H7.4 top features

Top feature 0 in H7.4: (feature 12951)

TOP ACTIVATIONS
MAX = 7.870

z
Tokenz
Feature activation+0.000
og
Tokenog
Feature activation+0.000
and
Token and
Feature activation+0.077
Proxy
Token Proxy
Feature activation+0.937
Card
Token Card
Feature activation+0.382
,
Token,
Feature activation+7.870
but
Token but
Feature activation+0.000
an
Token an
Feature activation+0.000
ongoing
Token ongoing
Feature activation+0.000
relationship
Token relationship
Feature activation+0.000
between
Token between
Feature activation+0.000
to
Token to
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
be
Token be
Feature activation+0.221
accepted
Token accepted
Feature activation+2.115
,
Token,
Feature activation+7.583
but
Token but
Feature activation+0.000
welcomed
Token welcomed
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
women
Token women
Feature activation+4.935
âĢ
TokenâĢ
Feature activation+1.738
Ļ
TokenĻ
Feature activation+1.724
s
Tokens
Feature activation+1.317
health
Token health
Feature activation+3.239
,
Token,
Feature activation+7.232
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
moral
Token moral
Feature activation+0.000
and
Token and
Feature activation+0.000
just
Token just
Feature activation+0.000
people
Token people
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
of
Token of
Feature activation+0.000
Panama
Token Panama
Feature activation+3.491
,
Token,
Feature activation+6.809
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
world
Token world
Feature activation+0.000
only
Token only
Feature activation+0.000
to
Token to
Feature activation+0.016
sell
Token sell
Feature activation+0.849
Apple
Token Apple
Feature activation+0.132
devices
Token devices
Feature activation+1.786
,
Token,
Feature activation+6.656
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
to
Token to
Feature activation+0.000
entirely
Token entirely
Feature activation+0.000
eliminate
Token eliminate
Feature activation+0.000
owning
Token owning
Feature activation+0.000
not
Token not
Feature activation+0.000
just
Token just
Feature activation+0.000
search
Token search
Feature activation+0.437
capacities
Token capacities
Feature activation+4.833
,
Token,
Feature activation+6.501
video
Token video
Feature activation+0.000
and
Token and
Feature activation+0.000
maps
Token maps
Feature activation+2.245
,
Token,
Feature activation+4.498
but
Token but
Feature activation+0.000
him
Token him
Feature activation+0.765
up
Token up
Feature activation+0.529
on
Token on
Feature activation+0.000
federal
Token federal
Feature activation+0.014
charges
Token charges
Feature activation+5.066
,
Token,
Feature activation+6.457
they
Token they
Feature activation+0.000
granted
Token granted
Feature activation+0.000
immunity
Token immunity
Feature activation+0.000
to
Token to
Feature activation+0.000
his
Token his
Feature activation+0.000
to
Token to
Feature activation+0.000
people
Token people
Feature activation+0.996
with
Token with
Feature activation+0.000
bipolar
Token bipolar
Feature activation+0.000
disorder
Token disorder
Feature activation+0.580
,
Token,
Feature activation+6.381
but
Token but
Feature activation+0.000
then
Token then
Feature activation+0.000
also
Token also
Feature activation+0.000
managing
Token managing
Feature activation+0.000
to
Token to
Feature activation+0.000
demands
Token demands
Feature activation+2.360
for
Token for
Feature activation+0.000
more
Token more
Feature activation+0.000
of
Token of
Feature activation+0.000
me
Token me
Feature activation+0.143
,
Token,
Feature activation+6.345
but
Token but
Feature activation+0.069
I
Token I
Feature activation+0.000
can
Token can
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
money
Token money
Feature activation+1.289
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
suit
Token suit
Feature activation+2.703
cases
Tokencases
Feature activation+1.691
,
Token,
Feature activation+6.257
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
having
Token having
Feature activation+0.000
attempted
Token attempted
Feature activation+0.000
to
Token to
Feature activation+0.000
not
Token not
Feature activation+0.000
just
Token just
Feature activation+0.000
for
Token for
Feature activation+0.000
us
Token us
Feature activation+4.414
today
Token today
Feature activation+5.260
,
Token,
Feature activation+6.153
but
Token but
Feature activation+0.000
for
Token for
Feature activation+0.000
the
Token the
Feature activation+0.000
generations
Token generations
Feature activation+0.000
to
Token to
Feature activation+0.000
wars
Token wars
Feature activation+1.431
in
Token in
Feature activation+0.000
Afghanistan
Token Afghanistan
Feature activation+0.928
and
Token and
Feature activation+0.000
Iraq
Token Iraq
Feature activation+3.465
,
Token,
Feature activation+6.131
but
Token but
Feature activation+0.167
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
Ŀ
TokenĿ
Feature activation+0.219
between
Token between
Feature activation+0.688
all
Token all
Feature activation+0.000
the
Token the
Feature activation+0.000
screens
Token screens
Feature activation+2.153
,
Token,
Feature activation+6.069
but
Token but
Feature activation+0.122
also
Token also
Feature activation+0.000
a
Token a
Feature activation+0.000
system
Token system
Feature activation+0.000
which
Token which
Feature activation+0.000
wasn
Token wasn
Feature activation+0.016
't
Token't
Feature activation+0.031
his
Token his
Feature activation+0.019
MVP
Token MVP
Feature activation+2.152
self
Token self
Feature activation+1.996
,
Token,
Feature activation+6.052
he
Token he
Feature activation+0.000
was
Token was
Feature activation+0.000
reduced
Token reduced
Feature activation+0.000
to
Token to
Feature activation+0.000
a
Token a
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+0.000
powerful
Token powerful
Feature activation+2.285
,
Token,
Feature activation+6.049
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
syntax
Token syntax
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
in
Token in
Feature activation+0.044
our
Token our
Feature activation+0.336
economy
Token economy
Feature activation+5.768
,
Token,
Feature activation+6.014
but
Token but
Feature activation+0.000
in
Token in
Feature activation+0.000
our
Token our
Feature activation+0.000
democracy
Token democracy
Feature activation+0.000
.
Token.
Feature activation+0.000
is
Token is
Feature activation+0.000
that
Token that
Feature activation+0.000
negativity
Token negativity
Feature activation+0.075
unw
Token unw
Feature activation+0.000
arranted
Tokenarranted
Feature activation+2.724
,
Token,
Feature activation+5.995
the
Token the
Feature activation+0.000
position
Token position
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
complaint
Token complaint
Feature activation+0.000
in
Tokenin
Feature activation+0.000
on
Token on
Feature activation+0.000
the
Token the
Feature activation+0.000
Tolkien
Token Tolkien
Feature activation+0.000
themes
Token themes
Feature activation+3.607
,
Token,
Feature activation+5.903
but
Token but
Feature activation+0.000
it
Token it
Feature activation+0.000
also
Token also
Feature activation+0.000
had
Token had
Feature activation+0.000
space
Token space
Feature activation+0.000
ig
Tokenig
Feature activation+0.000
ua
Tokenua
Feature activation+0.439
and
Token and
Feature activation+0.000
Barb
Token Barb
Feature activation+0.000
uda
Tokenuda
Feature activation+3.789
,
Token,
Feature activation+5.868
but
Token but
Feature activation+0.000
the
Token the
Feature activation+0.000
other
Token other
Feature activation+0.000
countries
Token countries
Feature activation+0.000
in
Token in
Feature activation+0.000
not
Token not
Feature activation+0.000
just
Token just
Feature activation+0.000
the
Token the
Feature activation+0.000
standard
Token standard
Feature activation+0.450
Medicare
Token Medicare
Feature activation+1.437
,
Token,
Feature activation+5.803
but
Token but
Feature activation+0.257
also
Token also
Feature activation+0.000
the
Token the
Feature activation+0.000
[
Token [
Feature activation+0.000
private
Tokenprivate
Feature activation+0.000

Top DFA by src position
MAX = 8.138

an
Token an
Feature activation-0.003
Top resid features:
incredible
Token incredible
Feature activation-0.001
Top resid features:
partnership
Token partnership
Feature activation+0.001
Top resid features:
âĢĶ
Token âĢĶ
Feature activation-0.076
Top resid features:
not
Token not
Feature activation+0.231
Top resid features:
only
Token only
Feature activation+8.099
Top resid features:
between
Token between
Feature activation+0.224
Top resid features:
Mr
Token Mr
Feature activation-0.025
Top resid features:
.
Token.
Feature activation-0.014
Top resid features:
Hert
Token Hert
Feature activation+0.002
Top resid features:
z
Tokenz
Feature activation-0.001
Top resid features:
kind
Token kind
Feature activation-0.004
Top resid features:
of
Token of
Feature activation+0.005
Top resid features:
shaming
Token shaming
Feature activation+0.001
Top resid features:
to
Token to
Feature activation-0.001
Top resid features:
not
Token not
Feature activation+0.050
Top resid features:
only
Token only
Feature activation+8.138
Top resid features:
be
Token be
Feature activation+0.128
Top resid features:
accepted
Token accepted
Feature activation+0.066
Top resid features:
,
Token,
Feature activation+0.060
Top resid features:
but
Token but
Feature activation+0.000
Top resid features:
welcomed
Token welcomed
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
abortions
Token abortions
Feature activation+0.000
Top resid features:
might
Token might
Feature activation+0.003
Top resid features:
be
Token be
Feature activation-0.001
Top resid features:
not
Token not
Feature activation+0.128
Top resid features:
only
Token only
Feature activation+7.696
Top resid features:
good
Token good
Feature activation+0.053
Top resid features:
for
Token for
Feature activation+0.040
Top resid features:
women
Token women
Feature activation+0.028
Top resid features:
âĢ
TokenâĢ
Feature activation+0.041
Top resid features:
Ļ
TokenĻ
Feature activation+0.021
Top resid features:
to
Token to
Feature activation+0.003
Top resid features:
educate
Token educate
Feature activation+0.006
Top resid features:
the
Token the
Feature activation-0.002
Top resid features:
people
Token people
Feature activation+0.005
Top resid features:
not
Token not
Feature activation+0.044
Top resid features:
only
Token only
Feature activation+7.100
Top resid features:
of
Token of
Feature activation+0.395
Top resid features:
Panama
Token Panama
Feature activation+0.007
Top resid features:
,
Token,
Feature activation-0.031
Top resid features:
but
Token but
Feature activation+0.000
Top resid features:
also
Token also
Feature activation+0.000
Top resid features:
company
Token company
Feature activation-0.005
Top resid features:
unveiled
Token unveiled
Feature activation+0.011
Top resid features:
its
Token its
Feature activation-0.005
Top resid features:
plan
Token plan
Feature activation-0.009
Top resid features:
not
Token not
Feature activation+0.076
Top resid features:
only
Token only
Feature activation+6.976
Top resid features:
to
Token to
Feature activation+0.200
Top resid features:
sell
Token sell
Feature activation+0.060
Top resid features:
Apple
Token Apple
Feature activation-0.000
Top resid features:
devices
Token devices
Feature activation+0.017
Top resid features:
,
Token,
Feature activation-0.010
Top resid features:
become
Token become
Feature activation+0.005
Top resid features:
greedy
Token greedy
Feature activation-0.019
Top resid features:
about
Token about
Feature activation-0.037
Top resid features:
owning
Token owning
Feature activation+0.016
Top resid features:
not
Token not
Feature activation+0.121
Top resid features:
just
Token just
Feature activation+7.070
Top resid features:
search
Token search
Feature activation+0.021
Top resid features:
capacities
Token capacities
Feature activation+0.026
Top resid features:
,
Token,
Feature activation+0.019
Top resid features:
video
Token video
Feature activation+0.000
Top resid features:
and
Token and
Feature activation+0.000
Top resid features:
Meanwhile
Token Meanwhile
Feature activation-0.014
Top resid features:
,
Token,
Feature activation-0.001
Top resid features:
the
Token the
Feature activation-0.019
Top resid features:
feds
Token feds
Feature activation-0.039
Top resid features:
not
Token not
Feature activation+0.020
Top resid features:
only
Token only
Feature activation+7.077
Top resid features:
promised
Token promised
Feature activation+0.039
Top resid features:
not
Token not
Feature activation+0.015
Top resid features:
to
Token to
Feature activation+0.026
Top resid features:
ring
Token ring
Feature activation+0.001
Top resid features:
him
Token him
Feature activation+0.010
Top resid features:
make
Token make
Feature activation-0.002
Top resid features:
an
Token an
Feature activation+0.001
Top resid features:
apology
Token apology
Feature activation+0.009
Top resid features:
after
Token after
Feature activation-0.042
Top resid features:
not
Token not
Feature activation-0.022
Top resid features:
only
Token only
Feature activation+7.099
Top resid features:
stocking
Token stocking
Feature activation+0.012
Top resid features:
a
Token a
Feature activation+0.009
Top resid features:
product
Token product
Feature activation+0.013
Top resid features:
offensive
Token offensive
Feature activation-0.003
Top resid features:
to
Token to
Feature activation+0.002
Top resid features:
is
Token is
Feature activation-0.003
Top resid features:
that
Token that
Feature activation+0.004
Top resid features:
now
Token now
Feature activation+0.016
Top resid features:
I
Token I
Feature activation-0.021
Top resid features:
not
Token not
Feature activation+0.013
Top resid features:
only
Token only
Feature activation+6.506
Top resid features:
have
Token have
Feature activation+0.115
Top resid features:
to
Token to
Feature activation+0.156
Top resid features:
deal
Token deal
Feature activation+0.051
Top resid features:
with
Token with
Feature activation+0.049
Top resid features:
the
Token the
Feature activation+0.001
Top resid features:
You
TokenYou
Feature activation+0.000
Top resid features:
have
Token have
Feature activation+0.012
Top resid features:
been
Token been
Feature activation-0.006
Top resid features:
caught
Token caught
Feature activation-0.018
Top resid features:
not
Token not
Feature activation+0.032
Top resid features:
only
Token only
Feature activation+6.831
Top resid features:
red
Token red
Feature activation+0.009
Top resid features:
-
Token-
Feature activation+0.003
Top resid features:
handed
Tokenhanded
Feature activation+0.020
Top resid features:
with
Token with
Feature activation+0.058
Top resid features:
the
Token the
Feature activation+0.016
Top resid features:
rights
Token rights
Feature activation+0.004
Top resid features:
will
Token will
Feature activation+0.004
Top resid features:
remain
Token remain
Feature activation+0.009
Top resid features:
intact
Token intact
Feature activation+0.003
Top resid features:
not
Token not
Feature activation+0.050
Top resid features:
just
Token just
Feature activation+5.780
Top resid features:
for
Token for
Feature activation+0.933
Top resid features:
us
Token us
Feature activation+0.097
Top resid features:
today
Token today
Feature activation+0.032
Top resid features:
,
Token,
Feature activation-0.069
Top resid features:
but
Token but
Feature activation+0.000
Top resid features:
ann
Tokenann
Feature activation-0.004
Top resid features:
elled
Tokenelled
Feature activation+0.008
Top resid features:
the
Token the
Feature activation-0.013
Top resid features:
opposition
Token opposition
Feature activation-0.023
Top resid features:
not
Token not
Feature activation+0.096
Top resid features:
only
Token only
Feature activation+6.144
Top resid features:
to
Token to
Feature activation+0.349
Top resid features:
the
Token the
Feature activation+0.132
Top resid features:
wars
Token wars
Feature activation+0.044
Top resid features:
in
Token in
Feature activation+0.022
Top resid features:
Afghanistan
Token Afghanistan
Feature activation-0.001
Top resid features:
Ċ
TokenĊ
Feature activation-0.010
Top resid features:
We
TokenWe
Feature activation-0.026
Top resid features:
have
Token have
Feature activation-0.014
Top resid features:
thus
Token thus
Feature activation-0.016
Top resid features:
not
Token not
Feature activation-0.003
Top resid features:
only
Token only
Feature activation+6.527
Top resid features:
an
Token an
Feature activation+0.110
Top resid features:
âĢ
Token âĢ
Feature activation+0.003
Top resid features:
ľ
Tokenľ
Feature activation-0.001
Top resid features:
air
Tokenair
Feature activation-0.003
Top resid features:
de
Token de
Feature activation-0.014
Top resid features:
Ċ
TokenĊ
Feature activation-0.007
Top resid features:
Ċ
TokenĊ
Feature activation-0.005
Top resid features:
Rod
TokenRod
Feature activation+0.021
Top resid features:
gers
Tokengers
Feature activation-0.008
Top resid features:
not
Token not
Feature activation-0.010
Top resid features:
only
Token only
Feature activation+6.605
Top resid features:
wasn
Token wasn
Feature activation+0.124
Top resid features:
't
Token't
Feature activation+0.029
Top resid features:
his
Token his
Feature activation-0.023
Top resid features:
MVP
Token MVP
Feature activation+0.013
Top resid features:
self
Token self
Feature activation-0.030
Top resid features:
called
Token called
Feature activation-0.001
Top resid features:
Swift
Token Swift
Feature activation-0.021
Top resid features:
that
Token that
Feature activation-0.008
Top resid features:
is
Token is
Feature activation+0.020
Top resid features:
not
Token not
Feature activation+0.106
Top resid features:
only
Token only
Feature activation+6.097
Top resid features:
more
Token more
Feature activation+0.399
Top resid features:
powerful
Token powerful
Feature activation+0.049
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:
problem
Token problem
Feature activation+0.003
Top resid features:
with
Token with
Feature activation+0.007
Top resid features:
monopol
Token monopol
Feature activation+0.008
Top resid features:
ies
Tokenies
Feature activation+0.002
Top resid features:
not
Token not
Feature activation+0.045
Top resid features:
only
Token only
Feature activation+5.658
Top resid features:
in
Token in
Feature activation+1.003
Top resid features:
our
Token our
Feature activation+0.094
Top resid features:
economy
Token economy
Feature activation-0.169
Top resid features:
,
Token,
Feature activation+0.036
Top resid features:
but
Token but
Feature activation+0.000
Top resid features:
and
Token and
Feature activation+0.013
Top resid features:
spite
Token spite
Feature activation-0.000
Top resid features:
ful
Tokenful
Feature activation+0.000
Top resid features:
.
Token.
Feature activation-0.089
Top resid features:
Not
Token Not
Feature activation+0.008
Top resid features:
only
Token only
Feature activation+5.551
Top resid features:
is
Token is
Feature activation+1.184
Top resid features:
that
Token that
Feature activation+0.018
Top resid features:
negativity
Token negativity
Feature activation-0.009
Top resid features:
unw
Token unw
Feature activation+0.018
Top resid features:
arranted
Tokenarranted
Feature activation-0.010
Top resid features:
Min
Token Min
Feature activation+0.020
Top resid features:
as
Tokenas
Feature activation-0.015
Top resid features:
Morg
Token Morg
Feature activation-0.008
Top resid features:
ul
Tokenul
Feature activation-0.005
Top resid features:
not
Token not
Feature activation+0.038
Top resid features:
only
Token only
Feature activation+6.261
Top resid features:
was
Token was
Feature activation+0.329
Top resid features:
all
Token all
Feature activation+0.026
Top resid features:
-
Token-
Feature activation-0.007
Top resid features:
in
Tokenin
Feature activation-0.004
Top resid features:
on
Token on
Feature activation+0.006
Top resid features:
,
Token,
Feature activation+0.007
Top resid features:
is
Token is
Feature activation+0.002
Top resid features:
assist
Token assist
Feature activation+0.001
Top resid features:
,
Token,
Feature activation+0.002
Top resid features:
not
Token not
Feature activation+0.206
Top resid features:
only
Token only
Feature activation+6.211
Top resid features:
Ant
Token Ant
Feature activation+0.024
Top resid features:
ig
Tokenig
Feature activation-0.014
Top resid features:
ua
Tokenua
Feature activation+0.040
Top resid features:
and
Token and
Feature activation-0.005
Top resid features:
Barb
Token Barb
Feature activation+0.006
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
take
Token take
Feature activation+0.005
Top resid features:
advantage
Token advantage
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.001
Top resid features:
not
Token not
Feature activation+0.049
Top resid features:
just
Token just
Feature activation+5.332
Top resid features:
the
Token the
Feature activation+0.831
Top resid features:
standard
Token standard
Feature activation+0.217
Top resid features:
Medicare
Token Medicare
Feature activation+0.063
Top resid features:
,
Token,
Feature activation+0.005
Top resid features:
but
Token but
Feature activation+0.000
Top resid features:

Decoder Weights Distribution

Head 0: 0.06

Head 1: 0.02

Head 2: 0.02

Head 3: 0.06

Head 4: 0.63

Head 5: 0.04

Head 6: 0.02

Head 7: 0.03

Head 8: 0.04

Head 9: 0.06

Head 10: 0.01

Head 11: 0.02

Positive logits

but4.17

but4.01

But3.14

But3.04

BUT2.91

}{2.48

ancock2.39

guiIcon2.34

However2.26

})2.15

anymore2.15

However2.11

though2.11

butt2.10

orst2.07

nor2.05

%,2.05

BUT2.04

guiActive2.01

})2.00

Negative logits

0004-2.26

etheless-2.22

answered-2.13

interstitial-1.98

-1.96

nonetheless-1.92

obin-1.90

Reviewed-1.85

.''.-1.85

"))-1.84

********************************-1.82

Quote-1.79

Thu-1.79

Thu-1.78

ebin-1.78

-1.77

"}],"-1.76

summary-1.75

itialized-1.75

TAMADRA-1.74

INTERVAL 7.083 - 7.870
CONTAINS 0.000%

to
Token to
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
be
Token be
Feature activation+0.221
accepted
Token accepted
Feature activation+2.115
,
Token,
Feature activation+7.583
but
Token but
Feature activation+0.000
welcomed
Token welcomed
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
z
Tokenz
Feature activation+0.000
og
Tokenog
Feature activation+0.000
and
Token and
Feature activation+0.077
Proxy
Token Proxy
Feature activation+0.937
Card
Token Card
Feature activation+0.382
,
Token,
Feature activation+7.870
but
Token but
Feature activation+0.000
an
Token an
Feature activation+0.000
ongoing
Token ongoing
Feature activation+0.000
relationship
Token relationship
Feature activation+0.000
between
Token between
Feature activation+0.000
women
Token women
Feature activation+4.935
âĢ
TokenâĢ
Feature activation+1.738
Ļ
TokenĻ
Feature activation+1.724
s
Tokens
Feature activation+1.317
health
Token health
Feature activation+3.239
,
Token,
Feature activation+7.232
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
moral
Token moral
Feature activation+0.000
and
Token and
Feature activation+0.000
just
Token just
Feature activation+0.000

INTERVAL 6.296 - 7.083
CONTAINS 0.001%

people
Token people
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
of
Token of
Feature activation+0.000
Panama
Token Panama
Feature activation+3.491
,
Token,
Feature activation+6.809
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
world
Token world
Feature activation+0.000
demands
Token demands
Feature activation+2.360
for
Token for
Feature activation+0.000
more
Token more
Feature activation+0.000
of
Token of
Feature activation+0.000
me
Token me
Feature activation+0.143
,
Token,
Feature activation+6.345
but
Token but
Feature activation+0.069
I
Token I
Feature activation+0.000
can
Token can
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
him
Token him
Feature activation+0.765
up
Token up
Feature activation+0.529
on
Token on
Feature activation+0.000
federal
Token federal
Feature activation+0.014
charges
Token charges
Feature activation+5.066
,
Token,
Feature activation+6.457
they
Token they
Feature activation+0.000
granted
Token granted
Feature activation+0.000
immunity
Token immunity
Feature activation+0.000
to
Token to
Feature activation+0.000
his
Token his
Feature activation+0.000
to
Token to
Feature activation+0.000
people
Token people
Feature activation+0.996
with
Token with
Feature activation+0.000
bipolar
Token bipolar
Feature activation+0.000
disorder
Token disorder
Feature activation+0.580
,
Token,
Feature activation+6.381
but
Token but
Feature activation+0.000
then
Token then
Feature activation+0.000
also
Token also
Feature activation+0.000
managing
Token managing
Feature activation+0.000
to
Token to
Feature activation+0.000
owning
Token owning
Feature activation+0.000
not
Token not
Feature activation+0.000
just
Token just
Feature activation+0.000
search
Token search
Feature activation+0.437
capacities
Token capacities
Feature activation+4.833
,
Token,
Feature activation+6.501
video
Token video
Feature activation+0.000
and
Token and
Feature activation+0.000
maps
Token maps
Feature activation+2.245
,
Token,
Feature activation+4.498
but
Token but
Feature activation+0.000

INTERVAL 5.509 - 6.296
CONTAINS 0.002%

ig
Tokenig
Feature activation+0.000
ua
Tokenua
Feature activation+0.439
and
Token and
Feature activation+0.000
Barb
Token Barb
Feature activation+0.000
uda
Tokenuda
Feature activation+3.789
,
Token,
Feature activation+5.868
but
Token but
Feature activation+0.000
the
Token the
Feature activation+0.000
other
Token other
Feature activation+0.000
countries
Token countries
Feature activation+0.000
in
Token in
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+0.000
powerful
Token powerful
Feature activation+2.285
,
Token,
Feature activation+6.049
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
syntax
Token syntax
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
in
Token in
Feature activation+0.044
our
Token our
Feature activation+0.336
economy
Token economy
Feature activation+5.768
,
Token,
Feature activation+6.014
but
Token but
Feature activation+0.000
in
Token in
Feature activation+0.000
our
Token our
Feature activation+0.000
democracy
Token democracy
Feature activation+0.000
.
Token.
Feature activation+0.000
inspect
Token inspect
Feature activation+0.401
the
Token the
Feature activation+0.000
photographs
Token photographs
Feature activation+2.298
and
Token and
Feature activation+0.131
videos
Token videos
Feature activation+3.409
,
Token,
Feature activation+5.529
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
Ŀ
TokenĿ
Feature activation+0.219
between
Token between
Feature activation+0.688
all
Token all
Feature activation+0.000
the
Token the
Feature activation+0.000
screens
Token screens
Feature activation+2.153
,
Token,
Feature activation+6.069
but
Token but
Feature activation+0.122
also
Token also
Feature activation+0.000
a
Token a
Feature activation+0.000
system
Token system
Feature activation+0.000
which
Token which
Feature activation+0.000

INTERVAL 4.722 - 5.509
CONTAINS 0.002%

towards
Token towards
Feature activation+0.000
the
Token the
Feature activation+0.000
creation
Token creation
Feature activation+0.406
of
Token of
Feature activation+0.000
jobs
Token jobs
Feature activation+2.239
,
Token,
Feature activation+4.772
but
Token but
Feature activation+0.209
,
Token,
Feature activation+0.000
essentially
Token essentially
Feature activation+0.000
,
Token,
Feature activation+0.000
because
Token because
Feature activation+0.000
put
Token put
Feature activation+0.000
private
Token private
Feature activation+0.000
property
Token property
Feature activation+0.243
into
Token into
Feature activation+0.154
question
Token question
Feature activation+5.400
,
Token,
Feature activation+5.470
but
Token but
Feature activation+0.297
also
Token also
Feature activation+0.000
demonstrated
Token demonstrated
Feature activation+0.000
that
Token that
Feature activation+0.000
capitalists
Token capitalists
Feature activation+0.000
about
Token about
Feature activation+0.000
owning
Token owning
Feature activation+0.000
not
Token not
Feature activation+0.000
just
Token just
Feature activation+0.000
search
Token search
Feature activation+0.437
capacities
Token capacities
Feature activation+4.833
,
Token,
Feature activation+6.501
video
Token video
Feature activation+0.000
and
Token and
Feature activation+0.000
maps
Token maps
Feature activation+2.245
,
Token,
Feature activation+4.498
the
Token the
Feature activation+0.061
walls
Token walls
Feature activation+3.210
of
Token of
Feature activation+0.252
the
Token the
Feature activation+0.108
university
Token university
Feature activation+3.714
,
Token,
Feature activation+4.891
but
Token but
Feature activation+0.000
to
Token to
Feature activation+0.000
focus
Token focus
Feature activation+0.000
externally
Token externally
Feature activation+0.000
around
Token around
Feature activation+0.000
ineffective
Token ineffective
Feature activation+3.600
at
Token at
Feature activation+0.637
preventing
Token preventing
Feature activation+0.303
teenage
Token teenage
Feature activation+0.000
pregnancy
Token pregnancy
Feature activation+2.493
,
Token,
Feature activation+4.813
it
Token it
Feature activation+0.000
contributes
Token contributes
Feature activation+0.000
to
Token to
Feature activation+0.000
a
Token a
Feature activation+0.000
higher
Token higher
Feature activation+0.000

INTERVAL 3.935 - 4.722
CONTAINS 0.003%

phy
Tokenphy
Feature activation+0.000
x
Tokenx
Feature activation+0.000
iate
Tokeniate
Feature activation+0.463
countless
Token countless
Feature activation+0.000
people
Token people
Feature activation+1.864
,
Token,
Feature activation+4.127
make
Token make
Feature activation+0.000
travel
Token travel
Feature activation+0.000
over
Token over
Feature activation+0.000
a
Token a
Feature activation+0.000
vast
Token vast
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
U
Token U
Feature activation+0.000
.
Token.
Feature activation+0.000
S
TokenS
Feature activation+1.140
.,
Token.,
Feature activation+4.429
but
Token but
Feature activation+0.000
widely
Token widely
Feature activation+0.000
supported
Token supported
Feature activation+0.000
by
Token by
Feature activation+0.000
governments
Token governments
Feature activation+0.000
and
Token and
Feature activation+0.000
Edwin
Token Edwin
Feature activation+0.000
Enc
Token Enc
Feature activation+0.000
arn
Tokenarn
Feature activation+0.000
acion
Tokenacion
Feature activation+0.000
),
Token),
Feature activation+4.533
it
Token it
Feature activation+0.111
will
Token will
Feature activation+0.635
help
Token help
Feature activation+0.000
them
Token them
Feature activation+0.000
to
Token to
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
affects
Token affects
Feature activation+0.251
overseas
Token overseas
Feature activation+0.891
talent
Token talent
Feature activation+1.724
,
Token,
Feature activation+4.251
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+1.815
she
Token she
Feature activation+0.000
said
Token said
Feature activation+1.009
,
Token,
Feature activation+3.630
and
Token and
Feature activation+0.060
U
Token U
Feature activation+0.000
H
TokenH
Feature activation+0.000
H
Token H
Feature activation+0.000
ilo
Tokenilo
Feature activation+0.492
,
Token,
Feature activation+4.708
but
Token but
Feature activation+0.145
several
Token several
Feature activation+0.000
community
Token community
Feature activation+0.000
colleges
Token colleges
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+4.764

INTERVAL 3.148 - 3.935
CONTAINS 0.005%

't
Token't
Feature activation+0.000
only
Token only
Feature activation+0.000
loved
Token loved
Feature activation+0.607
in
Token in
Feature activation+0.000
Boston
Token Boston
Feature activation+2.601
,
Token,
Feature activation+3.525
as
Token as
Feature activation+1.119
he
Token he
Feature activation+0.000
trend
Token trend
Feature activation+0.000
ed
Tokened
Feature activation+0.000
worldwide
Token worldwide
Feature activation+0.000
was
Token was
Feature activation+0.000
not
Token not
Feature activation+0.000
just
Token just
Feature activation+0.000
grand
Token grand
Feature activation+1.615
i
Tokeni
Feature activation+0.000
ose
Tokenose
Feature activation+3.163
but
Token but
Feature activation+0.000
genuinely
Token genuinely
Feature activation+0.000
grand
Token grand
Feature activation+0.000
.
Token.
Feature activation+0.000
At
Token At
Feature activation+0.000
she
Token she
Feature activation+0.000
didn
Token didn
Feature activation+0.352
't
Token't
Feature activation+0.000
like
Token like
Feature activation+0.505
dirty
Token dirty
Feature activation+1.027
movies
Token movies
Feature activation+3.550
,
Token,
Feature activation+3.972
but
Token but
Feature activation+0.214
because
Token because
Feature activation+0.000
they
Token they
Feature activation+0.000
aroused
Token aroused
Feature activation+0.000
or
Tokenor
Feature activation+0.000
over
Token over
Feature activation+0.000
in
Tokenin
Feature activation+0.000
fl
Tokenfl
Feature activation+0.000
ated
Tokenated
Feature activation+0.000
,
Token,
Feature activation+3.641
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
meaningless
Token meaningless
Feature activation+0.000
satisfy
Token satisfy
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
the
Token the
Feature activation+0.000
story
Token story
Feature activation+2.564
buffs
Token buffs
Feature activation+3.725
,
Token,
Feature activation+5.152
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
the
Token the
Feature activation+0.000
experience
Token experience
Feature activation+0.000

INTERVAL 2.361 - 3.148
CONTAINS 0.005%

do
Token do
Feature activation+0.163
not
Token not
Feature activation+0.000
have
Token have
Feature activation+0.000
long
Token long
Feature activation+0.000
term
Token term
Feature activation+0.285
plans
Token plans
Feature activation+2.746
.
Token.
Feature activation+1.293
We
Token We
Feature activation+0.000
want
Token want
Feature activation+0.000
to
Token to
Feature activation+0.000
tell
Token tell
Feature activation+0.000
wasn
Token wasn
Feature activation+0.463
âĢ
TokenâĢ
Feature activation+0.348
Ļ
TokenĻ
Feature activation+0.370
t
Tokent
Feature activation+0.022
working
Token working
Feature activation+2.023
anymore
Token anymore
Feature activation+2.829
,
Token,
Feature activation+3.440
mind
Token mind
Feature activation+0.000
you
Token you
Feature activation+0.269
,
Token,
Feature activation+2.169
but
Token but
Feature activation+0.000
be
Token be
Feature activation+0.000
difficult
Token difficult
Feature activation+0.019
or
Token or
Feature activation+0.000
feel
Token feel
Feature activation+0.000
important
Token important
Feature activation+0.446
,
Token,
Feature activation+2.661
but
Token but
Feature activation+0.110
because
Token because
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
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
aimed
Token aimed
Feature activation+0.000
not
Token not
Feature activation+0.000
at
Token at
Feature activation+0.000
Latinos
Token Latinos
Feature activation+2.747
,
Token,
Feature activation+3.685
but
Token but
Feature activation+0.000
at
Token at
Feature activation+0.000
stupid
Token stupid
Feature activation+0.000
Republicans
Token Republicans
Feature activation+0.000
to
Token to
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
feature
Token feature
Feature activation+1.659
my
Token my
Feature activation+0.000
music
Token music
Feature activation+3.137
but
Token but
Feature activation+0.000
to
Token to
Feature activation+0.000
also
Token also
Feature activation+0.000
do
Token do
Feature activation+0.000
an
Token an
Feature activation+0.000

INTERVAL 1.574 - 2.361
CONTAINS 0.010%

âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+0.000
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
jail
Token jail
Feature activation+1.637
-
Token-
Feature activation+0.000
like
Tokenlike
Feature activation+1.938
,
Token,
Feature activation+4.139
but
Token but
Feature activation+0.000
a
Token a
Feature activation+0.000
just
Token just
Feature activation+0.000
how
Token how
Feature activation+0.000
loud
Token loud
Feature activation+0.571
the
Token the
Feature activation+0.000
noise
Token noise
Feature activation+0.016
is
Token is
Feature activation+2.168
,
Token,
Feature activation+4.709
but
Token but
Feature activation+0.000
how
Token how
Feature activation+0.000
far
Token far
Feature activation+0.000
away
Token away
Feature activation+0.000
only
Token only
Feature activation+0.000
by
Token by
Feature activation+0.000
single
Token single
Feature activation+0.000
-
Token-
Feature activation+0.000
letter
Tokenletter
Feature activation+0.000
changes
Token changes
Feature activation+1.667
in
Token in
Feature activation+0.000
their
Token their
Feature activation+0.000
genetic
Token genetic
Feature activation+0.104
codes
Token codes
Feature activation+3.856
,
Token,
Feature activation+5.797
only
Token only
Feature activation+0.000
going
Token going
Feature activation+0.278
to
Token to
Feature activation+0.252
focus
Token focus
Feature activation+0.130
on
Token on
Feature activation+0.128
immigration
Token immigration
Feature activation+2.211
reform
Token reform
Feature activation+3.107
in
Token in
Feature activation+0.278
Congress
Token Congress
Feature activation+5.794
.
Token.
Feature activation+0.745
He
Token He
Feature activation+0.779
not
Token not
Feature activation+0.000
necessarily
Token necessarily
Feature activation+0.000
because
Token because
Feature activation+0.000
of
Token of
Feature activation+0.000
Trump
Token Trump
Feature activation+0.809
,
Token,
Feature activation+1.835
but
Token but
Feature activation+0.000
that
Token that
Feature activation+0.000
Trump
Token Trump
Feature activation+0.000
has
Token has
Feature activation+0.000
accelerated
Token accelerated
Feature activation+0.000

INTERVAL 0.787 - 1.574
CONTAINS 0.018%

education
Token education
Feature activation+0.000
not
Token not
Feature activation+0.000
simply
Token simply
Feature activation+0.000
as
Token as
Feature activation+0.000
a
Token a
Feature activation+0.000
means
Token means
Feature activation+1.317
to
Token to
Feature activation+0.000
an
Token an
Feature activation+0.000
end
Token end
Feature activation+1.850
,
Token,
Feature activation+3.369
the
Token the
Feature activation+0.317
will
Token will
Feature activation+0.000
bring
Token bring
Feature activation+0.000
the
Token the
Feature activation+0.000
sides
Token sides
Feature activation+0.000
together
Token together
Feature activation+0.517
,
Token,
Feature activation+0.831
but
Token but
Feature activation+0.049
aides
Token aides
Feature activation+0.000
were
Token were
Feature activation+0.000
ordered
Token ordered
Feature activation+0.000
to
Token to
Feature activation+0.000
not
Token not
Feature activation+0.000
for
Token for
Feature activation+0.000
simply
Token simply
Feature activation+0.000
showing
Token showing
Feature activation+0.000
up
Token up
Feature activation+0.000
,
Token,
Feature activation+1.561
or
Token or
Feature activation+0.642
mind
Token mind
Feature activation+0.000
lessly
Tokenlessly
Feature activation+0.000
driving
Token driving
Feature activation+0.000
the
Token the
Feature activation+0.000
might
Token might
Feature activation+0.000
seem
Token seem
Feature activation+0.000
like
Token like
Feature activation+0.000
mindless
Token mindless
Feature activation+0.000
entertainment
Token entertainment
Feature activation+0.489
to
Token to
Feature activation+0.813
some
Token some
Feature activation+0.373
(
Token (
Feature activation+0.208
i
Tokeni
Feature activation+0.000
.
Token.
Feature activation+0.000
e
Tokene
Feature activation+0.000
the
Token the
Feature activation+0.000
company
Token company
Feature activation+0.000
over
Token over
Feature activation+0.000
a
Token a
Feature activation+0.000
cliff
Token cliff
Feature activation+0.000
,
Token,
Feature activation+0.900
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
said
Token said
Feature activation+0.000
Dave
Token Dave
Feature activation+0.000
Blank
Token Blank
Feature activation+0.000

INTERVAL 0.000 - 0.787
CONTAINS 99.953%

and
Token and
Feature activation+0.000
get
Token get
Feature activation+0.000
as
Token as
Feature activation+0.000
much
Token much
Feature activation+0.000
information
Token information
Feature activation+0.046
as
Token as
Feature activation+0.000
they
Token they
Feature activation+0.000
can
Token can
Feature activation+0.000
about
Token about
Feature activation+0.000
them
Token them
Feature activation+0.000
.
Token.
Feature activation+0.000
should
Token should
Feature activation+0.000
know
Token know
Feature activation+0.000
the
Token the
Feature activation+0.000
device
Token device
Feature activation+0.000
outper
Token outper
Feature activation+0.000
formed
Tokenformed
Feature activation+0.000
the
Token the
Feature activation+0.000
Ex
Token Ex
Feature activation+0.000
yn
Tokenyn
Feature activation+0.000
os
Tokenos
Feature activation+0.000
-
Token-
Feature activation+0.000
body
Token body
Feature activation+0.000
and
Token and
Feature activation+0.000
head
Token head
Feature activation+0.000
are
Token are
Feature activation+0.000
diagonal
Token diagonal
Feature activation+0.000
,
Token,
Feature activation+0.000
not
Token not
Feature activation+0.000
vertical
Token vertical
Feature activation+0.000
and
Token and
Feature activation+0.000
not
Token not
Feature activation+0.000
horizontal
Token horizontal
Feature activation+0.000
accelerated
Token accelerated
Feature activation+0.000
the
Token the
Feature activation+0.000
pace
Token pace
Feature activation+0.000
.
Token.
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
Democratic
Token Democratic
Feature activation+0.000
Party
Token Party
Feature activation+0.000
here
Token here
Feature activation+0.000
was
Token was
Feature activation+0.000
hypothesis
Token hypothesis
Feature activation+0.000
would
Token would
Feature activation+0.000
be
Token be
Feature activation+0.000
that
Token that
Feature activation+0.000
aggress
Token aggress
Feature activation+0.000
iveness
Tokeniveness
Feature activation+0.000
as
Token as
Feature activation+0.000
well
Token well
Feature activation+0.000
as
Token as
Feature activation+0.000
bad
Token bad
Feature activation+0.000
mood
Token mood
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 H7.4: (feature 4312)

TOP ACTIVATIONS
MAX = 4.040

would
Token would
Feature activation+0.000
make
Token make
Feature activation+0.000
getting
Token getting
Feature activation+0.000
at
Token at
Feature activation+0.233
customer
Token customer
Feature activation+0.319
data
Token data
Feature activation+4.040
possible
Token possible
Feature activation+0.000
."
Token."
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
A
TokenA
Feature activation+0.000
take
Token take
Feature activation+0.000
the
Token the
Feature activation+0.000
security
Token security
Feature activation+0.000
of
Token of
Feature activation+0.000
their
Token their
Feature activation+0.000
data
Token data
Feature activation+3.787
very
Token very
Feature activation+0.000
seriously
Token seriously
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
to
Token to
Feature activation+0.000
support
Token support
Feature activation+0.152
Aram
Token Aram
Feature activation+0.650
is
Tokenis
Feature activation+3.236
Ay
Token Ay
Feature activation+0.000
ala
Tokenala
Feature activation+3.773
based
Token based
Feature activation+0.000
on
Token on
Feature activation+0.285
her
Token her
Feature activation+0.542
agreeing
Token agreeing
Feature activation+0.292
to
Token to
Feature activation+0.276
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
macro
Token macro
Feature activation+0.571
-
Token-
Feature activation+0.000
economic
Tokeneconomic
Feature activation+0.144
impact
Token impact
Feature activation+3.745
hard
Token hard
Feature activation+0.000
to
Token to
Feature activation+0.000
gauge
Token gauge
Feature activation+2.624
-
Token -
Feature activation+0.234
ranging
Token ranging
Feature activation+0.000
,
Token,
Feature activation+0.460
installing
Token installing
Feature activation+0.000
,
Token,
Feature activation+0.033
and
Token and
Feature activation+0.000
activating
Token activating
Feature activation+0.790
plugins
Token plugins
Feature activation+3.730
simple
Token simple
Feature activation+0.000
-
Token-
Feature activation+0.000
ish
Tokenish
Feature activation+0.317
,
Token,
Feature activation+0.000
but
Token but
Feature activation+0.000
ities
Tokenities
Feature activation+0.000
that
Token that
Feature activation+0.000
made
Token made
Feature activation+0.000
using
Token using
Feature activation+0.000
Mark
Token Mark
Feature activation+0.550
down
Tokendown
Feature activation+3.662
so
Token so
Feature activation+0.497
difficult
Token difficult
Feature activation+0.000
.
Token.
Feature activation+0.000
Did
Token Did
Feature activation+0.000
we
Token we
Feature activation+0.000
make
Token make
Feature activation+0.000
choosing
Token choosing
Feature activation+0.000
exactly
Token exactly
Feature activation+0.258
the
Token the
Feature activation+0.992
right
Token right
Feature activation+1.318
word
Token word
Feature activation+3.612
seem
Token seem
Feature activation+0.000
so
Token so
Feature activation+0.420
effort
Token effort
Feature activation+0.000
less
Tokenless
Feature activation+0.420
.
Token.
Feature activation+0.000
made
Token made
Feature activation+0.000
support
Token support
Feature activation+0.000
for
Token for
Feature activation+0.000
ending
Token ending
Feature activation+0.000
cannabis
Token cannabis
Feature activation+1.894
prohibition
Token prohibition
Feature activation+3.557
center
Token center
Feature activation+0.000
pieces
Tokenpieces
Feature activation+0.969
of
Token of
Feature activation+0.000
their
Token their
Feature activation+0.000
campaigns
Token campaigns
Feature activation+0.998
making
Token making
Feature activation+0.000
any
Token any
Feature activation+0.000
extension
Token extension
Feature activation+1.151
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
line
Token line
Feature activation+3.454
a
Token a
Feature activation+0.237
risky
Token risky
Feature activation+0.000
proposition
Token proposition
Feature activation+0.071
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
is
Token is
Feature activation+0.000
our
Token our
Feature activation+0.000
reliance
Token reliance
Feature activation+0.000
on
Token on
Feature activation+0.000
search
Token search
Feature activation+1.221
engines
Token engines
Feature activation+3.448
making
Token making
Feature activation+0.335
it
Token it
Feature activation+0.034
harder
Token harder
Feature activation+0.493
for
Token for
Feature activation+0.270
us
Token us
Feature activation+0.233
Se
TokenSe
Feature activation+0.000
am
Tokenam
Feature activation+0.000
makes
Token makes
Feature activation+0.000
creating
Token creating
Feature activation+0.000
themed
Token themed
Feature activation+0.012
towels
Token towels
Feature activation+3.433
an
Token an
Feature activation+0.000
easy
Token easy
Feature activation+0.000
no
Token no
Feature activation+0.000
-
Token-
Feature activation+0.000
se
Tokense
Feature activation+0.000
his
Token his
Feature activation+0.000
decision
Token decision
Feature activation+0.000
to
Token to
Feature activation+0.000
support
Token support
Feature activation+0.152
Aram
Token Aram
Feature activation+0.650
is
Tokenis
Feature activation+3.236
Ay
Token Ay
Feature activation+0.000
ala
Tokenala
Feature activation+3.773
based
Token based
Feature activation+0.000
on
Token on
Feature activation+0.285
her
Token her
Feature activation+0.542
M
TokenM
Feature activation+0.000
akes
Tokenakes
Feature activation+0.000
G
Token G
Feature activation+0.000
aining
Tokenaining
Feature activation+0.000
Popular
Token Popular
Feature activation+0.000
ity
Tokenity
Feature activation+3.174
Over
Token Over
Feature activation+0.934
Time
Token Time
Feature activation+0.364
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
on
Token on
Feature activation+0.000
and
Token and
Feature activation+0.000
they
Token they
Feature activation+0.000
make
Token make
Feature activation+0.000
getting
Token getting
Feature activation+0.000
smashed
Token smashed
Feature activation+3.094
more
Token more
Feature activation+0.023
fun
Token fun
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
.
Token.
Feature activation+0.000
finance
Token finance
Feature activation+0.084
ruling
Token ruling
Feature activation+1.368
,
Token,
Feature activation+0.261
Citizens
Token Citizens
Feature activation+0.015
United
Token United
Feature activation+1.674
,
Token,
Feature activation+3.051
seem
Token seem
Feature activation+1.454
like
Token like
Feature activation+0.430
a
Token a
Feature activation+0.000
tr
Token tr
Feature activation+0.000
ifle
Tokenifle
Feature activation+0.000
things
Token things
Feature activation+0.000
designed
Token designed
Feature activation+0.000
to
Token to
Feature activation+0.000
make
Token make
Feature activation+0.000
buying
Token buying
Feature activation+0.000
electronics
Token electronics
Feature activation+3.001
more
Token more
Feature activation+0.142
fun
Token fun
Feature activation+0.149
.
Token.
Feature activation+0.000
There
Token There
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
-
Token-
Feature activation+0.000
brother
Tokenbrother
Feature activation+0.000
Prince
Token Prince
Feature activation+0.000
Mu
Token Mu
Feature activation+0.144
q
Tokenq
Feature activation+0.000
rin
Tokenrin
Feature activation+2.978
the
Token the
Feature activation+0.000
deputy
Token deputy
Feature activation+0.000
prime
Token prime
Feature activation+0.000
minister
Token minister
Feature activation+0.000
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
charges
Token charges
Feature activation+0.000
against
Token against
Feature activation+0.000
Mr
Token Mr
Feature activation+0.000
.
Token.
Feature activation+0.000
Anderson
Token Anderson
Feature activation+2.908
dismissed
Token dismissed
Feature activation+0.000
,
Token,
Feature activation+0.000
acknowledging
Token acknowledging
Feature activation+0.000
that
Token that
Feature activation+0.000
the
Token the
Feature activation+0.000
to
Token to
Feature activation+0.296
AFC
Token AFC
Feature activation+0.480
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.210
s
Tokens
Feature activation+0.699
plans
Token plans
Feature activation+2.847
clear
Token clear
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Sam
TokenSam
Feature activation+0.000
goddamn
Token goddamn
Feature activation+0.000
good
Token good
Feature activation+0.000
makes
Token makes
Feature activation+0.000
watching
Token watching
Feature activation+0.000
this
Token this
Feature activation+0.982
band
Token band
Feature activation+2.833
an
Token an
Feature activation+0.000
unbelievably
Token unbelievably
Feature activation+0.000
fun
Token fun
Feature activation+0.000
time
Token time
Feature activation+0.035
.
Token.
Feature activation+0.000

