TOP ACTIVATIONS
MAX = 0.124

hard work , bravery and cle vern ess ! From now
friends . Everyone praised her cle vern ess and eventually Jan
. Jack had used his cle vern ess and creat iveness
got cle vere r and cle vere r . Eventually the
mouse was known as the cle ve rest mouse in the
day and the rabbit got cle vere r and cle vere
engine was known as the cle ve rest engine in all
John chose some carrots and cuc umbers to take home .
praised for his obedience and cle vern ess . <|endoftext|> Once
animals said , " The cle ve rest sheep ever !"
fountain burst out with the cle arest water they had ever
a moment of independence and cle vern ess , she used
continued to scatter the bread cr umbs . <|endoftext|> Once upon
two friends laughed at their cle vern ess and went off
and down her street . Everywhere she went , people smiled
So , he used his cle vern ess to find his
lin g to be the cle ve rest . But in
each and every country . Everywhere it visited , it left
having lots of fun . Everywhere she went , she showed
woman became insepar able . Everywhere the woman went , the

MIN ACTIVATIONS
MIN = -0.095

A ww , okay . â
?" The re ind eer nods again and win
Once there was a re ind eer named S illy .
â C ind y said yes , so
there was a poor re ind eer . He didn 't
I play with your re ind eer ?" Ben asks Lily
merry way . <|endoftext|> C ind y and her mom my
. " E ww , this fish is sm
s okay sweet heart . How about I give
. The re ind eer and the girl became
. She hugs the re ind eer . "
She picks up the re ind eer . She looks at
!" " A ww !" Lily says . "
But I like your re ind eer . It is so
at it said " ew ww !" One
a sle igh and re ind eer who fly in the
a big group of re ind eer . It was a
eer away . The re ind eer said goodbye to the
there was a happy re ind eer named Rudolph . He
began . C ind y tried to make Peter

INTERVAL 0.069 - 0.124
CONTAINS 0.002%

for someone to add . Everywhere he looked there were no
was overwhelmed with joy ! Everywhere she looked , she could
ribbon trailing behind her . Everywhere she went , the ribbon
and took it everywhere . Everywhere the little girl went she
, became best friends . Everywhere the driver went , Sack

INTERVAL 0.014 - 0.069
CONTAINS 13.057%

! Hur ry up !" Tim shouted . He was almost
find some cheap steam . His friends got very excited and
in herself , and it showed in her painting . She
she smiled right back . From then on , Hannah was
go for a ride , so he went to the st

INTERVAL -0.041 - 0.014
CONTAINS 86.528%

to our show !" They start to act like a pirate
Lily 's mom my asked her to write her name on
and ran to the grass . They saw another boy playing
, the rabbit went back to the forest and looked for
the girl became still . Instead of noise , she made

INTERVAL -0.095 - -0.041
CONTAINS 0.413%

noise . " Stop it , Tom !" L
, there was a k ang aroo . He was very
but she was very hard working . She loved to pick
and Tom . S ara and Tom are scared
the rocket . They count down from ten to zero and
Token hard
Token ID1327
Feature activation-1.21e-04

Pos logit contributions
ha+0.003
Little+0.002
 few+0.002
www+0.002
unt+0.002
Neg logit contributions
ishes-0.003
unes-0.003
anks-0.003
ware-0.003
 fault-0.003
Token work
Token ID670
Feature activation-3.23e-03

Pos logit contributions
ha+0.000
Little+0.000
www+0.000
Time+0.000
ww+0.000
Neg logit contributions
ware-0.000
anks-0.000
ishes-0.000
unes-0.000
 fault-0.000
Token,
Token ID11
Feature activation+4.02e-03

Pos logit contributions
ha+0.005
Little+0.005
www+0.005
ww+0.004
Time+0.004
Neg logit contributions
ishes-0.007
ware-0.007
anks-0.007
unes-0.007
 fault-0.007
Token bravery
Token ID37808
Feature activation+2.29e-03

Pos logit contributions
ware+0.008
ishes+0.008
unes+0.008
anks+0.008
 fault+0.007
Neg logit contributions
ha-0.007
www-0.006
Little-0.006
ww-0.006
unt-0.006
Token and
Token ID290
Feature activation+6.29e-03

Pos logit contributions
ishes+0.005
anks+0.005
ware+0.005
unes+0.005
 fault+0.005
Neg logit contributions
ha-0.004
Little-0.003
www-0.003
ww-0.003
Time-0.003
Token cle
Token ID1190
Feature activation+8.30e-02

Pos logit contributions
ishes+0.013
anks+0.013
ware+0.013
unes+0.013
 fault+0.013
Neg logit contributions
ha-0.010
Little-0.009
www-0.009
ww-0.009
unt-0.009
Tokenvern
Token ID933
Feature activation+6.66e-02

Pos logit contributions
ishes+0.124
ware+0.120
unes+0.120
anks+0.118
 fault+0.115
Neg logit contributions
ha-0.195
Little-0.175
www-0.171
 few-0.169
ww-0.168
Tokeness
Token ID408
Feature activation-4.00e-02

Pos logit contributions
ishes+0.143
ware+0.141
unes+0.138
anks+0.135
 fault+0.131
Neg logit contributions
ha-0.112
www-0.093
Little-0.093
ww-0.092
 few-0.092
Token!
Token ID0
Feature activation+1.96e-02

Pos logit contributions
ha+0.068
Little+0.059
www+0.057
 few+0.056
ww+0.056
Neg logit contributions
ishes-0.082
ware-0.082
anks-0.082
unes-0.081
 fault-0.079
Token From
Token ID3574
Feature activation+2.41e-02

Pos logit contributions
anks+0.043
ishes+0.043
unes+0.043
ware+0.043
 fault+0.042
Neg logit contributions
ha-0.030
Little-0.026
www-0.025
ww-0.024
 few-0.024
Token now
Token ID783
Feature activation+1.11e-02

Pos logit contributions
ishes+0.049
anks+0.048
ware+0.048
unes+0.047
 fault+0.046
Neg logit contributions
ha-0.040
Little-0.034
Time-0.034
www-0.033
 few-0.033
Token friends
Token ID2460
Feature activation+9.35e-03

Pos logit contributions
ishes+0.017
ware+0.017
unes+0.017
anks+0.017
 fault+0.017
Neg logit contributions
ha-0.015
Little-0.013
www-0.012
 few-0.012
ww-0.012
Token.
Token ID13
Feature activation+1.06e-02

Pos logit contributions
ishes+0.020
ware+0.020
anks+0.020
unes+0.020
 fault+0.019
Neg logit contributions
ha-0.015
Little-0.013
www-0.013
 few-0.012
ww-0.012
Token Everyone
Token ID11075
Feature activation-2.06e-02

Pos logit contributions
ware+0.026
ishes+0.026
unes+0.025
anks+0.025
 fault+0.024
Neg logit contributions
ha-0.014
Little-0.013
Time-0.012
unt-0.012
 few-0.011
Token praised
Token ID15342
Feature activation+1.36e-02

Pos logit contributions
ha+0.030
Little+0.026
www+0.025
ww+0.024
 few+0.024
Neg logit contributions
ishes-0.047
anks-0.047
unes-0.047
ware-0.047
 fault-0.046
Token her
Token ID607
Feature activation+3.67e-03

Pos logit contributions
ishes+0.032
ware+0.032
unes+0.032
anks+0.032
 fault+0.031
Neg logit contributions
ha-0.019
Little-0.016
www-0.015
 few-0.015
Time-0.015
Token cle
Token ID1190
Feature activation+8.20e-02

Pos logit contributions
ishes+0.007
anks+0.007
ware+0.007
unes+0.007
 fault+0.007
Neg logit contributions
ha-0.007
Little-0.006
www-0.006
Time-0.006
ww-0.006
Tokenvern
Token ID933
Feature activation+7.25e-02

Pos logit contributions
ishes+0.143
unes+0.137
ware+0.135
anks+0.133
 fault+0.129
Neg logit contributions
ha-0.178
Little-0.155
www-0.154
 few-0.151
ww-0.150
Tokeness
Token ID408
Feature activation-3.95e-02

Pos logit contributions
ishes+0.151
ware+0.149
unes+0.146
anks+0.144
 fault+0.139
Neg logit contributions
ha-0.124
Little-0.105
www-0.104
 few-0.103
ww-0.103
Token and
Token ID290
Feature activation+9.48e-03

Pos logit contributions
ha+0.074
Little+0.065
www+0.063
 few+0.062
ww+0.062
Neg logit contributions
ishes-0.075
ware-0.074
anks-0.074
unes-0.074
 fault-0.072
Token eventually
Token ID4191
Feature activation+1.14e-02

Pos logit contributions
ishes+0.021
anks+0.021
ware+0.021
unes+0.021
 fault+0.020
Neg logit contributions
ha-0.015
Little-0.013
www-0.012
ww-0.012
Time-0.012
Token Jan
Token ID2365
Feature activation-4.12e-03

Pos logit contributions
anks+0.024
ishes+0.024
unes+0.024
 fault+0.023
ware+0.023
Neg logit contributions
ha-0.017
Little-0.017
unt-0.015
Time-0.015
www-0.015
Token.
Token ID13
Feature activation+1.28e-02

Pos logit contributions
ishes+0.028
ware+0.028
anks+0.028
unes+0.027
 fault+0.027
Neg logit contributions
ha-0.021
Little-0.018
www-0.017
 few-0.017
ww-0.017
Token Jack
Token ID3619
Feature activation+1.27e-02

Pos logit contributions
ishes+0.032
ware+0.032
anks+0.031
unes+0.031
 fault+0.030
Neg logit contributions
ha-0.017
Little-0.015
Time-0.013
www-0.013
 few-0.013
Token had
Token ID550
Feature activation-1.35e-02

Pos logit contributions
ishes+0.026
ware+0.026
anks+0.026
unes+0.025
 fault+0.025
Neg logit contributions
ha-0.021
Little-0.018
www-0.018
Time-0.018
ww-0.018
Token used
Token ID973
Feature activation+1.64e-02

Pos logit contributions
ha+0.022
Little+0.020
unt+0.019
www+0.019
ww+0.019
Neg logit contributions
 fault-0.028
ware-0.027
anks-0.027
unes-0.027
ishes-0.027
Token his
Token ID465
Feature activation-1.34e-04

Pos logit contributions
ishes+0.041
anks+0.040
ware+0.040
unes+0.040
 fault+0.039
Neg logit contributions
ha-0.021
Little-0.018
 few-0.017
www-0.016
Time-0.016
Token cle
Token ID1190
Feature activation+8.04e-02

Pos logit contributions
ha+0.000
Little+0.000
 few+0.000
www+0.000
ww+0.000
Neg logit contributions
ishes-0.000
anks-0.000
unes-0.000
ware-0.000
 fault-0.000
Tokenvern
Token ID933
Feature activation+6.61e-02

Pos logit contributions
ishes+0.119
unes+0.113
ware+0.113
anks+0.109
 fault+0.105
Neg logit contributions
ha-0.192
www-0.170
Little-0.170
 few-0.169
ww-0.165
Tokeness
Token ID408
Feature activation-4.38e-02

Pos logit contributions
ishes+0.142
ware+0.140
unes+0.137
anks+0.133
els+0.130
Neg logit contributions
ha-0.110
www-0.093
 few-0.093
ww-0.092
Little-0.091
Token and
Token ID290
Feature activation+6.81e-03

Pos logit contributions
ha+0.067
Little+0.058
 few+0.056
www+0.055
ww+0.054
Neg logit contributions
anks-0.096
ware-0.096
ishes-0.096
unes-0.095
 fault-0.093
Token creat
Token ID1827
Feature activation+1.52e-02

Pos logit contributions
ishes+0.015
ware+0.015
anks+0.015
unes+0.015
 fault+0.015
Neg logit contributions
ha-0.010
Little-0.009
www-0.008
ww-0.008
Time-0.008
Tokeniveness
Token ID6517
Feature activation+1.10e-02

Pos logit contributions
anks+0.028
unes+0.027
ishes+0.027
ware+0.027
 fault+0.026
Neg logit contributions
ha-0.030
Little-0.026
www-0.026
ww-0.026
unt-0.026
Token got
Token ID1392
Feature activation+3.57e-03

Pos logit contributions
ishes+0.028
ware+0.028
anks+0.027
unes+0.027
 fault+0.027
Neg logit contributions
ha-0.018
Little-0.016
www-0.015
ww-0.015
Time-0.015
Token cle
Token ID1190
Feature activation+7.97e-02

Pos logit contributions
ware+0.008
unes+0.008
ishes+0.008
anks+0.008
 fault+0.008
Neg logit contributions
ha-0.005
 few-0.004
Little-0.004
Time-0.004
www-0.004
Tokenvere
Token ID4119
Feature activation+9.78e-03

Pos logit contributions
ishes+0.126
unes+0.120
ware+0.119
anks+0.116
 fault+0.114
Neg logit contributions
ha-0.183
Little-0.165
www-0.161
 few-0.159
ww-0.157
Tokenr
Token ID81
Feature activation+1.97e-02

Pos logit contributions
ishes+0.017
ware+0.017
anks+0.017
unes+0.017
 fault+0.017
Neg logit contributions
ha-0.020
Little-0.018
www-0.017
 few-0.017
ww-0.017
Token and
Token ID290
Feature activation+2.86e-03

Pos logit contributions
anks+0.041
ishes+0.041
ware+0.041
unes+0.041
 fault+0.041
Neg logit contributions
ha-0.033
Little-0.029
www-0.028
Time-0.027
ww-0.027
Token cle
Token ID1190
Feature activation+8.02e-02

Pos logit contributions
ishes+0.007
ware+0.006
anks+0.006
unes+0.006
 fault+0.006
Neg logit contributions
ha-0.004
Little-0.004
www-0.003
ww-0.003
Time-0.003
Tokenvere
Token ID4119
Feature activation+9.87e-03

Pos logit contributions
ishes+0.119
unes+0.114
ware+0.114
anks+0.108
 fault+0.107
Neg logit contributions
ha-0.194
Little-0.176
www-0.172
 few-0.170
ww-0.168
Tokenr
Token ID81
Feature activation+2.07e-02

Pos logit contributions
ishes+0.016
anks+0.016
ware+0.016
unes+0.016
 fault+0.016
Neg logit contributions
ha-0.021
Little-0.019
www-0.018
ww-0.018
Time-0.018
Token.
Token ID13
Feature activation+5.63e-03

Pos logit contributions
anks+0.042
ishes+0.041
 fault+0.041
ware+0.041
unes+0.041
Neg logit contributions
ha-0.036
Little-0.032
www-0.031
Time-0.031
ww-0.030
Token Eventually
Token ID16178
Feature activation+1.67e-02

Pos logit contributions
ishes+0.013
ware+0.013
unes+0.013
anks+0.013
 fault+0.012
Neg logit contributions
ha-0.008
Little-0.007
 few-0.007
Time-0.007
uren-0.007
Token the
Token ID262
Feature activation-3.09e-03

Pos logit contributions
ware+0.029
anks+0.029
unes+0.028
kins+0.028
ishes+0.027
Neg logit contributions
ha-0.028
While-0.026
Alice-0.026
Little-0.026
unt-0.026
Token mouse
Token ID10211
Feature activation+1.72e-02

Pos logit contributions
ishes+0.002
ware+0.002
anks+0.002
unes+0.002
 fault+0.002
Neg logit contributions
ha-0.002
www-0.001
Little-0.001
ww-0.001
Time-0.001
Token was
Token ID373
Feature activation+1.30e-02

Pos logit contributions
unes+0.041
anks+0.041
ware+0.041
ishes+0.041
 fault+0.039
Neg logit contributions
ha-0.024
Little-0.020
www-0.019
 few-0.019
ww-0.019
Token known
Token ID1900
Feature activation+6.73e-03

Pos logit contributions
 fault+0.024
ishes+0.024
anks+0.023
unes+0.023
ware+0.023
Neg logit contributions
ha-0.024
Little-0.021
www-0.021
ww-0.021
uren-0.020
Token as
Token ID355
Feature activation-1.34e-02

Pos logit contributions
ishes+0.014
ware+0.014
anks+0.014
unes+0.014
 fault+0.014
Neg logit contributions
ha-0.011
Little-0.009
www-0.009
 few-0.009
ww-0.009
Token the
Token ID262
Feature activation+5.14e-04

Pos logit contributions
ha+0.017
Little+0.014
www+0.013
ww+0.013
Time+0.013
Neg logit contributions
ishes-0.034
ware-0.033
anks-0.033
unes-0.033
 fault-0.033
Token cle
Token ID1190
Feature activation+8.01e-02

Pos logit contributions
ishes+0.001
anks+0.001
ware+0.001
unes+0.001
 fault+0.001
Neg logit contributions
ha-0.001
Little-0.001
www-0.001
ww-0.001
unt-0.001
Tokenve
Token ID303
Feature activation-4.87e-02

Pos logit contributions
ishes+0.155
ware+0.151
unes+0.150
anks+0.149
 fault+0.146
Neg logit contributions
ha-0.155
Little-0.133
www-0.132
 few-0.129
ww-0.128
Tokenrest
Token ID2118
Feature activation-3.09e-02

Pos logit contributions
ha+0.098
Little+0.093
uren+0.091
unt+0.091
Time+0.090
Neg logit contributions
ishes-0.078
ware-0.076
unes-0.075
aces-0.074
anks-0.074
Token mouse
Token ID10211
Feature activation+1.91e-02

Pos logit contributions
ha+0.053
Little+0.047
www+0.046
Time+0.045
uren+0.045
Neg logit contributions
ishes-0.064
anks-0.063
unes-0.062
ware-0.062
 fault-0.060
Token in
Token ID287
Feature activation+3.30e-02

Pos logit contributions
ishes+0.042
ware+0.042
anks+0.041
unes+0.041
 fault+0.040
Neg logit contributions
ha-0.030
Little-0.026
www-0.025
ww-0.024
Time-0.024
Token the
Token ID262
Feature activation+5.90e-04

Pos logit contributions
ishes+0.081
anks+0.081
ware+0.081
unes+0.080
 fault+0.079
Neg logit contributions
ha-0.043
Little-0.035
www-0.034
ww-0.033
unt-0.032
Token day
Token ID1110
Feature activation-1.44e-02

Pos logit contributions
ware+0.034
 fault+0.032
anks+0.031
unes+0.031
ishes+0.031
Neg logit contributions
ha-0.017
Little-0.015
T-0.015
Alice-0.014
www-0.014
Token and
Token ID290
Feature activation+3.15e-03

Pos logit contributions
ha+0.024
ww+0.021
www+0.021
unt+0.020
Little+0.020
Neg logit contributions
anks-0.029
ware-0.029
unes-0.029
ishes-0.028
aces-0.028
Token the
Token ID262
Feature activation-2.54e-03

Pos logit contributions
ishes+0.007
ware+0.007
anks+0.007
 fault+0.007
unes+0.006
Neg logit contributions
ha-0.005
Little-0.004
unt-0.004
ww-0.004
www-0.004
Token rabbit
Token ID22746
Feature activation+1.23e-02

Pos logit contributions
ha+0.004
Little+0.004
www+0.004
ww+0.004
Time+0.003
Neg logit contributions
ishes-0.005
ware-0.005
anks-0.005
unes-0.005
 fault-0.005
Token got
Token ID1392
Feature activation+3.57e-03

Pos logit contributions
ishes+0.028
ware+0.028
anks+0.027
unes+0.027
 fault+0.027
Neg logit contributions
ha-0.018
Little-0.016
www-0.015
ww-0.015
Time-0.015
Token cle
Token ID1190
Feature activation+7.97e-02

Pos logit contributions
ware+0.008
unes+0.008
ishes+0.008
anks+0.008
 fault+0.008
Neg logit contributions
ha-0.005
 few-0.004
Little-0.004
Time-0.004
www-0.004
Tokenvere
Token ID4119
Feature activation+9.78e-03

Pos logit contributions
ishes+0.126
unes+0.120
ware+0.119
anks+0.116
 fault+0.114
Neg logit contributions
ha-0.183
Little-0.165
www-0.161
 few-0.159
ww-0.157
Tokenr
Token ID81
Feature activation+1.97e-02

Pos logit contributions
ishes+0.017
ware+0.017
anks+0.017
unes+0.017
 fault+0.017
Neg logit contributions
ha-0.020
Little-0.018
www-0.017
 few-0.017
ww-0.017
Token and
Token ID290
Feature activation+2.86e-03

Pos logit contributions
anks+0.041
ishes+0.041
ware+0.041
unes+0.041
 fault+0.041
Neg logit contributions
ha-0.033
Little-0.029
www-0.028
Time-0.027
ww-0.027
Token cle
Token ID1190
Feature activation+8.02e-02

Pos logit contributions
ishes+0.007
ware+0.006
anks+0.006
unes+0.006
 fault+0.006
Neg logit contributions
ha-0.004
Little-0.004
www-0.003
ww-0.003
Time-0.003
Tokenvere
Token ID4119
Feature activation+9.87e-03

Pos logit contributions
ishes+0.119
unes+0.114
ware+0.114
anks+0.108
 fault+0.107
Neg logit contributions
ha-0.194
Little-0.176
www-0.172
 few-0.170
ww-0.168
Token engine
Token ID3113
Feature activation-1.70e-04

Pos logit contributions
ha+0.004
www+0.004
ww+0.004
uren+0.004
Alice+0.004
Neg logit contributions
 fault-0.003
ishes-0.003
anks-0.003
ware-0.003
 wounds-0.003
Token was
Token ID373
Feature activation+1.51e-02

Pos logit contributions
ha+0.000
Little+0.000
www+0.000
ww+0.000
Time+0.000
Neg logit contributions
ishes-0.000
anks-0.000
ware-0.000
 fault-0.000
unes-0.000
Token known
Token ID1900
Feature activation+1.02e-02

Pos logit contributions
 fault+0.028
ishes+0.026
anks+0.024
unes+0.024
ric+0.024
Neg logit contributions
ha-0.029
www-0.025
uren-0.025
ww-0.025
Little-0.024
Token as
Token ID355
Feature activation-1.59e-02

Pos logit contributions
anks+0.020
ishes+0.020
unes+0.020
 fault+0.020
ware+0.020
Neg logit contributions
ha-0.018
Little-0.016
www-0.016
Time-0.015
ww-0.015
Token the
Token ID262
Feature activation-1.75e-03

Pos logit contributions
ha+0.020
Little+0.016
www+0.016
ww+0.015
unt+0.015
Neg logit contributions
ishes-0.040
anks-0.039
ware-0.039
unes-0.039
 fault-0.039
Token cle
Token ID1190
Feature activation+7.94e-02

Pos logit contributions
ha+0.003
www+0.003
ww+0.003
uren+0.003
Little+0.003
Neg logit contributions
 fault-0.003
anks-0.003
ishes-0.003
unes-0.003
ware-0.003
Tokenve
Token ID303
Feature activation-4.62e-02

Pos logit contributions
ishes+0.131
ware+0.129
anks+0.128
unes+0.128
 fault+0.124
Neg logit contributions
ha-0.170
Little-0.153
www-0.148
ww-0.146
 few-0.146
Tokenrest
Token ID2118
Feature activation-3.21e-02

