TOP ACTIVATIONS
MAX = 0.062

way out . If only she had listened to her mom
But when she opened her closet , she was in for
's door and asked if she could borrow some jam .
and looked at the ost ric hes . They saw one
catch it . But when she caught it , Tommy saw
When they reached the sand pit , Jessie could see that
friends . But when her friends arrived , they hugged her
should not tease the ost ric hes . They are not
tried to open the new bottle , she accidentally tore the
mom and to the ost ric hes . They learned their
she tried to imagine a drink to go along with it
a secret code . If she broke the code , she
Mrs . Smith , if she could borrow it . But
on the horizon . Before she could investigate , the fairy
Just when she was about to give up
the door . But when she got there , it was
to him and asked if she could play too . The
began to think . Before she knew it , she had
pinch it . But when she did , the bug flew
her bright clothes . Before she knew it , her closet

MIN ACTIVATIONS
MIN = -0.074

wrap some tape around the prope ller and attach it to
, fifteen , sixteen , seventeen , eighteen , nineteen ,
had been made many years ago , but it still looked
, the nose , the prope ller . They work hard
, fifteen , sixteen , seventeen , eighteen , nineteen ,
ac orn a few days ago so I could have some
I built it many years ago and I am very proud
. The helicopter had big prope llers that spun around and
touch the wings and the prope ller and the wheels .
, twenty - four , twenty - five , twenty -
wing came off and the prope ller fell down . Sam
books , clothes from long ago , and special family keeps
, three , four , five , six , seven ,
" It was hidden long ago , and only seems fancy
light , Lily . Only twenty pounds ," Mom said .
Thank you , God . Amen ." He opened
, forty - five , forty - six , forty -
, forty - four , forty - five , forty -
her birthday a few days ago . She liked to practice
, eighteen , nineteen , twenty . They are happy .

INTERVAL 0.028 - 0.062
CONTAINS 0.856%

The little girl 's wish was to play like the best
. Not wanting her furry friend to get hurt , she
came and asked them what they wanted . "
When he thought he was almost done , a fox
didn 't want to be friends with Tim my anymore .

INTERVAL -0.006 - 0.028
CONTAINS 72.713%

delicious ! He enjoyed the tea and the
Tom asks Mia . " Let 's ask the
, " M ummy look ! I found a coin ".
it . He asks his mom and dad , " Where
was so happy with her computer . It was her favorite

INTERVAL -0.040 - -0.006
CONTAINS 26.286%

the caul iflower with some spicy sauce . The caul iflower
bathroom . There , on the sink , was a new
lived in a big garden with many other flowers . One
. Jim looked over the horizon and could see the sun
happy . Everyone loved the king because he was so compassionate

INTERVAL -0.074 - -0.040
CONTAINS 0.145%

are a good friend . Can you forgive me ?" Tom
named Lily . Lily was three years old and lived in
," she said . " Do you want to see ?"
make me feel better . Can you use your magical powers
library . She was only three years old , but she
Token way
Token ID835
Feature activation+3.15e-02

Pos logit contributions
 narrative+0.025
 volcan+0.023
 museums+0.022
 islands+0.022
 maybe+0.021
Neg logit contributions
run-0.019
imm-0.019
icycle-0.018
ay-0.018
itter-0.018
Token out
Token ID503
Feature activation+1.58e-02

Pos logit contributions
icycle+0.036
imm+0.036
aker+0.035
run+0.033
ay+0.031
Neg logit contributions
 they-0.055
 narrative-0.054
 volcan-0.053
 museums-0.051
 maybe-0.051
Token.
Token ID13
Feature activation+2.11e-04

Pos logit contributions
icycle+0.022
imm+0.020
itten+0.020
ippy+0.020
itter+0.020
Neg logit contributions
 narrative-0.026
 volcan-0.024
 they-0.024
 museums-0.024
 maybe-0.023
Token If
Token ID1002
Feature activation-1.19e-03