Top DFA by src position
MAX = 4.585

to
Token to
Feature activation-0.008
Top resid features:
say
Token say
Feature activation-0.004
Top resid features:
it
Token it
Feature activation-0.001
Top resid features:
would
Token would
Feature activation-0.014
Top resid features:
make
Token make
Feature activation-0.003
Top resid features:
getting
Token getting
Feature activation+4.585
Top resid features:
at
Token at
Feature activation+0.027
Top resid features:
customer
Token customer
Feature activation-0.014
Top resid features:
data
Token data
Feature activation+0.008
Top resid features:
possible
Token possible
Feature activation+0.000
Top resid features:
."
Token."
Feature activation+0.000
Top resid features:
that
Token that
Feature activation-0.008
Top resid features:
we
Token we
Feature activation-0.013
Top resid features:
take
Token take
Feature activation-0.038
Top resid features:
the
Token the
Feature activation+0.059
Top resid features:
security
Token security
Feature activation+0.411
Top resid features:
of
Token of
Feature activation+3.673
Top resid features:
their
Token their
Feature activation+0.190
Top resid features:
data
Token data
Feature activation+0.028
Top resid features:
very
Token very
Feature activation+0.000
Top resid features:
seriously
Token seriously
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
?
Token?
Feature activation-0.003
Top resid features:
Ċ
TokenĊ
Feature activation-0.001
Top resid features:
Ċ
TokenĊ
Feature activation-0.009
Top resid features:
Was
TokenWas
Feature activation+0.002
Top resid features:
his
Token his
Feature activation+0.318
Top resid features:
decision
Token decision
Feature activation+2.495
Top resid features:
to
Token to
Feature activation+1.294
Top resid features:
support
Token support
Feature activation+0.273
Top resid features:
Aram
Token Aram
Feature activation+0.029
Top resid features:
is
Tokenis
Feature activation+0.006
Top resid features:
Ay
Token Ay
Feature activation+0.005
Top resid features:
Sche
Token Sche
Feature activation+0.002
Top resid features:
ng
Tokenng
Feature activation-0.000
Top resid features:
en
Tokenen
Feature activation-0.002
Top resid features:
closure
Token closure
Feature activation+0.013
Top resid features:
makes
Token makes
Feature activation+0.058
Top resid features:
estimates
Token estimates
Feature activation+2.944
Top resid features:
of
Token of
Feature activation+0.920
Top resid features:
the
Token the
Feature activation+0.296
Top resid features:
macro
Token macro
Feature activation+0.020
Top resid features:
-
Token-
Feature activation+0.004
Top resid features:
economic
Tokeneconomic
Feature activation+0.023
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.148
Top resid features:
PA
TokenPA
Feature activation-0.006
Top resid features:
makes
Token makes
Feature activation-0.022
Top resid features:
the
Token the
Feature activation+0.012
Top resid features:
process
Token process
Feature activation+1.347
Top resid features:
of
Token of
Feature activation+2.630
Top resid features:
downloading
Token downloading
Feature activation+0.188
Top resid features:
,
Token,
Feature activation+0.008
Top resid features:
installing
Token installing
Feature activation+0.009
Top resid features:
,
Token,
Feature activation+0.001
Top resid features:
and
Token and
Feature activation+0.028
Top resid features:
ambig
Token ambig
Feature activation+0.004
Top resid features:
u
Tokenu
Feature activation+0.000
Top resid features:
ities
Tokenities
Feature activation+0.001
Top resid features:
that
Token that
Feature activation-0.080
Top resid features:
made
Token made
Feature activation+0.030
Top resid features:
using
Token using
Feature activation+4.230
Top resid features:
Mark
Token Mark
Feature activation+0.002
Top resid features:
down
Tokendown
Feature activation-0.008
Top resid features:
so
Token so
Feature activation+0.000
Top resid features:
difficult
Token difficult
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.011
Top resid features:
and
Token and
Feature activation+0.005
Top resid features:
how
Token how
Feature activation-0.028
Top resid features:
you
Token you
Feature activation-0.001
Top resid features:
make
Token make
Feature activation-0.038
Top resid features:
choosing
Token choosing
Feature activation+3.625
Top resid features:
exactly
Token exactly
Feature activation+0.294
Top resid features:
the
Token the
Feature activation+0.236
Top resid features:
right
Token right
Feature activation+0.063
Top resid features:
word
Token word
Feature activation+0.031
Top resid features:
seem
Token seem
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.001
Top resid features:
whom
Token whom
Feature activation-0.017
Top resid features:
have
Token have
Feature activation-0.014
Top resid features:
made
Token made
Feature activation+0.001
Top resid features:
support
Token support
Feature activation+1.677
Top resid features:
for
Token for
Feature activation+1.767
Top resid features:
ending
Token ending
Feature activation+0.653
Top resid features:
cannabis
Token cannabis
Feature activation+0.010
Top resid features:
prohibition
Token prohibition
Feature activation+0.016
Top resid features:
center
Token center
Feature activation+0.000
Top resid features:
pieces
Tokenpieces
Feature activation+0.000
Top resid features:
by
Token by
Feature activation+0.006
Top resid features:
20
Token 20
Feature activation+0.003
Top resid features:
31
Token31
Feature activation-0.006
Top resid features:
,
Token,
Feature activation-0.006
Top resid features:
making
Token making
Feature activation+0.119
Top resid features:
any
Token any
Feature activation+2.080
Top resid features:
extension
Token extension
Feature activation+0.883
Top resid features:
of
Token of
Feature activation+0.919
Top resid features:
the
Token the
Feature activation+0.093
Top resid features:
line
Token line
Feature activation-0.022
Top resid features:
a
Token a
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.002
Top resid features:
Ċ
TokenĊ
Feature activation-0.010
Top resid features:
But
TokenBut
Feature activation-0.005
Top resid features:
is
Token is
Feature activation+0.018
Top resid features:
our
Token our
Feature activation+0.142
Top resid features:
reliance
Token reliance
Feature activation+2.446
Top resid features:
on
Token on
Feature activation+1.283
Top resid features:
search
Token search
Feature activation+0.007
Top resid features:
engines
Token engines
Feature activation+0.038
Top resid features:
making
Token making
Feature activation+0.000
Top resid features:
it
Token it
Feature activation+0.000
Top resid features:
A
TokenA
Feature activation-0.007
Top resid features:
-
Token-
Feature activation+0.001
Top resid features:
Se
TokenSe
Feature activation-0.006
Top resid features:
am
Tokenam
Feature activation-0.012
Top resid features:
makes
Token makes
Feature activation-0.080
Top resid features:
creating
Token creating
Feature activation+3.798
Top resid features:
themed
Token themed
Feature activation+0.201
Top resid features:
towels
Token towels
Feature activation+0.029
Top resid features:
an
Token an
Feature activation+0.000
Top resid features:
easy
Token easy
Feature activation+0.000
Top resid features:
no
Token no
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation-0.002
Top resid features:
Ċ
TokenĊ
Feature activation-0.008
Top resid features:
Was
TokenWas
Feature activation-0.005
Top resid features:
his
Token his
Feature activation+0.201
Top resid features:
decision
Token decision
Feature activation+1.556
Top resid features:
to
Token to
Feature activation+1.751
Top resid features:
support
Token support
Feature activation+0.254
Top resid features:
Aram
Token Aram
Feature activation+0.082
Top resid features:
is
Tokenis
Feature activation-0.001
Top resid features:
Ay
Token Ay
Feature activation+0.000
Top resid features:
ala
Tokenala
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation-0.001
Top resid features:
Ċ
TokenĊ
Feature activation-0.016
Top resid features:
M
TokenM
Feature activation-0.011
Top resid features:
akes
Tokenakes
Feature activation-0.017
Top resid features:
G
Token G
Feature activation+0.095
Top resid features:
aining
Tokenaining
Feature activation+3.543
Top resid features:
Popular
Token Popular
Feature activation+0.021
Top resid features:
ity
Tokenity
Feature activation+0.043
Top resid features:
Over
Token Over
Feature activation+0.000
Top resid features:
Time
Token Time
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
s
Tokens
Feature activation+0.002
Top resid features:
on
Token on
Feature activation+0.012
Top resid features:
and
Token and
Feature activation-0.004
Top resid features:
they
Token they
Feature activation-0.017
Top resid features:
make
Token make
Feature activation-0.011
Top resid features:
getting
Token getting
Feature activation+3.588
Top resid features:
smashed
Token smashed
Feature activation+0.035
Top resid features:
more
Token more
Feature activation+0.000
Top resid features:
fun
Token fun
Feature activation+0.000
Top resid features:
âĢ
TokenâĢ
Feature activation+0.000
Top resid features:
Ŀ
TokenĿ
Feature activation+0.000
Top resid features:
's
Token's
Feature activation+0.070
Top resid features:
landmark
Token landmark
Feature activation+0.190
Top resid features:
campaign
Token campaign
Feature activation+0.040
Top resid features:
finance
Token finance
Feature activation+0.009
Top resid features:
ruling
Token ruling
Feature activation+0.485
Top resid features:
,
Token,
Feature activation+2.649
Top resid features:
Citizens
Token Citizens
Feature activation+0.008
Top resid features:
United
Token United
Feature activation+0.013
Top resid features:
,
Token,
Feature activation+0.007
Top resid features:
seem
Token seem
Feature activation+0.000
Top resid features:
like
Token like
Feature activation+0.000
Top resid features:
other
Token other
Feature activation-0.001
Top resid features:
things
Token things
Feature activation+0.001
Top resid features:
designed
Token designed
Feature activation+0.005
Top resid features:
to
Token to
Feature activation-0.020
Top resid features:
make
Token make
Feature activation+0.051
Top resid features:
buying
Token buying
Feature activation+3.442
Top resid features:
electronics
Token electronics
Feature activation+0.088
Top resid features:
more
Token more
Feature activation+0.000
Top resid features:
fun
Token fun
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
There
Token There
Feature activation+0.000
Top resid features:
he
Token he
Feature activation-0.028
Top resid features:
made
Token made
Feature activation-0.085
Top resid features:
his
Token his
Feature activation+0.028
Top resid features:
half
Token half
Feature activation+0.084
Top resid features:
-
Token-
Feature activation+0.007
Top resid features:
brother
Tokenbrother
Feature activation+3.059
Top resid features:
Prince
Token Prince
Feature activation+0.529
Top resid features:
Mu
Token Mu
Feature activation+0.011
Top resid features:
q
Tokenq
Feature activation+0.001
Top resid features:
rin
Tokenrin
Feature activation-0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
moved
Token moved
Feature activation-0.006
Top resid features:
to
Token to
Feature activation+0.009
Top resid features:
have
Token have
Feature activation-0.044
Top resid features:
the
Token the
Feature activation+0.122
Top resid features:
charges
Token charges
Feature activation+0.378
Top resid features:
against
Token against
Feature activation+3.213
Top resid features:
Mr
Token Mr
Feature activation-0.009
Top resid features:
.
Token.
Feature activation+0.008
Top resid features:
Anderson
Token Anderson
Feature activation+0.008
Top resid features:
dismissed
Token dismissed
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
residents
Token residents
Feature activation-0.002
Top resid features:
have
Token have
Feature activation-0.049
Top resid features:
made
Token made
Feature activation-0.028
Top resid features:
their
Token their
Feature activation+0.069
Top resid features:
objections
Token objections
Feature activation+1.384
Top resid features:
to
Token to
Feature activation+2.015
Top resid features:
AFC
Token AFC
Feature activation+0.007
Top resid features:
âĢ
TokenâĢ
Feature activation+0.004
Top resid features:
Ļ
TokenĻ
Feature activation+0.019
Top resid features:
s
Tokens
Feature activation+0.033
Top resid features:
plans
Token plans
Feature activation+0.002
Top resid features:
is
Token is
Feature activation+0.005
Top resid features:
so
Token so
Feature activation-0.007
Top resid features:
goddamn
Token goddamn
Feature activation+0.008
Top resid features:
good
Token good
Feature activation+0.005
Top resid features:
makes
Token makes
Feature activation+0.034
Top resid features:
watching
Token watching
Feature activation+2.951
Top resid features:
this
Token this
Feature activation+0.288
Top resid features:
band
Token band
Feature activation+0.036
Top resid features:
an
Token an
Feature activation+0.000
Top resid features:
unbelievably
Token unbelievably
Feature activation+0.000
Top resid features:
fun
Token fun
Feature activation+0.000
Top resid features:

Decoder Weights Distribution

Head 0: 0.07

Head 1: 0.02

Head 2: 0.02

Head 3: 0.07

Head 4: 0.57

Head 5: 0.03

Head 6: 0.02

Head 7: 0.02

Head 8: 0.07

Head 9: 0.08

Head 10: 0.01

Head 11: 0.02

Positive logits

obsolete3.00

solete2.60

alive2.54

moot2.47

afloat2.46

ivation2.45

ドラ2.40

coincide2.34

squarely2.33

hostage2.28

ibel2.25

irrelevant2.21

bitters2.21

appealing2.19

seriously2.18

ウス2.18

uzz2.14

ueless2.14

izzle2.13

ITCH2.11

Negative logits

collects-2.39

learns-2.30

Shares-2.22

amy-2.20

eh-2.20

gathers-2.19

Emin-2.18

cknowled-2.14

MN-2.08

intercept-2.06

surrounds-2.06

discovers-2.05

wd-2.04

sleeps-2.00

ichita-1.99

Regulatory-1.99

axes-1.96

frag-1.94

Carlson-1.93

hugs-1.92

INTERVAL 3.636 - 4.040
CONTAINS 0.001%

would
Token would
Feature activation+0.000
make
Token make
Feature activation+0.000
getting
Token getting
Feature activation+0.000
at
Token at
Feature activation+0.233
customer
Token customer
Feature activation+0.319
data
Token data
Feature activation+4.040
possible
Token possible
Feature activation+0.000
."
Token."
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
A
TokenA
Feature activation+0.000
take
Token take
Feature activation+0.000
the
Token the
Feature activation+0.000
security
Token security
Feature activation+0.000
of
Token of
Feature activation+0.000
their
Token their
Feature activation+0.000
data
Token data
Feature activation+3.787
very
Token very
Feature activation+0.000
seriously
Token seriously
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
macro
Token macro
Feature activation+0.571
-
Token-
Feature activation+0.000
economic
Tokeneconomic
Feature activation+0.144
impact
Token impact
Feature activation+3.745
hard
Token hard
Feature activation+0.000
to
Token to
Feature activation+0.000
gauge
Token gauge
Feature activation+2.624
-
Token -
Feature activation+0.234
ranging
Token ranging
Feature activation+0.000
,
Token,
Feature activation+0.460
installing
Token installing
Feature activation+0.000
,
Token,
Feature activation+0.033
and
Token and
Feature activation+0.000
activating
Token activating
Feature activation+0.790
plugins
Token plugins
Feature activation+3.730
simple
Token simple
Feature activation+0.000
-
Token-
Feature activation+0.000
ish
Tokenish
Feature activation+0.317
,
Token,
Feature activation+0.000
but
Token but
Feature activation+0.000
to
Token to
Feature activation+0.000
support
Token support
Feature activation+0.152
Aram
Token Aram
Feature activation+0.650
is
Tokenis
Feature activation+3.236
Ay
Token Ay
Feature activation+0.000
ala
Tokenala
Feature activation+3.773
based
Token based
Feature activation+0.000
on
Token on
Feature activation+0.285
her
Token her
Feature activation+0.542
agreeing
Token agreeing
Feature activation+0.292
to
Token to
Feature activation+0.276

INTERVAL 3.232 - 3.636
CONTAINS 0.001%

make
Token make
Feature activation+0.000
choosing
Token choosing
Feature activation+0.000
exactly
Token exactly
Feature activation+0.258
the
Token the
Feature activation+0.992
right
Token right
Feature activation+1.318
word
Token word
Feature activation+3.612
seem
Token seem
Feature activation+0.000
so
Token so
Feature activation+0.420
effort
Token effort
Feature activation+0.000
less
Tokenless
Feature activation+0.420
.
Token.
Feature activation+0.000
Se
TokenSe
Feature activation+0.000
am
Tokenam
Feature activation+0.000
makes
Token makes
Feature activation+0.000
creating
Token creating
Feature activation+0.000
themed
Token themed
Feature activation+0.012
towels
Token towels
Feature activation+3.433
an
Token an
Feature activation+0.000
easy
Token easy
Feature activation+0.000
no
Token no
Feature activation+0.000
-
Token-
Feature activation+0.000
se
Tokense
Feature activation+0.000
made
Token made
Feature activation+0.000
support
Token support
Feature activation+0.000
for
Token for
Feature activation+0.000
ending
Token ending
Feature activation+0.000
cannabis
Token cannabis
Feature activation+1.894
prohibition
Token prohibition
Feature activation+3.557
center
Token center
Feature activation+0.000
pieces
Tokenpieces
Feature activation+0.969
of
Token of
Feature activation+0.000
their
Token their
Feature activation+0.000
campaigns
Token campaigns
Feature activation+0.998
his
Token his
Feature activation+0.000
decision
Token decision
Feature activation+0.000
to
Token to
Feature activation+0.000
support
Token support
Feature activation+0.152
Aram
Token Aram
Feature activation+0.650
is
Tokenis
Feature activation+3.236
Ay
Token Ay
Feature activation+0.000
ala
Tokenala
Feature activation+3.773
based
Token based
Feature activation+0.000
on
Token on
Feature activation+0.285
her
Token her
Feature activation+0.542
making
Token making
Feature activation+0.000
any
Token any
Feature activation+0.000
extension
Token extension
Feature activation+1.151
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
line
Token line
Feature activation+3.454
a
Token a
Feature activation+0.237
risky
Token risky
Feature activation+0.000
proposition
Token proposition
Feature activation+0.071
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000

INTERVAL 2.828 - 3.232
CONTAINS 0.001%

goddamn
Token goddamn
Feature activation+0.000
good
Token good
Feature activation+0.000
makes
Token makes
Feature activation+0.000
watching
Token watching
Feature activation+0.000
this
Token this
Feature activation+0.982
band
Token band
Feature activation+2.833
an
Token an
Feature activation+0.000
unbelievably
Token unbelievably
Feature activation+0.000
fun
Token fun
Feature activation+0.000
time
Token time
Feature activation+0.035
.
Token.
Feature activation+0.000
-
Token-
Feature activation+0.000
brother
Tokenbrother
Feature activation+0.000
Prince
Token Prince
Feature activation+0.000
Mu
Token Mu
Feature activation+0.144
q
Tokenq
Feature activation+0.000
rin
Tokenrin
Feature activation+2.978
the
Token the
Feature activation+0.000
deputy
Token deputy
Feature activation+0.000
prime
Token prime
Feature activation+0.000
minister
Token minister
Feature activation+0.000
,
Token,
Feature activation+0.000
on
Token on
Feature activation+0.000
and
Token and
Feature activation+0.000
they
Token they
Feature activation+0.000
make
Token make
Feature activation+0.000
getting
Token getting
Feature activation+0.000
smashed
Token smashed
Feature activation+3.094
more
Token more
Feature activation+0.023
fun
Token fun
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
.
Token.
Feature activation+0.000
things
Token things
Feature activation+0.000
designed
Token designed
Feature activation+0.000
to
Token to
Feature activation+0.000
make
Token make
Feature activation+0.000
buying
Token buying
Feature activation+0.000
electronics
Token electronics
Feature activation+3.001
more
Token more
Feature activation+0.142
fun
Token fun
Feature activation+0.149
.
Token.
Feature activation+0.000
There
Token There
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
the
Token the
Feature activation+0.000
charges
Token charges
Feature activation+0.000
against
Token against
Feature activation+0.000
Mr
Token Mr
Feature activation+0.000
.
Token.
Feature activation+0.000
Anderson
Token Anderson
Feature activation+2.908
dismissed
Token dismissed
Feature activation+0.000
,
Token,
Feature activation+0.000
acknowledging
Token acknowledging
Feature activation+0.000
that
Token that
Feature activation+0.000
the
Token the
Feature activation+0.000

INTERVAL 2.424 - 2.828
CONTAINS 0.001%

,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
makes
Token makes
Feature activation+0.000
playing
Token playing
Feature activation+0.000
every
Token every
Feature activation+0.063
class
Token class
Feature activation+2.809
an
Token an
Feature activation+0.000
absolute
Token absolute
Feature activation+0.000
blast
Token blast
Feature activation+0.000
.
Token.
Feature activation+0.000
Don
Token Don
Feature activation+0.000
-
Token-
Feature activation+0.000
economic
Tokeneconomic
Feature activation+0.144
impact
Token impact
Feature activation+3.745
hard
Token hard
Feature activation+0.000
to
Token to
Feature activation+0.000
gauge
Token gauge
Feature activation+2.624
-
Token -
Feature activation+0.234
ranging
Token ranging
Feature activation+0.000
from
Token from
Feature activation+0.000
an
Token an
Feature activation+0.000
immediate
Token immediate
Feature activation+0.000
that
Token that
Feature activation+0.000
legislators
Token legislators
Feature activation+0.000
make
Token make
Feature activation+0.000
passing
Token passing
Feature activation+0.000
a
Token a
Feature activation+0.000
bill
Token bill
Feature activation+2.521
banning
Token banning
Feature activation+0.046
F
Token F
Feature activation+0.000
GM
TokenGM
Feature activation+2.743
/
Token/
Feature activation+0.000
C
TokenC
Feature activation+2.886
about
Token about
Feature activation+0.000
whether
Token whether
Feature activation+0.035
or
Token or
Feature activation+0.000
not
Token not
Feature activation+0.076
to
Token to
Feature activation+0.036
prosecute
Token prosecute
Feature activation+2.752
alleged
Token alleged
Feature activation+0.162
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
hate
Tokenhate
Feature activation+0.054
crimes
Token crimes
Feature activation+2.337
of
Token of
Feature activation+0.000
choosing
Token choosing
Feature activation+0.073
man
Token man
Feature activation+1.312
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
match
Token match
Feature activation+2.529
a
Token a
Feature activation+0.188
lot
Token lot
Feature activation+0.137
more
Token more
Feature activation+0.136
difficult
Token difficult
Feature activation+0.000
,
Token,
Feature activation+0.000

INTERVAL 2.020 - 2.424
CONTAINS 0.003%

sustainable
Token sustainable
Feature activation+0.000
growth
Token growth
Feature activation+0.361
and
Token and
Feature activation+0.000
the
Token the
Feature activation+0.000
welfare
Token welfare
Feature activation+0.686
state
Token state
Feature activation+2.344
incompatible
Token incompatible
Feature activation+0.000
?
Token?
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Since
TokenSince
Feature activation+0.000
and
Token and
Feature activation+0.076
live
Token live
Feature activation+0.459
around
Token around
Feature activation+0.000
distant
Token distant
Feature activation+0.000
stars
Token stars
Feature activation+1.968
,
Token,
Feature activation+2.278
and
Token and
Feature activation+1.100
dissect
Token dissect
Feature activation+0.000
s
Tokens
Feature activation+0.000
it
Token it
Feature activation+0.000
with
Token with
Feature activation+0.000
the
Token the
Feature activation+0.000
nuisance
Token nuisance
Feature activation+0.434
ab
Token ab
Feature activation+0.000
at
Tokenat
Feature activation+0.000
ement
Tokenement
Feature activation+2.087
program
Token program
Feature activation+2.364
âĢ
TokenâĢ
Feature activation+1.725
Ŀ
TokenĿ
Feature activation+1.454
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
Department
Token Department
Feature activation+0.000
prosecute
Token prosecute
Feature activation+2.752
alleged
Token alleged
Feature activation+0.162
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
hate
Tokenhate
Feature activation+0.054
crimes
Token crimes
Feature activation+2.337
âĢ
TokenâĢ
Feature activation+1.224
Ŀ
TokenĿ
Feature activation+1.001
particularly
Token particularly
Feature activation+2.178
susceptible
Token susceptible
Feature activation+0.629
to
Token to
Feature activation+0.336
of
Token of
Feature activation+0.000
Trump
Token Trump
Feature activation+0.511
âĢ
TokenâĢ
Feature activation+0.063
Ļ
TokenĻ
Feature activation+0.030
s
Tokens
Feature activation+0.000
win
Token win
Feature activation+2.248
entirely
Token entirely
Feature activation+0.080
on
Token on
Feature activation+0.000
feminism
Token feminism
Feature activation+0.605
âĢ
TokenâĢ
Feature activation+0.185
Ļ
TokenĻ
Feature activation+0.109

INTERVAL 1.616 - 2.020
CONTAINS 0.004%

ĺ
Tokenĺ
Feature activation+0.000
May
TokenMay
Feature activation+0.000
whatever
Token whatever
Feature activation+0.000
animals
Token animals
Feature activation+0.000
live
Token live
Feature activation+0.169
here
Token here
Feature activation+1.649
feed
Token feed
Feature activation+0.000
on
Token on
Feature activation+0.379
this
Token this
Feature activation+0.136
,
Token,
Feature activation+0.385
âĢ
TokenâĢ
Feature activation+0.000
connection
Token connection
Feature activation+0.254
with
Token with
Feature activation+0.000
a
Token a
Feature activation+0.000
particular
Token particular
Feature activation+0.000
team
Token team
Feature activation+0.482
,
Token,
Feature activation+1.621
to
Token to
Feature activation+0.372
be
Token be
Feature activation+0.045
an
Token an
Feature activation+0.000
important
Token important
Feature activation+0.000
aspect
Token aspect
Feature activation+0.000
this
Token this
Feature activation+0.000
perfect
Token perfect
Feature activation+0.000
process
Token process
Feature activation+0.000
of
Token of
Feature activation+0.000
regener
Token regener
Feature activation+0.000
ating
Tokenating
Feature activation+1.622
a
Token a
Feature activation+1.982
limb
Token limb
Feature activation+2.342
into
Token into
Feature activation+0.272
a
Token a
Feature activation+0.017
failure
Token failure
Feature activation+0.000
finds
Token finds
Feature activation+0.000
the
Token the
Feature activation+0.000
quiet
Token quiet
Feature activation+0.212
nuances
Token nuances
Feature activation+0.666
of
Token of
Feature activation+0.016
character
Token character
Feature activation+1.721
in
Token in
Feature activation+0.430
some
Token some
Feature activation+0.084
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
most
Token most
Feature activation+0.000
all
Token all
Feature activation+0.000
of
Token of
Feature activation+0.000
you
Token you
Feature activation+0.386
in
Token in
Feature activation+0.438
this
Token this
Feature activation+0.000
room
Token room
Feature activation+1.842
deeply
Token deeply
Feature activation+0.000
fearful
Token fearful
Feature activation+0.000
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
so
Token so
Feature activation+0.000

INTERVAL 1.212 - 1.616
CONTAINS 0.010%

the
Token the
Feature activation+0.000
'
Token '
Feature activation+0.000
bl
Tokenbl
Feature activation+0.000
onde
Tokenonde
Feature activation+0.000
'
Token'
Feature activation+0.000
comments
Token comments
Feature activation+1.217
as
Token as
Feature activation+0.000
'
Token '
Feature activation+0.000
dis
Tokendis
Feature activation+0.000
gr
Tokengr
Feature activation+0.000
ace
Tokenace
Feature activation+0.000
declared
Token declared
Feature activation+0.000
sizable
Token sizable
Feature activation+0.000
parts
Token parts
Feature activation+0.000
of
Token of
Feature activation+0.000
southern
Token southern
Feature activation+0.000
Mexico
Token Mexico
Feature activation+1.582
autonomous
Token autonomous
Feature activation+0.000
from
Token from
Feature activation+0.000
the
Token the
Feature activation+0.000
government
Token government
Feature activation+0.000
,
Token,
Feature activation+0.000
designed
Token designed
Feature activation+0.000
to
Token to
Feature activation+0.000
make
Token make
Feature activation+0.000
writing
Token writing
Feature activation+0.000
or
Token or
Feature activation+0.000
drawing
Token drawing
Feature activation+1.280
more
Token more
Feature activation+0.000
comfortable
Token comfortable
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
at
Token at
Feature activation+0.000
times
Token times
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
presidential
Token presidential
Feature activation+0.000
campaign
Token campaign
Feature activation+1.443
exhibited
Token exhibited
Feature activation+0.000
a
Token a
Feature activation+0.000
less
Token less
Feature activation+0.000
ideological
Token ideological
Feature activation+0.000
streak
Token streak
Feature activation+0.000
Pall
Token Pall
Feature activation+0.000
ister
Tokenister
Feature activation+0.000
and
Token and
Feature activation+0.000
the
Token the
Feature activation+0.000
Progressive
Token Progressive
Feature activation+0.000
Conservatives
Token Conservatives
Feature activation+1.307
(
Token (
Feature activation+0.639
PC
TokenPC
Feature activation+0.000
s
Tokens
Feature activation+0.000
)
Token)
Feature activation+0.936
in
Token in
Feature activation+0.000

INTERVAL 0.808 - 1.212
CONTAINS 0.023%

.
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
the
Token the
Feature activation+0.000
setting
Token setting
Feature activation+0.901
different
Token different
Feature activation+0.000
and
Token and
Feature activation+0.000
interesting
Token interesting
Feature activation+0.000
,
Token,
Feature activation+0.000
but
Token but
Feature activation+0.000
make
Token make
Feature activation+0.000
travel
Token travel
Feature activation+0.000
over
Token over
Feature activation+0.000
a
Token a
Feature activation+0.000
vast
Token vast
Feature activation+0.000
region
Token region
Feature activation+1.196
impossible
Token impossible
Feature activation+0.000
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
disrupt
Token disrupt
Feature activation+0.000
the
Token the
Feature activation+0.000
that
Token that
Feature activation+0.000
you
Token you
Feature activation+0.000
believe
Token believe
Feature activation+0.122
cares
Token cares
Feature activation+0.308
about
Token about
Feature activation+0.519
you
Token you
Feature activation+0.809
,
Token,
Feature activation+0.742
whether
Token whether
Feature activation+0.000
it
Token it
Feature activation+0.000
be
Token be
Feature activation+0.000
a
Token a
Feature activation+0.000
in
Token in
Feature activation+0.000
New
Token New
Feature activation+0.000
York
Token York
Feature activation+0.320
for
Token for
Feature activation+0.000
a
Token a
Feature activation+0.000
while
Token while
Feature activation+0.898
would
Token would
Feature activation+0.000
help
Token help
Feature activation+0.000
resolve
Token resolve
Feature activation+0.000
any
Token any
Feature activation+0.000
of
Token of
Feature activation+0.000
right
Token right
Feature activation+0.000
to
Token to
Feature activation+0.000
an
Token an
Feature activation+0.000
abortion
Token abortion
Feature activation+0.021
,
Token,
Feature activation+0.664
to
Token to
Feature activation+0.855
be
Token be
Feature activation+0.000
the
Token the
Feature activation+0.000
most
Token most
Feature activation+0.000
important
Token important
Feature activation+0.000
ruling
Token ruling
Feature activation+0.000

INTERVAL 0.404 - 0.808
CONTAINS 0.073%

protester
Token protester
Feature activation+0.000
at
Token at
Feature activation+0.000
Tian
Token Tian
Feature activation+0.000
an
Tokenan
Feature activation+0.000
men
Tokenmen
Feature activation+0.000
Square
Token Square
Feature activation+0.586
.
Token.
Feature activation+0.000
The
Token The
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.000
ĺ
Tokenĺ
Feature activation+0.000
Tank
TokenTank
Feature activation+0.000
text
Token text
Feature activation+0.050
message
Token message
Feature activation+0.128
accident
Token accident
Feature activation+0.296
cases
Token cases
Feature activation+0.420
going
Token going
Feature activation+0.018
forward
Token forward
Feature activation+0.615
:
Token:
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
We
TokenWe
Feature activation+0.000
affirm
Token affirm
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
t
Tokent
Feature activation+0.000
find
Token find
Feature activation+0.000
the
Token the
Feature activation+0.000
high
Token high
Feature activation+0.000
sensitivity
Token sensitivity
Feature activation+0.408
setting
Token setting
Feature activation+0.290
to
Token to
Feature activation+0.000
be
Token be
Feature activation+0.059
uncontroll
Token uncontroll
Feature activation+0.000
able
Tokenable
Feature activation+0.133
expect
Token expect
Feature activation+0.000
the
Token the
Feature activation+0.000
cost
Token cost
Feature activation+0.000
per
Token per
Feature activation+0.000
mile
Token mile
Feature activation+0.630
to
Token to
Feature activation+0.718
come
Token come
Feature activation+0.000
down
Token down
Feature activation+0.000
.
Token.
Feature activation+0.000
Fr
Token Fr
Feature activation+0.000
um
Tokenum
Feature activation+0.000
Guide
Token Guide
Feature activation+0.000
in
Token in
Feature activation+0.000
pdf
Token pdf
Feature activation+0.000
format
Token format
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.029
Ċ
TokenĊ
Feature activation+0.654
13
Token13
Feature activation+0.000
Essential
Token Essential
Feature activation+0.000
Words
Token Words
Feature activation+0.000
to
Token to
Feature activation+0.000
make
Token make
Feature activation+0.000

INTERVAL 0.000 - 0.404
CONTAINS 99.884%

,
Token,
Feature activation+0.000
these
Token these
Feature activation+0.000
people
Token people
Feature activation+0.000
,
Token,
Feature activation+0.000
even
Token even
Feature activation+0.000
without
Token without
Feature activation+0.000
knowing
Token knowing
Feature activation+0.000
me
Token me
Feature activation+0.000
,
Token,
Feature activation+0.000
they
Token they
Feature activation+0.000
've
Token've
Feature activation+0.000
a
Token a
Feature activation+0.000
serious
Token serious
Feature activation+0.000
real
Token real
Feature activation+0.000
-
Token-
Feature activation+0.000
world
Tokenworld
Feature activation+0.000
stumbling
Token stumbling
Feature activation+0.000
block
Token block
Feature activation+0.000
for
Token for
Feature activation+0.000
all
Token all
Feature activation+0.000
would
Token would
Feature activation+0.000
-
Token-
Feature activation+0.000
and
Token and
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
Nile
Token Nile
Feature activation+0.000
mars
Token mars
Feature activation+0.000
hes
Tokenhes
Feature activation+0.000
during
Token during
Feature activation+0.000
the
Token the
Feature activation+0.000
wet
Token wet
Feature activation+0.000
season
Token season
Feature activation+0.000
.
Token.
Feature activation+0.000
ari
Tokenari
Feature activation+0.000
ya
Tokenya
Feature activation+0.000
).
Token).
Feature activation+0.000
While
Token While
Feature activation+0.000
enough
Token enough
Feature activation+0.000
has
Token has
Feature activation+0.000
been
Token been
Feature activation+0.000
said
Token said
Feature activation+0.000
about
Token about
Feature activation+0.000
his
Token his
Feature activation+0.000
knack
Token knack
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
archaeological
Token archaeological
Feature activation+0.000
resource
Token resource
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
and
Token and
Feature activation+0.000
is
Token is
Feature activation+0.000
now
Token now
Feature activation+0.000
seeking
Token seeking
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 H7.4: (feature 12628)

TOP ACTIVATIONS
MAX = 2.838

s
Tokens
Feature activation+0.000
duty
Token duty
Feature activation+0.000
as
Token as
Feature activation+0.000
a
Token a
Feature activation+0.122
progressive
Token progressive
Feature activation+1.559
site
Token site
Feature activation+2.838
to
Token to
Feature activation+0.779
live
Token live
Feature activation+0.000
out
Token out
Feature activation+0.000
its
Token its
Feature activation+0.000
pro
Token pro
Feature activation+0.000
women
Token women
Feature activation+0.000
are
Token are
Feature activation+0.000
more
Token more
Feature activation+0.000
likely
Token likely
Feature activation+0.000
than
Token than
Feature activation+0.000
men
Token men
Feature activation+2.837
to
Token to
Feature activation+0.424
interrupt
Token interrupt
Feature activation+0.022
their
Token their
Feature activation+0.000
careers
Token careers
Feature activation+0.118
for
Token for
Feature activation+0.000
much
Token much
Feature activation+0.000
more
Token more
Feature activation+0.000
likely
Token likely
Feature activation+0.000
than
Token than
Feature activation+0.000
Romney
Token Romney
Feature activation+1.631
voters
Token voters
Feature activation+2.829
(
Token (
Feature activation+0.772
59
Token59
Feature activation+0.000
percent
Token percent
Feature activation+0.542
)
Token)
Feature activation+2.465
to
Token to
Feature activation+0.430
three
Token three
Feature activation+0.000
times
Token times
Feature activation+0.000
more
Token more
Feature activation+0.000
likely
Token likely
Feature activation+0.000
than
Token than
Feature activation+0.000
civilians
Token civilians
Feature activation+2.483
to
Token to
Feature activation+0.000
have
Token have
Feature activation+0.000
committed
Token committed
Feature activation+0.000
a
Token a
Feature activation+0.000
violent
Token violent
Feature activation+0.000
Romney
Token Romney
Feature activation+1.631
voters
Token voters
Feature activation+2.829
(
Token (
Feature activation+0.772
59
Token59
Feature activation+0.000
percent
Token percent
Feature activation+0.542
)
Token)
Feature activation+2.465
to
Token to
Feature activation+0.430
say
Token say
Feature activation+0.000
they
Token they
Feature activation+0.000
are
Token are
Feature activation+0.000
satisfied
Token satisfied
Feature activation+0.000
it
Token it
Feature activation+0.000
is
Token is
Feature activation+0.000
our
Token our
Feature activation+0.000
job
Token job
Feature activation+0.000
as
Token as
Feature activation+0.000
parents
Token parents
Feature activation+2.404
to
Token to
Feature activation+0.597
ded
Token ded
Feature activation+0.000
uce
Tokenuce
Feature activation+0.000
,
Token,
Feature activation+0.000
Sherlock
Token Sherlock
Feature activation+0.000
are
Token are
Feature activation+0.000
less
Token less
Feature activation+0.000
likely
Token likely
Feature activation+0.000
than
Token than
Feature activation+0.000
American
Token American
Feature activation+0.000
men
Token men
Feature activation+2.361
to
Token to
Feature activation+0.000
ride
Token ride
Feature activation+0.000
bikes
Token bikes
Feature activation+0.000
in
Token in
Feature activation+0.000
cities
Token cities
Feature activation+0.000
his
Token his
Feature activation+0.000
status
Token status
Feature activation+0.101
as
Token as
Feature activation+0.000
an
Token an
Feature activation+0.000
art
Token art
Feature activation+0.477
student
Token student
Feature activation+2.234
to
Token to
Feature activation+0.657
get
Token get
Feature activation+0.000
close
Token close
Feature activation+0.000
to
Token to
Feature activation+0.000
J
Token J
Feature activation+0.000
actually
Token actually
Feature activation+0.000
know
Token know
Feature activation+0.000
something
Token something
Feature activation+0.000
about
Token about
Feature activation+0.000
Friedman
Token Friedman
Feature activation+0.705
to
Token to
Feature activation+2.179
know
Token know
Feature activation+0.000
that
Token that
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
alt
Tokenalt
Feature activation+0.000
a
Tokena
Feature activation+1.828
in
Token in
Feature activation+0.859
Bain
Token Bain
Feature activation+0.935
's
Token's
Feature activation+0.315
blood
Token blood
Feature activation+2.101
to
Token to
Feature activation+0.000
suggest
Token suggest
Feature activation+0.000
that
Token that
Feature activation+0.000
Bain
Token Bain
Feature activation+0.000
was
Token was
Feature activation+0.000
likely
Token likely
Feature activation+0.000
than
Token than
Feature activation+0.000
the
Token the
Feature activation+0.000
average
Token average
Feature activation+0.875
internet
Token internet
Feature activation+0.000
user
Token user
Feature activation+2.079
to
Token to
Feature activation+0.000
watch
Token watch
Feature activation+0.000
ABC
Token ABC
Feature activation+0.000
programs
Token programs
Feature activation+0.000
like
Token like
Feature activation+0.000
Libertarian
Token Libertarian
Feature activation+0.000
John
Token John
Feature activation+0.000
Hosp
Token Hosp
Feature activation+0.000
ers
Tokeners
Feature activation+0.950
in
Token in
Feature activation+0.000
1972
Token 1972
Feature activation+2.044
that
Token that
Feature activation+0.390
a
Token a
Feature activation+0.000
third
Token third
Feature activation+0.000
-
Token-
Feature activation+0.000
party
Tokenparty
Feature activation+0.000
perhaps
Token perhaps
Feature activation+0.000
the
Token the
Feature activation+0.000
first
Token first
Feature activation+0.000
person
Token person
Feature activation+0.000
besides
Token besides
Feature activation+0.000
Satoshi
Token Satoshi
Feature activation+1.962
to
Token to
Feature activation+0.431
really
Token really
Feature activation+0.000
throw
Token throw
Feature activation+0.000
his
Token his
Feature activation+0.000
weight
Token weight
Feature activation+0.000
banks
Token banks
Feature activation+0.000
whose
Token whose
Feature activation+0.000
deposits
Token deposits
Feature activation+0.000
are
Token are
Feature activation+0.000
federally
Token federally
Feature activation+0.000
insured
Token insured
Feature activation+1.958
from
Token from
Feature activation+0.000
trading
Token trading
Feature activation+0.000
for
Token for
Feature activation+0.000
their
Token their
Feature activation+0.000
own
Token own
Feature activation+0.000
consumed
Token consumed
Feature activation+0.000
by
Token by
Feature activation+0.000
me
Token me
Feature activation+1.283
gal
Tokengal
Feature activation+0.000
oman
Tokenoman
Feature activation+0.000
ia
Tokenia
Feature activation+1.847
to
Token to
Feature activation+0.000
share
Token share
Feature activation+0.000
his
Token his
Feature activation+0.000
body
Token body
Feature activation+0.000
with
Token with
Feature activation+0.000
lack
Token lack
Feature activation+0.000
of
Token of
Feature activation+0.000
Cy
Token Cy
Feature activation+0.000
mb
Tokenmb
Feature activation+0.000
alt
Tokenalt
Feature activation+0.000
a
Tokena
Feature activation+1.828
in
Token in
Feature activation+0.859
Bain
Token Bain
Feature activation+0.935
's
Token's
Feature activation+0.315
blood
Token blood
Feature activation+2.101
to
Token to
Feature activation+0.000
love
Token love
Feature activation+1.172
of
Token of
Feature activation+0.042
American
Token American
Feature activation+0.081
values
Token values
Feature activation+1.554
and
Token and
Feature activation+1.042
traditions
Token traditions
Feature activation+1.780
that
Token that
Feature activation+0.000
she
Token she
Feature activation+0.000
had
Token had
Feature activation+0.000
no
Token no
Feature activation+0.000
choice
Token choice
Feature activation+0.000
a
Token a
Feature activation+0.000
good
Token good
Feature activation+0.000
while
Token while
Feature activation+0.000
after
Token after
Feature activation+0.000
the
Token the
Feature activation+0.000
session
Token session
Feature activation+1.737
getting
Token getting
Feature activation+0.000
sand
Token sand
Feature activation+0.000
out
Token out
Feature activation+0.000
of
Token of
Feature activation+0.000
their
Token their
Feature activation+0.000
There
Token There
Feature activation+0.000
is
Token is
Feature activation+0.000
no
Token no
Feature activation+0.000
place
Token place
Feature activation+0.000
in
Token in
Feature activation+0.000
Minnesota
Token Minnesota
Feature activation+1.706
for
Token for
Feature activation+0.000
intolerance
Token intolerance
Feature activation+0.000
of
Token of
Feature activation+0.000
all
Token all
Feature activation+0.000
Americans
Token Americans
Feature activation+0.000
final
Token final
Feature activation+0.000
interview
Token interview
Feature activation+0.020
in
Token in
Feature activation+0.268
the
Token the
Feature activation+0.004
White
Token White
Feature activation+0.000
House
Token House
Feature activation+1.691
to
Token to
Feature activation+0.797
tell
Token tell
Feature activation+0.000
Oprah
Token Oprah
Feature activation+0.000
Win
Token Win
Feature activation+0.000
frey
Tokenfrey
Feature activation+0.000

