TOP ACTIVATIONS
MAX = 6.101

, fifteen , sixteen , seventeen , eighteen , nineteen ,
thought they looked like ac rob ats . She enjoyed the
, fifteen , sixteen , seventeen , eighteen , nineteen ,
sund ae with lots of topp ings . They
flavors they liked and what topp ings would go best .
friends didn 't stop qu arre ling and were getting louder
said , " Stop qu arre ling , boys . You
. The bears stopped qu arre ling and asked him for
on , they never qu arre led again over clothes .
â t qu arre ling , the furniture stayed
together . They never qu arre led again . <|endoftext|> Once
He smells the y ummy topp ings . He wants to
he would take the av oc ados and scatter them around
lively sound . The qu arre ling stopped and they heard
That â s ok , you â
That â s ok . We all
and wearing gloves . The paramed ic was very polite and
together happily and never qu arre lled again . <|endoftext|> Once
both the friends stopped qu arre ling . They decided to
of cheese and sauce and topp ings . The people who

MIN ACTIVATIONS
MIN = -4.107

new things with his friend Ch ir py . <|endoftext|> Once
deep paws , to no avail . He finally thought of
times , but to no avail . Jane 's
again ?" Spot n uzz led Tom . He
out , but to no avail . He felt sad and
had a pet rabbit named Snow y . Lily loved to
ens and a rabbit named Ho ppy , getting ready for
met a kind bird named Ch ir py . Ch ir
him a lick and a n uzzle . She learned her
to Ben and Lily and n uzz led their hands .
yt ale creature called a n ymph ! The
, a little bird named Ch ir py , flew by
him . But to no avail , they were too scared
Where are you ?" their n anny shouted . She saw
clothes , like her sister L ila . One
Tom a lick and a n udge . Tom smiles and
some friends . She wished Snow y was alive . She
, a little chick named Ch ir py . They played
to play in the park every day . True is a
to play in the park every day . They swing ,

INTERVAL 3.549 - 6.101
CONTAINS 0.001%

Don â t be scared , I
That â s an elevator . It
on , they never qu arre led again and played together
thought they looked like ac rob ats . She enjoyed the
That â s the bird that was

INTERVAL 0.997 - 3.549
CONTAINS 7.024%

to go , Lucy and her Dad brought the cooler back
you are using the right numbers . If the numbers match
. <|endoftext|> One day , a little boy named Tim went
it look nice again . Paul was very proud
The man and his wife agreed and so

INTERVAL -1.555 - 0.997
CONTAINS 91.634%

their mom , who is in the kitchen .
Lucy was playing in her yard . She saw a small
big cave in the distance . He became curious and wanted
castle and told the knight that he needed to prepare better
. " Can we still play with the blocks

INTERVAL -4.107 - -1.555
CONTAINS 1.341%

girl trusted the sailor and felt safe with him . They
she wouldn 't scare it away . She saw the bird
them . She wished she could let them go home .
help . He found a way to clean the p uddle
it was very soft and fluffy . Every day
Token,
Token ID11
Feature activation+4.42e-01

Pos logit contributions
 subt+0.056
Deal+0.055
 math+0.055
 cream+0.054
 topp+0.054
Neg logit contributions
en-0.029
 dot-0.028
el-0.028
fly-0.027
ening-0.026
Token fifteen
Token ID17280
Feature activation+1.64e+00

Pos logit contributions
 subt+0.011
Deal+0.011
 cream+0.010
 math+0.010
 topp+0.010
Neg logit contributions
 dot-0.014
en-0.014
el-0.014
fly-0.013
ir-0.013
Token,
Token ID11
Feature activation+6.50e-01

Pos logit contributions
 subt+0.059
Deal+0.059
 math+0.056
 cream+0.056
Fif+0.056
Neg logit contributions
en-0.034
el-0.033
 dot-0.032
ening-0.030
le-0.030
Token sixteen
Token ID27913
Feature activation+2.53e+00

Pos logit contributions
 subt+0.017
Deal+0.017
 math+0.017
 cream+0.017
 topp+0.016
Neg logit contributions
 dot-0.019
en-0.019
el-0.019
fly-0.018
ir-0.018
Token,
Token ID11
Feature activation+8.50e-01

Pos logit contributions
Deal+0.100
 subt+0.099
 cream+0.096
 math+0.095
bek+0.094
Neg logit contributions
en-0.044
el-0.043
 dot-0.042
ening-0.038
ir-0.038
Token seventeen
Token ID38741
Feature activation+4.53e+00

Pos logit contributions
 subt+0.023
Deal+0.023
 math+0.023
 cream+0.022
 subtract+0.022
Neg logit contributions
 dot-0.025
en-0.025
el-0.024
fly-0.024
ir-0.023
Token,
Token ID11
Feature activation+8.06e-01

Pos logit contributions
 subt+0.199
bek+0.193
Deal+0.193
Fif+0.188
 charity+0.187
Neg logit contributions
en-0.063
el-0.063
inkle-0.052
le-0.051
ink-0.050
Token eighteen
Token ID29095
Feature activation+2.94e+00

Pos logit contributions
 subt+0.021
Deal+0.021
 math+0.021
 cream+0.020
 topp+0.020
Neg logit contributions
 dot-0.024
en-0.024
el-0.024
fly-0.023
ir-0.023
Token,
Token ID11
Feature activation+6.72e-01

Pos logit contributions
 subt+0.119
Deal+0.117
 math+0.113
 cream+0.113
Fif+0.112
Neg logit contributions
en-0.048
el-0.047
 dot-0.045
inkle-0.041
ir-0.041
Token nineteen
Token ID43063
Feature activation+1.11e+00

Pos logit contributions
 subt+0.017
Deal+0.017
 math+0.017
 cream+0.017
 topp+0.016
Neg logit contributions
 dot-0.021
en-0.020
el-0.020
fly-0.020
ir-0.019
Token,
Token ID11
Feature activation+5.82e-01

Pos logit contributions
 subt+0.045
Deal+0.043
 cream+0.042
 subtract+0.042
 math+0.042
Neg logit contributions
en-0.019
el-0.019
 dot-0.017
inkle-0.017
le-0.017
Token thought
Token ID1807
Feature activation+6.16e-01

Pos logit contributions
 subt+0.015
 subtract+0.015
 math+0.014
 cream+0.014
Deal+0.013
Neg logit contributions
ir-0.024
fly-0.023
py-0.023
 predators-0.023
ays-0.023
Token they
Token ID484
Feature activation+4.84e-01

Pos logit contributions
 math+0.020
 subt+0.016
 cakes+0.016
 cream+0.015
 homework+0.015
Neg logit contributions
fly-0.021
ley-0.020
 dot-0.019
an-0.018
ga-0.018
Token looked
Token ID3114
Feature activation+1.15e+00

Pos logit contributions
Deal+0.012
 topp+0.011
 y+0.011
 subtract+0.011
 cream+0.011
Neg logit contributions
ug-0.016
ir-0.016
ga-0.016
inkle-0.015
ays-0.015
Token like
Token ID588
Feature activation+7.77e-01

Pos logit contributions
 subt+0.030
 math+0.030
 cream+0.029
 topp+0.029
Deal+0.029
Neg logit contributions
en-0.035
 dot-0.035
el-0.034
fly-0.034
ug-0.034
Token ac
Token ID936
Feature activation+3.40e-01

Pos logit contributions
 math+0.018
 subt+0.017
 cream+0.017
 topp+0.016
 subtract+0.016
Neg logit contributions
en-0.027
el-0.027
 dot-0.027
fly-0.026
ir-0.026
Tokenrob
Token ID22609
Feature activation+4.32e+00

Pos logit contributions
 cream+0.008
Deal+0.007
 subt+0.007
 math+0.007
 topp+0.007
Neg logit contributions
 dot-0.012
en-0.012
fly-0.012
el-0.011
ug-0.011
Tokenats
Token ID1381
Feature activation+5.80e-02

Pos logit contributions
 subt+0.146
 math+0.141
 cream+0.140
Deal+0.139
 subtract+0.139
Neg logit contributions
en-0.099
el-0.098
ir-0.096
ug-0.094
le-0.094
Token.
Token ID13
Feature activation+2.45e-01

Pos logit contributions
 cream+0.002
 math+0.002
 cakes+0.002
 pizza+0.002
 cookies+0.002
Neg logit contributions
inkle-0.002
 dot-0.002
ug-0.002
ir-0.002
el-0.001
Token She
Token ID1375
Feature activation+5.01e-01

Pos logit contributions
 subt+0.008
 cream+0.008
 math+0.008
 topp+0.008
Deal+0.008
Neg logit contributions
en-0.006
 dot-0.006
el-0.006
fly-0.006
ug-0.005
Token enjoyed
Token ID8359
Feature activation-2.63e-01

Pos logit contributions
 subt+0.012
 cream+0.011
 subtract+0.011
 math+0.011
 topp+0.011
Neg logit contributions
ir-0.017
fly-0.017
ug-0.017
ays-0.017
 dot-0.017
Token the
Token ID262
Feature activation+5.45e-01

Pos logit contributions
en+0.008
 dot+0.008
el+0.008
fly+0.007
ug+0.007
Neg logit contributions
 subt-0.007
 math-0.007
 cream-0.007
Deal-0.007
 topp-0.007
Token,
Token ID11
Feature activation+3.70e-01

Pos logit contributions
 subt+0.048
 cream+0.047
 math+0.047
 topp+0.047
 subtract+0.046
Neg logit contributions
 dot-0.028
en-0.028
fly-0.028
ug-0.027
ga-0.026
Token fifteen
Token ID17280
Feature activation+1.30e+00

Pos logit contributions
 subt+0.009
 math+0.009
Deal+0.009
 cream+0.009
 topp+0.009
Neg logit contributions
en-0.012
 dot-0.012
el-0.012
fly-0.011
ug-0.011
Token,
Token ID11
Feature activation+5.09e-01

Pos logit contributions
 subt+0.046
 math+0.046
 cream+0.046
 topp+0.045
 subtract+0.044
Neg logit contributions
 dot-0.027
en-0.026
fly-0.026
ug-0.026
el-0.025
Token sixteen
Token ID27913
Feature activation+2.10e+00

Pos logit contributions
 subt+0.013
 math+0.013
 cream+0.013
Deal+0.013
 topp+0.013
Neg logit contributions
en-0.015
 dot-0.015
el-0.015
fly-0.015
ug-0.015
Token,
Token ID11
Feature activation+6.76e-01

Pos logit contributions
 subt+0.073
 math+0.073
 cream+0.073
 topp+0.072
 subtract+0.071
Neg logit contributions
fly-0.045
 dot-0.045
ug-0.044
ley-0.043
en-0.043
Token seventeen
Token ID38741
Feature activation+4.32e+00

Pos logit contributions
 subt+0.018
 math+0.017
 cream+0.017
Deal+0.017
 topp+0.017
Neg logit contributions
en-0.020
 dot-0.020
el-0.020
fly-0.020
ug-0.019
Token,
Token ID11
Feature activation+7.05e-01

Pos logit contributions
 subt+0.180
Deal+0.174
 math+0.173
 subtract+0.170
Fif+0.169
Neg logit contributions
en-0.068
el-0.066
 dot-0.060
le-0.059
ir-0.058
Token eighteen
Token ID29095
Feature activation+2.77e+00

Pos logit contributions
 subt+0.018
 math+0.018
 cream+0.017
Deal+0.017
 topp+0.017
Neg logit contributions
en-0.022
 dot-0.022
el-0.021
fly-0.021
ug-0.021
Token,
Token ID11
Feature activation+6.18e-01

Pos logit contributions
 subt+0.105
 math+0.103
 cream+0.103
Deal+0.102
 topp+0.102
Neg logit contributions
en-0.051
 dot-0.050
el-0.049
fly-0.048
ug-0.047
Token nineteen
Token ID43063
Feature activation+1.21e+00

Pos logit contributions
 subt+0.016
 math+0.015
Deal+0.015
 cream+0.015
 topp+0.015
Neg logit contributions
en-0.019
 dot-0.019
el-0.019
fly-0.019
ug-0.018
Token,
Token ID11
Feature activation+5.50e-01

Pos logit contributions
 subt+0.045
 math+0.044
 cream+0.044
Deal+0.044
 topp+0.044
Neg logit contributions
en-0.023
 dot-0.023
el-0.022
fly-0.022
ug-0.022
Token sund
Token ID37437
Feature activation-4.21e-01

Pos logit contributions
 subt+0.023
Deal+0.022
 subtract+0.021
 math+0.020
Fif+0.019
Neg logit contributions
 dot-0.026
el-0.024
 wil-0.024
fly-0.024
inkle-0.024
Tokenae
Token ID3609
Feature activation-6.19e-01

Pos logit contributions
le+0.013
ink+0.013
el+0.013
ie+0.012
ug+0.012
Neg logit contributions
 subt-0.013
 disl-0.012
 math-0.012
 subtract-0.012
Pro-0.011
Token with
Token ID351
Feature activation+6.53e-01

Pos logit contributions
en+0.014
 dot+0.014
el+0.014
fly+0.014
le+0.013
Neg logit contributions
 subt-0.021
 math-0.020
Deal-0.020
 cream-0.020
 topp-0.020
Token lots
Token ID6041
Feature activation-6.63e-02

Pos logit contributions
 cream+0.014
 subt+0.014
 math+0.014
 topp+0.014
 subtract+0.013
Neg logit contributions
en-0.023
fly-0.023
 dot-0.023
ug-0.023
le-0.022
Token of
Token ID286
Feature activation+1.03e+00

Pos logit contributions
fly+0.002
 dot+0.002
en+0.002
el+0.001
ug+0.001
Neg logit contributions
 cream-0.002
 subt-0.002
 math-0.002
 topp-0.002
 subtract-0.002
Token topp
Token ID23126
Feature activation+4.28e+00

Pos logit contributions
 subt+0.021
Deal+0.019
 math+0.019
 subtract+0.018
Fif+0.018
Neg logit contributions
el-0.038
en-0.037
le-0.036
ening-0.036
 dot-0.035
Tokenings
Token ID654
Feature activation-4.22e-01

Pos logit contributions
 cream+0.174
 topp+0.173
 subt+0.172
 math+0.172
Deal+0.169
Neg logit contributions
 dot-0.066
en-0.064
fly-0.062
ug-0.060
el-0.060
Token.
Token ID13
Feature activation-1.13e-01

Pos logit contributions
 dot+0.011
en+0.011
fly+0.011
el+0.010
ug+0.010
Neg logit contributions
 cream-0.013
 subt-0.013
 math-0.013
 topp-0.012
Deal-0.012
Token
Token ID198
Feature activation-1.76e-01

Pos logit contributions
en+0.003
 dot+0.003
el+0.003
fly+0.003
ug+0.003
Neg logit contributions
 subt-0.003
 cream-0.003
 math-0.003
 topp-0.003
Deal-0.003
Token
Token ID198
Feature activation-2.89e-01

Pos logit contributions
fly+0.007
 dot+0.007
ening+0.007
en+0.007
el+0.007
Neg logit contributions
 math-0.003
 subt-0.003
 cream-0.003
 subtract-0.003
Deal-0.003
TokenThey
Token ID2990
Feature activation+2.43e-01

Pos logit contributions
en+0.008
 dot+0.008
ga+0.008
ug+0.008
ches+0.008
Neg logit contributions
 topp-0.008
 subt-0.008
 cream-0.007
Deal-0.007
 math-0.007
Token flavors
Token ID17361
Feature activation+7.07e-01

Pos logit contributions
 cream+0.046
 topp+0.044
 subt+0.044
 math+0.044
Deal+0.044
Neg logit contributions
 dot-0.029
en-0.029
fly-0.028
ug-0.028
el-0.028
Token they
Token ID484
Feature activation+2.44e-01

Pos logit contributions
 cream+0.023
 topp+0.022
Deal+0.021
 math+0.021
 subt+0.021
Neg logit contributions
 dot-0.018
en-0.018
fly-0.018
py-0.017
ir-0.017
Token liked
Token ID8288
Feature activation+2.73e-01

Pos logit contributions
 cream+0.007
 topp+0.007
Deal+0.007
 subt+0.007
 math+0.007
Neg logit contributions
 dot-0.007
en-0.007
ug-0.007
el-0.007
fly-0.007
Token and
Token ID290
Feature activation+3.30e-01

Pos logit contributions
 cream+0.008
 math+0.008
 topp+0.008
 subt+0.008
 cakes+0.008
Neg logit contributions
 dot-0.008
en-0.007
fly-0.007
ug-0.007
el-0.007
Token what
Token ID644
Feature activation+7.14e-01

Pos logit contributions
 subt+0.009
 math+0.009
 cream+0.009
Deal+0.009
 topp+0.009
Neg logit contributions
en-0.010
 dot-0.010
el-0.009
fly-0.009
ug-0.009
Token topp
Token ID23126
Feature activation+4.23e+00

Pos logit contributions
 subt+0.022
Deal+0.021
 subtract+0.020
Fif+0.020
 cream+0.020
Neg logit contributions
el-0.018
en-0.018
 dot-0.017
fly-0.017
 wil-0.017
Tokenings
Token ID654
Feature activation-3.16e-01

Pos logit contributions
 cream+0.146
 topp+0.146
 math+0.142
 subt+0.142
Deal+0.140
Neg logit contributions
 dot-0.094
en-0.090
fly-0.089
ug-0.087
el-0.086
Token would
Token ID561
Feature activation-1.62e+00

Pos logit contributions
 dot+0.009
en+0.009
fly+0.009
ug+0.009
el+0.009
Neg logit contributions
 cream-0.009
 math-0.009
 subt-0.009
 topp-0.009
Deal-0.008
Token go
Token ID467
Feature activation+2.78e-01

Pos logit contributions
 dot+0.054
en+0.054
el+0.053
fly+0.053
ug+0.052
Neg logit contributions
 subt-0.037
 cream-0.037
 math-0.037
 topp-0.036
Deal-0.036
Token best
Token ID1266
Feature activation-7.10e-01

Pos logit contributions
 cream+0.008
 math+0.008
 subt+0.008
 topp+0.007
Deal+0.007
Neg logit contributions
 dot-0.008
fly-0.008
py-0.007
ug-0.007
en-0.007
Token.
Token ID13
Feature activation+3.64e-01

Pos logit contributions
en+0.019
el+0.018
 dot+0.018
ug+0.018
le+0.018
Neg logit contributions
 subt-0.022
Deal-0.021
 math-0.021
 subtract-0.020
 topp-0.020
Token friends
Token ID2460
Feature activation+7.31e-01

Pos logit contributions
 cream+0.006
 math+0.006
 subt+0.006
 cakes+0.006
 topp+0.006
Neg logit contributions
fly-0.015
 dot-0.015
en-0.015
ga-0.014
inkle-0.014
Token didn
Token ID1422
Feature activation-6.86e-01

Pos logit contributions
 subt+0.017
 cream+0.017
Deal+0.016
 subtract+0.016
 math+0.016
Neg logit contributions
 dot-0.024
inkle-0.023
ays-0.023
fly-0.023
unks-0.022
Token't
Token ID470
Feature activation+7.61e-01

Pos logit contributions
 dot+0.015
en+0.015
fly+0.015
ley+0.015
el+0.015
Neg logit contributions
 math-0.024
 subt-0.023
 cream-0.023
 subtract-0.022
 topp-0.022
Token stop
Token ID2245
Feature activation+1.02e+00

Pos logit contributions
 subt+0.024
 topp+0.024
 math+0.023
Deal+0.023
 cream+0.023
Neg logit contributions
en-0.019
 dot-0.019
le-0.018
fly-0.018
ga-0.018
Token qu
Token ID627
Feature activation+2.42e+00

Pos logit contributions
 math+0.025
 subt+0.025
 subtract+0.024
 cream+0.024
 topp+0.023
Neg logit contributions
ches-0.031
fly-0.030
 dot-0.030
ley-0.030
ug-0.030
Tokenarre
Token ID9624
Feature activation+4.21e+00

Pos logit contributions
 subt+0.033
 math+0.032
 cream+0.031
 topp+0.031
 subtract+0.030
Neg logit contributions
 dot-0.103
en-0.102
fly-0.102
el-0.101
ug-0.099
Tokenling
Token ID1359
Feature activation+5.29e-01

Pos logit contributions
 subt+0.173
 math+0.171
Deal+0.169
 cream+0.169
 topp+0.168
Neg logit contributions
en-0.068
 dot-0.064
el-0.062
le-0.061
ug-0.061
Token and
Token ID290
Feature activation-1.21e-01

Pos logit contributions
 math+0.016
 cream+0.016
 subt+0.015
 cakes+0.015
 topp+0.015
Neg logit contributions
 dot-0.014
inkle-0.013
el-0.013
fly-0.013
ley-0.013
Token were
Token ID547
Feature activation+7.97e-01

Pos logit contributions
 dot+0.004
inkle+0.004
el+0.004
ug+0.003
ir+0.003
Neg logit contributions
 subt-0.003
 cream-0.003
 subtract-0.003
 math-0.003
Deal-0.003
Token getting
Token ID1972
Feature activation+3.30e-01

Pos logit contributions
 cream+0.021
 math+0.021
 subt+0.020
Deal+0.019
 subtract+0.019
Neg logit contributions
 dot-0.025
el-0.024
en-0.024
ley-0.024
fly-0.023
Token louder
Token ID27089
Feature activation+1.08e+00

Pos logit contributions
 subt+0.009
 cream+0.009
 math+0.009
 topp+0.009
 subtract+0.009
Neg logit contributions
en-0.009
 dot-0.009
fly-0.009
el-0.009
ug-0.009
Token said
Token ID531
Feature activation-7.55e-01

Pos logit contributions
 dot+0.023
en+0.023
el+0.022
fly+0.022
ug+0.022
Neg logit contributions
 subt-0.023
 cream-0.022
 math-0.022
 topp-0.022
Deal-0.022
Token,
Token ID11
Feature activation+4.00e-01

Pos logit contributions
 dot+0.015
en+0.015
el+0.015
fly+0.015
ug+0.014
Neg logit contributions
 subt-0.027
 math-0.027
 cream-0.027
 topp-0.027
Deal-0.026
Token "
Token ID366
Feature activation-3.65e-02

Pos logit contributions
 subt+0.011
Deal+0.011
 topp+0.011
 math+0.011
 cream+0.011
Neg logit contributions
 dot-0.011
en-0.011
el-0.011
fly-0.011
ga-0.011
TokenStop
Token ID19485
Feature activation+1.39e+00

Pos logit contributions
en+0.001
el+0.001
 dot+0.001
ga+0.001
ir+0.001
Neg logit contributions
 math-0.001
 subt-0.001
 topp-0.001
 cream-0.001
Deal-0.001
Token qu
Token ID627
Feature activation+2.34e+00

Pos logit contributions
 math+0.035
 subt+0.034
 subtract+0.034
 cream+0.034
 topp+0.033
Neg logit contributions
el-0.043
fly-0.043
ley-0.043
ir-0.043
 dot-0.043
Tokenarre
Token ID9624
Feature activation+3.97e+00

Pos logit contributions
 subt+0.037
 math+0.035
Deal+0.035
 cream+0.035
 topp+0.034
Neg logit contributions
en-0.097
el-0.095
 dot-0.095
ug-0.094
ir-0.093
Tokenling
Token ID1359
Feature activation+5.90e-01

Pos logit contributions
 subt+0.172
 math+0.169
Deal+0.169
 cream+0.167
 topp+0.167
Neg logit contributions
en-0.057
el-0.051
 dot-0.051
le-0.050
ug-0.050
Token,
Token ID11
Feature activation+7.44e-01

