TOP ACTIVATIONS
MAX = 0.195

steps - and see what other y ummy food they could
y and scared . " Ah h ! This is not
just need to help each other ." B ert
time , saying , " Ah - ch oo !" He
. You can help in other ways , like picking flowers
! The other animals were sad and didn
, toys and lots of other things . At
forest . Who knows what other new adventures they would find
said . The other bear said , " L
brave , he even visited other ponds and me adows !
, he was at the other side of the field .
were more competitive with each other . They wanted to prove
together , always understanding each other . It was really sweet
't do it to any other sm elly skirts that you
and you will go the other way . How does that
ez ed again ! " Ah - ch oo !" He
you 'll have success in other things you try and sell
When he reached the other side , the girl said
from one foot to the other . He
. Soon , lots of other animals joined in , and

MIN ACTIVATIONS
MIN = -0.258

As they sn ugg led underneath the gear ,
to their room and sn ugg led in their beds .
you go , my d arl ings . Enjoy your lunch
?" Lily asked , sn ugg ling with her mom .
." They sn ugg led in their beds and
" Hello , my d arl ings . I 'm so
Don ny sn ugg led into his bed and
's wonderful , my d arl ings !" Mom my says
was over . He sn ugg led with them and listened
the cover . She sn ugg led with Mom and listened
and both of them sn ugg led together under a cozy
she pur red and sn ugg led up to her owner
name is Mr . Sn ugg les ." T
. Sn ugg les was so sad to
That night , Julie sn ugg led into her own soft
asleep . Fl uffy sn ugg led with her and pur
on the bed and sn ugg led together . They said
L ola smiled and sn ugg led in the warmth of
- just perfect for sn ugg ling . The baby lamb
his painting supplies , sn ugg led up in his bed

INTERVAL 0.081 - 0.195
CONTAINS 1.721%

lex ible . But then they had an idea
reward for studying was a massive plate of her favourite pasta
, you are brave . But you should not climb the
three , four ," she counts . " Look
ie , who wanted to reach a jar of cookies that

INTERVAL -0.032 - 0.081
CONTAINS 81.584%

mom smiled and said , " That 's a beautiful drawing
smile on her face . Lily loved to comb Anna 's
He ran to the kitchen and grabbed some lot ion .
hear the ball and the sticks outside . They wonder who
a monkey play with a ball . She cl apped and

INTERVAL -0.145 - -0.032
CONTAINS 16.582%

. They do not care about Dad . Mom
girl who liked to play with her toys . She had
" I 'm sorry , too . I was en vious
?" L ily thought for a minute , and
, Emily . Emily always put a leash on Max before

INTERVAL -0.258 - -0.145
CONTAINS 0.113%

She was surprised to see how much fun she had ,
Suddenly , Snow y said " My turn
flowers and hugs and said how wonderful she was . Lily
wave his arms to show how proud he was .
and were very careful with how they wrapped it .
Token steps
Token ID4831
Feature activation+2.79e-02

Pos logit contributions
pieces+0.126
Maria+0.120
od+0.119
iece+0.114
lar+0.110
Neg logit contributions
 cooler-0.079
robe-0.072
arl-0.070
melon-0.068
 Cream-0.064
Token -
Token ID532
Feature activation-2.02e-02

Pos logit contributions
pieces+0.046
od+0.041
Maria+0.040
one+0.036
enta+0.036
Neg logit contributions
robe-0.050
arl-0.046
melon-0.044
 cooler-0.043
ugg-0.042
Token and
Token ID290
Feature activation+1.90e-02

Pos logit contributions
robe+0.032
 cooler+0.029
arl+0.029
melon+0.029
ugg+0.027
Neg logit contributions
pieces-0.036
od-0.033
Maria-0.032
inda-0.030
lar-0.029
Token see
Token ID766
Feature activation+1.01e-01

Pos logit contributions
pieces+0.034
od+0.032
Maria+0.031
one+0.029
lier+0.029
Neg logit contributions
robe-0.028
 cooler-0.026
melon-0.026
arl-0.024
ugg-0.023
Token what
Token ID644
Feature activation-4.91e-02

Pos logit contributions
pieces+0.200
od+0.182
Maria+0.181
ata+0.170
enta+0.170
Neg logit contributions
robe-0.137
 cooler-0.128
melon-0.121
arl-0.119
 cabin-0.115
Token other
Token ID584
Feature activation+1.95e-01

Pos logit contributions
robe+0.071
 cooler+0.064
melon+0.063
 Cream+0.059
 Sleep+0.058
Neg logit contributions
pieces-0.094
od-0.083
Maria-0.081
ata-0.077
enta-0.077
Token y
Token ID331
Feature activation+1.36e-02

Pos logit contributions
pieces+0.374
od+0.341
Maria+0.337
enta+0.320
ata+0.319
Neg logit contributions
robe-0.268
 cooler-0.255
melon-0.236
 cabin-0.220
 Cream-0.218
Tokenummy
Token ID13513
Feature activation+1.05e-02

Pos logit contributions
pieces+0.027
od+0.025
Maria+0.024
lier+0.024
ill+0.023
Neg logit contributions
robe-0.020
arl-0.017
ugg-0.015
 cooler-0.015
 ranc-0.015
Token food
Token ID2057
Feature activation+9.78e-03

Pos logit contributions
pieces+0.021
Maria+0.019
od+0.019
enta+0.018
inda+0.017
Neg logit contributions
robe-0.014
 cooler-0.014
melon-0.013
 Cream-0.012
arl-0.012
Token they
Token ID484
Feature activation+3.04e-02

Pos logit contributions
pieces+0.017
od+0.015
Maria+0.015
enta+0.013
one+0.013
Neg logit contributions
robe-0.017
melon-0.015
 cooler-0.015
arl-0.015
ugg-0.014
Token could
Token ID714
Feature activation+1.64e-02

Pos logit contributions
pieces+0.056
Maria+0.049
od+0.048
one+0.047
 circles+0.046
Neg logit contributions
robe-0.047
melon-0.043
 cooler-0.042
arl-0.040
ugg-0.039
Tokeny
Token ID88
Feature activation-2.15e-02

Pos logit contributions
robe+0.011
arl+0.011
melon+0.010
 cooler+0.009
ugg+0.009
Neg logit contributions
pieces-0.008
od-0.007
Maria-0.007
enta-0.006
one-0.006
Token and
Token ID290
Feature activation+6.73e-03

Pos logit contributions
robe+0.040
arl+0.037
 cooler+0.035
melon+0.035
ugg+0.033
Neg logit contributions
pieces-0.034
Maria-0.030
od-0.030
enta-0.027
one-0.026
Token scared
Token ID12008
Feature activation-3.10e-03

Pos logit contributions
pieces+0.013
od+0.012
 circles+0.011
Maria+0.011
lier+0.011
Neg logit contributions
robe-0.010
arl-0.009
melon-0.008
ugg-0.008
 Cream-0.008
Token.
Token ID13
Feature activation+1.41e-02

Pos logit contributions
robe+0.004
arl+0.004
 cooler+0.004
melon+0.004
 sorry+0.003
Neg logit contributions
pieces-0.006
Maria-0.005
enta-0.005
od-0.005
lar-0.005
Token "
Token ID366
Feature activation+1.75e-03

Pos logit contributions
pieces+0.028
od+0.025
Maria+0.024
lier+0.023
one+0.023
Neg logit contributions
robe-0.021
arl-0.019
ugg-0.017
 cooler-0.017
melon-0.017
TokenAh
Token ID10910
Feature activation+1.94e-01

Pos logit contributions
pieces+0.003
od+0.003
Maria+0.003
lier+0.003
inda+0.002
Neg logit contributions
robe-0.003
arl-0.003
ugg-0.002
 cooler-0.002
 Cream-0.002
Tokenh
Token ID71
Feature activation+2.88e-03

Pos logit contributions
pieces+0.332
Maria+0.299
od+0.267
Rachel+0.257
 created+0.244
Neg logit contributions
arl-0.320
robe-0.300
 cooler-0.285
ugg-0.284
 sorry-0.274
Token!
Token ID0
Feature activation+3.34e-02

Pos logit contributions
pieces+0.004
od+0.004
Maria+0.004
lier+0.003
one+0.003
Neg logit contributions
robe-0.006
arl-0.005
melon-0.005
 cooler-0.005
ugg-0.005
Token This
Token ID770
Feature activation-1.14e-03

Pos logit contributions
pieces+0.061
od+0.055
Maria+0.053
lier+0.049
one+0.049
Neg logit contributions
robe-0.054
arl-0.049
melon-0.046
 cooler-0.045
ugg-0.045
Token is
Token ID318
Feature activation-7.49e-04

Pos logit contributions
robe+0.002
 cooler+0.001
arl+0.001
melon+0.001
 mall+0.001
Neg logit contributions
pieces-0.002
Maria-0.002
od-0.002
lar-0.002
inda-0.002
Token not
Token ID407
Feature activation-8.19e-02

Pos logit contributions
robe+0.001
arl+0.001
melon+0.001
ugg+0.001
 cooler+0.001
Neg logit contributions
pieces-0.002
od-0.001
Maria-0.001
one-0.001
lier-0.001
Token just
Token ID655
Feature activation+5.03e-02

Pos logit contributions
pieces+0.064
od+0.056
Maria+0.056
lier+0.050
enta+0.050
Neg logit contributions
robe-0.060
arl-0.054
 cooler-0.052
melon-0.050
ugg-0.049
Token need
Token ID761
Feature activation+5.34e-02

Pos logit contributions
pieces+0.099
Maria+0.090
od+0.089
Rachel+0.081
lier+0.080
Neg logit contributions
robe-0.072
arl-0.065
melon-0.063
 cooler-0.062
 Sleep-0.060
Token to
Token ID284
Feature activation+4.00e-02

Pos logit contributions
pieces+0.118
od+0.109
Maria+0.102
iece+0.099
ill+0.096
Neg logit contributions
robe-0.064
arl-0.061
 cooler-0.059
 Sleep-0.059
ugg-0.052
Token help
Token ID1037
Feature activation+1.41e-02

Pos logit contributions
pieces+0.067
od+0.060
Maria+0.058
one+0.053
lier+0.053
Neg logit contributions
robe-0.070
arl-0.065
 cooler-0.061
melon-0.060
ugg-0.058
Token each
Token ID1123
Feature activation-2.49e-02

Pos logit contributions
pieces+0.028
od+0.025
Maria+0.024
 rect+0.023
ez+0.023
Neg logit contributions
robe-0.019
arl-0.017
 cooler-0.016
 Sleep-0.016
 Cream-0.015
Token other
Token ID584
Feature activation+1.94e-01

Pos logit contributions
robe+0.032
 pack+0.029
 cooler+0.029
 waitress+0.029
 festival+0.028
Neg logit contributions
pieces-0.045
Maria-0.044
od-0.044
enta-0.043
ill-0.043
Token."
Token ID526
Feature activation+2.60e-03

Pos logit contributions
pieces+0.318
od+0.282
Maria+0.279
enta+0.250
one+0.247
Neg logit contributions
robe-0.349
arl-0.314
 cooler-0.301
melon-0.295
ugg-0.282
Token
Token ID198
Feature activation+3.90e-02

Pos logit contributions
pieces+0.006
lier+0.005
ets+0.005
ata+0.005
lar+0.005
Neg logit contributions
robe-0.003
ugg-0.003
arl-0.003
 Cream-0.003
melon-0.003
Token
Token ID198
Feature activation+3.84e-02

Pos logit contributions
pieces+0.084
lier+0.073
od+0.071
ets+0.071
asing+0.069
Neg logit contributions
robe-0.051
arl-0.046
ugg-0.043
 Cream-0.039
melon-0.038
TokenB
Token ID33
Feature activation+6.50e-04

Pos logit contributions
pieces+0.062
ets+0.054
lier+0.052
od+0.050
 pad+0.050
Neg logit contributions
robe-0.065
arl-0.059
ugg-0.058
 cabin-0.056
 ranc-0.055
Tokenert
Token ID861
Feature activation+1.52e-01

Pos logit contributions
pieces+0.001
od+0.001
Maria+0.001
lier+0.001
 minutes+0.001
Neg logit contributions
robe-0.001
arl-0.001
ugg-0.001
melon-0.001
 Cream-0.001
Token time
Token ID640
Feature activation+5.02e-02

Pos logit contributions
pieces+0.121
Maria+0.110
od+0.108
enta+0.101
asing+0.098
Neg logit contributions
robe-0.094
 cooler-0.088
arl-0.086
melon-0.081
ugg-0.075
Token,
Token ID11
Feature activation+6.38e-03

Pos logit contributions
pieces+0.094
Maria+0.085
od+0.079
enta+0.075
Rachel+0.075
Neg logit contributions
robe-0.077
arl-0.069
 cooler-0.069
melon-0.068
ugg-0.063
Token saying
Token ID2282
Feature activation+3.56e-02

Pos logit contributions
pieces+0.011
od+0.010
Maria+0.009
lier+0.008
one+0.008
Neg logit contributions
robe-0.011
 cooler-0.010
arl-0.010
melon-0.010
ugg-0.009
Token,
Token ID11
Feature activation+6.94e-03

Pos logit contributions
pieces+0.071
od+0.064
Maria+0.062
lier+0.058
one+0.058
Neg logit contributions
robe-0.052
arl-0.047
melon-0.043
 cooler-0.043
ugg-0.042
Token "
Token ID366
Feature activation+5.56e-03

Pos logit contributions
pieces+0.015
od+0.013
Maria+0.013
lier+0.012
one+0.012
Neg logit contributions
robe-0.009
arl-0.008
melon-0.008
 cooler-0.008
ugg-0.007
TokenAh
Token ID10910
Feature activation+1.94e-01

Pos logit contributions
pieces+0.011
od+0.010
Maria+0.009
ets+0.009
inda+0.009
Neg logit contributions
robe-0.008
arl-0.007
ugg-0.007
 Sleep-0.007
 Cream-0.007
Token-
Token ID12
Feature activation-1.23e-02

Pos logit contributions
pieces+0.290
Maria+0.264
od+0.239
Rachel+0.225
iece+0.219
Neg logit contributions
arl-0.333
robe-0.323
ugg-0.321
 cooler-0.307
melon-0.293
Tokench
Token ID354
Feature activation+3.79e-02

Pos logit contributions
robe+0.024
ugg+0.023
arl+0.023
 Cream+0.021
melon+0.021
Neg logit contributions
pieces-0.017
od-0.015
Maria-0.015
 circles-0.013
enta-0.013
Tokenoo
Token ID2238
Feature activation-2.50e-02

Pos logit contributions
pieces+0.068
od+0.061
Maria+0.059
one+0.055
lier+0.054
Neg logit contributions
robe-0.062
arl-0.056
 cooler-0.053
melon-0.052
ugg-0.051
Token!"
Token ID2474
Feature activation+2.22e-02

Pos logit contributions
robe+0.055
arl+0.051
melon+0.048
 cooler+0.048
ugg+0.048
Neg logit contributions
pieces-0.031
od-0.027
Maria-0.025
lier-0.023
one-0.023
Token He
Token ID679
Feature activation-1.21e-02

Pos logit contributions
pieces+0.045
od+0.041
Maria+0.040
enta+0.038
lar+0.038
Neg logit contributions
robe-0.032
arl-0.028
 cooler-0.027
ugg-0.027
melon-0.026
Token.
Token ID13
Feature activation+2.42e-02

Pos logit contributions
pieces+0.038
od+0.033
Maria+0.032
one+0.028
lier+0.028
Neg logit contributions
robe-0.049
arl-0.046
 cooler-0.043
melon-0.043
ugg-0.042
Token You
Token ID921
Feature activation+3.51e-02

Pos logit contributions
pieces+0.040
od+0.035
Maria+0.034
one+0.032
lier+0.031
Neg logit contributions
robe-0.043
arl-0.040
 cooler-0.037
melon-0.037
ugg-0.036
Token can
Token ID460
Feature activation+1.92e-02

Pos logit contributions
pieces+0.057
od+0.051
Maria+0.049
lier+0.045
one+0.044
Neg logit contributions
robe-0.064
arl-0.059
 cooler-0.055
melon-0.055
ugg-0.054
Token help
Token ID1037
Feature activation+1.32e-02

Pos logit contributions
pieces+0.034
od+0.031
Maria+0.030
lier+0.027
one+0.027
Neg logit contributions
robe-0.032
arl-0.029
melon-0.027
 cooler-0.027
ugg-0.027
Token in
Token ID287
Feature activation+1.84e-02

Pos logit contributions
pieces+0.029
od+0.025
Maria+0.025
ata+0.024
iece+0.024
Neg logit contributions
robe-0.015
arl-0.014
 cooler-0.013
 Cream-0.013
 Sleep-0.013
Token other
Token ID584
Feature activation+1.94e-01

Pos logit contributions
pieces+0.036
od+0.034
Maria+0.033
ata+0.031
ill+0.030
Neg logit contributions
robe-0.025
 cooler-0.023
arl-0.023
 cabin-0.021
 Sleep-0.021
Token ways
Token ID2842
Feature activation+3.46e-02

Pos logit contributions
pieces+0.428
ata+0.398
ez+0.392
od+0.388
Maria+0.376
Neg logit contributions
 cabin-0.187
 old-0.170
 refrigerator-0.166
 parents-0.160
 crib-0.158
Token,
Token ID11
Feature activation+5.36e-03

Pos logit contributions
pieces+0.069
Maria+0.060
od+0.060
Rachel+0.055
ata+0.055
Neg logit contributions
robe-0.052
arl-0.048
melon-0.043
ugg-0.043
 Sleep-0.043
Token like
Token ID588
Feature activation-1.73e-03

Pos logit contributions
pieces+0.008
od+0.008
Maria+0.007
enta+0.007
one+0.007
Neg logit contributions
robe-0.010
arl-0.009
 cooler-0.009
melon-0.009
 Sleep-0.008
Token picking
Token ID10868
Feature activation+2.87e-02

Pos logit contributions
robe+0.002
 cooler+0.002
arl+0.002
melon+0.002
 Sleep+0.002
Neg logit contributions
pieces-0.003
od-0.003
Maria-0.003
ill-0.003
ata-0.003
Token flowers
Token ID12734
Feature activation+9.59e-02

Pos logit contributions
pieces+0.061
Maria+0.056
od+0.054
one+0.051
enta+0.051
Neg logit contributions
robe-0.037
 cooler-0.034
arl-0.032
melon-0.031
 Cream-0.030
Token!
Token ID0
Feature activation+3.01e-02

Pos logit contributions
pieces+0.012
od+0.010
Maria+0.010
enta+0.009
ata+0.009
Neg logit contributions
robe-0.011
arl-0.010
 cooler-0.010
melon-0.010
ugg-0.009
Token
Token ID220
Feature activation-1.37e-02

Pos logit contributions
pieces+0.063
od+0.057
Maria+0.053
enta+0.052
asing+0.052
Neg logit contributions
robe-0.042
arl-0.036
ugg-0.034
 Cream-0.034
melon-0.034
Token
Token ID198
Feature activation+3.98e-02

Pos logit contributions
robe+0.022
arl+0.020
 cooler+0.018
ugg+0.018
melon+0.018
Neg logit contributions
pieces-0.026
od-0.023
Maria-0.022
lier-0.021
enta-0.020
Token
Token ID198
Feature activation+3.95e-02

Pos logit contributions
pieces+0.080
od+0.070
 rect+0.065
ets+0.065
lier+0.065
Neg logit contributions
robe-0.056
arl-0.054
ugg-0.049
 Cream-0.048
 cooler-0.046
TokenThe
Token ID464
Feature activation+2.05e-02

Pos logit contributions
pieces+0.065
od+0.060
ets+0.052
enta+0.051
ples+0.051
Neg logit contributions
robe-0.066
arl-0.063
ugg-0.060
 cooler-0.058
 Cream-0.058
Token other
Token ID584
Feature activation+1.94e-01

Pos logit contributions
pieces+0.033
Maria+0.028
 circles+0.028
od+0.028
 hours+0.027
Neg logit contributions
robe-0.036
arl-0.034
ugg-0.033
melon-0.031
 Cream-0.030
Token animals
Token ID4695
Feature activation+4.27e-02

Pos logit contributions
pieces+0.363
od+0.326
Maria+0.313
enta+0.288
lier+0.288
Neg logit contributions
robe-0.309
arl-0.276
 cooler-0.266
melon-0.259
ugg-0.252
Token were
Token ID547
Feature activation+4.39e-02

Pos logit contributions
pieces+0.077
od+0.068
Maria+0.066
enta+0.066
 rect+0.064
Neg logit contributions
robe-0.063
melon-0.060
 cooler-0.057
ugg-0.054
 tired-0.053
Token sad
Token ID6507
Feature activation-4.38e-02

Pos logit contributions
pieces+0.080
od+0.072
Maria+0.070
lier+0.065
one+0.064
Neg logit contributions
robe-0.073
arl-0.066
ugg-0.061
melon-0.059
 cooler-0.058
Token and
Token ID290
Feature activation+1.30e-02

Pos logit contributions
robe+0.072
arl+0.067
 cooler+0.065
melon+0.064
ugg+0.060
Neg logit contributions
pieces-0.080
Maria-0.068
od-0.068
enta-0.064
ata-0.062
Token didn
Token ID1422
Feature activation-4.95e-02

Pos logit contributions
pieces+0.022
od+0.019
Maria+0.019
lier+0.017
one+0.017
Neg logit contributions
robe-0.023
arl-0.021
ugg-0.019
 cooler-0.019
melon-0.019
Token,
Token ID11
Feature activation+3.89e-03

Pos logit contributions
pieces+0.001
od+0.001
Maria+0.001
one+0.001
lier+0.001
Neg logit contributions
robe-0.001
arl-0.001
 cooler-0.001
melon-0.001
ugg-0.001
Token toys
Token ID14958
Feature activation+1.69e-02

Pos logit contributions
pieces+0.007
od+0.007
Maria+0.007
one+0.006
lier+0.006
Neg logit contributions
robe-0.006
 cooler-0.005
arl-0.005
melon-0.005
ugg-0.005
Token and
Token ID290
Feature activation+1.67e-02

Pos logit contributions
pieces+0.028
od+0.024
Maria+0.024
lier+0.021
one+0.021
Neg logit contributions
robe-0.031
arl-0.028
 cooler-0.027
melon-0.027
ugg-0.026
Token lots
Token ID6041
Feature activation+4.29e-02

Pos logit contributions
pieces+0.033
od+0.031
Maria+0.030
ill+0.029
asing+0.029
Neg logit contributions
robe-0.023
 cooler-0.021
melon-0.018
 ranc-0.018
 mall-0.018
Token of
Token ID286
Feature activation+2.81e-02

Pos logit contributions
pieces+0.065
Maria+0.060
od+0.058
Rachel+0.055
enta+0.054
Neg logit contributions
robe-0.077
 cooler-0.073
arl-0.069
melon-0.069
ugg-0.066
Token other
Token ID584
Feature activation+1.93e-01

Pos logit contributions
pieces+0.058
od+0.055
Maria+0.055
enta+0.053
one+0.053
Neg logit contributions
robe-0.033
 cooler-0.031
 mall-0.028
 cabin-0.027
 ranc-0.026
Token things
Token ID1243
Feature activation-1.90e-02

Pos logit contributions
pieces+0.401
Maria+0.371
od+0.367
Rachel+0.351
one+0.348
Neg logit contributions
robe-0.241
 cooler-0.226
 cabin-0.215
 ranc-0.193
 mall-0.190
Token.
Token ID13
Feature activation+2.79e-02

Pos logit contributions
robe+0.032
arl+0.028
 cooler+0.028
melon+0.027
ugg+0.026
Neg logit contributions
pieces-0.033
od-0.029
Maria-0.029
one-0.027
enta-0.027
Token
Token ID198
Feature activation+4.17e-02