Top DFA by src position
MAX = 3.176

Salon
Token Salon
Feature activation+0.011
Top resid features:
âĢ
TokenâĢ
Feature activation+0.006
Top resid features:
Ļ
TokenĻ
Feature activation+0.001
Top resid features:
s
Tokens
Feature activation+0.038
Top resid features:
duty
Token duty
Feature activation+0.053
Top resid features:
as
Token as
Feature activation+2.487
Top resid features:
a
Token a
Feature activation+0.531
Top resid features:
progressive
Token progressive
Feature activation+0.033
Top resid features:
site
Token site
Feature activation+0.036
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
live
Token live
Feature activation+0.000
Top resid features:
because
Token because
Feature activation+0.012
Top resid features:
women
Token women
Feature activation-0.022
Top resid features:
are
Token are
Feature activation-0.015
Top resid features:
more
Token more
Feature activation-0.011
Top resid features:
likely
Token likely
Feature activation-0.025
Top resid features:
than
Token than
Feature activation+3.176
Top resid features:
men
Token men
Feature activation+0.087
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
interrupt
Token interrupt
Feature activation+0.000
Top resid features:
their
Token their
Feature activation+0.000
Top resid features:
careers
Token careers
Feature activation+0.000
Top resid features:
)
Token)
Feature activation-0.013
Top resid features:
are
Token are
Feature activation-0.025
Top resid features:
much
Token much
Feature activation-0.007
Top resid features:
more
Token more
Feature activation-0.028
Top resid features:
likely
Token likely
Feature activation-0.021
Top resid features:
than
Token than
Feature activation+3.123
Top resid features:
Romney
Token Romney
Feature activation+0.110
Top resid features:
voters
Token voters
Feature activation+0.003
Top resid features:
(
Token (
Feature activation+0.000
Top resid features:
59
Token59
Feature activation+0.000
Top resid features:
percent
Token percent
Feature activation+0.000
Top resid features:
about
Token about
Feature activation-0.013
Top resid features:
three
Token three
Feature activation-0.004
Top resid features:
times
Token times
Feature activation-0.006
Top resid features:
more
Token more
Feature activation-0.007
Top resid features:
likely
Token likely
Feature activation-0.005
Top resid features:
than
Token than
Feature activation+2.800
Top resid features:
civilians
Token civilians
Feature activation+0.117
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
have
Token have
Feature activation+0.000
Top resid features:
committed
Token committed
Feature activation+0.000
Top resid features:
a
Token a
Feature activation+0.000
Top resid features:
)
Token)
Feature activation-0.002
Top resid features:
are
Token are
Feature activation-0.020
Top resid features:
much
Token much
Feature activation-0.007
Top resid features:
more
Token more
Feature activation-0.034
Top resid features:
likely
Token likely
Feature activation-0.020
Top resid features:
than
Token than
Feature activation+2.740
Top resid features:
Romney
Token Romney
Feature activation+0.071
Top resid features:
voters
Token voters
Feature activation-0.046
Top resid features:
(
Token (
Feature activation+0.160
Top resid features:
59
Token59
Feature activation-0.001
Top resid features:
percent
Token percent
Feature activation-0.013
Top resid features:
which
Token which
Feature activation-0.000
Top resid features:
it
Token it
Feature activation-0.001
Top resid features:
is
Token is
Feature activation+0.011
Top resid features:
our
Token our
Feature activation+0.048
Top resid features:
job
Token job
Feature activation+0.073
Top resid features:
as
Token as
Feature activation+2.586
Top resid features:
parents
Token parents
Feature activation+0.055
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
ded
Token ded
Feature activation+0.000
Top resid features:
uce
Tokenuce
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
American
Token American
Feature activation+0.002
Top resid features:
women
Token women
Feature activation-0.007
Top resid features:
are
Token are
Feature activation-0.012
Top resid features:
less
Token less
Feature activation-0.023
Top resid features:
likely
Token likely
Feature activation-0.034
Top resid features:
than
Token than
Feature activation+2.757
Top resid features:
American
Token American
Feature activation+0.011
Top resid features:
men
Token men
Feature activation+0.043
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
ride
Token ride
Feature activation+0.000
Top resid features:
bikes
Token bikes
Feature activation+0.000
Top resid features:
said
Token said
Feature activation-0.009
Top resid features:
his
Token his
Feature activation-0.010
Top resid features:
uncle
Token uncle
Feature activation-0.001
Top resid features:
used
Token used
Feature activation-0.010
Top resid features:
his
Token his
Feature activation+0.669
Top resid features:
status
Token status
Feature activation+1.067
Top resid features:
as
Token as
Feature activation+0.548
Top resid features:
an
Token an
Feature activation+0.267
Top resid features:
art
Token art
Feature activation+0.030
Top resid features:
student
Token student
Feature activation+0.012
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.010
Top resid features:
you
Token you
Feature activation-0.000
Top resid features:
'd
Token'd
Feature activation-0.004
Top resid features:
have
Token have
Feature activation-0.044
Top resid features:
to
Token to
Feature activation+0.077
Top resid features:
have
Token have
Feature activation+2.297
Top resid features:
actually
Token actually
Feature activation+0.012
Top resid features:
know
Token know
Feature activation+0.189
Top resid features:
something
Token something
Feature activation-0.009
Top resid features:
about
Token about
Feature activation-0.002
Top resid features:
Friedman
Token Friedman
Feature activation+0.002
Top resid features:
other
Token other
Feature activation-0.005
Top resid features:
hand
Token hand
Feature activation-0.014
Top resid features:
,
Token,
Feature activation-0.007
Top resid features:
used
Token used
Feature activation-0.011
Top resid features:
the
Token the
Feature activation+0.078
Top resid features:
lack
Token lack
Feature activation+1.867
Top resid features:
of
Token of
Feature activation+0.367
Top resid features:
Cy
Token Cy
Feature activation+0.003
Top resid features:
mb
Tokenmb
Feature activation-0.000
Top resid features:
alt
Tokenalt
Feature activation+0.002
Top resid features:
a
Tokena
Feature activation-0.010
Top resid features:
cut
Tokencut
Feature activation-0.002
Top resid features:
ters
Tokenters
Feature activation-0.002
Top resid features:
are
Token are
Feature activation-0.007
Top resid features:
more
Token more
Feature activation-0.014
Top resid features:
likely
Token likely
Feature activation-0.048
Top resid features:
than
Token than
Feature activation+2.407
Top resid features:
the
Token the
Feature activation+0.126
Top resid features:
average
Token average
Feature activation-0.010
Top resid features:
internet
Token internet
Feature activation-0.001
Top resid features:
user
Token user
Feature activation+0.001
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
would
Token would
Feature activation-0.001
Top resid features:
be
Token be
Feature activation-0.001
Top resid features:
the
Token the
Feature activation-0.002
Top resid features:
first
Token first
Feature activation-0.000
Top resid features:
time
Token time
Feature activation+0.006
Top resid features:
since
Token since
Feature activation+2.340
Top resid features:
a
Token a
Feature activation+0.028
Top resid features:
m
Token m
Feature activation-0.001
Top resid features:
aver
Tokenaver
Feature activation-0.001
Top resid features:
ick
Tokenick
Feature activation-0.000
Top resid features:
elector
Token elector
Feature activation-0.001
Top resid features:
,
Token,
Feature activation-0.009
Top resid features:
perhaps
Token perhaps
Feature activation-0.024
Top resid features:
the
Token the
Feature activation-0.010
Top resid features:
first
Token first
Feature activation+0.017
Top resid features:
person
Token person
Feature activation-0.013
Top resid features:
besides
Token besides
Feature activation+2.335
Top resid features:
Satoshi
Token Satoshi
Feature activation-0.002
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
really
Token really
Feature activation+0.000
Top resid features:
throw
Token throw
Feature activation+0.000
Top resid features:
his
Token his
Feature activation+0.000
Top resid features:
rict
Tokenrict
Feature activation-0.004
Top resid features:
the
Token the
Feature activation-0.003
Top resid features:
ability
Token ability
Feature activation-0.016
Top resid features:
of
Token of
Feature activation+0.025
Top resid features:
banks
Token banks
Feature activation+0.045
Top resid features:
whose
Token whose
Feature activation+2.190
Top resid features:
deposits
Token deposits
Feature activation-0.020
Top resid features:
are
Token are
Feature activation+0.086
Top resid features:
federally
Token federally
Feature activation+0.008
Top resid features:
insured
Token insured
Feature activation-0.016
Top resid features:
from
Token from
Feature activation+0.000
Top resid features:
I
Token I
Feature activation+0.005
Top resid features:
believe
Token believe
Feature activation+0.006
Top resid features:
he
Token he
Feature activation-0.009
Top resid features:
was
Token was
Feature activation-0.039
Top resid features:
too
Token too
Feature activation-0.012
Top resid features:
consumed
Token consumed
Feature activation+1.648
Top resid features:
by
Token by
Feature activation+0.584
Top resid features:
me
Token me
Feature activation+0.045
Top resid features:
gal
Tokengal
Feature activation-0.053
Top resid features:
oman
Tokenoman
Feature activation-0.010
Top resid features:
ia
Tokenia
Feature activation+0.001
Top resid features:
other
Token other
Feature activation-0.007
Top resid features:
hand
Token hand
Feature activation-0.012
Top resid features:
,
Token,
Feature activation-0.005
Top resid features:
used
Token used
Feature activation-0.012
Top resid features:
the
Token the
Feature activation+0.075
Top resid features:
lack
Token lack
Feature activation+1.426
Top resid features:
of
Token of
Feature activation+0.725
Top resid features:
Cy
Token Cy
Feature activation-0.012
Top resid features:
mb
Tokenmb
Feature activation-0.002
Top resid features:
alt
Tokenalt
Feature activation-0.019
Top resid features:
a
Tokena
Feature activation-0.039
Top resid features:
wife
Tokenwife
Feature activation+0.001
Top resid features:
who
Token who
Feature activation-0.011
Top resid features:
was
Token was
Feature activation+0.025
Top resid features:
so
Token so
Feature activation+0.017
Top resid features:
driven
Token driven
Feature activation+0.207
Top resid features:
by
Token by
Feature activation+1.731
Top resid features:
her
Token her
Feature activation+0.073
Top resid features:
love
Token love
Feature activation+0.023
Top resid features:
of
Token of
Feature activation+0.029
Top resid features:
American
Token American
Feature activation+0.008
Top resid features:
values
Token values
Feature activation+0.010
Top resid features:
undoubtedly
Token undoubtedly
Feature activation-0.038
Top resid features:
spent
Token spent
Feature activation-0.020
Top resid features:
a
Token a
Feature activation+0.005
Top resid features:
good
Token good
Feature activation+0.009
Top resid features:
while
Token while
Feature activation+0.052
Top resid features:
after
Token after
Feature activation+2.073
Top resid features:
the
Token the
Feature activation+0.008
Top resid features:
session
Token session
Feature activation+0.008
Top resid features:
getting
Token getting
Feature activation+0.000
Top resid features:
sand
Token sand
Feature activation+0.000
Top resid features:
out
Token out
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.006
Top resid features:
There
Token There
Feature activation+0.005
Top resid features:
is
Token is
Feature activation-0.006
Top resid features:
no
Token no
Feature activation-0.005
Top resid features:
place
Token place
Feature activation-0.005
Top resid features:
in
Token in
Feature activation+2.014
Top resid features:
Minnesota
Token Minnesota
Feature activation+0.048
Top resid features:
for
Token for
Feature activation+0.000
Top resid features:
intolerance
Token intolerance
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
all
Token all
Feature activation+0.000
Top resid features:
Lady
Token Lady
Feature activation-0.013
Top resid features:
used
Token used
Feature activation-0.005
Top resid features:
her
Token her
Feature activation+0.139
Top resid features:
final
Token final
Feature activation+0.121
Top resid features:
interview
Token interview
Feature activation+0.746
Top resid features:
in
Token in
Feature activation+0.853
Top resid features:
the
Token the
Feature activation+0.153
Top resid features:
White
Token White
Feature activation+0.005
Top resid features:
House
Token House
Feature activation+0.027
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
tell
Token tell
Feature activation+0.000
Top resid features:

Decoder Weights Distribution

Head 0: 0.07

Head 1: 0.02

Head 2: 0.02

Head 3: 0.07

Head 4: 0.57

Head 5: 0.04

Head 6: 0.03

Head 7: 0.02

Head 8: 0.07

Head 9: 0.05

Head 10: 0.01

Head 11: 0.02

Positive logits

kefeller3.08

ammy2.73

ucket2.52

iggins2.48

qt2.44

dds2.41

umpy2.38

dissu2.34

izo2.28

ennett2.22

Moz2.19

imes2.18

rael2.16

ripp2.14

persuade2.10

bler2.09

uka2.08

imar2.07

monkey2.07

merce2.07

Negative logits

Cooldown-2.92

???-2.70

)?-2.44

JP-2.34

"?-2.26

.?-2.26

Tro-2.25

WARD-2.25

??-2.20

hy-2.11

-2.06

Phase-2.03

Reply-2.02

recy-2.02

-2.01

Trooper-2.01

ensible-1.98

amy-1.93

cooldown-1.92

Directive-1.90

INTERVAL 2.554 - 2.838
CONTAINS 0.000%

much
Token much
Feature activation+0.000
more
Token more
Feature activation+0.000
likely
Token likely
Feature activation+0.000
than
Token than
Feature activation+0.000
Romney
Token Romney
Feature activation+1.631
voters
Token voters
Feature activation+2.829
(
Token (
Feature activation+0.772
59
Token59
Feature activation+0.000
percent
Token percent
Feature activation+0.542
)
Token)
Feature activation+2.465
to
Token to
Feature activation+0.430
s
Tokens
Feature activation+0.000
duty
Token duty
Feature activation+0.000
as
Token as
Feature activation+0.000
a
Token a
Feature activation+0.122
progressive
Token progressive
Feature activation+1.559
site
Token site
Feature activation+2.838
to
Token to
Feature activation+0.779
live
Token live
Feature activation+0.000
out
Token out
Feature activation+0.000
its
Token its
Feature activation+0.000
pro
Token pro
Feature activation+0.000
women
Token women
Feature activation+0.000
are
Token are
Feature activation+0.000
more
Token more
Feature activation+0.000
likely
Token likely
Feature activation+0.000
than
Token than
Feature activation+0.000
men
Token men
Feature activation+2.837
to
Token to
Feature activation+0.424
interrupt
Token interrupt
Feature activation+0.022
their
Token their
Feature activation+0.000
careers
Token careers
Feature activation+0.118
for
Token for
Feature activation+0.000

INTERVAL 2.270 - 2.554
CONTAINS 0.000%

Romney
Token Romney
Feature activation+1.631
voters
Token voters
Feature activation+2.829
(
Token (
Feature activation+0.772
59
Token59
Feature activation+0.000
percent
Token percent
Feature activation+0.542
)
Token)
Feature activation+2.465
to
Token to
Feature activation+0.430
say
Token say
Feature activation+0.000
they
Token they
Feature activation+0.000
are
Token are
Feature activation+0.000
satisfied
Token satisfied
Feature activation+0.000
three
Token three
Feature activation+0.000
times
Token times
Feature activation+0.000
more
Token more
Feature activation+0.000
likely
Token likely
Feature activation+0.000
than
Token than
Feature activation+0.000
civilians
Token civilians
Feature activation+2.483
to
Token to
Feature activation+0.000
have
Token have
Feature activation+0.000
committed
Token committed
Feature activation+0.000
a
Token a
Feature activation+0.000
violent
Token violent
Feature activation+0.000
are
Token are
Feature activation+0.000
less
Token less
Feature activation+0.000
likely
Token likely
Feature activation+0.000
than
Token than
Feature activation+0.000
American
Token American
Feature activation+0.000
men
Token men
Feature activation+2.361
to
Token to
Feature activation+0.000
ride
Token ride
Feature activation+0.000
bikes
Token bikes
Feature activation+0.000
in
Token in
Feature activation+0.000
cities
Token cities
Feature activation+0.000
it
Token it
Feature activation+0.000
is
Token is
Feature activation+0.000
our
Token our
Feature activation+0.000
job
Token job
Feature activation+0.000
as
Token as
Feature activation+0.000
parents
Token parents
Feature activation+2.404
to
Token to
Feature activation+0.597
ded
Token ded
Feature activation+0.000
uce
Tokenuce
Feature activation+0.000
,
Token,
Feature activation+0.000
Sherlock
Token Sherlock
Feature activation+0.000

INTERVAL 1.987 - 2.270
CONTAINS 0.001%

alt
Tokenalt
Feature activation+0.000
a
Tokena
Feature activation+1.828
in
Token in
Feature activation+0.859
Bain
Token Bain
Feature activation+0.935
's
Token's
Feature activation+0.315
blood
Token blood
Feature activation+2.101
to
Token to
Feature activation+0.000
suggest
Token suggest
Feature activation+0.000
that
Token that
Feature activation+0.000
Bain
Token Bain
Feature activation+0.000
was
Token was
Feature activation+0.000
Libertarian
Token Libertarian
Feature activation+0.000
John
Token John
Feature activation+0.000
Hosp
Token Hosp
Feature activation+0.000
ers
Tokeners
Feature activation+0.950
in
Token in
Feature activation+0.000
1972
Token 1972
Feature activation+2.044
that
Token that
Feature activation+0.390
a
Token a
Feature activation+0.000
third
Token third
Feature activation+0.000
-
Token-
Feature activation+0.000
party
Tokenparty
Feature activation+0.000
actually
Token actually
Feature activation+0.000
know
Token know
Feature activation+0.000
something
Token something
Feature activation+0.000
about
Token about
Feature activation+0.000
Friedman
Token Friedman
Feature activation+0.705
to
Token to
Feature activation+2.179
know
Token know
Feature activation+0.000
that
Token that
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
his
Token his
Feature activation+0.000
status
Token status
Feature activation+0.101
as
Token as
Feature activation+0.000
an
Token an
Feature activation+0.000
art
Token art
Feature activation+0.477
student
Token student
Feature activation+2.234
to
Token to
Feature activation+0.657
get
Token get
Feature activation+0.000
close
Token close
Feature activation+0.000
to
Token to
Feature activation+0.000
J
Token J
Feature activation+0.000
likely
Token likely
Feature activation+0.000
than
Token than
Feature activation+0.000
the
Token the
Feature activation+0.000
average
Token average
Feature activation+0.875
internet
Token internet
Feature activation+0.000
user
Token user
Feature activation+2.079
to
Token to
Feature activation+0.000
watch
Token watch
Feature activation+0.000
ABC
Token ABC
Feature activation+0.000
programs
Token programs
Feature activation+0.000
like
Token like
Feature activation+0.000

INTERVAL 1.703 - 1.987
CONTAINS 0.001%

a
Token a
Feature activation+0.000
good
Token good
Feature activation+0.000
while
Token while
Feature activation+0.000
after
Token after
Feature activation+0.000
the
Token the
Feature activation+0.000
session
Token session
Feature activation+1.737
getting
Token getting
Feature activation+0.000
sand
Token sand
Feature activation+0.000
out
Token out
Feature activation+0.000
of
Token of
Feature activation+0.000
their
Token their
Feature activation+0.000
consumed
Token consumed
Feature activation+0.000
by
Token by
Feature activation+0.000
me
Token me
Feature activation+1.283
gal
Tokengal
Feature activation+0.000
oman
Tokenoman
Feature activation+0.000
ia
Tokenia
Feature activation+1.847
to
Token to
Feature activation+0.000
share
Token share
Feature activation+0.000
his
Token his
Feature activation+0.000
body
Token body
Feature activation+0.000
with
Token with
Feature activation+0.000
love
Token love
Feature activation+1.172
of
Token of
Feature activation+0.042
American
Token American
Feature activation+0.081
values
Token values
Feature activation+1.554
and
Token and
Feature activation+1.042
traditions
Token traditions
Feature activation+1.780
that
Token that
Feature activation+0.000
she
Token she
Feature activation+0.000
had
Token had
Feature activation+0.000
no
Token no
Feature activation+0.000
choice
Token choice
Feature activation+0.000
perhaps
Token perhaps
Feature activation+0.000
the
Token the
Feature activation+0.000
first
Token first
Feature activation+0.000
person
Token person
Feature activation+0.000
besides
Token besides
Feature activation+0.000
Satoshi
Token Satoshi
Feature activation+1.962
to
Token to
Feature activation+0.431
really
Token really
Feature activation+0.000
throw
Token throw
Feature activation+0.000
his
Token his
Feature activation+0.000
weight
Token weight
Feature activation+0.000
banks
Token banks
Feature activation+0.000
whose
Token whose
Feature activation+0.000
deposits
Token deposits
Feature activation+0.000
are
Token are
Feature activation+0.000
federally
Token federally
Feature activation+0.000
insured
Token insured
Feature activation+1.958
from
Token from
Feature activation+0.000
trading
Token trading
Feature activation+0.000
for
Token for
Feature activation+0.000
their
Token their
Feature activation+0.000
own
Token own
Feature activation+0.000

INTERVAL 1.419 - 1.703
CONTAINS 0.002%

people
Token people
Feature activation+0.444
who
Token who
Feature activation+0.000
live
Token live
Feature activation+0.736
in
Token in
Feature activation+0.380
this
Token this
Feature activation+0.381
region
Token region
Feature activation+1.434
to
Token to
Feature activation+0.326
support
Token support
Feature activation+0.000
a
Token a
Feature activation+0.000
Yes
Token Yes
Feature activation+0.000
vote
Token vote
Feature activation+0.000
second
Token second
Feature activation+0.000
most
Token most
Feature activation+0.000
important
Token important
Feature activation+0.000
supplier
Token supplier
Feature activation+0.000
after
Token after
Feature activation+0.000
Russia
Token Russia
Feature activation+1.430
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
rc
Tokenrc
Feature activation+0.000
/
Token/
Feature activation+0.000
service
Token service
Feature activation+1.362
,
Token,
Feature activation+0.372
in
Token in
Feature activation+0.510
January
Token January
Feature activation+0.437
2015
Token 2015
Feature activation+1.313
,
Token,
Feature activation+1.556
that
Token that
Feature activation+0.000
Z
Token Z
Feature activation+0.000
abl
Tokenabl
Feature activation+0.000
ud
Tokenud
Feature activation+0.000
ovsky
Tokenovsky
Feature activation+0.000
by
Token by
Feature activation+0.000
her
Token her
Feature activation+0.007
love
Token love
Feature activation+1.172
of
Token of
Feature activation+0.042
American
Token American
Feature activation+0.081
values
Token values
Feature activation+1.554
and
Token and
Feature activation+1.042
traditions
Token traditions
Feature activation+1.780
that
Token that
Feature activation+0.000
she
Token she
Feature activation+0.000
had
Token had
Feature activation+0.000
oth
Tokenoth
Feature activation+0.067
of
Token of
Feature activation+0.089
the
Token the
Feature activation+0.061
Indigenous
Token Indigenous
Feature activation+0.045
Environmental
Token Environmental
Feature activation+0.000
Network
Token Network
Feature activation+1.640
,
Token,
Feature activation+0.829
this
Token this
Feature activation+0.000
clip
Token clip
Feature activation+0.000
from
Token from
Feature activation+0.000
the
Token the
Feature activation+0.000

INTERVAL 1.135 - 1.419
CONTAINS 0.005%

and
Token and
Feature activation+0.189
live
Token live
Feature activation+0.336
around
Token around
Feature activation+0.000
distant
Token distant
Feature activation+0.000
stars
Token stars
Feature activation+1.063
,
Token,
Feature activation+1.356
and
Token and
Feature activation+0.982
dissect
Token dissect
Feature activation+0.000
s
Tokens
Feature activation+0.000
it
Token it
Feature activation+0.000
with
Token with
Feature activation+0.000
the
Token the
Feature activation+0.000
nuisance
Token nuisance
Feature activation+0.160
ab
Token ab
Feature activation+0.000
at
Tokenat
Feature activation+0.000
ement
Tokenement
Feature activation+1.137
program
Token program
Feature activation+1.253
âĢ
TokenâĢ
Feature activation+0.845
Ŀ
TokenĿ
Feature activation+0.637
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
Department
Token Department
Feature activation+0.000
's
Token's
Feature activation+0.000
first
Token first
Feature activation+0.000
world
Token world
Feature activation+0.000
war
Token war
Feature activation+0.602
effort
Token effort
Feature activation+0.969
to
Token to
Feature activation+1.243
help
Token help
Feature activation+0.000
improve
Token improve
Feature activation+0.000
community
Token community
Feature activation+0.000
relations
Token relations
Feature activation+0.000
during
Token during
Feature activation+0.000
the
Token the
Feature activation+0.000
only
Token only
Feature activation+0.000
one
Token one
Feature activation+0.000
at
Token at
Feature activation+0.000
D
Token D
Feature activation+0.152
under
Tokenunder
Feature activation+1.237
-
Token-
Feature activation+0.033
M
TokenM
Feature activation+0.002
iff
Tokeniff
Feature activation+0.000
lin
Tokenlin
Feature activation+1.576
who
Token who
Feature activation+0.000
at
Token at
Feature activation+0.000
times
Token times
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
presidential
Token presidential
Feature activation+0.000
campaign
Token campaign
Feature activation+1.266
exhibited
Token exhibited
Feature activation+0.000
a
Token a
Feature activation+0.000
less
Token less
Feature activation+0.000
ideological
Token ideological
Feature activation+0.000
streak
Token streak
Feature activation+0.000

INTERVAL 0.851 - 1.135
CONTAINS 0.011%

live
Token live
Feature activation+0.336
around
Token around
Feature activation+0.000
distant
Token distant
Feature activation+0.000
stars
Token stars
Feature activation+1.063
,
Token,
Feature activation+1.356
and
Token and
Feature activation+0.982
dissect
Token dissect
Feature activation+0.000
s
Tokens
Feature activation+0.000
it
Token it
Feature activation+0.000
with
Token with
Feature activation+0.000
startling
Token startling
Feature activation+0.000
our
Token our
Feature activation+0.000
court
Token court
Feature activation+0.000
rooms
Tokenrooms
Feature activation+0.203
or
Token or
Feature activation+0.000
our
Token our
Feature activation+0.000
jails
Token jails
Feature activation+0.941
filled
Token filled
Feature activation+0.000
with
Token with
Feature activation+0.188
people
Token people
Feature activation+0.000
caught
Token caught
Feature activation+0.000
on
Token on
Feature activation+0.000
,
Token,
Feature activation+0.000
ranging
Token ranging
Feature activation+0.000
from
Token from
Feature activation+0.000
time
Token time
Feature activation+0.000
between
Token between
Feature activation+0.000
promotions
Token promotions
Feature activation+0.868
,
Token,
Feature activation+0.543
time
Token time
Feature activation+0.000
at
Token at
Feature activation+0.000
current
Token current
Feature activation+0.000
job
Token job
Feature activation+0.117
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
From
Token From
Feature activation+0.000
beating
Token beating
Feature activation+0.000
Jamaica
Token Jamaica
Feature activation+0.266
at
Token at
Feature activation+0.000
home
Token home
Feature activation+1.042
to
Token to
Feature activation+0.416
nearly
Token nearly
Feature activation+0.000
getting
Token getting
Feature activation+0.000
out
Token out
Feature activation+0.000
of
Token of
Feature activation+0.000
t
Tokent
Feature activation+0.000
until
Token until
Feature activation+0.000
10
Token 10
Feature activation+0.000
years
Token years
Feature activation+0.000
after
Token after
Feature activation+0.000
starting
Token starting
Feature activation+0.949
the
Token the
Feature activation+0.213
laundry
Token laundry
Feature activation+0.935
service
Token service
Feature activation+1.362
,
Token,
Feature activation+0.372
in
Token in
Feature activation+0.510

INTERVAL 0.568 - 0.851
CONTAINS 0.033%

âĢ
TokenâĢ
Feature activation+0.253
Ļ
TokenĻ
Feature activation+0.973
s
Tokens
Feature activation+0.022
engineering
Token engineering
Feature activation+0.000
talent
Token talent
Feature activation+1.100
to
Token to
Feature activation+0.639
draw
Token draw
Feature activation+0.000
more
Token more
Feature activation+0.000
than
Token than
Feature activation+0.000
a
Token a
Feature activation+0.000
dozen
Token dozen
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Use
TokenUse
Feature activation+0.000
the
Token the
Feature activation+0.000
stats
Token stats
Feature activation+0.000
we
Token we
Feature activation+0.000
provide
Token provide
Feature activation+0.790
you
Token you
Feature activation+0.489
in
Token in
Feature activation+0.186
order
Token order
Feature activation+0.214
to
Token to
Feature activation+0.000
improve
Token improve
Feature activation+0.000
compare
Token compare
Feature activation+0.000
the
Token the
Feature activation+0.000
result
Token result
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
computation
Token computation
Feature activation+0.572
to
Token to
Feature activation+0.000
nature
Token nature
Feature activation+0.000
,
Token,
Feature activation+0.000
with
Token with
Feature activation+0.000
experiment
Token experiment
Feature activation+0.000
F
Token F
Feature activation+0.000
ours
Tokenours
Feature activation+0.000
qu
Tokenqu
Feature activation+0.000
are
Tokenare
Feature activation+0.787
Dist
Token Dist
Feature activation+0.000
illery
Tokenillery
Feature activation+0.654
in
Token in
Feature activation+0.162
a
Token a
Feature activation+0.000
couple
Token couple
Feature activation+0.000
of
Token of
Feature activation+0.000
particularly
Token particularly
Feature activation+0.000
state
Tokenstate
Feature activation+0.000
companies
Token companies
Feature activation+0.000
with
Token with
Feature activation+0.000
no
Token no
Feature activation+0.000
physical
Token physical
Feature activation+0.000
presence
Token presence
Feature activation+0.717
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
state
Token state
Feature activation+0.522
to
Token to
Feature activation+0.365
collect
Token collect
Feature activation+0.000

INTERVAL 0.284 - 0.568
CONTAINS 0.098%

as
Token as
Feature activation+0.000
digital
Token digital
Feature activation+0.000
ste
Token ste
Feature activation+0.000
gan
Tokengan
Feature activation+0.000
ography
Tokenography
Feature activation+0.476
,
Token,
Feature activation+0.554
in
Token in
Feature activation+0.335
which
Token which
Feature activation+0.000
an
Token an
Feature activation+0.000
image
Token image
Feature activation+0.000
file
Token file
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
ll
Tokenll
Feature activation+0.000
have
Token have
Feature activation+0.000
until
Token until
Feature activation+0.000
Wednesday
Token Wednesday
Feature activation+0.000
night
Token night
Feature activation+0.367
!
Token!
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
On
TokenOn
Feature activation+0.000
a
Token a
Feature activation+0.000
a
Token a
Feature activation+0.000
quiet
Token quiet
Feature activation+0.000
section
Token section
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
car
Token car
Feature activation+0.303
park
Token park
Feature activation+0.404
,
Token,
Feature activation+0.000
sweep
Token sweep
Feature activation+0.000
aside
Token aside
Feature activation+0.000
the
Token the
Feature activation+0.000
and
Token and
Feature activation+0.000
taking
Token taking
Feature activation+0.000
care
Token care
Feature activation+0.000
of
Token of
Feature activation+0.011
my
Token my
Feature activation+0.000
body
Token body
Feature activation+0.384
and
Token and
Feature activation+0.000
trying
Token trying
Feature activation+0.000
to
Token to
Feature activation+0.000
get
Token get
Feature activation+0.000
back
Token back
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
good
Tokengood
Feature activation+0.000
name
Token name
Feature activation+0.697
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.383
before
Token before
Feature activation+0.416
signing
Token signing
Feature activation+0.000
off
Token off
Feature activation+0.000
that
Token that
Feature activation+0.000
he
Token he
Feature activation+0.000
had
Token had
Feature activation+0.000

INTERVAL 0.000 - 0.284
CONTAINS 99.848%

next
Token next
Feature activation+0.000
target
Token target
Feature activation+0.000
for
Token for
Feature activation+0.000
Iraq
Token Iraq
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
armed
Token armed
Feature activation+0.000
forces
Token forces
Feature activation+0.000
after
Token after
Feature activation+0.000
they
Token they
Feature activation+0.000
work
Token work
Feature activation+0.127
support
Token support
Feature activation+0.000
our
Token our
Feature activation+0.000
efforts
Token efforts
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
"
Token"
Feature activation+0.000
The
TokenThe
Feature activation+0.000
State
Token State
Feature activation+0.000
of
Token of
Feature activation+0.000
Texas
Token Texas
Feature activation+0.000
visiting
Token visiting
Feature activation+0.000
Union
Token Union
Feature activation+0.000
in
Token in
Feature activation+0.000
connection
Token connection
Feature activation+0.000
with
Token with
Feature activation+0.000
any
Token any
Feature activation+0.000
such
Token such
Feature activation+0.000
futile
Token futile
Feature activation+0.000
visit
Token visit
Feature activation+0.000
shall
Token shall
Feature activation+0.000
be
Token be
Feature activation+0.000
has
Token has
Feature activation+0.000
promised
Token promised
Feature activation+0.000
his
Token his
Feature activation+0.000
team
Token team
Feature activation+0.000
will
Token will
Feature activation+0.000
attack
Token attack
Feature activation+0.000
Stoke
Token Stoke
Feature activation+0.000
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
every
Token every
Feature activation+0.000
other
Token other
Feature activation+0.000
being
Token being
Feature activation+0.000
smashed
Token smashed
Feature activation+0.000
like
Token like
Feature activation+0.000
that
Token that
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
car
Token car
Feature activation+0.000
was
Token was
Feature activation+0.000
completely
Token completely
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 H7.4: (feature 5552)

TOP ACTIVATIONS
MAX = 4.319

them
Token them
Feature activation+0.000
students
Token students
Feature activation+1.098
at
Token at
Feature activation+0.000
British
Token British
Feature activation+0.000
universities
Token universities
Feature activation+2.409
,
Token,
Feature activation+4.319
were
Token were
Feature activation+0.126
lined
Token lined
Feature activation+0.000
up
Token up
Feature activation+0.000
along
Token along
Feature activation+0.000
the
Token the
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
m
Tokenm
Feature activation+0.000
omp
Tokenomp
Feature activation+0.000
rene
Tokenrene
Feature activation+0.000
urs
Tokenurs
Feature activation+0.000
,"
Token,"
Feature activation+3.450
met
Token met
Feature activation+0.000
by
Token by
Feature activation+0.000
visiting
Token visiting
Feature activation+0.000
the
Token the
Feature activation+0.000
same
Token same
Feature activation+0.000
.
Token.
Feature activation+0.000
C
Token C
Feature activation+0.000
.
Token.
Feature activation+0.000
Se
Token Se
Feature activation+0.000
gar
Tokengar
Feature activation+0.000
,
Token,
Feature activation+3.301
is
Token is
Feature activation+0.010
set
Token set
Feature activation+0.000
around
Token around
Feature activation+0.000
the
Token the
Feature activation+0.000
fictional
Token fictional
Feature activation+0.000
he
Token he
Feature activation+0.000
once
Token once
Feature activation+0.000
thought
Token thought
Feature activation+0.000
of
Token of
Feature activation+0.000
becoming
Token becoming
Feature activation+0.000
,
Token,
Feature activation+3.091
looked
Token looked
Feature activation+0.000
at
Token at
Feature activation+0.000
the
Token the
Feature activation+0.000
spot
Token spot
Feature activation+0.000
where
Token where
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
This
TokenThis
Feature activation+0.000
,
Token,
Feature activation+0.000
it
Token it
Feature activation+0.000
turns
Token turns
Feature activation+0.894
out
Token out
Feature activation+3.089
,
Token,
Feature activation+0.418
offended
Token offended
Feature activation+0.000
some
Token some
Feature activation+0.000
of
Token of
Feature activation+0.000
his
Token his
Feature activation+0.000
to
Token to
Feature activation+0.086
be
Token be
Feature activation+0.000
released
Token released
Feature activation+0.083
this
Token this
Feature activation+0.000
summer
Token summer
Feature activation+0.689
,
Token,
Feature activation+3.082
is
Token is
Feature activation+0.139
the
Token the
Feature activation+0.000
Phoenix
Token Phoenix
Feature activation+0.000
and
Token and
Feature activation+0.000
Cycl
Token Cycl
Feature activation+0.000
they
Token they
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
ident
Tokenident
Feature activation+0.000
ify
Tokenify
Feature activation+0.000
,
Token,
Feature activation+3.066
âĢ
TokenâĢ
Feature activation+0.770
Ŀ
TokenĿ
Feature activation+1.144
are
Token are
Feature activation+0.217
posting
Token posting
Feature activation+0.000
lists
Token lists
Feature activation+0.000
military
Token military
Feature activation+0.258
again
Token again
Feature activation+0.216
threatening
Token threatening
Feature activation+0.000
international
Token international
Feature activation+0.000
peace
Token peace
Feature activation+0.140
,
Token,
Feature activation+3.004
was
Token was
Feature activation+0.000
simply
Token simply
Feature activation+0.000
unacceptable
Token unacceptable
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
f
Tokenf
Feature activation+0.000
rak
Tokenrak
Feature activation+0.000
{
Token {
Feature activation+0.000
c
Tokenc
Feature activation+0.015
}}}
Token}}}
Feature activation+0.129
,
Token ,
Feature activation+2.957
implies
Token implies
Feature activation+0.000
that
Token that
Feature activation+0.000
c
Token c
Feature activation+0.000
=
Token =
Feature activation+0.000
â
Token â
Feature activation+0.000
of
Token of
Feature activation+0.000
them
Token them
Feature activation+0.000
,
Token,
Feature activation+0.000
it
Token it
Feature activation+0.000
turns
Token turns
Feature activation+0.000
out
Token out
Feature activation+2.944
,
Token,
Feature activation+0.000
can
Token can
Feature activation+0.000
be
Token be
Feature activation+0.000
driven
Token driven
Feature activation+0.000
by
Token by
Feature activation+0.000
der
Tokender
Feature activation+0.000
all
Tokenall
Feature activation+2.051
and
Token and
Feature activation+0.197
R
Token R
Feature activation+0.000
ital
Tokenital
Feature activation+0.000
in
Tokenin
Feature activation+2.915
,
Token,
Feature activation+1.487
are
Token are
Feature activation+0.000
primarily
Token primarily
Feature activation+0.000
prescribed
Token prescribed
Feature activation+0.000
to
Token to
Feature activation+0.000
,
Token,
Feature activation+0.000
like
Token like
Feature activation+0.000
Sp
Token Sp
Feature activation+0.000
os
Tokenos
Feature activation+0.000
ato
Tokenato
Feature activation+0.896
,
Token,
Feature activation+2.913
promised
Token promised
Feature activation+0.000
to
Token to
Feature activation+0.000
meet
Token meet
Feature activation+0.000
with
Token with
Feature activation+0.000
city
Token city
Feature activation+0.000
by
Tokenby
Feature activation+0.000
,
Token,
Feature activation+0.000
having
Token having
Feature activation+0.000
seen
Token seen
Feature activation+0.000
enough
Token enough
Feature activation+0.000
,
Token,
Feature activation+2.886
sa
Token sa
Feature activation+0.000
un
Tokenun
Feature activation+0.000
tered
Tokentered
Feature activation+0.000
out
Token out
Feature activation+0.000
of
Token of
Feature activation+0.000
employees
Token employees
Feature activation+0.000
sitting
Token sitting
Feature activation+0.000
around
Token around
Feature activation+0.014
a
Token a
Feature activation+0.000
desk
Token desk
Feature activation+0.756
,
Token,
Feature activation+2.868
appeared
Token appeared
Feature activation+0.000
un
Token un
Feature activation+0.000
al
Tokenal
Feature activation+0.000
tered
Tokentered
Feature activation+0.000
on
Token on
Feature activation+0.000
"
Token "
Feature activation+0.000
s
Tokens
Feature activation+0.000
ensible
Tokenensible
Feature activation+0.000
"
Token"
Feature activation+0.000
positions
Token positions
Feature activation+0.217
,
Token,
Feature activation+2.849
is
Token is
Feature activation+0.000
at
Token at
Feature activation+0.000
the
Token the
Feature activation+0.000
heart
Token heart
Feature activation+0.000
of
Token of
Feature activation+0.000
oured
Tokenoured
Feature activation+0.000
of
Token of
Feature activation+0.000
their
Token their
Feature activation+0.000
tough
Token tough
Feature activation+0.000
guys
Token guys
Feature activation+1.310
,
Token,
Feature activation+2.848
appreciate
Token appreciate
Feature activation+0.000
.
Token.
Feature activation+0.000
In
Token In
Feature activation+0.000
the
Token the
Feature activation+0.000
midst
Token midst
Feature activation+0.000
mind
Token mind
Feature activation+0.206
by
Token by
Feature activation+0.000
Wood
Token Wood
Feature activation+0.000
in
Tokenin
Feature activation+0.000
ville
Tokenville
Feature activation+0.000
,
Token,
Feature activation+2.791
Wash
Token Wash
Feature activation+0.000
.-
Token.-
Feature activation+0.055
based
Tokenbased
Feature activation+0.000
Red
Token Red
Feature activation+0.000
hook
Tokenhook
Feature activation+0.141
-
Token-
Feature activation+0.000
or
Token or
Feature activation+0.000
64
Token 64
Feature activation+0.060
-
Token-
Feature activation+0.776
bit
Tokenbit
Feature activation+1.527
,
Token,
Feature activation+2.771
Fo
Token Fo
Feature activation+0.000
obar
Tokenobar
Feature activation+0.000
2000
Token2000
Feature activation+0.191
supports
Token supports
Feature activation+0.000
the
Token the
Feature activation+0.000
its
Token its
Feature activation+0.000
Adren
Token Adren
Feature activation+0.000
o
Tokeno
Feature activation+0.000
330
Token 330
Feature activation+0.126
GPU
Token GPU
Feature activation+2.292
,
Token,
Feature activation+2.765
is
Token is
Feature activation+0.452
considerably
Token considerably
Feature activation+0.000
more
Token more
Feature activation+0.000
powerful
Token powerful
Feature activation+0.000
than
Token than
Feature activation+0.000
citing
Token citing
Feature activation+0.000
friends
Token friends
Feature activation+0.774
of
Token of
Feature activation+0.000
Ly
Token Ly
Feature activation+0.000
ne
Tokenne
Feature activation+0.000
,
Token,
Feature activation+2.738
reported
Token reported
Feature activation+0.000
that
Token that
Feature activation+0.000
she
Token she
Feature activation+0.000
had
Token had
Feature activation+0.000
met
Token met
Feature activation+0.000

Top DFA by src position
MAX = 3.926

By
Token By
Feature activation+0.009
Top resid features:
9
Token 9
Feature activation-0.001
Top resid features:
am
Tokenam
Feature activation-0.024
Top resid features:
the
Token the
Feature activation-0.014
Top resid features:
crowd
Token crowd
Feature activation+0.064
Top resid features:
,
Token,
Feature activation+3.926
Top resid features:
many
Token many
Feature activation+0.459
Top resid features:
of
Token of
Feature activation+0.026
Top resid features:
them
Token them
Feature activation+0.210
Top resid features:
students
Token students
Feature activation+0.024
Top resid features:
at
Token at
Feature activation-0.008
Top resid features:
.
Token.
Feature activation-0.008
Top resid features:
Ċ
TokenĊ
Feature activation-0.002
Top resid features:
Ċ
TokenĊ
Feature activation-0.005
Top resid features:
The
TokenThe
Feature activation+0.002
Top resid features:
women
Token women
Feature activation-0.014
Top resid features:
,
Token,
Feature activation+3.142
Top resid features:
who
Token who
Feature activation+0.764
Top resid features:
refer
Token refer
Feature activation+0.010
Top resid features:
to
Token to
Feature activation+0.012
Top resid features:
themselves
Token themselves
Feature activation+0.025
Top resid features:
as
Token as
Feature activation+0.019
Top resid features:
June
Token June
Feature activation+0.002
Top resid features:
19
Token 19
Feature activation+0.002
Top resid features:
.
Token.
Feature activation-0.010
Top resid features:
The
Token The
Feature activation+0.002
Top resid features:
film
Token film
Feature activation+0.110
Top resid features:
,
Token,
Feature activation+3.123
Top resid features:
based
Token based
Feature activation+0.320
Top resid features:
on
Token on
Feature activation+0.097
Top resid features:
the
Token the
Feature activation+0.020
Top resid features:
comic
Token comic
Feature activation+0.032
Top resid features:
strips
Token strips
Feature activation+0.048
Top resid features:
This
TokenThis
Feature activation-0.003
Top resid features:
was
Token was
Feature activation-0.002
Top resid features:
1995
Token 1995
Feature activation-0.002
Top resid features:
and
Token and
Feature activation-0.057
Top resid features:
Vaughan
Token Vaughan
Feature activation-0.004
Top resid features:
,
Token,
Feature activation+3.181
Top resid features:
a
Token a
Feature activation+0.274
Top resid features:
glac
Token glac
Feature activation+0.001
Top resid features:
iologist
Tokeniologist
Feature activation-0.018
Top resid features:
with
Token with
Feature activation-0.021
Top resid features:
the
Token the
Feature activation-0.004
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.116
Top resid features:
Ċ
TokenĊ
Feature activation-0.012
Top resid features:
Ċ
TokenĊ
Feature activation-0.003
Top resid features:
This
TokenThis
Feature activation-0.014
Top resid features:
,
Token,
Feature activation+1.894
Top resid features:
it
Token it
Feature activation+1.720
Top resid features:
turns
Token turns
Feature activation+0.018
Top resid features:
out
Token out
Feature activation-0.022
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
offended
Token offended
Feature activation+0.000
Top resid features:
releases
Token releases
Feature activation+0.040
Top resid features:
as
Token as
Feature activation-0.029
Top resid features:
the
Token the
Feature activation+0.004
Top resid features:
next
Token next
Feature activation-0.017
Top resid features:
set
Token set
Feature activation+0.027
Top resid features:
,
Token,
Feature activation+3.228
Top resid features:
to
Token to
Feature activation+0.028
Top resid features:
be
Token be
Feature activation+0.019
Top resid features:
released
Token released
Feature activation+0.054
Top resid features:
this
Token this
Feature activation+0.018
Top resid features:
summer
Token summer
Feature activation+0.132
Top resid features:
other
Tokenother
Feature activation+0.014
Top resid features:
kin
Tokenkin
Feature activation+0.004
Top resid features:
âĢ
TokenâĢ
Feature activation+0.001
Top resid features:
Ŀ
TokenĿ
Feature activation+0.007
Top resid features:
movement
Token movement
Feature activation+0.044
Top resid features:
,
Token,
Feature activation+1.774
Top resid features:
in
Token in
Feature activation+0.318
Top resid features:
which
Token which
Feature activation+1.086
Top resid features:
various
Token various
Feature activation+0.023
Top resid features:
losers
Token losers
Feature activation+0.016
Top resid features:
redef
Token redef
Feature activation+0.014
Top resid features:
the
Token the
Feature activation-0.011
Top resid features:
pre
Token pre
Feature activation+0.002
Top resid features:
war
Tokenwar
Feature activation-0.003
Top resid features:
status
Token status
Feature activation+0.001
Top resid features:
quo
Token quo
Feature activation-0.002
Top resid features:
,
Token,
Feature activation+3.018
Top resid features:
with
Token with
Feature activation+0.489
Top resid features:
a
Token a
Feature activation+0.027
Top resid features:
bellig
Token bellig
Feature activation+0.003
Top resid features:
erent
Tokenerent
Feature activation+0.003
Top resid features:
Japanese
Token Japanese
Feature activation-0.002
Top resid features:
).
Token).
Feature activation+0.002
Top resid features:
The
Token The
Feature activation-0.018
Top resid features:
continuum
Token continuum
Feature activation+0.017
Top resid features:
hypothesis
Token hypothesis
Feature activation+0.074
Top resid features:
,
Token,
Feature activation+1.426
Top resid features:
which
Token which
Feature activation+1.984
Top resid features:
asserts
Token asserts
Feature activation-0.026
Top resid features:
that
Token that
Feature activation-0.011
Top resid features:
there
Token there
Feature activation-0.001
Top resid features:
are
Token are
Feature activation-0.007
Top resid features:
no
Token no
Feature activation+0.011
Top resid features:
And
Token And
Feature activation-0.006
Top resid features:
both
Token both
Feature activation+0.004
Top resid features:
of
Token of
Feature activation-0.006
Top resid features:
them
Token them
Feature activation+0.021
Top resid features:
,
Token,
Feature activation+1.598
Top resid features:
it
Token it
Feature activation+1.813
Top resid features:
turns
Token turns
Feature activation+0.020
Top resid features:
out
Token out
Feature activation-0.011
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
can
Token can
Feature activation+0.000
Top resid features:
be
Token be
Feature activation+0.000
Top resid features:
.
Token.
Feature activation-0.003
Top resid features:
Stim
Token Stim
Feature activation+0.002
Top resid features:
ul
Tokenul
Feature activation+0.002
Top resid features:
ants
Tokenants
Feature activation+0.023
Top resid features:
,
Token,
Feature activation+0.316
Top resid features:
like
Token like
Feature activation+3.020
Top resid features:
Ad
Token Ad
Feature activation+0.008
Top resid features:
der
Tokender
Feature activation+0.002
Top resid features:
all
Tokenall
Feature activation+0.006
Top resid features:
and
Token and
Feature activation-0.003
Top resid features:
R
Token R
Feature activation+0.014
Top resid features:
Ċ
TokenĊ
Feature activation-0.010
Top resid features:
Ċ
TokenĊ
Feature activation-0.022
Top resid features:
Aren
TokenAren
Feature activation-0.003
Top resid features:
a
Tokena
Feature activation+0.005
Top resid features:
,
Token,
Feature activation+0.308
Top resid features:
like
Token like
Feature activation+3.122
Top resid features:
Sp
Token Sp
Feature activation-0.001
Top resid features:
os
Tokenos
Feature activation+0.008
Top resid features:
ato
Tokenato
Feature activation+0.014
Top resid features:
,
Token,
Feature activation+0.036
Top resid features:
promised
Token promised
Feature activation+0.000
Top resid features:
sitting
Token sitting
Feature activation+0.009
Top resid features:
.
Token.
Feature activation-0.032
Top resid features:
Bart
Token Bart
Feature activation+0.014
Top resid features:
le
Tokenle
Feature activation+0.004
Top resid features:
by
Tokenby
Feature activation+0.025
Top resid features:
,
Token,
Feature activation+2.598
Top resid features:
having
Token having
Feature activation+0.571
Top resid features:
seen
Token seen
Feature activation+0.017
Top resid features:
enough
Token enough
Feature activation+0.025
Top resid features:
,
Token,
Feature activation+0.022
Top resid features:
sa
Token sa
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.003
Top resid features:
Ċ
TokenĊ
Feature activation-0.002
Top resid features:
Ċ
TokenĊ
Feature activation-0.001
Top resid features:
The
TokenThe
Feature activation-0.014
Top resid features:
picture
Token picture
Feature activation+0.065
Top resid features:
,
Token,
Feature activation+1.837
Top resid features:
showing
Token showing
Feature activation+1.434
Top resid features:
employees
Token employees
Feature activation+0.006
Top resid features:
sitting
Token sitting
Feature activation+0.024
Top resid features:
around
Token around
Feature activation-0.033
Top resid features:
a
Token a
Feature activation+0.004
Top resid features:
Ċ
TokenĊ
Feature activation-0.013
Top resid features:
That
TokenThat
Feature activation-0.062
Top resid features:
last
Token last
Feature activation+0.048
Top resid features:
point
Token point
Feature activation+0.078
Top resid features:
,
Token,
Feature activation+1.184
Top resid features:
about
Token about
Feature activation+1.943
Top resid features:
Romney
Token Romney
Feature activation+0.024
Top resid features:
having
Token having
Feature activation+0.067
Top resid features:
"
Token "
Feature activation+0.010
Top resid features:
re
Tokenre
Feature activation+0.003
Top resid features:
vers
Tokenvers
Feature activation-0.001
Top resid features:
of
Token of
Feature activation-0.000
Top resid features:
logic
Token logic
Feature activation-0.005
Top resid features:
that
Token that
Feature activation-0.092
Top resid features:
Leafs
Token Leafs
Feature activation-0.001
Top resid features:
fans
Token fans
Feature activation+0.031
Top resid features:
,
Token,
Feature activation+2.909
Top resid features:
always
Token always
Feature activation+0.199
Top resid features:
en
Token en
Feature activation+0.018
Top resid features:
am
Tokenam
Feature activation+0.004
Top resid features:
oured
Tokenoured
Feature activation+0.116
Top resid features:
of
Token of
Feature activation+0.067
Top resid features:
representative
Token representative
Feature activation-0.014
Top resid features:
clarified
Token clarified
Feature activation-0.021
Top resid features:
that
Token that
Feature activation-0.026
Top resid features:
the
Token the
Feature activation-0.006
Top resid features:
beer
Token beer
Feature activation+0.004
Top resid features:
,
Token,
Feature activation+3.000
Top resid features:
developed
Token developed
Feature activation+0.259
Top resid features:
with
Token with
Feature activation+0.016
Top resid features:
Buffalo
Token Buffalo
Feature activation+0.008
Top resid features:
Wild
Token Wild
Feature activation+0.011
Top resid features:
Wings
Token Wings
Feature activation+0.002
Top resid features:
Windows
Token Windows
Feature activation+0.015
Top resid features:
versions
Token versions
Feature activation+0.055
Top resid features:
on
Token on
Feature activation+0.027
Top resid features:
the
Token the
Feature activation+0.017
Top resid features:
market
Token market
Feature activation+0.015
Top resid features:
,
Token,
Feature activation+2.867
Top resid features:
be
Token be
Feature activation+0.056
Top resid features:
they
Token they
Feature activation+0.168
Top resid features:
32
Token 32
Feature activation-0.004
Top resid features:
-
Token-
Feature activation+0.001
Top resid features:
or
Token or
Feature activation-0.007
Top resid features:
quad
Token quad
Feature activation+0.020
Top resid features:
-
Token-
Feature activation+0.001
Top resid features:
core
Tokencore
Feature activation+0.009
Top resid features:
So
Token So
Feature activation+0.000
Top resid features:
C
TokenC
Feature activation-0.001
Top resid features:
,
Token,
Feature activation+1.602
Top resid features:
with
Token with
Feature activation+1.151
Top resid features:
its
Token its
Feature activation+0.370
Top resid features:
Adren
Token Adren
Feature activation+0.036
Top resid features:
o
Tokeno
Feature activation-0.002
Top resid features:
330
Token 330
Feature activation-0.004
Top resid features:
.
Token.
Feature activation-0.012
Top resid features:
Ċ
TokenĊ
Feature activation-0.009
Top resid features:
Ċ
TokenĊ
Feature activation-0.016
Top resid features:
The
TokenThe
Feature activation-0.007
Top resid features:
station
Token station
Feature activation-0.018
Top resid features:
,
Token,
Feature activation+1.938
Top resid features:
citing
Token citing
Feature activation+1.470
Top resid features:
friends
Token friends
Feature activation+0.025
Top resid features:
of
Token of
Feature activation-0.049
Top resid features:
Ly
Token Ly
Feature activation+0.010
Top resid features:
ne
Tokenne
Feature activation-0.003
Top resid features:

Decoder Weights Distribution

Head 0: 0.05

Head 1: 0.02

Head 2: 0.03

Head 3: 0.09

Head 4: 0.57

Head 5: 0.03

Head 6: 0.03

Head 7: 0.02

Head 8: 0.06

Head 9: 0.05

Head 10: 0.02

Head 11: 0.03

Positive logits

,,3.15

*,3.03

?,2.76

,...2.68

,2.66

,[2.59

!,2.57

,.2.30

%,2.28

wo2.17

+,2.14

®,2.12

appears2.03

,2.03

suffers2.02

.,2.01

(),2.00

appeared1.98

othes1.92

(),1.88

Negative logits

anwhile-2.84

apest-2.63

hedon-2.33

',"-2.27

rahim-2.23

ESA-2.18

uador-2.18

inki-2.13

igslist-2.12

estern-2.10

iliar-2.06

uncture-2.04

clusive-2.04

othal-2.02

unal-2.02

udeau-2.00

],"-1.99

apist-1.99

FTWARE-1.98

),"-1.97

INTERVAL 3.887 - 4.319
CONTAINS 0.000%

them
Token them
Feature activation+0.000
students
Token students
Feature activation+1.098
at
Token at
Feature activation+0.000
British
Token British
Feature activation+0.000
universities
Token universities
Feature activation+2.409
,
Token,
Feature activation+4.319
were
Token were
Feature activation+0.126
lined
Token lined
Feature activation+0.000
up
Token up
Feature activation+0.000
along
Token along
Feature activation+0.000
the
Token the
Feature activation+0.000

INTERVAL 3.455 - 3.887
CONTAINS 0.000%

INTERVAL 3.023 - 3.455
CONTAINS 0.001%

he
Token he
Feature activation+0.000
once
Token once
Feature activation+0.000
thought
Token thought
Feature activation+0.000
of
Token of
Feature activation+0.000
becoming
Token becoming
Feature activation+0.000
,
Token,
Feature activation+3.091
looked
Token looked
Feature activation+0.000
at
Token at
Feature activation+0.000
the
Token the
Feature activation+0.000
spot
Token spot
Feature activation+0.000
where
Token where
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
This
TokenThis
Feature activation+0.000
,
Token,
Feature activation+0.000
it
Token it
Feature activation+0.000
turns
Token turns
Feature activation+0.894
out
Token out
Feature activation+3.089
,
Token,
Feature activation+0.418
offended
Token offended
Feature activation+0.000
some
Token some
Feature activation+0.000
of
Token of
Feature activation+0.000
his
Token his
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
m
Tokenm
Feature activation+0.000
omp
Tokenomp
Feature activation+0.000
rene
Tokenrene
Feature activation+0.000
urs
Tokenurs
Feature activation+0.000
,"
Token,"
Feature activation+3.450
met
Token met
Feature activation+0.000
by
Token by
Feature activation+0.000
visiting
Token visiting
Feature activation+0.000
the
Token the
Feature activation+0.000
same
Token same
Feature activation+0.000
.
Token.
Feature activation+0.000
C
Token C
Feature activation+0.000
.
Token.
Feature activation+0.000
Se
Token Se
Feature activation+0.000
gar
Tokengar
Feature activation+0.000
,
Token,
Feature activation+3.301
is
Token is
Feature activation+0.010
set
Token set
Feature activation+0.000
around
Token around
Feature activation+0.000
the
Token the
Feature activation+0.000
fictional
Token fictional
Feature activation+0.000
to
Token to
Feature activation+0.086
be
Token be
Feature activation+0.000
released
Token released
Feature activation+0.083
this
Token this
Feature activation+0.000
summer
Token summer
Feature activation+0.689
,
Token,
Feature activation+3.082
is
Token is
Feature activation+0.139
the
Token the
Feature activation+0.000
Phoenix
Token Phoenix
Feature activation+0.000
and
Token and
Feature activation+0.000
Cycl
Token Cycl
Feature activation+0.000

INTERVAL 2.591 - 3.023
CONTAINS 0.003%

Order
Token Order
Feature activation+0.000
,
Token,
Feature activation+0.000
on
Token on
Feature activation+0.000
the
Token the
Feature activation+0.010
other
Token other
Feature activation+0.837
hand
Token hand
Feature activation+2.646
,
Token,
Feature activation+0.252
is
Token is
Feature activation+0.000
far
Token far
Feature activation+0.000
bigger
Token bigger
Feature activation+0.000
than
Token than
Feature activation+0.000
-
Token-
Feature activation+0.000
or
Token or
Feature activation+0.000
64
Token 64
Feature activation+0.060
-
Token-
Feature activation+0.776
bit
Tokenbit
Feature activation+1.527
,
Token,
Feature activation+2.771
Fo
Token Fo
Feature activation+0.000
obar
Tokenobar
Feature activation+0.000
2000
Token2000
Feature activation+0.191
supports
Token supports
Feature activation+0.000
the
Token the
Feature activation+0.000
"
Token "
Feature activation+0.000
s
Tokens
Feature activation+0.000
ensible
Tokenensible
Feature activation+0.000
"
Token"
Feature activation+0.000
positions
Token positions
Feature activation+0.217
,
Token,
Feature activation+2.849
is
Token is
Feature activation+0.000
at
Token at
Feature activation+0.000
the
Token the
Feature activation+0.000
heart
Token heart
Feature activation+0.000
of
Token of
Feature activation+0.000
value
Token value
Feature activation+0.000
for
Token for
Feature activation+0.000
the
Token the
Feature activation+0.000
new
Token new
Feature activation+0.000
tax
Token tax
Feature activation+0.000
,
Token,
Feature activation+2.663
have
Token have
Feature activation+0.000
also
Token also
Feature activation+0.000
recruited
Token recruited
Feature activation+0.000
billionaire
Token billionaire
Feature activation+0.000
Jim
Token Jim
Feature activation+0.000
f
Tokenf
Feature activation+0.000
rak
Tokenrak
Feature activation+0.000
{
Token {
Feature activation+0.000
c
Tokenc
Feature activation+0.015
}}}
Token}}}
Feature activation+0.129
,
Token ,
Feature activation+2.957
implies
Token implies
Feature activation+0.000
that
Token that
Feature activation+0.000
c
Token c
Feature activation+0.000
=
Token =
Feature activation+0.000
â
Token â
Feature activation+0.000

INTERVAL 2.159 - 2.591
CONTAINS 0.009%

people
Token people
Feature activation+1.253
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+1.899
in
Token in
Feature activation+1.072
English
Token English
Feature activation+0.716
,
Token,
Feature activation+2.369
was
Token was
Feature activation+0.089
one
Token one
Feature activation+0.000
of
Token of
Feature activation+0.000
two
Token two
Feature activation+0.000
forwarded
Token forwarded
Feature activation+0.000
ĺ
Tokenĺ
Feature activation+0.080
Cho
TokenCho
Feature activation+0.000
osing
Tokenosing
Feature activation+0.000
wisely
Token wisely
Feature activation+0.155
âĢ
TokenâĢ
Feature activation+1.384
Ļ
TokenĻ
Feature activation+2.502
,
Token,
Feature activation+0.609
invited
Token invited
Feature activation+0.000
professional
Token professional
Feature activation+0.000
societies
Token societies
Feature activation+0.000
to
Token to
Feature activation+0.000
wealth
Token wealth
Feature activation+1.190
of
Token of
Feature activation+0.074
protected
Token protected
Feature activation+0.000
rivers
Token rivers
Feature activation+1.574
and
Token and
Feature activation+0.129
waterways
Token waterways
Feature activation+2.422
inland
Token inland
Feature activation+1.484
,
Token,
Feature activation+2.320
were
Token were
Feature activation+0.150
ideal
Token ideal
Feature activation+0.000
for
Token for
Feature activation+0.000
s
Tokens
Feature activation+0.000
son
Token son
Feature activation+0.000
plays
Token plays
Feature activation+0.000
for
Token for
Feature activation+0.000
City
Token City
Feature activation+0.454
,
Token,
Feature activation+2.337
jumped
Token jumped
Feature activation+0.000
to
Token to
Feature activation+0.000
his
Token his
Feature activation+0.000
defence
Token defence
Feature activation+0.000
.
Token.
Feature activation+0.000
it
Token it
Feature activation+0.000
falls
Token falls
Feature activation+0.000
on
Token on
Feature activation+0.000
small
Token small
Feature activation+0.000
businesses
Token businesses
Feature activation+0.618
,
Token,
Feature activation+2.188
may
Token may
Feature activation+0.000
turn
Token turn
Feature activation+0.000
out
Token out
Feature activation+0.000
to
Token to
Feature activation+0.000
be
Token be
Feature activation+0.000

INTERVAL 1.727 - 2.159
CONTAINS 0.018%

national
Token national
Feature activation+0.000
salary
Token salary
Feature activation+0.000
,
Token,
Feature activation+0.000
with
Token with
Feature activation+0.000
bonuses
Token bonuses
Feature activation+0.825
,
Token,
Feature activation+1.775
profit
Token profit
Feature activation+0.000
sharing
Token sharing
Feature activation+0.000
and
Token and
Feature activation+0.000
commissions
Token commissions
Feature activation+1.023
for
Token for
Feature activation+0.000
ideas
Token ideas
Feature activation+0.000
,
Token,
Feature activation+0.000
no
Token no
Feature activation+0.000
matter
Token matter
Feature activation+0.000
how
Token how
Feature activation+0.000
insane
Token insane
Feature activation+1.963
,
Token,
Feature activation+1.697
enter
Token enter
Feature activation+0.114
the
Token the
Feature activation+0.000
mainstream
Token mainstream
Feature activation+0.000
debate
Token debate
Feature activation+0.000
even
Token even
Feature activation+0.000
negative
Token negative
Feature activation+0.000
-
Token-
Feature activation+0.000
em
Tokenem
Feature activation+0.000
ission
Tokenission
Feature activation+1.722
technologies
Token technologies
Feature activation+1.910
,
Token,
Feature activation+1.189
received
Token received
Feature activation+0.000
no
Token no
Feature activation+0.000
direct
Token direct
Feature activation+0.000
reference
Token reference
Feature activation+0.000
on
Token on
Feature activation+0.148
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.020
Mal
TokenMal
Feature activation+0.000
colm
Tokencolm
Feature activation+0.000
,
Token,
Feature activation+2.025
âĢ
TokenâĢ
Feature activation+0.321
Ŀ
TokenĿ
Feature activation+1.292
said
Token said
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
violence
Token violence
Feature activation+0.000
.
Token.
Feature activation+0.000
That
Token That
Feature activation+0.000
,
Token,
Feature activation+0.000
it
Token it
Feature activation+0.000
seems
Token seems
Feature activation+1.783
,
Token,
Feature activation+0.722
is
Token is
Feature activation+0.000
insufficient
Token insufficient
Feature activation+0.000
ly
Tokenly
Feature activation+0.000
ideological
Token ideological
Feature activation+0.000

INTERVAL 1.296 - 1.727
CONTAINS 0.038%

to
Token to
Feature activation+0.000
acknowledge
Token acknowledge
Feature activation+0.000
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
even
Token even
Feature activation+0.000
consider
Token consider
Feature activation+1.598
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
importance
Token importance
Feature activation+0.000
of
Token of
Feature activation+0.306
leading
Token leading
Feature activation+0.000
anti
Token anti
Feature activation+0.000
-
Token-
Feature activation+0.000
gay
Tokengay
Feature activation+0.000
marriage
Token marriage
Feature activation+0.000
bill
Token bill
Feature activation+0.304
,
Token,
Feature activation+1.365
demonstrate
Token demonstrate
Feature activation+0.000
outside
Token outside
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
9
Token 9
Feature activation+0.000
the
Token the
Feature activation+0.020
Winchester
Token Winchester
Feature activation+0.000
expert
Token expert
Feature activation+0.902
on
Token on
Feature activation+0.000
rifles
Token rifles
Feature activation+0.006
,
Token,
Feature activation+1.527
discussing
Token discussing
Feature activation+0.000
the
Token the
Feature activation+0.000
finer
Token finer
Feature activation+0.000
points
Token points
Feature activation+0.172
of
Token of
Feature activation+0.000
tax
Token tax
Feature activation+0.888
to
Token to
Feature activation+0.000
40
Token 40
Feature activation+0.000
%
Token%
Feature activation+0.058
each
Token each
Feature activation+0.407
,
Token,
Feature activation+1.495
would
Token would
Feature activation+0.000
fund
Token fund
Feature activation+0.000
this
Token this
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Introduction
TokenIntroduction
Feature activation+0.007
to
Token to
Feature activation+0.000
Computer
Token Computer
Feature activation+0.000
Science
Token Science
Feature activation+0.000
I
Token I
Feature activation+0.000
,
Token,
Feature activation+1.669
âĢ
TokenâĢ
Feature activation+0.437
Ŀ
TokenĿ
Feature activation+0.995
attracted
Token attracted
Feature activation+0.000
a
Token a
Feature activation+0.000
record
Token record
Feature activation+0.000

INTERVAL 0.864 - 1.296
CONTAINS 0.083%

Dallas
Token Dallas
Feature activation+0.316
and
Token and
Feature activation+0.362
Attorney
Token Attorney
Feature activation+0.000
General
Token General
Feature activation+0.000
Nathan
Token Nathan
Feature activation+0.000
Clifford
Token Clifford
Feature activation+0.899
paid
Token paid
Feature activation+0.000
a
Token a
Feature activation+0.000
visit
Token visit
Feature activation+0.000
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
s
Tokens
Feature activation+0.000
conservative
Token conservative
Feature activation+0.000
attack
Token attack
Feature activation+0.000
machine
Token machine
Feature activation+0.000
,
Token,
Feature activation+0.000
,
Token ,
Feature activation+1.064
was
Token was
Feature activation+0.000
born
Token born
Feature activation+0.000
.
Token.
Feature activation+0.000
Bonus
Token Bonus
Feature activation+0.000
:
Token:
Feature activation+0.000
housing
Token housing
Feature activation+0.000
,
Token,
Feature activation+0.000
Martin
Token Martin
Feature activation+0.000
Co
Token Co
Feature activation+0.000
ite
Tokenite
Feature activation+0.000
ux
Tokenux
Feature activation+0.885
,
Token,
Feature activation+0.170
will
Token will
Feature activation+0.000
complete
Token complete
Feature activation+0.000
his
Token his
Feature activation+0.000
revision
Token revision
Feature activation+0.000
suspicious
Token suspicious
Feature activation+0.000
of
Token of
Feature activation+0.000
abusive
Token abusive
Feature activation+0.000
treatment
Token treatment
Feature activation+0.945
towards
Token towards
Feature activation+0.000
prisoners
Token prisoners
Feature activation+1.028
,
Token,
Feature activation+0.918
confiscated
Token confiscated
Feature activation+0.000
one
Token one
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
California
Token California
Feature activation+0.000
-
Token-
Feature activation+0.000
wide
Tokenwide
Feature activation+0.000
student
Token student
Feature activation+0.000
group
Token group
Feature activation+0.550
,
Token,
Feature activation+1.130
said
Token said
Feature activation+0.000
the
Token the
Feature activation+0.000
UC
Token UC
Feature activation+0.000
system
Token system
Feature activation+0.000
had
Token had
Feature activation+0.000

INTERVAL 0.432 - 0.864
CONTAINS 0.188%

before
Token before
Feature activation+0.026
the
Token the
Feature activation+0.000
mon
Token mon
Feature activation+0.000
so
Tokenso
Feature activation+0.000
ons
Tokenons
Feature activation+0.000
,
Token,
Feature activation+0.625
has
Token has
Feature activation+0.000
just
Token just
Feature activation+0.000
begun
Token begun
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
period
Token period
Feature activation+0.000
,
Token,
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
1950
Token 1950
Feature activation+0.000
s
Tokens
Feature activation+0.467
and
Token and
Feature activation+0.000
1960
Token 1960
Feature activation+0.000
s
Tokens
Feature activation+1.158
,
Token,
Feature activation+0.499
when
Token when
Feature activation+0.000
we
Token we
Feature activation+0.000
want
Token want
Feature activation+0.000
to
Token to
Feature activation+0.000
face
Token face
Feature activation+0.000
together
Token together
Feature activation+0.000
,
Token,
Feature activation+0.529
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
Ms
Token Ms
Feature activation+0.000
K
Token K
Feature activation+0.000
oci
Tokenoci
Feature activation+0.000
director
Token director
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
International
Token International
Feature activation+0.000
Monetary
Token Monetary
Feature activation+0.000
Fund
Token Fund
Feature activation+0.468
,
Token,
Feature activation+0.000
has
Token has
Feature activation+0.000
called
Token called
Feature activation+0.000
for
Token for
Feature activation+0.000
a
Token a
Feature activation+0.000
at
Token at
Feature activation+0.000
Col
Token Col
Feature activation+0.000
chester
Tokenchester
Feature activation+0.000
general
Token general
Feature activation+0.000
hospital
Token hospital
Feature activation+0.595
,
Token,
Feature activation+0.589
said
Token said
Feature activation+0.000
his
Token his
Feature activation+0.000
daughter
Token daughter
Feature activation+0.000
had
Token had
Feature activation+0.000
spent
Token spent
Feature activation+0.000

INTERVAL 0.000 - 0.432
CONTAINS 99.661%

"
Token "
Feature activation+0.000
ge
Tokenge
Feature activation+0.000
ek
Tokenek
Feature activation+0.000
break
Token break
Feature activation+0.000
f
Tokenf
Feature activation+0.000
asts
Tokenasts
Feature activation+0.000
"
Token"
Feature activation+0.000
or
Token or
Feature activation+0.000
in
Token in
Feature activation+0.000
new
Token new
Feature activation+0.000
"
Token "
Feature activation+0.000
chart
Token chart
Feature activation+0.000
and
Token and
Feature activation+0.000
info
Token info
Feature activation+0.000
here
Token here
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
H
TokenH
Feature activation+0.000
ulk
Tokenulk
Feature activation+0.000
Hogan
Token Hogan
Feature activation+0.000
:
Token:
Feature activation+0.000
one
Token one
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
t
Tokent
Feature activation+0.000
entirely
Token entirely
Feature activation+0.000
see
Token see
Feature activation+0.000
eye
Token eye
Feature activation+0.000
-
Token-
Feature activation+0.000
to
Tokento
Feature activation+0.000
-
Token-
Feature activation+0.000
eye
Tokeneye
Feature activation+0.000
on
Token on
Feature activation+0.000
threats
Token threats
Feature activation+0.000
k
Tokenk
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
S
TokenS
Feature activation+0.000
ai
Tokenai
Feature activation+0.000
is
Token is
Feature activation+0.000
unstoppable
Token unstoppable
Feature activation+0.000
during
Token during
Feature activation+0.000
both
Token both
Feature activation+0.000
the
Token the
Feature activation+0.000
cast
Token cast
Feature activation+0.000
be
Token be
Feature activation+0.000
able
Token able
Feature activation+0.000
to
Token to
Feature activation+0.000
control
Token control
Feature activation+0.000
the
Token the
Feature activation+0.000
temperature
Token temperature
Feature activation+0.000
of
Token of
Feature activation+0.000
any
Token any
Feature activation+0.000
coil
Token coil
Feature activation+0.000
called
Token called
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 4 in H7.4: (feature 38779)

TOP ACTIVATIONS
MAX = 4.624

by
Token by
Feature activation+0.000
a
Token a
Feature activation+0.000
23
Token 23
Feature activation+0.000
-
Token-
Feature activation+0.000
9
Token9
Feature activation+1.060
vote
Token vote
Feature activation+4.624
legislation
Token legislation
Feature activation+0.412
out
Token out
Feature activation+0.000
-
Token-
Feature activation+0.000
of
Tokenof
Feature activation+0.000
-
Token-
Feature activation+0.000
by
Token by
Feature activation+0.000
a
Token a
Feature activation+0.000
23
Token 23
Feature activation+0.000
-
Token-
Feature activation+0.000
9
Token9
Feature activation+0.723
vote
Token vote
Feature activation+4.433
legislation
Token legislation
Feature activation+0.324
out
Token out
Feature activation+0.000
-
Token-
Feature activation+0.000
of
Tokenof
Feature activation+0.000
-
Token-
Feature activation+0.000
and
Token and
Feature activation+0.000
it
Token it
Feature activation+0.000
has
Token has
Feature activation+0.000
removed
Token removed
Feature activation+0.000
from
Token from
Feature activation+0.000
education
Token education
Feature activation+4.283
the
Token the
Feature activation+0.168
,
Token,
Feature activation+0.000
I
Token I
Feature activation+0.000
guess
Token guess
Feature activation+0.000
what
Token what
Feature activation+0.000
High
Token High
Feature activation+0.000
Court
Token Court
Feature activation+3.922
(
Token (
Feature activation+0.000
I
TokenI
Feature activation+0.000
HC
TokenHC
Feature activation+0.000
)
Token)
Feature activation+4.170
a
Token a
Feature activation+0.000
four
Token four
Feature activation+0.000
-
Token-
Feature activation+0.000
phase
Tokenphase
Feature activation+0.000
plan
Token plan
Feature activation+0.000
submitted
Token submitted
Feature activation+0.000
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
Islamabad
Token Islamabad
Feature activation+0.000
High
Token High
Feature activation+0.000
Court
Token Court
Feature activation+3.922
(
Token (
Feature activation+0.000
I
TokenI
Feature activation+0.000
HC
TokenHC
Feature activation+0.000
)
Token)
Feature activation+4.170
a
Token a
Feature activation+0.000
razor
Token razor
Feature activation+0.000
:
Token:
Feature activation+0.000
Never
Token Never
Feature activation+0.000
attribute
Token attribute
Feature activation+0.000
to
Token to
Feature activation+0.000
malice
Token malice
Feature activation+3.917
that
Token that
Feature activation+0.497
which
Token which
Feature activation+0.000
can
Token can
Feature activation+0.000
be
Token be
Feature activation+0.000
adequately
Token adequately
Feature activation+0.047
can
Token can
Feature activation+0.000
eventually
Token eventually
Feature activation+0.000
help
Token help
Feature activation+0.000
recreate
Token recreate
Feature activation+0.000
in
Token in
Feature activation+0.000
monkeys
Token monkeys
Feature activation+3.842
such
Token such
Feature activation+0.075
diseases
Token diseases
Feature activation+0.822
as
Token as
Feature activation+0.000
Parkinson
Token Parkinson
Feature activation+0.000
's
Token's
Feature activation+0.000
also
Token also
Feature activation+0.000
imported
Token imported
Feature activation+0.000
from
Token from
Feature activation+0.000
the
Token the
Feature activation+0.000
phone
Token phone
Feature activation+0.110
ecosystem
Token ecosystem
Feature activation+3.823
the
Token the
Feature activation+0.000
power
Token power
Feature activation+0.000
to
Token to
Feature activation+0.000
regulate
Token regulate
Feature activation+0.000
rates
Token rates
Feature activation+0.000
reaff
Token reaff
Feature activation+0.000
irm
Tokenirm
Feature activation+0.000
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
American
Token American
Feature activation+0.000
people
Token people
Feature activation+3.683
the
Token the
Feature activation+0.000
party
Token party
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
included
Token included
Feature activation+0.000
as
Token as
Feature activation+0.000
a
Token a
Feature activation+0.000
sub
Token sub
Feature activation+0.000
-
Token-
Feature activation+0.000
category
Tokencategory
Feature activation+3.678
which
Token which
Feature activation+0.414
of
Token of
Feature activation+0.000
those
Token those
Feature activation+0.000
constituted
Token constituted
Feature activation+0.000
major
Token major
Feature activation+0.000
5
Token 5
Feature activation+0.000
-
Token-
Feature activation+0.000
year
Tokenyear
Feature activation+0.000
-
Token-
Feature activation+0.000
old
Tokenold
Feature activation+2.149
students
Token students
Feature activation+3.480
a
Token a
Feature activation+0.000
sense
Token sense
Feature activation+0.000
of
Token of
Feature activation+0.000
patriotism
Token patriotism
Feature activation+0.000
with
Token with
Feature activation+0.000
-
Token-
Feature activation+0.000
sym
Tokensym
Feature activation+0.000
?
Token?
Feature activation+0.000
accepts
Token accepts
Feature activation+0.000
as
Token as
Feature activation+0.000
arguments
Token arguments
Feature activation+3.265
an
Token an
Feature activation+0.000
S
Token S
Feature activation+0.000
-
Token-
Feature activation+0.000
expression
Tokenexpression
Feature activation+0.018
(
Token (
Feature activation+0.000
the
Token the
Feature activation+0.000
justices
Token justices
Feature activation+0.000
dismissed
Token dismissed
Feature activation+0.000
,
Token,
Feature activation+0.000
without
Token without
Feature activation+0.000
comment
Token comment
Feature activation+3.200
,
Token,
Feature activation+0.602
Pf
Token Pf
Feature activation+0.000
izer
Tokenizer
Feature activation+0.312
's
Token's
Feature activation+0.000
claim
Token claim
Feature activation+0.225
to
Token to
Feature activation+0.000
show
Token show
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
Sample
Token Sample
Feature activation+0.074
Analysis
Token Analysis
Feature activation+3.030
at
Token at
Feature activation+0.000
Mars
Token Mars
Feature activation+1.852
instrument
Token instrument
Feature activation+2.628
that
Token that
Feature activation+0.000
there
Token there
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
incre
Tokenincre
Feature activation+0.000
ase
Tokenase
Feature activation+0.000
among
Token among
Feature activation+0.000
Catholics
Token Catholics
Feature activation+2.911
the
Token the
Feature activation+0.000
awareness
Token awareness
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
importance
Token importance
Feature activation+0.000
is
Token is
Feature activation+0.000
to
Token to
Feature activation+0.000
take
Token take
Feature activation+0.000
as
Token as
Feature activation+0.000
the
Token the
Feature activation+0.890
target
Token target
Feature activation+2.802
the
Token the
Feature activation+0.762
present
Token present
Feature activation+0.380
demand
Token demand
Feature activation+0.684
.
Token.
Feature activation+0.493
This
Token This
Feature activation+0.000
has
Token has
Feature activation+0.000
discussed
Token discussed
Feature activation+0.000
with
Token with
Feature activation+0.000
several
Token several
Feature activation+0.000
petrol
Token petrol
Feature activation+0.000
stations
Token stations
Feature activation+2.790
working
Token working
Feature activation+0.313
across
Token across
Feature activation+0.000
the
Token the
Feature activation+0.000
country
Token country
Feature activation+1.127
cance
Token cance
Feature activation+0.000
today
Token today
Feature activation+0.000
,
Token,
Feature activation+0.000
brings
Token brings
Feature activation+0.000
to
Token to
Feature activation+0.000
a
Token a
Feature activation+0.000
close
Token close
Feature activation+2.722
months
Token months
Feature activation+0.000
of
Token of
Feature activation+0.000
behind
Token behind
Feature activation+0.000
-
Token-
Feature activation+0.000
the
Tokenthe
Feature activation+0.000
to
Token to
Feature activation+0.000
discuss
Token discuss
Feature activation+0.000
with
Token with
Feature activation+0.000
their
Token their
Feature activation+0.000
Chinese
Token Chinese
Feature activation+0.000
counterparts
Token counterparts
Feature activation+2.718
what
Token what
Feature activation+0.000
happens
Token happens
Feature activation+0.000
when
Token when
Feature activation+0.000
,
Token,
Feature activation+0.000
for
Token for
Feature activation+0.000
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
presence
Token presence
Feature activation+0.000
in
Token in
Feature activation+0.000
my
Token my
Feature activation+0.000
community
Token community
Feature activation+2.683
of
Token of
Feature activation+0.000
a
Token a
Feature activation+0.000
company
Token company
Feature activation+0.000
that
Token that
Feature activation+0.000
participates
Token participates
Feature activation+0.000

Top DFA by src position
MAX = 4.912

The
Token The
Feature activation-0.015
Top resid features:
Senate
Token Senate
Feature activation+0.008
Top resid features:
on
Token on
Feature activation+0.020
Top resid features:
Monday
Token Monday
Feature activation+0.043
Top resid features:
approved
Token approved
Feature activation+0.030
Top resid features:
by
Token by
Feature activation+4.912
Top resid features:
a
Token a
Feature activation+0.222
Top resid features:
23
Token 23
Feature activation+0.002
Top resid features:
-
Token-
Feature activation-0.005
Top resid features:
9
Token9
Feature activation+0.056
Top resid features:
vote
Token vote
Feature activation+0.006
Top resid features:
The
TokenThe
Feature activation-0.005
Top resid features:
Senate
Token Senate
Feature activation+0.011
Top resid features:
on
Token on
Feature activation+0.035
Top resid features:
Monday
Token Monday
Feature activation+0.035
Top resid features:
approved
Token approved
Feature activation+0.033
Top resid features:
by
Token by
Feature activation+4.808
Top resid features:
a
Token a
Feature activation+0.186
Top resid features:
23
Token 23
Feature activation+0.006
Top resid features:
-
Token-
Feature activation-0.005
Top resid features:
9
Token9
Feature activation+0.049
Top resid features:
vote
Token vote
Feature activation+0.002
Top resid features:
,
Token,
Feature activation+0.005
Top resid features:
and
Token and
Feature activation-0.014
Top resid features:
it
Token it
Feature activation-0.006
Top resid features:
has
Token has
Feature activation-0.010
Top resid features:
removed
Token removed
Feature activation+0.101
Top resid features:
from
Token from
Feature activation+4.847
Top resid features:
education
Token education
Feature activation+0.029
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
I
Token I
Feature activation+0.000
Top resid features:
guess
Token guess
Feature activation+0.000
Top resid features:
DA
TokenDA
Feature activation-0.001
Top resid features:
)
Token)
Feature activation-0.000
Top resid features:
on
Token on
Feature activation+0.008
Top resid features:
Friday
Token Friday
Feature activation-0.018
Top resid features:
submitted
Token submitted
Feature activation+0.088
Top resid features:
to
Token to
Feature activation+4.134
Top resid features:
the
Token the
Feature activation+0.476
Top resid features:
Islamabad
Token Islamabad
Feature activation+0.042
Top resid features:
High
Token High
Feature activation+0.002
Top resid features:
Court
Token Court
Feature activation+0.077
Top resid features:
(
Token (
Feature activation+0.023
Top resid features:
DA
TokenDA
Feature activation-0.001
Top resid features:
)
Token)
Feature activation-0.004
Top resid features:
on
Token on
Feature activation+0.017
Top resid features:
Friday
Token Friday
Feature activation-0.010
Top resid features:
submitted
Token submitted
Feature activation+0.064
Top resid features:
to
Token to
Feature activation+3.821
Top resid features:
the
Token the
Feature activation+0.502
Top resid features:
Islamabad
Token Islamabad
Feature activation+0.083
Top resid features:
High
Token High
Feature activation+0.007
Top resid features:
Court
Token Court
Feature activation+0.066
Top resid features:
(
Token (
Feature activation+0.000
Top resid features:
s
Tokens
Feature activation-0.007
Top resid features:
razor
Token razor
Feature activation-0.003
Top resid features:
:
Token:
Feature activation-0.050
Top resid features:
Never
Token Never
Feature activation+0.021
Top resid features:
attribute
Token attribute
Feature activation+0.005
Top resid features:
to
Token to
Feature activation+4.512
Top resid features:
malice
Token malice
Feature activation+0.067
Top resid features:
that
Token that
Feature activation+0.000
Top resid features:
which
Token which
Feature activation+0.000
Top resid features:
can
Token can
Feature activation+0.000
Top resid features:
be
Token be
Feature activation+0.000
Top resid features:
method
Token method
Feature activation-0.003
Top resid features:
can
Token can
Feature activation-0.016
Top resid features:
eventually
Token eventually
Feature activation-0.027
Top resid features:
help
Token help
Feature activation-0.036
Top resid features:
recreate
Token recreate
Feature activation+0.084
Top resid features:
in
Token in
Feature activation+4.354
Top resid features:
monkeys
Token monkeys
Feature activation+0.053
Top resid features:
such
Token such
Feature activation+0.000
Top resid features:
diseases
Token diseases
Feature activation+0.000
Top resid features:
as
Token as
Feature activation+0.000
Top resid features:
Parkinson
Token Parkinson
Feature activation+0.000
Top resid features:
.
Token.
Feature activation-0.025
Top resid features:
But
Token But
Feature activation-0.039
Top resid features:
it
Token it
Feature activation+0.000
Top resid features:
also
Token also
Feature activation-0.002
Top resid features:
imported
Token imported
Feature activation+0.043
Top resid features:
from
Token from
Feature activation+4.152
Top resid features:
the
Token the
Feature activation+0.292
Top resid features:
phone
Token phone
Feature activation+0.006
Top resid features:
ecosystem
Token ecosystem
Feature activation-0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
power
Token power
Feature activation+0.000
Top resid features:
an
Token an
Feature activation+0.014
Top resid features:
attempt
Token attempt
Feature activation+0.010
Top resid features:
to
Token to
Feature activation+0.002
Top resid features:
reaff
Token reaff
Feature activation-0.008
Top resid features:
irm
Tokenirm
Feature activation-0.005
Top resid features:
to
Token to
Feature activation+3.868
Top resid features:
the
Token the
Feature activation+0.328
Top resid features:
American
Token American
Feature activation+0.027
Top resid features:
people
Token people
Feature activation+0.063
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
party
Token party
Feature activation+0.000
Top resid features:
years
Token years
Feature activation+0.004
Top resid features:
,
Token,
Feature activation-0.001
Top resid features:
and
Token and
Feature activation-0.034
Top resid features:
also
Token also
Feature activation+0.000
Top resid features:
included
Token included
Feature activation+0.109
Top resid features:
as
Token as
Feature activation+3.985
Top resid features:
a
Token a
Feature activation+0.175
Top resid features:
sub
Token sub
Feature activation+0.048
Top resid features:
-
Token-
Feature activation-0.002
Top resid features:
category
Tokencategory
Feature activation+0.028
Top resid features:
which
Token which
Feature activation+0.000
Top resid features:
kindergarten
Token kindergarten
Feature activation-0.002
Top resid features:
aims
Token aims
Feature activation-0.057
Top resid features:
to
Token to
Feature activation-0.131
Top resid features:
inst
Token inst
Feature activation-0.002
Top resid features:
ill
Tokenill
Feature activation+0.012
Top resid features:
in
Token in
Feature activation+3.511
Top resid features:
its
Token its
Feature activation+0.775
Top resid features:
3
Token 3
Feature activation+0.004
Top resid features:
-
Token-
Feature activation+0.000
Top resid features:
to
Token to
Feature activation+0.009
Top resid features:
5
Token 5
Feature activation-0.001
Top resid features:
has
Token has
Feature activation-0.021
Top resid features:
-
Token-
Feature activation+0.025
Top resid features:
sym
Tokensym
Feature activation-0.001
Top resid features:
?
Token?
Feature activation+0.012
Top resid features:
accepts
Token accepts
Feature activation+0.008
Top resid features:
as
Token as
Feature activation+3.729
Top resid features:
arguments
Token arguments
Feature activation+0.194
Top resid features:
an
Token an
Feature activation+0.000
Top resid features:
S
Token S
Feature activation+0.000
Top resid features:
-
Token-
Feature activation+0.000
Top resid features:
expression
Tokenexpression
Feature activation+0.000
Top resid features:
on
Token on
Feature activation+0.013
Top resid features:
Tuesday
Token Tuesday
Feature activation-0.023
Top resid features:
the
Token the
Feature activation-0.000
Top resid features:
justices
Token justices
Feature activation-0.015
Top resid features:
dismissed
Token dismissed
Feature activation+0.017
Top resid features:
,
Token,
Feature activation+2.974
Top resid features:
without
Token without
Feature activation+0.802
Top resid features:
comment
Token comment
Feature activation+0.068
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
Pf
Token Pf
Feature activation+0.000
Top resid features:
izer
Tokenizer
Feature activation+0.000
Top resid features:
y
Tokeny
Feature activation-0.001
Top resid features:
was
Token was
Feature activation+0.014
Top resid features:
able
Token able
Feature activation-0.009
Top resid features:
to
Token to
Feature activation-0.003
Top resid features:
show
Token show
Feature activation+0.000
Top resid features:
in
Token in
Feature activation+3.126
Top resid features:
the
Token the
Feature activation+0.533
Top resid features:
Sample
Token Sample
Feature activation+0.146
Top resid features:
Analysis
Token Analysis
Feature activation+0.077
Top resid features:
at
Token at
Feature activation+0.000
Top resid features:
Mars
Token Mars
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.007
Top resid features:
Ċ
TokenĊ
Feature activation+0.010
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
incre
Tokenincre
Feature activation+0.001
Top resid features:
ase
Tokenase
Feature activation+0.011
Top resid features:
among
Token among
Feature activation+3.260
Top resid features:
Catholics
Token Catholics
Feature activation+0.073
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
awareness
Token awareness
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.223
Top resid features:
approach
Token approach
Feature activation-0.003
Top resid features:
is
Token is
Feature activation-0.006
Top resid features:
to
Token to
Feature activation-0.001
Top resid features:
take
Token take
Feature activation+0.018
Top resid features:
as
Token as
Feature activation+2.499
Top resid features:
the
Token the
Feature activation+0.748
Top resid features:
target
Token target
Feature activation+0.151
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
present
Token present
Feature activation+0.000
Top resid features:
demand
Token demand
Feature activation+0.000
Top resid features:
Ministry
Token Ministry
Feature activation+0.007
Top resid features:
of
Token of
Feature activation-0.001
Top resid features:
Economy
Token Economy
Feature activation-0.001
Top resid features:
has
Token has
Feature activation-0.023
Top resid features:
discussed
Token discussed
Feature activation+0.079
Top resid features:
with
Token with
Feature activation+3.301
Top resid features:
several
Token several
Feature activation+0.123
Top resid features:
petrol
Token petrol
Feature activation+0.012
Top resid features:
stations
Token stations
Feature activation+0.049
Top resid features:
working
Token working
Feature activation+0.000
Top resid features:
across
Token across
Feature activation+0.000
Top resid features:
announced
Token announced
Feature activation+0.005
Top resid features:
later
Token later
Feature activation+0.004
Top resid features:
today
Token today
Feature activation+0.014
Top resid features:
,
Token,
Feature activation+0.007
Top resid features:
brings
Token brings
Feature activation+0.439
Top resid features:
to
Token to
Feature activation+2.636
Top resid features:
a
Token a
Feature activation+0.278
Top resid features:
close
Token close
Feature activation+0.044
Top resid features:
months
Token months
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
behind
Token behind
Feature activation+0.000
Top resid features:
decades
Token decades
Feature activation-0.008
Top resid features:
,
Token,
Feature activation-0.017
Top resid features:
tried
Token tried
Feature activation-0.010
Top resid features:
to
Token to
Feature activation-0.046
Top resid features:
discuss
Token discuss
Feature activation+0.045
Top resid features:
with
Token with
Feature activation+3.079
Top resid features:
their
Token their
Feature activation+0.189
Top resid features:
Chinese
Token Chinese
Feature activation-0.001
Top resid features:
counterparts
Token counterparts
Feature activation+0.074
Top resid features:
what
Token what
Feature activation+0.000
Top resid features:
happens
Token happens
Feature activation+0.000
Top resid features:
Durham
Token Durham
Feature activation+0.004
Top resid features:
resident
Token resident
Feature activation-0.013
Top resid features:
,
Token,
Feature activation+0.001
Top resid features:
the
Token the
Feature activation+0.012
Top resid features:
presence
Token presence
Feature activation-0.124
Top resid features:
in
Token in
Feature activation+2.686
Top resid features:
my
Token my
Feature activation+0.731
Top resid features:
community
Token community
Feature activation+0.070
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
a
Token a
Feature activation+0.000
Top resid features:
company
Token company
Feature activation+0.000
Top resid features:

Decoder Weights Distribution

Head 0: 0.08

Head 1: 0.02

Head 2: 0.02

Head 3: 0.06

Head 4: 0.55

Head 5: 0.03

Head 6: 0.03

Head 7: 0.02

Head 8: 0.09

Head 9: 0.06

Head 10: 0.02

Head 11: 0.02

Positive logits

DERR2.52

ample2.38

what2.12

periodic2.07

earance2.05

thora2.04

abies2.01

clipboard1.98

initions1.98

AAF1.97

everal1.96

ÍÍ1.93

geries1.90

objective1.90

gradual1.88

what1.87

rious1.82

truthful1.82

flattering1.81

hin1.80

Negative logits

],"-2.17

umn-2.10

Cannot-2.05

kefeller-2.03

Shepherd-2.02

の魔-2.01

Quincy-2.01

Otherwise-1.97

Origin-1.90

Num-1.90

ioned-1.88

wash-1.87

"}-1.87

]),-1.86

NC-1.86

nect-1.85

ust-1.84

Nasa-1.84

powder-1.81

Lamb-1.81

INTERVAL 4.161 - 4.624
CONTAINS 0.000%

and
Token and
Feature activation+0.000
it
Token it
Feature activation+0.000
has
Token has
Feature activation+0.000
removed
Token removed
Feature activation+0.000
from
Token from
Feature activation+0.000
education
Token education
Feature activation+4.283
the
Token the
Feature activation+0.168
,
Token,
Feature activation+0.000
I
Token I
Feature activation+0.000
guess
Token guess
Feature activation+0.000
what
Token what
Feature activation+0.000
by
Token by
Feature activation+0.000
a
Token a
Feature activation+0.000
23
Token 23
Feature activation+0.000
-
Token-
Feature activation+0.000
9
Token9
Feature activation+0.723
vote
Token vote
Feature activation+4.433
legislation
Token legislation
Feature activation+0.324
out
Token out
Feature activation+0.000
-
Token-
Feature activation+0.000
of
Tokenof
Feature activation+0.000
-
Token-
Feature activation+0.000
High
Token High
Feature activation+0.000
Court
Token Court
Feature activation+3.922
(
Token (
Feature activation+0.000
I
TokenI
Feature activation+0.000
HC
TokenHC
Feature activation+0.000
)
Token)
Feature activation+4.170
a
Token a
Feature activation+0.000
four
Token four
Feature activation+0.000
-
Token-
Feature activation+0.000
phase
Tokenphase
Feature activation+0.000
plan
Token plan
Feature activation+0.000
by
Token by
Feature activation+0.000
a
Token a
Feature activation+0.000
23
Token 23
Feature activation+0.000
-
Token-
Feature activation+0.000
9
Token9
Feature activation+1.060
vote
Token vote
Feature activation+4.624
legislation
Token legislation
Feature activation+0.412
out
Token out
Feature activation+0.000
-
Token-
Feature activation+0.000
of
Tokenof
Feature activation+0.000
-
Token-
Feature activation+0.000

INTERVAL 3.699 - 4.161
CONTAINS 0.000%

razor
Token razor
Feature activation+0.000
:
Token:
Feature activation+0.000
Never
Token Never
Feature activation+0.000
attribute
Token attribute
Feature activation+0.000
to
Token to
Feature activation+0.000
malice
Token malice
Feature activation+3.917
that
Token that
Feature activation+0.497
which
Token which
Feature activation+0.000
can
Token can
Feature activation+0.000
be
Token be
Feature activation+0.000
adequately
Token adequately
Feature activation+0.047
submitted
Token submitted
Feature activation+0.000
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
Islamabad
Token Islamabad
Feature activation+0.000
High
Token High
Feature activation+0.000
Court
Token Court
Feature activation+3.922
(
Token (
Feature activation+0.000
I
TokenI
Feature activation+0.000
HC
TokenHC
Feature activation+0.000
)
Token)
Feature activation+4.170
a
Token a
Feature activation+0.000
can
Token can
Feature activation+0.000
eventually
Token eventually
Feature activation+0.000
help
Token help
Feature activation+0.000
recreate
Token recreate
Feature activation+0.000
in
Token in
Feature activation+0.000
monkeys
Token monkeys
Feature activation+3.842
such
Token such
Feature activation+0.075
diseases
Token diseases
Feature activation+0.822
as
Token as
Feature activation+0.000
Parkinson
Token Parkinson
Feature activation+0.000
's
Token's
Feature activation+0.000
also
Token also
Feature activation+0.000
imported
Token imported
Feature activation+0.000
from
Token from
Feature activation+0.000
the
Token the
Feature activation+0.000
phone
Token phone
Feature activation+0.110
ecosystem
Token ecosystem
Feature activation+3.823
the
Token the
Feature activation+0.000
power
Token power
Feature activation+0.000
to
Token to
Feature activation+0.000
regulate
Token regulate
Feature activation+0.000
rates
Token rates
Feature activation+0.000

INTERVAL 3.237 - 3.699
CONTAINS 0.000%

included
Token included
Feature activation+0.000
as
Token as
Feature activation+0.000
a
Token a
Feature activation+0.000
sub
Token sub
Feature activation+0.000
-
Token-
Feature activation+0.000
category
Tokencategory
Feature activation+3.678
which
Token which
Feature activation+0.414
of
Token of
Feature activation+0.000
those
Token those
Feature activation+0.000
constituted
Token constituted
Feature activation+0.000
major
Token major
Feature activation+0.000
5
Token 5
Feature activation+0.000
-
Token-
Feature activation+0.000
year
Tokenyear
Feature activation+0.000
-
Token-
Feature activation+0.000
old
Tokenold
Feature activation+2.149
students
Token students
Feature activation+3.480
a
Token a
Feature activation+0.000
sense
Token sense
Feature activation+0.000
of
Token of
Feature activation+0.000
patriotism
Token patriotism
Feature activation+0.000
with
Token with
Feature activation+0.000
reaff
Token reaff
Feature activation+0.000
irm
Tokenirm
Feature activation+0.000
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
American
Token American
Feature activation+0.000
people
Token people
Feature activation+3.683
the
Token the
Feature activation+0.000
party
Token party
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
-
Token-
Feature activation+0.000
sym
Tokensym
Feature activation+0.000
?
Token?
Feature activation+0.000
accepts
Token accepts
Feature activation+0.000
as
Token as
Feature activation+0.000
arguments
Token arguments
Feature activation+3.265
an
Token an
Feature activation+0.000
S
Token S
Feature activation+0.000
-
Token-
Feature activation+0.000
expression
Tokenexpression
Feature activation+0.018
(
Token (
Feature activation+0.000

