TOP ACTIVATIONS
MAX = 0.253

the bird and said , "' Hello , my name is
Sam . He said , "' We should share our special
my smiled and said , "' Let 's go and do
and said excited ly , "' Wow , the buildings here
?" Joe said :" Don 't worry . I
sad . He said , "" I hope something will change
lesson never be rude . Fortunately
scared she felt happy . She
important the doll had a price
peace she just needs to remember
hugged his cat and said :" Thank you for the delicious
at her Mum and asked :" Is it ready , Mum
special things â little stones , happy flowers
strange the k itty was wearing
him . She says , " I am sorry , Ben
about him , he said :" My name is L enny
day to always be careful .
friends it 's easy then !"
her . He says : " I am sorry , Lily
rocked higher and higher . His

MIN ACTIVATIONS
MIN = -0.188

's mom my . " Looks like the wave destroyed all
the owner smiled . " Looks like you two are friends
print so badly . Fr ust rated tears were well ing
laughed and said , " Looks like that apple made you
was too strong . Fr ust rated , he started to
back . Fr ust rated , Jane went back
but couldn 't . Fr ust rated , he started to
or too hard . Fr ust rated , she kept searching
he said no . Fr ust rated , Bob decided to
smiled and said , " Looks like you like the wardrobe
a little crab named Cr ust y . Cr ust y
was too dangerous . Fr ust rated , Sam turned away
around the door . Fr ust rated , Jack put the
answer for him . Fr ust rated , he decided to
ge . Fr ust rated , the little ant
. One day , Cr ust y found a shiny shell
Cr ust y . Cr ust y lived in a big
stick . Fr ust rated , the lion went
table . " Looks like mom has some mail
something just because someone d ares you . Always remember to

INTERVAL 0.143 - 0.253
CONTAINS 1.068%

smiled , and replied , " I will ! I 'll
His mom my said , " It 's time for a
Dad said , " Yes , we did !"
?" Her mom said , " Yes , you can organize
a hug . " Are you okay ?" her

INTERVAL 0.032 - 0.143
CONTAINS 26.747%

day , Tim my 's friend came over to play .
, they were able to build the house in no time
Her mom came to help her and said , "
animals . Would you like to play too ?"
ily didn 't like bugs , so she asked her mom

INTERVAL -0.078 - 0.032
CONTAINS 67.561%

made her flower . The teacher says , "
the ball rolled away . " Tom , you
them hot chocolate and cookies . She was kind and loving
when Sarah was using her brush , she heard a loud
away from him . But Tom was stronger and

INTERVAL -0.188 - -0.078
CONTAINS 4.624%

One day , they saw a new boy in
wet . When she got out of the pool , she
, long sleep . <|endoftext|> Once there was a boy named
! <|endoftext|> Once upon a time there lived a baby .
When they got on , they saw a big box in
Token the
Token ID262
Feature activation+5.92e-02

Pos logit contributions
 mill+0.079
ting+0.078
ini+0.078
 painter+0.076
 paramed+0.075
Neg logit contributions
 much-0.060
 drift-0.060
 nap-0.059
 dives-0.057
 swim-0.056
Token bird
Token ID6512
Feature activation+4.22e-02

Pos logit contributions
cles+0.109
 laund+0.108
ting+0.107
rom+0.107
years+0.106
Neg logit contributions
 nap-0.080
 much-0.075
 long-0.071
 swim-0.069
 drift-0.068
Token and
Token ID290
Feature activation-2.65e-02

Pos logit contributions
ting+0.074
 mill+0.074
 painter+0.071
ini+0.070
 office+0.068
Neg logit contributions
 much-0.054
 dives-0.053
 drift-0.053
 nap-0.053
irm-0.052
Token said
Token ID531
Feature activation-4.22e-02

Pos logit contributions
 much+0.033
 drift+0.033
 dives+0.033
 nap+0.032
irm+0.031
Neg logit contributions
 mill-0.048
ting-0.047
ini-0.046
 painter-0.046
 office-0.045
Token,
Token ID11
Feature activation+5.13e-02

Pos logit contributions
 much+0.046
 drift+0.045
 dives+0.045
 nap+0.045
irm+0.043
Neg logit contributions
ting-0.083
 mill-0.083
 painter-0.080
ini-0.079
 paramed-0.077
Token "'
Token ID24018
Feature activation+2.53e-01

Pos logit contributions
ini+0.077
 mill+0.076
ting+0.074
 painter+0.073
 office+0.070
Neg logit contributions
 much-0.083
 drift-0.080
 dives-0.078
 nap-0.076
 swim-0.075
TokenHello
Token ID15496
Feature activation+7.11e-03

Pos logit contributions
 office+0.360
 mill+0.330
 desks+0.329
 porch+0.329
 floors+0.328
Neg logit contributions
 drift-0.434
 much-0.431
 nap-0.414
 swim-0.404
 dives-0.401
Token,
Token ID11
Feature activation+4.98e-02

Pos logit contributions
ting+0.013
 mill+0.013
ini+0.013
 painter+0.013
 laund+0.012
Neg logit contributions
 much-0.009
 nap-0.008
 drift-0.008
 dives-0.008
 dive-0.008
Token my
Token ID616
Feature activation-4.39e-03

Pos logit contributions
ting+0.086
keeping+0.086
cles+0.086
organized+0.085
 mill+0.085
Neg logit contributions
 much-0.067
 nap-0.064
 drift-0.061
 swim-0.059
 dive-0.059
Token name
Token ID1438
Feature activation-1.58e-02

Pos logit contributions
 much+0.005
 nap+0.005
 drift+0.005
 dives+0.005
irm+0.005
Neg logit contributions
ting-0.009
 mill-0.008
 painter-0.008
ini-0.008
 laund-0.008
Token is
Token ID318
Feature activation+3.10e-02

Pos logit contributions
 much+0.020
 drift+0.020
 dives+0.020
irm+0.020
 nap+0.019
Neg logit contributions
ting-0.028
 mill-0.028
 painter-0.027
ini-0.026
 office-0.026
Token Sam
Token ID3409
Feature activation+1.28e-02

Pos logit contributions
 mill+0.075
ting+0.075
 painter+0.072
ini+0.072
 paramed+0.070
Neg logit contributions
 much-0.054
 nap-0.053
 drift-0.053
 dives-0.052
irm-0.051
Token.
Token ID13
Feature activation-6.96e-03

Pos logit contributions
ting+0.024
 mill+0.024
 painter+0.023
ini+0.023
 office+0.022
Neg logit contributions
 much-0.015
 drift-0.015
 dives-0.015
 nap-0.014
irm-0.014
Token He
Token ID679
Feature activation-9.04e-03

Pos logit contributions
 much+0.007
 dives+0.007
 drift+0.007
irm+0.007
 nap+0.007
Neg logit contributions
 mill-0.014
ting-0.014
ini-0.014
 painter-0.014
cut-0.013
Token said
Token ID531
Feature activation-4.62e-02

Pos logit contributions
 much+0.012
 drift+0.012
 dives+0.012
 nap+0.012
irm+0.012
Neg logit contributions
 mill-0.015
ting-0.015
 painter-0.014
ini-0.014
 paramed-0.014
Token,
Token ID11
Feature activation+4.99e-02

Pos logit contributions
 much+0.049
 dives+0.043
 nap+0.042
 drift+0.041
 swim+0.039
Neg logit contributions
ini-0.098
 mill-0.098
ting-0.096
 painter-0.093
rom-0.092
Token "'
Token ID24018
Feature activation+2.49e-01

Pos logit contributions
ini+0.078
 mill+0.076
ting+0.073
 painter+0.072
 floors+0.070
Neg logit contributions
 much-0.080
 drift-0.076
 dives-0.073
 nap-0.072
 swim-0.071
TokenWe
Token ID1135
Feature activation-2.67e-02

Pos logit contributions
 mill+0.343
 office+0.332
 porch+0.327
 floors+0.313
 paramed+0.302
Neg logit contributions
 much-0.437
irm-0.410
 nap-0.401
ust-0.400
 drift-0.399
Token should
Token ID815
Feature activation+9.13e-02

Pos logit contributions
 drift+0.033
 dives+0.033
 much+0.033
irm+0.032
 nap+0.031
Neg logit contributions
 mill-0.048
ting-0.047
 painter-0.046
ini-0.045
cut-0.044
Token share
Token ID2648
Feature activation+2.49e-02

Pos logit contributions
 mill+0.178
ting+0.177
 painter+0.172
ini+0.169
 office+0.165
Neg logit contributions
 much-0.100
 dives-0.099
 drift-0.098
 nap-0.096
irm-0.095
Token our
Token ID674
Feature activation-2.63e-02

Pos logit contributions
 mill+0.051
ini+0.050
ting+0.049
 paramed+0.049
 painter+0.049
Neg logit contributions
 much-0.028
 nap-0.027
 dives-0.026
 drift-0.025
irm-0.024
Token special
Token ID2041
Feature activation-1.16e-02

Pos logit contributions
 dives+0.048
 nap+0.047
 drift+0.047
 much+0.047
irm+0.046
Neg logit contributions
ting-0.034
 mill-0.033
 paramed-0.031
 painter-0.031
cut-0.030
Tokenmy
Token ID1820
Feature activation+3.12e-02

Pos logit contributions
 mill+0.138
ting+0.136
 painter+0.124
ini+0.123
 paramed+0.122
Neg logit contributions
 much-0.181
 drift-0.180
 nap-0.179
 dives-0.177
irm-0.175
Token smiled
Token ID13541
Feature activation+9.43e-03

Pos logit contributions
ting+0.056
 mill+0.056
ini+0.054
 office+0.053
 painter+0.053
Neg logit contributions
 dives-0.038
 drift-0.038
 much-0.037
 nap-0.037
 swim-0.036
Token and
Token ID290
Feature activation-3.01e-02

Pos logit contributions
ting+0.016
 mill+0.015
 painter+0.015
ini+0.014
 paramed+0.014
Neg logit contributions
 dives-0.013
 nap-0.013
 drift-0.013
 much-0.012
irm-0.012
Token said
Token ID531
Feature activation-4.51e-02

Pos logit contributions
 much+0.036
 nap+0.035
 dives+0.035
 drift+0.034
irm+0.034
Neg logit contributions
 mill-0.057
ting-0.057
 painter-0.054
 paramed-0.053
ini-0.052
Token,
Token ID11
Feature activation+4.96e-02

Pos logit contributions
 much+0.046
 dives+0.044
 drift+0.044
 nap+0.044
irm+0.042
Neg logit contributions
 mill-0.093
ting-0.093
ini-0.090
 painter-0.090
 office-0.086
Token "'
Token ID24018
Feature activation+2.48e-01

Pos logit contributions
ini+0.076
 mill+0.075
ting+0.073
 painter+0.072
 office+0.070
Neg logit contributions
 much-0.078
 drift-0.076
 dives-0.073
 nap-0.071
 swim-0.070
TokenLet
Token ID5756
Feature activation+7.77e-02

Pos logit contributions
 office+0.372
 mill+0.352
 floors+0.332
cut+0.331
 homeowner+0.326
Neg logit contributions
 much-0.434
 drift-0.385
ust-0.371
 dives-0.369
 nap-0.368
Token's
Token ID338
Feature activation-2.68e-02

Pos logit contributions
ting+0.160
 mill+0.159
 painter+0.154
ini+0.151
 paramed+0.148
Neg logit contributions
 much-0.078
 dives-0.077
 drift-0.077
 nap-0.076
irm-0.075
Token go
Token ID467
Feature activation-6.44e-02

Pos logit contributions
 nap+0.030
 dive+0.030
 drift+0.029
 swim+0.029
 much+0.029
Neg logit contributions
ting-0.054
smith-0.053
cut-0.052
years-0.052
ini-0.052
Token and
Token ID290
Feature activation-3.04e-02

Pos logit contributions
 much+0.084
 nap+0.083
 dives+0.083
 drift+0.083
irm+0.079
Neg logit contributions
ting-0.112
 mill-0.112
 painter-0.109
ini-0.107
cut-0.106
Token do
Token ID466
Feature activation+6.60e-02

Pos logit contributions
 nap+0.036
 dives+0.035
 dive+0.035
 drift+0.035
 swim+0.035
Neg logit contributions
ting-0.059
ini-0.058
cut-0.058
smith-0.058
 mill-0.057
Token and
Token ID290
Feature activation-2.94e-02

Pos logit contributions
 mill+0.075
ting+0.075
ini+0.073
 painter+0.073
 paramed+0.070
Neg logit contributions
 much-0.049
 drift-0.047
 nap-0.047
 dives-0.047
irm-0.045
Token said
Token ID531
Feature activation-4.56e-02

Pos logit contributions
 much+0.047
 dives+0.046
 drift+0.046
 nap+0.045
irm+0.044
Neg logit contributions
ting-0.043
 mill-0.043
ini-0.042
 painter-0.041
 paramed-0.040
Token excited
Token ID6568
Feature activation-4.55e-03

Pos logit contributions
 much+0.048
 dives+0.046
 nap+0.044
 drift+0.044
irm+0.042
Neg logit contributions
ting-0.094
 mill-0.093
ini-0.091
 painter-0.090
cut-0.086
Tokenly
Token ID306
Feature activation-1.64e-02

Pos logit contributions
 much+0.008
 dives+0.008
 drift+0.008
 nap+0.008
irm+0.008
Neg logit contributions
ting-0.006
 mill-0.006
 painter-0.006
ini-0.006
 office-0.006
Token,
Token ID11
Feature activation+4.84e-02

Pos logit contributions
 much+0.018
 drift+0.017
 dives+0.017
 nap+0.017
irm+0.017
Neg logit contributions
ting-0.032
 mill-0.032
 painter-0.031
ini-0.031
 paramed-0.030
Token "'
Token ID24018
Feature activation+2.47e-01

Pos logit contributions
ting+0.057
 mill+0.056
 painter+0.053
ini+0.050
 paramed+0.049
Neg logit contributions
 much-0.091
 nap-0.091
 dives-0.090
 drift-0.090
irm-0.089
TokenWow
Token ID22017
Feature activation-8.43e-02

Pos logit contributions
 mill+0.287
 office+0.273
cut+0.264
 painter+0.255
ting+0.253
Neg logit contributions
 much-0.483
 drift-0.462
 nap-0.455
ust-0.445
 dives-0.445
Token,
Token ID11
Feature activation+4.94e-02

Pos logit contributions
irm+0.136
 drift+0.129
 dives+0.127
 hops+0.125
 much+0.125
Neg logit contributions
 mill-0.122
ting-0.122
 painter-0.114
ter-0.108
 office-0.108
Token the
Token ID262
Feature activation+5.97e-02

Pos logit contributions
ting+0.089
 mill+0.087
ini+0.086
 painter+0.084
cut+0.082
Neg logit contributions
 much-0.065
 nap-0.063
 dives-0.062
 drift-0.061
 dive-0.059
Token buildings
Token ID6832
Feature activation+4.39e-02

Pos logit contributions
ting+0.087
ini+0.080
cut+0.078
years+0.075
 mill+0.074
Neg logit contributions
 much-0.108
 nap-0.101
 drift-0.100
 dives-0.097
 time-0.093
Token here
Token ID994
Feature activation-2.24e-02

Pos logit contributions
ting+0.084
 mill+0.084
 painter+0.081
ini+0.080
 office+0.077
Neg logit contributions
 much-0.050
 drift-0.049
 dives-0.049
 nap-0.048
irm-0.047
Token?"
Token ID1701
Feature activation+7.24e-02

Pos logit contributions
 mill+0.022
ting+0.022
ini+0.021
 painter+0.021
rom+0.020
Neg logit contributions
 much-0.014
 nap-0.014
 drift-0.014
 dives-0.014
irm-0.013
Token
Token ID198
Feature activation-4.73e-02

Pos logit contributions
 mill+0.143
ting+0.142
 painter+0.138
ini+0.135
 paramed+0.132
Neg logit contributions
 much-0.078
 drift-0.077
 dives-0.077
 nap-0.076
irm-0.074
Token
Token ID198
Feature activation-4.63e-02

Pos logit contributions
 much+0.043
 nap+0.040
 drift+0.040
 dives+0.040
irm+0.038
Neg logit contributions
ting-0.104
 mill-0.103
 painter-0.099
ini-0.099
 paramed-0.096
TokenJoe
Token ID19585
Feature activation+2.98e-02

Pos logit contributions
 much+0.074
 nap+0.071
 drift+0.071
 dives+0.071
irm+0.069
Neg logit contributions
ting-0.071
 mill-0.069
ini-0.065
 painter-0.064
 office-0.062
Token said
Token ID531
Feature activation-4.72e-02

Pos logit contributions
ting+0.053
 mill+0.053
 painter+0.051
ini+0.050
 paramed+0.049
Neg logit contributions
 much-0.038
 drift-0.038
 dives-0.038
 nap-0.037
irm-0.036
Token:"
Token ID11097
Feature activation+2.43e-01

Pos logit contributions
 drift+0.058
 nap+0.058
irm+0.058
 dives+0.058
 much+0.056
Neg logit contributions
ting-0.084
 mill-0.083
 painter-0.082
 paramed-0.079
cut-0.078
TokenDon
Token ID3987
Feature activation+5.06e-02

Pos logit contributions
 mill+0.219
ting+0.218
 painter+0.204
ini+0.195
 paramed+0.185
Neg logit contributions
 much-0.522
 dives-0.520
 drift-0.519
 nap-0.515
irm-0.511
Token't
Token ID470
Feature activation-5.81e-02

Pos logit contributions
ting+0.072
 mill+0.071
ini+0.069
 painter+0.068
rom+0.067
Neg logit contributions
irm-0.087
 dives-0.084
 nap-0.084
 drift-0.081
 hops-0.080
Token worry
Token ID5490
Feature activation+2.98e-02

Pos logit contributions
 dives+0.090
 much+0.087
ls+0.087
irm+0.086
 drift+0.085
Neg logit contributions
 mill-0.087
ting-0.085
 painter-0.084
 paramed-0.079
ini-0.079
Token.
Token ID13
Feature activation+4.20e-03

Pos logit contributions
 mill+0.047
 painter+0.046
ting+0.046
ini+0.045
 office+0.044
Neg logit contributions
 dives-0.041
irm-0.041
 drift-0.041
 nap-0.039
 hops-0.039
Token I
Token ID314
Feature activation+4.28e-02

Pos logit contributions
ting+0.007
 mill+0.007
 painter+0.007
ini+0.007
 paramed+0.007
Neg logit contributions
 much-0.005
 dives-0.005
 drift-0.005
 nap-0.005
irm-0.005
Token sad
Token ID6507
Feature activation+5.25e-02

Pos logit contributions
 dives+0.057
 drift+0.056
 nap+0.055
irm+0.053
 hops+0.053
Neg logit contributions
 mill-0.079
 painter-0.078
ting-0.076
ini-0.076
 paramed-0.075
Token.
Token ID13
Feature activation-4.99e-03

Pos logit contributions
ting+0.089
 mill+0.086
 painter+0.082
 office+0.081
Why+0.079
Neg logit contributions
 drift-0.073
 dives-0.073
irm-0.070
 much-0.070
 nap-0.069
Token He
Token ID679
Feature activation-8.46e-03

Pos logit contributions
 much+0.004
 drift+0.003
 dives+0.003
 nap+0.003
irm+0.003
Neg logit contributions
ting-0.012
 mill-0.012
ini-0.011
 painter-0.011
 paramed-0.011
Token said
Token ID531
Feature activation-4.70e-02

Pos logit contributions
 much+0.013
 dives+0.013
 drift+0.013
 nap+0.012
irm+0.012
Neg logit contributions
 mill-0.013
ting-0.013
 painter-0.012
ini-0.012
 paramed-0.012
Token,
Token ID11
Feature activation+5.09e-02

Pos logit contributions
 much+0.048
 dives+0.046
 nap+0.046
 drift+0.046
irm+0.044
Neg logit contributions
 mill-0.097
ting-0.097
ini-0.094
 painter-0.093
rom-0.090
Token ""
Token ID13538
Feature activation+2.41e-01

Pos logit contributions
ini+0.084
 mill+0.081
ting+0.078
 painter+0.076
keeping+0.076
Neg logit contributions
 much-0.079
 drift-0.074
 nap-0.071
 dives-0.070
 swim-0.069
TokenI
Token ID40
Feature activation+1.98e-02

Pos logit contributions
 mill+0.266
ting+0.254
 office+0.231
 painter+0.230
ini+0.229
Neg logit contributions
 much-0.483
 drift-0.469
 nap-0.463
irm-0.461
 dives-0.460
Token hope
Token ID2911
Feature activation+2.87e-02

Pos logit contributions
 mill+0.033
 painter+0.033
ini+0.032
 paramed+0.031
ting+0.031
Neg logit contributions
 drift-0.026
 dives-0.026
irm-0.026
 much-0.025
ls-0.025
Token something
Token ID1223
Feature activation-6.12e-02

Pos logit contributions
ting+0.056
 mill+0.056
 painter+0.053
ini+0.053
 paramed+0.052
Neg logit contributions
 much-0.033
 drift-0.032
 nap-0.032
 dives-0.031
irm-0.030
Token will
Token ID481
Feature activation+8.18e-02

Pos logit contributions
irm+0.075
 drift+0.070
 dives+0.069
 fins+0.068
inn+0.067
Neg logit contributions
 mill-0.107
ting-0.106
 painter-0.105
Why-0.102
 office-0.100
Token change
Token ID1487
Feature activation-4.11e-02

Pos logit contributions
 mill+0.155
ting+0.153
 painter+0.149
ini+0.147
 office+0.143
Neg logit contributions
 dives-0.094
 much-0.093
irm-0.091
 drift-0.091
 nap-0.090
Token lesson
Token ID11483
Feature activation-9.59e-03

Pos logit contributions
 much+0.048
 nap+0.047
 dives+0.047
 drift+0.047
irm+0.045
Neg logit contributions
ting-0.074
 mill-0.073
 painter-0.070
ini-0.069
 paramed-0.067
Token
Token ID6184
Feature activation-2.24e-02

Pos logit contributions
 drift+0.013
 dives+0.013
 nap+0.013
 much+0.012
irm+0.012
Neg logit contributions
ting-0.017
 mill-0.017
 painter-0.016
ini-0.016
 office-0.015
Token
Token ID95
Feature activation+5.11e-02

Pos logit contributions
 much+0.031
 nap+0.030
 dives+0.030
 drift+0.030
irm+0.029
Neg logit contributions
 mill-0.038
ting-0.038
 painter-0.036
ini-0.036
 paramed-0.035
Token
Token ID26391
Feature activation-1.22e-02

Pos logit contributions
ting+0.076
 mill+0.075
 painter+0.073
ini+0.068
cut+0.067
Neg logit contributions
 much-0.081
 dives-0.077
 drift-0.077
irm-0.077
 nap-0.076
Token
Token ID447
Feature activation-1.62e-02

Pos logit contributions
 drift+0.016
 much+0.016
 dives+0.016
 nap+0.016
irm+0.016
Neg logit contributions
 mill-0.021
ting-0.020
 painter-0.020
ini-0.020
 office-0.019
Token
Token ID250
Feature activation+2.40e-01

Pos logit contributions
 much+0.020
 drift+0.018
 nap+0.018
 dives+0.018
irm+0.017
Neg logit contributions
ting-0.030
 painter-0.030
 mill-0.029
ini-0.029
 floors-0.028
Token never
Token ID1239
Feature activation-2.84e-02

Pos logit contributions
 mill+0.310
ting+0.302
 painter+0.288
 office+0.277
ini+0.277
Neg logit contributions
 much-0.432
 drift-0.410
 nap-0.408
irm-0.407
 dives-0.406
Token be
Token ID307
Feature activation+8.38e-03

Pos logit contributions
 drift+0.032
 much+0.032
irm+0.031
 dives+0.031
ls+0.031
Neg logit contributions
ting-0.053
 mill-0.053
 painter-0.051
 office-0.049
 paramed-0.048
Token rude
Token ID22066
Feature activation+3.85e-02