Pos logit contributions
pieces+0.057
od+0.054
asing+0.049
enta+0.049
 rect+0.049
Neg logit contributions
robe-0.037
arl-0.032
 Sleep-0.030
melon-0.030
ugg-0.030
Token
Token ID198
Feature activation+4.11e-02

Pos logit contributions
pieces+0.091
od+0.084
 rect+0.083
asing+0.080
ets+0.080
Neg logit contributions
robe-0.046
arl-0.044
ugg-0.041
 Sleep-0.038
melon-0.035
TokenAt
Token ID2953
Feature activation+4.72e-02

Pos logit contributions
pieces+0.066
od+0.064
ets+0.058
 rect+0.057
iece+0.057
Neg logit contributions
robe-0.062
arl-0.060
ugg-0.058
 Sleep-0.057
melon-0.054
Token forest
Token ID8222
Feature activation-1.07e-02

Pos logit contributions
pieces+0.013
od+0.012
Maria+0.011
lier+0.010
one+0.010
Neg logit contributions
robe-0.008
arl-0.007
 cooler-0.007
melon-0.007
ugg-0.006
Token.
Token ID13
Feature activation+2.64e-02

Pos logit contributions
robe+0.017
 cooler+0.015
melon+0.015
 cabin+0.014
arl+0.013
Neg logit contributions
pieces-0.019
Maria-0.017
od-0.017
ez-0.017
one-0.016
Token Who
Token ID5338
Feature activation+1.00e-01

Pos logit contributions
pieces+0.055
od+0.052
asing+0.049
lar+0.048
ez+0.048
Neg logit contributions
robe-0.033
arl-0.029
melon-0.028
 Cream-0.027
ugg-0.027
Token knows
Token ID4206
Feature activation+5.94e-02

Pos logit contributions
pieces+0.175
Maria+0.142
od+0.139
 circles+0.134
enta+0.133
Neg logit contributions
robe-0.166
 cooler-0.144
arl-0.137
 Sleep-0.137
ugg-0.137
Token what
Token ID644
Feature activation-4.45e-02

Pos logit contributions
pieces+0.124
od+0.110
Maria+0.109
ata+0.104
ez+0.104
Neg logit contributions
robe-0.077
 cooler-0.068
arl-0.065
melon-0.064
 Sleep-0.063
Token other
Token ID584
Feature activation+1.93e-01

Pos logit contributions
robe+0.057
 cooler+0.051
 Sleep+0.050
 cabin+0.049
melon+0.048
Neg logit contributions
pieces-0.091
iece-0.079
od-0.078
ata-0.078
ill-0.078
Token new
Token ID649
Feature activation+9.28e-02

Pos logit contributions
pieces+0.370
od+0.332
Maria+0.327
ata+0.313
enta+0.309
Neg logit contributions
robe-0.278
 cooler-0.252
melon-0.236
arl-0.234
 cabin-0.226
Token adventures
Token ID17545
Feature activation-5.18e-02

Pos logit contributions
pieces+0.187
od+0.169
Maria+0.164
ata+0.159
enta+0.157
Neg logit contributions
robe-0.130
 cooler-0.115
melon-0.106
arl-0.106
 cabin-0.105
Token they
Token ID484
Feature activation+3.28e-02

Pos logit contributions
robe+0.088
arl+0.077
 cooler+0.077
melon+0.077
ugg+0.070
Neg logit contributions
pieces-0.090
od-0.079
Maria-0.079
enta-0.071
one-0.071
Token would
Token ID561
Feature activation+1.20e-02

Pos logit contributions
pieces+0.057
Maria+0.050
od+0.049
one+0.046
enta+0.046
Neg logit contributions
robe-0.054
 cooler-0.048
arl-0.047
melon-0.047
 Sleep-0.044
Token find
Token ID1064
Feature activation+4.90e-02

Pos logit contributions
pieces+0.023
Maria+0.020
od+0.020
ata+0.019
enta+0.019
Neg logit contributions
robe-0.017
arl-0.016
 cooler-0.016
melon-0.015
 Sleep-0.014
Token said
Token ID531
Feature activation+3.28e-02

Pos logit contributions
pieces+0.043
one+0.036
 conclusions+0.036
od+0.034
iece+0.034
Neg logit contributions
robe-0.047
 cooler-0.042
melon-0.040
 bar-0.038
 Cream-0.038
Token.
Token ID13
Feature activation+1.91e-02

Pos logit contributions
pieces+0.054
od+0.047
Maria+0.046
one+0.042
enta+0.042
Neg logit contributions
robe-0.060
arl-0.054
melon-0.051
 cooler-0.051
ugg-0.050
Token
Token ID198
Feature activation+2.96e-02

Pos logit contributions
pieces+0.033
od+0.030
Maria+0.029
lier+0.027
one+0.027
Neg logit contributions
robe-0.032
arl-0.029
melon-0.027
 cooler-0.027
ugg-0.027
Token
Token ID198
Feature activation+2.89e-02

Pos logit contributions
pieces+0.057
od+0.051
Maria+0.048
lier+0.047
one+0.047
Neg logit contributions
robe-0.046
arl-0.042
ugg-0.038
 cooler-0.037
melon-0.037
TokenThe
Token ID464
Feature activation+1.76e-02

Pos logit contributions
pieces+0.040
od+0.034
Maria+0.033
lier+0.029
one+0.029
Neg logit contributions
robe-0.061
arl-0.056
melon-0.053
 cooler-0.053
ugg-0.052
Token other
Token ID584
Feature activation+1.93e-01

Pos logit contributions
pieces+0.030
od+0.026
Maria+0.025
 circles+0.024
lier+0.023
Neg logit contributions
robe-0.030
arl-0.029
ugg-0.027
melon-0.026
 cooler-0.025
Token bear
Token ID6842
Feature activation-2.58e-02

Pos logit contributions
pieces+0.369
od+0.334
Maria+0.327
one+0.300
ill+0.299
Neg logit contributions
robe-0.300
 cooler-0.253
arl-0.252
melon-0.247
ugg-0.236
Token said
Token ID531
Feature activation+2.98e-02

Pos logit contributions
robe+0.040
melon+0.037
 cooler+0.036
ugg+0.035
 Cream+0.032
Neg logit contributions
pieces-0.046
one-0.039
 conclusions-0.039
Maria-0.039
od-0.038
Token,
Token ID11
Feature activation+2.94e-03

Pos logit contributions
pieces+0.051
od+0.046
Maria+0.044
lier+0.041
one+0.040
Neg logit contributions
robe-0.052
arl-0.048
 cooler-0.044
melon-0.044
ugg-0.044
Token "
Token ID366
Feature activation-7.11e-03

Pos logit contributions
pieces+0.006
od+0.006
Maria+0.005
lier+0.005
one+0.005
Neg logit contributions
robe-0.004
arl-0.004
 cooler-0.003
melon-0.003
ugg-0.003
TokenL
Token ID43
Feature activation+3.56e-02

Pos logit contributions
robe+0.010
ugg+0.008
 Sleep+0.008
 Cream+0.008
arl+0.008
Neg logit contributions
pieces-0.015
od-0.013
Maria-0.012
 tal-0.012
ets-0.012
Token brave
Token ID14802
Feature activation+2.87e-02

Pos logit contributions
pieces+0.026
od+0.024
Maria+0.023
lier+0.021
one+0.021
Neg logit contributions
robe-0.024
arl-0.021
 cooler-0.020
melon-0.020
ugg-0.019
Token,
Token ID11
Feature activation+3.19e-03

Pos logit contributions
pieces+0.053
Maria+0.046
od+0.045
enta+0.041
ill+0.040
Neg logit contributions
robe-0.046
arl-0.042
 cooler-0.041
melon-0.040
ugg-0.038
Token he
Token ID339
Feature activation+2.53e-02

Pos logit contributions
pieces+0.006
od+0.006
Maria+0.006
enta+0.005
asing+0.005
Neg logit contributions
robe-0.005
 cooler-0.004
melon-0.004
arl-0.004
 Cream-0.004
Token even
Token ID772
Feature activation+6.62e-02

Pos logit contributions
pieces+0.043
od+0.037
Maria+0.036
enta+0.033
asing+0.033
Neg logit contributions
robe-0.044
 cooler-0.040
melon-0.038
ugg-0.036
 ranc-0.035
Token visited
Token ID8672
Feature activation-5.20e-02

Pos logit contributions
pieces+0.113
od+0.103
Maria+0.095
enta+0.090
INK+0.089
Neg logit contributions
robe-0.102
 cooler-0.099
melon-0.091
 tired-0.089
chard-0.086
Token other
Token ID584
Feature activation+1.93e-01

Pos logit contributions
robe+0.069
 cooler+0.059
melon+0.056
arl+0.054
 Cream+0.052
Neg logit contributions
pieces-0.108
od-0.099
Maria-0.097
enta-0.092
asing-0.092
Token ponds
Token ID47339
Feature activation+1.53e-03

Pos logit contributions
pieces+0.393
Maria+0.379
ez+0.360
od+0.355
enta+0.351
Neg logit contributions
 cooler-0.217
robe-0.216
 cabin-0.199
 ranc-0.191
 mall-0.180
Token and
Token ID290
Feature activation+1.89e-02

Pos logit contributions
pieces+0.003
od+0.002
Maria+0.002
enta+0.002
one+0.002
Neg logit contributions
robe-0.003
melon-0.002
 cooler-0.002
arl-0.002
ugg-0.002
Token me
Token ID502
Feature activation+3.66e-02

Pos logit contributions
pieces+0.033
od+0.031
lier+0.028
Maria+0.028
ill+0.027
Neg logit contributions
robe-0.030
 cooler-0.027
melon-0.026
 cabin-0.025
arl-0.024
Tokenadows
Token ID9797
Feature activation+5.17e-02

Pos logit contributions
pieces+0.059
od+0.052
Maria+0.050
 rect+0.049
one+0.045
Neg logit contributions
robe-0.068
ugg-0.057
arl-0.057
 cooler-0.056
 cabin-0.053
Token!
Token ID0
Feature activation+3.30e-02

Pos logit contributions
pieces+0.085
od+0.075
Maria+0.074
enta+0.067
one+0.066
Neg logit contributions
robe-0.093
arl-0.083
melon-0.081
 cooler-0.080
ugg-0.077
Token,
Token ID11
Feature activation+9.58e-03

Pos logit contributions
pieces+0.028
od+0.024
Maria+0.024
enta+0.022
 rect+0.022
Neg logit contributions
robe-0.023
 cooler-0.020
arl-0.019
melon-0.018
 cabin-0.018
Token he
Token ID339
Feature activation+3.16e-02

Pos logit contributions
pieces+0.018
od+0.017
Maria+0.016
lier+0.015
one+0.015
Neg logit contributions
robe-0.015
arl-0.013
 cooler-0.012
melon-0.012
ugg-0.012
Token was
Token ID373
Feature activation+3.64e-02

Pos logit contributions
pieces+0.056
od+0.047
Maria+0.046
enta+0.043
 rect+0.042
Neg logit contributions
robe-0.055
 cooler-0.050
melon-0.047
arl-0.044
ugg-0.044
Token at
Token ID379
Feature activation+8.77e-03

Pos logit contributions
pieces+0.069
od+0.062
Maria+0.060
lier+0.056
one+0.056
Neg logit contributions
robe-0.056
arl-0.050
 cooler-0.048
melon-0.047
ugg-0.045
Token the
Token ID262
Feature activation+9.91e-03

Pos logit contributions
pieces+0.017
od+0.015
Maria+0.015
lier+0.014
one+0.014
Neg logit contributions
robe-0.013
arl-0.012
 cooler-0.011
melon-0.011
ugg-0.011
Token other
Token ID584
Feature activation+1.93e-01

Pos logit contributions
pieces+0.022
od+0.020
Maria+0.020
enta+0.019
asing+0.018
Neg logit contributions
robe-0.012
 cooler-0.010
arl-0.010
melon-0.010
 cabin-0.009
Token side
Token ID1735
Feature activation+3.78e-02

Pos logit contributions
pieces+0.434
od+0.406
Maria+0.405
enta+0.386
ata+0.384
Neg logit contributions
robe-0.216
 cooler-0.198
 cabin-0.193
 mall-0.181
 refrigerator-0.177
Token of
Token ID286
Feature activation+2.71e-02

Pos logit contributions
pieces+0.062
od+0.056
Maria+0.056
enta+0.053
Rachel+0.049
Neg logit contributions
robe-0.063
melon-0.058
 cooler-0.058
arl-0.054
ugg-0.052
Token the
Token ID262
Feature activation+9.93e-03

Pos logit contributions
pieces+0.050
od+0.045
Maria+0.043
lier+0.041
one+0.040
Neg logit contributions
robe-0.044
arl-0.040
 cooler-0.038
melon-0.037
ugg-0.036
Token field
Token ID2214
Feature activation+2.02e-02

Pos logit contributions
pieces+0.022
od+0.021
Maria+0.020
ill+0.019
asing+0.019
Neg logit contributions
robe-0.012
 cooler-0.011
arl-0.011
melon-0.010
 cabin-0.010
Token.
Token ID13
Feature activation+3.26e-02

Pos logit contributions
pieces+0.035
od+0.031
Maria+0.031
enta+0.029
asing+0.027
Neg logit contributions
robe-0.035
melon-0.031
 cooler-0.031
arl-0.029
ugg-0.028
Token were
Token ID547
Feature activation+4.19e-02

Pos logit contributions
pieces+0.068
Maria+0.061
od+0.058
Rachel+0.055
ata+0.054
Neg logit contributions
robe-0.060
melon-0.058
 cooler-0.053
ugg-0.052
 cabin-0.049
Token more
Token ID517
Feature activation+6.31e-02

Pos logit contributions
pieces+0.073
Maria+0.066
od+0.066
ata+0.062
ill+0.061
Neg logit contributions
robe-0.063
 cooler-0.059
melon-0.059
arl-0.054
 tired-0.053
Token competitive
Token ID7606
Feature activation+6.88e-02

Pos logit contributions
pieces+0.141
Maria+0.126
one+0.124
od+0.123
lar+0.122
Neg logit contributions
 cooler-0.064
robe-0.062
 tired-0.058
arl-0.055
 thirsty-0.052
Token with
Token ID351
Feature activation-2.10e-02

Pos logit contributions
pieces+0.121
Maria+0.116
od+0.106
Jud+0.103
enta+0.103
Neg logit contributions
robe-0.103
 cooler-0.096
melon-0.085
 mall-0.084
 cabin-0.084
Token each
Token ID1123
Feature activation-2.70e-02

Pos logit contributions
robe+0.027
melon+0.025
 cooler+0.025
 Cream+0.023
arl+0.023
Neg logit contributions
pieces-0.041
od-0.037
Maria-0.036
ata-0.035
inda-0.035
Token other
Token ID584
Feature activation+1.92e-01

Pos logit contributions
 job+0.020
 bar+0.019
 well+0.019
 tight+0.018
 big+0.018
Neg logit contributions
Maria-0.058
ill-0.055
ples-0.054
older-0.053
light-0.053
Token.
Token ID13
Feature activation+2.98e-02

Pos logit contributions
pieces+0.321
Maria+0.298
od+0.286
enta+0.269
one+0.262
Neg logit contributions
robe-0.335
 cooler-0.280
melon-0.275
arl-0.274
 cabin-0.263
Token They
Token ID1119
Feature activation-1.10e-02

Pos logit contributions
pieces+0.065
od+0.058
inda+0.056
Maria+0.056
enta+0.055
Neg logit contributions
robe-0.034
 Cream-0.029
melon-0.029
arl-0.029
 Sleep-0.028
Token wanted
Token ID2227
Feature activation-1.87e-02

Pos logit contributions
robe+0.016
melon+0.015
 cooler+0.014
ugg+0.013
 tired+0.012
Neg logit contributions
pieces-0.020
Maria-0.018
od-0.018
one-0.017
lier-0.017
Token to
Token ID284
Feature activation+4.01e-02

Pos logit contributions
melon+0.018
 cooler+0.017
robe+0.017
 Cream+0.016
 not+0.016
Neg logit contributions
pieces-0.042
ata-0.037
inda-0.037
od-0.036
enta-0.036
Token prove
Token ID5879
Feature activation+6.73e-02

Pos logit contributions
pieces+0.076
Maria+0.069
od+0.068
one+0.065
ata+0.065
Neg logit contributions
robe-0.056
 cooler-0.050
melon-0.050
arl-0.050
 name-0.047
Token together
Token ID1978
Feature activation+4.34e-02

Pos logit contributions
robe+0.040
 cooler+0.036
melon+0.034
arl+0.033
 Cream+0.033
Neg logit contributions
pieces-0.048
Maria-0.044
od-0.044
asing-0.043
one-0.042
Token,
Token ID11
Feature activation-5.31e-03

Pos logit contributions
pieces+0.079
Maria+0.073
od+0.069
enta+0.068
one+0.067
Neg logit contributions
robe-0.067
melon-0.058
 cooler-0.057
arl-0.056
ugg-0.052
Token always
Token ID1464
Feature activation-1.11e-02

Pos logit contributions
robe+0.008
 cooler+0.008
melon+0.007
arl+0.007
 Sleep+0.007
Neg logit contributions
pieces-0.009
od-0.008
Maria-0.008
one-0.008
enta-0.008
Token understanding
Token ID4547
Feature activation+2.14e-02

Pos logit contributions
robe+0.016
 cooler+0.014
arl+0.014
melon+0.014
 Sleep+0.014
Neg logit contributions
pieces-0.021
od-0.020
Maria-0.019
ata-0.018
one-0.018
Token each
Token ID1123
Feature activation-2.92e-02

Pos logit contributions
pieces+0.041
od+0.037
Maria+0.037
enta+0.036
ata+0.035
Neg logit contributions
robe-0.029
arl-0.028
 cooler-0.026
melon-0.026
 cabin-0.024
Token other
Token ID584
Feature activation+1.92e-01

Pos logit contributions
 name+0.034
robe+0.034
 bar+0.034
 waitress+0.033
 refrigerator+0.033
Neg logit contributions
pieces-0.053
enta-0.052
od-0.051
Maria-0.051
CR-0.049
Token.
Token ID13
Feature activation+3.10e-02

Pos logit contributions
pieces+0.275
od+0.239
Maria+0.234
one+0.208
enta+0.208
Neg logit contributions
robe-0.384
arl-0.350
melon-0.333
 cooler-0.333
ugg-0.325
Token It
Token ID632
Feature activation+7.47e-03

Pos logit contributions
pieces+0.062
od+0.056
asing+0.053
enta+0.053
ced+0.053
Neg logit contributions
robe-0.043
arl-0.037
 Cream-0.036
ugg-0.036
melon-0.036
Token was
Token ID373
Feature activation+3.12e-02

Pos logit contributions
pieces+0.012
od+0.011
Maria+0.010
ill+0.009
enta+0.009
Neg logit contributions
robe-0.014
 cooler-0.012
melon-0.012
arl-0.011
 Sleep-0.011
Token really
Token ID1107
Feature activation+3.47e-02

Pos logit contributions
pieces+0.056
od+0.050
Maria+0.049
lier+0.045
one+0.045
Neg logit contributions
robe-0.051
arl-0.047
ugg-0.043
melon-0.043
 cooler-0.043
Token sweet
Token ID6029
Feature activation-2.20e-02

Pos logit contributions
pieces+0.072
od+0.066
Maria+0.064
lier+0.061
inda+0.059
Neg logit contributions
robe-0.047
arl-0.042
ugg-0.038
melon-0.038
 cooler-0.037
Token't
Token ID470
Feature activation-5.99e-03

Pos logit contributions
robe+0.076
arl+0.066
ugg+0.061
melon+0.061
 cooler+0.060
Neg logit contributions
pieces-0.117
od-0.108
Maria-0.103
one-0.099
lier-0.097
Token do
Token ID466
Feature activation+3.24e-02

Pos logit contributions
robe+0.009
arl+0.008
melon+0.008
 cooler+0.008
ugg+0.007
Neg logit contributions
pieces-0.011
od-0.010
Maria-0.010
enta-0.009
ata-0.009
Token it
Token ID340
Feature activation+9.22e-03

Pos logit contributions
pieces+0.065
Maria+0.059
od+0.059
ill+0.055
inda+0.054
Neg logit contributions
robe-0.044
arl-0.040
 cooler-0.038
melon-0.037
 Cream-0.035
Token to
Token ID284
Feature activation+3.63e-02

Pos logit contributions
pieces+0.015
Maria+0.014
od+0.013
Rachel+0.012
Jul+0.012
Neg logit contributions
robe-0.017
 cooler-0.015
arl-0.014
melon-0.014
ugg-0.013
Token any
Token ID597
Feature activation-3.67e-03

Pos logit contributions
pieces+0.060
od+0.053
Maria+0.051
one+0.047
lier+0.047
Neg logit contributions
robe-0.065
arl-0.060
melon-0.056
 cooler-0.056
ugg-0.055
Token other
Token ID584
Feature activation+1.92e-01

Pos logit contributions
 cooler+0.006
robe+0.006
arl+0.005
 crib+0.005
 dirty+0.005
Neg logit contributions
od-0.006
Maria-0.006
pieces-0.006
enta-0.005
ill-0.005
Token sm
Token ID895
Feature activation-1.21e-02

Pos logit contributions
Maria+0.372
pieces+0.352
od+0.349
ris+0.345
enta+0.343
Neg logit contributions
robe-0.244
 cooler-0.236
 mall-0.198
 crib-0.186
 dirty-0.184
Tokenelly
Token ID6148
Feature activation-9.16e-04

Pos logit contributions
robe+0.021
arl+0.020
melon+0.019
ugg+0.019
 cooler+0.018
Neg logit contributions
pieces-0.021
od-0.018
Maria-0.017
enta-0.016
one-0.016
Token skirts
Token ID42370
Feature activation+8.00e-03

Pos logit contributions
robe+0.002
 cooler+0.001
arl+0.001
melon+0.001
 mall+0.001
Neg logit contributions
pieces-0.002
Maria-0.002
od-0.002
enta-0.001
one-0.001
Token that
Token ID326
Feature activation-9.18e-02

Pos logit contributions
pieces+0.014
Maria+0.013
od+0.012
enta+0.011
Rachel+0.011
Neg logit contributions
robe-0.014
arl-0.012
melon-0.012
 cooler-0.012
ugg-0.011
Token you
Token ID345
Feature activation+5.79e-02

Pos logit contributions
robe+0.157
arl+0.135
 cooler+0.135
melon+0.135
ugg+0.126
Neg logit contributions
pieces-0.161
od-0.144
Maria-0.142
enta-0.129
lier-0.127
Token and
Token ID290
Feature activation+1.49e-02

Pos logit contributions
pieces+0.160
od+0.147
Maria+0.141
enta+0.134
ez+0.133
Neg logit contributions
robe-0.133
arl-0.126
 cooler-0.115
melon-0.111
ugg-0.107
Token you
Token ID345
Feature activation+5.54e-02

Pos logit contributions
pieces+0.026
od+0.023
Maria+0.022
lier+0.021
one+0.021
Neg logit contributions
robe-0.025
arl-0.023
 cooler-0.022
melon-0.022
ugg-0.021
Token will
Token ID481
Feature activation-2.82e-02

Pos logit contributions
pieces+0.090
od+0.080
Maria+0.077
lier+0.071
one+0.071
Neg logit contributions
robe-0.100
arl-0.092
melon-0.087
 cooler-0.086
ugg-0.085
Token go
Token ID467
Feature activation-2.72e-03

