feature activation+1.232
not
Token not
Feature activation+0.736
accept
Token accept
Feature activation+0.370
Coverage
Token Coverage
Feature activation+0.000
Aden
Token Aden
Feature activation+2.076
governor
Token governor
Feature activation+1.654
accuses
Token accuses
Feature activation+1.812
Yemen
Token Yemen
Feature activation+1.192
's
Token's
Feature activation+2.433
Houth
Token Houth
Feature activation+0.099
is
Tokenis
Feature activation+1.642
of
Token of
Feature activation+1.060
air
Token air
Feature activation+0.428
strike
Token strike
Feature activation+0.312

INTERVAL 1.851 - 2.222
CONTAINS 0.007%

Start
Token Start
Feature activation+2.040
ups
Token ups
Feature activation+2.425
and
Token and
Feature activation+1.179
intolerance
Token intolerance
Feature activation+1.620
can
Token can
Feature activation+2.854
't
Token't
Feature activation+1.908
go
Token go
Feature activation+0.473
together
Token together
Feature activation+0.000
,
Token,
Feature activation+0.537
says
Token says
Feature activation+0.228
Rahul
Token Rahul
Feature activation+0.154
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Previous
TokenPrevious
Feature activation+0.000
Coverage
Token Coverage
Feature activation+0.000
:
Token:
Feature activation+0.233
Tampa
Token Tampa
Feature activation+1.958
's
Token's
Feature activation+3.367
University
Token University
Feature activation+1.369
Mall
Token Mall
Feature activation+1.265
to
Token to
Feature activation+1.201
get
Token get
Feature activation+0.302
ideshow
Tokenideshow
Feature activation+0.000
:
Token:
Feature activation+0.624
25
Token 25
Feature activation+1.217
RI
Token RI
Feature activation+1.244
Super
Token Super
Feature activation+1.101
markets
Tokenmarkets
Feature activation+2.075
with
Token with
Feature activation+2.097
the
Token the
Feature activation+1.434
Highest
Token Highest
Feature activation+1.139
Number
Token Number
Feature activation+1.061
of
Token of
Feature activation+0.630
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Related
TokenRelated
Feature activation+0.000
:
Token:
Feature activation+0.350
Host
Token Host
Feature activation+2.161
ess
Tokeness
Feature activation+1.987
jobs
Token jobs
Feature activation+2.170
-
Token -
Feature activation+1.810
'
Token '
Feature activation+0.615
Great
TokenGreat
Feature activation+0.559
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Read
TokenRead
Feature activation+0.000
:
Token:
Feature activation+0.000
Here
Token Here
Feature activation+2.040
âĢ
TokenâĢ
Feature activation+0.475
Ļ
TokenĻ
Feature activation+0.340
s
Tokens
Feature activation+0.998
what
Token what
Feature activation+0.465
bitcoin
Token bitcoin
Feature activation+0.756

INTERVAL 1.481 - 1.851
CONTAINS 0.011%

RELATED
TokenRELATED
Feature activation+0.000
:
Token:
Feature activation+0.672
Students
Token Students
Feature activation+3.102
fear
Token fear
Feature activation+2.863
âĢ
Token âĢ
Feature activation+1.165
ĺ
Tokenĺ
Feature activation+1.544
m
Tokenm
Feature activation+0.492
ilit
Tokenilit
Feature activation+0.461
ar
Tokenar
Feature activation+0.167
ized
Tokenized
Feature activation+1.024
environment
Token environment
Feature activation+1.186
Ċ
TokenĊ
Feature activation+0.000
Related
TokenRelated
Feature activation+0.000
:
Token:
Feature activation+0.903
White
Token White
Feature activation+1.550
House
Token House
Feature activation+2.691
,
Token,
Feature activation+1.493
GOP
Token GOP
Feature activation+1.601
Intel
Token Intel
Feature activation+0.921
Committee
Token Committee
Feature activation+1.911
Heads
Token Heads
Feature activation+0.658
Push
Token Push
Feature activation+0.518
Read
Token Read
Feature activation+0.000
:
Token:
Feature activation+0.073
Start
Token Start
Feature activation+2.040
ups
Token ups
Feature activation+2.425
and
Token and
Feature activation+1.179
intolerance
Token intolerance
Feature activation+1.620
can
Token can
Feature activation+2.854
't
Token't
Feature activation+1.908
go
Token go
Feature activation+0.473
together
Token together
Feature activation+0.000
,
Token,
Feature activation+0.537
:
Token:
Feature activation+0.000
Cookies
Token Cookies
Feature activation+3.571
n
Token n
Feature activation+2.393
âĢ
TokenâĢ
Feature activation+0.340
Ļ
TokenĻ
Feature activation+1.067
C
Token C
Feature activation+1.522
reme
Tokenreme
Feature activation+0.566
Ore
Token Ore
Feature activation+0.424
os
Tokenos
Feature activation+2.958
:
Token:
Feature activation+1.451
The
Token The
Feature activation+1.041
Related
TokenRelated
Feature activation+0.000
:
Token:
Feature activation+0.579
Why
Token Why
Feature activation+1.881
tax
Token tax
Feature activation+1.516
reform
Token reform
Feature activation+2.621
and
Token and
Feature activation+1.521
tax
Token tax
Feature activation+0.207
cuts
Token cuts
Feature activation+0.900
are
Token are
Feature activation+1.853
not
Token not
Feature activation+1.499
the
Token the
Feature activation+1.227