Pos logit contributions
ting+0.016
 mill+0.015
 painter+0.015
ini+0.015
 office+0.014
Neg logit contributions
 drift-0.010
 nap-0.010
 dives-0.010
irm-0.010
 much-0.010
Token.
Token ID13
Feature activation-1.82e-03

Pos logit contributions
 mill+0.079
ini+0.075
ting+0.075
 painter+0.074
 laund+0.073
Neg logit contributions
 much-0.042
 nap-0.039
 dives-0.038
 drift-0.037
irm-0.037
Token Fortunately
Token ID20525
Feature activation+2.11e-03

Pos logit contributions
 drift+0.002
 dives+0.002
 nap+0.002
 much+0.002
irm+0.002
Neg logit contributions
ting-0.003
 mill-0.003
 painter-0.003
ini-0.003
 paramed-0.003
Token scared
Token ID12008
Feature activation+2.40e-03

Pos logit contributions
 dives+0.072
irm+0.070
 drift+0.069
 nap+0.069
ls+0.066
Neg logit contributions
ting-0.105
 painter-0.100
 mill-0.100
 paramed-0.094
 office-0.094
Token
Token ID6184
Feature activation-2.17e-02

Pos logit contributions
ting+0.005
 painter+0.004
 mill+0.004
ini+0.004
Why+0.004
Neg logit contributions
 dives-0.003
 drift-0.003
irm-0.003
 nap-0.003
 hops-0.003
Token
Token ID95
Feature activation+5.14e-02

Pos logit contributions
 much+0.034
 dives+0.033
 drift+0.033
 nap+0.033
irm+0.033
Neg logit contributions
 mill-0.033
ting-0.033
 painter-0.031
ini-0.030
 office-0.030
Token
Token ID26391
Feature activation-1.19e-02

Pos logit contributions
 mill+0.075
ting+0.074
 painter+0.071
ini+0.070
cut+0.067
Neg logit contributions
 much-0.082
 drift-0.081
 dives-0.080
 nap-0.080
irm-0.080
Token
Token ID447
Feature activation-1.88e-02

Pos logit contributions
 drift+0.018
 dives+0.018
 much+0.017
 nap+0.017
irm+0.017
Neg logit contributions
 mill-0.019
ting-0.019
 painter-0.018
ini-0.018
 office-0.017
Token
Token ID250
Feature activation+2.40e-01

Pos logit contributions
 much+0.023
 drift+0.022
 nap+0.022
 dives+0.021
irm+0.021
Neg logit contributions
ting-0.034
 painter-0.034
 mill-0.034
ini-0.033
 office-0.032
Token she
Token ID673
Feature activation-4.91e-02

Pos logit contributions
 mill+0.298
ting+0.278
 painter+0.273
 office+0.266
ini+0.259
Neg logit contributions
 much-0.452
 drift-0.434
ust-0.429
 nap-0.429
irm-0.426
Token felt
Token ID2936
Feature activation+4.13e-02

Pos logit contributions
 drift+0.060
 dives+0.060
 nap+0.060
 much+0.060
irm+0.060
Neg logit contributions
 mill-0.087
ting-0.086
 painter-0.083
ini-0.082
 office-0.081
Token happy
Token ID3772
Feature activation+1.89e-02

Pos logit contributions
ting+0.075
 mill+0.073
 painter+0.070
ini+0.069
cut+0.067
Neg logit contributions
 dives-0.051
 drift-0.051
 nap-0.050
irm-0.049
 much-0.049
Token.
Token ID13
Feature activation-3.66e-03

Pos logit contributions
ting+0.034
 mill+0.034
 painter+0.033
ini+0.032
 paramed+0.031
Neg logit contributions
 much-0.024
 drift-0.023
 dives-0.023
 nap-0.023
irm-0.023
Token She
Token ID1375
Feature activation+3.27e-04

Pos logit contributions
 much+0.005
 drift+0.004
 dives+0.004
 nap+0.004
irm+0.004
Neg logit contributions
 mill-0.007
ting-0.007
 painter-0.006
ini-0.006
 paramed-0.006
Token important
Token ID1593
Feature activation+4.07e-02

Pos logit contributions
 dives+0.070
 drift+0.070
 much+0.069
 nap+0.069
irm+0.069
Neg logit contributions
 mill-0.108
ting-0.108
 painter-0.104
ini-0.101
 office-0.099
Token
Token ID6184
Feature activation-2.29e-02

Pos logit contributions
ini+0.083
 mill+0.083
 paramed+0.083
 painter+0.081
ting+0.081
Neg logit contributions
 much-0.043
 nap-0.039
 drift-0.037
 dives-0.037
 hops-0.035
Token
Token ID95
Feature activation+4.96e-02

Pos logit contributions
 nap+0.033
 much+0.033
 dives+0.033
 drift+0.033
 dive+0.032
Neg logit contributions
ting-0.036
 mill-0.035
 painter-0.034
ini-0.034
cut-0.034
Token
Token ID26391
Feature activation-1.16e-02

Pos logit contributions
 mill+0.078
ting+0.078
cut+0.075
 painter+0.073
ini+0.071
Neg logit contributions
 much-0.070
 nap-0.070
 drift-0.069
irm-0.069
 dives-0.069
Token
Token ID447
Feature activation-1.74e-02

Pos logit contributions
 much+0.016
 dives+0.016
 nap+0.016
 drift+0.016
irm+0.015
Neg logit contributions
ting-0.020
 mill-0.020
 painter-0.019
ini-0.018
cut-0.018
Token
Token ID250
Feature activation+2.40e-01

Pos logit contributions
 much+0.022
 nap+0.021
 drift+0.021
 dives+0.021
ls+0.020
Neg logit contributions
ting-0.031
 mill-0.030
 painter-0.030
cut-0.030
ini-0.030
Token the
Token ID262
Feature activation+6.21e-02

Pos logit contributions
 mill+0.306
ting+0.288
 painter+0.278
 office+0.277
cut+0.270
Neg logit contributions
 much-0.448
ust-0.428
 nap-0.417
 drift-0.415
irm-0.409
Token doll
Token ID3654
Feature activation+1.87e-02

Pos logit contributions
ting+0.105
ini+0.098
 mill+0.097
rom+0.095
years+0.093
Neg logit contributions
 nap-0.099
 much-0.093
 time-0.086
 drift-0.085
 dives-0.083
Token had
Token ID550
Feature activation+4.26e-02

Pos logit contributions
ting+0.037
 mill+0.036
 painter+0.035
ini+0.034
 paramed+0.034
Neg logit contributions
 much-0.021
 drift-0.020
 nap-0.020
 dives-0.020
irm-0.019
Token a
Token ID257
Feature activation-6.89e-02

Pos logit contributions
 mill+0.077
ting+0.077
 painter+0.074
ini+0.072
 office+0.071
Neg logit contributions
 drift-0.052
 dives-0.052
 much-0.051
irm-0.051
 nap-0.050
Token price
Token ID2756
Feature activation+8.79e-03

Pos logit contributions
 much+0.080
 nap+0.078
 drift+0.077
 dives+0.076
irm+0.072
Neg logit contributions
ting-0.135
 mill-0.131
 painter-0.126
ini-0.126
cut-0.122
Token peace
Token ID4167
Feature activation-5.81e-02

Pos logit contributions
 nap+0.044
 much+0.044
 drift+0.043
 dives+0.043
irm+0.041
Neg logit contributions
ting-0.052
 mill-0.050
ini-0.049
 painter-0.048
 paramed-0.047
Token
Token ID6184
Feature activation-2.58e-02

Pos logit contributions
 much+0.070
 dives+0.069
 drift+0.069
 nap+0.067
irm+0.067
Neg logit contributions
ting-0.107
 mill-0.106
 painter-0.102
ini-0.100
 office-0.098
Token
Token ID95
Feature activation+5.01e-02

Pos logit contributions
 much+0.041
 drift+0.040
 nap+0.040
 dives+0.040
irm+0.039
Neg logit contributions
 mill-0.038
ting-0.038
 painter-0.037
ini-0.036
cut-0.034
Token
Token ID26391
Feature activation-9.53e-03

Pos logit contributions
 mill+0.071
 painter+0.068
ting+0.068
cut+0.066
ini+0.065
Neg logit contributions
 much-0.081
 drift-0.078
irm-0.076
 nap-0.076
 dives-0.075
Token
Token ID447
Feature activation-1.96e-02

Pos logit contributions
 much+0.011
 drift+0.011
 dives+0.011
 nap+0.010
irm+0.010
Neg logit contributions
ting-0.018
 mill-0.018
 painter-0.018
ini-0.017
 paramed-0.017
Token
Token ID250
Feature activation+2.39e-01

Pos logit contributions
 much+0.020
 nap+0.019
 drift+0.019
 dives+0.018
irm+0.018
Neg logit contributions
ting-0.041
 mill-0.040
 painter-0.039
ini-0.039
 floors-0.038
Token she
Token ID673
Feature activation-4.48e-02

Pos logit contributions
 mill+0.302
ting+0.296
 painter+0.279
ini+0.273
 office+0.265
Neg logit contributions
 much-0.438
 drift-0.426
 nap-0.426
 dives-0.420
irm-0.417
Token just
Token ID655
Feature activation-1.16e-01

Pos logit contributions
 drift+0.053
 dives+0.053
irm+0.053
 nap+0.052
 much+0.052
Neg logit contributions
 mill-0.081
ting-0.080
 painter-0.078
ini-0.077
 office-0.076
Token needs
Token ID2476
Feature activation+3.72e-02

Pos logit contributions
 much+0.116
 dives+0.116
 nap+0.115
 drift+0.115
irm+0.114
Neg logit contributions
ting-0.231
 mill-0.229
 painter-0.220
ini-0.213
 office-0.212
Token to
Token ID284
Feature activation+4.21e-02

Pos logit contributions
 mill+0.071
ting+0.071
ini+0.069
 painter+0.069
 paramed+0.066
Neg logit contributions
 much-0.045
 nap-0.043
 drift-0.043
 dives-0.042
irm-0.040
Token remember
Token ID3505
Feature activation-1.01e-02

Pos logit contributions
 mill+0.083
ting+0.083
 painter+0.080
ini+0.079
 office+0.077
Neg logit contributions
 much-0.045
 dives-0.044
 drift-0.043
 nap-0.043
irm-0.042
Token hugged
Token ID41130
Feature activation-9.27e-03

Pos logit contributions
ting+0.050
 mill+0.049
 painter+0.048
ini+0.046
 office+0.045
Neg logit contributions
 much-0.039
 drift-0.039
 dives-0.039
 nap-0.038
irm-0.038
Token his
Token ID465
Feature activation-3.86e-02

Pos logit contributions
 much+0.012
 nap+0.011
 drift+0.011
irm+0.011
 dives+0.011
Neg logit contributions
 mill-0.017
ini-0.017
 painter-0.017
ting-0.016
 office-0.016
Token cat
Token ID3797
Feature activation+8.62e-02

Pos logit contributions
 much+0.046
 nap+0.045
 drift+0.044
 dives+0.044
irm+0.042
Neg logit contributions
ting-0.075
 mill-0.073
 painter-0.070
ini-0.070
 paramed-0.068
Token and
Token ID290
Feature activation-3.06e-02

Pos logit contributions
ting+0.137
 mill+0.131
 painter+0.127
 office+0.126
cut+0.125
Neg logit contributions
 dives-0.126
 drift-0.123
irm-0.119
ust-0.118
ls-0.116
Token said
Token ID531
Feature activation-4.20e-02

Pos logit contributions
 much+0.035
 drift+0.035
 dives+0.035
 nap+0.035
irm+0.034
Neg logit contributions
 mill-0.058
ting-0.057
 painter-0.055
ini-0.054
 paramed-0.053
Token:"
Token ID11097
Feature activation+2.39e-01

Pos logit contributions
irm+0.043
 drift+0.043
 nap+0.043
 dives+0.042
ust+0.040
Neg logit contributions
ting-0.084
 mill-0.083
 painter-0.081
 paramed-0.079
cut-0.079
TokenThank
Token ID10449
Feature activation+2.52e-02

Pos logit contributions
ting+0.245
 mill+0.236
 painter+0.225
ini+0.219
 paramed+0.205
Neg logit contributions
 dives-0.489
 much-0.488
 drift-0.484
 nap-0.483
irm-0.479
Token you
Token ID345
Feature activation-6.19e-02

Pos logit contributions
 mill+0.054
 painter+0.052
ting+0.052
ini+0.051
 paramed+0.050
Neg logit contributions
 much-0.024
 nap-0.023
 dives-0.023
 drift-0.023
irm-0.021
Token for
Token ID329
Feature activation+3.97e-02

Pos logit contributions
 drift+0.066
 dives+0.066
 much+0.065
 nap+0.065
irm+0.064
Neg logit contributions
 mill-0.122
ting-0.122
 painter-0.118
ini-0.116
 office-0.114
Token the
Token ID262
Feature activation+6.11e-02

Pos logit contributions
ini+0.076
smith+0.073
rom+0.071
 mill+0.070
 pavement+0.069
Neg logit contributions
 much-0.056
 nap-0.050
 swim-0.049
 good-0.048
 dive-0.047
Token delicious
Token ID12625
Feature activation-1.47e-02

Pos logit contributions
ting+0.125
 mill+0.124
 painter+0.120
ini+0.119
rom+0.116
Neg logit contributions
 much-0.063
 nap-0.062
 drift-0.061
 dives-0.061
irm-0.058
Token at
Token ID379
Feature activation-3.75e-02

Pos logit contributions
 dives+0.089
 drift+0.087
 much+0.087
 nap+0.087
irm+0.085
Neg logit contributions
ting-0.157
 mill-0.155
 painter-0.149
ini-0.142
 office-0.141
Token her
Token ID607
Feature activation-3.48e-02

Pos logit contributions
 much+0.040
 nap+0.038
 time+0.034
 drift+0.033
irm+0.032
Neg logit contributions
 paramed-0.081
ini-0.080
 mill-0.076
cles-0.076
cut-0.076
Token Mum
Token ID38367
Feature activation+7.47e-02

Pos logit contributions
 nap+0.049
 much+0.048
 dives+0.048
 drift+0.047
irm+0.045
Neg logit contributions
ting-0.061
 mill-0.057
ini-0.056
 paramed-0.055
cut-0.055
Token and
Token ID290
Feature activation-3.01e-02

Pos logit contributions
 mill+0.144
ting+0.143
 painter+0.139
ini+0.137
 paramed+0.133
Neg logit contributions
 much-0.084
 drift-0.083
 dives-0.082
 nap-0.082
irm-0.079
Token asked
Token ID1965
Feature activation+2.35e-02

Pos logit contributions
 much+0.035
 drift+0.035
 dives+0.034
 nap+0.034
irm+0.032
Neg logit contributions
ting-0.057
 mill-0.057
ini-0.057
 painter-0.054
 office-0.054
Token:"
Token ID11097
Feature activation+2.39e-01

Pos logit contributions
ini+0.053
 mill+0.052
 painter+0.049
 pavement+0.049
ting+0.048
Neg logit contributions
 much-0.025
 nap-0.021
 drift-0.020
 okay-0.020
 dives-0.019
TokenIs
Token ID3792
Feature activation+3.60e-02

Pos logit contributions
 mill+0.240
ini+0.231
 office+0.228
 floors+0.216
 painter+0.208
Neg logit contributions
 much-0.505
 drift-0.486
 nap-0.486
 dives-0.472
 OK-0.464
Token it
Token ID340
Feature activation-1.58e-02

Pos logit contributions
ting+0.061
 mill+0.060
ini+0.060
 painter+0.058
cut+0.057
Neg logit contributions
 much-0.051
 nap-0.049
 dives-0.049
 drift-0.047
ls-0.046
Token ready
Token ID3492
Feature activation-9.84e-02

Pos logit contributions
irm+0.021
 dives+0.020
 drift+0.020
 hops+0.019
 much+0.019
Neg logit contributions
 mill-0.028
ting-0.027
 painter-0.027
ini-0.026
 office-0.026
Token,
Token ID11
Feature activation+5.03e-02

Pos logit contributions
 much+0.126
 drift+0.124
 dives+0.124
 nap+0.123
irm+0.120
Neg logit contributions
ting-0.174
 mill-0.174
 painter-0.167
ini-0.165
 paramed-0.160
Token Mum
Token ID38367
Feature activation+7.28e-02

Pos logit contributions
ting+0.085
 mill+0.085
 painter+0.082
ini+0.081
 paramed+0.078
Neg logit contributions
 much-0.068
 drift-0.067
 dives-0.067
 nap-0.067
irm-0.065
Token special
Token ID2041
Feature activation-1.16e-02

Pos logit contributions
 much+0.032
 drift+0.032
 dives+0.032
 nap+0.032
irm+0.031
Neg logit contributions
 mill-0.033
ting-0.033
 painter-0.031
ini-0.031
 office-0.030
Token things
Token ID1243
Feature activation+9.36e-03

Pos logit contributions
 much+0.019
 dives+0.019
 nap+0.018
 drift+0.018
irm+0.018
Neg logit contributions
ting-0.017
 mill-0.017
 painter-0.016
ini-0.016
 paramed-0.016
Tokenâ
Token ID22940
Feature activation+1.83e-02

Pos logit contributions
ting+0.016
 mill+0.015
 painter+0.014
cut+0.014
 office+0.014
Neg logit contributions
 dives-0.013
irm-0.013
 drift-0.013
 nap-0.013
 much-0.012
Token
Token ID26391
Feature activation-1.26e-02

Pos logit contributions
ting+0.028
 mill+0.027
 painter+0.026
ini+0.026
 office+0.025
Neg logit contributions
 much-0.029
 dives-0.028
 drift-0.028
 nap-0.028
irm-0.027
Token
Token ID447
Feature activation-1.84e-02

Pos logit contributions
 drift+0.016
 dives+0.016
 much+0.016
irm+0.016
 nap+0.016
Neg logit contributions
 mill-0.022
ting-0.022
 painter-0.021
ini-0.021
 office-0.021
Token
Token ID250
Feature activation+2.39e-01

Pos logit contributions
 much+0.026
 drift+0.025
 dives+0.025
 nap+0.025
irm+0.024
Neg logit contributions
ting-0.030
 mill-0.030
 painter-0.029
ini-0.029
cut-0.028
Token little
Token ID1310
Feature activation+2.15e-02

Pos logit contributions
 mill+0.310
ting+0.288
 painter+0.284
 office+0.277
 paramed+0.273
Neg logit contributions
 much-0.435
 drift-0.416
 nap-0.414
ust-0.411
irm-0.408
Token stones
Token ID14966
Feature activation-1.78e-02

Pos logit contributions
 mill+0.037
ting+0.037
 painter+0.036
ini+0.036
 paramed+0.035
Neg logit contributions
 much-0.028
 nap-0.028
 dives-0.028
 drift-0.028
irm-0.027
Token,
Token ID11
Feature activation+5.26e-02

Pos logit contributions
 much+0.020
 drift+0.020
 nap+0.020
 dives+0.020
irm+0.018
Neg logit contributions
 mill-0.034
ting-0.034
 painter-0.033
ini-0.033
 paramed-0.032
Token happy
Token ID3772
Feature activation+2.21e-02

Pos logit contributions
ting+0.100
 mill+0.099
 painter+0.096
ini+0.095
 paramed+0.093
Neg logit contributions
 much-0.063
 nap-0.061
 drift-0.060
 dives-0.060
irm-0.057
Token flowers
Token ID12734
Feature activation+1.60e-02

Pos logit contributions
 mill+0.039
ting+0.039
ini+0.038
 painter+0.038
 paramed+0.037
Neg logit contributions
 much-0.029
 dives-0.028
 nap-0.028
 drift-0.028
 hops-0.027
Token strange
Token ID6283
Feature activation+1.50e-02

Pos logit contributions
 dives+0.073
 drift+0.072
irm+0.072
 nap+0.072
 much+0.072
Neg logit contributions
ting-0.106
 mill-0.106
 painter-0.101
ini-0.099
cut-0.096
Token
Token ID6184
Feature activation-2.16e-02

Pos logit contributions
 mill+0.029
ting+0.029
ini+0.028
 painter+0.028
 paramed+0.028
Neg logit contributions
 much-0.017
 drift-0.016
 dives-0.016
 nap-0.016
irm-0.015
Token
Token ID95
Feature activation+4.96e-02

Pos logit contributions
 dives+0.032
 drift+0.031
 nap+0.031
 much+0.031
 dive+0.031
Neg logit contributions
ting-0.033
 mill-0.033
 painter-0.033
ini-0.033
cut-0.032
Token
Token ID26391
Feature activation-1.33e-02

Pos logit contributions
 mill+0.077
 painter+0.075
ting+0.074
ini+0.072
cut+0.072
Neg logit contributions
 dives-0.072
 drift-0.071
 much-0.070
irm-0.069
 nap-0.069
Token
Token ID447
Feature activation-1.84e-02

Pos logit contributions
 much+0.020
 drift+0.020
 dives+0.020
 nap+0.020
irm+0.020
Neg logit contributions
 mill-0.020
ting-0.020
 painter-0.019
ini-0.019
 office-0.018
Token
Token ID250
Feature activation+2.39e-01

Pos logit contributions
 much+0.026
 drift+0.025
 dives+0.025
 nap+0.025
irm+0.024
Neg logit contributions
ting-0.031
 mill-0.030
 painter-0.030
ini-0.029
cut-0.028
Token the
Token ID262
Feature activation+6.22e-02

Pos logit contributions
 mill+0.307
ting+0.289
 painter+0.282
 office+0.278
cut+0.271
Neg logit contributions
 much-0.437
 drift-0.423
ust-0.420
 nap-0.419
 dives-0.414
Token k
Token ID479
Feature activation-1.96e-03

Pos logit contributions
ini+0.099
ting+0.099
 mill+0.098
 painter+0.097
rom+0.094
Neg logit contributions
 nap-0.093
 much-0.093
 drift-0.089
 dives-0.087
 hops-0.086
Tokenitty
Token ID9760
Feature activation+4.76e-02

Pos logit contributions
 much+0.003
 dives+0.003
 nap+0.003
 hops+0.003
ls+0.003
Neg logit contributions
 mill-0.003
 painter-0.003
ting-0.003
 floors-0.003
 paramed-0.003
Token was
Token ID373
Feature activation+3.41e-02

Pos logit contributions
ting+0.091
 mill+0.090
 painter+0.087
ini+0.086
 paramed+0.084
Neg logit contributions
 much-0.055
 drift-0.053
 dives-0.053
 nap-0.053
irm-0.051
Token wearing
Token ID5762
Feature activation-1.63e-02

Pos logit contributions
ting+0.064
 mill+0.064
 painter+0.061
ini+0.060
 office+0.059
Neg logit contributions
 drift-0.039
 dives-0.039
 much-0.039
 nap-0.039
irm-0.038
Token him
Token ID683
Feature activation-1.23e-02

Pos logit contributions
 much+0.007
 nap+0.007
 drift+0.007
 dives+0.007
irm+0.006
Neg logit contributions
 mill-0.007
ini-0.007
 painter-0.007
ting-0.007
 office-0.007
Token.
Token ID13
Feature activation+5.60e-03

Pos logit contributions
 dives+0.017
 drift+0.017
 nap+0.016
irm+0.016
 much+0.016
Neg logit contributions
ting-0.022
 painter-0.020
 mill-0.019
 office-0.019
cut-0.019
Token She
Token ID1375
Feature activation+1.23e-02

Pos logit contributions
 mill+0.011
ting+0.011
ini+0.011
 painter+0.011
cut+0.010
Neg logit contributions
 much-0.006
 nap-0.006
 drift-0.006
 dives-0.006
irm-0.006
Token says
Token ID1139
Feature activation-4.74e-02

Pos logit contributions
 mill+0.021
ting+0.020
 office+0.019
 painter+0.019
 paramed+0.019
Neg logit contributions
ust-0.017
 much-0.017
 drift-0.016
irm-0.016
 nap-0.015
Token,
Token ID11
Feature activation+5.51e-02

