TOP ACTIVATIONS
MAX = 0.151

Rudy . He 's a birthday present from my grandma .
twins . They have a birthday today . Mom and Dad
, but sometimes , the endings are not . <|endoftext|> Once
, there was a huge birthday cake and a lot of
time , there was a birthday party in a big hall
time , there was a birthday . It was unusual because
, " Jack has a birthday today and I don 't
girl . She had a birthday . On her birthday ,
Jane . Jane had a birthday coming up , and she
a time there was a birthday . A girl had a
" Hey , hey , hey !" he yelled at the
Suddenly , the birthday girl 's mom said it
Lily . It was her birthday and she was very happy
day , it was his birthday . He was very excited
" Hey , hey , hey ! What is going on
! There was a big birthday party waiting for her with
do ? This is a birthday cake for Mrs . Lee
twins . They have a birthday today . Mom and Dad
day , it was their birthday . They were very happy
games , and a giant birthday cake . The kids were

MIN ACTIVATIONS
MIN = -0.161

share her own l oll ip ops with others . <|endoftext|>
and share her l oll ip ops with others , just
number all her l oll ip ops . She had two
they could buy l oll ip ops for them .
ball and the dog fet ches it . They push the
itty wanted to introduce Dog gy to her ball . She
for lots more l oll ip ops like the one she
get her own l oll ip ops . Betty be amed
ball and the dog fet ches it . They have fun
ball and the dog fet ches it . They hug the
him for remembering his pig gy bank . From that day
count how many l oll ip ops she had . She
walk and give l oll ip ops to children ." The
some money for her pig gy bank . One
ily learned that l oll ip ops are fragile and can
care of her l oll ip ops and other fragile things
money away in her pig gy bank every week . This
I will name him Sno opy . Because he is nos
't want to get anything stuck in his teeth again .
all valued their l oll ip ops the same . They

INTERVAL 0.073 - 0.151
CONTAINS 0.883%

. The next day , the bird was
have together tomorrow . The end . <|endoftext|> Once upon a
and it made a loud noise . Her
, she wants to paint something new . She thinks and
friendly smile . Everyone in the park was so

INTERVAL -0.005 - 0.073
CONTAINS 59.860%

was the day the weird witch learned that discussing can be
tomorrow !" The man felt sorry for her
each other . They nod . They say , " OK
. One day , his mom gave him a present .
, " Stop fighting , children . Why don 't you

INTERVAL -0.083 - -0.005
CONTAINS 38.875%

One day , a new green truck came to play .
� Let â € ™ s play a game ! â
to always listen . She was so happy with her silly
girl wanted to bake something y ummy . They found flour
He was so excited that he started to shake with happiness

INTERVAL -0.161 - -0.083
CONTAINS 0.382%

asked her mom if she could have it , but her
that she thought the pie might be in there . Josh
up to the te ash op and saw a nice lady
hair and you wasted the leather !" Tom knew
her hand . It rub s its head against her hand
Token Rudy
Token ID37602
Feature activation-4.80e-02

Pos logit contributions
 relief+0.020
 gentleman+0.018
 woman+0.018
 cal+0.018
 lady+0.018
Neg logit contributions
ug-0.016
issors-0.015
ors-0.014
om-0.014
utes-0.014
Token.
Token ID13
Feature activation+1.01e-02

Pos logit contributions
ug+0.071
ors+0.068
eds+0.067
issors+0.066
bowl+0.065
Neg logit contributions
 relief-0.089
 gentleman-0.081
 Heidi-0.076
 cal-0.076
 stranger-0.076
Token He
Token ID679
Feature activation+5.27e-03

Pos logit contributions
 relief+0.017
 lady+0.016
 cal+0.016
 service+0.015
 woman+0.015
Neg logit contributions
ug-0.016
issors-0.015
ors-0.015
bar-0.015
bowl-0.014
Token's
Token ID338
Feature activation-3.27e-02

Pos logit contributions
 relief+0.010
 gentleman+0.010
 woman+0.009
 Heidi+0.009
 stranger+0.009
Neg logit contributions
 fet-0.007
om-0.007
ug-0.007
ors-0.007
bowl-0.007
Token a
Token ID257
Feature activation+1.12e-02

Pos logit contributions
ug+0.047
issors+0.046
utes+0.045
bar+0.045
om+0.045
Neg logit contributions
 relief-0.057
 gentleman-0.053
 cal-0.052
 woman-0.050
 stranger-0.050
Token birthday
Token ID10955
Feature activation+1.51e-01

Pos logit contributions
 relief+0.016
 gentleman+0.015
 Heidi+0.015
 cal+0.015
 woman+0.014
Neg logit contributions
ug-0.019
utes-0.018
bar-0.018
bowl-0.017
issors-0.017
Token present
Token ID1944
Feature activation+4.17e-02

Pos logit contributions
 relief+0.266
 cal+0.252
 Heidi+0.248
 gentleman+0.237
 Lord+0.224
Neg logit contributions
ug-0.211
bar-0.200
issors-0.196
bowl-0.195
 clubs-0.195
Token from
Token ID422
Feature activation+1.93e-02

Pos logit contributions
 relief+0.084
 cal+0.077
 gentleman+0.072
 woman+0.070
 Heidi+0.070
Neg logit contributions
ug-0.057
ors-0.055
bar-0.055
issors-0.053
bed-0.051
Token my
Token ID616
Feature activation+1.61e-02

Pos logit contributions
 relief+0.033
 gentleman+0.032
 cal+0.031
 lady+0.030
 woman+0.030
Neg logit contributions
ug-0.030
issors-0.028
utes-0.027
bed-0.027
eds-0.027
Token grandma
Token ID49890
Feature activation+3.04e-02

Pos logit contributions
 relief+0.026
 Heidi+0.025
 gentleman+0.025
 cal+0.025
 stranger+0.024
Neg logit contributions
ug-0.025
utes-0.023
bed-0.023
 clubs-0.022
issors-0.022
Token.
Token ID13
Feature activation+3.23e-03

Pos logit contributions
 relief+0.054
 gentleman+0.049
 cal+0.048
 stranger+0.047
 lady+0.046
Neg logit contributions
ug-0.048
ors-0.046
issors-0.046
eds-0.044
usted-0.043
Token twins
Token ID20345
Feature activation+4.33e-02

Pos logit contributions
 gentleman+0.089
 relief+0.089
 woman+0.084
 lady+0.084
 stranger+0.081
Neg logit contributions
ug-0.103
issors-0.099
usted-0.096
bar-0.095
utes-0.095
Token.
Token ID13
Feature activation+1.22e-02

Pos logit contributions
 relief+0.067
 gentleman+0.065
 lady+0.062
 stranger+0.061
 Heidi+0.060
Neg logit contributions
ug-0.074
usted-0.068
bed-0.068
issors-0.067
ors-0.067
Token They
Token ID1119
Feature activation+1.77e-02

Pos logit contributions
 relief+0.022
 lady+0.021
 gentleman+0.021
 woman+0.020
 stranger+0.020
Neg logit contributions
ug-0.018
issors-0.017
usted-0.017
bar-0.016
bed-0.016
Token have
Token ID423
Feature activation+3.61e-02

Pos logit contributions
 relief+0.032
 gentleman+0.032
 Heidi+0.031
 woman+0.031
 stranger+0.030
Neg logit contributions
ug-0.022
issors-0.021
 built-0.021
usted-0.021
aved-0.021
Token a
Token ID257
Feature activation+5.60e-03

Pos logit contributions
 relief+0.053
 gentleman+0.050
 lady+0.050
 woman+0.049
 cal+0.048
Neg logit contributions
ug-0.065
issors-0.063
usted-0.061
bar-0.060
utes-0.060
Token birthday
Token ID10955
Feature activation+1.46e-01

Pos logit contributions
 relief+0.009
 gentleman+0.008
 cal+0.008
 Heidi+0.008
 lady+0.008
Neg logit contributions
issors-0.009
ug-0.009
utes-0.009
om-0.008
 clubs-0.008
Token today
Token ID1909
Feature activation+4.59e-02

Pos logit contributions
 relief+0.304
 cal+0.298
 Heidi+0.293
 gentleman+0.291
 stranger+0.272
Neg logit contributions
ug-0.151
usted-0.148
issors-0.146
utes-0.145
bed-0.142
Token.
Token ID13
Feature activation+1.08e-02

Pos logit contributions
 relief+0.097
 gentleman+0.095
 Heidi+0.090
 cal+0.089
 stranger+0.086
Neg logit contributions
ug-0.049
utes-0.048
 often-0.046
usted-0.046
 everywhere-0.045
Token Mom
Token ID11254
Feature activation+3.50e-02

Pos logit contributions
 relief+0.021
 gentleman+0.020
 stranger+0.020
 cal+0.020
 lady+0.020
Neg logit contributions
issors-0.014
ug-0.013
utes-0.012
usted-0.012
bowl-0.012
Token and
Token ID290
Feature activation+2.75e-02

Pos logit contributions
 relief+0.068
 gentleman+0.062
 gru+0.060
 stranger+0.057
 Heidi+0.057
Neg logit contributions
usted-0.048
om-0.048
angles-0.047
ors-0.046
 built-0.045
Token Dad
Token ID17415
Feature activation+3.55e-02

Pos logit contributions
 relief+0.047
 gentleman+0.043
 cal+0.043
 gen+0.040
 service+0.039
Neg logit contributions
issors-0.040
usted-0.039
ug-0.038
bar-0.037
bowl-0.037
Token,
Token ID11
Feature activation-1.45e-02

Pos logit contributions
 gentleman+0.017
 relief+0.017
 lady+0.016
 woman+0.016
 Heidi+0.016
Neg logit contributions
ug-0.009
bed-0.009
ors-0.009
 often-0.008
bowl-0.008
Token but
Token ID475
Feature activation+7.13e-03

Pos logit contributions
ug+0.020
usted+0.018
bar+0.018
issors+0.018
bed+0.017
Neg logit contributions
 gentleman-0.027
 relief-0.026
 lady-0.025
 stranger-0.025
 woman-0.025
Token sometimes
Token ID3360
Feature activation-1.13e-02

Pos logit contributions
 gentleman+0.013
 relief+0.013
 lady+0.012
 woman+0.012
 stranger+0.012
Neg logit contributions
ug-0.010
bar-0.009
issors-0.009
bed-0.009
utes-0.009
Token,
Token ID11
Feature activation+1.22e-03

Pos logit contributions
ug+0.015
 often+0.014
bed+0.013
ors+0.013
igs+0.012
Neg logit contributions
 gentleman-0.023
 relief-0.021
 lady-0.019
 woman-0.019
 stranger-0.019
Token the
Token ID262
Feature activation+2.19e-02

Pos logit contributions
 gentleman+0.002
 relief+0.002
 lady+0.002
 stranger+0.002
 woman+0.002
Neg logit contributions
ug-0.002
bed-0.002
usted-0.002
issors-0.002
ors-0.002
Token endings
Token ID38168
Feature activation+1.45e-01

Pos logit contributions
 gentleman+0.036
 Heidi+0.033
 relief+0.030
 Debbie+0.030
 cal+0.030
Neg logit contributions
ug-0.031
 hamm-0.030
bar-0.030
bed-0.029
usted-0.029
Token are
Token ID389
Feature activation+3.08e-02

Pos logit contributions
 gentleman+0.294
 woman+0.273
 Heidi+0.273
 relief+0.272
 lady+0.270
Neg logit contributions
ug-0.193
ors-0.168
 often-0.163
ir-0.157
ert-0.153
Token not
Token ID407
Feature activation+2.42e-02

Pos logit contributions
 gentleman+0.060
 relief+0.058
 woman+0.054
 lady+0.054
 stranger+0.053
Neg logit contributions
ug-0.040
issors-0.037
usted-0.037
 often-0.037
bar-0.036
Token.
Token ID13
Feature activation+3.22e-03

Pos logit contributions
 gentleman+0.046
 relief+0.044
 lady+0.042
 woman+0.041
 stranger+0.039
Neg logit contributions
ug-0.032
bar-0.030
usted-0.030
 often-0.028
issors-0.028
Token<|endoftext|>
Token ID50256
Feature activation+1.40e-02

Pos logit contributions
 relief+0.006
 gentleman+0.005
 lady+0.005
 stranger+0.005
 woman+0.005
Neg logit contributions
ug-0.005
issors-0.005
bar-0.004
usted-0.004
bowl-0.004
TokenOnce
Token ID7454
Feature activation-1.62e-02

Pos logit contributions
 relief+0.021
 gentleman+0.020
 lady+0.019
 woman+0.019
 cal+0.019
Neg logit contributions
ug-0.025
bowl-0.023
bed-0.023
usted-0.023
issors-0.023
Token,
Token ID11
Feature activation-7.50e-03

Pos logit contributions
 relief+0.075
 gentleman+0.071
 woman+0.069
 lady+0.067
 stranger+0.067
Neg logit contributions
usted-0.057
ug-0.056
bar-0.056
bed-0.055
ors-0.051
Token there
Token ID612
Feature activation-2.07e-02

Pos logit contributions
ug+0.010
usted+0.010
bar+0.009
 accidentally+0.009
 everywhere+0.009
Neg logit contributions
 relief-0.014
 cal-0.013
 MacDonald-0.012
 Lord-0.012
 service-0.012
Token was
Token ID373
Feature activation-8.44e-03

Pos logit contributions
ug+0.025
bed+0.024
bar+0.023
usted+0.022
rolled+0.021
Neg logit contributions
 relief-0.041
 gentleman-0.039
 cal-0.038
 service-0.038
 stranger-0.037
Token a
Token ID257
Feature activation+1.12e-02

Pos logit contributions
bar+0.012
 hamm+0.012
usted+0.011
ug+0.011
bed+0.010
Neg logit contributions
 gentleman-0.014
 relief-0.014
vation-0.013
 service-0.013
 woman-0.013
Token huge
Token ID3236
Feature activation+1.25e-02

Pos logit contributions
 Heidi+0.017
ment+0.016
 relief+0.016
 gentleman+0.015
 Eva+0.015
Neg logit contributions
 hamm-0.016
om-0.015
 hammer-0.015
ug-0.015
bar-0.015
Token birthday
Token ID10955
Feature activation+1.43e-01

Pos logit contributions
Sleep+0.019
 Heidi+0.019
ment+0.019
 cal+0.018
 Debbie+0.018
Neg logit contributions
ug-0.017
 hammer-0.017
om-0.016
 clubs-0.016
bar-0.016
Token cake
Token ID12187
Feature activation+7.09e-02

Pos logit contributions
 cal+0.261
 relief+0.251
 Heidi+0.250
 nods+0.233
 bluff+0.229
Neg logit contributions
 everywhere-0.165
om-0.157
bar-0.153
 clubs-0.149
 built-0.146
Token and
Token ID290
Feature activation-3.46e-03

Pos logit contributions
 relief+0.150
 cal+0.144
 gentleman+0.130
 Lord+0.130
 Heidi+0.126
Neg logit contributions
bar-0.080
bed-0.080
 everywhere-0.079
usted-0.074
ors-0.073
Token a
Token ID257
Feature activation+3.99e-03

Pos logit contributions
bar+0.004
usted+0.004
 lots+0.004
 built+0.004
bed+0.004
Neg logit contributions
 relief-0.006
 cal-0.006
 gentleman-0.006
 service-0.005
 Lord-0.005
Token lot
Token ID1256
Feature activation+1.56e-02

Pos logit contributions
 Heidi+0.006
 relief+0.006
 cal+0.006
 gentleman+0.005
 service+0.005
Neg logit contributions
om-0.006
bar-0.006
utes-0.006
 hamm-0.006
bowl-0.006
Token of
Token ID286
Feature activation+1.12e-02

Pos logit contributions
 relief+0.040
 gentleman+0.038
 Heidi+0.037
 woman+0.037
 MacDonald+0.036
Neg logit contributions
usted-0.010
bar-0.009
ug-0.008
om-0.008
issors-0.008
Token time
Token ID640
Feature activation+3.67e-02

Pos logit contributions
 relief+0.010
 lady+0.009
 gentleman+0.009
 woman+0.009
 cal+0.009
Neg logit contributions
ug-0.013
issors-0.012
usted-0.012
ors-0.012
bar-0.012
Token,
Token ID11
Feature activation-3.78e-03

Pos logit contributions
 relief+0.060
 gentleman+0.056
 woman+0.055
 lady+0.055
 cal+0.054
Neg logit contributions
ug-0.063
issors-0.058
bed-0.057
ors-0.056
bar-0.056
Token there
Token ID612
Feature activation-2.94e-02

Pos logit contributions
ug+0.007
issors+0.007
usted+0.007
ors+0.007
bar+0.007
Neg logit contributions
 relief-0.005
 lady-0.005
 gentleman-0.005
 woman-0.005
 stranger-0.005
Token was
Token ID373
Feature activation+2.83e-03

Pos logit contributions
ug+0.051
bed+0.048
issors+0.046
bar+0.046
usted+0.046
Neg logit contributions
 relief-0.045
 gentleman-0.043
 lady-0.042
 woman-0.042
 cal-0.042
Token a
Token ID257
Feature activation+1.36e-02

Pos logit contributions
 relief+0.003
 lady+0.003
 gentleman+0.003
 woman+0.003
 cal+0.003
Neg logit contributions
ug-0.006
issors-0.006
usted-0.006
ors-0.006
bar-0.006
Token birthday
Token ID10955
Feature activation+1.43e-01

Pos logit contributions
 relief+0.019
 Heidi+0.017
 gentleman+0.017
 service+0.017
 cal+0.017
Neg logit contributions
ug-0.024
utes-0.023
om-0.022
bowl-0.021
issors-0.021
Token party
Token ID2151
Feature activation+7.50e-02

Pos logit contributions
 relief+0.286
 cal+0.277
 Heidi+0.258
 confusion+0.251
 gentleman+0.247
Neg logit contributions
ug-0.159
utes-0.150
om-0.150
 everywhere-0.140
issors-0.136
Token in
Token ID287
Feature activation-3.04e-02

Pos logit contributions
 relief+0.168
 cal+0.159
 gentleman+0.145
 sigh+0.140
 Heidi+0.140
Neg logit contributions
ug-0.083
utes-0.078
bed-0.077
issors-0.075
usted-0.074
Token a
Token ID257
Feature activation-8.77e-03

Pos logit contributions
 both+0.032
 lots+0.031
issors+0.028
 hamm+0.027
 everywhere+0.026
Neg logit contributions
 gentleman-0.061
 cal-0.059
 relief-0.055
 stranger-0.054
 therapist-0.054
Token big
Token ID1263
Feature activation+1.47e-02

Pos logit contributions
ug+0.016
issors+0.016
utes+0.015
usted+0.015
bar+0.015
Neg logit contributions
 relief-0.012
 gentleman-0.011
 cal-0.011
 lady-0.011
 woman-0.011
Token hall
Token ID6899
Feature activation+4.55e-02

Pos logit contributions
 relief+0.025
 gentleman+0.024
 cal+0.024
 Heidi+0.023
 lady+0.022
Neg logit contributions
ug-0.021
bowl-0.020
issors-0.020
eds-0.020
bar-0.019
Token time
Token ID640
Feature activation+3.02e-02

Pos logit contributions
 relief+0.001
 lady+0.001
 gentleman+0.001
 woman+0.001
 cal+0.001
Neg logit contributions
ug-0.001
issors-0.001
usted-0.001
ors-0.001
bar-0.001
Token,
Token ID11
Feature activation-7.17e-03

Pos logit contributions
 relief+0.045
 lady+0.042
 gentleman+0.042
 woman+0.041
 cal+0.041
Neg logit contributions
ug-0.055
issors-0.052
bed-0.050
usted-0.050
ors-0.050
Token there
Token ID612
Feature activation-3.11e-02

Pos logit contributions
issors+0.015
ug+0.015
ors+0.014
usted+0.014
bar+0.014
Neg logit contributions
 relief-0.008
 lady-0.008
 woman-0.008
 stranger-0.007
 gentleman-0.007
Token was
Token ID373
Feature activation-1.03e-03

Pos logit contributions
ug+0.057
issors+0.054
bed+0.053
usted+0.053
bar+0.052
Neg logit contributions
 relief-0.044
 gentleman-0.042
 lady-0.042
 woman-0.041
 cal-0.040
Token a
Token ID257
Feature activation+1.23e-02

Pos logit contributions
ug+0.003
issors+0.003
ors+0.003
bar+0.002
eds+0.002
Neg logit contributions
 relief-0.001
 lady-0.001
 woman-0.001
 stranger-0.001
 Heidi-0.001
Token birthday
Token ID10955
Feature activation+1.41e-01

Pos logit contributions
 relief+0.016
 gentleman+0.014
 cal+0.014
 service+0.014
 Heidi+0.014
Neg logit contributions
ug-0.024
utes-0.022
issors-0.022
bowl-0.021
bed-0.021
Token.
Token ID13
Feature activation-3.87e-03

Pos logit contributions
 relief+0.270
 cal+0.261
 Heidi+0.245
 gentleman+0.231
 confusion+0.230
Neg logit contributions
ug-0.174
utes-0.167
om-0.165
 everywhere-0.153
issors-0.152
Token It
Token ID632
Feature activation+1.05e-02

Pos logit contributions
ug+0.005
issors+0.005
usted+0.005
bed+0.005
bowl+0.005
Neg logit contributions
 cal-0.007
 relief-0.007
 gentleman-0.007
 stranger-0.006
 lady-0.006
Token was
Token ID373
Feature activation+1.45e-03

Pos logit contributions
 relief+0.022
 gentleman+0.021
 cal+0.020
 service+0.019
 stranger+0.019
Neg logit contributions
bed-0.012
bowl-0.011
 built-0.011
 often-0.011
 accidentally-0.010
Token unusual
Token ID8468
Feature activation+1.74e-02

Pos logit contributions
 relief+0.002
 lady+0.002
 gentleman+0.002
 woman+0.002
 stranger+0.002
Neg logit contributions
ug-0.003
issors-0.003
ors-0.003
usted-0.003
bar-0.003
Token because
Token ID780
Feature activation+1.36e-02

Pos logit contributions
 relief+0.031
 cal+0.029
 gentleman+0.028
 Heidi+0.027
 woman+0.026
Neg logit contributions
ug-0.027
ors-0.025
usted-0.025
issors-0.025
utes-0.025
Token,
Token ID11
Feature activation+3.94e-03

Pos logit contributions
ug+0.009
bed+0.009
bar+0.008
usted+0.008
ors+0.008
Neg logit contributions
 relief-0.010
 gentleman-0.010
 cal-0.010
 stranger-0.010
 service-0.010
Token "
Token ID366
Feature activation-5.58e-02

Pos logit contributions
 relief+0.007
 gentleman+0.007
 cal+0.007
 service+0.007
 stranger+0.006
Neg logit contributions
ug-0.005
bar-0.005
issors-0.005
usted-0.005
bed-0.005
TokenJack
Token ID14295
Feature activation-1.12e-02

Pos logit contributions
ug+0.087
bar+0.086
issors+0.085
utes+0.083
bed+0.083
Neg logit contributions
 relief-0.094
 cal-0.087
 witch-0.083
 stranger-0.083
 service-0.083
Token has
Token ID468
Feature activation+6.32e-03

Pos logit contributions
ug+0.015
ors+0.014
om+0.014
bed+0.014
bar+0.013
Neg logit contributions
 relief-0.022
 gentleman-0.021
 stranger-0.021
 woman-0.020
 pastor-0.020
Token a
Token ID257
Feature activation+1.05e-02

Pos logit contributions
 relief+0.009
 gentleman+0.009
 lady+0.009
 woman+0.009
 cal+0.008
Neg logit contributions
ug-0.011
issors-0.011
usted-0.011
bar-0.011
ors-0.010
Token birthday
Token ID10955
Feature activation+1.41e-01

Pos logit contributions
 relief+0.015
 gentleman+0.014
 cal+0.014
 lady+0.014
 Heidi+0.014
Neg logit contributions
issors-0.017
ug-0.017
bar-0.017
utes-0.016
bowl-0.016
Token today
Token ID1909
Feature activation+3.78e-02

Pos logit contributions
 relief+0.265
 cal+0.259
 Heidi+0.248
 gentleman+0.247
 woman+0.238
Neg logit contributions
ug-0.170
bar-0.169
issors-0.158
usted-0.158
bowl-0.154
Token and
Token ID290
Feature activation+3.75e-03