Pos logit contributions
run+0.000
icycle+0.000
-+0.000
!+0.000
ow+0.000
Neg logit contributions
 narrative-0.000
 museums-0.000
 volcan-0.000
 islands-0.000
 Mercury-0.000
Token only
Token ID691
Feature activation+2.80e-02

Pos logit contributions
 narrative+0.002
 islands+0.002
 volcan+0.002
adies+0.002
ooks+0.002
Neg logit contributions
!-0.002
 too-0.002
 really-0.002
 very-0.002
 away-0.002
Token she
Token ID673
Feature activation+6.17e-02

Pos logit contributions
 too+0.031
 away+0.031
run+0.030
 very+0.030
 really+0.030
Neg logit contributions
 narrative-0.061
 islands-0.052
ighters-0.052
 museums-0.050
 volcan-0.050
Token had
Token ID550
Feature activation+2.04e-02

Pos logit contributions
run+0.078
icycle+0.077
imm+0.071
ay+0.071
aker+0.070
Neg logit contributions
 narrative-0.109
 museums-0.094
 volcan-0.093
 maybe-0.092
adies-0.091
Token listened
Token ID16399
Feature activation+1.23e-02

Pos logit contributions
icycle+0.029
run+0.027
itten+0.027
ay+0.027
imm+0.026
Neg logit contributions
 narrative-0.031
 museums-0.031
 volcan-0.030
 they-0.029
adies-0.029
Token to
Token ID284
Feature activation+2.08e-02

Pos logit contributions
run+0.016
icycle+0.015
itten+0.015
imm+0.015
ay+0.014
Neg logit contributions
 narrative-0.021
 they-0.019
 museums-0.019
 volcan-0.019
 maybe-0.019
Token her
Token ID607
Feature activation+1.75e-02

Pos logit contributions
run+0.028
!+0.026
-+0.026
 away+0.026
 really+0.025
Neg logit contributions
 narrative-0.038
 museums-0.032
 islands-0.032
 volcan-0.031
 Mercury-0.031
Token mom
Token ID1995
Feature activation+2.33e-02

Pos logit contributions
run+0.029
imm+0.028
icycle+0.028
chie+0.028
acked+0.027
Neg logit contributions
 narrative-0.025
 maybe-0.022
 volcan-0.021
 museums-0.021
 God-0.020
Token But
Token ID887
Feature activation-1.63e-02

Pos logit contributions
run+0.002
-+0.002
icycle+0.002
imm+0.002
 away+0.002
Neg logit contributions
 narrative-0.002
 museums-0.002
 volcan-0.002
 maybe-0.001
 islands-0.001
Token when
Token ID618
Feature activation-1.80e-02

Pos logit contributions
 narrative+0.022
 museums+0.019
 volcan+0.018
 islands+0.018
 Mercury+0.017
Neg logit contributions
run-0.028
icycle-0.027
--0.026
!-0.026
ay-0.026
Token she
Token ID673
Feature activation+3.28e-02

Pos logit contributions
 narrative+0.025
 museums+0.020
 volcan+0.020
 islands+0.019
 Mercury+0.019
Neg logit contributions
run-0.031
--0.029
!-0.029
icycle-0.029
 away-0.029
Token opened
Token ID4721
Feature activation+5.28e-03

Pos logit contributions
run+0.046
icycle+0.044
chie+0.044
imm+0.044
ay+0.044
Neg logit contributions
 narrative-0.053
 museums-0.047
 maybe-0.046
 volcan-0.045
 islands-0.045
Token her
Token ID607
Feature activation+1.19e-02

Pos logit contributions
run+0.010
icycle+0.010
itter+0.010
aker+0.010
imm+0.010
Neg logit contributions
 narrative-0.006
 museums-0.005
 volcan-0.005
 they-0.005
 islands-0.005
Token closet
Token ID21615
Feature activation+5.81e-02

Pos logit contributions
aker+0.020
imm+0.019
chie+0.019
it+0.019
atter+0.019
Neg logit contributions
 narrative-0.016
 museums-0.015
 planets-0.015
 volcan-0.015
 maybe-0.014
Token,
Token ID11
Feature activation+1.45e-02