Pos logit contributions
ha+0.096
uren+0.094
Little+0.094
www+0.092
unt+0.091
Neg logit contributions
ishes-0.066
aces-0.065
unes-0.063
els-0.063
anks-0.062
Token engine
Token ID3113
Feature activation+2.32e-04

Pos logit contributions
ha+0.058
www+0.053
Little+0.053
uren+0.052
Alice+0.052
Neg logit contributions
ishes-0.061
anks-0.060
unes-0.060
ware-0.058
 fault-0.057
Token in
Token ID287
Feature activation+3.08e-02

Pos logit contributions
anks+0.000
ishes+0.000
ware+0.000
unes+0.000
 fault+0.000
Neg logit contributions
ha-0.000
Little-0.000
www-0.000
ww-0.000
Time-0.000
Token all
Token ID477
Feature activation-1.07e-02

Pos logit contributions
 fault+0.070
ware+0.069
anks+0.069
ishes+0.069
unes+0.067
Neg logit contributions
ha-0.045
www-0.038
ww-0.037
Little-0.036
uren-0.036
Token John
Token ID1757
Feature activation-4.12e-02

Pos logit contributions
ishes+0.014
ware+0.013
anks+0.013
unes+0.013
aces+0.012
Neg logit contributions
ha-0.015
Little-0.013
ww-0.013
Time-0.013
www-0.013
Token chose
Token ID7690
Feature activation-6.47e-03

Pos logit contributions
ha+0.062
Little+0.054
www+0.052
ww+0.052
Time+0.052
Neg logit contributions
ishes-0.093
ware-0.089
anks-0.089
unes-0.089
 fault-0.088
Token some
Token ID617
Feature activation+1.01e-02

Pos logit contributions
ha+0.008
 few+0.006
Little+0.006
www+0.006
Time+0.006
Neg logit contributions
anks-0.017
ishes-0.016
unes-0.016
ware-0.016
 fault-0.016
Token carrots
Token ID34397
Feature activation+1.97e-02

Pos logit contributions
anks+0.022
ishes+0.021
unes+0.020
 fault+0.020
ware+0.020
Neg logit contributions
ha-0.017
 few-0.015
Little-0.015
www-0.014
ww-0.014
Token and
Token ID290
Feature activation+3.87e-03

Pos logit contributions
anks+0.055
 fault+0.054
ishes+0.052
unes+0.052
 wounds+0.052
Neg logit contributions
ha-0.017
 few-0.015
"-0.014
ww-0.012
 long-0.012
Token cuc
Token ID38421
Feature activation+7.94e-02

Pos logit contributions
ishes+0.008
anks+0.008
ware+0.008
unes+0.008
 fault+0.008
Neg logit contributions
ha-0.006
Little-0.005
www-0.005
 few-0.005
ww-0.005
Tokenumbers
Token ID17024
Feature activation+1.93e-02

Pos logit contributions
ses+0.120
 fault+0.117
 wounds+0.103
neys+0.102
anks+0.100
Neg logit contributions
ha-0.188
 few-0.170
ww-0.165
www-0.162
T-0.159
Token to
Token ID284
Feature activation+1.33e-04

Pos logit contributions
anks+0.048
 fault+0.047
ishes+0.046
unes+0.046
ses+0.045
Neg logit contributions
ha-0.024
 few-0.020
"-0.018
Little-0.017
www-0.017
Token take
Token ID1011
Feature activation-2.76e-02

Pos logit contributions
anks+0.000
unes+0.000
ishes+0.000
 fault+0.000
ware+0.000
Neg logit contributions
ha-0.000
Little-0.000
 few-0.000
www-0.000
ww-0.000
Token home
Token ID1363
Feature activation-1.52e-02

Pos logit contributions
ha+0.045
Little+0.040
www+0.038
Time+0.038
ww+0.038
Neg logit contributions
ishes-0.058
ware-0.057
unes-0.057
anks-0.057
 fault-0.055
Token.
Token ID13
Feature activation+1.12e-02

Pos logit contributions
ha+0.023
 few+0.020
Little+0.020
www+0.018
T+0.018
Neg logit contributions
anks-0.034
unes-0.033
ishes-0.033
 fault-0.033
ware-0.033
Token praised
Token ID15342
Feature activation+1.19e-02

Pos logit contributions
ha+0.023
unt+0.019
Little+0.019
www+0.019
uren+0.019
Neg logit contributions
ishes-0.026
 fault-0.024
anks-0.024
ware-0.023
 depending-0.023
Token for
Token ID329
Feature activation+8.73e-03

Pos logit contributions
ishes+0.025
ware+0.025
anks+0.024
unes+0.024
 fault+0.024
Neg logit contributions
ha-0.020
Little-0.018
www-0.017
 few-0.017
ww-0.017
Token his
Token ID465
Feature activation-8.77e-04

Pos logit contributions
anks+0.016
ishes+0.016
ware+0.016
 fault+0.016
unes+0.016
Neg logit contributions
ha-0.017
Little-0.014
www-0.014
ww-0.014
unt-0.014
Token obedience
Token ID28777
Feature activation-7.54e-03

Pos logit contributions
ha+0.002
www+0.001
Little+0.001
ww+0.001
Alice+0.001
Neg logit contributions
 fault-0.002
anks-0.002
ishes-0.002
unes-0.002
ware-0.002
Token and
Token ID290
Feature activation+8.32e-03

Pos logit contributions
ha+0.012
Little+0.010
www+0.010
ww+0.010
Time+0.010
Neg logit contributions
ishes-0.016
anks-0.016
ware-0.016
unes-0.016
 fault-0.016
Token cle
Token ID1190
Feature activation+7.81e-02

Pos logit contributions
ishes+0.017
anks+0.017
ware+0.017
unes+0.017
 fault+0.017
Neg logit contributions
ha-0.014
Little-0.012
www-0.012
ww-0.011
Time-0.011
Tokenvern
Token ID933
Feature activation+6.84e-02

Pos logit contributions
ishes+0.109
ware+0.107
unes+0.105
anks+0.100
 fault+0.097
Neg logit contributions
ha-0.194
Little-0.174
www-0.173
 few-0.170
ww-0.168
Tokeness
Token ID408
Feature activation-3.91e-02

Pos logit contributions
ishes+0.144
ware+0.144
unes+0.140
anks+0.136
 fault+0.132
Neg logit contributions
ha-0.115
www-0.098
Little-0.097
ww-0.097
 few-0.096
Token.
Token ID13
Feature activation+1.27e-02

Pos logit contributions
ha+0.064
Little+0.056
www+0.053
 few+0.053
ww+0.052
Neg logit contributions
ishes-0.083
ware-0.083
anks-0.082
unes-0.082
 fault-0.080
Token<|endoftext|>
Token ID50256
Feature activation+1.37e-02

Pos logit contributions
ware+0.031
ishes+0.030
anks+0.030
unes+0.030
 fault+0.029
Neg logit contributions
ha-0.018
Little-0.015
www-0.014
unt-0.014
Time-0.014
TokenOnce
Token ID7454
Feature activation+1.03e-02

Pos logit contributions
ware+0.027
anks+0.027
unes+0.027
ishes+0.026
ests+0.026
Neg logit contributions
ha-0.023
www-0.021
ww-0.021
Little-0.020
uren-0.020
Token animals
Token ID4695
Feature activation-5.66e-03

Pos logit contributions
ha+0.000
www+0.000
ww+0.000
Little+0.000
unt+0.000
Neg logit contributions
ware-0.000
ishes-0.000
unes-0.000
 fault-0.000
anks-0.000
Token said
Token ID531
Feature activation-8.43e-03

Pos logit contributions
ha+0.009
Little+0.008
www+0.008
 few+0.008
ww+0.008
Neg logit contributions
ishes-0.012
anks-0.012
ware-0.012
unes-0.012
 fault-0.011
Token,
Token ID11
Feature activation+5.03e-03

Pos logit contributions
ha+0.013
Little+0.011
www+0.011
ww+0.010
Time+0.010
Neg logit contributions
ware-0.019
ishes-0.019
anks-0.019
unes-0.019
 fault-0.018
Token "
Token ID366
Feature activation+7.11e-03

Pos logit contributions
ware+0.010
unes+0.010
ishes+0.009
anks+0.009
 depending+0.009
Neg logit contributions
ha-0.009
www-0.008
ww-0.008
Little-0.008
Time-0.007
TokenThe
Token ID464
Feature activation+1.96e-02

Pos logit contributions
ishes+0.015
ware+0.015
unes+0.014
anks+0.014
ests+0.014
Neg logit contributions
ha-0.011
www-0.009
unt-0.009
 few-0.009
Little-0.009
Token cle
Token ID1190
Feature activation+7.80e-02

Pos logit contributions
ishes+0.041
anks+0.040
ware+0.040
unes+0.040
 fault+0.040
Neg logit contributions
ha-0.033
Little-0.028
www-0.027
ww-0.027
Time-0.026
Tokenve
Token ID303
Feature activation-5.22e-02

Pos logit contributions
ishes+0.151
ware+0.147
unes+0.146
anks+0.144
 fault+0.142
Neg logit contributions
ha-0.151
Little-0.131
www-0.130
ww-0.126
 few-0.126
Tokenrest
Token ID2118
Feature activation-3.06e-02

Pos logit contributions
ha+0.089
Little+0.085
uren+0.083
T+0.083
Time+0.083
Neg logit contributions
ware-0.094
ishes-0.094
anks-0.093
unes-0.093
aces-0.092
Token sheep
Token ID15900
Feature activation-1.24e-02

Pos logit contributions
ha+0.045
Little+0.040
Time+0.038
T+0.037
Alice+0.037
Neg logit contributions
ishes-0.068
anks-0.068
ware-0.068
unes-0.067
 fault-0.066
Token ever
Token ID1683
Feature activation+6.56e-03

Pos logit contributions
ha+0.019
Little+0.016
www+0.015
Time+0.015
ww+0.015
Neg logit contributions
ishes-0.028
anks-0.028
ware-0.027
unes-0.027
 fault-0.027
Token!"
Token ID2474
Feature activation+1.29e-02

Pos logit contributions
ishes+0.014
anks+0.014
unes+0.014
 fault+0.014
ware+0.013
Neg logit contributions
ha-0.010
Little-0.009
unt-0.009
Alice-0.009
uren-0.008
Token fountain
Token ID29420
Feature activation+3.94e-03

Pos logit contributions
ha+0.010
Little+0.009
www+0.009
ww+0.009
T+0.009
Neg logit contributions
ishes-0.007
ware-0.007
unes-0.007
anks-0.007
 fault-0.007
Token burst
Token ID11173
Feature activation+3.62e-03

Pos logit contributions
ware+0.009
ishes+0.009
unes+0.009
anks+0.009
 fault+0.008
Neg logit contributions
ha-0.006
Little-0.005
www-0.005
ww-0.005
unt-0.005
Token out
Token ID503
Feature activation-5.10e-03

Pos logit contributions
unes+0.007
ishes+0.007
ware+0.007
anks+0.007
 fault+0.007
Neg logit contributions
ha-0.006
Little-0.006
www-0.006
ww-0.005
unt-0.005
Token with
Token ID351
Feature activation+2.69e-02

Pos logit contributions
ha+0.009
www+0.007
Little+0.007
ww+0.007
unt+0.007
Neg logit contributions
ishes-0.011
ware-0.011
unes-0.011
anks-0.010
 fault-0.010
Token the
Token ID262
Feature activation-3.75e-03

Pos logit contributions
ishes+0.065
ware+0.065
anks+0.065
unes+0.064
 fault+0.063
Neg logit contributions
ha-0.036
Little-0.031
www-0.029
 few-0.029
ww-0.029
Token cle
Token ID1190
Feature activation+7.80e-02

Pos logit contributions
ha+0.006
Little+0.005
 few+0.005
www+0.005
ww+0.005
Neg logit contributions
anks-0.008
ishes-0.008
ware-0.008
unes-0.008
 fault-0.008
Tokenarest
Token ID12423
Feature activation-2.12e-03

Pos logit contributions
ishes+0.171
ware+0.165
unes+0.165
anks+0.165
 fault+0.160
Neg logit contributions
ha-0.132
www-0.111
Little-0.107
ww-0.107
 few-0.106
Token water
Token ID1660
Feature activation-5.75e-03

Pos logit contributions
ha+0.003
Little+0.003
 few+0.003
www+0.003
Time+0.003
Neg logit contributions
anks-0.004
ishes-0.004
ware-0.004
unes-0.004
 fault-0.004
Token they
Token ID484
Feature activation+1.15e-02

Pos logit contributions
ha+0.008
Little+0.007
www+0.007
 few+0.007
ww+0.007
Neg logit contributions
anks-0.013
ishes-0.013
unes-0.013
ware-0.013
 fault-0.013
Token had
Token ID550
Feature activation-6.59e-03

Pos logit contributions
anks+0.027
ware+0.027
unes+0.027
 fault+0.027
ishes+0.027
Neg logit contributions
ha-0.017
Little-0.014
www-0.013
 few-0.013
Time-0.013
Token ever
Token ID1683
Feature activation+7.40e-03

Pos logit contributions
ha+0.010
Little+0.009
www+0.009
ww+0.008
 few+0.008
Neg logit contributions
ishes-0.015
ware-0.014
anks-0.014
unes-0.014
 fault-0.014
Token a
Token ID257
Feature activation-6.13e-03

Pos logit contributions
ware+0.012
ishes+0.011
anks+0.011
unes+0.011
 fault+0.011
Neg logit contributions
ha-0.006
Little-0.005
www-0.005
ww-0.005
Time-0.004
Token moment
Token ID2589
Feature activation+3.62e-03

Pos logit contributions
ha+0.007
www+0.006
ww+0.006
 Kelly+0.006
Alice+0.005
Neg logit contributions
ware-0.013
kins-0.012
 mouths-0.012
ishes-0.012
bugs-0.012
Token of
Token ID286
Feature activation+8.55e-03

Pos logit contributions
ware+0.006
unes+0.006
anks+0.006
kins+0.006
els+0.006
Neg logit contributions
ha-0.006
www-0.006
ww-0.006
T-0.005
Charlie-0.005
Token independence
Token ID10404
Feature activation-1.08e-03

Pos logit contributions
ware+0.020
ishes+0.020
unes+0.019
els+0.019
anks+0.019
Neg logit contributions
ha-0.012
ww-0.010
www-0.010
Little-0.009
unt-0.009
Token and
Token ID290
Feature activation-1.82e-04

Pos logit contributions
ha+0.002
Little+0.001
www+0.001
ww+0.001
Time+0.001
Neg logit contributions
ware-0.002
ishes-0.002
unes-0.002
anks-0.002
 fault-0.002
Token cle
Token ID1190
Feature activation+7.73e-02

Pos logit contributions
ha+0.000
Little+0.000
Time+0.000
ww+0.000
www+0.000
Neg logit contributions
ishes-0.000
anks-0.000
ware-0.000
unes-0.000
 fault-0.000
Tokenvern
Token ID933
Feature activation+6.30e-02

Pos logit contributions
ishes+0.127
ware+0.120
unes+0.120
anks+0.115
 fault+0.115
Neg logit contributions
ha-0.180
Little-0.157
www-0.155
 few-0.153
unt-0.150
Tokeness
Token ID408
Feature activation-4.82e-02

Pos logit contributions
ishes+0.135
ware+0.134
unes+0.130
anks+0.127
 fault+0.124
Neg logit contributions
ha-0.104
www-0.088
 few-0.086
Little-0.086
ww-0.086
Token,
Token ID11
Feature activation-5.50e-03

Pos logit contributions
ha+0.075
Little+0.065
www+0.062
ww+0.061
 few+0.061
Neg logit contributions
ishes-0.106
anks-0.105
ware-0.105
unes-0.105
 fault-0.103
Token she
Token ID673
Feature activation+9.60e-05

Pos logit contributions
ha+0.008
www+0.007
ww+0.007
Little+0.007
Time+0.007
Neg logit contributions
ishes-0.012
ware-0.012
anks-0.012
unes-0.012
 fault-0.012
Token used
Token ID973
Feature activation+1.65e-02

Pos logit contributions
ishes+0.000
ware+0.000
anks+0.000
unes+0.000
 fault+0.000
Neg logit contributions
ha-0.000
Little-0.000
www-0.000
ww-0.000
 few-0.000
Token continued
Token ID3767
Feature activation+1.66e-02

Pos logit contributions
ha+0.008
Little+0.006
www+0.005
 few+0.005
T+0.005
Neg logit contributions
 fault-0.023
ishes-0.022
ware-0.022
unes-0.022
anks-0.022
Token to
Token ID284
Feature activation+1.17e-02

Pos logit contributions
ishes+0.034
ware+0.034
anks+0.034
unes+0.034
 fault+0.033
Neg logit contributions
ha-0.028
Little-0.025
www-0.024
ww-0.023
Time-0.023
Token scatter
Token ID41058
Feature activation+5.65e-03

Pos logit contributions
 fault+0.024
ware+0.024
ishes+0.024
anks+0.024
unes+0.023
Neg logit contributions
ha-0.020
www-0.017
Little-0.017
unt-0.016
ww-0.016
Token the
Token ID262
Feature activation+6.64e-03

Pos logit contributions
anks+0.014
ware+0.014
unes+0.014
ishes+0.014
 fault+0.014
Neg logit contributions
ha-0.007
 few-0.006
Little-0.006
Time-0.005
www-0.005
Token bread
Token ID8509
Feature activation+1.33e-02

Pos logit contributions
anks+0.013
unes+0.012
ishes+0.012
ware+0.012
 fault+0.012
Neg logit contributions
ha-0.012
Little-0.011
 few-0.011
ww-0.010
www-0.010
Tokencr
Token ID6098
Feature activation+7.72e-02

Pos logit contributions
anks+0.030
ware+0.030
unes+0.030
ishes+0.029
 fault+0.029
Neg logit contributions
ha-0.018
 few-0.016
 long-0.015
ww-0.014
Little-0.014
Tokenumbs
Token ID18146
Feature activation+2.77e-02

Pos logit contributions
ware+0.115
 fault+0.112
anks+0.110
ishes+0.108
unes+0.105
Neg logit contributions
ha-0.178
Little-0.159
 few-0.159
ww-0.158
www-0.156
Token.
Token ID13
Feature activation+1.59e-02

Pos logit contributions
ishes+0.061
ware+0.061
anks+0.060
unes+0.060
 fault+0.059
Neg logit contributions
ha-0.043
Little-0.037
www-0.036
 few-0.035
ww-0.035
Token<|endoftext|>
Token ID50256
Feature activation+1.74e-02

Pos logit contributions
unes+0.032
ishes+0.032
anks+0.032
ware+0.032
 fault+0.031
Neg logit contributions
ha-0.027
Little-0.024
www-0.023
ww-0.023
 few-0.022
TokenOnce
Token ID7454
Feature activation-2.09e-03

Pos logit contributions
anks+0.037
ishes+0.037
ware+0.037
unes+0.036
 fault+0.036
Neg logit contributions
ha-0.028
Little-0.024
www-0.024
ww-0.023
 few-0.023
Token upon
Token ID2402
Feature activation-1.17e-02

Pos logit contributions
ha+0.004
Little+0.003
www+0.003
ww+0.003
Time+0.003
Neg logit contributions
ishes-0.004
ware-0.004
anks-0.004
unes-0.004
 fault-0.004
Token two
Token ID734
Feature activation-1.95e-02

Pos logit contributions
ware+0.044
ishes+0.044
unes+0.043
anks+0.043
 fault+0.043
Neg logit contributions
ha-0.036
Little-0.031
www-0.030
ww-0.030
T-0.030
Token friends
Token ID2460
Feature activation+7.82e-03

Pos logit contributions
ha+0.035
Little+0.031
www+0.029
 few+0.029
ww+0.029
Neg logit contributions
ishes-0.039
ware-0.039
anks-0.039
unes-0.038
 fault-0.037
Token laughed
Token ID13818
Feature activation+1.65e-02

Pos logit contributions
ishes+0.015
ware+0.014
anks+0.014
unes+0.014
 fault+0.014
Neg logit contributions
ha-0.014
Little-0.013
www-0.012
ww-0.012
Time-0.012
Token at
Token ID379
Feature activation-3.82e-03

Pos logit contributions
anks+0.034
ishes+0.034
ware+0.034
unes+0.033
 fault+0.033
Neg logit contributions
ha-0.028
Little-0.024
www-0.024
ww-0.023
Time-0.023
Token their
Token ID511
Feature activation-7.73e-03

Pos logit contributions
ha+0.006
Little+0.006
 few+0.005
www+0.005
Time+0.005
Neg logit contributions
anks-0.008
ware-0.008
unes-0.008
ishes-0.008
 fault-0.008
Token cle
Token ID1190
Feature activation+7.68e-02

Pos logit contributions
ha+0.014
Little+0.012
www+0.012
ww+0.012
Time+0.012
Neg logit contributions
ishes-0.016
anks-0.015
ware-0.015
unes-0.015
 fault-0.015
Tokenvern
Token ID933
Feature activation+6.68e-02

Pos logit contributions
ishes+0.113
ware+0.110
unes+0.109
anks+0.106
 fault+0.100
Neg logit contributions
ha-0.184
Little-0.163
www-0.162
 few-0.159
ww-0.158
Tokeness
Token ID408
Feature activation-4.15e-02

Pos logit contributions
ishes+0.142
ware+0.141
unes+0.138
anks+0.135
 fault+0.130
Neg logit contributions
ha-0.111
www-0.093
 few-0.092
Little-0.092
ww-0.092
Token and
Token ID290
Feature activation+4.95e-03

Pos logit contributions
ha+0.058
Little+0.049
www+0.046
 few+0.046
ww+0.045
Neg logit contributions
ishes-0.098
ware-0.098
anks-0.097
unes-0.097
 fault-0.095
Token went
Token ID1816
Feature activation+1.79e-02

Pos logit contributions
ishes+0.010
anks+0.010
ware+0.009
 fault+0.009
unes+0.009
Neg logit contributions
ha-0.008
Little-0.008
Time-0.007
www-0.007
ww-0.007
Token off
Token ID572
Feature activation-4.46e-02

Pos logit contributions
anks+0.032
unes+0.030
ishes+0.030
 fault+0.029
ware+0.029
Neg logit contributions
ha-0.036
Little-0.035
ww-0.032
www-0.032
T-0.032
Token and
Token ID290
Feature activation+9.59e-03

Pos logit contributions
ha+0.024
Little+0.021
www+0.020
ww+0.020
unt+0.020
Neg logit contributions
ware-0.033
ishes-0.033
anks-0.033
unes-0.033
 fault-0.033
Token down
Token ID866
Feature activation-3.77e-02

Pos logit contributions
ishes+0.017
anks+0.017
ware+0.017
 fault+0.017
unes+0.017
Neg logit contributions
ha-0.019
Little-0.016
ww-0.016
www-0.016
T-0.016
Token her
Token ID607
Feature activation+1.51e-03