Pos logit contributions
 drift+0.051
 much+0.051
 dives+0.051
 nap+0.050
irm+0.050
Neg logit contributions
ting-0.093
 mill-0.093
 painter-0.090
ini-0.088
 paramed-0.087
Token "
Token ID366
Feature activation+2.39e-01

Pos logit contributions
ini+0.094
 mill+0.089
ting+0.088
cles+0.087
 floors+0.086
Neg logit contributions
 much-0.081
 drift-0.078
 nap-0.077
 dives-0.074
 OK-0.073
TokenI
Token ID40
Feature activation+2.76e-02

Pos logit contributions
 mill+0.301
ini+0.282
 office+0.279
 painter+0.277
ting+0.275
Neg logit contributions
 nap-0.448
 drift-0.436
 much-0.433
 dives-0.418
irm-0.417
Token am
Token ID716
Feature activation+5.68e-02

Pos logit contributions
 mill+0.043
 painter+0.042
 paramed+0.041
ting+0.041
ini+0.040
Neg logit contributions
 drift-0.041
irm-0.040
 dives-0.040
 much-0.039
ust-0.039
Token sorry
Token ID7926
Feature activation-7.61e-04

Pos logit contributions
ting+0.113
 mill+0.113
 painter+0.110
 paramed+0.108
 office+0.107
Neg logit contributions
 dives-0.062
 drift-0.061
irm-0.058
 hops-0.057
 nap-0.057
Token,
Token ID11
Feature activation+5.24e-02

Pos logit contributions
irm+0.001
 drift+0.001
 dives+0.001
ust+0.001
 nap+0.001
Neg logit contributions
ting-0.001
 painter-0.001
 mill-0.001
 office-0.001
Why-0.001
Token Ben
Token ID3932
Feature activation-1.96e-03

Pos logit contributions
ting+0.092
 mill+0.092
 painter+0.088
 paramed+0.083
ter+0.081
Neg logit contributions
irm-0.069
 dives-0.069
 much-0.068
 drift-0.068
ust-0.065
Token about
Token ID546
Feature activation+3.00e-03

Pos logit contributions
 much+0.047
 drift+0.047
 dives+0.046
 nap+0.046
irm+0.045
Neg logit contributions
ting-0.075
 mill-0.074
 painter-0.072
ini-0.070
 office-0.068
Token him
Token ID683
Feature activation-1.90e-02

Pos logit contributions
 mill+0.005
ini+0.005
ting+0.005
 painter+0.005
 paramed+0.005
Neg logit contributions
 much-0.004
 nap-0.004
 drift-0.004
 dives-0.004
 swim-0.004
Token,
Token ID11
Feature activation+5.07e-02

Pos logit contributions
 much+0.022
 drift+0.022
 dives+0.022
 nap+0.022
irm+0.021
Neg logit contributions
ting-0.036
 mill-0.036
 painter-0.034
ini-0.034
 office-0.033
Token he
Token ID339
Feature activation-5.51e-02

Pos logit contributions
 mill+0.094
ting+0.093
 painter+0.090
ini+0.089
 paramed+0.086
Neg logit contributions
 much-0.062
 drift-0.060
 dives-0.060
 nap-0.060
irm-0.058
Token said
Token ID531
Feature activation-4.59e-02

Pos logit contributions
 much+0.078
 drift+0.077
 dives+0.077
 nap+0.077
irm+0.076
Neg logit contributions
ting-0.090
 mill-0.089
 painter-0.086
ini-0.084
 office-0.081
Token:"
Token ID11097
Feature activation+2.39e-01

Pos logit contributions
 much+0.050
 drift+0.049
 dives+0.049
 nap+0.049
irm+0.047
Neg logit contributions
ting-0.090
 mill-0.090
 painter-0.087
ini-0.085
 paramed-0.083
TokenMy
Token ID3666
Feature activation-3.18e-02

Pos logit contributions
ting+0.267
 mill+0.262
 painter+0.248
ini+0.242
 paramed+0.228
Neg logit contributions
 much-0.464
 dives-0.461
 drift-0.460
 nap-0.458
irm-0.451
Token name
Token ID1438
Feature activation-1.74e-02

Pos logit contributions
 much+0.043
 nap+0.042
 dives+0.041
 drift+0.041
irm+0.039
Neg logit contributions
ting-0.056
 mill-0.054
ini-0.052
 painter-0.052
 paramed-0.050
Token is
Token ID318
Feature activation+2.94e-02

Pos logit contributions
 much+0.020
 drift+0.020
 nap+0.020
 dives+0.020
irm+0.019
Neg logit contributions
 mill-0.033
ting-0.033
 painter-0.032
ini-0.031
 paramed-0.031
Token L
Token ID406
Feature activation+1.51e-02

Pos logit contributions
ting+0.053
 mill+0.053
 painter+0.051
ini+0.050
 paramed+0.049
Neg logit contributions
 drift-0.036
 dives-0.036
 much-0.036
irm-0.036
 nap-0.036
Tokenenny
Token ID11870
Feature activation+2.94e-02

Pos logit contributions
ting+0.028
 mill+0.028
 painter+0.027
ini+0.027
 paramed+0.026
Neg logit contributions
 much-0.018
 drift-0.018
 dives-0.018
 nap-0.018
irm-0.017
Token day
Token ID1110
Feature activation-2.69e-02

Pos logit contributions
 mill+0.052
ting+0.052
 painter+0.050
ini+0.048
 office+0.048
Neg logit contributions
 dives-0.039
 drift-0.039
irm-0.039
 much-0.039
 nap-0.039
Token
Token ID6184
Feature activation-2.79e-02

Pos logit contributions
 much+0.033
 drift+0.032
 dives+0.032
 nap+0.032
irm+0.031
Neg logit contributions
ting-0.049
 mill-0.049
 painter-0.047
ini-0.046
 office-0.045
Token
Token ID95
Feature activation+4.64e-02

Pos logit contributions
 much+0.041
 drift+0.040
 dives+0.040
 nap+0.040
irm+0.039
Neg logit contributions
ting-0.045
 mill-0.044
 painter-0.042
ini-0.042
 paramed-0.040
Token
Token ID26391
Feature activation-1.32e-02

Pos logit contributions
 mill+0.077
ting+0.077
 painter+0.072
ini+0.070
cut+0.070
Neg logit contributions
 much-0.065
irm-0.062
 drift-0.061
 dives-0.061
 nap-0.061
Token
Token ID447
Feature activation-2.26e-02

Pos logit contributions
 much+0.018
 dives+0.017
 drift+0.017
 nap+0.017
irm+0.017
Neg logit contributions
ting-0.023
 mill-0.023
 painter-0.022
ini-0.022
 paramed-0.021
Token
Token ID250
Feature activation+2.39e-01

Pos logit contributions
 much+0.025
 nap+0.023
 drift+0.022
irm+0.022
 dives+0.021
Neg logit contributions
ting-0.045
 mill-0.044
 painter-0.044
 floors-0.043
ini-0.043
Token to
Token ID284
Feature activation+4.40e-02

Pos logit contributions
 mill+0.287
ting+0.274
 painter+0.258
ini+0.247
 office+0.247
Neg logit contributions
 much-0.457
 drift-0.433
 nap-0.433
irm-0.432
ust-0.430
Token always
Token ID1464
Feature activation-8.30e-03

Pos logit contributions
ting+0.085
 mill+0.084
ini+0.081
 painter+0.080
 paramed+0.078
Neg logit contributions
 much-0.052
 nap-0.051
 drift-0.051
 dives-0.049
irm-0.048
Token be
Token ID307
Feature activation+7.84e-03

Pos logit contributions
ls+0.009
 dives+0.009
 much+0.009
 drift+0.009
irm+0.009
Neg logit contributions
 mill-0.015
ting-0.015
 painter-0.015
 office-0.015
ini-0.014
Token careful
Token ID8161
Feature activation+4.53e-02

Pos logit contributions
 mill+0.016
ting+0.016
 painter+0.015
ini+0.015
rom+0.015
Neg logit contributions
 much-0.009
 dives-0.008
 drift-0.008
irm-0.008
 nap-0.008
Token.
Token ID13
Feature activation-2.48e-03

Pos logit contributions
 mill+0.090
ting+0.089
 painter+0.085
ini+0.084
 laund+0.083
Neg logit contributions
 much-0.053
 drift-0.048
 nap-0.047
 dives-0.047
irm-0.046
Token friends
Token ID2460
Feature activation+1.71e-02

Pos logit contributions
ting+0.009
 mill+0.009
ini+0.009
 painter+0.008
rom+0.008
Neg logit contributions
 nap-0.007
 much-0.007
 dives-0.007
 drift-0.007
irm-0.006
Token
Token ID6184
Feature activation-2.19e-02

Pos logit contributions
 mill+0.036
ini+0.035
ting+0.035
 painter+0.035
 paramed+0.034
Neg logit contributions
 much-0.018
 nap-0.017
 drift-0.016
 dives-0.015
irm-0.015
Token
Token ID95
Feature activation+5.02e-02

Pos logit contributions
 much+0.033
 drift+0.032
 dives+0.032
 nap+0.032
irm+0.031
Neg logit contributions
ting-0.034
 mill-0.034
 painter-0.033
ini-0.032
 paramed-0.031
Token
Token ID26391
Feature activation-1.37e-02

Pos logit contributions
 mill+0.076
ting+0.074
 painter+0.070
ini+0.068
 floors+0.068
Neg logit contributions
 much-0.078
irm-0.075
 dives-0.074
 nap-0.074
 drift-0.074
Token
Token ID447
Feature activation-1.94e-02

Pos logit contributions
 much+0.019
 dives+0.019
 drift+0.019
 nap+0.019
irm+0.018
Neg logit contributions
ting-0.023
 mill-0.023
 painter-0.022
ini-0.021
 paramed-0.021
Token
Token ID250
Feature activation+2.39e-01

Pos logit contributions
 much+0.019
 nap+0.017
 drift+0.017
 dives+0.017
irm+0.016
Neg logit contributions
ting-0.042
 mill-0.040
 painter-0.040
ini-0.039
keeping-0.039
Token it
Token ID340
Feature activation-1.57e-02

Pos logit contributions
 mill+0.305
ting+0.292
 painter+0.276
 office+0.271
ini+0.268
Neg logit contributions
 much-0.440
 nap-0.415
ust-0.413
irm-0.413
 drift-0.412
Token's
Token ID338
Feature activation-2.42e-02

Pos logit contributions
 nap+0.020
 drift+0.019
irm+0.019
 dives+0.019
 much+0.019
Neg logit contributions
ting-0.027
 mill-0.027
 painter-0.026
ini-0.026
cut-0.026
Token easy
Token ID2562
Feature activation-1.31e-02

Pos logit contributions
 dives+0.023
 drift+0.022
 nap+0.021
ls+0.021
irm+0.021
Neg logit contributions
 mill-0.051
ting-0.050
 painter-0.049
 office-0.048
 paramed-0.047
Token then
Token ID788
Feature activation+2.42e-02

Pos logit contributions
 much+0.015
 nap+0.014
 dives+0.014
 drift+0.014
irm+0.013
Neg logit contributions
ting-0.026
 mill-0.026
ini-0.025
 painter-0.025
 paramed-0.024
Token!"
Token ID2474
Feature activation+6.94e-02

Pos logit contributions
ting+0.045
 mill+0.045
 painter+0.043
ini+0.043
 office+0.042
Neg logit contributions
 much-0.029
 drift-0.028
 dives-0.028
 nap-0.028
irm-0.027
Token her
Token ID607
Feature activation-2.81e-02

Pos logit contributions
 much+0.005
 nap+0.005
 drift+0.004
irm+0.004
 swim+0.004
Neg logit contributions
 mill-0.010
ini-0.009
 painter-0.009
 laund-0.009
cut-0.008
Token.
Token ID13
Feature activation+6.32e-03

Pos logit contributions
 dives+0.044
 drift+0.041
 dive+0.041
irm+0.040
ust+0.040
Neg logit contributions
ting-0.044
 painter-0.040
cut-0.040
 office-0.039
 mill-0.039
Token He
Token ID679
Feature activation+9.88e-04

Pos logit contributions
 mill+0.013
ting+0.013
 painter+0.013
ini+0.013
 paramed+0.012
Neg logit contributions
 much-0.006
 drift-0.006
 dives-0.006
 nap-0.006
irm-0.006
Token says
Token ID1139
Feature activation-4.72e-02

Pos logit contributions
 mill+0.002
ting+0.002
 office+0.002
 painter+0.001
 paramed+0.001
Neg logit contributions
ust-0.001
 much-0.001
 drift-0.001
irm-0.001
ver-0.001
Token:
Token ID25
Feature activation+1.04e-01

Pos logit contributions
 much+0.049
 drift+0.049
 dives+0.048
 nap+0.048
irm+0.047
Neg logit contributions
ting-0.095
 mill-0.095
 painter-0.092
ini-0.090
 paramed-0.088
Token "
Token ID366
Feature activation+2.39e-01

Pos logit contributions
ting+0.187
 mill+0.187
ini+0.179
 painter+0.178
 office+0.171
Neg logit contributions
 much-0.132
 drift-0.128
 dives-0.128
 nap-0.128
irm-0.122
TokenI
Token ID40
Feature activation+2.80e-02

Pos logit contributions
 mill+0.260
ting+0.257
 painter+0.241
ini+0.235
 office+0.224
Neg logit contributions
 much-0.475
 drift-0.470
 nap-0.468
 dives-0.468
irm-0.461
Token am
Token ID716
Feature activation+5.60e-02

Pos logit contributions
 mill+0.040
 paramed+0.039
 painter+0.038
ini+0.037
ting+0.036
Neg logit contributions
 drift-0.043
irm-0.043
ls-0.043
 dives-0.043
ust-0.042
Token sorry
Token ID7926
Feature activation-1.67e-03

Pos logit contributions
ting+0.110
 mill+0.110
 painter+0.107
 paramed+0.106
 office+0.105
Neg logit contributions
 dives-0.061
 drift-0.059
irm-0.057
ust-0.057
 nap-0.056
Token,
Token ID11
Feature activation+5.13e-02

Pos logit contributions
 drift+0.002
 dives+0.002
irm+0.002
 nap+0.002
ust+0.002
Neg logit contributions
ting-0.003
 painter-0.003
 mill-0.003
 office-0.003
ini-0.002
Token Lily
Token ID20037
Feature activation+1.71e-02

Pos logit contributions
ting+0.101
 mill+0.101
 painter+0.097
ini+0.096
 paramed+0.093
Neg logit contributions
 much-0.057
 drift-0.056
 dives-0.055
 nap-0.055
irm-0.053
Token rocked
Token ID36872
Feature activation-5.28e-02

Pos logit contributions
 much+0.047
ust+0.047
 dives+0.044
irm+0.044
 nap+0.043
Neg logit contributions
ting-0.065
 mill-0.061
 painter-0.058
cut-0.057
 pavement-0.056
Token
Token ID6184
Feature activation-3.16e-02

Pos logit contributions
 much+0.050
 drift+0.041
 nap+0.041
irm+0.038
 dives+0.037
Neg logit contributions
 mill-0.125
 painter-0.121
ini-0.120
 paramed-0.116
 office-0.115
Token
Token ID95
Feature activation+4.54e-02

Pos logit contributions
 much+0.048
 dives+0.046
 drift+0.046
ls+0.046
 nap+0.045
Neg logit contributions
ting-0.049
 mill-0.049
 painter-0.046
 paramed-0.045
rom-0.045
Token
Token ID26391
Feature activation-1.51e-02

Pos logit contributions
 mill+0.074
ting+0.073
 painter+0.070
ini+0.069
cut+0.067
Neg logit contributions
 much-0.065
 dives-0.064
 drift-0.064
 nap-0.063
irm-0.063
Token
Token ID447
Feature activation-2.11e-02

Pos logit contributions
 drift+0.019
 dives+0.019
 nap+0.019
irm+0.019
 much+0.019
Neg logit contributions
 mill-0.026
ting-0.026
 painter-0.025
ini-0.025
 office-0.025
Token
Token ID250
Feature activation+2.39e-01

Pos logit contributions
 much+0.028
 drift+0.028
 nap+0.027
 dives+0.027
irm+0.027
Neg logit contributions
ting-0.036
 mill-0.036
 painter-0.035
ini-0.035
 office-0.033
Token higher
Token ID2440
Feature activation-7.22e-02

Pos logit contributions
 mill+0.298
ting+0.297
 painter+0.281
ini+0.275
 paramed+0.262
Neg logit contributions
 much-0.430
 drift-0.425
 dives-0.424
 nap-0.422
irm-0.415
Token and
Token ID290
Feature activation-3.50e-02

Pos logit contributions
 much+0.086
 drift+0.084
 dives+0.084
 nap+0.084
irm+0.081
Neg logit contributions
ting-0.134
 mill-0.134
 painter-0.129
ini-0.128
 paramed-0.124
Token higher
Token ID2440
Feature activation-7.32e-02

Pos logit contributions
 much+0.040
 nap+0.039
 drift+0.039
 dives+0.039
irm+0.039
Neg logit contributions
ting-0.066
 mill-0.065
 painter-0.062
ini-0.061
cut-0.060
Token.
Token ID13
Feature activation-5.53e-04

Pos logit contributions
 much+0.088
 drift+0.086
 dives+0.086
 nap+0.085
irm+0.083
Neg logit contributions
ting-0.135
 mill-0.135
 painter-0.130
ini-0.129
 paramed-0.125
Token His
Token ID2399
Feature activation-1.65e-02

Pos logit contributions
 much+0.001
 drift+0.001
 dives+0.001
 nap+0.001
irm+0.001
Neg logit contributions
ting-0.001
 mill-0.001
 painter-0.001
ini-0.001
 paramed-0.001
Token's
Token ID338
Feature activation-2.58e-02

Pos logit contributions
 dives+0.025
 drift+0.024
 nap+0.024
 much+0.024
irm+0.024
Neg logit contributions
ting-0.037
 mill-0.036
 painter-0.035
cut-0.034
ini-0.034
Token mom
Token ID1995
Feature activation+1.03e-01

Pos logit contributions
 dives+0.037
 drift+0.037
 much+0.036
irm+0.036
 nap+0.036
Neg logit contributions
 mill-0.041
ting-0.041
 painter-0.040
ini-0.039
 office-0.038
Tokenmy
Token ID1820
Feature activation+2.78e-02

Pos logit contributions
 mill+0.150
ting+0.142
 painter+0.134
ini+0.133
 laund+0.132
Neg logit contributions
 much-0.174
 drift-0.169
 nap-0.165
irm-0.158
ust-0.157
Token.
Token ID13
Feature activation-4.75e-03

Pos logit contributions
ting+0.053
 mill+0.053
 painter+0.051
ini+0.050
 paramed+0.049
Neg logit contributions
 dives-0.032
 much-0.032
 nap-0.031
 drift-0.031
irm-0.031
Token "
Token ID366
Feature activation+2.28e-01

Pos logit contributions
 much+0.007
 drift+0.006
 nap+0.006
 dives+0.006
irm+0.006
Neg logit contributions
ini-0.008
 mill-0.008
ting-0.008
 painter-0.008
cut-0.007
TokenLooks
Token ID41102
Feature activation-1.88e-01

Pos logit contributions
 mill+0.330
 office+0.314
ting+0.306
 painter+0.291
cles+0.291
Neg logit contributions
 much-0.393
ust-0.377
 drift-0.351
irm-0.351
 nap-0.347
Token like
Token ID588
Feature activation+4.10e-02

Pos logit contributions
 drift+0.266
 nap+0.256
ls+0.256
 dives+0.252
 dive+0.247
Neg logit contributions
ting-0.291
 painter-0.275
 mill-0.273
 pavement-0.264
 floors-0.264
Token the
Token ID262
Feature activation+5.96e-02

Pos logit contributions
ting+0.078
ini+0.076
 mill+0.075
cut+0.074
rom+0.073
Neg logit contributions
 much-0.050
 nap-0.048
 dives-0.047
 drift-0.046
 dive-0.045
Token wave
Token ID6769
Feature activation-5.95e-02

Pos logit contributions
ting+0.080
 mill+0.079
 painter+0.075
ini+0.074
 paramed+0.071
Neg logit contributions
 much-0.103
 drift-0.102
 dives-0.102
 nap-0.101
irm-0.099
Token destroyed
Token ID6572
Feature activation+7.59e-02

Pos logit contributions
 much+0.069
 drift+0.069
 dives+0.068
 nap+0.068
irm+0.068
Neg logit contributions
ting-0.111
 mill-0.111
 painter-0.107
ini-0.105
 office-0.102
Token all
Token ID477
Feature activation-2.25e-02

Pos logit contributions
ting+0.140
 mill+0.139
 painter+0.139
ini+0.138
 paramed+0.137
Neg logit contributions
 much-0.092
 nap-0.089
 dives-0.087
 drift-0.086
 dive-0.081
Token the
Token ID262
Feature activation+5.77e-02

Pos logit contributions
ust+0.044
 dives+0.042
 nap+0.042
 drift+0.042
ares+0.041
Neg logit contributions
ting-0.052
 mill-0.049
 painter-0.046
 paramed-0.045
Why-0.044
Token owner
Token ID4870
Feature activation+8.02e-02

Pos logit contributions
ting+0.079
 mill+0.078
ini+0.075
 painter+0.075
rom+0.072
Neg logit contributions
 nap-0.100
 much-0.100
 drift-0.097
 dives-0.097
irm-0.095
Token smiled
Token ID13541
Feature activation+6.50e-03

Pos logit contributions
ting+0.127
 mill+0.122
 painter+0.122
 paramed+0.117
ini+0.116
Neg logit contributions
 dives-0.118
 drift-0.117
 much-0.115
irm-0.115
 nap-0.113
Token.
Token ID13
Feature activation-7.45e-03

Pos logit contributions
ting+0.011
 painter+0.010
 mill+0.010
ini+0.010
ter+0.010
Neg logit contributions
 dives-0.009
 nap-0.009
 drift-0.009
 dive-0.009
 hops-0.009
Token "
Token ID366
Feature activation+2.28e-01

Pos logit contributions
 much+0.010
 drift+0.010
 dives+0.009
 nap+0.009
irm+0.009
Neg logit contributions
 mill-0.013
ting-0.013
 painter-0.013
ini-0.013
 paramed-0.012
TokenLooks
Token ID41102
Feature activation-1.88e-01

Pos logit contributions
 mill+0.267
ting+0.255
 painter+0.241
 office+0.238
ini+0.233
Neg logit contributions
 much-0.445
 drift-0.431
 nap-0.427
 dives-0.425
irm-0.421
Token like
Token ID588
Feature activation+4.12e-02

Pos logit contributions
ls+0.311
 drift+0.311
 dive+0.298
 nap+0.296
ust+0.291
Neg logit contributions
ting-0.241
 pavement-0.215
 painter-0.211
 floors-0.198
Why-0.196
Token you
Token ID345
Feature activation-6.39e-02

Pos logit contributions
ting+0.074
 mill+0.073
 painter+0.071
ini+0.068
 office+0.067
Neg logit contributions
 drift-0.051
 dives-0.051
 much-0.051
irm-0.050
 nap-0.050
Token two
Token ID734
Feature activation-6.52e-02

Pos logit contributions
 much+0.070
 nap+0.067
 drift+0.066
 dives+0.065
irm+0.062
Neg logit contributions
 mill-0.131
ting-0.129
ini-0.125
 painter-0.125
 office-0.121
Token are
Token ID389
Feature activation+2.66e-02

Pos logit contributions
irm+0.089
 dives+0.088
 drift+0.088
ust+0.086
 much+0.085
Neg logit contributions
ting-0.109
 mill-0.104
 painter-0.099
ini-0.096
 paramed-0.096
Token friends
Token ID2460
Feature activation+1.10e-02

Pos logit contributions
ting+0.044
 mill+0.042
 painter+0.041
 paramed+0.039
ini+0.039
Neg logit contributions
 dives-0.038
irm-0.038
 drift-0.037
 nap-0.036
ust-0.036
Token print
Token ID3601
Feature activation-2.44e-02