Pos logit contributions
run+0.106
chie+0.105
aker+0.101
eting+0.098
hold+0.096
Neg logit contributions
 they-0.067
 narrative-0.061
 maybe-0.061
 museums-0.058
adays-0.056
Token she
Token ID673
Feature activation+4.58e-03

Pos logit contributions
run+0.022
!+0.021
 really+0.021
-+0.021
 away+0.020
Neg logit contributions
 narrative-0.023
 volcan-0.019
 islands-0.019
 museums-0.019
lan-0.018
Token was
Token ID373
Feature activation+2.16e-03

Pos logit contributions
run+0.007
ay+0.006
itten+0.006
icycle+0.006
chie+0.006
Neg logit contributions
 narrative-0.007
 museums-0.007
 islands-0.006
 volcan-0.006
 maybe-0.006
Token in
Token ID287
Feature activation-4.93e-03

Pos logit contributions
ay+0.003
icycle+0.003
itten+0.003
imm+0.003
run+0.003
Neg logit contributions
 narrative-0.003
 volcan-0.003
 maybe-0.003
ighters-0.003
 knights-0.003
Token for
Token ID329
Feature activation-1.86e-04

Pos logit contributions
 narrative+0.007
 museums+0.006
 islands+0.006
 volcan+0.006
 planets+0.005
Neg logit contributions
imm-0.009
itter-0.009
run-0.008
icycle-0.008
atter-0.008
Token's
Token ID338
Feature activation+9.93e-03

Pos logit contributions
run+0.005
icycle+0.005
imm+0.005
chie+0.005
ay+0.005
Neg logit contributions
 narrative-0.007
 museums-0.006
 volcan-0.006
 islands-0.006
 Mercury-0.006
Token door
Token ID3420
Feature activation+1.55e-02

Pos logit contributions
imm+0.016
chie+0.015
icycle+0.015
run+0.014
acked+0.014
Neg logit contributions
 narrative-0.015
lan-0.013
 volcan-0.013
 knights-0.013
 museums-0.012
Token and
Token ID290
Feature activation+6.64e-03

Pos logit contributions
chie+0.024
ippy+0.024
icycle+0.024
itten+0.023
ish+0.023
Neg logit contributions
 they-0.020
 narrative-0.020
 maybe-0.018
 museums-0.018
 sometimes-0.017
Token asked
Token ID1965
Feature activation-7.48e-03

Pos logit contributions
run+0.009
icycle+0.009
-+0.009
!+0.009
ow+0.008
Neg logit contributions
 narrative-0.011
 museums-0.010
 volcan-0.010
 islands-0.009
 Mercury-0.009
Token if
Token ID611
Feature activation+5.14e-03

Pos logit contributions
 narrative+0.012
 museums+0.010
 volcan+0.010
 islands+0.010
 Mercury+0.009
Neg logit contributions
run-0.012
--0.011
!-0.011
 away-0.011
 really-0.011
Token she
Token ID673
Feature activation+5.73e-02

Pos logit contributions
!+0.007
 away+0.007
-+0.007
 really+0.007
 too+0.007
Neg logit contributions
 narrative-0.009
 islands-0.007
ooks-0.007
 volcan-0.007
ighters-0.007
Token could
Token ID714
Feature activation+4.62e-03

Pos logit contributions
run+0.079
icycle+0.077
imm+0.074
 away+0.073
itten+0.073
Neg logit contributions
 narrative-0.094
 museums-0.082
 maybe-0.081
 volcan-0.081
lan-0.078
Token borrow
Token ID8804
Feature activation+1.01e-02

Pos logit contributions
run+0.009
icycle+0.009
imm+0.009
chie+0.008
itten+0.008
Neg logit contributions
 narrative-0.005
 museums-0.004
 maybe-0.004
 volcan-0.004
 they-0.004
Token some
Token ID617
Feature activation-6.33e-03