INTERVAL 2.774 - 3.237
CONTAINS 0.001%

has
Token has
Feature activation+0.000
discussed
Token discussed
Feature activation+0.000
with
Token with
Feature activation+0.000
several
Token several
Feature activation+0.000
petrol
Token petrol
Feature activation+0.000
stations
Token stations
Feature activation+2.790
working
Token working
Feature activation+0.313
across
Token across
Feature activation+0.000
the
Token the
Feature activation+0.000
country
Token country
Feature activation+1.127
cance
Token cance
Feature activation+0.000
is
Token is
Feature activation+0.000
to
Token to
Feature activation+0.000
take
Token take
Feature activation+0.000
as
Token as
Feature activation+0.000
the
Token the
Feature activation+0.890
target
Token target
Feature activation+2.802
the
Token the
Feature activation+0.762
present
Token present
Feature activation+0.380
demand
Token demand
Feature activation+0.684
.
Token.
Feature activation+0.493
This
Token This
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
incre
Tokenincre
Feature activation+0.000
ase
Tokenase
Feature activation+0.000
among
Token among
Feature activation+0.000
Catholics
Token Catholics
Feature activation+2.911
the
Token the
Feature activation+0.000
awareness
Token awareness
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
importance
Token importance
Feature activation+0.000
to
Token to
Feature activation+0.000
show
Token show
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
Sample
Token Sample
Feature activation+0.074
Analysis
Token Analysis
Feature activation+3.030
at
Token at
Feature activation+0.000
Mars
Token Mars
Feature activation+1.852
instrument
Token instrument
Feature activation+2.628
that
Token that
Feature activation+0.000
there
Token there
Feature activation+0.000
the
Token the
Feature activation+0.000
justices
Token justices
Feature activation+0.000
dismissed
Token dismissed
Feature activation+0.000
,
Token,
Feature activation+0.000
without
Token without
Feature activation+0.000
comment
Token comment
Feature activation+3.200
,
Token,
Feature activation+0.602
Pf
Token Pf
Feature activation+0.000
izer
Tokenizer
Feature activation+0.312
's
Token's
Feature activation+0.000
claim
Token claim
Feature activation+0.225

INTERVAL 2.312 - 2.774
CONTAINS 0.002%

to
Token to
Feature activation+0.000
address
Token address
Feature activation+0.000
with
Token with
Feature activation+0.000
a
Token a
Feature activation+0.000
blog
Token blog
Feature activation+0.814
post
Token post
Feature activation+2.398
.
Token.
Feature activation+0.000
Interestingly
Token Interestingly
Feature activation+0.000
enough
Token enough
Feature activation+0.000
,
Token,
Feature activation+0.000
I
Token I
Feature activation+0.000
parents
Token parents
Feature activation+0.000
alike
Token alike
Feature activation+0.000
greeted
Token greeted
Feature activation+0.000
with
Token with
Feature activation+0.000
aston
Token aston
Feature activation+0.000
ishment
Tokenishment
Feature activation+2.426
his
Token his
Feature activation+0.000
unprecedented
Token unprecedented
Feature activation+0.000
call
Token call
Feature activation+0.000
to
Token to
Feature activation+0.000
drop
Token drop
Feature activation+0.000
made
Token made
Feature activation+0.000
no
Token no
Feature activation+0.000
secret
Token secret
Feature activation+0.000
in
Token in
Feature activation+0.000
public
Token public
Feature activation+0.341
appearances
Token appearances
Feature activation+2.464
that
Token that
Feature activation+0.167
she
Token she
Feature activation+0.000
isn
Token isn
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
I
TokenI
Feature activation+0.000
read
Token read
Feature activation+0.000
with
Token with
Feature activation+0.000
dismay
Token dismay
Feature activation+2.612
the
Token the
Feature activation+0.000
article
Token article
Feature activation+0.000
on
Token on
Feature activation+0.000
Politico
Token Politico
Feature activation+0.062
that
Token that
Feature activation+0.000
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
presence
Token presence
Feature activation+0.000
in
Token in
Feature activation+0.000
my
Token my
Feature activation+0.000
community
Token community
Feature activation+2.683
of
Token of
Feature activation+0.000
a
Token a
Feature activation+0.000
company
Token company
Feature activation+0.000
that
Token that
Feature activation+0.000
participates
Token participates
Feature activation+0.000

INTERVAL 1.850 - 2.312
CONTAINS 0.002%

instead
Token instead
Feature activation+0.000
may
Token may
Feature activation+0.000
weigh
Token weigh
Feature activation+0.000
with
Token with
Feature activation+0.000
some
Token some
Feature activation+0.937
people
Token people
Feature activation+2.137
,
Token,
Feature activation+0.206
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
he
Token he
Feature activation+0.000
said
Token said
Feature activation+0.000
do
Token do
Feature activation+0.000
we
Token we
Feature activation+0.000
have
Token have
Feature activation+0.000
in
Token in
Feature activation+0.000
my
Token my
Feature activation+0.000
view
Token view
Feature activation+1.850
strong
Token strong
Feature activation+0.000
evidence
Token evidence
Feature activation+0.000
of
Token of
Feature activation+0.000
corpor
Token corpor
Feature activation+0.000
ates
Tokenates
Feature activation+0.000
explaining
Token explaining
Feature activation+0.000
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
middle
Token middle
Feature activation+0.000
-
Token-
Feature activation+0.000
man
Tokenman
Feature activation+1.902
that
Token that
Feature activation+0.000
I
Token I
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
m
Tokenm
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
Sample
Token Sample
Feature activation+0.074
Analysis
Token Analysis
Feature activation+3.030
at
Token at
Feature activation+0.000
Mars
Token Mars
Feature activation+1.852
instrument
Token instrument
Feature activation+2.628
that
Token that
Feature activation+0.000
there
Token there
Feature activation+0.000
was
Token was
Feature activation+0.000
a
Token a
Feature activation+0.000
for
Token for
Feature activation+0.000
the
Token the
Feature activation+0.000
oil
Token oil
Feature activation+0.000
and
Token and
Feature activation+0.000
gas
Token gas
Feature activation+0.000
business
Token business
Feature activation+1.984
,
Token,
Feature activation+1.590
at
Token at
Feature activation+0.000
least
Token least
Feature activation+0.293
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000

INTERVAL 1.387 - 1.850
CONTAINS 0.003%

uss
Tokenuss
Feature activation+0.000
les
Tokenles
Feature activation+2.428
with
Token with
Feature activation+0.054
suicidal
Token suicidal
Feature activation+0.000
wishes
Token wishes
Feature activation+0.916
a
Token a
Feature activation+1.659
drama
Token drama
Feature activation+0.000
of
Token of
Feature activation+0.000
love
Token love
Feature activation+0.000
and
Token and
Feature activation+0.000
honor
Token honor
Feature activation+0.000
-
Token-
Feature activation+0.000
authored
Tokenauthored
Feature activation+0.000
,
Token,
Feature activation+0.000
with
Token with
Feature activation+0.000
Bill
Token Bill
Feature activation+0.000
Robinson
Token Robinson
Feature activation+1.517
and
Token and
Feature activation+0.000
the
Token the
Feature activation+0.000
late
Token late
Feature activation+0.000
Desmond
Token Desmond
Feature activation+0.000
Ball
Token Ball
Feature activation+0.447
fifth
Token fifth
Feature activation+0.000
assessment
Token assessment
Feature activation+1.221
on
Token on
Feature activation+0.324
climate
Token climate
Feature activation+0.000
change
Token change
Feature activation+0.505
report
Token report
Feature activation+1.417
that
Token that
Feature activation+0.000
the
Token the
Feature activation+0.000
earth
Token earth
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
assess
Token assess
Feature activation+0.000
with
Token with
Feature activation+0.000
varying
Token varying
Feature activation+0.000
degrees
Token degrees
Feature activation+0.308
of
Token of
Feature activation+0.000
confidence
Token confidence
Feature activation+1.731
that
Token that
Feature activation+0.000
the
Token the
Feature activation+0.000
Syrian
Token Syrian
Feature activation+0.000
regime
Token regime
Feature activation+0.000
has
Token has
Feature activation+0.000
and
Token and
Feature activation+0.000
the
Token the
Feature activation+0.000
late
Token late
Feature activation+0.000
Desmond
Token Desmond
Feature activation+0.000
Ball
Token Ball
Feature activation+0.447
,
Token,
Feature activation+1.426
several
Token several
Feature activation+0.000
detailed
Token detailed
Feature activation+0.000
reports
Token reports
Feature activation+0.000
about
Token about
Feature activation+0.000
the
Token the
Feature activation+0.000

INTERVAL 0.925 - 1.387
CONTAINS 0.011%

them
Token them
Feature activation+0.000
to
Token to
Feature activation+0.000
believe
Token believe
Feature activation+0.000
as
Token as
Feature activation+0.000
they
Token they
Feature activation+0.000
do
Token do
Feature activation+1.104
.
Token.
Feature activation+0.000
There
Token There
Feature activation+0.000
is
Token is
Feature activation+0.000
some
Token some
Feature activation+0.000
value
Token value
Feature activation+0.000
rather
Token rather
Feature activation+0.000
than
Token than
Feature activation+0.000
be
Token be
Feature activation+0.000
victimized
Token victimized
Feature activation+0.618
by
Token by
Feature activation+0.936
,
Token,
Feature activation+1.243
their
Token their
Feature activation+0.000
pain
Token pain
Feature activation+0.000
.
Token.
Feature activation+0.000
For
Token For
Feature activation+0.000
example
Token example
Feature activation+0.000
at
Token at
Feature activation+0.000
times
Token times
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
presidential
Token presidential
Feature activation+0.000
campaign
Token campaign
Feature activation+1.188
exhibited
Token exhibited
Feature activation+0.000
a
Token a
Feature activation+0.000
less
Token less
Feature activation+0.000
ideological
Token ideological
Feature activation+0.000
streak
Token streak
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
also
Token also
Feature activation+0.000
counts
Token counts
Feature activation+0.000
as
Token as
Feature activation+0.000
cash
Token cash
Feature activation+0.668
in
Token in
Feature activation+1.157
the
Token the
Feature activation+0.402
bank
Token bank
Feature activation+0.853
and
Token and
Feature activation+0.138
can
Token can
Feature activation+0.000
thus
Token thus
Feature activation+0.000
kids
Token kids
Feature activation+0.000
by
Token by
Feature activation+0.000
posting
Token posting
Feature activation+0.000
on
Token on
Feature activation+0.000
social
Token social
Feature activation+0.000
media
Token media
Feature activation+1.126
is
Token is
Feature activation+0.000
not
Token not
Feature activation+0.000
that
Token that
Feature activation+0.000
different
Token different
Feature activation+0.000
from
Token from
Feature activation+0.000

INTERVAL 0.462 - 0.925
CONTAINS 0.028%

If
Token If
Feature activation+0.000
one
Token one
Feature activation+0.000
can
Token can
Feature activation+0.000
make
Token make
Feature activation+0.000
of
Token of
Feature activation+0.000
one
Token one
Feature activation+0.593
's
Token's
Feature activation+0.112
t
Token t
Feature activation+0.000
uss
Tokenuss
Feature activation+0.000
les
Tokenles
Feature activation+2.428
with
Token with
Feature activation+0.054
on
Token on
Feature activation+0.000
the
Token the
Feature activation+0.000
30
Token 30
Feature activation+0.000
th
Tokenth
Feature activation+0.198
of
Token of
Feature activation+0.035
June
Token June
Feature activation+0.478
.
Token.
Feature activation+0.000
Cop
Token Cop
Feature activation+0.000
tic
Tokentic
Feature activation+0.000
women
Token women
Feature activation+0.000
who
Token who
Feature activation+0.000
For
Token For
Feature activation+0.000
example
Token example
Feature activation+0.000
,
Token,
Feature activation+0.000
leaving
Token leaving
Feature activation+0.000
or
Token or
Feature activation+0.000
charging
Token charging
Feature activation+0.464
a
Token a
Feature activation+0.000
battery
Token battery
Feature activation+0.000
in
Token in
Feature activation+0.000
a
Token a
Feature activation+0.000
hot
Token hot
Feature activation+0.000
and
Token and
Feature activation+0.000
match
Token match
Feature activation+0.000
as
Token as
Feature activation+0.000
a
Token a
Feature activation+0.000
man
Token man
Feature activation+0.534
defender
Token defender
Feature activation+0.668
.
Token.
Feature activation+0.000
Cover
Token Cover
Feature activation+0.000
ing
Tokening
Feature activation+0.000
for
Token for
Feature activation+0.000
longer
Token longer
Feature activation+0.000
between
Token between
Feature activation+0.225
3
Token 3
Feature activation+0.000
,
Token,
Feature activation+0.000
5
Token 5
Feature activation+0.000
or
Token or
Feature activation+0.000
10
Token 10
Feature activation+0.528
videos
Token videos
Feature activation+0.024
.
Token.
Feature activation+0.000
This
Token This
Feature activation+0.000
type
Token type
Feature activation+0.000
of
Token of
Feature activation+0.000

INTERVAL 0.000 - 0.462
CONTAINS 99.952%

efforts
Token efforts
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
House
Token House
Feature activation+0.000
,"
Token,"
Feature activation+0.000
she
Token she
Feature activation+0.000
said
Token said
Feature activation+0.000
,
Token,
Feature activation+0.000
referring
Token referring
Feature activation+0.000
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
won
Token won
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
t
Tokent
Feature activation+0.000
find
Token find
Feature activation+0.000
a
Token a
Feature activation+0.000
goods
Token goods
Feature activation+0.000
-
Token-
Feature activation+0.000
car
Tokencar
Feature activation+0.000
riage
Tokenriage
Feature activation+0.000
to
Token to
Feature activation+0.000
's
Token's
Feature activation+0.000
truth
Token truth
Feature activation+0.000
as
Token as
Feature activation+0.000
much
Token much
Feature activation+0.000
as
Token as
Feature activation+0.000
it
Token it
Feature activation+0.000
exists
Token exists
Feature activation+0.000
."
Token."
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
He
TokenHe
Feature activation+0.000
cunt
Token cunt
Feature activation+0.000
to
Token to
Feature activation+0.000
them
Token them
Feature activation+0.000
too
Token too
Feature activation+0.000
?
Token?
Feature activation+0.000
Are
Token Are
Feature activation+0.000
they
Token they
Feature activation+0.000
m
Token m
Feature activation+0.000
olly
Tokenolly
Feature activation+0.000
c
Tokenc
Feature activation+0.000
odd
Tokenodd
Feature activation+0.000
and
Token and
Feature activation+0.000
Samsung
Token Samsung
Feature activation+0.000
Precision
Token Precision
Feature activation+0.000
Co
Token Co
Feature activation+0.000
.
Token.
Feature activation+0.000
(
Token (
Feature activation+0.000
now
Tokennow
Feature activation+0.000
called
Token called
Feature activation+0.000
Samsung
Token Samsung
Feature activation+0.000
Tech
Token Tech
Feature activation+0.000
win
Tokenwin
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 H7.4: (feature 22461)

TOP ACTIVATIONS
MAX = 4.959

giving
Token giving
Feature activation+0.000
walk
Token walk
Feature activation+0.000
ers
Tokeners
Feature activation+0.000
greater
Token greater
Feature activation+0.000
visibility
Token visibility
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+4.959
and
Token and
Feature activation+0.000
traffic
Token traffic
Feature activation+0.000
lights
Token lights
Feature activation+0.000
with
Token with
Feature activation+0.000
left
Token left
Feature activation+0.000
Ļ
TokenĻ
Feature activation+2.846
s
Tokens
Feature activation+1.817
ret
Token ret
Feature activation+0.000
rench
Tokenrench
Feature activation+0.000
ment
Tokenment
Feature activation+1.029
âĢĶ
TokenâĢĶ
Feature activation+4.492
is
Tokenis
Feature activation+0.274
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
the
Tokenthe
Feature activation+0.000
future
Token future
Feature activation+0.023
are
Token are
Feature activation+0.000
announced
Token announced
Feature activation+0.301
Jan
Token Jan
Feature activation+0.000
.
Token.
Feature activation+0.000
18
Token 18
Feature activation+0.034
âĢĶ
Token âĢĶ
Feature activation+4.341
but
Token but
Feature activation+0.000
for
Token for
Feature activation+0.000
the
Token the
Feature activation+0.000
first
Token first
Feature activation+0.000
time
Token time
Feature activation+0.000
âĢĶ
TokenâĢĶ
Feature activation+0.000
both
Tokenboth
Feature activation+0.000
men
Token men
Feature activation+0.000
and
Token and
Feature activation+0.000
women
Token women
Feature activation+2.087
!
Token!
Feature activation+4.235
âĢĶ
TokenâĢĶ
Feature activation+0.792
sp
Tokensp
Feature activation+0.000
urned
Tokenurned
Feature activation+0.000
watching
Token watching
Feature activation+0.000
pornographic
Token pornographic
Feature activation+0.000
two
Token two
Feature activation+0.000
countries
Token countries
Feature activation+0.000
-
Token -
Feature activation+0.000
Russia
Token Russia
Feature activation+0.096
and
Token and
Feature activation+0.424
Cyprus
Token Cyprus
Feature activation+3.530
.
Token.
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
He
Token He
Feature activation+0.000
said
Token said
Feature activation+0.000
Iraq
Token Iraq
Feature activation+0.021
and
Token and
Feature activation+0.000
who
Token who
Feature activation+0.000
thank
Token thank
Feature activation+0.000
God
Token God
Feature activation+0.000
!
Token!
Feature activation+3.405
survived
Token survived
Feature activation+0.076
âĢĶ
Token âĢĶ
Feature activation+3.401
will
Token will
Feature activation+0.060
die
Token die
Feature activation+0.000
,
Token,
Feature activation+0.000
who
Token who
Feature activation+0.000
thank
Token thank
Feature activation+0.000
God
Token God
Feature activation+0.000
!
Token!
Feature activation+3.405
survived
Token survived
Feature activation+0.076
âĢĶ
Token âĢĶ
Feature activation+3.401
will
Token will
Feature activation+0.060
die
Token die
Feature activation+0.000
,
Token,
Feature activation+0.000
be
Token be
Feature activation+0.000
m
Token m
Feature activation+0.000
Michael
Token Michael
Feature activation+0.000
Is
Token Is
Feature activation+0.000
ik
Tokenik
Feature activation+0.000
off
Tokenoff
Feature activation+0.000
piece
Token piece
Feature activation+0.078
?
Token?
Feature activation+3.325
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
A
TokenA
Feature activation+0.000
lawyer
Token lawyer
Feature activation+0.000
familiar
Token familiar
Feature activation+0.000
ale
Tokenale
Feature activation+0.000
ed
Tokened
Feature activation+0.000
bin
Token bin
Feature activation+0.000
Tal
Token Tal
Feature activation+0.000
al
Tokenal
Feature activation+0.000
âĢĶ
TokenâĢĶ
Feature activation+3.298
is
Tokenis
Feature activation+0.344
timed
Token timed
Feature activation+0.000
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
Paradise
Token Paradise
Feature activation+0.000
only
Token only
Feature activation+0.000
before
Token before
Feature activation+0.000
going
Token going
Feature activation+0.000
to
Token to
Feature activation+0.000
prison
Token prison
Feature activation+0.000
--
Token --
Feature activation+3.261
on
Token on
Feature activation+0.000
Tuesday
Token Tuesday
Feature activation+0.000
at
Token at
Feature activation+0.000
the
Token the
Feature activation+0.000
Ren
Token Ren
Feature activation+0.000
s
Tokens
Feature activation+0.603
always
Token always
Feature activation+0.000
sunny
Token sunny
Feature activation+1.040
in
Token in
Feature activation+0.161
LA
Token LA
Feature activation+0.644
âĢĶ
Token âĢĶ
Feature activation+3.202
but
Token but
Feature activation+0.002
the
Token the
Feature activation+0.000
drivers
Token drivers
Feature activation+0.000
are
Token are
Feature activation+0.000
her
Token her
Feature activation+0.000
bad
Token bad
Feature activation+0.156
teams
Token teams
Feature activation+0.277
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
NHL
Token NHL
Feature activation+0.141
âĢĵ
Token âĢĵ
Feature activation+3.167
would
Token would
Feature activation+0.179
sink
Token sink
Feature activation+0.000
like
Token like
Feature activation+0.000
a
Token a
Feature activation+0.000
stone
Token stone
Feature activation+0.000
beings
Token beings
Feature activation+0.000
âĢĶ
TokenâĢĶ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
G
Token G
Feature activation+0.000
entry
Tokenentry
Feature activation+0.295
âĢĶ
TokenâĢĶ
Feature activation+3.137
who
Tokenwho
Feature activation+0.000
threaten
Token threaten
Feature activation+0.000
to
Token to
Feature activation+0.000
corrupt
Token corrupt
Feature activation+0.000
their
Token their
Feature activation+0.000
Word
Token Word
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Year
Token Year
Feature activation+0.000
2013
Token 2013
Feature activation+0.000
âĢĵ
Token âĢĵ
Feature activation+3.135
is
Token is
Feature activation+0.000
not
Token not
Feature activation+0.000
just
Token just
Feature activation+0.000
the
Token the
Feature activation+0.000
preserve
Token preserve
Feature activation+0.000
eyes
Token eyes
Feature activation+0.000
of
Token of
Feature activation+0.000
a
Token a
Feature activation+0.000
concerned
Token concerned
Feature activation+0.000
government
Token government
Feature activation+0.000
-
Token -
Feature activation+3.073
exhibition
Token exhibition
Feature activation+0.000
,
Token,
Feature activation+0.000
expressing
Token expressing
Feature activation+0.000
new
Token new
Feature activation+0.000
and
Token and
Feature activation+0.000
Greens
Token Greens
Feature activation+0.000
pan
Tokenpan
Feature activation+0.000
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.180
Bern
Token Bern
Feature activation+0.000
anke
Tokenanke
Feature activation+3.052
âĢĶ
TokenâĢĶ
Feature activation+1.428
have
Tokenhave
Feature activation+0.000
been
Token been
Feature activation+0.000
Jewish
Token Jewish
Feature activation+0.000
.
Token.
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
t
Tokent
Feature activation+0.000
record
Token record
Feature activation+0.000
any
Token any
Feature activation+0.000
stats
Token stats
Feature activation+0.761
âĢĶ
Token âĢĶ
Feature activation+3.048
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
season
Token season
Feature activation+0.000
finale
Token finale
Feature activation+0.023
.
Token.
Feature activation+0.000
like
Tokenlike
Feature activation+0.000
a
Token a
Feature activation+0.000
shipping
Token shipping
Feature activation+0.000
container
Token container
Feature activation+1.772
âĢ
TokenâĢ
Feature activation+2.772
Ŀ
TokenĿ
Feature activation+3.042
âĢĶ
Token âĢĶ
Feature activation+0.030
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
jail
Token jail
Feature activation+0.000
-
Token-
Feature activation+0.000
Dist
Token Dist
Feature activation+0.000
ributed
Tokenributed
Feature activation+0.000
Aut
Token Aut
Feature activation+0.000
onomous
Tokenonomous
Feature activation+0.000
Organization
Token Organization
Feature activation+0.843
âĢ
Token âĢ
Feature activation+2.979
Ĵ
TokenĴ
Feature activation+0.392
is
Token is
Feature activation+0.000
similar
Token similar
Feature activation+0.000
to
Token to
Feature activation+0.000
an
Token an
Feature activation+0.000
to
Token to
Feature activation+0.000
be
Token be
Feature activation+0.000
reconnect
Token reconnect
Feature activation+0.000
ed
Tokened
Feature activation+0.000
twice
Token twice
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+2.974
Obama
Token Obama
Feature activation+0.000
told
Token told
Feature activation+0.000
Schumer
Token Schumer
Feature activation+0.000
that
Token that
Feature activation+0.000
the
Token the
Feature activation+0.000

Top DFA by src position
MAX = 5.150

-
Token-
Feature activation+0.003
Top resid features:
friendly
Tokenfriendly
Feature activation+0.004
Top resid features:
signals
Token signals
Feature activation+0.007
Top resid features:
âĢ
TokenâĢ
Feature activation+0.000
Top resid features:
Ŀ
TokenĿ
Feature activation+0.010
Top resid features:
âĢĶ
Token âĢĶ
Feature activation+5.150
Top resid features:
walk
Token walk
Feature activation+0.013
Top resid features:
signs
Token signs
Feature activation+0.008
Top resid features:
that
Token that
Feature activation+0.004
Top resid features:
allow
Token allow
Feature activation+0.010
Top resid features:
pedestrians
Token pedestrians
Feature activation+0.002
Top resid features:
âĢ
Token âĢ
Feature activation+0.162
Top resid features:
ľ
Tokenľ
Feature activation+0.059
Top resid features:
revolution
Tokenrevolution
Feature activation+0.030
Top resid features:
âĢ
TokenâĢ
Feature activation+0.019
Top resid features:
Ŀ
TokenĿ
Feature activation+0.085
Top resid features:
âĢĶ
TokenâĢĶ
Feature activation+4.271
Top resid features:
as
Tokenas
Feature activation+0.413
Top resid features:
opposed
Token opposed
Feature activation+0.066
Top resid features:
to
Token to
Feature activation+0.031
Top resid features:
Paul
Token Paul
Feature activation-0.000
Top resid features:
Ryan
Token Ryan
Feature activation+0.024
Top resid features:
of
Token of
Feature activation-0.003
Top resid features:
the
Token the
Feature activation-0.008
Top resid features:
votes
Token votes
Feature activation-0.006
Top resid features:
are
Token are
Feature activation+0.013
Top resid features:
counted
Token counted
Feature activation+0.007
Top resid features:
âĢĶ
Token âĢĶ
Feature activation+4.604
Top resid features:
with
Token with
Feature activation+0.130
Top resid features:
Tim
Token Tim
Feature activation-0.004
Top resid features:
Rain
Token Rain
Feature activation-0.004
Top resid features:
es
Tokenes
Feature activation+0.003
Top resid features:
and
Token and
Feature activation+0.003
Top resid features:
.
Token.
Feature activation-0.001
Top resid features:
Researchers
Token Researchers
Feature activation+0.047
Top resid features:
found
Token found
Feature activation+0.037
Top resid features:
that
Token that
Feature activation+0.009
Top resid features:
couples
Token couples
Feature activation+0.014
Top resid features:
âĢĶ
TokenâĢĶ
Feature activation+4.537
Top resid features:
both
Tokenboth
Feature activation+0.114
Top resid features:
men
Token men
Feature activation+0.028
Top resid features:
and
Token and
Feature activation+0.023
Top resid features:
women
Token women
Feature activation+0.008
Top resid features:
!
Token!
Feature activation+0.093
Top resid features:
citizen
Token citizen
Feature activation-0.004
Top resid features:
of
Token of
Feature activation+0.008
Top resid features:
the
Token the
Feature activation-0.003
Top resid features:
two
Token two
Feature activation+0.019
Top resid features:
countries
Token countries
Feature activation+0.008
Top resid features:
-
Token -
Feature activation+4.039
Top resid features:
Russia
Token Russia
Feature activation-0.005
Top resid features:
and
Token and
Feature activation+0.075
Top resid features:
Cyprus
Token Cyprus
Feature activation+0.009
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
âĢ
TokenâĢ
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.055
Top resid features:
where
Token where
Feature activation+0.034
Top resid features:
our
Token our
Feature activation-0.000
Top resid features:
military
Token military
Feature activation+0.028
Top resid features:
personnel
Token personnel
Feature activation+0.021
Top resid features:
âĢĶ
Token âĢĶ
Feature activation+3.730
Top resid features:
people
Token people
Feature activation+0.136
Top resid features:
like
Token like
Feature activation-0.000
Top resid features:
my
Token my
Feature activation+0.010
Top resid features:
Marine
Token Marine
Feature activation+0.017
Top resid features:
son
Token son
Feature activation+0.018
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.072
Top resid features:
where
Token where
Feature activation+0.045
Top resid features:
our
Token our
Feature activation+0.008
Top resid features:
military
Token military
Feature activation+0.025
Top resid features:
personnel
Token personnel
Feature activation+0.033
Top resid features:
âĢĶ
Token âĢĶ
Feature activation+3.560
Top resid features:
people
Token people
Feature activation+0.063
Top resid features:
like
Token like
Feature activation+0.075
Top resid features:
my
Token my
Feature activation+0.017
Top resid features:
Marine
Token Marine
Feature activation+0.018
Top resid features:
son
Token son
Feature activation+0.023
Top resid features:
could
Token could
Feature activation+0.004
Top resid features:
find
Token find
Feature activation-0.013
Top resid features:
repeating
Token repeating
Feature activation+0.006
Top resid features:
that
Token that
Feature activation+0.008
Top resid features:
citation
Token citation
Feature activation+0.016
Top resid features:
âĢĵ
TokenâĢĵ
Feature activation+3.638
Top resid features:
would
Tokenwould
Feature activation+0.060
Top resid features:
you
Token you
Feature activation+0.009
Top resid features:
believe
Token believe
Feature activation+0.012
Top resid features:
it
Token it
Feature activation+0.009
Top resid features:
,
Token,
Feature activation+0.029
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.078
Top resid features:
terror
Token terror
Feature activation+0.001
Top resid features:
faction
Token faction
Feature activation-0.161
Top resid features:
âĢĶ
TokenâĢĶ
Feature activation+2.149
Top resid features:
including
Tokenincluding
Feature activation+1.711
Top resid features:
[
Token [
Feature activation+0.005
Top resid features:
Os
TokenOs
Feature activation-0.000
Top resid features:
ama
Tokenama
Feature activation+0.001
Top resid features:
Bin
Token Bin
Feature activation-0.001
Top resid features:
exclusive
Token exclusive
Feature activation+0.014
Top resid features:
sit
Token sit
Feature activation-0.013
Top resid features:
-
Token-
Feature activation+0.004
Top resid features:
down
Tokendown
Feature activation+0.001
Top resid features:
interview
Token interview
Feature activation-0.047
Top resid features:
--
Token --
Feature activation+4.011
Top resid features:
his
Token his
Feature activation+0.047
Top resid features:
only
Token only
Feature activation+0.009
Top resid features:
before
Token before
Feature activation+0.012
Top resid features:
going
Token going
Feature activation-0.003
Top resid features:
to
Token to
Feature activation-0.002
Top resid features:
It
Token It
Feature activation+0.014
Top resid features:
âĢ
TokenâĢ
Feature activation+0.008
Top resid features:
Ļ
TokenĻ
Feature activation+0.229
Top resid features:
s
Tokens
Feature activation+0.111
Top resid features:
sunny
Token sunny
Feature activation+0.043
Top resid features:
âĢĶ
Token âĢĶ
Feature activation+3.219
Top resid features:
it
Token it
Feature activation+0.106
Top resid features:
âĢ
TokenâĢ
Feature activation-0.006
Top resid features:
Ļ
TokenĻ
Feature activation+0.020
Top resid features:
s
Tokens
Feature activation+0.060
Top resid features:
always
Token always
Feature activation+0.026
Top resid features:
the
Token the
Feature activation-0.005
Top resid features:
surface
Token surface
Feature activation+0.007
Top resid features:
but
Token but
Feature activation+0.042
Top resid features:
the
Token the
Feature activation+0.026
Top resid features:
bad
Token bad
Feature activation-0.169
Top resid features:
âĢĵ
Token âĢĵ
Feature activation+2.660
Top resid features:
and
Token and
Feature activation+1.074
Top resid features:
there
Token there
Feature activation+0.021
Top resid features:
's
Token's
Feature activation+0.020
Top resid features:
plenty
Token plenty
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.001
Top resid features:
lish
Tokenlish
Feature activation+0.034
Top resid features:
,
Token,
Feature activation-0.001
Top resid features:
nihil
Token nihil
Feature activation-0.001
Top resid features:
istic
Tokenistic
Feature activation+0.022
Top resid features:
beings
Token beings
Feature activation-0.227
Top resid features:
âĢĶ
TokenâĢĶ
Feature activation+3.853
Top resid features:
The
TokenThe
Feature activation+0.015
Top resid features:
G
Token G
Feature activation-0.001
Top resid features:
entry
Tokenentry
Feature activation-0.009
Top resid features:
âĢĶ
TokenâĢĶ
Feature activation+0.034
Top resid features:
who
Tokenwho
Feature activation+0.000
Top resid features:
act
Token act
Feature activation-0.001
Top resid features:
that
Token that
Feature activation-0.020
Top resid features:
proves
Token proves
Feature activation+0.049
Top resid features:
the
Token the
Feature activation+0.018
Top resid features:
selfie
Token selfie
Feature activation-0.037
Top resid features:
âĢĵ
Token âĢĵ
Feature activation+3.634
Top resid features:
a
Token a
Feature activation+0.083
Top resid features:
word
Token word
Feature activation+0.011
Top resid features:
that
Token that
Feature activation+0.007
Top resid features:
happened
Token happened
Feature activation-0.002
Top resid features:
to
Token to
Feature activation+0.002
Top resid features:
began
Token began
Feature activation+0.009
Top resid features:
attending
Token attending
Feature activation+0.017
Top resid features:
the
Token the
Feature activation-0.045
Top resid features:
increasingly
Token increasingly
Feature activation-0.042
Top resid features:
prestigious
Token prestigious
Feature activation+0.009
Top resid features:
-
Token -
Feature activation+3.744
Top resid features:
and
Token and
Feature activation+0.133
Top resid features:
increasingly
Token increasingly
Feature activation+0.007
Top resid features:
bizarre
Token bizarre
Feature activation+0.003
Top resid features:
and
Token and
Feature activation+0.007
Top resid features:
difficult
Token difficult
Feature activation+0.006
Top resid features:
on
Token on
Feature activation-0.002
Top resid features:
this
Token this
Feature activation+0.000
Top resid features:
list
Token list
Feature activation-0.023
Top resid features:
,
Token,
Feature activation-0.006
Top resid features:
five
Token five
Feature activation+0.006
Top resid features:
âĢĶ
TokenâĢĶ
Feature activation+3.474
Top resid features:
M
TokenM
Feature activation+0.023
Top resid features:
ey
Tokeney
Feature activation-0.001
Top resid features:
er
Tokener
Feature activation+0.006
Top resid features:
,
Token,
Feature activation+0.039
Top resid features:
Burns
Token Burns
Feature activation+0.004
Top resid features:
over
Token over
Feature activation-0.032
Top resid features:
to
Token to
Feature activation-0.037
Top resid features:
Buffalo
Token Buffalo
Feature activation-0.020
Top resid features:
and
Token and
Feature activation+0.036
Top resid features:
played
Token played
Feature activation+0.032
Top resid features:
âĢĶ
Token âĢĶ
Feature activation+3.436
Top resid features:
though
Token though
Feature activation+0.073
Top resid features:
he
Token he
Feature activation+0.011
Top resid features:
didn
Token didn
Feature activation+0.007
Top resid features:
âĢ
TokenâĢ
Feature activation-0.008
Top resid features:
Ļ
TokenĻ
Feature activation+0.012
Top resid features:
z
Tokenz
Feature activation-0.000
Top resid features:
as
Tokenas
Feature activation+0.001
Top resid features:
called
Token called
Feature activation-0.004
Top resid features:
the
Token the
Feature activation-0.008
Top resid features:
trailer
Token trailer
Feature activation-0.004
Top resid features:
âĢĶ
Token âĢĶ
Feature activation+2.266
Top resid features:
which
Token which
Feature activation+0.920
Top resid features:
looks
Token looks
Feature activation+0.059
Top resid features:
âĢ
Token âĢ
Feature activation+0.053
Top resid features:
ľ
Tokenľ
Feature activation+0.078
Top resid features:
like
Tokenlike
Feature activation+0.242
Top resid features:
.
Token.
Feature activation-0.003
Top resid features:
The
Token The
Feature activation-0.008
Top resid features:
DA
Token DA
Feature activation-0.007
Top resid features:
O
TokenO
Feature activation-0.007
Top resid features:
âĢ
Token âĢ
Feature activation+0.188
Top resid features:
Ĵ
TokenĴ
Feature activation+3.117
Top resid features:
an
Token an
Feature activation+0.271
Top resid features:
acronym
Token acronym
Feature activation+0.085
Top resid features:
for
Token for
Feature activation+0.019
Top resid features:
Dist
Token Dist
Feature activation-0.005
Top resid features:
ributed
Tokenributed
Feature activation-0.004
Top resid features:
Ċ
TokenĊ
Feature activation-0.027
Top resid features:
Over
TokenOver
Feature activation-0.064
Top resid features:
a
Token a
Feature activation+0.026
Top resid features:
shaky
Token shaky
Feature activation+0.075
Top resid features:
line
Token line
Feature activation+0.018
Top resid features:
âĢĶ
Token âĢĶ
Feature activation+3.406
Top resid features:
they
Token they
Feature activation+0.081
Top resid features:
had
Token had
Feature activation+0.083
Top resid features:
to
Token to
Feature activation+0.026
Top resid features:
be
Token be
Feature activation+0.007
Top resid features:
reconnect
Token reconnect
Feature activation-0.016
Top resid features:

Decoder Weights Distribution

Head 0: 0.05

Head 1: 0.02

Head 2: 0.02

Head 3: 0.09

Head 4: 0.55

Head 5: 0.04

Head 6: 0.03

Head 7: 0.03

Head 8: 0.05

Head 9: 0.06

Head 10: 0.02

Head 11: 0.04

Positive logits

--3.89

--3.76

)—3.72

*)3.58

3.54

3.52

)--3.51

-)3.42

)-3.32

-3.31

3.18

3.17

3.12

––2.96

—"2.93

---2.90

2.87

)2.84

2.82

2.78

Negative logits

Skydragon-2.22

cryptoc-2.09

utenberg-2.00

confir-1.98

sqor-1.97

avanaugh-1.97

guiName-1.95

atan-1.88

Mutant-1.82

azeera-1.81

IDES-1.80

xit-1.78

Inquis-1.77

certs-1.76

pestic-1.75

nil-1.75

Volcano-1.75

Nicarag-1.74

uner-1.74

izoph-1.73

INTERVAL 4.463 - 4.959
CONTAINS 0.000%

Ļ
TokenĻ
Feature activation+2.846
s
Tokens
Feature activation+1.817
ret
Token ret
Feature activation+0.000
rench
Tokenrench
Feature activation+0.000
ment
Tokenment
Feature activation+1.029
âĢĶ
TokenâĢĶ
Feature activation+4.492
is
Tokenis
Feature activation+0.274
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
the
Tokenthe
Feature activation+0.000
future
Token future
Feature activation+0.023
giving
Token giving
Feature activation+0.000
walk
Token walk
Feature activation+0.000
ers
Tokeners
Feature activation+0.000
greater
Token greater
Feature activation+0.000
visibility
Token visibility
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+4.959
and
Token and
Feature activation+0.000
traffic
Token traffic
Feature activation+0.000
lights
Token lights
Feature activation+0.000
with
Token with
Feature activation+0.000
left
Token left
Feature activation+0.000

INTERVAL 3.967 - 4.463
CONTAINS 0.000%

are
Token are
Feature activation+0.000
announced
Token announced
Feature activation+0.301
Jan
Token Jan
Feature activation+0.000
.
Token.
Feature activation+0.000
18
Token 18
Feature activation+0.034
âĢĶ
Token âĢĶ
Feature activation+4.341
but
Token but
Feature activation+0.000
for
Token for
Feature activation+0.000
the
Token the
Feature activation+0.000
first
Token first
Feature activation+0.000
time
Token time
Feature activation+0.000
âĢĶ
TokenâĢĶ
Feature activation+0.000
both
Tokenboth
Feature activation+0.000
men
Token men
Feature activation+0.000
and
Token and
Feature activation+0.000
women
Token women
Feature activation+2.087
!
Token!
Feature activation+4.235
âĢĶ
TokenâĢĶ
Feature activation+0.792
sp
Tokensp
Feature activation+0.000
urned
Tokenurned
Feature activation+0.000
watching
Token watching
Feature activation+0.000
pornographic
Token pornographic
Feature activation+0.000

INTERVAL 3.471 - 3.967
CONTAINS 0.000%

two
Token two
Feature activation+0.000
countries
Token countries
Feature activation+0.000
-
Token -
Feature activation+0.000
Russia
Token Russia
Feature activation+0.096
and
Token and
Feature activation+0.424
Cyprus
Token Cyprus
Feature activation+3.530
.
Token.
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
He
Token He
Feature activation+0.000
said
Token said
Feature activation+0.000

INTERVAL 2.975 - 3.471
CONTAINS 0.001%

s
Tokens
Feature activation+0.603
always
Token always
Feature activation+0.000
sunny
Token sunny
Feature activation+1.040
in
Token in
Feature activation+0.161
LA
Token LA
Feature activation+0.644
âĢĶ
Token âĢĶ
Feature activation+3.202
but
Token but
Feature activation+0.002
the
Token the
Feature activation+0.000
drivers
Token drivers
Feature activation+0.000
are
Token are
Feature activation+0.000
her
Token her
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
t
Tokent
Feature activation+0.000
record
Token record
Feature activation+0.000
any
Token any
Feature activation+0.000
stats
Token stats
Feature activation+0.761
âĢĶ
Token âĢĶ
Feature activation+3.048
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
season
Token season
Feature activation+0.000
finale
Token finale
Feature activation+0.023
.
Token.
Feature activation+0.000
Dist
Token Dist
Feature activation+0.000
ributed
Tokenributed
Feature activation+0.000
Aut
Token Aut
Feature activation+0.000
onomous
Tokenonomous
Feature activation+0.000
Organization
Token Organization
Feature activation+0.843
âĢ
Token âĢ
Feature activation+2.979
Ĵ
TokenĴ
Feature activation+0.392
is
Token is
Feature activation+0.000
similar
Token similar
Feature activation+0.000
to
Token to
Feature activation+0.000
an
Token an
Feature activation+0.000
Word
Token Word
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Year
Token Year
Feature activation+0.000
2013
Token 2013
Feature activation+0.000
âĢĵ
Token âĢĵ
Feature activation+3.135
is
Token is
Feature activation+0.000
not
Token not
Feature activation+0.000
just
Token just
Feature activation+0.000
the
Token the
Feature activation+0.000
preserve
Token preserve
Feature activation+0.000
Michael
Token Michael
Feature activation+0.000
Is
Token Is
Feature activation+0.000
ik
Tokenik
Feature activation+0.000
off
Tokenoff
Feature activation+0.000
piece
Token piece
Feature activation+0.078
?
Token?
Feature activation+3.325
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
A
TokenA
Feature activation+0.000
lawyer
Token lawyer
Feature activation+0.000
familiar
Token familiar
Feature activation+0.000

INTERVAL 2.480 - 2.975
CONTAINS 0.002%

Small
Token Small
Feature activation+0.000
Arms
Token Arms
Feature activation+0.000
F
Token F
Feature activation+0.000
iring
Tokeniring
Feature activation+0.000
School
Token School
Feature activation+0.550
âĢĶ
TokenâĢĶ
Feature activation+2.736
where
Tokenwhere
Feature activation+0.000
specially
Token specially
Feature activation+0.000
selected
Token selected
Feature activation+0.000
soldiers
Token soldiers
Feature activation+0.000
would
Token would
Feature activation+0.000
of
Token of
Feature activation+0.000
fantasy
Token fantasy
Feature activation+0.000
sports
Token sports
Feature activation+0.000
-
Token -
Feature activation+0.000
the
Token the
Feature activation+0.000
draft
Token draft
Feature activation+2.522
-
Token -
Feature activation+1.234
and
Token and
Feature activation+0.000
lets
Token lets
Feature activation+0.000
you
Token you
Feature activation+0.000
do
Token do
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.081
t
Tokent
Feature activation+0.000
even
Token even
Feature activation+0.000
deer
Token deer
Feature activation+0.000
season
Token season
Feature activation+0.139
âĢĵ
Token âĢĵ
Feature activation+2.938
but
Token but
Feature activation+0.000
Curtis
Token Curtis
Feature activation+0.000
Henry
Token Henry
Feature activation+0.000
trusted
Token trusted
Feature activation+0.000
that
Token that
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
like
Tokenlike
Feature activation+0.000
a
Token a
Feature activation+0.000
shipping
Token shipping
Feature activation+0.000
container
Token container
Feature activation+1.772
âĢ
TokenâĢ
Feature activation+2.772
Ŀ
TokenĿ
Feature activation+3.042
âĢĶ
Token âĢĶ
Feature activation+0.030
not
Token not
Feature activation+0.000
only
Token only
Feature activation+0.000
jail
Token jail
Feature activation+0.000
s
Tokens
Feature activation+0.000
trademark
Token trademark
Feature activation+0.000
two
Token two
Feature activation+0.000
years
Token years
Feature activation+0.000
later
Token later
Feature activation+0.063
âĢĶ
TokenâĢĶ
Feature activation+2.802
and
Tokenand
Feature activation+0.000
that
Token that
Feature activation+0.000
the
Token the
Feature activation+0.000
company
Token company
Feature activation+0.000
also
Token also
Feature activation+0.000

INTERVAL 1.984 - 2.480
CONTAINS 0.006%

products
Token products
Feature activation+0.000
âĢĵ
Token âĢĵ
Feature activation+0.000
both
Token both
Feature activation+0.000
legal
Token legal
Feature activation+0.384
and
Token and
Feature activation+0.000
illegal
Token illegal
Feature activation+2.473
.
Token.
Feature activation+0.000
We
Token We
Feature activation+0.000
have
Token have
Feature activation+0.000
a
Token a
Feature activation+0.000
significant
Token significant
Feature activation+0.000
and
Token and
Feature activation+0.000
its
Token its
Feature activation+0.000
pickup
Token pickup
Feature activation+0.000
-
Token-
Feature activation+0.000
and
Token and
Feature activation+0.000
SUV
Token SUV
Feature activation+2.478
-
Token-
Feature activation+0.450
focused
Tokenfocused
Feature activation+0.000
Texas
Token Texas
Feature activation+0.000
Truck
Token Truck
Feature activation+0.000
Inv
Token Inv
Feature activation+0.000
72
Token 72
Feature activation+0.000
percent
Token percent
Feature activation+0.338
of
Token of
Feature activation+0.000
likely
Token likely
Feature activation+0.000
voters
Token voters
Feature activation+0.900
âĢĶ
TokenâĢĶ
Feature activation+2.252
think
Tokenthink
Feature activation+0.000
Wall
Token Wall
Feature activation+0.000
Street
Token Street
Feature activation+0.000
has
Token has
Feature activation+0.000
too
Token too
Feature activation+0.000
S
Token S
Feature activation+0.000
&
Token&
Feature activation+0.000
P
TokenP
Feature activation+0.000
and
Token and
Feature activation+0.526
F
Token F
Feature activation+0.000
itch
Tokenitch
Feature activation+2.219
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
All
TokenAll
Feature activation+0.000
investors
Token investors
Feature activation+0.000
schools
Token schools
Feature activation+0.000
to
Token to
Feature activation+0.000
cease
Token cease
Feature activation+0.000
outdoor
Token outdoor
Feature activation+0.000
activities
Token activities
Feature activation+0.212
--
Token --
Feature activation+2.322
while
Token while
Feature activation+0.000
the
Token the
Feature activation+0.000
Beijing
Token Beijing
Feature activation+0.000
Municipal
Token Municipal
Feature activation+0.000
Environmental
Token Environmental
Feature activation+0.000

INTERVAL 1.488 - 1.984
CONTAINS 0.014%

,
Token,
Feature activation+0.220
or
Token or
Feature activation+0.000
just
Token just
Feature activation+0.000
plain
Token plain
Feature activation+0.000
bad
Token bad
Feature activation+0.000
luck
Token luck
Feature activation+1.575
âĢĶ
Token âĢĶ
Feature activation+1.420
but
Token but
Feature activation+0.000
a
Token a
Feature activation+0.000
hot
Token hot
Feature activation+0.000
goaltender
Token goaltender
Feature activation+0.000
âĢĶ
TokenâĢĶ
Feature activation+0.000
including
Tokenincluding
Feature activation+0.000
weapons
Token weapons
Feature activation+0.000
of
Token of
Feature activation+0.000
mass
Token mass
Feature activation+0.000
destruction
Token destruction
Feature activation+1.502
âĢĶ
TokenâĢĶ
Feature activation+0.038
their
Tokentheir
Feature activation+0.000
goal
Token goal
Feature activation+0.000
is
Token is
Feature activation+0.000
not
Token not
Feature activation+0.000
is
Token is
Feature activation+0.000
kept
Token kept
Feature activation+0.365
anonymous
Token anonymous
Feature activation+2.397
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
indictment
Token indictment
Feature activation+1.599
âĢĵ
TokenâĢĵ
Feature activation+1.457
to
Tokento
Feature activation+0.000
provide
Token provide
Feature activation+0.000
false
Token false
Feature activation+0.000
information
Token information
Feature activation+0.000
assistants
Token assistants
Feature activation+0.552
,
Token,
Feature activation+0.000
public
Token public
Feature activation+0.000
relations
Token relations
Feature activation+0.000
staff
Token staff
Feature activation+0.839
âĢĵ
Token âĢĵ
Feature activation+1.882
all
Token all
Feature activation+0.245
working
Token working
Feature activation+0.000
in
Token in
Feature activation+0.000
a
Token a
Feature activation+0.000
tight
Token tight
Feature activation+0.000
the
Token the
Feature activation+0.000
Egyptian
Token Egyptian
Feature activation+0.000
god
Token god
Feature activation+0.000
of
Token of
Feature activation+0.000
medicine
Token medicine
Feature activation+0.000
âĢĶ
TokenâĢĶ
Feature activation+1.679
issued
Tokenissued
Feature activation+0.000
this
Token this
Feature activation+0.000
statement
Token statement
Feature activation+0.000
:
Token:
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000