Pos logit contributions
ha+0.060
Little+0.052
www+0.051
ww+0.050
unt+0.049
Neg logit contributions
ware-0.080
anks-0.080
ishes-0.079
unes-0.079
 fault-0.079
Token street
Token ID4675
Feature activation-2.82e-02

Pos logit contributions
ishes+0.003
ware+0.003
anks+0.003
unes+0.003
 fault+0.003
Neg logit contributions
ha-0.002
Little-0.002
www-0.002
ww-0.002
Time-0.002
Token.
Token ID13
Feature activation+1.73e-02

Pos logit contributions
ha+0.049
Little+0.043
www+0.042
ww+0.041
Time+0.040
Neg logit contributions
ware-0.057
ishes-0.057
anks-0.057
unes-0.056
 fault-0.056
Token Everywhere
Token ID48201
Feature activation+7.64e-02

Pos logit contributions
unes+0.034
anks+0.033
ware+0.033
 fault+0.032
ishes+0.032
Neg logit contributions
ha-0.031
www-0.027
ww-0.027
Little-0.026
 few-0.025
Token she
Token ID673
Feature activation-5.22e-03

Pos logit contributions
ishes+0.158
ware+0.151
 fault+0.145
unes+0.144
 depending+0.140
Neg logit contributions
ha-0.109
www-0.098
Time-0.098
Little-0.097
ww-0.095
Token went
Token ID1816
Feature activation+1.28e-02

Pos logit contributions
ha+0.007
Little+0.006
www+0.006
ww+0.006
 few+0.006
Neg logit contributions
ishes-0.012
ware-0.012
anks-0.012
unes-0.012
 fault-0.012
Token,
Token ID11
Feature activation-4.21e-03

Pos logit contributions
unes+0.023
anks+0.023
aces+0.022
ware+0.021
els+0.021
Neg logit contributions
ha-0.021
T-0.020
ww-0.020
Little-0.020
www-0.019
Token people
Token ID661
Feature activation-1.31e-02

Pos logit contributions
ha+0.006
www+0.005
ww+0.005
unt+0.005
Little+0.005
Neg logit contributions
ishes-0.009
unes-0.009
ware-0.009
anks-0.009
aces-0.009
Token smiled
Token ID13541
Feature activation+6.16e-03

Pos logit contributions
ha+0.021
Little+0.019
www+0.019
ww+0.018
Time+0.018
Neg logit contributions
ishes-0.028
ware-0.027
anks-0.027
unes-0.027
 fault-0.026
Token So
Token ID1406
Feature activation+2.74e-02

Pos logit contributions
ware+0.035
ishes+0.034
unes+0.034
anks+0.033
 fault+0.033
Neg logit contributions
ha-0.024
Little-0.021
unt-0.020
www-0.020
 few-0.019
Token,
Token ID11
Feature activation+9.23e-04

Pos logit contributions
ishes+0.058
anks+0.055
ware+0.054
unes+0.052
kins+0.052
Neg logit contributions
ha-0.045
Time-0.038
www-0.037
Little-0.037
Alice-0.036
Token he
Token ID339
Feature activation+1.00e-02

Pos logit contributions
ishes+0.002
anks+0.002
ware+0.002
unes+0.002
rier+0.002
Neg logit contributions
ha-0.002
www-0.001
ww-0.001
Time-0.001
Little-0.001
Token used
Token ID973
Feature activation+1.43e-02

Pos logit contributions
ishes+0.022
ware+0.022
anks+0.022
unes+0.021
 fault+0.021
Neg logit contributions
ha-0.015
Little-0.013
www-0.013
ww-0.012
Time-0.012
Token his
Token ID465
Feature activation-8.42e-04

Pos logit contributions
ishes+0.035
anks+0.035
ware+0.035
unes+0.034
 fault+0.034
Neg logit contributions
ha-0.019
Little-0.016
 few-0.016
www-0.015
Time-0.015
Token cle
Token ID1190
Feature activation+7.60e-02

Pos logit contributions
ha+0.001
Little+0.001
 few+0.001
www+0.001
ww+0.001
Neg logit contributions
ishes-0.002
anks-0.002
unes-0.002
ware-0.002
 fault-0.002
Tokenvern
Token ID933
Feature activation+6.42e-02

Pos logit contributions
ishes+0.107
unes+0.102
ware+0.102
anks+0.095
 fault+0.091
Neg logit contributions
ha-0.190
www-0.170
 few-0.167
Little-0.166
ww-0.164
Tokeness
Token ID408
Feature activation-4.37e-02

Pos logit contributions
ishes+0.140
ware+0.139
unes+0.135
anks+0.129
els+0.128
Neg logit contributions
ha-0.106
www-0.089
 few-0.088
ww-0.088
unt-0.087
Token to
Token ID284
Feature activation+5.54e-03

Pos logit contributions
ha+0.064
Little+0.054
 few+0.054
www+0.050
ww+0.050
Neg logit contributions
ware-0.099
anks-0.099
unes-0.098
ishes-0.097
 fault-0.095
Token find
Token ID1064
Feature activation+1.59e-02

Pos logit contributions
anks+0.010
ishes+0.010
ware+0.010
unes+0.010
 fault+0.010
Neg logit contributions
ha-0.010
Little-0.009
ww-0.009
www-0.009
Time-0.009
Token his
Token ID465
Feature activation-3.58e-05

Pos logit contributions
ishes+0.037
ware+0.037
anks+0.037
unes+0.036
 fault+0.036
Neg logit contributions
ha-0.023
Little-0.020
www-0.019
ww-0.018
Time-0.018
Tokenlin
Token ID2815
Feature activation-8.72e-03

Pos logit contributions
ishes+0.016
ware+0.015
anks+0.015
unes+0.015
 fault+0.014
Neg logit contributions
ha-0.017
Little-0.015
www-0.015
Time-0.015
 few-0.015
Token g
Token ID308
Feature activation+7.55e-03

Pos logit contributions
ha+0.015
Little+0.013
www+0.013
ww+0.012
 few+0.012
Neg logit contributions
ishes-0.018
ware-0.018
anks-0.018
unes-0.018
 fault-0.017
Tokento
Token ID1462
Feature activation-4.11e-02

Pos logit contributions
ishes+0.011
ware+0.011
 fault+0.010
anks+0.010
unes+0.010
Neg logit contributions
ha-0.017
www-0.015
ww-0.015
Little-0.015
 few-0.014
Token be
Token ID307
Feature activation-1.61e-02

Pos logit contributions
ha+0.081
Little+0.075
Time+0.073
unt+0.072
www+0.071
Neg logit contributions
anks-0.073
ishes-0.071
ware-0.071
unes-0.070
aces-0.068
Token the
Token ID262
Feature activation-1.59e-03

Pos logit contributions
ha+0.028
unt+0.025
Little+0.025
www+0.025
ww+0.024
Neg logit contributions
ishes-0.031
aces-0.031
anks-0.030
kins-0.029
unes-0.028
Token cle
Token ID1190
Feature activation+7.54e-02

Pos logit contributions
ha+0.003
Little+0.003
 few+0.003
www+0.003
ww+0.003
Neg logit contributions
ishes-0.003
anks-0.003
unes-0.003
ware-0.003
 fault-0.003
Tokenve
Token ID303
Feature activation-5.11e-02

Pos logit contributions
ishes+0.159
unes+0.155
ware+0.155
anks+0.154
 fault+0.150
Neg logit contributions
ha-0.131
Little-0.112
www-0.111
ww-0.108
 few-0.107
Tokenrest
Token ID2118
Feature activation-3.46e-02

Pos logit contributions
ha+0.097
Little+0.090
Time+0.086
ww+0.085
www+0.085
Neg logit contributions
ware-0.091
ishes-0.091
anks-0.090
unes-0.090
 fault-0.087
Token.
Token ID13
Feature activation+6.05e-03

Pos logit contributions
ha+0.058
Little+0.052
Time+0.049
www+0.049
ww+0.048
Neg logit contributions
ware-0.071
ishes-0.071
anks-0.071
unes-0.070
 fault-0.069
Token But
Token ID887
Feature activation+4.18e-03

Pos logit contributions
ishes+0.014
ware+0.014
anks+0.014
unes+0.014
 fault+0.014
Neg logit contributions
ha-0.008
Little-0.008
 few-0.007
Time-0.007
www-0.007
Token in
Token ID287
Feature activation+2.59e-02

Pos logit contributions
ishes+0.010
ware+0.010
anks+0.010
unes+0.010
 fault+0.009
Neg logit contributions
ha-0.006
Little-0.005
www-0.005
ww-0.005
Time-0.005
Token each
Token ID1123
Feature activation-1.00e-02

Pos logit contributions
ha+0.005
Little+0.004
www+0.004
ww+0.004
Time+0.004
Neg logit contributions
ishes-0.008
ware-0.008
anks-0.008
unes-0.008
 fault-0.008
Token and
Token ID290
Feature activation+3.45e-03

Pos logit contributions
ha+0.017
uren+0.015
www+0.015
ww+0.015
unt+0.015
Neg logit contributions
ware-0.020
ishes-0.019
els-0.018
anks-0.018
 fault-0.018
Token every
Token ID790
Feature activation+1.58e-02

Pos logit contributions
ishes+0.007
anks+0.007
ware+0.007
unes+0.007
 fault+0.007
Neg logit contributions
ha-0.006
Little-0.005
www-0.005
ww-0.005
Time-0.004
Token country
Token ID1499
Feature activation-7.90e-03

Pos logit contributions
ware+0.039
 fault+0.037
ishes+0.036
anks+0.035
unes+0.035
Neg logit contributions
ha-0.019
Little-0.016
T-0.016
uren-0.016
www-0.015
Token.
Token ID13
Feature activation+8.02e-03

Pos logit contributions
ha+0.012
Little+0.009
 few+0.009
www+0.009
Time+0.009
Neg logit contributions
unes-0.018
ishes-0.018
anks-0.018
ware-0.018
 fault-0.017
Token Everywhere
Token ID48201
Feature activation+7.45e-02

Pos logit contributions
ishes+0.018
ware+0.018
anks+0.017
unes+0.017
 fault+0.017
Neg logit contributions
ha-0.012
Little-0.011
Time-0.010
 few-0.010
unt-0.010
Token it
Token ID340
Feature activation+6.46e-03

Pos logit contributions
ishes+0.181
ware+0.178
unes+0.173
 fault+0.173
anks+0.172
Neg logit contributions
ha-0.095
Little-0.081
www-0.079
Time-0.077
ww-0.077
Token visited
Token ID8672
Feature activation+1.42e-02

Pos logit contributions
ishes+0.015
anks+0.015
ware+0.015
unes+0.015
 fault+0.015
Neg logit contributions
ha-0.009
Little-0.008
www-0.007
ww-0.007
Time-0.007
Token,
Token ID11
Feature activation-2.83e-03

Pos logit contributions
ishes+0.035
ware+0.035
anks+0.035
unes+0.034
 fault+0.034
Neg logit contributions
ha-0.019
Little-0.016
www-0.015
ww-0.015
Time-0.015
Token it
Token ID340
Feature activation+6.61e-03

Pos logit contributions
ha+0.004
Little+0.003
www+0.003
ww+0.003
Time+0.003
Neg logit contributions
ishes-0.007
ware-0.006
anks-0.006
unes-0.006
 fault-0.006
Token left
Token ID1364
Feature activation+4.34e-03

Pos logit contributions
ishes+0.015
anks+0.015
unes+0.015
ware+0.015
 fault+0.015
Neg logit contributions
ha-0.010
Little-0.008
www-0.008
 few-0.008
Time-0.008
Token having
Token ID1719
Feature activation-3.81e-03

Pos logit contributions
anks+0.011
ishes+0.010
ware+0.010
 fault+0.010
unes+0.010
Neg logit contributions
ha-0.009
Little-0.008
www-0.007
ww-0.007
Time-0.007
Token lots
Token ID6041
Feature activation-2.31e-02

Pos logit contributions
ha+0.006
www+0.005
Little+0.005
ww+0.005
unt+0.005
Neg logit contributions
 fault-0.008
aces-0.008
anks-0.008
ishes-0.008
unes-0.008
Token of
Token ID286
Feature activation+1.42e-02

Pos logit contributions
ha+0.040
Little+0.035
www+0.035
ww+0.034
uren+0.033
Neg logit contributions
anks-0.045
unes-0.045
ishes-0.045
ware-0.045
 fault-0.044
Token fun
Token ID1257
Feature activation+1.09e-02

Pos logit contributions
ishes+0.033
anks+0.033
ware+0.033
unes+0.033
 fault+0.033
Neg logit contributions
ha-0.020
Little-0.017
www-0.016
ww-0.016
Time-0.015
Token.
Token ID13
Feature activation+1.14e-02

Pos logit contributions
ishes+0.021
ware+0.021
anks+0.021
unes+0.021
 fault+0.021
Neg logit contributions
ha-0.020
Little-0.017
www-0.017
ww-0.016
Time-0.016
Token Everywhere
Token ID48201
Feature activation+7.44e-02

Pos logit contributions
anks+0.024
unes+0.024
ishes+0.023
ware+0.023
 fault+0.023
Neg logit contributions
ha-0.019
Little-0.017
www-0.016
ww-0.016
 few-0.015
Token she
Token ID673
Feature activation-9.67e-04

Pos logit contributions
ishes+0.166
ware+0.164
 fault+0.157
unes+0.156
anks+0.154
Neg logit contributions
ha-0.106
www-0.089
Little-0.088
unt-0.087
ww-0.087
Token went
Token ID1816
Feature activation+1.51e-02

Pos logit contributions
ha+0.001
Little+0.001
www+0.001
ww+0.001
unt+0.001
Neg logit contributions
ware-0.002
ishes-0.002
anks-0.002
unes-0.002
 fault-0.002
Token,
Token ID11
Feature activation-3.73e-03

Pos logit contributions
unes+0.030
anks+0.030
ware+0.029
ishes+0.028
aces+0.028
Neg logit contributions
ha-0.024
T-0.021
Little-0.021
ww-0.021
www-0.021
Token she
Token ID673
Feature activation-5.32e-04

Pos logit contributions
ha+0.005
www+0.004
ww+0.004
Little+0.004
unt+0.004
Neg logit contributions
ishes-0.009
ware-0.009
anks-0.009
 fault-0.009
unes-0.008
Token showed
Token ID3751
Feature activation+4.86e-02

Pos logit contributions
ha+0.001
Little+0.001
www+0.001
ww+0.001
Time+0.001
Neg logit contributions
ishes-0.001
ware-0.001
anks-0.001
unes-0.001
 fault-0.001
Token woman
Token ID2415
Feature activation-2.47e-02

Pos logit contributions
ha+0.003
Little+0.002
www+0.002
ww+0.002
Time+0.002
Neg logit contributions
ishes-0.004
ware-0.004
anks-0.004
unes-0.004
 fault-0.003
Token became
Token ID2627
Feature activation+1.93e-02

Pos logit contributions
ha+0.040
Little+0.034
 few+0.032
www+0.032
ww+0.032
Neg logit contributions
ware-0.053
anks-0.053
unes-0.053
ishes-0.053
 fault-0.051
Token insepar
Token ID50140
Feature activation+1.90e-03

Pos logit contributions
ware+0.041
ishes+0.041
anks+0.041
unes+0.041
 fault+0.040
Neg logit contributions
ha-0.032
Little-0.028
www-0.026
 few-0.026
Time-0.026
Tokenable
Token ID540
Feature activation-1.20e-02

Pos logit contributions
ishes+0.004
ware+0.004
anks+0.004
unes+0.004
 fault+0.004
Neg logit contributions
ha-0.003
Little-0.003
www-0.003
ww-0.003
 few-0.003
Token.
Token ID13
Feature activation+1.26e-02

Pos logit contributions
ha+0.017
 few+0.015
Little+0.014
www+0.013
T+0.013
Neg logit contributions
ware-0.028
unes-0.028
anks-0.027
ishes-0.027
 fault-0.026
Token Everywhere
Token ID48201
Feature activation+7.43e-02

Pos logit contributions
ware+0.028
ishes+0.028
anks+0.027
unes+0.027
rier+0.026
Neg logit contributions
ha-0.020
Little-0.018
unt-0.017
Time-0.016
 few-0.016
Token the
Token ID262
Feature activation-3.22e-03

Pos logit contributions
ishes+0.163
ware+0.159
unes+0.156
anks+0.153
 fault+0.152
Neg logit contributions
ha-0.115
Little-0.101
www-0.098
Time-0.096
ww-0.095
Token woman
Token ID2415
Feature activation-2.56e-02

Pos logit contributions
ha+0.006
www+0.005
ww+0.005
Alice+0.005
uren+0.005
Neg logit contributions
ishes-0.006
ware-0.006
unes-0.006
anks-0.006
rier-0.005
Token went
Token ID1816
Feature activation+1.51e-02

Pos logit contributions
ha+0.036
Little+0.030
www+0.028
 few+0.028
unt+0.028
Neg logit contributions
anks-0.061
unes-0.061
ishes-0.060
ware-0.060
 fault-0.059
Token,
Token ID11
Feature activation-1.30e-03

Pos logit contributions
ishes+0.033
anks+0.033
unes+0.033
ware+0.032
 fault+0.031
Neg logit contributions
ha-0.023
Little-0.020
www-0.019
T-0.019
ww-0.019
Token the
Token ID262
Feature activation-2.63e-03

Pos logit contributions
ha+0.002
Little+0.001
www+0.001
ww+0.001
Time+0.001
Neg logit contributions
ishes-0.003
ware-0.003
anks-0.003
unes-0.003
 fault-0.003
Token
Token ID95
Feature activation-4.71e-03

Pos logit contributions
ha+0.024
 few+0.021
Time+0.021
Little+0.021
Charlie+0.020
Neg logit contributions
anks-0.022
unes-0.022
ware-0.022
rier-0.021
ishes-0.021
Token
Token ID26391
Feature activation-2.98e-02

Pos logit contributions
ha+0.009
 few+0.008
Little+0.008
ww+0.007
 Kelly+0.007
Neg logit contributions
anks-0.009
unes-0.008
ware-0.008
aces-0.008
 fault-0.008
Token
Token ID129
Feature activation-1.95e-02

Pos logit contributions
ha+0.053
 few+0.049
Time+0.046
Little+0.044
www+0.043
Neg logit contributions
anks-0.055
unes-0.054
ishes-0.053
 fault-0.052
aces-0.052
Token
Token ID241
Feature activation+1.01e-03

Pos logit contributions
ha+0.038
Little+0.033
 few+0.033
ww+0.031
Time+0.031
Neg logit contributions
anks-0.037
unes-0.036
 fault-0.035
aces-0.033
ware-0.033
TokenA
Token ID32
Feature activation+2.31e-03

Pos logit contributions
 fault+0.002
ware+0.002
unes+0.002
 depending+0.002
anks+0.002
Neg logit contributions
ha-0.001
Little-0.001
unt-0.001
ww-0.001
uren-0.001
Tokenww
Token ID1383
Feature activation-9.01e-02

Pos logit contributions
ware+0.005
 fault+0.005
unes+0.004
rier+0.004
anks+0.004
Neg logit contributions
 Kelly-0.003
Charlie-0.003
Kelly-0.002
Sarah-0.002
T-0.002
Token,
Token ID11
Feature activation-1.73e-02

Pos logit contributions
Time+0.122
While+0.122
ha+0.121
Charlie+0.119
Sarah+0.117
Neg logit contributions
ishes-0.190
unes-0.189
anks-0.187
ware-0.185
aces-0.180
Token okay
Token ID8788
Feature activation-1.70e-02

Pos logit contributions
ha+0.029
Little+0.025
Charlie+0.024
T+0.024
While+0.024
Neg logit contributions
unes-0.033
ware-0.033
 fault-0.033
ishes-0.032
 depending-0.031
Token.
Token ID13
Feature activation-2.12e-03

Pos logit contributions
ha+0.034
Little+0.030
uren+0.030
Time+0.029
Charlie+0.029
Neg logit contributions
 fault-0.027
ishes-0.027
unes-0.027
anks-0.027
ware-0.026
Tokenâ
Token ID22940
Feature activation-1.79e-02

Pos logit contributions
ha+0.004
Little+0.004
www+0.003
ww+0.003
 few+0.003
Neg logit contributions
ishes-0.004
anks-0.004
ware-0.004
unes-0.004
 fault-0.004
Token
Token ID26391
Feature activation-3.03e-02

Pos logit contributions
ha+0.033
 few+0.031
T+0.030
Little+0.028
While+0.028
Neg logit contributions
anks-0.032
unes-0.031
ishes-0.029
aces-0.029
ware-0.029
Token?"
Token ID1701
Feature activation-1.48e-03

Pos logit contributions
ha+0.002
Little+0.002
www+0.002
ww+0.002
 few+0.002
Neg logit contributions
anks-0.004
ishes-0.004
unes-0.004
ware-0.004
 fault-0.004
Token
Token ID198
Feature activation+5.87e-04

Pos logit contributions
ha+0.002
Little+0.002
unt+0.001
www+0.001
 few+0.001
Neg logit contributions
ishes-0.004
anks-0.004
ware-0.004
unes-0.004
 fault-0.004
Token
Token ID198
Feature activation-1.32e-03

Pos logit contributions
ishes+0.001
ware+0.001
anks+0.001
unes+0.001
 fault+0.001
Neg logit contributions
ha-0.001
Little-0.001
www-0.001
ww-0.001
 few-0.001
TokenThe
Token ID464
Feature activation+9.80e-03

Pos logit contributions
ha+0.001
www+0.001
 few+0.001
ww+0.001
Time+0.001
Neg logit contributions
ware-0.003
ests-0.003
ishes-0.003
els-0.003
ases-0.003
Token re
Token ID302
Feature activation-3.73e-02

Pos logit contributions
ware+0.022
ishes+0.022
unes+0.021
anks+0.021
 fault+0.021
Neg logit contributions
ha-0.015
Little-0.013
www-0.012
ww-0.012
Time-0.012
Tokenind
Token ID521
Feature activation-8.86e-02

Pos logit contributions
ha+0.057
Little+0.047
www+0.047
unt+0.046
ww+0.046
Neg logit contributions
unes-0.086
anks-0.084
ishes-0.082
 fault-0.082
ware-0.080
Tokeneer
Token ID28153
Feature activation-2.27e-02

Pos logit contributions
ha+0.211
Little+0.186
unt+0.184
www+0.184
 few+0.182
Neg logit contributions
anks-0.122
unes-0.118
 fault-0.115
ishes-0.115
ware-0.110
Token nods
Token ID34833
Feature activation-1.68e-02

Pos logit contributions
ha+0.036
Little+0.031
www+0.029
 few+0.029
uren+0.029
Neg logit contributions
anks-0.050
unes-0.049
ishes-0.049
 fault-0.048
ware-0.048
Token again
Token ID757
Feature activation+7.01e-03

Pos logit contributions
ha+0.022
Little+0.019
 few+0.018
T+0.018
Time+0.018
Neg logit contributions
ishes-0.042
unes-0.041
anks-0.041
 fault-0.040
ware-0.040
Token and
Token ID290
Feature activation-5.00e-03