Pos logit contributions
 math+0.019
 cream+0.019
 subt+0.018
 job+0.018
 topp+0.018
Neg logit contributions
 dot-0.014
fly-0.014
 wil-0.014
inkle-0.013
el-0.013
Token boys
Token ID6510
Feature activation+1.22e+00

Pos logit contributions
 subt+0.020
 math+0.020
Deal+0.019
 topp+0.019
 cream+0.019
Neg logit contributions
en-0.022
 dot-0.022
el-0.022
ug-0.021
fly-0.021
Token.
Token ID13
Feature activation+4.65e-01

Pos logit contributions
 math+0.034
 cream+0.034
Deal+0.033
 subt+0.032
 subtract+0.032
Neg logit contributions
 dot-0.035
fly-0.034
en-0.033
py-0.033
el-0.033
Token You
Token ID921
Feature activation-8.13e-02

Pos logit contributions
 subt+0.013
 math+0.013
 cream+0.013
 topp+0.013
Deal+0.013
Neg logit contributions
en-0.013
 dot-0.013
el-0.013
fly-0.013
le-0.012
Token.
Token ID13
Feature activation+2.61e-01

Pos logit contributions
 dot+0.018
ir+0.018
ug+0.018
py+0.018
en+0.018
Neg logit contributions
 cream-0.018
 topp-0.016
 math-0.016
 cakes-0.015
 pizza-0.015
Token The
Token ID383
Feature activation+8.90e-02

Pos logit contributions
 subt+0.007
 math+0.007
 cream+0.007
Deal+0.007
 topp+0.007
Neg logit contributions
en-0.008
 dot-0.008
el-0.008
fly-0.008
ug-0.007
Token bears
Token ID13062
Feature activation+1.29e+00

Pos logit contributions
 subt+0.002
 math+0.002
 cream+0.002
Deal+0.002
 topp+0.002
Neg logit contributions
 dot-0.003
en-0.003
el-0.003
fly-0.003
ug-0.003
Token stopped
Token ID5025
Feature activation+6.42e-01

Pos logit contributions
 cream+0.033
 topp+0.033
Deal+0.032
 subtract+0.031
 subt+0.031
Neg logit contributions
 dot-0.038
en-0.037
ays-0.037
py-0.037
ug-0.037
Token qu
Token ID627
Feature activation+2.02e+00

Pos logit contributions
 math+0.018
 subt+0.018
 cream+0.018
 subtract+0.018
 topp+0.017
Neg logit contributions
ches-0.017
 dot-0.017
le-0.017
ley-0.017
ug-0.017
Tokenarre
Token ID9624
Feature activation+3.95e+00

Pos logit contributions
 subt+0.048
 cream+0.048
 topp+0.047
 math+0.047
 subtract+0.046
Neg logit contributions
 dot-0.064
fly-0.063
en-0.063
ug-0.061
el-0.061
Tokenling
Token ID1359
Feature activation+3.48e-01

Pos logit contributions
 subt+0.175
 math+0.173
 cream+0.171
Deal+0.171
 topp+0.170
Neg logit contributions
en-0.050
 dot-0.047
el-0.046
fly-0.045
ug-0.043
Token and
Token ID290
Feature activation+3.05e-01

Pos logit contributions
 cream+0.013
 math+0.013
 subt+0.012
 topp+0.012
 pizza+0.012
Neg logit contributions
 dot-0.007
 wil-0.006
ug-0.006
inkle-0.006
fly-0.006
Token asked
Token ID1965
Feature activation+1.84e-01

Pos logit contributions
 subt+0.009
 math+0.009
 cream+0.009
 topp+0.009
Deal+0.009
Neg logit contributions
en-0.008
 dot-0.008
el-0.008
fly-0.008
ug-0.008
Token him
Token ID683
Feature activation+1.37e-01

Pos logit contributions
 math+0.005
 subt+0.005
 cream+0.005
Deal+0.005
 topp+0.005
Neg logit contributions
 dot-0.005
en-0.005
el-0.005
fly-0.005
 wil-0.005
Token for
Token ID329
Feature activation+1.25e-01

Pos logit contributions
 subt+0.004
 cream+0.004
 math+0.004
Deal+0.004
 topp+0.004
Neg logit contributions
en-0.004
 dot-0.004
fly-0.004
el-0.004
ug-0.004
Token on
Token ID319
Feature activation+4.13e-01

Pos logit contributions
 subt+0.020
 math+0.020
Deal+0.020
 cream+0.020
 topp+0.019
Neg logit contributions
en-0.019
 dot-0.019
fly-0.018
el-0.018
 wil-0.018
Token,
Token ID11
Feature activation+2.31e-01

Pos logit contributions
 math+0.012
 job+0.011
 subt+0.011
 cream+0.010
 pizza+0.010
Neg logit contributions
 dot-0.014
fly-0.013
ga-0.013
 wil-0.012
ala-0.012
Token they
Token ID484
Feature activation+6.41e-01

Pos logit contributions
 subt+0.006
 cream+0.006
Deal+0.006
 topp+0.006
 math+0.006
Neg logit contributions
en-0.007
 dot-0.007
el-0.006
fly-0.006
ug-0.006
Token never
Token ID1239
Feature activation-1.48e-01

Pos logit contributions
Deal+0.018
 subt+0.018
 math+0.017
 cream+0.017
 topp+0.016
Neg logit contributions
 dot-0.018
ening-0.018
en-0.017
fly-0.017
el-0.017
Token qu
Token ID627
Feature activation+1.81e+00

Pos logit contributions
en+0.004
 dot+0.004
fly+0.004
el+0.004
ening+0.004
Neg logit contributions
 subt-0.004
 math-0.004
 cream-0.004
Deal-0.004
 topp-0.004
Tokenarre
Token ID9624
Feature activation+3.95e+00

Pos logit contributions
 subt+0.020
 math+0.020
 topp+0.019
 cream+0.019
 subtract+0.018
Neg logit contributions
 dot-0.081
fly-0.081
en-0.080
ley-0.079
el-0.079
Tokenled
Token ID992
Feature activation+5.92e-01

Pos logit contributions
 subt+0.162
 math+0.159
 cream+0.158
Deal+0.158
 topp+0.157
Neg logit contributions
en-0.062
 dot-0.060
el-0.058
fly-0.057
ug-0.056
Token again
Token ID757
Feature activation+3.05e-01

Pos logit contributions
 math+0.018
 cream+0.016
 subt+0.016
Deal+0.016
 topp+0.015
Neg logit contributions
 dot-0.016
el-0.016
ley-0.016
ug-0.016
fly-0.016
Token over
Token ID625
Feature activation+1.65e-01

Pos logit contributions
 subt+0.011
 math+0.011
 cream+0.011
 topp+0.010
Deal+0.010
Neg logit contributions
 dot-0.006
en-0.006
el-0.006
fly-0.006
ug-0.006
Token clothes
Token ID8242
Feature activation+1.53e+00

Pos logit contributions
 math+0.004
 subt+0.004
 cream+0.004
 topp+0.004
 subtract+0.004
Neg logit contributions
 dot-0.005
ches-0.005
fly-0.005
le-0.005
ley-0.005
Token.
Token ID13
Feature activation-4.54e-02

Pos logit contributions
 cream+0.051
 math+0.050
 subt+0.049
Deal+0.048
 topp+0.048
Neg logit contributions
 dot-0.034
py-0.034
el-0.033
fly-0.033
ug-0.033
Tokenâ
Token ID22940
Feature activation+3.33e-01

Pos logit contributions
 math+0.006
 subt+0.005
 cream+0.005
 subtract+0.005
Fif+0.005
Neg logit contributions
fly-0.005
en-0.005
 dot-0.005
ley-0.005
ening-0.005
Token
Token ID26391
Feature activation+8.08e-01

Pos logit contributions
 math+0.008
 topp+0.008
 subt+0.008
 cream+0.008
 cakes+0.008
Neg logit contributions
en-0.011
el-0.010
le-0.010
ga-0.010
ug-0.010
Token
Token ID8151
Feature activation+2.73e+00

Pos logit contributions
 subt+0.022
 math+0.021
 cream+0.021
Deal+0.020
 subtract+0.020
Neg logit contributions
en-0.024
 dot-0.024
el-0.024
fly-0.024
ening-0.023
Tokent
Token ID83
Feature activation+7.89e-01

Pos logit contributions
 cream+0.076
 subt+0.075
 math+0.075
 topp+0.072
Fif+0.072
Neg logit contributions
 dot-0.077
en-0.077
fly-0.073
ening-0.072
el-0.070
Token qu
Token ID627
Feature activation+2.19e+00

Pos logit contributions
 subt+0.021
 cream+0.020
 math+0.020
 subtract+0.020
 topp+0.020
Neg logit contributions
el-0.023
 dot-0.023
en-0.023
fly-0.023
ug-0.022
Tokenarre
Token ID9624
Feature activation+3.94e+00

Pos logit contributions
 math+0.024
 subt+0.024
 cream+0.023
 subtract+0.023
 topp+0.023
Neg logit contributions
 dot-0.096
fly-0.095
ley-0.094
en-0.093
ga-0.093
Tokenling
Token ID1359
Feature activation+1.69e-01

Pos logit contributions
 cream+0.149
 subt+0.147
 topp+0.146
 math+0.145
 subtract+0.142
Neg logit contributions
 dot-0.072
fly-0.069
el-0.067
ir-0.065
en-0.064
Token,
Token ID11
Feature activation+4.23e-01

Pos logit contributions
 math+0.005
 cream+0.004
 job+0.004
 cakes+0.004
 subt+0.004
Neg logit contributions
 dot-0.005
fly-0.005
el-0.005
ley-0.005
inkle-0.005
Token the
Token ID262
Feature activation+4.19e-01

Pos logit contributions
 cream+0.011
 subt+0.010
 math+0.010
 cakes+0.010
 subtract+0.010
Neg logit contributions
fly-0.014
py-0.013
 dot-0.013
en-0.013
el-0.013
Token furniture
Token ID13091
Feature activation+2.04e-01

Pos logit contributions
 subt+0.011
 math+0.011
 cream+0.011
Deal+0.011
 topp+0.011
Neg logit contributions
en-0.013
 dot-0.012
el-0.012
fly-0.012
ug-0.012
Token stayed
Token ID9658
Feature activation+6.05e-02

Pos logit contributions
 cream+0.006
 math+0.006
 cakes+0.006
 subt+0.005
 topp+0.005
Neg logit contributions
 dot-0.006
en-0.006
ir-0.006
el-0.005
ga-0.005
Token together
Token ID1978
Feature activation-8.41e-01

Pos logit contributions
en+0.037
 dot+0.037
el+0.036
fly+0.036
ug+0.035
Neg logit contributions
 subt-0.031
 math-0.030
 cream-0.030
 topp-0.030
Deal-0.030
Token.
Token ID13
Feature activation+1.92e-01

Pos logit contributions
 dot+0.022
el+0.022
en+0.022
fly+0.022
ug+0.021
Neg logit contributions
 math-0.025
 subt-0.025
 cream-0.024
 topp-0.024
Deal-0.024
Token They
Token ID1119
Feature activation+1.76e-01

Pos logit contributions
 cream+0.006
 subt+0.006
 math+0.006
Deal+0.006
 topp+0.006
Neg logit contributions
en-0.005
 dot-0.005
el-0.005
fly-0.005
ug-0.005
Token never
Token ID1239
Feature activation-4.32e-01

Pos logit contributions
 subt+0.005
Deal+0.005
 math+0.005
 cream+0.005
 subtract+0.005
Neg logit contributions
 dot-0.005
en-0.005
el-0.005
fly-0.005
inkle-0.005
Token qu
Token ID627
Feature activation+1.63e+00

Pos logit contributions
en+0.011
 dot+0.011
fly+0.011
el+0.010
le+0.010
Neg logit contributions
 cream-0.013
 subt-0.013
 math-0.013
Deal-0.013
 topp-0.013
Tokenarre
Token ID9624
Feature activation+3.94e+00

Pos logit contributions
 subt+0.017
 math+0.017
 topp+0.017
 cream+0.016
 subtract+0.016
Neg logit contributions
 dot-0.074
fly-0.073
ley-0.072
en-0.072
el-0.072
Tokenled
Token ID992
Feature activation+7.72e-01

Pos logit contributions
 subt+0.156
 math+0.154
 cream+0.153
 topp+0.152
Deal+0.152
Neg logit contributions
en-0.066
 dot-0.066
el-0.064
fly-0.063
ug-0.061
Token again
Token ID757
Feature activation+2.83e-01

Pos logit contributions
 math+0.023
 cream+0.022
 topp+0.021
 pizza+0.021
 subt+0.020
Neg logit contributions
el-0.021
 dot-0.021
ug-0.021
ley-0.021
inkle-0.021
Token.
Token ID13
Feature activation+3.49e-01

Pos logit contributions
 subt+0.011
 math+0.011
 cream+0.010
 topp+0.010
 subtract+0.010
Neg logit contributions
 dot-0.005
en-0.005
el-0.005
fly-0.005
ug-0.005
Token<|endoftext|>
Token ID50256
Feature activation+3.75e-01

Pos logit contributions
 cream+0.010
 subt+0.010
 math+0.010
 topp+0.010
Deal+0.010
Neg logit contributions
en-0.010
 dot-0.009
el-0.009
fly-0.009
ug-0.009
TokenOnce
Token ID7454
Feature activation+1.60e-01

Pos logit contributions
 math+0.007
 subt+0.007
 cream+0.007
 topp+0.007
 las+0.007
Neg logit contributions
 wil-0.014
 dot-0.014
ening-0.014
en-0.014
ug-0.013
Token He
Token ID679
Feature activation+4.85e-01

Pos logit contributions
 subt+0.007
 cream+0.007
Deal+0.007
 topp+0.007
 subtract+0.007
Neg logit contributions
 dot-0.006
en-0.006
el-0.006
ug-0.005
ays-0.005
Token smells
Token ID25760
Feature activation+5.43e-01

Pos logit contributions
 topp+0.014
 math+0.014
 subt+0.013
 cream+0.013
 las+0.013
Neg logit contributions
ug-0.014
ir-0.014
 dot-0.014
fly-0.013
el-0.013
Token the
Token ID262
Feature activation+3.02e-01

Pos logit contributions
 subt+0.014
 math+0.014
 cream+0.014
 topp+0.014
Deal+0.014
Neg logit contributions
en-0.016
 dot-0.016
el-0.016
fly-0.016
ug-0.016
Token y
Token ID331
Feature activation+1.51e+00

Pos logit contributions
 subt+0.006
Deal+0.005
 subtract+0.005
 math+0.005
bek+0.005
Neg logit contributions
 dot-0.012
el-0.011
fly-0.011
en-0.011
le-0.011
Tokenummy
Token ID13513
Feature activation+1.16e+00

Pos logit contributions
 subt+0.054
 math+0.052
Deal+0.052
 subtract+0.051
 cream+0.050
Neg logit contributions
en-0.034
el-0.034
 dot-0.033
le-0.031
ening-0.031
Token topp
Token ID23126
Feature activation+3.94e+00

Pos logit contributions
 subt+0.023
Deal+0.021
 subtract+0.020
 math+0.019
Fif+0.019
Neg logit contributions
le-0.046
en-0.046
el-0.045
 dot-0.045
ening-0.044
Tokenings
Token ID654
Feature activation-4.24e-01

Pos logit contributions
 subt+0.173
 math+0.168
Deal+0.167
 subtract+0.165
 cream+0.163
Neg logit contributions
en-0.054
el-0.053
 dot-0.051
fly-0.049
le-0.048
Token.
Token ID13
Feature activation+1.37e-01

Pos logit contributions
 dot+0.011
en+0.010
fly+0.010
el+0.010
ug+0.010
Neg logit contributions
 cream-0.013
 math-0.013
 subt-0.013
 topp-0.013
Deal-0.013
Token He
Token ID679
Feature activation+4.38e-01

Pos logit contributions
 subt+0.004
Deal+0.004
 cream+0.004
 topp+0.004
 subtract+0.004
Neg logit contributions
 dot-0.003
en-0.003
el-0.003
ays-0.003
ug-0.003
Token wants
Token ID3382
Feature activation-2.35e-01

Pos logit contributions
 topp+0.013
 cream+0.012
 math+0.012
 subt+0.012
 las+0.012
Neg logit contributions
ug-0.013
 dot-0.012
ir-0.012
fly-0.012
el-0.012
Token to
Token ID284
Feature activation+6.60e-01

Pos logit contributions
en+0.007
 dot+0.007
el+0.007
fly+0.007
le+0.007
Neg logit contributions
 subt-0.006
Deal-0.006
 math-0.006
 cream-0.006
 topp-0.006
Token he
Token ID339
Feature activation-4.35e-01

Pos logit contributions
 math+0.007
 cream+0.007
 subt+0.007
 topp+0.007
Deal+0.007
Neg logit contributions
 dot-0.011
en-0.011
ir-0.011
ug-0.011
ening-0.011
Token would
Token ID561
Feature activation-3.37e-01

Pos logit contributions
 dot+0.014
ir+0.013
ays+0.013
ug+0.013
el+0.013
Neg logit contributions
Deal-0.011
 math-0.011
 cream-0.011
 topp-0.011
 subt-0.010
Token take
Token ID1011
Feature activation+9.38e-01

Pos logit contributions
 dot+0.012
ays+0.012
ug+0.012
el+0.012
ening+0.011
Neg logit contributions
 math-0.007
 cream-0.007
 subt-0.007
 subtract-0.007
 topp-0.007
Token the
Token ID262
Feature activation+3.12e-01

Pos logit contributions
 subt+0.024
 math+0.024
 cream+0.023
 topp+0.023
 subtract+0.022
Neg logit contributions
 dot-0.029
en-0.029
fly-0.029
el-0.028
ug-0.028
Token av
Token ID1196
Feature activation+3.77e+00

Pos logit contributions
 subt+0.008
 math+0.008
 cream+0.008
 topp+0.008
Deal+0.008
Neg logit contributions
en-0.010
 dot-0.009
el-0.009
fly-0.009
ug-0.009
Tokenoc
Token ID420
Feature activation+3.93e+00

Pos logit contributions
 subt+0.117
 math+0.117
Deal+0.115
 cream+0.115
 cakes+0.112
Neg logit contributions
en-0.101
el-0.100
fly-0.094
le-0.093
ir-0.093
Tokenados
Token ID22484
Feature activation+1.65e+00

Pos logit contributions
Deal+0.110
 subt+0.110
 math+0.110
 cream+0.107
Fif+0.107
Neg logit contributions
el-0.117
en-0.111
fly-0.106
ors-0.105
ug-0.105
Token and
Token ID290
Feature activation+8.36e-01

Pos logit contributions
 subt+0.045
 math+0.044
 cream+0.044
Deal+0.043
 topp+0.043
Neg logit contributions
en-0.048
 dot-0.047
el-0.047
fly-0.046
ug-0.045
Token scatter
Token ID41058
Feature activation-3.80e-01

Pos logit contributions
Deal+0.021
 subt+0.020
 math+0.020
Fif+0.019
 cream+0.019
Neg logit contributions
en-0.026
fly-0.025
el-0.025
 dot-0.025
ley-0.024
Token them
Token ID606
Feature activation+8.06e-02

Pos logit contributions
 dot+0.008
fly+0.008
ug+0.008
ga+0.007
en+0.007
Neg logit contributions
 cream-0.014
 topp-0.014
 math-0.014
 subt-0.013
 cakes-0.013
Token around
Token ID1088
Feature activation-1.00e+00

Pos logit contributions
 cream+0.003
 math+0.003
 topp+0.002
 subt+0.002
Deal+0.002
Neg logit contributions
 dot-0.002
en-0.002
ug-0.002
el-0.002
ga-0.002
Token lively
Token ID29696
Feature activation+2.13e-01

Pos logit contributions
 subt+0.012
 math+0.012
 cream+0.012
 topp+0.012
Deal+0.012
Neg logit contributions
en-0.016
 dot-0.016
el-0.016
fly-0.015
ug-0.015
Token sound
Token ID2128
Feature activation-2.69e-01

Pos logit contributions
 math+0.005
 cream+0.005
 subt+0.005
 topp+0.005
 subtract+0.005
Neg logit contributions
el-0.007
fly-0.007
pt-0.006
en-0.006
ir-0.006
Token.
Token ID13
Feature activation+1.59e-01

Pos logit contributions
el+0.007
ley+0.007
fly+0.007
 dot+0.007
ir+0.007
Neg logit contributions
 cream-0.008
 math-0.008
 subt-0.008
 topp-0.007
 subtract-0.007
Token The
Token ID383
Feature activation+5.56e-01

Pos logit contributions
 subt+0.004
 math+0.004
 subtract+0.004
 topp+0.004
Deal+0.004
Neg logit contributions
en-0.005
 dot-0.005
le-0.005
el-0.005
fly-0.005
Token qu
Token ID627
Feature activation+1.41e+00

Pos logit contributions
 subt+0.016
 math+0.016
 cream+0.016
Deal+0.015
 topp+0.015
Neg logit contributions
en-0.015
 dot-0.015
el-0.015
fly-0.015
ug-0.015
Tokenarre
Token ID9624
Feature activation+3.91e+00

Pos logit contributions
 topp+0.039
 cream+0.039
 subtract+0.038
 subt+0.038
 math+0.037
Neg logit contributions
ley-0.038
fly-0.037
 dot-0.037
ala-0.037
 wil-0.037
Tokenling
Token ID1359
Feature activation+2.44e-01

Pos logit contributions
 cream+0.152
 topp+0.151
 subt+0.148
 math+0.146
Deal+0.144
Neg logit contributions
 dot-0.066
fly-0.063
 wil-0.061
el-0.061
ala-0.059
Token stopped
Token ID5025
Feature activation+8.66e-01

Pos logit contributions
 cream+0.007
 topp+0.006
 math+0.006
 subt+0.006
 job+0.006
Neg logit contributions
unks-0.007
ays-0.007
inkle-0.006
ir-0.006
 wil-0.006
Token and
Token ID290
Feature activation+3.83e-01

Pos logit contributions
 math+0.024
 subtract+0.024
 subt+0.024
 cream+0.023
 topp+0.022
Neg logit contributions
ches-0.024
ley-0.024
le-0.023
 wil-0.022
ays-0.022
Token they
Token ID484
Feature activation+2.82e-01

Pos logit contributions
 subt+0.011
 math+0.011
 cream+0.011
 topp+0.011
Deal+0.011
Neg logit contributions
en-0.010
 dot-0.010
el-0.010
fly-0.010
ug-0.010
Token heard
Token ID2982
Feature activation+6.98e-01

Pos logit contributions
 subt+0.007
 cream+0.007
Deal+0.007
 topp+0.007
 subtract+0.007
Neg logit contributions
el-0.008
en-0.008
 dot-0.008
ening-0.008
fly-0.008
Token
Token ID129
Feature activation+1.01e+00

Pos logit contributions
 subt+0.030
 math+0.029
 cream+0.029
 topp+0.029
Deal+0.029
Neg logit contributions
en-0.031
 dot-0.031
el-0.031
fly-0.030
ug-0.030
Token
Token ID241
Feature activation+9.60e-01

Pos logit contributions
 subt+0.028
 math+0.028
 cream+0.027
Deal+0.027
 subtract+0.027
Neg logit contributions
 dot-0.029
fly-0.029
en-0.028
el-0.028
ug-0.028
TokenThat
Token ID2504
Feature activation+1.03e+00

Pos logit contributions
 subt+0.024
 math+0.024
 cream+0.023
Deal+0.023
 topp+0.023
Neg logit contributions
en-0.030
 dot-0.030
el-0.029
fly-0.029
ug-0.029
Tokenâ
Token ID22940
Feature activation+5.88e-02