Pos logit contributions
 relief+0.078
 gentleman+0.075
 woman+0.071
 cal+0.070
 stranger+0.068
Neg logit contributions
ug-0.043
ors-0.039
usted-0.039
 often-0.039
bar-0.038
Token I
Token ID314
Feature activation+2.51e-02

Pos logit contributions
 relief+0.006
 gentleman+0.006
 cal+0.006
 lady+0.005
 woman+0.005
Neg logit contributions
ug-0.006
bed-0.005
usted-0.005
bar-0.005
issors-0.005
Token don
Token ID836
Feature activation+1.17e-03

Pos logit contributions
 relief+0.048
 gentleman+0.048
 Heidi+0.048
 woman+0.048
 lady+0.047
Neg logit contributions
ug-0.029
bowl-0.029
aved-0.028
 built-0.027
bar-0.026
Token't
Token ID470
Feature activation-3.95e-02

Pos logit contributions
 lady+0.002
 relief+0.002
 gentleman+0.002
 stranger+0.002
 cal+0.002
Neg logit contributions
ug-0.002
issors-0.002
leys-0.002
bowl-0.002
eds-0.002
Token girl
Token ID2576
Feature activation+2.87e-02

Pos logit contributions
 relief+0.008
 cal+0.007
 Heidi+0.006
 gentleman+0.006
 service+0.006
Neg logit contributions
ug-0.019
bowl-0.018
bar-0.018
utes-0.018
issors-0.018
Token.
Token ID13
Feature activation-5.90e-03

Pos logit contributions
 lady+0.047
 woman+0.046
 relief+0.044
 gentleman+0.042
 witch+0.042
Neg logit contributions
issors-0.048
ug-0.046
usted-0.046
bar-0.046
leys-0.045
Token She
Token ID1375
Feature activation+1.47e-02

Pos logit contributions
ug+0.010
issors+0.010
bed+0.010
usted+0.010
ors+0.010
Neg logit contributions
 relief-0.009
 gentleman-0.008
 lady-0.008
 cal-0.008
 stranger-0.008
Token had
Token ID550
Feature activation-2.71e-02

Pos logit contributions
 relief+0.029
 gentleman+0.028
 cal+0.028
 lady+0.027
 stranger+0.027
Neg logit contributions
ug-0.019
bed-0.018
issors-0.017
bowl-0.016
eds-0.016
Token a
Token ID257
Feature activation+6.41e-03

Pos logit contributions
ug+0.061
issors+0.059
ors+0.057
usted+0.057
bar+0.056
Neg logit contributions
 relief-0.029
 lady-0.027
 woman-0.026
 gentleman-0.026
 stranger-0.025
Token birthday
Token ID10955
Feature activation+1.40e-01

Pos logit contributions
 relief+0.010
 gentleman+0.009
 cal+0.009
 lady+0.009
 service+0.009
Neg logit contributions
ug-0.011
issors-0.010
utes-0.010
bed-0.010
bar-0.010
Token.
Token ID13
Feature activation+3.86e-03

Pos logit contributions
 relief+0.268
 cal+0.266
 Heidi+0.250
 gentleman+0.246
 gru+0.231
Neg logit contributions
issors-0.164
ug-0.163
bed-0.158
om-0.157
bowl-0.156
Token On
Token ID1550
Feature activation-1.21e-02

Pos logit contributions
 relief+0.007
 gentleman+0.006
 cal+0.006
 lady+0.006
 stranger+0.006
Neg logit contributions
ug-0.006
issors-0.006
bed-0.005
usted-0.005
bowl-0.005
Token her
Token ID607
Feature activation+3.77e-02

Pos logit contributions
issors+0.019
bed+0.019
ug+0.019
bar+0.019
utes+0.018
Neg logit contributions
 relief-0.019
 gentleman-0.019
 cal-0.018
 stranger-0.018
 woman-0.017
Token birthday
Token ID10955
Feature activation+1.17e-01

Pos logit contributions
 gentleman+0.068
 relief+0.067
 cal+0.066
 stranger+0.063
 Heidi+0.062
Neg logit contributions
bed-0.051
issors-0.048
utes-0.047
ug-0.046
usted-0.045
Token,
Token ID11
Feature activation-4.93e-03

Pos logit contributions
 relief+0.254
 cal+0.245
 gentleman+0.234
 gru+0.218
 Heidi+0.218
Neg logit contributions
ug-0.121
bed-0.115
om-0.112
issors-0.108
bowl-0.106
Token Jane
Token ID12091
Feature activation+1.77e-02

Pos logit contributions
ug+0.033
issors+0.032
bed+0.030
usted+0.030
ors+0.029
Neg logit contributions
 relief-0.027
 cal-0.026
 gentleman-0.026
 stranger-0.024
 woman-0.024
Token.
Token ID13
Feature activation+2.14e-03

Pos logit contributions
 lady+0.017
 woman+0.017
 relief+0.017
 gentleman+0.015
 stranger+0.015
Neg logit contributions
ug-0.041
issors-0.040
bar-0.039
usted-0.038
utes-0.038
Token Jane
Token ID12091
Feature activation+2.35e-02

Pos logit contributions
 relief+0.003
 gentleman+0.003
 woman+0.003
 cal+0.003
 lady+0.003
Neg logit contributions
ug-0.004
issors-0.004
usted-0.004
bed-0.004
ors-0.004
Token had
Token ID550
Feature activation-2.67e-02

Pos logit contributions
 relief+0.038
 lady+0.037
 gentleman+0.036
 woman+0.036
 cal+0.035
Neg logit contributions
ug-0.038
issors-0.037
usted-0.035
bar-0.035
ors-0.035
Token a
Token ID257
Feature activation+7.79e-03

Pos logit contributions
ug+0.057
issors+0.055
usted+0.053
bar+0.053
ors+0.053
Neg logit contributions
 relief-0.031
 lady-0.029
 gentleman-0.028
 woman-0.028
 cal-0.027
Token birthday
Token ID10955
Feature activation+1.40e-01

Pos logit contributions
 relief+0.012
 gentleman+0.011
 cal+0.011
 lady+0.011
 woman+0.010
Neg logit contributions
ug-0.013
issors-0.013
utes-0.013
bed-0.012
bar-0.012
Token coming
Token ID2406
Feature activation+6.81e-02

Pos logit contributions
 relief+0.257
 cal+0.256
 Heidi+0.239
 gentleman+0.237
 pastor+0.222
Neg logit contributions
issors-0.177
ug-0.174
bowl-0.169
bed-0.168
om-0.167
Token up
Token ID510
Feature activation+4.09e-02

Pos logit contributions
 relief+0.131
 lady+0.130
 gentleman+0.126
 woman+0.126
 stranger+0.124
Neg logit contributions
ug-0.091
issors-0.085
usted-0.081
bar-0.081
ors-0.080
Token,
Token ID11
Feature activation-1.48e-02

Pos logit contributions
 relief+0.069
 gentleman+0.065
 lady+0.064
 woman+0.063
 cal+0.063
Neg logit contributions
ug-0.066
issors-0.063
usted-0.061
utes-0.060
bed-0.060
Token and
Token ID290
Feature activation+8.00e-03

Pos logit contributions
bed+0.024
usted+0.024
issors+0.023
ug+0.023
 everywhere+0.023
Neg logit contributions
 relief-0.022
 cal-0.021
 gentleman-0.021
 service-0.020
 payoff-0.019
Token she
Token ID673
Feature activation-5.56e-03

Pos logit contributions
 relief+0.010
 gentleman+0.010
 cal+0.009
 stranger+0.009
 lady+0.009
Neg logit contributions
bed-0.015
usted-0.015
ug-0.014
 often-0.014
bar-0.014
Token a
Token ID257
Feature activation+8.97e-03

Pos logit contributions
ug+0.009
bed+0.009
bowl+0.009
usted+0.009
leys+0.008
Neg logit contributions
 gentleman-0.011
 relief-0.010
 service-0.010
 lady-0.010
 cal-0.010
Token time
Token ID640
Feature activation+3.62e-02

Pos logit contributions
 relief+0.013
 gentleman+0.013
 lady+0.013
 woman+0.012
 cal+0.012
Neg logit contributions
ug-0.016
issors-0.015
usted-0.015
bed-0.015
utes-0.015
Token there
Token ID612
Feature activation-3.18e-02

Pos logit contributions
 relief+0.061
 gentleman+0.057
 woman+0.056
 lady+0.056
 cal+0.056
Neg logit contributions
ug-0.060
issors-0.055
bed-0.055
ors-0.054
bar-0.053
Token was
Token ID373
Feature activation+8.53e-05

Pos logit contributions
ug+0.047
bed+0.046
bowl+0.043
bar+0.042
ors+0.042
Neg logit contributions
 relief-0.056
 gentleman-0.054
 woman-0.053
 cal-0.053
 lady-0.052
Token a
Token ID257
Feature activation+1.17e-02

Pos logit contributions
 relief+0.000
 gentleman+0.000
 lady+0.000
 woman+0.000
 cal+0.000
Neg logit contributions
ug-0.000
issors-0.000
usted-0.000
bed-0.000
utes-0.000
Token birthday
Token ID10955
Feature activation+1.40e-01

Pos logit contributions
 relief+0.016
 gentleman+0.015
 Heidi+0.015
 service+0.015
 cal+0.015
Neg logit contributions
ug-0.020
utes-0.019
om-0.019
issors-0.018
bowl-0.018
Token.
Token ID13
Feature activation-4.27e-03

Pos logit contributions
 relief+0.276
 cal+0.264
 Heidi+0.250
 confusion+0.244
 gentleman+0.242
Neg logit contributions
ug-0.156
utes-0.151
om-0.147
 everywhere-0.140
 often-0.138
Token A
Token ID317
Feature activation+3.06e-02

Pos logit contributions
issors+0.005
ug+0.005
bed+0.004
usted+0.004
om+0.004
Neg logit contributions
 cal-0.009
 gentleman-0.008
 relief-0.008
 stranger-0.008
 man-0.008
Token girl
Token ID2576
Feature activation+3.97e-02

Pos logit contributions
 relief+0.050
 Heidi+0.049
 cal+0.048
ment+0.045
handed+0.044
Neg logit contributions
utes-0.040
om-0.039
 accidentally-0.039
 hamm-0.037
 hammer-0.036
Token had
Token ID550
Feature activation-2.27e-02

Pos logit contributions
 relief+0.092
 cal+0.088
 gentleman+0.085
 stranger+0.083
 Heidi+0.082
Neg logit contributions
ug-0.038
bed-0.037
bowl-0.034
om-0.033
issors-0.032
Token a
Token ID257
Feature activation+7.91e-03

Pos logit contributions
ug+0.043
issors+0.042
usted+0.041
bed+0.041
bar+0.040
Neg logit contributions
 relief-0.031
 gentleman-0.029
 cal-0.029
 lady-0.027
 stranger-0.027
Token"
Token ID1
Feature activation-9.76e-04

Pos logit contributions
 relief+0.043
 lady+0.041
 gentleman+0.040
 woman+0.040
 stranger+0.039
Neg logit contributions
ug-0.048
issors-0.046
ors-0.045
usted-0.045
bed-0.044
TokenHey
Token ID10814
Feature activation+8.39e-02

Pos logit contributions
bar+0.001
issors+0.001
utes+0.001
bowl+0.001
usted+0.001
Neg logit contributions
 relief-0.002
 cal-0.002
 service-0.001
 witch-0.001
 prayer-0.001
Token,
Token ID11
Feature activation+2.21e-02

Pos logit contributions
 relief+0.162
 cal+0.157
 Lord+0.156
 woman+0.156
 gentleman+0.155
Neg logit contributions
ug-0.098
bar-0.082
bed-0.081
om-0.080
ors-0.079
Token hey
Token ID17207
Feature activation+1.39e-01

Pos logit contributions
 relief+0.033
 gentleman+0.031
 service+0.030
 cal+0.030
 title+0.029
Neg logit contributions
ug-0.035
bar-0.034
issors-0.033
bowl-0.031
usted-0.031
Token,
Token ID11
Feature activation+3.16e-02

Pos logit contributions
 relief+0.272
 gentleman+0.270
 woman+0.258
 stranger+0.255
 cal+0.255
Neg logit contributions
ug-0.186
bed-0.162
ors-0.159
eds-0.155
usted-0.152
Token hey
Token ID17207
Feature activation+1.40e-01

Pos logit contributions
 gentleman+0.046
 woman+0.044
 relief+0.042
 Heidi+0.041
 lady+0.041
Neg logit contributions
ug-0.053
bar-0.050
issors-0.050
usted-0.048
 hamm-0.047
Token!"
Token ID2474
Feature activation+2.28e-02

Pos logit contributions
 gentleman+0.285
 relief+0.278
 woman+0.269
 stranger+0.266
 lady+0.261
Neg logit contributions
ug-0.181
bed-0.153
eds-0.152
ors-0.151
usted-0.145
Token he
Token ID339
Feature activation-1.10e-02

Pos logit contributions
 relief+0.038
 gentleman+0.037
 cal+0.037
 service+0.036
 stranger+0.035
Neg logit contributions
ug-0.033
issors-0.031
bed-0.031
bar-0.030
usted-0.029
Token yelled
Token ID22760
Feature activation+8.09e-03

Pos logit contributions
ug+0.018
usted+0.017
issors+0.017
bed+0.016
ors+0.016
Neg logit contributions
 relief-0.018
 gentleman-0.017
 service-0.016
 woman-0.016
 lady-0.016
Token at
Token ID379
Feature activation+1.28e-02

Pos logit contributions
 relief+0.014
 gentleman+0.013
 woman+0.013
 service+0.013
 stranger+0.013
Neg logit contributions
ug-0.013
bed-0.012
usted-0.012
ors-0.012
issors-0.012
Token the
Token ID262
Feature activation+2.46e-02

Pos logit contributions
 relief+0.024
 gentleman+0.023
 cal+0.023
 service+0.022
 lady+0.022
Neg logit contributions
ug-0.018
bed-0.016
issors-0.016
ors-0.016
ir-0.015
Token
Token ID198
Feature activation-1.22e-02

Pos logit contributions
ug+0.111
issors+0.105
bar+0.103
usted+0.102
ors+0.102
Neg logit contributions
 relief-0.065
 lady-0.065
 gentleman-0.061
 woman-0.061
 cal-0.060
Token
Token ID198
Feature activation+3.22e-03

Pos logit contributions
ug+0.026
issors+0.025
bar+0.025
ors+0.025
usted+0.025
Neg logit contributions
 lady-0.014
 relief-0.013
 gentleman-0.013
 woman-0.013
 stranger-0.012
TokenSuddenly
Token ID38582
Feature activation+2.82e-02

Pos logit contributions
 relief+0.005
 lady+0.005
 gentleman+0.005
 woman+0.005
 stranger+0.005
Neg logit contributions
ug-0.006
issors-0.005
ors-0.005
bar-0.005
utes-0.005
Token,
Token ID11
Feature activation+2.17e-02

Pos logit contributions
 gentleman+0.054
 relief+0.053
 stranger+0.050
 woman+0.050
 service+0.050
Neg logit contributions
ug-0.039
bed-0.036
usted-0.034
bar-0.033
ors-0.033
Token the
Token ID262
Feature activation+3.39e-02

Pos logit contributions
 gentleman+0.035
 cal+0.034
 relief+0.033
 service+0.032
 Lord+0.032
Neg logit contributions
bed-0.034
ug-0.032
 accidentally-0.032
usted-0.031
bar-0.030
Token birthday
Token ID10955
Feature activation+1.39e-01

Pos logit contributions
 gentleman+0.052
 Heidi+0.050
 relief+0.050
 cal+0.049
 MacDonald+0.047
Neg logit contributions
usted-0.047
 bricks-0.046
 cubes-0.046
bar-0.046
 everywhere-0.045
Token girl
Token ID2576
Feature activation+1.62e-02

Pos logit contributions
 relief+0.224
 cal+0.221
 gentleman+0.216
 Heidi+0.205
 bluff+0.202
Neg logit contributions
usted-0.189
 accidentally-0.183
 built-0.181
om-0.178
ug-0.178
Token's
Token ID338
Feature activation-4.11e-02

Pos logit contributions
 relief+0.032
 gentleman+0.030
 service+0.029
 cal+0.029
 lady+0.028
Neg logit contributions
bed-0.021
ug-0.020
usted-0.020
bar-0.019
 accidentally-0.019
Token mom
Token ID1995
Feature activation+2.93e-02

Pos logit contributions
bed+0.057
ug+0.056
usted+0.056
om+0.054
issors+0.053
Neg logit contributions
 gentleman-0.070
 Heidi-0.066
 relief-0.066
 cal-0.065
 MacDonald-0.064
Token said
Token ID531
Feature activation-1.13e-02

Pos logit contributions
 relief+0.046
 gentleman+0.044
 lady+0.043
 woman+0.042
 stranger+0.042
Neg logit contributions
ug-0.050
issors-0.047
usted-0.047
bar-0.046
bed-0.046
Token it
Token ID340
Feature activation-5.68e-03

Pos logit contributions
ug+0.015
bed+0.015
usted+0.014
bar+0.014
issors+0.014
Neg logit contributions
 relief-0.020
 gentleman-0.020
 cal-0.019
 woman-0.019
 lady-0.018
Token Lily
Token ID20037
Feature activation-2.50e-04

Pos logit contributions
ug+0.095
bar+0.089
ors+0.088
utes+0.088
issors+0.088
Neg logit contributions
 lady-0.041
 relief-0.041
 woman-0.037
 gentleman-0.035
 stranger-0.035
Token.
Token ID13
Feature activation+2.36e-03

Pos logit contributions
ug+0.001
issors+0.001
usted+0.001
utes+0.001
bar+0.001
Neg logit contributions
 relief-0.000
 lady-0.000
 woman-0.000
 gentleman-0.000
 stranger-0.000
Token It
Token ID632
Feature activation+1.79e-02

Pos logit contributions
 lady+0.002
 relief+0.002
 woman+0.001
 gentleman+0.001
 girl+0.001
Neg logit contributions
ug-0.006
eds-0.006
utes-0.006
bar-0.006
ors-0.006
Token was
Token ID373
Feature activation+1.00e-02

Pos logit contributions
 relief+0.026
 gentleman+0.024
 lady+0.024
 woman+0.024
 cal+0.023
Neg logit contributions
ug-0.032
issors-0.031
bed-0.030
usted-0.030
bar-0.030
Token her
Token ID607
Feature activation+3.71e-02

Pos logit contributions
 relief+0.011
 lady+0.011
 woman+0.011
 cal+0.010
 gentleman+0.010
Neg logit contributions
ug-0.022
ors-0.021
issors-0.021
utes-0.020
bar-0.020
Token birthday
Token ID10955
Feature activation+1.39e-01

Pos logit contributions
 relief+0.061
 gentleman+0.058
 cal+0.057
 lady+0.056
 woman+0.055
Neg logit contributions
ug-0.060
issors-0.058
utes-0.056
usted-0.055
bed-0.055
Token and
Token ID290
Feature activation+2.03e-02

Pos logit contributions
 relief+0.270
 cal+0.262
 gentleman+0.255
 Heidi+0.243
 stranger+0.241
Neg logit contributions
ug-0.176
issors-0.166
bed-0.164
usted-0.160
bowl-0.153
Token she
Token ID673
Feature activation-4.91e-03

Pos logit contributions
 relief+0.027
 cal+0.026
 gentleman+0.026
 service+0.024
 stranger+0.024
Neg logit contributions
bed-0.036
ug-0.035
usted-0.035
issors-0.034
bar-0.034
Token was
Token ID373
Feature activation-1.63e-03

Pos logit contributions
ug+0.007
bed+0.007
issors+0.007
usted+0.007
ors+0.007
Neg logit contributions
 relief-0.009
 gentleman-0.008
 lady-0.008
 cal-0.008
 stranger-0.008
Token very
Token ID845
Feature activation+1.40e-02

Pos logit contributions
ug+0.003
issors+0.002
usted+0.002
ors+0.002
bar+0.002
Neg logit contributions
 relief-0.003
 lady-0.003
 gentleman-0.003
 woman-0.003
 cal-0.003
Token happy
Token ID3772
Feature activation+2.92e-02

Pos logit contributions
 relief+0.022
 gentleman+0.020
 cal+0.020
 Heidi+0.020
 woman+0.019
Neg logit contributions
ug-0.023
issors-0.021
utes-0.021
om-0.021
usted-0.020
Token day
Token ID1110
Feature activation+5.68e-02

Pos logit contributions
 relief+0.013
 cal+0.013
 gentleman+0.013
 Heidi+0.013
 MacDonald+0.012
Neg logit contributions
ug-0.006
bowl-0.004
utes-0.004
om-0.004
issors-0.004
Token,
Token ID11
Feature activation+1.78e-02

Pos logit contributions
 relief+0.120
 cal+0.114
 woman+0.110
 witch+0.110
 Fairy+0.108
Neg logit contributions
ug-0.071
usted-0.065
bed-0.062
issors-0.062
om-0.061
Token it
Token ID340
Feature activation+3.48e-02

Pos logit contributions
 relief+0.030
 cal+0.029
 witch+0.028
 Fairy+0.028
 gentleman+0.027
Neg logit contributions
issors-0.027
usted-0.027
ug-0.027
bar-0.026
bed-0.025
Token was
Token ID373
Feature activation-2.72e-03

Pos logit contributions
 relief+0.060
 gentleman+0.057
 woman+0.055
 lady+0.054
 service+0.054
Neg logit contributions
ug-0.052
usted-0.051
bed-0.050
issors-0.049
bar-0.048
Token his
Token ID465
Feature activation+8.90e-03

Pos logit contributions
ug+0.005
issors+0.005
ors+0.005
usted+0.005
utes+0.005
Neg logit contributions
 relief-0.004
 lady-0.004
 woman-0.004
 gentleman-0.004
 cal-0.004
Token birthday
Token ID10955
Feature activation+1.39e-01

Pos logit contributions
 relief+0.015
 gentleman+0.014
 lady+0.014
 cal+0.013
 woman+0.013
Neg logit contributions
ug-0.015
issors-0.014
utes-0.013
usted-0.013
bar-0.013
Token.
Token ID13
Feature activation+1.03e-02

Pos logit contributions
 relief+0.274
 cal+0.261
 Heidi+0.254
 gentleman+0.250
 gru+0.235
Neg logit contributions
ug-0.163
issors-0.158
bar-0.158
usted-0.157
 everywhere-0.155
Token He
Token ID679
Feature activation+7.57e-03

Pos logit contributions
 relief+0.019
 witch+0.018
 cal+0.018
 woman+0.018
 Fairy+0.018
Neg logit contributions
issors-0.015
usted-0.014
ug-0.014
bar-0.013
om-0.013
Token was
Token ID373
Feature activation+9.42e-03

Pos logit contributions
 relief+0.015
 gentleman+0.013
 cal+0.013
 service+0.013
 Heidi+0.013
Neg logit contributions
ug-0.010
usted-0.010
bed-0.009
issors-0.009
ors-0.009
Token very
Token ID845
Feature activation+1.79e-02

Pos logit contributions
 relief+0.018
 gentleman+0.017
 lady+0.017
 woman+0.017
 cal+0.017
Neg logit contributions
ug-0.013
issors-0.012
usted-0.012
bed-0.011
bar-0.011
Token excited
Token ID6568
Feature activation+6.34e-03

Pos logit contributions
 relief+0.028
 gentleman+0.026
 cal+0.026
 woman+0.025
 Heidi+0.025
Neg logit contributions
ug-0.030
issors-0.029
usted-0.028
utes-0.028
om-0.028
Token"
Token ID1
Feature activation+9.61e-03

Pos logit contributions
 relief+0.037
 lady+0.036
 gentleman+0.035
 woman+0.035
 cal+0.034
Neg logit contributions
ug-0.036
issors-0.034
usted-0.033
ors-0.033
bar-0.032
TokenHey
Token ID10814
Feature activation+8.12e-02

Pos logit contributions
 relief+0.015
 gentleman+0.014
 service+0.013
 cal+0.013
 prayer+0.013
Neg logit contributions
usted-0.016
issors-0.015
ug-0.015
bed-0.015
utes-0.015
Token,
Token ID11
Feature activation+1.83e-02

Pos logit contributions
 relief+0.146
 gentleman+0.146
 woman+0.144
 cal+0.140
 lady+0.137
Neg logit contributions
ug-0.110
bed-0.098
ors-0.095
issors-0.094
usted-0.092
Token hey
Token ID17207
Feature activation+1.32e-01