Pos logit contributions
ishes+0.015
anks+0.015
ware+0.015
unes+0.015
 fault+0.015
Neg logit contributions
ha-0.011
Little-0.010
www-0.009
 few-0.009
ww-0.009
Token win
Token ID1592
Feature activation-3.00e-03

Pos logit contributions
ha+0.008
Little+0.007
www+0.007
 few+0.007
Time+0.007
Neg logit contributions
unes-0.011
anks-0.011
ishes-0.010
ware-0.010
 fault-0.010
TokenOnce
Token ID7454
Feature activation-1.80e-03

Pos logit contributions
ishes+0.046
anks+0.046
ware+0.046
unes+0.046
 fault+0.045
Neg logit contributions
ha-0.030
Little-0.026
www-0.025
ww-0.025
 few-0.024
Token there
Token ID612
Feature activation-2.79e-03

Pos logit contributions
ha+0.003
Little+0.003
 few+0.002
www+0.002
ww+0.002
Neg logit contributions
ishes-0.004
anks-0.004
ware-0.004
unes-0.004
 fault-0.004
Token was
Token ID373
Feature activation+1.30e-05

Pos logit contributions
ha+0.003
Little+0.003
 few+0.002
Time+0.002
unt+0.002
Neg logit contributions
anks-0.008
ishes-0.008
unes-0.007
ware-0.007
aces-0.007
Token a
Token ID257
Feature activation-1.52e-02

Pos logit contributions
ishes+0.000
unes+0.000
ware+0.000
anks+0.000
 fault+0.000
Neg logit contributions
ha-0.000
Little-0.000
www-0.000
unt-0.000
ww-0.000
Token re
Token ID302
Feature activation-3.75e-02

Pos logit contributions
ha+0.020
Time+0.016
www+0.016
Alice+0.015
 Kelly+0.015
Neg logit contributions
ware-0.034
ishes-0.033
kins-0.032
rier-0.032
melon-0.032
Tokenind
Token ID521
Feature activation-8.74e-02

Pos logit contributions
ha+0.063
Little+0.055
www+0.053
Time+0.052
ww+0.052
Neg logit contributions
ishes-0.079
ware-0.078
anks-0.078
unes-0.077
 fault-0.076
Tokeneer
Token ID28153
Feature activation-2.45e-02

Pos logit contributions
ha+0.232
Little+0.212
www+0.207
ww+0.206
 few+0.205
Neg logit contributions
ishes-0.097
anks-0.097
ware-0.095
unes-0.095
 fault-0.091
Token named
Token ID3706
Feature activation-7.43e-03

Pos logit contributions
ha+0.041
Little+0.036
www+0.034
Time+0.034
ww+0.034
Neg logit contributions
ishes-0.051
ware-0.051
anks-0.051
unes-0.050
 fault-0.049
Token S
Token ID311
Feature activation+1.26e-02

Pos logit contributions
ha+0.015
Time+0.013
Little+0.013
www+0.013
ww+0.012
Neg logit contributions
ishes-0.013
unes-0.013
ware-0.013
anks-0.012
 fault-0.012
Tokenilly
Token ID6548
Feature activation-2.55e-02

Pos logit contributions
ishes+0.025
anks+0.025
ware+0.025
unes+0.024
 fault+0.024
Neg logit contributions
ha-0.023
Little-0.020
unt-0.019
www-0.019
T-0.018
Token.
Token ID13
Feature activation-1.22e-03

Pos logit contributions
ha+0.042
Little+0.037
www+0.035
 few+0.035
ww+0.035
Neg logit contributions
ishes-0.054
anks-0.053
unes-0.053
ware-0.053
 fault-0.052
Tokenâ
Token ID22940
Feature activation-1.78e-02

Pos logit contributions
ishes+0.027
ware+0.027
anks+0.027
unes+0.026
 fault+0.026
Neg logit contributions
ha-0.020
Little-0.018
www-0.017
 few-0.016
Time-0.016
Token
Token ID26391
Feature activation-2.58e-02

Pos logit contributions
ha+0.033
 few+0.029
Little+0.029
Time+0.027
ww+0.027
Neg logit contributions
anks-0.034
unes-0.033
ishes-0.033
ware-0.033
 fault-0.032
Token
Token ID198
Feature activation-3.68e-03

Pos logit contributions
ha+0.044
 few+0.040
Time+0.038
www+0.037
ww+0.037
Neg logit contributions
anks-0.051
unes-0.050
 fault-0.049
ishes-0.048
aces-0.047
Token
Token ID198
Feature activation-4.83e-03

Pos logit contributions
ha+0.006
ww+0.005
www+0.005
 few+0.005
uren+0.005
Neg logit contributions
unes-0.007
anks-0.007
ware-0.007
 fault-0.007
ests-0.007
TokenC
Token ID34
Feature activation-2.29e-02

Pos logit contributions
ha+0.005
 Kelly+0.005
ww+0.005
 Margaret+0.004
www+0.004
Neg logit contributions
ests-0.010
ware-0.010
 depending-0.010
unes-0.010
auc-0.010
Tokenind
Token ID521
Feature activation-8.73e-02

Pos logit contributions
ha+0.043
Little+0.038
 few+0.038
www+0.037
Time+0.037
Neg logit contributions
ishes-0.042
ware-0.042
anks-0.042
unes-0.042
 fault-0.040
Tokeny
Token ID88
Feature activation-3.00e-02

Pos logit contributions
ha+0.155
Little+0.139
www+0.134
Time+0.133
 few+0.132
Neg logit contributions
ishes-0.173
ware-0.171
unes-0.169
anks-0.168
 fault-0.164
Token said
Token ID531
Feature activation-1.96e-02

Pos logit contributions
ha+0.048
Time+0.044
Little+0.043
ww+0.043
www+0.042
Neg logit contributions
ware-0.059
ishes-0.058
anks-0.057
unes-0.056
aces-0.056
Token yes
Token ID3763
Feature activation+1.56e-02

Pos logit contributions
ha+0.028
Little+0.025
Time+0.024
www+0.024
ww+0.023
Neg logit contributions
ware-0.045
unes-0.044
anks-0.044
ishes-0.043
 fault-0.043
Token,
Token ID11
Feature activation-1.27e-02

Pos logit contributions
anks+0.033
unes+0.032
 fault+0.032
ware+0.031
ishes+0.031
Neg logit contributions
ha-0.025
Little-0.022
Time-0.022
ww-0.021
www-0.021
Token so
Token ID523
Feature activation+3.80e-02

Pos logit contributions
ha+0.023
Little+0.020
ww+0.020
www+0.019
Time+0.019
Neg logit contributions
unes-0.023
ware-0.023
anks-0.022
 depending-0.022
ishes-0.022
Token there
Token ID612
Feature activation-3.30e-03

Pos logit contributions
ha+0.016
www+0.015
Little+0.014
uren+0.014
unt+0.014
Neg logit contributions
ishes-0.020
ware-0.020
aces-0.019
 fault-0.019
kins-0.019
Token was
Token ID373
Feature activation+9.41e-04

Pos logit contributions
ha+0.005
Little+0.004
www+0.004
ww+0.004
Time+0.004
Neg logit contributions
ishes-0.008
anks-0.008
ware-0.008
unes-0.008
 fault-0.007
Token a
Token ID257
Feature activation-1.58e-02

Pos logit contributions
unes+0.002
ware+0.002
 fault+0.002
ishes+0.002
anks+0.002
Neg logit contributions
ha-0.001
www-0.001
Little-0.001
unt-0.001
uren-0.001
Token poor
Token ID3595
Feature activation-2.03e-02

Pos logit contributions
ha+0.021
Alice+0.017
Time+0.017
www+0.017
 Kelly+0.016
Neg logit contributions
ware-0.034
ishes-0.033
rier-0.032
kins-0.032
melon-0.031
Token re
Token ID302
Feature activation-3.53e-02

Pos logit contributions
ha+0.029
Little+0.026
Time+0.025
Alice+0.024
Charlie+0.023
Neg logit contributions
ware-0.048
ishes-0.046
unes-0.045
 fault-0.044
anks-0.044
Tokenind
Token ID521
Feature activation-8.69e-02

Pos logit contributions
ha+0.063
Little+0.055
Time+0.053
www+0.053
ww+0.052
Neg logit contributions
ishes-0.072
ware-0.070
anks-0.069
unes-0.069
 fault-0.067
Tokeneer
Token ID28153
Feature activation-2.41e-02

Pos logit contributions
ha+0.234
Little+0.215
www+0.210
ww+0.208
 few+0.208
Neg logit contributions
ishes-0.094
ware-0.092
anks-0.092
unes-0.091
 fault-0.087
Token.
Token ID13
Feature activation-1.09e-03

Pos logit contributions
ha+0.041
Little+0.035
www+0.034
Time+0.034
ww+0.033
Neg logit contributions
ishes-0.050
ware-0.050
anks-0.050
unes-0.049
 fault-0.048
Token He
Token ID679
Feature activation+4.90e-03

Pos logit contributions
ha+0.002
Little+0.002
www+0.002
ww+0.002
Time+0.002
Neg logit contributions
unes-0.002
ware-0.002
anks-0.002
ishes-0.002
 fault-0.002
Token didn
Token ID1422
Feature activation-1.24e-02

Pos logit contributions
ware+0.010
ishes+0.010
anks+0.010
unes+0.010
 fault+0.009
Neg logit contributions
ha-0.008
Little-0.007
www-0.007
Time-0.007
ww-0.007
Token't
Token ID470
Feature activation-3.45e-02

Pos logit contributions
ha+0.017
Little+0.015
www+0.014
ww+0.014
Time+0.013
Neg logit contributions
anks-0.029
ware-0.029
unes-0.029
ishes-0.029
 fault-0.028
Token I
Token ID314
Feature activation-3.07e-02

Pos logit contributions
ha+0.024
Little+0.020
unt+0.019
www+0.019
ww+0.019
Neg logit contributions
ishes-0.046
anks-0.046
ware-0.045
unes-0.045
 fault-0.044
Token play
Token ID711
Feature activation-1.72e-02

Pos logit contributions
ha+0.043
Little+0.036
 few+0.034
www+0.034
ww+0.033
Neg logit contributions
ishes-0.074
ware-0.074
unes-0.073
anks-0.073
 fault-0.072
Token with
Token ID351
Feature activation+1.24e-02

Pos logit contributions
ha+0.026
Little+0.022
 few+0.021
www+0.021
ww+0.020
Neg logit contributions
ishes-0.040
unes-0.040
anks-0.039
 fault-0.039
ware-0.039
Token your
Token ID534
Feature activation-2.42e-02

Pos logit contributions
ishes+0.029
ware+0.029
unes+0.029
anks+0.029
 fault+0.028
Neg logit contributions
ha-0.017
Little-0.015
www-0.014
Time-0.014
ww-0.014
Token re
Token ID302
Feature activation-3.51e-02

Pos logit contributions
ha+0.036
 few+0.033
 little+0.032
 long+0.031
ww+0.030
Neg logit contributions
anks-0.052
unes-0.052
ishes-0.051
ware-0.050
 fault-0.049
Tokenind
Token ID521
Feature activation-8.64e-02

Pos logit contributions
ha+0.052
ww+0.044
 few+0.043
unt+0.043
www+0.043
Neg logit contributions
unes-0.081
anks-0.079
ware-0.078
 fault-0.077
 depending-0.075
Tokeneer
Token ID28153
Feature activation-1.92e-02

Pos logit contributions
ha+0.202
ww+0.178
 few+0.177
unt+0.177
www+0.175
Neg logit contributions
anks-0.122
unes-0.117
 fault-0.114
ware-0.113
ishes-0.112
Token?"
Token ID1701
Feature activation-2.65e-03

Pos logit contributions
ha+0.032
Little+0.028
www+0.026
 few+0.026
ww+0.026
Neg logit contributions
ware-0.040
unes-0.040
ishes-0.040
anks-0.040
 fault-0.039
Token Ben
Token ID3932
Feature activation+2.15e-02

Pos logit contributions
ha+0.004
Little+0.003
www+0.003
ww+0.003
Time+0.003
Neg logit contributions
ishes-0.006
ware-0.006
anks-0.006
unes-0.006
 fault-0.006
Token asks
Token ID7893
Feature activation-2.01e-02

Pos logit contributions
unes+0.057
anks+0.056
ishes+0.056
 fault+0.055
ware+0.055
Neg logit contributions
ha-0.024
Little-0.020
unt-0.019
 few-0.018
uren-0.018
Token Lily
Token ID20037
Feature activation+9.84e-03

Pos logit contributions
ha+0.028
Little+0.024
 few+0.023
T+0.022
 Kelly+0.022
Neg logit contributions
ishes-0.048
 fault-0.047
anks-0.047
unes-0.047
ware-0.047
Token merry
Token ID44570
Feature activation-1.57e-02

Pos logit contributions
ishes+0.024
anks+0.023
ware+0.023
unes+0.023
 fault+0.023
Neg logit contributions
ha-0.018
Little-0.016
www-0.015
 few-0.015
ww-0.015
Token way
Token ID835
Feature activation-1.10e-03

Pos logit contributions
ha+0.025
Little+0.023
Time+0.022
T+0.021
www+0.021
Neg logit contributions
ishes-0.033
anks-0.033
ware-0.033
 fault-0.032
aces-0.032
Token.
Token ID13
Feature activation+1.50e-02

Pos logit contributions
ha+0.002
Little+0.002
www+0.002
Time+0.002
ww+0.002
Neg logit contributions
 fault-0.002
anks-0.002
unes-0.002
ishes-0.002
ware-0.002
Token<|endoftext|>
Token ID50256
Feature activation+2.05e-02

Pos logit contributions
unes+0.029
anks+0.029
ishes+0.029
ware+0.029
 fault+0.028
Neg logit contributions
ha-0.027
Little-0.023
www-0.023
ww-0.022
 few-0.022
TokenC
Token ID34
Feature activation-2.36e-02

Pos logit contributions
anks+0.045
ishes+0.045
ware+0.045
unes+0.045
 fault+0.043
Neg logit contributions
ha-0.031
Little-0.027
www-0.027
ww-0.026
 few-0.026
Tokenind
Token ID521
Feature activation-8.61e-02

Pos logit contributions
ha+0.041
Little+0.036
www+0.035
 few+0.034
ww+0.034
Neg logit contributions
ishes-0.048
ware-0.047
anks-0.047
unes-0.047
 fault-0.045
Tokeny
Token ID88
Feature activation-3.17e-02

Pos logit contributions
ha+0.181
Little+0.165
www+0.160
Time+0.158
ww+0.157
Neg logit contributions
ishes-0.144
ware-0.141
anks-0.139
unes-0.139
 fault-0.134
Token and
Token ID290
Feature activation-3.21e-03

Pos logit contributions
ha+0.053
Time+0.047
www+0.046
Little+0.045
ww+0.045
Neg logit contributions
ishes-0.063
ware-0.062
anks-0.061
unes-0.060
 fault-0.059
Token her
Token ID607
Feature activation-1.02e-02

Pos logit contributions
ha+0.005
Little+0.005
 few+0.004
Time+0.004
www+0.004
Neg logit contributions
ishes-0.006
unes-0.006
melon-0.006
anks-0.005
ocular-0.005
Token mom
Token ID1995
Feature activation-7.79e-06

Pos logit contributions
ha+0.017
Little+0.015
www+0.015
Time+0.014
ww+0.014
Neg logit contributions
ishes-0.021
ware-0.021
anks-0.021
unes-0.020
 fault-0.020
Tokenmy
Token ID1820
Feature activation-4.02e-03

Feature not active on prev token;
no predictions were affected.
Token.
Token ID13
Feature activation-7.06e-04

Pos logit contributions
ha+0.023
Little+0.020
www+0.019
 few+0.019
ww+0.019
Neg logit contributions
ishes-0.034
anks-0.034
unes-0.034
ware-0.034
 fault-0.033
Token
Token ID198
Feature activation-6.60e-03

Pos logit contributions
ha+0.001
Little+0.001
ww+0.001
www+0.001
Time+0.001
Neg logit contributions
unes-0.001
ware-0.001
anks-0.001
ishes-0.001
 fault-0.001
Token
Token ID198
Feature activation-7.58e-03

Pos logit contributions
ha+0.011
ww+0.011
www+0.010
uren+0.010
 few+0.010
Neg logit contributions
ware-0.011
ests-0.011
unes-0.011
 fault-0.011
anks-0.011
Token"
Token ID1
Feature activation-1.11e-02

Pos logit contributions
ha+0.011
 night+0.010
 thunder+0.010
ww+0.010
 evening+0.009
Neg logit contributions
ests-0.012
auc-0.012
bugs-0.011
ectar-0.011
self-0.011
TokenE
Token ID36
Feature activation-2.88e-02

Pos logit contributions
ha+0.016
www+0.013
 few+0.013
unt+0.013
ww+0.012
Neg logit contributions
ware-0.026
ishes-0.025
unes-0.024
anks-0.023
ests-0.023
Tokenww
Token ID1383
Feature activation-8.58e-02

Pos logit contributions
ha+0.044
Little+0.039
Time+0.037
 few+0.037
unt+0.037
Neg logit contributions
ishes-0.064
ware-0.063
unes-0.062
anks-0.062
 fault-0.060
Token,
Token ID11
Feature activation-1.33e-02

Pos logit contributions
Time+0.128
ha+0.125
Alice+0.125
Little+0.123
Sarah+0.121
Neg logit contributions
ishes-0.194
unes-0.187
anks-0.186
ware-0.180
aces-0.171
Token this
Token ID428
Feature activation+2.42e-02

Pos logit contributions
ha+0.021
Little+0.018
www+0.017
Time+0.017
uren+0.017
Neg logit contributions
ware-0.029
unes-0.028
ishes-0.028
 fault-0.027
anks-0.027
Token fish
Token ID5916
Feature activation+9.12e-03

Pos logit contributions
ishes+0.043
anks+0.043
unes+0.043
ware+0.043
 fault+0.041
Neg logit contributions
ha-0.048
Little-0.043
www-0.042
ww-0.041
 few-0.041
Token is
Token ID318
Feature activation+3.21e-03

Pos logit contributions
unes+0.018
anks+0.018
ware+0.018
ishes+0.018
 fault+0.018
Neg logit contributions
ha-0.016
www-0.014
ww-0.014
Little-0.013
 few-0.013
Token sm
Token ID895
Feature activation+5.15e-03

Pos logit contributions
ishes+0.007
ware+0.007
anks+0.007
unes+0.007
 fault+0.007
Neg logit contributions
ha-0.005
Little-0.004
www-0.004
 few-0.004
ww-0.004
Token
Token ID26391
Feature activation-3.50e-02

Pos logit contributions
ha+0.048
 few+0.043
Little+0.041
www+0.040
T+0.040
Neg logit contributions
anks-0.050
unes-0.049
ishes-0.048
aces-0.047
ware-0.046
Token
Token ID8151
Feature activation-2.83e-02

Pos logit contributions
ha+0.056
 few+0.051
Time+0.050
Little+0.049
www+0.047
Neg logit contributions
unes-0.068
anks-0.068
aces-0.065
 fault-0.063
ric-0.063
Tokens
Token ID82
Feature activation-3.86e-03

Pos logit contributions
ha+0.055
Little+0.048
www+0.047
ww+0.046
 few+0.046
Neg logit contributions
ishes-0.052
ware-0.052
anks-0.051
unes-0.051
 fault-0.049
Token okay
Token ID8788
Feature activation-1.83e-02

Pos logit contributions
ha+0.007
Little+0.006
www+0.006
uren+0.006
ww+0.006
Neg logit contributions
unes-0.008
ishes-0.008
 fault-0.007
anks-0.007
ware-0.007
Token sweet
Token ID6029
Feature activation-2.16e-02

Pos logit contributions
ha+0.032
Little+0.027
Time+0.026
www+0.026
uren+0.026
Neg logit contributions
 fault-0.037
ishes-0.036
unes-0.036
anks-0.036
ware-0.035
Tokenheart
Token ID11499
Feature activation-8.54e-02

Pos logit contributions
ha+0.036
Little+0.032
Alice+0.030
Time+0.030
Charlie+0.030
Neg logit contributions
ishes-0.044
ware-0.044
unes-0.044
anks-0.043
 fault-0.041
Token.
Token ID13
Feature activation-2.68e-03

Pos logit contributions
ha+0.133
Little+0.116
www+0.110
Time+0.109
ww+0.108
Neg logit contributions
ishes-0.186
anks-0.185
ware-0.185
unes-0.184
 fault-0.181
Token How
Token ID1374
Feature activation+1.58e-02

Pos logit contributions
ha+0.005
Little+0.004
www+0.004
ww+0.004
 few+0.004
Neg logit contributions
unes-0.005
ishes-0.005
anks-0.005
ware-0.005
 fault-0.005
Token about
Token ID546
Feature activation-3.95e-02

Pos logit contributions
ishes+0.031
ware+0.030
anks+0.030
unes+0.030
 fault+0.029
Neg logit contributions
ha-0.029
Little-0.026
www-0.025
ww-0.024
 few-0.024
Token I
Token ID314
Feature activation-1.95e-02

Pos logit contributions
ha+0.061
Little+0.051
www+0.049
Time+0.049
ww+0.048
Neg logit contributions
ishes-0.089
ware-0.087
unes-0.086
anks-0.086
 fault-0.084
Token give
Token ID1577
Feature activation+1.42e-02

Pos logit contributions
ha+0.033
Little+0.029
www+0.028
ww+0.028
Time+0.027
Neg logit contributions
anks-0.039
ware-0.039
ishes-0.039
 fault-0.039
unes-0.039
Token.
Token ID13
Feature activation-3.04e-03

Pos logit contributions
ha+0.013
Little+0.011
www+0.011
ww+0.011
uren+0.011
Neg logit contributions
ware-0.014
ishes-0.014
unes-0.014
 fault-0.014
anks-0.014
Token
Token ID198
Feature activation-6.42e-03

Pos logit contributions
ha+0.005
Little+0.004
www+0.004
ww+0.004
 few+0.004
Neg logit contributions
ishes-0.007
anks-0.007
unes-0.007
ware-0.007
 fault-0.006
Token
Token ID198
Feature activation-7.94e-03

Pos logit contributions
ha+0.011
ww+0.011
www+0.010
uren+0.010
 few+0.010
Neg logit contributions
els-0.010
ware-0.010
unes-0.010
ests-0.010
anks-0.010
TokenThe
Token ID464
Feature activation+1.03e-02

Pos logit contributions
ha+0.011
 woke+0.009
 night+0.009
zing+0.009
 few+0.009
Neg logit contributions
els-0.014
ases-0.014
bugs-0.013
neys-0.013
olds-0.013
Token re
Token ID302
Feature activation-3.74e-02

Pos logit contributions
ware+0.025
ishes+0.024
unes+0.024
 fault+0.024
anks+0.024
Neg logit contributions
ha-0.014
Little-0.011
www-0.011
ww-0.011
Time-0.011
Tokenind
Token ID521
Feature activation-8.52e-02