Pos logit contributions
robe+0.051
arl+0.046
melon+0.043
ugg+0.043
 cooler+0.043
Neg logit contributions
pieces-0.046
od-0.041
Maria-0.039
one-0.037
lier-0.037
Token the
Token ID262
Feature activation-5.43e-04

Pos logit contributions
robe+0.005
arl+0.004
 cooler+0.004
melon+0.004
ugg+0.004
Neg logit contributions
pieces-0.005
od-0.004
Maria-0.004
lier-0.004
enta-0.004
Token other
Token ID584
Feature activation+1.92e-01

Pos logit contributions
robe+0.001
arl+0.001
ugg+0.001
melon+0.001
 cooler+0.001
Neg logit contributions
pieces-0.001
od-0.001
Maria-0.001
one-0.001
lier-0.001
Token way
Token ID835
Feature activation+8.49e-02

Pos logit contributions
pieces+0.414
Maria+0.402
od+0.395
ez+0.369
inda+0.365
Neg logit contributions
robe-0.220
 cooler-0.220
 mall-0.204
 cabin-0.203
 festival-0.185
Token.
Token ID13
Feature activation+2.93e-02

Pos logit contributions
pieces+0.152
od+0.138
Maria+0.134
enta+0.123
ez+0.122
Neg logit contributions
robe-0.141
arl-0.130
 cooler-0.122
melon-0.119
ugg-0.115
Token How
Token ID1374
Feature activation-4.67e-02

Pos logit contributions
pieces+0.049
od+0.044
Maria+0.041
inda+0.038
enta+0.038
Neg logit contributions
robe-0.051
arl-0.048
 Sleep-0.044
ugg-0.044
 cooler-0.044
Token does
Token ID857
Feature activation-1.62e-02

Pos logit contributions
robe+0.070
arl+0.063
 cooler+0.061
ugg+0.056
 Cream+0.055
Neg logit contributions
pieces-0.089
Maria-0.080
od-0.078
enta-0.076
ez-0.074
Token that
Token ID326
Feature activation-9.34e-02

Pos logit contributions
robe+0.031
arl+0.029
ugg+0.027
melon+0.027
 cooler+0.024
Neg logit contributions
pieces-0.026
od-0.022
Maria-0.021
one-0.020
lier-0.020
Tokenez
Token ID8471
Feature activation+9.56e-03

Pos logit contributions
pieces+0.061
od+0.050
 rect+0.046
Maria+0.044
 whoever+0.041
Neg logit contributions
robe-0.090
ugg-0.090
arl-0.086
 Sleep-0.074
uggle-0.074
Tokened
Token ID276
Feature activation-3.55e-02

Pos logit contributions
pieces+0.019
Maria+0.016
od+0.016
enta+0.015
 rect+0.015
Neg logit contributions
robe-0.016
ugg-0.013
arl-0.013
 cooler-0.011
 Cream-0.011
Token again
Token ID757
Feature activation+3.30e-02

Pos logit contributions
robe+0.070
arl+0.063
 cooler+0.062
melon+0.061
ugg+0.058
Neg logit contributions
pieces-0.054
od-0.046
Maria-0.045
enta-0.040
one-0.040
Token!
Token ID0
Feature activation+2.60e-02

Pos logit contributions
pieces+0.069
Maria+0.061
od+0.060
one+0.058
enta+0.056
Neg logit contributions
robe-0.044
 cooler-0.042
arl-0.038
melon-0.037
 tired-0.036
Token "
Token ID366
Feature activation+1.68e-03

Pos logit contributions
pieces+0.050
od+0.045
Maria+0.043
lier+0.041
enta+0.041
Neg logit contributions
robe-0.040
arl-0.036
 cooler-0.033
melon-0.033
ugg-0.033
TokenAh
Token ID10910
Feature activation+1.92e-01

Pos logit contributions
pieces+0.003
od+0.003
Maria+0.003
ets+0.003
inda+0.003
Neg logit contributions
robe-0.002
arl-0.002
ugg-0.002
 Sleep-0.002
 Cream-0.002
Token-
Token ID12
Feature activation-1.37e-02

Pos logit contributions
pieces+0.304
Maria+0.275
od+0.248
Rachel+0.236
ill+0.232
Neg logit contributions
arl-0.315
ugg-0.305
robe-0.298
 cooler-0.287
 sorry-0.281
Tokench
Token ID354
Feature activation+3.62e-02

Pos logit contributions
ugg+0.026
robe+0.026
arl+0.025
melon+0.024
 Cream+0.023
Neg logit contributions
pieces-0.020
od-0.018
Maria-0.017
 circles-0.015
 rect-0.015
Tokenoo
Token ID2238
Feature activation-2.57e-02

Pos logit contributions
pieces+0.066
od+0.060
Maria+0.058
one+0.054
lier+0.054
Neg logit contributions
robe-0.058
arl-0.053
 cooler-0.049
melon-0.049
ugg-0.048
Token!"
Token ID2474
Feature activation+1.91e-02

Pos logit contributions
robe+0.058
arl+0.054
melon+0.051
 cooler+0.051
ugg+0.051
Neg logit contributions
pieces-0.030
od-0.026
Maria-0.024
lier-0.021
one-0.021
Token He
Token ID679
Feature activation-1.60e-02

Pos logit contributions
pieces+0.039
od+0.035
Maria+0.034
enta+0.033
one+0.032
Neg logit contributions
robe-0.027
arl-0.023
 cooler-0.023
melon-0.022
ugg-0.022
Token you
Token ID345
Feature activation+5.57e-02

Pos logit contributions
robe+0.022
arl+0.020
 cooler+0.019
ugg+0.019
melon+0.018
Neg logit contributions
pieces-0.025
od-0.022
Maria-0.022
enta-0.020
ata-0.020
Token'll
Token ID1183
Feature activation-1.22e-02

Pos logit contributions
pieces+0.093
Maria+0.084
od+0.082
 circles+0.073
 rect+0.073
Neg logit contributions
robe-0.099
arl-0.088
ugg-0.086
 cooler-0.086
melon-0.085
Token have
Token ID423
Feature activation+8.63e-02

Pos logit contributions
robe+0.022
arl+0.020
melon+0.019
 cooler+0.019
ugg+0.018
Neg logit contributions
pieces-0.020
od-0.018
Maria-0.017
lier-0.016
one-0.016
Token success
Token ID1943
Feature activation+4.16e-02

Pos logit contributions
pieces+0.182
od+0.176
Maria+0.171
iece+0.165
ill+0.161
Neg logit contributions
 cooler-0.102
robe-0.093
 Cream-0.091
 Sleep-0.089
 Rudolph-0.083
Token in
Token ID287
Feature activation+1.81e-02

Pos logit contributions
pieces+0.066
Maria+0.061
od+0.057
enta+0.052
one+0.051
Neg logit contributions
robe-0.076
arl-0.067
 cooler-0.065
 Cream-0.064
 Sleep-0.062
Token other
Token ID584
Feature activation+1.92e-01

Pos logit contributions
pieces+0.035
od+0.032
Maria+0.032
lier+0.029
ata+0.029
Neg logit contributions
robe-0.025
arl-0.024
 cooler-0.023
melon-0.022
 Sleep-0.022
Token things
Token ID1243
Feature activation-1.94e-02

Pos logit contributions
od+0.363
Maria+0.359
pieces+0.357
ata+0.345
ill+0.330
Neg logit contributions
 cooler-0.244
 cabin-0.227
robe-0.221
 mall-0.220
 crib-0.213
Token you
Token ID345
Feature activation+5.71e-02

Pos logit contributions
robe+0.036
arl+0.033
 cooler+0.032
melon+0.031
ugg+0.031
Neg logit contributions
pieces-0.031
Maria-0.027
od-0.027
enta-0.024
inda-0.023
Token try
Token ID1949
Feature activation+5.51e-02

Pos logit contributions
pieces+0.100
Maria+0.093
od+0.090
enta+0.080
Jul+0.080
Neg logit contributions
robe-0.093
arl-0.085
 cooler-0.082
melon-0.081
ugg-0.079
Token and
Token ID290
Feature activation+2.51e-02

Pos logit contributions
pieces+0.096
Maria+0.088
od+0.083
lar+0.080
inda+0.078
Neg logit contributions
robe-0.092
arl-0.080
 cooler-0.079
melon-0.076
 Cream-0.075
Token sell
Token ID3677
Feature activation-3.18e-02

Pos logit contributions
pieces+0.045
od+0.042
Maria+0.041
inda+0.037
ill+0.037
Neg logit contributions
robe-0.039
 cooler-0.035
arl-0.035
melon-0.035
ugg-0.033
Token
Token ID198
Feature activation+4.71e-02

Pos logit contributions
pieces+0.068
od+0.062
Maria+0.059
lier+0.056
one+0.056
Neg logit contributions
robe-0.052
arl-0.047
melon-0.043
 cooler-0.043
ugg-0.042
TokenWhen
Token ID2215
Feature activation-4.21e-02

Pos logit contributions
pieces+0.086
od+0.076
lier+0.069
asing+0.068
 rect+0.067
Neg logit contributions
robe-0.077
arl-0.073
ugg-0.067
 cooler-0.064
 Cream-0.063
Token he
Token ID339
Feature activation+3.00e-02

Pos logit contributions
robe+0.066
arl+0.059
 cooler+0.058
melon+0.057
ugg+0.054
Neg logit contributions
pieces-0.079
od-0.071
Maria-0.068
lier-0.064
one-0.063
Token reached
Token ID4251
Feature activation+6.05e-02

Pos logit contributions
pieces+0.065
Maria+0.054
asing+0.052
od+0.051
lier+0.049
Neg logit contributions
robe-0.040
 cooler-0.040
melon-0.036
ugg-0.034
 bar-0.034
Token the
Token ID262
Feature activation+1.96e-03

Pos logit contributions
pieces+0.117
od+0.104
Maria+0.100
lier+0.095
asing+0.093
Neg logit contributions
robe-0.094
arl-0.084
 cooler-0.083
melon-0.080
ugg-0.075
Token other
Token ID584
Feature activation+1.92e-01

Pos logit contributions
pieces+0.004
od+0.004
Maria+0.004
ez+0.004
enta+0.004
Neg logit contributions
robe-0.002
 cooler-0.002
arl-0.002
melon-0.002
 cabin-0.002
Token side
Token ID1735
Feature activation+3.72e-02

Pos logit contributions
pieces+0.473
Maria+0.469
od+0.457
ez+0.457
Rachel+0.434
Neg logit contributions
 cooler-0.163
robe-0.159
 cabin-0.139
melon-0.124
 mall-0.117
Token,
Token ID11
Feature activation+1.00e-03

Pos logit contributions
pieces+0.075
od+0.067
Maria+0.064
ez+0.064
enta+0.063
Neg logit contributions
robe-0.051
melon-0.047
 cooler-0.047
 cabin-0.041
arl-0.040
Token the
Token ID262
Feature activation+2.21e-03

Pos logit contributions
pieces+0.002
od+0.002
Maria+0.002
one+0.001
lier+0.001
Neg logit contributions
robe-0.002
arl-0.002
ugg-0.001
melon-0.001
 cooler-0.001
Token girl
Token ID2576
Feature activation+8.05e-02

Pos logit contributions
pieces+0.004
od+0.004
Maria+0.004
one+0.003
 circles+0.003
Neg logit contributions
robe-0.003
arl-0.003
ugg-0.003
melon-0.003
 cooler-0.003
Token said
Token ID531
Feature activation+3.22e-02

Pos logit contributions
pieces+0.151
od+0.126
Maria+0.124
 conclusions+0.121
asing+0.119
Neg logit contributions
 cooler-0.116
robe-0.116
 tired-0.107
ugg-0.103
melon-0.101
Token from
Token ID422
Feature activation-6.83e-02

Pos logit contributions
pieces+0.029
od+0.026
Maria+0.025
enta+0.023
one+0.023
Neg logit contributions
robe-0.026
arl-0.023
 cooler-0.023
melon-0.022
ugg-0.021
Token one
Token ID530
Feature activation+7.86e-03

Pos logit contributions
robe+0.089
 cooler+0.085
 cabin+0.077
melon+0.076
arl+0.075
Neg logit contributions
pieces-0.141
od-0.134
Maria-0.126
asing-0.122
ata-0.120
Token foot
Token ID2366
Feature activation+6.74e-02

Pos logit contributions
od+0.015
pieces+0.015
Maria+0.015
enta+0.015
ez+0.014
Neg logit contributions
 cooler-0.010
 cabin-0.009
robe-0.009
 crib-0.009
melon-0.009
Token to
Token ID284
Feature activation+4.19e-02

Pos logit contributions
pieces+0.124
Maria+0.114
od+0.112
enta+0.103
Rachel+0.102
Neg logit contributions
robe-0.107
 cooler-0.094
melon-0.092
arl-0.091
ugg-0.084
Token the
Token ID262
Feature activation+6.98e-03

Pos logit contributions
pieces+0.087
od+0.081
Maria+0.078
asing+0.075
lier+0.075
Neg logit contributions
robe-0.052
 cooler-0.049
arl-0.047
melon-0.046
 Sleep-0.042
Token other
Token ID584
Feature activation+1.92e-01

Pos logit contributions
pieces+0.014
od+0.013
Maria+0.013
enta+0.012
asing+0.012
Neg logit contributions
robe-0.010
 cooler-0.009
arl-0.009
melon-0.008
 cabin-0.008
Token.
Token ID13
Feature activation+2.49e-02

Pos logit contributions
pieces+0.425
Maria+0.390
od+0.379
enta+0.375
inda+0.359
Neg logit contributions
robe-0.227
 cooler-0.219
 cabin-0.215
 mall-0.199
 crib-0.189
Token
Token ID220
Feature activation-1.55e-02

Pos logit contributions
pieces+0.052
od+0.047
Maria+0.045
enta+0.043
lar+0.043
Neg logit contributions
robe-0.034
arl-0.030
melon-0.028
 cooler-0.028
 Cream-0.028
Token
Token ID198
Feature activation+3.88e-02

Pos logit contributions
robe+0.026
arl+0.023
 cooler+0.021
melon+0.021
ugg+0.021
Neg logit contributions
pieces-0.028
od-0.025
Maria-0.024
lier-0.022
enta-0.022
Token
Token ID198
Feature activation+3.82e-02

Pos logit contributions
pieces+0.081
od+0.071
asing+0.066
lier+0.066
 rect+0.066
Neg logit contributions
robe-0.052
arl-0.049
ugg-0.044
 Cream-0.044
 cooler-0.042
TokenHe
Token ID1544
Feature activation-2.43e-02

Pos logit contributions
pieces+0.062
od+0.057
Maria+0.049
enta+0.049
ets+0.049
Neg logit contributions
robe-0.065
arl-0.062
ugg-0.057
 Cream-0.057
 cooler-0.056
Token.
Token ID13
Feature activation+2.18e-02

Pos logit contributions
pieces+0.063
Maria+0.059
od+0.057
one+0.054
enta+0.054
Neg logit contributions
robe-0.058
 cooler-0.051
melon-0.049
arl-0.046
ugg-0.045
Token Soon
Token ID15894
Feature activation+9.38e-02

Pos logit contributions
pieces+0.044
od+0.039
lier+0.037
Maria+0.037
lar+0.036
Neg logit contributions
robe-0.032
arl-0.028
melon-0.027
ugg-0.026
 Cream-0.025
Token,
Token ID11
Feature activation+3.57e-03

Pos logit contributions
pieces+0.203
Maria+0.179
asing+0.172
od+0.171
enta+0.169
Neg logit contributions
 cooler-0.108
robe-0.101
melon-0.096
 tired-0.091
 Otto-0.088
Token lots
Token ID6041
Feature activation+4.22e-02

Pos logit contributions
pieces+0.008
od+0.007
lier+0.007
Maria+0.007
enta+0.007
Neg logit contributions
robe-0.004
 cooler-0.004
melon-0.004
 tired-0.004
 cabin-0.003
Token of
Token ID286
Feature activation+2.47e-02

Pos logit contributions
pieces+0.076
Maria+0.067
od+0.066
enta+0.065
Jul+0.062
Neg logit contributions
 cooler-0.061
robe-0.059
melon-0.058
arl-0.057
ugg-0.053
Token other
Token ID584
Feature activation+1.92e-01

Pos logit contributions
pieces+0.052
ill+0.051
od+0.051
one+0.048
Maria+0.048
Neg logit contributions
 cooler-0.026
robe-0.023
 old-0.022
 tired-0.022
 mall-0.021
Token animals
Token ID4695
Feature activation+4.14e-02

Pos logit contributions
pieces+0.416
Maria+0.397
od+0.385
Rachel+0.382
ill+0.376
Neg logit contributions
robe-0.188
 cooler-0.182
 old-0.171
 tired-0.166
 cabin-0.165
Token joined
Token ID5399
Feature activation-3.98e-02

Pos logit contributions
pieces+0.073
Maria+0.064
od+0.061
one+0.060
enta+0.058
Neg logit contributions
robe-0.065
melon-0.062
 cooler-0.059
arl-0.056
ugg-0.056
Token in
Token ID287
Feature activation+1.91e-02

Pos logit contributions
robe+0.040
melon+0.038
arl+0.036
 cooler+0.034
 name+0.031
Neg logit contributions
pieces-0.081
ata-0.078
one-0.077
enta-0.076
 mastered-0.075
Token,
Token ID11
Feature activation-1.80e-03

Pos logit contributions
pieces+0.036
od+0.033
Maria+0.032
one+0.032
asing+0.031
Neg logit contributions
robe-0.028
 cooler-0.025
arl-0.025
melon-0.024
 cabin-0.023
Token and
Token ID290
Feature activation+8.85e-03

Pos logit contributions
robe+0.003
melon+0.003
 cooler+0.003
arl+0.002
 tired+0.002
Neg logit contributions
pieces-0.003
od-0.003
Maria-0.003
one-0.003
lier-0.003
Token
Token ID198
Feature activation+5.97e-03

Pos logit contributions
robe+0.077
arl+0.069
melon+0.063
ugg+0.061
 cooler+0.058
Neg logit contributions
pieces-0.101
od-0.088
lar-0.087
enta-0.086
Maria-0.086
Token
Token ID198
Feature activation+7.66e-03

Pos logit contributions
pieces+0.012
od+0.011
ets+0.011
asing+0.011
 rect+0.011
Neg logit contributions
robe-0.008
arl-0.007
ugg-0.006
 Sleep-0.006
 Cream-0.006
TokenAs
Token ID1722
Feature activation-8.84e-02

Pos logit contributions
pieces+0.012
od+0.011
ets+0.010
enta+0.010
asing+0.010
Neg logit contributions
robe-0.013
arl-0.012
ugg-0.011
 Cream-0.011
 Sleep-0.011
Token they
Token ID484
Feature activation+3.04e-03

Pos logit contributions
robe+0.117
 cooler+0.106
melon+0.104
arl+0.102
ugg+0.097
Neg logit contributions
pieces-0.180
od-0.165
asing-0.155
enta-0.154
lier-0.153
Token sn
Token ID3013
Feature activation-6.10e-02

Pos logit contributions
pieces+0.005
od+0.004
Maria+0.004
one+0.004
lier+0.004
Neg logit contributions
robe-0.005
arl-0.005
 cooler-0.005
melon-0.005
ugg-0.005
Tokenugg
Token ID6837
Feature activation-2.23e-01

Pos logit contributions
robe+0.049
 ranc+0.036
arl+0.033
 waitress+0.032
 cooler+0.029
Neg logit contributions
pieces-0.152
od-0.134
Maria-0.134
ill-0.132
lier-0.131
Tokenled
Token ID992
Feature activation-4.27e-02

Pos logit contributions
robe+0.419
arl+0.396
melon+0.369
 cooler+0.367
ugg+0.363
Neg logit contributions
pieces-0.338
od-0.300
Maria-0.284
lier-0.264
one-0.255
Token underneath
Token ID14638
Feature activation+6.69e-02

Pos logit contributions
robe+0.069
arl+0.063
melon+0.058
 cooler+0.058
ugg+0.057
Neg logit contributions
pieces-0.078
od-0.070
Maria-0.068
one-0.063
lier-0.063
Token the
Token ID262
Feature activation-1.88e-02

Pos logit contributions
pieces+0.120
od+0.108
Maria+0.104
one+0.097
lier+0.096
Neg logit contributions
robe-0.112
arl-0.099
 cooler-0.096
melon-0.095
ugg-0.091
Token gear
Token ID7733
Feature activation-3.48e-02

Pos logit contributions
robe+0.026
arl+0.024
melon+0.021
ugg+0.021
 cooler+0.020
Neg logit contributions
pieces-0.037
od-0.033
Maria-0.033
 circles-0.032
lier-0.031
Token,
Token ID11
Feature activation-1.61e-02

Pos logit contributions
robe+0.060
arl+0.051
melon+0.051
 cooler+0.050
ugg+0.048
Neg logit contributions
pieces-0.063
od-0.055
Maria-0.054
enta-0.050
lier-0.050
Token to
Token ID284
Feature activation+1.54e-02

Pos logit contributions
robe+0.067
arl+0.062
 Sleep+0.061
melon+0.061
 Cream+0.059
Neg logit contributions
pieces-0.032
 circles-0.027
 minutes-0.024
Maria-0.024
od-0.023
Token their
Token ID511
Feature activation-1.62e-02

Pos logit contributions
pieces+0.030
od+0.027
Maria+0.026
one+0.025
lier+0.025
Neg logit contributions
robe-0.023
arl-0.021
 cooler-0.020
melon-0.019
ugg-0.019
Token room
Token ID2119
Feature activation-6.94e-02

Pos logit contributions
robe+0.024
arl+0.023
ugg+0.020
melon+0.019
 Sleep+0.018
Neg logit contributions
pieces-0.030
 circles-0.027
Maria-0.026
od-0.026
 hours-0.025
Token and
Token ID290
Feature activation-8.25e-03

Pos logit contributions
robe+0.157
arl+0.150
melon+0.142
 cooler+0.140
ugg+0.138
Neg logit contributions
pieces-0.077
od-0.065
Maria-0.060
lier-0.055
ill-0.054
Token sn
Token ID3013
Feature activation-4.19e-02

Pos logit contributions
robe+0.015
arl+0.013
 cooler+0.013
melon+0.013
ugg+0.012
Neg logit contributions
pieces-0.014
od-0.012
Maria-0.012
lier-0.011
one-0.011
Tokenugg
Token ID6837
Feature activation-2.21e-01

Pos logit contributions
robe+0.034
 ranc+0.026
melon+0.024
 Sleep+0.022
 cooler+0.022
Neg logit contributions
pieces-0.104
asing-0.091
e-0.089
ets-0.089
ez-0.089
Tokenled
Token ID992
Feature activation-4.30e-02

Pos logit contributions
robe+0.446
arl+0.419
melon+0.411
 cooler+0.398
ugg+0.396
Neg logit contributions
pieces-0.286
od-0.248
Maria-0.230
lier-0.224
ets-0.214
Token in
Token ID287
Feature activation+4.73e-04

Pos logit contributions
robe+0.080
arl+0.079
melon+0.072
ugg+0.069
 ranc+0.068
Neg logit contributions
pieces-0.061
od-0.055
 circles-0.053
Maria-0.052
lier-0.051
Token their
Token ID511
Feature activation-1.61e-02

Pos logit contributions
pieces+0.001
od+0.001
Maria+0.001
 circles+0.001
lier+0.001
Neg logit contributions
robe-0.001
arl-0.001
melon-0.001
ugg-0.001
 cooler-0.001
Token beds
Token ID20237
Feature activation-5.76e-02

Pos logit contributions
arl+0.028
robe+0.027
melon+0.025
ugg+0.023
chard+0.022
Neg logit contributions
pieces-0.021
 circles-0.021
 minutes-0.021
 hours-0.019
's-0.018
Token.
Token ID13
Feature activation-4.94e-03