Pos logit contributions
 gentleman+0.026
 relief+0.026
 woman+0.025
 lady+0.024
 cal+0.024
Neg logit contributions
ug-0.032
issors-0.031
usted-0.030
bar-0.030
bed-0.029
Token,
Token ID11
Feature activation+2.87e-02

Pos logit contributions
 relief+0.255
 gentleman+0.252
 woman+0.242
 stranger+0.239
 cal+0.237
Neg logit contributions
ug-0.176
bed-0.161
ors-0.159
eds-0.157
usted-0.153
Token hey
Token ID17207
Feature activation+1.39e-01

Pos logit contributions
 gentleman+0.038
 woman+0.037
 relief+0.036
 lady+0.036
 cal+0.034
Neg logit contributions
ug-0.053
issors-0.051
usted-0.049
bed-0.049
bar-0.048
Token!
Token ID0
Feature activation+2.61e-02

Pos logit contributions
 gentleman+0.270
 relief+0.270
 woman+0.259
 stranger+0.254
 lady+0.253
Neg logit contributions
ug-0.181
eds-0.164
ors-0.163
bed-0.163
usted-0.155
Token What
Token ID1867
Feature activation+1.02e-01

Pos logit contributions
 relief+0.039
 lady+0.037
 gentleman+0.037
 woman+0.036
 cal+0.036
Neg logit contributions
ug-0.046
issors-0.045
usted-0.043
ors-0.042
bar-0.042
Token is
Token ID318
Feature activation+2.17e-02

Pos logit contributions
 relief+0.144
 lady+0.138
 gentleman+0.134
 woman+0.133
 cal+0.132
Neg logit contributions
ug-0.190
issors-0.183
usted-0.176
bar-0.176
ors-0.175
Token going
Token ID1016
Feature activation+2.68e-02

Pos logit contributions
 relief+0.030
 lady+0.029
 gentleman+0.028
 woman+0.028
 stranger+0.027
Neg logit contributions
ug-0.041
issors-0.039
usted-0.038
bar-0.038
ors-0.038
Token on
Token ID319
Feature activation-4.02e-02

Pos logit contributions
 relief+0.050
 lady+0.048
 gentleman+0.048
 woman+0.047
 cal+0.047
Neg logit contributions
ug-0.037
issors-0.035
usted-0.033
bar-0.033
ors-0.033
Token!
Token ID0
Feature activation-1.39e-02

Pos logit contributions
 relief+0.223
 gentleman+0.219
 cal+0.218
 Heidi+0.214
 pastor+0.202
Neg logit contributions
usted-0.143
 everywhere-0.143
bed-0.142
bar-0.142
ug-0.137
Token There
Token ID1318
Feature activation+8.53e-03

Pos logit contributions
ug+0.023
issors+0.022
usted+0.021
bed+0.021
utes+0.021
Neg logit contributions
 relief-0.022
 gentleman-0.021
 lady-0.021
 stranger-0.020
 cal-0.020
Token was
Token ID373
Feature activation-7.89e-03

Pos logit contributions
 gentleman+0.017
 relief+0.017
 stranger+0.016
 lady+0.016
 cal+0.015
Neg logit contributions
ug-0.011
bed-0.010
usted-0.010
bar-0.010
bowl-0.010
Token a
Token ID257
Feature activation+9.94e-03

Pos logit contributions
usted+0.011
bed+0.011
bar+0.011
ug+0.011
 hamm+0.010
Neg logit contributions
 gentleman-0.013
 relief-0.013
 stranger-0.012
 cal-0.012
 woman-0.012
Token big
Token ID1263
Feature activation+2.58e-02

Pos logit contributions
 relief+0.014
 gentleman+0.014
 Heidi+0.014
 cal+0.014
 service+0.013
Neg logit contributions
ug-0.015
utes-0.015
bar-0.015
issors-0.015
bed-0.015
Token birthday
Token ID10955
Feature activation+1.39e-01

Pos logit contributions
 Heidi+0.039
 gentleman+0.039
 cal+0.039
 pastor+0.038
 relief+0.037
Neg logit contributions
ug-0.036
bed-0.036
bar-0.035
 clubs-0.035
om-0.034
Token party
Token ID2151
Feature activation+7.00e-02

Pos logit contributions
 cal+0.258
 relief+0.252
 Heidi+0.244
 gentleman+0.237
 Lord+0.228
Neg logit contributions
 everywhere-0.163
usted-0.151
 built-0.151
bar-0.150
om-0.150
Token waiting
Token ID4953
Feature activation+4.70e-02

Pos logit contributions
 relief+0.151
 cal+0.145
 gentleman+0.131
 Lord+0.127
 nods+0.127
Neg logit contributions
usted-0.076
ug-0.075
bed-0.074
 everywhere-0.071
 built-0.070
Token for
Token ID329
Feature activation+2.01e-03

Pos logit contributions
 relief+0.115
 gentleman+0.112
 cal+0.109
 woman+0.106
 Heidi+0.106
Neg logit contributions
usted-0.035
ug-0.034
utes-0.032
issors-0.031
 everywhere-0.030
Token her
Token ID607
Feature activation-1.24e-02

Pos logit contributions
 gentleman+0.004
 cal+0.003
 stranger+0.003
 relief+0.003
 woman+0.003
Neg logit contributions
ug-0.003
bed-0.003
ir-0.003
issors-0.003
eds-0.003
Token with
Token ID351
Feature activation+6.59e-03

Pos logit contributions
ug+0.017
bed+0.017
utes+0.017
usted+0.017
ors+0.017
Neg logit contributions
 relief-0.022
 gentleman-0.020
 cal-0.020
 Heidi-0.020
 stranger-0.019
Token do
Token ID466
Feature activation+5.54e-03

Pos logit contributions
 relief+0.032
 lady+0.029
 man+0.026
 woman+0.025
 witch+0.024
Neg logit contributions
ug-0.080
ors-0.076
utes-0.076
usted-0.073
ants-0.073
Token?
Token ID30
Feature activation+1.96e-02

Pos logit contributions
 relief+0.010
 gentleman+0.009
 woman+0.009
 cal+0.009
 lady+0.009
Neg logit contributions
ug-0.008
issors-0.008
ors-0.008
bed-0.008
utes-0.008
Token This
Token ID770
Feature activation+4.51e-02

Pos logit contributions
 relief+0.032
 stranger+0.030
 lady+0.030
 gentleman+0.030
 cal+0.030
Neg logit contributions
issors-0.033
ug-0.033
bed-0.030
usted-0.030
bar-0.030
Token is
Token ID318
Feature activation+2.56e-02

Pos logit contributions
 gentleman+0.087
 relief+0.085
 cal+0.077
 Heidi+0.073
 gru+0.071
Neg logit contributions
issors-0.056
om-0.056
 built-0.055
usted-0.054
angles-0.054
Token a
Token ID257
Feature activation-6.81e-03

Pos logit contributions
 gentleman+0.044
 relief+0.043
 woman+0.042
 lady+0.041
 stranger+0.041
Neg logit contributions
issors-0.040
ug-0.039
bar-0.037
usted-0.036
ors-0.035
Token birthday
Token ID10955
Feature activation+1.39e-01

Pos logit contributions
issors+0.012
ug+0.012
bar+0.011
utes+0.011
eds+0.011
Neg logit contributions
 gentleman-0.009
 relief-0.009
 cal-0.008
 Heidi-0.008
 lady-0.008
Token cake
Token ID12187
Feature activation+6.96e-02

Pos logit contributions
 relief+0.227
 cal+0.224
 gentleman+0.223
 Heidi+0.221
 gru+0.200
Neg logit contributions
issors-0.187
bar-0.183
ug-0.183
usted-0.177
om-0.174
Token for
Token ID329
Feature activation+3.50e-03

Pos logit contributions
 relief+0.138
 gentleman+0.128
 cal+0.128
 woman+0.122
 Heidi+0.120
Neg logit contributions
issors-0.087
bar-0.086
ug-0.086
ors-0.082
eds-0.081
Token Mrs
Token ID9074
Feature activation+1.30e-02

Pos logit contributions
 gentleman+0.006
 relief+0.006
 woman+0.006
 cal+0.006
 stranger+0.006
Neg logit contributions
ug-0.005
issors-0.005
utes-0.005
eds-0.005
usted-0.005
Token.
Token ID13
Feature activation-2.50e-03

Pos logit contributions
 relief+0.026
 cal+0.023
 relieved+0.021
 gentleman+0.021
vation+0.021
Neg logit contributions
ug-0.019
ors-0.018
issors-0.017
ants-0.017
ert-0.016
Token Lee
Token ID5741
Feature activation-1.17e-03

Pos logit contributions
bar+0.004
issors+0.004
om+0.004
ree+0.003
ert+0.003
Neg logit contributions
 cal-0.004
 relief-0.004
 stranger-0.004
 amused-0.003
 relieved-0.003
Token twins
Token ID20345
Feature activation+2.80e-02

Pos logit contributions
 lady+0.065
 relief+0.062
 woman+0.062
 stranger+0.059
 service+0.059
Neg logit contributions
ug-0.085
ors-0.084
issors-0.083
bar-0.082
bed-0.079
Token.
Token ID13
Feature activation-1.38e-02

Pos logit contributions
 relief+0.051
 gentleman+0.048
 Heidi+0.046
 stranger+0.046
 woman+0.044
Neg logit contributions
ug-0.038
bed-0.038
ors-0.036
usted-0.036
eds-0.036
Token They
Token ID1119
Feature activation+1.89e-03

Pos logit contributions
ug+0.020
issors+0.019
usted+0.018
bed+0.018
utes+0.018
Neg logit contributions
 relief-0.025
 lady-0.024
 gentleman-0.024
 woman-0.024
 cal-0.023
Token have
Token ID423
Feature activation+2.66e-02

Pos logit contributions
 relief+0.004
 gentleman+0.004
 cal+0.003
 woman+0.003
 stranger+0.003
Neg logit contributions
issors-0.002
ug-0.002
 built-0.002
 often-0.002
bowl-0.002
Token a
Token ID257
Feature activation+1.08e-03

Pos logit contributions
 relief+0.039
 gentleman+0.036
 cal+0.035
 lady+0.034
 stranger+0.033
Neg logit contributions
ug-0.048
issors-0.046
usted-0.046
utes-0.046
bar-0.046
Token birthday
Token ID10955
Feature activation+1.38e-01

Pos logit contributions
 relief+0.002
 gentleman+0.002
 cal+0.002
 Heidi+0.001
 lady+0.001
Neg logit contributions
utes-0.002
issors-0.002
ug-0.002
bowl-0.002
om-0.002
Token today
Token ID1909
Feature activation+3.82e-02

Pos logit contributions
 relief+0.278
 cal+0.264
 gentleman+0.256
 Heidi+0.253
 stranger+0.240
Neg logit contributions
ug-0.167
utes-0.164
issors-0.163
usted-0.161
bed-0.156
Token.
Token ID13
Feature activation-6.54e-03

Pos logit contributions
 relief+0.085
 gentleman+0.080
 cal+0.079
 gru+0.076
 Heidi+0.075
Neg logit contributions
utes-0.037
om-0.037
 often-0.037
ug-0.035
 everywhere-0.035
Token Mom
Token ID11254
Feature activation+3.65e-02

Pos logit contributions
ug+0.011
issors+0.011
usted+0.010
bed+0.010
utes+0.010
Neg logit contributions
 relief-0.010
 lady-0.010
 gentleman-0.010
 woman-0.010
 cal-0.010
Token and
Token ID290
Feature activation+1.81e-02

Pos logit contributions
 relief+0.066
 gentleman+0.060
 stranger+0.056
 gru+0.055
 woman+0.055
Neg logit contributions
usted-0.054
ors-0.053
angles-0.053
om-0.053
issors-0.052
Token Dad
Token ID17415
Feature activation+3.55e-02

Pos logit contributions
 relief+0.021
 lady+0.020
 gentleman+0.020
 woman+0.019
 cal+0.019
Neg logit contributions
ug-0.038
issors-0.037
usted-0.036
ors-0.035
bar-0.035
Token day
Token ID1110
Feature activation+6.04e-02

Pos logit contributions
 relief+0.022
 gentleman+0.021
 cal+0.021
 Heidi+0.021
 MacDonald+0.020
Neg logit contributions
ug-0.009
om-0.008
utes-0.007
bowl-0.007
issors-0.007
Token,
Token ID11
Feature activation+2.05e-02

Pos logit contributions
 relief+0.129
 gentleman+0.126
 cal+0.122
 woman+0.121
 witch+0.114
Neg logit contributions
ug-0.069
bed-0.069
usted-0.063
utes-0.059
om-0.058
Token it
Token ID340
Feature activation+3.80e-02

Pos logit contributions
 gentleman+0.040
 relief+0.039
 cal+0.039
 woman+0.036
 service+0.035
Neg logit contributions
bed-0.025
usted-0.024
utes-0.023
 both-0.023
 everywhere-0.022
Token was
Token ID373
Feature activation+1.16e-03

Pos logit contributions
 relief+0.062
 gentleman+0.060
 woman+0.057
 lady+0.056
 cal+0.055
Neg logit contributions
ug-0.061
usted-0.059
bed-0.058
issors-0.058
bar-0.056
Token their
Token ID511
Feature activation+2.99e-02

Pos logit contributions
 relief+0.002
 lady+0.002
 cal+0.001
 woman+0.001
 stranger+0.001
Neg logit contributions
ug-0.002
issors-0.002
ors-0.002
bar-0.002
utes-0.002
Token birthday
Token ID10955
Feature activation+1.38e-01

Pos logit contributions
 relief+0.046
 gentleman+0.042
 cal+0.042
 lady+0.040
 woman+0.040
Neg logit contributions
ug-0.052
issors-0.051
utes-0.050
usted-0.048
bed-0.047
Token.
Token ID13
Feature activation+1.17e-02

Pos logit contributions
 relief+0.276
 cal+0.256
 gentleman+0.256
 Heidi+0.255
 gru+0.238
Neg logit contributions
usted-0.158
ug-0.156
om-0.152
 everywhere-0.152
issors-0.151
Token They
Token ID1119
Feature activation+9.88e-03

Pos logit contributions
 gentleman+0.020
 service+0.020
 cal+0.020
 woman+0.020
 relief+0.019
Neg logit contributions
issors-0.016
bed-0.015
usted-0.015
ug-0.015
bowl-0.014
Token were
Token ID547
Feature activation+1.69e-02

Pos logit contributions
 relief+0.019
 gentleman+0.018
 cal+0.018
 lady+0.018
 woman+0.017
Neg logit contributions
ug-0.013
usted-0.012
issors-0.012
bed-0.011
ors-0.011
Token very
Token ID845
Feature activation+1.86e-02

Pos logit contributions
 relief+0.032
 lady+0.031
 gentleman+0.030
 woman+0.030
 stranger+0.030
Neg logit contributions
ug-0.023
issors-0.022
bar-0.021
ors-0.021
usted-0.021
Token happy
Token ID3772
Feature activation+3.31e-02

Pos logit contributions
 relief+0.027
 gentleman+0.027
 cal+0.025
 woman+0.025
 lady+0.025
Neg logit contributions
ug-0.033
issors-0.031
usted-0.031
utes-0.030
om-0.030
Token games
Token ID1830
Feature activation+5.08e-02

Pos logit contributions
usted+0.006
ug+0.006
 clubs+0.006
bar+0.006
bowl+0.006
Neg logit contributions
 relief-0.011
 gentleman-0.010
 cal-0.010
 Heidi-0.010
 lady-0.009
Token,
Token ID11
Feature activation-2.18e-02

Pos logit contributions
 relief+0.094
 cal+0.086
 gentleman+0.085
 woman+0.083
 Lord+0.082
Neg logit contributions
ug-0.072
 clubs-0.070
usted-0.070
issors-0.069
bed-0.069
Token and
Token ID290
Feature activation+9.07e-03

Pos logit contributions
usted+0.025
ug+0.024
 clubs+0.023
bed+0.023
bar+0.022
Neg logit contributions
 relief-0.044
 gentleman-0.043
 stranger-0.041
 woman-0.040
 Heidi-0.040
Token a
Token ID257
Feature activation+1.22e-02

Pos logit contributions
 relief+0.018
 gentleman+0.018
 cal+0.017
 Heidi+0.017
 lady+0.016
Neg logit contributions
ug-0.011
usted-0.011
bar-0.010
bed-0.009
 hamm-0.009
Token giant
Token ID6175
Feature activation+4.24e-03

Pos logit contributions
 relief+0.019
 Heidi+0.018
 gentleman+0.018
 cal+0.018
 service+0.017
Neg logit contributions
bar-0.018
ug-0.018
utes-0.018
bowl-0.017
issors-0.017
Token birthday
Token ID10955
Feature activation+1.38e-01

Pos logit contributions
 relief+0.008
 cal+0.007
 Heidi+0.007
 gentleman+0.007
 Lord+0.007
Neg logit contributions
ug-0.006
bar-0.006
bowl-0.006
bed-0.006
 clubs-0.006
Token cake
Token ID12187
Feature activation+6.93e-02

Pos logit contributions
 relief+0.244
 cal+0.232
 Heidi+0.228
 gentleman+0.213
 Lord+0.204
Neg logit contributions
 clubs-0.186
usted-0.183
ug-0.183
bar-0.182
utes-0.178
Token.
Token ID13
Feature activation+1.18e-03

Pos logit contributions
 relief+0.140
 cal+0.129
 gentleman+0.119
 Heidi+0.117
 Lord+0.115
Neg logit contributions
bar-0.090
bed-0.089
usted-0.088
ug-0.087
utes-0.087
Token The
Token ID383
Feature activation+1.37e-02

Pos logit contributions
 relief+0.002
 lady+0.002
 gentleman+0.002
 woman+0.002
 cal+0.002
Neg logit contributions
ug-0.002
issors-0.002
usted-0.002
ors-0.002
bar-0.002
Token kids
Token ID3988
Feature activation+3.99e-02

Pos logit contributions
 Heidi+0.022
 relief+0.021
 cal+0.020
 gentleman+0.020
ment+0.019
Neg logit contributions
 cubes-0.020
usted-0.019
bowl-0.019
 bricks-0.018
 clubs-0.018
Token were
Token ID547
Feature activation+6.22e-03

Pos logit contributions
 relief+0.066
 lady+0.066
 gentleman+0.064
 woman+0.064
 stranger+0.063
Neg logit contributions
ug-0.062
issors-0.061
utes-0.057
ors-0.057
eds-0.056
Token share
Token ID2648
Feature activation-8.14e-03

Pos logit contributions
ug+0.002
issors+0.002
bar+0.002
bed+0.002
usted+0.002
Neg logit contributions
 relief-0.003
 gentleman-0.002
 woman-0.002
 cal-0.002
 stranger-0.002
Token her
Token ID607
Feature activation-3.04e-02

Pos logit contributions
bed+0.010
usted+0.009
eds+0.009
ors+0.009
issors+0.009
Neg logit contributions
 gentleman-0.015
 relief-0.015
 prayer-0.014
 woman-0.014
 pastor-0.014
Token own
Token ID898
Feature activation-7.08e-02

Pos logit contributions
 clubs+0.044
bed+0.044
igs+0.043
 hamm+0.042
bowl+0.041
Neg logit contributions
 Heidi-0.047
 pastor-0.045
 gentleman-0.045
 relief-0.044
 cal-0.043
Token l
Token ID300
Feature activation-4.97e-02

Pos logit contributions
 clubs+0.109
bed+0.107
issors+0.103
igs+0.102
usted+0.101
Neg logit contributions
 relief-0.110
 cal-0.104
 gentleman-0.102
 Heidi-0.101
 woman-0.101
Tokenoll
Token ID692
Feature activation-6.53e-02

Pos logit contributions
ug+0.080
issors+0.078
usted+0.075
bar+0.074
ors+0.073
Neg logit contributions
 relief-0.087
 woman-0.081
 lady-0.078
 cal-0.078
 gentleman-0.078
Tokenip
Token ID541
Feature activation-1.61e-01

Pos logit contributions
ug+0.086
issors+0.085
ors+0.085
bed+0.079
usted+0.079
Neg logit contributions
 relief-0.134
 cal-0.123
 woman-0.121
 gentleman-0.119
 lady-0.117
Tokenops
Token ID2840
Feature activation-1.11e-01

Pos logit contributions
ors+0.129
bed+0.120
issors+0.115
ug+0.111
usted+0.109
Neg logit contributions
 relief-0.428
 woman-0.397
 cal-0.390
 stranger-0.387
 gentleman-0.386
Token with
Token ID351
Feature activation-1.69e-02

Pos logit contributions
ug+0.166
usted+0.162
ors+0.162
bed+0.160
issors+0.160
Neg logit contributions
 relief-0.207
 gentleman-0.189
 woman-0.187
 cal-0.183
 stranger-0.179
Token others
Token ID1854
Feature activation+8.09e-03

Pos logit contributions
ug+0.025
issors+0.025
bed+0.025
ors+0.024
usted+0.024
Neg logit contributions
 gentleman-0.028
 relief-0.028
 stranger-0.026
 cal-0.026
 lady-0.026
Token.
Token ID13
Feature activation-8.59e-03

Pos logit contributions
 relief+0.015
 gentleman+0.014
 cal+0.013
 woman+0.013
 lady+0.013
Neg logit contributions
usted-0.011
ors-0.011
bed-0.011
ug-0.011
bowl-0.011
Token<|endoftext|>
Token ID50256
Feature activation+1.12e-02

Pos logit contributions
ug+0.014
issors+0.013
usted+0.013
ors+0.013
bar+0.013
Neg logit contributions
 relief-0.014
 gentleman-0.013
 lady-0.013
 stranger-0.013
 woman-0.013
Token and
Token ID290
Feature activation-1.64e-02

Pos logit contributions
ug+0.034
usted+0.033
bed+0.032
issors+0.031
ors+0.031
Neg logit contributions
 relief-0.052
 gentleman-0.047
 cal-0.046
 woman-0.046
 Heidi-0.046
Token share
Token ID2648
Feature activation+1.94e-04

Pos logit contributions
ug+0.023
issors+0.023
usted+0.022
bed+0.022
bar+0.022
Neg logit contributions
 relief-0.029
 gentleman-0.028
 lady-0.028
 woman-0.027
 cal-0.027
Token her
Token ID607
Feature activation-3.02e-02

Pos logit contributions
 relief+0.000
 gentleman+0.000
 service+0.000
 cal+0.000
 pastor+0.000
Neg logit contributions
bed-0.000
issors-0.000
 clubs-0.000
usted-0.000
bowl-0.000
Token l
Token ID300
Feature activation-4.01e-02

Pos logit contributions
bed+0.045
 clubs+0.045
issors+0.045
bowl+0.045
igs+0.045
Neg logit contributions
 relief-0.046
 Heidi-0.043
 gentleman-0.043
 cal-0.042
 woman-0.041
Tokenoll
Token ID692
Feature activation-5.60e-02

Pos logit contributions
ug+0.069
issors+0.067
usted+0.064
bar+0.064
ors+0.063
Neg logit contributions
 relief-0.064
 woman-0.059
 lady-0.059
 gentleman-0.059
 cal-0.058
Tokenip
Token ID541
Feature activation-1.55e-01

Pos logit contributions
ug+0.078
issors+0.075
ors+0.072
usted+0.071
bar+0.070
Neg logit contributions
 relief-0.107
 lady-0.100
 gentleman-0.100
 woman-0.099
 cal-0.099
Tokenops
Token ID2840
Feature activation-9.68e-02

Pos logit contributions
ug+0.136
issors+0.134
ors+0.133
bed+0.130
usted+0.125
Neg logit contributions
 relief-0.382
 woman-0.356
 gentleman-0.352
 cal-0.352
 stranger-0.350
Token with
Token ID351
Feature activation-1.39e-02

Pos logit contributions
ug+0.149
issors+0.144
usted+0.141
bed+0.141
ors+0.140
Neg logit contributions
 relief-0.174
 gentleman-0.161
 woman-0.158
 cal-0.156
 stranger-0.154
Token others
Token ID1854
Feature activation+5.91e-03

Pos logit contributions
ug+0.022
issors+0.022
bed+0.021
usted+0.021
ors+0.021
Neg logit contributions
 relief-0.023
 gentleman-0.022
 lady-0.021
 cal-0.021
 woman-0.021
Token,
Token ID11
Feature activation-2.71e-02

