TOP ACTIVATIONS
MAX = 0.127

know , Lily . Cater pill ars need leaves and air
Did you know that cater pill ars turn into beautiful butterflies
careful , Lily . Cater pill ars can be tricky .
saw lots of colourful cater pill ars and an old lady
We have different likes and disl ikes . We can be
and watched as the cater pill ars grew bigger and bigger
he noticed some other cater pill ars stretching towards it too
you , Lily . Cater pill ars can leave and never
but he also knew cater pill ars were not pets .
ily liked to keep cater pill ars in a jar .
, " I love cater pill ars ! You can be
put it back . Cater pill ars need to eat leaves
Lily , no . Cater pill ars are not for keeping
was a box of ch ocol ates . Lily loved ch
was a box of ch ocol ates . " For me
and started eating the ch ocol ates . The shop keeper
not eat too many ch ocol ates and his friend learned
a big box of ch ocol ates . It was so
ates . Lily loved ch ocol ates . She thanked Ben
and lady bugs and cater pill ars ." Tim

MIN ACTIVATIONS
MIN = -0.155

the cars stopped and the pedestrian light turned green . Then
and books . They see alt ars and offerings and priests
so beautiful inside and the fringe from the cut shells made
yellow and green until the formerly color less walls seemed to
replied with a gentle ' yip '. The little
show . All the beautiful clot he es were marching down
You are under arrest for harassing and stealing from these children
stop . Anna and Ben unb uck led and got off
- a shirt and pants combo - and brushed his hair
her the pan with the unb aked cake . Mom was
So , with artificial so aps , they made
then on , the SPL ASH was always under control !
smiled and said , " Church is so nice mom my
make one too . He dips his brush in the red
. It was poison he ml ock . It was very
up a carrot stick and dips it in a white sauce
" We are the weapon creators ! We can stop the
It is blue . He dips his brush in the paint
and a wand . She dips the wand in the soap
chicken n ug get and dips it in a red sauce

INTERVAL 0.057 - 0.127
CONTAINS 0.165%

their mom . They saw many flowers , birds and bugs
saw a cute little puppy w agging its tail .
the little girl felt really proud of herself for becoming brave
week . The farmer had many animals that Sam loved to
I will give you the courage and strength to save yourself

INTERVAL -0.014 - 0.057
CONTAINS 81.980%

and Eric played the trumpet . The next day
P oppy and her dad threw the ball back
their mom if they could play some more . Their mom
not help the baby bird . <|endoftext|> Once upon a time
. They forget about the cat . The cat forg ets

INTERVAL -0.084 - -0.014
CONTAINS 17.841%

, mom . Tom put me in jail . But I
store . He was looking for a new hat to wear
plate . She saw the pastry and the note . She
day , the little boy decided to try and count the
the teacher wrote on the black board with chalk . Lily

INTERVAL -0.155 - -0.084
CONTAINS 0.014%

I want to see . Maybe it is a doll or
house . Billy stamped his feet and said ,
noise . " Crunch , crunch , crunch .
dream , Tim my . Maybe we can find a book
to them . He w ags his tail .
Token know
Token ID760
Feature activation+1.37e-02

Pos logit contributions
yles+0.016
 spill+0.016
yle+0.016
cases+0.016
 hairc+0.015
Neg logit contributions
age-0.020
>-0.018
amed-0.017
 Follow-0.017
man-0.017
Token,
Token ID11
Feature activation-3.51e-04

Pos logit contributions
cases+0.028
yles+0.028
 salon+0.026
yle+0.026
outing+0.025
Neg logit contributions
age-0.017
 back-0.016
 right-0.015
 welcome-0.015
est-0.015
Token Lily
Token ID20037
Feature activation+1.34e-02

Pos logit contributions
age+0.001
 back+0.001
 welcome+0.001
 right+0.001
 home+0.001
Neg logit contributions
yles-0.001
cases-0.001
ls-0.001
yle-0.001
outing-0.001
Token.
Token ID13
Feature activation+1.91e-03

Pos logit contributions
cases+0.021
yles+0.021
ls+0.021
yle+0.019
 salon+0.019
Neg logit contributions
age-0.024
amed-0.020
>-0.020
 Follow-0.019
 welcome-0.019
Token Cater
Token ID40742
Feature activation+4.31e-03

Pos logit contributions
yles+0.004
cases+0.004
yle+0.003
ls+0.003
 salon+0.003
Neg logit contributions
age-0.003
amed-0.002
>-0.002
 home-0.002
 back-0.002
Tokenpill
Token ID27215
Feature activation+1.27e-01

Pos logit contributions
yles+0.005
ls+0.005
cases+0.005
yle+0.005
 salon+0.004
Neg logit contributions
age-0.010
 back-0.009
>-0.009
 home-0.008
 welcome-0.008
Tokenars
Token ID945
Feature activation+1.50e-02

Pos logit contributions
cases+0.221
ls+0.221
yles+0.218
yle+0.202
outing+0.201
Neg logit contributions
age-0.193
 home-0.170
gy-0.168
 back-0.167
>-0.167
Token need
Token ID761
Feature activation-8.70e-04

Pos logit contributions
yles+0.027
cases+0.025
yle+0.024
 salon+0.023
ries+0.023
Neg logit contributions
age-0.026
 back-0.023
 welcome-0.023
 right-0.023
est-0.022
Token leaves
Token ID5667
Feature activation+6.60e-03

Pos logit contributions
age+0.001
>+0.001
gy+0.001
amed+0.001
est+0.001
Neg logit contributions
yles-0.002
cases-0.002
ls-0.002
 hairc-0.001
yle-0.001
Token and
Token ID290
Feature activation-7.63e-04

Pos logit contributions
yles+0.013
cases+0.013
ls+0.012
yle+0.012
 salon+0.012
Neg logit contributions
age-0.010
 back-0.008
>-0.008
 welcome-0.008
 home-0.008
Token air
Token ID1633
Feature activation-1.48e-02

Pos logit contributions
age+0.001
 back+0.001
 welcome+0.001
>+0.001
 home+0.001
Neg logit contributions
yles-0.001
cases-0.001
yle-0.001
ls-0.001
 salon-0.001
Token Did
Token ID7731
Feature activation-1.03e-02

Pos logit contributions
age+0.003
 back+0.003
 home+0.003
amed+0.002
>+0.002
Neg logit contributions
yles-0.003
yle-0.003
cases-0.003
 hairc-0.003
 salon-0.003
Token you
Token ID345
Feature activation-4.23e-03

Pos logit contributions
age+0.013
 back+0.012
 right+0.012
 welcome+0.011
 home+0.011
Neg logit contributions
cases-0.021
yles-0.021
outing-0.020
ls-0.019
 tant-0.019
Token know
Token ID760
Feature activation+1.19e-02

Pos logit contributions
age+0.007
 back+0.006
 welcome+0.005
>+0.005
 home+0.005
Neg logit contributions
yles-0.008
cases-0.008
ls-0.007
yle-0.007
 salon-0.007
Token that
Token ID326
Feature activation+6.17e-03

Pos logit contributions
cases+0.021
yles+0.021
yle+0.019
ls+0.019
 salon+0.019
Neg logit contributions
age-0.019
 back-0.017
 welcome-0.017
 home-0.017
 right-0.016
Token cater
Token ID20825
Feature activation+1.03e-02

Pos logit contributions
yles+0.011
cases+0.010
ls+0.010
yle+0.010
 salon+0.010
Neg logit contributions
age-0.011
>-0.009
 back-0.009
 welcome-0.008
amed-0.008
Tokenpill
Token ID27215
Feature activation+1.26e-01

Pos logit contributions
yles+0.011
cases+0.011
ls+0.010
yle+0.010
 salon+0.010
Neg logit contributions
age-0.024
 back-0.021
>-0.021
 welcome-0.021
 home-0.021
Tokenars
Token ID945
Feature activation+1.38e-02

Pos logit contributions
cases+0.214
ls+0.211
yles+0.211
 else+0.196
yle+0.194
Neg logit contributions
age-0.194
 home-0.176
 back-0.176
gy-0.175
>-0.170
Token turn
Token ID1210
Feature activation+1.72e-02

Pos logit contributions
yles+0.027
cases+0.026
yle+0.025
 salon+0.025
 hairc+0.023
Neg logit contributions
age-0.021
 back-0.019
 welcome-0.018
 right-0.017
 home-0.017
Token into
Token ID656
Feature activation+1.35e-03

Pos logit contributions
yles+0.036
cases+0.036
ls+0.034
yle+0.033
 salon+0.033
Neg logit contributions
age-0.023
 back-0.018
>-0.018
 welcome-0.018
 home-0.017
Token beautiful
Token ID4950
Feature activation-8.50e-03

Pos logit contributions
yles+0.003
cases+0.003
ls+0.002
yle+0.002
 salon+0.002
Neg logit contributions
age-0.002
 back-0.002
>-0.002
 welcome-0.002
 home-0.001
Token butterflies
Token ID44571
Feature activation-3.22e-03

Pos logit contributions
age+0.015
 back+0.012
>+0.012
 welcome+0.012
 home+0.012
Neg logit contributions
yles-0.014
cases-0.014
ls-0.013
yle-0.013
 salon-0.013
Token careful
Token ID8161
Feature activation+3.33e-03

Pos logit contributions
eling+0.044
outing+0.041
 else+0.037
 frowned+0.036
ls+0.036
Neg logit contributions
age-0.047
est-0.043
ney-0.042
gy-0.042
inea-0.041
Token,
Token ID11
Feature activation-8.76e-04

Pos logit contributions
cases+0.006
 spill+0.005
outing+0.005
yles+0.005
 salon+0.005
Neg logit contributions
age-0.005
 Follow-0.004
amed-0.004
>-0.004
 welcome-0.004
Token Lily
Token ID20037
Feature activation+9.98e-03

Pos logit contributions
age+0.001
 right+0.001
 back+0.001
 welcome+0.001
 ok+0.001
Neg logit contributions
yles-0.002
cases-0.002
yle-0.001
ls-0.001
 salon-0.001
Token.
Token ID13
Feature activation+5.07e-04

Pos logit contributions
cases+0.017
yles+0.017
ls+0.016
 salon+0.016
 spill+0.015
Neg logit contributions
age-0.016
amed-0.013
 Follow-0.013
>-0.013
 back-0.013
Token Cater
Token ID40742
Feature activation+2.24e-03

Pos logit contributions
yles+0.001
cases+0.001
yle+0.001
 salon+0.001
ls+0.001
Neg logit contributions
age-0.001
 back-0.001
 home-0.001
>-0.001
est-0.001
Tokenpill
Token ID27215
Feature activation+1.26e-01

Pos logit contributions
yles+0.003
cases+0.003
ls+0.003
yle+0.003
 salon+0.002
Neg logit contributions
age-0.005
 back-0.004
>-0.004
 home-0.004
 welcome-0.004
Tokenars
Token ID945
Feature activation+1.42e-02

Pos logit contributions
ls+0.208
cases+0.206
yles+0.195
 else+0.188
outing+0.186
Neg logit contributions
age-0.193
 home-0.184
gy-0.181
 back-0.178
>-0.173
Token can
Token ID460
Feature activation+6.45e-03

Pos logit contributions
yles+0.024
cases+0.023
yle+0.021
 salon+0.021
ls+0.021
Neg logit contributions
age-0.026
 back-0.023
 welcome-0.022
>-0.022
 home-0.022
Token be
Token ID307
Feature activation+2.36e-02

Pos logit contributions
yles+0.011
cases+0.011
yle+0.010
ls+0.010
 salon+0.010
Neg logit contributions
age-0.011
 back-0.009
 welcome-0.009
 home-0.009
>-0.009
Token tricky
Token ID17198
Feature activation+1.15e-02

Pos logit contributions
yles+0.038
cases+0.037
ls+0.036
 spill+0.035
ries+0.034
Neg logit contributions
age-0.042
>-0.035
est-0.034
 Follow-0.034
 back-0.034
Token.
Token ID13
Feature activation+1.72e-03

Pos logit contributions
cases+0.020
yles+0.020
ls+0.019
 hairc+0.019
 spill+0.018
Neg logit contributions
age-0.019
>-0.015
amed-0.015
 Follow-0.015
gy-0.015
Token saw
Token ID2497
Feature activation-2.34e-02

Pos logit contributions
yles+0.005
cases+0.005
ls+0.005
yle+0.005
ries+0.005
Neg logit contributions
age-0.005
 back-0.005
>-0.005
 home-0.005
 welcome-0.004
Token lots
Token ID6041
Feature activation+5.23e-02

Pos logit contributions
age+0.029
 back+0.023
 home+0.022
 welcome+0.022
>+0.022
Neg logit contributions
yles-0.052
cases-0.051
ls-0.049
yle-0.048
 salon-0.048
Token of
Token ID286
Feature activation+3.06e-02

Pos logit contributions
yles+0.121
cases+0.113
yle+0.110
 salon+0.109
outing+0.108
Neg logit contributions
est-0.057
 welcome-0.055
age-0.055
 back-0.053
 right-0.052
Token colourful
Token ID44799
Feature activation+9.71e-03

Pos logit contributions
yles+0.057
cases+0.056
 salon+0.052
yle+0.052
outing+0.050
Neg logit contributions
age-0.047
 welcome-0.043
 home-0.042
 right-0.042
est-0.042
Token cater
Token ID20825
Feature activation+1.27e-02

Pos logit contributions
yles+0.019
cases+0.019
 salon+0.018
yle+0.018
outing+0.018
Neg logit contributions
age-0.013
est-0.012
 home-0.012
 back-0.012
 welcome-0.011
Tokenpill
Token ID27215
Feature activation+1.24e-01

Pos logit contributions
yles+0.008
cases+0.008
yle+0.006
ls+0.006
 salon+0.006
Neg logit contributions
age-0.036
 back-0.032
>-0.032
 welcome-0.032
 home-0.031
Tokenars
Token ID945
Feature activation+1.53e-02

Pos logit contributions
cases+0.218
ls+0.216
yles+0.207
 else+0.191
l+0.190
Neg logit contributions
age-0.189
 back-0.173
est-0.169
 home-0.169
gy-0.169
Token and
Token ID290
Feature activation+1.67e-03

Pos logit contributions
yles+0.032
cases+0.031
yle+0.030
 salon+0.030
ls+0.029
Neg logit contributions
age-0.020
 back-0.017
 home-0.016
 right-0.016
 across-0.016
Token an
Token ID281
Feature activation+4.84e-03

Pos logit contributions
yles+0.003
cases+0.003
yle+0.003
outing+0.003
 salon+0.003
Neg logit contributions
age-0.003
 back-0.003
 welcome-0.002
 home-0.002
 across-0.002
Token old
Token ID1468
Feature activation+2.38e-02

Pos logit contributions
yles+0.008
cases+0.007
yle+0.007
ries+0.007
ls+0.007
Neg logit contributions
age-0.008
 Follow-0.007
 back-0.007
>-0.007
amed-0.006
Token lady
Token ID10846
Feature activation+1.62e-02

Pos logit contributions
yles+0.036
 salon+0.036
cases+0.035
ls+0.035
yle+0.034
Neg logit contributions
age-0.042
>-0.037
 Follow-0.036
amed-0.036
 back-0.035
Token We
Token ID775
Feature activation-1.56e-02

Pos logit contributions
yles+0.013
cases+0.012
yle+0.012
ls+0.012
 salon+0.011
Neg logit contributions
age-0.009
 back-0.008
>-0.008
 home-0.007
est-0.007
Token have
Token ID423
Feature activation+1.92e-02

Pos logit contributions
age+0.023
 right+0.019
 back+0.019
 welcome+0.018
 home+0.018
Neg logit contributions
yles-0.033
cases-0.032
ries-0.030
yle-0.029
ls-0.029
Token different
Token ID1180
Feature activation-1.76e-02

Pos logit contributions
yles+0.042
cases+0.041
yle+0.039
ls+0.038
outing+0.037
Neg logit contributions
age-0.024
 back-0.022
 home-0.021
 right-0.020
 welcome-0.020
Token likes
Token ID7832
Feature activation+1.48e-02

Pos logit contributions
age+0.035
 back+0.031
 welcome+0.030
>+0.030
est+0.030
Neg logit contributions
yles-0.025
cases-0.025
yle-0.023
ls-0.023
 salon-0.022
Token and
Token ID290
Feature activation+1.11e-03

Pos logit contributions
cases+0.028
yles+0.027
yle+0.025
outing+0.025
ls+0.025
Neg logit contributions
age-0.023
 back-0.021
 home-0.020
 welcome-0.020
>-0.019
Token disl
Token ID19621
Feature activation+1.22e-01

Pos logit contributions
yles+0.002
cases+0.002
yle+0.002
ls+0.002
outing+0.002
Neg logit contributions
age-0.002
 back-0.002
 welcome-0.002
 right-0.001
 home-0.001
Tokenikes
Token ID7938
Feature activation-5.06e-02

Pos logit contributions
yles+0.187
cases+0.185
ls+0.177
 salon+0.177
yle+0.174
Neg logit contributions
age-0.221
est-0.189
>-0.185
 back-0.185
 right-0.182
Token.
Token ID13
Feature activation+5.62e-03

Pos logit contributions
age+0.073
 back+0.060
 welcome+0.058
>+0.058
 home+0.057
Neg logit contributions
yles-0.102
cases-0.100
yle-0.094
ls-0.094
 salon-0.092
Token We
Token ID775
Feature activation-1.62e-02

Pos logit contributions
yles+0.011
cases+0.010
yle+0.010
ls+0.009
 salon+0.009
Neg logit contributions
age-0.009
 home-0.007
 back-0.007
>-0.007
est-0.007
Token can
Token ID460
Feature activation+1.09e-02

Pos logit contributions
age+0.025
 back+0.021
 right+0.020
 welcome+0.020
 home+0.020
Neg logit contributions
yles-0.033
cases-0.032
ries-0.030
yle-0.029
ls-0.029
Token be
Token ID307
Feature activation+2.73e-02

Pos logit contributions
yles+0.021
cases+0.021
ls+0.019
yle+0.019
ries+0.019
Neg logit contributions
age-0.017
 back-0.014
 welcome-0.014
 home-0.013
>-0.013
Token and
Token ID290
Feature activation+8.41e-04

Pos logit contributions
yles+0.056
cases+0.054
ls+0.052
yle+0.052
 salon+0.051
Neg logit contributions
age-0.035
 back-0.028
>-0.027
 home-0.027
 welcome-0.027
Token watched
Token ID7342
Feature activation-2.07e-03

Pos logit contributions
yles+0.002
cases+0.002
ls+0.002
outing+0.002
yle+0.002
Neg logit contributions
age-0.001
 back-0.001
 home-0.001
 right-0.001
 welcome-0.001
Token as
Token ID355
Feature activation+4.97e-02

Pos logit contributions
age+0.003
 back+0.003
>+0.003
 welcome+0.003
 home+0.003
Neg logit contributions
yles-0.004
cases-0.004
ls-0.003
yle-0.003
 salon-0.003
Token the
Token ID262
Feature activation-1.22e-02

Pos logit contributions
cases+0.098
outing+0.094
yles+0.093
eling+0.090
 salon+0.088
Neg logit contributions
 home-0.066
 back-0.065
 right-0.061
 welcome-0.059
 across-0.058
Token cater
Token ID20825
Feature activation+1.28e-02

Pos logit contributions
age+0.023
 back+0.019
>+0.019
 welcome+0.019
 home+0.018
Neg logit contributions
yles-0.020
cases-0.019
ls-0.018
yle-0.018
 salon-0.017
Tokenpill
Token ID27215
Feature activation+1.21e-01

Pos logit contributions
cases+0.025
yles+0.024
yle+0.022
 salon+0.022
outing+0.022
Neg logit contributions
age-0.021
 back-0.016
est-0.016
 welcome-0.016
>-0.016
Tokenars
Token ID945
Feature activation+1.44e-02

Pos logit contributions
cases+0.235
yles+0.231
ls+0.223
yle+0.213
 salon+0.210
Neg logit contributions
age-0.173
 back-0.145
 home-0.144
>-0.143
est-0.141
Token grew
Token ID6348
Feature activation+1.61e-02

Pos logit contributions
yles+0.026
yle+0.024
cases+0.024
 hairc+0.023
outing+0.023
Neg logit contributions
age-0.024
 back-0.023
 across-0.021
 home-0.021
 right-0.021
Token bigger
Token ID5749
Feature activation+2.62e-02

Pos logit contributions
yles+0.033
cases+0.033
ls+0.032
ries+0.030
yle+0.030
Neg logit contributions
age-0.023
>-0.018
 welcome-0.017
 Follow-0.017
est-0.017
Token and
Token ID290
Feature activation+2.62e-04

Pos logit contributions
yles+0.054
cases+0.053
yle+0.050
 salon+0.050
outing+0.049
Neg logit contributions
age-0.036
 back-0.030
 home-0.029
est-0.028
 right-0.028
Token bigger
Token ID5749
Feature activation+2.61e-02

Pos logit contributions
yles+0.001
cases+0.001
ls+0.001
yle+0.001
 salon+0.001
Neg logit contributions
age-0.000
 back-0.000
>-0.000
 welcome-0.000
 home-0.000
Token he
Token ID339
Feature activation+2.51e-02

Pos logit contributions
yles+0.005
cases+0.005
ls+0.004
 salon+0.004
yle+0.004
Neg logit contributions
age-0.006
 back-0.005
 home-0.005
 welcome-0.005
 right-0.005
Token noticed
Token ID6810
Feature activation-3.94e-02

Pos logit contributions
yles+0.041
cases+0.040
yle+0.037
ls+0.037
 salon+0.036
Neg logit contributions
age-0.045
 back-0.040
>-0.038
 home-0.038
 right-0.038
Token some
Token ID617
Feature activation+3.99e-02

Pos logit contributions
age+0.054
 back+0.048
 home+0.046
 right+0.045
 welcome+0.043
Neg logit contributions
yles-0.079
cases-0.079
ls-0.074
 salon-0.073
yle-0.072
Token other
Token ID584
Feature activation+2.94e-02

Pos logit contributions
yles+0.082
cases+0.079
yle+0.073
 salon+0.071
ls+0.070
Neg logit contributions
age-0.054
 back-0.053
 right-0.051
 welcome-0.051
est-0.051
Token cater
Token ID20825
Feature activation+1.28e-02

Pos logit contributions
yles+0.050
cases+0.050
ls+0.045
outing+0.045
 salon+0.044
Neg logit contributions
age-0.049
est-0.046
 back-0.045
 welcome-0.045
 home-0.044
Tokenpill
Token ID27215
Feature activation+1.21e-01

Pos logit contributions
cases+0.013
yles+0.012
 salon+0.011
yle+0.011
outing+0.010
Neg logit contributions
age-0.032
est-0.028
 back-0.027
 welcome-0.027
>-0.027
Tokenars
Token ID945
Feature activation+1.48e-02

Pos logit contributions
cases+0.225
yles+0.220
ls+0.216
yle+0.202
outing+0.199
Neg logit contributions
age-0.178
 back-0.155
 home-0.155
>-0.151
gy-0.150
Token stretching
Token ID20880
Feature activation-7.95e-03