Pos logit contributions
robe+0.118
arl+0.110
melon+0.105
 cooler+0.103
ugg+0.103
Neg logit contributions
pieces-0.078
od-0.068
Maria-0.064
lier-0.059
one-0.058
Token you
Token ID345
Feature activation+1.61e-03

Pos logit contributions
pieces+0.115
Maria+0.102
od+0.099
one+0.092
 circles+0.089
Neg logit contributions
robe-0.162
arl-0.156
ugg-0.141
melon-0.136
 cooler-0.133
Token go
Token ID467
Feature activation-5.31e-02

Pos logit contributions
pieces+0.003
Maria+0.003
od+0.003
enta+0.002
one+0.002
Neg logit contributions
robe-0.003
arl-0.002
 cooler-0.002
melon-0.002
ugg-0.002
Token,
Token ID11
Feature activation-3.08e-02

Pos logit contributions
robe+0.087
arl+0.079
 cooler+0.075
melon+0.074
ugg+0.072
Neg logit contributions
pieces-0.096
od-0.087
Maria-0.085
lier-0.078
one-0.077
Token my
Token ID616
Feature activation-4.25e-03

Pos logit contributions
robe+0.050
arl+0.045
 cooler+0.043
melon+0.042
ugg+0.040
Neg logit contributions
pieces-0.056
od-0.051
Maria-0.049
enta-0.046
lier-0.045
Token d
Token ID288
Feature activation-5.51e-03

Pos logit contributions
robe+0.006
arl+0.005
melon+0.005
 cooler+0.005
ugg+0.004
Neg logit contributions
pieces-0.009
od-0.008
Maria-0.008
lier-0.007
one-0.007
Tokenarl
Token ID7063
Feature activation-2.20e-01

Pos logit contributions
robe+0.004
arl+0.004
melon+0.003
ugg+0.003
 cooler+0.003
Neg logit contributions
pieces-0.015
od-0.014
Maria-0.014
enta-0.013
lier-0.013
Tokenings
Token ID654
Feature activation-5.24e-02

Pos logit contributions
robe+0.357
arl+0.321
 cooler+0.306
melon+0.303
ugg+0.297
Neg logit contributions
pieces-0.388
od-0.352
Maria-0.330
one-0.316
lier-0.312
Token.
Token ID13
Feature activation+1.11e-02

Pos logit contributions
robe+0.099
arl+0.091
melon+0.086
 cooler+0.085
ugg+0.084
Neg logit contributions
pieces-0.084
od-0.073
Maria-0.071
enta-0.064
one-0.063
Token Enjoy
Token ID18179
Feature activation-5.09e-02

Pos logit contributions
pieces+0.018
od+0.016
Maria+0.015
one+0.014
lier+0.014
Neg logit contributions
robe-0.020
arl-0.018
melon-0.018
 cooler-0.017
ugg-0.017
Token your
Token ID534
Feature activation+8.21e-03

Pos logit contributions
robe+0.094
arl+0.085
melon+0.080
 cooler+0.079
ugg+0.079
Neg logit contributions
pieces-0.082
od-0.073
Maria-0.069
lier-0.064
one-0.064
Token lunch
Token ID9965
Feature activation-3.35e-02

Pos logit contributions
pieces+0.016
od+0.014
Maria+0.014
 circles+0.013
lier+0.013
Neg logit contributions
robe-0.011
arl-0.010
ugg-0.009
melon-0.009
 Cream-0.008
Token?"
Token ID1701
Feature activation+2.59e-02

Pos logit contributions
robe+0.036
arl+0.033
 cooler+0.030
melon+0.030
ugg+0.029
Neg logit contributions
pieces-0.050
od-0.044
Maria-0.043
one-0.040
enta-0.040
Token Lily
Token ID20037
Feature activation+3.54e-02

Pos logit contributions
pieces+0.053
od+0.048
Maria+0.046
lier+0.044
one+0.044
Neg logit contributions
robe-0.037
arl-0.033
melon-0.030
 cooler-0.030
ugg-0.029
Token asked
Token ID1965
Feature activation+2.04e-04

Pos logit contributions
pieces+0.071
od+0.056
Maria+0.054
one+0.053
lar+0.052
Neg logit contributions
robe-0.051
 cooler-0.047
 bar-0.045
 ranc-0.044
 tired-0.044
Token,
Token ID11
Feature activation-1.65e-02

Pos logit contributions
pieces+0.000
od+0.000
Maria+0.000
ata+0.000
one+0.000
Neg logit contributions
robe-0.000
arl-0.000
 cooler-0.000
melon-0.000
ugg-0.000
Token sn
Token ID3013
Feature activation-4.05e-02

Pos logit contributions
robe+0.030
arl+0.027
ugg+0.026
melon+0.025
 cooler+0.024
Neg logit contributions
pieces-0.027
od-0.024
Maria-0.024
 circles-0.022
lier-0.022
Tokenugg
Token ID6837
Feature activation-2.20e-01

Pos logit contributions
robe+0.062
 cooler+0.055
 ranc+0.053
 cabin+0.052
 Sleep+0.052
Neg logit contributions
pieces-0.073
od-0.067
ill-0.066
Maria-0.065
asing-0.064
Tokenling
Token ID1359
Feature activation-2.50e-02

Pos logit contributions
robe+0.455
melon+0.423
arl+0.423
ugg+0.418
 cooler+0.418
Neg logit contributions
pieces-0.254
od-0.233
iece-0.208
Maria-0.207
lier-0.203
Token with
Token ID351
Feature activation-4.22e-02

Pos logit contributions
robe+0.049
arl+0.048
ugg+0.044
melon+0.043
 Cream+0.042
Neg logit contributions
pieces-0.034
od-0.031
Maria-0.030
 circles-0.029
lier-0.029
Token her
Token ID607
Feature activation-1.89e-02

Pos logit contributions
ugg+0.099
robe+0.099
arl+0.097
melon+0.089
 ranc+0.088
Neg logit contributions
pieces-0.041
 circles-0.038
Maria-0.031
 Rachel-0.029
ill-0.029
Token mom
Token ID1995
Feature activation-5.46e-02

Pos logit contributions
robe+0.023
arl+0.021
ugg+0.020
melon+0.020
 Sleep+0.018
Neg logit contributions
pieces-0.040
od-0.036
Maria-0.035
 circles-0.035
ill-0.033
Token.
Token ID13
Feature activation-1.05e-02

Pos logit contributions
robe+0.098
arl+0.088
 cooler+0.086
melon+0.085
ugg+0.082
Neg logit contributions
pieces-0.088
od-0.079
Maria-0.077
one-0.069
enta-0.069
Token."
Token ID526
Feature activation-2.02e-02

Pos logit contributions
pieces+0.023
od+0.020
Maria+0.019
one+0.017
lier+0.017
Neg logit contributions
robe-0.028
arl-0.026
 cooler-0.025
melon-0.025
ugg-0.024
Token
Token ID198
Feature activation+8.36e-03

Pos logit contributions
robe+0.037
arl+0.033
melon+0.032
 cooler+0.031
ugg+0.030
Neg logit contributions
pieces-0.031
Maria-0.028
od-0.027
 circles-0.025
ill-0.025
Token
Token ID198
Feature activation+8.51e-03

Pos logit contributions
pieces+0.014
Maria+0.013
od+0.013
 circles+0.012
ill+0.011
Neg logit contributions
robe-0.014
arl-0.013
melon-0.012
 cooler-0.012
ugg-0.011
TokenThey
Token ID2990
Feature activation-7.06e-02

Pos logit contributions
pieces+0.012
od+0.010
Maria+0.010
lier+0.008
one+0.008
Neg logit contributions
robe-0.018
arl-0.017
melon-0.016
 cooler-0.016
ugg-0.015
Token sn
Token ID3013
Feature activation-4.17e-02

Pos logit contributions
robe+0.131
arl+0.124
melon+0.115
 cooler+0.112
ugg+0.109
Neg logit contributions
pieces-0.107
od-0.096
Maria-0.092
lier-0.084
ill-0.084
Tokenugg
Token ID6837
Feature activation-2.19e-01

Pos logit contributions
robe+0.029
 ranc+0.019
melon+0.019
arl+0.018
 cooler+0.018
Neg logit contributions
pieces-0.111
Maria-0.098
e-0.098
lier-0.098
asing-0.098
Tokenled
Token ID992
Feature activation-4.03e-02

Pos logit contributions
robe+0.412
arl+0.382
melon+0.365
 cooler+0.361
ugg+0.351
Neg logit contributions
pieces-0.333
od-0.293
Maria-0.279
lier-0.261
one-0.255
Token in
Token ID287
Feature activation+2.56e-04

Pos logit contributions
robe+0.075
arl+0.072
melon+0.065
ugg+0.063
 Cream+0.062
Neg logit contributions
pieces-0.058
od-0.053
Maria-0.051
lier-0.050
 circles-0.050
Token their
Token ID511
Feature activation-1.47e-02

Pos logit contributions
pieces+0.000
od+0.000
Maria+0.000
 circles+0.000
 hours+0.000
Neg logit contributions
robe-0.000
arl-0.000
melon-0.000
ugg-0.000
 cooler-0.000
Token beds
Token ID20237
Feature activation-5.61e-02

Pos logit contributions
robe+0.025
arl+0.025
melon+0.022
ugg+0.021
 Cream+0.020
Neg logit contributions
pieces-0.020
 circles-0.020
 hours-0.018
 rect-0.018
 minutes-0.018
Token and
Token ID290
Feature activation-4.83e-03

Pos logit contributions
robe+0.118
arl+0.110
 cooler+0.104
melon+0.104
ugg+0.102
Neg logit contributions
pieces-0.075
od-0.065
Maria-0.062
one-0.055
lier-0.055
Token"
Token ID1
Feature activation-4.44e-02

Pos logit contributions
pieces+0.011
od+0.010
Maria+0.009
lier+0.009
one+0.008
Neg logit contributions
robe-0.012
arl-0.011
 cooler-0.010
ugg-0.010
 Cream-0.010
TokenHello
Token ID15496
Feature activation-3.20e-02

Pos logit contributions
robe+0.075
arl+0.069
ugg+0.063
 cooler+0.063
melon+0.063
Neg logit contributions
pieces-0.079
od-0.070
Maria-0.066
lier-0.063
enta-0.061
Token,
Token ID11
Feature activation-2.07e-02

Pos logit contributions
robe+0.050
arl+0.046
 cooler+0.046
 Cream+0.043
melon+0.042
Neg logit contributions
pieces-0.059
od-0.052
Maria-0.051
asing-0.047
ez-0.047
Token my
Token ID616
Feature activation-4.82e-03

Pos logit contributions
robe+0.032
arl+0.027
melon+0.027
ugg+0.026
 cooler+0.025
Neg logit contributions
pieces-0.040
od-0.036
Maria-0.035
one-0.033
lier-0.032
Token d
Token ID288
Feature activation-3.91e-03

Pos logit contributions
robe+0.007
arl+0.007
melon+0.007
 cooler+0.006
ugg+0.006
Neg logit contributions
pieces-0.009
od-0.008
Maria-0.008
lier-0.008
one-0.007
Tokenarl
Token ID7063
Feature activation-2.18e-01

Pos logit contributions
robe+0.001
 cooler+0.000
melon+0.000
arl+0.000
ugg+0.000
Neg logit contributions
pieces-0.012
od-0.011
Maria-0.011
one-0.011
ill-0.011
Tokenings
Token ID654
Feature activation-4.99e-02

Pos logit contributions
robe+0.344
chard+0.335
 cooler+0.334
melon+0.333
 ranc+0.322
Neg logit contributions
pieces-0.310
od-0.306
one-0.281
ata-0.274
ill-0.272
Token.
Token ID13
Feature activation-5.73e-03

Pos logit contributions
robe+0.089
arl+0.085
 cooler+0.078
 Cream+0.076
ugg+0.075
Neg logit contributions
pieces-0.086
od-0.074
Maria-0.072
enta-0.068
 rect-0.065
Token I
Token ID314
Feature activation+1.83e-02

Pos logit contributions
robe+0.011
arl+0.010
melon+0.009
 cooler+0.009
ugg+0.009
Neg logit contributions
pieces-0.009
od-0.008
Maria-0.008
one-0.007
lier-0.007
Token'm
Token ID1101
Feature activation+4.57e-02

Pos logit contributions
pieces+0.031
od+0.027
Maria+0.026
one+0.024
lier+0.024
Neg logit contributions
robe-0.033
arl-0.030
 cooler-0.028
ugg-0.028
melon-0.028
Token so
Token ID523
Feature activation-5.19e-02

Pos logit contributions
pieces+0.082
od+0.076
Maria+0.072
 circles+0.065
enta+0.065
Neg logit contributions
robe-0.077
arl-0.069
ugg-0.066
melon-0.063
 cooler-0.059
Token
Token ID198
Feature activation+5.90e-03

Pos logit contributions
robe+0.081
arl+0.074
melon+0.065
ugg+0.064
 Cream+0.062
Neg logit contributions
pieces-0.109
od-0.095
Maria-0.091
lar-0.090
enta-0.090
Token
Token ID198
Feature activation+7.35e-03

Pos logit contributions
pieces+0.012
od+0.011
ets+0.010
enta+0.010
asing+0.010
Neg logit contributions
robe-0.008
arl-0.007
ugg-0.007
 Cream-0.007
melon-0.006
TokenDon
Token ID3987
Feature activation-9.24e-02

Pos logit contributions
pieces+0.012
od+0.011
ets+0.009
enta+0.009
one+0.009
Neg logit contributions
robe-0.013
arl-0.012
ugg-0.011
 Cream-0.011
melon-0.011
Tokenny
Token ID3281
Feature activation-4.06e-02

Pos logit contributions
robe+0.114
arl+0.101
 cooler+0.089
melon+0.089
ugg+0.088
Neg logit contributions
pieces-0.206
od-0.187
Maria-0.182
lier-0.172
enta-0.171
Token sn
Token ID3013
Feature activation-5.83e-02

Pos logit contributions
robe+0.071
arl+0.064
melon+0.060
ugg+0.059
 cooler+0.059
Neg logit contributions
pieces-0.072
od-0.062
Maria-0.062
enta-0.057
 circles-0.056
Tokenugg
Token ID6837
Feature activation-2.17e-01

Pos logit contributions
 ranc+0.079
robe+0.078
 cabin+0.074
 waitress+0.073
 cooler+0.071
Neg logit contributions
pieces-0.101
ill-0.098
od-0.093
Maria-0.091
e-0.090
Tokenled
Token ID992
Feature activation-4.03e-02

Pos logit contributions
robe+0.457
arl+0.432
melon+0.421
 cooler+0.420
ugg+0.418
Neg logit contributions
pieces-0.252
od-0.222
Maria-0.201
lier-0.189
inda-0.187
Token into
Token ID656
Feature activation+2.19e-02

Pos logit contributions
robe+0.069
arl+0.063
 cooler+0.059
melon+0.059
ugg+0.057
Neg logit contributions
pieces-0.070
od-0.063
Maria-0.061
one-0.056
lier-0.055
Token his
Token ID465
Feature activation-2.87e-02

Pos logit contributions
pieces+0.043
od+0.039
Maria+0.037
lier+0.036
one+0.035
Neg logit contributions
robe-0.033
arl-0.030
 cooler-0.029
melon-0.027
 Sleep-0.026
Token bed
Token ID3996
Feature activation-1.20e-02

Pos logit contributions
robe+0.039
arl+0.036
melon+0.032
ugg+0.031
 Cream+0.029
Neg logit contributions
pieces-0.059
od-0.051
Maria-0.051
 circles-0.050
lier-0.048
Token and
Token ID290
Feature activation-2.34e-03

Pos logit contributions
robe+0.022
arl+0.021
melon+0.019
 cooler+0.019
ugg+0.019
Neg logit contributions
pieces-0.019
od-0.017
Maria-0.016
lier-0.015
one-0.014
Token's
Token ID338
Feature activation-1.07e-03

Pos logit contributions
robe+0.193
arl+0.179
melon+0.169
 cooler+0.168
ugg+0.165
Neg logit contributions
pieces-0.142
od-0.124
Maria-0.119
lier-0.107
one-0.107
Token wonderful
Token ID7932
Feature activation+4.51e-02

Pos logit contributions
robe+0.002
ugg+0.002
arl+0.001
melon+0.001
 ranc+0.001
Neg logit contributions
pieces-0.002
od-0.002
Maria-0.002
lier-0.002
one-0.002
Token,
Token ID11
Feature activation-2.21e-02

Pos logit contributions
pieces+0.069
od+0.060
Maria+0.058
lier+0.053
one+0.053
Neg logit contributions
robe-0.087
arl-0.081
melon-0.076
 cooler-0.075
ugg-0.074
Token my
Token ID616
Feature activation-1.94e-03

Pos logit contributions
robe+0.032
arl+0.029
 cooler+0.027
melon+0.027
ugg+0.025
Neg logit contributions
pieces-0.044
od-0.040
Maria-0.039
one-0.036
enta-0.036
Token d
Token ID288
Feature activation-3.28e-03

Pos logit contributions
robe+0.003
arl+0.002
melon+0.002
 cooler+0.002
ugg+0.002
Neg logit contributions
pieces-0.004
od-0.004
Maria-0.004
lier-0.003
one-0.003
Tokenarl
Token ID7063
Feature activation-2.17e-01

Pos logit contributions
robe+0.003
arl+0.003
melon+0.002
 cooler+0.002
ugg+0.002
Neg logit contributions
pieces-0.008
od-0.008
Maria-0.008
lier-0.007
one-0.007
Tokenings
Token ID654
Feature activation-4.74e-02

Pos logit contributions
robe+0.314
arl+0.282
 cooler+0.262
melon+0.261
ugg+0.255
Neg logit contributions
pieces-0.427
od-0.388
Maria-0.373
one-0.352
lier-0.350
Token!"
Token ID2474
Feature activation+2.86e-02

Pos logit contributions
robe+0.095
arl+0.089
melon+0.084
 cooler+0.083
ugg+0.082
Neg logit contributions
pieces-0.070
od-0.060
Maria-0.058
enta-0.052
one-0.051
Token Mom
Token ID11254
Feature activation-2.22e-02

Pos logit contributions
pieces+0.054
od+0.049
enta+0.046
Maria+0.046
one+0.045
Neg logit contributions
robe-0.044
melon-0.039
arl-0.038
 cooler-0.038
 Cream-0.037
Tokenmy
Token ID1820
Feature activation+3.41e-02

Pos logit contributions
robe+0.036
melon+0.035
arl+0.034
 cooler+0.034
ugg+0.032
Neg logit contributions
pieces-0.038
od-0.032
Maria-0.030
 hours-0.030
 rect-0.030
Token says
Token ID1139
Feature activation+2.84e-02

Pos logit contributions
pieces+0.057
od+0.047
Maria+0.046
one+0.046
enta+0.045
Neg logit contributions
robe-0.057
melon-0.056
 cooler-0.053
 Cream-0.050
 waitress-0.049
Token was
Token ID373
Feature activation+4.99e-03

Pos logit contributions
robe+0.004
arl+0.004
 cooler+0.003
melon+0.003
ugg+0.003
Neg logit contributions
pieces-0.003
od-0.003
Maria-0.003
asing-0.002
inda-0.002
Token over
Token ID625
Feature activation-3.34e-02

Pos logit contributions
pieces+0.008
Maria+0.007
od+0.007
lier+0.007
one+0.007
Neg logit contributions
robe-0.009
ugg-0.008
arl-0.008
melon-0.007
 Cream-0.007
Token.
Token ID13
Feature activation+6.50e-03

Pos logit contributions
robe+0.051
 cooler+0.046
arl+0.046
 cabin+0.043
melon+0.043
Neg logit contributions
pieces-0.068
od-0.058
Maria-0.055
ata-0.053
asing-0.053
Token He
Token ID679
Feature activation-3.93e-02

Pos logit contributions
pieces+0.014
od+0.012
Maria+0.012
lier+0.011
one+0.011
Neg logit contributions
robe-0.009
arl-0.008
melon-0.007
 cooler-0.007
ugg-0.007
Token sn
Token ID3013
Feature activation-3.87e-02

Pos logit contributions
robe+0.065
 cooler+0.059
ugg+0.055
 ranc+0.054
 cabin+0.054
Neg logit contributions
pieces-0.074
od-0.065
Maria-0.061
asing-0.056
one-0.056
Tokenugg
Token ID6837
Feature activation-2.16e-01

Pos logit contributions
robe+0.028
 ranc+0.023
melon+0.022
 cooler+0.020
 waitress+0.019
Neg logit contributions
pieces-0.097
Maria-0.089
ill-0.088
ez-0.088
asing-0.087
Tokenled
Token ID992
Feature activation-3.99e-02

Pos logit contributions
robe+0.414
arl+0.395
melon+0.381
 cooler+0.370
ugg+0.365
Neg logit contributions
pieces-0.302
od-0.267
Maria-0.254
lier-0.237
lar-0.230
Token with
Token ID351
Feature activation-4.16e-02

Pos logit contributions
robe+0.065
arl+0.061
melon+0.054
ugg+0.054
 Cream+0.051
Neg logit contributions
pieces-0.069
od-0.063
Maria-0.062
 circles-0.060
lier-0.060
Token them
Token ID606
Feature activation-4.03e-02

Pos logit contributions
robe+0.070
arl+0.064
melon+0.059
ugg+0.059
 cooler+0.058
Neg logit contributions
pieces-0.074
od-0.066
Maria-0.064
one-0.059
lier-0.059
Token and
Token ID290
Feature activation+1.27e-04

Pos logit contributions
robe+0.076
arl+0.071
melon+0.067
 cooler+0.065
ugg+0.065
Neg logit contributions
pieces-0.061
od-0.054
Maria-0.052
lier-0.048
one-0.047
Token listened
Token ID16399
Feature activation-6.97e-02

Pos logit contributions
pieces+0.000
od+0.000
Maria+0.000
lier+0.000
one+0.000
Neg logit contributions
robe-0.000
arl-0.000
melon-0.000
 cooler-0.000
ugg-0.000
Token the
Token ID262
Feature activation-2.05e-02

Pos logit contributions
pieces+0.071
od+0.063
Maria+0.061
one+0.056
lier+0.056
Neg logit contributions
robe-0.071
arl-0.064
 cooler-0.060
melon-0.060
ugg-0.059
Token cover
Token ID3002
Feature activation+1.68e-02

Pos logit contributions
robe+0.025
arl+0.022
melon+0.020
 cooler+0.020
ugg+0.019
Neg logit contributions
pieces-0.046
od-0.042
Maria-0.041
lier-0.038
one-0.038
Token.
Token ID13
Feature activation-1.63e-03

Pos logit contributions
pieces+0.029
od+0.025
Maria+0.025
enta+0.024
asing+0.023
Neg logit contributions
robe-0.031
arl-0.026
melon-0.026
 cooler-0.026
ugg-0.024
Token She
Token ID1375
Feature activation-2.63e-02

Pos logit contributions
robe+0.003
arl+0.002
melon+0.002
 cooler+0.002
ugg+0.002
Neg logit contributions
pieces-0.003
od-0.003
Maria-0.003
lier-0.002
one-0.002
Token sn
Token ID3013
Feature activation-3.73e-02

Pos logit contributions
robe+0.044
 cooler+0.039
ugg+0.036
melon+0.036
 Cream+0.035
Neg logit contributions
pieces-0.049
od-0.040
Maria-0.039
one-0.038
asing-0.037
Tokenugg
Token ID6837
Feature activation-2.16e-01

Pos logit contributions
robe+0.028
 ranc+0.021
 cabin+0.020
 cooler+0.020
melon+0.017
Neg logit contributions
pieces-0.097
od-0.088
lier-0.088
ill-0.087
Maria-0.087
Tokenled
Token ID992
Feature activation-3.91e-02