Pos logit contributions
 math+0.033
 subt+0.033
 cream+0.033
 topp+0.033
Deal+0.032
Neg logit contributions
 dot-0.025
en-0.024
ir-0.024
ug-0.024
el-0.024
Token
Token ID26391
Feature activation+8.28e-01

Pos logit contributions
 math+0.001
 topp+0.001
 subt+0.001
 cream+0.001
 cakes+0.001
Neg logit contributions
en-0.002
el-0.002
le-0.002
ug-0.002
 dot-0.002
Token
Token ID8151
Feature activation+3.89e+00

Pos logit contributions
 subt+0.023
 math+0.023
 cream+0.023
 topp+0.023
Deal+0.023
Neg logit contributions
en-0.023
 dot-0.023
el-0.023
fly-0.022
ug-0.022
Tokens
Token ID82
Feature activation+1.33e+00

Pos logit contributions
Deal+0.111
 cream+0.111
 subt+0.110
 math+0.109
 topp+0.109
Neg logit contributions
 dot-0.112
en-0.107
ening-0.103
fly-0.102
py-0.101
Token ok
Token ID12876
Feature activation+8.96e-01

Pos logit contributions
 math+0.041
Deal+0.040
 cream+0.040
 subt+0.040
 topp+0.039
Neg logit contributions
 dot-0.035
en-0.034
ir-0.033
fly-0.033
ug-0.033
Token,
Token ID11
Feature activation+1.07e+00

Pos logit contributions
 math+0.028
 cream+0.028
 subt+0.028
Deal+0.027
 topp+0.027
Neg logit contributions
 dot-0.023
en-0.022
 wil-0.022
ir-0.021
fly-0.021
Token you
Token ID345
Feature activation+9.45e-01

Pos logit contributions
 math+0.030
 subt+0.030
 cream+0.029
 topp+0.029
 subtract+0.028
Neg logit contributions
 dot-0.030
en-0.030
el-0.030
fly-0.030
ir-0.029
Tokenâ
Token ID22940
Feature activation+1.55e-01

Pos logit contributions
 job+0.021
Deal+0.020
 math+0.020
 pay+0.020
 list+0.020
Neg logit contributions
ening-0.030
 dot-0.030
ays-0.029
eds-0.029
ga-0.029
Token
Token ID129
Feature activation+7.85e-01

Pos logit contributions
 subt+0.029
 math+0.028
 cream+0.028
Deal+0.027
 topp+0.027
Neg logit contributions
en-0.031
 dot-0.031
el-0.031
fly-0.030
ug-0.030
Token
Token ID241
Feature activation+4.34e-01

Pos logit contributions
 subt+0.022
 math+0.021
 cream+0.021
Deal+0.021
 subtract+0.021
Neg logit contributions
 dot-0.023
fly-0.022
en-0.022
el-0.022
ug-0.021
TokenThat
Token ID2504
Feature activation+8.09e-01

Pos logit contributions
 subt+0.011
 math+0.010
 cream+0.010
Deal+0.010
 topp+0.010
Neg logit contributions
en-0.014
 dot-0.014
el-0.014
fly-0.014
ug-0.013
Tokenâ
Token ID22940
Feature activation-1.83e-02

Pos logit contributions
 math+0.024
 cream+0.024
 subt+0.024
 topp+0.024
Deal+0.023
Neg logit contributions
ir-0.022
 dot-0.021
ug-0.021
en-0.021
fly-0.021
Token
Token ID26391
Feature activation+8.75e-01

Pos logit contributions
en+0.001
el+0.001
le+0.001
ug+0.001
 dot+0.001
Neg logit contributions
 math-0.000
 subt-0.000
 topp-0.000
 cream-0.000
 cakes-0.000
Token
Token ID8151
Feature activation+3.89e+00

Pos logit contributions
 subt+0.025
 math+0.024
 cream+0.024
 topp+0.024
Deal+0.024
Neg logit contributions
en-0.024
 dot-0.024
el-0.024
fly-0.024
ug-0.023
Tokens
Token ID82
Feature activation+1.17e+00

Pos logit contributions
 cream+0.114
Deal+0.113
 math+0.112
 subt+0.111
 topp+0.111
Neg logit contributions
 dot-0.110
en-0.105
ening-0.101
fly-0.100
py-0.099
Token ok
Token ID12876
Feature activation+8.22e-01

Pos logit contributions
Deal+0.036
 math+0.036
 cream+0.035
 subt+0.035
 topp+0.034
Neg logit contributions
 dot-0.031
en-0.030
ir-0.029
fly-0.029
ug-0.029
Token.
Token ID13
Feature activation+3.60e-01

Pos logit contributions
 math+0.027
 subt+0.026
 cream+0.026
Deal+0.026
 topp+0.026
Neg logit contributions
 dot-0.020
en-0.019
 wil-0.019
fly-0.019
el-0.019
Token We
Token ID775
Feature activation+8.43e-01

Pos logit contributions
 subt+0.010
 math+0.010
 cream+0.010
Deal+0.010
 topp+0.010
Neg logit contributions
en-0.010
 dot-0.010
el-0.010
fly-0.010
ug-0.009
Token all
Token ID477
Feature activation-2.60e-01

Pos logit contributions
Deal+0.021
 cream+0.019
 subt+0.019
 job+0.019
 topp+0.019
Neg logit contributions
ening-0.025
 dot-0.025
en-0.024
ays-0.024
 predators-0.024
Token and
Token ID290
Feature activation-2.70e-01

Pos logit contributions
en+0.018
 dot+0.018
py+0.017
fly+0.017
ir+0.017
Neg logit contributions
 cream-0.019
 math-0.017
 topp-0.017
 pizza-0.017
 subtract-0.017
Token wearing
Token ID5762
Feature activation+2.20e-02

Pos logit contributions
en+0.008
 dot+0.008
el+0.008
fly+0.008
ug+0.008
Neg logit contributions
 subt-0.007
 math-0.007
Deal-0.007
 cream-0.006
 topp-0.006
Token gloves
Token ID18051
Feature activation-1.03e+00

Pos logit contributions
 subt+0.001
Deal+0.001
 math+0.001
 subtract+0.001
 cream+0.001
Neg logit contributions
en-0.001
el-0.001
 dot-0.001
le-0.001
ening-0.001
Token.
Token ID13
Feature activation+1.23e+00

Pos logit contributions
 dot+0.032
ir+0.031
inkle+0.031
el+0.031
en+0.031
Neg logit contributions
 cream-0.026
 math-0.024
 topp-0.024
 pizza-0.023
Deal-0.023
Token The
Token ID383
Feature activation+8.82e-01

Pos logit contributions
 cream+0.039
Deal+0.038
 subt+0.038
 math+0.037
 topp+0.036
Neg logit contributions
en-0.031
el-0.031
 dot-0.030
fly-0.029
ening-0.028
Token paramed
Token ID34581
Feature activation+3.87e+00

Pos logit contributions
 subt+0.024
 math+0.023
Deal+0.023
 topp+0.022
 subtract+0.022
Neg logit contributions
el-0.026
 wil-0.026
ga-0.025
 dot-0.025
ley-0.024
Tokenic
Token ID291
Feature activation+5.37e-01

Pos logit contributions
 math+0.157
 subt+0.145
 topp+0.145
 las+0.144
Fif+0.141
Neg logit contributions
en-0.080
el-0.078
ie-0.073
le-0.070
ening-0.070
Token was
Token ID373
Feature activation+2.46e-02

Pos logit contributions
 subt+0.017
 cream+0.017
 topp+0.017
 math+0.017
Deal+0.017
Neg logit contributions
 dot-0.013
en-0.013
fly-0.013
el-0.012
ug-0.012
Token very
Token ID845
Feature activation+6.51e-01

Pos logit contributions
 cream+0.001
 math+0.001
 subt+0.001
 topp+0.001
 subtract+0.001
Neg logit contributions
 dot-0.001
en-0.001
fly-0.001
ley-0.001
ir-0.001
Token polite
Token ID23507
Feature activation+1.07e+00

Pos logit contributions
 subt+0.016
 math+0.015
 subtract+0.015
Deal+0.015
 topp+0.015
Neg logit contributions
 wil-0.020
inkle-0.020
el-0.020
ug-0.020
en-0.020
Token and
Token ID290
Feature activation+6.86e-01

Pos logit contributions
 math+0.040
 cream+0.040
 subt+0.039
 topp+0.039
 subtract+0.038
Neg logit contributions
ir-0.021
 dot-0.020
inkle-0.020
el-0.020
fly-0.020
Token together
Token ID1978
Feature activation+2.91e-02

Pos logit contributions
en+0.007
ug+0.006
 dot+0.006
le+0.006
 freely+0.006
Neg logit contributions
 subt-0.009
 topp-0.009
 las-0.009
 subtract-0.009
Fif-0.008
Token happily
Token ID18177
Feature activation-2.70e-01

Pos logit contributions
 subt+0.001
 cream+0.001
 math+0.001
 topp+0.001
Deal+0.001
Neg logit contributions
en-0.001
 dot-0.001
el-0.001
fly-0.001
ug-0.001
Token and
Token ID290
Feature activation+1.40e-01

Pos logit contributions
en+0.005
 dot+0.005
fly+0.005
ug+0.005
el+0.004
Neg logit contributions
 subt-0.010
 topp-0.010
 math-0.010
 cream-0.010
 las-0.010
Token never
Token ID1239
Feature activation-3.06e-01

Pos logit contributions
 cream+0.005
 subt+0.005
 math+0.005
 topp+0.005
Deal+0.005
Neg logit contributions
en-0.003
 dot-0.003
fly-0.003
el-0.003
ga-0.003
Token qu
Token ID627
Feature activation+1.89e+00

Pos logit contributions
en+0.008
 dot+0.007
fly+0.007
pt+0.007
ening+0.007
Neg logit contributions
 math-0.010
Deal-0.010
 cream-0.010
 subt-0.010
 las-0.010
Tokenarre
Token ID9624
Feature activation+3.85e+00

Pos logit contributions
 subt+0.027
 math+0.026
 cream+0.026
 topp+0.025
Deal+0.025
Neg logit contributions
en-0.079
 dot-0.079
el-0.078
fly-0.078
ug-0.077
Tokenlled
Token ID3353
Feature activation+7.04e-01

Pos logit contributions
 subt+0.166
 math+0.163
Deal+0.162
 cream+0.160
 subtract+0.159
Neg logit contributions
en-0.057
 dot-0.051
le-0.051
ug-0.051
el-0.050
Token again
Token ID757
Feature activation+2.21e-01

Pos logit contributions
 math+0.024
 cream+0.024
 subt+0.024
 topp+0.024
Deal+0.024
Neg logit contributions
 dot-0.015
fly-0.015
el-0.015
en-0.015
ug-0.014
Token.
Token ID13
Feature activation+2.43e-01

Pos logit contributions
 subt+0.008
Deal+0.007
 cream+0.007
 math+0.007
 topp+0.007
Neg logit contributions
en-0.005
 dot-0.005
el-0.005
fly-0.005
le-0.005
Token<|endoftext|>
Token ID50256
Feature activation+2.94e-01

Pos logit contributions
 subt+0.007
 cream+0.007
 topp+0.007
 math+0.007
Deal+0.007
Neg logit contributions
en-0.006
 dot-0.006
el-0.006
fly-0.006
ug-0.006
TokenOnce
Token ID7454
Feature activation+1.49e-01

Pos logit contributions
 math+0.006
 cream+0.006
 subt+0.006
 topp+0.006
Deal+0.006
Neg logit contributions
 dot-0.010
en-0.010
fly-0.010
ug-0.010
ening-0.010
Token both
Token ID1111
Feature activation+1.14e-01

Pos logit contributions
en+0.002
 dot+0.002
el+0.002
ches+0.002
fly+0.002
Neg logit contributions
 math-0.002
 subt-0.002
 cream-0.002
 topp-0.002
Deal-0.002
Token the
Token ID262
Feature activation-4.57e-01

Pos logit contributions
 cream+0.003
 math+0.003
 subt+0.003
 topp+0.003
 cakes+0.003
Neg logit contributions
en-0.003
 dot-0.003
ga-0.003
el-0.003
ir-0.003
Token friends
Token ID2460
Feature activation+9.71e-01

Pos logit contributions
en+0.015
 dot+0.015
el+0.015
fly+0.015
ug+0.015
Neg logit contributions
 subt-0.010
 math-0.010
 cream-0.010
 topp-0.010
Deal-0.010
Token stopped
Token ID5025
Feature activation+4.60e-01

Pos logit contributions
 cream+0.021
 subt+0.021
 subtract+0.020
Deal+0.020
 cakes+0.019
Neg logit contributions
 dot-0.031
ays-0.031
py-0.030
unks-0.030
inkle-0.029
Token qu
Token ID627
Feature activation+1.79e+00

Pos logit contributions
 subt+0.013
 math+0.013
 subtract+0.012
 cream+0.012
 topp+0.012
Neg logit contributions
ches-0.013
ley-0.013
le-0.012
 dot-0.012
ug-0.012
Tokenarre
Token ID9624
Feature activation+3.84e+00

Pos logit contributions
 subt+0.040
 cream+0.039
 subtract+0.039
 topp+0.039
 math+0.039
Neg logit contributions
fly-0.057
 dot-0.057
ley-0.057
 wil-0.055
ug-0.055
Tokenling
Token ID1359
Feature activation-6.57e-02

Pos logit contributions
 subt+0.163
 cream+0.162
 math+0.160
 topp+0.160
Deal+0.158
Neg logit contributions
 dot-0.052
en-0.050
el-0.049
fly-0.049
ug-0.047
Token.
Token ID13
Feature activation+1.56e-01

Pos logit contributions
 dot+0.002
unks+0.001
ug+0.001
inkle+0.001
ley+0.001
Neg logit contributions
 cream-0.002
 math-0.002
 subt-0.002
 topp-0.002
 pizza-0.002
Token They
Token ID1119
Feature activation+3.35e-01

Pos logit contributions
 subt+0.004
 math+0.004
 subtract+0.004
 topp+0.004
Deal+0.003
Neg logit contributions
le-0.005
en-0.005
ie-0.005
inkle-0.005
 dot-0.005
Token decided
Token ID3066
Feature activation-3.99e-01

Pos logit contributions
 cream+0.008
 subt+0.008
Deal+0.008
 subtract+0.007
 topp+0.007
Neg logit contributions
 dot-0.011
ays-0.010
en-0.010
ug-0.010
el-0.010
Token to
Token ID284
Feature activation+1.05e-01

Pos logit contributions
en+0.013
 dot+0.012
el+0.012
fly+0.012
ug+0.012
Neg logit contributions
 subt-0.010
 math-0.010
 cream-0.010
Deal-0.010
 topp-0.010
Token of
Token ID286
Feature activation+1.71e+00

Pos logit contributions
 dot+0.029
fly+0.028
en+0.028
el+0.027
ley+0.027
Neg logit contributions
 cream-0.037
 math-0.036
 topp-0.036
 subt-0.035
 subtract-0.034
Token cheese
Token ID9891
Feature activation-3.34e-02

Pos logit contributions
 subt+0.027
 cream+0.027
 math+0.027
 topp+0.026
Deal+0.025
Neg logit contributions
 dot-0.070
en-0.070
fly-0.069
el-0.069
ug-0.068
Token and
Token ID290
Feature activation-5.18e-01

Pos logit contributions
en+0.001
el+0.001
 dot+0.001
ir+0.001
ening+0.001
Neg logit contributions
 cream-0.001
 pizza-0.001
 topp-0.001
 math-0.001
 cakes-0.001
Token sauce
Token ID10746
Feature activation-5.96e-02

Pos logit contributions
en+0.018
 dot+0.018
el+0.018
fly+0.018
ug+0.017
Neg logit contributions
 subt-0.011
 math-0.010
Deal-0.010
 cream-0.010
 subtract-0.010
Token and
Token ID290
Feature activation-4.83e-01

Pos logit contributions
 dot+0.002
en+0.002
el+0.002
fly+0.002
ir+0.002
Neg logit contributions
 cream-0.002
 pizza-0.002
 topp-0.002
 math-0.002
 cakes-0.002
Token topp
Token ID23126
Feature activation+3.84e+00

Pos logit contributions
en+0.017
 dot+0.017
el+0.017
fly+0.017
ug+0.016
Neg logit contributions
 subt-0.010
 math-0.009
Deal-0.009
 cream-0.009
 topp-0.009
Tokenings
Token ID654
Feature activation-9.70e-01

Pos logit contributions
 cream+0.152
 topp+0.151
 math+0.144
 cakes+0.143
Deal+0.141
Neg logit contributions
 dot-0.070
fly-0.067
en-0.065
 wil-0.064
ga-0.062
Token.
Token ID13
Feature activation+7.82e-02

Pos logit contributions
 dot+0.026
en+0.025
fly+0.025
el+0.024
ir+0.024
Neg logit contributions
 cream-0.030
 topp-0.029
 math-0.029
 subt-0.028
 cakes-0.028
Token The
Token ID383
Feature activation-1.66e-01

Pos logit contributions
 subt+0.002
 cream+0.002
 math+0.002
Deal+0.002
 topp+0.002
Neg logit contributions
 dot-0.002
en-0.002
el-0.002
fly-0.002
ug-0.002
Token people
Token ID661
Feature activation+2.14e+00

Pos logit contributions
 dot+0.006
en+0.006
el+0.006
fly+0.006
ug+0.006
Neg logit contributions
 subt-0.004
 math-0.003
Deal-0.003
 cream-0.003
 subtract-0.003
Token who
Token ID508
Feature activation+1.61e+00

Pos logit contributions
 cream+0.058
 topp+0.056
 math+0.054
Deal+0.053
 cakes+0.052
Neg logit contributions
 dot-0.064
en-0.063
el-0.061
ug-0.061
py-0.061
Token new
Token ID649
Feature activation-5.82e-01

Pos logit contributions
en+0.051
 dot+0.050
el+0.049
fly+0.049
ug+0.048
Neg logit contributions
 subt-0.037
 math-0.036
Deal-0.036
 cream-0.036
 topp-0.036
Token things
Token ID1243
Feature activation-1.44e+00

Pos logit contributions
en+0.020
 dot+0.020
el+0.020
fly+0.020
ug+0.020
Neg logit contributions
 subt-0.013
 math-0.012
 cream-0.012
Deal-0.012
 topp-0.012
Token with
Token ID351
Feature activation-1.99e-01

Pos logit contributions
 dot+0.038
en+0.037
el+0.037
ug+0.036
fly+0.036
Neg logit contributions
 math-0.043
 cream-0.043
 subt-0.043
Deal-0.042
 topp-0.042
Token his
Token ID465
Feature activation-1.81e+00

Pos logit contributions
 dot+0.007
ga+0.006
fly+0.006
ug+0.006
ays+0.006
Neg logit contributions
 math-0.005
 cream-0.005
 topp-0.005
 pizza-0.005
 cakes-0.005
Token friend
Token ID1545
Feature activation-9.94e-01

Pos logit contributions
en+0.063
le+0.062
ug+0.061
 dot+0.061
ga+0.060
Neg logit contributions
 cream-0.040
 math-0.040
 topp-0.039
Deal-0.038
 subt-0.038
Token Ch
Token ID609
Feature activation-4.11e+00

Pos logit contributions
 dot+0.026
en+0.026
el+0.025
ir+0.025
ug+0.025
Neg logit contributions
 math-0.030
 subt-0.029
 cream-0.029
 topp-0.029
Deal-0.029
Tokenir
Token ID343
Feature activation-1.57e+00

Pos logit contributions
 dot+0.080
 lands+0.074
 landed+0.067
ley+0.061
ening+0.061
Neg logit contributions
 subtract-0.135
igsaw-0.134
 subt-0.132
 topp-0.130
Deal-0.129
Tokenpy
Token ID9078
Feature activation-1.09e+00

Pos logit contributions
 dot+0.020
ays+0.015
 wil+0.015
inkle+0.015
ink+0.014
Neg logit contributions
 math-0.067
 cream-0.066
 topp-0.066
 subtract-0.066
 cakes-0.064
Token.
Token ID13
Feature activation-6.47e-01

Pos logit contributions
en+0.027
 dot+0.027
el+0.027
fly+0.026
ug+0.026
Neg logit contributions
 subt-0.034
 math-0.034
 cream-0.033
Deal-0.033
 topp-0.033
Token<|endoftext|>
Token ID50256
Feature activation-2.39e-02

Pos logit contributions
en+0.020
 dot+0.020
el+0.019
ening+0.019
le+0.019
Neg logit contributions
 subt-0.016
 math-0.016
Deal-0.016
 cream-0.016
 topp-0.016
TokenOnce
Token ID7454
Feature activation-3.80e-01

Pos logit contributions
ening+0.001
 dot+0.001
 wil+0.001
en+0.001
fly+0.001
Neg logit contributions
 math-0.000
 cream-0.000
 subt-0.000
 topp-0.000
 las-0.000
Token deep
Token ID2769
Feature activation-3.37e-01

Pos logit contributions
 subt+0.009
 math+0.009
 cream+0.009
 topp+0.009
Deal+0.009
Neg logit contributions
en-0.012
 dot-0.012
el-0.012
fly-0.012
ug-0.011
Token paws
Token ID49193
Feature activation-3.97e-01

Pos logit contributions
fly+0.009
ches+0.009
 dot+0.009
ays+0.009
ala+0.009
Neg logit contributions
 math-0.010
 subt-0.010
 job-0.010
 cream-0.009
 topp-0.009
Token,
Token ID11
Feature activation+2.29e-01

Pos logit contributions
inkle+0.012
 dot+0.012
en+0.012
py+0.012
fly+0.012
Neg logit contributions
 cream-0.010
 math-0.010
 subtract-0.010
 topp-0.009
 subt-0.009
Token to
Token ID284
Feature activation+1.17e+00

Pos logit contributions
 subt+0.007
 math+0.007
 cream+0.007
 topp+0.007
 subtract+0.007
Neg logit contributions
en-0.006
 dot-0.006
el-0.005
fly-0.005
ir-0.005
Token no
Token ID645
Feature activation+5.55e-02

Pos logit contributions
 subt+0.032
Deal+0.032
 topp+0.032
 cream+0.031
 las+0.030
Neg logit contributions
en-0.034
 dot-0.033
el-0.033
fly-0.032
le-0.031
Token avail
Token ID29107
Feature activation-3.91e+00

Pos logit contributions
 math+0.001
 subt+0.001
 cream+0.001
 topp+0.001
 subtract+0.001
Neg logit contributions
fly-0.002
 dot-0.002
en-0.002
ug-0.002
el-0.002
Token.
Token ID13
Feature activation+2.19e-01

Pos logit contributions
 dot+0.129
 wil+0.124
inkle+0.122
ga+0.120
ches+0.119
Neg logit contributions
 math-0.084
 topp-0.081
 cream-0.080
Deal-0.078
 job-0.075
Token He
Token ID679
Feature activation+6.07e-01

Pos logit contributions
 subt+0.006
 cream+0.006
 math+0.006
 topp+0.006
Deal+0.006
Neg logit contributions
en-0.006
 dot-0.006
el-0.006
fly-0.006
ug-0.006
Token finally
Token ID3443
Feature activation-7.78e-01

Pos logit contributions
 subt+0.016
Deal+0.016
 cream+0.015
 math+0.015
 topp+0.015
Neg logit contributions
 dot-0.018
ir-0.018
ays-0.018
ug-0.018
en-0.017
Token thought
Token ID1807
Feature activation+2.52e-01

Pos logit contributions
en+0.020
 dot+0.019
el+0.019
fly+0.019
ug+0.019
Neg logit contributions
 subt-0.024
 math-0.024
 cream-0.024
Deal-0.024
 topp-0.024
Token of
Token ID286
Feature activation+9.02e-01