Pos logit contributions
 relief+0.012
 gentleman+0.010
 woman+0.010
 cal+0.010
 Heidi+0.010
Neg logit contributions
bed-0.008
ors-0.008
usted-0.008
ug-0.008
bowl-0.007
Token just
Token ID655
Feature activation-1.23e-02

Pos logit contributions
ug+0.041
issors+0.039
usted+0.038
bar+0.037
bed+0.037
Neg logit contributions
 relief-0.047
 gentleman-0.045
 lady-0.044
 woman-0.044
 stranger-0.044
Token number
Token ID1271
Feature activation-1.58e-02

Pos logit contributions
ug+0.008
bar+0.008
bed+0.008
issors+0.008
 drib+0.008
Neg logit contributions
 woman-0.011
 gentleman-0.011
 relief-0.011
 stranger-0.011
 lady-0.010
Token all
Token ID477
Feature activation-8.34e-02

Pos logit contributions
ug+0.024
usted+0.024
issors+0.023
bar+0.023
ors+0.023
Neg logit contributions
 relief-0.029
 gentleman-0.026
 woman-0.025
 stranger-0.025
 lady-0.025
Token her
Token ID607
Feature activation-1.69e-02

Pos logit contributions
ug+0.122
usted+0.120
issors+0.116
eds+0.116
ors+0.115
Neg logit contributions
 relief-0.146
 gentleman-0.137
 woman-0.132
 cal-0.130
 gen-0.128
Token l
Token ID300
Feature activation-3.45e-02

Pos logit contributions
issors+0.026
ug+0.026
usted+0.025
bed+0.025
ors+0.025
Neg logit contributions
 relief-0.029
 gentleman-0.027
 woman-0.027
 stranger-0.027
 lady-0.026
Tokenoll
Token ID692
Feature activation-4.98e-02

Pos logit contributions
ug+0.061
issors+0.059
usted+0.057
bar+0.056
ors+0.056
Neg logit contributions
 relief-0.053
 lady-0.050
 woman-0.050
 gentleman-0.049
 stranger-0.048
Tokenip
Token ID541
Feature activation-1.52e-01

Pos logit contributions
ug+0.025
issors+0.023
ors+0.022
usted+0.020
bed+0.019
Neg logit contributions
 relief-0.141
 woman-0.134
 gentleman-0.133
 cal-0.133
 lady-0.133
Tokenops
Token ID2840
Feature activation-9.44e-02

Pos logit contributions
ug+0.118
issors+0.116
ors+0.116
bed+0.114
usted+0.114
Neg logit contributions
 relief-0.393
 woman-0.365
 stranger-0.361
 gentleman-0.360
 cal-0.360
Token.
Token ID13
Feature activation-1.37e-04

Pos logit contributions
ug+0.154
usted+0.151
issors+0.150
bar+0.149
ors+0.148
Neg logit contributions
 relief-0.163
 gentleman-0.146
 cal-0.144
 woman-0.144
 stranger-0.139
Token She
Token ID1375
Feature activation+1.32e-02

Pos logit contributions
ug+0.000
issors+0.000
usted+0.000
bed+0.000
ors+0.000
Neg logit contributions
 relief-0.000
 lady-0.000
 gentleman-0.000
 woman-0.000
 stranger-0.000
Token had
Token ID550
Feature activation-4.13e-02

Pos logit contributions
 relief+0.020
 lady+0.019
 gentleman+0.019
 woman+0.018
 cal+0.018
Neg logit contributions
ug-0.024
issors-0.023
usted-0.022
ors-0.022
bar-0.021
Token two
Token ID734
Feature activation-3.45e-02

Pos logit contributions
ug+0.058
issors+0.057
usted+0.057
bar+0.056
bed+0.054
Neg logit contributions
 relief-0.074
 gentleman-0.071
 lady-0.070
 woman-0.069
 stranger-0.068
Token they
Token ID484
Feature activation-3.51e-02

Pos logit contributions
ug+0.027
bar+0.026
ert+0.025
bed+0.025
leys+0.025
Neg logit contributions
 relief-0.032
 gentleman-0.029
 cal-0.028
 woman-0.026
 lady-0.026
Token could
Token ID714
Feature activation-9.22e-02

Pos logit contributions
ug+0.031
issors+0.029
usted+0.028
bar+0.027
ors+0.027
Neg logit contributions
 relief-0.083
 lady-0.079
 gentleman-0.079
 woman-0.078
 cal-0.077
Token buy
Token ID2822
Feature activation+2.86e-02

Pos logit contributions
ug+0.140
issors+0.139
bar+0.132
usted+0.129
bed+0.129
Neg logit contributions
 relief-0.159
 gentleman-0.148
 woman-0.147
 lady-0.145
 cal-0.145
Token l
Token ID300
Feature activation-4.04e-02

Pos logit contributions
 relief+0.055
 gentleman+0.051
 woman+0.050
 lady+0.050
 stranger+0.050
Neg logit contributions
ug-0.038
bar-0.037
issors-0.037
ors-0.036
bed-0.035
Tokenoll
Token ID692
Feature activation-5.64e-02

Pos logit contributions
ug+0.071
issors+0.068
usted+0.066
bar+0.065
ors+0.065
Neg logit contributions
 relief-0.062
 lady-0.059
 gentleman-0.058
 woman-0.057
 cal-0.056
Tokenip
Token ID541
Feature activation-1.51e-01

Pos logit contributions
ug+0.038
issors+0.035
ors+0.034
usted+0.031
bar+0.031
Neg logit contributions
 relief-0.152
 woman-0.143
 gentleman-0.142
 lady-0.142
 cal-0.141
Tokenops
Token ID2840
Feature activation-1.04e-01

Pos logit contributions
ug+0.143
issors+0.143
ors+0.139
bed+0.134
bar+0.133
Neg logit contributions
 relief-0.364
 woman-0.337
 stranger-0.333
 gentleman-0.332
 cal-0.330
Token for
Token ID329
Feature activation-1.31e-02

Pos logit contributions
ug+0.172
issors+0.169
bar+0.169
ors+0.163
usted+0.163
Neg logit contributions
 relief-0.181
 woman-0.160
 gentleman-0.159
 cal-0.157
 stranger-0.153
Token them
Token ID606
Feature activation-5.15e-03

Pos logit contributions
ug+0.020
issors+0.019
bed+0.018
ors+0.018
bar+0.018
Neg logit contributions
 relief-0.023
 gentleman-0.022
 lady-0.022
 woman-0.022
 cal-0.022
Token.
Token ID13
Feature activation-1.39e-02

Pos logit contributions
ug+0.009
issors+0.009
bar+0.009
ors+0.009
usted+0.009
Neg logit contributions
 relief-0.008
 gentleman-0.007
 woman-0.007
 lady-0.007
 cal-0.007
Token
Token ID198
Feature activation-8.18e-03

Pos logit contributions
ug+0.023
issors+0.022
usted+0.021
bed+0.021
ors+0.021
Neg logit contributions
 relief-0.023
 lady-0.021
 gentleman-0.021
 woman-0.021
 cal-0.021
Token ball
Token ID2613
Feature activation-4.85e-02

Pos logit contributions
 relief+0.008
 gentleman+0.008
 cal+0.008
 service+0.008
 lady+0.008
Neg logit contributions
issors-0.008
ug-0.008
usted-0.008
utes-0.008
bar-0.008
Token and
Token ID290
Feature activation-1.09e-02

Pos logit contributions
ug+0.096
issors+0.093
utes+0.092
ors+0.090
ants+0.089
Neg logit contributions
 lady-0.065
 cal-0.058
 man-0.057
 stranger-0.055
 relief-0.055
Token the
Token ID262
Feature activation+1.10e-02

Pos logit contributions
bed+0.014
 clubs+0.014
ug+0.014
issors+0.014
 fet+0.014
Neg logit contributions
 gentleman-0.019
 relief-0.019
 stranger-0.018
 woman-0.018
 Lord-0.018
Token dog
Token ID3290
Feature activation-1.78e-02

Pos logit contributions
 relief+0.022
 gentleman+0.022
 cal+0.021
 service+0.021
 Heidi+0.021
Neg logit contributions
issors-0.013
ug-0.013
 clubs-0.012
bar-0.012
usted-0.012
Token fet
Token ID11351
Feature activation-9.65e-02

Pos logit contributions
ug+0.028
issors+0.027
usted+0.026
bed+0.026
ors+0.026
Neg logit contributions
 relief-0.030
 gentleman-0.029
 woman-0.028
 lady-0.028
 stranger-0.027
Tokenches
Token ID2052
Feature activation-1.50e-01

Pos logit contributions
bar+0.128
ogram+0.124
bed+0.120
bowl+0.119
years+0.115
Neg logit contributions
 relief-0.170
 lady-0.167
 man-0.164
 cal-0.163
 stranger-0.150
Token it
Token ID340
Feature activation-3.57e-02

Pos logit contributions
ug+0.314
issors+0.311
bar+0.311
utes+0.309
usted+0.305
Neg logit contributions
 lady-0.160
 man-0.150
 relief-0.148
 cal-0.142
 stranger-0.134
Token.
Token ID13
Feature activation-5.73e-03

Pos logit contributions
ug+0.083
issors+0.080
utes+0.079
bar+0.076
eds+0.076
Neg logit contributions
 lady-0.034
 relief-0.031
 cal-0.030
 man-0.029
 stranger-0.028
Token They
Token ID1119
Feature activation-3.17e-03

Pos logit contributions
ug+0.008
issors+0.008
usted+0.008
utes+0.008
bed+0.007
Neg logit contributions
 relief-0.010
 gentleman-0.010
 lady-0.010
 stranger-0.010
 woman-0.010
Token push
Token ID4574
Feature activation-1.16e-02

Pos logit contributions
ug+0.005
issors+0.005
usted+0.005
bed+0.004
bar+0.004
Neg logit contributions
 relief-0.005
 gentleman-0.005
 woman-0.005
 lady-0.005
 stranger-0.005
Token the
Token ID262
Feature activation+1.25e-03

Pos logit contributions
issors+0.022
ug+0.022
bar+0.022
utes+0.022
leys+0.022
Neg logit contributions
 relief-0.015
 lady-0.015
 woman-0.014
 cal-0.014
 man-0.014
Tokenitty
Token ID9760
Feature activation-2.73e-02

Pos logit contributions
 relief+0.003
 stranger+0.003
 cal+0.003
 lady+0.003
 woman+0.003
Neg logit contributions
ug-0.003
bar-0.003
bed-0.003
ors-0.003
issors-0.003
Token wanted
Token ID2227
Feature activation-2.97e-02

Pos logit contributions
ug+0.041
bed+0.041
bar+0.038
ors+0.038
usted+0.037
Neg logit contributions
 relief-0.050
 gentleman-0.045
 stranger-0.045
 service-0.044
 cal-0.043
Token to
Token ID284
Feature activation+1.32e-02

Pos logit contributions
ug+0.018
bar+0.017
issors+0.016
ors+0.015
bed+0.014
Neg logit contributions
 relief-0.078
 gentleman-0.075
 cal-0.072
 lady-0.071
 woman-0.071
Token introduce
Token ID10400
Feature activation-4.76e-03

Pos logit contributions
 relief+0.023
 gentleman+0.023
 woman+0.022
 cal+0.021
 lady+0.021
Neg logit contributions
ug-0.019
issors-0.018
bar-0.018
bed-0.017
bowl-0.017
Token Dog
Token ID8532
Feature activation-5.69e-02

Pos logit contributions
ug+0.008
bed+0.007
issors+0.007
ors+0.007
bar+0.007
Neg logit contributions
 relief-0.008
 gentleman-0.007
 cal-0.007
 woman-0.007
 lady-0.007
Tokengy
Token ID1360
Feature activation-1.50e-01

Pos logit contributions
ug+0.055
issors+0.049
bar+0.049
ors+0.049
bed+0.048
Neg logit contributions
 relief-0.136
 gentleman-0.125
 lady-0.124
 cal-0.122
 woman-0.122
Token to
Token ID284
Feature activation-1.68e-02

Pos logit contributions
ug+0.188
issors+0.177
usted+0.166
ors+0.164
bed+0.164
Neg logit contributions
 relief-0.296
 lady-0.293
 gentleman-0.287
 woman-0.285
 stranger-0.283
Token her
Token ID607
Feature activation+2.63e-06

Pos logit contributions
ug+0.027
issors+0.024
bed+0.024
bar+0.024
ors+0.023
Neg logit contributions
 relief-0.027
 gentleman-0.027
 lady-0.025
 woman-0.025
 stranger-0.025
Token ball
Token ID2613
Feature activation-7.68e-02

Feature not active on prev token;
no predictions were affected.
Token.
Token ID13
Feature activation-1.41e-02

Pos logit contributions
ug+0.134
issors+0.128
bar+0.124
usted+0.123
ors+0.122
Neg logit contributions
 relief-0.118
 gentleman-0.110
 lady-0.110
 woman-0.108
 cal-0.107
Token She
Token ID1375
Feature activation+1.34e-02

Pos logit contributions
ug+0.023
issors+0.022
bar+0.021
usted+0.021
ors+0.021
Neg logit contributions
 relief-0.023
 cal-0.021
 stranger-0.021
 gentleman-0.021
 lady-0.021
Token for
Token ID329
Feature activation+2.75e-03

Pos logit contributions
ug+0.013
bar+0.012
bed+0.012
usted+0.012
ors+0.012
Neg logit contributions
 relief-0.016
 gentleman-0.015
 stranger-0.015
 cal-0.014
 woman-0.014
Token lots
Token ID6041
Feature activation-5.03e-02

Pos logit contributions
 gentleman+0.005
 stranger+0.005
 lady+0.004
 woman+0.004
 relief+0.004
Neg logit contributions
ug-0.004
bed-0.004
 hamm-0.003
issors-0.003
igs-0.003
Token more
Token ID517
Feature activation-2.38e-03

Pos logit contributions
ug+0.041
ors+0.036
usted+0.035
 often+0.026
ir+0.024
Neg logit contributions
 relief-0.128
 gentleman-0.122
 woman-0.119
 Heidi-0.118
 gen-0.116
Token l
Token ID300
Feature activation-3.67e-02

Pos logit contributions
ug+0.003
 often+0.003
ors+0.003
usted+0.003
igs+0.003
Neg logit contributions
 gentleman-0.004
 relief-0.004
 Heidi-0.004
 lady-0.004
 woman-0.004
Tokenoll
Token ID692
Feature activation-4.74e-02

Pos logit contributions
ug+0.061
issors+0.058
usted+0.056
bar+0.055
ors+0.055
Neg logit contributions
 relief-0.062
 woman-0.059
 lady-0.058
 gentleman-0.058
 stranger-0.057
Tokenip
Token ID541
Feature activation-1.49e-01

Pos logit contributions
ug+0.050
ors+0.048
issors+0.046
igs+0.044
ants+0.043
Neg logit contributions
 relief-0.113
 cal-0.104
 gentleman-0.103
 woman-0.103
 lady-0.101
Tokenops
Token ID2840
Feature activation-1.01e-01

Pos logit contributions
ors+0.105
ug+0.105
issors+0.095
bed+0.092
usted+0.091
Neg logit contributions
 relief-0.403
 woman-0.373
 cal-0.371
 gentleman-0.370
 stranger-0.368
Token like
Token ID588
Feature activation-5.69e-03

Pos logit contributions
ug+0.162
bar+0.153
ors+0.153
usted+0.150
issors+0.149
Neg logit contributions
 relief-0.184
 gentleman-0.166
 cal-0.163
 woman-0.160
 lady-0.156
Token the
Token ID262
Feature activation-8.13e-03

Pos logit contributions
ug+0.009
bar+0.009
usted+0.008
bed+0.008
issors+0.008
Neg logit contributions
 relief-0.009
 gentleman-0.009
 cal-0.008
 woman-0.008
 stranger-0.008
Token one
Token ID530
Feature activation-8.79e-03

Pos logit contributions
bar+0.012
usted+0.011
 drib+0.011
issors+0.011
 targets+0.011
Neg logit contributions
 relief-0.012
 Heidi-0.011
 gentleman-0.011
 cal-0.010
 service-0.010
Token she
Token ID673
Feature activation-3.44e-02

Pos logit contributions
ug+0.011
bar+0.011
 often+0.010
bowl+0.010
ogram+0.010
Neg logit contributions
 relief-0.017
 gentleman-0.016
 Heidi-0.015
 woman-0.014
 cal-0.014
Token get
Token ID651
Feature activation+1.64e-02

Pos logit contributions
ug+0.168
issors+0.158
bed+0.156
bar+0.155
usted+0.152
Neg logit contributions
 relief-0.185
 gentleman-0.179
 lady-0.176
 woman-0.175
 stranger-0.172
Token her
Token ID607
Feature activation-5.12e-03

Pos logit contributions
 gentleman+0.033
 relief+0.032
 woman+0.031
 lady+0.031
 witch+0.030
Neg logit contributions
ug-0.020
usted-0.019
bed-0.019
bar-0.019
bowl-0.018
Token own
Token ID898
Feature activation-4.47e-02

Pos logit contributions
ug+0.007
issors+0.007
bed+0.007
utes+0.007
bar+0.007
Neg logit contributions
 gentleman-0.009
 relief-0.009
 cal-0.008
 woman-0.008
 lady-0.008
Token l
Token ID300
Feature activation-3.20e-02

Pos logit contributions
ug+0.064
bar+0.062
 clubs+0.062
bed+0.062
usted+0.062
Neg logit contributions
 relief-0.073
 gentleman-0.072
 cal-0.067
 woman-0.067
 Heidi-0.067
Tokenoll
Token ID692
Feature activation-4.41e-02

Pos logit contributions
ug+0.056
issors+0.053
usted+0.051
bar+0.051
utes+0.050
Neg logit contributions
 relief-0.051
 lady-0.048
 woman-0.047
 gentleman-0.047
 cal-0.046
Tokenip
Token ID541
Feature activation-1.49e-01

Pos logit contributions
ug+0.061
issors+0.058
ors+0.058
bar+0.056
usted+0.055
Neg logit contributions
 relief-0.087
 cal-0.081
 gentleman-0.080
 woman-0.079
 lady-0.079
Tokenops
Token ID2840
Feature activation-9.90e-02

Pos logit contributions
ug+0.157
ors+0.152
bar+0.149
issors+0.148
bed+0.146
Neg logit contributions
 relief-0.341
 woman-0.313
 gentleman-0.312
 cal-0.311
 stranger-0.307
Token.
Token ID13
Feature activation-9.96e-03

Pos logit contributions
ug+0.163
bar+0.155
issors+0.154
usted+0.153
ors+0.152
Neg logit contributions
 relief-0.168
 gentleman-0.155
 woman-0.151
 cal-0.150
 lady-0.150
Token Betty
Token ID29504
Feature activation-2.84e-02

Pos logit contributions
ug+0.017
issors+0.016
usted+0.016
ors+0.015
bar+0.015
Neg logit contributions
 relief-0.016
 lady-0.015
 gentleman-0.015
 woman-0.014
 cal-0.014
Token be
Token ID307
Feature activation+6.03e-02

Pos logit contributions
ug+0.047
bed+0.045
usted+0.045
bar+0.043
ors+0.043
Neg logit contributions
 relief-0.046
 gentleman-0.043
 stranger-0.042
 lady-0.042
 woman-0.042
Tokenamed
Token ID2434
Feature activation-5.81e-03

Pos logit contributions
 relief+0.120
 gentleman+0.118
 lady+0.117
 stranger+0.115
 woman+0.115
Neg logit contributions
ug-0.076
issors-0.069
usted-0.069
bar-0.068
bed-0.067
Token ball
Token ID2613
Feature activation-4.79e-02

Pos logit contributions
 relief+0.013
 cal+0.012
 gentleman+0.012
 Lord+0.012
 service+0.012
Neg logit contributions
issors-0.012
ug-0.012
 clubs-0.011
usted-0.011
eds-0.011
Token and
Token ID290
Feature activation-7.40e-03

Pos logit contributions
ug+0.094
issors+0.094
utes+0.091
ors+0.090
ants+0.088
Neg logit contributions
 lady-0.062
 cal-0.056
 man-0.055
 stranger-0.053
 relief-0.053
Token the
Token ID262
Feature activation+1.36e-02

Pos logit contributions
ug+0.009
issors+0.009
bed+0.009
 clubs+0.009
 fet+0.009
Neg logit contributions
 gentleman-0.013
 relief-0.013
 stranger-0.013
 Lord-0.013
 woman-0.013
Token dog
Token ID3290
Feature activation-1.71e-02

Pos logit contributions
 relief+0.028
 gentleman+0.027
 cal+0.027
 Lord+0.026
 Heidi+0.026
Neg logit contributions
issors-0.015
ug-0.015
 clubs-0.015
bar-0.014
bowl-0.014
Token fet
Token ID11351
Feature activation-9.51e-02

Pos logit contributions
ug+0.029
issors+0.028
usted+0.027
bed+0.027
ors+0.027
Neg logit contributions
 relief-0.027
 gentleman-0.026
 lady-0.025
 woman-0.025
 stranger-0.025
Tokenches
Token ID2052
Feature activation-1.48e-01

Pos logit contributions
bar+0.124
ogram+0.120
bed+0.117
bowl+0.116
ug+0.112
Neg logit contributions
 relief-0.172
 lady-0.166
 cal-0.166
 man-0.163
 sigh-0.148
Token it
Token ID340
Feature activation-3.21e-02

Pos logit contributions
issors+0.305
bar+0.305
ug+0.304
utes+0.299
usted+0.297
Neg logit contributions
 lady-0.161
 relief-0.154
 man-0.152
 cal-0.148
 stranger-0.138
Token.
Token ID13
Feature activation-2.40e-03

Pos logit contributions
ug+0.075
issors+0.072
utes+0.071
usted+0.069
bar+0.069
Neg logit contributions
 lady-0.031
 relief-0.029
 cal-0.027
 man-0.026
 stranger-0.026
Token They
Token ID1119
Feature activation-2.47e-03

Pos logit contributions
ug+0.003
issors+0.003
bed+0.003
utes+0.003
usted+0.003
Neg logit contributions
 relief-0.005
 gentleman-0.005
 stranger-0.005
 lady-0.005
 woman-0.005
Token have
Token ID423
Feature activation+1.58e-02

Pos logit contributions
ug+0.004
issors+0.003
bed+0.003
bowl+0.003
bar+0.003
Neg logit contributions
 relief-0.004
 gentleman-0.004
 woman-0.004
 lady-0.004
 stranger-0.004
Token fun
Token ID1257
Feature activation+3.67e-02

Pos logit contributions
 relief+0.021
 gentleman+0.020
 lady+0.020
 woman+0.019
 stranger+0.019
Neg logit contributions
ug-0.030
issors-0.029
usted-0.029
bar-0.028
bed-0.028
Token ball
Token ID2613
Feature activation-3.70e-02

Pos logit contributions
ug+0.018
ors+0.018
issors+0.017
bed+0.017
leys+0.016
Neg logit contributions
 lady-0.022
 man-0.021
 stranger-0.021
 woman-0.021
 relief-0.021
Token and
Token ID290
Feature activation+4.39e-03

Pos logit contributions
leys+0.086
ants+0.086
utes+0.086
issors+0.085
ug+0.084
Neg logit contributions
 lady-0.037
 man-0.031
 cal-0.028
 ranger-0.027
 stranger-0.027
Token the
Token ID262
Feature activation+1.82e-02

Pos logit contributions
 gentleman+0.008
 relief+0.008
 stranger+0.008
 Lord+0.008
 woman+0.008
Neg logit contributions
ug-0.005
issors-0.005
 clubs-0.005
bar-0.005
bed-0.005
Token dog
Token ID3290
Feature activation-1.35e-02

Pos logit contributions
 relief+0.035
 gentleman+0.033
 cal+0.033
 service+0.032
 Lord+0.032
Neg logit contributions
ug-0.023
issors-0.023
bar-0.022
 clubs-0.022
usted-0.021
Token fet
Token ID11351
Feature activation-9.31e-02

Pos logit contributions
ug+0.024
issors+0.023
usted+0.022
ors+0.022
bed+0.022
Neg logit contributions
 relief-0.021
 gentleman-0.020
 woman-0.019
 lady-0.019
 stranger-0.019
Tokenches
Token ID2052
Feature activation-1.48e-01

Pos logit contributions
bar+0.119
ogram+0.116
bed+0.112
bowl+0.111
ug+0.109
Neg logit contributions
 relief-0.169
 cal-0.165
 lady-0.165
 man-0.161
 stranger-0.148
Token it
Token ID340
Feature activation-3.19e-02