Pos logit contributions
robe+0.409
arl+0.386
melon+0.367
 cooler+0.363
ugg+0.355
Neg logit contributions
pieces-0.320
od-0.284
Maria-0.269
lier-0.252
one-0.246
Token with
Token ID351
Feature activation-3.87e-02

Pos logit contributions
robe+0.067
arl+0.064
ugg+0.057
melon+0.055
 ranc+0.054
Neg logit contributions
pieces-0.065
od-0.060
Maria-0.057
 circles-0.056
lier-0.055
Token Mom
Token ID11254
Feature activation-2.94e-02

Pos logit contributions
robe+0.077
arl+0.073
ugg+0.069
melon+0.067
 ranc+0.066
Neg logit contributions
pieces-0.056
od-0.049
Maria-0.048
 circles-0.046
ill-0.042
Token and
Token ID290
Feature activation-7.95e-03

Pos logit contributions
robe+0.058
arl+0.053
 cooler+0.051
melon+0.051
ugg+0.050
Neg logit contributions
pieces-0.043
od-0.038
Maria-0.036
one-0.032
enta-0.032
Token listened
Token ID16399
Feature activation-6.68e-02

Pos logit contributions
robe+0.016
arl+0.015
 cooler+0.014
melon+0.014
ugg+0.013
Neg logit contributions
pieces-0.011
od-0.010
Maria-0.009
enta-0.008
lier-0.008
Token and
Token ID290
Feature activation-4.84e-02

Pos logit contributions
pieces+0.025
Maria+0.021
od+0.020
enta+0.019
one+0.018
Neg logit contributions
robe-0.034
arl-0.031
 cooler-0.031
melon-0.029
 cabin-0.029
Token both
Token ID1111
Feature activation-1.83e-02

Pos logit contributions
robe+0.082
arl+0.073
 cooler+0.072
melon+0.071
ugg+0.066
Neg logit contributions
pieces-0.083
od-0.077
Maria-0.073
one-0.068
lier-0.068
Token of
Token ID286
Feature activation+2.95e-03

Pos logit contributions
robe+0.030
 cooler+0.027
melon+0.026
arl+0.026
ugg+0.025
Neg logit contributions
pieces-0.031
od-0.028
Maria-0.027
enta-0.026
lier-0.025
Token them
Token ID606
Feature activation-5.79e-02

Pos logit contributions
pieces+0.005
od+0.005
ata+0.005
Maria+0.005
asing+0.005
Neg logit contributions
robe-0.004
 cooler-0.004
arl-0.004
melon-0.004
 cabin-0.004
Token sn
Token ID3013
Feature activation-4.92e-02

Pos logit contributions
robe+0.099
 cooler+0.086
melon+0.085
arl+0.083
ugg+0.080
Neg logit contributions
pieces-0.099
Maria-0.091
od-0.089
one-0.082
enta-0.081
Tokenugg
Token ID6837
Feature activation-2.16e-01

Pos logit contributions
robe+0.065
arl+0.054
 cooler+0.053
melon+0.052
 ranc+0.049
Neg logit contributions
pieces-0.103
od-0.092
Maria-0.089
lier-0.088
enta-0.085
Tokenled
Token ID992
Feature activation-4.21e-02

Pos logit contributions
robe+0.376
arl+0.342
 cooler+0.318
melon+0.317
ugg+0.311
Neg logit contributions
pieces-0.379
od-0.337
Maria-0.331
one-0.305
enta-0.298
Token together
Token ID1978
Feature activation+3.19e-02

Pos logit contributions
robe+0.068
arl+0.060
 cooler+0.060
melon+0.058
ugg+0.056
Neg logit contributions
pieces-0.077
od-0.070
Maria-0.067
one-0.063
enta-0.063
Token under
Token ID739
Feature activation+5.40e-02

Pos logit contributions
pieces+0.053
od+0.048
Maria+0.047
enta+0.043
one+0.042
Neg logit contributions
robe-0.057
arl-0.050
 cooler-0.050
melon-0.049
ugg-0.047
Token a
Token ID257
Feature activation-3.03e-03

Pos logit contributions
pieces+0.103
od+0.093
Maria+0.089
asing+0.084
one+0.084
Neg logit contributions
robe-0.083
arl-0.075
 cooler-0.073
melon-0.072
ugg-0.067
Token cozy
Token ID37438
Feature activation-4.05e-02

Pos logit contributions
robe+0.004
arl+0.003
 cooler+0.003
melon+0.003
ugg+0.003
Neg logit contributions
pieces-0.007
od-0.006
Maria-0.006
one-0.006
lier-0.006
Token she
Token ID673
Feature activation+2.24e-02

Pos logit contributions
robe+0.209
arl+0.195
melon+0.176
ugg+0.174
 cooler+0.174
Neg logit contributions
pieces-0.241
od-0.217
Maria-0.212
one-0.195
lier-0.194
Token pur
Token ID1308
Feature activation-4.43e-02

Pos logit contributions
pieces+0.037
od+0.032
Maria+0.031
one+0.029
 circles+0.028
Neg logit contributions
robe-0.042
 cooler-0.036
melon-0.036
ugg-0.035
arl-0.035
Tokenred
Token ID445
Feature activation-1.08e-01

Pos logit contributions
robe+0.093
arl+0.087
ugg+0.083
melon+0.080
 Sleep+0.076
Neg logit contributions
pieces-0.065
od-0.053
Maria-0.052
one-0.048
 circles-0.046
Token and
Token ID290
Feature activation-1.60e-02

Pos logit contributions
robe+0.207
arl+0.189
melon+0.179
 cooler+0.179
ugg+0.176
Neg logit contributions
pieces-0.167
od-0.146
Maria-0.141
one-0.129
lier-0.127
Token sn
Token ID3013
Feature activation-4.90e-02

Pos logit contributions
robe+0.028
arl+0.025
 cooler+0.025
melon+0.024
ugg+0.024
Neg logit contributions
pieces-0.027
od-0.024
Maria-0.023
one-0.022
lier-0.021
Tokenugg
Token ID6837
Feature activation-2.16e-01

Pos logit contributions
robe+0.038
 ranc+0.027
 cooler+0.027
 cabin+0.027
 refrigerator+0.024
Neg logit contributions
pieces-0.125
ez-0.111
Maria-0.111
lier-0.110
asing-0.110
Tokenled
Token ID992
Feature activation-3.83e-02

Pos logit contributions
robe+0.424
arl+0.398
melon+0.388
 cooler+0.382
ugg+0.366
Neg logit contributions
pieces-0.291
od-0.255
Maria-0.240
lier-0.229
inda-0.218
Token up
Token ID510
Feature activation-7.98e-02

Pos logit contributions
robe+0.059
arl+0.052
 cooler+0.049
melon+0.049
ugg+0.048
Neg logit contributions
pieces-0.074
od-0.067
Maria-0.064
one-0.061
enta-0.060
Token to
Token ID284
Feature activation+2.16e-02

Pos logit contributions
robe+0.130
arl+0.117
 cooler+0.112
melon+0.111
ugg+0.109
Neg logit contributions
pieces-0.150
od-0.134
Maria-0.129
one-0.123
enta-0.121
Token her
Token ID607
Feature activation-1.74e-02

Pos logit contributions
pieces+0.048
od+0.045
Maria+0.043
one+0.042
lier+0.041
Neg logit contributions
robe-0.027
arl-0.023
 cooler-0.022
 Sleep-0.021
melon-0.021
Token owner
Token ID4870
Feature activation-1.99e-02

Pos logit contributions
robe+0.023
arl+0.020
melon+0.018
 cooler+0.018
ugg+0.018
Neg logit contributions
pieces-0.037
od-0.034
Maria-0.033
lier-0.031
one-0.031
Token name
Token ID1438
Feature activation-7.91e-02

Pos logit contributions
robe+0.020
arl+0.020
ugg+0.018
melon+0.017
 cooler+0.014
Neg logit contributions
pieces-0.051
od-0.047
 circles-0.045
Maria-0.044
 pad-0.044
Token is
Token ID318
Feature activation-1.90e-02

Pos logit contributions
robe+0.135
arl+0.122
 cooler+0.114
melon+0.114
ugg+0.112
Neg logit contributions
pieces-0.139
od-0.124
Maria-0.120
one-0.110
lier-0.110
Token Mr
Token ID1770
Feature activation-7.38e-02

Pos logit contributions
robe+0.026
arl+0.023
melon+0.021
ugg+0.021
 cooler+0.021
Neg logit contributions
pieces-0.040
od-0.036
Maria-0.035
lier-0.033
one-0.033
Token.
Token ID13
Feature activation-8.39e-03

Pos logit contributions
robe+0.146
arl+0.135
 cooler+0.127
melon+0.127
ugg+0.124
Neg logit contributions
pieces-0.110
od-0.096
Maria-0.092
one-0.083
lier-0.083
Token Sn
Token ID5489
Feature activation-1.59e-01

Pos logit contributions
robe+0.010
arl+0.010
melon+0.010
ugg+0.009
 cooler+0.008
Neg logit contributions
pieces-0.016
od-0.015
Maria-0.015
 circles-0.015
one-0.015
Tokenugg
Token ID6837
Feature activation-2.16e-01

Pos logit contributions
robe+0.139
arl+0.118
 cooler+0.117
melon+0.113
 ranc+0.100
Neg logit contributions
pieces-0.386
od-0.363
lier-0.346
Maria-0.342
asing-0.339
Tokenles
Token ID829
Feature activation+4.57e-02

Pos logit contributions
robe+0.547
arl+0.542
melon+0.539
 ranc+0.531
chard+0.530
Neg logit contributions
pieces-0.113
od-0.093
ets-0.076
one-0.067
 circles-0.066
Token."
Token ID526
Feature activation-2.75e-02

Pos logit contributions
pieces+0.067
od+0.058
Maria+0.056
lier+0.051
one+0.051
Neg logit contributions
robe-0.090
arl-0.084
melon-0.079
 cooler-0.079
ugg-0.078
Token
Token ID198
Feature activation+3.28e-03

Pos logit contributions
robe+0.054
arl+0.051
melon+0.050
 cooler+0.047
ugg+0.045
Neg logit contributions
pieces-0.037
Maria-0.032
od-0.032
 circles-0.029
lier-0.028
Token
Token ID198
Feature activation+3.84e-03

Pos logit contributions
pieces+0.006
od+0.005
Maria+0.005
one+0.005
lier+0.005
Neg logit contributions
robe-0.005
arl-0.005
ugg-0.005
 cooler-0.005
melon-0.005
TokenT
Token ID51
Feature activation-7.39e-02

Pos logit contributions
pieces+0.004
od+0.003
Maria+0.003
lier+0.003
ata+0.003
Neg logit contributions
robe-0.009
arl-0.008
melon-0.008
 cooler-0.008
ugg-0.008
Token.
Token ID13
Feature activation-1.70e-02

Pos logit contributions
pieces+0.064
od+0.060
Maria+0.058
enta+0.053
ata+0.053
Neg logit contributions
robe-0.058
arl-0.052
 cooler-0.051
melon-0.050
ugg-0.047
Token
Token ID220
Feature activation-4.73e-02

Pos logit contributions
robe+0.025
arl+0.023
melon+0.021
 cooler+0.021
ugg+0.021
Neg logit contributions
pieces-0.033
od-0.030
Maria-0.029
enta-0.027
asing-0.027
Token
Token ID198
Feature activation+9.35e-03

Pos logit contributions
robe+0.078
arl+0.071
 cooler+0.064
melon+0.064
ugg+0.064
Neg logit contributions
pieces-0.086
od-0.077
Maria-0.073
lier-0.069
enta-0.069
Token
Token ID198
Feature activation+1.00e-02

Pos logit contributions
pieces+0.018
od+0.017
asing+0.016
lier+0.016
enta+0.015
Neg logit contributions
robe-0.013
arl-0.012
ugg-0.011
 Cream-0.011
 cooler-0.011
TokenSn
Token ID16501
Feature activation-1.13e-01

Pos logit contributions
pieces+0.016
od+0.014
enta+0.012
lier+0.012
Maria+0.012
Neg logit contributions
robe-0.018
arl-0.017
ugg-0.016
 Cream-0.016
 cooler-0.016
Tokenugg
Token ID6837
Feature activation-2.16e-01

Pos logit contributions
robe+0.180
arl+0.161
 cooler+0.155
melon+0.150
 ranc+0.145
Neg logit contributions
pieces-0.204
od-0.185
Maria-0.175
lier-0.169
one-0.164
Tokenles
Token ID829
Feature activation+4.78e-02

Pos logit contributions
robe+0.483
arl+0.475
melon+0.446
 cooler+0.445
ugg+0.434
Neg logit contributions
pieces-0.212
od-0.182
Maria-0.154
lier-0.151
lar-0.143
Token was
Token ID373
Feature activation+7.70e-03

Pos logit contributions
pieces+0.077
od+0.067
Maria+0.066
one+0.060
enta+0.059
Neg logit contributions
robe-0.090
 cooler-0.077
arl-0.077
melon-0.077
ugg-0.076
Token so
Token ID523
Feature activation-5.81e-02

Pos logit contributions
pieces+0.014
od+0.013
Maria+0.013
lier+0.012
 circles+0.011
Neg logit contributions
robe-0.012
arl-0.011
melon-0.010
ugg-0.010
 cooler-0.010
Token sad
Token ID6507
Feature activation-6.21e-02

Pos logit contributions
robe+0.099
arl+0.087
melon+0.081
 ranc+0.077
 cooler+0.075
Neg logit contributions
pieces-0.097
Maria-0.091
od-0.090
 realised-0.082
lier-0.081
Token to
Token ID284
Feature activation+1.78e-02

Pos logit contributions
robe+0.097
arl+0.086
 cooler+0.083
melon+0.080
ugg+0.078
Neg logit contributions
pieces-0.121
od-0.106
Maria-0.106
enta-0.099
one-0.095
TokenThat
Token ID2504
Feature activation-8.01e-02

Pos logit contributions
pieces+0.031
lier+0.027
od+0.025
 pad+0.025
one+0.025
Neg logit contributions
robe-0.034
arl-0.030
ugg-0.029
 Sleep-0.029
 Cream-0.027
Token night
Token ID1755
Feature activation+3.04e-02

Pos logit contributions
robe+0.128
arl+0.111
 cooler+0.108
melon+0.107
ugg+0.106
Neg logit contributions
pieces-0.149
od-0.133
Maria-0.129
lier-0.120
one-0.119
Token,
Token ID11
Feature activation-1.40e-02

Pos logit contributions
pieces+0.056
od+0.049
Maria+0.047
lier+0.045
one+0.045
Neg logit contributions
robe-0.050
 cooler-0.041
arl-0.041
 Sleep-0.040
ugg-0.040
Token Julie
Token ID21946
Feature activation+6.48e-02

Pos logit contributions
robe+0.025
 Sleep+0.022
 cooler+0.021
 cabin+0.021
 tired+0.020
Neg logit contributions
pieces-0.022
lier-0.019
od-0.018
one-0.018
ez-0.018
Token sn
Token ID3013
Feature activation-3.89e-02

Pos logit contributions
pieces+0.108
od+0.093
Maria+0.091
asing+0.083
one+0.081
Neg logit contributions
robe-0.116
 cooler-0.103
ugg-0.103
 Sleep-0.101
 ranc-0.098
Tokenugg
Token ID6837
Feature activation-2.16e-01

Pos logit contributions
robe+0.032
 ranc+0.028
 cooler+0.025
melon+0.025
 cabin+0.024
Neg logit contributions
pieces-0.091
lier-0.083
 mastered-0.082
asing-0.082
aw-0.082
Tokenled
Token ID992
Feature activation-3.74e-02

Pos logit contributions
robe+0.410
arl+0.394
melon+0.375
 cooler+0.369
ugg+0.357
Neg logit contributions
pieces-0.309
od-0.274
Maria-0.259
lier-0.240
inda-0.234
Token into
Token ID656
Feature activation+2.49e-02

Pos logit contributions
robe+0.058
arl+0.056
melon+0.049
ugg+0.048
 Cream+0.046
Neg logit contributions
pieces-0.068
od-0.062
Maria-0.060
lier-0.057
 circles-0.057
Token her
Token ID607
Feature activation-1.63e-02

Pos logit contributions
pieces+0.042
od+0.037
Maria+0.036
 circles+0.033
lier+0.032
Neg logit contributions
robe-0.044
arl-0.041
melon-0.038
ugg-0.038
 cooler-0.037
Token own
Token ID898
Feature activation+7.38e-02

Pos logit contributions
arl+0.029
melon+0.029
robe+0.025
ugg+0.024
chard+0.023
Neg logit contributions
 minutes-0.022
 hours-0.021
 circles-0.021
pieces-0.020
's-0.020
Token soft
Token ID2705
Feature activation-1.15e-03

Pos logit contributions
pieces+0.120
 circles+0.113
 hours+0.108
 minutes+0.106
's+0.106
Neg logit contributions
arl-0.116
melon-0.115
robe-0.106
ugg-0.100
chard-0.095
Token asleep
Token ID16039
Feature activation-9.81e-02

Pos logit contributions
pieces+0.039
 circles+0.037
od+0.037
Maria+0.036
ill+0.035
Neg logit contributions
robe-0.029
melon-0.028
arl-0.027
 Cream-0.026
ugg-0.026
Token.
Token ID13
Feature activation+1.10e-02

Pos logit contributions
robe+0.191
arl+0.177
 cooler+0.166
melon+0.166
ugg+0.163
Neg logit contributions
pieces-0.148
od-0.129
Maria-0.124
one-0.113
lier-0.112
Token Fl
Token ID1610
Feature activation-9.07e-02

Pos logit contributions
Maria+0.013
pieces+0.013
od+0.013
 circles+0.012
ill+0.012
Neg logit contributions
robe-0.022
melon-0.021
arl-0.020
 cooler-0.019
 yacht-0.018
Tokenuffy
Token ID15352
Feature activation-5.56e-02

Pos logit contributions
robe+0.107
melon+0.092
 cooler+0.086
arl+0.086
 ranc+0.083
Neg logit contributions
od-0.182
pieces-0.177
ill-0.172
asing-0.168
lier-0.167
Token sn
Token ID3013
Feature activation-4.07e-02

Pos logit contributions
robe+0.101
arl+0.094
melon+0.088
 cooler+0.087
ugg+0.085
Neg logit contributions
pieces-0.089
od-0.079
Maria-0.076
lier-0.070
one-0.070
Tokenugg
Token ID6837
Feature activation-2.15e-01

Pos logit contributions
robe+0.029
 ranc+0.025
melon+0.023
 cooler+0.022
 cabin+0.021
Neg logit contributions
pieces-0.098
lier-0.092
aw-0.091
e-0.090
ill-0.090
Tokenled
Token ID992
Feature activation-3.86e-02

Pos logit contributions
robe+0.438
arl+0.420
melon+0.409
 cooler+0.396
ugg+0.389
Neg logit contributions
pieces-0.261
od-0.230
Maria-0.214
lier-0.209
one-0.198
Token with
Token ID351
Feature activation-3.89e-02

Pos logit contributions
robe+0.080
arl+0.078
melon+0.073
ugg+0.070
 Cream+0.069
Neg logit contributions
pieces-0.046
od-0.042
lier-0.042
Maria-0.040
 circles-0.040
Token her
Token ID607
Feature activation-1.86e-02

Pos logit contributions
robe+0.080
melon+0.076
arl+0.075
ugg+0.068
 ranc+0.068
Neg logit contributions
pieces-0.046
 circles-0.043
Maria-0.043
od-0.042
e-0.040
Token and
Token ID290
Feature activation-1.42e-03

Pos logit contributions
robe+0.035
arl+0.034
melon+0.033
ugg+0.032
 Cream+0.031
Neg logit contributions
pieces-0.026
od-0.022
 circles-0.021
Maria-0.021
lier-0.020
Token pur
Token ID1308
Feature activation-3.61e-02

Pos logit contributions
robe+0.002
arl+0.002
melon+0.002
ugg+0.002
 Cream+0.002
Neg logit contributions
pieces-0.002
Maria-0.002
od-0.002
 imagined-0.002
 circles-0.002
Token on
Token ID319
Feature activation-1.39e-04

Pos logit contributions
pieces+0.053
od+0.046
Maria+0.045
one+0.041
lier+0.040
Neg logit contributions
robe-0.070
arl-0.065
 cooler-0.061
melon-0.061
ugg-0.060
Token the
Token ID262
Feature activation-5.70e-02

Pos logit contributions
robe+0.000
 cooler+0.000
 crib+0.000
arl+0.000
 cabin+0.000
Neg logit contributions
pieces-0.000
od-0.000
asing-0.000
ata-0.000
lier-0.000
Token bed
Token ID3996
Feature activation-3.77e-02

Pos logit contributions
robe+0.078
arl+0.067
 cooler+0.065
melon+0.064
ugg+0.058
Neg logit contributions
pieces-0.120
od-0.112
Maria-0.108
lier-0.102
one-0.102
Token and
Token ID290
Feature activation-3.04e-02

Pos logit contributions
robe+0.072
arl+0.064
 cooler+0.060
melon+0.060
ugg+0.059
Neg logit contributions
pieces-0.061
od-0.053
Maria-0.052
one-0.047
enta-0.047
Token sn
Token ID3013
Feature activation-5.44e-02

Pos logit contributions
robe+0.053
 cooler+0.047
arl+0.047
melon+0.045
ugg+0.044
Neg logit contributions
pieces-0.052
od-0.047
Maria-0.045
one-0.042
lier-0.042
Tokenugg
Token ID6837
Feature activation-2.15e-01

Pos logit contributions
robe+0.055
 ranc+0.050
 cabin+0.045
 cooler+0.043
 waitress+0.043
Neg logit contributions
pieces-0.121
od-0.108
ill-0.107
Maria-0.105
 mastered-0.104
Tokenled
Token ID992
Feature activation-4.13e-02

Pos logit contributions
robe+0.410
arl+0.386
melon+0.368
 cooler+0.361
ugg+0.354
Neg logit contributions
pieces-0.312
od-0.277
Maria-0.261
lier-0.245
one-0.238
Token together
Token ID1978
Feature activation+2.72e-02

Pos logit contributions
robe+0.064
arl+0.057
 cooler+0.054
melon+0.053
ugg+0.052
Neg logit contributions
pieces-0.079
od-0.072
Maria-0.069
one-0.065
enta-0.064
Token.
Token ID13
Feature activation+4.17e-03

Pos logit contributions
pieces+0.047
od+0.041
Maria+0.041
enta+0.038
one+0.037
Neg logit contributions
robe-0.048
 cooler-0.042
arl-0.041
melon-0.041
ugg-0.039
Token They
Token ID1119
Feature activation-3.20e-02

Pos logit contributions
pieces+0.008
od+0.007
Maria+0.007
lier+0.006
one+0.006
Neg logit contributions
robe-0.007
arl-0.006
 cooler-0.005
melon-0.005
ugg-0.005
Token said
Token ID531
Feature activation+3.02e-03

Pos logit contributions
robe+0.059
arl+0.055
melon+0.051
 cooler+0.051
ugg+0.050
Neg logit contributions
pieces-0.051
od-0.045
Maria-0.043
lier-0.040
one-0.039
TokenL
Token ID43
Feature activation-1.39e-02

Pos logit contributions
robe+0.013
arl+0.013
ugg+0.011
 cooler+0.011
 Cream+0.011
Neg logit contributions
pieces-0.012
od-0.011
lier-0.010
enta-0.010
ets-0.009
Tokenola
Token ID5708
Feature activation+1.64e-02