INTERVAL 1.111 - 1.481
CONTAINS 0.018%

MORE
TokenMORE
Feature activation+0.000
:
Token:
Feature activation+0.000
Theresa
Token Theresa
Feature activation+0.762
May
Token May
Feature activation+1.557
âĢ
TokenâĢ
Feature activation+1.464
Ļ
TokenĻ
Feature activation+1.207
s
Tokens
Feature activation+2.095
gamble
Token gamble
Feature activation+0.987
ends
Token ends
Feature activation+0.765
in
Token in
Feature activation+0.536
hung
Token hung
Feature activation+0.008
PHOTOS
TokenPHOTOS
Feature activation+0.000
:
Token:
Feature activation+0.000
35
Token 35
Feature activation+0.835
of
Token of
Feature activation+1.448
2014
Token 2014
Feature activation+1.988
's
Token's
Feature activation+1.199
Most
Token Most
Feature activation+0.135
Ant
Token Ant
Feature activation+0.329
icipated
Tokenicipated
Feature activation+0.613
Movies
Token Movies
Feature activation+2.020
:
Token:
Feature activation+0.764
Bo
Token Bo
Feature activation+0.506
X
Token X
Feature activation+0.492
il
Tokenil
Feature activation+0.000
ai
Tokenai
Feature activation+1.442
's
Token's
Feature activation+1.029
wife
Token wife
Feature activation+1.175
arrested
Token arrested
Feature activation+0.925
in
Token in
Feature activation+0.434
Brit
Token Brit
Feature activation+0.179
on
Tokenon
Feature activation+0.217
death
Token death
Feature activation+0.091
World
TokenWorld
Feature activation+0.957
Views
Token Views
Feature activation+0.875
:
Token:
Feature activation+1.305
Afghan
Token Afghan
Feature activation+2.030
villagers
Token villagers
Feature activation+2.147
recall
Token recall
Feature activation+1.433
when
Token when
Feature activation+0.740
Ber
Token Ber
Feature activation+0.554
gd
Tokengd
Feature activation+0.000
ahl
Tokenahl
Feature activation+0.431
stumbled
Token stumbled
Feature activation+0.014
Also
TokenAlso
Feature activation+0.000
Read
Token Read
Feature activation+0.000
:
Token:
Feature activation+0.073
Start
Token Start
Feature activation+2.040
ups
Token ups
Feature activation+2.425
and
Token and
Feature activation+1.179
intolerance
Token intolerance
Feature activation+1.620
can
Token can
Feature activation+2.854
't
Token't
Feature activation+1.908
go
Token go
Feature activation+0.473
together
Token together
Feature activation+0.000

INTERVAL 0.741 - 1.111
CONTAINS 0.026%

Ļ
TokenĻ
Feature activation+0.387
s
Tokens
Feature activation+0.100
monster
Token monster
Feature activation+0.226
2017
Token 2017
Feature activation+0.279
gain
Token gain
Feature activation+0.000
looks
Token looks
Feature activation+0.960
like
Token like
Feature activation+0.128
in
Token in
Feature activation+0.767
one
Token one
Feature activation+0.303
hum
Token hum
Feature activation+0.000
ong
Tokenong
Feature activation+0.000
but
Token but
Feature activation+1.200
NZ
Token NZ
Feature activation+1.438
may
Token may
Feature activation+1.029
not
Token not
Feature activation+1.008
get
Token get
Feature activation+0.920
the
Token the
Feature activation+0.990
spo
Token spo
Feature activation+0.323
ils
Tokenils
Feature activation+0.069
Ċ
TokenĊ
Feature activation+2.121
Ċ
TokenĊ
Feature activation+0.980
*
Token*
Feature activation+0.000
see
Tokensee
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
B
TokenB
Feature activation+0.307
inary
Tokeninary
Feature activation+0.576
star
Token star
Feature activation+0.881
apocalypse
Token apocalypse
Feature activation+1.143
âĢ
TokenâĢ
Feature activation+0.318
ľ
Tokenľ
Feature activation+0.154
).
Token).
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
hopes
Token hopes
Feature activation+0.460
you
Token you
Feature activation+0.568
'll
Token'll
Feature activation+0.794
wear
Token wear
Feature activation+0.794
its
Token its
Feature activation+0.594
'
Token '
Feature activation+0.781
PC
TokenPC
Feature activation+0.000
on
Token on
Feature activation+0.000
your
Token your
Feature activation+0.000
wrist
Token wrist
Feature activation+0.000
'
Token'
Feature activation+1.220
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
JUST
TokenJUST
Feature activation+0.000
WATCHED
Token WATCHED
Feature activation+0.000
Booker
Token Booker
Feature activation+0.439
on
Token on
Feature activation+1.003
GOP
Token GOP
Feature activation+0.512
health
Token health
Feature activation+0.522
strategy
Token strategy
Feature activation+0.634
:
Token:
Feature activation+0.491
'
Token '
Feature activation+0.324