Pos logit contributions
yles+0.032
cases+0.030
yle+0.030
 salon+0.029
ls+0.029
Neg logit contributions
age-0.019
 back-0.018
 right-0.016
 across-0.016
 home-0.015
Token towards
Token ID3371
Feature activation-1.01e-02

Pos logit contributions
age+0.011
>+0.009
 welcome+0.009
 back+0.009
 Follow+0.009
Neg logit contributions
yles-0.016
cases-0.016
ls-0.015
yle-0.015
 salon-0.015
Token it
Token ID340
Feature activation+1.06e-02

Pos logit contributions
age+0.018
 back+0.016
 home+0.015
 welcome+0.015
 right+0.015
Neg logit contributions
cases-0.017
yles-0.017
ls-0.016
 salon-0.015
yle-0.015
Token too
Token ID1165
Feature activation+3.74e-02

Pos logit contributions
cases+0.019
yles+0.019
ls+0.018
yle+0.017
 salon+0.017
Neg logit contributions
age-0.017
>-0.014
 welcome-0.014
 Follow-0.013
amed-0.013
Token you
Token ID345
Feature activation-2.59e-03

Pos logit contributions
age+0.004
 back+0.004
 home+0.004
 right+0.004
 welcome+0.003
Neg logit contributions
yles-0.006
cases-0.006
ls-0.005
 salon-0.005
outing-0.005
Token,
Token ID11
Feature activation+1.10e-03

Pos logit contributions
age+0.004
 back+0.003
 welcome+0.003
>+0.003
 home+0.003
Neg logit contributions
yles-0.005
cases-0.005
yle-0.004
ls-0.004
 salon-0.004
Token Lily
Token ID20037
Feature activation+7.03e-03

Pos logit contributions
yles+0.002
cases+0.002
ls+0.002
yle+0.002
outing+0.002
Neg logit contributions
age-0.002
 right-0.001
 welcome-0.001
 back-0.001
 home-0.001
Token.
Token ID13
Feature activation+4.36e-03

Pos logit contributions
cases+0.011
yles+0.011
ls+0.011
 salon+0.010
yle+0.010
Neg logit contributions
age-0.012
amed-0.010
 Follow-0.010
>-0.010
est-0.009
Token Cater
Token ID40742
Feature activation+3.13e-03

Pos logit contributions
yles+0.008
cases+0.008
yle+0.008
 salon+0.008
 hairc+0.007
Neg logit contributions
age-0.006
 back-0.005
amed-0.005
 home-0.005
>-0.005
Tokenpill
Token ID27215
Feature activation+1.19e-01

Pos logit contributions
yles+0.002
cases+0.002
ls+0.002
yle+0.002
 salon+0.001
Neg logit contributions
age-0.009
 back-0.008
>-0.008
 home-0.008
 welcome-0.008
Tokenars
Token ID945
Feature activation+1.30e-02

Pos logit contributions
ls+0.190
cases+0.188
 else+0.176
yles+0.172
l+0.169
Neg logit contributions
age-0.185
gy-0.183
 home-0.178
wood-0.175
 back-0.170
Token can
Token ID460
Feature activation+8.04e-03

Pos logit contributions
yles+0.021
cases+0.020
yle+0.019
ls+0.019
 salon+0.019
Neg logit contributions
age-0.024
 back-0.021
 welcome-0.021
 right-0.020
>-0.020
Token leave
Token ID2666
Feature activation+1.45e-02

Pos logit contributions
yles+0.016
cases+0.016
ls+0.015
yle+0.015
 salon+0.014
Neg logit contributions
age-0.012
 back-0.010
 welcome-0.010
>-0.010
 home-0.009
Token and
Token ID290
Feature activation+1.95e-04

Pos logit contributions
yles+0.027
cases+0.026
ls+0.024
yle+0.024
 salon+0.024
Neg logit contributions
age-0.023
 back-0.020
 welcome-0.019
 home-0.019
>-0.019
Token never
Token ID1239
Feature activation-5.06e-02

Pos logit contributions
yles+0.000
cases+0.000
ls+0.000
yle+0.000
 salon+0.000
Neg logit contributions
age-0.000
 back-0.000
 welcome-0.000
 home-0.000
>-0.000
Token but
Token ID475
Feature activation-1.16e-02

Pos logit contributions
yles+0.000
cases+0.000
ls+0.000
yle+0.000
 salon+0.000
Neg logit contributions
age-0.000
 right-0.000
 back-0.000
 welcome-0.000
 home-0.000
Token he
Token ID339
Feature activation+2.49e-02

Pos logit contributions
 right+0.020
 back+0.019
age+0.019
 welcome+0.019
 home+0.018
Neg logit contributions
cases-0.019
yles-0.019
yle-0.017
ls-0.017
outing-0.017
Token also
Token ID635
Feature activation+1.51e-02

Pos logit contributions
yles+0.048
cases+0.046
yle+0.042
ries+0.041
 salon+0.041
Neg logit contributions
age-0.040
 back-0.035
 right-0.033
ney-0.032
man-0.032
Token knew
Token ID2993
Feature activation-6.33e-03

Pos logit contributions
yles+0.032
cases+0.031
ries+0.029
ls+0.028
yle+0.028
Neg logit contributions
age-0.022
 back-0.020
 right-0.020
 welcome-0.019
amed-0.018
Token cater
Token ID20825
Feature activation+1.08e-02

Pos logit contributions
age+0.010
 back+0.010
 right+0.009
 home+0.009
 best+0.009
Neg logit contributions
cases-0.011
yles-0.011
outing-0.010
yle-0.010
 spill-0.010
Tokenpill
Token ID27215
Feature activation+1.19e-01

Pos logit contributions
cases+0.012
yles+0.011
 salon+0.010
yle+0.010
 hairc+0.009
Neg logit contributions
age-0.026
est-0.022
 welcome-0.022
 Follow-0.022
 back-0.022
Tokenars
Token ID945
Feature activation+1.27e-02

Pos logit contributions
yles+0.241
cases+0.238
ls+0.226
yle+0.223
 salon+0.219
Neg logit contributions
age-0.167
 back-0.134
>-0.132
 home-0.130
 welcome-0.130
Token were
Token ID547
Feature activation+3.26e-02

Pos logit contributions
yles+0.029
yle+0.027
cases+0.027
 salon+0.026
ls+0.025
Neg logit contributions
age-0.015
 back-0.015
 right-0.013
 home-0.012
est-0.012
Token not
Token ID407
Feature activation+4.78e-04

Pos logit contributions
yles+0.069
cases+0.068
yle+0.064
 salon+0.063
ls+0.062
Neg logit contributions
age-0.044
 back-0.040
 right-0.038
 welcome-0.038
 home-0.036
Token pets
Token ID17252
Feature activation+1.53e-02

Pos logit contributions
yles+0.001
cases+0.001
yle+0.001
ls+0.001
 salon+0.001
Neg logit contributions
age-0.001
 back-0.001
 welcome-0.001
 home-0.001
 right-0.001
Token.
Token ID13
Feature activation+3.55e-03

Pos logit contributions
yles+0.033
cases+0.032
yle+0.030
 salon+0.030
ls+0.029
Neg logit contributions
age-0.020
 back-0.018
 welcome-0.017
 right-0.017
 home-0.016
Tokenily
Token ID813
Feature activation+1.27e-02

Pos logit contributions
yles+0.031
cases+0.031
yle+0.028
ls+0.028
 salon+0.028
Neg logit contributions
age-0.029
 back-0.025
 home-0.025
>-0.025
 welcome-0.025
Token liked
Token ID8288
Feature activation+7.38e-03

Pos logit contributions
yles+0.020
cases+0.020
ls+0.019
yle+0.018
 salon+0.018
Neg logit contributions
age-0.023
 back-0.019
>-0.019
 welcome-0.019
est-0.018
Token to
Token ID284
Feature activation+1.43e-02

Pos logit contributions
cases+0.017
yles+0.017
yle+0.016
outing+0.016
ls+0.016
Neg logit contributions
age-0.007
 home-0.007
 back-0.007
 welcome-0.006
 right-0.006
Token keep
Token ID1394
Feature activation+1.28e-02

Pos logit contributions
yles+0.029
cases+0.028
ls+0.027
outing+0.026
yle+0.026
Neg logit contributions
age-0.021
 welcome-0.018
 back-0.017
 home-0.017
>-0.017
Token cater
Token ID20825
Feature activation+7.53e-03

Pos logit contributions
cases+0.028
outing+0.025
yles+0.025
 salon+0.025
yle+0.025
Neg logit contributions
 home-0.016
 back-0.016
age-0.016
 right-0.015
 welcome-0.014
Tokenpill
Token ID27215
Feature activation+1.19e-01

Pos logit contributions
cases+0.008
yles+0.008
 salon+0.007
 spill+0.006
yle+0.006
Neg logit contributions
age-0.018
 welcome-0.016
est-0.016
 back-0.015
 Follow-0.015
Tokenars
Token ID945
Feature activation+1.19e-02

Pos logit contributions
ls+0.203
cases+0.203
yles+0.191
 else+0.185
l+0.182
Neg logit contributions
age-0.182
gy-0.167
 back-0.165
 home-0.164
est-0.164
Token in
Token ID287
Feature activation-1.39e-02

Pos logit contributions
cases+0.022
yles+0.022
ls+0.022
yle+0.021
 salon+0.020
Neg logit contributions
age-0.018
 welcome-0.014
 Follow-0.014
>-0.014
est-0.014
Token a
Token ID257
Feature activation+5.97e-04

Pos logit contributions
age+0.022
 welcome+0.021
 back+0.021
 home+0.021
 right+0.020
Neg logit contributions
cases-0.026
yles-0.025
ls-0.024
outing-0.023
yle-0.022
Token jar
Token ID17379
Feature activation+5.81e-03

Pos logit contributions
yles+0.001
cases+0.001
ls+0.001
 salon+0.001
yle+0.001
Neg logit contributions
age-0.001
est-0.001
>-0.001
amed-0.001
 welcome-0.001
Token.
Token ID13
Feature activation+2.81e-03

Pos logit contributions
cases+0.010
yles+0.009
ls+0.009
 salon+0.009
yle+0.009
Neg logit contributions
age-0.010
>-0.008
 welcome-0.008
 Follow-0.008
est-0.008
Token,
Token ID11
Feature activation+2.74e-03

Pos logit contributions
yles+0.009
cases+0.009
yle+0.009
ls+0.009
 salon+0.009
Neg logit contributions
age-0.005
 back-0.004
 welcome-0.004
>-0.004
 home-0.004
Token "
Token ID366
Feature activation+1.54e-02

Pos logit contributions
yles+0.006
cases+0.005
ls+0.005
yle+0.005
 salon+0.005
Neg logit contributions
age-0.004
 back-0.003
>-0.003
 welcome-0.003
 home-0.003
TokenI
Token ID40
Feature activation-1.59e-02

Pos logit contributions
yles+0.024
cases+0.023
yle+0.022
ls+0.022
 salon+0.022
Neg logit contributions
age-0.029
 back-0.025
 welcome-0.024
>-0.024
 home-0.024
Token love
Token ID1842
Feature activation-1.92e-02

Pos logit contributions
age+0.028
 back+0.024
>+0.024
 welcome+0.023
est+0.023
Neg logit contributions
yles-0.026
cases-0.026
ls-0.024
yle-0.024
 salon-0.024
Token cater
Token ID20825
Feature activation+1.11e-02

Pos logit contributions
age+0.032
 back+0.026
>+0.026
 welcome+0.026
 home+0.025
Neg logit contributions
yles-0.035
cases-0.034
ls-0.032
yle-0.032
 salon-0.031
Tokenpill
Token ID27215
Feature activation+1.18e-01

Pos logit contributions
yles+0.009
cases+0.008
yle+0.007
 salon+0.007
ls+0.007
Neg logit contributions
age-0.030
 back-0.027
 welcome-0.026
>-0.026
est-0.026
Tokenars
Token ID945
Feature activation+1.34e-02

Pos logit contributions
ls+0.194
cases+0.194
yles+0.185
 else+0.175
yle+0.174
Neg logit contributions
age-0.190
gy-0.171
 back-0.171
 home-0.170
>-0.169
Token!
Token ID0
Feature activation-1.06e-02

Pos logit contributions
cases+0.027
yles+0.027
ls+0.026
yle+0.025
 salon+0.025
Neg logit contributions
age-0.019
>-0.015
 Follow-0.014
amed-0.014
 back-0.014
Token You
Token ID921
Feature activation-2.28e-02

Pos logit contributions
age+0.020
amed+0.018
 home+0.018
 back+0.017
>+0.017
Neg logit contributions
yles-0.015
cases-0.015
yle-0.013
 salon-0.013
 hairc-0.013
Token can
Token ID460
Feature activation+9.53e-03

Pos logit contributions
age+0.045
>+0.039
 back+0.037
 home+0.037
amed+0.037
Neg logit contributions
cases-0.032
yles-0.032
ls-0.030
yle-0.029
 spill-0.029
Token be
Token ID307
Feature activation+2.77e-02

Pos logit contributions
yles+0.016
cases+0.016
yle+0.015
ls+0.015
 salon+0.015
Neg logit contributions
age-0.016
>-0.013
 back-0.013
est-0.013
amed-0.013
Token put
Token ID1234
Feature activation-7.81e-03

Pos logit contributions
 spill+0.023
 salon+0.022
yles+0.022
yle+0.022
cases+0.022
Neg logit contributions
age-0.024
amed-0.020
 Follow-0.020
>-0.020
est-0.020
Token it
Token ID340
Feature activation+8.90e-03

Pos logit contributions
age+0.017
>+0.014
est+0.014
ney+0.014
 Follow+0.014
Neg logit contributions
yles-0.010
ls-0.009
ries-0.009
yle-0.009
cases-0.009
Token back
Token ID736
Feature activation-4.61e-02

Pos logit contributions
cases+0.017
ls+0.017
 spill+0.016
yles+0.016
outing+0.015
Neg logit contributions
age-0.013
est-0.010
 Follow-0.010
 impact-0.010
 effect-0.010
Token.
Token ID13
Feature activation+3.72e-03

Pos logit contributions
age+0.062
>+0.047
 welcome+0.047
 back+0.047
est+0.046
Neg logit contributions
yles-0.096
cases-0.095
ls-0.090
yle-0.089
 salon-0.088
Token Cater
Token ID40742
Feature activation+2.07e-03

Pos logit contributions
yles+0.007
yle+0.006
cases+0.006
 hairc+0.006
 salon+0.006
Neg logit contributions
age-0.006
amed-0.005
 back-0.005
>-0.005
sy-0.005
Tokenpill
Token ID27215
Feature activation+1.17e-01

Pos logit contributions
yles+0.002
ls+0.001
cases+0.001
yle+0.001
 salon+0.001
Neg logit contributions
age-0.006
 back-0.005
>-0.005
 home-0.005
 welcome-0.005
Tokenars
Token ID945
Feature activation+1.23e-02

Pos logit contributions
ls+0.193
cases+0.187
yles+0.179
 else+0.176
l+0.170
Neg logit contributions
age-0.184
gy-0.178
 home-0.168
est-0.166
wood-0.165
Token need
Token ID761
Feature activation+8.65e-06

Pos logit contributions
yles+0.022
cases+0.021
yle+0.020
 salon+0.019
ries+0.019
Neg logit contributions
age-0.021
 back-0.019
 welcome-0.018
 right-0.018
 home-0.017
Token to
Token ID284
Feature activation+1.32e-02

Feature not active on prev token;
no predictions were affected.
Token eat
Token ID4483
Feature activation+1.16e-02

Pos logit contributions
yles+0.026
cases+0.025
ls+0.024
yle+0.023
 salon+0.023
Neg logit contributions
age-0.020
 back-0.017
 welcome-0.017
>-0.016
 home-0.016
Token leaves
Token ID5667
Feature activation+4.49e-03

Pos logit contributions
yles+0.021
cases+0.021
ls+0.020
yle+0.020
 salon+0.019
Neg logit contributions
age-0.019
>-0.015
 back-0.015
 welcome-0.014
est-0.014
Token Lily
Token ID20037
Feature activation+8.09e-03

Pos logit contributions
age+0.002
 back+0.002
 welcome+0.002
 right+0.002
 home+0.002
Neg logit contributions
cases-0.003
yles-0.003
ls-0.003
outing-0.003
yle-0.003
Token,
Token ID11
Feature activation-1.61e-03

Pos logit contributions
cases+0.014
ls+0.014
yles+0.014
 wra+0.013
yle+0.013
Neg logit contributions
age-0.013
amed-0.011
>-0.011
 Follow-0.010
 welcome-0.010
Token no
Token ID645
Feature activation-1.80e-03

Pos logit contributions
age+0.003
 back+0.002
 welcome+0.002
 right+0.002
 home+0.002
Neg logit contributions
cases-0.003
yles-0.003
ls-0.003
yle-0.003
outing-0.003
Token.
Token ID13
Feature activation+3.03e-03

Pos logit contributions
age+0.003
>+0.002
 back+0.002
amed+0.002
 Follow+0.002
Neg logit contributions
yles-0.004
cases-0.004
ls-0.003
 salon-0.003
yle-0.003
Token Cater
Token ID40742
Feature activation+1.64e-03

Pos logit contributions
yles+0.006
cases+0.006
yle+0.006
ls+0.006
 salon+0.006
Neg logit contributions
age-0.004
amed-0.003
>-0.003
 home-0.003
 back-0.003
Tokenpill
Token ID27215
Feature activation+1.17e-01

Pos logit contributions
yles+0.001
ls+0.001
cases+0.001
yle+0.001
 salon+0.001
Neg logit contributions
age-0.005
 back-0.004
>-0.004
 home-0.004
 welcome-0.004
Tokenars
Token ID945
Feature activation+1.15e-02

Pos logit contributions
yles+0.242
cases+0.239
ls+0.226
yle+0.224
 salon+0.221
Neg logit contributions
age-0.160
 back-0.127
>-0.125
 welcome-0.124
 home-0.122
Token are
Token ID389
Feature activation+2.35e-02

Pos logit contributions
yles+0.022
cases+0.020
yle+0.019
 salon+0.019
ries+0.018
Neg logit contributions
age-0.019
 back-0.017
 welcome-0.017
 right-0.016
est-0.015
Token not
Token ID407
Feature activation-5.29e-04

Pos logit contributions
yles+0.046
cases+0.046
yle+0.043
 salon+0.042
ls+0.042
Neg logit contributions
age-0.035
 back-0.030
 welcome-0.029
 right-0.028
 home-0.028
Token for
Token ID329
Feature activation-2.33e-02

Pos logit contributions
age+0.001
>+0.001
 back+0.001
 welcome+0.001
 Follow+0.001
Neg logit contributions
yles-0.001
cases-0.001
ls-0.001
yle-0.001
 salon-0.001
Token keeping
Token ID5291
Feature activation-2.09e-02

Pos logit contributions
 home+0.030
 welcome+0.029
age+0.029
 back+0.029
 right+0.028
Neg logit contributions
cases-0.050
yles-0.046
 salon-0.043
yle-0.042
ls-0.041
Token was
Token ID373
Feature activation+7.38e-03

Pos logit contributions
yles+0.030
cases+0.030
ls+0.028
yle+0.027
 salon+0.027
Neg logit contributions
age-0.037
 back-0.031
>-0.031
 welcome-0.031
 home-0.031
Token a
Token ID257
Feature activation+4.90e-04

Pos logit contributions
yles+0.014
cases+0.014
ls+0.014
yle+0.013
 salon+0.013
Neg logit contributions
age-0.011
>-0.009
est-0.008
 back-0.008
 welcome-0.008
Token box
Token ID3091
Feature activation-4.67e-02

Pos logit contributions
yles+0.001
cases+0.001
 salon+0.001
ls+0.001
 spill+0.001
Neg logit contributions
age-0.001
>-0.001
est-0.001
amed-0.001
 Follow-0.001
Token of
Token ID286
Feature activation+2.92e-02

Pos logit contributions
age+0.072
>+0.058
 welcome+0.057
 Follow+0.057
 back+0.056
Neg logit contributions
cases-0.088
yles-0.084
ls-0.081
yle-0.081
 salon-0.080
Token ch
Token ID442
Feature activation+9.07e-03

Pos logit contributions
yles+0.048
cases+0.047
yle+0.043
ls+0.043
 salon+0.042
Neg logit contributions
age-0.053
 back-0.046
 welcome-0.046
 home-0.045
>-0.044
Tokenocol
Token ID4668
Feature activation+1.15e-01

Pos logit contributions
yles+0.008
cases+0.008
 salon+0.007
yle+0.006
ls+0.006
Neg logit contributions
age-0.024
 back-0.021
>-0.021
 welcome-0.021
 home-0.021
Tokenates
Token ID689
Feature activation-4.71e-02

Pos logit contributions
yles+0.261
cases+0.253
ls+0.243
ries+0.243
yle+0.241
Neg logit contributions
age-0.135
 back-0.105
 welcome-0.105
>-0.101
est-0.100
Token.
Token ID13
Feature activation+1.31e-03

Pos logit contributions
age+0.063
 back+0.053
 home+0.050
 right+0.049
>+0.049
Neg logit contributions
yles-0.099
cases-0.095
yle-0.091
 salon-0.091
ls-0.090
Token Lily
Token ID20037
Feature activation+1.32e-02

Pos logit contributions
yles+0.002
cases+0.002
yle+0.002
ls+0.002
 salon+0.002
Neg logit contributions
age-0.002
est-0.002
amed-0.002
 home-0.002
>-0.002
Token loved
Token ID6151
Feature activation-1.60e-03

Pos logit contributions
yles+0.022
cases+0.020
yle+0.020
ls+0.020
outing+0.019
Neg logit contributions
age-0.024
 back-0.021
>-0.021
 right-0.020
gy-0.020
Token ch
Token ID442
Feature activation+1.11e-02

Pos logit contributions
 home+0.002
 back+0.002
 right+0.002
 welcome+0.002
 for+0.002
Neg logit contributions
cases-0.003
outing-0.003
ls-0.003
yles-0.003
yle-0.003
Token was
Token ID373
Feature activation+9.88e-03

Pos logit contributions
yles+0.027
cases+0.026
yle+0.024
outing+0.024
 salon+0.024
Neg logit contributions
age-0.021
 back-0.019
 home-0.018
 welcome-0.018
est-0.018
Token a
Token ID257
Feature activation-7.67e-05

Pos logit contributions
yles+0.023
cases+0.023
ls+0.022
yle+0.022
 salon+0.021
Neg logit contributions
age-0.011
 back-0.008
 welcome-0.008
>-0.008
 home-0.008
Token box
Token ID3091
Feature activation-4.34e-02

Pos logit contributions
age+0.000
>+0.000
 back+0.000
 welcome+0.000
est+0.000
Neg logit contributions
yles-0.000
cases-0.000
ls-0.000
 salon-0.000
yle-0.000
Token of
Token ID286
Feature activation+3.02e-02