Pos logit contributions
robe+0.019
arl+0.018
ugg+0.015
melon+0.012
 waitress+0.012
Neg logit contributions
pieces-0.030
 rect-0.025
enta-0.025
lier-0.025
ples-0.025
Token smiled
Token ID13541
Feature activation-2.30e-02

Pos logit contributions
pieces+0.029
Maria+0.024
od+0.024
one+0.023
asing+0.022
Neg logit contributions
robe-0.027
ugg-0.024
melon-0.024
 cooler-0.023
 Cream-0.022
Token and
Token ID290
Feature activation-3.08e-03

Pos logit contributions
robe+0.051
arl+0.048
melon+0.046
ugg+0.045
 cooler+0.045
Neg logit contributions
pieces-0.027
od-0.024
Maria-0.022
lier-0.021
 circles-0.020
Token sn
Token ID3013
Feature activation-4.76e-02

Pos logit contributions
robe+0.006
 cooler+0.005
arl+0.005
ugg+0.005
melon+0.005
Neg logit contributions
pieces-0.005
od-0.004
Maria-0.004
one-0.004
lier-0.004
Tokenugg
Token ID6837
Feature activation-2.15e-01

Pos logit contributions
robe+0.032
 cooler+0.021
 ranc+0.021
 cabin+0.021
arl+0.020
Neg logit contributions
pieces-0.126
od-0.117
Maria-0.115
lier-0.112
enta-0.112
Tokenled
Token ID992
Feature activation-3.68e-02

Pos logit contributions
robe+0.397
arl+0.370
melon+0.349
 cooler+0.347
ugg+0.339
Neg logit contributions
pieces-0.334
od-0.298
Maria-0.283
lier-0.261
one-0.258
Token in
Token ID287
Feature activation+1.64e-03

Pos logit contributions
robe+0.063
arl+0.057
 cooler+0.053
melon+0.053
ugg+0.052
Neg logit contributions
pieces-0.064
od-0.057
Maria-0.055
one-0.051
lier-0.051
Token the
Token ID262
Feature activation-1.32e-02

Pos logit contributions
pieces+0.003
od+0.003
Maria+0.003
one+0.002
lier+0.002
Neg logit contributions
robe-0.003
arl-0.002
 cooler-0.002
melon-0.002
ugg-0.002
Token warmth
Token ID23125
Feature activation+1.35e-02

Pos logit contributions
robe+0.020
arl+0.019
melon+0.017
ugg+0.016
 ranc+0.016
Neg logit contributions
pieces-0.024
Maria-0.021
 circles-0.021
od-0.020
lier-0.019
Token of
Token ID286
Feature activation+1.35e-02

Pos logit contributions
pieces+0.022
od+0.019
Maria+0.019
enta+0.017
one+0.016
Neg logit contributions
robe-0.026
arl-0.023
 cooler-0.023
melon-0.023
ugg-0.022
Token -
Token ID532
Feature activation-4.79e-02

Pos logit contributions
robe+0.053
arl+0.049
melon+0.047
 cooler+0.046
ugg+0.045
Neg logit contributions
pieces-0.042
od-0.036
Maria-0.036
enta-0.032
one-0.032
Token just
Token ID655
Feature activation+2.54e-02

Pos logit contributions
robe+0.079
arl+0.074
melon+0.067
ugg+0.066
 cooler+0.063
Neg logit contributions
pieces-0.084
od-0.075
Maria-0.073
 circles-0.069
one-0.069
Token perfect
Token ID2818
Feature activation+6.62e-02

Pos logit contributions
pieces+0.047
od+0.042
Maria+0.041
lier+0.037
one+0.037
Neg logit contributions
robe-0.041
arl-0.037
 cooler-0.036
melon-0.035
ugg-0.034
Token for
Token ID329
Feature activation+3.59e-02

Pos logit contributions
pieces+0.110
od+0.098
Maria+0.094
lier+0.088
one+0.088
Neg logit contributions
robe-0.115
arl-0.107
melon-0.099
ugg-0.098
 cooler-0.098
Token sn
Token ID3013
Feature activation-4.16e-02

Pos logit contributions
pieces+0.068
od+0.061
Maria+0.059
one+0.056
lier+0.056
Neg logit contributions
robe-0.055
arl-0.050
 cooler-0.048
melon-0.047
ugg-0.045
Tokenugg
Token ID6837
Feature activation-2.14e-01

Pos logit contributions
robe+0.065
melon+0.056
 cooler+0.055
 waitress+0.054
 ranc+0.054
Neg logit contributions
pieces-0.074
lier-0.063
ets-0.062
od-0.062
inda-0.061
Tokenling
Token ID1359
Feature activation-2.35e-02

Pos logit contributions
robe+0.437
arl+0.411
melon+0.390
 cooler+0.386
ugg+0.383
Neg logit contributions
pieces-0.291
od-0.255
Maria-0.239
lier-0.224
one-0.216
Token.
Token ID13
Feature activation-1.15e-02

Pos logit contributions
robe+0.040
arl+0.036
 cooler+0.035
melon+0.034
ugg+0.033
Neg logit contributions
pieces-0.041
od-0.037
Maria-0.035
one-0.033
enta-0.032
Token The
Token ID383
Feature activation-8.64e-03

Pos logit contributions
robe+0.019
arl+0.018
melon+0.017
 cooler+0.016
ugg+0.016
Neg logit contributions
pieces-0.020
od-0.018
Maria-0.018
one-0.016
lier-0.016
Token baby
Token ID5156
Feature activation+3.58e-02

Pos logit contributions
robe+0.015
arl+0.014
melon+0.013
ugg+0.013
 cooler+0.011
Neg logit contributions
pieces-0.014
Maria-0.012
 hours-0.012
 pad-0.012
 circles-0.012
Token lamb
Token ID19343
Feature activation-3.13e-02

Pos logit contributions
pieces+0.063
od+0.056
Maria+0.054
one+0.049
lier+0.049
Neg logit contributions
robe-0.061
arl-0.054
 cooler-0.052
melon-0.052
ugg-0.051
Token his
Token ID465
Feature activation-3.47e-02

Pos logit contributions
robe+0.028
arl+0.028
melon+0.025
ugg+0.024
 ranc+0.023
Neg logit contributions
pieces-0.024
od-0.021
Maria-0.021
 circles-0.020
 hours-0.019
Token painting
Token ID12036
Feature activation+6.20e-02

Pos logit contributions
robe+0.048
arl+0.045
melon+0.041
ugg+0.040
 cooler+0.039
Neg logit contributions
pieces-0.070
od-0.062
Maria-0.061
 circles-0.057
lier-0.057
Token supplies
Token ID9416
Feature activation-5.43e-03

Pos logit contributions
pieces+0.099
od+0.087
Maria+0.084
one+0.076
lier+0.076
Neg logit contributions
robe-0.118
arl-0.105
 cooler-0.101
melon-0.100
ugg-0.099
Token,
Token ID11
Feature activation-1.87e-02

Pos logit contributions
robe+0.012
arl+0.011
melon+0.010
 cooler+0.010
ugg+0.010
Neg logit contributions
pieces-0.007
od-0.006
Maria-0.005
lier-0.005
one-0.005
Token sn
Token ID3013
Feature activation-4.23e-02

Pos logit contributions
robe+0.035
arl+0.031
 cooler+0.031
melon+0.030
ugg+0.029
Neg logit contributions
pieces-0.030
od-0.027
Maria-0.026
lier-0.024
one-0.023
Tokenugg
Token ID6837
Feature activation-2.14e-01

Pos logit contributions
robe+0.042
 ranc+0.035
 cooler+0.032
melon+0.032
 waitress+0.032
Neg logit contributions
pieces-0.097
lier-0.087
od-0.086
oser-0.086
ets-0.085
Tokenled
Token ID992
Feature activation-3.92e-02

Pos logit contributions
robe+0.428
arl+0.412
melon+0.398
ugg+0.385
 cooler+0.382
Neg logit contributions
pieces-0.265
od-0.240
Maria-0.220
lier-0.218
ets-0.211
Token up
Token ID510
Feature activation-7.25e-02

Pos logit contributions
robe+0.066
arl+0.064
melon+0.056
ugg+0.056
 Cream+0.053
Neg logit contributions
pieces-0.066
od-0.060
Maria-0.058
 circles-0.056
lier-0.056
Token in
Token ID287
Feature activation-1.13e-03

Pos logit contributions
robe+0.131
arl+0.123
melon+0.112
 cooler+0.111
ugg+0.111
Neg logit contributions
pieces-0.116
od-0.104
Maria-0.100
lier-0.091
 circles-0.090
Token his
Token ID465
Feature activation-3.48e-02

Pos logit contributions
robe+0.002
arl+0.002
 cooler+0.002
melon+0.002
ugg+0.001
Neg logit contributions
pieces-0.002
od-0.002
Maria-0.002
lier-0.002
one-0.002
Token bed
Token ID3996
Feature activation-1.76e-02

Pos logit contributions
robe+0.047
arl+0.046
melon+0.041
ugg+0.038
 Cream+0.037
Neg logit contributions
pieces-0.069
od-0.060
Maria-0.059
 circles-0.059
 pad-0.056
Tokenlex
Token ID2588
Feature activation+3.96e-02

Pos logit contributions
robe+0.006
arl+0.005
 cooler+0.005
melon+0.005
 ranc+0.005
Neg logit contributions
pieces-0.005
od-0.004
lier-0.004
Maria-0.004
ill-0.004
Tokenible
Token ID856
Feature activation+6.27e-02

Pos logit contributions
pieces+0.072
Maria+0.059
od+0.056
 circles+0.051
 rect+0.050
Neg logit contributions
robe-0.072
ugg-0.061
melon-0.061
arl-0.058
 Cream-0.058
Token.
Token ID13
Feature activation-1.48e-02

Pos logit contributions
pieces+0.106
Maria+0.098
od+0.092
Rachel+0.088
enta+0.085
Neg logit contributions
robe-0.110
arl-0.096
 cooler-0.096
melon-0.090
 Cream-0.089
Token
Token ID198
Feature activation-3.24e-04

Pos logit contributions
robe+0.021
arl+0.019
 Cream+0.018
ugg+0.018
melon+0.018
Neg logit contributions
pieces-0.030
od-0.027
Maria-0.026
inda-0.025
one-0.024
Token
Token ID198
Feature activation+6.25e-04

Pos logit contributions
robe+0.000
arl+0.000
 Cream+0.000
ugg+0.000
 Sleep+0.000
Neg logit contributions
pieces-0.001
od-0.001
asing-0.001
 rect-0.001
lier-0.001
TokenBut
Token ID1537
Feature activation+8.40e-02

Pos logit contributions
pieces+0.001
od+0.001
Maria+0.001
ill+0.001
lier+0.001
Neg logit contributions
robe-0.001
arl-0.001
 cooler-0.001
 Cream-0.001
melon-0.001
Token then
Token ID788
Feature activation-4.99e-03

Pos logit contributions
pieces+0.121
od+0.105
Maria+0.101
one+0.091
lier+0.091
Neg logit contributions
robe-0.169
arl-0.157
melon-0.148
 cooler-0.147
ugg-0.145
Token they
Token ID484
Feature activation+2.77e-03

Pos logit contributions
robe+0.008
arl+0.007
 cooler+0.007
melon+0.007
ugg+0.007
Neg logit contributions
pieces-0.009
od-0.008
Maria-0.008
lier-0.007
one-0.007
Token had
Token ID550
Feature activation+1.33e-02

Pos logit contributions
pieces+0.005
od+0.004
Maria+0.004
enta+0.004
asing+0.003
Neg logit contributions
robe-0.005
 cooler-0.005
 tired-0.004
melon-0.004
 ranc-0.004
Token an
Token ID281
Feature activation+1.09e-02

Pos logit contributions
pieces+0.025
od+0.023
Maria+0.022
lier+0.020
one+0.020
Neg logit contributions
robe-0.021
arl-0.019
 cooler-0.018
melon-0.017
ugg-0.017
Token idea
Token ID2126
Feature activation-1.54e-02

Pos logit contributions
pieces+0.022
od+0.020
Maria+0.019
one+0.018
lier+0.018
Neg logit contributions
robe-0.016
arl-0.014
ugg-0.013
melon-0.013
 cooler-0.012
Token reward
Token ID6721
Feature activation+2.82e-02

Pos logit contributions
robe+0.006
arl+0.006
melon+0.005
 cooler+0.005
ugg+0.005
Neg logit contributions
pieces-0.011
od-0.011
Maria-0.010
lier-0.010
one-0.010
Token for
Token ID329
Feature activation+4.98e-02

Pos logit contributions
pieces+0.047
Maria+0.042
od+0.042
enta+0.038
 rect+0.038
Neg logit contributions
robe-0.053
 cooler-0.044
melon-0.043
arl-0.041
ugg-0.040
Token studying
Token ID11065
Feature activation+1.81e-02

Pos logit contributions
pieces+0.096
od+0.087
Maria+0.084
lier+0.081
one+0.081
Neg logit contributions
robe-0.072
arl-0.065
 cooler-0.064
melon-0.061
 Sleep-0.058
Token was
Token ID373
Feature activation+2.62e-02

Pos logit contributions
pieces+0.031
od+0.028
Maria+0.028
enta+0.026
one+0.025
Neg logit contributions
robe-0.031
 cooler-0.026
arl-0.026
melon-0.024
 Cream-0.024
Token a
Token ID257
Feature activation+1.90e-02

Pos logit contributions
pieces+0.049
od+0.044
Maria+0.043
enta+0.039
one+0.039
Neg logit contributions
robe-0.041
 cooler-0.036
arl-0.035
melon-0.035
 Cream-0.032
Token massive
Token ID4858
Feature activation+8.46e-02

Pos logit contributions
pieces+0.040
od+0.038
Maria+0.036
asing+0.034
one+0.034
Neg logit contributions
robe-0.025
arl-0.022
 cooler-0.021
melon-0.020
ugg-0.019
Token plate
Token ID7480
Feature activation-4.31e-02

Pos logit contributions
pieces+0.167
od+0.152
Maria+0.148
one+0.138
enta+0.136
Neg logit contributions
robe-0.124
arl-0.111
 cooler-0.106
melon-0.104
ugg-0.099
Token of
Token ID286
Feature activation+1.89e-02

Pos logit contributions
robe+0.089
arl+0.080
melon+0.078
 cooler+0.077
ugg+0.074
Neg logit contributions
pieces-0.062
od-0.053
Maria-0.052
enta-0.047
one-0.046
Token her
Token ID607
Feature activation-5.70e-03

Pos logit contributions
pieces+0.041
od+0.038
Maria+0.037
one+0.035
enta+0.035
Neg logit contributions
robe-0.024
 cooler-0.021
arl-0.020
melon-0.020
 Cream-0.018
Token favourite
Token ID12507
Feature activation+2.91e-02

Pos logit contributions
robe+0.007
arl+0.006
 cooler+0.006
melon+0.006
ugg+0.005
Neg logit contributions
pieces-0.012
od-0.012
Maria-0.012
one-0.011
enta-0.011
Token pasta
Token ID26296
Feature activation+4.75e-04

Pos logit contributions
pieces+0.061
od+0.057
Maria+0.056
one+0.052
asing+0.052
Neg logit contributions
robe-0.040
 cooler-0.034
melon-0.031
arl-0.031
 refrigerator-0.030
Token,
Token ID11
Feature activation-1.58e-02

Pos logit contributions
pieces+0.077
od+0.069
Maria+0.065
lier+0.061
one+0.061
Neg logit contributions
robe-0.103
arl-0.097
melon-0.090
 cooler-0.089
ugg-0.088
Token you
Token ID345
Feature activation+4.59e-02

Pos logit contributions
robe+0.030
arl+0.028
melon+0.026
ugg+0.026
 cooler+0.025
Neg logit contributions
pieces-0.024
od-0.021
Maria-0.020
one-0.019
 circles-0.018
Token are
Token ID389
Feature activation+2.73e-02

Pos logit contributions
pieces+0.056
od+0.049
Maria+0.044
lier+0.043
ata+0.042
Neg logit contributions
robe-0.100
arl-0.094
melon-0.092
ugg-0.088
 cooler-0.087
Token brave
Token ID14802
Feature activation+2.43e-02

Pos logit contributions
pieces+0.042
od+0.040
Maria+0.037
 circles+0.036
lier+0.035
Neg logit contributions
robe-0.052
arl-0.049
ugg-0.048
melon-0.044
 Cream-0.042
Token.
Token ID13
Feature activation+7.39e-04

Pos logit contributions
pieces+0.041
od+0.035
Maria+0.035
enta+0.032
one+0.030
Neg logit contributions
robe-0.044
arl-0.041
 cooler-0.039
melon-0.036
 cabin-0.036
Token But
Token ID887
Feature activation+9.97e-02

Pos logit contributions
pieces+0.001
od+0.001
Maria+0.001
one+0.001
 circles+0.001
Neg logit contributions
robe-0.001
arl-0.001
melon-0.001
 cooler-0.001
ugg-0.001
Token you
Token ID345
Feature activation+4.86e-02

Pos logit contributions
pieces+0.144
od+0.129
Maria+0.122
one+0.118
 circles+0.112
Neg logit contributions
robe-0.194
arl-0.178
ugg-0.167
melon-0.165
 cooler-0.160
Token should
Token ID815
Feature activation+3.07e-03

Pos logit contributions
pieces+0.065
od+0.058
Maria+0.053
ata+0.050
lier+0.049
Neg logit contributions
robe-0.102
arl-0.095
melon-0.089
ugg-0.088
 cooler-0.087
Token not
Token ID407
Feature activation-8.97e-02

Pos logit contributions
pieces+0.005
od+0.005
Maria+0.004
e+0.004
one+0.004
Neg logit contributions
robe-0.005
arl-0.004
ugg-0.004
melon-0.004
 Cream-0.004
Token climb
Token ID12080
Feature activation+7.36e-02

Pos logit contributions
robe+0.139
arl+0.132
melon+0.123
 cooler+0.122
ugg+0.118
Neg logit contributions
pieces-0.169
od-0.149
Maria-0.147
enta-0.134
inda-0.133
Token the
Token ID262
Feature activation-1.18e-02

Pos logit contributions
pieces+0.134
od+0.120
Maria+0.118
one+0.106
inda+0.106
Neg logit contributions
robe-0.120
arl-0.110
 cooler-0.105
melon-0.102
ugg-0.097
Token three
Token ID1115
Feature activation+3.68e-02

Pos logit contributions
robe+0.016
 cooler+0.015
arl+0.015
 Cream+0.013
 Sleep+0.013
Neg logit contributions
pieces-0.025
Maria-0.024
od-0.023
lier-0.022
lar-0.022
Token,
Token ID11
Feature activation-1.22e-02

Pos logit contributions
pieces+0.066
Maria+0.063
Rachel+0.058
od+0.058
iece+0.057
Neg logit contributions
 cooler-0.050
robe-0.050
arl-0.049
melon-0.044
 Cream-0.044
Token four
Token ID1440
Feature activation+8.36e-02

Pos logit contributions
robe+0.016
arl+0.015
 cooler+0.015
 Cream+0.013
melon+0.013
Neg logit contributions
pieces-0.024
Maria-0.023
od-0.022
lar-0.021
lier-0.021
Token,"
Token ID553
Feature activation+7.43e-03

Pos logit contributions
Maria+0.157
pieces+0.154
Rachel+0.148
od+0.142
Jud+0.140
Neg logit contributions
 cooler-0.110
robe-0.106
arl-0.103
melon-0.092
 ranc-0.092
Token she
Token ID673
Feature activation+3.28e-02

Pos logit contributions
pieces+0.019
od+0.017
Maria+0.017
lier+0.017
lar+0.016
Neg logit contributions
robe-0.006
arl-0.005
 cooler-0.005
melon-0.005
ugg-0.004
Token counts
Token ID9853
Feature activation+1.02e-01

Pos logit contributions
pieces+0.040
od+0.034
Maria+0.032
lier+0.028
one+0.028
Neg logit contributions
robe-0.073
arl-0.068
 cooler-0.065
melon-0.065
ugg-0.063
Token.
Token ID13
Feature activation+5.01e-03

Pos logit contributions
pieces+0.178
Maria+0.155
od+0.155
lier+0.144
lar+0.141
Neg logit contributions
robe-0.170
arl-0.160
 cooler-0.150
melon-0.148
 Cream-0.138
Token
Token ID198
Feature activation+1.64e-02

Pos logit contributions
pieces+0.011
od+0.009
lar+0.009
Maria+0.009
lier+0.009
Neg logit contributions
robe-0.007
arl-0.006
 cooler-0.006
 Cream-0.006
melon-0.006
Token
Token ID198
Feature activation+1.58e-02

Pos logit contributions
pieces+0.040
od+0.034
asing+0.033
lar+0.033
lier+0.033
Neg logit contributions
arl-0.018
robe-0.018
 Cream-0.015
ugg-0.015
 Sleep-0.014
Token"
Token ID1
Feature activation-3.37e-02

Pos logit contributions
pieces+0.030
od+0.026
Maria+0.024
ez+0.024
lar+0.023
Neg logit contributions
robe-0.024
arl-0.023
 cooler-0.021
 Cream-0.021
melon-0.021
TokenLook
Token ID8567
Feature activation+4.57e-02

Pos logit contributions
robe+0.046
arl+0.043
 Cream+0.041
 cooler+0.037
melon+0.036
Neg logit contributions
pieces-0.069
ez-0.061
od-0.059
 tal-0.057
lar-0.057
Tokenie
Token ID494
Feature activation+4.42e-02

Pos logit contributions
pieces+0.060
 mastered+0.051
Maria+0.049
 circles+0.049
 rect+0.048
Neg logit contributions
robe-0.044
arl-0.038
melon-0.036
ugg-0.035
 cabin-0.033
Token,
Token ID11
Feature activation-2.03e-02

Pos logit contributions
pieces+0.057
od+0.050
Maria+0.047
lier+0.042
one+0.042
Neg logit contributions
robe-0.095
arl-0.090
 cooler-0.084
melon-0.084
ugg-0.082
Token who
Token ID508
Feature activation-1.68e-02

Pos logit contributions
robe+0.042
arl+0.039
melon+0.037
 cooler+0.037
ugg+0.036
Neg logit contributions
pieces-0.028
od-0.024
Maria-0.023
lier-0.021
one-0.021
Token wanted
Token ID2227
Feature activation-3.91e-02

Pos logit contributions
arl+0.043
rier+0.037
melon+0.037
robe+0.036
 cooler+0.034
Neg logit contributions
e-0.017
�-0.016
's-0.015
od-0.015
ez-0.015
Token to
Token ID284
Feature activation+2.42e-02

Pos logit contributions
robe+0.064
 cooler+0.056
arl+0.056
melon+0.055
ugg+0.053
Neg logit contributions
pieces-0.070
od-0.062
Maria-0.060
lier-0.056
enta-0.055
Token reach
Token ID3151
Feature activation+9.31e-02

Pos logit contributions
pieces+0.041
od+0.036
Maria+0.035
one+0.032
 circles+0.032
Neg logit contributions
robe-0.043
arl-0.040
melon-0.037
ugg-0.036
 cooler-0.036
Token a
Token ID257
Feature activation-1.94e-03

Pos logit contributions
pieces+0.180
od+0.162
Maria+0.160
lier+0.149
enta+0.147
Neg logit contributions
robe-0.136
 cooler-0.124
arl-0.121
melon-0.116
 cabin-0.109
Token jar
Token ID17379
Feature activation-3.22e-02

Pos logit contributions
robe+0.003
arl+0.002
ugg+0.002
melon+0.002
 Sleep+0.002
Neg logit contributions
pieces-0.004
od-0.004
Maria-0.003
lier-0.003
one-0.003
Token of
Token ID286
Feature activation+9.34e-03