Pos logit contributions
 math+0.007
 cream+0.006
 topp+0.006
 pizza+0.006
 subt+0.006
Neg logit contributions
fly-0.009
 dot-0.009
en-0.008
ley-0.008
el-0.008
Token times
Token ID1661
Feature activation-6.43e-01

Pos logit contributions
 math+0.005
 cream+0.005
 subt+0.005
 topp+0.005
 subtract+0.005
Neg logit contributions
fly-0.004
 dot-0.004
en-0.004
el-0.004
ley-0.004
Token,
Token ID11
Feature activation-5.63e-02

Pos logit contributions
 dot+0.018
fly+0.018
py+0.018
eds+0.018
en+0.018
Neg logit contributions
 math-0.019
 subt-0.018
 subtract-0.018
 cream-0.017
 job-0.016
Token but
Token ID475
Feature activation-1.47e+00

Pos logit contributions
fly+0.001
 dot+0.001
en+0.001
ug+0.001
el+0.001
Neg logit contributions
 math-0.002
 subt-0.002
 cream-0.002
 topp-0.002
 subtract-0.002
Token to
Token ID284
Feature activation+1.41e+00

Pos logit contributions
en+0.037
 dot+0.037
el+0.036
fly+0.036
ug+0.035
Neg logit contributions
 subt-0.046
 math-0.046
 cream-0.045
Deal-0.045
 topp-0.044
Token no
Token ID645
Feature activation+4.11e-01

Pos logit contributions
 topp+0.043
 subt+0.043
Deal+0.041
 cream+0.041
 las+0.041
Neg logit contributions
en-0.038
el-0.036
 dot-0.036
fly-0.035
le-0.035
Token avail
Token ID29107
Feature activation-3.81e+00

Pos logit contributions
 math+0.010
 subt+0.010
 cream+0.009
 topp+0.009
Deal+0.009
Neg logit contributions
 dot-0.013
en-0.013
fly-0.013
ug-0.013
el-0.013
Token.
Token ID13
Feature activation+5.17e-01

Pos logit contributions
 dot+0.123
ga+0.116
inkle+0.116
 hawk+0.116
ug+0.116
Neg logit contributions
 math-0.087
Deal-0.086
 cream-0.080
 topp-0.079
 job-0.077
Token
Token ID220
Feature activation+1.57e+00

Pos logit contributions
 subt+0.014
 cream+0.014
 topp+0.014
 math+0.014
Deal+0.014
Neg logit contributions
en-0.015
el-0.015
 dot-0.015
fly-0.014
ir-0.014
Token
Token ID628
Feature activation+6.21e-01

Pos logit contributions
 math+0.040
 subt+0.039
Deal+0.038
 cream+0.038
 subtract+0.037
Neg logit contributions
 dot-0.050
en-0.050
fly-0.048
el-0.048
ening-0.048
Token Jane
Token ID12091
Feature activation+2.39e-01

Pos logit contributions
 subt+0.020
 topp+0.020
 subtract+0.019
 cream+0.019
Deal+0.019
Neg logit contributions
 dot-0.014
 wil-0.014
el-0.014
ir-0.014
en-0.014
Token's
Token ID338
Feature activation+9.86e-01

Pos logit contributions
 subt+0.005
 cream+0.004
 math+0.004
 subtract+0.004
Deal+0.004
Neg logit contributions
ays-0.008
 dot-0.008
py-0.008
ug-0.008
ir-0.008
Token again
Token ID757
Feature activation-1.06e+00

Pos logit contributions
ening+0.011
py+0.011
 dot+0.011
ley+0.011
 wil+0.011
Neg logit contributions
 math-0.011
 cream-0.010
Deal-0.009
 quiz-0.009
 homework-0.009
Token?"
Token ID1701
Feature activation-7.39e-01

Pos logit contributions
 dot+0.028
py+0.028
ening+0.028
fly+0.027
 wil+0.027
Neg logit contributions
 math-0.033
 cream-0.032
 job-0.030
 dessert-0.029
Deal-0.029
Token
Token ID198
Feature activation-9.00e-01

Pos logit contributions
en+0.026
 dot+0.026
ening+0.025
ches+0.025
le+0.025
Neg logit contributions
 math-0.017
 subt-0.016
Deal-0.016
 cream-0.015
 subtract-0.015
Token
Token ID198
Feature activation-1.29e+00

Pos logit contributions
en+0.035
ening+0.035
 dot+0.034
fly+0.033
 wil+0.032
Neg logit contributions
 math-0.017
 cream-0.017
 subt-0.016
Deal-0.016
 subtract-0.015
TokenSpot
Token ID32565
Feature activation-1.56e+00

Pos logit contributions
en+0.042
 dot+0.042
el+0.042
fly+0.042
ug+0.041
Neg logit contributions
 subt-0.030
 math-0.029
 cream-0.029
 topp-0.029
Deal-0.029
Token n
Token ID299
Feature activation-3.67e+00

Pos logit contributions
 dot+0.051
ug+0.051
ir+0.050
el+0.049
en+0.049
Neg logit contributions
Deal-0.037
 cream-0.037
 subt-0.037
 math-0.037
 topp-0.036
Tokenuzz
Token ID4715
Feature activation-1.30e+00

Pos logit contributions
 dot+0.101
ches+0.099
ening+0.098
en+0.093
 predators+0.091
Neg logit contributions
Deal-0.104
 cream-0.103
 math-0.099
 topp-0.098
Fif-0.096
Tokenled
Token ID992
Feature activation-4.06e-01

Pos logit contributions
 dot+0.030
ug+0.029
 predators+0.028
ga+0.028
ir+0.027
Neg logit contributions
 cream-0.045
Deal-0.042
 topp-0.042
 math-0.039
 las-0.038
Token Tom
Token ID4186
Feature activation-4.98e-01

Pos logit contributions
 dot+0.011
ug+0.011
fly+0.011
 wil+0.011
en+0.011
Neg logit contributions
 cream-0.012
 math-0.011
 subt-0.011
 topp-0.011
 las-0.011
Token.
Token ID13
Feature activation-9.59e-01

Pos logit contributions
 dot+0.013
ir+0.013
ug+0.013
inkle+0.012
 wil+0.012
Neg logit contributions
 cream-0.015
 subtract-0.015
 topp-0.015
 subt-0.015
 math-0.014
Token He
Token ID679
Feature activation+7.24e-02

Pos logit contributions
en+0.028
 dot+0.028
el+0.028
fly+0.027
ug+0.027
Neg logit contributions
 subt-0.026
 math-0.025
 cream-0.025
Deal-0.025
 topp-0.025
Token out
Token ID503
Feature activation+2.40e-01

Pos logit contributions
ink+0.038
inkle+0.038
 wil+0.037
fly+0.037
 dot+0.036
Neg logit contributions
 math-0.025
 cream-0.024
 job-0.024
 subtract-0.020
 ready-0.020
Token,
Token ID11
Feature activation+3.51e-02

Pos logit contributions
 math+0.006
 subt+0.006
 subtract+0.006
 cream+0.006
 topp+0.006
Neg logit contributions
ga-0.007
 dot-0.007
ley-0.007
inkle-0.007
ays-0.007
Token but
Token ID475
Feature activation-1.39e+00

Pos logit contributions
 math+0.001
 cream+0.001
 subt+0.001
 subtract+0.001
 topp+0.001
Neg logit contributions
el-0.001
ug-0.001
ir-0.001
 dot-0.001
en-0.001
Token to
Token ID284
Feature activation+1.60e+00

Pos logit contributions
 dot+0.047
en+0.047
el+0.046
fly+0.045
ir+0.045
Neg logit contributions
 math-0.032
 cream-0.031
 subt-0.031
Deal-0.030
 topp-0.029
Token no
Token ID645
Feature activation+2.90e-01

Pos logit contributions
 subt+0.042
 math+0.041
 cream+0.041
Deal+0.040
 topp+0.040
Neg logit contributions
en-0.048
 dot-0.048
el-0.047
fly-0.046
ug-0.046
Token avail
Token ID29107
Feature activation-3.64e+00

Pos logit contributions
 math+0.007
 cream+0.007
 subt+0.007
 topp+0.007
 subtract+0.007
Neg logit contributions
fly-0.009
 dot-0.009
ug-0.009
ga-0.009
el-0.009
Token.
Token ID13
Feature activation+1.89e-01

Pos logit contributions
 dot+0.117
 wil+0.112
inkle+0.111
ga+0.110
ir+0.108
Neg logit contributions
 math-0.082
 topp-0.079
Deal-0.079
 cream-0.078
 job-0.073
Token He
Token ID679
Feature activation+3.64e-01

Pos logit contributions
 subt+0.005
 cream+0.005
 math+0.005
 topp+0.005
Deal+0.005
Neg logit contributions
en-0.006
 dot-0.006
el-0.006
fly-0.005
ug-0.005
Token felt
Token ID2936
Feature activation-9.35e-01

Pos logit contributions
 cream+0.009
 subt+0.009
Deal+0.008
 topp+0.008
 y+0.008
Neg logit contributions
ays-0.012
 dot-0.011
ir-0.011
inkle-0.011
ors-0.011
Token sad
Token ID6507
Feature activation+5.46e-01

Pos logit contributions
fly+0.033
ley+0.032
 dot+0.032
el+0.031
en+0.031
Neg logit contributions
 math-0.021
 cream-0.020
 y-0.019
 subt-0.019
 las-0.018
Token and
Token ID290
Feature activation+2.50e-01

Pos logit contributions
 math+0.017
 cream+0.015
 subt+0.015
 job+0.015
 subtract+0.014
Neg logit contributions
ir-0.015
el-0.014
ga-0.014
fly-0.014
inkle-0.014
Token had
Token ID550
Feature activation-4.23e-01

Pos logit contributions
 subt+0.005
 cream+0.005
 math+0.005
Deal+0.005
 topp+0.005
Neg logit contributions
en-0.005
 dot-0.005
el-0.005
fly-0.005
ug-0.005
Token a
Token ID257
Feature activation+1.71e-01

Pos logit contributions
 dot+0.012
el+0.012
 wil+0.012
en+0.012
ga+0.011
Neg logit contributions
 subt-0.012
Deal-0.011
 subtract-0.011
 topp-0.011
 cream-0.011
Token pet
Token ID4273
Feature activation-3.27e-01

Pos logit contributions
 subt+0.005
 subtract+0.005
 topp+0.005
Fif+0.005
Deal+0.005
Neg logit contributions
 dot-0.005
 wil-0.004
fly-0.004
el-0.004
inkle-0.004
Token rabbit
Token ID22746
Feature activation-7.57e-01

Pos logit contributions
en+0.010
ening+0.010
inkle+0.010
ug+0.009
ir+0.009
Neg logit contributions
 math-0.009
 cream-0.009
Deal-0.009
 subt-0.008
 pizza-0.008
Token named
Token ID3706
Feature activation-1.20e+00

Pos logit contributions
ening+0.022
en+0.022
inkle+0.021
 dot+0.021
ays+0.021
Neg logit contributions
 math-0.022
 cream-0.021
 pizza-0.021
Deal-0.020
 cakes-0.019
Token Snow
Token ID7967
Feature activation-3.56e+00

Pos logit contributions
ening+0.040
 dot+0.040
ug+0.040
en+0.040
le+0.040
Neg logit contributions
 math-0.029
 subt-0.026
 pizza-0.025
 subtract-0.025
 cream-0.025
Tokeny
Token ID88
Feature activation-1.14e+00

Pos logit contributions
 lands+0.066
 dot+0.064
 predators+0.064
ening+0.063
ug+0.063
Neg logit contributions
 cream-0.120
Deal-0.114
 topp-0.107
 pizza-0.105
 las-0.105
Token.
Token ID13
Feature activation-4.02e-01

Pos logit contributions
en+0.038
ening+0.037
ir+0.037
ug+0.037
 dot+0.037
Neg logit contributions
 math-0.027
Deal-0.027
 cream-0.026
 pizza-0.025
 job-0.024
Token Lily
Token ID20037
Feature activation-4.58e-01

Pos logit contributions
fly+0.014
ening+0.013
ug+0.013
le+0.013
ir+0.013
Neg logit contributions
 snack-0.009
 math-0.009
 Cere-0.009
 y-0.009
 cookies-0.009
Token loved
Token ID6151
Feature activation+7.57e-03

Pos logit contributions
ir+0.022
ays+0.022
ug+0.021
 dot+0.021
 predators+0.020
Neg logit contributions
Deal-0.006
 y-0.006
 cream-0.006
 responsible-0.004
 creative-0.004
Token to
Token ID284
Feature activation-5.50e-02

Pos logit contributions
 subt+0.000
 math+0.000
 cream+0.000
 topp+0.000
Deal+0.000
Neg logit contributions
en-0.000
 dot-0.000
el-0.000
fly-0.000
ug-0.000
Tokenens
Token ID641
Feature activation-2.03e+00

Pos logit contributions
 dot+0.056
en+0.056
el+0.055
fly+0.055
ug+0.053
Neg logit contributions
 subt-0.059
Deal-0.059
 cream-0.059
 math-0.059
 topp-0.057
Token and
Token ID290
Feature activation-1.26e-01

Pos logit contributions
 dot+0.069
inkle+0.066
en+0.065
el+0.064
ening+0.064
Neg logit contributions
 math-0.046
 pizza-0.044
Deal-0.043
 cream-0.043
 cakes-0.042
Token a
Token ID257
Feature activation-3.66e-01

Pos logit contributions
fly+0.003
 dot+0.003
en+0.003
py+0.003
pt+0.003
Neg logit contributions
 subt-0.004
 topp-0.004
 math-0.004
 subtract-0.004
Deal-0.004
Token rabbit
Token ID22746
Feature activation-1.58e+00

Pos logit contributions
 dot+0.009
fly+0.009
 wil+0.008
en+0.008
el+0.008
Neg logit contributions
 subt-0.012
Deal-0.011
 topp-0.011
 subtract-0.011
 math-0.011
Token named
Token ID3706
Feature activation-1.25e+00

Pos logit contributions
en+0.044
ening+0.043
 dot+0.043
inkle+0.043
el+0.042
Neg logit contributions
 math-0.044
Deal-0.043
 cream-0.042
 subt-0.041
 topp-0.041
Token Ho
Token ID9544
Feature activation-3.55e+00

Pos logit contributions
en+0.036
 dot+0.036
el+0.035
fly+0.035
ug+0.034
Neg logit contributions
 subt-0.034
 math-0.034
 cream-0.034
 topp-0.033
Deal-0.033
Tokenppy
Token ID14097
Feature activation-2.14e+00

Pos logit contributions
ening+0.065
 dot+0.065
 wil+0.062
en+0.061
el+0.060
Neg logit contributions
Deal-0.135
 topp-0.124
 cream-0.123
 math-0.122
 subt-0.121
Token,
Token ID11
Feature activation-1.25e+00

Pos logit contributions
 dot+0.069
en+0.067
inkle+0.067
el+0.067
ening+0.066
Neg logit contributions
 math-0.050
Deal-0.047
 cream-0.047
 subt-0.047
 pizza-0.046
Token getting
Token ID1972
Feature activation+2.13e-01

Pos logit contributions
inkle+0.047
 dot+0.045
ir+0.045
ug+0.045
le+0.045
Neg logit contributions
 y-0.027
 hamb-0.026
 cream-0.026
 pizza-0.025
 math-0.024
Token ready
Token ID3492
Feature activation-2.40e-01

Pos logit contributions
 subt+0.006
 math+0.006
Deal+0.006
 topp+0.006
 cream+0.006
Neg logit contributions
en-0.006
el-0.006
 dot-0.006
fly-0.006
ug-0.006
Token for
Token ID329
Feature activation+1.93e-02

Pos logit contributions
 dot+0.008
en+0.008
fly+0.008
el+0.008
ug+0.008
Neg logit contributions
 math-0.005
 cream-0.005
 subt-0.005
 topp-0.005
 las-0.005
Token met
Token ID1138
Feature activation+3.95e-01

Pos logit contributions
ays+0.013
ir+0.013
inkle+0.013
 dot+0.013
ug+0.013
Neg logit contributions
 subt-0.007
 cream-0.006
 math-0.006
 subtract-0.006
Deal-0.006
Token a
Token ID257
Feature activation-6.06e-01

Pos logit contributions
 math+0.007
 cream+0.007
Deal+0.007
 subt+0.006
 topp+0.006
Neg logit contributions
 dot-0.016
en-0.015
ga-0.015
ening-0.015
ug-0.015
Token kind
Token ID1611
Feature activation-7.48e-01

Pos logit contributions
en+0.019
 dot+0.019
el+0.019
fly+0.019
ug+0.019
Neg logit contributions
 subt-0.015
 math-0.014
 cream-0.014
Deal-0.014
 topp-0.014
Token bird
Token ID6512
Feature activation-4.86e-01

Pos logit contributions
ays+0.023
ches+0.023
unks+0.023
ening+0.022
en+0.022
Neg logit contributions
 cream-0.019
 math-0.019
 pizza-0.018
 customer-0.018
 job-0.017
Token named
Token ID3706
Feature activation-5.37e-01

Pos logit contributions
 dot+0.013
en+0.013
ga+0.013
ening+0.013
inkle+0.013
Neg logit contributions
 math-0.013
 cream-0.013
Deal-0.013
 pizza-0.012
 topp-0.012
Token Ch
Token ID609
Feature activation-3.53e+00

Pos logit contributions
en+0.012
 dot+0.012
el+0.012
ug+0.012
le+0.012
Neg logit contributions
 math-0.018
 subt-0.018
 cream-0.018
 topp-0.017
Deal-0.017
Tokenir
Token ID343
Feature activation-6.38e-01

Pos logit contributions
 dot+0.059
 landed+0.057
 prope+0.056
 lands+0.055
ley+0.055
Neg logit contributions
 math-0.114
igsaw-0.113
Fif-0.107
 topp-0.106
 las-0.106
Tokenpy
Token ID9078
Feature activation-8.46e-01

Pos logit contributions
 dot+0.007
inkle+0.007
ley+0.007
 wil+0.006
ga+0.006
Neg logit contributions
 math-0.027
 cream-0.026
 subtract-0.026
 cakes-0.026
 topp-0.025
Token.
Token ID13
Feature activation-1.02e+00

Pos logit contributions
inkle+0.027
 dot+0.025
ga+0.025
ink+0.025
en+0.025
Neg logit contributions
 math-0.020
Deal-0.019
 cream-0.019
 topp-0.019
 subt-0.019
Token Ch
Token ID609
Feature activation-2.97e+00

Pos logit contributions
en+0.025
el+0.024
le+0.024
 wil+0.024
 dot+0.024
Neg logit contributions
 subt-0.033
 math-0.032
Deal-0.031
 subtract-0.031
 topp-0.031
Tokenir
Token ID343
Feature activation-7.92e-01

Pos logit contributions
 dot+0.054
hole+0.052
 lands+0.051
 landed+0.050
ley+0.050
Neg logit contributions
igsaw-0.098
 math-0.093
 topp-0.088
 dessert-0.087
Fif-0.086
Token him
Token ID683
Feature activation+2.93e-02

Pos logit contributions
 dot+0.011
fly+0.011
en+0.011
ug+0.011
inkle+0.011
Neg logit contributions
 math-0.013
 cream-0.012
 subt-0.012
 topp-0.012
 subtract-0.012
Token a
Token ID257
Feature activation-5.80e-01

Pos logit contributions
 subt+0.001
 subtract+0.001
Deal+0.001
 topp+0.001
bek+0.001
Neg logit contributions
el-0.001
ches-0.001
en-0.001
pt-0.001
ley-0.001
Token lick
Token ID36728
Feature activation+1.63e-01

Pos logit contributions
 dot+0.015
fly+0.014
 wil+0.014
pt+0.014
ches+0.014
Neg logit contributions
 subt-0.019
Deal-0.018
 subtract-0.017
 topp-0.017
Fif-0.016
Token and
Token ID290
Feature activation-7.27e-01

Pos logit contributions
 cream+0.006
 math+0.005
 topp+0.005
 pizza+0.005
 las+0.005
Neg logit contributions
inkle-0.004
en-0.004
fly-0.004
py-0.004
 dot-0.004
Token a
Token ID257
Feature activation-2.71e-01

Pos logit contributions
 dot+0.020
en+0.019
fly+0.019
 wil+0.019
el+0.019
Neg logit contributions
Deal-0.021
 subt-0.020
 math-0.020
Fif-0.020
 subtract-0.019
Token n
Token ID299
Feature activation-3.52e+00

Pos logit contributions
 dot+0.008
fly+0.007
 wing+0.007
 wil+0.007
inkle+0.006
Neg logit contributions
 subt-0.009
Deal-0.008
 subtract-0.008
Fif-0.007
 topp-0.007
Tokenuzzle
Token ID9625
Feature activation-1.39e+00

Pos logit contributions
ening+0.090
ches+0.088
 dot+0.086
en+0.086
 predators+0.082
Neg logit contributions
 cream-0.100
Deal-0.099
 topp-0.097
 math-0.095
bek-0.093
Token.
Token ID13
Feature activation-3.56e-01

Pos logit contributions
 dot+0.044
inkle+0.043
unks+0.041
en+0.041
 predators+0.041
Neg logit contributions
 cream-0.036
 math-0.032
 cakes-0.031
 topp-0.031
 pizza-0.031
Token She
Token ID1375
Feature activation-7.20e-03

Pos logit contributions
en+0.010
le+0.010
fly+0.010
el+0.010
 dot+0.010
Neg logit contributions
 math-0.010
 subt-0.010
 subtract-0.010
Deal-0.009
 cream-0.009
Token learned
Token ID4499
Feature activation-9.53e-01

Pos logit contributions
 dot+0.000
ening+0.000
ug+0.000
en+0.000
ir+0.000
Neg logit contributions
 cream-0.000
Deal-0.000
 subt-0.000
 topp-0.000
 math-0.000
Token her
Token ID607
Feature activation+6.51e-01

Pos logit contributions
en+0.022
el+0.021
 dot+0.020
pt+0.020
fly+0.020
Neg logit contributions
 subt-0.032
Deal-0.031
 topp-0.031
 subtract-0.031
Fif-0.030
Token to
Token ID284
Feature activation-2.29e-01

Pos logit contributions
 cream+0.019
 math+0.019
 topp+0.018
 subt+0.018
Deal+0.018
Neg logit contributions
ug-0.017
 dot-0.016
el-0.016
ley-0.016
ir-0.016
Token Ben
Token ID3932
Feature activation-1.13e+00

Pos logit contributions
ug+0.006
 dot+0.005
ening+0.005
en+0.005
el+0.005
Neg logit contributions
 math-0.007
 cream-0.007
 subt-0.007
 las-0.007
 topp-0.007
Token and
Token ID290
Feature activation-8.36e-01

Pos logit contributions
ug+0.028
 dot+0.027
 wil+0.027
ir+0.027
py+0.026
Neg logit contributions
 cream-0.034
 math-0.034
 pizza-0.033
 topp-0.033
 list-0.031
Token Lily
Token ID20037
Feature activation-1.27e-01

Pos logit contributions
fly+0.022
 dot+0.022
en+0.022
ches+0.021
 wil+0.021
Neg logit contributions
 subt-0.025
Deal-0.025
 math-0.025
 topp-0.024
 subtract-0.024
Token and
Token ID290
Feature activation+1.59e-01

Pos logit contributions
ug+0.004
 dot+0.004
ir+0.003
el+0.003
py+0.003
Neg logit contributions
 cream-0.003
 topp-0.003
 math-0.003
 subtract-0.003
 subt-0.003
Token n
Token ID299
Feature activation-3.50e+00