Pos logit contributions
icycle+0.019
run+0.018
imm+0.018
chie+0.018
itten+0.018
Neg logit contributions
 narrative-0.012
 museums-0.011
 volcan-0.011
 maybe-0.010
 they-0.010
Token jam
Token ID16853
Feature activation+1.54e-02

Pos logit contributions
 narrative+0.007
 museums+0.006
 volcan+0.006
 islands+0.005
 maybe+0.005
Neg logit contributions
run-0.012
icycle-0.012
chie-0.012
imm-0.012
itter-0.011
Token.
Token ID13
Feature activation-1.71e-03

Pos logit contributions
chie+0.023
icycle+0.023
run+0.022
eting+0.020
ivered+0.020
Neg logit contributions
 narrative-0.024
 they-0.024
 maybe-0.022
 museums-0.021
 sometimes-0.020
Token and
Token ID290
Feature activation+2.16e-02

Pos logit contributions
 narrative+0.039
 museums+0.033
adies+0.032
 Mercury+0.032
 volcan+0.032
Neg logit contributions
run-0.025
--0.023
!-0.023
ow-0.022
 away-0.022
Token looked
Token ID3114
Feature activation-1.29e-02

Pos logit contributions
run+0.024
-+0.022
icycle+0.022
!+0.022
 away+0.021
Neg logit contributions
 narrative-0.043
 museums-0.038
 volcan-0.037
 islands-0.037
 Mercury-0.036
Token at
Token ID379
Feature activation+6.68e-03

Pos logit contributions
 narrative+0.021
 volcan+0.019
 they+0.019
 maybe+0.019
 islands+0.019
Neg logit contributions
icycle-0.018
run-0.017
itten-0.017
ay-0.017
imm-0.016
Token the
Token ID262
Feature activation+1.05e-02

Pos logit contributions
run+0.011
icycle+0.011
-+0.011
itten+0.010
ay+0.010
Neg logit contributions
 narrative-0.010
 museums-0.008
 volcan-0.008
 islands-0.008
 Mercury-0.007
Token ost
Token ID23619
Feature activation+1.20e-03

Pos logit contributions
run+0.021
icycle+0.019
imm+0.019
ay+0.018
ow+0.018
Neg logit contributions
 narrative-0.013
 volcan-0.012
 museums-0.011
 islands-0.011
 whales-0.010
Tokenric
Token ID1173
Feature activation+5.72e-02

Pos logit contributions
run+0.002
 away+0.002
 really+0.002
 package+0.002
 leash+0.002
Neg logit contributions
 narrative-0.002
 volcan-0.002
 islands-0.002
lan-0.002
 museums-0.002
Tokenhes
Token ID956
Feature activation+8.67e-03

Pos logit contributions
run+0.102
icycle+0.099
itten+0.094
chie+0.093
 away+0.093
Neg logit contributions
 narrative-0.076
 volcan-0.064
 museums-0.061
ighters-0.060
adies-0.059
Token.
Token ID13
Feature activation-6.60e-03

Pos logit contributions
icycle+0.013
run+0.011
imm+0.011
chie+0.011
itten+0.011
Neg logit contributions
 narrative-0.015
 they-0.015
 volcan-0.013
 maybe-0.013
 sometimes-0.012
Token They
Token ID1119
Feature activation-1.40e-03

Pos logit contributions
 narrative+0.010
 museums+0.007
 islands+0.007
adies+0.007
 volcan+0.007
Neg logit contributions
run-0.011
 really-0.010
!-0.010
ow-0.010
 too-0.010
Token saw
Token ID2497
Feature activation-2.68e-02

Pos logit contributions
 narrative+0.002
 museums+0.002
 volcan+0.002
 islands+0.002
 maybe+0.002
Neg logit contributions
run-0.002
icycle-0.002
ay-0.002
--0.002
itten-0.002
Token one
Token ID530
Feature activation+9.06e-03