Pos logit contributions
robe+0.068
arl+0.064
melon+0.060
ugg+0.060
 cooler+0.059
Neg logit contributions
pieces-0.042
od-0.036
Maria-0.033
one-0.031
lier-0.031
Token cookies
Token ID14746
Feature activation-1.72e-02

Pos logit contributions
pieces+0.017
od+0.015
Maria+0.015
 circles+0.014
one+0.014
Neg logit contributions
robe-0.015
arl-0.014
melon-0.013
ugg-0.013
 cooler-0.012
Token that
Token ID326
Feature activation-1.07e-01

Pos logit contributions
robe+0.033
arl+0.031
melon+0.029
 cooler+0.029
ugg+0.028
Neg logit contributions
pieces-0.026
od-0.023
Maria-0.022
lier-0.020
one-0.020
Token mom
Token ID1995
Feature activation-4.89e-02

Pos logit contributions
pieces+0.014
od+0.012
Maria+0.012
 circles+0.011
enta+0.011
Neg logit contributions
robe-0.007
arl-0.006
melon-0.006
ugg-0.005
 cooler-0.005
Token smiled
Token ID13541
Feature activation-2.51e-02

Pos logit contributions
robe+0.080
melon+0.073
 cooler+0.072
ugg+0.072
 Cream+0.068
Neg logit contributions
pieces-0.083
od-0.070
Maria-0.068
one-0.065
 conclusions-0.064
Token and
Token ID290
Feature activation+3.98e-04

Pos logit contributions
robe+0.062
arl+0.059
melon+0.057
ugg+0.056
 cooler+0.054
Neg logit contributions
pieces-0.022
od-0.020
Maria-0.018
lier-0.017
ill-0.016
Token said
Token ID531
Feature activation+1.69e-02

Pos logit contributions
pieces+0.001
od+0.001
Maria+0.001
one+0.000
lier+0.000
Neg logit contributions
robe-0.001
 cooler-0.001
arl-0.001
melon-0.001
ugg-0.001
Token,
Token ID11
Feature activation-1.58e-02

Pos logit contributions
pieces+0.026
od+0.023
Maria+0.021
lier+0.020
one+0.019
Neg logit contributions
robe-0.033
arl-0.030
 cooler-0.029
melon-0.028
ugg-0.028
Token "
Token ID366
Feature activation-2.18e-02

Pos logit contributions
robe+0.022
arl+0.020
 cooler+0.019
melon+0.018
ugg+0.018
Neg logit contributions
pieces-0.032
od-0.029
Maria-0.028
one-0.026
lier-0.026
TokenThat
Token ID2504
Feature activation-7.91e-02

Pos logit contributions
robe+0.028
arl+0.024
ugg+0.024
 Sleep+0.024
 Cream+0.023
Neg logit contributions
pieces-0.046
od-0.042
ets-0.037
inations-0.037
ata-0.037
Token's
Token ID338
Feature activation+2.29e-02

Pos logit contributions
robe+0.176
 cooler+0.155
melon+0.153
 Sleep+0.153
 Cream+0.152
Neg logit contributions
pieces-0.100
Maria-0.082
od-0.075
ill-0.072
asing-0.072
Token a
Token ID257
Feature activation+7.02e-03

Pos logit contributions
pieces+0.036
od+0.033
Maria+0.032
lier+0.029
inda+0.029
Neg logit contributions
robe-0.043
arl-0.040
ugg-0.039
melon-0.038
 cooler-0.035
Token beautiful
Token ID4950
Feature activation+3.46e-02

Pos logit contributions
pieces+0.013
od+0.012
Maria+0.011
lier+0.011
 circles+0.011
Neg logit contributions
robe-0.011
arl-0.010
ugg-0.009
melon-0.009
 cooler-0.009
Token drawing
Token ID8263
Feature activation+3.30e-02

Pos logit contributions
pieces+0.067
Maria+0.059
od+0.058
lier+0.056
asing+0.055
Neg logit contributions
robe-0.053
 cooler-0.044
arl-0.044
melon-0.044
ugg-0.041
Token smile
Token ID8212
Feature activation-1.03e-03

Pos logit contributions
robe+0.005
arl+0.005
melon+0.004
 cooler+0.004
ugg+0.004
Neg logit contributions
pieces-0.009
od-0.008
Maria-0.008
lier-0.007
one-0.007
Token on
Token ID319
Feature activation+3.87e-02

Pos logit contributions
robe+0.002
arl+0.002
melon+0.002
 cooler+0.002
ugg+0.002
Neg logit contributions
pieces-0.002
od-0.001
Maria-0.001
lier-0.001
one-0.001
Token her
Token ID607
Feature activation-2.51e-02

Pos logit contributions
pieces+0.062
od+0.055
Maria+0.052
lier+0.048
one+0.048
Neg logit contributions
robe-0.072
arl-0.066
melon-0.062
ugg-0.062
 cooler-0.062
Token face
Token ID1986
Feature activation+8.70e-03

Pos logit contributions
robe+0.031
arl+0.026
 cooler+0.024
melon+0.023
ugg+0.022
Neg logit contributions
pieces-0.056
od-0.052
Maria-0.052
one-0.049
lier-0.048
Token.
Token ID13
Feature activation+1.26e-03

Pos logit contributions
pieces+0.014
Maria+0.013
od+0.012
enta+0.011
one+0.011
Neg logit contributions
robe-0.016
arl-0.014
 cooler-0.014
melon-0.014
ugg-0.013
Token Lily
Token ID20037
Feature activation+3.16e-02

Pos logit contributions
pieces+0.002
od+0.002
Maria+0.002
one+0.002
lier+0.002
Neg logit contributions
robe-0.002
arl-0.002
 cooler-0.002
melon-0.002
ugg-0.002
Token loved
Token ID6151
Feature activation-3.91e-02

Pos logit contributions
pieces+0.062
od+0.053
Maria+0.053
one+0.049
asing+0.048
Neg logit contributions
robe-0.048
 cooler-0.040
melon-0.039
ugg-0.039
 Cream-0.038
Token to
Token ID284
Feature activation+2.01e-02

Pos logit contributions
robe+0.070
arl+0.064
melon+0.060
 cooler+0.059
ugg+0.059
Neg logit contributions
pieces-0.065
od-0.058
Maria-0.056
one-0.051
lier-0.051
Token comb
Token ID1974
Feature activation-1.90e-03

Pos logit contributions
pieces+0.041
od+0.038
Maria+0.038
lier+0.036
one+0.036
Neg logit contributions
robe-0.027
arl-0.024
 cooler-0.023
melon-0.022
ugg-0.021
Token Anna
Token ID11735
Feature activation+2.18e-02

Pos logit contributions
robe+0.003
arl+0.003
melon+0.003
 cooler+0.003
ugg+0.003
Neg logit contributions
pieces-0.003
od-0.003
Maria-0.003
lier-0.003
one-0.003
Token's
Token ID338
Feature activation+1.44e-02

Pos logit contributions
pieces+0.021
Maria+0.019
od+0.018
one+0.016
lar+0.014
Neg logit contributions
robe-0.052
 cooler-0.047
melon-0.046
arl-0.046
ugg-0.045
Token He
Token ID679
Feature activation-2.20e-02

Pos logit contributions
pieces+0.046
od+0.042
Maria+0.041
enta+0.039
lier+0.039
Neg logit contributions
robe-0.024
arl-0.021
 cooler-0.018
melon-0.018
ugg-0.018
Token ran
Token ID4966
Feature activation+2.82e-03

Pos logit contributions
robe+0.034
 cooler+0.032
melon+0.029
 ranc+0.029
 cabin+0.028
Neg logit contributions
pieces-0.043
od-0.036
Maria-0.034
iece-0.033
 conclusions-0.033
Token to
Token ID284
Feature activation+3.20e-02

Pos logit contributions
pieces+0.004
od+0.004
Maria+0.004
lier+0.003
 circles+0.003
Neg logit contributions
robe-0.005
arl-0.005
melon-0.005
ugg-0.005
 cooler-0.005
Token the
Token ID262
Feature activation-3.74e-03

Pos logit contributions
pieces+0.060
od+0.054
Maria+0.052
lier+0.049
one+0.049
Neg logit contributions
robe-0.051
arl-0.046
 cooler-0.043
melon-0.043
ugg-0.041
Token kitchen
Token ID9592
Feature activation-3.56e-02

Pos logit contributions
robe+0.005
arl+0.005
ugg+0.004
melon+0.004
 cooler+0.004
Neg logit contributions
pieces-0.008
od-0.007
Maria-0.007
 circles-0.006
one-0.006
Token and
Token ID290
Feature activation+6.84e-03

Pos logit contributions
robe+0.078
arl+0.074
ugg+0.068
melon+0.068
 cooler+0.068
Neg logit contributions
pieces-0.043
od-0.038
Maria-0.036
lier-0.031
ill-0.031
Token grabbed
Token ID13176
Feature activation+2.65e-02

Pos logit contributions
pieces+0.011
od+0.010
Maria+0.009
lier+0.009
one+0.008
Neg logit contributions
robe-0.013
 cooler-0.011
melon-0.011
arl-0.011
ugg-0.011
Token some
Token ID617
Feature activation-2.95e-02

Pos logit contributions
pieces+0.049
od+0.045
Maria+0.043
lier+0.040
one+0.040
Neg logit contributions
robe-0.042
arl-0.038
 cooler-0.035
melon-0.035
ugg-0.034
Token lot
Token ID1256
Feature activation+6.98e-02

Pos logit contributions
robe+0.040
 cooler+0.038
arl+0.036
melon+0.034
 refrigerator+0.032
Neg logit contributions
pieces-0.060
Maria-0.057
od-0.057
enta-0.052
Rachel-0.052
Tokenion
Token ID295
Feature activation-9.17e-02

Pos logit contributions
pieces+0.121
od+0.105
Maria+0.104
enta+0.093
lar+0.093
Neg logit contributions
robe-0.119
arl-0.111
melon-0.107
 cooler-0.105
ugg-0.101
Token.
Token ID13
Feature activation+1.81e-02

Pos logit contributions
robe+0.181
arl+0.170
ugg+0.155
 cooler+0.154
melon+0.152
Neg logit contributions
pieces-0.131
od-0.115
Maria-0.105
lier-0.103
one-0.101
Token hear
Token ID3285
Feature activation+7.10e-02

Pos logit contributions
robe+0.080
arl+0.075
melon+0.070
ugg+0.068
 Cream+0.067
Neg logit contributions
pieces-0.056
od-0.048
Maria-0.044
lier-0.043
ata-0.043
Token the
Token ID262
Feature activation-1.84e-02

Pos logit contributions
pieces+0.140
od+0.127
Maria+0.123
one+0.115
lier+0.115
Neg logit contributions
robe-0.105
arl-0.095
 cooler-0.087
melon-0.087
ugg-0.085
Token ball
Token ID2613
Feature activation-8.05e-03

Pos logit contributions
robe+0.027
arl+0.025
ugg+0.024
melon+0.024
 Cream+0.022
Neg logit contributions
pieces-0.033
 circles-0.029
od-0.028
Maria-0.028
lier-0.026
Token and
Token ID290
Feature activation-2.91e-03

Pos logit contributions
robe+0.013
arl+0.012
 cooler+0.012
melon+0.012
ugg+0.011
Neg logit contributions
pieces-0.015
od-0.013
Maria-0.013
enta-0.012
one-0.012
Token the
Token ID262
Feature activation-2.05e-02

Pos logit contributions
robe+0.005
arl+0.004
 cooler+0.004
melon+0.004
ugg+0.004
Neg logit contributions
pieces-0.005
od-0.005
Maria-0.005
one-0.004
lier-0.004
Token sticks
Token ID16461
Feature activation+4.60e-02

Pos logit contributions
robe+0.030
arl+0.028
ugg+0.026
melon+0.026
 cooler+0.025
Neg logit contributions
pieces-0.039
od-0.034
Maria-0.033
 circles-0.032
lier-0.031
Token outside
Token ID2354
Feature activation+2.84e-03

Pos logit contributions
pieces+0.076
od+0.067
Maria+0.067
enta+0.060
one+0.059
Neg logit contributions
robe-0.083
arl-0.077
 cooler-0.074
melon-0.073
ugg-0.068
Token.
Token ID13
Feature activation+1.78e-03

Pos logit contributions
pieces+0.005
od+0.005
Maria+0.004
enta+0.004
one+0.004
Neg logit contributions
robe-0.005
arl-0.004
 cooler-0.004
melon-0.004
 cabin-0.004
Token They
Token ID1119
Feature activation-3.74e-02

Pos logit contributions
pieces+0.004
od+0.003
Maria+0.003
lier+0.003
one+0.003
Neg logit contributions
robe-0.003
arl-0.002
melon-0.002
 cooler-0.002
ugg-0.002
Token wonder
Token ID4240
Feature activation+2.56e-02

Pos logit contributions
robe+0.073
arl+0.066
melon+0.062
ugg+0.062
 Cream+0.060
Neg logit contributions
pieces-0.056
od-0.049
Maria-0.047
 circles-0.044
ill-0.044
Token who
Token ID508
Feature activation-2.10e-02

Pos logit contributions
pieces+0.057
Maria+0.052
od+0.051
enta+0.050
ez+0.049
Neg logit contributions
robe-0.029
arl-0.028
 cooler-0.027
melon-0.024
 cabin-0.024
Token a
Token ID257
Feature activation+8.69e-04

Pos logit contributions
pieces+0.083
od+0.074
Maria+0.072
one+0.066
lier+0.066
Neg logit contributions
robe-0.076
arl-0.069
melon-0.064
 cooler-0.064
ugg-0.063
Token monkey
Token ID21657
Feature activation+4.03e-02

Pos logit contributions
pieces+0.002
od+0.002
Maria+0.002
asing+0.002
one+0.002
Neg logit contributions
robe-0.001
 cooler-0.001
arl-0.001
melon-0.001
 cabin-0.001
Token play
Token ID711
Feature activation-4.82e-02

Pos logit contributions
pieces+0.066
Maria+0.061
od+0.058
Rachel+0.053
inda+0.053
Neg logit contributions
robe-0.070
 cooler-0.064
arl-0.062
melon-0.062
 cabin-0.058
Token with
Token ID351
Feature activation-4.39e-02

Pos logit contributions
robe+0.077
arl+0.068
 cooler+0.065
melon+0.065
ugg+0.062
Neg logit contributions
pieces-0.089
od-0.080
Maria-0.079
one-0.073
enta-0.072
Token a
Token ID257
Feature activation+7.72e-05

Pos logit contributions
robe+0.062
 cooler+0.055
arl+0.052
melon+0.052
 Cream+0.047
Neg logit contributions
pieces-0.087
od-0.077
Maria-0.077
inda-0.073
asing-0.073
Token ball
Token ID2613
Feature activation-7.46e-03

Pos logit contributions
pieces+0.000
Maria+0.000
od+0.000
asing+0.000
ill+0.000
Neg logit contributions
robe-0.000
 cooler-0.000
arl-0.000
 cabin-0.000
melon-0.000
Token.
Token ID13
Feature activation+5.76e-03

Pos logit contributions
robe+0.014
arl+0.012
melon+0.012
 cooler+0.012
ugg+0.011
Neg logit contributions
pieces-0.012
Maria-0.011
od-0.011
enta-0.010
Rachel-0.009
Token She
Token ID1375
Feature activation-1.96e-02

Pos logit contributions
pieces+0.012
od+0.011
Maria+0.011
one+0.010
enta+0.010
Neg logit contributions
robe-0.008
arl-0.007
 cooler-0.006
melon-0.006
ugg-0.006
Token cl
Token ID537
Feature activation+4.01e-02

Pos logit contributions
robe+0.030
 cooler+0.027
melon+0.026
 bar+0.026
 Cream+0.025
Neg logit contributions
pieces-0.038
iece-0.031
od-0.031
Maria-0.030
asing-0.029
Tokenapped
Token ID6320
Feature activation-6.92e-02

Pos logit contributions
pieces+0.079
od+0.071
Maria+0.069
one+0.064
lier+0.064
Neg logit contributions
robe-0.060
arl-0.055
melon-0.050
 cooler-0.050
ugg-0.049
Token and
Token ID290
Feature activation-6.38e-04

Pos logit contributions
robe+0.121
arl+0.111
 cooler+0.104
melon+0.104
ugg+0.101
Neg logit contributions
pieces-0.118
od-0.105
Maria-0.101
one-0.093
lier-0.093
Token.
Token ID13
Feature activation+1.18e-02

Pos logit contributions
pieces+0.041
od+0.035
Maria+0.035
enta+0.032
one+0.031
Neg logit contributions
robe-0.040
arl-0.035
 cooler-0.034
melon-0.032
ugg-0.032
Token They
Token ID1119
Feature activation-2.41e-02

Pos logit contributions
pieces+0.024
od+0.022
Maria+0.021
one+0.020
lier+0.020
Neg logit contributions
robe-0.017
arl-0.015
 cooler-0.014
melon-0.014
ugg-0.014
Token do
Token ID466
Feature activation+2.31e-02

Pos logit contributions
robe+0.047
arl+0.043
melon+0.040
ugg+0.040
 Cream+0.039
Neg logit contributions
pieces-0.036
od-0.032
Maria-0.030
lier-0.028
ata-0.028
Token not
Token ID407
Feature activation-8.80e-02

Pos logit contributions
pieces+0.054
od+0.051
Maria+0.048
ata+0.047
one+0.046
Neg logit contributions
robe-0.026
arl-0.023
melon-0.021
ugg-0.020
 cooler-0.020
Token care
Token ID1337
Feature activation-3.52e-02

Pos logit contributions
robe+0.160
arl+0.148
melon+0.138
 cooler+0.138
ugg+0.136
Neg logit contributions
pieces-0.144
od-0.127
Maria-0.122
lier-0.112
one-0.112
Token about
Token ID546
Feature activation-5.26e-02

Pos logit contributions
robe+0.054
arl+0.050
 cooler+0.042
ugg+0.042
melon+0.041
Neg logit contributions
pieces-0.069
Maria-0.060
od-0.059
enta-0.057
lier-0.055
Token Dad
Token ID17415
Feature activation-1.42e-02

Pos logit contributions
robe+0.082
arl+0.074
 cooler+0.070
melon+0.070
ugg+0.066
Neg logit contributions
pieces-0.099
od-0.088
Maria-0.086
one-0.081
lier-0.080
Token.
Token ID13
Feature activation+1.22e-02

Pos logit contributions
robe+0.030
 cooler+0.028
melon+0.027
arl+0.027
 Cream+0.026
Neg logit contributions
pieces-0.018
Maria-0.016
od-0.015
inda-0.013
enta-0.013
Token
Token ID198
Feature activation+2.47e-02

Pos logit contributions
pieces+0.023
od+0.021
Maria+0.021
lier+0.019
one+0.019
Neg logit contributions
robe-0.019
arl-0.017
melon-0.016
 cooler-0.016
ugg-0.015
Token
Token ID198
Feature activation+2.40e-02

Pos logit contributions
pieces+0.041
od+0.036
Maria+0.035
one+0.032
lier+0.032
Neg logit contributions
robe-0.044
arl-0.041
melon-0.038
 cooler-0.038
ugg-0.037
TokenMom
Token ID29252
Feature activation-1.96e-02

Pos logit contributions
pieces+0.038
od+0.034
Maria+0.032
lier+0.030
one+0.029
Neg logit contributions
robe-0.045
arl-0.042
 cooler-0.039
ugg-0.039
melon-0.038
Token girl
Token ID2576
Feature activation+4.82e-02

Pos logit contributions
robe+0.031
arl+0.027
melon+0.025
ugg+0.025
 Sleep+0.023
Neg logit contributions
pieces-0.031
od-0.028
lier-0.026
Maria-0.025
one-0.024
Token who
Token ID508
Feature activation-3.07e-02

Pos logit contributions
pieces+0.085
Maria+0.075
od+0.074
one+0.067
ata+0.066
Neg logit contributions
robe-0.079
 cooler-0.069
ugg-0.068
arl-0.068
melon-0.068
Token liked
Token ID8288
Feature activation-4.14e-02

Pos logit contributions
arl+0.059
robe+0.057
melon+0.053
 cooler+0.050
ugg+0.047
Neg logit contributions
pieces-0.042
od-0.041
Maria-0.038
e-0.037
lier-0.036
Token to
Token ID284
Feature activation+1.20e-02

Pos logit contributions
robe+0.087
arl+0.083
ugg+0.075
melon+0.075
 cooler+0.074
Neg logit contributions
pieces-0.054
od-0.047
Maria-0.045
 circles-0.042
lier-0.039
Token play
Token ID711
Feature activation-5.43e-02

Pos logit contributions
pieces+0.024
od+0.021
Maria+0.021
lier+0.020
one+0.020
Neg logit contributions
robe-0.018
arl-0.015
 cooler-0.015
melon-0.014
ugg-0.014
Token with
Token ID351
Feature activation-4.83e-02

Pos logit contributions
robe+0.085
arl+0.077
melon+0.071
 cooler+0.071
ugg+0.069
Neg logit contributions
pieces-0.103
od-0.093
Maria-0.090
one-0.084
lier-0.084
Token her
Token ID607
Feature activation-3.38e-02

Pos logit contributions
robe+0.076
arl+0.068
 cooler+0.064
melon+0.062
ugg+0.060
Neg logit contributions
pieces-0.092
od-0.081
Maria-0.079
one-0.075
lier-0.075
Token toys
Token ID14958
Feature activation-1.00e-02

Pos logit contributions
robe+0.048
arl+0.044
melon+0.040
ugg+0.040
 cooler+0.039
Neg logit contributions
pieces-0.068
od-0.061
Maria-0.059
lier-0.055
one-0.055
Token.
Token ID13
Feature activation-1.58e-02

Pos logit contributions
robe+0.019
arl+0.017
 cooler+0.016
melon+0.016
ugg+0.016
Neg logit contributions
pieces-0.016
od-0.014
Maria-0.014
one-0.012
lier-0.012
Token She
Token ID1375
Feature activation-3.20e-02

Pos logit contributions
robe+0.027
arl+0.026
melon+0.024
ugg+0.023
 cooler+0.023
Neg logit contributions
pieces-0.026
od-0.023
Maria-0.023
lier-0.022
one-0.021
Token had
Token ID550
Feature activation+6.53e-03

Pos logit contributions
robe+0.058
 cooler+0.050
ugg+0.048
melon+0.048
arl+0.047
Neg logit contributions
pieces-0.056
od-0.047
Maria-0.046
lier-0.043
one-0.042
Token "
Token ID366
Feature activation-1.27e-02

Pos logit contributions
robe+0.019
arl+0.016
 cooler+0.016
ugg+0.016
melon+0.015
Neg logit contributions
pieces-0.020
od-0.019
Maria-0.018
enta-0.017
 circles-0.017
TokenI
Token ID40
Feature activation+9.91e-03

Pos logit contributions
robe+0.022
arl+0.020
melon+0.019
 cooler+0.019
ugg+0.018
Neg logit contributions
pieces-0.022
od-0.020
Maria-0.019
lier-0.018
one-0.018
Token'm
Token ID1101
Feature activation+6.52e-02

Pos logit contributions
pieces+0.018
od+0.016
Maria+0.016
enta+0.014
one+0.014
Neg logit contributions
robe-0.017
arl-0.015
ugg-0.014
 cooler-0.014
melon-0.014
Token sorry
Token ID7926
Feature activation-7.68e-02

Pos logit contributions
pieces+0.151
od+0.144
Maria+0.140
 circles+0.131
enta+0.131
Neg logit contributions
robe-0.075
ugg-0.054
arl-0.052
melon-0.051
 ranc-0.047
Token,
Token ID11
Feature activation-8.55e-03