Pos logit contributions
age+0.068
 back+0.055
>+0.055
 welcome+0.055
 home+0.054
Neg logit contributions
yles-0.081
cases-0.080
ls-0.075
yle-0.075
 salon-0.073
Token ch
Token ID442
Feature activation+1.61e-02

Pos logit contributions
yles+0.051
cases+0.050
ls+0.046
yle+0.046
outing+0.045
Neg logit contributions
age-0.052
 welcome-0.047
 back-0.047
 home-0.046
 right-0.045
Tokenocol
Token ID4668
Feature activation+1.15e-01

Pos logit contributions
cases+0.017
yles+0.017
 salon+0.015
yle+0.013
 wra+0.012
Neg logit contributions
age-0.039
 back-0.035
 home-0.035
 welcome-0.035
est-0.034
Tokenates
Token ID689
Feature activation-4.76e-02

Pos logit contributions
yles+0.259
cases+0.253
ls+0.241
yle+0.240
ries+0.238
Neg logit contributions
age-0.137
 back-0.106
 welcome-0.105
>-0.103
 home-0.101
Token.
Token ID13
Feature activation+2.68e-03

Pos logit contributions
age+0.058
 back+0.050
 home+0.047
 right+0.046
 across+0.044
Neg logit contributions
yles-0.106
cases-0.100
 salon-0.097
yle-0.097
ls-0.096
Token "
Token ID366
Feature activation+1.77e-02

Pos logit contributions
yles+0.006
cases+0.006
yle+0.005
ls+0.005
 salon+0.005
Neg logit contributions
age-0.004
 back-0.003
>-0.003
 welcome-0.003
est-0.003
TokenFor
Token ID1890
Feature activation+1.48e-04

Pos logit contributions
yles+0.028
cases+0.028
yle+0.026
ls+0.026
 salon+0.025
Neg logit contributions
age-0.033
 back-0.028
>-0.027
 welcome-0.027
 home-0.027
Token me
Token ID502
Feature activation-3.35e-02

Pos logit contributions
yles+0.000
cases+0.000
ls+0.000
yle+0.000
 salon+0.000
Neg logit contributions
age-0.000
 back-0.000
 welcome-0.000
>-0.000
 home-0.000
Token and
Token ID290
Feature activation-2.35e-04

Pos logit contributions
age+0.064
 back+0.053
 home+0.051
>+0.050
 welcome+0.050
Neg logit contributions
yles-0.091
cases-0.087
yle-0.083
ls-0.083
 salon-0.081
Token started
Token ID2067
Feature activation+8.13e-03

Pos logit contributions
age+0.000
 back+0.000
 home+0.000
 right+0.000
>+0.000
Neg logit contributions
yles-0.000
cases-0.000
outing-0.000
ls-0.000
yle-0.000
Token eating
Token ID6600
Feature activation+2.27e-03

Pos logit contributions
yles+0.018
cases+0.018
ls+0.018
yle+0.017
outing+0.017
Neg logit contributions
age-0.009
 back-0.009
 home-0.009
 right-0.008
 across-0.008
Token the
Token ID262
Feature activation-1.19e-02

Pos logit contributions
yles+0.005
cases+0.005
yle+0.004
 salon+0.004
outing+0.004
Neg logit contributions
 back-0.003
age-0.003
 right-0.003
 home-0.003
 welcome-0.003
Token ch
Token ID442
Feature activation+1.65e-02

Pos logit contributions
age+0.023
>+0.019
amed+0.019
 Follow+0.019
 welcome+0.018
Neg logit contributions
cases-0.019
yles-0.017
ls-0.017
 wra-0.017
 spill-0.016
Tokenocol
Token ID4668
Feature activation+1.15e-01

Pos logit contributions
yles+0.014
cases+0.013
 salon+0.011
yle+0.010
ls+0.010
Neg logit contributions
age-0.044
est-0.039
 welcome-0.039
 back-0.039
 home-0.038
Tokenates
Token ID689
Feature activation-4.77e-02

Pos logit contributions
yles+0.268
cases+0.263
ls+0.251
yle+0.249
 salon+0.245
Neg logit contributions
age-0.126
 back-0.095
 welcome-0.093
>-0.092
 home-0.090
Token.
Token ID13
Feature activation+2.17e-03

Pos logit contributions
age+0.061
 back+0.053
 home+0.051
 right+0.050
est+0.047
Neg logit contributions
yles-0.103
cases-0.098
yle-0.094
 salon-0.093
ls-0.092
Token The
Token ID383
Feature activation-5.96e-04

Pos logit contributions
yles+0.005
cases+0.005
ls+0.005
yle+0.005
 salon+0.005
Neg logit contributions
age-0.002
 back-0.002
 welcome-0.002
>-0.002
 home-0.002
Token shop
Token ID6128
Feature activation+3.14e-02

Pos logit contributions
age+0.001
 back+0.001
 welcome+0.001
 right+0.001
 home+0.001
Neg logit contributions
yles-0.001
cases-0.001
ls-0.001
yle-0.001
outing-0.001
Tokenkeeper
Token ID13884
Feature activation+1.09e-02

Pos logit contributions
yles+0.062
cases+0.056
ls+0.055
yle+0.055
 spill+0.053
Neg logit contributions
age-0.048
 back-0.042
 home-0.039
man-0.039
 right-0.039
Token not
Token ID407
Feature activation+9.27e-05

Pos logit contributions
age+0.033
 back+0.027
>+0.027
 welcome+0.026
 home+0.026
Neg logit contributions
yles-0.043
cases-0.042
ls-0.040
yle-0.040
 salon-0.039
Token eat
Token ID4483
Feature activation+1.13e-02

Pos logit contributions
yles+0.000
cases+0.000
ls+0.000
yle+0.000
ries+0.000
Neg logit contributions
age-0.000
 back-0.000
 welcome-0.000
 home-0.000
 right-0.000
Token too
Token ID1165
Feature activation+3.18e-02

Pos logit contributions
yles+0.022
cases+0.022
yle+0.020
 salon+0.020
outing+0.020
Neg logit contributions
 back-0.016
 home-0.015
age-0.015
 right-0.015
 welcome-0.015
Token many
Token ID867
Feature activation+5.29e-02

Pos logit contributions
yles+0.084
cases+0.079
yle+0.078
 salon+0.077
outing+0.072
Neg logit contributions
age-0.028
 home-0.026
 welcome-0.024
 back-0.023
 right-0.021
Token ch
Token ID442
Feature activation+1.24e-02

Pos logit contributions
cases+0.110
yles+0.109
outing+0.105
 salon+0.105
yle+0.103
Neg logit contributions
est-0.066
 back-0.064
age-0.063
 welcome-0.063
 home-0.063
Tokenocol
Token ID4668
Feature activation+1.15e-01

Pos logit contributions
cases+0.015
yles+0.014
 salon+0.012
 wra+0.011
yle+0.011
Neg logit contributions
age-0.028
 welcome-0.026
 home-0.026
 back-0.026
 advantage-0.025
Tokenates
Token ID689
Feature activation-4.88e-02

Pos logit contributions
yles+0.264
cases+0.256
ls+0.246
yle+0.245
ries+0.243
Neg logit contributions
age-0.130
 back-0.099
 welcome-0.099
>-0.097
 Follow-0.093
Token and
Token ID290
Feature activation-2.93e-03

Pos logit contributions
age+0.062
 back+0.056
 home+0.052
 right+0.051
 welcome+0.049
Neg logit contributions
yles-0.106
cases-0.101
 salon-0.096
yle-0.096
ls-0.094
Token his
Token ID465
Feature activation+2.64e-02

Pos logit contributions
age+0.005
 back+0.004
 home+0.004
 welcome+0.004
 right+0.004
Neg logit contributions
yles-0.005
cases-0.005
yle-0.005
ls-0.005
outing-0.005
Token friend
Token ID1545
Feature activation-1.85e-02

Pos logit contributions
yles+0.047
cases+0.046
ls+0.043
yle+0.043
 salon+0.042
Neg logit contributions
age-0.044
 back-0.037
 welcome-0.036
>-0.036
 home-0.036
Token learned
Token ID4499
Feature activation+1.74e-02

Pos logit contributions
age+0.034
>+0.028
 back+0.028
 welcome+0.028
 home+0.028
Neg logit contributions
yles-0.030
cases-0.029
ls-0.027
yle-0.027
 salon-0.026
Token a
Token ID257
Feature activation-8.68e-04

Pos logit contributions
age+0.033
 back+0.028
 home+0.026
 right+0.025
 welcome+0.025
Neg logit contributions
yles-0.064
cases-0.062
ls-0.060
yle-0.059
 salon-0.058
Token big
Token ID1263
Feature activation-5.33e-03

Pos logit contributions
age+0.001
 back+0.001
 welcome+0.001
 home+0.001
 right+0.001
Neg logit contributions
yles-0.002
cases-0.002
ls-0.001
yle-0.001
ries-0.001
Token box
Token ID3091
Feature activation-4.68e-02

Pos logit contributions
age+0.009
>+0.007
 back+0.007
 welcome+0.007
 Follow+0.007
Neg logit contributions
yles-0.010
cases-0.010
ls-0.009
 salon-0.009
yle-0.009
Token of
Token ID286
Feature activation+2.93e-02

Pos logit contributions
age+0.065
 back+0.054
 right+0.051
 home+0.051
>+0.050
Neg logit contributions
yles-0.098
cases-0.094
ls-0.090
yle-0.089
 salon-0.087
Token ch
Token ID442
Feature activation+1.11e-02

Pos logit contributions
yles+0.052
cases+0.051
outing+0.046
yle+0.046
ls+0.046
Neg logit contributions
age-0.048
 back-0.044
 welcome-0.044
 home-0.043
 right-0.043
Tokenocol
Token ID4668
Feature activation+1.14e-01

Pos logit contributions
cases+0.013
yles+0.012
 salon+0.011
yle+0.010
 hairc+0.009
Neg logit contributions
age-0.026
 home-0.023
 back-0.023
>-0.023
 welcome-0.023
Tokenates
Token ID689
Feature activation-4.91e-02

Pos logit contributions
yles+0.260
cases+0.252
ls+0.242
ries+0.240
yle+0.240
Neg logit contributions
age-0.133
 welcome-0.103
 back-0.103
>-0.099
est-0.098
Token.
Token ID13
Feature activation+1.92e-03

Pos logit contributions
age+0.060
 back+0.053
 right+0.050
 home+0.050
 across+0.046
Neg logit contributions
yles-0.108
cases-0.101
yle-0.098
 salon-0.098
ls-0.097
Token It
Token ID632
Feature activation+1.96e-02

Pos logit contributions
cases+0.004
yles+0.004
yle+0.004
 salon+0.004
ls+0.003
Neg logit contributions
age-0.003
 back-0.002
>-0.002
 welcome-0.002
est-0.002
Token was
Token ID373
Feature activation+9.01e-03

Pos logit contributions
yles+0.037
cases+0.035
yle+0.034
 salon+0.033
outing+0.033
Neg logit contributions
age-0.032
 back-0.027
 right-0.026
est-0.026
gy-0.025
Token so
Token ID523
Feature activation+2.06e-02

Pos logit contributions
yles+0.019
cases+0.019
ls+0.018
yle+0.018
 salon+0.018
Neg logit contributions
age-0.012
 back-0.009
 welcome-0.009
>-0.009
 home-0.009
Tokenates
Token ID689
Feature activation-4.71e-02

Pos logit contributions
yles+0.261
cases+0.253
ls+0.243
ries+0.243
yle+0.241
Neg logit contributions
age-0.135
 back-0.105
 welcome-0.105
>-0.101
est-0.100
Token.
Token ID13
Feature activation+1.31e-03

Pos logit contributions
age+0.063
 back+0.053
 home+0.050
 right+0.049
>+0.049
Neg logit contributions
yles-0.099
cases-0.095
yle-0.091
 salon-0.091
ls-0.090
Token Lily
Token ID20037
Feature activation+1.32e-02

Pos logit contributions
yles+0.002
cases+0.002
yle+0.002
ls+0.002
 salon+0.002
Neg logit contributions
age-0.002
est-0.002
amed-0.002
 home-0.002
>-0.002
Token loved
Token ID6151
Feature activation-1.60e-03

Pos logit contributions
yles+0.022
cases+0.020
yle+0.020
ls+0.020
outing+0.019
Neg logit contributions
age-0.024
 back-0.021
>-0.021
 right-0.020
gy-0.020
Token ch
Token ID442
Feature activation+1.11e-02

Pos logit contributions
 home+0.002
 back+0.002
 right+0.002
 welcome+0.002
 for+0.002
Neg logit contributions
cases-0.003
outing-0.003
ls-0.003
yles-0.003
yle-0.003
Tokenocol
Token ID4668
Feature activation+1.14e-01

Pos logit contributions
cases+0.015
 salon+0.013
yles+0.013
 rang+0.011
 frowned+0.011
Neg logit contributions
age-0.024
 welcome-0.022
 home-0.021
est-0.021
 back-0.021
Tokenates
Token ID689
Feature activation-4.66e-02

Pos logit contributions
yles+0.269
cases+0.263
ls+0.251
yle+0.250
 salon+0.245
Neg logit contributions
age-0.125
 back-0.093
 welcome-0.091
>-0.090
 home-0.088
Token.
Token ID13
Feature activation+2.30e-03

Pos logit contributions
age+0.063
 back+0.053
 home+0.050
 welcome+0.049
 right+0.049
Neg logit contributions
yles-0.098
cases-0.095
yle-0.090
 salon-0.090
ls-0.089
Token She
Token ID1375
Feature activation-3.77e-03

Pos logit contributions
yles+0.004
cases+0.003
ls+0.003
yle+0.003
 salon+0.003
Neg logit contributions
age-0.005
amed-0.004
est-0.004
>-0.004
 back-0.004
Token thanked
Token ID26280
Feature activation+3.22e-02

Pos logit contributions
age+0.006
 back+0.006
>+0.005
 right+0.005
sy+0.005
Neg logit contributions
yles-0.007
cases-0.006
yle-0.006
outing-0.006
ries-0.006
Token Ben
Token ID3932
Feature activation+3.44e-02

Pos logit contributions
cases+0.059
yles+0.055
ls+0.052
outing+0.051
 salon+0.050
Neg logit contributions
age-0.050
 back-0.049
 welcome-0.047
 home-0.047
 right-0.046
Token and
Token ID290
Feature activation-3.84e-03

Pos logit contributions
age+0.010
 back+0.008
 welcome+0.008
 home+0.008
>+0.008
Neg logit contributions
yles-0.014
cases-0.013
yle-0.013
ls-0.012
 salon-0.012
Token lady
Token ID10846
Feature activation+7.94e-03

Pos logit contributions
age+0.006
 back+0.005
 welcome+0.005
 right+0.005
 home+0.005
Neg logit contributions
cases-0.007
yles-0.007
yle-0.007
outing-0.006
 salon-0.006
Tokenbugs
Token ID32965
Feature activation-5.45e-03

Pos logit contributions
cases+0.014
yles+0.014
yle+0.013
 salon+0.012
ls+0.012
Neg logit contributions
age-0.014
 back-0.012
 welcome-0.011
 home-0.011
>-0.011
Token and
Token ID290
Feature activation-4.21e-03

Pos logit contributions
age+0.008
 back+0.007
 welcome+0.006
 home+0.006
>+0.006
Neg logit contributions
yles-0.011
cases-0.011
yle-0.010
ls-0.010
 salon-0.010
Token cater
Token ID20825
Feature activation+5.36e-03

Pos logit contributions
age+0.007
 back+0.006
 welcome+0.006
 right+0.006
 home+0.006
Neg logit contributions
cases-0.008
yles-0.008
yle-0.007
outing-0.007
 salon-0.007
Tokenpill
Token ID27215
Feature activation+1.14e-01

Pos logit contributions
cases+0.005
yles+0.005
yle+0.004
 salon+0.004
outing+0.004
Neg logit contributions
age-0.013
 back-0.012
>-0.012
 welcome-0.012
 home-0.012
Tokenars
Token ID945
Feature activation+8.25e-03

Pos logit contributions
yles+0.211
cases+0.210
ls+0.207
yle+0.194
 spill+0.192
Neg logit contributions
age-0.168
 back-0.142
>-0.141
 home-0.139
est-0.138
Token."
Token ID526
Feature activation+1.33e-02

Pos logit contributions
yles+0.017
cases+0.017
yle+0.016
 salon+0.016
ls+0.016
Neg logit contributions
age-0.011
 back-0.010
 welcome-0.009
 home-0.009
>-0.009
Token
Token ID198
Feature activation-1.12e-02

Pos logit contributions
yles+0.028
cases+0.027
ls+0.026
yle+0.026
 salon+0.025
Neg logit contributions
age-0.018
 back-0.014
>-0.014
 welcome-0.014
 home-0.014
Token
Token ID198
Feature activation-1.12e-02

Pos logit contributions
age+0.017
est+0.013
 back+0.013
 welcome+0.013
 home+0.013
Neg logit contributions
yles-0.022
cases-0.021
ls-0.021
yle-0.020
 salon-0.020
TokenTim
Token ID14967
Feature activation-8.57e-03

Pos logit contributions
age+0.019
 back+0.016
>+0.016
 welcome+0.016
 home+0.015
Neg logit contributions
yles-0.020
cases-0.019
ls-0.018
yle-0.018
 salon-0.018
Token the
Token ID262
Feature activation-1.30e-02

Pos logit contributions
yles+0.064
cases+0.064
ls+0.060
outing+0.059
yle+0.058
Neg logit contributions
age-0.058
 home-0.056
 back-0.056
 right-0.053
 welcome-0.051
Token cars
Token ID5006
Feature activation-1.40e-02

Pos logit contributions
age+0.022
 back+0.019
 welcome+0.018
>+0.018
 home+0.018
Neg logit contributions
yles-0.023
cases-0.022
ls-0.021
yle-0.021
 salon-0.020
Token stopped
Token ID5025
Feature activation-2.68e-04

Pos logit contributions
age+0.026
 back+0.023
 home+0.022
 right+0.021
 across+0.021
Neg logit contributions
yles-0.025
yle-0.022
ls-0.021
outing-0.021
ries-0.021
Token and
Token ID290
Feature activation-2.89e-03

Pos logit contributions
age+0.000
 back+0.000
>+0.000
 welcome+0.000
 home+0.000
Neg logit contributions
yles-0.001
cases-0.001
ls-0.000
yle-0.000
 salon-0.000
Token the
Token ID262
Feature activation-1.20e-02

Pos logit contributions
age+0.005
 back+0.005
 right+0.004
 home+0.004
 across+0.004
Neg logit contributions
yles-0.005
cases-0.005
yle-0.005
ls-0.005
outing-0.005
Token pedestrian
Token ID22382
Feature activation-1.55e-01

Pos logit contributions
age+0.020
>+0.017
 back+0.017
 welcome+0.017
est+0.016
Neg logit contributions
cases-0.020
yles-0.020
 salon-0.019
ls-0.019
yle-0.019
Token light
Token ID1657
Feature activation+3.41e-02

Pos logit contributions
age+0.234
>+0.194
 Follow+0.186
 welcome+0.184
 home+0.182
Neg logit contributions
cases-0.284
yles-0.274
ls-0.265
yle-0.254
outing-0.253
Token turned
Token ID2900
Feature activation+1.45e-02

Pos logit contributions
yles+0.070
yle+0.064
cases+0.063
 salon+0.062
ls+0.061
Neg logit contributions
age-0.053
 back-0.044
est-0.043
 home-0.041
 right-0.041
Token green
Token ID4077
Feature activation-1.27e-02

Pos logit contributions
yles+0.029
yle+0.027
 salon+0.026
cases+0.026
ls+0.025
Neg logit contributions
age-0.022
 back-0.020
 right-0.018
 home-0.018
 across-0.018
Token.
Token ID13
Feature activation+1.90e-03

Pos logit contributions
age+0.022
est+0.019
 back+0.018
 home+0.018
 across+0.017
Neg logit contributions
yles-0.023
yle-0.022
cases-0.022
ls-0.021
 salon-0.021
Token Then
Token ID3244
Feature activation+1.50e-02

Pos logit contributions
yles+0.004
cases+0.004
yle+0.003
 salon+0.003
ls+0.003
Neg logit contributions
age-0.003
est-0.002
 back-0.002
 welcome-0.002
>-0.002
Token and
Token ID290
Feature activation-3.82e-03

Pos logit contributions
age+0.013
 back+0.011
 right+0.010
 welcome+0.010
 home+0.010
Neg logit contributions
yles-0.023
yle-0.022
 salon-0.022
cases-0.021
ls-0.021
Token books
Token ID3835
Feature activation-4.44e-02

Pos logit contributions
age+0.006
 back+0.006
 welcome+0.005
 right+0.005
 across+0.005
Neg logit contributions
yle-0.007
yles-0.007
outing-0.007
cases-0.007
ls-0.006
Token.
Token ID13
Feature activation+2.27e-03

Pos logit contributions
age+0.043
 back+0.042
 right+0.039
 across+0.034
 home+0.033
Neg logit contributions
yles-0.111
yle-0.107
outing-0.102
 tant-0.101
 hairc-0.101
Token They
Token ID1119
Feature activation-7.26e-03

Pos logit contributions
yles+0.004
cases+0.004
yle+0.004
ls+0.004
 salon+0.004
Neg logit contributions
age-0.003
 back-0.003
 welcome-0.003
>-0.003
 home-0.003
Token see
Token ID766
Feature activation-1.20e-02

Pos logit contributions
 welcome+0.009
 back+0.008
 right+0.008
age+0.008
 home+0.007
Neg logit contributions
yles-0.016
cases-0.015
ls-0.015
yle-0.015
ries-0.015
Token alt
Token ID5988
Feature activation-1.51e-01

Pos logit contributions
 back+0.017
age+0.016
 right+0.016
 home+0.015
 welcome+0.015
Neg logit contributions
yles-0.023
cases-0.022
yle-0.021
outing-0.021
 salon-0.020
Tokenars
Token ID945
Feature activation+7.66e-03

Pos logit contributions
age+0.202
 back+0.157
>+0.153
 welcome+0.149
 home+0.146
Neg logit contributions
yles-0.328
cases-0.317
yle-0.306
ls-0.302
 salon-0.297
Token and
Token ID290
Feature activation-4.71e-03

Pos logit contributions
yles+0.018
yle+0.017
cases+0.017
 salon+0.016
ls+0.016
Neg logit contributions
age-0.008
 back-0.008
 right-0.007
 home-0.007
 across-0.007
Token offerings
Token ID18369
Feature activation-2.20e-02

Pos logit contributions
age+0.007
 back+0.007
 welcome+0.007
 right+0.007
 across+0.007
Neg logit contributions
yle-0.008
yles-0.008
cases-0.008
outing-0.008
ls-0.008
Token and
Token ID290
Feature activation-5.00e-03

Pos logit contributions
 back+0.019
 right+0.016
age+0.016
 home+0.016
 welcome+0.015
Neg logit contributions
yles-0.058
yle-0.055
 salon-0.054
cases-0.052
 tant-0.052
Token priests
Token ID17345
Feature activation-3.88e-02