INTERVAL 0.370 - 0.741
CONTAINS 0.050%

Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
[
Token[
Feature activation+0.000
Heat
TokenHeat
Feature activation+0.916
âĢ
Token âĢ
Feature activation+0.394
ĺ
Tokenĺ
Feature activation+0.575
possibly
Tokenpossibly
Feature activation+0.503
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.142
ready
Token ready
Feature activation+0.000
to
Token to
Feature activation+0.000
Climate
Token Climate
Feature activation+0.377
and
Token and
Feature activation+0.300
Ice
Token Ice
Feature activation+0.143
Sheet
Token Sheet
Feature activation+0.540
Configuration
Token Configuration
Feature activation+0.672
During
Token During
Feature activation+0.381
Early
Token Early
Feature activation+0.000
Pl
Token Pl
Feature activation+0.000
i
Tokeni
Feature activation+0.106
ocene
Tokenocene
Feature activation+0.000
Inter
Token Inter
Feature activation+0.000
ands
Tokenands
Feature activation+0.000
'
Token'
Feature activation+0.253
Found
Token Found
Feature activation+0.563
to
Token to
Feature activation+0.498
Be
Token Be
Feature activation+0.296
New
Token New
Feature activation+0.575
Species
Token Species
Feature activation+0.592
.")
Token.")
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
By
TokenBy
Feature activation+0.000
,
Token,
Feature activation+0.983
trade
Token trade
Feature activation+0.875
-
Token-
Feature activation+0.262
block
Tokenblock
Feature activation+0.767
buzz
Token buzz
Feature activation+0.895
from
Token from
Feature activation+0.625
around
Token around
Feature activation+0.656
league
Token league
Feature activation+0.362
Ċ
TokenĊ
Feature activation+0.434
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
biggest
Token biggest
Feature activation+0.842
gun
Token gun
Feature activation+0.327
problem
Token problem
Feature activation+0.942
is
Token is
Feature activation+1.074
the
Token the
Feature activation+0.759
one
Token one
Feature activation+0.590
we
Token we
Feature activation+0.000
don
Token don
Feature activation+0.000
't
Token't
Feature activation+0.000
talk
Token talk
Feature activation+0.000
about
Token about
Feature activation+0.000

INTERVAL 0.000 - 0.370
CONTAINS 99.881%

austerity
Token austerity
Feature activation+0.000
,
Token,
Feature activation+0.000
standing
Token standing
Feature activation+0.000
up
Token up
Feature activation+0.000
for
Token for
Feature activation+0.000
communities
Token communities
Feature activation+0.000
being
Token being
Feature activation+0.000
damaged
Token damaged
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
compound
Token compound
Feature activation+0.000
panel
Token panel
Feature activation+0.000
,
Token,
Feature activation+0.000
Fox
Token Fox
Feature activation+0.000
analyst
Token analyst
Feature activation+0.000
Juan
Token Juan
Feature activation+0.000
Williams
Token Williams
Feature activation+0.000
,
Token,
Feature activation+0.000
who
Token who
Feature activation+0.000
is
Token is
Feature activation+0.000
African
Token African
Feature activation+0.000
-
Token-
Feature activation+0.000
in
Token in
Feature activation+0.000
Arkansas
Token Arkansas
Feature activation+0.000
during
Token during
Feature activation+0.000
the
Token the
Feature activation+0.000
time
Token time
Feature activation+0.000
Bill
Token Bill
Feature activation+0.000
Clinton
Token Clinton
Feature activation+0.000
was
Token was
Feature activation+0.000
governor
Token governor
Feature activation+0.000
,
Token,
Feature activation+0.000
to
Token to
Feature activation+0.000
com
Tokencom
Feature activation+0.000
contributor
Token contributor
Feature activation+0.000
Laura
Token Laura
Feature activation+0.000
Miller
Token Miller
Feature activation+0.000
wrote
Token wrote
Feature activation+0.000
an
Token an
Feature activation+0.000
article
Token article
Feature activation+0.000
called
Token called
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
In
TokenIn
Feature activation+0.000
That
Token That
Feature activation+0.000
's
Token's
Feature activation+0.000
quite
Token quite
Feature activation+0.000
different
Token different
Feature activation+0.000
than
Token than
Feature activation+0.000
with
Token with
Feature activation+0.000
TPP
Token TPP
Feature activation+0.000
,
Token,
Feature activation+0.000
certainly
Token certainly
Feature activation+0.000
.
Token.
Feature activation+0.000
Oh
Token Oh
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