Pos logit contributions
ha+0.058
Little+0.050
www+0.048
ww+0.047
 few+0.047
Neg logit contributions
anks-0.083
unes-0.083
ishes-0.082
ware-0.082
 fault-0.081
Tokeneer
Token ID28153
Feature activation-2.26e-02

Pos logit contributions
ha+0.222
Little+0.203
www+0.198
 few+0.196
ww+0.196
Neg logit contributions
anks-0.099
ishes-0.097
unes-0.097
ware-0.096
 fault-0.094
Token and
Token ID290
Feature activation-7.87e-03

Pos logit contributions
ha+0.034
Little+0.030
www+0.029
ww+0.028
Time+0.028
Neg logit contributions
ishes-0.050
ware-0.050
anks-0.049
unes-0.049
 fault-0.048
Token the
Token ID262
Feature activation-9.09e-03

Pos logit contributions
ha+0.011
Little+0.010
www+0.009
ww+0.009
Time+0.009
Neg logit contributions
ware-0.018
ishes-0.018
unes-0.018
anks-0.018
 fault-0.017
Token girl
Token ID2576
Feature activation-1.89e-02

Pos logit contributions
ha+0.014
Little+0.012
www+0.011
ww+0.011
Time+0.011
Neg logit contributions
ware-0.021
ishes-0.020
unes-0.020
anks-0.020
 fault-0.020
Token became
Token ID2627
Feature activation+1.08e-02

Pos logit contributions
ha+0.029
Little+0.025
www+0.024
 few+0.024
Time+0.024
Neg logit contributions
ishes-0.042
ware-0.041
anks-0.041
unes-0.041
 fault-0.040
Token.
Token ID13
Feature activation-6.63e-03

Pos logit contributions
ha+0.006
 few+0.005
Little+0.005
T+0.005
unt+0.005
Neg logit contributions
ishes-0.011
ware-0.011
unes-0.011
anks-0.010
 fault-0.010
Token She
Token ID1375
Feature activation-8.07e-03

Pos logit contributions
ha+0.008
Little+0.007
Time+0.007
unt+0.007
 few+0.007
Neg logit contributions
ishes-0.017
ware-0.017
anks-0.016
unes-0.016
 fault-0.016
Token hugs
Token ID40657
Feature activation-1.07e-02

Pos logit contributions
ha+0.015
Little+0.013
Time+0.012
www+0.012
 few+0.012
Neg logit contributions
unes-0.015
anks-0.015
ishes-0.015
ware-0.014
 fault-0.014
Token the
Token ID262
Feature activation-7.39e-03

Pos logit contributions
ha+0.016
Little+0.014
www+0.013
Time+0.013
 few+0.013
Neg logit contributions
ishes-0.024
anks-0.024
ware-0.024
unes-0.024
 fault-0.024
Token re
Token ID302
Feature activation-3.30e-02

Pos logit contributions
ha+0.013
Little+0.012
www+0.011
ww+0.011
Time+0.011
Neg logit contributions
ware-0.015
ishes-0.015
unes-0.014
anks-0.014
 fault-0.014
Tokenind
Token ID521
Feature activation-8.41e-02

Pos logit contributions
ha+0.044
unt+0.037
Little+0.036
www+0.035
T+0.035
Neg logit contributions
unes-0.080
anks-0.079
 fault-0.077
ware-0.076
ishes-0.076
Tokeneer
Token ID28153
Feature activation-1.95e-02

Pos logit contributions
ha+0.205
Little+0.182
unt+0.181
 few+0.180
www+0.180
Neg logit contributions
anks-0.112
unes-0.107
 fault-0.106
ishes-0.105
ware-0.102
Token.
Token ID13
Feature activation-6.51e-03

Pos logit contributions
ha+0.029
Little+0.025
 few+0.024
www+0.023
ww+0.023
Neg logit contributions
ishes-0.044
ware-0.044
anks-0.044
unes-0.044
 fault-0.043
Token
Token ID198
Feature activation-6.03e-03

Pos logit contributions
ha+0.008
Little+0.007
Time+0.007
unt+0.007
www+0.007
Neg logit contributions
ishes-0.016
ware-0.016
anks-0.016
unes-0.016
 fault-0.016
Token
Token ID198
Feature activation-7.29e-03

Pos logit contributions
ha+0.008
Little+0.007
www+0.006
Time+0.006
unt+0.006
Neg logit contributions
ishes-0.015
ware-0.015
anks-0.015
unes-0.015
 fault-0.014
Token"
Token ID1
Feature activation-1.23e-02

Pos logit contributions
ha+0.009
www+0.008
 few+0.007
zing+0.007
ww+0.007
Neg logit contributions
els-0.015
ishes-0.015
ases-0.015
ware-0.015
kins-0.015
Token She
Token ID1375
Feature activation-1.12e-02

Pos logit contributions
ishes+0.023
ware+0.023
anks+0.023
unes+0.023
 fault+0.022
Neg logit contributions
ha-0.018
Little-0.016
www-0.015
Time-0.015
ww-0.015
Token picks
Token ID11103
Feature activation+2.20e-03

Pos logit contributions
ha+0.019
Little+0.016
www+0.015
 few+0.015
Time+0.015
Neg logit contributions
anks-0.024
unes-0.024
ishes-0.024
ware-0.023
 fault-0.023
Token up
Token ID510
Feature activation-2.48e-02

Pos logit contributions
ishes+0.004
ware+0.004
anks+0.004
unes+0.004
 fault+0.004
Neg logit contributions
ha-0.004
Little-0.004
www-0.004
Time-0.003
ww-0.003
Token the
Token ID262
Feature activation-4.97e-03

Pos logit contributions
ha+0.033
Little+0.028
www+0.027
Time+0.026
ww+0.026
Neg logit contributions
ishes-0.061
ware-0.060
unes-0.060
anks-0.059
 fault-0.058
Token re
Token ID302
Feature activation-3.39e-02

Pos logit contributions
ha+0.009
Little+0.008
 few+0.008
www+0.007
ww+0.007
Neg logit contributions
anks-0.010
ishes-0.010
unes-0.010
ware-0.010
 fault-0.009
Tokenind
Token ID521
Feature activation-8.37e-02

Pos logit contributions
ha+0.050
Little+0.042
www+0.041
 few+0.041
ww+0.041
Neg logit contributions
unes-0.078
anks-0.078
 fault-0.075
ware-0.075
ishes-0.075
Tokeneer
Token ID28153
Feature activation-1.83e-02

Pos logit contributions
ha+0.204
Little+0.184
 few+0.181
www+0.180
ww+0.179
Neg logit contributions
anks-0.112
unes-0.107
 fault-0.105
ishes-0.105
ware-0.102
Token.
Token ID13
Feature activation-2.86e-03

Pos logit contributions
ha+0.026
Little+0.023
www+0.021
 few+0.021
ww+0.021
Neg logit contributions
ishes-0.042
anks-0.042
ware-0.042
unes-0.042
 fault-0.041
Token She
Token ID1375
Feature activation-1.12e-02

Pos logit contributions
ha+0.005
Little+0.004
www+0.004
ww+0.004
 few+0.004
Neg logit contributions
ishes-0.006
ware-0.006
anks-0.006
unes-0.006
 fault-0.006
Token looks
Token ID3073
Feature activation+5.86e-03

Pos logit contributions
ha+0.020
Little+0.018
 few+0.017
uren+0.017
unt+0.016
Neg logit contributions
unes-0.022
anks-0.022
ishes-0.021
 fault-0.020
rier-0.020
Token at
Token ID379
Feature activation-9.25e-03

Pos logit contributions
ware+0.013
anks+0.013
ishes+0.013
unes+0.013
 fault+0.013
Neg logit contributions
ha-0.009
Little-0.008
 few-0.007
www-0.007
ww-0.007
Token!"
Token ID2474
Feature activation-3.44e-03

Pos logit contributions
ishes+0.008
ware+0.008
anks+0.008
unes+0.008
 fault+0.007
Neg logit contributions
ha-0.006
Little-0.005
www-0.005
Time-0.005
 few-0.005
Token
Token ID198
Feature activation-1.03e-02

Pos logit contributions
ha+0.005
Little+0.004
 few+0.004
www+0.004
Time+0.004
Neg logit contributions
ishes-0.009
anks-0.008
unes-0.008
ware-0.008
 fault-0.008
Token
Token ID198
Feature activation-1.13e-02

Pos logit contributions
ha+0.014
Little+0.012
www+0.011
ww+0.011
Time+0.011
Neg logit contributions
ishes-0.024
ware-0.024
anks-0.024
unes-0.024
 fault-0.024
Token"
Token ID1
Feature activation-1.51e-02

Pos logit contributions
ha+0.013
www+0.010
unt+0.010
 Kelly+0.010
ww+0.010
Neg logit contributions
ware-0.027
kins-0.027
ishes-0.027
ests-0.027
 wounds-0.026
TokenA
Token ID32
Feature activation-3.66e-04

Pos logit contributions
ha+0.021
Little+0.017
www+0.017
unt+0.017
 few+0.016
Neg logit contributions
ware-0.035
ishes-0.034
unes-0.034
anks-0.034
 fault-0.033
Tokenww
Token ID1383
Feature activation-8.34e-02

Pos logit contributions
Little+0.001
Sarah+0.001
John+0.001
ha+0.001
Kelly+0.001
Neg logit contributions
ware-0.001
rier-0.001
ishes-0.001
 fault-0.001
anks-0.001
Token!"
Token ID2474
Feature activation-5.58e-03

Pos logit contributions
Sarah+0.127
Little+0.126
Alice+0.124
Time+0.124
T+0.124
Neg logit contributions
ishes-0.175
anks-0.172
unes-0.168
ware-0.166
aces-0.155
Token Lily
Token ID20037
Feature activation+1.15e-02

Pos logit contributions
ha+0.008
Little+0.007
www+0.007
ww+0.007
Time+0.007
Neg logit contributions
ware-0.013
ishes-0.013
anks-0.013
unes-0.012
 fault-0.012
Token says
Token ID1139
Feature activation-6.27e-03

Pos logit contributions
ishes+0.020
unes+0.020
ware+0.020
anks+0.019
rier+0.019
Neg logit contributions
ha-0.024
Little-0.021
unt-0.020
www-0.020
uren-0.020
Token.
Token ID13
Feature activation-1.02e-02

Pos logit contributions
ha+0.009
Little+0.007
 few+0.007
T+0.007
 long+0.007
Neg logit contributions
ishes-0.015
ware-0.015
unes-0.014
anks-0.014
 fault-0.014
Token "
Token ID366
Feature activation-1.79e-02

Pos logit contributions
ha+0.015
Little+0.013
www+0.012
ww+0.012
Time+0.012
Neg logit contributions
ishes-0.023
ware-0.023
anks-0.023
unes-0.023
 fault-0.023
TokenBut
Token ID1537
Feature activation-1.09e-02

Pos logit contributions
ha+0.016
Little+0.014
www+0.013
Time+0.013
ww+0.012
Neg logit contributions
ishes-0.031
anks-0.031
unes-0.031
ware-0.030
 fault-0.030
Token I
Token ID314
Feature activation-2.34e-02

Pos logit contributions
ha+0.015
Little+0.013
www+0.012
ww+0.012
 few+0.012
Neg logit contributions
anks-0.026
ishes-0.025
unes-0.025
ware-0.025
 fault-0.025
Token like
Token ID588
Feature activation+5.59e-03

Pos logit contributions
ha+0.045
Little+0.040
www+0.039
ww+0.038
unt+0.038
Neg logit contributions
ishes-0.042
 fault-0.042
anks-0.042
ware-0.041
unes-0.041
Token your
Token ID534
Feature activation-2.27e-02

Pos logit contributions
anks+0.014
ishes+0.014
unes+0.014
ware+0.014
 fault+0.014
Neg logit contributions
ha-0.007
Little-0.006
 few-0.006
www-0.005
ww-0.005
Token re
Token ID302
Feature activation-3.19e-02

Pos logit contributions
ha+0.029
 few+0.027
 little+0.025
unt+0.025
 long+0.025
Neg logit contributions
unes-0.054
anks-0.053
ishes-0.053
ware-0.052
aces-0.051
Tokenind
Token ID521
Feature activation-8.31e-02

Pos logit contributions
ha+0.041
unt+0.034
Little+0.034
www+0.033
ww+0.033
Neg logit contributions
unes-0.078
anks-0.076
ware-0.076
 fault-0.074
ishes-0.074
Tokeneer
Token ID28153
Feature activation-1.96e-02

Pos logit contributions
ha+0.197
unt+0.176
ww+0.174
 few+0.173
www+0.173
Neg logit contributions
anks-0.110
ware-0.108
unes-0.107
ishes-0.106
 fault-0.102
Token.
Token ID13
Feature activation-7.40e-03

Pos logit contributions
ha+0.028
 few+0.023
Little+0.023
ww+0.023
www+0.023
Neg logit contributions
ware-0.045
unes-0.045
anks-0.044
ishes-0.044
 fault-0.043
Token It
Token ID632
Feature activation+1.15e-02

Pos logit contributions
ha+0.011
Little+0.010
unt+0.010
www+0.009
ww+0.009
Neg logit contributions
ishes-0.017
ware-0.016
anks-0.016
 fault-0.016
unes-0.016
Token is
Token ID318
Feature activation+1.82e-03

Pos logit contributions
anks+0.028
unes+0.027
uggets+0.026
ishes+0.026
ware+0.026
Neg logit contributions
ha-0.017
ww-0.014
 few-0.013
www-0.013
unt-0.013
Token so
Token ID523
Feature activation+3.82e-02

Pos logit contributions
ishes+0.004
ware+0.004
anks+0.004
unes+0.004
 fault+0.004
Neg logit contributions
ha-0.003
Little-0.002
www-0.002
 few-0.002
ww-0.002
Token at
Token ID379
Feature activation-5.27e-03

Pos logit contributions
ha+0.005
Little+0.004
www+0.004
ww+0.004
unt+0.004
Neg logit contributions
ishes-0.005
 fault-0.005
anks-0.005
unes-0.005
ware-0.005
Token it
Token ID340
Feature activation+1.50e-03

Pos logit contributions
ha+0.006
Little+0.005
 few+0.005
www+0.005
unt+0.005
Neg logit contributions
anks-0.014
ware-0.013
unes-0.013
ishes-0.013
 fault-0.013
Token said
Token ID531
Feature activation-2.15e-02

Pos logit contributions
ishes+0.003
anks+0.003
unes+0.003
ware+0.003
 fault+0.003
Neg logit contributions
ha-0.002
Little-0.002
 few-0.002
www-0.002
ww-0.002
Token "
Token ID366
Feature activation-5.67e-03

Pos logit contributions
ha+0.030
Little+0.026
www+0.024
ww+0.024
Time+0.024
Neg logit contributions
ishes-0.051
ware-0.050
anks-0.050
unes-0.050
 fault-0.049
Tokenew
Token ID413
Feature activation-2.04e-02

Pos logit contributions
ha+0.010
unt+0.008
www+0.008
Little+0.008
uren+0.008
Neg logit contributions
ishes-0.011
unes-0.011
ware-0.011
anks-0.011
aces-0.010
Tokenww
Token ID1383
Feature activation-8.30e-02

Pos logit contributions
ha+0.033
Time+0.030
Alice+0.030
Little+0.029
unt+0.028
Neg logit contributions
ishes-0.043
unes-0.041
anks-0.040
ware-0.040
els-0.039
Token!"
Token ID2474
Feature activation+1.00e-03

Pos logit contributions
Sarah+0.143
Time+0.142
Alice+0.141
ha+0.138
John+0.136
Neg logit contributions
ishes-0.171
unes-0.162
anks-0.162
ware-0.152
sters-0.148
Token
Token ID220
Feature activation-4.19e-03

Pos logit contributions
ishes+0.002
ware+0.002
anks+0.002
unes+0.002
 fault+0.002
Neg logit contributions
ha-0.002
Little-0.001
 few-0.001
www-0.001
unt-0.001
Token
Token ID198
Feature activation-1.31e-02

Pos logit contributions
ha+0.004
Little+0.004
unt+0.003
Alice+0.003
T+0.003
Neg logit contributions
ishes-0.012
ware-0.012
unes-0.012
anks-0.011
 fault-0.011
Token
Token ID198
Feature activation-1.28e-02

Pos logit contributions
ha+0.023
ww+0.021
www+0.021
Time+0.019
Little+0.019
Neg logit contributions
ware-0.024
anks-0.023
ests-0.023
unes-0.023
ishes-0.022
TokenOne
Token ID3198
Feature activation+7.74e-03

Pos logit contributions
ha+0.017
ww+0.015
www+0.015
uren+0.013
zing+0.013
Neg logit contributions
neys-0.026
ware-0.026
ests-0.026
ocular-0.025
 preferences-0.025
Token a
Token ID257
Feature activation-1.10e-02

Pos logit contributions
ha+0.024
Little+0.020
www+0.019
ww+0.018
 few+0.018
Neg logit contributions
ishes-0.045
ware-0.044
anks-0.044
unes-0.044
 fault-0.043
Token sle
Token ID3133
Feature activation-2.46e-02

Pos logit contributions
ha+0.015
Little+0.012
www+0.012
ww+0.011
Time+0.011
Neg logit contributions
ishes-0.027
ware-0.027
anks-0.026
unes-0.026
 fault-0.026
Tokenigh
Token ID394
Feature activation-1.49e-02

Pos logit contributions
ha+0.053
Little+0.047
www+0.046
ww+0.045
 few+0.045
Neg logit contributions
ware-0.040
anks-0.039
unes-0.039
ishes-0.039
 fault-0.038
Token and
Token ID290
Feature activation-6.79e-03

Pos logit contributions
ha+0.026
Little+0.023
www+0.022
Time+0.022
ww+0.022
Neg logit contributions
ware-0.030
ishes-0.030
anks-0.030
unes-0.030
aces-0.028
Token re
Token ID302
Feature activation-3.04e-02

Pos logit contributions
ha+0.011
Little+0.010
www+0.009
Time+0.009
ww+0.009
Neg logit contributions
ware-0.015
ishes-0.015
anks-0.014
unes-0.014
aces-0.014
Tokenind
Token ID521
Feature activation-8.29e-02

Pos logit contributions
ha+0.047
Little+0.041
www+0.039
ww+0.038
 few+0.038
Neg logit contributions
anks-0.067
ware-0.067
unes-0.067
ishes-0.066
 fault-0.065
Tokeneer
Token ID28153
Feature activation-1.89e-02

Pos logit contributions
ha+0.216
Little+0.195
www+0.192
 few+0.190
ww+0.190
Neg logit contributions
anks-0.094
ware-0.094
ishes-0.094
unes-0.093
 fault-0.090
Token who
Token ID508
Feature activation-2.76e-02

Pos logit contributions
ha+0.032
Little+0.029
www+0.028
Time+0.027
ww+0.027
Neg logit contributions
ishes-0.039
anks-0.039
ware-0.039
unes-0.038
 fault-0.037
Token fly
Token ID6129
Feature activation+1.29e-02

Pos logit contributions
ha+0.051
Little+0.046
www+0.044
ww+0.043
 few+0.043
Neg logit contributions
ishes-0.052
ware-0.052
anks-0.052
unes-0.052
 fault-0.050
Token in
Token ID287
Feature activation+2.16e-02

Pos logit contributions
ishes+0.027
ware+0.026
anks+0.026
unes+0.026
 fault+0.026
Neg logit contributions
ha-0.022
Little-0.019
www-0.018
ww-0.018
Time-0.018
Token the
Token ID262
Feature activation-2.41e-03

Pos logit contributions
ware+0.051
ishes+0.050
 fault+0.049
anks+0.049
unes+0.049
Neg logit contributions
ha-0.030
www-0.026
ww-0.026
Little-0.025
uren-0.024
Token a
Token ID257
Feature activation-1.07e-02

Pos logit contributions
ware+0.008
ishes+0.008
unes+0.008
anks+0.008
 fault+0.008
Neg logit contributions
ha-0.004
Little-0.003
www-0.003
ww-0.003
Time-0.003
Token big
Token ID1263
Feature activation-1.05e-02

Pos logit contributions
ha+0.018
www+0.015
Little+0.015
Alice+0.015
uren+0.015
Neg logit contributions
ware-0.021
 fault-0.019
ishes-0.019
endant-0.019
rier-0.019
Token group
Token ID1448
Feature activation-1.21e-02

Pos logit contributions
ha+0.020
T+0.018
Little+0.018
While+0.017
Time+0.017
Neg logit contributions
ware-0.019
ishes-0.018
unes-0.018
 fault-0.018
melon-0.018
Token of
Token ID286
Feature activation+9.71e-04

Pos logit contributions
ha+0.022
Little+0.019
www+0.019
ww+0.018
T+0.018
Neg logit contributions
ware-0.024
unes-0.023
anks-0.023
ishes-0.023
aces-0.022
Token re
Token ID302
Feature activation-3.39e-02

Pos logit contributions
ishes+0.002
ware+0.002
unes+0.002
anks+0.002
 fault+0.002
Neg logit contributions
ha-0.002
Little-0.002
www-0.002
ww-0.002
uren-0.001
Tokenind
Token ID521
Feature activation-8.28e-02

Pos logit contributions
ha+0.060
Little+0.053
www+0.051
Time+0.051
ww+0.050
Neg logit contributions
ishes-0.069
ware-0.067
anks-0.067
unes-0.066
 fault-0.065
Tokeneer
Token ID28153
Feature activation-2.12e-02

Pos logit contributions
ha+0.219
Little+0.200
www+0.196
ww+0.194
 few+0.193
Neg logit contributions
ishes-0.092
anks-0.092
ware-0.092
unes-0.090
 fault-0.088
Token.
Token ID13
Feature activation-2.99e-03

Pos logit contributions
ha+0.036
Little+0.032
www+0.031
ww+0.031
unt+0.031
Neg logit contributions
ishes-0.044
anks-0.043
ware-0.043
unes-0.043
 fault-0.041
Token It
Token ID632
Feature activation+1.14e-02

Pos logit contributions
ha+0.005
Little+0.005
www+0.004
ww+0.004
Time+0.004
Neg logit contributions
ishes-0.006
ware-0.006
unes-0.006
anks-0.006
 fault-0.006
Token was
Token ID373
Feature activation+2.01e-03

Pos logit contributions
ware+0.024
ishes+0.023
unes+0.022
anks+0.022
 fault+0.022
Neg logit contributions
ha-0.019
Little-0.016
www-0.016
ww-0.016
Time-0.016
Token a
Token ID257
Feature activation-1.10e-02

Pos logit contributions
ishes+0.005
ware+0.005
anks+0.005
unes+0.005
 fault+0.005
Neg logit contributions
ha-0.003
Little-0.002
www-0.002
ww-0.002
Time-0.002
Tokeneer
Token ID28153
Feature activation-1.76e-02

Pos logit contributions
ha+0.226
Little+0.207
www+0.202
 few+0.200
ww+0.200
Neg logit contributions
anks-0.103
ishes-0.103
unes-0.102
ware-0.102
 fault-0.098
Token away
Token ID1497
Feature activation-2.18e-02