Pos logit contributions
 much+0.085
 drift+0.084
 dives+0.083
irm+0.083
ust+0.083
Neg logit contributions
ting-0.115
 mill-0.114
 office-0.106
 painter-0.106
ini-0.104
Token so
Token ID523
Feature activation-4.52e-02

Pos logit contributions
 much+0.028
 nap+0.027
 drift+0.026
 dives+0.026
irm+0.025
Neg logit contributions
 mill-0.048
ting-0.047
 painter-0.046
ini-0.046
 paramed-0.044
Token badly
Token ID11234
Feature activation-3.59e-03

Pos logit contributions
 much+0.046
 drift+0.044
 dives+0.044
 nap+0.044
irm+0.043
Neg logit contributions
ting-0.093
 mill-0.093
 painter-0.090
ini-0.089
 paramed-0.086
Token.
Token ID13
Feature activation-4.91e-03

Pos logit contributions
 much+0.004
 drift+0.004
 dives+0.004
 nap+0.004
irm+0.004
Neg logit contributions
ting-0.007
 mill-0.007
 painter-0.006
ini-0.006
 office-0.006
Token Fr
Token ID1305
Feature activation-1.64e-02

Pos logit contributions
 much+0.005
irm+0.005
 drift+0.005
 nap+0.005
 dives+0.005
Neg logit contributions
ini-0.010
cut-0.010
 painter-0.010
ting-0.010
 mill-0.010
Tokenust
Token ID436
Feature activation-1.87e-01

Pos logit contributions
 drift+0.012
 dives+0.012
 nap+0.011
 much+0.011
 hops+0.010
Neg logit contributions
 mill-0.039
ini-0.038
ting-0.038
 painter-0.038
rom-0.037
Tokenrated
Token ID4111
Feature activation-1.25e-02

Pos logit contributions
 nap+0.270
 much+0.261
 drift+0.258
 dives+0.257
irm+0.249
Neg logit contributions
 mill-0.295
ting-0.292
cut-0.287
ini-0.284
 painter-0.280
Token tears
Token ID10953
Feature activation-4.41e-02

Pos logit contributions
 much+0.013
 dives+0.012
 drift+0.012
 nap+0.012
irm+0.011
Neg logit contributions
 mill-0.026
ting-0.026
 painter-0.025
ini-0.025
 paramed-0.024
Token were
Token ID547
Feature activation+2.15e-02

Pos logit contributions
 much+0.054
 drift+0.053
 dives+0.053
 nap+0.053
irm+0.052
Neg logit contributions
ting-0.080
 mill-0.080
 painter-0.077
ini-0.076
 paramed-0.074
Token well
Token ID880
Feature activation-3.37e-02

Pos logit contributions
ini+0.044
ting+0.044
 painter+0.044
 mill+0.044
smith+0.043
Neg logit contributions
 much-0.026
 drift-0.021
 nap-0.021
 okay-0.021
irm-0.020
Tokening
Token ID278
Feature activation-7.26e-04

Pos logit contributions
 dives+0.043
 drift+0.042
 much+0.042
irm+0.042
 nap+0.042
Neg logit contributions
ting-0.060
 mill-0.058
 painter-0.054
 office-0.054
ini-0.052
Token laughed
Token ID13818
Feature activation-3.46e-02

Pos logit contributions
 mill+0.056
ini+0.054
ting+0.054
 pavement+0.052
 office+0.052
Neg logit contributions
 much-0.033
 drift-0.033
 dives-0.032
 nap-0.032
 swim-0.030
Token and
Token ID290
Feature activation-2.85e-02

Pos logit contributions
 dives+0.051
 drift+0.050
 nap+0.050
irm+0.049
 dive+0.049
Neg logit contributions
ting-0.057
 mill-0.052
 painter-0.051
Why-0.049
 paramed-0.048
Token said
Token ID531
Feature activation-4.47e-02

Pos logit contributions
 much+0.033
 drift+0.032
 dives+0.032
 nap+0.032
irm+0.031
Neg logit contributions
 mill-0.055
ting-0.054
 painter-0.052
ini-0.052
 office-0.051
Token,
Token ID11
Feature activation+4.82e-02

Pos logit contributions
 much+0.046
 nap+0.041
 dives+0.041
 drift+0.041
 dive+0.037
Neg logit contributions
 mill-0.095
ting-0.094
ini-0.094
 painter-0.091
 office-0.090
Token "
Token ID366
Feature activation+2.29e-01

Pos logit contributions
ini+0.076
 mill+0.074
ting+0.071
keeping+0.071
 painter+0.070
Neg logit contributions
 much-0.075
 drift-0.073
 dives-0.069
 nap-0.069
 swim-0.067
TokenLooks
Token ID41102
Feature activation-1.87e-01

Pos logit contributions
 mill+0.328
 office+0.302
ting+0.289
 painter+0.283
cut+0.279
Neg logit contributions
 much-0.406
 drift-0.386
ust-0.385
 nap-0.381
irm-0.378
Token like
Token ID588
Feature activation+4.14e-02

Pos logit contributions
 drift+0.307
ls+0.305
 dive+0.291
ust+0.286
 nap+0.286
Neg logit contributions
ting-0.241
 painter-0.222
 pavement-0.216
 floors-0.207
 mill-0.203
Token that
Token ID326
Feature activation+2.61e-02

Pos logit contributions
 mill+0.078
ting+0.078
ini+0.075
 painter+0.075
 paramed+0.073
Neg logit contributions
 much-0.049
 nap-0.048
 drift-0.047
 dives-0.047
irm-0.045
Token apple
Token ID17180
Feature activation+1.88e-02

Pos logit contributions
 mill+0.046
ting+0.045
 painter+0.045
ini+0.043
 office+0.042
Neg logit contributions
 dives-0.033
 drift-0.033
irm-0.033
 much-0.032
 nap-0.032
Token made
Token ID925
Feature activation+3.99e-02

Pos logit contributions
 mill+0.034
ting+0.034
 painter+0.033
ini+0.032
 office+0.031
Neg logit contributions
 drift-0.023
irm-0.023
 dives-0.023
 much-0.023
 nap-0.022
Token you
Token ID345
Feature activation-6.52e-02

Pos logit contributions
ting+0.066
 mill+0.064
 painter+0.062
ini+0.061
 office+0.059
Neg logit contributions
 dives-0.056
 drift-0.056
irm-0.055
 nap-0.054
 much-0.054
Token was
Token ID373
Feature activation+3.33e-02

Pos logit contributions
 mill+0.058
ting+0.056
cut+0.055
ini+0.053
 painter+0.051
Neg logit contributions
irm-0.076
 nap-0.075
 much-0.074
ust-0.073
 dives-0.072
Token too
Token ID1165
Feature activation-9.23e-02

Pos logit contributions
 mill+0.061
ting+0.060
 painter+0.057
cut+0.056
 paramed+0.055
Neg logit contributions
 dives-0.040
 nap-0.039
 drift-0.039
irm-0.039
ls-0.038
Token strong
Token ID1913
Feature activation+1.51e-02

Pos logit contributions
 nap+0.114
 drift+0.113
 dives+0.112
irm+0.112
 hops+0.109
Neg logit contributions
 mill-0.170
 painter-0.152
ini-0.151
ting-0.145
cut-0.145
Token.
Token ID13
Feature activation-4.03e-03

Pos logit contributions
ting+0.027
 mill+0.024
 painter+0.023
 office+0.022
cut+0.022
Neg logit contributions
 dives-0.021
 drift-0.020
irm-0.020
ust-0.020
 nap-0.020
Token Fr
Token ID1305
Feature activation-1.45e-02

Pos logit contributions
 drift+0.005
 much+0.005
 dives+0.005
 nap+0.005
irm+0.005
Neg logit contributions
ting-0.007
 mill-0.007
 painter-0.007
ini-0.006
 office-0.006
Tokenust
Token ID436
Feature activation-1.86e-01

Pos logit contributions
 drift+0.014
 nap+0.013
 dives+0.012
 hops+0.011
irm+0.011
Neg logit contributions
ini-0.034
 painter-0.032
 mill-0.032
years-0.031
rom-0.031
Tokenrated
Token ID4111
Feature activation-1.05e-02

Pos logit contributions
 nap+0.288
irm+0.270
 much+0.267
 drift+0.265
 dives+0.263
Neg logit contributions
cut-0.283
 mill-0.271
ting-0.267
ini-0.263
ter-0.256
Token,
Token ID11
Feature activation+5.25e-02

Pos logit contributions
 much+0.011
 drift+0.011
 dives+0.011
 nap+0.011
irm+0.011
Neg logit contributions
ting-0.021
 mill-0.021
 painter-0.020
ini-0.020
 office-0.019
Token he
Token ID339
Feature activation-6.12e-02

Pos logit contributions
ting+0.115
 mill+0.115
 painter+0.110
 office+0.105
ini+0.105
Neg logit contributions
irm-0.047
 drift-0.045
ust-0.045
 nap-0.044
 dives-0.044
Token started
Token ID2067
Feature activation+5.47e-02

Pos logit contributions
 much+0.092
 nap+0.091
 drift+0.089
irm+0.088
 dives+0.087
Neg logit contributions
ting-0.095
 mill-0.095
 painter-0.089
ini-0.086
cut-0.086
Token to
Token ID284
Feature activation+4.37e-02

Pos logit contributions
 mill+0.089
ting+0.088
 painter+0.084
 office+0.081
 paramed+0.080
Neg logit contributions
 much-0.077
 nap-0.076
ust-0.076
irm-0.075
 dives-0.075
Token back
Token ID736
Feature activation+3.16e-02

Pos logit contributions
 dives+0.026
 drift+0.026
 hops+0.025
irm+0.024
 dive+0.024
Neg logit contributions
ting-0.029
 mill-0.028
 office-0.027
 painter-0.027
ini-0.026
Token.
Token ID13
Feature activation-3.12e-03

Pos logit contributions
ting+0.062
 mill+0.062
 painter+0.059
ini+0.059
 paramed+0.057
Neg logit contributions
 much-0.035
 drift-0.034
 dives-0.034
 nap-0.034
irm-0.033
Token
Token ID198
Feature activation-5.26e-02

Pos logit contributions
 much+0.003
 drift+0.003
 dives+0.003
 nap+0.003
irm+0.003
Neg logit contributions
ting-0.007
 mill-0.007
ini-0.006
 painter-0.006
 paramed-0.006
Token
Token ID198
Feature activation-5.20e-02

Pos logit contributions
 much+0.051
 drift+0.050
 dives+0.050
 nap+0.050
irm+0.048
Neg logit contributions
 mill-0.109
ting-0.109
 painter-0.106
ini-0.104
 paramed-0.102
TokenFr
Token ID6732
Feature activation-6.49e-03

Pos logit contributions
 much+0.081
ust+0.073
 dives+0.070
 nap+0.070
 drift+0.069
Neg logit contributions
ting-0.089
 mill-0.086
ini-0.082
 painter-0.080
 office-0.078
Tokenust
Token ID436
Feature activation-1.85e-01

Pos logit contributions
 drift+0.003
 nap+0.003
 dives+0.003
 hops+0.002
 much+0.002
Neg logit contributions
 mill-0.017
ini-0.017
 painter-0.017
ting-0.017
rom-0.017
Tokenrated
Token ID4111
Feature activation-9.89e-03

Pos logit contributions
 nap+0.261
 drift+0.250
 much+0.248
 dives+0.245
irm+0.241
Neg logit contributions
 mill-0.301
ting-0.298
 painter-0.293
cut-0.291
ini-0.289
Token,
Token ID11
Feature activation+5.09e-02

Pos logit contributions
 much+0.010
 dives+0.010
 drift+0.009
 nap+0.009
irm+0.009
Neg logit contributions
 mill-0.020
ting-0.020
 painter-0.020
ini-0.020
 paramed-0.019
Token Jane
Token ID12091
Feature activation-2.34e-02

Pos logit contributions
 mill+0.087
ting+0.087
 painter+0.084
ini+0.081
 paramed+0.079
Neg logit contributions
 dives-0.068
 drift-0.068
 much-0.068
 nap-0.067
irm-0.066
Token went
Token ID1816
Feature activation-3.95e-02

Pos logit contributions
 much+0.029
 drift+0.029
 dives+0.028
 nap+0.028
irm+0.027
Neg logit contributions
 mill-0.043
ting-0.042
ini-0.041
 painter-0.041
 paramed-0.040
Token back
Token ID736
Feature activation+3.20e-02

Pos logit contributions
 much+0.050
 dives+0.049
 drift+0.049
 nap+0.048
irm+0.047
Neg logit contributions
 mill-0.071
ting-0.071
 painter-0.069
ini-0.068
 paramed-0.066
Token but
Token ID475
Feature activation+9.68e-03

Pos logit contributions
ting+0.095
 mill+0.094
 painter+0.091
ini+0.089
 office+0.087
Neg logit contributions
 much-0.051
 dives-0.050
 drift-0.050
 nap-0.050
irm-0.049
Token couldn
Token ID3521
Feature activation+1.09e-01

Pos logit contributions
ting+0.018
 mill+0.018
 painter+0.016
cut+0.016
 office+0.016
Neg logit contributions
 drift-0.012
 dives-0.012
irm-0.012
 nap-0.012
 much-0.011
Token't
Token ID470
Feature activation-5.82e-02

Pos logit contributions
ting+0.149
 mill+0.149
 painter+0.132
Why+0.131
rom+0.131
Neg logit contributions
irm-0.188
ust-0.179
 dives-0.179
 drift-0.173
 nap-0.173
Token.
Token ID13
Feature activation-7.64e-03

Pos logit contributions
irm+0.077
 dives+0.077
 much+0.075
ls+0.073
 drift+0.073
Neg logit contributions
ting-0.100
 mill-0.099
 painter-0.098
 office-0.090
ini-0.089
Token Fr
Token ID1305
Feature activation-1.27e-02

Pos logit contributions
 much+0.007
 dives+0.007
 drift+0.007
 nap+0.007
irm+0.007
Neg logit contributions
ting-0.016
 mill-0.016
 painter-0.016
ini-0.016
 paramed-0.015
Tokenust
Token ID436
Feature activation-1.85e-01

Pos logit contributions
 drift+0.011
 dives+0.010
 nap+0.010
 much+0.010
 hops+0.009
Neg logit contributions
ini-0.030
 mill-0.029
 painter-0.028
ting-0.028
cut-0.028
Tokenrated
Token ID4111
Feature activation-1.04e-02

Pos logit contributions
 nap+0.281
 drift+0.261
 dives+0.260
 much+0.259
irm+0.258
Neg logit contributions
cut-0.281
 mill-0.274
ting-0.273
ini-0.267
 painter-0.258
Token,
Token ID11
Feature activation+4.76e-02

Pos logit contributions
 drift+0.013
 nap+0.013
 dives+0.013
irm+0.013
 much+0.013
Neg logit contributions
ting-0.019
 mill-0.019
 painter-0.018
ini-0.017
 office-0.017
Token he
Token ID339
Feature activation-6.36e-02

Pos logit contributions
ting+0.110
 mill+0.110
 painter+0.105
 office+0.101
cut+0.100
Neg logit contributions
irm-0.037
ust-0.036
 dives-0.035
 drift-0.035
 nap-0.035
Token started
Token ID2067
Feature activation+5.37e-02

Pos logit contributions
 much+0.094
 drift+0.092
 nap+0.091
 dives+0.091
irm+0.090
Neg logit contributions
 mill-0.101
ting-0.101
 painter-0.096
ini-0.093
 office-0.091
Token to
Token ID284
Feature activation+4.11e-02

Pos logit contributions
ting+0.103
 mill+0.103
ini+0.101
 painter+0.099
cut+0.096
Neg logit contributions
 much-0.063
 drift-0.062
 nap-0.061
 dives-0.061
irm-0.058
Token or
Token ID393
Feature activation-1.80e-02

Pos logit contributions
 much+0.020
 drift+0.020
 nap+0.020
 dives+0.019
irm+0.019
Neg logit contributions
 mill-0.034
ting-0.033
 painter-0.033
ini-0.033
 paramed-0.032
Token too
Token ID1165
Feature activation-9.51e-02

Pos logit contributions
 dives+0.020
 much+0.020
 nap+0.020
 drift+0.020
irm+0.019
Neg logit contributions
ting-0.034
 mill-0.033
 painter-0.031
cut-0.031
ini-0.031
Token hard
Token ID1327
Feature activation-1.73e-02

Pos logit contributions
 dives+0.112
irm+0.112
 drift+0.107
ares+0.105
 nap+0.104
Neg logit contributions
 mill-0.172
ting-0.164
 painter-0.158
ini-0.156
cut-0.152
Token.
Token ID13
Feature activation-7.54e-03

Pos logit contributions
 dives+0.022
irm+0.022
 drift+0.021
 nap+0.021
 much+0.021
Neg logit contributions
ting-0.031
 mill-0.030
 painter-0.028
 office-0.028
ini-0.027
Token Fr
Token ID1305
Feature activation-1.73e-02

Pos logit contributions
 much+0.008
 nap+0.008
 drift+0.008
 dives+0.007
irm+0.007
Neg logit contributions
ting-0.015
 mill-0.015
 painter-0.015
ini-0.015
 paramed-0.014
Tokenust
Token ID436
Feature activation-1.85e-01

Pos logit contributions
 drift+0.021
 dives+0.020
 nap+0.020
 much+0.020
irm+0.019
Neg logit contributions
 mill-0.032
ting-0.032
ini-0.032
 painter-0.031
rom-0.030
Tokenrated
Token ID4111
Feature activation-1.06e-02

Pos logit contributions
 nap+0.285
 dives+0.278
 drift+0.277
 much+0.277
irm+0.274
Neg logit contributions
 mill-0.273
ting-0.272
ini-0.261
cut-0.260
 painter-0.258
Token,
Token ID11
Feature activation+4.71e-02

Pos logit contributions
 much+0.010
 dives+0.010
 drift+0.009
 nap+0.009
irm+0.009
Neg logit contributions
 mill-0.023
 painter-0.022
ting-0.022
ini-0.022
 paramed-0.022
Token she
Token ID673
Feature activation-5.20e-02

Pos logit contributions
 mill+0.096
ting+0.095
ini+0.094
 painter+0.092
 paramed+0.090
Neg logit contributions
 much-0.050
 drift-0.047
 nap-0.047
 dives-0.046
 hops-0.043
Token kept
Token ID4030
Feature activation+7.06e-02

Pos logit contributions
 much+0.076
 drift+0.075
 dives+0.074
 nap+0.074
irm+0.073
Neg logit contributions
ting-0.083
 mill-0.083
 painter-0.079
ini-0.078
 office-0.075
Token searching
Token ID10342
Feature activation-5.32e-02

Pos logit contributions
ting+0.118
 mill+0.114
 painter+0.108
 office+0.105
ini+0.103
Neg logit contributions
 much-0.101
 dives-0.098
 nap-0.098
 drift-0.098
ust-0.097
Token he
Token ID339
Feature activation-6.22e-02

Pos logit contributions
ting+0.017
 mill+0.017
 painter+0.016
cut+0.016
ini+0.016
Neg logit contributions
 drift-0.012
 dives-0.012
 nap-0.012
 much-0.012
irm-0.011
Token said
Token ID531
Feature activation-4.81e-02

Pos logit contributions
 drift+0.095
 dives+0.093
 nap+0.091
 much+0.091
irm+0.090
Neg logit contributions
 mill-0.097
ting-0.095
 painter-0.093
ini-0.090
cut-0.088
Token no
Token ID645
Feature activation+4.80e-03

Pos logit contributions
 much+0.047
 nap+0.040
 dives+0.039
 drift+0.039
 swim+0.037
Neg logit contributions
ini-0.105
 mill-0.105
ting-0.104
 painter-0.101
 laund-0.099
Token.
Token ID13
Feature activation-9.08e-03

Pos logit contributions
 mill+0.010
ting+0.010
 painter+0.010
ini+0.010
 paramed+0.009
Neg logit contributions
 much-0.005
 nap-0.005
 drift-0.005
 dives-0.005
irm-0.004
Token Fr
Token ID1305
Feature activation-1.50e-02

Pos logit contributions
 much+0.011
 dives+0.011
 drift+0.011
 nap+0.011
irm+0.011
Neg logit contributions
ting-0.017
 mill-0.016
 painter-0.016
ini-0.016
 paramed-0.015
Tokenust
Token ID436
Feature activation-1.85e-01

Pos logit contributions
 drift+0.014
 nap+0.013
 dives+0.012
 much+0.012
 hops+0.012
Neg logit contributions
ini-0.035
 mill-0.033
 painter-0.033
ting-0.033
rom-0.032
Tokenrated
Token ID4111
Feature activation-1.11e-02

Pos logit contributions
 nap+0.281
 drift+0.265
 much+0.254
 dives+0.253
irm+0.247
Neg logit contributions
cut-0.287
 mill-0.280
ting-0.280
ini-0.274
 painter-0.265
Token,
Token ID11
Feature activation+4.89e-02

Pos logit contributions
 much+0.012
 drift+0.012
 dives+0.012
 nap+0.012
irm+0.011
Neg logit contributions
ting-0.022
 mill-0.022
 painter-0.021
ini-0.021
 paramed-0.020
Token Bob
Token ID5811
Feature activation-2.05e-02

Pos logit contributions
ting+0.090
 mill+0.089
Why+0.083
 painter+0.080
cut+0.080
Neg logit contributions
irm-0.065
ust-0.062
 nap-0.061
 drift-0.060
 dives-0.060
Token decided
Token ID3066
Feature activation+3.45e-02

Pos logit contributions
 drift+0.029
 much+0.029
 nap+0.028
 dives+0.028
irm+0.028
Neg logit contributions
 mill-0.033
ting-0.033
 painter-0.032
 paramed-0.031
 office-0.030
Token to
Token ID284
Feature activation+4.24e-02

Pos logit contributions
 mill+0.071
ini+0.071
ting+0.071
 painter+0.069
rom+0.067
Neg logit contributions
 much-0.036
 dives-0.033
 nap-0.033
 drift-0.032
irm-0.031
Token smiled
Token ID13541
Feature activation+9.19e-03

Pos logit contributions
ting+0.168
 mill+0.168
ini+0.160
 painter+0.157
 office+0.155
Neg logit contributions
 much-0.107
 nap-0.104
 drift-0.103
 dives-0.103
irm-0.096
Token and
Token ID290
Feature activation-3.00e-02

Pos logit contributions
ting+0.014
 painter+0.013
 mill+0.013
ini+0.012
Why+0.012
Neg logit contributions
 dives-0.015
 nap-0.014
 drift-0.014
 dive-0.014
irm-0.014
Token said
Token ID531
Feature activation-4.62e-02

Pos logit contributions
 dives+0.039
irm+0.039
 much+0.039
 drift+0.038
ust+0.038
Neg logit contributions
 mill-0.052
ting-0.052
 painter-0.050
 paramed-0.048
ini-0.047
Token,
Token ID11
Feature activation+5.19e-02

Pos logit contributions
 much+0.047
 dives+0.043
 drift+0.043
 nap+0.043
irm+0.040
Neg logit contributions
 mill-0.098
ting-0.097
ini-0.094
 painter-0.094
 office-0.091
Token "
Token ID366
Feature activation+2.30e-01

Pos logit contributions
 mill+0.078
ini+0.077
ting+0.077
 painter+0.073
 office+0.071
Neg logit contributions
 much-0.083
 drift-0.081
 nap-0.077
 dives-0.077
 swim-0.074
TokenLooks
Token ID41102
Feature activation-1.85e-01

Pos logit contributions
 mill+0.321
ting+0.304
 office+0.299
 painter+0.286
 paramed+0.280
Neg logit contributions
 much-0.405
 drift-0.380
 nap-0.375
ust-0.374
 dives-0.370
Token like
Token ID588
Feature activation+4.17e-02