Pos logit contributions
age+0.008
 back+0.008
 welcome+0.007
 right+0.007
 home+0.007
Neg logit contributions
yles-0.009
yle-0.009
cases-0.009
outing-0.009
ls-0.009
Token so
Token ID523
Feature activation+1.79e-02

Pos logit contributions
yles+0.060
cases+0.059
yle+0.055
ls+0.055
 salon+0.054
Neg logit contributions
age-0.042
 back-0.035
 welcome-0.034
>-0.034
 home-0.033
Token beautiful
Token ID4950
Feature activation-8.19e-03

Pos logit contributions
yles+0.039
 salon+0.037
yle+0.036
cases+0.036
ries+0.033
Neg logit contributions
age-0.023
 welcome-0.022
 back-0.021
 right-0.021
 home-0.021
Token inside
Token ID2641
Feature activation+1.47e-02

Pos logit contributions
age+0.011
 back+0.009
 home+0.008
 right+0.008
est+0.008
Neg logit contributions
yles-0.017
cases-0.017
yle-0.016
 salon-0.016
ls-0.016
Token and
Token ID290
Feature activation-3.12e-03

Pos logit contributions
yles+0.028
cases+0.027
yle+0.026
 salon+0.026
ls+0.025
Neg logit contributions
age-0.022
 back-0.021
 right-0.020
 welcome-0.020
 home-0.020
Token the
Token ID262
Feature activation-1.78e-02

Pos logit contributions
age+0.004
 back+0.004
 right+0.004
 welcome+0.004
 home+0.004
Neg logit contributions
yles-0.006
cases-0.006
yle-0.006
 salon-0.006
ls-0.006
Token fringe
Token ID26821
Feature activation-1.34e-01

Pos logit contributions
age+0.029
 back+0.027
 home+0.026
 welcome+0.026
 right+0.026
Neg logit contributions
yles-0.032
cases-0.031
 salon-0.028
yle-0.028
ls-0.028
Token from
Token ID422
Feature activation-9.66e-03

Pos logit contributions
age+0.187
 back+0.161
 home+0.149
 right+0.148
est+0.148
Neg logit contributions
yles-0.278
cases-0.267
yle-0.255
 salon-0.253
ls-0.249
Token the
Token ID262
Feature activation-1.81e-02

Pos logit contributions
age+0.013
 back+0.013
 home+0.013
>+0.012
 welcome+0.012
Neg logit contributions
cases-0.019
yles-0.018
 salon-0.017
ls-0.017
yle-0.017
Token cut
Token ID2005
Feature activation+1.26e-03

Pos logit contributions
age+0.031
 back+0.027
>+0.026
 welcome+0.026
 home+0.026
Neg logit contributions
yles-0.031
cases-0.031
ls-0.028
yle-0.028
 salon-0.028
Token shells
Token ID19679
Feature activation+4.52e-03

Pos logit contributions
yles+0.002
cases+0.002
 salon+0.002
yle+0.002
ls+0.002
Neg logit contributions
age-0.002
 back-0.002
est-0.002
 home-0.002
 right-0.002
Token made
Token ID925
Feature activation-4.03e-02

Pos logit contributions
yles+0.010
cases+0.009
yle+0.009
 salon+0.009
 hairc+0.009
Neg logit contributions
age-0.006
 back-0.005
 right-0.005
 home-0.005
 across-0.005
Token yellow
Token ID7872
Feature activation-2.54e-02

Pos logit contributions
age+0.021
 back+0.018
 welcome+0.018
>+0.018
 home+0.017
Neg logit contributions
yles-0.022
cases-0.021
yle-0.020
ls-0.020
 salon-0.020
Token and
Token ID290
Feature activation-1.08e-02

Pos logit contributions
age+0.038
>+0.031
 back+0.030
 welcome+0.030
 Follow+0.030
Neg logit contributions
yles-0.048
cases-0.048
ls-0.045
yle-0.045
 salon-0.044
Token green
Token ID4077
Feature activation-1.72e-02

Pos logit contributions
age+0.019
 back+0.016
 welcome+0.016
>+0.015
 home+0.015
Neg logit contributions
yles-0.019
cases-0.018
yle-0.017
ls-0.017
 salon-0.017
Token until
Token ID1566
Feature activation+2.68e-02

Pos logit contributions
age+0.027
 back+0.022
>+0.022
 welcome+0.022
 home+0.021
Neg logit contributions
yles-0.032
cases-0.032
ls-0.030
yle-0.030
 salon-0.029
Token the
Token ID262
Feature activation-1.39e-02

Pos logit contributions
cases+0.045
yles+0.041
outing+0.041
yle+0.040
ls+0.038
Neg logit contributions
 back-0.046
 home-0.045
age-0.043
 right-0.042
 across-0.042
Token formerly
Token ID15734
Feature activation-1.31e-01

Pos logit contributions
age+0.023
>+0.019
 Follow+0.018
 back+0.018
 welcome+0.018
Neg logit contributions
yles-0.024
cases-0.024
ls-0.022
yle-0.022
 salon-0.022
Token color
Token ID3124
Feature activation+2.48e-02

Pos logit contributions
age+0.245
 Follow+0.214
>+0.212
amed+0.198
gy+0.197
Neg logit contributions
cases-0.173
 spill-0.167
yles-0.166
 hairc-0.164
 salon-0.163
Tokenless
Token ID1203
Feature activation+1.04e-02

Pos logit contributions
yles+0.048
cases+0.043
yle+0.042
 salon+0.041
ls+0.041
Neg logit contributions
age-0.040
 back-0.036
 home-0.033
est-0.033
 right-0.032
Token walls
Token ID7714
Feature activation-6.08e-03

Pos logit contributions
yles+0.020
cases+0.020
ls+0.018
yle+0.018
 salon+0.018
Neg logit contributions
age-0.016
 back-0.013
>-0.013
 welcome-0.013
 home-0.012
Token seemed
Token ID3947
Feature activation-1.02e-02

Pos logit contributions
age+0.008
 back+0.007
 home+0.007
 right+0.006
est+0.006
Neg logit contributions
yles-0.013
cases-0.012
yle-0.012
ls-0.012
 salon-0.012
Token to
Token ID284
Feature activation+1.06e-02

Pos logit contributions
age+0.014
 back+0.012
 welcome+0.012
 home+0.011
>+0.011
Neg logit contributions
yles-0.021
cases-0.020
yle-0.019
ls-0.019
 salon-0.019
Token replied
Token ID8712
Feature activation-9.48e-03

Pos logit contributions
age+0.018
 back+0.016
 right+0.015
>+0.015
 home+0.015
Neg logit contributions
yles-0.016
cases-0.014
yle-0.014
ls-0.014
ries-0.013
Token with
Token ID351
Feature activation+4.25e-03

Pos logit contributions
age+0.016
ney+0.013
>+0.013
 Follow+0.012
 welcome+0.012
Neg logit contributions
cases-0.017
ls-0.016
yles-0.016
yle-0.015
outing-0.015
Token a
Token ID257
Feature activation-3.06e-03

Pos logit contributions
yles+0.008
cases+0.008
ls+0.008
yle+0.008
 spill+0.008
Neg logit contributions
age-0.006
ney-0.005
gy-0.005
>-0.005
 Follow-0.004
Token gentle
Token ID10296
Feature activation+1.47e-02

Pos logit contributions
age+0.006
>+0.005
 back+0.005
 welcome+0.005
 home+0.005
Neg logit contributions
yles-0.005
cases-0.005
ls-0.004
yle-0.004
 salon-0.004
Token '
Token ID705
Feature activation+2.30e-02

Pos logit contributions
cases+0.018
yle+0.018
yles+0.018
 spill+0.017
 hairc+0.017
Neg logit contributions
age-0.028
>-0.026
 Follow-0.025
 back-0.024
amed-0.024
Tokenyip
Token ID39666
Feature activation-1.30e-01

Pos logit contributions
cases+0.045
yles+0.042
ls+0.041
 salon+0.040
outing+0.039
Neg logit contributions
age-0.037
 back-0.029
 welcome-0.029
est-0.028
 Follow-0.028
Token'.
Token ID4458
Feature activation-2.97e-02

Pos logit contributions
age+0.159
>+0.125
 back+0.123
 home+0.119
 Follow+0.118
Neg logit contributions
cases-0.273
yles-0.272
yle-0.263
ls-0.261
 salon-0.255
Token
Token ID198
Feature activation-1.33e-02

Pos logit contributions
age+0.045
est+0.038
>+0.037
amed+0.037
 back+0.036
Neg logit contributions
cases-0.055
yles-0.054
ls-0.051
yle-0.051
 salon-0.050
Token
Token ID198
Feature activation-1.34e-02

Pos logit contributions
age+0.022
est+0.021
 effect+0.019
 welcome+0.019
 advantage+0.019
Neg logit contributions
yles-0.021
cases-0.021
ls-0.021
yle-0.019
l-0.019
TokenThe
Token ID464
Feature activation+3.83e-03

Pos logit contributions
age+0.023
est+0.021
 welcome+0.021
 home+0.021
 back+0.020
Neg logit contributions
cases-0.021
yles-0.020
 salon-0.020
ls-0.019
yle-0.019
Token little
Token ID1310
Feature activation+4.90e-03

Pos logit contributions
yles+0.007
cases+0.006
ls+0.006
yle+0.006
 salon+0.006
Neg logit contributions
age-0.007
 back-0.006
 welcome-0.006
>-0.006
 home-0.006
Token show
Token ID905
Feature activation+1.92e-02

Pos logit contributions
age+0.048
 back+0.041
 welcome+0.040
>+0.039
 home+0.039
Neg logit contributions
yles-0.045
cases-0.044
ls-0.041
yle-0.041
 salon-0.039
Token.
Token ID13
Feature activation-3.37e-03

Pos logit contributions
yles+0.036
cases+0.036
yle+0.034
ls+0.034
 salon+0.034
Neg logit contributions
age-0.029
 back-0.024
>-0.024
 welcome-0.023
 home-0.023
Token All
Token ID1439
Feature activation+3.37e-02

Pos logit contributions
age+0.004
 welcome+0.003
>+0.003
 back+0.003
 home+0.003
Neg logit contributions
yles-0.008
cases-0.007
ls-0.007
yle-0.007
 salon-0.007
Token the
Token ID262
Feature activation-1.15e-02

Pos logit contributions
cases+0.069
yles+0.068
outing+0.061
 salon+0.061
ls+0.061
Neg logit contributions
age-0.048
 welcome-0.043
 back-0.043
est-0.042
 home-0.041
Token beautiful
Token ID4950
Feature activation-9.65e-03

Pos logit contributions
 welcome+0.016
 best+0.015
 home+0.015
 right+0.014
 back+0.014
Neg logit contributions
cases-0.026
yles-0.024
outing-0.023
ls-0.023
ries-0.022
Token clot
Token ID5148
Feature activation-1.27e-01

Pos logit contributions
est+0.015
age+0.015
 home+0.015
 back+0.014
 welcome+0.014
Neg logit contributions
cases-0.018
yles-0.017
outing-0.017
ls-0.015
eling-0.015
Tokenhe
Token ID258
Feature activation+6.13e-02

Pos logit contributions
age+0.268
 back+0.230
>+0.228
 welcome+0.228
 home+0.225
Neg logit contributions
yles-0.172
cases-0.168
ls-0.155
yle-0.152
 salon-0.148
Tokenes
Token ID274
Feature activation+1.87e-02

Pos logit contributions
outing+0.116
cases+0.111
yles+0.111
iqu+0.108
ucker+0.101
Neg logit contributions
age-0.088
 day-0.078
 home-0.078
man-0.077
est-0.076
Token were
Token ID547
Feature activation+2.78e-02

Pos logit contributions
yles+0.034
cases+0.033
 salon+0.031
outing+0.031
yle+0.030
Neg logit contributions
age-0.030
 back-0.027
est-0.026
 home-0.026
 welcome-0.025
Token marching
Token ID26409
Feature activation-7.82e-03

Pos logit contributions
cases+0.060
yles+0.060
ls+0.057
yle+0.055
 salon+0.055
Neg logit contributions
age-0.036
 back-0.031
 home-0.030
 welcome-0.029
 right-0.029
Token down
Token ID866
Feature activation-1.50e-02

Pos logit contributions
age+0.012
>+0.009
 welcome+0.009
 Follow+0.009
ney+0.009
Neg logit contributions
cases-0.015
yles-0.015
yle-0.014
ls-0.014
 salon-0.013
TokenYou
Token ID1639
Feature activation-1.59e-02

Pos logit contributions
cases+0.034
yles+0.034
ls+0.032
ries+0.030
 salon+0.030
Neg logit contributions
age-0.033
 right-0.027
 back-0.027
>-0.026
 Follow-0.026
Token are
Token ID389
Feature activation+2.48e-02

Pos logit contributions
age+0.027
 welcome+0.023
 back+0.023
 right+0.022
 home+0.021
Neg logit contributions
yles-0.030
cases-0.028
 salon-0.027
yle-0.026
ls-0.026
Token under
Token ID739
Feature activation-1.61e-02

Pos logit contributions
yles+0.042
cases+0.041
ls+0.040
 wra+0.040
yle+0.040
Neg logit contributions
age-0.043
est-0.035
ney-0.033
gy-0.032
>-0.032
Token arrest
Token ID3251
Feature activation+8.39e-02

Pos logit contributions
age+0.033
 back+0.030
 home+0.030
>+0.030
 welcome+0.029
Neg logit contributions
yles-0.022
cases-0.022
outing-0.020
yle-0.020
ls-0.019
Token for
Token ID329
Feature activation-1.80e-02

Pos logit contributions
cases+0.143
ls+0.143
 spill+0.133
yles+0.132
 hairc+0.131
Neg logit contributions
age-0.132
ney-0.111
amed-0.110
 Follow-0.109
sy-0.108
Token harassing
Token ID34110
Feature activation-1.23e-01

Pos logit contributions
age+0.024
 back+0.019
>+0.018
 welcome+0.018
est+0.018
Neg logit contributions
yles-0.038
cases-0.038
ls-0.036
yle-0.036
 salon-0.035
Token and
Token ID290
Feature activation-3.45e-04

Pos logit contributions
age+0.194
>+0.157
 back+0.154
 Follow+0.152
 welcome+0.151
Neg logit contributions
cases-0.218
yles-0.214
ls-0.204
yle-0.202
 hairc-0.199
Token stealing
Token ID16568
Feature activation-1.18e-02

Pos logit contributions
age+0.001
 back+0.001
 welcome+0.001
>+0.001
 home+0.000
Neg logit contributions
yles-0.001
cases-0.001
ls-0.001
yle-0.001
 salon-0.001
Token from
Token ID422
Feature activation-1.14e-03

Pos logit contributions
age+0.016
 back+0.014
 welcome+0.013
>+0.013
 home+0.013
Neg logit contributions
yles-0.024
cases-0.024
yle-0.023
ls-0.022
 salon-0.022
Token these
Token ID777
Feature activation+4.27e-03

Pos logit contributions
age+0.002
 back+0.001
 home+0.001
 welcome+0.001
>+0.001
Neg logit contributions
cases-0.002
yles-0.002
ls-0.002
yle-0.002
 salon-0.002
Token children
Token ID1751
Feature activation+4.25e-03

Pos logit contributions
yles+0.008
cases+0.007
yle+0.007
outing+0.007
ls+0.006
Neg logit contributions
age-0.007
 back-0.007
 home-0.006
 welcome-0.006
 right-0.006
Token stop
Token ID2245
Feature activation+1.38e-02

Pos logit contributions
yles+0.025
cases+0.023
yle+0.023
ls+0.022
ries+0.022
Neg logit contributions
 home-0.013
 back-0.012
age-0.012
 right-0.011
>-0.010
Token.
Token ID13
Feature activation-1.50e-04

Pos logit contributions
yles+0.031
cases+0.030
yle+0.028
ls+0.028
 salon+0.028
Neg logit contributions
age-0.016
 back-0.015
 right-0.014
 home-0.014
 across-0.013
Token Anna
Token ID11735
Feature activation-1.97e-02

Pos logit contributions
age+0.000
 back+0.000
>+0.000
 welcome+0.000
 home+0.000
Neg logit contributions
yles-0.000
cases-0.000
ls-0.000
yle-0.000
 salon-0.000
Token and
Token ID290
Feature activation-5.59e-03

Pos logit contributions
age+0.020
 back+0.017
 right+0.015
 across+0.014
 home+0.014
Neg logit contributions
yles-0.049
ries-0.046
cases-0.046
outing-0.045
yle-0.045
Token Ben
Token ID3932
Feature activation+2.21e-02

Pos logit contributions
 back+0.009
 right+0.008
age+0.008
 home+0.008
 welcome+0.008
Neg logit contributions
cases-0.010
yles-0.010
outing-0.009
yle-0.009
ls-0.009
Token unb
Token ID22619
Feature activation-1.21e-01

Pos logit contributions
yles+0.040
ries+0.038
ocol+0.037
ls+0.036
 salon+0.035
Neg logit contributions
 back-0.038
age-0.035
 right-0.035
 across-0.032
 home-0.031
Tokenuck
Token ID1347
Feature activation-5.75e-02

Pos logit contributions
age+0.196
 welcome+0.162
 back+0.161
>+0.161
 home+0.157
Neg logit contributions
yles-0.219
cases-0.217
ls-0.202
yle-0.202
 salon-0.197
Tokenled
Token ID992
Feature activation-9.61e-03

Pos logit contributions
age+0.092
 back+0.090
 right+0.081
amed+0.081
est+0.080
Neg logit contributions
yles-0.105
 salon-0.097
cases-0.096
yle-0.095
outing-0.093
Token and
Token ID290
Feature activation-6.19e-03

Pos logit contributions
age+0.015
>+0.012
 welcome+0.012
 Follow+0.011
amed+0.011
Neg logit contributions
yles-0.018
cases-0.018
ls-0.017
yle-0.017
 hairc-0.016
Token got
Token ID1392
Feature activation-6.06e-03

Pos logit contributions
 back+0.011
 right+0.010
age+0.009
 welcome+0.009
 home+0.009
Neg logit contributions
yles-0.011
cases-0.011
outing-0.010
ls-0.010
ge-0.010
Token off
Token ID572
Feature activation+1.33e-03

Pos logit contributions
age+0.006
 back+0.005
>+0.004
 welcome+0.004
 home+0.004
Neg logit contributions
yles-0.015
cases-0.014
ls-0.014
yle-0.014
 salon-0.014
Token -
Token ID532
Feature activation-1.36e-02

Pos logit contributions
age+0.016
 back+0.013
>+0.012
 welcome+0.012
 home+0.012
Neg logit contributions
yles-0.022
cases-0.022
ls-0.021
yle-0.021
 salon-0.020
Token a
Token ID257
Feature activation-1.80e-03

Pos logit contributions
age+0.021
 back+0.017
 welcome+0.017
 home+0.016
 right+0.016
Neg logit contributions
yles-0.026
cases-0.026
ls-0.024
yle-0.024
 salon-0.024
Token shirt
Token ID10147
Feature activation-2.65e-02

Pos logit contributions
age+0.003
>+0.003
 back+0.003
 welcome+0.003
est+0.003
Neg logit contributions
yles-0.003
cases-0.003
 salon-0.003
ls-0.003
yle-0.003
Token and
Token ID290
Feature activation-3.99e-03

Pos logit contributions
age+0.040
>+0.032
 back+0.032
 welcome+0.032
 home+0.031
Neg logit contributions
cases-0.050
yles-0.050
ls-0.047
yle-0.047
 salon-0.046
Token pants
Token ID12581
Feature activation-1.80e-02

Pos logit contributions
age+0.006
 back+0.005
 welcome+0.004
 home+0.004
 right+0.004
Neg logit contributions
yles-0.008
cases-0.008
ls-0.008
outing-0.008
yle-0.008
Token combo
Token ID14831
Feature activation-1.19e-01

Pos logit contributions
age+0.025
 back+0.021
 right+0.019
 home+0.019
est+0.019
Neg logit contributions
yles-0.039
cases-0.036
yle-0.035
ls-0.035
outing-0.035
Token -
Token ID532
Feature activation-1.13e-02

Pos logit contributions
age+0.179
>+0.145
 welcome+0.143
 back+0.142
 Follow+0.138
Neg logit contributions
cases-0.226
yles-0.225
yle-0.212
ls-0.209
 salon-0.206
Token and
Token ID290
Feature activation-2.62e-03

Pos logit contributions
age+0.020
 back+0.018
 right+0.018
 welcome+0.017
 home+0.017
Neg logit contributions
yles-0.019
cases-0.017
ls-0.017
yle-0.017
ries-0.016
Token brushed
Token ID31782
Feature activation+2.04e-03

Pos logit contributions
age+0.005
 back+0.004
 welcome+0.004
 home+0.004
>+0.004
Neg logit contributions
yles-0.004
cases-0.004
yle-0.004
ls-0.004
 salon-0.004
Token his
Token ID465
Feature activation+2.54e-02

Pos logit contributions
yles+0.004
cases+0.004
yle+0.004
ls+0.004
 salon+0.004
Neg logit contributions
age-0.003
 back-0.002
est-0.002
 home-0.002
 right-0.002
Token hair
Token ID4190
Feature activation+7.85e-03

Pos logit contributions
yles+0.042
cases+0.040
ls+0.038
yle+0.037
 salon+0.037
Neg logit contributions
age-0.046
 back-0.041
 home-0.039
>-0.038
 welcome-0.038
Token her
Token ID607
Feature activation-6.11e-03

Pos logit contributions
yles+0.010
cases+0.010
yle+0.009
 salon+0.009
outing+0.009
Neg logit contributions
age-0.014
 back-0.014
 home-0.013
 right-0.013
 welcome-0.013
Token the
Token ID262
Feature activation-1.50e-02

Pos logit contributions
age+0.009
 back+0.008
 home+0.008
 right+0.008
>+0.008
Neg logit contributions
yles-0.012
cases-0.012
yle-0.011
ls-0.011
 salon-0.010
Token pan
Token ID3425
Feature activation-1.83e-03

Pos logit contributions
age+0.028
 back+0.025
 welcome+0.024
 right+0.024
 home+0.024
Neg logit contributions
yles-0.024
cases-0.023
yle-0.022
ls-0.021
 salon-0.020
Token with
Token ID351
Feature activation+4.63e-03

Pos logit contributions
age+0.002
 back+0.002
 right+0.002
 home+0.002
est+0.002
Neg logit contributions
yles-0.004
cases-0.004
yle-0.004
 salon-0.004
ls-0.004
Token the
Token ID262
Feature activation-1.53e-02

Pos logit contributions
cases+0.010
yles+0.010
yle+0.009
ls+0.009
 salon+0.009
Neg logit contributions
 back-0.006
age-0.006
 home-0.006
 right-0.005
 welcome-0.005
Token unb
Token ID22619
Feature activation-1.17e-01

Pos logit contributions
age+0.029
 back+0.025
 welcome+0.025
>+0.025
 home+0.024
Neg logit contributions
yles-0.024
cases-0.023
ls-0.021
yle-0.021
 salon-0.021
Tokenaked
Token ID4335
Feature activation-5.31e-02