INTERVAL 0.992 - 1.488
CONTAINS 0.032%

âĢĵ
Token âĢĵ
Feature activation+0.000
which
Token which
Feature activation+0.000
were
Token were
Feature activation+0.000
completely
Token completely
Feature activation+0.000
legal
Token legal
Feature activation+0.000
âĢĵ
Token âĢĵ
Feature activation+1.245
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
liberal
Token liberal
Feature activation+0.000
media
Token media
Feature activation+0.000
.
Token.
Feature activation+0.000
hunting
Token hunting
Feature activation+0.000
at
Token at
Feature activation+0.000
night
Token night
Feature activation+0.702
with
Token with
Feature activation+0.207
flash
Token flash
Feature activation+0.000
lights
Tokenlights
Feature activation+1.360
.
Token.
Feature activation+0.000
The
Token The
Feature activation+0.000
practice
Token practice
Feature activation+0.000
was
Token was
Feature activation+0.000
outlawed
Token outlawed
Feature activation+0.000
the
Token the
Feature activation+0.000
president
Token president
Feature activation+0.000
believes
Token believes
Feature activation+0.721
,
Token,
Feature activation+1.061
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+1.031
former
Token former
Feature activation+0.000
Education
Token Education
Feature activation+0.000
Secretary
Token Secretary
Feature activation+0.000
Ar
Token Ar
Feature activation+0.000
ne
Tokenne
Feature activation+0.000
its
Token its
Feature activation+0.000
low
Token low
Feature activation+0.000
-
Token-
Feature activation+0.000
impact
Tokenimpact
Feature activation+0.000
style
Token style
Feature activation+0.620
--
Token --
Feature activation+1.290
violated
Token violated
Feature activation+0.000
"
Token "
Feature activation+0.000
leave
Tokenleave
Feature activation+0.000
no
Token no
Feature activation+0.000
trace
Token trace
Feature activation+0.000
ieving
Tokenieving
Feature activation+0.000
men
Token men
Feature activation+0.000
âĢĶ
TokenâĢĶ
Feature activation+0.000
and
Tokenand
Feature activation+0.000
yes
Token yes
Feature activation+0.000
women
Token women
Feature activation+1.122
âĢĶ
TokenâĢĶ
Feature activation+1.207
would
Tokenwould
Feature activation+0.000
appreciate
Token appreciate
Feature activation+0.000
a
Token a
Feature activation+0.000
"
Token "
Feature activation+0.000

INTERVAL 0.496 - 0.992
CONTAINS 0.085%

the
Token the
Feature activation+0.000
pit
Token pit
Feature activation+0.000
bull
Token bull
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.105
s
Tokens
Feature activation+0.683
jaws
Token jaws
Feature activation+0.000
locked
Token locked
Feature activation+0.014
onto
Token onto
Feature activation+0.157
Yuri
Token Yuri
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.552
âĢĶ
Token âĢĶ
Feature activation+0.000
all
Token all
Feature activation+0.000
published
Token published
Feature activation+0.000
in
Token in
Feature activation+0.000
this
Token this
Feature activation+0.000
decade
Token decade
Feature activation+0.542
and
Token and
Feature activation+0.200
intended
Token intended
Feature activation+0.000
more
Token more
Feature activation+0.000
or
Token or
Feature activation+0.000
less
Token less
Feature activation+0.000
but
Token but
Feature activation+0.449
many
Token many
Feature activation+0.050
conservatives
Token conservatives
Feature activation+1.572
as
Token as
Feature activation+1.235
well
Token well
Feature activation+1.297
âĢĶ
Token âĢĶ
Feature activation+0.870
have
Token have
Feature activation+0.000
been
Token been
Feature activation+0.000
quick
Token quick
Feature activation+0.000
to
Token to
Feature activation+0.000
condemn
Token condemn
Feature activation+0.000
ik
Tokenik
Feature activation+0.000
arn
Tokenarn
Feature activation+0.000
ika
Tokenika
Feature activation+0.000
and
Token and
Feature activation+0.000
Raj
Token Raj
Feature activation+0.000
a
Tokena
Feature activation+0.620
Har
Token Har
Feature activation+0.000
ish
Tokenish
Feature activation+0.000
Chandra
Token Chandra
Feature activation+0.024
âĢĶ
Token âĢĶ
Feature activation+0.386
in
Token in
Feature activation+0.000
management
Token management
Feature activation+0.000
--
Token--
Feature activation+0.000
including
Tokenincluding
Feature activation+0.000
I
Token I
Feature activation+0.000
ve
Tokenve
Feature activation+0.000
--
Token--
Feature activation+0.901
to
Tokento
Feature activation+0.000
fix
Token fix
Feature activation+0.000
these
Token these
Feature activation+0.000
problems
Token problems
Feature activation+0.000
on
Token on
Feature activation+0.000

INTERVAL 0.000 - 0.496
CONTAINS 99.859%

year
Tokenyear
Feature activation+0.000
-
Token-
Feature activation+0.000
old
Tokenold
Feature activation+0.000
man
Token man
Feature activation+0.000
was
Token was
Feature activation+0.000
arrested
Token arrested
Feature activation+0.000
on
Token on
Feature activation+0.000
Wednesday
Token Wednesday
Feature activation+0.000
and
Token and
Feature activation+0.000
remains
Token remains
Feature activation+0.000
in
Token in
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Stra
TokenStra
Feature activation+0.000
uss
Tokenuss
Feature activation+0.000
-
Token-
Feature activation+0.000
K
TokenK
Feature activation+0.000
ahn
Tokenahn
Feature activation+0.000
told
Token told
Feature activation+0.000
a
Token a
Feature activation+0.000
Port
Token Port
Feature activation+0.000
Authority
Token Authority
Feature activation+0.000
detective
Token detective
Feature activation+0.000
down
Token down
Feature activation+0.000
.
Token.
Feature activation+0.000
Strength
Token Strength
Feature activation+0.000
training
Token training
Feature activation+0.000
cannot
Token cannot
Feature activation+0.000
delay
Token delay
Feature activation+0.000
the
Token the
Feature activation+0.000
breakdown
Token breakdown
Feature activation+0.000
of
Token of
Feature activation+0.000
and
Token and
Feature activation+0.000
rogen
Tokenrogen
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
For
TokenFor
Feature activation+0.000
those
Token those
Feature activation+0.000
looking
Token looking
Feature activation+0.000
to
Token to
Feature activation+0.000
get
Token get
Feature activation+0.000
in
Token in
Feature activation+0.000
with
Token with
Feature activation+0.000
the
Token the
Feature activation+0.000
a
Token a
Feature activation+0.000
deeper
Token deeper
Feature activation+0.000
attachment
Token attachment
Feature activation+0.000
to
Token to
Feature activation+0.000
Canada
Token Canada
Feature activation+0.000
and
Token and
Feature activation+0.000
feels
Token feels
Feature activation+0.000
a
Token a
Feature activation+0.000
stronger
Token stronger
Feature activation+0.000
sense
Token sense
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 6 in H7.4: (feature 1959)

TOP ACTIVATIONS
MAX = 4.016

ix
Tokenix
Feature activation+0.000
and
Token and
Feature activation+1.127
Team
Token Team
Feature activation+0.000
Liquid
Token Liquid
Feature activation+2.474
Academy
Token Academy
Feature activation+2.082
,
Token,
Feature activation+4.016
fielding
Token fielding
Feature activation+0.000
offers
Token offers
Feature activation+0.000
for
Token for
Feature activation+0.000
their
Token their
Feature activation+0.000
Challenger
Token Challenger
Feature activation+0.000
,
Token,
Feature activation+0.000
including
Token including
Feature activation+0.000
NR
Token NR
Feature activation+0.000
G
TokenG
Feature activation+0.581
eSports
Token eSports
Feature activation+1.161
,
Token,
Feature activation+3.517
Nova
Token Nova
Feature activation+0.000
eSports
Token eSports
Feature activation+0.000
,
Token,
Feature activation+0.000
E
Token E
Feature activation+0.000
an
Tokenan
Feature activation+0.000
Nos
Token Nos
Feature activation+0.000
.
Token.
Feature activation+0.268
1
Token 1
Feature activation+0.000
and
Token and
Feature activation+0.000
2
Token 2
Feature activation+0.934
,
Token,
Feature activation+2.832
in
Token in
Feature activation+0.000
a
Token a
Feature activation+0.000
civil
Token civil
Feature activation+0.000
lawsuit
Token lawsuit
Feature activation+0.000
against
Token against
Feature activation+0.000
Ghana
Token Ghana
Feature activation+0.000
ian
Tokenian
Feature activation+0.000
capital
Token capital
Feature activation+0.000
,
Token,
Feature activation+0.000
Acc
Token Acc
Feature activation+0.000
ra
Tokenra
Feature activation+2.692
,
Token,
Feature activation+0.000
who
Token who
Feature activation+0.000
said
Token said
Feature activation+0.000
they
Token they
Feature activation+0.000
wanted
Token wanted
Feature activation+0.000
the
Token the
Feature activation+0.000
New
Token New
Feature activation+0.000
Hampshire
Token Hampshire
Feature activation+0.000
Firearms
Token Firearms
Feature activation+0.000
Coalition
Token Coalition
Feature activation+0.041
,
Token,
Feature activation+2.617
when
Token when
Feature activation+0.000
he
Token he
Feature activation+0.000
asserted
Token asserted
Feature activation+0.000
that
Token that
Feature activation+0.000
violent
Token violent
Feature activation+0.000
AR
Token AR
Feature activation+0.000
-
Token-
Feature activation+0.000
15
Token15
Feature activation+0.000
r
Token r
Feature activation+0.000
affle
Tokenaffle
Feature activation+0.000
,
Token,
Feature activation+2.520
c
Token c
Feature activation+0.000
ospons
Tokenospons
Feature activation+0.000
ored
Tokenored
Feature activation+0.331
by
Token by
Feature activation+0.000
Gun
Token Gun
Feature activation+0.000
E
Token E
Feature activation+0.000
an
Tokenan
Feature activation+0.000
ix
Tokenix
Feature activation+0.000
and
Token and
Feature activation+1.127
Team
Token Team
Feature activation+0.000
Liquid
Token Liquid
Feature activation+2.474
Academy
Token Academy
Feature activation+2.082
,
Token,
Feature activation+4.016
fielding
Token fielding
Feature activation+0.000
offers
Token offers
Feature activation+0.000
for
Token for
Feature activation+0.000
counsel
Token counsel
Feature activation+0.000
,
Token,
Feature activation+0.000
Brian
Token Brian
Feature activation+0.000
H
Token H
Feature activation+0.000
aged
Tokenaged
Feature activation+0.000
orn
Tokenorn
Feature activation+2.438
,
Token,
Feature activation+1.204
an
Token an
Feature activation+0.000
appellate
Token appellate
Feature activation+0.000
court
Token court
Feature activation+0.023
judge
Token judge
Feature activation+0.235
Ŀ
TokenĿ
Feature activation+0.005
Jack
Token Jack
Feature activation+0.000
Mc
Token Mc
Feature activation+0.000
Man
TokenMan
Feature activation+0.000
us
Tokenus
Feature activation+0.000
,
Token,
Feature activation+2.435
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boxing
Token boxing
Feature activation+0.000
champion
Token champion
Feature activation+0.000
.
Token.
Feature activation+0.000
finance
Token finance
Feature activation+0.000
ruling
Token ruling
Feature activation+0.000
,
Token,
Feature activation+0.000
Citizens
Token Citizens
Feature activation+0.000
United
Token United
Feature activation+1.055
,
Token,
Feature activation+2.364
seem
Token seem
Feature activation+0.425
like
Token like
Feature activation+0.000
a
Token a
Feature activation+0.000
tr
Token tr
Feature activation+0.000
ifle
Tokenifle
Feature activation+0.000
ancient
Token ancient
Feature activation+0.000
,
Token,
Feature activation+1.862
much
Token much
Feature activation+0.000
less
Token less
Feature activation+0.044
universal
Token universal
Feature activation+0.415
,
Token,
Feature activation+2.339
will
Token will
Feature activation+0.000
fall
Token fall
Feature activation+0.000
in
Token in
Feature activation+0.000
turn
Token turn
Feature activation+0.000
?
Token?
Feature activation+0.000
almost
Token almost
Feature activation+0.000
17
Token 17
Feature activation+0.000
years
Token years
Feature activation+0.000
in
Token in
Feature activation+0.000
office
Token office
Feature activation+0.000
,
Token,
Feature activation+2.305
and
Token and
Feature activation+0.000
their
Token their
Feature activation+0.000
federal
Token federal
Feature activation+0.000
leader
Token leader
Feature activation+0.000
lost
Token lost
Feature activation+0.000
boys
Token boys
Feature activation+0.000
,
Token,
Feature activation+0.000
ages
Token ages
Feature activation+0.000
11
Token 11
Feature activation+0.000
and
Token and
Feature activation+0.000
13
Token 13
Feature activation+2.303
,
Token,
Feature activation+1.546
according
Token according
Feature activation+0.000
to
Token to
Feature activation+0.000
patch
Token patch
Feature activation+0.000
.
Token.
Feature activation+0.000
party
Tokenparty
Feature activation+0.000
clone
Token clone
Feature activation+0.000
,
Token,
Feature activation+0.000
Bon
Token Bon
Feature activation+0.000
fire
Tokenfire
Feature activation+0.000
,
Token,
Feature activation+2.232
to
Token to
Feature activation+0.000
employees
Token employees
Feature activation+0.000
and
Token and
Feature activation+0.000
by
Token by
Feature activation+0.000
early
Token early
Feature activation+0.000
interest
Token interest
Feature activation+0.000
,
Token,
Feature activation+0.000
much
Token much
Feature activation+0.000
less
Token less
Feature activation+0.000
zeal
Token zeal
Feature activation+1.609
,
Token,
Feature activation+2.195
in
Token in
Feature activation+0.000
getting
Token getting
Feature activation+0.000
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
truth
Token truth
Feature activation+0.000
family
Token family
Feature activation+0.000
,
Token,
Feature activation+0.000
including
Token including
Feature activation+0.000
young
Token young
Feature activation+0.000
William
Token William
Feature activation+0.028
,
Token,
Feature activation+2.185
then
Token then
Feature activation+0.000
age
Token age
Feature activation+0.000
9
Token 9
Feature activation+0.367
,
Token,
Feature activation+1.939
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
election
Token election
Feature activation+0.000
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
its
Token its
Feature activation+0.058
results
Token results
Feature activation+2.170
,
Token,
Feature activation+1.484
null
Token null
Feature activation+0.000
and
Token and
Feature activation+0.000
void
Token void
Feature activation+0.195
because
Token because
Feature activation+0.000
phen
Tokenphen
Feature activation+0.000
yle
Tokenyle
Feature activation+0.000
thy
Tokenthy
Feature activation+0.000
lam
Tokenlam
Feature activation+0.000
ine
Tokenine
Feature activation+1.705
,
Token,
Feature activation+2.156
in
Token in
Feature activation+0.235
readily
Token readily
Feature activation+0.000
available
Token available
Feature activation+0.000
weight
Token weight
Feature activation+0.000
loss
Token loss
Feature activation+0.000
-
Token-
Feature activation+0.000
powers
Tokenpowers
Feature activation+0.000
,
Token,
Feature activation+0.000
France
Token France
Feature activation+0.000
and
Token and
Feature activation+0.000
Britain
Token Britain
Feature activation+2.152
.
Token.
Feature activation+0.000
Washington
Token Washington
Feature activation+0.000
is
Token is
Feature activation+0.000
not
Token not
Feature activation+0.000
in
Token in
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
end
Token end
Feature activation+0.000
of
Token of
Feature activation+0.000
June
Token June
Feature activation+0.966
,
Token,
Feature activation+2.147
in
Token in
Feature activation+0.189
the
Token the
Feature activation+0.007
Texas
Token Texas
Feature activation+0.000
case
Token case
Feature activation+0.046
.
Token.
Feature activation+0.000

Top DFA by src position
MAX = 4.236

five
Token five
Feature activation+0.027
Top resid features:
of
Token of
Feature activation+0.014
Top resid features:
six
Token six
Feature activation+0.024
Top resid features:
Challenger
Token Challenger
Feature activation+0.001
Top resid features:
teams
Token teams
Feature activation+0.029
Top resid features:
,
Token,
Feature activation+4.236
Top resid features:
including
Token including
Feature activation+0.207
Top resid features:
NR
Token NR
Feature activation-0.003
Top resid features:
G
TokenG
Feature activation+0.000
Top resid features:
eSports
Token eSports
Feature activation-0.001
Top resid features:
,
Token,
Feature activation+0.029
Top resid features:
five
Token five
Feature activation+0.008
Top resid features:
of
Token of
Feature activation+0.016
Top resid features:
six
Token six
Feature activation+0.008
Top resid features:
Challenger
Token Challenger
Feature activation+0.005
Top resid features:
teams
Token teams
Feature activation+0.044
Top resid features:
,
Token,
Feature activation+3.841
Top resid features:
including
Token including
Feature activation+0.196
Top resid features:
NR
Token NR
Feature activation-0.018
Top resid features:
G
TokenG
Feature activation+0.019
Top resid features:
eSports
Token eSports
Feature activation+0.028
Top resid features:
,
Token,
Feature activation+0.085
Top resid features:
Edwards
Token Edwards
Feature activation-0.003
Top resid features:
have
Token have
Feature activation-0.012
Top resid features:
represented
Token represented
Feature activation-0.123
Top resid features:
two
Token two
Feature activation-0.001
Top resid features:
women
Token women
Feature activation+0.109
Top resid features:
,
Token,
Feature activation+2.875
Top resid features:
known
Token known
Feature activation+0.405
Top resid features:
only
Token only
Feature activation+0.019
Top resid features:
as
Token as
Feature activation+0.037
Top resid features:
Jane
Token Jane
Feature activation+0.012
Top resid features:
Does
Token Does
Feature activation+0.003
Top resid features:
in
Token in
Feature activation+0.035
Top resid features:
the
Token the
Feature activation+0.014
Top resid features:
Ghana
Token Ghana
Feature activation+0.005
Top resid features:
ian
Tokenian
Feature activation+0.016
Top resid features:
capital
Token capital
Feature activation+0.076
Top resid features:
,
Token,
Feature activation+3.237
Top resid features:
Acc
Token Acc
Feature activation+0.066
Top resid features:
ra
Tokenra
Feature activation-0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
who
Token who
Feature activation+0.000
Top resid features:
said
Token said
Feature activation+0.000
Top resid features:
an
Token an
Feature activation-0.001
Top resid features:
anti
Token anti
Feature activation+0.002
Top resid features:
-
Token-
Feature activation-0.000
Top resid features:
Brown
TokenBrown
Feature activation-0.001
Top resid features:
rally
Token rally
Feature activation+0.003
Top resid features:
,
Token,
Feature activation+2.749
Top resid features:
featuring
Token featuring
Feature activation+0.165
Top resid features:
an
Token an
Feature activation-0.003
Top resid features:
AR
Token AR
Feature activation+0.004
Top resid features:
-
Token-
Feature activation+0.000
Top resid features:
15
Token15
Feature activation+0.002
Top resid features:
an
Token an
Feature activation+0.004
Top resid features:
anti
Token anti
Feature activation+0.003
Top resid features:
-
Token-
Feature activation-0.001
Top resid features:
Brown
TokenBrown
Feature activation-0.002
Top resid features:
rally
Token rally
Feature activation+0.031
Top resid features:
,
Token,
Feature activation+2.809
Top resid features:
featuring
Token featuring
Feature activation+0.307
Top resid features:
an
Token an
Feature activation-0.016
Top resid features:
AR
Token AR
Feature activation+0.006
Top resid features:
-
Token-
Feature activation+0.002
Top resid features:
15
Token15
Feature activation+0.014
Top resid features:
five
Token five
Feature activation+0.002
Top resid features:
of
Token of
Feature activation+0.006
Top resid features:
six
Token six
Feature activation+0.001
Top resid features:
Challenger
Token Challenger
Feature activation+0.004
Top resid features:
teams
Token teams
Feature activation+0.037
Top resid features:
,
Token,
Feature activation+2.332
Top resid features:
including
Token including
Feature activation+0.653
Top resid features:
NR
Token NR
Feature activation-0.003
Top resid features:
G
TokenG
Feature activation+0.000
Top resid features:
eSports
Token eSports
Feature activation-0.001
Top resid features:
,
Token,
Feature activation+0.065
Top resid features:
making
Token making
Feature activation-0.004
Top resid features:
his
Token his
Feature activation+0.001
Top resid features:
own
Token own
Feature activation+0.001
Top resid features:
legal
Token legal
Feature activation+0.008
Top resid features:
counsel
Token counsel
Feature activation+0.032
Top resid features:
,
Token,
Feature activation+3.028
Top resid features:
Brian
Token Brian
Feature activation+0.162
Top resid features:
H
Token H
Feature activation+0.003
Top resid features:
aged
Tokenaged
Feature activation+0.004
Top resid features:
orn
Tokenorn
Feature activation+0.001
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
employ
Token employ
Feature activation+0.027
Top resid features:
an
Token an
Feature activation-0.003
Top resid features:
armed
Token armed
Feature activation+0.003
Top resid features:
boun
Token boun
Feature activation+0.001
Top resid features:
cer
Tokencer
Feature activation+0.006
Top resid features:
,
Token,
Feature activation+2.842
Top resid features:
âĢ
Token âĢ
Feature activation+0.057
Top resid features:
ľ
Tokenľ
Feature activation+0.023
Top resid features:
Eat
TokenEat
Feature activation-0.017
Top resid features:
-
Token-
Feature activation-0.000
Top resid features:
âĢ
TokenâĢ
Feature activation+0.001
Top resid features:
's
Token's
Feature activation-0.002
Top resid features:
landmark
Token landmark
Feature activation+0.017
Top resid features:
campaign
Token campaign
Feature activation-0.005
Top resid features:
finance
Token finance
Feature activation+0.001
Top resid features:
ruling
Token ruling
Feature activation+0.129
Top resid features:
,
Token,
Feature activation+2.873
Top resid features:
Citizens
Token Citizens
Feature activation+0.019
Top resid features:
United
Token United
Feature activation+0.017
Top resid features:
,
Token,
Feature activation+0.032
Top resid features:
seem
Token seem
Feature activation+0.000
Top resid features:
like
Token like
Feature activation+0.000
Top resid features:
oppose
Token oppose
Feature activation+0.004
Top resid features:
that
Token that
Feature activation+0.004
Top resid features:
other
Token other
Feature activation+0.051
Top resid features:
tab
Token tab
Feature activation+0.002
Top resid features:
oos
Tokenoos
Feature activation+0.105
Top resid features:
,
Token,
Feature activation+2.070
Top resid features:
much
Token much
Feature activation+0.271
Top resid features:
less
Token less
Feature activation+0.269
Top resid features:
ancient
Token ancient
Feature activation+0.045
Top resid features:
,
Token,
Feature activation+0.030
Top resid features:
much
Token much
Feature activation-0.036
Top resid features:
elections
Token elections
Feature activation+0.034
Top resid features:
in
Token in
Feature activation+0.022
Top resid features:
Saskatchewan
Token Saskatchewan
Feature activation-0.011
Top resid features:
and
Token and
Feature activation+0.145
Top resid features:
Manitoba
Token Manitoba
Feature activation-0.007
Top resid features:
,
Token,
Feature activation+2.254
Top resid features:
the
Token the
Feature activation+0.272
Top resid features:
latter
Token latter
Feature activation+0.047
Top resid features:
ending
Token ending
Feature activation+0.035
Top resid features:
almost
Token almost
Feature activation+0.003
Top resid features:
17
Token 17
Feature activation-0.005
Top resid features:
contact
Token contact
Feature activation-0.008
Top resid features:
with
Token with
Feature activation-0.019
Top resid features:
two
Token two
Feature activation-0.007
Top resid features:
young
Token young
Feature activation+0.020
Top resid features:
boys
Token boys
Feature activation+0.139
Top resid features:
,
Token,
Feature activation+1.994
Top resid features:
ages
Token ages
Feature activation+0.769
Top resid features:
11
Token 11
Feature activation+0.023
Top resid features:
and
Token and
Feature activation+0.099
Top resid features:
13
Token 13
Feature activation+0.020
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
demonstrating
Token demonstrating
Feature activation-0.024
Top resid features:
a
Token a
Feature activation-0.012
Top resid features:
House
Token House
Feature activation+0.008
Top resid features:
party
Tokenparty
Feature activation+0.020
Top resid features:
clone
Token clone
Feature activation+0.115
Top resid features:
,
Token,
Feature activation+2.613
Top resid features:
Bon
Token Bon
Feature activation+0.122
Top resid features:
fire
Tokenfire
Feature activation+0.006
Top resid features:
,
Token,
Feature activation+0.053
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
employees
Token employees
Feature activation+0.000
Top resid features:
have
Token have
Feature activation-0.013
Top resid features:
shown
Token shown
Feature activation-0.003
Top resid features:
no
Token no
Feature activation-0.020
Top resid features:
great
Token great
Feature activation+0.006
Top resid features:
interest
Token interest
Feature activation+0.063
Top resid features:
,
Token,
Feature activation+2.588
Top resid features:
much
Token much
Feature activation+0.221
Top resid features:
less
Token less
Feature activation+0.066
Top resid features:
zeal
Token zeal
Feature activation-0.028
Top resid features:
,
Token,
Feature activation+0.038
Top resid features:
in
Token in
Feature activation+0.000
Top resid features:
so
Token so
Feature activation-0.001
Top resid features:
Mark
Token Mark
Feature activation-0.034
Top resid features:
took
Token took
Feature activation+0.010
Top resid features:
his
Token his
Feature activation-0.020
Top resid features:
family
Token family
Feature activation+0.036
Top resid features:
,
Token,
Feature activation+2.358
Top resid features:
including
Token including
Feature activation+0.403
Top resid features:
young
Token young
Feature activation+0.037
Top resid features:
William
Token William
Feature activation+0.017
Top resid features:
,
Token,
Feature activation+0.094
Top resid features:
then
Token then
Feature activation+0.000
Top resid features:
Court
Token Court
Feature activation-0.001
Top resid features:
declared
Token declared
Feature activation-0.003
Top resid features:
the
Token the
Feature activation+0.019
Top resid features:
election
Token election
Feature activation+0.024
Top resid features:
,
Token,
Feature activation+0.197
Top resid features:
and
Token and
Feature activation+2.446
Top resid features:
its
Token its
Feature activation+0.133
Top resid features:
results
Token results
Feature activation+0.075
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
null
Token null
Feature activation+0.000
Top resid features:
and
Token and
Feature activation+0.000
Top resid features:
discouraged
Token discouraged
Feature activation+0.006
Top resid features:
-
Token-
Feature activation+0.003
Top resid features:
use
Tokenuse
Feature activation+0.001
Top resid features:
ingredients
Token ingredients
Feature activation+0.038
Top resid features:
,
Token,
Feature activation+0.504
Top resid features:
such
Token such
Feature activation+1.533
Top resid features:
as
Token as
Feature activation+0.455
Top resid features:
ep
Token ep
Feature activation-0.002
Top resid features:
hed
Tokenhed
Feature activation+0.000
Top resid features:
ra
Tokenra
Feature activation+0.003
Top resid features:
,
Token,
Feature activation+0.029
Top resid features:
of
Token of
Feature activation-0.026
Top resid features:
two
Token two
Feature activation-0.003
Top resid features:
veto
Token veto
Feature activation+0.056
Top resid features:
-
Token-
Feature activation+0.002
Top resid features:
powers
Tokenpowers
Feature activation+0.084
Top resid features:
,
Token,
Feature activation+2.780
Top resid features:
France
Token France
Feature activation+0.054
Top resid features:
and
Token and
Feature activation+0.039
Top resid features:
Britain
Token Britain
Feature activation+0.022
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
Washington
Token Washington
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.011
Top resid features:
justices
Token justices
Feature activation+0.005
Top resid features:
prepare
Token prepare
Feature activation-0.005
Top resid features:
a
Token a
Feature activation-0.006
Top resid features:
ruling
Token ruling
Feature activation+0.140
Top resid features:
,
Token,
Feature activation+2.324
Top resid features:
expected
Token expected
Feature activation+0.157
Top resid features:
by
Token by
Feature activation+0.032
Top resid features:
the
Token the
Feature activation+0.002
Top resid features:
end
Token end
Feature activation+0.032
Top resid features:
of
Token of
Feature activation+0.006
Top resid features:

Decoder Weights Distribution

Head 0: 0.07

Head 1: 0.02

Head 2: 0.02

Head 3: 0.09

Head 4: 0.54

Head 5: 0.04

Head 6: 0.03

Head 7: 0.03

Head 8: 0.05

Head 9: 0.06

Head 10: 0.03

Head 11: 0.03

Positive logits

%,2.81

%.2.71

.,2.45

*,2.42

onto2.36

/,2.33

$.2.31

-.2.26

into2.24

_.2.10

kes2.03

ar2.03

je2.02

Dresden2.01

.*1.99

*.1.98

Newsletter1.98

quet1.96

dar1.95

arty1.94

Negative logits

suffice-2.83

+++-2.34

avascript-2.16

UL-2.07

incorpor-1.92

doesnt-1.86

millenn-1.85

abound-1.81

ain-1.80

aspiration-1.79

????????-1.79

?????-1.78

++)-1.77

surely-1.77

FANT-1.71

huh-1.70

baugh-1.70

pegged-1.68

????-1.67

Technology-1.66

INTERVAL 3.615 - 4.016
CONTAINS 0.000%

ix
Tokenix
Feature activation+0.000
and
Token and
Feature activation+1.127
Team
Token Team
Feature activation+0.000
Liquid
Token Liquid
Feature activation+2.474
Academy
Token Academy
Feature activation+2.082
,
Token,
Feature activation+4.016
fielding
Token fielding
Feature activation+0.000
offers
Token offers
Feature activation+0.000
for
Token for
Feature activation+0.000
their
Token their
Feature activation+0.000
Challenger
Token Challenger
Feature activation+0.000

INTERVAL 3.213 - 3.615
CONTAINS 0.000%

,
Token,
Feature activation+0.000
including
Token including
Feature activation+0.000
NR
Token NR
Feature activation+0.000
G
TokenG
Feature activation+0.581
eSports
Token eSports
Feature activation+1.161
,
Token,
Feature activation+3.517
Nova
Token Nova
Feature activation+0.000
eSports
Token eSports
Feature activation+0.000
,
Token,
Feature activation+0.000
E
Token E
Feature activation+0.000
an
Tokenan
Feature activation+0.000

INTERVAL 2.811 - 3.213
CONTAINS 0.000%

Nos
Token Nos
Feature activation+0.000
.
Token.
Feature activation+0.268
1
Token 1
Feature activation+0.000
and
Token and
Feature activation+0.000
2
Token 2
Feature activation+0.934
,
Token,
Feature activation+2.832
in
Token in
Feature activation+0.000
a
Token a
Feature activation+0.000
civil
Token civil
Feature activation+0.000
lawsuit
Token lawsuit
Feature activation+0.000
against
Token against
Feature activation+0.000

INTERVAL 2.410 - 2.811
CONTAINS 0.001%

AR
Token AR
Feature activation+0.000
-
Token-
Feature activation+0.000
15
Token15
Feature activation+0.000
r
Token r
Feature activation+0.000
affle
Tokenaffle
Feature activation+0.000
,
Token,
Feature activation+2.520
c
Token c
Feature activation+0.000
ospons
Tokenospons
Feature activation+0.000
ored
Tokenored
Feature activation+0.331
by
Token by
Feature activation+0.000
Gun
Token Gun
Feature activation+0.000
the
Token the
Feature activation+0.000
New
Token New
Feature activation+0.000
Hampshire
Token Hampshire
Feature activation+0.000
Firearms
Token Firearms
Feature activation+0.000
Coalition
Token Coalition
Feature activation+0.041
,
Token,
Feature activation+2.617
when
Token when
Feature activation+0.000
he
Token he
Feature activation+0.000
asserted
Token asserted
Feature activation+0.000
that
Token that
Feature activation+0.000
violent
Token violent
Feature activation+0.000
Ghana
Token Ghana
Feature activation+0.000
ian
Tokenian
Feature activation+0.000
capital
Token capital
Feature activation+0.000
,
Token,
Feature activation+0.000
Acc
Token Acc
Feature activation+0.000
ra
Tokenra
Feature activation+2.692
,
Token,
Feature activation+0.000
who
Token who
Feature activation+0.000
said
Token said
Feature activation+0.000
they
Token they
Feature activation+0.000
wanted
Token wanted
Feature activation+0.000
counsel
Token counsel
Feature activation+0.000
,
Token,
Feature activation+0.000
Brian
Token Brian
Feature activation+0.000
H
Token H
Feature activation+0.000
aged
Tokenaged
Feature activation+0.000
orn
Tokenorn
Feature activation+2.438
,
Token,
Feature activation+1.204
an
Token an
Feature activation+0.000
appellate
Token appellate
Feature activation+0.000
court
Token court
Feature activation+0.023
judge
Token judge
Feature activation+0.235
Ŀ
TokenĿ
Feature activation+0.005
Jack
Token Jack
Feature activation+0.000
Mc
Token Mc
Feature activation+0.000
Man
TokenMan
Feature activation+0.000
us
Tokenus
Feature activation+0.000
,
Token,
Feature activation+2.435
a
Token a
Feature activation+0.000
former
Token former
Feature activation+0.000
boxing
Token boxing
Feature activation+0.000
champion
Token champion
Feature activation+0.000
.
Token.
Feature activation+0.000

INTERVAL 2.008 - 2.410
CONTAINS 0.002%

of
Token of
Feature activation+0.000
U
Token U
Feature activation+0.000
zi
Tokenzi
Feature activation+0.000
's
Token's
Feature activation+0.000
managers
Token managers
Feature activation+1.041
,
Token,
Feature activation+2.114
to
Token to
Feature activation+0.000
let
Token let
Feature activation+0.000
them
Token them
Feature activation+0.000
know
Token know
Feature activation+0.000
they
Token they
Feature activation+0.000
Ahmed
Token Ahmed
Feature activation+0.000
al
Token al
Feature activation+0.000
-
Token-
Feature activation+0.000
J
TokenJ
Feature activation+0.000
ab
Tokenab
Feature activation+0.000
ari
Tokenari
Feature activation+2.082
,
Token,
Feature activation+1.120
in
Token in
Feature activation+0.317
the
Token the
Feature activation+0.000
heart
Token heart
Feature activation+0.000
of
Token of
Feature activation+0.000
party
Tokenparty
Feature activation+0.000
clone
Token clone
Feature activation+0.000
,
Token,
Feature activation+0.000
Bon
Token Bon
Feature activation+0.000
fire
Tokenfire
Feature activation+0.000
,
Token,
Feature activation+2.232
to
Token to
Feature activation+0.000
employees
Token employees
Feature activation+0.000
and
Token and
Feature activation+0.000
by
Token by
Feature activation+0.000
early
Token early
Feature activation+0.000
,
Token,
Feature activation+0.000
P
Token P
Feature activation+0.000
rav
Tokenrav
Feature activation+0.000
in
Tokenin
Feature activation+0.000
Gord
Token Gord
Feature activation+0.000
han
Tokenhan
Feature activation+2.066
,
Token,
Feature activation+0.625
in
Token in
Feature activation+0.000
an
Token an
Feature activation+0.000
overnight
Token overnight
Feature activation+0.000
cabinet
Token cabinet
Feature activation+0.000
family
Token family
Feature activation+0.000
,
Token,
Feature activation+0.000
including
Token including
Feature activation+0.000
young
Token young
Feature activation+0.000
William
Token William
Feature activation+0.028
,
Token,
Feature activation+2.185
then
Token then
Feature activation+0.000
age
Token age
Feature activation+0.000
9
Token 9
Feature activation+0.367
,
Token,
Feature activation+1.939
to
Token to
Feature activation+0.000

INTERVAL 1.607 - 2.008
CONTAINS 0.005%

and
Token and
Feature activation+0.000
Customs
Token Customs
Feature activation+0.000
Enforcement
Token Enforcement
Feature activation+0.000
(
Token (
Feature activation+0.000
ICE
TokenICE
Feature activation+0.000
),
Token),
Feature activation+1.617
of
Token of
Feature activation+0.000
any
Token any
Feature activation+0.000
right
Token right
Feature activation+0.000
to
Token to
Feature activation+0.000
sue
Token sue
Feature activation+0.000
great
Token great
Feature activation+0.000
interest
Token interest
Feature activation+0.000
,
Token,
Feature activation+0.000
much
Token much
Feature activation+0.000
less
Token less
Feature activation+0.000
zeal
Token zeal
Feature activation+1.609
,
Token,
Feature activation+2.195
in
Token in
Feature activation+0.000
getting
Token getting
Feature activation+0.000
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
for
Token for
Feature activation+0.000
players
Token players
Feature activation+0.000
with
Token with
Feature activation+0.000
adjusted
Token adjusted
Feature activation+0.000
discipline
Token discipline
Feature activation+0.090
,
Token,
Feature activation+1.685
very
Token very
Feature activation+0.388
soon
Token soon
Feature activation+0.000
,"
Token,"
Feature activation+0.000
the
Token the
Feature activation+0.000
NFL
Token NFL
Feature activation+0.000
,
Token,
Feature activation+0.000
should
Token should
Feature activation+0.000
she
Token she
Feature activation+0.000
become
Token become
Feature activation+0.000
president
Token president
Feature activation+0.303
,
Token,
Feature activation+1.748
and
Token and
Feature activation+0.188
such
Token such
Feature activation+0.000
meetings
Token meetings
Feature activation+0.000
generate
Token generate
Feature activation+0.000
friendships
Token friendships
Feature activation+0.000
drug
Token drug
Feature activation+0.000
created
Token created
Feature activation+0.000
for
Token for
Feature activation+0.000
sleep
Token sleep
Feature activation+0.000
disorders
Token disorders
Feature activation+0.057
,
Token,
Feature activation+1.814
for
Token for
Feature activation+0.000
its
Token its
Feature activation+0.000
ability
Token ability
Feature activation+0.000
to
Token to
Feature activation+0.000
shar
Token shar
Feature activation+0.000

INTERVAL 1.205 - 1.607
CONTAINS 0.015%

Georgia
Token Georgia
Feature activation+0.000
commit
Token commit
Feature activation+0.367
since
Token since
Feature activation+0.000
July
Token July
Feature activation+0.000
2014
Token 2014
Feature activation+0.092
,
Token,
Feature activation+1.595
to
Token to
Feature activation+0.099
follow
Token follow
Feature activation+0.000
Rich
Token Rich
Feature activation+0.000
t
Tokent
Feature activation+0.000
to
Token to
Feature activation+0.000
,
Token,
Feature activation+0.000
instead
Token instead
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
tumor
Token tumor
Feature activation+0.056
,
Token,
Feature activation+1.426
in
Token in
Feature activation+0.013
the
Token the
Feature activation+0.000
hopes
Token hopes
Feature activation+0.000
of
Token of
Feature activation+0.000
generating
Token generating
Feature activation+0.000
him
Token him
Feature activation+0.000
,
Token,
Feature activation+0.000
with
Token with
Feature activation+0.000
good
Token good
Feature activation+0.000
reason
Token reason
Feature activation+0.085
,
Token,
Feature activation+1.381
except
Token except
Feature activation+0.000
for
Token for
Feature activation+0.000
the
Token the
Feature activation+0.000
morning
Token morning
Feature activation+0.000
of
Token of
Feature activation+0.000
for
Token for
Feature activation+0.000
$
Token $
Feature activation+0.000
275
Token275
Feature activation+0.000
Amazon
Token Amazon
Feature activation+0.000
cards
Token cards
Feature activation+0.000
,
Token,
Feature activation+1.356
on
Token on
Feature activation+0.000
how
Token how
Feature activation+0.000
and
Token and
Feature activation+0.000
why
Token why
Feature activation+0.000
they
Token they
Feature activation+0.000
the
Token the
Feature activation+0.000
end
Token end
Feature activation+0.000
of
Token of
Feature activation+0.000
his
Token his
Feature activation+0.000
twenties
Token twenties
Feature activation+1.138
,
Token,
Feature activation+1.436
when
Token when
Feature activation+0.000
he
Token he
Feature activation+0.000
was
Token was
Feature activation+0.000
just
Token just
Feature activation+0.000
one
Token one
Feature activation+0.000

INTERVAL 0.803 - 1.205
CONTAINS 0.039%

including
Token including
Feature activation+0.000
trim
Token trim
Feature activation+0.000
ming
Tokenming
Feature activation+0.000
their
Token their
Feature activation+0.000
size
Token size
Feature activation+0.000
,
Token,
Feature activation+1.018
and
Token and
Feature activation+0.000
he
Token he
Feature activation+0.000
warned
Token warned
Feature activation+0.000
that
Token that
Feature activation+0.000
otherwise
Token otherwise
Feature activation+0.000
the
Token the
Feature activation+0.000
bank
Token bank
Feature activation+0.000
and
Token and
Feature activation+0.000
insurance
Token insurance
Feature activation+0.000
sectors
Token sectors
Feature activation+0.879
,
Token,
Feature activation+0.807
which
Token which
Feature activation+0.000
are
Token are
Feature activation+0.000
able
Token able
Feature activation+0.000
to
Token to
Feature activation+0.000
control
Token control
Feature activation+0.000
beings
Token beings
Feature activation+0.000
âĢĶ
TokenâĢĶ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
G
Token G
Feature activation+0.000
entry
Tokenentry
Feature activation+0.000
âĢĶ
TokenâĢĶ
Feature activation+0.846
who
Tokenwho
Feature activation+0.000
threaten
Token threaten
Feature activation+0.000
to
Token to
Feature activation+0.000
corrupt
Token corrupt
Feature activation+0.000
their
Token their
Feature activation+0.000
a
Token a
Feature activation+0.000
particularly
Token particularly
Feature activation+0.000
vocal
Token vocal
Feature activation+0.000
Swedish
Token Swedish
Feature activation+0.000
woman
Token woman
Feature activation+0.000
,
Token,
Feature activation+1.127
stomp
Token stomp
Feature activation+0.000
ing
Tokening
Feature activation+0.000
around
Token around
Feature activation+0.000
the
Token the
Feature activation+0.000
building
Token building
Feature activation+0.000
said
Token said
Feature activation+0.116
to
Token to
Feature activation+0.000
be
Token be
Feature activation+0.000
from
Token from
Feature activation+0.000
China
Token China
Feature activation+0.178
,
Token,
Feature activation+0.849
that
Token that
Feature activation+0.000
hacked
Token hacked
Feature activation+0.000
Google
Token Google
Feature activation+0.000
in
Token in
Feature activation+0.000
2010
Token 2010
Feature activation+0.000

INTERVAL 0.402 - 0.803
CONTAINS 0.105%

high
Token high
Feature activation+0.000
-
Token-
Feature activation+0.000
TH
TokenTH
Feature activation+0.000
C
TokenC
Feature activation+0.000
levels
Token levels
Feature activation+0.000
,
Token,
Feature activation+0.683
in
Token in
Feature activation+0.009
March
Token March
Feature activation+0.000
.
Token.
Feature activation+0.000
Rivers
Token Rivers
Feature activation+0.000
said
Token said
Feature activation+0.000
whole
Token whole
Feature activation+0.000
,
Token,
Feature activation+0.000
J
Token J
Feature activation+0.000
arn
Tokenarn
Feature activation+0.000
agin
Tokenagin
Feature activation+0.000
says
Token says
Feature activation+0.564
,
Token,
Feature activation+0.000
meaning
Token meaning
Feature activation+0.000
at
Token at
Feature activation+0.000
least
Token least
Feature activation+0.000
30
Token 30
Feature activation+0.000
David
Token David
Feature activation+0.000
Hyde
Token Hyde
Feature activation+0.000
Pierce
Token Pierce
Feature activation+0.000
,
Token,
Feature activation+0.000
among
Token among
Feature activation+0.014
others
Token others
Feature activation+0.444
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
cent
Token cent
Feature activation+0.000
for
Token for
Feature activation+0.000
Sob
Token Sob
Feature activation+0.000
yan
Tokenyan
Feature activation+0.000
in
Tokenin
Feature activation+0.000
,
Token,
Feature activation+0.526
who
Token who
Feature activation+0.000
narrowly
Token narrowly
Feature activation+0.000
avoided
Token avoided
Feature activation+0.000
a
Token a
Feature activation+0.000
second
Token second
Feature activation+0.000
s
Tokens
Feature activation+0.026
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.068
daughter
Token daughter
Feature activation+0.000
,
Token,
Feature activation+0.482
and
Token and
Feature activation+0.416
each
Token each
Feature activation+0.000
har
Token har
Feature activation+0.000
bors
Tokenbors
Feature activation+0.000
its
Token its
Feature activation+0.000

INTERVAL 0.000 - 0.402
CONTAINS 99.833%

de
Token de
Feature activation+0.000
la
Token la
Feature activation+0.000
Rep
Token Rep
Feature activation+0.000
ú
Tokenú
Feature activation+0.000
b
Tokenb
Feature activation+0.000
lic
Tokenlic
Feature activation+0.000
a
Tokena
Feature activation+0.000
(
Token (
Feature activation+0.000
PG
TokenPG
Feature activation+0.000
R
TokenR
Feature activation+0.000
)
Token)
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
These
TokenThese
Feature activation+0.000
protests
Token protests
Feature activation+0.000
are
Token are
Feature activation+0.000
different
Token different
Feature activation+0.000
from
Token from
Feature activation+0.000
earlier
Token earlier
Feature activation+0.000
ones
Token ones
Feature activation+0.000
because
Token because
Feature activation+0.000
the
Token the
Feature activation+0.000
Kurdish
Token Kurdish
Feature activation+0.000
the
Token the
Feature activation+0.000
past
Token past
Feature activation+0.000
I
Token I
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
ve
Tokenve
Feature activation+0.000
had
Token had
Feature activation+0.000
trouble
Token trouble
Feature activation+0.000
with
Token with
Feature activation+0.000
taking
Token taking
Feature activation+0.000
on
Token on
Feature activation+0.000
day
Token day
Feature activation+0.000
from
Token from
Feature activation+0.000
people
Token people
Feature activation+0.000
in
Token in
Feature activation+0.000
London
Token London
Feature activation+0.000
,
Token,
Feature activation+0.000
Berkshire
Token Berkshire
Feature activation+0.000
,
Token,
Feature activation+0.000
Oxford
Token Oxford
Feature activation+0.000
shire
Tokenshire
Feature activation+0.121
,
Token,
Feature activation+0.000
Adelaide
Token Adelaide
Feature activation+0.000
Test
Token Test
Feature activation+0.000
against
Token against
Feature activation+0.000
South
Token South
Feature activation+0.000
Africa
Token Africa
Feature activation+0.000
will
Token will
Feature activation+0.000
be
Token be
Feature activation+0.000
a
Token a
Feature activation+0.000
day
Token day
Feature activation+0.000
-
Token-
Feature activation+0.000
night
Tokennight
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 H7.4: (feature 42471)