Pos logit contributions
 drift+0.320
ls+0.317
zag+0.306
 nap+0.300
 dive+0.300
Neg logit contributions
ting-0.209
 painter-0.198
 pavement-0.189
 floors-0.187
 artist-0.181
Token you
Token ID345
Feature activation-6.15e-02

Pos logit contributions
 mill+0.073
ting+0.073
 painter+0.070
ini+0.067
 office+0.067
Neg logit contributions
 drift-0.053
 dives-0.053
irm-0.053
 much-0.051
 nap-0.051
Token like
Token ID588
Feature activation+4.20e-02

Pos logit contributions
 much+0.069
 nap+0.066
 drift+0.065
 dives+0.065
irm+0.062
Neg logit contributions
ting-0.122
 mill-0.122
ini-0.118
 painter-0.117
 paramed-0.114
Token the
Token ID262
Feature activation+5.88e-02

Pos logit contributions
ting+0.075
 mill+0.074
ini+0.073
 painter+0.072
 paramed+0.069
Neg logit contributions
 much-0.055
 nap-0.053
 dives-0.053
 drift-0.053
irm-0.050
Token wardrobe
Token ID35593
Feature activation+4.09e-02

Pos logit contributions
ting+0.106
 mill+0.101
ini+0.096
rom+0.096
 painter+0.096
Neg logit contributions
 nap-0.080
 much-0.079
 drift-0.075
 dives-0.071
irm-0.068
Token a
Token ID257
Feature activation-7.49e-02

Pos logit contributions
 mill+0.051
ting+0.050
 painter+0.049
ini+0.047
 office+0.046
Neg logit contributions
 dives-0.035
 drift-0.035
irm-0.034
ls-0.033
 hops-0.033
Token little
Token ID1310
Feature activation+2.24e-02

Pos logit contributions
 much+0.093
 nap+0.090
 dives+0.088
 drift+0.088
 swim+0.083
Neg logit contributions
ting-0.141
 mill-0.135
ini-0.131
 painter-0.128
rom-0.128
Token crab
Token ID32202
Feature activation+2.30e-02

Pos logit contributions
ting+0.028
 mill+0.026
ini+0.025
rom+0.025
 painter+0.024
Neg logit contributions
 much-0.041
 nap-0.041
 dives-0.040
 drift-0.040
 swim-0.039
Token named
Token ID3706
Feature activation+4.03e-02

Pos logit contributions
ting+0.031
 mill+0.031
 painter+0.029
ini+0.029
 paramed+0.028
Neg logit contributions
 much-0.040
 drift-0.039
 nap-0.039
 dives-0.039
irm-0.038
Token Cr
Token ID3864
Feature activation-5.69e-02

Pos logit contributions
ting+0.056
 mill+0.055
 painter+0.053
ini+0.048
 pavement+0.048
Neg logit contributions
irm-0.067
 drift-0.066
 dives-0.066
ust-0.065
 much-0.065
Tokenust
Token ID436
Feature activation-1.85e-01

Pos logit contributions
 drift+0.051
 dives+0.049
 nap+0.046
 much+0.045
 hops+0.044
Neg logit contributions
ini-0.125
 painter-0.125
 mill-0.123
ting-0.120
rom-0.118
Tokeny
Token ID88
Feature activation+4.09e-03

Pos logit contributions
 drift+0.296
 nap+0.293
 much+0.292
 dives+0.291
irm+0.287
Neg logit contributions
 mill-0.265
ting-0.264
 painter-0.254
ini-0.249
cut-0.238
Token.
Token ID13
Feature activation-1.58e-02

Pos logit contributions
ting+0.009
 mill+0.009
 laund+0.009
ini+0.009
 paramed+0.009
Neg logit contributions
 much-0.003
 swim-0.003
 nap-0.003
 dive-0.003
ls-0.003
Token Cr
Token ID3864
Feature activation-5.48e-02

Pos logit contributions
 much+0.021
 dives+0.020
 drift+0.020
 nap+0.020
irm+0.019
Neg logit contributions
ting-0.028
 mill-0.028
 painter-0.027
ini-0.027
 paramed-0.026
Tokenust
Token ID436
Feature activation-1.84e-01

Pos logit contributions
 drift+0.039
 dives+0.035
 nap+0.034
 much+0.032
 hops+0.030
Neg logit contributions
ini-0.134
 painter-0.131
 mill-0.129
ting-0.127
rom-0.126
Tokeny
Token ID88
Feature activation+4.80e-03

Pos logit contributions
 drift+0.321
 nap+0.319
 dives+0.312
 much+0.308
irm+0.307
Neg logit contributions
 mill-0.230
ting-0.229
ini-0.224
 painter-0.222
cut-0.215
Token was
Token ID373
Feature activation+3.18e-02

Pos logit contributions
 much+0.030
 dives+0.030
 drift+0.030
 nap+0.029
irm+0.029
Neg logit contributions
 mill-0.031
ting-0.031
 painter-0.029
ini-0.029
 paramed-0.028
Token too
Token ID1165
Feature activation-9.40e-02

Pos logit contributions
 mill+0.058
ting+0.056
 office+0.055
 iron+0.055
Why+0.055
Neg logit contributions
 dives-0.039
ls-0.036
 drift-0.034
irm-0.034
 hops-0.034
Token dangerous
Token ID4923
Feature activation+5.16e-02

Pos logit contributions
 dives+0.130
 drift+0.127
 nap+0.126
irm+0.125
 hops+0.123
Neg logit contributions
 mill-0.165
 painter-0.149
ini-0.147
 iron-0.141
 paramed-0.140
Token.
Token ID13
Feature activation-5.37e-03

Pos logit contributions
ting+0.101
 mill+0.101
 painter+0.097
ini+0.096
 paramed+0.093
Neg logit contributions
 much-0.057
 drift-0.056
 dives-0.055
 nap-0.055
irm-0.053
Token Fr
Token ID1305
Feature activation-1.42e-02

Pos logit contributions
 much+0.005
 dives+0.005
 drift+0.005
 nap+0.005
irm+0.005
Neg logit contributions
ting-0.011
 mill-0.011
 painter-0.011
ini-0.011
 paramed-0.011
Tokenust
Token ID436
Feature activation-1.84e-01

Pos logit contributions
 drift+0.013
 nap+0.013
 dives+0.013
 much+0.012
 hops+0.012
Neg logit contributions
ini-0.032
 mill-0.031
 painter-0.031
ting-0.030
 office-0.030
Tokenrated
Token ID4111
Feature activation-1.03e-02

Pos logit contributions
 nap+0.296
 drift+0.281
 dives+0.279
 much+0.276
irm+0.272
Neg logit contributions
cut-0.269
 mill-0.261
ini-0.256
ting-0.252
 painter-0.249
Token,
Token ID11
Feature activation+4.87e-02

Pos logit contributions
 drift+0.011
 much+0.011
 dives+0.011
 nap+0.011
irm+0.011
Neg logit contributions
ting-0.020
 mill-0.020
 painter-0.019
ini-0.019
 office-0.018
Token Sam
Token ID3409
Feature activation+1.37e-02

Pos logit contributions
ting+0.093
 mill+0.092
 painter+0.090
 office+0.085
ini+0.085
Neg logit contributions
 drift-0.056
 dives-0.056
 nap-0.055
 much-0.055
irm-0.055
Token turned
Token ID2900
Feature activation-5.84e-02

Pos logit contributions
 mill+0.022
ting+0.022
 painter+0.022
ini+0.021
 office+0.021
Neg logit contributions
 much-0.020
 dives-0.019
 drift-0.019
 nap-0.019
irm-0.019
Token away
Token ID1497
Feature activation-9.79e-03

Pos logit contributions
 much+0.070
 drift+0.069
 dives+0.068
 nap+0.068
irm+0.066
Neg logit contributions
 mill-0.110
ting-0.109
 painter-0.106
ini-0.105
 office-0.101
Token around
Token ID1088
Feature activation+3.93e-02

Pos logit contributions
 much+0.033
 drift+0.033
 dives+0.032
 nap+0.032
irm+0.031
Neg logit contributions
 mill-0.055
ting-0.055
 painter-0.053
ini-0.053
 paramed-0.051
Token the
Token ID262
Feature activation+5.60e-02

Pos logit contributions
 paramed+0.088
 mill+0.085
ini+0.083
 laund+0.081
 painter+0.080
Neg logit contributions
 much-0.041
 nap-0.035
 drift-0.034
 time-0.033
 dives-0.032
Token door
Token ID3420
Feature activation+2.53e-03

Pos logit contributions
ting+0.102
 paramed+0.100
ican+0.094
years+0.093
rom+0.092
Neg logit contributions
 nap-0.090
 much-0.086
 time-0.084
 long-0.077
 close-0.074
Token.
Token ID13
Feature activation-3.01e-03

Pos logit contributions
 mill+0.005
ting+0.005
 painter+0.005
ini+0.005
 paramed+0.005
Neg logit contributions
 much-0.003
 drift-0.003
 nap-0.003
 dives-0.003
irm-0.002
Token Fr
Token ID1305
Feature activation-1.48e-02

Pos logit contributions
 much+0.003
 drift+0.003
 dives+0.003
 nap+0.003
irm+0.003
Neg logit contributions
ting-0.006
 mill-0.006
 painter-0.006
ini-0.006
 office-0.006
Tokenust
Token ID436
Feature activation-1.84e-01

Pos logit contributions
 drift+0.009
 nap+0.009
 dives+0.008
 hops+0.008
 much+0.008
Neg logit contributions
ini-0.038
 painter-0.037
 mill-0.037
rom-0.036
 office-0.035
Tokenrated
Token ID4111
Feature activation-1.21e-02

Pos logit contributions
 nap+0.266
 much+0.251
 drift+0.250
 dives+0.247
irm+0.243
Neg logit contributions
 mill-0.293
ting-0.291
cut-0.289
ini-0.287
 painter-0.286
Token,
Token ID11
Feature activation+4.86e-02

Pos logit contributions
 much+0.013
 dives+0.012
 drift+0.012
 nap+0.012
irm+0.012
Neg logit contributions
 mill-0.025
ting-0.024
 painter-0.024
ini-0.024
 paramed-0.023
Token Jack
Token ID3619
Feature activation-6.31e-03

Pos logit contributions
ting+0.090
 mill+0.089
 painter+0.087
Why+0.081
 office+0.080
Neg logit contributions
 drift-0.059
 nap-0.058
irm-0.058
 dives-0.057
ust-0.056
Token put
Token ID1234
Feature activation+3.54e-02

Pos logit contributions
 much+0.009
 dives+0.009
 drift+0.009
 nap+0.008
irm+0.008
Neg logit contributions
 mill-0.011
ting-0.011
ini-0.010
 painter-0.010
 paramed-0.010
Token the
Token ID262
Feature activation+5.64e-02

Pos logit contributions
 mill+0.065
ting+0.064
 painter+0.063
ini+0.063
 paramed+0.061
Neg logit contributions
 much-0.044
 drift-0.042
 nap-0.042
 dives-0.042
irm-0.041
Token answer
Token ID3280
Feature activation-2.62e-02

Pos logit contributions
 much+0.020
 drift+0.019
ls+0.019
ust+0.019
 dives+0.019
Neg logit contributions
 mill-0.034
 painter-0.033
 office-0.032
ting-0.031
 paramed-0.031
Token for
Token ID329
Feature activation+3.82e-02

Pos logit contributions
 much+0.026
 drift+0.024
 nap+0.024
 dives+0.024
irm+0.022
Neg logit contributions
 mill-0.055
ting-0.054
ini-0.054
 painter-0.053
 paramed-0.052
Token him
Token ID683
Feature activation-2.01e-02

Pos logit contributions
 mill+0.061
ting+0.061
ini+0.060
 painter+0.060
 paramed+0.058
Neg logit contributions
 much-0.057
 drift-0.055
 nap-0.055
 dives-0.054
irm-0.052
Token.
Token ID13
Feature activation-3.01e-03

Pos logit contributions
 much+0.022
 drift+0.021
 dives+0.021
 nap+0.021
irm+0.020
Neg logit contributions
 mill-0.040
ting-0.040
 painter-0.038
ini-0.038
 paramed-0.037
Token Fr
Token ID1305
Feature activation-1.40e-02

Pos logit contributions
 drift+0.004
 nap+0.003
 dives+0.003
 much+0.003
irm+0.003
Neg logit contributions
 mill-0.006
ting-0.006
 painter-0.005
 office-0.005
 paramed-0.005
Tokenust
Token ID436
Feature activation-1.84e-01

Pos logit contributions
 drift+0.013
 nap+0.013
 dives+0.013
 much+0.012
 hops+0.012
Neg logit contributions
 mill-0.030
ini-0.030
 painter-0.030
ting-0.030
rom-0.029
Tokenrated
Token ID4111
Feature activation-8.49e-03

Pos logit contributions
 nap+0.287
 drift+0.266
 much+0.264
 dives+0.263
irm+0.260
Neg logit contributions
cut-0.279
 mill-0.278
ting-0.273
ini-0.269
 painter-0.265
Token,
Token ID11
Feature activation+5.05e-02

Pos logit contributions
 much+0.009
 drift+0.009
 dives+0.009
 nap+0.009
irm+0.008
Neg logit contributions
 mill-0.017
ting-0.017
 painter-0.016
ini-0.016
 paramed-0.016
Token he
Token ID339
Feature activation-6.17e-02

Pos logit contributions
ting+0.108
 mill+0.107
 painter+0.104
 office+0.100
 paramed+0.099
Neg logit contributions
 drift-0.047
 dives-0.047
 nap-0.046
irm-0.046
 much-0.044
Token decided
Token ID3066
Feature activation+3.58e-02

Pos logit contributions
 much+0.095
 nap+0.095
 drift+0.092
irm+0.092
 dives+0.091
Neg logit contributions
 mill-0.093
ting-0.092
 painter-0.088
cut-0.083
 office-0.083
Token to
Token ID284
Feature activation+4.37e-02

Pos logit contributions
 mill+0.070
ting+0.070
ini+0.068
 painter+0.068
 paramed+0.065
Neg logit contributions
 much-0.040
 dives-0.039
 drift-0.039
 nap-0.038
irm-0.037
Tokenge
Token ID469
Feature activation-4.18e-02

Pos logit contributions
ting+0.007
 mill+0.007
ini+0.007
 painter+0.007
rom+0.007
Neg logit contributions
 dives-0.004
 drift-0.004
 much-0.004
 nap-0.004
irm-0.004
Token.
Token ID13
Feature activation-7.17e-03

Pos logit contributions
 dives+0.049
 drift+0.048
 much+0.048
 nap+0.047
irm+0.046
Neg logit contributions
ting-0.079
 mill-0.078
 painter-0.075
ini-0.074
 office-0.072
Token
Token ID198
Feature activation-5.79e-02

Pos logit contributions
 much+0.006
 nap+0.005
 drift+0.005
 dives+0.004
irm+0.004
Neg logit contributions
ting-0.017
 mill-0.017
ini-0.017
 painter-0.016
cut-0.016
Token
Token ID198
Feature activation-5.77e-02

Pos logit contributions
 dives+0.063
 drift+0.063
 nap+0.061
irm+0.061
 much+0.060
Neg logit contributions
 mill-0.112
ting-0.111
 painter-0.108
ini-0.107
 office-0.104
TokenFr
Token ID6732
Feature activation-6.33e-03

Pos logit contributions
 much+0.083
ust+0.070
 nap+0.068
 drift+0.065
irm+0.064
Neg logit contributions
ting-0.110
 mill-0.107
 painter-0.100
ini-0.100
keeping-0.099
Tokenust
Token ID436
Feature activation-1.84e-01

Pos logit contributions
 drift+0.004
 dives+0.003
 nap+0.003
 much+0.003
 hops+0.003
Neg logit contributions
 mill-0.016
ini-0.016
 painter-0.016
ting-0.015
rom-0.015
Tokenrated
Token ID4111
Feature activation-1.02e-02

Pos logit contributions
 nap+0.262
 drift+0.248
 much+0.248
 dives+0.246
irm+0.241
Neg logit contributions
 mill-0.295
ting-0.292
cut-0.288
ini-0.284
 painter-0.282
Token,
Token ID11
Feature activation+4.90e-02

Pos logit contributions
 much+0.011
 drift+0.011
 dives+0.010
 nap+0.010
irm+0.010
Neg logit contributions
 mill-0.021
ting-0.021
 painter-0.020
ini-0.020
 paramed-0.019
Token the
Token ID262
Feature activation+5.83e-02

Pos logit contributions
ting+0.093
 mill+0.093
ini+0.090
 painter+0.089
 paramed+0.086
Neg logit contributions
 much-0.058
 drift-0.056
 nap-0.056
 dives-0.055
irm-0.053
Token little
Token ID1310
Feature activation+2.14e-02

Pos logit contributions
ting+0.076
rom+0.074
keeping+0.073
stick+0.072
 paramed+0.071
Neg logit contributions
 nap-0.107
 much-0.106
 dives-0.101
 dive-0.100
 swim-0.100
Token ant
Token ID1885
Feature activation+7.88e-02

Pos logit contributions
rom+0.027
ting+0.027
 paramed+0.026
keeping+0.026
cut+0.026
Neg logit contributions
 nap-0.040
 dives-0.038
 hops-0.038
 dive-0.037
 much-0.037
Token.
Token ID13
Feature activation-1.47e-02

Pos logit contributions
ting+0.028
 mill+0.028
 painter+0.027
ini+0.026
cut+0.026
Neg logit contributions
 drift-0.018
 much-0.018
 nap-0.018
irm-0.018
 dives-0.018
Token One
Token ID1881
Feature activation-4.55e-02

Pos logit contributions
 much+0.012
 drift+0.012
 dives+0.012
 nap+0.012
irm+0.012
Neg logit contributions
 mill-0.032
ting-0.032
 painter-0.031
ini-0.031
 paramed-0.030
Token day
Token ID1110
Feature activation-2.95e-02

Pos logit contributions
irm+0.080
 drift+0.079
ares+0.079
ls+0.078
ust+0.077
Neg logit contributions
 mill-0.062
 painter-0.057
ini-0.056
ting-0.050
 office-0.049
Token,
Token ID11
Feature activation+4.56e-02

Pos logit contributions
 drift+0.042
irm+0.041
ust+0.041
 nap+0.039
 dives+0.039
Neg logit contributions
ting-0.049
 mill-0.048
 painter-0.046
 paramed-0.043
years-0.042
Token Cr
Token ID3864
Feature activation-5.41e-02

Pos logit contributions
ting+0.088
 mill+0.088
 painter+0.085
ini+0.083
 paramed+0.081
Neg logit contributions
 much-0.051
 drift-0.051
 dives-0.051
 nap-0.050
irm-0.049
Tokenust
Token ID436
Feature activation-1.84e-01

Pos logit contributions
 drift+0.042
 dives+0.037
 nap+0.035
 much+0.033
 dive+0.033
Neg logit contributions
ini-0.130
 painter-0.127
 mill-0.124
ting-0.122
rom-0.122
Tokeny
Token ID88
Feature activation+4.59e-03

Pos logit contributions
 nap+0.344
 drift+0.339
 much+0.333
irm+0.328
 dives+0.323
Neg logit contributions
ting-0.205
 mill-0.204
ini-0.195
 painter-0.192
cut-0.184
Token found
Token ID1043
Feature activation+7.62e-02

Pos logit contributions
ting+0.007
 mill+0.007
 painter+0.007
ini+0.006
 paramed+0.006
Neg logit contributions
 much-0.007
 dives-0.007
 nap-0.007
 drift-0.007
irm-0.007
Token a
Token ID257
Feature activation-7.38e-02

Pos logit contributions
 mill+0.178
ting+0.176
ini+0.176
 painter+0.174
rom+0.170
Neg logit contributions
 much-0.065
 dives-0.052
 nap-0.052
irm-0.047
 swim-0.047
Token shiny
Token ID22441
Feature activation+1.63e-03

Pos logit contributions
 much+0.085
 drift+0.084
 dives+0.084
 nap+0.082
irm+0.082
Neg logit contributions
 mill-0.140
ting-0.138
 painter-0.135
ini-0.132
 office-0.128
Token shell
Token ID7582
Feature activation-3.11e-02

Pos logit contributions
ting+0.003
 paramed+0.003
years+0.003
 mill+0.003
ini+0.003
Neg logit contributions
 nap-0.002
 dives-0.002
 much-0.002
 dive-0.002
 swim-0.002
Token Cr
Token ID3864
Feature activation-5.69e-02

Pos logit contributions
ting+0.056
 mill+0.055
 painter+0.053
ini+0.048
 pavement+0.048
Neg logit contributions
irm-0.067
 drift-0.066
 dives-0.066
ust-0.065
 much-0.065
Tokenust
Token ID436
Feature activation-1.85e-01

Pos logit contributions
 drift+0.051
 dives+0.049
 nap+0.046
 much+0.045
 hops+0.044
Neg logit contributions
ini-0.125
 painter-0.125
 mill-0.123
ting-0.120
rom-0.118
Tokeny
Token ID88
Feature activation+4.09e-03

Pos logit contributions
 drift+0.296
 nap+0.293
 much+0.292
 dives+0.291
irm+0.287
Neg logit contributions
 mill-0.265
ting-0.264
 painter-0.254
ini-0.249
cut-0.238
Token.
Token ID13
Feature activation-1.58e-02

Pos logit contributions
ting+0.009
 mill+0.009
 laund+0.009
ini+0.009
 paramed+0.009
Neg logit contributions
 much-0.003
 swim-0.003
 nap-0.003
 dive-0.003
ls-0.003
Token Cr
Token ID3864
Feature activation-5.48e-02

Pos logit contributions
 much+0.021
 dives+0.020
 drift+0.020
 nap+0.020
irm+0.019
Neg logit contributions
ting-0.028
 mill-0.028
 painter-0.027
ini-0.027
 paramed-0.026
Tokenust
Token ID436
Feature activation-1.84e-01

Pos logit contributions
 drift+0.039
 dives+0.035
 nap+0.034
 much+0.032
 hops+0.030
Neg logit contributions
ini-0.134
 painter-0.131
 mill-0.129
ting-0.127
rom-0.126
Tokeny
Token ID88
Feature activation+4.80e-03

Pos logit contributions
 drift+0.321
 nap+0.319
 dives+0.312
 much+0.308
irm+0.307
Neg logit contributions
 mill-0.230
ting-0.229
ini-0.224
 painter-0.222
cut-0.215
Token lived
Token ID5615
Feature activation+6.29e-02

Pos logit contributions
ting+0.008
 mill+0.007
 painter+0.007
ini+0.007
 paramed+0.007
Neg logit contributions
 much-0.007
 nap-0.007
 dives-0.007
 drift-0.007
 hops-0.007
Token in
Token ID287
Feature activation-4.82e-02

Pos logit contributions
ting+0.103
 mill+0.101
 painter+0.100
ini+0.095
years+0.094
Neg logit contributions
 drift-0.090
 dives-0.086
 nap-0.085
irm-0.083
 much-0.083
Token a
Token ID257
Feature activation-7.47e-02

Pos logit contributions
 drift+0.057
 dives+0.057
irm+0.056
ust+0.055
 much+0.055
Neg logit contributions
 mill-0.090
ting-0.089
 painter-0.086
ini-0.083
 office-0.083
Token big
Token ID1263
Feature activation+1.37e-02

Pos logit contributions
 much+0.073
 nap+0.070
 dives+0.067
 dive+0.063
 swim+0.062
Neg logit contributions
ting-0.171
cles-0.158
rom-0.157
ini-0.157
 paramed-0.157
Token stick
Token ID4859
Feature activation+1.55e-02

Pos logit contributions
irm+0.100
ls+0.096
 dives+0.093
 hops+0.093
oll+0.089
Neg logit contributions
 iron-0.078
 painter-0.076
 mill-0.073
cut-0.072
ting-0.072
Token.
Token ID13
Feature activation-5.03e-03

Pos logit contributions
ting+0.028
 mill+0.026
 painter+0.025
cut+0.024
ini+0.024
Neg logit contributions
 dives-0.019
 drift-0.019
 nap-0.019
irm-0.019
 hops-0.018
Token
Token ID198
Feature activation-5.70e-02