Pos logit contributions
 subt+0.005
Deal+0.005
 math+0.005
 subtract+0.005
 topp+0.005
Neg logit contributions
fly-0.004
en-0.004
 landed-0.004
 dot-0.004
 wil-0.004
Tokenuzz
Token ID4715
Feature activation-1.40e+00

Pos logit contributions
ches+0.088
 dot+0.088
ening+0.084
en+0.080
 hawk+0.079
Neg logit contributions
Deal-0.103
 cream-0.103
 topp-0.101
bek-0.098
Fif-0.097
Tokenled
Token ID992
Feature activation-8.81e-01

Pos logit contributions
 dot+0.031
ug+0.029
py+0.028
inkle+0.028
ga+0.028
Neg logit contributions
 cream-0.049
 topp-0.047
Deal-0.044
 cakes-0.042
 pizza-0.042
Token their
Token ID511
Feature activation+3.36e-01

Pos logit contributions
 dot+0.023
en+0.023
ug+0.023
fly+0.022
el+0.022
Neg logit contributions
 cream-0.026
 math-0.026
 subt-0.026
 topp-0.025
Deal-0.025
Token hands
Token ID2832
Feature activation-9.41e-01

Pos logit contributions
 subt+0.011
 math+0.011
 cream+0.011
 topp+0.011
Deal+0.011
Neg logit contributions
en-0.008
 dot-0.008
el-0.008
fly-0.008
ug-0.007
Token.
Token ID13
Feature activation-2.55e-01

Pos logit contributions
en+0.027
 dot+0.027
ug+0.027
fly+0.026
py+0.026
Neg logit contributions
 cream-0.025
 topp-0.025
 math-0.025
 subt-0.024
Deal-0.023
Tokenyt
Token ID20760
Feature activation+5.95e-01

Pos logit contributions
el+0.031
ches+0.030
en+0.030
ug+0.030
ir+0.030
Neg logit contributions
 cream-0.032
 math-0.031
 topp-0.030
 subt-0.029
 job-0.029
Tokenale
Token ID1000
Feature activation-9.68e-01

Pos logit contributions
 math+0.020
 cream+0.020
 topp+0.020
 subt+0.019
Deal+0.019
Neg logit contributions
 dot-0.012
 wil-0.012
fly-0.012
en-0.011
ink-0.011
Token creature
Token ID7185
Feature activation-1.10e+00

Pos logit contributions
en+0.024
 dot+0.024
el+0.024
ug+0.023
ir+0.023
Neg logit contributions
 cream-0.029
 math-0.029
 topp-0.029
 subt-0.028
 subtract-0.027
Token called
Token ID1444
Feature activation-9.08e-01

Pos logit contributions
ug+0.032
en+0.032
ir+0.031
 dot+0.031
el+0.031
Neg logit contributions
 math-0.028
 cream-0.028
 topp-0.027
 pizza-0.026
 subt-0.026
Token a
Token ID257
Feature activation-1.21e+00

Pos logit contributions
en+0.027
 dot+0.027
el+0.026
le+0.026
 wil+0.026
Neg logit contributions
 math-0.024
 subt-0.023
 cream-0.023
 topp-0.023
 las-0.022
Token n
Token ID299
Feature activation-3.49e+00

Pos logit contributions
en+0.035
el+0.034
ug+0.033
 dot+0.033
le+0.033
Neg logit contributions
 math-0.034
 cream-0.033
 subt-0.033
 topp-0.033
 las-0.032
Tokenymph
Token ID20896
Feature activation-1.74e+00

Pos logit contributions
ening+0.086
ches+0.083
 landed+0.081
 wil+0.079
 dot+0.077
Neg logit contributions
bek-0.096
Fif-0.091
 las-0.090
 subt-0.087
Deal-0.087
Token!
Token ID0
Feature activation-1.75e-01

Pos logit contributions
el+0.048
ug+0.047
 dot+0.047
en+0.047
ays+0.047
Neg logit contributions
 math-0.045
 cream-0.045
 topp-0.045
 las-0.044
 subt-0.044
Token
Token ID198
Feature activation+2.61e-01

Pos logit contributions
en+0.005
 dot+0.005
el+0.005
fly+0.005
le+0.005
Neg logit contributions
 subt-0.005
 math-0.005
 cream-0.005
 topp-0.005
Deal-0.005
Token
Token ID198
Feature activation-5.22e-02

Pos logit contributions
 subt+0.005
 math+0.005
 cream+0.004
 subtract+0.004
Fif+0.004
Neg logit contributions
en-0.010
ening-0.010
 dot-0.010
fly-0.010
 wil-0.010
TokenThe
Token ID464
Feature activation+4.52e-01

Pos logit contributions
en+0.002
 dot+0.002
el+0.002
fly+0.002
ug+0.002
Neg logit contributions
 subt-0.001
 math-0.001
 cream-0.001
 topp-0.001
Deal-0.001
Token,
Token ID11
Feature activation-1.33e+00

Pos logit contributions
 dot+0.014
 wil+0.013
inkle+0.013
ening+0.013
ir+0.013
Neg logit contributions
 math-0.013
 subt-0.013
 cream-0.013
 pizza-0.013
 las-0.012
Token a
Token ID257
Feature activation+2.23e-01

Pos logit contributions
en+0.044
 dot+0.044
el+0.043
fly+0.043
ug+0.042
Neg logit contributions
 subt-0.031
 math-0.030
 cream-0.030
Deal-0.030
 topp-0.030
Token little
Token ID1310
Feature activation-6.83e-01

Pos logit contributions
 subt+0.006
Deal+0.006
 topp+0.006
 subtract+0.006
 math+0.006
Neg logit contributions
 dot-0.007
fly-0.006
ug-0.006
 hawk-0.006
el-0.006
Token bird
Token ID6512
Feature activation-8.59e-01

Pos logit contributions
en+0.017
 dot+0.016
el+0.016
fly+0.016
ug+0.016
Neg logit contributions
 subt-0.022
 math-0.022
 cream-0.021
 topp-0.021
Deal-0.021
Token named
Token ID3706
Feature activation-6.56e-01

Pos logit contributions
 dot+0.021
en+0.021
ening+0.021
ga+0.020
ays+0.020
Neg logit contributions
 cream-0.025
 math-0.024
Deal-0.024
 topp-0.023
 subt-0.023
Token Ch
Token ID609
Feature activation-3.48e+00

Pos logit contributions
en+0.016
 dot+0.016
el+0.016
fly+0.015
ug+0.015
Neg logit contributions
 subt-0.021
 math-0.021
 cream-0.020
Deal-0.020
 topp-0.020
Tokenir
Token ID343
Feature activation-7.32e-01

Pos logit contributions
 wil+0.059
 dot+0.053
ening+0.052
 lands+0.052
 landed+0.047
Neg logit contributions
 math-0.116
igsaw-0.111
Fif-0.111
Deal-0.111
 subtract-0.111
Tokenpy
Token ID9078
Feature activation-1.28e+00

Pos logit contributions
 dot+0.006
 wil+0.006
inkle+0.006
ley+0.005
en+0.005
Neg logit contributions
 math-0.033
 cream-0.033
 subtract-0.032
 subt-0.032
 topp-0.031
Token,
Token ID11
Feature activation-7.11e-01

Pos logit contributions
inkle+0.036
 dot+0.036
ays+0.035
ga+0.035
 wil+0.034
Neg logit contributions
 math-0.032
 subt-0.031
 cream-0.030
 las-0.029
 topp-0.029
Token flew
Token ID13112
Feature activation-8.48e-01

Pos logit contributions
inkle+0.027
 dot+0.025
le+0.025
en+0.024
ir+0.024
Neg logit contributions
 cream-0.014
 hamb-0.014
 subt-0.014
 las-0.014
 y-0.014
Token by
Token ID416
Feature activation+5.94e-01

Pos logit contributions
 dot+0.025
en+0.025
inkle+0.024
el+0.024
fly+0.024
Neg logit contributions
 cream-0.023
 math-0.022
 subt-0.022
 topp-0.021
Deal-0.021
Token him
Token ID683
Feature activation-2.55e-01

Pos logit contributions
 math+0.007
 topp+0.007
 cream+0.006
 subt+0.006
 cakes+0.006
Neg logit contributions
 dot-0.011
ga-0.011
ley-0.011
le-0.011
fly-0.011
Token.
Token ID13
Feature activation+3.27e-01

Pos logit contributions
 dot+0.008
en+0.008
ga+0.007
ug+0.007
el+0.007
Neg logit contributions
 math-0.007
 cream-0.006
 topp-0.006
 subt-0.006
Deal-0.006
Token But
Token ID887
Feature activation-1.02e+00

Pos logit contributions
 subt+0.009
 math+0.008
 cream+0.008
 topp+0.008
Deal+0.008
Neg logit contributions
en-0.010
 dot-0.010
el-0.010
fly-0.010
ug-0.009
Token to
Token ID284
Feature activation+1.92e+00

Pos logit contributions
 dot+0.029
en+0.029
el+0.028
fly+0.028
ir+0.028
Neg logit contributions
 math-0.028
 subt-0.028
 cream-0.027
 topp-0.027
Deal-0.027
Token no
Token ID645
Feature activation+4.43e-01

Pos logit contributions
 topp+0.055
 subt+0.055
Deal+0.055
 cream+0.054
 las+0.051
Neg logit contributions
en-0.054
el-0.052
fly-0.050
ir-0.050
 dot-0.050
Token avail
Token ID29107
Feature activation-3.48e+00

Pos logit contributions
 math+0.010
 subt+0.010
 cream+0.010
 topp+0.010
Deal+0.009
Neg logit contributions
 dot-0.015
en-0.015
fly-0.015
el-0.015
ug-0.015
Token,
Token ID11
Feature activation-1.97e-01

Pos logit contributions
 dot+0.109
 wil+0.105
 horizon+0.103
ga+0.103
 hawk+0.102
Neg logit contributions
 math-0.079
 topp-0.078
Deal-0.077
 cream-0.073
 job-0.072
Token they
Token ID484
Feature activation+1.75e-01

Pos logit contributions
 dot+0.006
en+0.006
el+0.006
fly+0.006
ir+0.006
Neg logit contributions
 math-0.005
 subt-0.005
 cream-0.005
 topp-0.005
Deal-0.005
Token were
Token ID547
Feature activation+4.01e-01

Pos logit contributions
Deal+0.004
 cream+0.004
 y+0.004
 math+0.004
 topp+0.004
Neg logit contributions
 dot-0.005
ays-0.005
ug-0.005
ga-0.005
en-0.005
Token too
Token ID1165
Feature activation+4.17e-01

Pos logit contributions
 math+0.011
 cream+0.010
 subt+0.010
 topp+0.009
 pizza+0.009
Neg logit contributions
 dot-0.013
el-0.012
ley-0.012
en-0.012
ug-0.012
Token scared
Token ID12008
Feature activation+6.64e-01

Pos logit contributions
 cream+0.010
 math+0.009
 subt+0.009
 topp+0.009
 job+0.009
Neg logit contributions
en-0.014
 dot-0.014
fly-0.014
el-0.013
py-0.013
Token Where
Token ID6350
Feature activation-1.11e+00

Pos logit contributions
 subt+0.019
 math+0.019
 topp+0.019
 cream+0.019
 subtract+0.018
Neg logit contributions
en-0.023
 dot-0.022
el-0.022
ug-0.022
fly-0.022
Token are
Token ID389
Feature activation+1.45e+00

Pos logit contributions
en+0.026
 dot+0.026
el+0.026
fly+0.025
ug+0.025
Neg logit contributions
 subt-0.036
 math-0.035
 cream-0.035
Deal-0.035
 topp-0.035
Token you
Token ID345
Feature activation-1.17e-01

Pos logit contributions
Deal+0.057
 subt+0.056
 subtract+0.053
bek+0.053
Fif+0.051
Neg logit contributions
el-0.029
 us-0.024
ie-0.024
pt-0.023
en-0.023
Token?"
Token ID1701
Feature activation-9.31e-01

Pos logit contributions
 dot+0.004
ug+0.004
fly+0.003
ley+0.003
ga+0.003
Neg logit contributions
 math-0.003
 cream-0.003
 las-0.003
 subt-0.003
 topp-0.003
Token their
Token ID511
Feature activation+1.43e+00

Pos logit contributions
en+0.035
 dot+0.035
ug+0.034
fly+0.034
ening+0.034
Neg logit contributions
 math-0.019
 subt-0.018
 cream-0.017
 subtract-0.016
 las-0.016
Token n
Token ID299
Feature activation-3.47e+00

Pos logit contributions
 math+0.027
 cream+0.026
 subt+0.025
 cakes+0.023
 topp+0.023
Neg logit contributions
en-0.055
ug-0.053
le-0.053
ches-0.053
ening-0.053
Tokenanny
Token ID7737
Feature activation-1.11e+00

Pos logit contributions
ening+0.104
 dot+0.104
ches+0.103
 hawk+0.099
en+0.098
Neg logit contributions
 subt-0.081
 math-0.080
 cream-0.080
 topp-0.077
Fif-0.077
Token shouted
Token ID17293
Feature activation-3.87e-01

Pos logit contributions
 dot+0.036
ug+0.036
ga+0.035
ir+0.035
 predators+0.034
Neg logit contributions
 subt-0.025
 subtract-0.024
 math-0.024
 cream-0.024
 topp-0.023
Token.
Token ID13
Feature activation-3.95e-01

Pos logit contributions
en+0.010
 dot+0.010
fly+0.010
el+0.010
ug+0.010
Neg logit contributions
 subt-0.012
 math-0.011
 cream-0.011
 subtract-0.011
 topp-0.011
Token She
Token ID1375
Feature activation-7.62e-01

Pos logit contributions
en+0.015
le+0.015
ening+0.014
ug+0.014
el+0.014
Neg logit contributions
 math-0.008
 subt-0.008
 subtract-0.007
 las-0.007
 cream-0.007
Token saw
Token ID2497
Feature activation+1.01e+00

Pos logit contributions
ug+0.024
 dot+0.023
ir+0.023
en+0.023
el+0.023
Neg logit contributions
 cream-0.020
 subt-0.020
 topp-0.019
 math-0.019
 las-0.018
Token clothes
Token ID8242
Feature activation+7.97e-01

Pos logit contributions
 subt+0.023
 math+0.023
 cream+0.023
Deal+0.023
 topp+0.022
Neg logit contributions
 dot-0.019
fly-0.019
en-0.019
el-0.019
ug-0.019
Token,
Token ID11
Feature activation-6.08e-01

Pos logit contributions
 cream+0.027
 math+0.026
 subt+0.026
Deal+0.025
 subtract+0.025
Neg logit contributions
fly-0.019
 dot-0.018
ug-0.018
py-0.018
ir-0.018
Token like
Token ID588
Feature activation+9.12e-01

Pos logit contributions
en+0.018
 dot+0.018
fly+0.017
ug+0.017
el+0.017
Neg logit contributions
 math-0.017
 subt-0.017
 cream-0.016
 topp-0.016
 subtract-0.016
Token her
Token ID607
Feature activation-6.33e-02

Pos logit contributions
 math+0.023
 subt+0.022
 cream+0.022
 subtract+0.021
 cakes+0.021
Neg logit contributions
fly-0.030
 dot-0.030
inkle-0.030
en-0.029
ley-0.029
Token sister
Token ID6621
Feature activation-5.62e-01

Pos logit contributions
fly+0.002
ug+0.002
en+0.002
ches+0.002
ir+0.002
Neg logit contributions
 math-0.001
 cream-0.001
 subt-0.001
 job-0.001
 cakes-0.001
Token L
Token ID406
Feature activation-3.47e+00

Pos logit contributions
ir+0.016
ug+0.016
 dot+0.016
fly+0.016
 hawk+0.016
Neg logit contributions
 math-0.016
 subt-0.015
 cream-0.015
 cakes-0.015
 subtract-0.014
Tokenila
Token ID10102
Feature activation-8.71e-01

Pos logit contributions
 dot+0.092
fly+0.087
ening+0.084
en+0.081
ley+0.080
Neg logit contributions
 cream-0.109
Deal-0.109
 math-0.109
 subt-0.103
 cakes-0.102
Token.
Token ID13
Feature activation-2.12e-01

Pos logit contributions
ir+0.029
 hawk+0.028
ug+0.028
 dot+0.028
el+0.028
Neg logit contributions
 math-0.021
 cream-0.019
 cakes-0.019
 homework-0.018
 subt-0.018
Token
Token ID198
Feature activation-1.76e-01

Pos logit contributions
en+0.005
 dot+0.005
fly+0.005
el+0.005
ug+0.005
Neg logit contributions
 subt-0.006
 math-0.006
Deal-0.006
 cream-0.006
 subtract-0.006
Token
Token ID198
Feature activation-5.45e-01

Pos logit contributions
ening+0.007
 dot+0.007
fly+0.007
en+0.007
ley+0.007
Neg logit contributions
 math-0.003
Deal-0.003
 subt-0.003
 cream-0.002
Fif-0.002
TokenOne
Token ID3198
Feature activation+3.80e-01

Pos logit contributions
en+0.015
el+0.015
 dot+0.015
ir+0.015
ug+0.015
Neg logit contributions
 subt-0.016
 topp-0.015
 cream-0.015
 math-0.015
 subtract-0.015
Token Tom
Token ID4186
Feature activation+1.49e-01

Pos logit contributions
 dot+0.003
fly+0.003
inkle+0.003
ink+0.003
 wil+0.003
Neg logit contributions
 math-0.002
 cream-0.002
 subt-0.002
 pizza-0.002
 topp-0.002
Token a
Token ID257
Feature activation-3.14e-01

Pos logit contributions
 subt+0.003
 math+0.003
 cream+0.003
 topp+0.003
Deal+0.003
Neg logit contributions
 dot-0.005
en-0.005
el-0.005
fly-0.005
ug-0.005
Token lick
Token ID36728
Feature activation+1.81e-01

Pos logit contributions
 dot+0.009
fly+0.009
ches+0.008
pt+0.008
inkle+0.008
Neg logit contributions
 subt-0.009
Deal-0.009
 subtract-0.008
 topp-0.008
Fif-0.008
Token and
Token ID290
Feature activation-3.43e-01

Pos logit contributions
 cream+0.006
 math+0.006
 topp+0.006
 pizza+0.005
 las+0.005
Neg logit contributions
inkle-0.004
 dot-0.004
 wil-0.004
en-0.004
fly-0.004
Token a
Token ID257
Feature activation-8.53e-02

Pos logit contributions
en+0.009
 dot+0.009
fly+0.009
el+0.009
le+0.008
Neg logit contributions
Deal-0.010
 subt-0.010
bek-0.010
 cream-0.010
Fif-0.009
Token n
Token ID299
Feature activation-3.46e+00

Pos logit contributions
 dot+0.002
fly+0.002
 wil+0.002
inkle+0.002
 wing+0.002
Neg logit contributions
 subt-0.003
Deal-0.003
 subtract-0.002
 math-0.002
Fif-0.002
Tokenudge
Token ID12587
Feature activation-7.44e-01

Pos logit contributions
ening+0.093
 dot+0.089
ches+0.088
en+0.087
 predators+0.085
Neg logit contributions
Deal-0.097
 cream-0.096
 math-0.093
 topp-0.091
Fif-0.089
Token.
Token ID13
Feature activation-6.29e-01

Pos logit contributions
en+0.019
 dot+0.018
el+0.018
fly+0.018
ug+0.018
Neg logit contributions
 cream-0.023
 subt-0.023
 math-0.023
 topp-0.023
Deal-0.022
Token Tom
Token ID4186
Feature activation-4.05e-02

Pos logit contributions
 dot+0.018
en+0.017
el+0.017
fly+0.017
ug+0.017
Neg logit contributions
 subt-0.018
 cream-0.017
 math-0.017
Deal-0.017
 topp-0.017
Token smiles
Token ID21845
Feature activation+4.25e-01

Pos logit contributions
ug+0.001
ir+0.001
 dot+0.001
el+0.001
fly+0.001
Neg logit contributions
 math-0.001
 cream-0.001
 subt-0.001
 topp-0.001
 subtract-0.001
Token and
Token ID290
Feature activation+4.45e-01

Pos logit contributions
 subt+0.018
 math+0.017
 cream+0.017
Deal+0.017
 topp+0.017
Neg logit contributions
en-0.006
 dot-0.006
el-0.006
fly-0.006
ug-0.006
Token some
Token ID617
Feature activation-7.01e-01

Pos logit contributions
 wil+0.032
 landed+0.031
pt+0.030
en+0.030
fly+0.030
Neg logit contributions
 subt-0.042
 topp-0.040
 las-0.040
 subtract-0.039
igsaw-0.039
Token friends
Token ID2460
Feature activation-2.91e-01

Pos logit contributions
en+0.028
fly+0.028
 dot+0.028
el+0.027
ening+0.027
Neg logit contributions
 cream-0.013
 math-0.012
Deal-0.012
 topp-0.012
 subt-0.011
Token.
Token ID13
Feature activation-1.99e-01

Pos logit contributions
 dot+0.009
en+0.009
el+0.008
ug+0.008
ir+0.008
Neg logit contributions
 cream-0.008
 math-0.008
Deal-0.008
 subt-0.008
 topp-0.007
Token She
Token ID1375
Feature activation-4.54e-01

Pos logit contributions
en+0.007
le+0.007
ening+0.007
ir+0.007
inkle+0.007
Neg logit contributions
 math-0.004
 subt-0.004
Deal-0.004
 subtract-0.004
Fif-0.004
Token wished
Token ID16555
Feature activation-1.19e+00

Pos logit contributions
ir+0.016
ug+0.016
ening+0.016
ays+0.016
ors+0.016
Neg logit contributions
Deal-0.010
 cream-0.010
 y-0.009
 subt-0.009
 cash-0.009
Token Snow
Token ID7967
Feature activation-3.45e+00

Pos logit contributions
 dot+0.049
en+0.049
fly+0.048
le+0.047
ening+0.047
Neg logit contributions
 math-0.020
 cream-0.019
 subt-0.018
 cakes-0.017
 las-0.017
Tokeny
Token ID88
Feature activation-1.03e+00

Pos logit contributions
 predators+0.079
 prope+0.071
inkle+0.071
ening+0.070
 dot+0.070
Neg logit contributions
 cream-0.107
Deal-0.099
 topp-0.091
 las-0.087
Un-0.086
Token was
Token ID373
Feature activation-1.38e+00

Pos logit contributions
ir+0.034
 dot+0.033
ug+0.033
inkle+0.032
ga+0.032
Neg logit contributions
 cream-0.028
 math-0.026
Deal-0.026
 job-0.024
 topp-0.024
Token alive
Token ID6776
Feature activation-1.11e+00

Pos logit contributions
en+0.046
 dot+0.046
el+0.045
fly+0.045
ug+0.044
Neg logit contributions
 math-0.031
 subt-0.031
 cream-0.031
Deal-0.030
 topp-0.030
Token.
Token ID13
Feature activation-4.77e-01

Pos logit contributions
ir+0.041
 dot+0.041
el+0.040
unks+0.040
inkle+0.040
Neg logit contributions
 math-0.024
 cream-0.023
Deal-0.022
 y-0.020
 pizza-0.020
Token She
Token ID1375
Feature activation-4.79e-01

Pos logit contributions
en+0.017
ening+0.016
ir+0.016
le+0.016
fly+0.016
Neg logit contributions
Deal-0.011
 math-0.011
 subt-0.011
 subtract-0.010
Fif-0.009
Token,
Token ID11
Feature activation-2.37e+00

Pos logit contributions
 dot+0.045
ening+0.043
 wil+0.042
en+0.042
ga+0.041
Neg logit contributions
 math-0.049
 cream-0.048
 pizza-0.047
 cakes-0.046
 las-0.046
Token a
Token ID257
Feature activation-6.05e-01