Pos logit contributions
ha+0.028
Little+0.025
www+0.024
ww+0.023
Time+0.023
Neg logit contributions
ishes-0.038
ware-0.037
anks-0.037
unes-0.037
 fault-0.036
Token.
Token ID13
Feature activation+1.10e-02

Pos logit contributions
ha+0.037
Little+0.032
www+0.031
ww+0.030
Time+0.030
Neg logit contributions
ishes-0.045
ware-0.045
anks-0.045
unes-0.044
 fault-0.044
Token The
Token ID383
Feature activation+1.65e-02

Pos logit contributions
unes+0.022
ishes+0.022
anks+0.022
ware+0.022
 fault+0.021
Neg logit contributions
ha-0.019
Little-0.017
www-0.016
ww-0.016
 few-0.016
Token re
Token ID302
Feature activation-2.87e-02

Pos logit contributions
ware+0.035
ishes+0.035
unes+0.034
anks+0.033
 fault+0.033
Neg logit contributions
ha-0.027
Little-0.023
www-0.022
ww-0.022
Time-0.022
Tokenind
Token ID521
Feature activation-8.27e-02

Pos logit contributions
ha+0.042
Little+0.036
www+0.035
 few+0.034
unt+0.034
Neg logit contributions
unes-0.066
anks-0.066
ware-0.065
ishes-0.064
 fault-0.064
Tokeneer
Token ID28153
Feature activation-1.70e-02

Pos logit contributions
ha+0.213
Little+0.194
www+0.189
 few+0.188
ww+0.188
Neg logit contributions
anks-0.098
unes-0.096
ishes-0.096
ware-0.095
 fault-0.092
Token said
Token ID531
Feature activation-5.23e-03

Pos logit contributions
ha+0.025
Little+0.021
www+0.020
 few+0.020
ww+0.020
Neg logit contributions
anks-0.039
unes-0.039
ishes-0.039
ware-0.039
 fault-0.038
Token goodbye
Token ID24829
Feature activation-1.01e-02

Pos logit contributions
ha+0.008
Little+0.007
www+0.006
Time+0.006
ww+0.006
Neg logit contributions
ware-0.012
unes-0.012
ishes-0.012
anks-0.012
 fault-0.012
Token to
Token ID284
Feature activation+1.63e-03

Pos logit contributions
ha+0.016
Little+0.014
www+0.013
Time+0.013
ww+0.013
Neg logit contributions
ishes-0.022
anks-0.022
ware-0.022
unes-0.021
 fault-0.021
Token the
Token ID262
Feature activation-1.35e-03

Pos logit contributions
ishes+0.004
anks+0.004
ware+0.004
unes+0.004
 fault+0.003
Neg logit contributions
ha-0.003
Little-0.002
www-0.002
ww-0.002
Time-0.002
Token there
Token ID612
Feature activation-1.25e-03

Pos logit contributions
ha+0.019
Little+0.016
www+0.016
ww+0.016
unt+0.015
Neg logit contributions
ishes-0.027
ware-0.027
unes-0.026
 fault-0.026
anks-0.026
Token was
Token ID373
Feature activation+9.79e-04

Pos logit contributions
ha+0.002
Little+0.002
www+0.002
ww+0.001
Time+0.001
Neg logit contributions
ishes-0.003
ware-0.003
anks-0.003
unes-0.003
 fault-0.003
Token a
Token ID257
Feature activation-1.41e-02

Pos logit contributions
ishes+0.002
ware+0.002
unes+0.002
 fault+0.002
anks+0.002
Neg logit contributions
ha-0.001
Little-0.001
www-0.001
unt-0.001
ww-0.001
Token happy
Token ID3772
Feature activation+2.96e-03

Pos logit contributions
ha+0.019
Time+0.015
www+0.015
Alice+0.015
Little+0.014
Neg logit contributions
ware-0.032
ishes-0.031
rier-0.030
kins-0.030
melon-0.029
Token re
Token ID302
Feature activation-3.27e-02

Pos logit contributions
ware+0.006
ishes+0.006
melon+0.006
unes+0.005
ectar+0.005
Neg logit contributions
ha-0.005
Little-0.004
Alice-0.004
Time-0.004
www-0.004
Tokenind
Token ID521
Feature activation-8.26e-02

Pos logit contributions
ha+0.057
Little+0.050
Time+0.048
www+0.048
ww+0.047
Neg logit contributions
ishes-0.067
ware-0.066
anks-0.066
unes-0.065
 fault-0.064
Tokeneer
Token ID28153
Feature activation-2.14e-02

Pos logit contributions
ha+0.226
Little+0.209
www+0.203
Time+0.202
ww+0.201
Neg logit contributions
ishes-0.086
ware-0.085
anks-0.084
unes-0.083
 fault-0.079
Token named
Token ID3706
Feature activation-5.87e-03

Pos logit contributions
ha+0.037
Little+0.033
www+0.032
Time+0.031
ww+0.031
Neg logit contributions
ishes-0.043
ware-0.043
anks-0.043
unes-0.042
 fault-0.041
Token Rudolph
Token ID48684
Feature activation+5.41e-04

Pos logit contributions
ha+0.011
Time+0.011
Little+0.010
www+0.010
Sarah+0.010
Neg logit contributions
unes-0.010
ishes-0.010
ware-0.010
anks-0.009
ocular-0.009
Token.
Token ID13
Feature activation-1.25e-03

Pos logit contributions
ishes+0.001
anks+0.001
ware+0.001
unes+0.001
 fault+0.001
Neg logit contributions
ha-0.001
Little-0.001
www-0.001
Time-0.001
ww-0.001
Token He
Token ID679
Feature activation+7.23e-03

Pos logit contributions
ha+0.003
Little+0.002
www+0.002
Sarah+0.002
Time+0.002
Neg logit contributions
ware-0.002
ishes-0.002
anks-0.002
unes-0.002
ests-0.002
Token began
Token ID2540
Feature activation-2.33e-02

Pos logit contributions
ishes+0.034
ware+0.033
anks+0.033
unes+0.033
 fault+0.032
Neg logit contributions
ha-0.024
Little-0.021
www-0.020
 few-0.020
ww-0.019
Token.
Token ID13
Feature activation+9.29e-03

Pos logit contributions
ha+0.036
Little+0.031
 few+0.030
www+0.030
Time+0.029
Neg logit contributions
ishes-0.052
ware-0.051
anks-0.051
unes-0.051
 fault-0.049
Token
Token ID198
Feature activation+2.47e-03

Pos logit contributions
unes+0.019
ware+0.019
anks+0.019
ishes+0.019
 fault+0.018
Neg logit contributions
ha-0.016
Little-0.013
www-0.013
ww-0.013
Time-0.013
Token
Token ID198
Feature activation-1.14e-04

Pos logit contributions
ware+0.005
anks+0.005
unes+0.005
ests+0.005
 fault+0.004
Neg logit contributions
ha-0.004
ww-0.004
www-0.004
uren-0.003
Little-0.003
TokenC
Token ID34
Feature activation-1.95e-02

Pos logit contributions
ha+0.000
ww+0.000
www+0.000
 Kelly+0.000
unt+0.000
Neg logit contributions
ests-0.000
ware-0.000
neys-0.000
ases-0.000
auc-0.000
Tokenind
Token ID521
Feature activation-8.26e-02

Pos logit contributions
ha+0.038
Little+0.034
www+0.033
 few+0.033
ww+0.032
Neg logit contributions
ishes-0.036
ware-0.035
anks-0.035
unes-0.035
 fault-0.034
Tokeny
Token ID88
Feature activation-2.51e-02

Pos logit contributions
ha+0.146
Little+0.131
www+0.126
Time+0.124
ww+0.124
Neg logit contributions
ishes-0.165
ware-0.163
anks-0.161
unes-0.161
 fault-0.157
Token tried
Token ID3088
Feature activation-7.21e-03

Pos logit contributions
ha+0.041
Little+0.036
Time+0.036
ww+0.035
www+0.035
Neg logit contributions
ware-0.050
ishes-0.050
anks-0.049
unes-0.047
 fault-0.047
Token to
Token ID284
Feature activation-5.17e-03

Pos logit contributions
ha+0.013
Little+0.011
www+0.011
ww+0.011
unt+0.010
Neg logit contributions
 fault-0.014
anks-0.014
ware-0.014
ishes-0.014
unes-0.014
Token make
Token ID787
Feature activation+1.16e-02

Pos logit contributions
ha+0.009
ww+0.008
www+0.008
Little+0.008
unt+0.008
Neg logit contributions
anks-0.010
ware-0.010
ishes-0.010
 fault-0.010
unes-0.010
Token Peter
Token ID5613
Feature activation-4.98e-03

Pos logit contributions
anks+0.026
ishes+0.026
 fault+0.026
ware+0.026
unes+0.026
Neg logit contributions
ha-0.018
Little-0.015
www-0.014
ww-0.014
Time-0.014
Token for
Token ID329
Feature activation+8.67e-03

Pos logit contributions
ha+0.025
Little+0.022
 few+0.022
www+0.021
Time+0.021
Neg logit contributions
anks-0.029
ware-0.029
ishes-0.029
unes-0.029
 fault-0.029
Token someone
Token ID2130
Feature activation+4.76e-03

Pos logit contributions
ishes+0.018
ware+0.018
anks+0.018
unes+0.018
 fault+0.018
Neg logit contributions
ha-0.014
Little-0.012
www-0.012
ww-0.012
unt-0.012
Token to
Token ID284
Feature activation+6.50e-03

Pos logit contributions
anks+0.012
ware+0.012
unes+0.012
ishes+0.011
rier+0.011
Neg logit contributions
ha-0.006
 few-0.005
Little-0.005
www-0.005
unt-0.005
Token add
Token ID751
Feature activation+3.63e-02

Pos logit contributions
ishes+0.015
ware+0.015
unes+0.015
anks+0.014
 fault+0.014
Neg logit contributions
ha-0.010
Little-0.009
 few-0.008
www-0.008
Time-0.008
Token.
Token ID13
Feature activation+1.25e-02

Pos logit contributions
ware+0.090
anks+0.090
ishes+0.087
unes+0.087
ses+0.085
Neg logit contributions
ha-0.045
 few-0.040
 long-0.036
 time-0.035
Little-0.033
Token Everywhere
Token ID48201
Feature activation+7.27e-02

Pos logit contributions
ishes+0.030
ware+0.030
anks+0.029
unes+0.029
 fault+0.029
Neg logit contributions
ha-0.017
Little-0.015
 few-0.014
Time-0.014
unt-0.013
Token he
Token ID339
Feature activation+8.42e-03

Pos logit contributions
ishes+0.152
ware+0.148
unes+0.143
anks+0.141
 fault+0.141
Neg logit contributions
ha-0.116
www-0.102
Little-0.100
ww-0.100
unt-0.098
Token looked
Token ID3114
Feature activation+4.99e-04

Pos logit contributions
ishes+0.020
anks+0.020
unes+0.020
ware+0.020
 fault+0.020
Neg logit contributions
ha-0.012
Little-0.010
www-0.009
ww-0.009
 few-0.009
Token there
Token ID612
Feature activation+8.35e-03

Pos logit contributions
ishes+0.001
ware+0.001
anks+0.001
unes+0.001
 fault+0.001
Neg logit contributions
ha-0.001
Little-0.001
www-0.000
ww-0.000
 few-0.000
Token were
Token ID547
Feature activation+4.69e-03

Pos logit contributions
anks+0.020
ishes+0.019
unes+0.019
ware+0.019
 fault+0.019
Neg logit contributions
ha-0.012
Little-0.010
 few-0.010
Time-0.010
 Kelly-0.009
Token no
Token ID645
Feature activation-2.03e-02

Pos logit contributions
anks+0.010
ishes+0.010
unes+0.010
ware+0.010
 fault+0.010
Neg logit contributions
ha-0.007
 few-0.006
Little-0.006
Time-0.006
www-0.006
Token was
Token ID373
Feature activation+6.94e-03

Pos logit contributions
ishes+0.010
ware+0.010
anks+0.009
unes+0.009
 fault+0.009
Neg logit contributions
ha-0.006
Little-0.005
www-0.005
ww-0.005
Time-0.005
Token overwhelmed
Token ID20974
Feature activation+5.50e-04

Pos logit contributions
 fault+0.012
anks+0.012
ishes+0.012
unes+0.012
ware+0.011
Neg logit contributions
ha-0.014
Little-0.013
ww-0.012
www-0.012
T-0.012
Token with
Token ID351
Feature activation+2.05e-02

Pos logit contributions
ishes+0.001
ware+0.001
anks+0.001
unes+0.001
 fault+0.001
Neg logit contributions
ha-0.001
Little-0.001
www-0.001
ww-0.001
Time-0.001
Token joy
Token ID8716
Feature activation-7.51e-03

Pos logit contributions
ishes+0.048
anks+0.048
ware+0.048
unes+0.047
 fault+0.046
Neg logit contributions
ha-0.030
Little-0.025
www-0.024
 few-0.024
Time-0.023
Token!
Token ID0
Feature activation+8.12e-03

Pos logit contributions
ha+0.014
Little+0.012
www+0.012
ww+0.011
Time+0.011
Neg logit contributions
ishes-0.015
ware-0.015
anks-0.015
unes-0.015
 fault-0.015
Token Everywhere
Token ID48201
Feature activation+7.01e-02

Pos logit contributions
unes+0.017
anks+0.017
ishes+0.017
ware+0.016
 fault+0.016
Neg logit contributions
ha-0.014
Little-0.012
www-0.012
ww-0.012
Time-0.011
Token she
Token ID673
Feature activation-3.25e-03

Pos logit contributions
ishes+0.147
ware+0.143
unes+0.136
els+0.134
aces+0.133
Neg logit contributions
ha-0.102
ww-0.091
www-0.091
Time-0.090
Little-0.089
Token looked
Token ID3114
Feature activation-4.12e-03

Pos logit contributions
ha+0.005
Little+0.004
ww+0.004
www+0.004
unt+0.004
Neg logit contributions
ware-0.007
ishes-0.007
unes-0.007
anks-0.007
els-0.006
Token,
Token ID11
Feature activation-8.36e-03

Pos logit contributions
ha+0.005
Little+0.004
www+0.004
ww+0.004
Time+0.004
Neg logit contributions
ishes-0.010
ware-0.010
unes-0.010
anks-0.010
 fault-0.010
Token she
Token ID673
Feature activation-3.28e-03

Pos logit contributions
ha+0.012
www+0.010
ww+0.009
Little+0.009
unt+0.009
Neg logit contributions
ishes-0.020
ware-0.020
unes-0.020
anks-0.019
 fault-0.019
Token could
Token ID714
Feature activation+1.65e-02

Pos logit contributions
ha+0.004
Little+0.004
www+0.004
ww+0.004
Time+0.004
Neg logit contributions
ishes-0.008
ware-0.008
anks-0.008
unes-0.008
 fault-0.008
Token ribbon
Token ID29092
Feature activation-1.63e-02

Pos logit contributions
ha+0.005
Little+0.004
www+0.004
 few+0.004
ww+0.004
Neg logit contributions
ishes-0.006
anks-0.006
ware-0.006
unes-0.006
 fault-0.006
Token trailing
Token ID25462
Feature activation-9.39e-03

Pos logit contributions
ha+0.025
Little+0.022
www+0.021
Time+0.020
ww+0.020
Neg logit contributions
ware-0.036
ishes-0.036
anks-0.036
unes-0.036
 fault-0.035
Token behind
Token ID2157
Feature activation-5.80e-03

Pos logit contributions
ha+0.018
Little+0.016
www+0.015
ww+0.015
Time+0.015
Neg logit contributions
ware-0.018
ishes-0.018
anks-0.018
unes-0.017
 fault-0.017
Token her
Token ID607
Feature activation-8.44e-04

Pos logit contributions
ha+0.009
www+0.008
ww+0.008
Little+0.008
unt+0.007
Neg logit contributions
ware-0.013
unes-0.013
anks-0.012
ishes-0.012
aces-0.012
Token.
Token ID13
Feature activation+6.18e-03

Pos logit contributions
ha+0.001
Little+0.001
 few+0.001
www+0.001
uren+0.001
Neg logit contributions
ishes-0.002
ware-0.002
unes-0.002
anks-0.002
 fault-0.002
Token Everywhere
Token ID48201
Feature activation+6.93e-02

Pos logit contributions
anks+0.012
unes+0.012
 fault+0.011
ware+0.011
aces+0.011
Neg logit contributions
ha-0.012
ww-0.011
www-0.010
Little-0.010
Time-0.009
Token she
Token ID673
Feature activation-1.28e-03

Pos logit contributions
ishes+0.131
ware+0.126
 fault+0.126
els+0.125
 depending+0.124
Neg logit contributions
ha-0.103
ww-0.098
www-0.097
T-0.092
While-0.092
Token went
Token ID1816
Feature activation+1.33e-02

Pos logit contributions
ha+0.002
ww+0.002
unt+0.002
www+0.002
Little+0.002
Neg logit contributions
ware-0.003
ishes-0.003
aces-0.002
unes-0.002
els-0.002
Token,
Token ID11
Feature activation-6.95e-03

Pos logit contributions
unes+0.026
ware+0.025
anks+0.025
aces+0.024
els+0.024
Neg logit contributions
ha-0.021
ww-0.020
www-0.020
Little-0.019
T-0.019
Token the
Token ID262
Feature activation-3.90e-03

Pos logit contributions
ha+0.010
www+0.009
ww+0.009
unt+0.008
Little+0.008
Neg logit contributions
ishes-0.015
unes-0.015
anks-0.015
ware-0.015
 fault-0.015
Token ribbon
Token ID29092
Feature activation-1.60e-02

Pos logit contributions
ha+0.007
Little+0.006
www+0.006
ww+0.006
unt+0.006
Neg logit contributions
ware-0.008
ishes-0.007
 fault-0.007
aces-0.007
unes-0.007
Token and
Token ID290
Feature activation+2.94e-03

Pos logit contributions
ha+0.039
Little+0.033
www+0.032
ww+0.031
Time+0.031
Neg logit contributions
ishes-0.057
ware-0.056
anks-0.056
unes-0.056
 fault-0.055
Token took
Token ID1718
Feature activation+1.56e-03

Pos logit contributions
ishes+0.006
anks+0.006
ware+0.006
aces+0.005
 fault+0.005
Neg logit contributions
ha-0.005
Time-0.004
Little-0.004
unt-0.004
ww-0.004
Token it
Token ID340
Feature activation+4.71e-03

Pos logit contributions
anks+0.003
aces+0.003
ware+0.003
els+0.003
 fault+0.003
Neg logit contributions
unt-0.002
ww-0.002
Little-0.002
www-0.002
Time-0.002
Token everywhere
Token ID8347
Feature activation+2.25e-02

Pos logit contributions
anks+0.009
unes+0.009
 fault+0.009
ishes+0.009
ware+0.009
Neg logit contributions
ha-0.009
Little-0.008
www-0.008
Time-0.007
ww-0.007
Token.
Token ID13
Feature activation+7.76e-03

Pos logit contributions
anks+0.042
ware+0.042
unes+0.041
rier+0.040
 fault+0.040
Neg logit contributions
ha-0.041
ww-0.036
Time-0.035
www-0.035
Little-0.035
Token Everywhere
Token ID48201
Feature activation+7.13e-02

Pos logit contributions
anks+0.016
unes+0.016
ishes+0.016
ware+0.015
 fault+0.015
Neg logit contributions
ha-0.014
Little-0.012
www-0.011
ww-0.011
Time-0.011
Token the
Token ID262
Feature activation-4.13e-03

Pos logit contributions
ishes+0.150
ware+0.145
unes+0.138
 fault+0.137
els+0.134
Neg logit contributions
ha-0.103
Time-0.091
ww-0.091
www-0.090
While-0.090
Token little
Token ID1310
Feature activation-1.18e-02

Pos logit contributions
ha+0.006
www+0.005
Time+0.005
ww+0.005
T+0.005
Neg logit contributions
ware-0.010
ishes-0.009
rier-0.009
unes-0.009
kins-0.009
Token girl
Token ID2576
Feature activation-1.23e-02

Pos logit contributions
ha+0.017
Little+0.014
www+0.014
ww+0.013
Time+0.013
Neg logit contributions
ishes-0.028
ware-0.028
unes-0.027
anks-0.027
 fault-0.026
Token went
Token ID1816
Feature activation+1.31e-02

Pos logit contributions
ha+0.018
Little+0.016
www+0.015
ww+0.015
Time+0.015
Neg logit contributions
ishes-0.028
ware-0.028
anks-0.027
unes-0.027
 fault-0.027
Token she
Token ID673
Feature activation-3.45e-04

Pos logit contributions
anks+0.028
unes+0.027
ware+0.027
ishes+0.026
aces+0.026
Neg logit contributions
ha-0.020
Little-0.018
ww-0.018
T-0.018
www-0.018
Token,
Token ID11
Feature activation-5.02e-03

Pos logit contributions
ishes+0.031
anks+0.030
ware+0.030
unes+0.030
 fault+0.030
Neg logit contributions
ha-0.021
Little-0.019
www-0.019
ww-0.018
Time-0.018
Token became
Token ID2627
Feature activation+1.34e-02

Pos logit contributions
ha+0.008
www+0.007
Little+0.007
ww+0.007
unt+0.007
Neg logit contributions
ishes-0.011
ware-0.010
unes-0.010
anks-0.010
 fault-0.010
Token best
Token ID1266
Feature activation+6.62e-04

Pos logit contributions
ishes+0.028
anks+0.028
ware+0.028
unes+0.028
 fault+0.027
Neg logit contributions
ha-0.023
Little-0.020
www-0.019
ww-0.019
Time-0.018
Token friends
Token ID2460
Feature activation-2.95e-03

Pos logit contributions
ishes+0.001
anks+0.001
unes+0.001
 fault+0.001
ware+0.001
Neg logit contributions
ha-0.001
Little-0.001
www-0.001
ww-0.001
Time-0.001
Token.
Token ID13
Feature activation+5.09e-03

Pos logit contributions
ha+0.005
Little+0.004
www+0.004
ww+0.004
Time+0.004
Neg logit contributions
ishes-0.006
anks-0.006
ware-0.006
unes-0.006
 fault-0.006
Token Everywhere
Token ID48201
Feature activation+7.03e-02

Pos logit contributions
ware+0.010
ishes+0.010
unes+0.010
anks+0.010
 fault+0.009
Neg logit contributions
ha-0.009
Little-0.008
 few-0.008
unt-0.008
Time-0.008
Token the
Token ID262
Feature activation-5.38e-03

Pos logit contributions
ishes+0.142
ware+0.135
anks+0.132
unes+0.130
 fault+0.126
Neg logit contributions
ha-0.117
Little-0.105
www-0.104
Time-0.101
Alice-0.101
Token driver
Token ID4639
Feature activation-1.36e-03

Pos logit contributions
ha+0.009
www+0.008
ww+0.008
Little+0.008
T+0.008
Neg logit contributions
ishes-0.011
ware-0.010
anks-0.010
unes-0.010
 fault-0.010
Token went
Token ID1816
Feature activation+1.24e-02