Pos logit contributions
 drift+0.005
 dives+0.005
 much+0.005
 nap+0.005
irm+0.005
Neg logit contributions
 mill-0.010
ting-0.010
 painter-0.010
ini-0.009
 office-0.009
Token
Token ID198
Feature activation-5.66e-02

Pos logit contributions
 drift+0.064
 dives+0.063
 nap+0.062
irm+0.061
 much+0.060
Neg logit contributions
 mill-0.109
ting-0.108
 painter-0.105
ini-0.104
 office-0.102
TokenFr
Token ID6732
Feature activation-5.96e-03

Pos logit contributions
 much+0.082
 nap+0.075
 dives+0.075
 drift+0.074
ust+0.073
Neg logit contributions
ting-0.098
 mill-0.097
 painter-0.093
ini-0.091
 paramed-0.088
Tokenust
Token ID436
Feature activation-1.83e-01

Pos logit contributions
 drift+0.004
 nap+0.004
 dives+0.004
 hops+0.003
 much+0.003
Neg logit contributions
 mill-0.015
ini-0.015
 painter-0.014
cut-0.014
ting-0.014
Tokenrated
Token ID4111
Feature activation-8.03e-03

Pos logit contributions
 nap+0.269
 drift+0.258
 much+0.253
 dives+0.252
irm+0.250
Neg logit contributions
 mill-0.288
ting-0.285
cut-0.282
 painter-0.280
ini-0.278
Token,
Token ID11
Feature activation+4.84e-02

Pos logit contributions
 much+0.009
 drift+0.008
 dives+0.008
 nap+0.008
irm+0.008
Neg logit contributions
ting-0.016
 mill-0.016
 painter-0.015
ini-0.015
 paramed-0.015
Token the
Token ID262
Feature activation+5.77e-02

Pos logit contributions
ting+0.087
 mill+0.087
 painter+0.083
 office+0.078
cut+0.078
Neg logit contributions
irm-0.062
 drift-0.061
 dives-0.061
ust-0.059
 nap-0.059
Token lion
Token ID18744
Feature activation+2.28e-02

Pos logit contributions
ting+0.088
rom+0.086
 mill+0.085
ini+0.085
 laund+0.084
Neg logit contributions
 much-0.092
 nap-0.092
 dives-0.089
 swim-0.086
 drift-0.086
Token went
Token ID1816
Feature activation-4.00e-02

Pos logit contributions
ting+0.036
 mill+0.036
 painter+0.035
ini+0.033
cut+0.033
Neg logit contributions
 much-0.034
 drift-0.033
 dives-0.033
 nap-0.033
irm-0.033
Token table
Token ID3084
Feature activation+1.99e-02

Pos logit contributions
years+0.083
ting+0.078
ini+0.077
 paramed+0.076
 mill+0.074
Neg logit contributions
 nap-0.103
 drift-0.096
 much-0.095
 swim-0.091
 time-0.089
Token.
Token ID13
Feature activation-2.16e-03

Pos logit contributions
ting+0.043
 mill+0.042
ini+0.041
 painter+0.041
 paramed+0.040
Neg logit contributions
 much-0.019
 nap-0.018
 drift-0.018
 dives-0.017
irm-0.017
Token
Token ID198
Feature activation-5.15e-02

Pos logit contributions
 much+0.002
 dives+0.002
 nap+0.002
 drift+0.002
irm+0.002
Neg logit contributions
ting-0.005
 mill-0.005
 painter-0.005
ini-0.005
cut-0.005
Token
Token ID198
Feature activation-5.15e-02

Pos logit contributions
 much+0.044
 nap+0.038
 dives+0.037
 drift+0.036
irm+0.034
Neg logit contributions
ting-0.122
 mill-0.119
 painter-0.116
ini-0.115
 paramed-0.111
Token"
Token ID1
Feature activation+2.25e-01

Pos logit contributions
 much+0.073
 nap+0.064
 dives+0.063
ls+0.061
 drift+0.060
Neg logit contributions
ting-0.096
 mill-0.093
ini-0.089
 painter-0.088
stick-0.086
TokenLooks
Token ID41102
Feature activation-1.82e-01

Pos logit contributions
 mill+0.261
ting+0.251
ini+0.249
 office+0.236
rom+0.229
Neg logit contributions
 nap-0.433
 much-0.420
 dives-0.415
 drift-0.414
irm-0.399
Token like
Token ID588
Feature activation+4.30e-02

Pos logit contributions
 drift+0.267
 nap+0.255
ust+0.250
 dive+0.250
 dives+0.249
Neg logit contributions
ting-0.278
 painter-0.266
 mill-0.264
 floors-0.252
 pavement-0.248
Token mom
Token ID1995
Feature activation+1.01e-01

Pos logit contributions
ting+0.080
ini+0.078
 mill+0.078
 painter+0.075
cut+0.075
Neg logit contributions
 much-0.056
 nap-0.054
 dives-0.052
 drift-0.051
 dive-0.049
Token has
Token ID468
Feature activation+3.99e-02

Pos logit contributions
ting+0.194
 mill+0.190
ini+0.183
 laund+0.180
 painter+0.179
Neg logit contributions
 much-0.121
 nap-0.112
 dives-0.108
 drift-0.108
ls-0.104
Token some
Token ID617
Feature activation-6.47e-02

Pos logit contributions
ting+0.077
ini+0.075
 mill+0.074
 painter+0.072
cut+0.072
Neg logit contributions
 much-0.051
 nap-0.048
 dives-0.046
 drift-0.045
irm-0.042
Token mail
Token ID6920
Feature activation-8.96e-03

Pos logit contributions
 much+0.085
 nap+0.083
 dives+0.079
 drift+0.079
ls+0.074
Neg logit contributions
ting-0.120
 mill-0.115
ini-0.113
 painter-0.112
rom-0.110
Token something
Token ID1223
Feature activation-6.21e-02

Pos logit contributions
ting+0.109
 mill+0.108
 painter+0.103
 office+0.100
ini+0.099
Neg logit contributions
 drift-0.089
 dives-0.088
 nap-0.087
irm-0.087
 much-0.086
Token just
Token ID655
Feature activation-1.15e-01

Pos logit contributions
 drift+0.082
irm+0.081
 dives+0.078
 nap+0.077
 hops+0.076
Neg logit contributions
 mill-0.107
ting-0.105
 painter-0.102
 office-0.100
ini-0.097
Token because
Token ID780
Feature activation-3.50e-02

Pos logit contributions
 much+0.139
 drift+0.138
 dives+0.138
 nap+0.137
irm+0.135
Neg logit contributions
ting-0.211
 mill-0.207
 painter-0.199
 office-0.192
ini-0.192
Token someone
Token ID2130
Feature activation-3.74e-02

Pos logit contributions
 much+0.040
 drift+0.039
 dives+0.039
 nap+0.039
irm+0.038
Neg logit contributions
 mill-0.067
ting-0.067
 painter-0.064
ini-0.063
 office-0.062
Token d
Token ID288
Feature activation-4.83e-02

Pos logit contributions
 drift+0.050
irm+0.049
 much+0.048
 nap+0.048
 dives+0.048
Neg logit contributions
ting-0.064
 mill-0.064
 painter-0.062
 office-0.059
ini-0.059
Tokenares
Token ID3565
Feature activation-1.82e-01

Pos logit contributions
 much+0.063
 drift+0.062
 dives+0.062
 nap+0.062
irm+0.060
Neg logit contributions
ting-0.085
 mill-0.085
 painter-0.081
ini-0.080
 paramed-0.078
Token you
Token ID345
Feature activation-6.37e-02

Pos logit contributions
 much+0.212
 drift+0.212
 dives+0.210
 nap+0.209
irm+0.204
Neg logit contributions
ting-0.340
 mill-0.337
 painter-0.325
ini-0.319
 office-0.311
Token.
Token ID13
Feature activation-4.66e-03

Pos logit contributions
 much+0.073
 dives+0.071
 nap+0.071
 drift+0.070
irm+0.067
Neg logit contributions
 mill-0.123
ting-0.122
 painter-0.119
ini-0.118
 paramed-0.115
Token Always
Token ID16622
Feature activation-7.37e-03

Pos logit contributions
 much+0.006
 drift+0.006
 dives+0.006
 nap+0.006
irm+0.006
Neg logit contributions
 mill-0.008
ting-0.008
 painter-0.007
ini-0.007
 paramed-0.007
Token remember
Token ID3505
Feature activation-8.99e-03

Pos logit contributions
ls+0.008
 much+0.008
 dives+0.008
 drift+0.008
irm+0.008
Neg logit contributions
 mill-0.014
ting-0.014
 painter-0.013
 office-0.013
ini-0.013
Token to
Token ID284
Feature activation+4.30e-02

Pos logit contributions
 much+0.011
 nap+0.010
 dives+0.010
 drift+0.010
 dive+0.009
Neg logit contributions
ini-0.017
ting-0.017
 mill-0.017
 painter-0.017
 paramed-0.017
Token smiled
Token ID13541
Feature activation+8.70e-03

Pos logit contributions
 much+0.012
 drift+0.012
 dives+0.012
 nap+0.012
irm+0.012
Neg logit contributions
ting-0.016
 mill-0.016
 painter-0.015
ini-0.015
 paramed-0.015
Token,
Token ID11
Feature activation+4.99e-02

Pos logit contributions
ting+0.014
 mill+0.013
 painter+0.013
ini+0.012
rom+0.012
Neg logit contributions
 dives-0.013
 nap-0.013
 drift-0.013
 much-0.012
irm-0.012
Token and
Token ID290
Feature activation-3.00e-02

Pos logit contributions
 mill+0.080
ting+0.080
 painter+0.077
ini+0.076
 paramed+0.073
Neg logit contributions
 much-0.073
 drift-0.071
 dives-0.071
 nap-0.071
irm-0.069
Token replied
Token ID8712
Feature activation-3.09e-02

Pos logit contributions
 nap+0.037
 much+0.036
 drift+0.036
 dives+0.036
irm+0.036
Neg logit contributions
ting-0.056
 mill-0.055
 painter-0.053
 paramed-0.050
cut-0.050
Token,
Token ID11
Feature activation+4.99e-02

Pos logit contributions
 much+0.033
 drift+0.033
 dives+0.033
 nap+0.033
irm+0.032
Neg logit contributions
ting-0.061
 mill-0.061
 painter-0.059
ini-0.057
 paramed-0.056
Token "
Token ID366
Feature activation+2.27e-01

Pos logit contributions
 mill+0.069
ting+0.068
ini+0.068
 painter+0.065
 office+0.062
Neg logit contributions
 much-0.085
 drift-0.083
 dives-0.081
 nap-0.080
irm-0.077
TokenI
Token ID40
Feature activation+1.69e-02

Pos logit contributions
 mill+0.280
 office+0.260
ting+0.255
 painter+0.247
 floors+0.243
Neg logit contributions
 much-0.425
 drift-0.417
ust-0.411
 dives-0.405
irm-0.402
Token will
Token ID481
Feature activation+8.23e-02

Pos logit contributions
 mill+0.026
 painter+0.024
ini+0.023
 paramed+0.023
cut+0.023
Neg logit contributions
ls-0.025
irm-0.025
 much-0.024
 drift-0.024
 dives-0.024
Token!
Token ID0
Feature activation+1.12e-02

Pos logit contributions
 mill+0.154
ting+0.154
 painter+0.148
ini+0.144
 office+0.142
Neg logit contributions
 much-0.095
 dives-0.095
irm-0.094
 drift-0.091
 nap-0.091
Token I
Token ID314
Feature activation+4.07e-02

Pos logit contributions
ting+0.019
 mill+0.019
 painter+0.018
ini+0.018
 paramed+0.017
Neg logit contributions
 much-0.016
 drift-0.015
 dives-0.015
 nap-0.015
irm-0.015
Token'll
Token ID1183
Feature activation+8.60e-02

Pos logit contributions
 mill+0.067
ting+0.063
 painter+0.063
ini+0.063
cut+0.062
Neg logit contributions
ls-0.053
 much-0.053
irm-0.053
 dives-0.052
 drift-0.051
Token His
Token ID2399
Feature activation-1.42e-02

Pos logit contributions
 drift+0.008
 nap+0.008
 much+0.008
 dives+0.008
irm+0.008
Neg logit contributions
 mill-0.009
ting-0.009
 painter-0.009
ini-0.008
 paramed-0.008
Token mom
Token ID1995
Feature activation+1.02e-01

Pos logit contributions
 much+0.020
 nap+0.019
 drift+0.019
 dives+0.019
irm+0.019
Neg logit contributions
ting-0.024
 mill-0.024
 painter-0.023
ini-0.023
 paramed-0.022
Tokenmy
Token ID1820
Feature activation+2.86e-02

Pos logit contributions
 mill+0.156
ting+0.148
 pavement+0.143
 laund+0.141
ini+0.141
Neg logit contributions
 much-0.158
 drift-0.158
 nap-0.154
 swim-0.147
 dives-0.147
Token said
Token ID531
Feature activation-4.71e-02

Pos logit contributions
 mill+0.053
ting+0.053
ini+0.051
 painter+0.051
 office+0.049
Neg logit contributions
 drift-0.034
 much-0.034
 nap-0.033
 dives-0.033
irm-0.031
Token,
Token ID11
Feature activation+5.08e-02

Pos logit contributions
 much+0.049
 drift+0.047
 dives+0.047
 nap+0.046
irm+0.044
Neg logit contributions
 mill-0.096
ting-0.096
 painter-0.093
ini-0.093
 office-0.089
Token "
Token ID366
Feature activation+2.29e-01

Pos logit contributions
ini+0.080
 mill+0.078
 painter+0.075
ting+0.075
 office+0.071
Neg logit contributions
 much-0.080
 drift-0.076
 dives-0.072
 nap-0.072
 swim-0.071
TokenIt
Token ID1026
Feature activation-1.63e-02

Pos logit contributions
 mill+0.286
ting+0.272
 office+0.268
 painter+0.264
ini+0.253
Neg logit contributions
 much-0.419
 drift-0.409
 dives-0.408
irm-0.403
 nap-0.401
Token's
Token ID338
Feature activation-2.78e-02

Pos logit contributions
 nap+0.023
irm+0.023
 much+0.022
 drift+0.022
inn+0.021
Neg logit contributions
ting-0.026
 mill-0.026
ini-0.025
 painter-0.025
cut-0.025
Token time
Token ID640
Feature activation-1.01e-01

Pos logit contributions
 dives+0.025
ls+0.024
 drift+0.023
 hops+0.022
 nap+0.022
Neg logit contributions
 mill-0.061
ting-0.060
 office-0.058
 painter-0.057
Why-0.057
Token for
Token ID329
Feature activation+3.76e-02

Pos logit contributions
 much+0.118
 drift+0.118
 dives+0.117
 nap+0.115
irm+0.115
Neg logit contributions
 mill-0.186
ting-0.185
 painter-0.178
ini-0.173
 office-0.172
Token a
Token ID257
Feature activation-7.17e-02

Pos logit contributions
ting+0.062
 mill+0.062
 painter+0.058
 office+0.057
 paramed+0.055
Neg logit contributions
ust-0.051
irm-0.051
 dives-0.050
 drift-0.050
 much-0.050
Token
Token ID220
Feature activation-9.47e-03

Pos logit contributions
ting+0.129
 mill+0.121
 painter+0.115
 paramed+0.111
rom+0.110
Neg logit contributions
ust-0.095
 dives-0.094
 nap-0.093
irm-0.091
 drift-0.091
Token
Token ID198
Feature activation-5.03e-02

Pos logit contributions
 nap+0.012
 drift+0.012
 dives+0.012
irm+0.012
 much+0.012
Neg logit contributions
 mill-0.016
ting-0.015
rom-0.015
 painter-0.015
 office-0.015
TokenDad
Token ID46270
Feature activation+6.86e-02

Pos logit contributions
 much+0.082
 drift+0.080
 dives+0.080
 nap+0.080
irm+0.078
Neg logit contributions
ting-0.074
 mill-0.073
 painter-0.070
ini-0.068
 office-0.066
Token said
Token ID531
Feature activation-4.58e-02

Pos logit contributions
 mill+0.121
ting+0.121
 painter+0.117
ini+0.115
 paramed+0.112
Neg logit contributions
 much-0.088
 drift-0.086
 dives-0.086
 nap-0.085
irm-0.084
Token,
Token ID11
Feature activation+5.19e-02

Pos logit contributions
 much+0.049
 drift+0.049
 dives+0.049
 nap+0.049
irm+0.048
Neg logit contributions
ting-0.090
 mill-0.090
 painter-0.087
ini-0.085
 paramed-0.083
Token "
Token ID366
Feature activation+2.30e-01

Pos logit contributions
 mill+0.076
ini+0.075
ting+0.075
 painter+0.071
 office+0.070
Neg logit contributions
 much-0.084
 drift-0.082
 dives-0.080
 nap-0.079
 swim-0.077
TokenYes
Token ID5297
Feature activation-9.28e-02

Pos logit contributions
 mill+0.300
ting+0.279
 office+0.278
 paramed+0.267
 painter+0.261
Neg logit contributions
 much-0.424
 drift-0.401
ust-0.399
 dives-0.396
 nap-0.391
Token,
Token ID11
Feature activation+5.27e-02

Pos logit contributions
 drift+0.106
 much+0.106
 dives+0.106
 nap+0.104
irm+0.103
Neg logit contributions
ting-0.177
 mill-0.176
 painter-0.170
ini-0.166
 paramed-0.162
Token we
Token ID356
Feature activation-2.19e-02

Pos logit contributions
ting+0.103
 mill+0.103
 painter+0.099
ini+0.098
 paramed+0.095
Neg logit contributions
 much-0.058
 drift-0.057
 dives-0.057
 nap-0.056
irm-0.054
Token did
Token ID750
Feature activation+3.23e-02

Pos logit contributions
 much+0.023
 nap+0.022
 drift+0.022
 dives+0.021
 dive+0.020
Neg logit contributions
ting-0.046
 mill-0.045
ini-0.044
 paramed-0.043
 painter-0.043
Token!"
Token ID2474
Feature activation+7.16e-02

Pos logit contributions
 mill+0.055
ting+0.055
 painter+0.053
ini+0.051
 office+0.051
Neg logit contributions
 drift-0.042
 dives-0.042
irm-0.042
 nap-0.041
 much-0.041
Token?"
Token ID1701
Feature activation+6.65e-02

Pos logit contributions
ting+0.031
 mill+0.030
 painter+0.029
ini+0.029
 paramed+0.028
Neg logit contributions
 much-0.024
 nap-0.024
 dives-0.023
 drift-0.023
irm-0.023
Token Her
Token ID2332
Feature activation-5.31e-03

Pos logit contributions
ting+0.123
 mill+0.123
 painter+0.118
ini+0.117
 paramed+0.113
Neg logit contributions
 much-0.081
 drift-0.080
 dives-0.079
 nap-0.079
irm-0.077
Token mom
Token ID1995
Feature activation+1.01e-01

Pos logit contributions
 much+0.007
 nap+0.007
 drift+0.006
 dives+0.006
irm+0.006
Neg logit contributions
ting-0.010
 mill-0.010
ini-0.009
 painter-0.009
cut-0.009
Token said
Token ID531
Feature activation-5.01e-02

Pos logit contributions
 mill+0.189
ting+0.189
 painter+0.183
ini+0.180
 paramed+0.175
Neg logit contributions
 much-0.121
 dives-0.119
 drift-0.118
 nap-0.117
irm-0.115
Token,
Token ID11
Feature activation+4.72e-02

Pos logit contributions
 much+0.051
 drift+0.048
 nap+0.048
 dives+0.048
irm+0.046
Neg logit contributions
 mill-0.104
ting-0.104
 painter-0.100
ini-0.100
 office-0.096
Token "
Token ID366
Feature activation+2.29e-01

Pos logit contributions
ini+0.079
 mill+0.077
keeping+0.073
ting+0.072
 painter+0.072
Neg logit contributions
 much-0.071
 drift-0.067
 nap-0.063
 they-0.062
 dives-0.062
TokenYes
Token ID5297
Feature activation-9.35e-02

Pos logit contributions
 mill+0.374
ting+0.371
 painter+0.354
ini+0.349
 office+0.347
Neg logit contributions
 much-0.341
 nap-0.320
 drift-0.318
 dives-0.314
irm-0.314
Token,
Token ID11
Feature activation+4.89e-02

Pos logit contributions
 dives+0.104
 much+0.103
 drift+0.103
 nap+0.101
irm+0.100
Neg logit contributions
 mill-0.182
ting-0.182
 painter-0.175
ini-0.171
 paramed-0.168
Token you
Token ID345
Feature activation-6.34e-02

Pos logit contributions
 mill+0.093
ting+0.092
 painter+0.090
ini+0.087
 paramed+0.085
Neg logit contributions
 dives-0.056
 drift-0.056
irm-0.055
 much-0.054
 nap-0.053
Token can
Token ID460
Feature activation+1.15e-01

Pos logit contributions
 dives+0.092
 drift+0.091
irm+0.091
ls+0.088
 nap+0.087
Neg logit contributions
 mill-0.101
 painter-0.098
ting-0.097
ini-0.092
 office-0.091
Token organize
Token ID16481
Feature activation+2.04e-02

Pos logit contributions
 mill+0.207
ting+0.206
 painter+0.199
ini+0.196
 office+0.190
Neg logit contributions
 much-0.146
 dives-0.144
 drift-0.143
 nap-0.141
irm-0.140
Token a
Token ID257
Feature activation-7.04e-02

Pos logit contributions
 dives+0.038
 drift+0.038
 much+0.037
irm+0.036
 nap+0.036
Neg logit contributions
 mill-0.065
ting-0.065
 painter-0.062
ini-0.061
 office-0.061
Token hug
Token ID16225
Feature activation+1.07e-02

Pos logit contributions
irm+0.095
ust+0.094
 dives+0.090
 drift+0.090
ls+0.088
Neg logit contributions
 mill-0.118
 painter-0.117
 paramed-0.114
 iron-0.106
 laund-0.106
Token.
Token ID13
Feature activation-3.35e-03

Pos logit contributions
 mill+0.025
ini+0.024
 painter+0.024
years+0.023
ting+0.023
Neg logit contributions
 much-0.009
 nap-0.008
 drift-0.008
irm-0.007
 dives-0.007
Token
Token ID198
Feature activation-5.30e-02

Pos logit contributions
 much+0.004
 nap+0.004
 drift+0.004
 dives+0.004
irm+0.004
Neg logit contributions
ting-0.006
 mill-0.006
 painter-0.006
ini-0.006
cut-0.006
Token
Token ID198
Feature activation-5.23e-02

Pos logit contributions
 much+0.050
 drift+0.048
 nap+0.048
 dives+0.048
irm+0.045
Neg logit contributions
ting-0.113
 mill-0.113
 painter-0.109
ini-0.108
 paramed-0.105
Token"
Token ID1
Feature activation+2.26e-01

Pos logit contributions
 much+0.071
 nap+0.061
 drift+0.059
 dives+0.057
irm+0.057
Neg logit contributions
 mill-0.101
ting-0.101
smith-0.097
 painter-0.093
 office-0.093
TokenAre
Token ID8491
Feature activation+6.75e-02

Pos logit contributions
 mill+0.276
ting+0.259
 office+0.254
 painter+0.242
ini+0.240
Neg logit contributions
 much-0.433
 nap-0.429
 drift-0.422
irm-0.410
 dives-0.404
Token you
Token ID345
Feature activation-6.19e-02

Pos logit contributions
ting+0.101
Why+0.089
years+0.088
 mill+0.086
ter+0.085
Neg logit contributions
 drift-0.112
 dives-0.108
irm-0.106
 nap-0.101
 hops-0.101
Token okay
Token ID8788
Feature activation-2.91e-02