Pos logit contributions
en+0.078
 dot+0.076
el+0.075
ening+0.074
ug+0.074
Neg logit contributions
 math-0.056
 subt-0.056
 cream-0.056
Deal-0.055
 topp-0.054
Token little
Token ID1310
Feature activation-1.09e+00

Pos logit contributions
en+0.019
 dot+0.019
el+0.019
fly+0.018
ug+0.018
Neg logit contributions
 subt-0.015
 math-0.015
 cream-0.015
Deal-0.014
 topp-0.014
Token chick
Token ID17943
Feature activation-1.87e+00

Pos logit contributions
en+0.035
inkle+0.034
ches+0.033
ening+0.033
ink+0.033
Neg logit contributions
 cream-0.027
 math-0.026
Deal-0.025
 pizza-0.025
 topp-0.025
Token named
Token ID3706
Feature activation-1.64e+00

Pos logit contributions
en+0.049
ening+0.049
 dot+0.047
el+0.046
inkle+0.046
Neg logit contributions
 cream-0.054
Deal-0.053
 math-0.051
 pizza-0.050
 job-0.050
Token Ch
Token ID609
Feature activation-3.43e+00

Pos logit contributions
en+0.049
ening+0.049
le+0.047
 dot+0.047
inkle+0.046
Neg logit contributions
 math-0.043
 pizza-0.041
 las-0.041
 y-0.041
 cream-0.041
Tokenir
Token ID343
Feature activation-1.36e+00

Pos logit contributions
 lands+0.070
ening+0.067
 dot+0.061
fly+0.060
 landed+0.059
Neg logit contributions
igsaw-0.099
 lunch-0.097
 dinner-0.094
 math-0.091
 dessert-0.089
Tokenpy
Token ID9078
Feature activation-1.49e+00

Pos logit contributions
inkle+0.017
 dot+0.017
 wil+0.015
ays+0.015
ening+0.014
Neg logit contributions
 cream-0.051
 math-0.050
 subtract-0.049
 cakes-0.048
 pies-0.048
Token.
Token ID13
Feature activation-1.12e+00

Pos logit contributions
inkle+0.046
 dot+0.045
en+0.045
ening+0.044
el+0.044
Neg logit contributions
Deal-0.036
 math-0.035
 cream-0.035
 topp-0.033
 subt-0.033
Token They
Token ID1119
Feature activation-1.91e-01

Pos logit contributions
 wil+0.036
le+0.035
ening+0.034
en+0.033
inkle+0.033
Neg logit contributions
 math-0.028
 subt-0.027
 las-0.026
 Grad-0.026
 subtract-0.026
Token played
Token ID2826
Feature activation-9.91e-01

Pos logit contributions
ays+0.008
inkle+0.007
en+0.007
ening+0.007
el+0.007
Neg logit contributions
Deal-0.004
 y-0.003
 cream-0.003
 topp-0.003
 cakes-0.003
Token to
Token ID284
Feature activation+6.75e-01

Pos logit contributions
 subt+0.026
Deal+0.025
bek+0.024
 subtract+0.024
Fif+0.024
Neg logit contributions
 dot-0.018
fly-0.017
en-0.016
ga-0.016
el-0.016
Token play
Token ID711
Feature activation-9.14e-01

Pos logit contributions
 subt+0.013
 math+0.013
 cream+0.013
Deal+0.013
 topp+0.013
Neg logit contributions
en-0.025
 dot-0.025
el-0.024
fly-0.024
ening-0.024
Token in
Token ID287
Feature activation-1.08e+00

Pos logit contributions
en+0.025
 dot+0.025
el+0.024
fly+0.024
ug+0.024
Neg logit contributions
 subt-0.027
 math-0.026
 topp-0.026
 cream-0.026
 subtract-0.026
Token the
Token ID262
Feature activation-1.00e+00

Pos logit contributions
en+0.032
 dot+0.032
el+0.031
fly+0.031
ug+0.031
Neg logit contributions
 subt-0.028
 math-0.028
 cream-0.028
 topp-0.027
Deal-0.027
Token park
Token ID3952
Feature activation-8.06e-01

Pos logit contributions
 dot+0.028
en+0.028
el+0.028
fly+0.026
 wil+0.026
Neg logit contributions
 subt-0.029
Deal-0.028
 topp-0.028
 subtract-0.028
 cream-0.026
Token every
Token ID790
Feature activation-3.41e+00

Pos logit contributions
en+0.021
 dot+0.021
el+0.020
fly+0.020
ening+0.020
Neg logit contributions
 math-0.024
Deal-0.024
 cream-0.024
 subt-0.024
 topp-0.024
Token day
Token ID1110
Feature activation-1.29e+00

Pos logit contributions
en+0.131
 dot+0.129
fly+0.128
el+0.127
ug+0.127
Neg logit contributions
 cream-0.062
 math-0.061
 subt-0.058
Deal-0.056
 topp-0.056
Token.
Token ID13
Feature activation-1.50e+00

Pos logit contributions
fly+0.040
en+0.040
 dot+0.039
ley+0.038
py+0.038
Neg logit contributions
 math-0.034
 cream-0.032
 subt-0.032
 topp-0.031
Deal-0.030
Token True
Token ID6407
Feature activation-1.59e+00

Pos logit contributions
en+0.043
el+0.042
ug+0.042
 dot+0.041
fly+0.041
Neg logit contributions
 math-0.043
 subt-0.042
Deal-0.041
 topp-0.040
 cream-0.040
Token is
Token ID318
Feature activation-7.36e-01

Pos logit contributions
ir+0.059
ug+0.057
 dot+0.055
el+0.053
py+0.050
Neg logit contributions
Deal-0.039
 math-0.039
 cream-0.035
 pizza-0.033
 topp-0.033
Token a
Token ID257
Feature activation-4.18e-01

Pos logit contributions
en+0.023
 dot+0.023
el+0.023
fly+0.023
ug+0.022
Neg logit contributions
 subt-0.018
 math-0.018
 cream-0.018
 topp-0.018
Deal-0.018
Token to
Token ID284
Feature activation+8.60e-01

Pos logit contributions
Deal+0.041
 subt+0.041
bek+0.040
 subtract+0.038
Fif+0.038
Neg logit contributions
 dot-0.024
el-0.022
en-0.022
fly-0.021
ga-0.021
Token play
Token ID711
Feature activation-5.94e-01

Pos logit contributions
 subt+0.017
 math+0.017
 cream+0.017
 topp+0.016
Deal+0.016
Neg logit contributions
en-0.031
 dot-0.031
el-0.030
fly-0.030
ug-0.030
Token in
Token ID287
Feature activation-7.76e-01

Pos logit contributions
en+0.013
 dot+0.013
le+0.013
ug+0.013
el+0.013
Neg logit contributions
 subt-0.021
 subtract-0.020
 topp-0.019
Fif-0.019
Deal-0.019
Token the
Token ID262
Feature activation-7.35e-01

Pos logit contributions
en+0.021
 dot+0.021
el+0.021
fly+0.020
ug+0.020
Neg logit contributions
 subt-0.023
Deal-0.023
 topp-0.022
 cream-0.022
 math-0.022
Token park
Token ID3952
Feature activation-5.23e-01

Pos logit contributions
el+0.018
 dot+0.018
en+0.017
 wil+0.016
fly+0.016
Neg logit contributions
 subt-0.024
Deal-0.024
 subtract-0.024
 topp-0.023
bek-0.022
Token every
Token ID790
Feature activation-3.40e+00

Pos logit contributions
en+0.012
 dot+0.012
el+0.012
fly+0.012
ug+0.011
Neg logit contributions
 subt-0.018
 math-0.017
 cream-0.017
 topp-0.017
Deal-0.017
Token day
Token ID1110
Feature activation-1.09e+00

Pos logit contributions
en+0.125
 dot+0.124
fly+0.122
el+0.121
ug+0.120
Neg logit contributions
 cream-0.066
 math-0.066
 subt-0.065
Deal-0.062
 topp-0.062
Token.
Token ID13
Feature activation-7.63e-01

Pos logit contributions
fly+0.031
en+0.031
 dot+0.031
ley+0.030
py+0.030
Neg logit contributions
 math-0.030
 subt-0.030
 cream-0.030
 topp-0.029
Deal-0.028
Token They
Token ID1119
Feature activation-5.25e-01

Pos logit contributions
en+0.018
 dot+0.018
el+0.017
fly+0.017
le+0.016
Neg logit contributions
 subt-0.025
 cream-0.025
Deal-0.025
 math-0.025
 topp-0.024
Token swing
Token ID9628
Feature activation-8.56e-01

Pos logit contributions
el+0.018
en+0.018
 dot+0.018
ening+0.018
fly+0.018
Neg logit contributions
Deal-0.012
 math-0.011
 subt-0.011
 cream-0.011
 topp-0.011
Token,
Token ID11
Feature activation-1.20e+00

Pos logit contributions
 dot+0.026
ening+0.025
en+0.024
inkle+0.024
fly+0.024
Neg logit contributions
 cream-0.026
 math-0.025
Deal-0.024
 cakes-0.023
 pizza-0.022
Token
Token ID129
Feature activation+4.33e-01

Pos logit contributions
 subt+0.019
 cream+0.018
 math+0.018
 topp+0.018
Deal+0.018
Neg logit contributions
 dot-0.017
en-0.017
fly-0.017
el-0.017
ug-0.017
Token
Token ID241
Feature activation+4.61e-01

Pos logit contributions
 subt+0.012
 math+0.012
 cream+0.012
 subtract+0.012
Deal+0.012
Neg logit contributions
 dot-0.012
fly-0.012
en-0.012
el-0.012
ug-0.012
TokenDon
Token ID3987
Feature activation-7.78e-01

Pos logit contributions
 subt+0.011
 math+0.011
 cream+0.011
Deal+0.011
 topp+0.011
Neg logit contributions
en-0.015
 dot-0.015
el-0.015
fly-0.014
ug-0.014
Tokenâ
Token ID22940
Feature activation+3.07e-01

Pos logit contributions
 dot+0.016
en+0.016
fly+0.016
el+0.015
ening+0.015
Neg logit contributions
 subt-0.028
 math-0.028
 cream-0.027
Deal-0.027
 subtract-0.027
Token
Token ID26391
Feature activation+8.64e-01

Pos logit contributions
 math+0.008
 topp+0.008
 subt+0.008
 cakes+0.007
 cream+0.007
Neg logit contributions
en-0.010
el-0.009
le-0.009
ga-0.009
ug-0.009
Token
Token ID8151
Feature activation+3.64e+00

Pos logit contributions
 subt+0.025
 cream+0.025
 math+0.025
 topp+0.025
Deal+0.025
Neg logit contributions
 dot-0.023
fly-0.022
en-0.022
el-0.022
ug-0.022
Tokent
Token ID83
Feature activation+1.22e+00

Pos logit contributions
 math+0.101
 subt+0.101
Deal+0.100
 cream+0.100
 topp+0.098
Neg logit contributions
 dot-0.105
en-0.104
ening-0.099
le-0.097
fly-0.096
Token be
Token ID307
Feature activation+1.09e+00

Pos logit contributions
 math+0.030
 subtract+0.029
 subt+0.029
 cream+0.028
 job+0.028
Neg logit contributions
ays-0.038
 dot-0.038
fly-0.038
ley-0.037
el-0.037
Token scared
Token ID12008
Feature activation+1.14e+00

Pos logit contributions
 subt+0.030
 math+0.030
 cream+0.029
 subtract+0.029
 topp+0.028
Neg logit contributions
 dot-0.032
el-0.031
en-0.031
fly-0.031
ley-0.031
Token,
Token ID11
Feature activation+1.10e+00

Pos logit contributions
 math+0.030
 cream+0.029
 job+0.027
 subt+0.026
 subtract+0.026
Neg logit contributions
fly-0.036
 dot-0.036
inkle-0.035
 wil-0.035
ittered-0.034
Token I
Token ID314
Feature activation-4.19e-01

Pos logit contributions
 math+0.031
 subt+0.031
 cream+0.031
 subtract+0.029
 topp+0.029
Neg logit contributions
fly-0.031
en-0.031
el-0.031
 dot-0.031
ening-0.030
Token
Token ID129
Feature activation+5.43e-01

Pos logit contributions
 subt+0.019
 math+0.019
 cream+0.019
 topp+0.018
Deal+0.018
Neg logit contributions
en-0.021
 dot-0.021
el-0.020
fly-0.020
ug-0.020
Token
Token ID241
Feature activation+1.41e-01

Pos logit contributions
 subt+0.015
 math+0.014
 cream+0.014
Deal+0.014
 subtract+0.014
Neg logit contributions
 dot-0.016
fly-0.016
en-0.016
el-0.016
ug-0.015
TokenThat
Token ID2504
Feature activation+7.61e-01

Pos logit contributions
 subt+0.004
 math+0.003
 cream+0.003
 topp+0.003
Deal+0.003
Neg logit contributions
en-0.004
 dot-0.004
el-0.004
fly-0.004
ug-0.004
Tokenâ
Token ID22940
Feature activation-2.64e-01

Pos logit contributions
 math+0.022
 cream+0.022
 subt+0.022
 topp+0.022
 subtract+0.021
Neg logit contributions
ir-0.021
 dot-0.021
ug-0.021
fly-0.020
en-0.020
Token
Token ID26391
Feature activation+6.20e-01

Pos logit contributions
en+0.008
el+0.008
le+0.008
ug+0.008
ches+0.008
Neg logit contributions
 math-0.007
 subt-0.006
 topp-0.006
 cream-0.006
 subtract-0.006
Token
Token ID8151
Feature activation+3.58e+00

Pos logit contributions
 subt+0.017
 math+0.017
 cream+0.017
Deal+0.017
 topp+0.017
Neg logit contributions
en-0.018
 dot-0.018
el-0.017
fly-0.017
ug-0.017
Tokens
Token ID82
Feature activation+9.66e-01

Pos logit contributions
 cream+0.106
Deal+0.106
 math+0.106
 subt+0.105
 topp+0.104
Neg logit contributions
 dot-0.100
en-0.095
ening-0.090
fly-0.090
 predators-0.089
Token an
Token ID281
Feature activation+7.60e-01

Pos logit contributions
 math+0.030
Deal+0.030
 cream+0.030
 subt+0.029
 topp+0.029
Neg logit contributions
 dot-0.025
en-0.024
fly-0.024
ir-0.024
ug-0.023
Token elevator
Token ID20932
Feature activation+1.41e+00

Pos logit contributions
 math+0.017
 cream+0.016
 job+0.016
 subt+0.016
 list+0.016
Neg logit contributions
fly-0.028
 dot-0.026
ays-0.025
en-0.025
ches-0.025
Token.
Token ID13
Feature activation+5.33e-01

Pos logit contributions
 cream+0.045
 subt+0.045
 math+0.045
Deal+0.043
 topp+0.043
Neg logit contributions
 dot-0.035
en-0.034
el-0.033
ir-0.033
ug-0.033
Token It
Token ID632
Feature activation-8.29e-02

Pos logit contributions
 subt+0.015
 math+0.015
 subtract+0.015
Deal+0.014
 topp+0.014
Neg logit contributions
en-0.015
fly-0.015
 dot-0.015
 wil-0.015
le-0.015
Token on
Token ID319
Feature activation+5.38e-01

Pos logit contributions
 subt+0.033
 topp+0.033
 cream+0.033
 math+0.033
 subtract+0.033
Neg logit contributions
el-0.021
ug-0.021
en-0.020
 dot-0.020
ir-0.019
Token,
Token ID11
Feature activation+6.11e-01

Pos logit contributions
 math+0.017
 cream+0.016
 subt+0.016
 job+0.016
 pizza+0.015
Neg logit contributions
 dot-0.015
fly-0.015
en-0.014
 wil-0.014
ga-0.014
Token they
Token ID484
Feature activation+4.61e-01

Pos logit contributions
 subt+0.022
Deal+0.022
 subtract+0.021
bek+0.021
Fif+0.021
Neg logit contributions
en-0.012
 or-0.011
el-0.011
 dot-0.011
ug-0.010
Token never
Token ID1239
Feature activation-2.83e-01

Pos logit contributions
 subt+0.014
 math+0.013
 cream+0.013
Deal+0.013
 topp+0.013
Neg logit contributions
en-0.012
 dot-0.012
el-0.012
fly-0.012
ug-0.012
Token qu
Token ID627
Feature activation+1.77e+00

Pos logit contributions
en+0.008
 dot+0.007
fly+0.007
el+0.007
le+0.007
Neg logit contributions
 math-0.009
 subt-0.009
 cream-0.009
Deal-0.008
 topp-0.008
Tokenarre
Token ID9624
Feature activation+3.78e+00

Pos logit contributions
 subt+0.021
 math+0.020
 cream+0.020
 topp+0.020
 subtract+0.019
Neg logit contributions
 dot-0.078
en-0.077
fly-0.077
el-0.076
ley-0.076
Tokenled
Token ID992
Feature activation+5.93e-01

Pos logit contributions
 subt+0.160
 math+0.157
 cream+0.156
Deal+0.155
 topp+0.155
Neg logit contributions
en-0.056
 dot-0.053
el-0.052
ug-0.050
le-0.050
Token again
Token ID757
Feature activation+3.07e-01

Pos logit contributions
 math+0.018
 cream+0.017
Deal+0.016
 subt+0.016
 topp+0.016
Neg logit contributions
 dot-0.016
el-0.016
ley-0.016
ir-0.016
inkle-0.016
Token and
Token ID290
Feature activation+2.26e-01

Pos logit contributions
 subt+0.012
 math+0.011
 cream+0.011
 topp+0.011
Deal+0.011
Neg logit contributions
 dot-0.006
en-0.006
el-0.006
fly-0.006
ug-0.005
Token played
Token ID2826
Feature activation-5.63e-01

Pos logit contributions
 subt+0.007
 topp+0.007
 cream+0.007
 math+0.007
Deal+0.007
Neg logit contributions
en-0.006
fly-0.005
el-0.005
 dot-0.005
 wil-0.005
Token together
Token ID1978
Feature activation-7.05e-01

Pos logit contributions
en+0.015
le+0.014
 dot+0.014
ug+0.014
inkle+0.013
Neg logit contributions
 subt-0.017
 subtract-0.017
 topp-0.017
 las-0.016
Fif-0.016
Token thought
Token ID1807
Feature activation+6.16e-01

Pos logit contributions
 subt+0.015
 subtract+0.015
 math+0.014
 cream+0.014
Deal+0.013
Neg logit contributions
ir-0.024
fly-0.023
py-0.023
 predators-0.023
ays-0.023
Token they
Token ID484
Feature activation+4.84e-01

Pos logit contributions
 math+0.020
 subt+0.016
 cakes+0.016
 cream+0.015
 homework+0.015
Neg logit contributions
fly-0.021
ley-0.020
 dot-0.019
an-0.018
ga-0.018
Token looked
Token ID3114
Feature activation+1.15e+00

Pos logit contributions
Deal+0.012
 topp+0.011
 y+0.011
 subtract+0.011
 cream+0.011
Neg logit contributions
ug-0.016
ir-0.016
ga-0.016
inkle-0.015
ays-0.015
Token like
Token ID588
Feature activation+7.77e-01

Pos logit contributions
 subt+0.030
 math+0.030
 cream+0.029
 topp+0.029
Deal+0.029
Neg logit contributions
en-0.035
 dot-0.035
el-0.034
fly-0.034
ug-0.034
Token ac
Token ID936
Feature activation+3.40e-01

Pos logit contributions
 math+0.018
 subt+0.017
 cream+0.017
 topp+0.016
 subtract+0.016
Neg logit contributions
en-0.027
el-0.027
 dot-0.027
fly-0.026
ir-0.026
Tokenrob
Token ID22609
Feature activation+4.32e+00

Pos logit contributions
 cream+0.008
Deal+0.007
 subt+0.007
 math+0.007
 topp+0.007
Neg logit contributions
 dot-0.012
en-0.012
fly-0.012
el-0.011
ug-0.011
Tokenats
Token ID1381
Feature activation+5.80e-02

Pos logit contributions
 subt+0.146
 math+0.141
 cream+0.140
Deal+0.139
 subtract+0.139
Neg logit contributions
en-0.099
el-0.098
ir-0.096
ug-0.094
le-0.094
Token.
Token ID13
Feature activation+2.45e-01

Pos logit contributions
 cream+0.002
 math+0.002
 cakes+0.002
 pizza+0.002
 cookies+0.002
Neg logit contributions
inkle-0.002
 dot-0.002
ug-0.002
ir-0.002
el-0.001
Token She
Token ID1375
Feature activation+5.01e-01

Pos logit contributions
 subt+0.008
 cream+0.008
 math+0.008
 topp+0.008
Deal+0.008
Neg logit contributions
en-0.006
 dot-0.006
el-0.006
fly-0.006
ug-0.005
Token enjoyed
Token ID8359
Feature activation-2.63e-01

Pos logit contributions
 subt+0.012
 cream+0.011
 subtract+0.011
 math+0.011
 topp+0.011
Neg logit contributions
ir-0.017
fly-0.017
ug-0.017
ays-0.017
 dot-0.017
Token the
Token ID262
Feature activation+5.45e-01

Pos logit contributions
en+0.008
 dot+0.008
el+0.008
fly+0.007
ug+0.007
Neg logit contributions
 subt-0.007
 math-0.007
 cream-0.007
Deal-0.007
 topp-0.007
Token
Token ID129
Feature activation+8.28e-01

Pos logit contributions
 subt+0.022
 math+0.022
 cream+0.021
Deal+0.021
 topp+0.021
Neg logit contributions
en-0.026
 dot-0.026
el-0.026
fly-0.025
ening-0.025
Token
Token ID241
Feature activation+6.53e-01

Pos logit contributions
 subt+0.024
 math+0.023
 cream+0.023
 topp+0.023
Deal+0.023
Neg logit contributions
en-0.023
 dot-0.023
el-0.022
fly-0.022
ug-0.022
TokenThat
Token ID2504
Feature activation+8.96e-01

Pos logit contributions
 subt+0.016
 math+0.016
 cream+0.016
 topp+0.016
Deal+0.016
Neg logit contributions
en-0.020
 dot-0.020
el-0.020
fly-0.020
ug-0.019
Tokenâ
Token ID22940
Feature activation-2.07e-01

Pos logit contributions
 math+0.028
 cream+0.028
 subt+0.028
 topp+0.028
Deal+0.027
Neg logit contributions
 dot-0.022
en-0.022
ir-0.022
fly-0.022
el-0.022
Token
Token ID26391
Feature activation+7.10e-01

Pos logit contributions
en+0.007
el+0.006
le+0.006
ug+0.006
ening+0.006
Neg logit contributions
 math-0.005
 topp-0.005
 subt-0.005
 cream-0.005
 cakes-0.005
Token
Token ID8151
Feature activation+3.55e+00

Pos logit contributions
 subt+0.020
 math+0.020
 cream+0.020
Deal+0.019
 topp+0.019
Neg logit contributions
en-0.020
 dot-0.020
el-0.019
fly-0.019
ug-0.019
Tokens
Token ID82
Feature activation+1.18e+00

Pos logit contributions
Deal+0.106
 cream+0.104
 subt+0.103
 math+0.103
 topp+0.102
Neg logit contributions
 dot-0.099
en-0.095
fly-0.092
ening-0.091
py-0.090
Token the
Token ID262
Feature activation+4.58e-01

Pos logit contributions
 math+0.035
 subt+0.035
 cream+0.035
Deal+0.035
 topp+0.034
Neg logit contributions
 dot-0.032
en-0.031
fly-0.030
el-0.030
ir-0.030
Token bird
Token ID6512
Feature activation+1.17e+00