Pos logit contributions
ug+0.316
utes+0.313
issors+0.312
bar+0.312
leys+0.311
Neg logit contributions
 lady-0.151
 man-0.144
 relief-0.136
 cal-0.135
 stranger-0.124
Token.
Token ID13
Feature activation+2.68e-04

Pos logit contributions
ug+0.078
issors+0.075
utes+0.075
eds+0.073
leys+0.072
Neg logit contributions
 lady-0.028
 cal-0.023
 relief-0.023
 man-0.023
 stranger-0.021
Token They
Token ID1119
Feature activation+7.15e-04

Pos logit contributions
 relief+0.001
 gentleman+0.000
 lady+0.000
 woman+0.000
 stranger+0.000
Neg logit contributions
ug-0.000
issors-0.000
usted-0.000
bed-0.000
utes-0.000
Token hug
Token ID16225
Feature activation+2.91e-02

Pos logit contributions
 relief+0.001
 gentleman+0.001
 lady+0.001
 woman+0.001
 stranger+0.001
Neg logit contributions
ug-0.001
issors-0.001
usted-0.001
bar-0.001
bed-0.001
Token the
Token ID262
Feature activation+8.09e-06

Pos logit contributions
 relief+0.052
 gentleman+0.050
 lady+0.049
 cal+0.048
 woman+0.048
Neg logit contributions
ug-0.043
issors-0.041
bed-0.040
ors-0.039
usted-0.039
Token him
Token ID683
Feature activation-9.01e-03

Pos logit contributions
ug+0.021
bed+0.019
ir+0.018
 often+0.018
usted+0.018
Neg logit contributions
 gentleman-0.029
 woman-0.029
 relief-0.027
 lady-0.027
 stranger-0.027
Token for
Token ID329
Feature activation+2.63e-03

Pos logit contributions
ug+0.012
ors+0.010
usted+0.010
bed+0.010
bar+0.010
Neg logit contributions
 relief-0.017
 gentleman-0.016
 woman-0.016
 Heidi-0.016
 pastor-0.016
Token remembering
Token ID24865
Feature activation-1.37e-02

Pos logit contributions
 gentleman+0.005
 woman+0.005
 lady+0.004
 relief+0.004
 stranger+0.004
Neg logit contributions
ug-0.004
bed-0.004
bar-0.003
issors-0.003
usted-0.003
Token his
Token ID465
Feature activation-2.45e-02

Pos logit contributions
ug+0.016
usted+0.015
ors+0.015
bed+0.015
eds+0.015
Neg logit contributions
 relief-0.028
 gentleman-0.027
 woman-0.026
 lady-0.026
 stranger-0.025
Token pig
Token ID12967
Feature activation-9.49e-02

Pos logit contributions
ug+0.040
issors+0.039
utes+0.038
bar+0.038
 clubs+0.037
Neg logit contributions
 gentleman-0.034
 relief-0.034
 Heidi-0.033
 cal-0.033
 lady-0.032
Tokengy
Token ID1360
Feature activation-1.47e-01

Pos logit contributions
ug+0.136
bowl+0.132
bar+0.130
ors+0.129
eds+0.129
Neg logit contributions
 relief-0.172
 gentleman-0.159
 cal-0.154
 stranger-0.151
 woman-0.150
Token bank
Token ID3331
Feature activation-5.98e-02

Pos logit contributions
ug+0.233
issors+0.222
bar+0.218
usted+0.216
bed+0.216
Neg logit contributions
 relief-0.244
 gentleman-0.231
 lady-0.227
 woman-0.223
 cal-0.222
Token.
Token ID13
Feature activation-4.31e-03

Pos logit contributions
ug+0.093
usted+0.089
bar+0.089
issors+0.087
ors+0.084
Neg logit contributions
 relief-0.106
 gentleman-0.097
 woman-0.096
 Heidi-0.094
 witch-0.093
Token From
Token ID3574
Feature activation+1.35e-02

Pos logit contributions
ug+0.007
issors+0.007
bar+0.006
usted+0.006
ors+0.006
Neg logit contributions
 relief-0.007
 lady-0.007
 gentleman-0.007
 woman-0.007
 cal-0.007
Token that
Token ID326
Feature activation+7.31e-03

Pos logit contributions
 relief+0.025
 gentleman+0.023
 woman+0.023
 cal+0.023
 prayer+0.023
Neg logit contributions
ug-0.018
bed-0.018
bar-0.017
ors-0.016
issors-0.016
Token day
Token ID1110
Feature activation+2.31e-02

Pos logit contributions
 relief+0.010
 lady+0.010
 woman+0.010
 gentleman+0.010
 stranger+0.010
Neg logit contributions
ug-0.013
issors-0.013
usted-0.013
utes-0.013
ors-0.013
Token count
Token ID954
Feature activation+1.87e-02

Pos logit contributions
 relief+0.017
 gentleman+0.017
 woman+0.017
 cal+0.016
 lady+0.016
Neg logit contributions
ug-0.015
issors-0.014
bar-0.014
bed-0.013
usted-0.013
Token how
Token ID703
Feature activation-9.29e-03

Pos logit contributions
 relief+0.033
 gentleman+0.031
 lady+0.030
 woman+0.030
 cal+0.029
Neg logit contributions
ug-0.029
issors-0.028
usted-0.027
bar-0.027
ors-0.026
Token many
Token ID867
Feature activation-2.58e-02

Pos logit contributions
ug+0.012
issors+0.012
usted+0.012
utes+0.011
bar+0.011
Neg logit contributions
 relief-0.018
 gentleman-0.017
 lady-0.017
 woman-0.017
 cal-0.016
Token l
Token ID300
Feature activation-3.06e-02

Pos logit contributions
ug+0.035
igs+0.033
usted+0.033
 often+0.032
 clubs+0.031
Neg logit contributions
 relief-0.049
 gentleman-0.048
 cal-0.042
 woman-0.042
 Heidi-0.041
Tokenoll
Token ID692
Feature activation-4.18e-02

Pos logit contributions
ug+0.040
issors+0.037
usted+0.035
utes+0.034
bar+0.034
Neg logit contributions
 relief-0.064
 woman-0.060
 gentleman-0.060
 lady-0.059
 cal-0.059
Tokenip
Token ID541
Feature activation-1.46e-01

Pos logit contributions
ug+0.049
ors+0.046
issors+0.045
bed+0.043
usted+0.043
Neg logit contributions
 relief-0.093
 cal-0.087
 gentleman-0.086
 woman-0.085
 lady-0.085
Tokenops
Token ID2840
Feature activation-9.84e-02

Pos logit contributions
ug+0.126
ors+0.121
bed+0.120
issors+0.116
usted+0.115
Neg logit contributions
 relief-0.371
 gentleman-0.341
 woman-0.339
 cal-0.339
 stranger-0.338
Token she
Token ID673
Feature activation-4.16e-02

Pos logit contributions
ug+0.158
bar+0.154
usted+0.151
bed+0.148
ors+0.148
Neg logit contributions
 relief-0.178
 gentleman-0.158
 cal-0.154
 woman-0.153
 stranger-0.148
Token had
Token ID550
Feature activation-8.20e-02

Pos logit contributions
ug+0.041
bed+0.039
issors+0.039
usted+0.038
bar+0.037
Neg logit contributions
 relief-0.095
 gentleman-0.088
 lady-0.086
 woman-0.085
 stranger-0.085
Token.
Token ID13
Feature activation-2.46e-03

Pos logit contributions
usted+0.114
ug+0.113
issors+0.111
bar+0.110
leys+0.110
Neg logit contributions
 relief-0.154
 gentleman-0.142
 woman-0.136
 stranger-0.134
 cal-0.134
Token She
Token ID1375
Feature activation+1.54e-02

Pos logit contributions
issors+0.004
ug+0.004
usted+0.004
bar+0.004
ors+0.004
Neg logit contributions
 relief-0.004
 cal-0.004
 stranger-0.004
 lady-0.004
 gentleman-0.004
Token walk
Token ID2513
Feature activation+1.21e-02

Pos logit contributions
 relief+0.012
 woman+0.011
 gentleman+0.011
 stranger+0.011
 lady+0.011
Neg logit contributions
ug-0.011
issors-0.011
bar-0.010
bed-0.010
ors-0.010
Token and
Token ID290
Feature activation-1.29e-02

Pos logit contributions
 relief+0.022
 gentleman+0.021
 lady+0.021
 woman+0.020
 cal+0.020
Neg logit contributions
ug-0.018
issors-0.017
ors-0.016
bed-0.016
usted-0.016
Token give
Token ID1577
Feature activation+1.77e-02

Pos logit contributions
ug+0.019
issors+0.018
bar+0.018
bed+0.018
usted+0.018
Neg logit contributions
 relief-0.022
 gentleman-0.021
 lady-0.020
 woman-0.020
 cal-0.020
Token l
Token ID300
Feature activation-2.91e-02

Pos logit contributions
 relief+0.030
 gentleman+0.029
 lady+0.028
 woman+0.028
 stranger+0.027
Neg logit contributions
ug-0.027
issors-0.027
usted-0.026
ors-0.026
bed-0.026
Tokenoll
Token ID692
Feature activation-4.78e-02

Pos logit contributions
ug+0.053
issors+0.051
usted+0.049
ors+0.048
bar+0.048
Neg logit contributions
 relief-0.043
 lady-0.040
 gentleman-0.040
 woman-0.039
 cal-0.039
Tokenip
Token ID541
Feature activation-1.46e-01

Pos logit contributions
ug+0.042
issors+0.039
usted+0.036
ors+0.035
bar+0.035
Neg logit contributions
 relief-0.115
 lady-0.111
 gentleman-0.110
 woman-0.109
 cal-0.108
Tokenops
Token ID2840
Feature activation-8.86e-02

Pos logit contributions
ug+0.136
issors+0.129
ors+0.124
usted+0.122
bar+0.119
Neg logit contributions
 relief-0.349
 woman-0.327
 gentleman-0.327
 lady-0.324
 stranger-0.322
Token to
Token ID284
Feature activation-1.32e-02

Pos logit contributions
ug+0.125
issors+0.120
usted+0.118
bar+0.118
ors+0.115
Neg logit contributions
 relief-0.170
 gentleman-0.157
 woman-0.155
 cal-0.154
 lady-0.151
Token children
Token ID1751
Feature activation-2.55e-02

Pos logit contributions
ug+0.019
issors+0.019
bed+0.018
bar+0.018
ors+0.018
Neg logit contributions
 relief-0.023
 gentleman-0.023
 woman-0.022
 stranger-0.022
 cal-0.021
Token."
Token ID526
Feature activation-1.24e-02

Pos logit contributions
ug+0.035
issors+0.032
ors+0.032
bar+0.032
usted+0.032
Neg logit contributions
 relief-0.050
 gentleman-0.046
 cal-0.045
 woman-0.045
 stranger-0.044
Token The
Token ID383
Feature activation+7.22e-03

Pos logit contributions
ug+0.018
usted+0.018
issors+0.018
bed+0.017
ors+0.016
Neg logit contributions
 relief-0.022
 stranger-0.021
 gentleman-0.021
 cal-0.021
 woman-0.020
Token some
Token ID617
Feature activation+3.74e-03

Pos logit contributions
 relief+0.062
 gentleman+0.061
 woman+0.058
 Heidi+0.057
 lady+0.056
Neg logit contributions
usted-0.045
ug-0.043
issors-0.043
bar-0.040
utes-0.039
Token money
Token ID1637
Feature activation+2.16e-02

Pos logit contributions
 relief+0.006
 gentleman+0.006
 Heidi+0.006
 woman+0.005
 cal+0.005
Neg logit contributions
ug-0.006
issors-0.006
usted-0.006
 clubs-0.006
utes-0.006
Token for
Token ID329
Feature activation-5.77e-03

Pos logit contributions
 relief+0.042
 gentleman+0.038
 cal+0.037
 Heidi+0.036
 woman+0.036
Neg logit contributions
issors-0.031
ug-0.030
usted-0.029
ors-0.029
bar-0.029
Token her
Token ID607
Feature activation-1.46e-02

Pos logit contributions
ug+0.009
issors+0.009
bed+0.009
ors+0.009
usted+0.009
Neg logit contributions
 relief-0.009
 gentleman-0.009
 cal-0.009
 lady-0.009
 woman-0.009
Token pig
Token ID12967
Feature activation-9.28e-02

Pos logit contributions
ug+0.021
issors+0.020
 clubs+0.020
ors+0.020
utes+0.019
Neg logit contributions
 relief-0.026
 gentleman-0.024
 cal-0.024
 Heidi-0.023
 woman-0.022
Tokengy
Token ID1360
Feature activation-1.45e-01

Pos logit contributions
ug+0.167
issors+0.162
ors+0.156
usted+0.155
bar+0.155
Neg logit contributions
 relief-0.138
 gentleman-0.128
 lady-0.127
 woman-0.124
 cal-0.124
Token bank
Token ID3331
Feature activation-5.19e-02

Pos logit contributions
ug+0.234
issors+0.225
usted+0.216
bar+0.214
ors+0.214
Neg logit contributions
 relief-0.239
 lady-0.227
 gentleman-0.226
 woman-0.222
 cal-0.220
Token.
Token ID13
Feature activation-7.50e-03

Pos logit contributions
ug+0.093
issors+0.090
usted+0.088
bar+0.086
ors+0.086
Neg logit contributions
 relief-0.082
 gentleman-0.073
 woman-0.072
 cal-0.070
 Heidi-0.069
Token
Token ID198
Feature activation-3.61e-03

Pos logit contributions
ug+0.015
issors+0.014
eds+0.014
utes+0.014
ors+0.014
Neg logit contributions
 relief-0.010
 lady-0.010
 gentleman-0.009
 woman-0.009
 cal-0.009
Token
Token ID198
Feature activation+1.22e-02

Pos logit contributions
ug+0.009
issors+0.009
ors+0.009
bar+0.009
usted+0.009
Neg logit contributions
 lady-0.003
 woman-0.003
 man-0.002
 stranger-0.002
 gentleman-0.002
TokenOne
Token ID3198
Feature activation+4.53e-03

Pos logit contributions
 lady+0.020
 woman+0.018
 stranger+0.018
 relief+0.018
 gentleman+0.018
Neg logit contributions
ug-0.021
ors-0.020
eds-0.019
issors-0.019
bed-0.018
Tokenily
Token ID813
Feature activation+6.26e-03

Pos logit contributions
 relief+0.024
 lady+0.023
 gentleman+0.022
 woman+0.022
 cal+0.022
Neg logit contributions
ug-0.039
issors-0.038
bar-0.037
usted-0.036
bed-0.036
Token learned
Token ID4499
Feature activation-2.58e-02

Pos logit contributions
 lady+0.008
 relief+0.008
 cal+0.008
 woman+0.008
 man+0.008
Neg logit contributions
ug-0.013
issors-0.012
utes-0.012
ors-0.011
bar-0.011
Token that
Token ID326
Feature activation-3.59e-02

Pos logit contributions
ug+0.057
issors+0.054
bar+0.053
bowl+0.053
ors+0.052
Neg logit contributions
 relief-0.029
 lady-0.029
 woman-0.027
 stranger-0.027
 cal-0.027
Token l
Token ID300
Feature activation-2.22e-02

Pos logit contributions
ug+0.066
issors+0.065
leys+0.063
utes+0.060
ors+0.060
Neg logit contributions
 lady-0.054
 relief-0.051
 woman-0.051
 cal-0.049
 man-0.049
Tokenoll
Token ID692
Feature activation-4.72e-02

Pos logit contributions
ug+0.034
utes+0.033
bed+0.033
issors+0.033
leys+0.033
Neg logit contributions
 relief-0.035
 lady-0.033
 gentleman-0.033
 cal-0.032
 stranger-0.031
Tokenip
Token ID541
Feature activation-1.45e-01

Pos logit contributions
ug+0.036
issors+0.033
usted+0.030
ors+0.030
bar+0.029
Neg logit contributions
 relief-0.119
 lady-0.115
 gentleman-0.114
 woman-0.113
 cal-0.112
Tokenops
Token ID2840
Feature activation-9.25e-02

Pos logit contributions
ug+0.170
issors+0.161
usted+0.151
ors+0.151
bed+0.149
Neg logit contributions
 relief-0.306
 lady-0.294
 gentleman-0.293
 woman-0.290
 stranger-0.287
Token are
Token ID389
Feature activation+2.43e-03

Pos logit contributions
ug+0.139
issors+0.133
usted+0.128
ors+0.127
bar+0.126
Neg logit contributions
 relief-0.164
 gentleman-0.158
 lady-0.156
 woman-0.154
 stranger-0.151
Token fragile
Token ID21049
Feature activation-5.24e-02

Pos logit contributions
 relief+0.004
 lady+0.003
 cal+0.003
 woman+0.003
 gentleman+0.003
Neg logit contributions
ug-0.004
issors-0.004
ors-0.004
utes-0.004
bed-0.004
Token and
Token ID290
Feature activation+5.95e-03

Pos logit contributions
ug+0.085
issors+0.081
usted+0.079
ors+0.079
bed+0.078
Neg logit contributions
 relief-0.088
 gentleman-0.084
 lady-0.083
 woman-0.082
 stranger-0.080
Token can
Token ID460
Feature activation-3.31e-02

Pos logit contributions
 relief+0.009
 lady+0.009
 woman+0.009
 gentleman+0.009
 cal+0.009
Neg logit contributions
ug-0.010
issors-0.010
usted-0.009
utes-0.009
bar-0.009
Token care
Token ID1337
Feature activation+9.78e-04

Pos logit contributions
 relief+0.001
 gentleman+0.001
 lady+0.001
 woman+0.001
 stranger+0.001
Neg logit contributions
ug-0.001
issors-0.001
usted-0.001
bed-0.001
ors-0.001
Token of
Token ID286
Feature activation-4.19e-02

Pos logit contributions
 relief+0.002
 lady+0.002
 woman+0.002
 gentleman+0.002
 cal+0.001
Neg logit contributions
ug-0.002
issors-0.002
utes-0.001
ors-0.001
bar-0.001
Token her
Token ID607
Feature activation-1.58e-02

Pos logit contributions
ug+0.073
issors+0.071
bed+0.069
ors+0.068
usted+0.067
Neg logit contributions
 relief-0.064
 gentleman-0.062
 stranger-0.060
 lady-0.059
 woman-0.058
Token l
Token ID300
Feature activation-3.27e-02

Pos logit contributions
ug+0.025
issors+0.025
bed+0.024
usted+0.024
ors+0.024
Neg logit contributions
 relief-0.026
 gentleman-0.025
 stranger-0.024
 Heidi-0.024
 woman-0.023
Tokenoll
Token ID692
Feature activation-4.94e-02

Pos logit contributions
ug+0.062
issors+0.059
bed+0.057
utes+0.057
usted+0.057
Neg logit contributions
 relief-0.044
 lady-0.043
 gentleman-0.042
 woman-0.041
 cal-0.040
Tokenip
Token ID541
Feature activation-1.44e-01

Pos logit contributions
ug+0.030
issors+0.027
ors+0.024
usted+0.024
bar+0.023
Neg logit contributions
 relief-0.133
 lady-0.128
 gentleman-0.128
 woman-0.127
 cal-0.126
Tokenops
Token ID2840
Feature activation-9.44e-02

Pos logit contributions
ug+0.145
issors+0.139
ors+0.132
usted+0.131
bed+0.130
Neg logit contributions
 relief-0.333
 gentleman-0.317
 lady-0.314
 woman-0.313
 stranger-0.312
Token and
Token ID290
Feature activation-1.52e-02

Pos logit contributions
ug+0.155
issors+0.148
usted+0.145
bed+0.144
ors+0.144
Neg logit contributions
 relief-0.157
 gentleman-0.148
 lady-0.145
 woman-0.144
 cal-0.142
Token other
Token ID584
Feature activation-2.28e-02

Pos logit contributions
ug+0.023
issors+0.022
bed+0.021
usted+0.021
bar+0.021
Neg logit contributions
 relief-0.026
 gentleman-0.025
 lady-0.025
 woman-0.024
 stranger-0.024
Token fragile
Token ID21049
Feature activation-3.80e-02

Pos logit contributions
ug+0.039
usted+0.038
issors+0.037
ors+0.037
bar+0.036
Neg logit contributions
 relief-0.035
 gentleman-0.033
 lady-0.032
 stranger-0.031
 woman-0.031
Token things
Token ID1243
Feature activation+4.47e-02

Pos logit contributions
ug+0.067
usted+0.064
ors+0.064
bed+0.064
issors+0.063
Neg logit contributions
 relief-0.058
 gentleman-0.054
 woman-0.053
 stranger-0.053
 lady-0.052
Token money
Token ID1637
Feature activation+6.20e-03

Pos logit contributions
ug+0.065
issors+0.063
usted+0.061
bar+0.060
bed+0.059
Neg logit contributions
 relief-0.084
 gentleman-0.080
 lady-0.078
 woman-0.078
 cal-0.077
Token away
Token ID1497
Feature activation-3.59e-02

Pos logit contributions
 relief+0.012
 gentleman+0.011
 cal+0.011
 woman+0.011
 lady+0.011
Neg logit contributions
ug-0.008
issors-0.008
usted-0.008
bar-0.008
ors-0.008
Token in
Token ID287
Feature activation-5.44e-02

Pos logit contributions
ug+0.053
bed+0.052
ors+0.050
usted+0.050
issors+0.050
Neg logit contributions
 relief-0.064
 gentleman-0.061
 cal-0.058
 woman-0.058
 stranger-0.057
Token her
Token ID607
Feature activation+5.76e-03

Pos logit contributions
ug+0.091
issors+0.089
bed+0.084
bar+0.083
usted+0.082
Neg logit contributions
 gentleman-0.082
 relief-0.082
 cal-0.077
 lady-0.076
 stranger-0.076
Token pig
Token ID12967
Feature activation-9.31e-02

Pos logit contributions
 Heidi+0.011
 relief+0.011
 gentleman+0.010
 cal+0.010
 stranger+0.010
Neg logit contributions
issors-0.007
 clubs-0.007
ug-0.007
utes-0.007
bed-0.007
Tokengy
Token ID1360
Feature activation-1.44e-01

Pos logit contributions
ug+0.132
issors+0.126
bar+0.120
ors+0.120
usted+0.120
Neg logit contributions
 relief-0.174
 gentleman-0.164
 lady-0.164
 cal-0.161
 woman-0.161
Tokenbank
Token ID17796
Feature activation-4.44e-02

Pos logit contributions
ug+0.182
issors+0.173
bar+0.167
bed+0.165
usted+0.165
Neg logit contributions
 relief-0.289
 gentleman-0.273
 lady-0.272
 cal-0.269
 woman-0.267
Token every
Token ID790
Feature activation-5.49e-02

Pos logit contributions
ug+0.077
issors+0.074
usted+0.074
bed+0.073
bar+0.073
Neg logit contributions
 relief-0.071
 gentleman-0.064
 cal-0.063
 woman-0.063
 stranger-0.062
Token week
Token ID1285
Feature activation-1.67e-02

Pos logit contributions
ug+0.072
bar+0.063
issors+0.062
bowl+0.061
bed+0.061
Neg logit contributions
 relief-0.104
 gentleman-0.101
 cal-0.099
 lady-0.096
 woman-0.096
Token.
Token ID13
Feature activation-1.20e-02

Pos logit contributions
ug+0.024
usted+0.023
ors+0.022
bed+0.022
issors+0.022
Neg logit contributions
 relief-0.031
 gentleman-0.028
 cal-0.028
 stranger-0.026
 gen-0.026
Token This
Token ID770
Feature activation+4.23e-02

Pos logit contributions
ug+0.018
issors+0.017
usted+0.017
bed+0.017
bowl+0.016
Neg logit contributions
 relief-0.021
 gentleman-0.020
 lady-0.020
 woman-0.019
 stranger-0.019
TokenI
Token ID40
Feature activation+3.94e-02

Pos logit contributions
bar+0.057
bed+0.054
utes+0.052
issors+0.050
amb+0.050
Neg logit contributions
 relief-0.072
 cal-0.064
 service-0.062
 witch-0.061
 prayer-0.060
Token will
Token ID481
Feature activation+1.34e-02