Pos logit contributions
ls+0.054
 dives+0.048
 common+0.046
 drift+0.046
 Zac+0.044
Neg logit contributions
ting-0.138
ter-0.131
ini-0.130
 painter-0.130
 mill-0.129
Token?"
Token ID1701
Feature activation+5.90e-02

Pos logit contributions
 drift+0.037
 dives+0.037
 much+0.036
irm+0.036
 nap+0.035
Neg logit contributions
 mill-0.052
ting-0.052
 painter-0.052
ini-0.049
 office-0.049
Token her
Token ID607
Feature activation-3.37e-02

Pos logit contributions
 mill+0.110
ting+0.110
 painter+0.107
ini+0.104
 paramed+0.102
Neg logit contributions
 much-0.070
 dives-0.069
 drift-0.069
 nap-0.068
irm-0.067
Token day
Token ID1110
Feature activation-2.82e-02

Pos logit contributions
 dives+0.110
 drift+0.108
irm+0.107
ls+0.107
ares+0.104
Neg logit contributions
 painter-0.083
 mill-0.082
ini-0.076
 office-0.070
ting-0.069
Token,
Token ID11
Feature activation+4.62e-02

Pos logit contributions
ust+0.043
irm+0.043
 drift+0.043
 dives+0.041
 hops+0.040
Neg logit contributions
ting-0.044
 mill-0.044
 painter-0.041
 paramed-0.040
 laund-0.038
Token Tim
Token ID5045
Feature activation+3.81e-02

Pos logit contributions
 mill+0.073
ting+0.073
 painter+0.069
ini+0.066
 paramed+0.065
Neg logit contributions
 drift-0.069
 dives-0.069
irm-0.068
 nap-0.067
 much-0.067
Tokenmy
Token ID1820
Feature activation+2.63e-02

Pos logit contributions
ting+0.050
 mill+0.049
 painter+0.045
 paramed+0.045
cut+0.044
Neg logit contributions
 much-0.067
 nap-0.067
 dives-0.066
 drift-0.066
irm-0.065
Token's
Token ID338
Feature activation-2.67e-02

Pos logit contributions
ting+0.046
 mill+0.046
 painter+0.044
ini+0.044
 paramed+0.043
Neg logit contributions
 much-0.034
 dives-0.034
 drift-0.033
 nap-0.033
irm-0.032
Token friend
Token ID1545
Feature activation+4.97e-02

Pos logit contributions
 nap+0.030
 much+0.028
 dives+0.026
 drift+0.026
 swim+0.026
Neg logit contributions
ting-0.056
cut-0.052
 mill-0.051
years-0.051
ini-0.051
Token came
Token ID1625
Feature activation+7.46e-03

Pos logit contributions
ting+0.084
 mill+0.083
 painter+0.082
 paramed+0.078
ini+0.077
Neg logit contributions
 drift-0.069
irm-0.068
 dives-0.068
 much-0.067
 nap-0.066
Token over
Token ID625
Feature activation-3.44e-02

Pos logit contributions
ting+0.010
 painter+0.010
 mill+0.009
 office+0.009
ini+0.009
Neg logit contributions
 drift-0.013
 nap-0.012
ls-0.012
 dives-0.012
irm-0.012
Token to
Token ID284
Feature activation+4.06e-02

Pos logit contributions
 much+0.039
 nap+0.038
 dives+0.038
 drift+0.036
irm+0.035
Neg logit contributions
 mill-0.068
ting-0.067
ini-0.066
 painter-0.065
cut-0.064
Token play
Token ID711
Feature activation+2.59e-03

Pos logit contributions
 mill+0.084
ini+0.084
ting+0.083
 floors+0.081
rom+0.080
Neg logit contributions
 swim-0.045
 nap-0.045
 drift-0.043
 dive-0.042
 dives-0.041
Token.
Token ID13
Feature activation-6.92e-03

Pos logit contributions
ting+0.005
 mill+0.005
 painter+0.005
ini+0.005
 office+0.005
Neg logit contributions
 much-0.003
 drift-0.003
 dives-0.003
 nap-0.003
irm-0.003
Token,
Token ID11
Feature activation+5.18e-02

Pos logit contributions
 much+0.160
 drift+0.154
 nap+0.152
 dives+0.151
irm+0.144
Neg logit contributions
ting-0.301
 mill-0.299
ini-0.290
 painter-0.289
 paramed-0.281
Token they
Token ID484
Feature activation-7.71e-02

Pos logit contributions
ting+0.117
 mill+0.117
 painter+0.113
ini+0.112
 paramed+0.109
Neg logit contributions
 much-0.041
 drift-0.040
 dives-0.040
 nap-0.040
irm-0.038
Token were
Token ID547
Feature activation+2.38e-02

Pos logit contributions
 much+0.095
 drift+0.092
 nap+0.088
 dives+0.087
 dive+0.084
Neg logit contributions
ting-0.149
 mill-0.144
ini-0.141
 paramed-0.137
 painter-0.136
Token able
Token ID1498
Feature activation-3.41e-02

Pos logit contributions
ting+0.050
 mill+0.049
ini+0.048
 painter+0.047
 paramed+0.046
Neg logit contributions
 much-0.025
 drift-0.023
 dives-0.023
 nap-0.022
irm-0.021
Token to
Token ID284
Feature activation+4.52e-02

Pos logit contributions
 drift+0.051
 dives+0.050
irm+0.050
 nap+0.049
 hops+0.048
Neg logit contributions
 mill-0.053
ting-0.051
 painter-0.051
 office-0.048
ini-0.048
Token build
Token ID1382
Feature activation+6.91e-02

Pos logit contributions
ting+0.088
 mill+0.079
years+0.078
ini+0.078
 paramed+0.077
Neg logit contributions
 nap-0.059
 much-0.059
 drift-0.058
 dive-0.058
 swim-0.057
Token the
Token ID262
Feature activation+5.91e-02

Pos logit contributions
 mill+0.121
ting+0.121
 painter+0.116
ini+0.114
 office+0.112
Neg logit contributions
 drift-0.088
 much-0.088
 dives-0.088
 nap-0.087
irm-0.086
Token house
Token ID2156
Feature activation+6.85e-02

Pos logit contributions
ting+0.085
years+0.071
stick+0.071
cut+0.067
 paramed+0.067
Neg logit contributions
 much-0.111
 nap-0.107
 drift-0.100
 dives-0.098
 time-0.097
Token in
Token ID287
Feature activation-4.44e-02

Pos logit contributions
ting+0.133
 mill+0.133
 painter+0.128
ini+0.126
 paramed+0.123
Neg logit contributions
 much-0.076
 drift-0.074
 dives-0.074
 nap-0.074
irm-0.072
Token no
Token ID645
Feature activation+5.64e-03

Pos logit contributions
 much+0.062
 nap+0.056
 drift+0.054
 dives+0.051
 dive+0.050
Neg logit contributions
ting-0.086
ini-0.080
 mill-0.079
 paramed-0.078
 painter-0.077
Token time
Token ID640
Feature activation-1.04e-01

Pos logit contributions
 mill+0.013
 painter+0.013
ting+0.013
 office+0.012
 floors+0.012
Neg logit contributions
ls-0.004
ust-0.004
 dives-0.004
 drift-0.004
irm-0.004
Token
Token ID198
Feature activation-5.31e-02

Pos logit contributions
 much+0.049
 nap+0.046
 drift+0.045
 dives+0.044
irm+0.043
Neg logit contributions
ting-0.119
 mill-0.119
 painter-0.115
ini-0.114
cut-0.110
TokenHer
Token ID9360
Feature activation-1.18e-02

Pos logit contributions
 much+0.084
 nap+0.080
irm+0.077
ust+0.076
 drift+0.076
Neg logit contributions
ting-0.086
 mill-0.084
ini-0.078
 painter-0.077
cut-0.075
Token mom
Token ID1995
Feature activation+1.03e-01

Pos logit contributions
 much+0.016
 nap+0.016
 dives+0.016
 drift+0.016
irm+0.015
Neg logit contributions
ting-0.021
 mill-0.020
ini-0.019
 painter-0.019
cut-0.018
Token came
Token ID1625
Feature activation+1.05e-02

Pos logit contributions
 mill+0.197
ini+0.190
 painter+0.185
keeping+0.184
ting+0.183
Neg logit contributions
 much-0.124
 drift-0.121
 nap-0.118
 dives-0.114
irm-0.109
Token to
Token ID284
Feature activation+4.28e-02

Pos logit contributions
 mill+0.020
ini+0.020
ting+0.019
cut+0.019
 painter+0.019
Neg logit contributions
 much-0.012
 nap-0.012
 dives-0.012
 drift-0.012
irm-0.011
Token help
Token ID1037
Feature activation+3.80e-02

Pos logit contributions
ini+0.089
 mill+0.088
ting+0.085
cles+0.085
cut+0.085
Neg logit contributions
 nap-0.050
 drift-0.046
 swim-0.046
 much-0.042
 dive-0.041
Token her
Token ID607
Feature activation-3.30e-02

Pos logit contributions
 mill+0.068
ting+0.066
ini+0.065
 painter+0.064
cut+0.062
Neg logit contributions
 much-0.052
 nap-0.051
 drift-0.050
 dives-0.048
irm-0.047
Token and
Token ID290
Feature activation-2.97e-02

Pos logit contributions
 much+0.041
 dives+0.040
 drift+0.040
 nap+0.040
irm+0.039
Neg logit contributions
 mill-0.060
ting-0.060
 painter-0.058
ini-0.057
 office-0.056
Token said
Token ID531
Feature activation-4.57e-02

Pos logit contributions
 much+0.030
 drift+0.028
 nap+0.027
 dives+0.027
irm+0.026
Neg logit contributions
ting-0.062
 mill-0.062
ini-0.060
 painter-0.060
 pavement-0.059
Token,
Token ID11
Feature activation+4.97e-02

Pos logit contributions
 much+0.046
 nap+0.043
 drift+0.043
 dives+0.042
irm+0.040
Neg logit contributions
ting-0.096
 mill-0.096
ini-0.094
 painter-0.093
 office-0.089
Token "
Token ID366
Feature activation+2.31e-01

Pos logit contributions
ini+0.082
 mill+0.079
 painter+0.076
ting+0.076
keeping+0.074
Neg logit contributions
 much-0.076
 drift-0.071
 nap-0.067
 they-0.067
 okay-0.067
Token animals
Token ID4695
Feature activation+2.19e-02

Pos logit contributions
ting+0.020
 mill+0.020
 painter+0.020
ini+0.020
 paramed+0.019
Neg logit contributions
 much-0.011
 nap-0.011
 dives-0.010
 drift-0.010
 dive-0.010
Token.
Token ID13
Feature activation-6.17e-03

Pos logit contributions
 mill+0.042
ting+0.041
 painter+0.040
ini+0.040
 paramed+0.039
Neg logit contributions
 much-0.026
 drift-0.025
 nap-0.025
 dives-0.025
irm-0.023
Token Would
Token ID10928
Feature activation+3.12e-02

Pos logit contributions
 much+0.008
 drift+0.008
 dives+0.008
 nap+0.008
irm+0.008
Neg logit contributions
 mill-0.011
ting-0.011
 painter-0.010
ini-0.010
 paramed-0.010
Token you
Token ID345
Feature activation-6.25e-02

Pos logit contributions
ting+0.059
 mill+0.058
 painter+0.056
ini+0.055
 office+0.053
Neg logit contributions
 drift-0.036
 much-0.036
 dives-0.036
 nap-0.036
irm-0.036
Token like
Token ID588
Feature activation+3.79e-02

Pos logit contributions
ls+0.090
irm+0.088
 dives+0.088
 drift+0.087
 much+0.084
Neg logit contributions
 painter-0.095
ting-0.094
 mill-0.092
ini-0.088
 office-0.085
Token to
Token ID284
Feature activation+4.25e-02

Pos logit contributions
ini+0.079
ting+0.078
 mill+0.078
 paramed+0.075
 painter+0.075
Neg logit contributions
 much-0.041
 nap-0.041
 dives-0.038
 drift-0.037
 dive-0.037
Token play
Token ID711
Feature activation+4.56e-03

Pos logit contributions
 mill+0.082
ting+0.082
 painter+0.079
ini+0.077
 office+0.075
Neg logit contributions
 much-0.047
 dives-0.047
 drift-0.046
irm-0.045
 nap-0.045
Token too
Token ID1165
Feature activation-9.55e-02

Pos logit contributions
 mill+0.009
ting+0.009
ini+0.009
 painter+0.009
 paramed+0.009
Neg logit contributions
 much-0.005
 nap-0.005
 dives-0.005
 drift-0.005
irm-0.004
Token?"
Token ID1701
Feature activation+6.87e-02

Pos logit contributions
 drift+0.146
 dives+0.145
 much+0.145
 nap+0.143
irm+0.142
Neg logit contributions
 mill-0.145
ting-0.144
 painter-0.138
ini-0.135
 office-0.131
Token
Token ID198
Feature activation-5.26e-02

Pos logit contributions
ting+0.140
 mill+0.140
 painter+0.136
ini+0.134
 paramed+0.130
Neg logit contributions
 much-0.070
 drift-0.068
 dives-0.068
 nap-0.068
irm-0.066
Token
Token ID198
Feature activation-5.17e-02

Pos logit contributions
 much+0.048
 drift+0.045
 nap+0.045
 dives+0.045
irm+0.043
Neg logit contributions
ting-0.115
 mill-0.114
 painter-0.110
ini-0.109
 office-0.106
Tokenily
Token ID813
Feature activation+3.55e-02

Pos logit contributions
 drift+0.004
 much+0.004
 dives+0.004
 nap+0.004
irm+0.004
Neg logit contributions
ting-0.003
 mill-0.003
 painter-0.003
ini-0.003
 office-0.003
Token didn
Token ID1422
Feature activation+6.18e-02

Pos logit contributions
ting+0.062
 mill+0.060
ini+0.058
 painter+0.058
 paramed+0.056
Neg logit contributions
 much-0.049
 dives-0.048
 drift-0.047
 nap-0.047
irm-0.046
Token't
Token ID470
Feature activation-5.94e-02

Pos logit contributions
 mill+0.088
ting+0.087
 painter+0.083
rom+0.082
 paramed+0.081
Neg logit contributions
irm-0.100
 dives-0.099
 drift-0.096
 hops-0.096
 nap-0.093
Token like
Token ID588
Feature activation+4.11e-02

Pos logit contributions
ls+0.074
 dives+0.071
irm+0.069
 drift+0.069
 hops+0.064
Neg logit contributions
 painter-0.111
ting-0.105
 mill-0.101
 owl-0.101
 paramed-0.098
Token bugs
Token ID11316
Feature activation+2.96e-02

Pos logit contributions
ini+0.084
rom+0.081
cut+0.080
 paramed+0.079
cles+0.078
Neg logit contributions
 much-0.052
 nap-0.048
 dives-0.046
 dive-0.044
 time-0.044
Token,
Token ID11
Feature activation+4.84e-02

Pos logit contributions
ting+0.056
 mill+0.054
 painter+0.052
ini+0.050
cut+0.049
Neg logit contributions
 dives-0.035
 much-0.035
 drift-0.035
irm-0.034
 nap-0.033
Token so
Token ID523
Feature activation-4.43e-02

Pos logit contributions
ini+0.106
 mill+0.105
ting+0.104
 painter+0.104
 paramed+0.100
Neg logit contributions
 much-0.046
 nap-0.043
 drift-0.043
 dives-0.042
irm-0.039
Token she
Token ID673
Feature activation-5.11e-02

Pos logit contributions
 much+0.051
 dives+0.050
 drift+0.050
 nap+0.050
irm+0.049
Neg logit contributions
ting-0.085
 mill-0.085
 painter-0.082
ini-0.080
 paramed-0.078
Token asked
Token ID1965
Feature activation+2.13e-02

Pos logit contributions
 much+0.080
 drift+0.079
 nap+0.078
 dives+0.077
irm+0.076
Neg logit contributions
 mill-0.076
ting-0.076
 painter-0.073
ini-0.070
 paramed-0.069
Token her
Token ID607
Feature activation-3.66e-02

Pos logit contributions
ini+0.047
 mill+0.044
 laund+0.041
ting+0.041
rom+0.041
Neg logit contributions
 much-0.026
 nap-0.023
 drift-0.022
 okay-0.021
 OK-0.021
Token mom
Token ID1995
Feature activation+1.02e-01

Pos logit contributions
 much+0.046
 nap+0.045
 dives+0.043
 drift+0.043
 swim+0.041
Neg logit contributions
ting-0.070
 mill-0.067
ini-0.066
rom-0.063
 pavement-0.063
Token made
Token ID925
Feature activation+3.79e-02

Pos logit contributions
ting+0.023
 painter+0.023
 mill+0.023
ini+0.021
 office+0.021
Neg logit contributions
 much-0.021
 drift-0.021
 dives-0.021
irm-0.020
ust-0.020
Token her
Token ID607
Feature activation-3.54e-02

Pos logit contributions
ting+0.065
 mill+0.064
 painter+0.064
 office+0.061
ini+0.060
Neg logit contributions
 drift-0.049
 dives-0.048
irm-0.047
ust-0.047
 nap-0.046
Token flower
Token ID15061
Feature activation-1.30e-03

Pos logit contributions
 dives+0.049
irm+0.049
 drift+0.049
ust+0.049
 much+0.046
Neg logit contributions
 painter-0.061
 office-0.057
 mill-0.055
ting-0.052
ini-0.052
Token.
Token ID13
Feature activation+6.18e-05

Pos logit contributions
 dives+0.002
irm+0.002
ust+0.002
 drift+0.002
 much+0.002
Neg logit contributions
 painter-0.002
 mill-0.002
 office-0.002
ting-0.002
ini-0.002
Token
Token ID198
Feature activation-5.64e-02

Pos logit contributions
 painter+0.000
 mill+0.000
ting+0.000
Why+0.000
 paramed+0.000
Neg logit contributions
 drift-0.000
 dives-0.000
 nap-0.000
 dive-0.000
irm-0.000
Token
Token ID198
Feature activation-5.64e-02

Pos logit contributions
 much+0.056
 drift+0.055
 dives+0.055
 nap+0.054
irm+0.053
Neg logit contributions
 mill-0.115
ting-0.115
 painter-0.111
ini-0.110
 office-0.107
TokenThe
Token ID464
Feature activation+7.36e-02

Pos logit contributions
 much+0.088
 nap+0.084
 drift+0.084
 dives+0.084
irm+0.082
Neg logit contributions
 mill-0.089
ting-0.088
 painter-0.084
ini-0.082
 office-0.080
Token teacher
Token ID4701
Feature activation+9.82e-02

Pos logit contributions
ting+0.104
 mill+0.100
ini+0.097
rom+0.094
cut+0.093
Neg logit contributions
 nap-0.131
 much-0.129
 dives-0.123
 drift-0.120
 hops-0.117
Token says
Token ID1139
Feature activation-5.32e-02

Pos logit contributions
 mill+0.181
ting+0.181
 painter+0.180
 office+0.172
 paramed+0.164
Neg logit contributions
 much-0.119
irm-0.116
 drift-0.116
 nap-0.112
ust-0.112
Token,
Token ID11
Feature activation+5.00e-02

Pos logit contributions
 drift+0.059
 much+0.059
 dives+0.058
irm+0.058
 nap+0.058
Neg logit contributions
 mill-0.103
ting-0.103
 painter-0.100
ini-0.096
 paramed-0.096
Token "
Token ID366
Feature activation+2.31e-01

Pos logit contributions
ini+0.102
keeping+0.093
 mill+0.092
 floors+0.090
 pavement+0.088
Neg logit contributions
 much-0.064
 they-0.060
 nap-0.058
 swim-0.057
 dives-0.056
Token the
Token ID262
Feature activation+5.97e-02

Pos logit contributions
 much+0.034
 dives+0.033
 nap+0.033
 drift+0.032
 swim+0.032
Neg logit contributions
ini-0.071
ting-0.068
 mill-0.067
 painter-0.066
rom-0.065
Token ball
Token ID2613
Feature activation+1.02e-02

Pos logit contributions
ting+0.110
years+0.108
 paramed+0.107
 mill+0.106
ini+0.106
Neg logit contributions
 nap-0.080
 much-0.076
 dives-0.074
 drift-0.072
 swim-0.072
Token rolled
Token ID11686
Feature activation-3.05e-02

Pos logit contributions
 mill+0.016
ting+0.016
 painter+0.015
cut+0.014
 office+0.014
Neg logit contributions
 drift-0.015
irm-0.015
 dives-0.015
 much-0.015
 nap-0.014
Token away
Token ID1497
Feature activation-7.63e-03

Pos logit contributions
 much+0.033
 dives+0.031
 nap+0.030
irm+0.030
 drift+0.030
Neg logit contributions
 mill-0.063
ini-0.063
 painter-0.062
ting-0.061
 paramed-0.060
Token.
Token ID13
Feature activation-2.49e-03

Pos logit contributions
 dives+0.010
 drift+0.010
irm+0.010
 nap+0.010
 hops+0.009
Neg logit contributions
ting-0.013
 mill-0.013
 painter-0.012
cut-0.012
 office-0.012
Token
Token ID198
Feature activation-5.71e-02

Pos logit contributions
 drift+0.003
 dives+0.003
 nap+0.003
 much+0.003
irm+0.003
Neg logit contributions
 mill-0.005
ting-0.005
 painter-0.005
 paramed-0.004
ini-0.004
Token
Token ID198
Feature activation-5.66e-02

Pos logit contributions
 dives+0.077
 drift+0.077
irm+0.075
 hops+0.073
 nap+0.072
Neg logit contributions
 mill-0.096
 painter-0.092
ting-0.091
ini-0.090
rom-0.090
Token"
Token ID1
Feature activation+2.24e-01

Pos logit contributions
 drift+0.077
 dives+0.077
 nap+0.076
 much+0.075
irm+0.075
Neg logit contributions
 mill-0.096
ting-0.094
 painter-0.091
ini-0.089
 paramed-0.088
TokenTom
Token ID13787
Feature activation+4.84e-02

Pos logit contributions
 mill+0.211
ting+0.200
ini+0.195
 office+0.192
 painter+0.182
Neg logit contributions
 much-0.480
 nap-0.476
 dives-0.468
 drift-0.467
irm-0.458
Token,
Token ID11
Feature activation+5.01e-02

Pos logit contributions
ting+0.112
 mill+0.111
ini+0.107
 laund+0.107
 floors+0.106
Neg logit contributions
 much-0.037
 nap-0.035
 dives-0.035
 drift-0.032
 hops-0.032
Token you
Token ID345
Feature activation-6.40e-02

Pos logit contributions
 mill+0.093
ting+0.093
 painter+0.090
ini+0.087
 paramed+0.086
Neg logit contributions
 much-0.060
 drift-0.059
 dives-0.059
irm-0.058
 nap-0.057
Token them
Token ID606
Feature activation-2.03e-02

Pos logit contributions
 much+0.051
 dives+0.050
 drift+0.050
 nap+0.050
irm+0.049
Neg logit contributions
 mill-0.049
ting-0.049
 painter-0.047
ini-0.046
 paramed-0.044
Token hot
Token ID3024
Feature activation-4.22e-03

Pos logit contributions
 drift+0.028
 dives+0.028
irm+0.028
 much+0.027
 hops+0.027
Neg logit contributions
ting-0.033
 mill-0.033
 painter-0.032
 paramed-0.030
 office-0.030
Token chocolate
Token ID11311
Feature activation-1.73e-02

Pos logit contributions
 nap+0.006
 much+0.006
 dives+0.006
 drift+0.006
irm+0.006
Neg logit contributions
ting-0.007
 mill-0.007
 painter-0.007
ini-0.007
years-0.007
Token and
Token ID290
Feature activation-3.44e-02

Pos logit contributions
 dives+0.024
 much+0.023
 drift+0.023
irm+0.023
 nap+0.023
Neg logit contributions
 mill-0.029
ting-0.029
 painter-0.027
ini-0.027
 paramed-0.027
Token cookies
Token ID14746
Feature activation-2.89e-02