Pos logit contributions
age+0.206
 back+0.172
>+0.168
 welcome+0.167
 home+0.166
Neg logit contributions
yles-0.202
cases-0.198
ls-0.184
yle-0.183
 salon-0.180
Token cake
Token ID12187
Feature activation-6.01e-02

Pos logit contributions
age+0.073
 back+0.064
 right+0.060
 home+0.059
est+0.058
Neg logit contributions
yles-0.114
cases-0.106
yle-0.103
ls-0.101
 salon-0.100
Token.
Token ID13
Feature activation+1.67e-03

Pos logit contributions
 back+0.064
age+0.064
 right+0.056
 home+0.054
 across+0.048
Neg logit contributions
yles-0.152
yle-0.139
eling-0.133
cases-0.132
Get-0.129
Token Mom
Token ID11254
Feature activation+2.21e-03

Pos logit contributions
yles+0.004
cases+0.004
yle+0.003
eling+0.003
ls+0.003
Neg logit contributions
age-0.002
 back-0.002
 welcome-0.002
 right-0.002
 home-0.002
Token was
Token ID373
Feature activation+6.58e-03

Pos logit contributions
yles+0.004
yle+0.004
cases+0.004
 hairc+0.004
ls+0.004
Neg logit contributions
age-0.003
 back-0.003
 right-0.003
man-0.003
amed-0.003
Token
Token ID198
Feature activation-1.29e-02

Pos logit contributions
yles+0.005
cases+0.005
ls+0.004
yle+0.004
 salon+0.004
Neg logit contributions
age-0.003
 back-0.003
>-0.003
 welcome-0.003
 home-0.003
Token
Token ID198
Feature activation-1.30e-02

Pos logit contributions
age+0.017
est+0.013
 back+0.013
 welcome+0.013
>+0.013
Neg logit contributions
yles-0.027
cases-0.026
ls-0.025
yle-0.025
 salon-0.025
TokenSo
Token ID2396
Feature activation-4.12e-02

Pos logit contributions
age+0.022
>+0.018
 back+0.018
 welcome+0.018
 right+0.018
Neg logit contributions
yles-0.023
cases-0.023
ls-0.021
yle-0.021
ries-0.021
Token,
Token ID11
Feature activation-4.59e-03

Pos logit contributions
age+0.051
 back+0.046
 home+0.045
 welcome+0.045
 right+0.044
Neg logit contributions
yles-0.089
cases-0.086
yle-0.080
outing-0.079
 salon-0.079
Token with
Token ID351
Feature activation+2.74e-03

Pos logit contributions
 right+0.007
 back+0.007
 home+0.007
 welcome+0.007
age+0.007
Neg logit contributions
yles-0.008
cases-0.008
outing-0.007
ls-0.007
yle-0.007
Token artificial
Token ID11666
Feature activation-1.17e-01

Pos logit contributions
yles+0.006
cases+0.006
yle+0.005
ls+0.005
 salon+0.005
Neg logit contributions
age-0.003
 back-0.003
 home-0.003
 welcome-0.003
>-0.003
Token so
Token ID523
Feature activation+1.64e-02

Pos logit contributions
age+0.188
 back+0.156
 welcome+0.154
>+0.153
 home+0.151
Neg logit contributions
yles-0.215
cases-0.211
ls-0.196
yle-0.196
 salon-0.194
Tokenaps
Token ID1686
Feature activation-3.69e-02

Pos logit contributions
yles+0.035
cases+0.034
yle+0.032
 salon+0.031
ries+0.031
Neg logit contributions
age-0.022
 welcome-0.019
 back-0.018
 right-0.018
 home-0.018
Token,
Token ID11
Feature activation-5.38e-03

Pos logit contributions
age+0.053
 back+0.046
 welcome+0.044
est+0.044
 home+0.043
Neg logit contributions
yles-0.076
cases-0.073
yle-0.068
 salon-0.068
ls-0.066
Token they
Token ID484
Feature activation-8.29e-03

Pos logit contributions
 right+0.009
 back+0.009
 home+0.009
 welcome+0.009
 best+0.008
Neg logit contributions
yles-0.009
cases-0.009
outing-0.008
ls-0.008
yle-0.008
Token made
Token ID925
Feature activation-4.09e-02

Pos logit contributions
age+0.013
 back+0.012
 home+0.011
amed+0.011
>+0.011
Neg logit contributions
yles-0.015
cases-0.015
 salon-0.014
ls-0.014
ries-0.014
Token then
Token ID788
Feature activation+3.30e-02

Pos logit contributions
age+0.002
est+0.002
 welcome+0.002
>+0.002
 Follow+0.002
Neg logit contributions
yles-0.002
cases-0.002
ls-0.002
yle-0.002
 salon-0.002
Token on
Token ID319
Feature activation-3.24e-03

Pos logit contributions
yles+0.058
cases+0.057
ls+0.056
yle+0.054
 spill+0.052
Neg logit contributions
age-0.051
>-0.042
gy-0.042
est-0.041
 back-0.040
Token,
Token ID11
Feature activation-1.84e-03

Pos logit contributions
age+0.005
>+0.004
 Follow+0.004
amed+0.004
 welcome+0.004
Neg logit contributions
cases-0.006
yles-0.006
ls-0.006
yle-0.006
 salon-0.006
Token the
Token ID262
Feature activation-1.86e-02

Pos logit contributions
age+0.003
 back+0.003
>+0.003
 welcome+0.003
 home+0.003
Neg logit contributions
yles-0.003
cases-0.003
ls-0.003
yle-0.003
 salon-0.003
Token SPL
Token ID46341
Feature activation+2.54e-02

Pos logit contributions
age+0.035
>+0.028
 Follow+0.028
 back+0.028
 welcome+0.027
Neg logit contributions
yles-0.029
cases-0.029
ls-0.027
yle-0.027
 salon-0.027
TokenASH
Token ID11211
Feature activation-1.16e-01

Pos logit contributions
ries+0.052
yles+0.050
outing+0.048
yle+0.046
 hairc+0.046
Neg logit contributions
age-0.031
 welcome-0.028
 back-0.027
 Follow-0.026
 home-0.026
Token was
Token ID373
Feature activation+5.80e-03

Pos logit contributions
age+0.183
>+0.154
 welcome+0.151
 back+0.150
 Follow+0.148
Neg logit contributions
cases-0.212
yles-0.210
yle-0.194
ls-0.194
 salon-0.192
Token always
Token ID1464
Feature activation-7.70e-03

Pos logit contributions
yles+0.011
cases+0.010
 spill+0.010
ls+0.010
 salon+0.010
Neg logit contributions
age-0.009
 Follow-0.008
est-0.007
>-0.007
 welcome-0.007
Token under
Token ID739
Feature activation-1.87e-02

Pos logit contributions
age+0.013
 Follow+0.011
>+0.010
gy+0.009
gen+0.009
Neg logit contributions
yles-0.014
cases-0.013
 spill-0.013
 salon-0.012
ries-0.012
Token control
Token ID1630
Feature activation-1.03e-02

Pos logit contributions
age+0.029
 back+0.026
 home+0.025
 welcome+0.024
>+0.024
Neg logit contributions
yles-0.035
cases-0.035
yle-0.033
ls-0.032
 salon-0.032
Token!
Token ID0
Feature activation-1.43e-02

Pos logit contributions
age+0.014
 back+0.012
 home+0.012
est+0.012
 welcome+0.011
Neg logit contributions
yles-0.022
cases-0.021
yle-0.020
ls-0.020
 salon-0.019
Token smiled
Token ID13541
Feature activation+2.95e-03

Pos logit contributions
age+0.001
 back+0.000
 home+0.000
>+0.000
 right+0.000
Neg logit contributions
yles-0.001
cases-0.001
ls-0.001
ries-0.001
yle-0.001
Token and
Token ID290
Feature activation-6.93e-04

Pos logit contributions
cases+0.005
yles+0.005
ls+0.005
yle+0.005
 salon+0.005
Neg logit contributions
age-0.005
>-0.004
 welcome-0.004
 Follow-0.004
amed-0.004
Token said
Token ID531
Feature activation+2.50e-03

Pos logit contributions
age+0.001
 back+0.001
 home+0.001
 welcome+0.001
 right+0.001
Neg logit contributions
yles-0.001
cases-0.001
outing-0.001
ries-0.001
ls-0.001
Token,
Token ID11
Feature activation+1.42e-03

Pos logit contributions
yles+0.005
cases+0.005
ls+0.005
yle+0.005
 salon+0.004
Neg logit contributions
age-0.004
>-0.003
 back-0.003
 welcome-0.003
 home-0.003
Token "
Token ID366
Feature activation+1.67e-02

Pos logit contributions
yles+0.003
cases+0.003
ls+0.003
yle+0.003
 salon+0.003
Neg logit contributions
age-0.002
 back-0.002
>-0.002
 welcome-0.002
 home-0.002
TokenChurch
Token ID46686
Feature activation-1.16e-01

Pos logit contributions
yles+0.028
cases+0.028
ls+0.026
yle+0.026
 salon+0.025
Neg logit contributions
age-0.029
 back-0.024
 welcome-0.024
>-0.024
est-0.024
Token is
Token ID318
Feature activation+1.73e-03

Pos logit contributions
age+0.176
>+0.140
 back+0.138
 welcome+0.136
 Follow+0.135
Neg logit contributions
cases-0.220
yles-0.219
ls-0.205
yle-0.202
 salon-0.202
Token so
Token ID523
Feature activation+2.02e-02

Pos logit contributions
yles+0.003
cases+0.003
ls+0.003
 salon+0.003
yle+0.003
Neg logit contributions
age-0.003
>-0.002
 Follow-0.002
est-0.002
 back-0.002
Token nice
Token ID3621
Feature activation-2.31e-02

Pos logit contributions
yles+0.047
yle+0.044
cases+0.044
ries+0.043
 salon+0.043
Neg logit contributions
age-0.024
 welcome-0.022
 back-0.021
 right-0.020
 home-0.020
Token mom
Token ID1995
Feature activation-8.06e-03

Pos logit contributions
age+0.034
>+0.027
amed+0.026
 back+0.026
 Follow+0.026
Neg logit contributions
cases-0.045
yles-0.045
ls-0.042
 salon-0.041
yle-0.041
Tokenmy
Token ID1820
Feature activation-2.35e-02

Pos logit contributions
age+0.008
>+0.006
 Follow+0.006
amed+0.005
 back+0.005
Neg logit contributions
yles-0.019
cases-0.019
 salon-0.018
yle-0.018
ls-0.018
Token make
Token ID787
Feature activation-5.96e-03

Pos logit contributions
cases+0.024
yles+0.023
ls+0.022
yle+0.021
outing+0.021
Neg logit contributions
age-0.018
 back-0.015
 welcome-0.015
 home-0.015
 right-0.014
Token one
Token ID530
Feature activation-1.61e-02

Pos logit contributions
age+0.010
 back+0.008
 welcome+0.008
 home+0.008
>+0.008
Neg logit contributions
yles-0.011
cases-0.011
ls-0.010
yle-0.010
 salon-0.010
Token too
Token ID1165
Feature activation+3.32e-02

Pos logit contributions
age+0.028
>+0.022
gy+0.022
amed+0.021
 Follow+0.021
Neg logit contributions
yles-0.028
ls-0.027
cases-0.026
yle-0.026
 spill-0.026
Token.
Token ID13
Feature activation+1.70e-03

Pos logit contributions
yles+0.078
cases+0.074
 salon+0.072
yle+0.071
ls+0.070
Neg logit contributions
age-0.035
 back-0.034
 right-0.033
 home-0.032
 welcome-0.031
Token He
Token ID679
Feature activation-3.32e-03

Pos logit contributions
yles+0.004
cases+0.004
yle+0.004
ls+0.004
 salon+0.004
Neg logit contributions
age-0.002
 back-0.002
 welcome-0.002
>-0.002
 home-0.002
Token dips
Token ID48904
Feature activation-1.15e-01

Pos logit contributions
age+0.006
 back+0.005
 right+0.005
>+0.005
 welcome+0.005
Neg logit contributions
yles-0.005
cases-0.005
yle-0.005
outing-0.005
ries-0.004
Token his
Token ID465
Feature activation+1.86e-02

Pos logit contributions
age+0.189
 back+0.169
 home+0.162
 welcome+0.162
 right+0.160
Neg logit contributions
cases-0.207
yles-0.203
ls-0.185
 salon-0.184
yle-0.182
Token brush
Token ID14093
Feature activation+7.72e-03

Pos logit contributions
yles+0.028
cases+0.027
ls+0.026
yle+0.025
 salon+0.025
Neg logit contributions
age-0.036
>-0.030
 welcome-0.030
 Follow-0.029
 back-0.029
Token in
Token ID287
Feature activation-1.73e-02

Pos logit contributions
cases+0.015
yles+0.015
ls+0.014
yle+0.014
 salon+0.014
Neg logit contributions
age-0.011
 welcome-0.009
>-0.009
 Follow-0.009
est-0.008
Token the
Token ID262
Feature activation-1.71e-02

Pos logit contributions
age+0.023
 back+0.022
 right+0.021
 home+0.021
 welcome+0.020
Neg logit contributions
cases-0.037
yles-0.035
ls-0.033
 salon-0.032
yle-0.032
Token red
Token ID2266
Feature activation-3.45e-02

Pos logit contributions
age+0.032
gy+0.028
 Follow+0.027
>+0.026
wood+0.026
Neg logit contributions
 salon-0.025
 spill-0.025
 brushes-0.024
ls-0.022
 wra-0.022
Token.
Token ID13
Feature activation-1.15e-02

Pos logit contributions
age+0.002
>+0.002
 welcome+0.002
 back+0.002
 Follow+0.002
Neg logit contributions
yles-0.003
cases-0.003
ls-0.003
yle-0.003
 salon-0.003
Token It
Token ID632
Feature activation+1.61e-02

Pos logit contributions
age+0.017
 back+0.014
>+0.014
est+0.014
 home+0.014
Neg logit contributions
yles-0.022
cases-0.022
yle-0.020
ls-0.020
 salon-0.020
Token was
Token ID373
Feature activation+6.77e-03

Pos logit contributions
yles+0.029
cases+0.027
yle+0.026
outing+0.025
ls+0.025
Neg logit contributions
age-0.028
 back-0.024
 right-0.023
>-0.022
est-0.022
Token poison
Token ID8764
Feature activation+1.36e-02

Pos logit contributions
yles+0.013
cases+0.013
ls+0.012
yle+0.012
 salon+0.012
Neg logit contributions
age-0.010
 back-0.008
>-0.008
 welcome-0.008
est-0.008
Token he
Token ID339
Feature activation+2.82e-02

Pos logit contributions
cases+0.025
yles+0.025
ls+0.024
 spill+0.023
yle+0.023
Neg logit contributions
age-0.021
>-0.017
 welcome-0.017
 Follow-0.017
 home-0.016
Tokenml
Token ID4029
Feature activation-1.14e-01

Pos logit contributions
yles+0.050
cases+0.048
yle+0.045
 salon+0.045
ries+0.043
Neg logit contributions
age-0.050
 back-0.042
>-0.041
 welcome-0.041
 right-0.040
Tokenock
Token ID735
Feature activation+4.45e-03

Pos logit contributions
age+0.197
 back+0.167
 welcome+0.164
>+0.164
 Follow+0.163
Neg logit contributions
cases-0.189
yles-0.189
yle-0.176
ls-0.175
 salon-0.174
Token.
Token ID13
Feature activation-3.22e-03

Pos logit contributions
yles+0.009
cases+0.008
yle+0.008
 salon+0.008
ls+0.008
Neg logit contributions
age-0.007
 back-0.006
>-0.005
 home-0.005
est-0.005
Token It
Token ID632
Feature activation+1.71e-02

Pos logit contributions
age+0.005
 back+0.004
>+0.004
est+0.004
 home+0.004
Neg logit contributions
yles-0.006
cases-0.006
yle-0.006
ls-0.006
 salon-0.006
Token was
Token ID373
Feature activation+6.92e-03

Pos logit contributions
yles+0.032
cases+0.031
yle+0.029
 salon+0.029
outing+0.028
Neg logit contributions
age-0.029
 back-0.024
 right-0.023
>-0.023
gy-0.023
Token very
Token ID845
Feature activation-2.24e-02

Pos logit contributions
yles+0.014
cases+0.013
yle+0.013
ls+0.012
 salon+0.012
Neg logit contributions
age-0.010
 back-0.009
 welcome-0.008
>-0.008
 home-0.008
Token up
Token ID510
Feature activation+1.68e-02

Pos logit contributions
yles+0.009
cases+0.009
ls+0.008
yle+0.008
 salon+0.008
Neg logit contributions
age-0.005
>-0.003
 welcome-0.003
 Follow-0.003
 back-0.003
Token a
Token ID257
Feature activation+2.18e-04

Pos logit contributions
yles+0.041
cases+0.041
yle+0.038
 salon+0.037
outing+0.037
Neg logit contributions
 back-0.016
 right-0.016
age-0.015
 home-0.014
 welcome-0.013
Token carrot
Token ID40562
Feature activation-1.18e-02

Pos logit contributions
yles+0.000
cases+0.000
yle+0.000
ls+0.000
 salon+0.000
Neg logit contributions
age-0.000
 back-0.000
 welcome-0.000
 home-0.000
>-0.000
Token stick
Token ID4859
Feature activation+1.35e-02

Pos logit contributions
age+0.019
 welcome+0.017
>+0.017
 Follow+0.017
amed+0.017
Neg logit contributions
cases-0.019
ls-0.019
yles-0.019
 spill-0.018
 salon-0.018
Token and
Token ID290
Feature activation-2.59e-03

Pos logit contributions
yles+0.026
cases+0.026
ls+0.024
yle+0.024
 salon+0.024
Neg logit contributions
age-0.020
>-0.016
 welcome-0.016
 back-0.016
 home-0.016
Token dips
Token ID48904
Feature activation-1.14e-01

Pos logit contributions
 back+0.004
 right+0.004
age+0.004
 welcome+0.004
 Follow+0.004
Neg logit contributions
cases-0.004
yles-0.004
yle-0.004
outing-0.004
 salon-0.004
Token it
Token ID340
Feature activation+3.72e-03

Pos logit contributions
age+0.228
 back+0.198
 welcome+0.195
>+0.194
 home+0.193
Neg logit contributions
yles-0.166
cases-0.165
ls-0.149
yle-0.148
 salon-0.147
Token in
Token ID287
Feature activation-1.63e-02

Pos logit contributions
ls+0.007
yles+0.007
cases+0.007
outing+0.007
 spill+0.007
Neg logit contributions
age-0.006
 welcome-0.004
>-0.004
est-0.004
 Follow-0.004
Token a
Token ID257
Feature activation-6.31e-04

Pos logit contributions
age+0.022
 back+0.020
 right+0.020
 home+0.020
 welcome+0.020
Neg logit contributions
cases-0.035
yles-0.033
ls-0.031
 salon-0.030
yle-0.030
Token white
Token ID2330
Feature activation-2.28e-02

Pos logit contributions
age+0.001
>+0.001
 back+0.001
 welcome+0.001
est+0.001
Neg logit contributions
yles-0.001
cases-0.001
ls-0.001
yle-0.001
 salon-0.001
Token sauce
Token ID10746
Feature activation-2.56e-03

Pos logit contributions
age+0.043
 Follow+0.039
amed+0.038
 ok+0.038
 welcome+0.038
Neg logit contributions
 spill-0.029
 wra-0.029
ls-0.028
yles-0.027
 hairc-0.027
Token "
Token ID366
Feature activation+1.69e-02

Pos logit contributions
yles+0.004
cases+0.004
yle+0.004
ls+0.004
 salon+0.004
Neg logit contributions
age-0.003
 back-0.002
>-0.002
 welcome-0.002
 right-0.002
TokenWe
Token ID1135
Feature activation-4.27e-02

Pos logit contributions
yles+0.030
cases+0.029
yle+0.028
ls+0.027
 salon+0.027
Neg logit contributions
age-0.028
 back-0.023
 home-0.023
 welcome-0.023
>-0.023
Token are
Token ID389
Feature activation+2.63e-02

Pos logit contributions
age+0.084
>+0.071
 back+0.071
 home+0.069
est+0.069
Neg logit contributions
yles-0.063
cases-0.061
yle-0.057
ls-0.056
 spill-0.055
Token the
Token ID262
Feature activation-1.04e-02

Pos logit contributions
yles+0.040
ls+0.035
yle+0.034
 typing+0.034
cases+0.034
Neg logit contributions
age-0.056
est-0.046
ney-0.044
 Follow-0.042
nob-0.041
Token weapon
Token ID4282
Feature activation+4.15e-03

Pos logit contributions
age+0.012
>+0.009
 Follow+0.009
amed+0.008
est+0.008
Neg logit contributions
yles-0.024
 salon-0.023
yle-0.023
cases-0.023
 spill-0.022
Token creators
Token ID16294
Feature activation-1.14e-01

Pos logit contributions
cases+0.008
ls+0.007
yles+0.007
 salon+0.007
 spill+0.007
Neg logit contributions
age-0.006
>-0.005
amed-0.005
 welcome-0.005
 Follow-0.005
Token!
Token ID0
Feature activation-1.31e-02

Pos logit contributions
age+0.156
>+0.121
 back+0.120
 welcome+0.120
 Follow+0.118
Neg logit contributions
yles-0.229
cases-0.225
yle-0.214
 salon-0.212
ls-0.211
Token We
Token ID775
Feature activation-1.57e-02

Pos logit contributions
age+0.022
 home+0.019
 back+0.019
amed+0.019
>+0.018
Neg logit contributions
yles-0.022
cases-0.021
yle-0.020
ls-0.019
Get-0.019
Token can
Token ID460
Feature activation+1.09e-02

Pos logit contributions
age+0.030
>+0.025
 back+0.025
est+0.025
 home+0.025
Neg logit contributions
yles-0.024
cases-0.023
ls-0.022
yle-0.022
 spill-0.021
Token stop
Token ID2245
Feature activation+1.98e-02

Pos logit contributions
yles+0.019
cases+0.019
yle+0.018
ls+0.018
 salon+0.018
Neg logit contributions
age-0.018
>-0.015
 back-0.015
 welcome-0.015
est-0.015
Token the
Token ID262
Feature activation-1.10e-02

Pos logit contributions
yles+0.031
cases+0.028
ls+0.028
yle+0.027
 spill+0.027
Neg logit contributions
age-0.039
est-0.032
 welcome-0.032
amed-0.031
>-0.031
Token It
Token ID632
Feature activation+2.09e-02

Pos logit contributions
yles+0.018
cases+0.017
yle+0.017
ls+0.017
 salon+0.016
Neg logit contributions
age-0.008
 welcome-0.006
 back-0.006
 right-0.006
 Follow-0.006
Token is
Token ID318
Feature activation+3.85e-03

Pos logit contributions
yles+0.041
cases+0.040
yle+0.039
outing+0.036
ls+0.036
Neg logit contributions
age-0.032
 back-0.028
 right-0.026
 welcome-0.025
est-0.025
Token blue
Token ID4171
Feature activation-7.80e-03