Pos logit contributions
 relief+0.070
 gentleman+0.069
 Heidi+0.068
 stranger+0.067
 woman+0.067
Neg logit contributions
ug-0.052
om-0.050
bowl-0.049
 built-0.047
issors-0.047
Token name
Token ID1438
Feature activation+5.22e-02

Pos logit contributions
 relief+0.022
 gentleman+0.021
 cal+0.021
 woman+0.020
 stranger+0.020
Neg logit contributions
bed-0.019
ug-0.019
issors-0.019
bar-0.018
 often-0.018
Token him
Token ID683
Feature activation-3.13e-03

Pos logit contributions
 relief+0.107
 gentleman+0.101
 woman+0.097
 cal+0.094
 witch+0.093
Neg logit contributions
ug-0.059
ors-0.056
bar-0.056
issors-0.055
bed-0.054
Token Sno
Token ID42578
Feature activation-2.29e-02

Pos logit contributions
ug+0.005
issors+0.005
bed+0.005
ors+0.005
usted+0.004
Neg logit contributions
 relief-0.005
 gentleman-0.005
 woman-0.005
 stranger-0.005
 cal-0.005
Tokenopy
Token ID11081
Feature activation-1.43e-01

Pos logit contributions
ug+0.021
ors+0.020
issors+0.019
utes+0.018
om+0.018
Neg logit contributions
 relief-0.062
 gentleman-0.055
 cal-0.055
 lady-0.054
 woman-0.054
Token.
Token ID13
Feature activation+1.04e-02

Pos logit contributions
ug+0.223
ors+0.215
issors+0.213
bed+0.207
usted+0.204
Neg logit contributions
 relief-0.251
 cal-0.224
 gentleman-0.217
 Heidi-0.210
 woman-0.208
Token Because
Token ID4362
Feature activation+1.13e-02

Pos logit contributions
 relief+0.020
 service+0.018
 woman+0.018
 stranger+0.018
 lady+0.018
Neg logit contributions
bar-0.013
issors-0.012
ug-0.012
bowl-0.012
ors-0.011
Token he
Token ID339
Feature activation-2.06e-02

Pos logit contributions
 relief+0.018
 gentleman+0.016
 cal+0.015
 lady+0.015
 woman+0.015
Neg logit contributions
ug-0.021
bar-0.019
issors-0.019
ors-0.019
utes-0.019
Token is
Token ID318
Feature activation+2.26e-02

Pos logit contributions
om+0.022
ors+0.022
 often+0.021
bed+0.021
 fet+0.020
Neg logit contributions
 relief-0.045
 gentleman-0.041
 woman-0.039
 stranger-0.039
 lady-0.038
Token nos
Token ID43630
Feature activation+1.37e-02

Pos logit contributions
 relief+0.043
 gentleman+0.043
 service+0.040
vation+0.039
ment+0.038
Neg logit contributions
 often-0.027
ug-0.025
 both-0.025
om-0.025
 everywhere-0.025
Token't
Token ID470
Feature activation-5.80e-02

Pos logit contributions
ug+0.079
leys+0.079
issors+0.077
eds+0.075
bowl+0.075
Neg logit contributions
 cal-0.090
 lady-0.089
 relief-0.088
 stranger-0.086
 gentleman-0.084
Token want
Token ID765
Feature activation+5.87e-03

Pos logit contributions
ug+0.101
issors+0.097
usted+0.094
ors+0.093
bar+0.093
Neg logit contributions
 relief-0.089
 lady-0.085
 gentleman-0.084
 woman-0.082
 cal-0.081
Token to
Token ID284
Feature activation+2.76e-03

Pos logit contributions
 relief+0.011
 cal+0.011
 lady+0.011
 stranger+0.011
 gentleman+0.011
Neg logit contributions
ug-0.008
issors-0.007
utes-0.007
usted-0.007
bar-0.007
Token get
Token ID651
Feature activation+2.13e-02

Pos logit contributions
 relief+0.005
 gentleman+0.004
 woman+0.004
 lady+0.004
 stranger+0.004
Neg logit contributions
ug-0.004
issors-0.004
bed-0.004
bar-0.004
ors-0.004
Token anything
Token ID1997
Feature activation+4.97e-02

Pos logit contributions
 relief+0.032
 lady+0.031
 gentleman+0.030
 cal+0.030
 woman+0.030
Neg logit contributions
ug-0.038
issors-0.036
utes-0.035
usted-0.035
ors-0.034
Token stuck
Token ID7819
Feature activation-1.43e-01

Pos logit contributions
 gentleman+0.089
 relief+0.089
 woman+0.087
 lady+0.086
 stranger+0.081
Neg logit contributions
ug-0.071
issors-0.066
ors-0.066
usted-0.066
bar-0.066
Token in
Token ID287
Feature activation-4.97e-02

Pos logit contributions
issors+0.269
utes+0.266
ug+0.258
usted+0.256
ants+0.254
Neg logit contributions
 lady-0.178
 relief-0.172
 man-0.167
 stranger-0.166
 gentleman-0.162
Token his
Token ID465
Feature activation-6.16e-03

Pos logit contributions
ug+0.107
usted+0.105
issors+0.105
utes+0.105
leys+0.103
Neg logit contributions
 relief-0.056
 lady-0.050
 cal-0.049
 woman-0.047
 stranger-0.047
Token teeth
Token ID9941
Feature activation-5.63e-02

Pos logit contributions
ug+0.009
issors+0.009
bar+0.008
bed+0.008
usted+0.008
Neg logit contributions
 relief-0.011
 gentleman-0.011
 lady-0.011
 woman-0.010
 cal-0.010
Token again
Token ID757
Feature activation+2.03e-02

Pos logit contributions
ug+0.115
issors+0.111
usted+0.107
utes+0.106
ors+0.106
Neg logit contributions
 relief-0.069
 lady-0.065
 gentleman-0.064
 woman-0.062
 cal-0.062
Token.
Token ID13
Feature activation-2.17e-03

Pos logit contributions
 relief+0.026
 gentleman+0.024
 lady+0.024
 woman+0.024
 cal+0.023
Neg logit contributions
ug-0.041
issors-0.039
usted-0.038
bed-0.038
ors-0.038
Token all
Token ID477
Feature activation-4.63e-02

Pos logit contributions
ug+0.023
issors+0.022
usted+0.022
bar+0.021
bed+0.021
Neg logit contributions
 relief-0.027
 lady-0.025
 gentleman-0.025
 woman-0.025
 cal-0.025
Token valued
Token ID17560
Feature activation-4.68e-02

Pos logit contributions
ug+0.065
usted+0.059
bar+0.058
bed+0.058
bowl+0.057
Neg logit contributions
 relief-0.087
 gentleman-0.083
 lady-0.082
 woman-0.081
 cal-0.080
Token their
Token ID511
Feature activation-1.65e-02

Pos logit contributions
ug+0.068
issors+0.064
bed+0.064
ors+0.061
usted+0.061
Neg logit contributions
 relief-0.084
 gentleman-0.082
 lady-0.081
 woman-0.081
 stranger-0.079
Token l
Token ID300
Feature activation-3.26e-02

Pos logit contributions
bed+0.024
 clubs+0.024
ug+0.024
issors+0.024
usted+0.024
Neg logit contributions
 relief-0.026
 gentleman-0.025
 Heidi-0.024
 woman-0.024
 cal-0.024
Tokenoll
Token ID692
Feature activation-4.38e-02

Pos logit contributions
ug+0.056
issors+0.053
usted+0.051
bar+0.051
ors+0.050
Neg logit contributions
 relief-0.054
 woman-0.050
 lady-0.050
 gentleman-0.049
 cal-0.049
Tokenip
Token ID541
Feature activation-1.42e-01

Pos logit contributions
ug+0.052
ors+0.048
issors+0.048
bed+0.046
bar+0.045
Neg logit contributions
 relief-0.096
 cal-0.090
 gentleman-0.089
 woman-0.089
 lady-0.089
Tokenops
Token ID2840
Feature activation-8.96e-02

Pos logit contributions
bed+0.109
ug+0.108
ors+0.108
issors+0.101
usted+0.099
Neg logit contributions
 relief-0.363
 gentleman-0.342
 woman-0.339
 cal-0.338
 stranger-0.337
Token the
Token ID262
Feature activation-2.46e-02

Pos logit contributions
ug+0.138
bowl+0.130
bed+0.130
bar+0.129
usted+0.129
Neg logit contributions
 relief-0.161
 gentleman-0.153
 woman-0.147
 cal-0.144
 stranger-0.144
Token same
Token ID976
Feature activation-5.98e-03

Pos logit contributions
issors+0.031
bar+0.029
 clubs+0.029
bowl+0.029
ug+0.028
Neg logit contributions
 relief-0.045
 gentleman-0.043
 cal-0.041
 Heidi-0.041
 woman-0.041
Token.
Token ID13
Feature activation-6.61e-03

Pos logit contributions
ug+0.009
bed+0.009
usted+0.009
issors+0.008
bowl+0.008
Neg logit contributions
 relief-0.010
 gentleman-0.009
 cal-0.009
 Heidi-0.009
 woman-0.009
Token They
Token ID1119
Feature activation+2.45e-03

Pos logit contributions
ug+0.011
issors+0.010
usted+0.010
bar+0.010
ors+0.010
Neg logit contributions
 relief-0.011
 lady-0.010
 gentleman-0.010
 woman-0.010
 cal-0.010
Token.
Token ID13
Feature activation-1.40e-03

Pos logit contributions
 relief+0.049
 gentleman+0.047
 stranger+0.044
 Heidi+0.044
 woman+0.044
Neg logit contributions
bed-0.037
ug-0.037
usted-0.035
bar-0.033
eds-0.033
Token
Token ID220
Feature activation-4.74e-02

Pos logit contributions
ug+0.002
issors+0.002
usted+0.002
bar+0.002
bed+0.002
Neg logit contributions
 relief-0.003
 stranger-0.003
 lady-0.003
 gentleman-0.003
 woman-0.003
Token
Token ID198
Feature activation-3.24e-03

Pos logit contributions
ug+0.074
issors+0.070
usted+0.068
utes+0.067
bed+0.066
Neg logit contributions
 relief-0.084
 gentleman-0.080
 stranger-0.080
 woman-0.080
 lady-0.079
Token
Token ID198
Feature activation+1.07e-02

Pos logit contributions
ug+0.007
issors+0.007
ors+0.006
bar+0.006
usted+0.006
Neg logit contributions
 relief-0.004
 lady-0.004
 gentleman-0.004
 cal-0.004
 woman-0.004
TokenThe
Token ID464
Feature activation+2.40e-02

Pos logit contributions
 relief+0.016
 lady+0.016
 gentleman+0.015
 woman+0.015
 cal+0.015
Neg logit contributions
ug-0.019
issors-0.018
ors-0.018
usted-0.017
bar-0.017
Token next
Token ID1306
Feature activation+8.74e-02

Pos logit contributions
 Debbie+0.036
 gentleman+0.035
 MacDonald+0.035
 Heidi+0.035
ment+0.035
Neg logit contributions
ug-0.032
 everywhere-0.031
 clubs-0.030
 bricks-0.030
bed-0.029
Token day
Token ID1110
Feature activation+4.25e-02

Pos logit contributions
 gentleman+0.203
 relief+0.202
 Heidi+0.202
 cal+0.197
 amused+0.195
Neg logit contributions
ug-0.075
utes-0.056
 often-0.048
bar-0.047
inn-0.047
Token,
Token ID11
Feature activation+5.27e-03

Pos logit contributions
 gentleman+0.087
 relief+0.085
 pastor+0.080
 stranger+0.079
 woman+0.078
Neg logit contributions
ug-0.053
usted-0.045
bed-0.042
 built-0.041
leys-0.040
Token the
Token ID262
Feature activation+2.90e-02

Pos logit contributions
 gentleman+0.010
 witch+0.009
 title+0.009
 woman+0.009
 pastor+0.009
Neg logit contributions
ug-0.006
bed-0.006
usted-0.006
bowl-0.005
bar-0.005
Token bird
Token ID6512
Feature activation-1.24e-02

Pos logit contributions
 MacDonald+0.049
ment+0.048
 gentleman+0.046
 Debbie+0.046
 Heidi+0.044
Neg logit contributions
 everywhere-0.035
ug-0.033
 bricks-0.032
 brick-0.031
bed-0.031
Token was
Token ID373
Feature activation-1.08e-02

Pos logit contributions
ug+0.018
bed+0.017
bar+0.016
eds+0.016
usted+0.016
Neg logit contributions
 relief-0.023
 gentleman-0.022
 Heidi-0.020
 stranger-0.020
 lady-0.020
Token have
Token ID423
Feature activation-3.24e-02

Pos logit contributions
ug+0.108
issors+0.103
bar+0.100
usted+0.099
bed+0.099
Neg logit contributions
 relief-0.093
 lady-0.089
 gentleman-0.089
 woman-0.087
 cal-0.086
Token together
Token ID1978
Feature activation-4.45e-02

Pos logit contributions
ug+0.034
bed+0.033
eds+0.030
ors+0.030
leys+0.030
Neg logit contributions
 relief-0.073
 gentleman-0.072
 woman-0.070
 lady-0.070
 stranger-0.069
Token tomorrow
Token ID9439
Feature activation+3.05e-02

Pos logit contributions
ug+0.075
bed+0.071
usted+0.069
issors+0.069
bar+0.069
Neg logit contributions
 relief-0.071
 gentleman-0.068
 woman-0.066
 lady-0.066
 stranger-0.064
Token.
Token ID13
Feature activation-4.56e-03

Pos logit contributions
 relief+0.047
 gentleman+0.044
 lady+0.043
 woman+0.042
 cal+0.042
Neg logit contributions
ug-0.055
issors-0.051
ors-0.051
bed-0.051
usted-0.050
Token The
Token ID383
Feature activation+1.52e-02

Pos logit contributions
ug+0.009
issors+0.008
usted+0.008
ors+0.008
bar+0.008
Neg logit contributions
 relief-0.006
 lady-0.006
 gentleman-0.006
 woman-0.006
 cal-0.006
Token end
Token ID886
Feature activation+8.75e-02

Pos logit contributions
 gentleman+0.024
 Heidi+0.024
 Eva+0.024
 Debbie+0.023
 cal+0.023
Neg logit contributions
issors-0.020
usted-0.020
 clubs-0.019
ug-0.019
bar-0.019
Token.
Token ID13
Feature activation+7.63e-03

Pos logit contributions
 gentleman+0.159
 relief+0.150
 woman+0.146
 Heidi+0.145
 stranger+0.145
Neg logit contributions
ug-0.128
eds-0.117
ors-0.113
utes-0.113
om-0.111
Token<|endoftext|>
Token ID50256
Feature activation+1.79e-02

Pos logit contributions
 relief+0.011
 lady+0.010
 gentleman+0.010
 woman+0.010
 cal+0.010
Neg logit contributions
ug-0.014
issors-0.013
usted-0.013
ors-0.013
bar-0.013
TokenOnce
Token ID7454
Feature activation-9.35e-03

Pos logit contributions
 relief+0.026
 lady+0.025
 gentleman+0.025
 woman+0.024
 cal+0.024
Neg logit contributions
ug-0.032
issors-0.031
usted-0.030
ors-0.029
bar-0.029
Token upon
Token ID2402
Feature activation-2.26e-03

Pos logit contributions
issors+0.015
bar+0.015
ug+0.015
bed+0.014
usted+0.014
Neg logit contributions
 lady-0.015
 man-0.014
 woman-0.014
 stranger-0.014
 relief-0.014
Token a
Token ID257
Feature activation+1.17e-02

Pos logit contributions
ug+0.004
issors+0.004
bed+0.004
usted+0.004
bowl+0.004
Neg logit contributions
 relief-0.003
 gentleman-0.003
 lady-0.003
 woman-0.003
 cal-0.003
Token and
Token ID290
Feature activation-4.04e-03

Pos logit contributions
ug+0.009
issors+0.009
utes+0.009
ors+0.008
usted+0.008
Neg logit contributions
 relief-0.005
 lady-0.005
 gentleman-0.005
 cal-0.005
 woman-0.005
Token it
Token ID340
Feature activation+6.28e-03

Pos logit contributions
ug+0.006
issors+0.006
bed+0.006
usted+0.006
ors+0.005
Neg logit contributions
 relief-0.007
 gentleman-0.007
 lady-0.007
 woman-0.007
 cal-0.007
Token made
Token ID925
Feature activation-5.15e-02

Pos logit contributions
 lady+0.009
 relief+0.009
 man+0.008
 cal+0.008
 woman+0.008
Neg logit contributions
ug-0.012
issors-0.011
utes-0.011
ants-0.010
ors-0.010
Token a
Token ID257
Feature activation-3.15e-02

Pos logit contributions
ug+0.100
issors+0.096
usted+0.096
bed+0.094
ors+0.092
Neg logit contributions
 relief-0.066
 gentleman-0.066
 lady-0.063
 woman-0.062
 stranger-0.061
Token loud
Token ID7812
Feature activation+1.18e-02

Pos logit contributions
ug+0.066
issors+0.064
ors+0.062
usted+0.061
bed+0.061
Neg logit contributions
 relief-0.039
 lady-0.036
 gentleman-0.035
 woman-0.035
 cal-0.035
Token noise
Token ID7838
Feature activation+7.91e-02

Pos logit contributions
 gentleman+0.011
 relief+0.011
 woman+0.010
 lady+0.010
 stranger+0.010
Neg logit contributions
ug-0.027
bar-0.025
bed-0.025
ors-0.025
usted-0.025
Token.
Token ID13
Feature activation-5.59e-03

Pos logit contributions
 relief+0.136
 gentleman+0.132
 Heidi+0.127
 woman+0.124
 lady+0.124
Neg logit contributions
ug-0.128
ors-0.118
bed-0.116
usted-0.115
bar-0.111
Token
Token ID220
Feature activation-4.94e-02

Pos logit contributions
ug+0.012
issors+0.011
bar+0.011
eds+0.011
ors+0.011
Neg logit contributions
 lady-0.007
 relief-0.007
 woman-0.006
 gentleman-0.006
 man-0.006
Token
Token ID198
Feature activation-4.74e-03

Pos logit contributions
ug+0.111
bar+0.107
issors+0.105
ors+0.104
usted+0.103
Neg logit contributions
 lady-0.046
 relief-0.045
 woman-0.041
 man-0.040
 cal-0.040
Token
Token ID198
Feature activation+1.02e-02

Pos logit contributions
ug+0.012
bar+0.012
issors+0.012
ors+0.011
usted+0.011
Neg logit contributions
 lady-0.004
 woman-0.003
 man-0.003
 Heidi-0.003
 relief-0.003
TokenHer
Token ID9360
Feature activation+3.23e-02

Pos logit contributions
 lady+0.014
 relief+0.013
 woman+0.013
 gentleman+0.013
 stranger+0.013
Neg logit contributions
ug-0.020
bar-0.019
issors-0.018
ors-0.018
eds-0.018
Token,
Token ID11
Feature activation+3.29e-04

Pos logit contributions
 relief+0.055
 gentleman+0.054
 woman+0.050
 cal+0.049
 stranger+0.048
Neg logit contributions
ug-0.037
usted-0.035
ors-0.034
bed-0.034
om-0.034
Token she
Token ID673
Feature activation-1.35e-03

Pos logit contributions
 relief+0.001
 cal+0.001
 gentleman+0.001
 gru+0.000
vation+0.000
Neg logit contributions
usted-0.000
bed-0.000
ants-0.000
 everywhere-0.000
 often-0.000
Token wants
Token ID3382
Feature activation+1.56e-02

Pos logit contributions
ug+0.003
issors+0.003
bar+0.002
usted+0.002
bed+0.002
Neg logit contributions
 relief-0.002
 lady-0.002
 woman-0.002
 cal-0.002
 gentleman-0.002
Token to
Token ID284
Feature activation+1.84e-02

Pos logit contributions
 lady+0.032
 relief+0.032
 woman+0.031
 stranger+0.031
 gentleman+0.031
Neg logit contributions
ug-0.019
issors-0.017
ors-0.016
usted-0.016
bed-0.016
Token paint
Token ID7521
Feature activation+1.62e-02

Pos logit contributions
 lady+0.027
 relief+0.027
 woman+0.024
 service+0.024
 cal+0.024
Neg logit contributions
ug-0.034
ors-0.032
eds-0.032
leys-0.032
issors-0.031
Token something
Token ID1223
Feature activation+7.40e-02

Pos logit contributions
 relief+0.028
 gentleman+0.027
 woman+0.025
 stranger+0.025
 lady+0.024
Neg logit contributions
ug-0.026
issors-0.025
bar-0.025
bed-0.024
ors-0.024
Token new
Token ID649
Feature activation-1.25e-02

Pos logit contributions
 relief+0.118
 gentleman+0.109
 woman+0.102
 cal+0.099
 Heidi+0.099
Neg logit contributions
ug-0.125
utes-0.122
issors-0.121
usted-0.119
ors-0.118
Token.
Token ID13
Feature activation-6.32e-03

Pos logit contributions
ug+0.021
bed+0.020
issors+0.020
ors+0.020
usted+0.019
Neg logit contributions
 relief-0.020
 gentleman-0.019
 cal-0.018
 woman-0.018
 lady-0.018
Token She
Token ID1375
Feature activation+8.32e-03

Pos logit contributions
issors+0.010
usted+0.010
ug+0.010
bar+0.009
bed+0.009
Neg logit contributions
 relief-0.010
 cal-0.009
 service-0.009
 gentleman-0.009
 woman-0.009
Token thinks
Token ID6834
Feature activation+3.38e-02

Pos logit contributions
 lady+0.012
 relief+0.012
 cal+0.012
 woman+0.012
 man+0.012
Neg logit contributions
ug-0.014
bar-0.014
issors-0.014
bed-0.013
bowl-0.013
Token and
Token ID290
Feature activation-9.04e-03

Pos logit contributions
 relief+0.049
 lady+0.048
 woman+0.046
 cal+0.045
 stranger+0.044
Neg logit contributions
ug-0.058
issors-0.057
eds-0.056
bowl-0.056
bar-0.056
Token friendly
Token ID8030
Feature activation+1.14e-02

Pos logit contributions
ug+0.053
utes+0.053
issors+0.052
usted+0.051
bar+0.051
Neg logit contributions
 relief-0.049
 gentleman-0.046
 Heidi-0.044
 cal-0.043
 lady-0.042
Token smile
Token ID8212
Feature activation+2.41e-02

Pos logit contributions
 relief+0.014
 cal+0.014
 Heidi+0.013
 Eva+0.013
 gentleman+0.013
Neg logit contributions
ug-0.020
utes-0.019
issors-0.018
 clubs-0.018
usted-0.018
Token.
Token ID13
Feature activation-9.31e-03

Pos logit contributions
 relief+0.046
 cal+0.042
 gentleman+0.041
 lady+0.040
 woman+0.040
Neg logit contributions
ug-0.035
bar-0.033
usted-0.032
issors-0.032
ors-0.032
Token
Token ID198
Feature activation-1.09e-02

Pos logit contributions
ug+0.014
issors+0.012
bowl+0.012
bar+0.012
utes+0.012
Neg logit contributions
 relief-0.017
 lady-0.016
 stranger-0.016
 gentleman-0.015
 woman-0.015
Token
Token ID198
Feature activation+3.22e-03

Pos logit contributions
issors+0.017
ug+0.017
usted+0.016
ors+0.016
bar+0.015
Neg logit contributions
 relief-0.019
 lady-0.019
 gentleman-0.018
 woman-0.018
 stranger-0.018
TokenEveryone
Token ID16190
Feature activation+8.05e-02

Pos logit contributions
 relief+0.005
 lady+0.005
 gentleman+0.005
 woman+0.005
 stranger+0.005
Neg logit contributions
ug-0.005
issors-0.005
ors-0.005
bed-0.005
usted-0.005
Token in
Token ID287
Feature activation-2.01e-02

Pos logit contributions
 relief+0.158
 gentleman+0.144
 lady+0.142
 Heidi+0.141
 woman+0.140
Neg logit contributions
ug-0.109
bed-0.100
usted-0.099
bowl-0.099
ors-0.093
Token the
Token ID262
Feature activation+2.95e-02

Pos logit contributions
ug+0.029
ir+0.026
utes+0.024
om+0.024
 targets+0.024
Neg logit contributions
 gentleman-0.034
 relief-0.033
 lady-0.033
 cal-0.032
 stranger-0.031
Token park
Token ID3952
Feature activation+5.34e-02