TOP ACTIVATIONS
MAX = 8.126

,
Token,
Feature activation+0.000
Fra
Token Fra
Feature activation+0.000
ule
Tokenule
Feature activation+0.000
in
Tokenin
Feature activation+0.000
Maria
Token Maria
Feature activation+0.000
?
Token?
Feature activation+8.126
It
Token It
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
the
Token the
Feature activation+0.000
her
Token her
Feature activation+0.000
body
Token body
Feature activation+0.000
scrub
Token scrub
Feature activation+0.000
bers
Tokenbers
Feature activation+0.000
went
Token went
Feature activation+0.112
.
Token.
Feature activation+3.969
also
Tokenalso
Feature activation+0.000
they
Token they
Feature activation+0.000
compress
Token compress
Feature activation+0.000
,
Token,
Feature activation+0.000
have
Tokenhave
Feature activation+0.000
did
Token did
Feature activation+0.000
not
Token not
Feature activation+0.000
report
Token report
Feature activation+0.000
a
Token a
Feature activation+0.000
flare
Token flare
Feature activation+0.000
.
Token.
Feature activation+3.262
Ċ
TokenĊ
Feature activation+0.224
Ċ
TokenĊ
Feature activation+2.350
Materials
TokenMaterials
Feature activation+0.000
and
Token and
Feature activation+0.000
Methods
Token Methods
Feature activation+0.206
which
Token which
Feature activation+0.000
offers
Token offers
Feature activation+0.000
infinite
Token infinite
Feature activation+0.000
design
Token design
Feature activation+0.000
possibilities
Token possibilities
Feature activation+0.000
!
Token!
Feature activation+3.222
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.397
Everything
TokenEverything
Feature activation+0.000
start
Token start
Feature activation+0.000
with
Token with
Feature activation+0.000
ISON
TokenISON
Feature activation+1.915
PHOTO
Token PHOTO
Feature activation+2.345
SP
Token SP
Feature activation+0.000
E
TokenE
Feature activation+0.000
AK
TokenAK
Feature activation+0.000
S
TokenS
Feature activation+2.964
OUT
Token OUT
Feature activation+1.663
Ċ
TokenĊ
Feature activation+2.498
Ċ
TokenĊ
Feature activation+1.076
Y
TokenY
Feature activation+0.000
ELL
TokenELL
Feature activation+0.000
o
Tokeno
Feature activation+0.000
vs
Token vs
Feature activation+0.000
Montreal
Token Montreal
Feature activation+0.832
All
Token All
Feature activation+0.000
-
Token-
Feature activation+0.000
Stars
TokenStars
Feature activation+2.755
at
Token at
Feature activation+0.000
Van
Token Van
Feature activation+0.000
ier
Tokenier
Feature activation+0.000
College
Token College
Feature activation+0.026
Ċ
TokenĊ
Feature activation+2.176
stupid
Token stupid
Feature activation+0.000
long
Token long
Feature activation+0.000
range
Token range
Feature activation+0.000
land
Token land
Feature activation+0.000
movement
Token movement
Feature activation+0.000
-
Token -
Feature activation+2.722
The
Token The
Feature activation+0.000
Bishop
Token Bishop
Feature activation+0.000
ric
Tokenric
Feature activation+0.000
of
Token of
Feature activation+0.000
Rome
Token Rome
Feature activation+0.000
Î
TokenÎ
Feature activation+0.000
ķ
Tokenķ
Feature activation+0.935
Î
TokenÎ
Feature activation+0.000
Ķ
TokenĶ
Feature activation+0.920
Î
TokenÎ
Feature activation+0.000
¥
Token¥
Feature activation+2.570
Î
TokenÎ
Feature activation+0.000
£
Token£
Feature activation+1.199
;
Token;
Feature activation+2.277
Att
Token Att
Feature activation+0.000
ic
Tokenic
Feature activation+0.000
SP
Token SP
Feature activation+0.000
E
TokenE
Feature activation+0.000
AK
TokenAK
Feature activation+0.000
S
TokenS
Feature activation+2.964
OUT
Token OUT
Feature activation+1.663
Ċ
TokenĊ
Feature activation+2.498
Ċ
TokenĊ
Feature activation+1.076
Y
TokenY
Feature activation+0.000
ELL
TokenELL
Feature activation+0.000
OW
TokenOW
Feature activation+0.000
ST
TokenST
Feature activation+0.000
report
Token report
Feature activation+0.000
a
Token a
Feature activation+0.000
flare
Token flare
Feature activation+0.000
.
Token.
Feature activation+3.262
Ċ
TokenĊ
Feature activation+0.224
Ċ
TokenĊ
Feature activation+2.350
Materials
TokenMaterials
Feature activation+0.000
and
Token and
Feature activation+0.000
Methods
Token Methods
Feature activation+0.206
:
Token:
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
V
Token V
Feature activation+0.000
IR
TokenIR
Feature activation+0.000
AL
TokenAL
Feature activation+0.000
B
Token B
Feature activation+0.000
ISON
TokenISON
Feature activation+1.915
PHOTO
Token PHOTO
Feature activation+2.345
SP
Token SP
Feature activation+0.000
E
TokenE
Feature activation+0.000
AK
TokenAK
Feature activation+0.000
S
TokenS
Feature activation+2.964
OUT
Token OUT
Feature activation+1.663
Ķ
TokenĶ
Feature activation+0.920
Î
TokenÎ
Feature activation+0.000
¥
Token¥
Feature activation+2.570
Î
TokenÎ
Feature activation+0.000
£
Token£
Feature activation+1.199
;
Token;
Feature activation+2.277
Att
Token Att
Feature activation+0.000
ic
Tokenic
Feature activation+0.000
:
Token:
Feature activation+0.435
á
Token á
Feature activation+0.000
¼
Token¼
Feature activation+0.000
Stars
TokenStars
Feature activation+2.755
at
Token at
Feature activation+0.000
Van
Token Van
Feature activation+0.000
ier
Tokenier
Feature activation+0.000
College
Token College
Feature activation+0.026
Ċ
TokenĊ
Feature activation+2.176
Ċ
TokenĊ
Feature activation+1.694
Sun
TokenSun
Feature activation+0.000
Aug
Token Aug
Feature activation+0.000
9
Token 9
Feature activation+0.000
:
Token:
Feature activation+0.000
pl
Tokenpl
Feature activation+0.000
os
Tokenos
Feature activation+0.000
.
Token.
Feature activation+0.000
org
Tokenorg
Feature activation+0.109
Ċ
TokenĊ
Feature activation+0.031
Ċ
TokenĊ
Feature activation+2.135
Image
TokenImage
Feature activation+0.000
courtesy
Token courtesy
Feature activation+0.000
of
Token of
Feature activation+0.000
Wikimedia
Token Wikimedia
Feature activation+0.000
Commons
Token Commons
Feature activation+0.000
themselves
Token themselves
Feature activation+0.000
to
Token to
Feature activation+0.339
ext
Token ext
Feature activation+0.000
ran
Tokenran
Feature activation+0.000
ational
Tokenational
Feature activation+0.000
perception
Token perception
Feature activation+2.060
,
Token,
Feature activation+0.000
thinking
Token thinking
Feature activation+0.000
and
Token and
Feature activation+0.000
activity
Token activity
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.233
Mobile
TokenMobile
Feature activation+0.000
Lep
Token Lep
Feature activation+0.000
re
Tokenre
Feature activation+0.000
ch
Tokench
Feature activation+0.000
aun
Tokenaun
Feature activation+0.000
."
Token."
Feature activation+1.950
Ċ
TokenĊ
Feature activation+0.362
Ċ
TokenĊ
Feature activation+1.081
At
TokenAt
Feature activation+0.000
the
Token the
Feature activation+0.000
center
Token center
Feature activation+0.000
OK
TokenOK
Feature activation+0.000
V
Token V
Feature activation+0.000
IR
TokenIR
Feature activation+0.000
AL
TokenAL
Feature activation+0.000
B
Token B
Feature activation+0.000
ISON
TokenISON
Feature activation+1.915
PHOTO
Token PHOTO
Feature activation+2.345
SP
Token SP
Feature activation+0.000
E
TokenE
Feature activation+0.000
AK
TokenAK
Feature activation+0.000
S
TokenS
Feature activation+2.964
E
Token E
Feature activation+0.000
UTH
TokenUTH
Feature activation+0.000
AN
TokenAN
Feature activation+0.000
IZ
TokenIZ
Feature activation+0.000
ED
TokenED
Feature activation+0.000
Ċ
TokenĊ
Feature activation+1.800
Ċ
TokenĊ
Feature activation+0.597
Y
TokenY
Feature activation+0.000
ELL
TokenELL
Feature activation+0.000
OW
TokenOW
Feature activation+0.000
ST
TokenST
Feature activation+0.000
at
Token at
Feature activation+0.000
Van
Token Van
Feature activation+0.000
ier
Tokenier
Feature activation+0.000
College
Token College
Feature activation+0.026
Ċ
TokenĊ
Feature activation+2.176
Ċ
TokenĊ
Feature activation+1.694
Sun
TokenSun
Feature activation+0.000
Aug
Token Aug
Feature activation+0.000
9
Token 9
Feature activation+0.000
:
Token:
Feature activation+0.000
Val
Token Val
Feature activation+0.000
PHOTO
Token PHOTO
Feature activation+2.345
SP
Token SP
Feature activation+0.000
E
TokenE
Feature activation+0.000
AK
TokenAK
Feature activation+0.000
S
TokenS
Feature activation+2.964
OUT
Token OUT
Feature activation+1.663
Ċ
TokenĊ
Feature activation+2.498
Ċ
TokenĊ
Feature activation+1.076
Y
TokenY
Feature activation+0.000
ELL
TokenELL
Feature activation+0.000
OW
TokenOW
Feature activation+0.000

Top DFA by src position
MAX = 9.077

<|endoftext|>
Token<|endoftext|>
Feature activation+0.384
Top resid features:
:
Token:
Feature activation+9.077
Top resid features:
Where
Token Where
Feature activation+0.059
Top resid features:
are
Token are
Feature activation-0.009
Top resid features:
we
Token we
Feature activation-0.001
Top resid features:
going
Token going
Feature activation+0.031
Top resid features:
,
Token,
Feature activation+0.008
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.559
Top resid features:
:
Token:
Feature activation+4.310
Top resid features:
Not
TokenNot
Feature activation+0.228
Top resid features:
pretty
Token pretty
Feature activation+0.052
Top resid features:
if
Token if
Feature activation-0.022
Top resid features:
your
Token your
Feature activation+0.060
Top resid features:
girlfriend
Token girlfriend
Feature activation+0.099
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.560
Top resid features:
:
Token:
Feature activation+3.004
Top resid features:
Ċ
TokenĊ
Feature activation+0.092
Top resid features:
Ċ
TokenĊ
Feature activation+0.086
Top resid features:
We
TokenWe
Feature activation+0.115
Top resid features:
compared
Token compared
Feature activation+0.167
Top resid features:
culture
Token culture
Feature activation-0.004
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.471
Top resid features:
:
Token:
Feature activation+3.896
Top resid features:
Ċ
TokenĊ
Feature activation+0.093
Top resid features:
Ċ
TokenĊ
Feature activation+0.069
Top resid features:
We
TokenWe
Feature activation+0.041
Top resid features:
have
Token have
Feature activation+0.023
Top resid features:
thus
Token thus
Feature activation+0.024
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.214
Top resid features:
:
Token:
Feature activation+3.628
Top resid features:
LOC
Token LOC
Feature activation+0.032
Top resid features:
AL
TokenAL
Feature activation+0.014
Top resid features:
WOM
Token WOM
Feature activation+0.013
Top resid features:
AN
TokenAN
Feature activation+0.011
Top resid features:
WHO
Token WHO
Feature activation+0.103
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.398
Top resid features:
:
Token:
Feature activation+3.864
Top resid features:
Val
Token Val
Feature activation+0.028
Top resid features:
par
Tokenpar
Feature activation+0.004
Top resid features:
ais
Tokenais
Feature activation+0.008
Top resid features:
o
Tokeno
Feature activation+0.019
Top resid features:
vs
Token vs
Feature activation-0.295
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.542
Top resid features:
:
Token:
Feature activation+3.378
Top resid features:
Fixed
Token Fixed
Feature activation+0.019
Top resid features:
an
Token an
Feature activation+0.066
Top resid features:
issue
Token issue
Feature activation+0.059
Top resid features:
with
Token with
Feature activation+0.049
Top resid features:
stupid
Token stupid
Feature activation+0.019
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.462
Top resid features:
:
Token:
Feature activation+2.985
Top resid features:
Î
Token Î
Feature activation+0.389
Top resid features:
ķ
Tokenķ
Feature activation+0.047
Top resid features:
Î
TokenÎ
Feature activation+0.022
Top resid features:
£
Token£
Feature activation+0.011
Top resid features:
Î
TokenÎ
Feature activation-0.015
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.492
Top resid features:
:
Token:
Feature activation+3.405
Top resid features:
LOC
Token LOC
Feature activation+0.027
Top resid features:
AL
TokenAL
Feature activation+0.006
Top resid features:
WOM
Token WOM
Feature activation+0.014
Top resid features:
AN
TokenAN
Feature activation+0.043
Top resid features:
WHO
Token WHO
Feature activation+0.097
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.443
Top resid features:
:
Token:
Feature activation+2.213
Top resid features:
Ċ
TokenĊ
Feature activation+0.308
Top resid features:
Ċ
TokenĊ
Feature activation+0.044
Top resid features:
We
TokenWe
Feature activation+0.030
Top resid features:
compared
Token compared
Feature activation+0.071
Top resid features:
culture
Token culture
Feature activation-0.004
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.429
Top resid features:
:
Token:
Feature activation+3.309
Top resid features:
LOC
Token LOC
Feature activation+0.043
Top resid features:
AL
TokenAL
Feature activation+0.019
Top resid features:
WOM
Token WOM
Feature activation+0.017
Top resid features:
AN
TokenAN
Feature activation+0.021
Top resid features:
WHO
Token WHO
Feature activation+0.085
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.486
Top resid features:
:
Token:
Feature activation+2.514
Top resid features:
Î
Token Î
Feature activation+0.303
Top resid features:
ķ
Tokenķ
Feature activation+0.044
Top resid features:
Î
TokenÎ
Feature activation+0.018
Top resid features:
£
Token£
Feature activation+0.032
Top resid features:
Î
TokenÎ
Feature activation+0.008
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.535
Top resid features:
:
Token:
Feature activation+2.751
Top resid features:
Val
Token Val
Feature activation+0.018
Top resid features:
par
Tokenpar
Feature activation-0.012
Top resid features:
ais
Tokenais
Feature activation+0.001
Top resid features:
o
Tokeno
Feature activation+0.016
Top resid features:
vs
Token vs
Feature activation+0.012
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.498
Top resid features:
:
Token:
Feature activation+2.867
Top resid features:
www
Token www
Feature activation+0.004
Top resid features:
.
Token.
Feature activation-0.001
Top resid features:
j
Tokenj
Feature activation+0.002
Top resid features:
ournals
Tokenournals
Feature activation+0.022
Top resid features:
.
Token.
Feature activation+0.013
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.437
Top resid features:
:
Token:
Feature activation+2.893
Top resid features:
Ċ
TokenĊ
Feature activation+0.028
Top resid features:
Ċ
TokenĊ
Feature activation+0.025
Top resid features:
âĢ
TokenâĢ
Feature activation+0.031
Top resid features:
ľ
Tokenľ
Feature activation+0.020
Top resid features:
Let
TokenLet
Feature activation+0.060
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.540
Top resid features:
:
Token:
Feature activation+2.551
Top resid features:
the
Token the
Feature activation+0.325
Top resid features:
"
Token "
Feature activation+0.020
Top resid features:
Mobile
TokenMobile
Feature activation+0.019
Top resid features:
Lep
Token Lep
Feature activation-0.020
Top resid features:
re
Tokenre
Feature activation-0.016
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.318
Top resid features:
:
Token:
Feature activation+2.890
Top resid features:
LOC
Token LOC
Feature activation+0.039
Top resid features:
AL
TokenAL
Feature activation+0.014
Top resid features:
WOM
Token WOM
Feature activation+0.022
Top resid features:
AN
TokenAN
Feature activation+0.063
Top resid features:
WHO
Token WHO
Feature activation+0.246
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.540
Top resid features:
:
Token:
Feature activation+2.595
Top resid features:
LOC
Token LOC
Feature activation+0.005
Top resid features:
AL
TokenAL
Feature activation+0.001
Top resid features:
WOM
Token WOM
Feature activation+0.001
Top resid features:
AN
TokenAN
Feature activation+0.016
Top resid features:
WHO
Token WHO
Feature activation+0.036
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.535
Top resid features:
:
Token:
Feature activation+2.336
Top resid features:
Val
Token Val
Feature activation+0.024
Top resid features:
par
Tokenpar
Feature activation-0.032
Top resid features:
ais
Tokenais
Feature activation+0.000
Top resid features:
o
Tokeno
Feature activation+0.014
Top resid features:
vs
Token vs
Feature activation+0.034
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.253
Top resid features:
:
Token:
Feature activation+2.299
Top resid features:
LOC
Token LOC
Feature activation+0.027
Top resid features:
AL
TokenAL
Feature activation+0.009
Top resid features:
WOM
Token WOM
Feature activation+0.009
Top resid features:
AN
TokenAN
Feature activation+0.018
Top resid features:
WHO
Token WHO
Feature activation+0.124
Top resid features:

Decoder Weights Distribution

Head 0: 0.04

Head 1: 0.02

Head 2: 0.01

Head 3: 0.05

Head 4: 0.54

Head 5: 0.06

Head 6: 0.06

Head 7: 0.02

Head 8: 0.04

Head 9: 0.13

Head 10: 0.01

Head 11: 0.02

Positive logits

)]3.54

·3.07

));2.99

][2.80

])2.76

),2.74

);2.73

)2.66

];2.61

................2.57

):2.54

]2.54

],2.53

))2.47

)))2.46

2.46

Likes2.31

..............2.31

}}2.21

  2.21

Negative logits

bleacher-2.05

outwe-2.02

momentarily-1.98

morp-1.94

dipped-1.87

Magikarp-1.85

outweigh-1.85

UV-1.81

plummet-1.80

tert-1.80

dipping-1.76

staggered-1.76

halves-1.76

commuting-1.75

temporarily-1.74

submar-1.73

unden-1.72

insignificant-1.71

advoc-1.71

umble-1.70

INTERVAL 7.314 - 8.126
CONTAINS 0.000%

,
Token,
Feature activation+0.000
Fra
Token Fra
Feature activation+0.000
ule
Tokenule
Feature activation+0.000
in
Tokenin
Feature activation+0.000
Maria
Token Maria
Feature activation+0.000
?
Token?
Feature activation+8.126
It
Token It
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
the
Token the
Feature activation+0.000

INTERVAL 6.501 - 7.314
CONTAINS 0.000%

INTERVAL 5.688 - 6.501
CONTAINS 0.000%

INTERVAL 4.876 - 5.688
CONTAINS 0.000%

INTERVAL 4.063 - 4.876
CONTAINS 0.000%

INTERVAL 3.251 - 4.063
CONTAINS 0.000%

her
Token her
Feature activation+0.000
body
Token body
Feature activation+0.000
scrub
Token scrub
Feature activation+0.000
bers
Tokenbers
Feature activation+0.000
went
Token went
Feature activation+0.112
.
Token.
Feature activation+3.969
also
Tokenalso
Feature activation+0.000
they
Token they
Feature activation+0.000
compress
Token compress
Feature activation+0.000
,
Token,
Feature activation+0.000
have
Tokenhave
Feature activation+0.000
did
Token did
Feature activation+0.000
not
Token not
Feature activation+0.000
report
Token report
Feature activation+0.000
a
Token a
Feature activation+0.000
flare
Token flare
Feature activation+0.000
.
Token.
Feature activation+3.262
Ċ
TokenĊ
Feature activation+0.224
Ċ
TokenĊ
Feature activation+2.350
Materials
TokenMaterials
Feature activation+0.000
and
Token and
Feature activation+0.000
Methods
Token Methods
Feature activation+0.206

INTERVAL 2.438 - 3.251
CONTAINS 0.001%

o
Tokeno
Feature activation+0.000
vs
Token vs
Feature activation+0.000
Montreal
Token Montreal
Feature activation+0.832
All
Token All
Feature activation+0.000
-
Token-
Feature activation+0.000
Stars
TokenStars
Feature activation+2.755
at
Token at
Feature activation+0.000
Van
Token Van
Feature activation+0.000
ier
Tokenier
Feature activation+0.000
College
Token College
Feature activation+0.026
Ċ
TokenĊ
Feature activation+2.176
stupid
Token stupid
Feature activation+0.000
long
Token long
Feature activation+0.000
range
Token range
Feature activation+0.000
land
Token land
Feature activation+0.000
movement
Token movement
Feature activation+0.000
-
Token -
Feature activation+2.722
The
Token The
Feature activation+0.000
Bishop
Token Bishop
Feature activation+0.000
ric
Tokenric
Feature activation+0.000
of
Token of
Feature activation+0.000
Rome
Token Rome
Feature activation+0.000
SP
Token SP
Feature activation+0.000
E
TokenE
Feature activation+0.000
AK
TokenAK
Feature activation+0.000
S
TokenS
Feature activation+2.964
OUT
Token OUT
Feature activation+1.663
Ċ
TokenĊ
Feature activation+2.498
Ċ
TokenĊ
Feature activation+1.076
Y
TokenY
Feature activation+0.000
ELL
TokenELL
Feature activation+0.000
OW
TokenOW
Feature activation+0.000
ST
TokenST
Feature activation+0.000
Î
TokenÎ
Feature activation+0.000
ķ
Tokenķ
Feature activation+0.935
Î
TokenÎ
Feature activation+0.000
Ķ
TokenĶ
Feature activation+0.920
Î
TokenÎ
Feature activation+0.000
¥
Token¥
Feature activation+2.570
Î
TokenÎ
Feature activation+0.000
£
Token£
Feature activation+1.199
;
Token;
Feature activation+2.277
Att
Token Att
Feature activation+0.000
ic
Tokenic
Feature activation+0.000
which
Token which
Feature activation+0.000
offers
Token offers
Feature activation+0.000
infinite
Token infinite
Feature activation+0.000
design
Token design
Feature activation+0.000
possibilities
Token possibilities
Feature activation+0.000
!
Token!
Feature activation+3.222
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.397
Everything
TokenEverything
Feature activation+0.000
start
Token start
Feature activation+0.000
with
Token with
Feature activation+0.000

INTERVAL 1.625 - 2.438
CONTAINS 0.001%

Ķ
TokenĶ
Feature activation+0.920
Î
TokenÎ
Feature activation+0.000
¥
Token¥
Feature activation+2.570
Î
TokenÎ
Feature activation+0.000
£
Token£
Feature activation+1.199
;
Token;
Feature activation+2.277
Att
Token Att
Feature activation+0.000
ic
Tokenic
Feature activation+0.000
:
Token:
Feature activation+0.435
á
Token á
Feature activation+0.000
¼
Token¼
Feature activation+0.000
at
Token at
Feature activation+0.000
Van
Token Van
Feature activation+0.000
ier
Tokenier
Feature activation+0.000
College
Token College
Feature activation+0.026
Ċ
TokenĊ
Feature activation+2.176
Ċ
TokenĊ
Feature activation+1.694
Sun
TokenSun
Feature activation+0.000
Aug
Token Aug
Feature activation+0.000
9
Token 9
Feature activation+0.000
:
Token:
Feature activation+0.000
Val
Token Val
Feature activation+0.000
themselves
Token themselves
Feature activation+0.000
to
Token to
Feature activation+0.339
ext
Token ext
Feature activation+0.000
ran
Tokenran
Feature activation+0.000
ational
Tokenational
Feature activation+0.000
perception
Token perception
Feature activation+2.060
,
Token,
Feature activation+0.000
thinking
Token thinking
Feature activation+0.000
and
Token and
Feature activation+0.000
activity
Token activity
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.233
V
Token V
Feature activation+0.000
IR
TokenIR
Feature activation+0.000
AL
TokenAL
Feature activation+0.000
B
Token B
Feature activation+0.000
ISON
TokenISON
Feature activation+1.915
PHOTO
Token PHOTO
Feature activation+2.345
SP
Token SP
Feature activation+0.000
E
TokenE
Feature activation+0.000
AK
TokenAK
Feature activation+0.000
S
TokenS
Feature activation+2.964
OUT
Token OUT
Feature activation+1.663
pl
Tokenpl
Feature activation+0.000
os
Tokenos
Feature activation+0.000
.
Token.
Feature activation+0.000
org
Tokenorg
Feature activation+0.109
Ċ
TokenĊ
Feature activation+0.031
Ċ
TokenĊ
Feature activation+2.135
Image
TokenImage
Feature activation+0.000
courtesy
Token courtesy
Feature activation+0.000
of
Token of
Feature activation+0.000
Wikimedia
Token Wikimedia
Feature activation+0.000
Commons
Token Commons
Feature activation+0.000

INTERVAL 0.813 - 1.625
CONTAINS 0.002%

Val
Token Val
Feature activation+0.000
par
Tokenpar
Feature activation+0.000
ais
Tokenais
Feature activation+0.000
o
Tokeno
Feature activation+0.000
vs
Token vs
Feature activation+0.000
Montreal
Token Montreal
Feature activation+0.832
All
Token All
Feature activation+0.000
-
Token-
Feature activation+0.000
Stars
TokenStars
Feature activation+2.755
at
Token at
Feature activation+0.000
Van
Token Van
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
:
Token:
Feature activation+0.000
American
Token American
Feature activation+0.000
Barbarian
Token Barbarian
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.081
Ċ
TokenĊ
Feature activation+1.202
Int
TokenInt
Feature activation+0.000
ro
Tokenro
Feature activation+0.000
[
Token [
Feature activation+0.000
0
Token0
Feature activation+0.000
:
Token:
Feature activation+0.000
=
Token=
Feature activation+0.592
|
Token|
Feature activation+0.000
|
Token |
Feature activation+0.000
=
Token=
Feature activation+0.000
----------------------------------------------------------------
Token----------------------------------------------------------------
Feature activation+0.000
------------
Token------------
Feature activation+0.943
=
Token=
Feature activation+1.326
|
Token|
Feature activation+0.404
|
Token |
Feature activation+0.000
=
Token=
Feature activation+0.000
----------------------------------------------------------------
Token----------------------------------------------------------------
Feature activation+0.000
Ï
TokenÏ
Feature activation+0.000
ľ
Tokenľ
Feature activation+1.017
Î
TokenÎ
Feature activation+0.000
ķ
Tokenķ
Feature activation+0.935
Î
TokenÎ
Feature activation+0.000
Ķ
TokenĶ
Feature activation+0.920
Î
TokenÎ
Feature activation+0.000
¥
Token¥
Feature activation+2.570
Î
TokenÎ
Feature activation+0.000
£
Token£
Feature activation+1.199
;
Token;
Feature activation+2.277
re
Tokenre
Feature activation+0.000
ch
Tokench
Feature activation+0.000
aun
Tokenaun
Feature activation+0.000
."
Token."
Feature activation+1.950
Ċ
TokenĊ
Feature activation+0.362
Ċ
TokenĊ
Feature activation+1.081
At
TokenAt
Feature activation+0.000
the
Token the
Feature activation+0.000
center
Token center
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000

INTERVAL 0.000 - 0.813
CONTAINS 99.996%

IB
TokenIB
Feature activation+0.000
legislative
Token legislative
Feature activation+0.000
affairs
Token affairs
Feature activation+0.000
manager
Token manager
Feature activation+0.000
Kevin
Token Kevin
Feature activation+0.000
Kuh
Token Kuh
Feature activation+0.000
l
Tokenl
Feature activation+0.000
man
Tokenman
Feature activation+0.000
said
Token said
Feature activation+0.000
in
Token in
Feature activation+0.000
a
Token a
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
title
Token title
Feature activation+0.000
of
Token of
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
Captain
TokenCaptain
Feature activation+0.000
Canada
Token Canada
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
.
Token.
Feature activation+0.000
Har
Token Har
Feature activation+0.000
po
Tokenpo
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
and
Token and
Feature activation+0.000
making
Token making
Feature activation+0.000
a
Token a
Feature activation+0.000
run
Token run
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
inform
Token inform
Feature activation+0.000
decision
Token decision
Feature activation+0.000
-
Token-
Feature activation+0.000
making
Tokenmaking
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
âĢĶ
TokenâĢĶ
Feature activation+0.000
a
Token a
Feature activation+0.000
deficit
Token deficit
Feature activation+0.000
that
Token that
Feature activation+0.000
,
Token,
Feature activation+0.000
junior
Token junior
Feature activation+0.000
rights
Token rights
Feature activation+0.000
before
Token before
Feature activation+0.000
the
Token the
Feature activation+0.000
January
Token January
Feature activation+0.000
trade
Token trade
Feature activation+0.000
deadline
Token deadline
Feature activation+0.000
last
Token last
Feature activation+0.000
season
Token season
Feature activation+0.000
.
Token.
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 8 in H7.4: (feature 48243)

TOP ACTIVATIONS
MAX = 4.843

Rescue
Token Rescue
Feature activation+0.000
Force
Token Force
Feature activation+0.000
(
Token (
Feature activation+0.000
N
Token N
Feature activation+0.000
DR
TokenDR
Feature activation+2.313
F
TokenF
Feature activation+4.843
),
Token ),
Feature activation+0.000
police
Token police
Feature activation+0.000
and
Token and
Feature activation+0.000
fire
Token fire
Feature activation+0.000
service
Token service
Feature activation+0.000
superhero
Token superhero
Feature activation+0.000
underwear
Token underwear
Feature activation+0.000
.
Token.
Feature activation+0.000
(
Token (
Feature activation+0.000
Supp
Token Supp
Feature activation+0.000
lied
Tokenlied
Feature activation+4.552
)
Token )
Feature activation+0.526
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
T
TokenT
Feature activation+0.000
iny
Tokeniny
Feature activation+0.000
park
Token park
Feature activation+0.000
.
Token.
Feature activation+0.000
(
Token (
Feature activation+0.000
Ron
Token Ron
Feature activation+0.000
Bull
Token Bull
Feature activation+3.400
/
Token /
Feature activation+4.006
Toronto
Token Toronto
Feature activation+0.000
Star
Token Star
Feature activation+2.671
Archives
Token Archives
Feature activation+2.979
)
Token )
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
(
Token (
Feature activation+0.000
CAR
Token CAR
Feature activation+0.000
LOS
TokenLOS
Feature activation+0.000
OS
Token OS
Feature activation+0.000
OR
TokenOR
Feature activation+2.678
IO
TokenIO
Feature activation+3.664
/
Token /
Feature activation+1.522
Toronto
Token Toronto
Feature activation+0.069
Star
Token Star
Feature activation+1.734
)
Token )
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
the
Token the
Feature activation+0.000
park
Token park
Feature activation+0.000
.
Token.
Feature activation+0.000
(
Token (
Feature activation+0.000
Ron
Token Ron
Feature activation+0.000
Bull
Token Bull
Feature activation+3.400
/
Token /
Feature activation+4.006
Toronto
Token Toronto
Feature activation+0.000
Star
Token Star
Feature activation+2.671
Archives
Token Archives
Feature activation+2.979
)
Token )
Feature activation+0.000
[
Token[
Feature activation+0.000
Sy
Token Sy
Feature activation+0.000
ll
Tokenll
Feature activation+0.000
abus
Tokenabus
Feature activation+1.932
and
Token and
Feature activation+0.000
Review
Token Review
Feature activation+3.316
|
Token |
Feature activation+0.001
|
Token |
Feature activation+0.000
|
Token |
Feature activation+0.000
--------------------------------
Token --------------------------------
Feature activation+0.000
---
Token---
Feature activation+0.000
young
Token young
Feature activation+0.000
,
Token,
Feature activation+0.000
poor
Token poor
Feature activation+0.103
Indian
Token Indian
Feature activation+0.000
women
Token women
Feature activation+1.309
.
Token.
Feature activation+3.002
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
While
TokenWhile
Feature activation+0.000
the
Token the
Feature activation+0.000
government
Token government
Feature activation+0.000
Ron
Token Ron
Feature activation+0.000
Bull
Token Bull
Feature activation+3.400
/
Token /
Feature activation+4.006
Toronto
Token Toronto
Feature activation+0.000
Star
Token Star
Feature activation+2.671
Archives
Token Archives
Feature activation+2.979
)
Token )
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
ille
Tokenille
Feature activation+0.110
as
Tokenas
Feature activation+1.172
Z
Token Z
Feature activation+0.025
av
Tokenav
Feature activation+0.000
all
Tokenall
Feature activation+1.710
is
Tokenis
Feature activation+2.964
/
Token /
Feature activation+0.975
AFP
Token AFP
Feature activation+0.000
/
Token/
Feature activation+0.000
Getty
TokenGetty
Feature activation+0.676
Images
Token Images
Feature activation+1.621
_
Token_
Feature activation+0.000
ts
Tokents
Feature activation+1.336
vector
Tokenvector
Feature activation+2.134
(
Token (
Feature activation+0.000
coales
Token coales
Feature activation+0.000
ce
Tokence
Feature activation+2.690
(
Token (
Feature activation+0.000
name
Token name
Feature activation+0.151
,
Token ,
Feature activation+0.000
''
Token ''
Feature activation+0.179
)
Token )
Feature activation+0.786
.
Token.
Feature activation+0.000
(
Token (
Feature activation+0.000
CAR
Token CAR
Feature activation+0.000
LOS
TokenLOS
Feature activation+0.000
OS
Token OS
Feature activation+0.000
OR
TokenOR
Feature activation+2.678
IO
TokenIO
Feature activation+3.664
/
Token /
Feature activation+1.522
Toronto
Token Toronto
Feature activation+0.069
Star
Token Star
Feature activation+1.734
)
Token )
Feature activation+0.000
(
Token (
Feature activation+0.000
Ron
Token Ron
Feature activation+0.000
Bull
Token Bull
Feature activation+3.400
/
Token /
Feature activation+4.006
Toronto
Token Toronto
Feature activation+0.000
Star
Token Star
Feature activation+2.671
Archives
Token Archives
Feature activation+2.979
)
Token )
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
seconds
Token seconds
Feature activation+0.000
(
Token (
Feature activation+0.000
echo
Tokenecho
Feature activation+0.000
"
Token "
Feature activation+0.000
$
Token $
Feature activation+0.000
@
Token@
Feature activation+2.448
";
Token";
Feature activation+0.000
sleep
Token sleep
Feature activation+0.000
${
Token ${
Feature activation+0.000
duration
Tokenduration
Feature activation+0.000
})
Token})
Feature activation+0.000
Disaster
Token Disaster
Feature activation+0.000
Rescue
Token Rescue
Feature activation+0.000
Force
Token Force
Feature activation+0.000
(
Token (
Feature activation+0.000
N
Token N
Feature activation+0.000
DR
TokenDR
Feature activation+2.313
F
TokenF
Feature activation+4.843
),
Token ),
Feature activation+0.000
police
Token police
Feature activation+0.000
and
Token and
Feature activation+0.000
fire
Token fire
Feature activation+0.000
:
Token:
Feature activation+0.726
F
Token F
Feature activation+0.000
ara
Tokenara
Feature activation+0.000
Shim
Token Shim
Feature activation+0.000
bo
Tokenbo
Feature activation+1.167
/
Token /
Feature activation+2.211
)
Token )
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
former
Token former
Feature activation+0.000
nn
Tokennn
Feature activation+0.000
ore
Tokenore
Feature activation+0.000
map
Tokenmap
Feature activation+0.000
<
Token <
Feature activation+0.000
F
Token F
Feature activation+0.000
5
Token5
Feature activation+2.209
>
Token >
Feature activation+0.000
:
Token :
Feature activation+0.000
Gund
Token Gund
Feature activation+0.000
o
Tokeno
Feature activation+0.119
T
TokenT
Feature activation+0.000
weight
Tokenweight
Feature activation+0.000
(
Token (
Feature activation+0.000
to
Token to
Feature activation+0.000
_
Token_
Feature activation+0.000
ts
Tokents
Feature activation+1.336
vector
Tokenvector
Feature activation+2.134
(
Token (
Feature activation+0.000
coales
Token coales
Feature activation+0.000
ce
Tokence
Feature activation+2.690
(
Token (
Feature activation+0.000
name
Token name
Feature activation+0.151
,
Token,
Feature activation+0.000
'
Token '
Feature activation+0.000
we
Token we
Feature activation+0.000
can
Token can
Feature activation+0.000
do
Token do
Feature activation+0.000
anything
Token anything
Feature activation+2.124
to
Token to
Feature activation+0.000
promote
Token promote
Feature activation+0.312
the
Token the
Feature activation+0.000
process
Token process
Feature activation+1.757
",
Token ",
Feature activation+0.000
=
Token=
Feature activation+0.000
----------------
Token----------------
Feature activation+0.000
-
Token-
Feature activation+0.000
=[
Token=[
Feature activation+0.000
Course
Token Course
Feature activation+0.000
materials
Token materials
Feature activation+2.077
available
Token available
Feature activation+0.957
on
Token on
Feature activation+0.000
GitHub
Token GitHub
Feature activation+1.021
]
Token ]
Feature activation+0.000
=
Token=
Feature activation+0.000
|
Token |
Feature activation+0.000
--
Token --
Feature activation+0.000
[
Token[
Feature activation+0.000
Sy
Token Sy
Feature activation+0.000
ll
Tokenll
Feature activation+0.000
abus
Tokenabus
Feature activation+1.932
and
Token and
Feature activation+0.000
Review
Token Review
Feature activation+3.316
|
Token |
Feature activation+0.001
|
Token |
Feature activation+0.000
|
Token |
Feature activation+0.000

Top DFA by src position
MAX = 5.408

National
Token National
Feature activation+0.017
Top resid features:
Disaster
Token Disaster
Feature activation-0.005
Top resid features:
Rescue
Token Rescue
Feature activation-0.002
Top resid features:
Force
Token Force
Feature activation+0.001
Top resid features:
(
Token (
Feature activation+0.031
Top resid features:
N
Token N
Feature activation+5.408
Top resid features:
DR
TokenDR
Feature activation+0.058
Top resid features:
F
TokenF
Feature activation+0.036
Top resid features:
),
Token ),
Feature activation+0.000
Top resid features:
police
Token police
Feature activation+0.000
Top resid features:
and
Token and
Feature activation+0.000
Top resid features:
and
Token and
Feature activation-0.001
Top resid features:
superhero
Token superhero
Feature activation+0.002
Top resid features:
underwear
Token underwear
Feature activation-0.000
Top resid features:
.
Token.
Feature activation+0.010
Top resid features:
(
Token (
Feature activation-0.030
Top resid features:
Supp
Token Supp
Feature activation+5.140
Top resid features:
lied
Tokenlied
Feature activation+0.094
Top resid features:
)
Token )
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
T
TokenT
Feature activation+0.000
Top resid features:
near
Token near
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
park
Token park
Feature activation+0.001
Top resid features:
.
Token.
Feature activation-0.004
Top resid features:
(
Token (
Feature activation-0.076
Top resid features:
Ron
Token Ron
Feature activation+4.794
Top resid features:
Bull
Token Bull
Feature activation+0.012
Top resid features:
/
Token /
Feature activation-0.008
Top resid features:
Toronto
Token Toronto
Feature activation+0.000
Top resid features:
Star
Token Star
Feature activation+0.000
Top resid features:
Archives
Token Archives
Feature activation+0.000
Top resid features:
2
Token 2
Feature activation+0.020
Top resid features:
,
Token,
Feature activation-0.004
Top resid features:
2010
Token 2010
Feature activation+0.001
Top resid features:
.
Token.
Feature activation+0.026
Top resid features:
(
Token (
Feature activation+0.065
Top resid features:
CAR
Token CAR
Feature activation+4.154
Top resid features:
LOS
TokenLOS
Feature activation+0.094
Top resid features:
OS
Token OS
Feature activation+0.016
Top resid features:
OR
TokenOR
Feature activation+0.014
Top resid features:
IO
TokenIO
Feature activation+0.013
Top resid features:
/
Token /
Feature activation+0.000
Top resid features:
near
Token near
Feature activation-0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
park
Token park
Feature activation+0.001
Top resid features:
.
Token.
Feature activation+0.002
Top resid features:
(
Token (
Feature activation+0.043
Top resid features:
Ron
Token Ron
Feature activation+4.043
Top resid features:
Bull
Token Bull
Feature activation+0.005
Top resid features:
/
Token /
Feature activation+0.000
Top resid features:
Toronto
Token Toronto
Feature activation+0.000
Top resid features:
Star
Token Star
Feature activation+0.000
Top resid features:
Archives
Token Archives
Feature activation+0.000
Top resid features:
/
Token/
Feature activation+0.006
Top resid features:
27
Token27
Feature activation+0.009
Top resid features:
|
Token |
Feature activation+0.014
Top resid features:
--
Token --
Feature activation+0.059
Top resid features:
[
Token[
Feature activation+0.031
Top resid features:
Sy
Token Sy
Feature activation+3.598
Top resid features:
ll
Tokenll
Feature activation+0.003
Top resid features:
abus
Tokenabus
Feature activation+0.092
Top resid features:
and
Token and
Feature activation+0.013
Top resid features:
Review
Token Review
Feature activation+0.027
Top resid features:
|
Token |
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.072
Top resid features:
"
Token"
Feature activation+0.003
Top resid features:
exploitation
Token exploitation
Feature activation+3.685
Top resid features:
of
Token of
Feature activation-0.013
Top resid features:
young
Token young
Feature activation+0.004
Top resid features:
,
Token,
Feature activation+0.004
Top resid features:
poor
Token poor
Feature activation-0.010
Top resid features:
Indian
Token Indian
Feature activation-0.003
Top resid features:
near
Token near
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
park
Token park
Feature activation-0.000
Top resid features:
.
Token.
Feature activation+0.002
Top resid features:
(
Token (
Feature activation-0.098
Top resid features:
Ron
Token Ron
Feature activation+3.572
Top resid features:
Bull
Token Bull
Feature activation+0.010
Top resid features:
/
Token /
Feature activation+0.041
Top resid features:
Toronto
Token Toronto
Feature activation+0.050
Top resid features:
Star
Token Star
Feature activation+0.110
Top resid features:
Archives
Token Archives
Feature activation+0.049
Top resid features:
of
Token of
Feature activation-0.001
Top resid features:
the
Token the
Feature activation-0.000
Top resid features:
fighters
Token fighters
Feature activation+0.000
Top resid features:
.
Token.
Feature activation-0.003
Top resid features:
(
Token (
Feature activation-0.039
Top resid features:
Ach
Token Ach
Feature activation+3.440
Top resid features:
ille
Tokenille
Feature activation+0.028
Top resid features:
as
Tokenas
Feature activation+0.182
Top resid features:
Z
Token Z
Feature activation+0.051
Top resid features:
av
Tokenav
Feature activation+0.008
Top resid features:
all
Tokenall
Feature activation+0.014
Top resid features:
to
Token to
Feature activation+0.599
Top resid features:
_
Token_
Feature activation+0.053
Top resid features:
ts
Tokents
Feature activation-0.004
Top resid features:
vector
Tokenvector
Feature activation+0.037
Top resid features:
(
Token (
Feature activation+0.483
Top resid features:
coales
Token coales
Feature activation+1.969
Top resid features:
ce
Tokence
Feature activation+0.063
Top resid features:
(
Token (
Feature activation+0.000
Top resid features:
name
Token name
Feature activation+0.000
Top resid features:
,
Token ,
Feature activation+0.000
Top resid features:
''
Token ''
Feature activation+0.000
Top resid features:
2
Token 2
Feature activation+0.014
Top resid features:
,
Token,
Feature activation-0.001
Top resid features:
2010
Token 2010
Feature activation+0.003
Top resid features:
.
Token.
Feature activation+0.007
Top resid features:
(
Token (
Feature activation+0.009
Top resid features:
CAR
Token CAR
Feature activation+3.017
Top resid features:
LOS
TokenLOS
Feature activation+0.339
Top resid features:
OS
Token OS
Feature activation-0.001
Top resid features:
OR
TokenOR
Feature activation-0.004
Top resid features:
IO
TokenIO
Feature activation+0.000
Top resid features:
/
Token /
Feature activation+0.000
Top resid features:
near
Token near
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
park
Token park
Feature activation-0.001
Top resid features:
.
Token.
Feature activation-0.005
Top resid features:
(
Token (
Feature activation-0.070
Top resid features:
Ron
Token Ron
Feature activation+3.357
Top resid features:
Bull
Token Bull
Feature activation+0.013
Top resid features:
/
Token /
Feature activation+0.086
Top resid features:
Toronto
Token Toronto
Feature activation+0.019
Top resid features:
Star
Token Star
Feature activation+0.045
Top resid features:
Archives
Token Archives
Feature activation+0.000
Top resid features:
3
Token 3
Feature activation+0.009
Top resid features:
seconds
Token seconds
Feature activation+0.010
Top resid features:
(
Token (
Feature activation+0.008
Top resid features:
echo
Tokenecho
Feature activation-0.010
Top resid features:
"
Token "
Feature activation-0.001
Top resid features:
$
Token $
Feature activation+3.033
Top resid features:
@
Token@
Feature activation+0.048
Top resid features:
";
Token";
Feature activation+0.000
Top resid features:
sleep
Token sleep
Feature activation+0.000
Top resid features:
${
Token ${
Feature activation+0.000
Top resid features:
duration
Tokenduration
Feature activation+0.000
Top resid features:
National
Token National
Feature activation+0.002
Top resid features:
Disaster
Token Disaster
Feature activation-0.005
Top resid features:
Rescue
Token Rescue
Feature activation-0.028
Top resid features:
Force
Token Force
Feature activation-0.031
Top resid features:
(
Token (
Feature activation-0.001
Top resid features:
N
Token N
Feature activation+3.081
Top resid features:
DR
TokenDR
Feature activation+0.048
Top resid features:
F
TokenF
Feature activation+0.000
Top resid features:
),
Token ),
Feature activation+0.000
Top resid features:
police
Token police
Feature activation+0.000
Top resid features:
and
Token and
Feature activation+0.000
Top resid features:
the
Token the
Feature activation-0.000
Top resid features:
Latino
Token Latino
Feature activation+0.001
Top resid features:
community
Token community
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.003
Top resid features:
(
Token (
Feature activation-0.060
Top resid features:
Artist
Token Artist
Feature activation+2.969
Top resid features:
rendition
Token rendition
Feature activation+0.019
Top resid features:
:
Token:
Feature activation+0.024
Top resid features:
F
Token F
Feature activation+0.004
Top resid features:
ara
Tokenara
Feature activation+0.006
Top resid features:
Shim
Token Shim
Feature activation-0.013
Top resid features:
Ċ
TokenĊ
Feature activation+0.002
Top resid features:
nn
Tokennn
Feature activation-0.020
Top resid features:
ore
Tokenore
Feature activation-0.046
Top resid features:
map
Tokenmap
Feature activation+0.019
Top resid features:
<
Token <
Feature activation+0.051
Top resid features:
F
Token F
Feature activation+2.771
Top resid features:
5
Token5
Feature activation+0.066
Top resid features:
>
Token >
Feature activation+0.000
Top resid features:
:
Token :
Feature activation+0.000
Top resid features:
Gund
Token Gund
Feature activation+0.000
Top resid features:
o
Tokeno
Feature activation+0.000
Top resid features:
sv
Tokensv
Feature activation+0.008
Top resid features:
=
Token =
Feature activation-0.019
Top resid features:
set
Token set
Feature activation+0.038
Top resid features:
weight
Tokenweight
Feature activation-0.001
Top resid features:
(
Token (
Feature activation+0.153
Top resid features:
to
Token to
Feature activation+2.420
Top resid features:
_
Token_
Feature activation+0.163
Top resid features:
ts
Tokents
Feature activation+0.006
Top resid features:
vector
Tokenvector
Feature activation+0.040
Top resid features:
(
Token (
Feature activation+0.000
Top resid features:
coales
Token coales
Feature activation+0.000
Top resid features:
said
Token said
Feature activation-0.001
Top resid features:
to
Token to
Feature activation+0.008
Top resid features:
them
Token them
Feature activation+0.008
Top resid features:
,
Token,
Feature activation-0.000
Top resid features:
'
Token '
Feature activation+0.010
Top resid features:
we
Token we
Feature activation+2.334
Top resid features:
can
Token can
Feature activation+0.055
Top resid features:
do
Token do
Feature activation-0.052
Top resid features:
anything
Token anything
Feature activation+0.033
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
promote
Token promote
Feature activation+0.000
Top resid features:
|
Token |
Feature activation+0.011
Top resid features:
=
Token=
Feature activation+0.005
Top resid features:
----------------
Token----------------
Feature activation+0.004
Top resid features:
-
Token-
Feature activation+0.001
Top resid features:
=[
Token=[
Feature activation+0.072
Top resid features:
Course
Token Course
Feature activation+2.569
Top resid features:
materials
Token materials
Feature activation+0.058
Top resid features:
available
Token available
Feature activation+0.000
Top resid features:
on
Token on
Feature activation+0.000
Top resid features:
GitHub
Token GitHub
Feature activation+0.000
Top resid features:
]
Token ]
Feature activation+0.000
Top resid features:
/
Token/
Feature activation+0.020
Top resid features:
27
Token27
Feature activation+0.014
Top resid features:
|
Token |
Feature activation+0.035
Top resid features:
--
Token --
Feature activation+0.076
Top resid features:
[
Token[
Feature activation+0.121
Top resid features:
Sy
Token Sy
Feature activation+2.076
Top resid features:
ll
Tokenll
Feature activation+0.019
Top resid features:
abus
Tokenabus
Feature activation+0.066
Top resid features:
and
Token and
Feature activation+0.000
Top resid features:
Review
Token Review
Feature activation+0.000
Top resid features:
|
Token |
Feature activation+0.000
Top resid features:

Decoder Weights Distribution

Head 0: 0.07

Head 1: 0.02

Head 2: 0.02

Head 3: 0.06

Head 4: 0.54

Head 5: 0.05

Head 6: 0.04

Head 7: 0.03

Head 8: 0.06

Head 9: 0.05

Head 10: 0.02

Head 11: 0.04

Positive logits

).6.38

),6.28

)6.09

].6.02

);5.88

,"5.72

):5.69

)]5.64

],5.59

]5.47

])5.43

."5.38

))5.24

",5.24

.)5.08

));4.97

];4.73

".4.65

());4.56

)))4.42

Negative logits

SPONSORED-3.18

[/-2.89

etheless-2.73

)=(-2.66

��-2.50

IPM-2.45

theless-2.28

��-2.20

-2.11

oples-2.07

ennes-2.07

Wiz-2.06

Argon-2.06

':-2.02

CTV-2.01

Sanford-1.98

Nib-1.96

Jaw-1.91

scl-1.91

awaru-1.91

INTERVAL 4.358 - 4.843
CONTAINS 0.000%

Rescue
Token Rescue
Feature activation+0.000
Force
Token Force
Feature activation+0.000
(
Token (
Feature activation+0.000
N
Token N
Feature activation+0.000
DR
TokenDR
Feature activation+2.313
F
TokenF
Feature activation+4.843
),
Token ),
Feature activation+0.000
police
Token police
Feature activation+0.000
and
Token and
Feature activation+0.000
fire
Token fire
Feature activation+0.000
service
Token service
Feature activation+0.000
superhero
Token superhero
Feature activation+0.000
underwear
Token underwear
Feature activation+0.000
.
Token.
Feature activation+0.000
(
Token (
Feature activation+0.000
Supp
Token Supp
Feature activation+0.000
lied
Tokenlied
Feature activation+4.552
)
Token )
Feature activation+0.526
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
T
TokenT
Feature activation+0.000
iny
Tokeniny
Feature activation+0.000

INTERVAL 3.874 - 4.358
CONTAINS 0.000%

park
Token park
Feature activation+0.000
.
Token.
Feature activation+0.000
(
Token (
Feature activation+0.000
Ron
Token Ron
Feature activation+0.000
Bull
Token Bull
Feature activation+3.400
/
Token /
Feature activation+4.006
Toronto
Token Toronto
Feature activation+0.000
Star
Token Star
Feature activation+2.671
Archives
Token Archives
Feature activation+2.979
)
Token )
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000

INTERVAL 3.390 - 3.874
CONTAINS 0.000%

(
Token (
Feature activation+0.000
CAR
Token CAR
Feature activation+0.000
LOS
TokenLOS
Feature activation+0.000
OS
Token OS
Feature activation+0.000
OR
TokenOR
Feature activation+2.678
IO
TokenIO
Feature activation+3.664
/
Token /
Feature activation+1.522
Toronto
Token Toronto
Feature activation+0.069
Star
Token Star
Feature activation+1.734
)
Token )
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
the
Token the
Feature activation+0.000
park
Token park
Feature activation+0.000
.
Token.
Feature activation+0.000
(
Token (
Feature activation+0.000
Ron
Token Ron
Feature activation+0.000
Bull
Token Bull
Feature activation+3.400
/
Token /
Feature activation+4.006
Toronto
Token Toronto
Feature activation+0.000
Star
Token Star
Feature activation+2.671
Archives
Token Archives
Feature activation+2.979
)
Token )
Feature activation+0.000

INTERVAL 2.906 - 3.390
CONTAINS 0.000%

ille
Tokenille
Feature activation+0.110
as
Tokenas
Feature activation+1.172
Z
Token Z
Feature activation+0.025
av
Tokenav
Feature activation+0.000
all
Tokenall
Feature activation+1.710
is
Tokenis
Feature activation+2.964
/
Token /
Feature activation+0.975
AFP
Token AFP
Feature activation+0.000
/
Token/
Feature activation+0.000
Getty
TokenGetty
Feature activation+0.676
Images
Token Images
Feature activation+1.621
[
Token[
Feature activation+0.000
Sy
Token Sy
Feature activation+0.000
ll
Tokenll
Feature activation+0.000
abus
Tokenabus
Feature activation+1.932
and
Token and
Feature activation+0.000
Review
Token Review
Feature activation+3.316
|
Token |
Feature activation+0.001
|
Token |
Feature activation+0.000
|
Token |
Feature activation+0.000
--------------------------------
Token --------------------------------
Feature activation+0.000
---
Token---
Feature activation+0.000
young
Token young
Feature activation+0.000
,
Token,
Feature activation+0.000
poor
Token poor
Feature activation+0.103
Indian
Token Indian
Feature activation+0.000
women
Token women
Feature activation+1.309
.
Token.
Feature activation+3.002
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
While
TokenWhile
Feature activation+0.000
the
Token the
Feature activation+0.000
government
Token government
Feature activation+0.000
Ron
Token Ron
Feature activation+0.000
Bull
Token Bull
Feature activation+3.400
/
Token /
Feature activation+4.006
Toronto
Token Toronto
Feature activation+0.000
Star
Token Star
Feature activation+2.671
Archives
Token Archives
Feature activation+2.979
)
Token )
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000

INTERVAL 2.421 - 2.906
CONTAINS 0.001%

.
Token.
Feature activation+0.000
(
Token (
Feature activation+0.000
CAR
Token CAR
Feature activation+0.000
LOS
TokenLOS
Feature activation+0.000
OS
Token OS
Feature activation+0.000
OR
TokenOR
Feature activation+2.678
IO
TokenIO
Feature activation+3.664
/
Token /
Feature activation+1.522
Toronto
Token Toronto
Feature activation+0.069
Star
Token Star
Feature activation+1.734
)
Token )
Feature activation+0.000
_
Token_
Feature activation+0.000
ts
Tokents
Feature activation+1.336
vector
Tokenvector
Feature activation+2.134
(
Token (
Feature activation+0.000
coales
Token coales
Feature activation+0.000
ce
Tokence
Feature activation+2.690
(
Token (
Feature activation+0.000
name
Token name
Feature activation+0.151
,
Token ,
Feature activation+0.000
''
Token ''
Feature activation+0.179
)
Token )
Feature activation+0.786
seconds
Token seconds
Feature activation+0.000
(
Token (
Feature activation+0.000
echo
Tokenecho
Feature activation+0.000
"
Token "
Feature activation+0.000
$
Token $
Feature activation+0.000
@
Token@
Feature activation+2.448
";
Token";
Feature activation+0.000
sleep
Token sleep
Feature activation+0.000
${
Token ${
Feature activation+0.000
duration
Tokenduration
Feature activation+0.000
})
Token})
Feature activation+0.000
(
Token (
Feature activation+0.000
Ron
Token Ron
Feature activation+0.000
Bull
Token Bull
Feature activation+3.400
/
Token /
Feature activation+4.006
Toronto
Token Toronto
Feature activation+0.000
Star
Token Star
Feature activation+2.671
Archives
Token Archives
Feature activation+2.979
)
Token )
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000

INTERVAL 1.937 - 2.421
CONTAINS 0.001%

Disaster
Token Disaster
Feature activation+0.000
Rescue
Token Rescue
Feature activation+0.000
Force
Token Force
Feature activation+0.000
(
Token (
Feature activation+0.000
N
Token N
Feature activation+0.000
DR
TokenDR
Feature activation+2.313
F
TokenF
Feature activation+4.843
),
Token ),
Feature activation+0.000
police
Token police
Feature activation+0.000
and
Token and
Feature activation+0.000
fire
Token fire
Feature activation+0.000
nn
Tokennn
Feature activation+0.000
ore
Tokenore
Feature activation+0.000
map
Tokenmap
Feature activation+0.000
<
Token <
Feature activation+0.000
F
Token F
Feature activation+0.000
5
Token5
Feature activation+2.209
>
Token >
Feature activation+0.000
:
Token :
Feature activation+0.000
Gund
Token Gund
Feature activation+0.000
o
Tokeno
Feature activation+0.119
T
TokenT
Feature activation+0.000
,
Token,
Feature activation+0.000
'
Token '
Feature activation+0.000
we
Token we
Feature activation+0.000
can
Token can
Feature activation+0.000
do
Token do
Feature activation+0.000
anything
Token anything
Feature activation+2.124
to
Token to
Feature activation+0.000
promote
Token promote
Feature activation+0.312
the
Token the
Feature activation+0.000
process
Token process
Feature activation+1.757
",
Token ",
Feature activation+0.000
=
Token=
Feature activation+0.000
----------------
Token----------------
Feature activation+0.000
-
Token-
Feature activation+0.000
=[
Token=[
Feature activation+0.000
Course
Token Course
Feature activation+0.000
materials
Token materials
Feature activation+2.077
available
Token available
Feature activation+0.957
on
Token on
Feature activation+0.000
GitHub
Token GitHub
Feature activation+1.021
]
Token ]
Feature activation+0.000
=
Token=
Feature activation+0.000
weight
Tokenweight
Feature activation+0.000
(
Token (
Feature activation+0.000
to
Token to
Feature activation+0.000
_
Token_
Feature activation+0.000
ts
Tokents
Feature activation+1.336
vector
Tokenvector
Feature activation+2.134
(
Token (
Feature activation+0.000
coales
Token coales
Feature activation+0.000
ce
Tokence
Feature activation+2.690
(
Token (
Feature activation+0.000
name
Token name
Feature activation+0.151

INTERVAL 1.453 - 1.937
CONTAINS 0.002%

|
Token |
Feature activation+0.000
--
Token --
Feature activation+0.000
[
Token[
Feature activation+0.000
Sy
Token Sy
Feature activation+0.000
ll
Tokenll
Feature activation+0.000
abus
Tokenabus
Feature activation+1.932
and
Token and
Feature activation+0.000
Review
Token Review
Feature activation+3.316
|
Token |
Feature activation+0.001
|
Token |
Feature activation+0.000
|
Token |
Feature activation+0.000
ur
Tokenur
Feature activation+0.000
.
Token.
Feature activation+0.000
«
Token «
Feature activation+0.000
N
Token N
Feature activation+0.000
OL
TokenOL
Feature activation+0.000
AF
TokenAF
Feature activation+1.790
ran
Tokenran
Feature activation+0.000
ça
Tokença
Feature activation+0.493
ise
Tokenise
Feature activation+1.342
ve
Token ve
Feature activation+0.000
ut
Tokenut
Feature activation+0.000
CAR
Token CAR
Feature activation+0.000
LOS
TokenLOS
Feature activation+0.000
OS
Token OS
Feature activation+0.000
OR
TokenOR
Feature activation+2.678
IO
TokenIO
Feature activation+3.664
/
Token /
Feature activation+1.522
Toronto
Token Toronto
Feature activation+0.069
Star
Token Star
Feature activation+1.734
)
Token )
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
an
Token an
Feature activation+0.000
assist
Token assist
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
final
Token final
Feature activation+0.000
.
Token.
Feature activation+1.604
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
So
TokenSo
Feature activation+0.000
,
Token,
Feature activation+0.000
many
Token many
Feature activation+0.000
WHERE
Token WHERE
Feature activation+0.000
(
Token (
Feature activation+0.000
t
Token t
Feature activation+0.000
sv
Tokensv
Feature activation+1.673
@@
Token @@
Feature activation+0.000
q
Token q
Feature activation+1.522
)
Token )
Feature activation+0.054
)
Token )
Feature activation+0.000
AS
Token AS
Feature activation+0.000
t
Token t
Feature activation+0.000
1
Token1
Feature activation+0.136

INTERVAL 0.969 - 1.453
CONTAINS 0.004%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
"
Token"
Feature activation+0.000
ARM
Token ARM
Feature activation+0.000
Cortex
Token Cortex
Feature activation+0.138
A
Token A
Feature activation+0.000
53
Token53
Feature activation+1.166
CPU
Token CPU
Feature activation+0.675
cores
Token cores
Feature activation+0.000
with
Token with
Feature activation+0.000
four
Token four
Feature activation+0.000
"
Token "
Feature activation+0.000
coales
Token coales
Feature activation+0.000
ce
Tokence
Feature activation+0.911
(
Token (
Feature activation+0.000
passage
Token passage
Feature activation+0.609
,
Token ,
Feature activation+0.443
''
Token ''
Feature activation+1.001
)
Token )
Feature activation+1.042
),
Token),
Feature activation+0.850
'
Token '
Feature activation+0.000
D
TokenD
Feature activation+0.000
'
Token'
Feature activation+0.790
N
Token N
Feature activation+0.000
OL
TokenOL
Feature activation+0.000
AF
TokenAF
Feature activation+1.790
ran
Tokenran
Feature activation+0.000
ça
Tokença
Feature activation+0.493
ise
Tokenise
Feature activation+1.342
ve
Token ve
Feature activation+0.000
ut
Tokenut
Feature activation+0.000
tend
Token tend
Feature activation+0.000
re
Tokenre
Feature activation+0.719
un
Token un
Feature activation+0.000
ts
Tokents
Feature activation+0.000
query
Tokenquery
Feature activation+0.000
(
Token (
Feature activation+0.000
'
Token '
Feature activation+0.000
river
Tokenriver
Feature activation+0.227
'
Token'
Feature activation+1.404
)
Token )
Feature activation+0.159
AS
Token AS
Feature activation+0.000
q
Token q
Feature activation+0.148
WHERE
Token WHERE
Feature activation+0.000
(
Token (
Feature activation+0.000
;
Token;
Feature activation+0.000
draft
Token draft
Feature activation+0.000
short
Token short
Feature activation+0.000
-
Token-
Feature activation+0.000
and
Token and
Feature activation+0.000
long
Token long
Feature activation+1.267
-
Token-
Feature activation+0.000
term
Tokenterm
Feature activation+0.000
career
Token career
Feature activation+0.000
plans
Token plans
Feature activation+0.000
;
Token;
Feature activation+0.000

INTERVAL 0.484 - 0.969
CONTAINS 0.010%

''
Token ''
Feature activation+1.001
)
Token )
Feature activation+1.042
),
Token),
Feature activation+0.850
'
Token '
Feature activation+0.000
D
TokenD
Feature activation+0.000
'
Token'
Feature activation+0.790
);
Token );
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Creating
TokenCreating
Feature activation+0.000
functions
Token functions
Feature activation+0.000
Grave
Token Grave
Feature activation+0.000
Rob
Token Rob
Feature activation+0.000
ber
Tokenber
Feature activation+0.034
Into
Token Into
Feature activation+0.000
the
Token the
Feature activation+0.000
Depths
Token Depths
Feature activation+0.518
Chapter
Token Chapter
Feature activation+0.000
3
Token 3
Feature activation+0.000
Dark
Token Dark
Feature activation+0.000
Des
Token Des
Feature activation+0.000
cent
Tokencent
Feature activation+0.000
Conservative
Token Conservative
Feature activation+0.000
Authent
Token Authent
Feature activation+0.000
ic
Tokenic
Feature activation+0.000
Respons
Token Respons
Feature activation+0.000
ive
Tokenive
Feature activation+0.225
Leadership
Token Leadership
Feature activation+0.586
for
Token for
Feature activation+0.000
You
Token You
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+0.000
and
Token and
Feature activation+0.000
for
Token for
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
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.591
Ge
TokenGe
Feature activation+0.000
op
Tokenop
Feature activation+0.000
olitics
Tokenolitics
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
and
Token and
Feature activation+0.000
there
Token there
Feature activation+0.000
are
Token are
Feature activation+0.000
no
Token no
Feature activation+0.000
shortcuts
Token shortcuts
Feature activation+0.000
.
Token.
Feature activation+0.570
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
W
TokenW
Feature activation+0.000
IRED
TokenIRED
Feature activation+0.000
:
Token:
Feature activation+0.000

INTERVAL 0.000 - 0.484
CONTAINS 99.981%

unresolved
Token unresolved
Feature activation+0.000
.
Token.
Feature activation+0.000
In
TokenIn
Feature activation+0.000
the
Token the
Feature activation+0.000
town
Token town
Feature activation+0.000
of
Token of
Feature activation+0.000
Nit
Token Nit
Feature activation+0.000
ro
Tokenro
Feature activation+0.000
,
Token,
Feature activation+0.000
West
Token West
Feature activation+0.000
Virginia
Token Virginia
Feature activation+0.000
that
Token that
Feature activation+0.000
area
Token area
Feature activation+0.000
.
Token.
Feature activation+0.000
Whatever
Token Whatever
Feature activation+0.000
he
Token he
Feature activation+0.000
can
Token can
Feature activation+0.000
give
Token give
Feature activation+0.000
will
Token will
Feature activation+0.000
undoubtedly
Token undoubtedly
Feature activation+0.000
be
Token be
Feature activation+0.000
more
Token more
Feature activation+0.000
few
Token few
Feature activation+0.000
ropes
Token ropes
Feature activation+0.000
and
Token and
Feature activation+0.000
hit
Token hit
Feature activation+0.000
the
Token the
Feature activation+0.000
road
Token road
Feature activation+0.000
,
Token,
Feature activation+0.000
in
Token in
Feature activation+0.000
hot
Token hot
Feature activation+0.000
pursuit
Token pursuit
Feature activation+0.000
of
Token of
Feature activation+0.000
Sunday
Token Sunday
Feature activation+0.000
,
Token,
Feature activation+0.000
Tom
Token Tom
Feature activation+0.000
a
Tokena
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
85
Token 85
Feature activation+0.000
th
Tokenth
Feature activation+0.000
birthday
Token birthday
Feature activation+0.000
,
Token,
Feature activation+0.000
be
Token be
Feature activation+0.000
done
Token done
Feature activation+0.000
either
Token either
Feature activation+0.000
by
Token by
Feature activation+0.000
running
Token running
Feature activation+0.000
with
Token with
Feature activation+0.000
the
Token the
Feature activation+0.000
ball
Token ball
Feature activation+0.000
or
Token or
Feature activation+0.000
by
Token by
Feature activation+0.000
a
Token a
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 H7.4: (feature 19694)

TOP ACTIVATIONS
MAX = 8.202

cover
Token cover
Feature activation+0.000
versus
Token versus
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
-
Token-
Feature activation+0.000
Advertisement
Token Advertisement
Feature activation+0.306
-
Token -
Feature activation+8.202
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.769
Let
TokenLet
Feature activation+0.000
's
Token's
Feature activation+0.000
watch
Token watch
Feature activation+0.000
battery
Token battery
Feature activation+0.000
replacement
Token replacement
Feature activation+0.365
by
Token by
Feature activation+0.201
$
Token $
Feature activation+0.000
50
Token50
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+6.996
from
Token from
Feature activation+0.000
$
Token $
Feature activation+0.000
79
Token79
Feature activation+0.000
to
Token to
Feature activation+0.000
$
Token $
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
Advertisement
Token Advertisement
Feature activation+0.296
-
Token -
Feature activation+6.938
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Much
TokenMuch
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Af
Token Af
Feature activation+0.000
raid
Tokenraid
Feature activation+0.000
to
Token to
Feature activation+0.000
go
Token go
Feature activation+0.000
public
Token public
Feature activation+1.161
-
Token -
Feature activation+6.831
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Beyond
TokenBeyond
Feature activation+0.000
Fiat
Token Fiat
Feature activation+0.000
's
Token's
Feature activation+0.000
hum
Token hum
Feature activation+0.000
pty
Tokenpty
Feature activation+0.000
-
Token-
Feature activation+0.000
d
Tokend
Feature activation+0.000
oo
Tokenoo
Feature activation+0.000
-
Token-
Feature activation+5.947
08
Token08
Feature activation+0.000
36
Token36
Feature activation+0.775
,
Token,
Feature activation+0.000
d
Token d
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
Tomorrow
Token Tomorrow
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+2.039
Ļ
TokenĻ
Feature activation+4.342
s
Tokens
Feature activation+0.000
Solutions
Token Solutions
Feature activation+0.000
--
Token --
Feature activation+5.476
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Former
TokenFormer
Feature activation+0.000
Mayor
Token Mayor
Feature activation+0.000
Lee
Token Lee
Feature activation+0.000
Christ
Token Christ
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+1.036
Ļ
TokenĻ
Feature activation+1.083
s
Tokens
Feature activation+0.000
cross
Token cross
Feature activation+0.000
.
Token.
Feature activation+4.659
The
Token The
Feature activation+0.118
pope
Token pope
Feature activation+0.000
had
Token had
Feature activation+0.000
indicated
Token indicated
Feature activation+0.000
in
Token in
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
--
Token--
Feature activation+0.000
Testing
Token Testing
Feature activation+0.000
Tomorrow
Token Tomorrow
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+2.039
Ļ
TokenĻ
Feature activation+4.342
s
Tokens
Feature activation+0.000
Solutions
Token Solutions
Feature activation+0.000
--
Token --
Feature activation+5.476
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
vir
Token vir
Feature activation+0.000
g
Tokeng
Feature activation+0.000
inal
Tokeninal
Feature activation+0.000
material
Token material
Feature activation+0.000
resource
Token resource
Feature activation+0.000
.
Token.
Feature activation+4.038
From
Token From
Feature activation+0.000
this
Token this
Feature activation+0.000
starting
Token starting
Feature activation+0.000
point
Token point
Feature activation+0.000
,
Token,
Feature activation+0.000
over
Token over
Feature activation+0.619
Orlando
Token Orlando
Feature activation+0.533
City
Token City
Feature activation+0.599
B
Token B
Feature activation+0.772
Ċ
TokenĊ
Feature activation+1.154
Ċ
TokenĊ
Feature activation+3.999
Eastern
TokenEastern
Feature activation+0.000
Conference
Token Conference
Feature activation+0.000
Sem
Token Sem
Feature activation+0.000
if
Tokenif
Feature activation+0.000
inal
Tokeninal
Feature activation+0.000
every
Token every
Feature activation+0.000
class
Token class
Feature activation+0.000
an
Token an
Feature activation+0.000
absolute
Token absolute
Feature activation+0.000
blast
Token blast
Feature activation+0.000
.
Token.
Feature activation+3.712
Don
Token Don
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
t
Tokent
Feature activation+0.000
dismiss
Token dismiss
Feature activation+0.000
of
Token of
Feature activation+0.000
Wall
Token Wall
Feature activation+0.000
Street
Token Street
Feature activation+0.228
's
Token's
Feature activation+0.000
attention
Token attention
Feature activation+0.767
.
Token.
Feature activation+3.631
C
Token C
Feature activation+0.000
ME
TokenME
Feature activation+0.000
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
world
Token world
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.126
Ċ
TokenĊ
Feature activation+2.343
Engine
TokenEngine
Feature activation+0.000
ering
Tokenering
Feature activation+0.000
News
Token News
Feature activation+0.000
-
Token-
Feature activation+3.623
Record
TokenRecord
Feature activation+0.000
reported
Token reported
Feature activation+0.000
in
Token in
Feature activation+0.000
1964
Token 1964
Feature activation+0.000
that
Token that
Feature activation+0.000
wood
Tokenwood
Feature activation+0.060
ruled
Token ruled
Feature activation+0.000
from
Token from
Feature activation+0.000
the
Token the
Feature activation+0.000
bench
Token bench
Feature activation+1.343
[
Token [
Feature activation+3.535
AP
TokenAP
Feature activation+0.000
report
Token report
Feature activation+0.154
]
Token]
Feature activation+1.487
after
Token after
Feature activation+0.000
a
Token a
Feature activation+0.000
from
Token from
Feature activation+0.000
large
Token large
Feature activation+0.000
commercial
Token commercial
Feature activation+0.000
jet
Token jet
Feature activation+1.094
aircraft
Token aircraft
Feature activation+0.916
.
Token.
Feature activation+3.436
Ċ
TokenĊ
Feature activation+0.126
Ċ
TokenĊ
Feature activation+2.343
Engine
TokenEngine
Feature activation+0.000
ering
Tokenering
Feature activation+0.000
News
Token News
Feature activation+0.000
$
Token $
Feature activation+0.000
79
Token79
Feature activation+0.000
to
Token to
Feature activation+0.000
$
Token $
Feature activation+0.000
29
Token29
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+3.434
for
Token for
Feature activation+0.000
anyone
Token anyone
Feature activation+0.000
with
Token with
Feature activation+0.000
an
Token an
Feature activation+0.000
iPhone
Token iPhone
Feature activation+0.000
-
Token-
Feature activation+0.000
up
Tokenup
Feature activation+0.000
of
Token of
Feature activation+0.000
a
Token a
Feature activation+0.000
resume
Token resume
Feature activation+0.705
.
Token.
Feature activation+3.347
Ċ
TokenĊ
Feature activation+0.169
Ċ
TokenĊ
Feature activation+1.375
We
TokenWe
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
skins
Token skins
Feature activation+0.000
to
Token to
Feature activation+0.000
share
Token share
Feature activation+0.000
with
Token with
Feature activation+0.000
others
Token others
Feature activation+0.000
.
Token.
Feature activation+3.313
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+2.200
In
TokenIn
Feature activation+0.000
1999
Token 1999
Feature activation+0.000
,
Token,
Feature activation+0.000
-
Token-
Feature activation+3.419
don
Tokendon
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+1.750
t
Tokent
Feature activation+0.962
-
Token-
Feature activation+3.268
we
Tokenwe
Feature activation+0.000
-
Token-
Feature activation+0.000
have
Tokenhave
Feature activation+0.000
-
Token-
Feature activation+0.000
a
Tokena
Feature activation+0.000
yellow
Tokenyellow
Feature activation+0.000
arrows
Token arrows
Feature activation+0.000
,
Token,
Feature activation+0.000
lower
Token lower
Feature activation+0.000
panels
Token panels
Feature activation+0.000
).
Token).
Feature activation+3.250
(
Token (
Feature activation+0.000
d
Tokend
Feature activation+0.000
)
Token)
Feature activation+0.000
Following
Token Following
Feature activation+0.000
injection
Token injection
Feature activation+0.000

Top DFA by src position
MAX = 9.170

a
Token a
Feature activation+0.000
Top resid features:
hard
Token hard
Feature activation+0.000
Top resid features:
cover
Token cover
Feature activation-0.000
Top resid features:
versus
Token versus
Feature activation-0.001
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.011
Top resid features:
-
Token-
Feature activation+9.170
Top resid features:
Advertisement
Token Advertisement
Feature activation-0.055
Top resid features:
-
Token -
Feature activation-0.029
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
Let
TokenLet
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.544
Top resid features:
-
Token-
Feature activation+7.784
Top resid features:
of
Tokenof
Feature activation+0.087
Top resid features:
-
Token-
Feature activation-0.004
Top resid features:
w
Tokenw
Feature activation+0.001
Top resid features:
arr
Tokenarr
Feature activation+0.001
Top resid features:
anty
Tokenanty
Feature activation+0.033
Top resid features:
HE
TokenHE
Feature activation-0.011
Top resid features:
AP
TokenAP
Feature activation-0.002
Top resid features:
.
Token.
Feature activation-0.001
Top resid features:
Ċ
TokenĊ
Feature activation+0.002
Top resid features:
Ċ
TokenĊ
Feature activation-0.008
Top resid features:
-
Token-
Feature activation+7.980
Top resid features:
Advertisement
Token Advertisement
Feature activation-0.070
Top resid features:
-
Token -
Feature activation+0.015
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
Much
TokenMuch
Feature activation+0.000
Top resid features:
to
Token to
Feature activation-0.000
Top resid features:
AFP
Token AFP
Feature activation-0.005
Top resid features:
.
Token.
Feature activation-0.004
Top resid features:
Ċ
TokenĊ
Feature activation-0.009
Top resid features:
Ċ
TokenĊ
Feature activation-0.054
Top resid features:
-
Token-
Feature activation+7.950
Top resid features:
Af
Token Af
Feature activation-0.010
Top resid features:
raid
Tokenraid
Feature activation-0.073
Top resid features:
to
Token to
Feature activation-0.013
Top resid features:
go
Token go
Feature activation-0.009
Top resid features:
public
Token public
Feature activation-0.017
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.395
Top resid features:
-
Token-
Feature activation+7.166
Top resid features:
justice
Tokenjustice
Feature activation+0.037
Top resid features:
,
Token,
Feature activation+0.013
Top resid features:
animals
Token animals
Feature activation+0.019
Top resid features:
,
Token,
Feature activation+0.017
Top resid features:
animal
Token animal
Feature activation+0.011
Top resid features:
in
Token in
Feature activation-0.004
Top resid features:
June
Token June
Feature activation-0.006
Top resid features:
.
Token.
Feature activation-0.003
Top resid features:
Ċ
TokenĊ
Feature activation-0.001
Top resid features:
Ċ
TokenĊ
Feature activation-0.021
Top resid features:
--
Token--
Feature activation+6.989
Top resid features:
Testing
Token Testing
Feature activation+0.008
Top resid features:
Tomorrow
Token Tomorrow
Feature activation-0.115
Top resid features:
âĢ
TokenâĢ
Feature activation-0.103
Top resid features:
Ļ
TokenĻ
Feature activation-0.127
Top resid features:
s
Tokens
Feature activation-0.146
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.636
Top resid features:
-
Token-
Feature activation+4.993
Top resid features:
century
Tokencentury
Feature activation+0.135
Top resid features:
discovery
Token discovery
Feature activation+0.038
Top resid features:
of
Token of
Feature activation+0.021
Top resid features:
remnants
Token remnants
Feature activation+0.014
Top resid features:
of
Token of
Feature activation+0.005
Top resid features:
in
Token in
Feature activation-0.003
Top resid features:
June
Token June
Feature activation-0.007
Top resid features:
.
Token.
Feature activation-0.006
Top resid features:
Ċ
TokenĊ
Feature activation-0.005
Top resid features:
Ċ
TokenĊ
Feature activation+0.002
Top resid features:
--
Token--
Feature activation+5.705
Top resid features:
Testing
Token Testing
Feature activation-0.010
Top resid features:
Tomorrow
Token Tomorrow
Feature activation-0.065
Top resid features:
âĢ
TokenâĢ
Feature activation-0.130
Top resid features:
Ļ
TokenĻ
Feature activation-0.054
Top resid features:
s
Tokens
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.574
Top resid features:
-
Token-
Feature activation+4.411
Top resid features:
products
Tokenproducts
Feature activation+0.152
Top resid features:
have
Token have
Feature activation+0.145
Top resid features:
just
Token just
Feature activation+0.044
Top resid features:
as
Token as
Feature activation-0.054
Top resid features:
valid
Token valid
Feature activation-0.058
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.590
Top resid features:
-
Token-
Feature activation+4.691
Top resid features:
0
Token0
Feature activation+0.201
Top resid features:
victory
Token victory
Feature activation+0.055
Top resid features:
over
Token over
Feature activation+0.027
Top resid features:
Orlando
Token Orlando
Feature activation+0.018
Top resid features:
City
Token City
Feature activation-0.006
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.609
Top resid features:
-
Token-
Feature activation+4.099
Top resid features:
squ
Tokensqu
Feature activation+0.032
Top resid features:
ad
Tokenad
Feature activation-0.004
Top resid features:
action
Token action
Feature activation+0.047
Top resid features:
is
Token is
Feature activation+0.049
Top resid features:
electr
Token electr
Feature activation-0.006
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.584
Top resid features:
-
Token-
Feature activation+3.791
Top resid features:
f
Tokenf
Feature activation+0.166
Top resid features:
ringe
Tokenringe
Feature activation+0.017
Top resid features:
item
Token item
Feature activation+0.014
Top resid features:
to
Token to
Feature activation+0.130
Top resid features:
an
Token an
Feature activation+0.015
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.485
Top resid features:
-
Token-
Feature activation+4.182
Top resid features:
speed
Tokenspeed
Feature activation+0.086
Top resid features:
collisions
Token collisions
Feature activation+0.005
Top resid features:
from
Token from
Feature activation+0.013
Top resid features:
large
Token large
Feature activation+0.002
Top resid features:
commercial
Token commercial
Feature activation-0.003
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.573
Top resid features:
-
Token-
Feature activation+4.072
Top resid features:
Gate
TokenGate
Feature activation+0.119
Top resid features:
wood
Tokenwood
Feature activation+0.037
Top resid features:
ruled
Token ruled
Feature activation+0.039
Top resid features:
from
Token from
Feature activation+0.053
Top resid features:
the
Token the
Feature activation+0.013
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.661
Top resid features:
-
Token-
Feature activation+3.965
Top resid features:
speed
Tokenspeed
Feature activation+0.051
Top resid features:
collisions
Token collisions
Feature activation+0.022
Top resid features:
from
Token from
Feature activation+0.060
Top resid features:
large
Token large
Feature activation+0.003
Top resid features:
commercial
Token commercial
Feature activation-0.024
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.331
Top resid features:
-
Token-
Feature activation+4.183
Top resid features:
of
Tokenof
Feature activation+0.086
Top resid features:
-
Token-
Feature activation-0.002
Top resid features:
w
Tokenw
Feature activation+0.001
Top resid features:
arr
Tokenarr
Feature activation+0.000
Top resid features:
anty
Tokenanty
Feature activation+0.051
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.597
Top resid features:
-
Token-
Feature activation+3.640
Top resid features:
up
Tokenup
Feature activation+0.209
Top resid features:
of
Token of
Feature activation+0.025
Top resid features:
a
Token a
Feature activation+0.019
Top resid features:
resume
Token resume
Feature activation+0.236
Top resid features:
.
Token.
Feature activation+0.052
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.530
Top resid features:
-
Token-
Feature activation+3.680
Top resid features:
show
Tokenshow
Feature activation+0.073
Top resid features:
-
Token-
Feature activation+0.012
Top resid features:
worthy
Tokenworthy
Feature activation+0.034
Top resid features:
visual
Token visual
Feature activation+0.002
Top resid features:
izations
Tokenizations
Feature activation+0.028
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.365
Top resid features:
-
Token-
Feature activation+3.882
Top resid features:
why
Tokenwhy
Feature activation+0.028
Top resid features:
-
Token-
Feature activation+0.004
Top resid features:
don
Tokendon
Feature activation-0.017
Top resid features:
âĢ
TokenâĢ
Feature activation-0.012
Top resid features:
Ļ
TokenĻ
Feature activation+0.035
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.560
Top resid features:
-
Token-
Feature activation+3.883
Top resid features:
positive
Tokenpositive
Feature activation+0.052
Top resid features:
stri
Token stri
Feature activation+0.002
Top resid features:
at
Tokenat
Feature activation-0.008
Top resid features:
op
Tokenop
Feature activation-0.004
Top resid features:
all
Tokenall
Feature activation-0.001
Top resid features:

Decoder Weights Distribution

Head 0: 0.05

Head 1: 0.03

Head 2: 0.03

Head 3: 0.07

Head 4: 0.53

Head 5: 0.06

Head 6: 0.07

Head 7: 0.03

Head 8: 0.04

Head 9: 0.06

Head 10: 0.01

Head 11: 0.02

Positive logits

·3.11

2.90

2.60

):2.41

-2.39

2.39

)*2.38

)2.37

-------2.36

2.32

*)2.30

)--2.28

)-2.25

2.22

├──2.17

2.17

---2.14

2.10

othing2.10

:2.08

Negative logits

',"-2.30

?'"-2.20

rift-2.12

halves-2.04

?".-2.03

,'"-1.97

CLASSIFIED-1.92

'."-1.91

priesthood-1.85

Sunder-1.84

REE-1.75

Thousand-1.75

.'"-1.75

poaching-1.72

holdings-1.71

stake-1.71

subsistence-1.71

dipping-1.68

?",-1.65

satur-1.65

INTERVAL 7.382 - 8.202
CONTAINS 0.000%

cover
Token cover
Feature activation+0.000
versus
Token versus
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
-
Token-
Feature activation+0.000
Advertisement
Token Advertisement
Feature activation+0.306
-
Token -
Feature activation+8.202
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.769
Let
TokenLet
Feature activation+0.000
's
Token's
Feature activation+0.000
watch
Token watch
Feature activation+0.000

INTERVAL 6.562 - 7.382
CONTAINS 0.000%

.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
-
Token-
Feature activation+0.000
Advertisement
Token Advertisement
Feature activation+0.296
-
Token -
Feature activation+6.938
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Much
TokenMuch
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
battery
Token battery
Feature activation+0.000
replacement
Token replacement
Feature activation+0.365
by
Token by
Feature activation+0.201
$
Token $
Feature activation+0.000
50
Token50
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+6.996
from
Token from
Feature activation+0.000
$
Token $
Feature activation+0.000
79
Token79
Feature activation+0.000
to
Token to
Feature activation+0.000
$
Token $
Feature activation+0.000
Af
Token Af
Feature activation+0.000
raid
Tokenraid
Feature activation+0.000
to
Token to
Feature activation+0.000
go
Token go
Feature activation+0.000
public
Token public
Feature activation+1.161
-
Token -
Feature activation+6.831
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Beyond
TokenBeyond
Feature activation+0.000
Fiat
Token Fiat
Feature activation+0.000
's
Token's
Feature activation+0.000

INTERVAL 5.742 - 6.562
CONTAINS 0.000%

hum
Token hum
Feature activation+0.000
pty
Tokenpty
Feature activation+0.000
-
Token-
Feature activation+0.000
d
Tokend
Feature activation+0.000
oo
Tokenoo
Feature activation+0.000
-
Token-
Feature activation+5.947
08
Token08
Feature activation+0.000
36
Token36
Feature activation+0.775
,
Token,
Feature activation+0.000
d
Token d
Feature activation+0.000
ar
Tokenar
Feature activation+0.000

INTERVAL 4.921 - 5.742
CONTAINS 0.000%

Tomorrow
Token Tomorrow
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+2.039
Ļ
TokenĻ
Feature activation+4.342
s
Tokens
Feature activation+0.000
Solutions
Token Solutions
Feature activation+0.000
--
Token --
Feature activation+5.476
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Former
TokenFormer
Feature activation+0.000
Mayor
Token Mayor
Feature activation+0.000
Lee
Token Lee
Feature activation+0.000

INTERVAL 4.101 - 4.921
CONTAINS 0.000%

Christ
Token Christ
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+1.036
Ļ
TokenĻ
Feature activation+1.083
s
Tokens
Feature activation+0.000
cross
Token cross
Feature activation+0.000
.
Token.
Feature activation+4.659
The
Token The
Feature activation+0.118
pope
Token pope
Feature activation+0.000
had
Token had
Feature activation+0.000
indicated
Token indicated
Feature activation+0.000
in
Token in
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
--
Token--
Feature activation+0.000
Testing
Token Testing
Feature activation+0.000
Tomorrow
Token Tomorrow
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+2.039
Ļ
TokenĻ
Feature activation+4.342
s
Tokens
Feature activation+0.000
Solutions
Token Solutions
Feature activation+0.000
--
Token --
Feature activation+5.476
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000

INTERVAL 3.281 - 4.101
CONTAINS 0.001%

$
Token $
Feature activation+0.000
79
Token79
Feature activation+0.000
to
Token to
Feature activation+0.000
$
Token $
Feature activation+0.000
29
Token29
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+3.434
for
Token for
Feature activation+0.000
anyone
Token anyone
Feature activation+0.000
with
Token with
Feature activation+0.000
an
Token an
Feature activation+0.000
iPhone
Token iPhone
Feature activation+0.000
skins
Token skins
Feature activation+0.000
to
Token to
Feature activation+0.000
share
Token share
Feature activation+0.000
with
Token with
Feature activation+0.000
others
Token others
Feature activation+0.000
.
Token.
Feature activation+3.313
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+2.200
In
TokenIn
Feature activation+0.000
1999
Token 1999
Feature activation+0.000
,
Token,
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.126
Ċ
TokenĊ
Feature activation+2.343
Engine
TokenEngine
Feature activation+0.000
ering
Tokenering
Feature activation+0.000
News
Token News
Feature activation+0.000
-
Token-
Feature activation+3.623
Record
TokenRecord
Feature activation+0.000
reported
Token reported
Feature activation+0.000
in
Token in
Feature activation+0.000
1964
Token 1964
Feature activation+0.000
that
Token that
Feature activation+0.000
over
Token over
Feature activation+0.619
Orlando
Token Orlando
Feature activation+0.533
City
Token City
Feature activation+0.599
B
Token B
Feature activation+0.772
Ċ
TokenĊ
Feature activation+1.154
Ċ
TokenĊ
Feature activation+3.999
Eastern
TokenEastern
Feature activation+0.000
Conference
Token Conference
Feature activation+0.000
Sem
Token Sem
Feature activation+0.000
if
Tokenif
Feature activation+0.000
inal
Tokeninal
Feature activation+0.000
of
Token of
Feature activation+0.000
Wall
Token Wall
Feature activation+0.000
Street
Token Street
Feature activation+0.228
's
Token's
Feature activation+0.000
attention
Token attention
Feature activation+0.767
.
Token.
Feature activation+3.631
C
Token C
Feature activation+0.000
ME
TokenME
Feature activation+0.000
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
world
Token world
Feature activation+0.000

INTERVAL 2.461 - 3.281
CONTAINS 0.003%

-
Token-
Feature activation+0.000
to
Token to
Feature activation+0.000
five
Token five
Feature activation+0.000
-
Token-
Feature activation+2.377
minute
Tokenminute
Feature activation+0.028
video
Token video
Feature activation+2.640
,
Token,
Feature activation+1.828
participants
Token participants
Feature activation+0.000
are
Token are
Feature activation+0.000
asked
Token asked
Feature activation+0.000
to
Token to
Feature activation+0.000
-
Token-
Feature activation+0.000
lighting
Tokenlighting
Feature activation+0.005
Bush
Token Bush
Feature activation+0.000
tax
Token tax
Feature activation+0.125
cuts
Token cuts
Feature activation+0.000
.
Token.
Feature activation+3.195
The
Token The
Feature activation+0.000
same
Token same
Feature activation+0.000
guy
Token guy
Feature activation+0.000
who
Token who
Feature activation+0.000
testified
Token testified
Feature activation+0.000
ar
Tokenar
Feature activation+0.000
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
Indy
Token Indy
Feature activation+0.000
500
Token 500
Feature activation+0.637
.
Token.
Feature activation+2.685
As
Token As
Feature activation+0.000
a
Token a
Feature activation+0.000
supplier
Token supplier
Feature activation+0.000
to
Token to
Feature activation+0.000
Rid
Token Rid
Feature activation+0.000
are
Token are
Feature activation+0.000
two
Token two
Feature activation+0.000
completely
Token completely
Feature activation+0.000
different
Token different
Feature activation+0.000
environments
Token environments
Feature activation+0.000
.
Token.
Feature activation+3.078
Sheffield
Token Sheffield
Feature activation+0.000
,
Token,
Feature activation+0.000
you
Token you
Feature activation+0.000
can
Token can
Feature activation+0.000
get
Token get
Feature activation+0.000
-
Token-
Feature activation+0.000
making
Tokenmaking
Feature activation+0.000
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
Senate
Token Senate
Feature activation+1.542
.
Token.
Feature activation+2.596
But
Token But
Feature activation+0.000
the
Token the
Feature activation+0.000
tea
Token tea
Feature activation+0.000
party
Token party
Feature activation+0.000
wing
Token wing
Feature activation+0.000

INTERVAL 1.640 - 2.461
CONTAINS 0.004%

to
Token to
Feature activation+0.000
five
Token five
Feature activation+0.000
-
Token-
Feature activation+2.377
minute
Tokenminute
Feature activation+0.028
video
Token video
Feature activation+2.640
,
Token,
Feature activation+1.828
participants
Token participants
Feature activation+0.000
are
Token are
Feature activation+0.000
asked
Token asked
Feature activation+0.000
to
Token to
Feature activation+0.000
make
Token make
Feature activation+0.000
every
Token every
Feature activation+0.000
hour
Token hour
Feature activation+0.000
of
Token of
Feature activation+0.000
every
Token every
Feature activation+0.000
day
Token day
Feature activation+0.000
.
Token.
Feature activation+2.352
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.658
One
TokenOne
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
iti
Tokeniti
Feature activation+0.000
ated
Tokenated
Feature activation+0.000
deadline
Token deadline
Feature activation+0.000
has
Token has
Feature activation+0.000
passed
Token passed
Feature activation+0.732
.
Token.
Feature activation+2.458
Stein
Token Stein
Feature activation+0.000
will
Token will
Feature activation+0.000
have
Token have
Feature activation+0.000
to
Token to
Feature activation+0.000
challenge
Token challenge
Feature activation+0.000
-
Token-
Feature activation+0.000
day
Tokenday
Feature activation+0.000
trial
Token trial
Feature activation+0.777
.
Token.
Feature activation+2.529
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+1.696
We
TokenWe
Feature activation+0.000
love
Token love
Feature activation+0.000
you
Token you
Feature activation+0.000
.
Token.
Feature activation+0.081
Ċ
TokenĊ
Feature activation+0.000
usterity
Tokenusterity
Feature activation+0.153
alternative
Token alternative
Feature activation+1.511
to
Token to
Feature activation+0.637
the
Token the
Feature activation+0.069
Westminster
Token Westminster
Feature activation+0.000
establishment
Token establishment
Feature activation+2.227
is
Token is
Feature activation+0.000
âĢĵ
Token âĢĵ
Feature activation+1.281
not
Token not
Feature activation+0.000
just
Token just
Feature activation+0.000
to
Token to
Feature activation+0.000

INTERVAL 0.820 - 1.640
CONTAINS 0.014%

âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
on
Token on
Feature activation+0.000
college
Token college
Feature activation+0.000
campuses
Token campuses
Feature activation+0.000
]
Token]
Feature activation+1.508
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
new
Token new
Feature activation+0.000
law
Token law
Feature activation+0.000
ribute
Tokenribute
Feature activation+0.000
Your
Token Your
Feature activation+0.000
Fan
Token Fan
Feature activation+0.000
The
Token The
Feature activation+0.000
ories
Tokenories
Feature activation+0.000
===
Token ===
Feature activation+1.600
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Join
TokenJoin
Feature activation+0.000
Fans
Token Fans
Feature activation+0.000
of
Token of
Feature activation+0.000
14
Token14
Feature activation+0.000
âĢ¢
Token âĢ¢
Feature activation+0.000
@
Token @
Feature activation+0.000
BF
TokenBF
Feature activation+0.000
itz
Tokenitz
Feature activation+0.000
gerald
Tokengerald
Feature activation+1.110
T
TokenT
Feature activation+0.147
U
TokenU
Feature activation+0.916
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
The
TokenThe
Feature activation+0.000
Los
Token Los
Feature activation+0.000
before
Token before
Feature activation+0.000
a
Token a
Feature activation+0.000
scheduled
Token scheduled
Feature activation+0.000
signing
Token signing
Feature activation+0.000
event
Token event
Feature activation+0.000
.
Token.
Feature activation+0.910
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Wiki
TokenWiki
Feature activation+0.000
Leaks
TokenLeaks
Feature activation+0.000
ends
Token ends
Feature activation+0.000
either
Token either
Feature activation+0.000
mobile
Token mobile
Feature activation+0.000
or
Token or
Feature activation+0.000
handheld
Token handheld
Feature activation+0.000
platforms
Token platforms
Feature activation+0.000
.
Token.
Feature activation+1.594
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
-
Token-
Feature activation+0.287
No
TokenNo
Feature activation+0.000
games
Token games
Feature activation+0.000

INTERVAL 0.000 - 0.820
CONTAINS 99.977%

who
Token who
Feature activation+0.000
red
Token red
Feature activation+0.000
sh
Tokensh
Feature activation+0.000
irted
Tokenirted
Feature activation+0.000
his
Token his
Feature activation+0.000
first
Token first
Feature activation+0.000
year
Token year
Feature activation+0.000
and
Token and
Feature activation+0.000
then
Token then
Feature activation+0.000
saw
Token saw
Feature activation+0.000
most
Token most
Feature activation+0.000
TA
Token TA
Feature activation+0.000
KE
TokenKE
Feature activation+0.000
THAT
Token THAT
Feature activation+0.000
CH
Token CH
Feature activation+0.000
ANCE
TokenANCE
Feature activation+0.000
,
Token,
Feature activation+0.000
DON
Token DON
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
T
TokenT
Feature activation+0.000
P
Token P
Feature activation+0.000
about
Token about
Feature activation+0.000
Eric
Token Eric
Feature activation+0.000
Holder
Token Holder
Feature activation+0.000
's
Token's
Feature activation+0.000
position
Token position
Feature activation+0.000
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
whether
Token whether
Feature activation+0.000
it
Token it
Feature activation+0.000
applied
Token applied
Feature activation+0.000
to
Token to
Feature activation+0.000
two
Token two
Feature activation+0.000
on
Token on
Feature activation+0.000
-
Token-
Feature activation+0.000
prem
Tokenprem
Feature activation+0.000
ises
Tokenises
Feature activation+0.000
servers
Token servers
Feature activation+0.000
.
Token.
Feature activation+0.000
The
Token The
Feature activation+0.000
maintenance
Token maintenance
Feature activation+0.000
,
Token,
Feature activation+0.000
frequent
Token frequent
Feature activation+0.000
a
Token a
Feature activation+0.000
girls
Token girls
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
novel
Token novel
Feature activation+0.000
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
mother
Token mother
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
romance
Token romance
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