Pos logit contributions
yles+0.007
cases+0.007
ls+0.007
yle+0.007
 salon+0.007
Neg logit contributions
age-0.006
 back-0.005
>-0.005
 welcome-0.005
 home-0.005
Token.
Token ID13
Feature activation+7.35e-03

Pos logit contributions
age+0.013
>+0.011
 Follow+0.010
 welcome+0.010
 back+0.010
Neg logit contributions
cases-0.014
yles-0.014
ls-0.013
 salon-0.013
ries-0.013
Token He
Token ID679
Feature activation+4.35e-03

Pos logit contributions
yles+0.017
cases+0.017
ls+0.016
 salon+0.016
yle+0.016
Neg logit contributions
age-0.008
 welcome-0.007
 back-0.006
 right-0.006
 Follow-0.006
Token dips
Token ID48904
Feature activation-1.13e-01

Pos logit contributions
cases+0.008
yles+0.008
yle+0.007
ge+0.007
outing+0.007
Neg logit contributions
 back-0.006
 right-0.006
age-0.006
 welcome-0.006
 best-0.006
Token his
Token ID465
Feature activation+2.50e-02

Pos logit contributions
age+0.174
 back+0.154
 welcome+0.147
 home+0.147
 right+0.144
Neg logit contributions
cases-0.215
yles-0.213
 salon-0.196
ls-0.194
yle-0.193
Token brush
Token ID14093
Feature activation+1.17e-02

Pos logit contributions
yles+0.040
cases+0.039
ls+0.036
yle+0.036
 salon+0.035
Neg logit contributions
age-0.046
 back-0.039
>-0.039
 welcome-0.039
 home-0.038
Token in
Token ID287
Feature activation-1.28e-02

Pos logit contributions
cases+0.024
yles+0.023
ls+0.022
 salon+0.022
yle+0.022
Neg logit contributions
age-0.016
 welcome-0.013
>-0.013
 Follow-0.013
est-0.012
Token the
Token ID262
Feature activation-1.12e-02

Pos logit contributions
age+0.019
 back+0.017
 home+0.016
 right+0.016
 welcome+0.016
Neg logit contributions
cases-0.025
yles-0.025
ls-0.023
yle-0.023
 salon-0.022
Token paint
Token ID7521
Feature activation+1.90e-02

Pos logit contributions
age+0.021
 Follow+0.018
>+0.018
gy+0.018
est+0.018
Neg logit contributions
cases-0.015
 salon-0.015
 spill-0.015
yles-0.014
 wra-0.014
Token and
Token ID290
Feature activation+1.68e-03

Pos logit contributions
yles+0.016
cases+0.015
ls+0.015
yle+0.014
 salon+0.014
Neg logit contributions
age-0.012
 back-0.010
>-0.010
 welcome-0.010
 home-0.010
Token a
Token ID257
Feature activation+2.21e-03

Pos logit contributions
yles+0.004
cases+0.004
yle+0.003
 salon+0.003
outing+0.003
Neg logit contributions
age-0.002
 back-0.002
 welcome-0.002
 right-0.002
est-0.002
Token wand
Token ID11569
Feature activation-6.40e-03

Pos logit contributions
yles+0.004
cases+0.004
ls+0.003
yle+0.003
 salon+0.003
Neg logit contributions
age-0.004
 back-0.003
>-0.003
 welcome-0.003
 home-0.003
Token.
Token ID13
Feature activation+5.00e-03

Pos logit contributions
age+0.009
 back+0.007
 home+0.007
 right+0.007
est+0.007
Neg logit contributions
yles-0.014
cases-0.013
yle-0.013
ls-0.012
 salon-0.012
Token She
Token ID1375
Feature activation-7.68e-03

Pos logit contributions
yles+0.009
cases+0.009
ls+0.008
yle+0.008
 salon+0.008
Neg logit contributions
age-0.009
 back-0.007
 welcome-0.007
>-0.007
 home-0.007
Token dips
Token ID48904
Feature activation-1.12e-01

Pos logit contributions
age+0.012
 back+0.011
 right+0.011
>+0.011
 welcome+0.011
Neg logit contributions
cases-0.013
yles-0.013
yle-0.012
ries-0.012
outing-0.012
Token the
Token ID262
Feature activation-1.08e-02

Pos logit contributions
age+0.218
 back+0.188
 welcome+0.185
>+0.185
 home+0.183
Neg logit contributions
yles-0.169
cases-0.167
ls-0.153
yle-0.152
 salon-0.150
Token wand
Token ID11569
Feature activation-6.76e-03

Pos logit contributions
age+0.022
>+0.019
 welcome+0.019
 Follow+0.019
 back+0.019
Neg logit contributions
yles-0.014
cases-0.014
ls-0.014
 spill-0.014
yle-0.013
Token in
Token ID287
Feature activation-1.28e-02

Pos logit contributions
age+0.010
 welcome+0.008
>+0.008
 Follow+0.008
gy+0.008
Neg logit contributions
cases-0.012
yles-0.012
ls-0.012
 spill-0.012
yle-0.011
Token the
Token ID262
Feature activation-1.11e-02

Pos logit contributions
age+0.019
 back+0.017
 home+0.016
 right+0.016
 welcome+0.016
Neg logit contributions
cases-0.025
yles-0.025
ls-0.023
yle-0.023
 salon-0.023
Token soap
Token ID19533
Feature activation+7.52e-03

Pos logit contributions
age+0.021
gy+0.019
>+0.019
 ok+0.018
 Follow+0.018
Neg logit contributions
 spill-0.015
 salon-0.014
ls-0.014
 wra-0.014
cases-0.014
Token chicken
Token ID9015
Feature activation-2.54e-02

Pos logit contributions
yles+0.007
cases+0.007
yle+0.007
ls+0.006
 salon+0.006
Neg logit contributions
age-0.007
 back-0.006
 welcome-0.006
 home-0.005
>-0.005
Token n
Token ID299
Feature activation+8.84e-03

Pos logit contributions
age+0.043
 welcome+0.037
amed+0.037
>+0.037
est+0.036
Neg logit contributions
ls-0.039
cases-0.039
yles-0.038
 salon-0.037
yle-0.037
Tokenug
Token ID1018
Feature activation-5.63e-02

Pos logit contributions
yles+0.014
cases+0.014
yle+0.013
ls+0.013
outing+0.013
Neg logit contributions
age-0.016
 back-0.014
 welcome-0.013
>-0.013
est-0.013
Tokenget
Token ID1136
Feature activation-9.52e-03

Pos logit contributions
age+0.103
 welcome+0.088
>+0.087
 Follow+0.086
 back+0.085
Neg logit contributions
yles-0.087
cases-0.087
yle-0.082
ls-0.081
 salon-0.079
Token and
Token ID290
Feature activation+9.49e-04

Pos logit contributions
age+0.015
 welcome+0.013
est+0.012
 Follow+0.012
>+0.012
Neg logit contributions
cases-0.017
yles-0.016
ls-0.016
 salon-0.016
yle-0.016
Token dips
Token ID48904
Feature activation-1.11e-01

Pos logit contributions
cases+0.002
yles+0.002
yle+0.002
outing+0.001
 salon+0.001
Neg logit contributions
age-0.002
 back-0.002
 right-0.001
 welcome-0.001
 Follow-0.001
Token it
Token ID340
Feature activation+8.01e-03

Pos logit contributions
age+0.210
 back+0.184
 welcome+0.180
 home+0.179
>+0.178
Neg logit contributions
cases-0.173
yles-0.173
 salon-0.155
ls-0.155
yle-0.154
Token in
Token ID287
Feature activation-1.32e-02

Pos logit contributions
ls+0.015
cases+0.015
yles+0.015
outing+0.014
 spill+0.014
Neg logit contributions
age-0.013
 welcome-0.010
est-0.010
>-0.009
 Follow-0.009
Token a
Token ID257
Feature activation+2.84e-03

Pos logit contributions
age+0.018
 back+0.017
 home+0.016
 right+0.016
 welcome+0.016
Neg logit contributions
cases-0.027
yles-0.026
ls-0.024
 salon-0.024
yle-0.024
Token red
Token ID2266
Feature activation-3.04e-02

Pos logit contributions
yles+0.005
cases+0.005
ls+0.005
yle+0.004
 salon+0.004
Neg logit contributions
age-0.005
>-0.004
 back-0.004
 welcome-0.004
est-0.004
Token sauce
Token ID10746
Feature activation+1.25e-03

Pos logit contributions
age+0.055
 Follow+0.049
 welcome+0.048
amed+0.048
>+0.048
Neg logit contributions
 spill-0.043
cases-0.043
 salon-0.042
yles-0.042
ls-0.042
Token their
Token ID511
Feature activation+1.60e-02

Pos logit contributions
cases+0.013
yles+0.012
outing+0.012
yle+0.011
 salon+0.011
Neg logit contributions
 back-0.011
age-0.011
 welcome-0.011
 home-0.011
 right-0.010
Token mom
Token ID1995
Feature activation-8.95e-03

Pos logit contributions
yles+0.033
cases+0.033
eling+0.030
outing+0.030
yle+0.029
Neg logit contributions
 back-0.023
 home-0.022
 best-0.021
 welcome-0.021
age-0.020
Token.
Token ID13
Feature activation+2.01e-03

Pos logit contributions
age+0.016
 Follow+0.013
>+0.013
amed+0.012
 welcome+0.012
Neg logit contributions
yles-0.015
cases-0.015
ls-0.014
 salon-0.014
yle-0.014
Token They
Token ID1119
Feature activation-6.71e-03

Pos logit contributions
yles+0.004
yle+0.004
cases+0.004
ls+0.004
 salon+0.004
Neg logit contributions
age-0.003
 back-0.002
 welcome-0.002
>-0.002
 home-0.002
Token saw
Token ID2497
Feature activation-2.13e-02

Pos logit contributions
age+0.009
 back+0.009
 welcome+0.009
 home+0.009
>+0.009
Neg logit contributions
yles-0.013
cases-0.013
ries-0.012
ls-0.012
outing-0.012
Token many
Token ID867
Feature activation+5.79e-02

Pos logit contributions
age+0.021
 back+0.017
 welcome+0.016
 home+0.016
 right+0.015
Neg logit contributions
yles-0.052
cases-0.051
yle-0.049
ls-0.048
 salon-0.048
Token flowers
Token ID12734
Feature activation+9.41e-03

Pos logit contributions
yles+0.129
cases+0.125
yle+0.119
 salon+0.119
oled+0.116
Neg logit contributions
 best-0.071
 welcome-0.067
est-0.067
 right-0.063
 back-0.058
Token,
Token ID11
Feature activation+1.12e-03

Pos logit contributions
yles+0.022
cases+0.021
yle+0.021
 salon+0.020
 hairc+0.020
Neg logit contributions
age-0.011
 back-0.009
 welcome-0.009
 right-0.008
 home-0.008
Token birds
Token ID10087
Feature activation-8.43e-03

Pos logit contributions
yles+0.003
yle+0.003
cases+0.003
 salon+0.002
outing+0.002
Neg logit contributions
age-0.001
 right-0.001
 back-0.001
 welcome-0.001
 across-0.001
Token and
Token ID290
Feature activation+1.72e-03

Pos logit contributions
age+0.007
 back+0.005
 right+0.005
 home+0.005
 welcome+0.005
Neg logit contributions
cases-0.023
yles-0.022
yle-0.022
 tant-0.021
 hairc-0.021
Token bugs
Token ID11316
Feature activation-2.12e-02

Pos logit contributions
yles+0.003
cases+0.003
yle+0.003
outing+0.003
 salon+0.003
Neg logit contributions
age-0.002
 back-0.002
 right-0.002
 welcome-0.002
 across-0.002
Token saw
Token ID2497
Feature activation-2.72e-02

Pos logit contributions
age+0.006
 back+0.006
 home+0.005
 right+0.005
 Follow+0.005
Neg logit contributions
yles-0.006
yle-0.005
outing-0.005
ls-0.005
cases-0.005
Token a
Token ID257
Feature activation-3.44e-03

Pos logit contributions
age+0.041
 back+0.034
 welcome+0.033
 home+0.033
>+0.032
Neg logit contributions
yles-0.053
cases-0.053
ls-0.049
yle-0.049
 salon-0.049
Token cute
Token ID13779
Feature activation-2.73e-02

Pos logit contributions
age+0.006
 back+0.005
 welcome+0.005
 home+0.005
 right+0.005
Neg logit contributions
yles-0.006
cases-0.006
ls-0.006
yle-0.006
ries-0.006
Token little
Token ID1310
Feature activation+3.65e-03

Pos logit contributions
age+0.046
 back+0.038
>+0.038
 welcome+0.038
 home+0.037
Neg logit contributions
yles-0.048
cases-0.047
ls-0.044
yle-0.044
 salon-0.043
Token puppy
Token ID26188
Feature activation-2.19e-02

Pos logit contributions
yles+0.006
cases+0.006
ls+0.006
 salon+0.006
yle+0.006
Neg logit contributions
age-0.006
>-0.005
amed-0.005
 welcome-0.005
 Follow-0.005
Token w
Token ID266
Feature activation+6.17e-02

Pos logit contributions
age+0.036
>+0.030
 welcome+0.029
amed+0.029
est+0.029
Neg logit contributions
yles-0.039
cases-0.038
ls-0.037
yle-0.036
 salon-0.035
Tokenagging
Token ID16406
Feature activation-5.86e-02

Pos logit contributions
yles+0.119
cases+0.116
 salon+0.108
yle+0.107
ls+0.107
Neg logit contributions
age-0.098
 back-0.080
 welcome-0.078
>-0.078
 home-0.077
Token its
Token ID663
Feature activation+2.28e-02

Pos logit contributions
age+0.092
 Follow+0.078
 welcome+0.077
>+0.073
est+0.072
Neg logit contributions
yles-0.103
cases-0.102
ls-0.097
yle-0.097
 spill-0.093
Token tail
Token ID7894
Feature activation-1.52e-02

Pos logit contributions
cases+0.037
 spill+0.036
 wra+0.036
ls+0.035
 tant+0.035
Neg logit contributions
 Follow-0.030
age-0.030
iddle-0.026
 welcome-0.026
wood-0.026
Token.
Token ID13
Feature activation-2.38e-03

Pos logit contributions
age+0.026
 welcome+0.022
 Follow+0.022
>+0.022
est+0.021
Neg logit contributions
cases-0.025
yles-0.024
ls-0.023
yle-0.022
 salon-0.022
Token
Token ID220
Feature activation+2.72e-02

Pos logit contributions
age+0.003
 back+0.002
>+0.002
 welcome+0.002
 home+0.002
Neg logit contributions
yles-0.005
cases-0.005
ls-0.005
yle-0.005
 salon-0.005
Token the
Token ID262
Feature activation-1.14e-02

Pos logit contributions
yles+0.004
cases+0.004
outing+0.003
yle+0.003
ls+0.003
Neg logit contributions
age-0.006
 back-0.006
 home-0.006
 welcome-0.006
 right-0.006
Token little
Token ID1310
Feature activation+9.96e-03

Pos logit contributions
age+0.020
 Follow+0.016
>+0.016
gy+0.016
amed+0.016
Neg logit contributions
cases-0.018
yles-0.018
 wra-0.017
 salon-0.017
ls-0.017
Token girl
Token ID2576
Feature activation-2.78e-02

Pos logit contributions
yles+0.017
cases+0.016
ls+0.016
 salon+0.016
ries+0.016
Neg logit contributions
age-0.017
>-0.014
 Follow-0.014
amed-0.014
 welcome-0.013
Token felt
Token ID2936
Feature activation+1.75e-03

Pos logit contributions
age+0.051
 back+0.043
>+0.043
 welcome+0.042
 home+0.042
Neg logit contributions
yles-0.044
cases-0.044
ls-0.040
yle-0.040
 salon-0.040
Token really
Token ID1107
Feature activation-1.38e-02

Pos logit contributions
yles+0.003
cases+0.003
yle+0.003
ls+0.003
 else+0.003
Neg logit contributions
age-0.003
gy-0.002
>-0.002
ney-0.002
man-0.002
Token proud
Token ID6613
Feature activation+6.06e-02

Pos logit contributions
age+0.020
 back+0.016
 welcome+0.016
>+0.015
 home+0.015
Neg logit contributions
yles-0.028
cases-0.028
ls-0.026
yle-0.026
 salon-0.026
Token of
Token ID286
Feature activation+3.00e-02

Pos logit contributions
yles+0.105
cases+0.104
ls+0.096
yle+0.096
 salon+0.094
Neg logit contributions
age-0.104
 back-0.088
 welcome-0.086
>-0.086
 home-0.085
Token herself
Token ID5223
Feature activation+4.63e-03

Pos logit contributions
yles+0.055
cases+0.054
ls+0.050
yle+0.050
 salon+0.049
Neg logit contributions
age-0.049
 back-0.041
 welcome-0.041
>-0.040
 home-0.040
Token for
Token ID329
Feature activation-1.84e-02

Pos logit contributions
cases+0.008
ls+0.008
yles+0.008
yle+0.007
 salon+0.007
Neg logit contributions
age-0.007
amed-0.006
>-0.006
 Follow-0.006
 welcome-0.006
Token becoming
Token ID5033
Feature activation+4.01e-03

Pos logit contributions
age+0.025
 back+0.023
 welcome+0.023
 right+0.022
 home+0.022
Neg logit contributions
cases-0.039
yles-0.037
ls-0.036
yle-0.034
 salon-0.034
Token brave
Token ID14802
Feature activation+3.49e-02

Pos logit contributions
yles+0.008
cases+0.007
yle+0.007
ls+0.007
ries+0.007
Neg logit contributions
age-0.006
>-0.005
 Follow-0.005
est-0.005
ney-0.005
Token week
Token ID1285
Feature activation+2.16e-02

Pos logit contributions
yles+0.065
cases+0.065
ls+0.063
yle+0.062
 salon+0.062
Neg logit contributions
age-0.021
>-0.014
 back-0.013
amed-0.013
 welcome-0.012
Token.
Token ID13
Feature activation+2.83e-03

Pos logit contributions
yles+0.048
cases+0.046
ls+0.045
yle+0.044
outing+0.044
Neg logit contributions
age-0.027
 back-0.023
 home-0.022
 right-0.021
 welcome-0.021
Token The
Token ID383
Feature activation-3.14e-03

Pos logit contributions
yles+0.005
cases+0.005
 salon+0.005
yle+0.005
 spill+0.005
Neg logit contributions
age-0.004
amed-0.003
>-0.003
 back-0.003
 home-0.003
Token farmer
Token ID18739
Feature activation+2.90e-02

Pos logit contributions
age+0.006
>+0.005
amed+0.005
 Follow+0.005
 back+0.005
Neg logit contributions
cases-0.005
yles-0.005
ls-0.005
 salon-0.004
yle-0.004
Token had
Token ID550
Feature activation-7.61e-03

Pos logit contributions
yles+0.050
cases+0.048
yle+0.045
ls+0.045
 salon+0.044
Neg logit contributions
age-0.051
 back-0.043
 welcome-0.042
>-0.042
 home-0.041
Token many
Token ID867
Feature activation+5.66e-02

Pos logit contributions
age+0.008
 back+0.007
 home+0.006
 welcome+0.006
 right+0.006
Neg logit contributions
yles-0.019
cases-0.018
yle-0.017
ls-0.017
outing-0.017
Token animals
Token ID4695
Feature activation+1.27e-03

Pos logit contributions
yles+0.123
 salon+0.115
yle+0.115
cases+0.115
outing+0.109
Neg logit contributions
age-0.076
est-0.066
 welcome-0.066
 home-0.065
 back-0.065
Token that
Token ID326
Feature activation+3.35e-03

Pos logit contributions
yles+0.003
cases+0.003
yle+0.003
ls+0.003
 salon+0.003
Neg logit contributions
age-0.002
 back-0.001
 welcome-0.001
 home-0.001
 right-0.001
Token Sam
Token ID3409
Feature activation+4.58e-02

Pos logit contributions
yles+0.006
cases+0.005
 salon+0.005
yle+0.005
outing+0.005
Neg logit contributions
age-0.006
 welcome-0.005
 back-0.005
 home-0.005
 right-0.005
Token loved
Token ID6151
Feature activation+4.19e-03

Pos logit contributions
yles+0.085
cases+0.080
ls+0.078
yle+0.078
outing+0.077
Neg logit contributions
age-0.074
 back-0.064
 home-0.060
 right-0.060
>-0.059
Token to
Token ID284
Feature activation+1.38e-02

Pos logit contributions
cases+0.008
yles+0.008
 salon+0.008
yle+0.008
ls+0.008
Neg logit contributions
age-0.006
 back-0.005
 home-0.005
 welcome-0.005
 right-0.005
Token I
Token ID314
Feature activation-2.20e-03

Pos logit contributions
yles+0.006
cases+0.006
yle+0.006
 hairc+0.006
 salon+0.006
Neg logit contributions
age-0.006
>-0.005
 back-0.005
 home-0.005
gy-0.005
Token will
Token ID481
Feature activation-4.63e-02

Pos logit contributions
age+0.004
>+0.003
 back+0.003
 welcome+0.003
 home+0.003
Neg logit contributions
yles-0.004
cases-0.004
ls-0.004
yle-0.004
 salon-0.004
Token give
Token ID1577
Feature activation+2.91e-02

Pos logit contributions
age+0.072
 back+0.059
>+0.059
est+0.057
 welcome+0.057
Neg logit contributions
yles-0.085
cases-0.084
yle-0.079
ls-0.078
 salon-0.078
Token you
Token ID345
Feature activation-1.51e-03

Pos logit contributions
yles+0.046
cases+0.044
ls+0.042
yle+0.042
 salon+0.040
Neg logit contributions
age-0.056
>-0.046
est-0.045
 welcome-0.045
 back-0.044
Token the
Token ID262
Feature activation-1.18e-02

Pos logit contributions
age+0.002
>+0.002
gy+0.002
 Follow+0.002
ney+0.002
Neg logit contributions
cases-0.003
yles-0.003
 hairc-0.003
ls-0.003
yle-0.003
Token courage
Token ID11917
Feature activation+6.02e-02

Pos logit contributions
age+0.023
 back+0.020
 welcome+0.020
 home+0.020
 right+0.020
Neg logit contributions
yles-0.018
cases-0.018
ls-0.017
yle-0.016
outing-0.016
Token and
Token ID290
Feature activation+2.00e-03

Pos logit contributions
cases+0.111
yles+0.110
ls+0.104
yle+0.103
 salon+0.102
Neg logit contributions
age-0.093
>-0.075
 welcome-0.073
 back-0.072
est-0.072
Token strength
Token ID4202
Feature activation+5.95e-02

Pos logit contributions
yles+0.004
cases+0.004
ls+0.003
yle+0.003
 salon+0.003
Neg logit contributions
age-0.003
 back-0.003
 welcome-0.003
>-0.003
 home-0.003
Token to
Token ID284
Feature activation+1.48e-02