Pos logit contributions
ha+0.002
Little+0.002
www+0.001
 few+0.001
unt+0.001
Neg logit contributions
anks-0.003
unes-0.003
 fault-0.003
ware-0.003
ishes-0.003
Token,
Token ID11
Feature activation-4.12e-03

Pos logit contributions
anks+0.027
ishes+0.026
unes+0.026
ware+0.026
 fault+0.024
Neg logit contributions
ha-0.019
Little-0.017
www-0.016
ww-0.016
T-0.016
Token Sack
Token ID49175
Feature activation+1.65e-02

Pos logit contributions
ha+0.006
Little+0.005
www+0.005
ww+0.005
unt+0.005
Neg logit contributions
ishes-0.009
ware-0.009
anks-0.009
unes-0.009
 fault-0.009
Token!
Token ID0
Feature activation+2.48e-02

Pos logit contributions
ware+0.000
anks+0.000
unes+0.000
ishes+0.000
aces+0.000
Neg logit contributions
ha-0.000
Little-0.000
 few-0.000
Time-0.000
www-0.000
Token Hur
Token ID10077
Feature activation-6.20e-03

Pos logit contributions
ishes+0.053
anks+0.052
ware+0.052
unes+0.052
 fault+0.051
Neg logit contributions
ha-0.040
Little-0.035
www-0.033
ww-0.033
 few-0.033
Tokenry
Token ID563
Feature activation+2.56e-02

Pos logit contributions
ha+0.013
Little+0.011
www+0.011
ww+0.011
Time+0.011
Neg logit contributions
ishes-0.011
anks-0.011
unes-0.011
ware-0.011
 fault-0.010
Token up
Token ID510
Feature activation-1.38e-02

Pos logit contributions
ishes+0.051
ware+0.051
anks+0.051
unes+0.051
 fault+0.049
Neg logit contributions
ha-0.045
Little-0.040
www-0.038
ww-0.038
Time-0.037
Token!"
Token ID2474
Feature activation+4.70e-03

Pos logit contributions
ha+0.022
Little+0.019
www+0.019
ww+0.018
Time+0.018
Neg logit contributions
ishes-0.030
ware-0.029
anks-0.029
unes-0.029
 fault-0.029
Token Tim
Token ID5045
Feature activation+1.49e-02

Pos logit contributions
ishes+0.010
anks+0.010
unes+0.010
ware+0.010
 fault+0.010
Neg logit contributions
ha-0.007
Little-0.006
www-0.006
ww-0.006
Time-0.006
Token shouted
Token ID17293
Feature activation-3.78e-04

Pos logit contributions
ishes+0.032
anks+0.032
ware+0.031
unes+0.031
 fault+0.031
Neg logit contributions
ha-0.024
Little-0.020
www-0.020
ww-0.020
Time-0.020
Token.
Token ID13
Feature activation+1.51e-02

Pos logit contributions
ha+0.001
Little+0.001
www+0.001
ww+0.001
Time+0.001
Neg logit contributions
anks-0.001
ware-0.001
unes-0.001
 fault-0.001
ishes-0.001
Token He
Token ID679
Feature activation+1.41e-02

Pos logit contributions
ishes+0.035
ware+0.035
anks+0.034
unes+0.034
 fault+0.034
Neg logit contributions
ha-0.022
Little-0.019
www-0.018
ww-0.018
Time-0.017
Token was
Token ID373
Feature activation+6.29e-03

Pos logit contributions
ware+0.030
ishes+0.029
anks+0.029
unes+0.028
 fault+0.028
Neg logit contributions
ha-0.021
Little-0.020
ww-0.019
www-0.019
Time-0.018
Token almost
Token ID2048
Feature activation+2.93e-03

Pos logit contributions
 fault+0.011
anks+0.011
unes+0.011
ishes+0.011
ware+0.011
Neg logit contributions
ha-0.012
Little-0.011
ww-0.011
www-0.011
unt-0.010
Token find
Token ID1064
Feature activation+1.62e-02

Pos logit contributions
ware+0.022
anks+0.022
unes+0.022
ishes+0.021
 fault+0.021
Neg logit contributions
ha-0.018
Little-0.015
 few-0.014
www-0.014
 rush-0.014
Token some
Token ID617
Feature activation+1.43e-02

Pos logit contributions
anks+0.040
ware+0.040
unes+0.039
ishes+0.039
 fault+0.038
Neg logit contributions
ha-0.022
Little-0.018
 few-0.017
www-0.017
ww-0.017
Token cheap
Token ID7026
Feature activation+5.44e-03

Pos logit contributions
anks+0.035
unes+0.034
ware+0.034
ishes+0.034
 fault+0.033
Neg logit contributions
ha-0.020
 few-0.017
Little-0.017
ww-0.016
Time-0.016
Token steam
Token ID13324
Feature activation-1.17e-02

Pos logit contributions
anks+0.012
unes+0.011
ware+0.011
ishes+0.011
 fault+0.011
Neg logit contributions
ha-0.009
 few-0.008
Little-0.007
ww-0.007
www-0.007
Token.
Token ID13
Feature activation+8.08e-03

Pos logit contributions
ha+0.018
Little+0.014
www+0.014
 few+0.014
ww+0.014
Neg logit contributions
ishes-0.026
ware-0.026
anks-0.026
unes-0.025
 fault-0.025
Token His
Token ID2399
Feature activation+2.05e-02

Pos logit contributions
ishes+0.017
ware+0.017
anks+0.016
unes+0.016
 fault+0.016
Neg logit contributions
ha-0.013
Little-0.012
 few-0.011
unt-0.011
uren-0.011
Token friends
Token ID2460
Feature activation+5.73e-03

Pos logit contributions
ware+0.040
unes+0.040
ishes+0.040
anks+0.040
 fault+0.038
Neg logit contributions
ha-0.037
Little-0.033
 few-0.032
www-0.032
ww-0.031
Token got
Token ID1392
Feature activation+1.83e-03

Pos logit contributions
ishes+0.011
ware+0.011
anks+0.011
unes+0.011
 fault+0.010
Neg logit contributions
ha-0.010
Little-0.009
www-0.009
ww-0.009
Time-0.009
Token very
Token ID845
Feature activation+1.42e-02

Pos logit contributions
ware+0.005
anks+0.005
unes+0.005
ishes+0.005
ests+0.005
Neg logit contributions
ha-0.002
 few-0.002
Time-0.002
Little-0.002
 long-0.002
Token excited
Token ID6568
Feature activation+2.77e-03

Pos logit contributions
ware+0.039
unes+0.039
anks+0.039
ishes+0.037
ests+0.037
Neg logit contributions
ha-0.014
 few-0.013
Little-0.012
ww-0.012
www-0.012
Token and
Token ID290
Feature activation+2.19e-03

Pos logit contributions
ishes+0.007
ware+0.006
anks+0.006
unes+0.006
 fault+0.006
Neg logit contributions
ha-0.004
Little-0.003
www-0.003
 few-0.003
Time-0.003
Token in
Token ID287
Feature activation+2.97e-02

Pos logit contributions
ishes+0.013
ware+0.013
anks+0.013
unes+0.013
 fault+0.012
Neg logit contributions
ha-0.008
Little-0.007
www-0.006
ww-0.006
Time-0.006
Token herself
Token ID5223
Feature activation+8.69e-03

Pos logit contributions
ware+0.059
anks+0.059
els+0.058
 fault+0.058
ishes+0.058
Neg logit contributions
ha-0.051
ww-0.043
www-0.042
unt-0.042
Little-0.041
Token,
Token ID11
Feature activation-2.93e-03

Pos logit contributions
ishes+0.019
ware+0.019
anks+0.019
unes+0.019
 fault+0.018
Neg logit contributions
ha-0.014
Little-0.012
www-0.011
ww-0.011
Time-0.011
Token and
Token ID290
Feature activation+5.41e-03

Pos logit contributions
ha+0.005
ww+0.004
www+0.004
Little+0.004
unt+0.004
Neg logit contributions
ware-0.006
 fault-0.006
ishes-0.006
anks-0.006
unes-0.005
Token it
Token ID340
Feature activation+4.82e-03

Pos logit contributions
ishes+0.011
ware+0.011
anks+0.011
unes+0.011
 fault+0.011
Neg logit contributions
ha-0.008
Little-0.007
ww-0.007
www-0.007
unt-0.007
Token showed
Token ID3751
Feature activation+4.87e-02

Pos logit contributions
ware+0.010
ishes+0.010
anks+0.010
 fault+0.009
unes+0.009
Neg logit contributions
ha-0.008
www-0.007
ww-0.007
Little-0.007
unt-0.007
Token in
Token ID287
Feature activation+3.19e-02

Pos logit contributions
ishes+0.115
anks+0.115
ware+0.113
unes+0.113
 fault+0.112
Neg logit contributions
ha-0.068
Little-0.058
www-0.056
ww-0.056
Time-0.055
Token her
Token ID607
Feature activation+1.54e-03

Pos logit contributions
ware+0.067
 fault+0.067
anks+0.066
ishes+0.065
els+0.064
Neg logit contributions
ha-0.052
www-0.044
ww-0.043
Little-0.042
unt-0.041
Token painting
Token ID12036
Feature activation+1.28e-02

Pos logit contributions
unes+0.003
ishes+0.003
ware+0.003
anks+0.003
 fault+0.003
Neg logit contributions
ha-0.002
Little-0.002
 few-0.002
www-0.002
ww-0.002
Token.
Token ID13
Feature activation+1.05e-02

Pos logit contributions
ishes+0.029
unes+0.028
ware+0.028
anks+0.028
 fault+0.028
Neg logit contributions
ha-0.019
Little-0.017
 few-0.016
www-0.016
unt-0.016
Token She
Token ID1375
Feature activation+1.35e-03

Pos logit contributions
anks+0.022
ishes+0.022
unes+0.022
ware+0.022
 fault+0.021
Neg logit contributions
ha-0.018
Little-0.015
www-0.015
ww-0.015
Time-0.014
Token she
Token ID673
Feature activation-7.43e-04

Pos logit contributions
ishes+0.012
anks+0.012
ware+0.012
unes+0.012
 fault+0.012
Neg logit contributions
ha-0.010
Little-0.008
www-0.008
ww-0.008
Time-0.008
Token smiled
Token ID13541
Feature activation+7.83e-03

Pos logit contributions
ha+0.001
Little+0.001
www+0.001
ww+0.001
Time+0.001
Neg logit contributions
ishes-0.002
ware-0.002
anks-0.002
 fault-0.001
unes-0.001
Token right
Token ID826
Feature activation+2.42e-03

Pos logit contributions
ishes+0.016
anks+0.016
ware+0.016
unes+0.016
 fault+0.016
Neg logit contributions
ha-0.013
Little-0.011
www-0.011
ww-0.011
Time-0.011
Token back
Token ID736
Feature activation+2.45e-02

Pos logit contributions
ishes+0.005
ware+0.005
anks+0.005
unes+0.005
 fault+0.005
Neg logit contributions
ha-0.004
Little-0.004
www-0.004
ww-0.004
Time-0.004
Token.
Token ID13
Feature activation+1.45e-02

Pos logit contributions
anks+0.047
ishes+0.046
 fault+0.045
unes+0.045
ware+0.045
Neg logit contributions
ha-0.047
Little-0.041
www-0.040
ww-0.040
Time-0.039
Token From
Token ID3574
Feature activation+2.45e-02

Pos logit contributions
ishes+0.032
anks+0.032
unes+0.032
ware+0.032
 fault+0.031
Neg logit contributions
ha-0.023
Little-0.019
www-0.019
ww-0.019
 few-0.018
Token then
Token ID788
Feature activation+8.94e-03

Pos logit contributions
ishes+0.044
ocular+0.042
els+0.042
ware+0.041
anks+0.041
Neg logit contributions
ha-0.046
Time-0.039
 few-0.038
www-0.038
T-0.038
Token on
Token ID319
Feature activation-1.14e-02

Pos logit contributions
anks+0.015
aces+0.014
unes+0.014
els+0.014
ishes+0.014
Neg logit contributions
ha-0.017
Little-0.015
uren-0.015
Charlie-0.015
Sarah-0.015
Token,
Token ID11
Feature activation+8.78e-04

Pos logit contributions
ha+0.017
Time+0.014
Little+0.014
www+0.014
T+0.014
Neg logit contributions
ware-0.025
ishes-0.025
unes-0.024
 fault-0.024
anks-0.024
Token Hannah
Token ID26013
Feature activation-1.35e-02

Pos logit contributions
ishes+0.002
ware+0.002
unes+0.002
anks+0.002
 fault+0.002
Neg logit contributions
ha-0.001
www-0.001
ww-0.001
Little-0.001
Time-0.001
Token was
Token ID373
Feature activation+1.13e-02

Pos logit contributions
ha+0.023
www+0.021
ww+0.021
Time+0.020
Little+0.020
Neg logit contributions
ishes-0.026
anks-0.025
ware-0.025
 fault-0.024
aces-0.024
Token go
Token ID467
Feature activation+1.18e-02

Pos logit contributions
anks+0.007
ware+0.007
ishes+0.007
unes+0.007
 fault+0.006
Neg logit contributions
ha-0.004
Little-0.004
www-0.003
ww-0.003
Time-0.003
Token for
Token ID329
Feature activation+5.22e-03

Pos logit contributions
ishes+0.026
anks+0.025
ware+0.025
unes+0.025
 fault+0.024
Neg logit contributions
ha-0.019
Little-0.017
www-0.016
ww-0.016
Time-0.015
Token a
Token ID257
Feature activation-5.93e-03

Pos logit contributions
anks+0.010
 fault+0.010
ware+0.009
unes+0.009
aces+0.009
Neg logit contributions
ha-0.009
Little-0.008
www-0.008
unt-0.008
ww-0.008
Token ride
Token ID6594
Feature activation-7.88e-03

Pos logit contributions
ha+0.007
Little+0.007
Time+0.007
www+0.006
Sarah+0.006
Neg logit contributions
ware-0.013
 fault-0.013
ishes-0.013
unes-0.012
ocular-0.012
Token,
Token ID11
Feature activation-9.05e-03

Pos logit contributions
ha+0.014
Little+0.012
Time+0.012
www+0.012
ww+0.012
Neg logit contributions
anks-0.016
ishes-0.016
ware-0.016
unes-0.015
 fault-0.015
Token so
Token ID523
Feature activation+4.29e-02

Pos logit contributions
ha+0.016
Little+0.014
www+0.013
 few+0.013
Time+0.013
Neg logit contributions
ishes-0.019
anks-0.018
ware-0.018
unes-0.018
 fault-0.018
Token he
Token ID339
Feature activation+7.71e-03

Pos logit contributions
ware+0.063
ishes+0.059
 mouths+0.058
anks+0.058
ocular+0.056
Neg logit contributions
ha-0.087
Time-0.082
Little-0.081
Alice-0.081
www-0.080
Token went
Token ID1816
Feature activation+1.31e-02

Pos logit contributions
ware+0.016
ishes+0.016
anks+0.015
unes+0.015
aces+0.015
Neg logit contributions
ha-0.011
Little-0.011
Time-0.010
www-0.010
Alice-0.010
Token to
Token ID284
Feature activation+3.00e-03

Pos logit contributions
anks+0.024
ware+0.023
unes+0.022
ishes+0.022
 fault+0.021
Neg logit contributions
ha-0.026
Little-0.025
ww-0.024
www-0.023
Time-0.022
Token the
Token ID262
Feature activation-4.48e-03

Pos logit contributions
ware+0.006
anks+0.006
 fault+0.006
ishes+0.006
unes+0.006
Neg logit contributions
ha-0.005
ww-0.004
www-0.004
Little-0.004
Time-0.004
Token st
Token ID336
Feature activation-3.86e-03

Pos logit contributions
ha+0.008
Little+0.007
www+0.007
ww+0.007
Time+0.007
Neg logit contributions
ware-0.009
anks-0.008
ishes-0.008
ocular-0.008
unes-0.008
Token to
Token ID284
Feature activation-1.09e-03

Pos logit contributions
ishes+0.006
unes+0.006
anks+0.006
ware+0.006
ocular+0.006
Neg logit contributions
ha-0.006
Little-0.005
Time-0.005
While-0.005
After-0.005
Token our
Token ID674
Feature activation-1.84e-02

Pos logit contributions
ha+0.002
Little+0.002
www+0.002
Time+0.002
ww+0.001
Neg logit contributions
anks-0.002
ishes-0.002
unes-0.002
ware-0.002
 fault-0.002
Token show
Token ID905
Feature activation+3.05e-02

Pos logit contributions
ha+0.034
Little+0.030
www+0.029
ww+0.028
Time+0.028
Neg logit contributions
ishes-0.036
ware-0.036
anks-0.035
unes-0.035
 fault-0.035
Token!"
Token ID2474
Feature activation-1.85e-03

Pos logit contributions
anks+0.065
ware+0.063
ishes+0.063
unes+0.063
 fault+0.062
Neg logit contributions
ha-0.048
Little-0.044
Time-0.041
www-0.040
T-0.040
Token They
Token ID1119
Feature activation-1.20e-02

Pos logit contributions
ha+0.003
uren+0.002
 few+0.002
unt+0.002
Little+0.002
Neg logit contributions
ishes-0.005
unes-0.004
ware-0.004
 fault-0.004
anks-0.004
Token start
Token ID923
Feature activation-1.40e-02

Pos logit contributions
ha+0.019
Little+0.016
www+0.015
ww+0.015
Time+0.015
Neg logit contributions
ishes-0.027
ware-0.027
anks-0.027
unes-0.026
 fault-0.026
Token to
Token ID284
Feature activation-4.24e-04

Pos logit contributions
ha+0.026
Little+0.023
www+0.022
ww+0.022
unt+0.021
Neg logit contributions
ware-0.028
anks-0.027
ishes-0.027
unes-0.027
 fault-0.027
Token act
Token ID719
Feature activation-3.80e-04

Pos logit contributions
ha+0.001
www+0.001
ww+0.001
Little+0.001
unt+0.001
Neg logit contributions
ware-0.001
ishes-0.001
anks-0.001
 fault-0.001
els-0.001
Token like
Token ID588
Feature activation+8.71e-03

Pos logit contributions
ha+0.001
Little+0.001
www+0.001
ww+0.000
 few+0.000
Neg logit contributions
ishes-0.001
ware-0.001
anks-0.001
unes-0.001
 fault-0.001
Token a
Token ID257
Feature activation-7.01e-03

Pos logit contributions
ishes+0.020
anks+0.020
ware+0.020
unes+0.020
 fault+0.019
Neg logit contributions
ha-0.013
Little-0.011
 few-0.010
www-0.010
Time-0.010
Token pirate
Token ID25868
Feature activation+1.16e-02

Pos logit contributions
ha+0.013
Little+0.012
www+0.011
Time+0.011
Alice+0.011
Neg logit contributions
ware-0.013
 fault-0.012
anks-0.012
ishes-0.012
rier-0.012
Token Lily
Token ID20037
Feature activation+1.31e-02

Pos logit contributions
ha+0.023
www+0.021
Little+0.020
ww+0.020
unt+0.018
Neg logit contributions
ware-0.026
ishes-0.023
unes-0.023
kins-0.022
melon-0.022
Token's
Token ID338
Feature activation+9.22e-03

Pos logit contributions
ishes+0.026
ware+0.024
 fault+0.023
anks+0.022
aces+0.022
Neg logit contributions
ha-0.023
uren-0.020
unt-0.020
Alice-0.020
www-0.020
Token mom
Token ID1995
Feature activation+1.06e-03

Pos logit contributions
ishes+0.022
ware+0.022
unes+0.022
anks+0.022
 fault+0.021
Neg logit contributions
ha-0.012
Little-0.011
 few-0.010
www-0.010
unt-0.010
Tokenmy
Token ID1820
Feature activation-3.55e-04

Pos logit contributions
ishes+0.002
unes+0.002
ware+0.002
anks+0.002
 fault+0.002
Neg logit contributions
ha-0.002
uren-0.002
www-0.002
Alice-0.002
Little-0.002
Token asked
Token ID1965
Feature activation-1.78e-02

Pos logit contributions
ha+0.001
Alice+0.001
Little+0.001
uren+0.001
www+0.001
Neg logit contributions
ishes-0.001
ware-0.001
anks-0.001
unes-0.001
 fault-0.001
Token her
Token ID607
Feature activation-6.40e-03

Pos logit contributions
ha+0.027
www+0.023
Little+0.023
ww+0.023
uren+0.022
Neg logit contributions
ishes-0.040
ware-0.039
unes-0.039
anks-0.038
 fault-0.036
Token to
Token ID284
Feature activation+1.63e-03

Pos logit contributions
ha+0.011
Little+0.010
www+0.010
ww+0.010
Time+0.010
Neg logit contributions
ware-0.013
ishes-0.013
anks-0.012
unes-0.012
 fault-0.012
Token write
Token ID3551
Feature activation+8.19e-03

Pos logit contributions
anks+0.003
ware+0.003
ishes+0.003
unes+0.003
 fault+0.003
Neg logit contributions
ha-0.003
Little-0.003
www-0.003
ww-0.003
Time-0.002
Token her
Token ID607
Feature activation-6.25e-03

Pos logit contributions
ware+0.018
anks+0.018
ishes+0.018
unes+0.018
aces+0.017
Neg logit contributions
ha-0.012
www-0.011
Little-0.011
ww-0.011
Time-0.011
Token name
Token ID1438
Feature activation-1.57e-02

Pos logit contributions
ha+0.008
Little+0.007
 few+0.007
www+0.007
unt+0.006
Neg logit contributions
unes-0.015
ishes-0.015
ware-0.015
anks-0.015
 fault-0.015
Token on
Token ID319
Feature activation-1.79e-02

Pos logit contributions
ha+0.026
Little+0.023
www+0.023
ww+0.022
Time+0.022
Neg logit contributions
ware-0.033
anks-0.032
unes-0.032
ishes-0.031
aces-0.031
Token and
Token ID290
Feature activation-8.72e-04

Pos logit contributions
ware+0.054
ishes+0.054
anks+0.054
unes+0.054
 fault+0.053
Neg logit contributions
ha-0.031
Little-0.026
 few-0.025
www-0.024
unt-0.024
Token ran
Token ID4966
Feature activation-2.05e-03

Pos logit contributions
ha+0.002
Little+0.001
www+0.001
ww+0.001
Time+0.001
Neg logit contributions
ishes-0.002
ware-0.002
unes-0.002
anks-0.002
 fault-0.002
Token to
Token ID284
Feature activation+3.94e-03

Pos logit contributions
Little+0.004
ha+0.004
ww+0.004
www+0.004
unt+0.004
Neg logit contributions
unes-0.003
anks-0.003
 fault-0.003
ocular-0.003
ishes-0.003
Token the
Token ID262
Feature activation-5.15e-03

Pos logit contributions
ware+0.009
 fault+0.009
anks+0.009
ishes+0.009
unes+0.008
Neg logit contributions
ha-0.006
ww-0.006
www-0.005
Little-0.005
unt-0.005
Token grass
Token ID8701
Feature activation-1.36e-02