Pos logit contributions
 subt+0.011
 math+0.011
Deal+0.011
 topp+0.011
 cream+0.011
Neg logit contributions
 dot-0.014
en-0.014
el-0.014
fly-0.014
ug-0.013
Token that
Token ID326
Feature activation+9.47e-01

Pos logit contributions
 cream+0.034
Deal+0.033
 topp+0.032
 math+0.032
 y+0.031
Neg logit contributions
 dot-0.033
ga-0.031
ir-0.030
ug-0.030
en-0.030
Token was
Token ID373
Feature activation-2.46e-01

Pos logit contributions
 math+0.028
 cream+0.028
 subt+0.027
 topp+0.027
Deal+0.027
Neg logit contributions
 dot-0.025
en-0.025
ir-0.025
fly-0.025
el-0.025
Token to
Token ID284
Feature activation+1.54e+00

Pos logit contributions
 cream+0.014
 math+0.014
 subt+0.014
 topp+0.013
 job+0.013
Neg logit contributions
 dot-0.028
en-0.027
fly-0.027
ug-0.026
py-0.026
Token go
Token ID467
Feature activation+7.04e-01

Pos logit contributions
 subt+0.050
Deal+0.049
 topp+0.048
 math+0.047
 subtract+0.047
Neg logit contributions
en-0.042
el-0.041
ir-0.039
 dot-0.038
le-0.038
Token,
Token ID11
Feature activation+4.10e-01

Pos logit contributions
 cream+0.018
 math+0.017
 job+0.016
 pizza+0.015
 subt+0.015
Neg logit contributions
 dot-0.024
fly-0.023
ening-0.022
 wil-0.022
en-0.022
Token Lucy
Token ID22162
Feature activation-7.75e-02

Pos logit contributions
 subt+0.013
 topp+0.013
 subtract+0.013
Deal+0.013
 math+0.012
Neg logit contributions
en-0.010
el-0.010
 dot-0.010
fly-0.009
le-0.009
Token and
Token ID290
Feature activation+2.82e+00

Pos logit contributions
ug+0.002
 dot+0.002
ir+0.002
py+0.002
ening+0.002
Neg logit contributions
 cream-0.002
 subt-0.002
 topp-0.002
 hamb-0.002
 las-0.002
Token her
Token ID607
Feature activation+1.69e+00

Pos logit contributions
 subt+0.111
 topp+0.108
Deal+0.105
 subtract+0.104
Fif+0.103
Neg logit contributions
el-0.064
en-0.060
ie-0.054
le-0.052
i-0.050
Token Dad
Token ID17415
Feature activation+7.08e-01

Pos logit contributions
 subt+0.045
 subtract+0.042
Deal+0.042
 math+0.041
 topp+0.041
Neg logit contributions
el-0.053
en-0.051
 wil-0.050
 dot-0.049
pt-0.048
Token brought
Token ID3181
Feature activation+4.14e-01

Pos logit contributions
 cream+0.018
 subt+0.018
Deal+0.018
 topp+0.018
 math+0.017
Neg logit contributions
 dot-0.022
en-0.021
fly-0.021
ening-0.021
el-0.020
Token the
Token ID262
Feature activation-1.13e+00

Pos logit contributions
 subt+0.012
 cream+0.011
 math+0.011
 topp+0.011
Deal+0.011
Neg logit contributions
 dot-0.012
en-0.012
fly-0.012
el-0.011
ug-0.011
Token cooler
Token ID19346
Feature activation+6.81e-01

Pos logit contributions
 dot+0.044
el+0.044
en+0.044
fly+0.044
le+0.043
Neg logit contributions
 subt-0.020
 subtract-0.018
 math-0.017
 topp-0.016
Deal-0.016
Token back
Token ID736
Feature activation+1.09e-01

Pos logit contributions
 subt+0.022
 math+0.021
 topp+0.021
Deal+0.021
 subtract+0.021
Neg logit contributions
en-0.017
 dot-0.016
el-0.016
fly-0.016
ug-0.016
Token you
Token ID345
Feature activation+2.43e-01

Pos logit contributions
en+0.031
 dot+0.031
el+0.031
ug+0.030
fly+0.030
Neg logit contributions
 subt-0.040
Deal-0.039
 topp-0.038
 math-0.038
 cream-0.038
Token are
Token ID389
Feature activation+2.69e-01

Pos logit contributions
 topp+0.006
 cream+0.006
 subtract+0.006
Deal+0.006
 math+0.006
Neg logit contributions
el-0.008
 dot-0.008
en-0.008
ening-0.007
fly-0.007
Token using
Token ID1262
Feature activation-6.53e-02

Pos logit contributions
 subt+0.007
Deal+0.006
 topp+0.006
 math+0.006
 cream+0.006
Neg logit contributions
en-0.009
 dot-0.009
el-0.008
fly-0.008
ug-0.008
Token the
Token ID262
Feature activation-2.42e-01

Pos logit contributions
fly+0.002
 dot+0.002
py+0.002
le+0.002
ley+0.002
Neg logit contributions
 math-0.002
 cream-0.002
 subt-0.002
 topp-0.002
 subtract-0.001
Token right
Token ID826
Feature activation-6.30e-01

Pos logit contributions
en+0.008
 dot+0.008
el+0.007
fly+0.007
ug+0.007
Neg logit contributions
 subt-0.006
 math-0.006
 cream-0.006
 topp-0.006
Deal-0.006
Token numbers
Token ID3146
Feature activation+1.59e+00

Pos logit contributions
fly+0.020
 dot+0.020
ir+0.020
el+0.019
en+0.019
Neg logit contributions
 math-0.016
 cream-0.015
 topp-0.015
 subt-0.015
 subtract-0.014
Token.
Token ID13
Feature activation-5.20e-02

Pos logit contributions
 math+0.047
 cream+0.047
 subt+0.046
 topp+0.046
 subtract+0.045
Neg logit contributions
 dot-0.042
en-0.042
fly-0.041
el-0.041
ir-0.041
Token If
Token ID1002
Feature activation-1.15e+00

Pos logit contributions
 dot+0.001
en+0.001
el+0.001
ug+0.001
fly+0.001
Neg logit contributions
 cream-0.002
 subt-0.002
 math-0.002
 topp-0.002
Deal-0.002
Token the
Token ID262
Feature activation+7.75e-01

Pos logit contributions
en+0.033
 dot+0.033
el+0.032
fly+0.032
ug+0.032
Neg logit contributions
 subt-0.031
 math-0.031
 cream-0.031
 topp-0.030
Deal-0.030
Token numbers
Token ID3146
Feature activation+9.94e-01

Pos logit contributions
 subt+0.021
Deal+0.021
 topp+0.019
Fif+0.019
 cream+0.019
Neg logit contributions
 dot-0.023
el-0.021
en-0.021
fly-0.021
ug-0.021
Token match
Token ID2872
Feature activation+3.74e-01

Pos logit contributions
 math+0.029
 cream+0.029
 subtract+0.028
 topp+0.028
 subt+0.027
Neg logit contributions
 dot-0.027
en-0.027
py-0.027
fly-0.027
ley-0.026
Token.
Token ID13
Feature activation-3.88e-01

Pos logit contributions
en+0.003
 dot+0.003
el+0.002
fly+0.002
ug+0.002
Neg logit contributions
 subt-0.004
 math-0.004
 cream-0.004
 topp-0.003
Deal-0.003
Token<|endoftext|>
Token ID50256
Feature activation+1.09e-01

Pos logit contributions
el+0.009
 dot+0.009
en+0.009
ug+0.009
fly+0.009
Neg logit contributions
 subt-0.013
 cream-0.012
 math-0.012
 topp-0.012
Deal-0.012
TokenOne
Token ID3198
Feature activation-1.95e-01

Pos logit contributions
 math+0.003
 cream+0.002
 subt+0.002
 topp+0.002
Deal+0.002
Neg logit contributions
en-0.004
 dot-0.004
fly-0.003
ening-0.003
el-0.003
Token day
Token ID1110
Feature activation+3.47e-01

Pos logit contributions
el+0.005
ie+0.005
le+0.005
y+0.005
 dot+0.005
Neg logit contributions
 subt-0.007
 subtract-0.006
 topp-0.006
Deal-0.006
Fif-0.006
Token,
Token ID11
Feature activation-1.78e-01

Pos logit contributions
 subt+0.014
Deal+0.014
 subtract+0.014
 math+0.014
 topp+0.014
Neg logit contributions
el-0.005
en-0.005
 dot-0.005
ug-0.005
ir-0.005
Token a
Token ID257
Feature activation+1.31e+00

Pos logit contributions
en+0.006
 dot+0.006
ening+0.006
ga+0.006
fly+0.006
Neg logit contributions
 cream-0.004
 math-0.004
Deal-0.004
 topp-0.004
 subt-0.003
Token little
Token ID1310
Feature activation+2.49e-01

Pos logit contributions
 subt+0.046
 subtract+0.044
 topp+0.041
 math+0.039
Fif+0.038
Neg logit contributions
 dot-0.035
fly-0.031
ug-0.030
 hawk-0.029
le-0.029
Token boy
Token ID2933
Feature activation-6.03e-01

Pos logit contributions
 subt+0.007
 math+0.006
 cream+0.006
 topp+0.006
Deal+0.006
Neg logit contributions
en-0.008
 dot-0.008
el-0.007
fly-0.007
ug-0.007
Token named
Token ID3706
Feature activation-3.89e-01

Pos logit contributions
en+0.011
 dot+0.011
fly+0.011
el+0.010
ug+0.010
Neg logit contributions
 subt-0.023
 math-0.022
 cream-0.022
 topp-0.022
Deal-0.022
Token Tim
Token ID5045
Feature activation-1.38e+00

Pos logit contributions
el+0.008
pt+0.007
 dot+0.007
fly+0.007
ug+0.007
Neg logit contributions
 subt-0.015
 subtract-0.014
Deal-0.014
 cooperative-0.013
 math-0.013
Token went
Token ID1816
Feature activation-1.56e-01

Pos logit contributions
 dot+0.043
ug+0.042
fly+0.042
ala+0.042
ir+0.042
Neg logit contributions
 cream-0.035
 topp-0.032
Deal-0.032
 math-0.032
 las-0.031
Token it
Token ID340
Feature activation-2.42e-01

Pos logit contributions
 subt+0.013
Deal+0.013
 cream+0.013
 math+0.013
 topp+0.012
Neg logit contributions
en-0.013
 dot-0.012
el-0.012
ug-0.012
le-0.012
Token look
Token ID804
Feature activation-4.76e-02

Pos logit contributions
en+0.007
 dot+0.007
el+0.007
fly+0.007
le+0.007
Neg logit contributions
 subt-0.007
Deal-0.007
 topp-0.006
 cream-0.006
 math-0.006
Token nice
Token ID3621
Feature activation+2.68e-01

Pos logit contributions
en+0.002
 dot+0.002
el+0.001
le+0.001
ening+0.001
Neg logit contributions
 subt-0.001
Deal-0.001
 cream-0.001
 math-0.001
 topp-0.001
Token again
Token ID757
Feature activation-1.26e-01

Pos logit contributions
 subt+0.009
 math+0.009
 cream+0.009
Deal+0.009
 topp+0.009
Neg logit contributions
en-0.006
 dot-0.006
el-0.006
fly-0.006
ug-0.006
Token.
Token ID13
Feature activation+8.71e-01

Pos logit contributions
en+0.002
el+0.002
le+0.002
 dot+0.002
ening+0.002
Neg logit contributions
 subt-0.005
Deal-0.005
 subtract-0.005
 topp-0.005
Fif-0.005
Token
Token ID220
Feature activation+1.99e+00

Pos logit contributions
 cream+0.025
 topp+0.024
 subt+0.024
 math+0.024
Deal+0.024
Neg logit contributions
en-0.024
el-0.024
 dot-0.024
ug-0.023
fly-0.023
Token
Token ID198
Feature activation+8.91e-01

Pos logit contributions
 subt+0.059
 cream+0.058
 math+0.057
Deal+0.057
 topp+0.057
Neg logit contributions
en-0.052
 dot-0.052
el-0.051
ug-0.051
fly-0.050
TokenPaul
Token ID12041
Feature activation+6.42e-01

Pos logit contributions
 subt+0.023
 cream+0.023
 topp+0.023
 math+0.023
Deal+0.022
Neg logit contributions
en-0.027
 dot-0.027
el-0.026
ug-0.026
fly-0.026
Token was
Token ID373
Feature activation-1.11e-01

Pos logit contributions
 subt+0.018
 cream+0.017
 topp+0.017
 subtract+0.017
 math+0.017
Neg logit contributions
 dot-0.019
ir-0.018
fly-0.018
 predators-0.018
py-0.018
Token very
Token ID845
Feature activation+2.86e-01

Pos logit contributions
en+0.003
 dot+0.003
el+0.003
fly+0.003
ug+0.003
Neg logit contributions
 subt-0.003
 topp-0.003
Deal-0.003
 math-0.003
 cream-0.003
Token proud
Token ID6613
Feature activation+1.05e+00

Pos logit contributions
 topp+0.009
 subt+0.009
 subtract+0.009
 las+0.008
Deal+0.008
Neg logit contributions
 wil-0.007
inkle-0.006
ug-0.006
 trapped-0.006
ubby-0.006
Token
Token ID26391
Feature activation+5.52e-01

Pos logit contributions
en+0.010
le+0.010
el+0.010
fly+0.010
ug+0.010
Neg logit contributions
 math-0.008
 topp-0.008
 subt-0.008
 cream-0.007
 subtract-0.007
Token
Token ID198
Feature activation+1.21e-02

Pos logit contributions
 subt+0.015
 math+0.015
 cream+0.015
Deal+0.015
 topp+0.014
Neg logit contributions
en-0.016
 dot-0.016
el-0.016
fly-0.015
ug-0.015
Token
Token ID198
Feature activation-2.96e-02

Pos logit contributions
 subt+0.000
 math+0.000
 cream+0.000
Deal+0.000
 subtract+0.000
Neg logit contributions
 dot-0.000
en-0.000
fly-0.000
ening-0.000
el-0.000
TokenThe
Token ID464
Feature activation+7.54e-01

Pos logit contributions
en+0.001
ug+0.001
el+0.001
 dot+0.001
ga+0.001
Neg logit contributions
 topp-0.001
 math-0.001
 subt-0.001
 cream-0.001
Deal-0.001
Token man
Token ID582
Feature activation+1.24e+00

Pos logit contributions
 subt+0.021
 math+0.020
 topp+0.020
 cream+0.020
Deal+0.020
Neg logit contributions
 dot-0.022
en-0.022
el-0.021
fly-0.021
ug-0.021
Token and
Token ID290
Feature activation+1.93e+00

Pos logit contributions
 cream+0.030
 subt+0.029
 topp+0.029
Deal+0.028
 math+0.028
Neg logit contributions
 dot-0.041
en-0.038
ir-0.038
ays-0.038
ening-0.038
Token his
Token ID465
Feature activation+1.32e+00

Pos logit contributions
 topp+0.061
Deal+0.060
 subt+0.060
 math+0.059
 cream+0.058
Neg logit contributions
en-0.049
el-0.048
ley-0.047
 dot-0.046
ug-0.046
Token wife
Token ID3656
Feature activation+1.02e+00

Pos logit contributions
 subt+0.025
 topp+0.024
Deal+0.024
 math+0.024
 subtract+0.023
Neg logit contributions
en-0.049
 dot-0.048
el-0.048
ug-0.047
fly-0.047
Token agreed
Token ID4987
Feature activation-1.15e+00

Pos logit contributions
 cream+0.026
Deal+0.025
 subt+0.025
 topp+0.024
 cakes+0.024
Neg logit contributions
 dot-0.033
ir-0.031
ays-0.031
ening-0.031
en-0.030
Token and
Token ID290
Feature activation+2.73e-01

Pos logit contributions
 dot+0.032
fly+0.031
en+0.031
el+0.030
ir+0.030
Neg logit contributions
 cream-0.034
 math-0.033
 subt-0.033
 topp-0.032
Deal-0.032
Token so
Token ID523
Feature activation+5.26e-01

Pos logit contributions
 math+0.008
 subt+0.008
Deal+0.008
 topp+0.008
 cream+0.008
Neg logit contributions
en-0.007
el-0.007
ug-0.007
fly-0.007
 dot-0.007
Token their
Token ID511
Feature activation+1.33e+00

Pos logit contributions
 subt+0.016
 math+0.016
 cream+0.016
Deal+0.016
 topp+0.016
Neg logit contributions
 dot-0.017
en-0.017
fly-0.017
el-0.017
ug-0.016
Token mom
Token ID1995
Feature activation-5.55e-02

Pos logit contributions
 subt+0.037
Deal+0.037
 topp+0.036
 subtract+0.035
 math+0.035
Neg logit contributions
 dot-0.038
el-0.037
en-0.037
fly-0.036
 wil-0.035
Token,
Token ID11
Feature activation+1.38e-01

Pos logit contributions
 dot+0.001
ug+0.001
en+0.001
fly+0.001
 wil+0.001
Neg logit contributions
 subt-0.002
 math-0.002
 cream-0.002
 topp-0.002
 subtract-0.002
Token who
Token ID508
Feature activation+1.09e+00

Pos logit contributions
 cream+0.003
 math+0.003
Deal+0.003
 subt+0.003
 cakes+0.003
Neg logit contributions
 dot-0.005
en-0.005
ug-0.005
ir-0.005
inkle-0.005
Token is
Token ID318
Feature activation-2.20e-02

Pos logit contributions
 cream+0.038
 subt+0.038
 math+0.038
Deal+0.038
 topp+0.037
Neg logit contributions
 dot-0.023
en-0.023
fly-0.022
ug-0.022
el-0.022
Token in
Token ID287
Feature activation+2.49e-01

Pos logit contributions
 dot+0.001
en+0.001
fly+0.001
ley+0.001
ug+0.001
Neg logit contributions
 cream-0.001
 math-0.001
 subt-0.001
 cakes-0.001
 topp-0.001
Token the
Token ID262
Feature activation+7.23e-01

Pos logit contributions
 math+0.006
 subt+0.006
 cream+0.006
 topp+0.006
 subtract+0.006
Neg logit contributions
en-0.008
 dot-0.008
fly-0.008
ug-0.008
el-0.008
Token kitchen
Token ID9592
Feature activation+1.35e+00

Pos logit contributions
 subt+0.022
 subtract+0.021
 topp+0.021
Deal+0.021
Fif+0.020
Neg logit contributions
el-0.019
 wil-0.019
 dot-0.019
 wing-0.017
ga-0.017
Token.
Token ID13
Feature activation-1.86e-01

Pos logit contributions
 cream+0.041
 math+0.041
 subt+0.040
 topp+0.040
Deal+0.039
Neg logit contributions
 dot-0.035
en-0.035
fly-0.034
ug-0.034
el-0.034
Token
Token ID198
Feature activation+4.20e-02

Pos logit contributions
en+0.006
le+0.006
 dot+0.006
ug+0.006
fly+0.006
Neg logit contributions
 math-0.004
 subt-0.004
Deal-0.004
 cream-0.004
 subtract-0.004
Token
Token ID198
Feature activation-3.57e-01

Pos logit contributions
 math+0.001
 cream+0.001
 subt+0.001
 pizza+0.001
 subtract+0.001
Neg logit contributions
en-0.002
ening-0.002
 dot-0.002
fly-0.002
le-0.002
Token Lucy
Token ID22162
Feature activation-1.61e+00

Pos logit contributions
en+0.003
 dot+0.003
el+0.003
fly+0.003
ir+0.003
Neg logit contributions
 subt-0.003
 topp-0.003
 cream-0.003
 math-0.003
Deal-0.003
Token was
Token ID373
Feature activation-3.69e-01

Pos logit contributions
ug+0.052
 dot+0.051
ir+0.051
fly+0.050
ches+0.050
Neg logit contributions
 cream-0.038
Deal-0.036
 las-0.035
 topp-0.034
 subt-0.034
Token playing
Token ID2712
Feature activation-2.62e-01

Pos logit contributions
el+0.009
ug+0.009
en+0.009
 dot+0.009
ga+0.009
Neg logit contributions
 topp-0.012
 subt-0.011
Deal-0.011
 subtract-0.011
 cream-0.011
Token in
Token ID287
Feature activation-8.59e-01

Pos logit contributions
 dot+0.004
le+0.004
en+0.004
bed+0.003
el+0.003
Neg logit contributions
 subt-0.012
 topp-0.011
 subtract-0.011
bek-0.011
 hamb-0.010
Token her
Token ID607
Feature activation-8.18e-01

Pos logit contributions
en+0.023
el+0.023
 dot+0.023
ir+0.021
fly+0.021
Neg logit contributions
Deal-0.026
 subt-0.026
 topp-0.025
 cream-0.024
 subtract-0.024
Token yard
Token ID12699
Feature activation-4.55e-01

Pos logit contributions
 dot+0.020
el+0.020
en+0.018
 wil+0.018
 wing+0.018
Neg logit contributions
 topp-0.028
 subt-0.028
Deal-0.027
 subtract-0.027
bek-0.025
Token.
Token ID13
Feature activation-1.08e-01

Pos logit contributions
en+0.011
 dot+0.010
el+0.010
fly+0.010
ug+0.010
Neg logit contributions
 subt-0.015
 math-0.015
 topp-0.015
Deal-0.015
 cream-0.015
Token She
Token ID1375
Feature activation+2.09e-01

Pos logit contributions
en+0.003
 dot+0.003
fly+0.003
ug+0.003
el+0.003
Neg logit contributions
 math-0.003
 subt-0.003
 cream-0.003
Deal-0.003
 topp-0.003
Token saw
Token ID2497
Feature activation+1.14e+00

Pos logit contributions
 cream+0.005
Deal+0.004
 math+0.004
 subt+0.004
 topp+0.004
Neg logit contributions
ir-0.008
en-0.008
fly-0.008
 dot-0.007
ening-0.007
Token a
Token ID257
Feature activation-3.71e-01

Pos logit contributions
 math+0.024
 subt+0.024
 cream+0.024
 topp+0.023
Deal+0.023
Neg logit contributions
en-0.040
 dot-0.040
fly-0.039
el-0.039
ug-0.038
Token small
Token ID1402
Feature activation-7.00e-01

Pos logit contributions
 dot+0.011
fly+0.010
en+0.010
el+0.010
 wil+0.010
Neg logit contributions
 subt-0.010
Deal-0.010
 subtract-0.010
 topp-0.010
 math-0.010
Token big
Token ID1263
Feature activation-6.27e-01

Pos logit contributions
en+0.023
el+0.023
 dot+0.023
ug+0.022
ir+0.022
Neg logit contributions
 math-0.018
 cream-0.018
 subt-0.018
 topp-0.017
Deal-0.017
Token cave
Token ID11527
Feature activation-7.05e-01

Pos logit contributions
ays+0.018
ala+0.018
ug+0.018
ga+0.018
en+0.018
Neg logit contributions
 math-0.018
 cream-0.018
 job-0.017
 pizza-0.017
 topp-0.016
Token in
Token ID287
Feature activation-1.59e+00

Pos logit contributions
en+0.017
 dot+0.016
el+0.016
ug+0.016
fly+0.016
Neg logit contributions
 cream-0.023
 math-0.023
 subt-0.023
 topp-0.022
Deal-0.022
Token the
Token ID262
Feature activation-1.23e+00

Pos logit contributions
 dot+0.050
ug+0.050
en+0.050
fly+0.050
el+0.049
Neg logit contributions
 math-0.039
 cream-0.038
 subt-0.038
 topp-0.037
 subtract-0.036
Token distance
Token ID5253
Feature activation+2.31e-01