Pos logit contributions
 gen+0.050
 relief+0.050
 Heidi+0.049
 cal+0.048
 gentleman+0.048
Neg logit contributions
 clubs-0.039
ug-0.038
issors-0.038
utes-0.036
 targets-0.035
Token was
Token ID373
Feature activation-2.62e-02

Pos logit contributions
 relief+0.111
 gentleman+0.101
 cal+0.100
 stranger+0.099
 Heidi+0.098
Neg logit contributions
ug-0.065
bed-0.065
 built-0.060
usted-0.059
bar-0.057
Token so
Token ID523
Feature activation-1.35e-02

Pos logit contributions
ug+0.035
bar+0.031
usted+0.031
 built+0.031
 often+0.031
Neg logit contributions
 gentleman-0.047
 relief-0.047
 woman-0.044
 stranger-0.044
 witch-0.044
Token was
Token ID373
Feature activation-1.52e-02

Pos logit contributions
 gentleman+0.064
 relief+0.064
 lady+0.059
 stranger+0.058
 cal+0.058
Neg logit contributions
ug-0.050
usted-0.047
bar-0.046
bowl-0.046
issors-0.046
Token the
Token ID262
Feature activation+3.22e-03

Pos logit contributions
ug+0.022
 often+0.020
bed+0.020
issors+0.020
usted+0.020
Neg logit contributions
 gentleman-0.027
 woman-0.025
 stranger-0.025
 lady-0.024
 service-0.024
Token day
Token ID1110
Feature activation+4.05e-02

Pos logit contributions
 relief+0.005
 gentleman+0.005
ment+0.005
 Heidi+0.005
 cal+0.005
Neg logit contributions
issors-0.004
 targets-0.004
utes-0.004
ug-0.004
bar-0.004
Token the
Token ID262
Feature activation+2.19e-03

Pos logit contributions
 gentleman+0.075
 relief+0.072
 stranger+0.069
 cal+0.068
 pastor+0.068
Neg logit contributions
ug-0.059
usted-0.051
bed-0.050
ors-0.048
 often-0.047
Token weird
Token ID7650
Feature activation+4.23e-03

Pos logit contributions
 relief+0.003
 gentleman+0.003
 Heidi+0.003
 cal+0.003
ment+0.003
Neg logit contributions
issors-0.004
ug-0.003
utes-0.003
 targets-0.003
 hamm-0.003
Token witch
Token ID16365
Feature activation+1.48e-02

Pos logit contributions
 relief+0.006
 cal+0.006
 gentleman+0.006
 Heidi+0.005
 Debbie+0.005
Neg logit contributions
ug-0.007
alk-0.007
bed-0.007
utes-0.007
 often-0.006
Token learned
Token ID4499
Feature activation-5.88e-02

Pos logit contributions
 relief+0.030
 gentleman+0.029
 cal+0.027
 stranger+0.027
 lady+0.026
Neg logit contributions
ug-0.018
bed-0.017
ors-0.016
bowl-0.016
bar-0.016
Token that
Token ID326
Feature activation-5.01e-02

Pos logit contributions
usted+0.083
ug+0.083
ors+0.079
issors+0.077
bed+0.077
Neg logit contributions
 relief-0.101
 gentleman-0.100
 woman-0.094
 lady-0.091
 stranger-0.089
Token discussing
Token ID11142
Feature activation-4.23e-03

Pos logit contributions
ug+0.076
bed+0.072
issors+0.070
usted+0.070
bar+0.069
Neg logit contributions
 gentleman-0.084
 relief-0.081
 lady-0.075
 woman-0.075
 stranger-0.074
Token can
Token ID460
Feature activation-6.84e-02

Pos logit contributions
ug+0.006
issors+0.006
ors+0.006
 clubs+0.006
 everywhere+0.006
Neg logit contributions
 gentleman-0.008
 relief-0.007
 woman-0.007
 lady-0.007
 stranger-0.006
Token be
Token ID307
Feature activation+4.07e-03

Pos logit contributions
ug+0.105
bar+0.101
issors+0.098
bed+0.096
ors+0.095
Neg logit contributions
 relief-0.119
 gentleman-0.116
 woman-0.109
 lady-0.107
 stranger-0.106
Token tomorrow
Token ID9439
Feature activation+3.03e-02

Pos logit contributions
ug+0.022
issors+0.021
usted+0.021
ors+0.020
bar+0.020
Neg logit contributions
 relief-0.019
 lady-0.018
 gentleman-0.018
 woman-0.017
 cal-0.017
Token!"
Token ID2474
Feature activation-1.50e-02

Pos logit contributions
 relief+0.065
 gentleman+0.061
 cal+0.059
 woman+0.058
 lady+0.058
Neg logit contributions
ug-0.036
issors-0.034
ors-0.033
usted-0.033
bar-0.031
Token
Token ID220
Feature activation-6.43e-02

Pos logit contributions
ug+0.021
bed+0.020
issors+0.020
usted+0.020
ors+0.019
Neg logit contributions
 relief-0.028
 gentleman-0.026
 stranger-0.026
 cal-0.025
 lady-0.025
Token
Token ID198
Feature activation-2.00e-02

Pos logit contributions
ug+0.117
issors+0.112
usted+0.108
bar+0.107
ors+0.107
Neg logit contributions
 relief-0.093
 lady-0.089
 gentleman-0.087
 woman-0.086
 cal-0.085
Token
Token ID198
Feature activation-2.00e-03

Pos logit contributions
ug+0.043
issors+0.043
bar+0.041
usted+0.041
ors+0.041
Neg logit contributions
 lady-0.023
 relief-0.022
 woman-0.021
 gentleman-0.021
 man-0.020
TokenThe
Token ID464
Feature activation+2.00e-02

Pos logit contributions
ug+0.004
issors+0.003
ors+0.003
usted+0.003
bar+0.003
Neg logit contributions
 relief-0.003
 lady-0.003
 gentleman-0.003
 woman-0.003
 stranger-0.003
Token man
Token ID582
Feature activation+2.95e-02

Pos logit contributions
 relief+0.023
 Heidi+0.022
 Eva+0.020
 Alice+0.020
 Debbie+0.020
Neg logit contributions
issors-0.036
 hamm-0.036
 clubs-0.034
 hammer-0.034
 drib-0.034
Token felt
Token ID2936
Feature activation-7.50e-03

Pos logit contributions
 relief+0.048
 gentleman+0.045
 service+0.043
 Heidi+0.043
 stranger+0.043
Neg logit contributions
usted-0.046
ug-0.046
bed-0.046
bar-0.044
bowl-0.041
Token sorry
Token ID7926
Feature activation+4.13e-02

Pos logit contributions
ug+0.010
usted+0.010
bowl+0.010
bed+0.009
bar+0.009
Neg logit contributions
 gentleman-0.012
 woman-0.012
 service-0.012
 lady-0.012
 witch-0.012
Token for
Token ID329
Feature activation+4.50e-03

Pos logit contributions
 relief+0.093
 gentleman+0.092
 woman+0.089
 lady+0.089
 stranger+0.088
Neg logit contributions
ug-0.040
usted-0.037
issors-0.035
bar-0.035
bed-0.034
Token her
Token ID607
Feature activation-5.64e-03

Pos logit contributions
 gentleman+0.008
 relief+0.008
 woman+0.008
 stranger+0.007
 prayer+0.007
Neg logit contributions
ug-0.006
bed-0.006
ir-0.005
 often-0.005
bar-0.005
Token each
Token ID1123
Feature activation-2.50e-02

Pos logit contributions
 relief+0.021
 lady+0.020
 gentleman+0.019
 woman+0.019
 cal+0.019
Neg logit contributions
ug-0.016
issors-0.015
usted-0.015
bar-0.015
ors-0.015
Token other
Token ID584
Feature activation-3.97e-02

Pos logit contributions
ug+0.047
ors+0.043
utes+0.043
bar+0.043
issors+0.042
Neg logit contributions
 lady-0.039
 relief-0.036
 woman-0.036
 stranger-0.035
 man-0.035
Token.
Token ID13
Feature activation+9.63e-03

Pos logit contributions
bed+0.054
usted+0.054
issors+0.052
ors+0.051
ug+0.050
Neg logit contributions
 relief-0.075
 gentleman-0.074
 woman-0.068
 stranger-0.067
 Heidi-0.067
Token They
Token ID1119
Feature activation+6.34e-03

Pos logit contributions
 relief+0.018
 gentleman+0.017
 lady+0.017
 woman+0.016
 stranger+0.016
Neg logit contributions
ug-0.014
issors-0.013
usted-0.013
utes-0.013
bed-0.013
Token nod
Token ID18666
Feature activation+4.93e-02

Pos logit contributions
 gentleman+0.010
 relief+0.010
 woman+0.010
 lady+0.010
 service+0.010
Neg logit contributions
issors-0.009
ug-0.009
usted-0.009
bed-0.009
bowl-0.008
Token.
Token ID13
Feature activation+9.92e-03

Pos logit contributions
 gentleman+0.082
 relief+0.080
 stranger+0.076
 woman+0.076
 lady+0.075
Neg logit contributions
ug-0.076
bed-0.075
utes-0.071
usted-0.071
issors-0.071
Token They
Token ID1119
Feature activation+4.66e-03

Pos logit contributions
 relief+0.018
 gentleman+0.017
 lady+0.017
 stranger+0.017
 woman+0.017
Neg logit contributions
ug-0.014
issors-0.014
utes-0.013
usted-0.013
bed-0.013
Token say
Token ID910
Feature activation+6.26e-02

Pos logit contributions
 gentleman+0.008
 woman+0.008
 relief+0.008
 lady+0.007
 service+0.007
Neg logit contributions
issors-0.007
usted-0.006
ug-0.006
bowl-0.006
 drib-0.006
Token,
Token ID11
Feature activation-7.65e-03

Pos logit contributions
 relief+0.102
 gentleman+0.101
 woman+0.095
 stranger+0.093
 lady+0.092
Neg logit contributions
ug-0.102
bed-0.094
issors-0.093
usted-0.093
utes-0.091
Token "
Token ID366
Feature activation-4.49e-02

Pos logit contributions
ug+0.012
issors+0.012
bed+0.012
usted+0.011
utes+0.011
Neg logit contributions
 relief-0.013
 gentleman-0.012
 stranger-0.012
 lady-0.012
 cal-0.012
TokenOK
Token ID11380
Feature activation+5.10e-02

Pos logit contributions
issors+0.074
ug+0.073
bed+0.073
usted+0.071
bar+0.071
Neg logit contributions
 relief-0.069
 gentleman-0.065
 cal-0.063
 stranger-0.063
 witch-0.063
Token.
Token ID13
Feature activation-3.88e-03

Pos logit contributions
 relief+0.068
 Heidi+0.064
 gentleman+0.061
 woman+0.059
 lady+0.058
Neg logit contributions
ug-0.041
 often-0.039
 everywhere-0.039
ors-0.039
bed-0.037
Token One
Token ID1881
Feature activation-1.15e-02

Pos logit contributions
ug+0.008
usted+0.007
eds+0.007
issors+0.007
bed+0.007
Neg logit contributions
 relief-0.005
 lady-0.005
 gentleman-0.005
 man-0.004
 Heidi-0.004
Token day
Token ID1110
Feature activation+5.65e-02

Pos logit contributions
ug+0.010
bowl+0.009
utes+0.008
om+0.008
ogram+0.008
Neg logit contributions
 relief-0.027
 cal-0.026
 gentleman-0.026
 MacDonald-0.025
 Heidi-0.025
Token,
Token ID11
Feature activation+1.31e-02

Pos logit contributions
 relief+0.088
 gentleman+0.084
 lady+0.084
 woman+0.083
 cal+0.083
Neg logit contributions
ug-0.097
issors-0.092
bed-0.091
usted-0.090
ors-0.088
Token his
Token ID465
Feature activation+2.74e-02

Pos logit contributions
 relief+0.019
 lady+0.018
 gentleman+0.017
 woman+0.017
 stranger+0.017
Neg logit contributions
ug-0.025
issors-0.024
usted-0.023
ors-0.023
bar-0.023
Token mom
Token ID1995
Feature activation+4.04e-02

Pos logit contributions
 relief+0.045
 cal+0.042
 gentleman+0.042
 Heidi+0.040
 lady+0.039
Neg logit contributions
ug-0.044
utes-0.043
issors-0.042
usted-0.041
bed-0.040
Token gave
Token ID2921
Feature activation-8.80e-03

Pos logit contributions
 relief+0.076
 gentleman+0.071
 lady+0.071
 woman+0.070
 cal+0.070
Neg logit contributions
ug-0.057
issors-0.054
usted-0.053
bar-0.052
ors-0.052
Token him
Token ID683
Feature activation+2.99e-03

Pos logit contributions
ug+0.016
issors+0.015
utes+0.014
bar+0.014
usted+0.014
Neg logit contributions
 relief-0.013
 lady-0.013
 stranger-0.012
 woman-0.012
 gentleman-0.012
Token a
Token ID257
Feature activation-2.04e-02

Pos logit contributions
 relief+0.005
 gentleman+0.004
 lady+0.004
 woman+0.004
 cal+0.004
Neg logit contributions
ug-0.005
issors-0.005
usted-0.005
bar-0.005
ors-0.005
Token present
Token ID1944
Feature activation+3.32e-02

Pos logit contributions
issors+0.036
ug+0.035
utes+0.035
bar+0.034
usted+0.034
Neg logit contributions
 relief-0.029
 gentleman-0.026
 cal-0.026
 lady-0.025
 Heidi-0.025
Token.
Token ID13
Feature activation-4.66e-03

Pos logit contributions
 relief+0.062
 gentleman+0.058
 cal+0.058
 lady+0.058
 woman+0.057
Neg logit contributions
ug-0.047
issors-0.045
bar-0.043
usted-0.043
bed-0.043
Token,
Token ID11
Feature activation-4.27e-03

Pos logit contributions
 relief+0.015
 gentleman+0.014
 woman+0.014
 cal+0.014
 lady+0.014
Neg logit contributions
ug-0.015
issors-0.014
bar-0.014
ors-0.014
usted-0.014
Token "
Token ID366
Feature activation-4.95e-02

Pos logit contributions
ug+0.007
issors+0.006
usted+0.006
bed+0.006
bar+0.006
Neg logit contributions
 relief-0.007
 gentleman-0.007
 cal-0.007
 stranger-0.007
 lady-0.007
TokenStop
Token ID19485
Feature activation+8.01e-02

Pos logit contributions
ug+0.095
issors+0.093
usted+0.089
bar+0.089
ors+0.088
Neg logit contributions
 relief-0.068
 gentleman-0.062
 lady-0.062
 cal-0.061
 woman-0.060
Token fighting
Token ID4330
Feature activation-3.01e-03

Pos logit contributions
 relief+0.154
 gentleman+0.146
 lady+0.144
 cal+0.143
 woman+0.143
Neg logit contributions
ug-0.111
ors-0.105
issors-0.103
bed-0.101
usted-0.097
Token,
Token ID11
Feature activation-1.17e-02

Pos logit contributions
ug+0.006
issors+0.006
bar+0.005
usted+0.005
utes+0.005
Neg logit contributions
 relief-0.004
 lady-0.004
 gentleman-0.004
 cal-0.004
 woman-0.004
Token children
Token ID1751
Feature activation+6.97e-03

Pos logit contributions
issors+0.019
ug+0.017
usted+0.017
bar+0.017
ors+0.016
Neg logit contributions
 relief-0.018
 cal-0.018
 gentleman-0.017
 prayer-0.017
ment-0.017
Token.
Token ID13
Feature activation+6.73e-03

Pos logit contributions
 relief+0.010
 gentleman+0.010
 lady+0.009
 woman+0.009
 cal+0.009
Neg logit contributions
ug-0.013
issors-0.012
ors-0.012
usted-0.012
bar-0.012
Token Why
Token ID4162
Feature activation+7.44e-02

Pos logit contributions
 relief+0.010
 lady+0.010
 gentleman+0.010
 woman+0.009
 cal+0.009
Neg logit contributions
ug-0.012
issors-0.012
usted-0.011
ors-0.011
bar-0.011
Token don
Token ID836
Feature activation+2.45e-03

Pos logit contributions
 relief+0.144
 gentleman+0.140
 woman+0.134
 cal+0.132
 lady+0.131
Neg logit contributions
ug-0.104
issors-0.102
ors-0.094
usted-0.093
utes-0.090
Token't
Token ID470
Feature activation-3.35e-02

Pos logit contributions
 relief+0.005
 lady+0.005
 gentleman+0.005
 woman+0.005
 stranger+0.005
Neg logit contributions
ug-0.003
issors-0.003
usted-0.003
bar-0.003
bed-0.003
Token you
Token ID345
Feature activation+3.84e-02

Pos logit contributions
ug+0.057
issors+0.056
ors+0.054
bar+0.053
usted+0.053
Neg logit contributions
 relief-0.053
 gentleman-0.049
 cal-0.048
 woman-0.048
 lady-0.048
TokenOne
Token ID3198
Feature activation+4.46e-03

Pos logit contributions
 lady+0.023
 stranger+0.021
 woman+0.021
 man+0.021
 gentleman+0.021
Neg logit contributions
ug-0.020
ors-0.019
issors-0.019
usted-0.019
eds-0.018
Token day
Token ID1110
Feature activation+5.55e-02

Pos logit contributions
 relief+0.009
 gentleman+0.008
 cal+0.008
 lady+0.008
 woman+0.008
Neg logit contributions
ug-0.006
issors-0.005
utes-0.005
bowl-0.005
bar-0.005
Token,
Token ID11
Feature activation+1.67e-02

Pos logit contributions
 relief+0.072
 lady+0.069
 gentleman+0.066
 woman+0.065
 stranger+0.065
Neg logit contributions
ug-0.109
issors-0.107
bar-0.103
usted-0.103
ors-0.102
Token a
Token ID257
Feature activation+2.15e-02

Pos logit contributions
 relief+0.023
 gentleman+0.022
 cal+0.022
 lady+0.021
 woman+0.021
Neg logit contributions
ug-0.031
issors-0.030
bed-0.029
usted-0.029
utes-0.029
Token new
Token ID649
Feature activation+5.26e-02

Pos logit contributions
 relief+0.030
 Heidi+0.030
 gentleman+0.029
ment+0.028
 cal+0.028
Neg logit contributions
ug-0.035
 hamm-0.033
utes-0.033
om-0.031
 accidentally-0.031
Token green
Token ID4077
Feature activation-2.87e-02

Pos logit contributions
 relief+0.065
 cal+0.059
 Heidi+0.056
 gentleman+0.056
 stranger+0.053
Neg logit contributions
ug-0.109
bar-0.103
issors-0.100
utes-0.099
bed-0.097
Token truck
Token ID7779
Feature activation+7.17e-03

Pos logit contributions
ug+0.041
bar+0.038
issors+0.037
bed+0.036
usted+0.036
Neg logit contributions
 relief-0.054
 cal-0.050
 gentleman-0.050
 lady-0.049
 Heidi-0.048
Token came
Token ID1625
Feature activation+2.57e-02

Pos logit contributions
 relief+0.011
 gentleman+0.010
 lady+0.010
 woman+0.010
 cal+0.010
Neg logit contributions
ug-0.013
issors-0.012
usted-0.012
ors-0.012
bar-0.012
Token to
Token ID284
Feature activation+2.94e-02

Pos logit contributions
 lady+0.036
 cal+0.035
 man+0.034
 woman+0.032
 stranger+0.032
Neg logit contributions
issors-0.048
igs-0.043
angles-0.043
ug-0.042
ors-0.042
Token play
Token ID711
Feature activation+8.36e-03

Pos logit contributions
 relief+0.044
 gentleman+0.044
 lady+0.041
 stranger+0.041
 cal+0.041
Neg logit contributions
ug-0.050
issors-0.048
bed-0.046
bar-0.046
utes-0.045
Token.
Token ID13
Feature activation+1.78e-03

Pos logit contributions
 relief+0.011
 lady+0.010
 gentleman+0.010
 woman+0.010
 cal+0.010
Neg logit contributions
ug-0.016
issors-0.016
usted-0.016
bar-0.015
ors-0.015
Token�
Token ID241
Feature activation-2.84e-02

Pos logit contributions
ug+0.005
issors+0.004
usted+0.004
bar+0.004
ors+0.004
Neg logit contributions
 relief-0.004
 lady-0.004
 gentleman-0.004
 woman-0.004
 cal-0.004
TokenLet
Token ID5756
Feature activation-3.08e-03

Pos logit contributions
ug+0.049
issors+0.047
usted+0.045
bed+0.045
utes+0.045
Neg logit contributions
 relief-0.044
 cal-0.041
 lady-0.041
 gentleman-0.040
 stranger-0.040
Tokenâ
Token ID22940
Feature activation-3.09e-02

Pos logit contributions
ug+0.003
usted+0.003
bar+0.003
ors+0.003
issors+0.003
Neg logit contributions
 relief-0.007
 gentleman-0.006
 woman-0.006
 service-0.006
 lady-0.006
Token€
Token ID26391
Feature activation-1.00e-02

Pos logit contributions
ug+0.052
issors+0.051
bar+0.049
bed+0.045
ors+0.045
Neg logit contributions
 woman-0.051
 stranger-0.051
 witch-0.049
 lady-0.049
 relief-0.048
Tokenâ„¢
Token ID8151
Feature activation-2.74e-02

Pos logit contributions
ug+0.013
issors+0.012
ors+0.012
om+0.012
utes+0.012
Neg logit contributions
 relief-0.021
 stranger-0.019
 cal-0.019
 lady-0.019
 woman-0.019
Tokens
Token ID82
Feature activation-8.19e-02

Pos logit contributions
ug+0.039
issors+0.039
utes+0.037
bed+0.037
igs+0.037
Neg logit contributions
 lady-0.048
 gentleman-0.046
 stranger-0.044
 man-0.043
 woman-0.043
Token play
Token ID711
Feature activation+1.17e-02

Pos logit contributions
ug+0.129
issors+0.121
bar+0.118
usted+0.117
bed+0.116
Neg logit contributions
 relief-0.139
 gentleman-0.134
 lady-0.131
 woman-0.131
 cal-0.128
Token a
Token ID257
Feature activation-2.97e-02

Pos logit contributions
 relief+0.019
 gentleman+0.019
 woman+0.018
 lady+0.018
 stranger+0.017
Neg logit contributions
ug-0.019
bed-0.018
bar-0.017
utes-0.017
bowl-0.017
Token game
Token ID983
Feature activation-3.79e-03

Pos logit contributions
ug+0.047
issors+0.045
usted+0.043
bar+0.043
utes+0.043
Neg logit contributions
 relief-0.049
 lady-0.047
 gentleman-0.047
 woman-0.046
 cal-0.045
Token!
Token ID0
Feature activation-1.67e-02

Pos logit contributions
ug+0.006
issors+0.006
bar+0.006
ors+0.006
usted+0.006
Neg logit contributions
 relief-0.006
 gentleman-0.006
 woman-0.006
 lady-0.006
 cal-0.006
Tokenâ
Token ID22940
Feature activation-1.41e-02

Pos logit contributions
ug+0.028
issors+0.027
usted+0.026
bar+0.026
ors+0.025
Neg logit contributions
 relief-0.027
 lady-0.025
 stranger-0.025
 woman-0.025
 gentleman-0.025
Token to
Token ID284
Feature activation+8.81e-03

Pos logit contributions
ug+0.017
usted+0.017
bed+0.016
bar+0.015
 everywhere+0.015
Neg logit contributions
 relief-0.043
 gentleman-0.042
 woman-0.041
 service-0.040
 lady-0.040
Token always
Token ID1464
Feature activation-9.84e-03

Pos logit contributions
 relief+0.014
 gentleman+0.014
 woman+0.014
 lady+0.014
 stranger+0.013
Neg logit contributions
ug-0.014
bar-0.013
issors-0.013
bed-0.013
bowl-0.012
Token listen
Token ID6004
Feature activation+3.48e-02

Pos logit contributions
bed+0.012
 built+0.012
 drib+0.012
bowl+0.012
eds+0.012
Neg logit contributions
 relief-0.017
 gentleman-0.017
 woman-0.017
 gen-0.017
 stranger-0.016
Token.
Token ID13
Feature activation-9.79e-03

Pos logit contributions
 relief+0.075
 gentleman+0.071
 woman+0.069
 lady+0.069
 cal+0.068
Neg logit contributions
ug-0.040
usted-0.038
bed-0.037
issors-0.037
ors-0.037
Token She
Token ID1375
Feature activation+7.53e-03