Pos logit contributions
 dives+0.047
 much+0.046
 drift+0.046
irm+0.046
ust+0.045
Neg logit contributions
 mill-0.057
ting-0.056
 painter-0.055
ini-0.052
 paramed-0.052
Token.
Token ID13
Feature activation-1.66e-03

Pos logit contributions
 dives+0.038
irm+0.037
 much+0.037
 drift+0.037
ust+0.036
Neg logit contributions
ting-0.050
 mill-0.050
 painter-0.047
cut-0.046
 paramed-0.046
Token She
Token ID1375
Feature activation+1.20e-03

Pos logit contributions
 drift+0.002
 dives+0.002
 much+0.002
 nap+0.002
irm+0.002
Neg logit contributions
 mill-0.003
ting-0.003
 painter-0.003
ini-0.003
 paramed-0.003
Token was
Token ID373
Feature activation+3.43e-02

Pos logit contributions
 mill+0.002
 painter+0.002
ting+0.002
 office+0.002
ini+0.002
Neg logit contributions
 dives-0.002
irm-0.002
 drift-0.002
 much-0.002
ust-0.002
Token kind
Token ID1611
Feature activation+6.07e-02

Pos logit contributions
 painter+0.060
ting+0.060
 mill+0.060
 paramed+0.058
 office+0.058
Neg logit contributions
 dives-0.043
 drift-0.042
 hops-0.041
 nap-0.041
irm-0.041
Token and
Token ID290
Feature activation-3.47e-02

Pos logit contributions
ting+0.110
 mill+0.109
 painter+0.105
ini+0.104
 paramed+0.100
Neg logit contributions
 much-0.076
 drift-0.075
 dives-0.075
 nap-0.074
irm-0.072
Token loving
Token ID14442
Feature activation+2.09e-02

Pos logit contributions
 dives+0.041
 drift+0.041
 nap+0.040
 much+0.040
irm+0.040
Neg logit contributions
 mill-0.064
ting-0.064
 painter-0.062
ini-0.061
 paramed-0.059
Token when
Token ID618
Feature activation-4.27e-02

Pos logit contributions
 mill+0.089
ting+0.089
ini+0.087
cut+0.085
 painter+0.083
Neg logit contributions
 much-0.065
 nap-0.060
 drift-0.058
 dives-0.056
irm-0.054
Token Sarah
Token ID10490
Feature activation+1.41e-02

Pos logit contributions
 much+0.059
 drift+0.058
 dives+0.058
 nap+0.057
irm+0.056
Neg logit contributions
ting-0.072
 mill-0.072
 painter-0.069
ini-0.068
 paramed-0.066
Token was
Token ID373
Feature activation+3.09e-02

Pos logit contributions
ting+0.027
 mill+0.027
cut+0.026
ini+0.026
 paramed+0.026
Neg logit contributions
 nap-0.017
 drift-0.016
 dives-0.015
 much-0.015
irm-0.015
Token using
Token ID1262
Feature activation+7.66e-03

Pos logit contributions
ting+0.047
 mill+0.047
 painter+0.045
ini+0.044
 paramed+0.042
Neg logit contributions
 much-0.048
 dives-0.047
 drift-0.047
 nap-0.047
irm-0.046
Token her
Token ID607
Feature activation-3.21e-02

Pos logit contributions
ting+0.013
 mill+0.013
ini+0.013
 painter+0.012
 paramed+0.012
Neg logit contributions
 much-0.011
 nap-0.011
 drift-0.011
 dives-0.010
irm-0.010
Token brush
Token ID14093
Feature activation+3.02e-02

Pos logit contributions
 nap+0.053
 much+0.051
 drift+0.051
 dives+0.050
irm+0.048
Neg logit contributions
ting-0.050
 mill-0.047
ini-0.046
cut-0.045
 paramed-0.044
Token,
Token ID11
Feature activation+4.94e-02

Pos logit contributions
ting+0.063
 mill+0.063
ini+0.062
 painter+0.061
 paramed+0.060
Neg logit contributions
 much-0.031
 nap-0.030
 drift-0.029
 dives-0.026
irm-0.026
Token she
Token ID673
Feature activation-4.64e-02

Pos logit contributions
ting+0.107
 mill+0.106
ini+0.105
cut+0.100
 paramed+0.100
Neg logit contributions
 much-0.051
 nap-0.044
 drift-0.042
 dives-0.038
 they-0.037
Token heard
Token ID2982
Feature activation+6.56e-02

Pos logit contributions
 much+0.061
 nap+0.058
 dives+0.057
 drift+0.055
 hops+0.053
Neg logit contributions
ting-0.088
 mill-0.083
cut-0.081
ican-0.080
 paramed-0.080
Token a
Token ID257
Feature activation-7.31e-02

Pos logit contributions
 mill+0.151
ini+0.150
ting+0.146
cut+0.144
 painter+0.141
Neg logit contributions
 much-0.065
 nap-0.052
 drift-0.049
 dives-0.045
 time-0.042
Token loud
Token ID7812
Feature activation+8.87e-03

Pos logit contributions
 much+0.087
 drift+0.083
 nap+0.082
 dives+0.080
 swim+0.077
Neg logit contributions
ting-0.144
 mill-0.137
ini-0.137
keeping-0.133
 painter-0.132
Token away
Token ID1497
Feature activation-9.50e-03

Pos logit contributions
ting+0.027
 office+0.024
stick+0.024
 mill+0.024
rom+0.023
Neg logit contributions
 dives-0.023
 drift-0.022
 hops-0.022
 much-0.021
irm-0.021
Token from
Token ID422
Feature activation-7.25e-02

Pos logit contributions
 dives+0.014
irm+0.013
 nap+0.013
 drift+0.013
 hops+0.012
Neg logit contributions
ting-0.017
 mill-0.016
 painter-0.015
cut-0.015
 office-0.015
Token him
Token ID683
Feature activation-1.70e-02

Pos logit contributions
 nap+0.078
 much+0.076
 drift+0.071
 swim+0.070
 dives+0.068
Neg logit contributions
ini-0.162
 paramed-0.152
 mill-0.147
 painter-0.143
 laund-0.141
Token.
Token ID13
Feature activation-2.14e-03

Pos logit contributions
 dives+0.021
 much+0.020
 drift+0.020
irm+0.020
 nap+0.019
Neg logit contributions
ting-0.032
 mill-0.031
 painter-0.030
 office-0.029
ini-0.029
Token
Token ID198
Feature activation-5.44e-02

Pos logit contributions
 much+0.002
 nap+0.002
 dives+0.002
 drift+0.002
irm+0.002
Neg logit contributions
ini-0.005
ting-0.005
 mill-0.005
 painter-0.004
cut-0.004
Token
Token ID198
Feature activation-5.37e-02

Pos logit contributions
 much+0.050
 drift+0.047
 nap+0.047
 dives+0.047
irm+0.044
Neg logit contributions
ting-0.119
 mill-0.118
 painter-0.115
ini-0.114
 paramed-0.110
TokenBut
Token ID1537
Feature activation+4.18e-02

Pos logit contributions
 much+0.077
 nap+0.070
 dives+0.068
 drift+0.067
irm+0.067
Neg logit contributions
ting-0.094
 mill-0.092
ini-0.092
 painter-0.089
smith-0.087
Token Tom
Token ID4186
Feature activation+1.28e-02

Pos logit contributions
 mill+0.078
ting+0.077
 painter+0.075
ini+0.074
 paramed+0.071
Neg logit contributions
 much-0.051
 drift-0.050
 dives-0.049
 nap-0.049
irm-0.048
Token was
Token ID373
Feature activation+3.03e-02

Pos logit contributions
 mill+0.024
ting+0.024
ini+0.023
 painter+0.023
 paramed+0.022
Neg logit contributions
 much-0.015
 nap-0.015
 dives-0.015
 drift-0.015
irm-0.014
Token stronger
Token ID7387
Feature activation-4.61e-02

Pos logit contributions
ting+0.057
 painter+0.055
years+0.055
 mill+0.054
Why+0.054
Neg logit contributions
 drift-0.037
 dives-0.035
irm-0.032
 hops-0.032
 nap-0.032
Token and
Token ID290
Feature activation-3.57e-02

Pos logit contributions
 much+0.054
 dives+0.052
 nap+0.052
 drift+0.052
irm+0.050
Neg logit contributions
 mill-0.088
ting-0.088
 painter-0.085
ini-0.085
 paramed-0.082
Token
Token ID198
Feature activation-5.99e-02

Pos logit contributions
 drift+0.007
 dives+0.007
 much+0.007
 nap+0.007
irm+0.007
Neg logit contributions
 mill-0.011
ting-0.011
 painter-0.011
ini-0.011
rom-0.010
Token
Token ID198
Feature activation-5.97e-02

Pos logit contributions
 dives+0.069
 drift+0.069
irm+0.067
 nap+0.066
 much+0.064
Neg logit contributions
 mill-0.113
ting-0.110
 painter-0.109
ini-0.108
rom-0.105
TokenOne
Token ID3198
Feature activation-5.95e-02

Pos logit contributions
 much+0.068
 drift+0.065
 nap+0.065
 dives+0.065
irm+0.062
Neg logit contributions
ting-0.117
 mill-0.116
 painter-0.112
ini-0.110
 office-0.108
Token day
Token ID1110
Feature activation-2.99e-02

Pos logit contributions
irm+0.118
ls+0.117
ares+0.115
 drift+0.113
 dives+0.112
Neg logit contributions
 mill-0.073
 painter-0.071
 office-0.069
ini-0.060
 paramed-0.055
Token,
Token ID11
Feature activation+4.78e-02

Pos logit contributions
ust+0.050
irm+0.049
 drift+0.048
 dives+0.045
 nap+0.044
Neg logit contributions
ting-0.043
 mill-0.042
 painter-0.040
 laund-0.037
ter-0.036
Token they
Token ID484
Feature activation-8.16e-02

Pos logit contributions
ting+0.097
 mill+0.097
 painter+0.094
 laund+0.089
Why+0.089
Neg logit contributions
 drift-0.050
 dives-0.049
irm-0.048
 hops-0.047
ust-0.046
Token saw
Token ID2497
Feature activation+1.18e-01

Pos logit contributions
 much+0.130
 nap+0.127
 drift+0.127
 dives+0.127
irm+0.123
Neg logit contributions
ting-0.121
 mill-0.120
ini-0.114
 painter-0.114
 paramed-0.109
Token a
Token ID257
Feature activation-7.34e-02

Pos logit contributions
ini+0.305
ting+0.301
cut+0.285
 painter+0.283
 mill+0.282
Neg logit contributions
 much-0.091
 dives-0.063
 nap-0.061
 they-0.059
 swim-0.059
Token new
Token ID649
Feature activation+1.14e-02

Pos logit contributions
 much+0.075
 nap+0.072
 dives+0.067
 swim+0.066
 dive+0.065
Neg logit contributions
ting-0.162
keeping-0.150
years-0.150
ini-0.148
cles-0.146
Token boy
Token ID2933
Feature activation+2.70e-02

Pos logit contributions
ting+0.017
 mill+0.016
ini+0.015
years+0.015
 painter+0.015
Neg logit contributions
 much-0.019
 nap-0.018
 dives-0.018
 drift-0.018
 dive-0.017
Token in
Token ID287
Feature activation-4.85e-02

Pos logit contributions
ting+0.041
 mill+0.040
 painter+0.039
ini+0.037
 office+0.037
Neg logit contributions
 drift-0.042
 much-0.042
 dives-0.041
 nap-0.041
irm-0.041
Token wet
Token ID9583
Feature activation+2.58e-02

Pos logit contributions
 dives+0.032
 drift+0.032
irm+0.032
 nap+0.032
ust+0.032
Neg logit contributions
 mill-0.031
ting-0.031
 painter-0.030
ini-0.030
 paramed-0.028
Token.
Token ID13
Feature activation-7.77e-03

Pos logit contributions
 mill+0.058
 painter+0.058
 paramed+0.056
ini+0.056
 office+0.056
Neg logit contributions
 much-0.022
 nap-0.021
 dives-0.020
 drift-0.020
irm-0.019
Token When
Token ID1649
Feature activation-5.52e-02

Pos logit contributions
 much+0.011
 nap+0.011
 drift+0.011
 dives+0.011
irm+0.011
Neg logit contributions
 mill-0.013
ting-0.013
ini-0.012
 painter-0.012
cut-0.012
Token she
Token ID673
Feature activation-4.90e-02

Pos logit contributions
 dives+0.063
 much+0.063
 drift+0.063
irm+0.062
 nap+0.061
Neg logit contributions
 mill-0.106
ting-0.105
 painter-0.101
ini-0.098
 office-0.097
Token got
Token ID1392
Feature activation+3.86e-03

Pos logit contributions
ust+0.093
 much+0.091
irm+0.090
 possible+0.086
 drift+0.086
Neg logit contributions
 mill-0.059
ting-0.054
 laund-0.050
Why-0.049
 painter-0.048
Token out
Token ID503
Feature activation-1.21e-01

Pos logit contributions
 mill+0.006
ting+0.006
 painter+0.005
 office+0.005
 laund+0.005
Neg logit contributions
 drift-0.006
ust-0.006
 dives-0.006
irm-0.005
 hops-0.005
Token of
Token ID286
Feature activation-1.45e-02

Pos logit contributions
ust+0.164
irm+0.154
 dives+0.153
 drift+0.149
 much+0.144
Neg logit contributions
 mill-0.211
ting-0.209
 painter-0.198
 office-0.196
ini-0.192
Token the
Token ID262
Feature activation+5.90e-02

Pos logit contributions
 nap+0.021
 much+0.020
 drift+0.020
 dives+0.019
irm+0.019
Neg logit contributions
ting-0.025
ini-0.024
 mill-0.024
 painter-0.024
rom-0.023
Token pool
Token ID5933
Feature activation-5.60e-02

Pos logit contributions
 mill+0.079
ting+0.073
ini+0.070
 floors+0.068
 laund+0.068
Neg logit contributions
ust-0.105
 drift-0.104
irm-0.104
 much-0.102
 hops-0.100
Token,
Token ID11
Feature activation+4.73e-02

Pos logit contributions
 drift+0.064
 much+0.064
 dives+0.063
irm+0.063
 nap+0.061
Neg logit contributions
 mill-0.107
ting-0.106
 painter-0.102
ini-0.099
 office-0.098
Token she
Token ID673
Feature activation-4.92e-02

Pos logit contributions
 mill+0.103
ini+0.103
ting+0.102
 painter+0.099
cles+0.099
Neg logit contributions
 much-0.045
 nap-0.044
 drift-0.041
 dives-0.039
irm-0.037
Token,
Token ID11
Feature activation+5.01e-02

Pos logit contributions
ares+0.001
ust+0.001
ls+0.001
ver+0.001
irm+0.001
Neg logit contributions
 mill-0.001
 office-0.001
 painter-0.001
 porch-0.001
 paramed-0.001
Token long
Token ID890
Feature activation-3.60e-02

Pos logit contributions
 mill+0.096
ting+0.095
 painter+0.093
ini+0.091
 paramed+0.089
Neg logit contributions
 dives-0.054
 much-0.053
irm-0.052
ust-0.052
ls-0.052
Token sleep
Token ID3993
Feature activation-6.60e-02

Pos logit contributions
 much+0.027
 dives+0.026
ust+0.026
irm+0.025
 drift+0.025
Neg logit contributions
 mill-0.082
ting-0.079
 painter-0.078
ini-0.077
 paramed-0.076
Token.
Token ID13
Feature activation-4.30e-03

Pos logit contributions
 much+0.074
 drift+0.072
 nap+0.071
 dives+0.069
irm+0.067
Neg logit contributions
 mill-0.129
ting-0.128
 painter-0.126
ini-0.124
 paramed-0.120
Token<|endoftext|>
Token ID50256
Feature activation+6.21e-02

Pos logit contributions
ls+0.007
 dives+0.006
irm+0.006
 hops+0.006
 nap+0.006
Neg logit contributions
ting-0.007
 mill-0.007
 office-0.007
 painter-0.006
Why-0.006
TokenOnce
Token ID7454
Feature activation-1.23e-01

Pos logit contributions
 mill+0.099
ting+0.099
 painter+0.095
ini+0.093
 paramed+0.090
Neg logit contributions
 much-0.091
 drift-0.089
 dives-0.089
 nap-0.089
irm-0.086
Token there
Token ID612
Feature activation-4.12e-02

Pos logit contributions
 dives+0.160
irm+0.159
 drift+0.156
 fins+0.146
 hops+0.146
Neg logit contributions
 mill-0.215
 painter-0.206
 office-0.198
ini-0.196
rom-0.194
Token was
Token ID373
Feature activation+2.77e-02

Pos logit contributions
 much+0.059
 nap+0.057
 drift+0.057
 dives+0.057
irm+0.055
Neg logit contributions
ting-0.068
 mill-0.068
 painter-0.065
ini-0.064
 paramed-0.062
Token a
Token ID257
Feature activation-7.69e-02

Pos logit contributions
 mill+0.053
ting+0.052
 painter+0.051
ini+0.050
 paramed+0.049
Neg logit contributions
 dives-0.031
 drift-0.031
irm-0.030
 nap-0.029
 much-0.029
Token boy
Token ID2933
Feature activation+2.67e-02

Pos logit contributions
 much+0.099
 nap+0.097
 drift+0.093
 dives+0.093
 swim+0.089
Neg logit contributions
ting-0.142
 mill-0.135
ini-0.130
rom-0.128
keeping-0.128
Token named
Token ID3706
Feature activation+4.04e-02

Pos logit contributions
ting+0.044
 mill+0.042
 pavement+0.041
ini+0.041
 floors+0.041
Neg logit contributions
 much-0.041
 nap-0.038
 swim-0.037
 dive-0.036
 drift-0.035
Token!
Token ID0
Feature activation+2.15e-02

Pos logit contributions
 much+0.003
 dives+0.003
 drift+0.003
 nap+0.003
irm+0.003
Neg logit contributions
ting-0.005
 mill-0.005
 painter-0.005
ini-0.005
 office-0.005
Token<|endoftext|>
Token ID50256
Feature activation+6.18e-02

Pos logit contributions
ting+0.037
 mill+0.036
 painter+0.035
rom+0.034
ini+0.034
Neg logit contributions
 dives-0.030
 drift-0.028
irm-0.028
 dive-0.028
 nap-0.027
TokenOnce
Token ID7454
Feature activation-1.24e-01

Pos logit contributions
 mill+0.107
ting+0.106
 painter+0.102
ini+0.101
 paramed+0.098
Neg logit contributions
 much-0.082
 drift-0.080
 dives-0.080
 nap-0.080
irm-0.078
Token upon
Token ID2402
Feature activation+3.68e-02

Pos logit contributions
 dives+0.163
 drift+0.161
irm+0.161
 hops+0.148
 fins+0.147
Neg logit contributions
 mill-0.213
 painter-0.208
ini-0.196
 office-0.194
rom-0.192
Token a
Token ID257
Feature activation-7.35e-02

Pos logit contributions
Why+0.052
ting+0.052
 painter+0.051
 mill+0.050
 office+0.049
Neg logit contributions
 dives-0.056
ls-0.056
irm-0.055
 drift-0.054
 hops-0.053
Token time
Token ID640
Feature activation-1.01e-01

Pos logit contributions
ls+0.092
ares+0.085
 dives+0.084
 drift+0.079
 disappears+0.077
Neg logit contributions
 mill-0.124
 painter-0.121
Why-0.119
 porch-0.116
 neighborhood-0.116
Token there
Token ID612
Feature activation-3.86e-02

Pos logit contributions
irm+0.176
 drift+0.172
 dives+0.161
fy+0.158
ust+0.157
Neg logit contributions
 mill-0.135
 painter-0.125
ting-0.122
 office-0.117
Why-0.113
Token lived
Token ID5615
Feature activation+6.51e-02

Pos logit contributions
 much+0.054
 nap+0.053
 drift+0.053
 dives+0.052
irm+0.050
Neg logit contributions
ting-0.065
 mill-0.064
 painter-0.062
ini-0.061
 paramed-0.060
Token a
Token ID257
Feature activation-7.35e-02

Pos logit contributions
ting+0.126
 mill+0.125
 painter+0.122
ini+0.118
 office+0.116
Neg logit contributions
 drift-0.073
 dives-0.072
 much-0.070
irm-0.069
 nap-0.068
Token baby
Token ID5156
Feature activation+1.92e-02

Pos logit contributions
 much+0.095
 drift+0.094
 dives+0.093
 nap+0.092
irm+0.091
Neg logit contributions
 mill-0.129
ting-0.129
 painter-0.125
ini-0.122
 paramed-0.118
Token.
Token ID13
Feature activation-1.36e-02

Pos logit contributions
 pavement+0.033
years+0.033
 laund+0.033
Fighting+0.033
cles+0.033
Neg logit contributions
 much-0.025
 nap-0.024
 swim-0.023
 they-0.022
ls-0.022
TokenWhen
Token ID2215
Feature activation-7.08e-02

Pos logit contributions
 much+0.071
 drift+0.065
 nap+0.064
 dives+0.063
ust+0.062
Neg logit contributions
ting-0.097
 mill-0.094
 painter-0.091
ini-0.089
 pavement-0.086
Token they
Token ID484
Feature activation-7.84e-02

Pos logit contributions
irm+0.097
 dives+0.093
ust+0.093
 hops+0.090
 nap+0.086
Neg logit contributions
 mill-0.127
ting-0.123
 laund-0.115
 painter-0.115
 office-0.115
Token got
Token ID1392
Feature activation+4.34e-03

Pos logit contributions
irm+0.129
ls+0.123
 much+0.122
ust+0.119
 dives+0.117
Neg logit contributions
ting-0.109
 mill-0.109
 painter-0.101
years-0.099
 laund-0.097
Token on
Token ID319
Feature activation-7.57e-02

Pos logit contributions
 mill+0.007
ting+0.006
 painter+0.006
 office+0.006
years+0.006
Neg logit contributions
 drift-0.007
ust-0.006
irm-0.006
 hops-0.006
 nap-0.006
Token,
Token ID11
Feature activation+4.74e-02

Pos logit contributions
 much+0.107
 drift+0.104
 nap+0.104
 dives+0.104
irm+0.098
Neg logit contributions
ting-0.127
ini-0.124
 mill-0.123
 painter-0.123
cut-0.117
Token they
Token ID484
Feature activation-7.86e-02

Pos logit contributions
 mill+0.080
ting+0.079
 painter+0.075
rom+0.074
 laund+0.073
Neg logit contributions
 dives-0.066
irm-0.065
 nap-0.063
 hops-0.063
 drift-0.063
Token saw
Token ID2497
Feature activation+1.19e-01

Pos logit contributions
 much+0.119
 drift+0.117
 dives+0.117
 nap+0.116
irm+0.114
Neg logit contributions
ting-0.122
 mill-0.122
 painter-0.116
ini-0.115
 paramed-0.110
Token a
Token ID257
Feature activation-7.23e-02

Pos logit contributions
ini+0.270
 painter+0.257
ting+0.253
 paramed+0.245
 mill+0.244
Neg logit contributions
 much-0.125
 nap-0.098
 dives-0.097
 they-0.095
 drift-0.090
Token big
Token ID1263
Feature activation+1.39e-02

Pos logit contributions
 much+0.085
 nap+0.083
 dives+0.075
 swim+0.074
 drift+0.074
Neg logit contributions
ting-0.153
keeping-0.141
ini-0.138
years-0.137
organized-0.137
Token box
Token ID3091
Feature activation+1.71e-03

Pos logit contributions
ting+0.026
ini+0.025
years+0.024
keeping+0.024
rom+0.024
Neg logit contributions
 nap-0.019
 dives-0.018
 much-0.018
 dive-0.018
 swim-0.017
Token in
Token ID287
Feature activation-4.63e-02

Pos logit contributions
 mill+0.003
ting+0.003
 painter+0.003
ini+0.003
 office+0.003
Neg logit contributions
 much-0.002
 dives-0.002
 drift-0.002
 nap-0.002
irm-0.002