Pos logit contributions
en+0.028
 dot+0.027
el+0.027
ug+0.027
fly+0.027
Neg logit contributions
 math-0.041
 subt-0.041
 cream-0.041
Deal-0.040
 topp-0.040
Token.
Token ID13
Feature activation-4.83e-01

Pos logit contributions
 cream+0.007
 math+0.007
 subt+0.007
 topp+0.007
 subtract+0.007
Neg logit contributions
 dot-0.006
en-0.006
fly-0.006
ug-0.006
el-0.006
Token He
Token ID679
Feature activation+6.42e-02

Pos logit contributions
 dot+0.013
en+0.013
el+0.013
fly+0.012
ug+0.012
Neg logit contributions
 subt-0.014
 cream-0.014
Deal-0.014
 math-0.014
 topp-0.014
Token became
Token ID2627
Feature activation-1.88e-01

Pos logit contributions
 subt+0.002
 cream+0.002
 math+0.002
Deal+0.002
 topp+0.002
Neg logit contributions
 dot-0.002
en-0.002
el-0.002
ir-0.002
fly-0.002
Token curious
Token ID11040
Feature activation+1.81e+00

Pos logit contributions
 dot+0.004
el+0.004
ug+0.004
fly+0.004
 wil+0.004
Neg logit contributions
 subt-0.006
 topp-0.006
 subtract-0.006
 cream-0.006
Deal-0.006
Token and
Token ID290
Feature activation+3.27e-02

Pos logit contributions
 math+0.057
 cream+0.054
 topp+0.054
 pizza+0.052
 cakes+0.052
Neg logit contributions
inkle-0.049
fly-0.049
ir-0.049
ga-0.048
ink-0.046
Token wanted
Token ID2227
Feature activation-7.03e-01

Pos logit contributions
 subt+0.001
 topp+0.001
 math+0.001
 cream+0.001
 subtract+0.001
Neg logit contributions
en-0.001
ie-0.001
ley-0.001
 landed-0.001
el-0.001
Token castle
Token ID16669
Feature activation+1.28e+00

Pos logit contributions
 dot+0.003
fly+0.003
en+0.003
el+0.003
 wil+0.003
Neg logit contributions
 subt-0.004
Deal-0.004
 subtract-0.004
 topp-0.004
 cream-0.004
Token and
Token ID290
Feature activation-3.62e-01

Pos logit contributions
 subt+0.049
 math+0.048
 cream+0.048
 topp+0.048
Deal+0.048
Neg logit contributions
 dot-0.023
en-0.023
el-0.022
fly-0.021
ug-0.021
Token told
Token ID1297
Feature activation-5.00e-01

Pos logit contributions
fly+0.009
en+0.009
 dot+0.009
el+0.008
ley+0.008
Neg logit contributions
 subt-0.012
 math-0.011
Deal-0.011
 topp-0.011
 cream-0.011
Token the
Token ID262
Feature activation-1.20e-02

Pos logit contributions
en+0.011
el+0.011
 dot+0.011
le+0.010
ir+0.010
Neg logit contributions
 subt-0.017
Deal-0.017
 cream-0.017
 topp-0.017
 subtract-0.016
Token knight
Token ID22062
Feature activation+2.17e-01

Pos logit contributions
 dot+0.000
en+0.000
fly+0.000
py+0.000
le+0.000
Neg logit contributions
 subt-0.000
Deal-0.000
 cream-0.000
 topp-0.000
 subtract-0.000
Token that
Token ID326
Feature activation+8.27e-01

Pos logit contributions
 subt+0.006
 math+0.006
 cream+0.006
Deal+0.006
 topp+0.006
Neg logit contributions
en-0.006
 dot-0.006
el-0.006
fly-0.006
ug-0.006
Token he
Token ID339
Feature activation-1.94e-01

Pos logit contributions
 subt+0.021
Deal+0.021
 cream+0.020
 topp+0.020
 math+0.020
Neg logit contributions
en-0.025
 dot-0.025
fly-0.024
el-0.024
ug-0.024
Token needed
Token ID2622
Feature activation-1.04e+00

Pos logit contributions
ir+0.007
ays+0.007
 dot+0.006
ga+0.006
en+0.006
Neg logit contributions
Deal-0.004
 topp-0.004
 math-0.004
 cream-0.004
 y-0.004
Token to
Token ID284
Feature activation+9.03e-01

Pos logit contributions
en+0.039
 dot+0.039
fly+0.038
el+0.038
ug+0.037
Neg logit contributions
 math-0.020
 subt-0.020
 cream-0.020
 topp-0.019
 subtract-0.019
Token prepare
Token ID8335
Feature activation+1.23e-01

Pos logit contributions
 subt+0.020
 math+0.019
 cream+0.019
Deal+0.019
 topp+0.019
Neg logit contributions
en-0.031
 dot-0.031
el-0.030
fly-0.030
ug-0.030
Token better
Token ID1365
Feature activation-7.46e-02

Pos logit contributions
 subt+0.003
 math+0.003
 cream+0.003
 topp+0.003
Deal+0.003
Neg logit contributions
 dot-0.004
en-0.004
fly-0.004
el-0.004
ug-0.004
Token.
Token ID13
Feature activation-4.91e-01

Pos logit contributions
 subt+0.007
 math+0.007
 cream+0.007
 topp+0.007
Deal+0.007
Neg logit contributions
en-0.005
 dot-0.005
el-0.004
fly-0.004
ug-0.004
Token
Token ID198
Feature activation-2.15e-01

Pos logit contributions
 dot+0.012
el+0.011
en+0.011
fly+0.011
ug+0.011
Neg logit contributions
 cream-0.016
 subt-0.016
 topp-0.015
Deal-0.015
 math-0.015
Token
Token ID198
Feature activation-6.29e-01

Pos logit contributions
en+0.008
 dot+0.008
ening+0.008
fly+0.008
ley+0.008
Neg logit contributions
 math-0.004
 cream-0.004
 subt-0.004
Deal-0.004
 subtract-0.004
Token"
Token ID1
Feature activation-1.52e-01

Pos logit contributions
ir+0.016
el+0.016
ug+0.016
en+0.016
 dot+0.015
Neg logit contributions
 topp-0.019
 subt-0.018
 math-0.018
 cream-0.018
 subtract-0.018
TokenCan
Token ID6090
Feature activation-4.29e-01

Pos logit contributions
 dot+0.004
le+0.004
ga+0.004
el+0.004
 wil+0.004
Neg logit contributions
 math-0.004
 job-0.004
 list-0.004
 subt-0.004
 topp-0.004
Token we
Token ID356
Feature activation-5.13e-01

Pos logit contributions
en+0.013
 dot+0.012
ening+0.012
fly+0.012
el+0.012
Neg logit contributions
 math-0.012
 subt-0.012
Deal-0.012
 cream-0.012
 topp-0.011
Token still
Token ID991
Feature activation-5.80e-01

Pos logit contributions
en+0.017
 dot+0.017
el+0.017
fly+0.016
ug+0.016
Neg logit contributions
 subt-0.012
 topp-0.012
 cream-0.012
 math-0.012
Deal-0.012
Token play
Token ID711
Feature activation-3.51e-01

Pos logit contributions
ening+0.021
ley+0.021
ches+0.021
 dot+0.021
en+0.021
Neg logit contributions
 math-0.011
 cream-0.011
 subtract-0.011
 subt-0.011
Deal-0.011
Token with
Token ID351
Feature activation+3.94e-01

Pos logit contributions
en+0.010
 dot+0.010
el+0.010
fly+0.010
ening+0.010
Neg logit contributions
 math-0.010
 subt-0.009
 cream-0.009
Deal-0.009
 topp-0.009
Token the
Token ID262
Feature activation-8.19e-01

Pos logit contributions
 math+0.011
 cream+0.010
 topp+0.010
 cakes+0.010
 pizza+0.010
Neg logit contributions
fly-0.013
 dot-0.013
py-0.012
en-0.012
ga-0.012
Token blocks
Token ID7021
Feature activation+1.05e+00

Pos logit contributions
 dot+0.025
en+0.024
el+0.024
fly+0.023
ening+0.023
Neg logit contributions
 subt-0.022
Deal-0.021
 subtract-0.020
 topp-0.020
 cream-0.020
Token girl
Token ID2576
Feature activation+6.45e-01

Pos logit contributions
 subt+0.037
 math+0.036
 cream+0.035
Deal+0.035
 topp+0.035
Neg logit contributions
en-0.046
 dot-0.046
el-0.045
fly-0.045
ug-0.044
Token trusted
Token ID13467
Feature activation-9.41e-01

Pos logit contributions
 cream+0.012
Deal+0.012
 topp+0.012
 hamb+0.011
 y+0.011
Neg logit contributions
ir-0.024
ays-0.024
 dot-0.024
ening-0.023
ala-0.023
Token the
Token ID262
Feature activation+5.51e-01

Pos logit contributions
en+0.026
el+0.025
fly+0.024
 dot+0.024
ir+0.024
Neg logit contributions
 subt-0.028
 topp-0.027
Deal-0.027
 cream-0.027
 math-0.027
Token sailor
Token ID43272
Feature activation+7.17e-01

Pos logit contributions
 subt+0.017
 math+0.017
Deal+0.017
 topp+0.017
 cream+0.017
Neg logit contributions
 dot-0.014
en-0.014
el-0.014
fly-0.013
ug-0.013
Token and
Token ID290
Feature activation+1.90e-01

Pos logit contributions
 math+0.020
 cream+0.020
 job+0.019
 pizza+0.019
Deal+0.019
Neg logit contributions
 dot-0.021
ays-0.020
ir-0.020
el-0.019
ug-0.019
Token felt
Token ID2936
Feature activation-2.00e+00

Pos logit contributions
 subt+0.006
 math+0.006
 topp+0.006
 subtract+0.006
Deal+0.006
Neg logit contributions
en-0.005
fly-0.005
ley-0.004
 dot-0.004
el-0.004
Token safe
Token ID3338
Feature activation-1.37e+00

Pos logit contributions
en+0.067
 dot+0.066
el+0.065
fly+0.065
ug+0.064
Neg logit contributions
 subt-0.046
 math-0.045
 cream-0.045
Deal-0.045
 topp-0.044
Token with
Token ID351
Feature activation+1.90e-01

Pos logit contributions
 dot+0.035
en+0.035
el+0.034
ir+0.034
ug+0.033
Neg logit contributions
 cream-0.042
 math-0.041
 subt-0.040
Deal-0.040
 topp-0.040
Token him
Token ID683
Feature activation+3.69e-01

Pos logit contributions
 math+0.004
 cream+0.004
 subt+0.004
 topp+0.004
Deal+0.004
Neg logit contributions
 dot-0.007
en-0.006
fly-0.006
ug-0.006
el-0.006
Token.
Token ID13
Feature activation+3.01e-01

Pos logit contributions
 math+0.010
 cream+0.010
Deal+0.010
 topp+0.010
 subt+0.010
Neg logit contributions
en-0.010
 dot-0.010
el-0.010
fly-0.010
ir-0.010
Token They
Token ID1119
Feature activation+2.31e-01

Pos logit contributions
 subt+0.009
 math+0.009
 cream+0.009
Deal+0.009
 topp+0.009
Neg logit contributions
en-0.008
 dot-0.007
el-0.007
fly-0.007
ug-0.007
Token she
Token ID673
Feature activation-6.35e-01

Pos logit contributions
en+0.002
 dot+0.002
fly+0.001
el+0.001
le+0.001
Neg logit contributions
 math-0.001
 subt-0.001
Deal-0.001
 cream-0.001
 subtract-0.001
Token wouldn
Token ID3636
Feature activation-1.86e+00

Pos logit contributions
ir+0.023
ays+0.022
 dot+0.022
ening+0.022
el+0.022
Neg logit contributions
 cream-0.015
 y-0.015
Deal-0.014
 topp-0.013
 treat-0.013
Token't
Token ID470
Feature activation-3.59e-01

Pos logit contributions
ening+0.070
ley+0.067
fly+0.067
 wil+0.066
maid+0.065
Neg logit contributions
 pay-0.032
 earn-0.031
 math-0.030
 job-0.029
 subtract-0.028
Token scare
Token ID19437
Feature activation+2.73e-01

Pos logit contributions
ays+0.014
el+0.013
ening+0.013
ink+0.013
ug+0.013
Neg logit contributions
 pay-0.006
 listen-0.006
 job-0.006
 cream-0.006
 treat-0.006
Token it
Token ID340
Feature activation-4.80e-01

Pos logit contributions
 math+0.007
 cream+0.007
 job+0.006
 subt+0.006
Fif+0.006
Neg logit contributions
 dot-0.008
inkle-0.008
ink-0.008
fly-0.008
ittered-0.008
Token away
Token ID1497
Feature activation-1.69e+00

Pos logit contributions
ink+0.015
ley+0.015
 dot+0.015
inkle+0.015
ug+0.015
Neg logit contributions
 math-0.011
 cream-0.011
 job-0.010
 y-0.010
 subtract-0.009
Token.
Token ID13
Feature activation-3.52e-01

Pos logit contributions
 dot+0.047
en+0.046
el+0.046
ug+0.045
ir+0.045
Neg logit contributions
 math-0.048
 cream-0.047
 subt-0.046
Deal-0.044
 topp-0.044
Token She
Token ID1375
Feature activation+4.24e-02

Pos logit contributions
en+0.009
el+0.009
 dot+0.009
fly+0.009
ir+0.009
Neg logit contributions
 subt-0.011
 cream-0.011
 topp-0.010
 math-0.010
Deal-0.010
Token saw
Token ID2497
Feature activation+1.05e+00

Pos logit contributions
 cream+0.001
 topp+0.001
Deal+0.001
 y+0.001
 subt+0.001
Neg logit contributions
ays-0.002
 dot-0.002
ening-0.002
ir-0.001
inkle-0.001
Token the
Token ID262
Feature activation-6.43e-01

Pos logit contributions
 math+0.020
 cream+0.019
 topp+0.017
 cakes+0.017
 subt+0.017
Neg logit contributions
inkle-0.041
en-0.041
ug-0.040
 dot-0.040
ches-0.040
Token bird
Token ID6512
Feature activation+5.29e-01

Pos logit contributions
en+0.018
ug+0.018
ches+0.018
ors+0.017
ley+0.017
Neg logit contributions
 cream-0.020
 math-0.019
 topp-0.019
 job-0.018
 cakes-0.018
Token them
Token ID606
Feature activation+8.87e-01

Pos logit contributions
en+0.002
 dot+0.002
el+0.002
fly+0.002
ug+0.002
Neg logit contributions
 subt-0.003
 math-0.003
 cream-0.003
 topp-0.003
Deal-0.003
Token.
Token ID13
Feature activation+4.86e-02

Pos logit contributions
 math+0.024
 cream+0.024
 subt+0.023
Deal+0.023
 topp+0.023
Neg logit contributions
 dot-0.026
en-0.026
fly-0.025
ug-0.025
inkle-0.025
Token She
Token ID1375
Feature activation-7.04e-01

Pos logit contributions
 subt+0.002
 math+0.002
 cream+0.002
 topp+0.002
Deal+0.002
Neg logit contributions
en-0.001
 dot-0.001
el-0.001
fly-0.001
ug-0.001
Token wished
Token ID16555
Feature activation-1.23e+00

Pos logit contributions
 dot+0.021
en+0.021
el+0.021
ug+0.021
fly+0.020
Neg logit contributions
 subt-0.019
 cream-0.019
Deal-0.018
 math-0.018
 topp-0.018
Token she
Token ID673
Feature activation-1.35e+00

Pos logit contributions
 dot+0.042
en+0.041
fly+0.041
el+0.040
ug+0.040
Neg logit contributions
 math-0.029
 subt-0.029
 cream-0.028
Deal-0.027
 subtract-0.027
Token could
Token ID714
Feature activation-1.74e+00

Pos logit contributions
 dot+0.037
en+0.037
ir+0.037
ening+0.035
el+0.035
Neg logit contributions
Deal-0.043
 cream-0.042
 subt-0.039
 math-0.039
 topp-0.039
Token let
Token ID1309
Feature activation-4.58e-01

Pos logit contributions
ening+0.064
 dot+0.063
en+0.063
el+0.061
ug+0.061
Neg logit contributions
 cream-0.037
Deal-0.036
 math-0.035
 subt-0.033
 job-0.033
Token them
Token ID606
Feature activation+1.61e-01

Pos logit contributions
en+0.009
 dot+0.008
el+0.008
fly+0.008
ug+0.008
Neg logit contributions
 subt-0.017
 math-0.017
 cream-0.017
Deal-0.017
 topp-0.017
Token go
Token ID467
Feature activation-5.67e-01

Pos logit contributions
 subt+0.005
 topp+0.004
 math+0.004
Deal+0.004
 subtract+0.004
Neg logit contributions
en-0.005
 dot-0.005
el-0.005
le-0.005
fly-0.005
Token home
Token ID1363
Feature activation+1.85e-01

Pos logit contributions
 dot+0.020
ening+0.020
fly+0.019
 wil+0.019
inkle+0.019
Neg logit contributions
 cream-0.013
 math-0.012
 job-0.011
 pay-0.011
 pizza-0.011
Token.
Token ID13
Feature activation-1.20e-01

Pos logit contributions
 subt+0.006
 math+0.006
 cream+0.006
Deal+0.006
 topp+0.005
Neg logit contributions
 dot-0.005
en-0.005
el-0.005
fly-0.005
ug-0.004
Token help
Token ID1037
Feature activation-6.55e-01

Pos logit contributions
 dot+0.015
ening+0.014
en+0.014
el+0.014
fly+0.014
Neg logit contributions
 cream-0.010
 math-0.010
 subt-0.009
Deal-0.009
 topp-0.009
Token.
Token ID13
Feature activation-3.28e-01

Pos logit contributions
 dot+0.020
fly+0.019
en+0.019
 wil+0.019
ening+0.019
Neg logit contributions
 math-0.018
 cream-0.017
 subt-0.017
 subtract-0.017
 topp-0.017
Token He
Token ID679
Feature activation+6.37e-02

Pos logit contributions
en+0.008
 dot+0.008
el+0.008
ug+0.008
fly+0.008
Neg logit contributions
 cream-0.010
Deal-0.010
 subt-0.010
 topp-0.010
 math-0.010
Token found
Token ID1043
Feature activation+4.81e-02

Pos logit contributions
 cream+0.002
 subt+0.002
 topp+0.002
Deal+0.001
 math+0.001
Neg logit contributions
 dot-0.002
ir-0.002
el-0.002
en-0.002
ening-0.002
Token a
Token ID257
Feature activation-9.16e-01

Pos logit contributions
 cream+0.001
 math+0.001
 topp+0.001
 subt+0.001
 subtract+0.001
Neg logit contributions
 dot-0.002
en-0.002
fly-0.002
el-0.002
ir-0.002
Token way
Token ID835
Feature activation-2.50e+00

Pos logit contributions
en+0.028
 dot+0.028
el+0.028
fly+0.027
ug+0.027
Neg logit contributions
 subt-0.023
 math-0.023
 cream-0.023
 topp-0.022
Deal-0.022
Token to
Token ID284
Feature activation-1.44e-01

Pos logit contributions
inkle+0.111
ink+0.108
le+0.107
 dot+0.105
fly+0.103
Neg logit contributions
 math-0.035
 cream-0.032
 job-0.028
 pizza-0.025
 subtract-0.023
Token clean
Token ID3424
Feature activation-5.44e-01

Pos logit contributions
en+0.005
 dot+0.005
el+0.005
fly+0.004
ug+0.004
Neg logit contributions
 subt-0.003
 math-0.003
 cream-0.003
Deal-0.003
 topp-0.003
Token the
Token ID262
Feature activation-1.03e+00

Pos logit contributions
fly+0.016
 dot+0.016
py+0.015
ug+0.015
 wil+0.015
Neg logit contributions
 cream-0.017
 math-0.016
 topp-0.015
 cakes-0.015
 pizza-0.015
Token p
Token ID279
Feature activation+1.90e-01

Pos logit contributions
ug+0.033
en+0.033
fly+0.033
ley+0.032
el+0.032
Neg logit contributions
 cream-0.026
 math-0.025
 topp-0.024
 subt-0.024
 cakes-0.023
Tokenuddle
Token ID24500
Feature activation+2.26e-02

Pos logit contributions
 cream+0.006
 math+0.005
 subt+0.005
Deal+0.005
 topp+0.005
Neg logit contributions
 dot-0.005
en-0.005
fly-0.005
ening-0.005
le-0.005
Token it
Token ID340
Feature activation-5.87e-01

Pos logit contributions
fly+0.004
en+0.004
ley+0.004
ga+0.004
 dot+0.004
Neg logit contributions
 math-0.005
 cream-0.005
 cakes-0.005
 subt-0.004
 topp-0.004
Token was
Token ID373
Feature activation-1.24e+00

Pos logit contributions
ir+0.017
ug+0.017
en+0.017
 dot+0.017
ening+0.016
Neg logit contributions
 cream-0.016
Deal-0.016
 subt-0.015
 math-0.015
 topp-0.015
Token very
Token ID845
Feature activation-4.83e-01

Pos logit contributions
 dot+0.048
le+0.047
ug+0.047
fly+0.047
en+0.046
Neg logit contributions
 cream-0.027
 math-0.026
 y-0.025
 cakes-0.023
 topp-0.022
Token soft
Token ID2705
Feature activation-7.50e-01

Pos logit contributions
el+0.014
 dot+0.014
 wil+0.014
en+0.014
ug+0.014
Neg logit contributions
 subt-0.013
 subtract-0.012
 topp-0.012
 math-0.012
Deal-0.012
Token and
Token ID290
Feature activation-1.63e-01

Pos logit contributions
ir+0.020
ug+0.019
inkle+0.019
fly+0.019
ala+0.018
Neg logit contributions
 cream-0.025
 math-0.024
 pizza-0.023
 cakes-0.023
 topp-0.022
Token fluffy
Token ID39145
Feature activation-1.62e+00

Pos logit contributions
en+0.005
 dot+0.005
el+0.005
fly+0.005
ug+0.005
Neg logit contributions
 subt-0.004
 cream-0.004
 math-0.004
Deal-0.004
 topp-0.004
Token.
Token ID13
Feature activation+3.83e-01

Pos logit contributions
ug+0.060
ir+0.059
inkle+0.058
fly+0.056
ala+0.055
Neg logit contributions
 cream-0.035
 math-0.033
 pizza-0.032
 dessert-0.031
 cakes-0.031
Token
Token ID220
Feature activation+1.24e+00

Pos logit contributions
 subt+0.011
 cream+0.011
 math+0.011
 topp+0.011
Deal+0.011
Neg logit contributions
en-0.010
 dot-0.010
el-0.010
fly-0.010
ug-0.010
Token
Token ID198
Feature activation+3.51e-01

Pos logit contributions
 math+0.025
Deal+0.024
 cream+0.023
 subt+0.023
 subtract+0.023
Neg logit contributions
en-0.046
 dot-0.046
ening-0.045
fly-0.045
ley-0.044
TokenEvery
Token ID6109
Feature activation+2.63e-01

Pos logit contributions
 math+0.007
Deal+0.007
 subt+0.006
 cream+0.006
 subtract+0.006
Neg logit contributions
 dot-0.013
fly-0.013
ening-0.013
 wil-0.013
en-0.013
Token day
Token ID1110
Feature activation+1.48e-01

Pos logit contributions
 math+0.004
 cream+0.004
Deal+0.004
 job+0.004
 subt+0.004
Neg logit contributions
en-0.011
 dot-0.010
ening-0.010
ug-0.010
fly-0.010