Pos logit contributions
ug+0.017
issors+0.017
eds+0.016
ors+0.016
bed+0.016
Neg logit contributions
 relief-0.014
 lady-0.014
 gentleman-0.013
 woman-0.013
 cal-0.013
Token was
Token ID373
Feature activation-1.21e-02

Pos logit contributions
 relief+0.014
 lady+0.013
 gentleman+0.013
 woman+0.013
 stranger+0.013
Neg logit contributions
ug-0.011
issors-0.010
bed-0.010
usted-0.010
bar-0.010
Token so
Token ID523
Feature activation+1.56e-03

Pos logit contributions
ug+0.015
usted+0.015
bed+0.014
bar+0.014
 often+0.013
Neg logit contributions
 relief-0.023
 gentleman-0.022
 woman-0.022
 stranger-0.022
 lady-0.021
Token happy
Token ID3772
Feature activation+2.49e-02

Pos logit contributions
 relief+0.003
 woman+0.003
 gentleman+0.003
 witch+0.003
 stranger+0.002
Neg logit contributions
ug-0.002
 often-0.002
om-0.002
 drib-0.002
usted-0.002
Token with
Token ID351
Feature activation+1.16e-02

Pos logit contributions
 relief+0.056
 cal+0.050
 gentleman+0.050
 woman+0.048
 gen+0.048
Neg logit contributions
bed-0.028
ors-0.027
ug-0.027
usted-0.026
utes-0.024
Token her
Token ID607
Feature activation+4.16e-03

Pos logit contributions
 gentleman+0.018
 relief+0.018
 lady+0.018
 woman+0.018
 stranger+0.018
Neg logit contributions
ug-0.019
issors-0.018
usted-0.017
bed-0.017
ors-0.017
Token silly
Token ID14397
Feature activation-2.69e-02

Pos logit contributions
 gentleman+0.007
 relief+0.007
 Heidi+0.007
 cal+0.007
 stranger+0.006
Neg logit contributions
issors-0.006
ug-0.006
utes-0.006
 clubs-0.006
bed-0.006
Token girl
Token ID2576
Feature activation+2.61e-02

Pos logit contributions
 Heidi+0.029
 relief+0.026
 Eva+0.025
 Sophia+0.024
 gentleman+0.024
Neg logit contributions
ug-0.021
 everywhere-0.021
 often-0.021
usted-0.019
bowl-0.019
Token wanted
Token ID2227
Feature activation-1.75e-02

Pos logit contributions
 relief+0.046
 gentleman+0.043
 stranger+0.041
 cal+0.041
 lady+0.041
Neg logit contributions
ug-0.039
usted-0.038
bed-0.038
bar-0.036
bowl-0.036
Token to
Token ID284
Feature activation+1.01e-02

Pos logit contributions
ug+0.010
issors+0.010
usted+0.010
bed+0.009
eds+0.009
Neg logit contributions
 relief-0.044
 gentleman-0.044
 woman-0.041
 cal-0.040
 lady-0.040
Token bake
Token ID28450
Feature activation+4.66e-02

Pos logit contributions
 gentleman+0.017
 relief+0.017
 woman+0.016
 cal+0.016
 stranger+0.016
Neg logit contributions
ug-0.014
issors-0.014
bowl-0.013
bar-0.013
bed-0.013
Token something
Token ID1223
Feature activation+5.71e-02

Pos logit contributions
 gentleman+0.101
 woman+0.091
 stranger+0.091
 relief+0.090
 pastor+0.084
Neg logit contributions
bed-0.050
usted-0.048
issors-0.048
ug-0.047
igs-0.046
Token y
Token ID331
Feature activation-6.88e-02

Pos logit contributions
 relief+0.113
 gentleman+0.110
 woman+0.104
 cal+0.097
 witch+0.096
Neg logit contributions
ug-0.065
bowl-0.065
usted-0.064
bar-0.063
utes-0.063
Tokenummy
Token ID13513
Feature activation-3.74e-02

Pos logit contributions
ug+0.100
issors+0.096
utes+0.093
usted+0.093
bar+0.091
Neg logit contributions
 relief-0.120
 lady-0.116
 gentleman-0.113
 stranger-0.112
 cal-0.110
Token.
Token ID13
Feature activation-1.71e-03

Pos logit contributions
bowl+0.046
ors+0.045
bed+0.044
ug+0.044
usted+0.043
Neg logit contributions
 relief-0.073
 gentleman-0.070
 woman-0.068
 cal-0.066
 pastor-0.064
Token They
Token ID1119
Feature activation-7.88e-05

Pos logit contributions
ug+0.003
issors+0.003
usted+0.003
bed+0.003
ors+0.003
Neg logit contributions
 relief-0.003
 gentleman-0.003
 lady-0.002
 stranger-0.002
 cal-0.002
Token found
Token ID1043
Feature activation-1.30e-02

Pos logit contributions
ug+0.000
utes+0.000
issors+0.000
ors+0.000
eds+0.000
Neg logit contributions
 lady-0.000
 woman-0.000
 relief-0.000
 man-0.000
 cal-0.000
Token flour
Token ID10601
Feature activation+9.29e-03

Pos logit contributions
ug+0.019
usted+0.019
issors+0.019
bar+0.018
bed+0.018
Neg logit contributions
 gentleman-0.022
 relief-0.022
 stranger-0.021
 lady-0.020
 woman-0.020
Token He
Token ID679
Feature activation-1.61e-05

Pos logit contributions
ug+0.010
eds+0.009
utes+0.009
issors+0.009
leys+0.009
Neg logit contributions
 lady-0.008
 relief-0.008
 gentleman-0.008
 man-0.008
 woman-0.008
Token was
Token ID373
Feature activation+1.34e-03

Pos logit contributions
ug+0.000
issors+0.000
utes+0.000
ors+0.000
bar+0.000
Neg logit contributions
 relief-0.000
 lady-0.000
 woman-0.000
 cal-0.000
 gentleman-0.000
Token so
Token ID523
Feature activation+1.32e-02

Pos logit contributions
 relief+0.002
 lady+0.002
 cal+0.002
 woman+0.002
 gentleman+0.002
Neg logit contributions
ug-0.003
ors-0.002
issors-0.002
utes-0.002
eds-0.002
Token excited
Token ID6568
Feature activation+2.87e-03

Pos logit contributions
 relief+0.014
 lady+0.013
 gentleman+0.012
 cal+0.012
 stranger+0.012
Neg logit contributions
ug-0.030
issors-0.029
ors-0.029
bar-0.028
eds-0.028
Token that
Token ID326
Feature activation-1.97e-02

Pos logit contributions
 relief+0.007
 cal+0.006
 gentleman+0.006
 woman+0.006
 lady+0.006
Neg logit contributions
ors-0.003
ug-0.003
usted-0.003
issors-0.003
bed-0.003
Token he
Token ID339
Feature activation-3.18e-02

Pos logit contributions
ug+0.044
issors+0.042
usted+0.041
bar+0.041
ors+0.041
Neg logit contributions
 relief-0.021
 lady-0.020
 gentleman-0.019
 woman-0.019
 cal-0.019
Token started
Token ID2067
Feature activation-6.70e-02

Pos logit contributions
ug+0.049
issors+0.047
utes+0.045
bar+0.045
ors+0.044
Neg logit contributions
 relief-0.055
 lady-0.053
 gentleman-0.052
 cal-0.051
 woman-0.051
Token to
Token ID284
Feature activation+1.25e-02

Pos logit contributions
ug+0.101
issors+0.093
utes+0.089
bar+0.087
ors+0.086
Neg logit contributions
 relief-0.123
 lady-0.120
 cal-0.117
 stranger-0.116
 woman-0.115
Token shake
Token ID13279
Feature activation+2.82e-03

Pos logit contributions
 relief+0.019
 lady+0.018
 gentleman+0.017
 woman+0.017
 cal+0.017
Neg logit contributions
ug-0.022
issors-0.021
usted-0.021
ors-0.020
bar-0.020
Token with
Token ID351
Feature activation+8.82e-03

Pos logit contributions
 relief+0.004
 lady+0.004
 gentleman+0.004
 woman+0.004
 cal+0.004
Neg logit contributions
ug-0.005
issors-0.005
utes-0.005
usted-0.005
bar-0.005
Token happiness
Token ID12157
Feature activation+4.86e-02

Pos logit contributions
 gentleman+0.013
 lady+0.013
 relief+0.013
 woman+0.013
 stranger+0.012
Neg logit contributions
ug-0.015
issors-0.015
usted-0.014
bed-0.014
bar-0.014
Token asked
Token ID1965
Feature activation+6.94e-03

Pos logit contributions
 lady+0.013
 woman+0.013
 relief+0.012
 man+0.012
 cal+0.012
Neg logit contributions
ug-0.021
issors-0.021
bar-0.019
ors-0.019
utes-0.019
Token her
Token ID607
Feature activation+2.64e-02

Pos logit contributions
 relief+0.009
 lady+0.008
 gentleman+0.008
 woman+0.008
 cal+0.008
Neg logit contributions
ug-0.014
issors-0.014
usted-0.013
ors-0.013
bar-0.013
Token mom
Token ID1995
Feature activation+1.47e-02

Pos logit contributions
 relief+0.038
 gentleman+0.035
 lady+0.034
 cal+0.034
 woman+0.034
Neg logit contributions
ug-0.048
issors-0.047
usted-0.046
utes-0.045
bed-0.045
Token if
Token ID611
Feature activation-1.82e-02

Pos logit contributions
 relief+0.025
 lady+0.024
 gentleman+0.023
 woman+0.023
 cal+0.023
Neg logit contributions
ug-0.024
issors-0.023
usted-0.021
ors-0.021
utes-0.021
Token she
Token ID673
Feature activation-3.68e-02

Pos logit contributions
 accidentally+0.029
bed+0.027
ug+0.027
 couldn+0.026
bar+0.026
Neg logit contributions
 relief-0.029
 gentleman-0.026
 cal-0.025
 title-0.024
vation-0.024
Token could
Token ID714
Feature activation-9.52e-02

Pos logit contributions
ug+0.028
issors+0.025
bed+0.025
usted+0.025
ors+0.024
Neg logit contributions
 relief-0.090
 gentleman-0.086
 lady-0.085
 service-0.084
 cal-0.084
Token have
Token ID423
Feature activation-1.79e-02

Pos logit contributions
ug+0.179
issors+0.170
ors+0.166
usted+0.166
bar+0.165
Neg logit contributions
 relief-0.134
 lady-0.131
 woman-0.123
 gentleman-0.123
 cal-0.122
Token it
Token ID340
Feature activation-1.22e-02

Pos logit contributions
ug+0.026
usted+0.026
bed+0.025
issors+0.025
bar+0.024
Neg logit contributions
 gentleman-0.030
 relief-0.029
 lady-0.028
 woman-0.028
 stranger-0.027
Token,
Token ID11
Feature activation-2.42e-02

Pos logit contributions
ug+0.018
bed+0.017
bar+0.017
issors+0.017
usted+0.017
Neg logit contributions
 relief-0.022
 gentleman-0.021
 woman-0.021
 stranger-0.020
 pastor-0.020
Token but
Token ID475
Feature activation-1.62e-04

Pos logit contributions
ug+0.041
issors+0.039
usted+0.038
ors+0.037
bar+0.037
Neg logit contributions
 relief-0.038
 lady-0.036
 gentleman-0.036
 woman-0.035
 cal-0.035
Token her
Token ID607
Feature activation+2.07e-02

Pos logit contributions
ug+0.000
issors+0.000
bed+0.000
usted+0.000
bar+0.000
Neg logit contributions
 relief-0.000
 gentleman-0.000
 cal-0.000
 stranger-0.000
 lady-0.000
Token that
Token ID326
Feature activation-2.06e-02

Pos logit contributions
ug+0.029
ors+0.027
bar+0.027
usted+0.027
bed+0.026
Neg logit contributions
 relief-0.040
 gentleman-0.036
 cal-0.035
 lady-0.035
 service-0.035
Token she
Token ID673
Feature activation-3.28e-02

Pos logit contributions
ug+0.031
bowl+0.028
bar+0.028
ors+0.028
bed+0.028
Neg logit contributions
 relief-0.036
 gentleman-0.035
 cal-0.032
 lady-0.032
 woman-0.031
Token thought
Token ID1807
Feature activation-4.14e-03

Pos logit contributions
ug+0.041
ors+0.037
bowl+0.037
bed+0.037
usted+0.037
Neg logit contributions
 relief-0.065
 gentleman-0.062
 lady-0.061
 service-0.060
 woman-0.059
Token the
Token ID262
Feature activation-2.32e-02

Pos logit contributions
ug+0.006
issors+0.005
ors+0.005
bar+0.005
usted+0.005
Neg logit contributions
 relief-0.008
 gentleman-0.008
 cal-0.007
 lady-0.007
 woman-0.007
Token pie
Token ID2508
Feature activation+3.70e-04

Pos logit contributions
issors+0.027
bar+0.027
bowl+0.025
usted+0.024
ug+0.023
Neg logit contributions
 relief-0.045
 gentleman-0.045
 cal-0.042
 woman-0.042
 stranger-0.040
Token might
Token ID1244
Feature activation-9.23e-02

Pos logit contributions
 relief+0.001
 gentleman+0.001
 cal+0.001
 woman+0.001
 pastor+0.001
Neg logit contributions
bowl-0.000
bar-0.000
ug-0.000
bed-0.000
issors-0.000
Token be
Token ID307
Feature activation+3.99e-03

Pos logit contributions
ug+0.143
issors+0.139
bar+0.138
bowl+0.130
bed+0.130
Neg logit contributions
 gentleman-0.151
 relief-0.148
 woman-0.143
 lady-0.140
 stranger-0.136
Token in
Token ID287
Feature activation-5.07e-02

Pos logit contributions
 gentleman+0.007
 woman+0.007
 relief+0.007
 lady+0.007
 stranger+0.006
Neg logit contributions
ug-0.006
issors-0.005
leys-0.005
 built-0.005
usted-0.005
Token there
Token ID612
Feature activation-6.21e-02

Pos logit contributions
ug+0.084
issors+0.082
bar+0.078
eds+0.077
bed+0.077
Neg logit contributions
 gentleman-0.075
 relief-0.071
 lady-0.069
 woman-0.069
 stranger-0.067
Token.
Token ID13
Feature activation-6.70e-03

Pos logit contributions
ug+0.099
issors+0.094
bar+0.092
bed+0.092
usted+0.091
Neg logit contributions
 relief-0.103
 gentleman-0.098
 lady-0.094
 woman-0.094
 cal-0.093
Token Josh
Token ID8518
Feature activation+3.59e-03

Pos logit contributions
ug+0.011
issors+0.011
usted+0.010
utes+0.010
bed+0.010
Neg logit contributions
 relief-0.011
 cal-0.010
 lady-0.010
 stranger-0.010
 gentleman-0.010
Token up
Token ID510
Feature activation+4.63e-02

Pos logit contributions
 lady+0.010
 gentleman+0.010
 woman+0.010
 relief+0.010
 stranger+0.010
Neg logit contributions
ug-0.007
issors-0.006
usted-0.006
ors-0.006
bar-0.006
Token to
Token ID284
Feature activation+1.49e-02

Pos logit contributions
 relief+0.094
 lady+0.092
 gentleman+0.090
 woman+0.090
 stranger+0.089
Neg logit contributions
ug-0.057
issors-0.054
usted-0.051
ors-0.051
bar-0.051
Token the
Token ID262
Feature activation+2.27e-02

Pos logit contributions
 relief+0.023
 gentleman+0.021
 cal+0.021
 lady+0.021
 woman+0.020
Neg logit contributions
ug-0.026
issors-0.025
utes-0.024
bed-0.023
ors-0.023
Token te
Token ID573
Feature activation-3.45e-02

Pos logit contributions
 relief+0.035
 cal+0.031
 Heidi+0.029
 gentleman+0.028
 prayer+0.027
Neg logit contributions
issors-0.038
bar-0.037
utes-0.037
bowl-0.036
ug-0.036
Tokenash
Token ID1077
Feature activation-2.89e-02

Pos logit contributions
ug+0.035
issors+0.032
usted+0.030
bar+0.030
ors+0.030
Neg logit contributions
 relief-0.078
 lady-0.076
 gentleman-0.075
 woman-0.074
 cal-0.074
Tokenop
Token ID404
Feature activation-8.36e-02

Pos logit contributions
ug+0.026
issors+0.022
bed+0.021
utes+0.021
ors+0.021
Neg logit contributions
 relief-0.073
 gentleman-0.068
 cal-0.068
 lady-0.067
 woman-0.067
Token and
Token ID290
Feature activation-1.19e-02

Pos logit contributions
ug+0.152
issors+0.149
utes+0.141
usted+0.141
bar+0.139
Neg logit contributions
 lady-0.115
 woman-0.111
 gentleman-0.108
 relief-0.107
 man-0.106
Token saw
Token ID2497
Feature activation+7.87e-03

Pos logit contributions
ug+0.018
bed+0.017
issors+0.017
bar+0.017
usted+0.017
Neg logit contributions
 relief-0.021
 gentleman-0.019
 cal-0.019
 lady-0.018
 stranger-0.018
Token a
Token ID257
Feature activation+4.93e-03

Pos logit contributions
 relief+0.012
 gentleman+0.011
 lady+0.011
 cal+0.011
 woman+0.010
Neg logit contributions
ug-0.014
bar-0.013
usted-0.013
bed-0.013
issors-0.013
Token nice
Token ID3621
Feature activation+6.59e-02

Pos logit contributions
 relief+0.008
 gentleman+0.007
 Heidi+0.007
 cal+0.007
 service+0.007
Neg logit contributions
utes-0.008
ug-0.007
om-0.007
bar-0.007
issors-0.007
Token lady
Token ID10846
Feature activation+3.36e-02

Pos logit contributions
 relief+0.105
 Heidi+0.095
 cal+0.094
 gentleman+0.088
 Lord+0.084
Neg logit contributions
om-0.107
ug-0.104
utes-0.101
bed-0.094
usted-0.094
Token hair
Token ID4190
Feature activation+2.18e-02

Pos logit contributions
issors+0.088
ug+0.086
usted+0.083
bed+0.083
bar+0.081
Neg logit contributions
 relief-0.112
 gentleman-0.110
 stranger-0.108
 cal-0.106
 Heidi-0.102
Token and
Token ID290
Feature activation+3.86e-04

Pos logit contributions
 relief+0.031
 gentleman+0.029
 lady+0.029
 woman+0.028
 stranger+0.028
Neg logit contributions
ug-0.041
issors-0.040
usted-0.038
bar-0.038
ors-0.038
Token you
Token ID345
Feature activation+5.31e-02

Pos logit contributions
 gentleman+0.001
 relief+0.001
 stranger+0.001
 cal+0.001
 Lord+0.001
Neg logit contributions
issors-0.001
bed-0.001
ug-0.001
usted-0.001
bar-0.001
Token wasted
Token ID18359
Feature activation-3.43e-02

Pos logit contributions
 relief+0.087
 lady+0.083
 gentleman+0.081
 woman+0.081
 cal+0.080
Neg logit contributions
ug-0.086
issors-0.082
utes-0.079
usted-0.078
ors-0.078
Token the
Token ID262
Feature activation-1.39e-02

Pos logit contributions
ug+0.052
usted+0.052
issors+0.051
bar+0.049
bed+0.049
Neg logit contributions
 gentleman-0.058
 woman-0.054
 relief-0.054
 stranger-0.053
 lady-0.052
Token leather
Token ID11620
Feature activation-9.13e-02

Pos logit contributions
issors+0.020
usted+0.019
bar+0.019
ug+0.019
om+0.018
Neg logit contributions
 gentleman-0.024
 relief-0.023
 Heidi-0.023
 stranger-0.022
 cal-0.022
Token!"
Token ID2474
Feature activation-6.19e-03

Pos logit contributions
ug+0.154
issors+0.150
bed+0.149
usted+0.148
bar+0.148
Neg logit contributions
 relief-0.148
 cal-0.135
 gentleman-0.132
 Heidi-0.128
 stranger-0.128
Token
Token ID198
Feature activation-1.82e-02

Pos logit contributions
issors+0.006
usted+0.006
om+0.006
ug+0.006
utes+0.005
Neg logit contributions
 relief-0.013
 gentleman-0.013
 traveler-0.012
 cal-0.012
 service-0.012
Token
Token ID198
Feature activation-4.61e-04

Pos logit contributions
ug+0.030
issors+0.029
usted+0.028
utes+0.028
bed+0.027
Neg logit contributions
 relief-0.029
 gentleman-0.028
 lady-0.027
 woman-0.027
 stranger-0.027
TokenTom
Token ID13787
Feature activation+2.09e-02

Pos logit contributions
issors+0.001
ug+0.001
usted+0.001
utes+0.001
eds+0.001
Neg logit contributions
 relief-0.001
 gentleman-0.001
 cal-0.001
 stranger-0.001
 prayer-0.001
Token knew
Token ID2993
Feature activation-1.09e-02

Pos logit contributions
 relief+0.033
 gentleman+0.030
 lady+0.030
 woman+0.030
 stranger+0.029
Neg logit contributions
ug-0.036
usted-0.035
issors-0.034
bed-0.034
bar-0.033
Token her
Token ID607
Feature activation-4.50e-04

Pos logit contributions
ug+0.062
issors+0.060
ors+0.058
bar+0.058
utes+0.058
Neg logit contributions
 relief-0.034
 lady-0.034
 woman-0.031
 gentleman-0.030
 cal-0.030
Token hand
Token ID1021
Feature activation-1.60e-02

Pos logit contributions
ug+0.001
bed+0.001
usted+0.001
issors+0.001
bar+0.001
Neg logit contributions
 gentleman-0.001
 relief-0.001
 service-0.001
 stranger-0.001
 woman-0.001
Token.
Token ID13
Feature activation+1.52e-03

Pos logit contributions
ug+0.026
usted+0.025
bed+0.025
bar+0.025
issors+0.025
Neg logit contributions
 relief-0.026
 gentleman-0.025
 stranger-0.024
 woman-0.024
 lady-0.023
Token It
Token ID632
Feature activation-9.30e-04

Pos logit contributions
 relief+0.002
 gentleman+0.002
 lady+0.002
 woman+0.002
 stranger+0.002
Neg logit contributions
ug-0.003
issors-0.003
usted-0.002
bed-0.002
utes-0.002
Token rub
Token ID6437
Feature activation-2.27e-02

Pos logit contributions
ug+0.002
bar+0.002
usted+0.002
bed+0.002
issors+0.001
Neg logit contributions
 relief-0.001
 gentleman-0.001
 woman-0.001
 lady-0.001
 stranger-0.001
Tokens
Token ID82
Feature activation-9.27e-02

Pos logit contributions
ug+0.026
issors+0.024
usted+0.023
bed+0.023
ors+0.023
Neg logit contributions
 relief-0.048
 gentleman-0.047
 lady-0.046
 woman-0.046
 stranger-0.045
Token its
Token ID663
Feature activation-3.86e-02

Pos logit contributions
ug+0.153
issors+0.147
usted+0.142
bed+0.140
bar+0.140
Neg logit contributions
 relief-0.150
 gentleman-0.144
 lady-0.143
 woman-0.141
 stranger-0.138
Token head
Token ID1182
Feature activation-1.28e-02

Pos logit contributions
ug+0.064
issors+0.061
usted+0.058
ors+0.058
bar+0.058
Neg logit contributions
 relief-0.063
 lady-0.060
 gentleman-0.059
 woman-0.059
 cal-0.058
Token against
Token ID1028
Feature activation-2.55e-02

Pos logit contributions
ug+0.023
issors+0.023
usted+0.022
ors+0.022
utes+0.022
Neg logit contributions
 relief-0.018
 lady-0.018
 gentleman-0.017
 woman-0.017
 cal-0.017
Token her
Token ID607
Feature activation-5.94e-03

Pos logit contributions
ug+0.047
issors+0.045
bed+0.044
usted+0.043
bar+0.043
Neg logit contributions
 relief-0.036
 gentleman-0.035
 lady-0.033
 cal-0.033
 woman-0.033
Token hand
Token ID1021
Feature activation-1.87e-02

Pos logit contributions
ug+0.009
issors+0.008
bed+0.008
bar+0.008
usted+0.008
Neg logit contributions
 relief-0.011
 gentleman-0.010
 lady-0.010
 cal-0.010
 stranger-0.010