Pos logit contributions
ha+0.009
Little+0.008
www+0.007
ww+0.007
Time+0.007
Neg logit contributions
ishes-0.011
ware-0.011
anks-0.011
unes-0.011
 fault-0.010
Token.
Token ID13
Feature activation+2.64e-04

Pos logit contributions
ha+0.019
Little+0.016
 few+0.016
unt+0.015
Time+0.015
Neg logit contributions
ishes-0.032
anks-0.031
unes-0.031
ware-0.031
 fault-0.030
Token They
Token ID1119
Feature activation-8.98e-03

Pos logit contributions
unes+0.000
anks+0.000
ware+0.000
ishes+0.000
 fault+0.000
Neg logit contributions
ha-0.001
Little-0.000
www-0.000
ww-0.000
Time-0.000
Token saw
Token ID2497
Feature activation+8.50e-03

Pos logit contributions
ha+0.013
Little+0.012
ww+0.012
www+0.012
Time+0.011
Neg logit contributions
ishes-0.021
ware-0.020
unes-0.019
anks-0.019
 fault-0.019
Token another
Token ID1194
Feature activation+1.35e-02

Pos logit contributions
anks+0.022
ishes+0.022
ware+0.022
unes+0.022
 fault+0.021
Neg logit contributions
ha-0.010
Little-0.008
 few-0.008
Time-0.008
www-0.007
Token boy
Token ID2933
Feature activation+2.76e-03

Pos logit contributions
ware+0.024
 fault+0.024
ishes+0.024
unes+0.023
anks+0.023
Neg logit contributions
ha-0.025
Little-0.022
uren-0.022
Time-0.022
T-0.022
Token playing
Token ID2712
Feature activation-2.38e-04

Pos logit contributions
ishes+0.006
ware+0.006
anks+0.006
unes+0.006
 fault+0.006
Neg logit contributions
ha-0.005
Little-0.004
www-0.004
ww-0.004
unt-0.004
Token,
Token ID11
Feature activation-3.77e-03

Pos logit contributions
ha+0.025
www+0.023
Alice+0.023
unt+0.023
uren+0.022
Neg logit contributions
ishes-0.032
unes-0.031
anks-0.030
ware-0.030
 mouths-0.028
Token the
Token ID262
Feature activation-3.33e-03

Pos logit contributions
ha+0.005
www+0.004
Little+0.004
unt+0.004
ww+0.004
Neg logit contributions
ishes-0.009
anks-0.009
ware-0.009
unes-0.009
 fault-0.009
Token rabbit
Token ID22746
Feature activation+9.83e-03

Pos logit contributions
ha+0.006
www+0.005
Alice+0.005
Little+0.005
ww+0.005
Neg logit contributions
ishes-0.007
ware-0.007
anks-0.007
unes-0.006
 fault-0.006
Token went
Token ID1816
Feature activation+1.52e-02

Pos logit contributions
ishes+0.022
ware+0.022
anks+0.022
unes+0.021
 fault+0.021
Neg logit contributions
ha-0.015
Little-0.013
www-0.012
ww-0.012
Time-0.012
Token back
Token ID736
Feature activation+1.98e-02

Pos logit contributions
anks+0.027
unes+0.026
ishes+0.025
ware+0.025
 fault+0.024
Neg logit contributions
ha-0.032
Little-0.030
www-0.028
ww-0.028
T-0.027
Token to
Token ID284
Feature activation+4.89e-03

Pos logit contributions
ishes+0.041
anks+0.040
ware+0.040
unes+0.040
 fault+0.039
Neg logit contributions
ha-0.035
Little-0.031
www-0.030
ww-0.029
Time-0.029
Token the
Token ID262
Feature activation-3.02e-03

Pos logit contributions
anks+0.010
ware+0.010
 fault+0.010
ishes+0.010
unes+0.010
Neg logit contributions
ha-0.008
Little-0.007
www-0.007
ww-0.007
unt-0.007
Token forest
Token ID8222
Feature activation+1.53e-03

Pos logit contributions
ha+0.006
www+0.005
Little+0.005
uren+0.005
ww+0.005
Neg logit contributions
ware-0.006
ishes-0.006
anks-0.005
 fault-0.005
melon-0.005
Token and
Token ID290
Feature activation+3.96e-03

Pos logit contributions
ishes+0.004
unes+0.004
ware+0.004
anks+0.004
 fault+0.003
Neg logit contributions
ha-0.002
Little-0.002
 few-0.002
www-0.002
Time-0.002
Token looked
Token ID3114
Feature activation+2.54e-03

Pos logit contributions
ishes+0.009
anks+0.008
ware+0.008
unes+0.008
 fault+0.008
Neg logit contributions
ha-0.006
Little-0.006
ww-0.005
www-0.005
unt-0.005
Token for
Token ID329
Feature activation+7.41e-03

Pos logit contributions
ishes+0.005
anks+0.005
ware+0.005
unes+0.005
 fault+0.005
Neg logit contributions
ha-0.005
Little-0.004
www-0.004
ww-0.004
Time-0.004
Token the
Token ID262
Feature activation-2.46e-03

Pos logit contributions
ha+0.006
www+0.005
ww+0.005
unt+0.005
Little+0.005
Neg logit contributions
ware-0.008
unes-0.008
ishes-0.008
anks-0.008
 fault-0.008
Token girl
Token ID2576
Feature activation-1.01e-02

Pos logit contributions
ha+0.004
Little+0.003
www+0.003
ww+0.003
T+0.003
Neg logit contributions
ware-0.005
 fault-0.005
ishes-0.005
unes-0.005
anks-0.005
Token became
Token ID2627
Feature activation+1.52e-02

Pos logit contributions
ha+0.015
ww+0.015
www+0.015
Time+0.014
unt+0.014
Neg logit contributions
ishes-0.021
ware-0.020
aces-0.019
anks-0.019
els-0.019
Token still
Token ID991
Feature activation+1.60e-02

Pos logit contributions
ware+0.031
ishes+0.031
anks+0.031
unes+0.031
 fault+0.030
Neg logit contributions
ha-0.025
Little-0.022
 few-0.022
www-0.021
Time-0.021
Token.
Token ID13
Feature activation+5.33e-03

Pos logit contributions
ishes+0.033
anks+0.033
ware+0.033
 fault+0.033
unes+0.032
Neg logit contributions
ha-0.028
Little-0.024
www-0.023
ww-0.023
unt-0.022
Token Instead
Token ID5455
Feature activation-2.07e-02

Pos logit contributions
ishes+0.012
ware+0.012
anks+0.012
unes+0.012
 fault+0.012
Neg logit contributions
ha-0.008
Little-0.007
www-0.007
 few-0.007
Time-0.006
Token of
Token ID286
Feature activation+1.37e-02

Pos logit contributions
ha+0.034
Little+0.031
www+0.030
ww+0.030
Time+0.030
Neg logit contributions
unes-0.041
ware-0.041
ishes-0.040
anks-0.040
 fault-0.040
Token noise
Token ID7838
Feature activation-1.70e-02

Pos logit contributions
 fault+0.028
ware+0.028
anks+0.028
ishes+0.028
unes+0.028
Neg logit contributions
ha-0.023
Little-0.020
www-0.020
ww-0.019
unt-0.019
Token,
Token ID11
Feature activation-4.76e-03

Pos logit contributions
ha+0.026
www+0.023
ww+0.023
Little+0.023
Time+0.022
Neg logit contributions
ware-0.036
unes-0.035
ishes-0.035
anks-0.035
 fault-0.033
Token she
Token ID673
Feature activation+5.60e-03

Pos logit contributions
ha+0.006
www+0.005
ww+0.005
Little+0.005
unt+0.005
Neg logit contributions
ware-0.012
ishes-0.011
unes-0.011
anks-0.011
 fault-0.011
Token made
Token ID925
Feature activation+1.40e-02

Pos logit contributions
ishes+0.012
ware+0.012
anks+0.012
unes+0.012
 fault+0.012
Neg logit contributions
ha-0.008
Little-0.008
www-0.007
ww-0.007
Time-0.007
Token noise
Token ID7838
Feature activation-1.82e-02

Pos logit contributions
ha+0.044
Little+0.039
www+0.037
Alice+0.036
ww+0.036
Neg logit contributions
ware-0.054
anks-0.053
unes-0.052
ishes-0.051
 fault-0.051
Token.
Token ID13
Feature activation+3.19e-03

Pos logit contributions
ha+0.037
www+0.033
Little+0.033
ww+0.032
uren+0.032
Neg logit contributions
ware-0.031
anks-0.031
unes-0.031
ishes-0.030
 fault-0.030
Token
Token ID198
Feature activation-4.11e-03

Pos logit contributions
ishes+0.008
anks+0.007
ware+0.007
unes+0.007
 fault+0.007
Neg logit contributions
ha-0.004
Little-0.004
Time-0.004
T-0.004
 few-0.004
Token
Token ID198
Feature activation-5.98e-03

Pos logit contributions
ha+0.006
www+0.005
ww+0.005
Little+0.005
 few+0.005
Neg logit contributions
ware-0.009
anks-0.009
ishes-0.009
unes-0.009
 fault-0.009
Token"
Token ID1
Feature activation-8.69e-03

Pos logit contributions
ha+0.007
 Margaret+0.007
uren+0.007
 Kelly+0.006
www+0.006
Neg logit contributions
kins-0.012
auc-0.012
ests-0.012
ases-0.012
els-0.011
TokenStop
Token ID19485
Feature activation-5.27e-02

Pos logit contributions
ha+0.012
www+0.010
unt+0.010
 few+0.010
uren+0.010
Neg logit contributions
ware-0.020
ishes-0.019
unes-0.019
ests-0.018
 fault-0.018
Token it
Token ID340
Feature activation+1.58e-03

Pos logit contributions
ha+0.084
Time+0.077
Little+0.076
Alice+0.074
unt+0.073
Neg logit contributions
 fault-0.109
ishes-0.106
anks-0.103
unes-0.103
ware-0.102
Token,
Token ID11
Feature activation-9.85e-03

Pos logit contributions
anks+0.003
unes+0.003
ishes+0.003
 fault+0.003
ware+0.003
Neg logit contributions
ha-0.002
uren-0.002
Time-0.002
T-0.002
Little-0.002
Token Tom
Token ID4186
Feature activation-1.13e-03

Pos logit contributions
ha+0.016
Little+0.014
www+0.014
unt+0.014
ww+0.014
Neg logit contributions
ishes-0.020
ware-0.020
unes-0.020
 fault-0.020
anks-0.020
Token!"
Token ID2474
Feature activation-6.89e-03

Pos logit contributions
ha+0.002
Little+0.001
Time+0.001
 few+0.001
uren+0.001
Neg logit contributions
ishes-0.003
anks-0.002
 fault-0.002
unes-0.002
ware-0.002
Token L
Token ID406
Feature activation+4.68e-03

Pos logit contributions
ha+0.009
Little+0.008
www+0.007
 few+0.007
Time+0.007
Neg logit contributions
ishes-0.017
anks-0.017
unes-0.017
ware-0.017
 fault-0.016
Token,
Token ID11
Feature activation-9.21e-03

Pos logit contributions
ishes+0.025
ware+0.025
anks+0.025
unes+0.025
 fault+0.024
Neg logit contributions
ha-0.012
Little-0.010
www-0.009
ww-0.009
Time-0.009
Token there
Token ID612
Feature activation+1.49e-03

Pos logit contributions
ha+0.014
www+0.012
Little+0.012
ww+0.011
uren+0.011
Neg logit contributions
ishes-0.020
ware-0.020
 fault-0.020
unes-0.020
anks-0.019
Token was
Token ID373
Feature activation+3.94e-03

Pos logit contributions
ishes+0.003
anks+0.003
ware+0.003
unes+0.003
 fault+0.003
Neg logit contributions
ha-0.002
Little-0.002
www-0.002
Time-0.002
ww-0.002
Token a
Token ID257
Feature activation-1.04e-02

Pos logit contributions
 fault+0.009
unes+0.009
ishes+0.009
ware+0.009
anks+0.009
Neg logit contributions
ha-0.006
www-0.005
Little-0.005
unt-0.005
ww-0.004
Token k
Token ID479
Feature activation-1.87e-02

Pos logit contributions
ha+0.015
www+0.012
Alice+0.011
Time+0.011
 Kelly+0.011
Neg logit contributions
ware-0.023
ishes-0.022
rier-0.021
kins-0.021
melon-0.021
Tokenang
Token ID648
Feature activation-4.14e-02

Pos logit contributions
ha+0.041
Little+0.037
www+0.036
Time+0.036
ww+0.035
Neg logit contributions
ishes-0.030
ware-0.029
anks-0.029
unes-0.029
 fault-0.027
Tokenaroo
Token ID38049
Feature activation-1.23e-02

Pos logit contributions
ha+0.095
Little+0.087
Time+0.086
www+0.084
ww+0.084
Neg logit contributions
ware-0.059
ishes-0.057
unes-0.057
anks-0.056
 fault-0.053
Token.
Token ID13
Feature activation+3.72e-04

Pos logit contributions
ha+0.020
Little+0.018
www+0.017
Time+0.017
ww+0.017
Neg logit contributions
ishes-0.026
ware-0.026
anks-0.026
unes-0.026
 fault-0.025
Token He
Token ID679
Feature activation+7.61e-03

Pos logit contributions
unes+0.001
anks+0.001
ishes+0.001
ware+0.001
aces+0.001
Neg logit contributions
ha-0.001
Little-0.001
www-0.001
Time-0.001
 few-0.001
Token was
Token ID373
Feature activation+3.45e-03

Pos logit contributions
ishes+0.016
ware+0.016
anks+0.016
unes+0.016
 fault+0.015
Neg logit contributions
ha-0.012
Little-0.011
www-0.010
Time-0.010
unt-0.010
Token very
Token ID845
Feature activation+6.03e-03

Pos logit contributions
ishes+0.007
anks+0.007
unes+0.007
ware+0.007
 fault+0.007
Neg logit contributions
ha-0.006
Little-0.005
www-0.005
Time-0.005
ww-0.005
Token but
Token ID475
Feature activation+1.51e-05

Pos logit contributions
ha+0.027
Little+0.023
www+0.023
ww+0.023
unt+0.022
Neg logit contributions
ware-0.024
 fault-0.024
unes-0.023
aces-0.023
ishes-0.023
Token she
Token ID673
Feature activation-2.30e-03

Pos logit contributions
ware+0.000
ishes+0.000
 fault+0.000
unes+0.000
anks+0.000
Neg logit contributions
ha-0.000
Little-0.000
www-0.000
ww-0.000
 few-0.000
Token was
Token ID373
Feature activation+5.10e-03

Pos logit contributions
ha+0.004
Little+0.003
www+0.003
ww+0.003
 few+0.003
Neg logit contributions
ware-0.005
ishes-0.005
anks-0.005
unes-0.005
 fault-0.005
Token very
Token ID845
Feature activation+7.84e-03

Pos logit contributions
 fault+0.009
ishes+0.009
anks+0.009
unes+0.009
ware+0.009
Neg logit contributions
ha-0.010
Little-0.009
www-0.009
ww-0.009
uren-0.008
Token hard
Token ID1327
Feature activation+6.25e-03

Pos logit contributions
ishes+0.014
unes+0.014
ware+0.014
anks+0.014
rier+0.014
Neg logit contributions
ha-0.015
Little-0.012
Time-0.012
uren-0.012
Charlie-0.012
Tokenworking
Token ID16090
Feature activation-5.49e-02

Pos logit contributions
anks+0.013
 fault+0.013
ishes+0.012
ware+0.012
unes+0.012
Neg logit contributions
ha-0.011
Little-0.010
www-0.010
ww-0.009
Time-0.009
Token.
Token ID13
Feature activation-3.21e-03

Pos logit contributions
ha+0.091
Little+0.079
www+0.076
 few+0.075
ww+0.075
Neg logit contributions
ishes-0.116
ware-0.114
anks-0.114
unes-0.114
 fault-0.111
Token She
Token ID1375
Feature activation-5.68e-03

Pos logit contributions
ha+0.004
Little+0.004
Time+0.004
 few+0.004
www+0.003
Neg logit contributions
ishes-0.008
anks-0.008
ware-0.008
unes-0.007
 fault-0.007
Token loved
Token ID6151
Feature activation-5.24e-03

Pos logit contributions
ha+0.009
Little+0.008
ww+0.007
www+0.007
Time+0.007
Neg logit contributions
ware-0.012
ishes-0.012
unes-0.012
 fault-0.012
anks-0.012
Token to
Token ID284
Feature activation+7.10e-04

Pos logit contributions
ha+0.010
Little+0.008
www+0.008
ww+0.008
uren+0.008
Neg logit contributions
ishes-0.010
ware-0.010
unes-0.010
anks-0.010
 fault-0.010
Token pick
Token ID2298
Feature activation-2.80e-03

Pos logit contributions
ware+0.001
anks+0.001
ishes+0.001
unes+0.001
 fault+0.001
Neg logit contributions
ha-0.001
ww-0.001
Little-0.001
Time-0.001
www-0.001
Token and
Token ID290
Feature activation-1.10e-02

Pos logit contributions
ha+0.008
Little+0.007
 few+0.006
unt+0.006
T+0.006
Neg logit contributions
ware-0.014
ishes-0.014
unes-0.014
anks-0.014
 fault-0.013
Token Tom
Token ID4186
Feature activation-5.06e-04

Pos logit contributions
ha+0.015
Little+0.013
www+0.012
ww+0.012
 few+0.012
Neg logit contributions
ishes-0.026
ware-0.026
anks-0.026
unes-0.026
 fault-0.025
Token.
Token ID13
Feature activation-5.58e-03

Pos logit contributions
ha+0.001
 few+0.001
Little+0.001
T+0.000
"+0.000
Neg logit contributions
ware-0.001
ishes-0.001
unes-0.001
anks-0.001
ests-0.001
Token
Token ID198
Feature activation-6.46e-03

Pos logit contributions
ha+0.007
Little+0.006
 few+0.005
Time+0.005
unt+0.005
Neg logit contributions
ishes-0.015
anks-0.014
ware-0.014
unes-0.014
 fault-0.014
Token
Token ID198
Feature activation-7.76e-03

Pos logit contributions
ha+0.009
Little+0.007
www+0.007
Time+0.007
 few+0.007
Neg logit contributions
ishes-0.016
ware-0.016
anks-0.016
unes-0.016
 fault-0.015
TokenS
Token ID50
Feature activation-4.36e-02

Pos logit contributions
ha+0.008
 Kelly+0.006
uren+0.006
www+0.006
Little+0.006
Neg logit contributions
ware-0.019
ishes-0.019
unes-0.019
rier-0.019
anks-0.019
Tokenara
Token ID3301
Feature activation-2.43e-02

Pos logit contributions
ha+0.059
Little+0.049
www+0.047
 few+0.046
ww+0.046
Neg logit contributions
ishes-0.106
ware-0.106
anks-0.105
unes-0.105
 fault-0.103
Token and
Token ID290
Feature activation-1.20e-02

Pos logit contributions
ha+0.034
Little+0.027
"+0.026
www+0.025
unt+0.025
Neg logit contributions
unes-0.057
anks-0.057
ware-0.057
ishes-0.056
 fault-0.055
Token Tom
Token ID4186
Feature activation-1.21e-03

Pos logit contributions
ha+0.014
Little+0.012
www+0.011
ww+0.010
Time+0.010
Neg logit contributions
ishes-0.031
ware-0.031
anks-0.031
unes-0.031
 fault-0.030
Token are
Token ID389
Feature activation+1.68e-03

Pos logit contributions
ha+0.002
Little+0.002
www+0.002
 few+0.002
Time+0.002
Neg logit contributions
ware-0.002
ishes-0.002
unes-0.002
anks-0.002
 fault-0.002
Token scared
Token ID12008
Feature activation-8.44e-03

Pos logit contributions
ware+0.004
unes+0.004
ishes+0.004
anks+0.004
 fault+0.004
Neg logit contributions
ha-0.002
 few-0.002
Little-0.002
Time-0.002
Alice-0.002
Token the
Token ID262
Feature activation-4.90e-03

Pos logit contributions
ha+0.026
Little+0.023
 few+0.022
Time+0.021
unt+0.021
Neg logit contributions
ishes-0.040
anks-0.039
unes-0.039
ware-0.039
 fault-0.038
Token rocket
Token ID10701
Feature activation+2.08e-02

Pos logit contributions
ha+0.010
Little+0.009
www+0.009
ww+0.009
Time+0.008
Neg logit contributions
ware-0.009
ishes-0.009
unes-0.008
anks-0.008
 fault-0.008
Token.
Token ID13
Feature activation-5.37e-03

Pos logit contributions
ishes+0.048
unes+0.048
anks+0.047
 fault+0.047
ware+0.047
Neg logit contributions
ha-0.030
Little-0.025
 few-0.024
unt-0.023
www-0.022
Token They
Token ID1119
Feature activation-1.21e-02

Pos logit contributions
ha+0.010
Little+0.009
Time+0.009
 few+0.009
unt+0.009
Neg logit contributions
ishes-0.010
ware-0.010
anks-0.010
unes-0.009
 fault-0.009
Token count
Token ID954
Feature activation+5.17e-03

Pos logit contributions
ha+0.019
Little+0.017
www+0.016
 few+0.016
ww+0.015
Neg logit contributions
ishes-0.026
anks-0.026
ware-0.026
unes-0.026
 fault-0.026
Token down
Token ID866
Feature activation-4.71e-02

Pos logit contributions
anks+0.013
ishes+0.013
unes+0.012
ware+0.012
 fault+0.012
Neg logit contributions
ha-0.007
Little-0.006
 few-0.006
Time-0.005
www-0.005
Token from
Token ID422
Feature activation+4.27e-02

Pos logit contributions
ha+0.074
Little+0.063
www+0.063
ww+0.062
Time+0.059
Neg logit contributions
ware-0.104
ishes-0.104
anks-0.104
unes-0.103
 fault-0.101
Token ten
Token ID3478
Feature activation-1.12e-02

Pos logit contributions
ishes+0.106
anks+0.106
ware+0.105
unes+0.105
 fault+0.103
Neg logit contributions
ha-0.055
Little-0.046
www-0.043
 few-0.042
ww-0.042
Token to
Token ID284
Feature activation+1.61e-04

Pos logit contributions
ha+0.018
Little+0.016
www+0.016
ww+0.016
Time+0.015
Neg logit contributions
ware-0.024
ishes-0.024
unes-0.023
anks-0.023
 fault-0.022
Token zero
Token ID6632
Feature activation+2.20e-03

Pos logit contributions
ware+0.000
ishes+0.000
anks+0.000
unes+0.000
 fault+0.000
Neg logit contributions
ha-0.000
www-0.000
ww-0.000
Little-0.000
unt-0.000
Token and
Token ID290
Feature activation-6.25e-03

Pos logit contributions
anks+0.005
ishes+0.005
unes+0.005
 fault+0.005
ware+0.005
Neg logit contributions
ha-0.003
Little-0.003
 few-0.003
www-0.003
Time-0.003