Pos logit contributions
robe+0.156
arl+0.141
melon+0.136
ugg+0.134
 cooler+0.133
Neg logit contributions
pieces-0.110
od-0.098
Maria-0.092
lier-0.085
one-0.085
Token too
Token ID1165
Feature activation-8.31e-02

Pos logit contributions
robe+0.014
arl+0.013
 cooler+0.012
melon+0.012
ugg+0.012
Neg logit contributions
pieces-0.015
od-0.014
Maria-0.013
lier-0.012
one-0.012
Token.
Token ID13
Feature activation+1.12e-03

Pos logit contributions
robe+0.181
arl+0.163
melon+0.158
ugg+0.158
 cooler+0.155
Neg logit contributions
pieces-0.104
od-0.095
Maria-0.088
lier-0.080
one-0.078
Token I
Token ID314
Feature activation+3.41e-02

Pos logit contributions
pieces+0.002
Maria+0.001
od+0.001
asing+0.001
 circles+0.001
Neg logit contributions
robe-0.002
 cooler-0.002
arl-0.002
melon-0.002
 ranc-0.002
Token was
Token ID373
Feature activation+2.06e-02

Pos logit contributions
pieces+0.058
od+0.050
Maria+0.049
one+0.045
enta+0.044
Neg logit contributions
robe-0.061
arl-0.056
 cooler-0.053
ugg-0.053
melon-0.052
Token en
Token ID551
Feature activation-5.46e-02

Pos logit contributions
pieces+0.050
od+0.046
Maria+0.045
one+0.042
lier+0.042
Neg logit contributions
robe-0.021
arl-0.018
 cooler-0.017
melon-0.017
ugg-0.015
Tokenvious
Token ID1442
Feature activation+6.31e-03

Pos logit contributions
arl+0.113
robe+0.109
 cooler+0.100
melon+0.099
ugg+0.096
Neg logit contributions
pieces-0.075
Maria-0.069
od-0.065
enta-0.060
Jo-0.057
Token?"
Token ID1701
Feature activation+5.07e-03

Pos logit contributions
pieces+0.114
od+0.102
Maria+0.100
lier+0.094
one+0.093
Neg logit contributions
robe-0.070
arl-0.062
melon-0.058
 cooler-0.057
ugg-0.055
Token
Token ID198
Feature activation+2.66e-02

Pos logit contributions
pieces+0.011
od+0.010
one+0.009
Maria+0.009
asing+0.009
Neg logit contributions
robe-0.007
arl-0.006
ugg-0.006
 Cream-0.005
 cooler-0.005
Token
Token ID198
Feature activation+2.65e-02

Pos logit contributions
pieces+0.059
od+0.052
asing+0.051
 rect+0.050
one+0.050
Neg logit contributions
robe-0.033
arl-0.030
ugg-0.028
 Cream-0.027
 cooler-0.026
TokenL
Token ID43
Feature activation+2.61e-02

Pos logit contributions
pieces+0.048
od+0.043
Maria+0.039
one+0.039
asing+0.038
Neg logit contributions
robe-0.043
arl-0.040
ugg-0.037
 cooler-0.036
 Cream-0.036
Tokenily
Token ID813
Feature activation+5.92e-02

Pos logit contributions
pieces+0.044
od+0.037
Maria+0.037
 rect+0.035
enta+0.035
Neg logit contributions
robe-0.048
arl-0.044
ugg-0.039
 cooler-0.039
melon-0.038
Token thought
Token ID1807
Feature activation-4.45e-02

Pos logit contributions
pieces+0.098
od+0.085
Maria+0.082
one+0.076
lier+0.075
Neg logit contributions
robe-0.107
arl-0.097
melon-0.092
 cooler-0.092
ugg-0.091
Token for
Token ID329
Feature activation+4.91e-02

Pos logit contributions
robe+0.067
arl+0.061
 cooler+0.057
melon+0.055
ugg+0.054
Neg logit contributions
pieces-0.084
Maria-0.076
od-0.076
lier-0.071
enta-0.070
Token a
Token ID257
Feature activation+1.30e-02

Pos logit contributions
pieces+0.074
od+0.067
Maria+0.067
 circles+0.061
one+0.060
Neg logit contributions
robe-0.092
arl-0.091
melon-0.086
ugg-0.082
 cooler-0.077
Token minute
Token ID5664
Feature activation+8.93e-02

Pos logit contributions
pieces+0.026
od+0.024
Maria+0.023
asing+0.022
enta+0.022
Neg logit contributions
robe-0.019
arl-0.017
 cooler-0.016
 ranc-0.015
melon-0.015
Token,
Token ID11
Feature activation-8.19e-03

Pos logit contributions
pieces+0.168
Maria+0.151
enta+0.143
od+0.142
lar+0.135
Neg logit contributions
robe-0.136
arl-0.117
 cooler-0.115
melon-0.113
ugg-0.108
Token and
Token ID290
Feature activation+9.87e-03

Pos logit contributions
robe+0.012
 cooler+0.011
arl+0.010
melon+0.010
 Sleep+0.010
Neg logit contributions
pieces-0.016
od-0.014
Maria-0.014
enta-0.013
ata-0.013
Token,
Token ID11
Feature activation-1.49e-02

Pos logit contributions
robe+0.037
arl+0.033
melon+0.032
 cooler+0.032
ugg+0.031
Neg logit contributions
pieces-0.035
Maria-0.031
od-0.031
enta-0.027
inda-0.027
Token Emily
Token ID17608
Feature activation+4.36e-02

Pos logit contributions
robe+0.021
arl+0.019
 cooler+0.018
melon+0.017
ugg+0.017
Neg logit contributions
pieces-0.030
od-0.028
Maria-0.027
lier-0.025
one-0.025
Token.
Token ID13
Feature activation+4.65e-03

Pos logit contributions
pieces+0.060
od+0.052
Maria+0.049
one+0.045
lier+0.045
Neg logit contributions
robe-0.090
arl-0.085
melon-0.079
 cooler-0.079
ugg-0.078
Token Emily
Token ID17608
Feature activation+4.41e-02

Pos logit contributions
pieces+0.010
od+0.009
Maria+0.008
one+0.008
lier+0.008
Neg logit contributions
robe-0.006
arl-0.006
 cooler-0.005
melon-0.005
 Sleep-0.005
Token always
Token ID1464
Feature activation-1.69e-02

Pos logit contributions
pieces+0.069
od+0.061
Maria+0.059
lier+0.053
one+0.053
Neg logit contributions
robe-0.083
arl-0.076
melon-0.072
 cooler-0.071
ugg-0.070
Token put
Token ID1234
Feature activation-3.93e-02

Pos logit contributions
robe+0.033
arl+0.032
melon+0.030
 cooler+0.029
ugg+0.028
Neg logit contributions
pieces-0.023
od-0.021
Maria-0.020
one-0.019
enta-0.018
Token a
Token ID257
Feature activation+4.92e-03

Pos logit contributions
robe+0.071
arl+0.067
melon+0.064
ugg+0.063
 cooler+0.060
Neg logit contributions
pieces-0.061
od-0.053
Maria-0.051
 circles-0.048
lier-0.048
Token leash
Token ID40529
Feature activation-3.37e-02

Pos logit contributions
pieces+0.010
od+0.008
 pad+0.008
Maria+0.008
 circles+0.008
Neg logit contributions
robe-0.007
arl-0.006
melon-0.006
 Sleep-0.005
ugg-0.005
Token on
Token ID319
Feature activation+4.70e-02

Pos logit contributions
arl+0.065
robe+0.064
melon+0.059
 cooler+0.059
 waitress+0.057
Neg logit contributions
pieces-0.043
od-0.037
one-0.035
 pad-0.035
ata-0.034
Token Max
Token ID5436
Feature activation-1.72e-02

Pos logit contributions
pieces+0.089
od+0.080
Maria+0.078
one+0.073
enta+0.072
Neg logit contributions
robe-0.074
arl-0.066
 cooler-0.061
melon-0.061
ugg-0.058
Token before
Token ID878
Feature activation-1.61e-03

Pos logit contributions
robe+0.045
arl+0.042
 cooler+0.041
melon+0.040
ugg+0.040
Neg logit contributions
pieces-0.014
od-0.011
Maria-0.010
one-0.008
lier-0.008
Token She
Token ID1375
Feature activation+3.05e-03

Pos logit contributions
pieces+0.050
od+0.045
Maria+0.043
lier+0.041
enta+0.041
Neg logit contributions
robe-0.034
arl-0.030
melon-0.028
 cooler-0.028
ugg-0.027
Token was
Token ID373
Feature activation+2.62e-02

Pos logit contributions
pieces+0.005
Maria+0.004
od+0.004
lier+0.004
ill+0.004
Neg logit contributions
robe-0.005
 cooler-0.005
ugg-0.005
melon-0.005
 cabin-0.004
Token surprised
Token ID6655
Feature activation-9.21e-03

Pos logit contributions
pieces+0.048
od+0.043
Maria+0.042
lier+0.039
one+0.039
Neg logit contributions
robe-0.042
arl-0.039
melon-0.036
 cooler-0.035
ugg-0.035
Token to
Token ID284
Feature activation+3.78e-02

Pos logit contributions
robe+0.015
arl+0.014
melon+0.013
 cooler+0.013
ugg+0.013
Neg logit contributions
pieces-0.017
od-0.014
Maria-0.014
enta-0.013
ata-0.013
Token see
Token ID766
Feature activation+9.89e-02

Pos logit contributions
pieces+0.075
lier+0.072
 rect+0.070
ata+0.069
od+0.069
Neg logit contributions
 not-0.044
robe-0.040
 name-0.039
 sorry-0.039
 mom-0.035
Token how
Token ID703
Feature activation-1.68e-01

Pos logit contributions
pieces+0.196
od+0.178
Maria+0.171
enta+0.166
ata+0.165
Neg logit contributions
robe-0.135
arl-0.124
melon-0.120
 cooler-0.119
 Cream-0.109
Token much
Token ID881
Feature activation+6.67e-02

Pos logit contributions
robe+0.248
arl+0.219
melon+0.200
 cooler+0.196
ugg+0.194
Neg logit contributions
pieces-0.340
od-0.308
Maria-0.303
one-0.282
lier-0.281
Token fun
Token ID1257
Feature activation+3.05e-02

Pos logit contributions
pieces+0.132
od+0.121
Maria+0.115
enta+0.109
ata+0.108
Neg logit contributions
robe-0.095
 cooler-0.088
arl-0.085
melon-0.083
 Sleep-0.076
Token she
Token ID673
Feature activation+5.01e-02

Pos logit contributions
pieces+0.062
od+0.056
Maria+0.055
ata+0.052
enta+0.052
Neg logit contributions
robe-0.041
 cooler-0.038
melon-0.035
arl-0.033
 cabin-0.033
Token had
Token ID550
Feature activation+4.50e-02

Pos logit contributions
pieces+0.094
Maria+0.082
od+0.081
ill+0.074
weet+0.074
Neg logit contributions
robe-0.081
 cooler-0.070
ugg-0.064
 Cream-0.063
 cabin-0.062
Token,
Token ID11
Feature activation+4.93e-03

Pos logit contributions
pieces+0.085
Maria+0.072
od+0.070
inda+0.065
 rect+0.064
Neg logit contributions
robe-0.071
 cooler-0.063
 Sleep-0.059
 not-0.059
 well-0.056
Token
Token ID220
Feature activation-3.29e-02

Pos logit contributions
pieces+0.036
od+0.033
Maria+0.031
enta+0.030
asing+0.029
Neg logit contributions
robe-0.027
arl-0.024
 cooler-0.022
melon-0.022
ugg-0.022
Token
Token ID198
Feature activation+2.53e-02

Pos logit contributions
robe+0.054
arl+0.049
melon+0.045
 cooler+0.045
ugg+0.044
Neg logit contributions
pieces-0.060
od-0.053
Maria-0.051
lier-0.048
enta-0.048
Token
Token ID198
Feature activation+2.59e-02

Pos logit contributions
pieces+0.049
od+0.044
 rect+0.040
asing+0.040
enta+0.040
Neg logit contributions
robe-0.037
arl-0.035
ugg-0.032
 cooler-0.031
 Cream-0.031
TokenSuddenly
Token ID38582
Feature activation+6.77e-02

Pos logit contributions
pieces+0.040
od+0.036
enta+0.031
Maria+0.031
ets+0.031
Neg logit contributions
robe-0.047
arl-0.045
ugg-0.042
 cooler-0.041
 Cream-0.041
Token,
Token ID11
Feature activation-1.59e-02

Pos logit contributions
pieces+0.126
od+0.110
Maria+0.106
enta+0.100
asing+0.099
Neg logit contributions
robe-0.109
 cooler-0.096
arl-0.095
melon-0.092
ugg-0.088
Token Snow
Token ID7967
Feature activation-1.55e-01

Pos logit contributions
robe+0.025
arl+0.023
 cooler+0.021
melon+0.021
ugg+0.021
Neg logit contributions
pieces-0.030
od-0.027
Maria-0.026
lier-0.024
one-0.024
Tokeny
Token ID88
Feature activation-2.90e-02

Pos logit contributions
robe+0.296
arl+0.264
melon+0.249
 cooler+0.247
ugg+0.244
Neg logit contributions
pieces-0.254
od-0.211
Maria-0.203
one-0.188
asing-0.187
Token said
Token ID531
Feature activation+1.70e-02

Pos logit contributions
robe+0.056
 cooler+0.050
melon+0.048
arl+0.046
 ranc+0.046
Neg logit contributions
pieces-0.045
od-0.037
Maria-0.036
asing-0.034
enta-0.034
Token "
Token ID366
Feature activation-1.92e-02

Pos logit contributions
pieces+0.026
od+0.024
Maria+0.022
lier+0.021
one+0.020
Neg logit contributions
robe-0.032
arl-0.030
 cooler-0.027
melon-0.027
ugg-0.027
TokenMy
Token ID3666
Feature activation+4.02e-02

Pos logit contributions
robe+0.029
arl+0.025
ugg+0.024
 Sleep+0.023
 Cream+0.023
Neg logit contributions
pieces-0.038
od-0.034
Maria-0.032
enta-0.031
 rect-0.031
Token turn
Token ID1210
Feature activation+1.77e-02

Pos logit contributions
pieces+0.085
od+0.077
Maria+0.074
one+0.071
lier+0.070
Neg logit contributions
robe-0.052
arl-0.048
ugg-0.043
melon-0.043
 cooler-0.041
Token flowers
Token ID12734
Feature activation+6.44e-02

Pos logit contributions
robe+0.051
arl+0.049
ugg+0.044
melon+0.043
 cooler+0.042
Neg logit contributions
pieces-0.051
od-0.044
Maria-0.042
lier-0.039
 circles-0.039
Token and
Token ID290
Feature activation-1.94e-02

Pos logit contributions
pieces+0.094
od+0.082
Maria+0.078
lier+0.071
one+0.070
Neg logit contributions
robe-0.127
arl-0.119
melon-0.110
 cooler-0.110
ugg-0.109
Token hugs
Token ID40657
Feature activation-7.77e-02

Pos logit contributions
robe+0.029
 cooler+0.027
melon+0.025
 ranc+0.023
arl+0.023
Neg logit contributions
pieces-0.037
od-0.035
Maria-0.033
one-0.033
lier-0.032
Token and
Token ID290
Feature activation-1.13e-02

Pos logit contributions
robe+0.138
arl+0.121
melon+0.116
 cooler+0.116
ugg+0.113
Neg logit contributions
pieces-0.134
od-0.118
Maria-0.116
enta-0.109
asing-0.107
Token said
Token ID531
Feature activation+6.50e-03

Pos logit contributions
robe+0.018
arl+0.016
 cooler+0.015
melon+0.015
ugg+0.015
Neg logit contributions
pieces-0.021
od-0.019
Maria-0.019
one-0.017
lier-0.017
Token how
Token ID703
Feature activation-1.80e-01

Pos logit contributions
pieces+0.011
od+0.010
Maria+0.010
lier+0.009
one+0.009
Neg logit contributions
robe-0.011
arl-0.010
 cooler-0.009
melon-0.009
ugg-0.009
Token wonderful
Token ID7932
Feature activation+7.54e-02

Pos logit contributions
robe+0.293
arl+0.266
melon+0.247
 cooler+0.245
ugg+0.238
Neg logit contributions
pieces-0.331
od-0.299
Maria-0.288
lier-0.266
one-0.264
Token she
Token ID673
Feature activation+3.38e-02

Pos logit contributions
pieces+0.149
od+0.133
Maria+0.130
enta+0.123
one+0.122
Neg logit contributions
robe-0.113
 cooler-0.100
melon-0.093
 Cream-0.090
arl-0.090
Token was
Token ID373
Feature activation+1.93e-02

Pos logit contributions
pieces+0.062
Maria+0.055
od+0.052
 circles+0.049
 rect+0.048
Neg logit contributions
robe-0.057
 cooler-0.049
ugg-0.046
 Sleep-0.045
melon-0.045
Token.
Token ID13
Feature activation-7.56e-03

Pos logit contributions
pieces+0.036
Maria+0.036
Jul+0.033
 rect+0.033
Rachel+0.033
Neg logit contributions
robe-0.025
melon-0.023
 cooler-0.022
 tired-0.021
 thirsty-0.020
Token Lily
Token ID20037
Feature activation+4.37e-02

Pos logit contributions
robe+0.011
arl+0.010
melon+0.009
ugg+0.009
 cooler+0.009
Neg logit contributions
pieces-0.015
od-0.013
Maria-0.013
one-0.012
lier-0.012
Token wave
Token ID6769
Feature activation+5.97e-02

Pos logit contributions
pieces+0.011
od+0.010
Maria+0.010
one+0.009
lier+0.009
Neg logit contributions
robe-0.008
 cooler-0.007
melon-0.007
arl-0.007
ugg-0.007
Token his
Token ID465
Feature activation-1.97e-02

Pos logit contributions
pieces+0.122
Maria+0.110
od+0.108
enta+0.105
lar+0.103
Neg logit contributions
robe-0.083
arl-0.071
 cooler-0.070
melon-0.067
ugg-0.062
Token arms
Token ID5101
Feature activation+4.45e-02

Pos logit contributions
robe+0.032
arl+0.029
 cooler+0.028
melon+0.027
ugg+0.025
Neg logit contributions
pieces-0.036
od-0.033
Maria-0.032
one-0.030
lier-0.029
Token to
Token ID284
Feature activation+3.12e-02

Pos logit contributions
pieces+0.072
od+0.063
Maria+0.061
lier+0.056
one+0.056
Neg logit contributions
robe-0.081
arl-0.075
melon-0.070
 cooler-0.070
ugg-0.069
Token show
Token ID905
Feature activation+9.09e-02

Pos logit contributions
pieces+0.062
Maria+0.060
od+0.059
asing+0.058
ata+0.057
Neg logit contributions
robe-0.034
 cooler-0.033
melon-0.032
arl-0.032
 bar-0.032
Token how
Token ID703
Feature activation-1.74e-01

Pos logit contributions
pieces+0.181
Maria+0.167
od+0.165
enta+0.159
ill+0.156
Neg logit contributions
robe-0.124
arl-0.108
 cooler-0.105
melon-0.101
 Sleep-0.092
Token proud
Token ID6613
Feature activation+4.88e-02

Pos logit contributions
robe+0.288
arl+0.250
melon+0.224
ugg+0.218
 cooler+0.216
Neg logit contributions
pieces-0.331
od-0.299
Maria-0.289
one-0.275
lier-0.273
Token he
Token ID339
Feature activation+1.62e-02

Pos logit contributions
pieces+0.104
Maria+0.093
od+0.090
enta+0.089
lar+0.088
Neg logit contributions
robe-0.060
 cooler-0.057
melon-0.054
arl-0.052
 Cream-0.046
Token was
Token ID373
Feature activation+1.97e-02

Pos logit contributions
pieces+0.032
Maria+0.028
od+0.025
ill+0.025
asing+0.025
Neg logit contributions
robe-0.024
 cooler-0.021
 Sleep-0.019
 Cream-0.018
 cabin-0.018
Token.
Token ID13
Feature activation+9.51e-03

Pos logit contributions
pieces+0.039
Maria+0.038
Jul+0.037
 rect+0.036
Ah+0.036
Neg logit contributions
robe-0.022
 cooler-0.021
melon-0.021
 tired-0.020
 thirsty-0.019
Token
Token ID198
Feature activation+2.75e-02

Pos logit contributions
pieces+0.016
od+0.014
Maria+0.014
lier+0.013
one+0.013
Neg logit contributions
robe-0.017
arl-0.015
melon-0.014
 cooler-0.014
ugg-0.014
Token and
Token ID290
Feature activation+4.10e-03

Pos logit contributions
pieces+0.087
od+0.076
Maria+0.075
lier+0.068
one+0.068
Neg logit contributions
robe-0.098
arl-0.088
melon-0.084
 cooler-0.084
ugg-0.081
Token were
Token ID547
Feature activation+3.07e-02

Pos logit contributions
pieces+0.008
od+0.007
Maria+0.007
ill+0.006
lier+0.006
Neg logit contributions
robe-0.007
 cooler-0.006
arl-0.005
melon-0.005
 ranc-0.005
Token very
Token ID845
Feature activation+1.56e-03

Pos logit contributions
pieces+0.059
od+0.053
Maria+0.051
lier+0.047
one+0.047
Neg logit contributions
robe-0.047
arl-0.043
melon-0.040
 cooler-0.039
ugg-0.039
Token careful
Token ID8161
Feature activation+2.21e-02

Pos logit contributions
pieces+0.003
od+0.003
Maria+0.003
enta+0.002
lier+0.002
Neg logit contributions
robe-0.002
arl-0.002
melon-0.002
ugg-0.002
 Sleep-0.002
Token with
Token ID351
Feature activation-3.01e-02

Pos logit contributions
pieces+0.037
od+0.033
Maria+0.032
lier+0.029
one+0.029
Neg logit contributions
robe-0.039
arl-0.036
melon-0.034
 cooler-0.034
ugg-0.033
Token how
Token ID703
Feature activation-1.72e-01

Pos logit contributions
robe+0.053
arl+0.048
melon+0.045
 cooler+0.045
ugg+0.044
Neg logit contributions
pieces-0.051
od-0.046
Maria-0.044
lier-0.040
one-0.040
Token they
Token ID484
Feature activation+2.34e-02

Pos logit contributions
robe+0.314
arl+0.287
melon+0.269
 cooler+0.259
ugg+0.258
Neg logit contributions
pieces-0.285
od-0.249
Maria-0.243
one-0.228
lier-0.224
Token wrapped
Token ID12908
Feature activation-8.41e-02

Pos logit contributions
pieces+0.035
od+0.031
Maria+0.029
lier+0.027
one+0.027
Neg logit contributions
robe-0.045
arl-0.043
melon-0.040
 cooler-0.039
ugg-0.039
Token it
Token ID340
Feature activation+1.87e-03

Pos logit contributions
robe+0.146
arl+0.128
melon+0.124
 cooler+0.123
ugg+0.117
Neg logit contributions
pieces-0.146
od-0.131
Maria-0.126
one-0.117
enta-0.117
Token.
Token ID13
Feature activation+1.41e-02

Pos logit contributions
pieces+0.003
od+0.003
Maria+0.003
enta+0.003
one+0.003
Neg logit contributions
robe-0.003
arl-0.003
 cooler-0.003
melon-0.003
ugg-0.002
Token
Token ID220
Feature activation-2.53e-02

Pos logit contributions
pieces+0.030
od+0.027
Maria+0.026
asing+0.026
lar+0.026
Neg logit contributions
robe-0.018
arl-0.016
 Cream-0.015
melon-0.015
ugg-0.015