Pos logit contributions
yles+0.115
cases+0.114
ls+0.107
yle+0.107
 salon+0.105
Neg logit contributions
age-0.088
>-0.070
 back-0.070
 welcome-0.069
est-0.068
Token save
Token ID3613
Feature activation+3.57e-02

Pos logit contributions
yles+0.031
cases+0.030
ls+0.029
yle+0.028
outing+0.028
Neg logit contributions
age-0.021
 welcome-0.017
 back-0.017
 home-0.017
 right-0.016
Token yourself
Token ID3511
Feature activation-3.54e-02

Pos logit contributions
yles+0.060
cases+0.059
ls+0.056
yle+0.055
 salon+0.054
Neg logit contributions
age-0.063
>-0.051
 back-0.051
 welcome-0.051
est-0.050
Token and
Token ID290
Feature activation-4.38e-03

Pos logit contributions
age+0.063
>+0.052
 Follow+0.051
 welcome+0.051
ney+0.048
Neg logit contributions
cases-0.071
yles-0.069
 salon-0.066
ls-0.065
yle-0.064
Token Eric
Token ID7651
Feature activation+9.93e-03

Pos logit contributions
age+0.008
 back+0.007
 home+0.007
 across+0.007
 right+0.007
Neg logit contributions
cases-0.007
yles-0.007
outing-0.006
yle-0.006
ries-0.006
Token played
Token ID2826
Feature activation-5.34e-03

Pos logit contributions
yles+0.017
cases+0.017
ries+0.016
ls+0.016
outing+0.015
Neg logit contributions
age-0.017
 back-0.015
est-0.014
 across-0.014
 home-0.014
Token the
Token ID262
Feature activation-1.28e-02

Pos logit contributions
age+0.009
>+0.007
 welcome+0.007
 Follow+0.007
ney+0.007
Neg logit contributions
yles-0.010
cases-0.009
ls-0.009
yle-0.009
 salon-0.009
Token trumpet
Token ID40334
Feature activation-4.85e-03

Pos logit contributions
age+0.026
 welcome+0.022
 Follow+0.022
>+0.022
 home+0.021
Neg logit contributions
cases-0.018
yles-0.018
ls-0.016
yle-0.016
 salon-0.016
Token.
Token ID13
Feature activation-1.89e-03

Pos logit contributions
age+0.007
 welcome+0.006
>+0.006
 Follow+0.006
amed+0.006
Neg logit contributions
cases-0.009
yles-0.009
ls-0.009
 salon-0.009
yle-0.008
Token
Token ID198
Feature activation-1.03e-02

Pos logit contributions
age+0.003
 back+0.002
>+0.002
 welcome+0.002
 home+0.002
Neg logit contributions
yles-0.004
cases-0.004
ls-0.004
yle-0.004
 salon-0.004
Token
Token ID198
Feature activation-9.37e-03

Pos logit contributions
age+0.014
 back+0.011
 welcome+0.011
>+0.011
est+0.011
Neg logit contributions
yles-0.021
cases-0.021
ls-0.020
yle-0.020
 salon-0.019
TokenThe
Token ID464
Feature activation+5.80e-03

Pos logit contributions
age+0.016
>+0.013
 back+0.013
 welcome+0.013
 home+0.013
Neg logit contributions
yles-0.017
cases-0.016
ls-0.015
yle-0.015
 salon-0.015
Token next
Token ID1306
Feature activation-8.68e-04

Pos logit contributions
yles+0.010
cases+0.010
ls+0.009
yle+0.009
 salon+0.009
Neg logit contributions
age-0.010
>-0.008
 back-0.008
 welcome-0.008
 Follow-0.008
Token day
Token ID1110
Feature activation+2.58e-03

Pos logit contributions
age+0.001
 Follow+0.001
>+0.001
gy+0.001
amed+0.001
Neg logit contributions
yles-0.002
cases-0.002
 salon-0.002
 hairc-0.002
yle-0.002
Token
Token ID198
Feature activation-9.55e-03

Pos logit contributions
yles+0.004
ls+0.004
cases+0.004
yle+0.003
s+0.003
Neg logit contributions
age-0.004
ney-0.003
est-0.003
>-0.003
gy-0.003
Token
Token ID198
Feature activation-9.53e-03

Pos logit contributions
age+0.016
est+0.014
 back+0.013
gy+0.013
 welcome+0.012
Neg logit contributions
yles-0.016
ls-0.016
cases-0.016
yle-0.015
l-0.015
TokenP
Token ID47
Feature activation-1.64e-02

Pos logit contributions
age+0.015
 back+0.013
 home+0.012
est+0.012
 welcome+0.012
Neg logit contributions
yles-0.017
ls-0.017
cases-0.016
yle-0.016
 spill-0.015
Tokenoppy
Token ID26696
Feature activation-7.71e-03

Pos logit contributions
age+0.033
 back+0.029
est+0.028
 right+0.028
>+0.028
Neg logit contributions
ls-0.023
yles-0.022
cases-0.021
yle-0.021
outing-0.020
Token and
Token ID290
Feature activation-2.25e-03

Pos logit contributions
age+0.014
 back+0.012
>+0.012
 welcome+0.011
 home+0.011
Neg logit contributions
yles-0.013
cases-0.012
yle-0.012
ls-0.011
 salon-0.011
Token her
Token ID607
Feature activation-3.91e-03

Pos logit contributions
age+0.003
 back+0.003
 home+0.003
 right+0.003
 welcome+0.003
Neg logit contributions
cases-0.004
yles-0.004
outing-0.004
ries-0.004
 salon-0.004
Token dad
Token ID9955
Feature activation+7.97e-03

Pos logit contributions
 back+0.006
 home+0.005
 best+0.005
 right+0.005
 welcome+0.004
Neg logit contributions
yles-0.009
cases-0.008
outing-0.008
ge-0.008
ries-0.008
Token threw
Token ID9617
Feature activation-1.46e-02

Pos logit contributions
yles+0.013
cases+0.012
ls+0.012
outing+0.012
 salon+0.012
Neg logit contributions
age-0.014
 back-0.013
 home-0.013
 right-0.012
amed-0.012
Token the
Token ID262
Feature activation-1.59e-02

Pos logit contributions
age+0.030
>+0.024
ney+0.024
 Follow+0.024
 welcome+0.024
Neg logit contributions
yles-0.021
ls-0.020
cases-0.020
yle-0.019
 spill-0.019
Token ball
Token ID2613
Feature activation-3.43e-02

Pos logit contributions
age+0.027
>+0.021
 welcome+0.021
 Follow+0.021
 back+0.021
Neg logit contributions
cases-0.028
yles-0.028
ls-0.027
yle-0.026
 salon-0.026
Token back
Token ID736
Feature activation-4.77e-02

Pos logit contributions
age+0.049
>+0.039
 welcome+0.039
 Follow+0.037
ney+0.037
Neg logit contributions
ls-0.067
cases-0.064
s-0.061
 spill-0.061
yles-0.060
Token their
Token ID511
Feature activation+1.22e-02

Pos logit contributions
age+0.014
>+0.011
amed+0.011
est+0.011
 Follow+0.011
Neg logit contributions
yles-0.014
cases-0.014
ls-0.013
yle-0.013
 salon-0.012
Token mom
Token ID1995
Feature activation-1.17e-02

Pos logit contributions
yles+0.022
cases+0.022
yle+0.021
ls+0.020
outing+0.019
Neg logit contributions
age-0.020
 back-0.018
 home-0.017
 welcome-0.017
>-0.017
Token if
Token ID611
Feature activation-1.75e-02

Pos logit contributions
age+0.014
 back+0.011
 welcome+0.010
>+0.010
 home+0.010
Neg logit contributions
yles-0.026
cases-0.026
ls-0.025
yle-0.025
 salon-0.024
Token they
Token ID484
Feature activation-3.49e-03

Pos logit contributions
 back+0.035
 right+0.034
 home+0.034
age+0.033
 welcome+0.033
Neg logit contributions
cases-0.023
yles-0.022
outing-0.022
 tant-0.021
yle-0.020
Token could
Token ID714
Feature activation+2.26e-03

Pos logit contributions
 back+0.005
age+0.005
 right+0.004
 welcome+0.004
 home+0.004
Neg logit contributions
yles-0.007
ries-0.007
cases-0.007
 salon-0.006
outing-0.006
Token play
Token ID711
Feature activation+1.59e-02

Pos logit contributions
yles+0.004
cases+0.004
ls+0.004
yle+0.004
 salon+0.004
Neg logit contributions
age-0.004
 back-0.003
>-0.003
 welcome-0.003
est-0.003
Token some
Token ID617
Feature activation+3.33e-02

Pos logit contributions
yles+0.030
cases+0.028
ls+0.028
 salon+0.028
 spill+0.027
Neg logit contributions
age-0.025
>-0.020
 Follow-0.019
amed-0.019
est-0.019
Token more
Token ID517
Feature activation-2.24e-02

Pos logit contributions
yles+0.079
cases+0.078
yle+0.074
ls+0.073
 salon+0.072
Neg logit contributions
age-0.035
 back-0.028
 welcome-0.027
 home-0.027
>-0.026
Token.
Token ID13
Feature activation+1.62e-03

Pos logit contributions
age+0.030
 back+0.025
 home+0.025
 welcome+0.025
 right+0.024
Neg logit contributions
yles-0.046
cases-0.046
yle-0.044
outing-0.042
 salon-0.042
Token Their
Token ID5334
Feature activation+1.70e-02

Pos logit contributions
yles+0.003
cases+0.003
yle+0.003
eling+0.003
ls+0.003
Neg logit contributions
age-0.003
 back-0.002
 home-0.002
 welcome-0.002
>-0.002
Token mom
Token ID1995
Feature activation-1.28e-02

Pos logit contributions
yles+0.033
cases+0.033
yle+0.031
ls+0.030
outing+0.029
Neg logit contributions
age-0.026
 back-0.024
 home-0.023
>-0.022
 welcome-0.022
Token not
Token ID407
Feature activation-1.71e-03

Pos logit contributions
yles+0.005
cases+0.005
ls+0.004
yle+0.004
 salon+0.004
Neg logit contributions
age-0.004
>-0.004
 back-0.004
 welcome-0.004
est-0.004
Token help
Token ID1037
Feature activation+6.10e-03

Pos logit contributions
age+0.003
 back+0.002
>+0.002
 welcome+0.002
 home+0.002
Neg logit contributions
yles-0.003
cases-0.003
ls-0.003
yle-0.003
 salon-0.003
Token the
Token ID262
Feature activation-1.57e-02

Pos logit contributions
yles+0.011
cases+0.011
ls+0.010
yle+0.010
 salon+0.010
Neg logit contributions
age-0.010
 back-0.009
 home-0.008
 welcome-0.008
 right-0.008
Token baby
Token ID5156
Feature activation+2.45e-03

Pos logit contributions
age+0.032
 back+0.029
 home+0.028
 welcome+0.028
>+0.028
Neg logit contributions
yles-0.022
cases-0.022
ls-0.019
yle-0.019
 salon-0.019
Token bird
Token ID6512
Feature activation-1.12e-02

Pos logit contributions
yles+0.005
cases+0.005
ls+0.004
yle+0.004
 salon+0.004
Neg logit contributions
age-0.004
 back-0.003
>-0.003
 welcome-0.003
 home-0.003
Token.
Token ID13
Feature activation+4.48e-04

Pos logit contributions
age+0.018
>+0.014
 welcome+0.014
 back+0.014
 home+0.014
Neg logit contributions
yles-0.021
cases-0.021
ls-0.019
yle-0.019
 salon-0.019
Token<|endoftext|>
Token ID50256
Feature activation+3.02e-02

Pos logit contributions
yles+0.001
cases+0.001
ls+0.001
yle+0.001
 salon+0.001
Neg logit contributions
age-0.001
est-0.001
 back-0.001
amed-0.001
 welcome-0.001
TokenOnce
Token ID7454
Feature activation+4.61e-02

Pos logit contributions
yles+0.047
ls+0.045
cases+0.045
yle+0.043
l+0.041
Neg logit contributions
age-0.055
est-0.045
 Follow-0.044
 advantage-0.044
amed-0.043
Token upon
Token ID2402
Feature activation+4.62e-05

Pos logit contributions
yles+0.075
cases+0.074
yle+0.068
 salon+0.067
 hairc+0.065
Neg logit contributions
age-0.080
 back-0.074
 home-0.073
 welcome-0.072
>-0.072
Token a
Token ID257
Feature activation-8.85e-05

Pos logit contributions
yles+0.000
cases+0.000
ls+0.000
yle+0.000
 salon+0.000
Neg logit contributions
age-0.000
 back-0.000
>-0.000
 welcome-0.000
 home-0.000
Token time
Token ID640
Feature activation-3.92e-03

Pos logit contributions
age+0.000
 back+0.000
 Follow+0.000
est+0.000
amed+0.000
Neg logit contributions
yles-0.000
cases-0.000
ls-0.000
 spill-0.000
 salon-0.000
Token.
Token ID13
Feature activation+6.69e-03

Pos logit contributions
age+0.012
 back+0.010
>+0.010
 welcome+0.010
 home+0.010
Neg logit contributions
yles-0.016
cases-0.016
ls-0.015
yle-0.015
 salon-0.015
Token They
Token ID1119
Feature activation+2.07e-03

Pos logit contributions
yles+0.013
cases+0.013
ls+0.012
yle+0.012
 salon+0.012
Neg logit contributions
age-0.010
 back-0.008
>-0.008
 welcome-0.008
est-0.008
Token forget
Token ID6044
Feature activation+3.86e-02

Pos logit contributions
yles+0.004
cases+0.004
ls+0.003
ries+0.003
yle+0.003
Neg logit contributions
age-0.003
 back-0.003
 welcome-0.003
 right-0.003
 home-0.003
Token about
Token ID546
Feature activation+2.06e-02

Pos logit contributions
yles+0.078
cases+0.076
yle+0.072
ls+0.072
 salon+0.071
Neg logit contributions
age-0.055
 back-0.046
>-0.045
 welcome-0.044
 home-0.044
Token the
Token ID262
Feature activation-8.44e-03

Pos logit contributions
cases+0.048
yles+0.043
ls+0.042
yle+0.041
outing+0.041
Neg logit contributions
 back-0.026
 home-0.025
 right-0.025
 welcome-0.023
 best-0.021
Token cat
Token ID3797
Feature activation+6.50e-03

Pos logit contributions
age+0.015
>+0.012
 back+0.012
 Follow+0.012
est+0.012
Neg logit contributions
yles-0.014
cases-0.014
ls-0.013
 salon-0.013
 spill-0.012
Token.
Token ID13
Feature activation+7.00e-03

Pos logit contributions
yles+0.013
cases+0.013
yle+0.012
ls+0.012
 salon+0.012
Neg logit contributions
age-0.009
 back-0.008
>-0.007
 home-0.007
 welcome-0.007
Token The
Token ID383
Feature activation+6.04e-03

Pos logit contributions
yles+0.014
cases+0.014
ls+0.013
yle+0.013
 salon+0.013
Neg logit contributions
age-0.010
 back-0.008
 welcome-0.008
>-0.008
 home-0.008
Token cat
Token ID3797
Feature activation+6.80e-03

Pos logit contributions
cases+0.009
 salon+0.009
yles+0.009
ls+0.009
 spill+0.009
Neg logit contributions
age-0.011
amed-0.009
est-0.009
 Follow-0.009
>-0.009
Token forg
Token ID5784
Feature activation+1.08e-02

Pos logit contributions
yles+0.013
cases+0.013
yle+0.012
outing+0.012
 salon+0.011
Neg logit contributions
age-0.011
 back-0.010
 right-0.009
 welcome-0.009
>-0.009
Tokenets
Token ID1039
Feature activation-1.87e-02

Pos logit contributions
yles+0.017
cases+0.017
ls+0.016
yle+0.015
 salon+0.015
Neg logit contributions
age-0.020
>-0.017
 back-0.017
 welcome-0.016
 home-0.016
Token,
Token ID11
Feature activation-6.45e-03

Pos logit contributions
yles+0.121
cases+0.115
yle+0.112
ls+0.110
 salon+0.107
Neg logit contributions
age-0.061
 back-0.054
 right-0.051
 home-0.049
 welcome-0.049
Token mom
Token ID1995
Feature activation-1.46e-02

Pos logit contributions
age+0.012
>+0.010
 back+0.010
 welcome+0.010
 home+0.009
Neg logit contributions
yles-0.011
cases-0.010
ls-0.010
yle-0.010
 salon-0.010
Token.
Token ID13
Feature activation-1.34e-03

Pos logit contributions
age+0.024
>+0.019
 Follow+0.019
 back+0.019
amed+0.019
Neg logit contributions
yles-0.026
cases-0.026
 salon-0.024
ls-0.024
yle-0.024
Token Tom
Token ID4186
Feature activation+2.63e-02

Pos logit contributions
age+0.003
amed+0.002
 home+0.002
sy+0.002
ench+0.002
Neg logit contributions
cases-0.002
yles-0.002
 hairc-0.002
 salon-0.002
yle-0.002
Token put
Token ID1234
Feature activation-1.34e-02

Pos logit contributions
yles+0.044
cases+0.044
ls+0.041
yle+0.040
 salon+0.040
Neg logit contributions
age-0.046
>-0.038
 back-0.038
 welcome-0.037
 home-0.037
Token me
Token ID502
Feature activation-4.27e-02

Pos logit contributions
age+0.021
>+0.017
est+0.016
 Follow+0.016
ney+0.016
Neg logit contributions
yles-0.025
cases-0.024
ls-0.024
ries-0.023
yle-0.023
Token in
Token ID287
Feature activation-1.90e-02

Pos logit contributions
age+0.056
>+0.046
est+0.045
 effect+0.044
 Follow+0.043
Neg logit contributions
cases-0.085
yles-0.082
ls-0.081
ries-0.080
 hairc-0.078
Token jail
Token ID7356
Feature activation+5.22e-02

Pos logit contributions
age+0.029
 back+0.023
>+0.023
 welcome+0.023
 home+0.023
Neg logit contributions
yles-0.037
cases-0.036
ls-0.034
yle-0.034
 salon-0.033
Token.
Token ID13
Feature activation-2.09e-03

Pos logit contributions
yles+0.121
yle+0.114
cases+0.110
ls+0.108
 spill+0.106
Neg logit contributions
age-0.058
 back-0.055
 right-0.051
 home-0.050
 welcome-0.046
Token But
Token ID887
Feature activation-1.40e-02

Pos logit contributions
age+0.003
amed+0.003
 home+0.003
 back+0.003
>+0.003
Neg logit contributions
cases-0.004
yles-0.004
 hairc-0.003
yle-0.003
 salon-0.003
Token I
Token ID314
Feature activation-1.14e-02

Pos logit contributions
age+0.024
 back+0.020
 welcome+0.020
 home+0.020
>+0.019
Neg logit contributions
yles-0.025
cases-0.024
ls-0.023
yle-0.022
 salon-0.022
Token store
Token ID3650
Feature activation+5.25e-02

Pos logit contributions
age+0.027
 back+0.023
 home+0.022
 welcome+0.022
>+0.022
Neg logit contributions
yles-0.028
cases-0.027
ls-0.026
yle-0.025
ries-0.024
Token.
Token ID13
Feature activation-4.11e-04

Pos logit contributions
cases+0.098
yles+0.095
yle+0.089
 salon+0.088
ls+0.088
Neg logit contributions
age-0.084
>-0.068
 welcome-0.067
est-0.067
 Follow-0.066
Token He
Token ID679
Feature activation-6.63e-03

Pos logit contributions
age+0.000
 back+0.000
 welcome+0.000
>+0.000
 home+0.000
Neg logit contributions
yles-0.001
cases-0.001
ls-0.001
yle-0.001
 salon-0.001
Token was
Token ID373
Feature activation+5.78e-03

Pos logit contributions
age+0.011
 back+0.010
>+0.009
 home+0.009
gy+0.009
Neg logit contributions
yles-0.012
cases-0.012
ls-0.011
ries-0.011
yle-0.011
Token looking
Token ID2045
Feature activation-1.42e-02

Pos logit contributions
yles+0.013
cases+0.013
yle+0.012
 salon+0.012
ls+0.012
Neg logit contributions
age-0.007
>-0.005
est-0.005
 back-0.005
 welcome-0.005
Token for
Token ID329
Feature activation-2.54e-02

Pos logit contributions
age+0.016
>+0.010
gy+0.010
amed+0.010
ney+0.010
Neg logit contributions
cases-0.034
yles-0.033
ls-0.033
yle-0.032
 hairc-0.031
Token a
Token ID257
Feature activation-2.26e-03

Pos logit contributions
age+0.026
 back+0.026
 home+0.026
 welcome+0.024
 right+0.024
Neg logit contributions
yles-0.059
cases-0.058
outing-0.056
ls-0.055
 salon-0.055
Token new
Token ID649
Feature activation+1.40e-02

Pos logit contributions
age+0.004
 back+0.003
 welcome+0.003
 home+0.003
>+0.003
Neg logit contributions
yles-0.004
cases-0.004
ls-0.004
yle-0.004
 salon-0.004
Token hat
Token ID6877
Feature activation-5.30e-03

Pos logit contributions
 salon+0.023
yles+0.023
cases+0.023
yle+0.022
 hairc+0.021
Neg logit contributions
age-0.024
>-0.021
amed-0.021
 Follow-0.020
 welcome-0.019
Token to
Token ID284
Feature activation+9.83e-03

Pos logit contributions
age+0.008
>+0.007
 welcome+0.007
 Follow+0.006
 back+0.006
Neg logit contributions
cases-0.010
yles-0.010
yle-0.009
ls-0.009
 salon-0.009
Token wear
Token ID5806
Feature activation-1.89e-02

Pos logit contributions
yles+0.019
cases+0.019
outing+0.017
ls+0.017
yle+0.016
Neg logit contributions
age-0.017
 welcome-0.014
 back-0.014
 home-0.014
 right-0.013
Token plate
Token ID7480
Feature activation-2.44e-02

Pos logit contributions
age+0.013
 back+0.012
 welcome+0.011
 right+0.011
 home+0.011
Neg logit contributions
yles-0.015
cases-0.015
ls-0.013
yle-0.013
outing-0.013
Token.
Token ID13
Feature activation+8.22e-03

Pos logit contributions
age+0.034
 back+0.029
 welcome+0.027
 home+0.027
 right+0.027
Neg logit contributions
yles-0.052
cases-0.049
yle-0.047
ls-0.047
 salon-0.046
Token She
Token ID1375
Feature activation-5.47e-04

Pos logit contributions
yles+0.016
cases+0.016
ls+0.015
yle+0.015
 salon+0.014
Neg logit contributions
age-0.012
 back-0.010
 welcome-0.010
>-0.010
 home-0.010
Token saw
Token ID2497
Feature activation-1.84e-02

Pos logit contributions
 back+0.001
amed+0.001
age+0.001
 right+0.001
>+0.001
Neg logit contributions
yles-0.001
cases-0.001
outing-0.001
 salon-0.001
 hairc-0.001
Token the
Token ID262
Feature activation-8.45e-03

Pos logit contributions
 back+0.019
 right+0.018
 home+0.017
 welcome+0.016
 out+0.015
Neg logit contributions
cases-0.042
yles-0.040
outing-0.040
eling-0.038
yle-0.038
Token pastry
Token ID41738
Feature activation-2.00e-02

Pos logit contributions
age+0.014
 back+0.013
 right+0.013
 home+0.012
 welcome+0.012
Neg logit contributions
yles-0.015
cases-0.014
ls-0.013
yle-0.013
outing-0.013
Token and
Token ID290
Feature activation+3.55e-03

Pos logit contributions
age+0.022
 back+0.021
 right+0.020
 home+0.018
 across+0.018
Neg logit contributions
yles-0.047
cases-0.044
yle-0.043
outing-0.041
ls-0.041
Token the
Token ID262
Feature activation-8.31e-03

Pos logit contributions
yles+0.008
cases+0.008
outing+0.007
yle+0.007
ls+0.007
Neg logit contributions
 back-0.005
 right-0.004
 across-0.004
 over-0.004
 welcome-0.004
Token note
Token ID3465
Feature activation-1.81e-02

Pos logit contributions
age+0.015
 back+0.013
 welcome+0.013
>+0.013
 home+0.013
Neg logit contributions
yles-0.013
cases-0.013
ls-0.012
yle-0.012
 salon-0.012
Token.
Token ID13
Feature activation+8.22e-03

Pos logit contributions
 back+0.017
 right+0.016
age+0.014
 across+0.012
 home+0.012
Neg logit contributions
yles-0.049
yle-0.045
 tant-0.044
eling-0.044
 hairc-0.043
Token She
Token ID1375
Feature activation-7.75e-04

Pos logit contributions
yles+0.017
cases+0.016
ls+0.015
yle+0.015
 salon+0.015
Neg logit contributions
age-0.012
 back-0.010
 welcome-0.010
>-0.009
 home-0.009
Token day
Token ID1110
Feature activation-2.67e-04

Pos logit contributions
age+0.044
 back+0.029
>+0.028
amed+0.028
 Follow+0.027
Neg logit contributions
yles-0.111
cases-0.110
ls-0.108
 salon-0.106
 spill-0.106
Token,
Token ID11
Feature activation-5.96e-03

Pos logit contributions
age+0.000
 back+0.000
 welcome+0.000
>+0.000
 home+0.000
Neg logit contributions
yles-0.001
cases-0.001
yle-0.001
ls-0.001
 salon-0.001
Token the
Token ID262
Feature activation-2.19e-02

Pos logit contributions
age+0.007
 back+0.007
 home+0.007
 right+0.007
 welcome+0.007
Neg logit contributions
yles-0.012
cases-0.012
 salon-0.011
outing-0.011
ls-0.011
Token little
Token ID1310
Feature activation+2.85e-03

Pos logit contributions
age+0.044
>+0.037
 back+0.037
 welcome+0.036
 home+0.036
Neg logit contributions
cases-0.032
yles-0.032
ls-0.029
yle-0.028
 salon-0.028
Token boy
Token ID2933
Feature activation-2.49e-02

Pos logit contributions
yles+0.003
cases+0.003
ls+0.003
 spill+0.003
 salon+0.003
Neg logit contributions
age-0.006
>-0.005
amed-0.005
 Follow-0.005
sy-0.005
Token decided
Token ID3066
Feature activation-2.57e-02

Pos logit contributions
age+0.045
 back+0.037
 welcome+0.037
>+0.037
 home+0.036
Neg logit contributions
yles-0.041
cases-0.041
ls-0.038
yle-0.037
 salon-0.037
Token to
Token ID284
Feature activation+6.02e-03

Pos logit contributions
age+0.038
 back+0.031
>+0.030
 welcome+0.030
 home+0.030
Neg logit contributions
yles-0.051
cases-0.050
ls-0.047
yle-0.047
 salon-0.046
Token try
Token ID1949
Feature activation+4.61e-02

Pos logit contributions
yles+0.011
cases+0.011
yle+0.010
 salon+0.010
ls+0.010
Neg logit contributions
age-0.009
 back-0.008
>-0.008
 welcome-0.008
est-0.007
Token and
Token ID290
Feature activation-5.83e-03

Pos logit contributions
yles+0.093
cases+0.092
ls+0.087
yle+0.087
 salon+0.085
Neg logit contributions
age-0.065
 back-0.053
>-0.051
 home-0.051
 welcome-0.051
Token count
Token ID954
Feature activation-3.17e-03

Pos logit contributions
age+0.008
 back+0.007
 welcome+0.007
 home+0.007
>+0.006
Neg logit contributions
yles-0.012
cases-0.012
ls-0.011
yle-0.011
outing-0.011
Token the
Token ID262
Feature activation-2.31e-02

Pos logit contributions
 back+0.004
age+0.004
 home+0.004
 right+0.004
 across+0.004
Neg logit contributions
yles-0.007
cases-0.007
yle-0.006
 salon-0.006
ls-0.006
Token the
Token ID262
Feature activation-1.25e-02

Pos logit contributions
yles+0.001
cases+0.001
outing+0.001
yle+0.001
ls+0.001
Neg logit contributions
age-0.001
 back-0.001
 home-0.001
 right-0.001
 welcome-0.001
Token teacher
Token ID4701
Feature activation+1.87e-02

Pos logit contributions
age+0.025
 back+0.023
 welcome+0.022
 right+0.022
 home+0.022
Neg logit contributions
yles-0.019
cases-0.017
ries-0.016
ls-0.016
yle-0.016
Token wrote
Token ID2630
Feature activation-1.60e-02

Pos logit contributions
yles+0.033
outing+0.031
yle+0.030
ls+0.030
ries+0.029
Neg logit contributions
age-0.030
 back-0.027
 right-0.026
 home-0.026
 across-0.024
Token on
Token ID319
Feature activation+8.11e-04

Pos logit contributions
age+0.020
 back+0.017
>+0.015
 home+0.015
 right+0.015
Neg logit contributions
yles-0.035
cases-0.034
yle-0.033
 salon-0.033
ls-0.033
Token the
Token ID262
Feature activation-1.28e-02

Pos logit contributions
yles+0.002
cases+0.002
yle+0.002
ls+0.002
outing+0.002
Neg logit contributions
age-0.001
 back-0.001
 right-0.001
 home-0.001
est-0.001
Token black
Token ID2042
Feature activation-1.92e-02

Pos logit contributions
age+0.022
 back+0.019
 right+0.018
 welcome+0.018
 home+0.018
Neg logit contributions
yles-0.023
cases-0.022
ls-0.021
yle-0.020
ries-0.020
Tokenboard
Token ID3526
Feature activation+6.21e-03

Pos logit contributions
age+0.031
 back+0.026
est+0.026
 welcome+0.025
 home+0.025
Neg logit contributions
yles-0.036
cases-0.034
ls-0.033
yle-0.032
outing-0.031
Token with
Token ID351
Feature activation+7.62e-03

Pos logit contributions
yles+0.013
cases+0.012
yle+0.012
ls+0.012
 salon+0.011
Neg logit contributions
age-0.009
 back-0.007
>-0.007
 welcome-0.007
 home-0.007
Token chalk
Token ID30860
Feature activation+4.06e-02

Pos logit contributions
yles+0.019
outing+0.018
 salon+0.018
cases+0.018
ls+0.018
Neg logit contributions
age-0.007
 back-0.007
 home-0.006
 welcome-0.006
 right-0.006
Token.
Token ID13
Feature activation+2.49e-03

Pos logit contributions
yles+0.092
yle+0.084
cases+0.084
eling+0.083
outing+0.082
Neg logit contributions
age-0.051
 back-0.043
 right-0.040
 home-0.039
 across-0.037
Token Lily
Token ID20037
Feature activation+5.32e-03

Pos logit contributions
yles+0.005
cases+0.005
ls+0.004
yle+0.004
 salon+0.004
Neg logit contributions
age-0.004
 back-0.003
>-0.003
 welcome-0.003
 home-0.003
Token I
Token ID314
Feature activation-8.11e-03

Pos logit contributions
age+0.017
 back+0.015
 welcome+0.015
 home+0.015
>+0.015
Neg logit contributions
yles-0.015
cases-0.015
ls-0.014
yle-0.014
 salon-0.013
Token want
Token ID765
Feature activation-5.79e-02

Pos logit contributions
age+0.014
 welcome+0.011
 back+0.011
>+0.011
 home+0.011
Neg logit contributions
yles-0.015
cases-0.015
ls-0.014
yle-0.014
ries-0.013
Token to
Token ID284
Feature activation+1.01e-02

Pos logit contributions
age+0.075
 back+0.066
 home+0.064
 welcome+0.061
 right+0.060
Neg logit contributions
yles-0.123
cases-0.121
yle-0.114
ls-0.113
outing-0.112
Token see
Token ID766
Feature activation-1.15e-02

Pos logit contributions
yles+0.020
cases+0.019
ls+0.018
yle+0.018
outing+0.017
Neg logit contributions
age-0.016
 back-0.013
 welcome-0.013
 home-0.013
>-0.013
Token.
Token ID13
Feature activation-4.19e-04

Pos logit contributions
age+0.017
 back+0.016
 home+0.015
 right+0.015
 welcome+0.015
Neg logit contributions
cases-0.021
yles-0.021
 salon-0.020
ls-0.020
outing-0.020
Token Maybe
Token ID6674
Feature activation-8.55e-02

Pos logit contributions
age+0.001
est+0.001
amed+0.001
 home+0.001
>+0.001
Neg logit contributions
cases-0.001
yles-0.001
 salon-0.001
yle-0.001
 hairc-0.001
Token it
Token ID340
Feature activation+2.78e-03

Pos logit contributions
 back+0.136
age+0.134
 home+0.132
 right+0.131
 welcome+0.127
Neg logit contributions
yles-0.142
cases-0.138
outing-0.135
yle-0.130
 tant-0.129
Token is
Token ID318
Feature activation-2.99e-03

Pos logit contributions
yles+0.006
cases+0.006
yle+0.006
 salon+0.005
ls+0.005
Neg logit contributions
age-0.004
 back-0.003
 right-0.003
 welcome-0.003
est-0.003
Token a
Token ID257
Feature activation-1.41e-03

Pos logit contributions
age+0.005
 Follow+0.004
est+0.004
>+0.004
amed+0.004
Neg logit contributions
yles-0.005
cases-0.005
ls-0.005
 spill-0.005
 salon-0.005
Token doll
Token ID3654
Feature activation-1.78e-02

Pos logit contributions
age+0.002
amed+0.002
>+0.002
 Follow+0.002
est+0.002
Neg logit contributions
yles-0.002
 salon-0.002
cases-0.002
 spill-0.002
ls-0.002
Token or
Token ID393
Feature activation-3.67e-02

Pos logit contributions
age+0.025
 back+0.020
>+0.020
 welcome+0.020
 home+0.019
Neg logit contributions
yles-0.036
cases-0.036
ls-0.033
yle-0.033
 salon-0.033
Tokenhouse
Token ID4803
Feature activation-4.69e-02

Pos logit contributions
cases+0.058
yles+0.056
 salon+0.054
ls+0.053
yle+0.053
Neg logit contributions
age-0.045
>-0.037
 welcome-0.036
 Follow-0.036
amed-0.036
Token.
Token ID13
Feature activation-1.15e-03

Pos logit contributions
age+0.072
>+0.059
 Follow+0.057
 welcome+0.056
amed+0.055
Neg logit contributions
cases-0.089
yles-0.086
ls-0.082
 salon-0.081
yle-0.080
Token
Token ID198
Feature activation-1.30e-02

Pos logit contributions
age+0.002
est+0.001
sy+0.001
 back+0.001
amed+0.001
Neg logit contributions
cases-0.002
yles-0.002
ls-0.002
yle-0.002
ries-0.002
Token
Token ID198
Feature activation-1.31e-02

Pos logit contributions
age+0.021
est+0.020
 back+0.018
 welcome+0.017
 effect+0.017
Neg logit contributions
yles-0.022
cases-0.022
ls-0.020
yle-0.020
l-0.019
TokenBilly
Token ID49640
Feature activation-1.11e-02

Pos logit contributions
age+0.021
est+0.019
 back+0.019
 welcome+0.019
 home+0.019
Neg logit contributions
cases-0.022
yles-0.021
ls-0.019
yle-0.019
outing-0.019
Token stamped
Token ID28305
Feature activation-9.80e-02

Pos logit contributions
age+0.022
 welcome+0.019
 back+0.019
>+0.018
 home+0.018
Neg logit contributions
cases-0.016
yles-0.016
ls-0.015
yle-0.014
 salon-0.014
Token his
Token ID465
Feature activation+1.87e-02

Pos logit contributions
age+0.166
 back+0.137
>+0.136
 welcome+0.135
 home+0.133
Neg logit contributions
yles-0.174
cases-0.171
ls-0.160
yle-0.159
 salon-0.156
Token feet
Token ID3625
Feature activation+4.03e-03

Pos logit contributions
cases+0.028
 salon+0.027
ls+0.027
yles+0.027
 spill+0.026
Neg logit contributions
age-0.034
 Follow-0.029
 welcome-0.029
amed-0.029
>-0.028
Token and
Token ID290
Feature activation-3.98e-03

Pos logit contributions
cases+0.008
yles+0.007
 salon+0.007
ls+0.007
yle+0.007
Neg logit contributions
age-0.006
 welcome-0.005
>-0.005
 Follow-0.005
est-0.005
Token said
Token ID531
Feature activation-2.13e-03

Pos logit contributions
age+0.008
 back+0.007
 home+0.007
 right+0.007
>+0.007
Neg logit contributions
yles-0.006
cases-0.005
yle-0.005
ls-0.005
 salon-0.005
Token,
Token ID11
Feature activation-4.46e-03

Pos logit contributions
age+0.003
>+0.002
 welcome+0.002
est+0.002
 Follow+0.002
Neg logit contributions
cases-0.004
yles-0.004
ls-0.004
yle-0.004
 salon-0.004
Token noise
Token ID7838
Feature activation+3.38e-03

Pos logit contributions
cases+0.022
yles+0.022
ls+0.022
 spill+0.022
 salon+0.020
Neg logit contributions
age-0.029
 Follow-0.026
>-0.026
amed-0.026
 welcome-0.026
Token.
Token ID13
Feature activation-3.49e-03

Pos logit contributions
yles+0.007
cases+0.007
ls+0.006
yle+0.006
 salon+0.006
Neg logit contributions
age-0.005
 back-0.004
>-0.004
 welcome-0.004
 home-0.004
Token
Token ID198
Feature activation-8.95e-03

Pos logit contributions
age+0.005
 back+0.004
 welcome+0.004
>+0.004
 home+0.004
Neg logit contributions
yles-0.007
cases-0.007
ls-0.007
yle-0.007
 salon-0.006
Token
Token ID198
Feature activation-8.23e-03

Pos logit contributions
age+0.012
 back+0.010
>+0.009
 welcome+0.009
est+0.009
Neg logit contributions
yles-0.019
cases-0.018
ls-0.018
yle-0.017
 salon-0.017
Token"
Token ID1
Feature activation-6.97e-04

Pos logit contributions
age+0.013
 back+0.010
>+0.010
 welcome+0.010
 home+0.010
Neg logit contributions
yles-0.016
cases-0.016
ls-0.015
yle-0.015
 salon-0.014
TokenCrunch
Token ID49384
Feature activation-8.50e-02

Pos logit contributions
age+0.001
 right+0.001
 back+0.001
>+0.001
amed+0.001
Neg logit contributions
cases-0.001
yles-0.001
ls-0.001
yle-0.001
 salon-0.001
Token,
Token ID11
Feature activation-5.28e-03

Pos logit contributions
age+0.135
 Follow+0.111
 back+0.106
 home+0.104
>+0.103
Neg logit contributions
ls-0.158
ries-0.154
cases-0.152
yles-0.147
 spill-0.143
Token crunch
Token ID23527
Feature activation-1.39e-02

Pos logit contributions
age+0.009
 welcome+0.008
 back+0.008
 right+0.008
>+0.007
Neg logit contributions
yles-0.010
cases-0.009
yle-0.009
ls-0.008
 salon-0.008
Token,
Token ID11
Feature activation-4.14e-03

Pos logit contributions
age+0.021
 back+0.017
>+0.017
 welcome+0.017
 home+0.017
Neg logit contributions
yles-0.026
cases-0.026
ls-0.025
yle-0.024
ries-0.024
Token crunch
Token ID23527
Feature activation-1.31e-02

Pos logit contributions
age+0.007
 back+0.006
 welcome+0.006
 right+0.006
>+0.006
Neg logit contributions
yles-0.007
cases-0.007
yle-0.007
 salon-0.007
ls-0.007
Token.
Token ID13
Feature activation+3.86e-03

Pos logit contributions
age+0.020
 back+0.016
>+0.016
 welcome+0.016
 home+0.016
Neg logit contributions
yles-0.025
cases-0.025
ls-0.024
yle-0.023
 salon-0.023
Token dream
Token ID4320
Feature activation-7.85e-03

Pos logit contributions
age+0.045
amed+0.040
>+0.037
 Follow+0.037
 back+0.036
Neg logit contributions
cases-0.062
yles-0.061
 salon-0.060
ls-0.058
 hairc-0.057
Token,
Token ID11
Feature activation-1.30e-03

Pos logit contributions
age+0.010
 back+0.008
>+0.008
 welcome+0.008
 home+0.008
Neg logit contributions
yles-0.017
cases-0.017
ls-0.016
yle-0.015
 salon-0.015
Token Tim
Token ID5045
Feature activation+1.75e-02

Pos logit contributions
age+0.002
 back+0.002
>+0.002
 welcome+0.002
 home+0.002
Neg logit contributions
yles-0.002
cases-0.002
ls-0.002
yle-0.002
 salon-0.002
Tokenmy
Token ID1820
Feature activation-2.30e-02

Pos logit contributions
yles+0.032
ries+0.031
 salon+0.031
 spill+0.030
cases+0.030
Neg logit contributions
age-0.025
 Follow-0.023
amed-0.021
>-0.020
 across-0.020
Token.
Token ID13
Feature activation+1.38e-03

Pos logit contributions
age+0.037
amed+0.032
>+0.031
 Follow+0.031
 welcome+0.029
Neg logit contributions
yles-0.040
cases-0.039
ls-0.037
yle-0.037
 hairc-0.036
Token Maybe
Token ID6674
Feature activation-8.53e-02

Pos logit contributions
yles+0.002
cases+0.002
yle+0.002
 hairc+0.002
 salon+0.002
Neg logit contributions
age-0.002
amed-0.002
 home-0.002
>-0.002
 back-0.002
Token we
Token ID356
Feature activation-2.38e-02

Pos logit contributions
age+0.129
 back+0.115
 home+0.112
 welcome+0.112
 right+0.110
Neg logit contributions
cases-0.159
yles-0.158
outing-0.148
ls-0.148
 salon-0.146
Token can
Token ID460
Feature activation+6.61e-03

Pos logit contributions
age+0.040
 back+0.033
 welcome+0.033
>+0.033
 home+0.032
Neg logit contributions
yles-0.042
cases-0.042
ls-0.039
yle-0.039
 salon-0.038
Token find
Token ID1064
Feature activation-3.26e-02

Pos logit contributions
yles+0.012
cases+0.012
yle+0.011
 salon+0.011
ls+0.011
Neg logit contributions
age-0.010
>-0.008
 back-0.008
 welcome-0.008
est-0.008
Token a
Token ID257
Feature activation-5.09e-04

Pos logit contributions
age+0.049
>+0.038
 back+0.037
 welcome+0.037
est+0.036
Neg logit contributions
yles-0.065
cases-0.064
ls-0.060
yle-0.060
 salon-0.058
Token book
Token ID1492
Feature activation-2.38e-02

Pos logit contributions
age+0.001
>+0.001
amed+0.001
est+0.001
 Follow+0.001
Neg logit contributions
yles-0.001
cases-0.001
 salon-0.001
yle-0.001
 hairc-0.001
Token to
Token ID284
Feature activation+1.90e-02

Pos logit contributions
age+0.030
 back+0.024
>+0.023
 welcome+0.023
 home+0.023
Neg logit contributions
yles-0.042
cases-0.041
ls-0.039
yle-0.039
 salon-0.038
Token them
Token ID606
Feature activation+1.56e-02

Pos logit contributions
cases+0.035
yles+0.034
outing+0.032
ls+0.031
ries+0.031
Neg logit contributions
 back-0.031
age-0.030
 welcome-0.029
 home-0.029
 right-0.028
Token.
Token ID13
Feature activation+8.71e-03

Pos logit contributions
cases+0.026
yles+0.025
ls+0.024
yle+0.023
 salon+0.023
Neg logit contributions
age-0.028
amed-0.023
 welcome-0.022
est-0.022
>-0.022
Token He
Token ID679
Feature activation+8.22e-03

Pos logit contributions
yles+0.022
cases+0.021
yle+0.020
 salon+0.020
ries+0.020
Neg logit contributions
age-0.009
 back-0.007
 welcome-0.007
 Follow-0.006
 right-0.006
Token w
Token ID266
Feature activation+6.92e-02

Pos logit contributions
yles+0.014
cases+0.014
yle+0.014
outing+0.013
 salon+0.012
Neg logit contributions
 back-0.013
age-0.013
 right-0.012
 Follow-0.012
 welcome-0.012
Tokenags
Token ID3775
Feature activation-9.38e-02

Pos logit contributions
yles+0.179
cases+0.177
yle+0.168
ls+0.168
 salon+0.166
Neg logit contributions
age-0.062
 back-0.042
 welcome-0.040
>-0.040
 home-0.039
Token his
Token ID465
Feature activation+2.68e-02

Pos logit contributions
age+0.126
 back+0.115
 home+0.108
 welcome+0.103
 right+0.101
Neg logit contributions
cases-0.201
yles-0.198
 salon-0.183
yle-0.182
ls-0.182
Token tail
Token ID7894
Feature activation-6.11e-03

Pos logit contributions
ls+0.046
cases+0.046
 salon+0.046
yles+0.045
 wra+0.045
Neg logit contributions
age-0.036
 Follow-0.031
wood-0.031
day-0.028
sy-0.028
Token.
Token ID13
Feature activation+8.64e-03

Pos logit contributions
age+0.011
>+0.009
 Follow+0.009
amed+0.009
 welcome+0.009
Neg logit contributions
cases-0.009
ls-0.008
yles-0.008
 salon-0.008
outing-0.008
Token
Token ID198
Feature activation-1.96e-03

Pos logit contributions
yles+0.020
cases+0.020
yle+0.019
 salon+0.019
ls+0.018
Neg logit contributions
age-0.010
 back-0.008
 welcome-0.008
>-0.008
 home-0.007
Token
Token ID198
Feature activation-2.01e-03

Pos logit contributions
age+0.003
est+0.002
 back+0.002
>+0.002
 welcome+0.002
Neg logit contributions
yles-0.004
cases-0.004
ls-0.004
yle-0.004
 salon-0.004