Pos logit contributions
 narrative+0.050
 museums+0.042
adies+0.040
 islands+0.039
ooks+0.039
Neg logit contributions
run-0.034
--0.033
 really-0.032
ow-0.032
 very-0.032
Token catch
Token ID4929
Feature activation+1.36e-02

Pos logit contributions
imm+0.012
run+0.012
icycle+0.012
itten+0.011
ay+0.011
Neg logit contributions
 narrative-0.009
 museums-0.009
 maybe-0.008
 volcan-0.008
 islands-0.008
Token it
Token ID340
Feature activation+1.03e-02

Pos logit contributions
imm+0.026
icycle+0.026
chie+0.026
run+0.025
itter+0.024
Neg logit contributions
 they-0.013
 narrative-0.013
 maybe-0.013
 museums-0.012
 volcan-0.012
Token.
Token ID13
Feature activation-2.87e-03

Pos logit contributions
chie+0.019
icycle+0.019
aker+0.018
itten+0.018
run+0.018
Neg logit contributions
 they-0.012
 sometimes-0.012
 maybe-0.011
 volcan-0.010
 narrative-0.010
Token But
Token ID887
Feature activation+1.12e-03

Pos logit contributions
 narrative+0.003
 museums+0.003
 maybe+0.003
 they+0.003
 volcan+0.003
Neg logit contributions
imm-0.005
chie-0.005
--0.005
aker-0.005
ow-0.005
Token when
Token ID618
Feature activation-5.36e-03

Pos logit contributions
run+0.002
icycle+0.002
imm+0.002
chie+0.002
itten+0.002
Neg logit contributions
 narrative-0.002
 museums-0.001
 maybe-0.001
 they-0.001
 volcan-0.001
Token she
Token ID673
Feature activation+5.71e-02

Pos logit contributions
 narrative+0.008
 museums+0.007
 volcan+0.006
 islands+0.006
 Mercury+0.006
Neg logit contributions
run-0.009
icycle-0.008
--0.008
!-0.008
ow-0.008
Token caught
Token ID4978
Feature activation+2.67e-02

Pos logit contributions
chie+0.090
eting+0.088
icycle+0.087
ooter+0.086
run+0.085
Neg logit contributions
 they-0.073
 maybe-0.068
 narrative-0.067
 museums-0.059
adays-0.058
Token it
Token ID340
Feature activation+2.83e-02

Pos logit contributions
imm+0.054
run+0.054
chie+0.054
eting+0.052
icycle+0.052
Neg logit contributions
 maybe-0.024
 they-0.023
 narrative-0.021
 museums-0.021
 volcan-0.020
Token,
Token ID11
Feature activation+2.12e-02

Pos logit contributions
aker+0.055
icycle+0.054
run+0.054
chie+0.053
eting+0.052
Neg logit contributions
 they-0.032
 maybe-0.029
 sometimes-0.024
adays-0.024
 narrative-0.023
Token Tommy
Token ID19919
Feature activation+7.99e-04

Pos logit contributions
run+0.037
icycle+0.035
imm+0.035
ow+0.035
-+0.035
Neg logit contributions
 narrative-0.029
 museums-0.024
 volcan-0.023
 islands-0.022
 maybe-0.022
Token saw
Token ID2497
Feature activation-1.24e-02

Pos logit contributions
ay+0.002
chie+0.001
run+0.001
imm+0.001
eting+0.001
Neg logit contributions
 they-0.001
 maybe-0.001
 sometimes-0.001
 islands-0.001
 knights-0.001
Token When
Token ID1649
Feature activation-7.93e-03

Pos logit contributions
run+0.008
-+0.008
icycle+0.008
 away+0.008
imm+0.008
Neg logit contributions
 narrative-0.007
 museums-0.006
 volcan-0.006
 islands-0.006
 maybe-0.006
Token they
Token ID484
Feature activation+3.45e-02

Pos logit contributions
 narrative+0.008
 volcan+0.006
 museums+0.006
 islands+0.006
 Mercury+0.005
Neg logit contributions
!-0.017
run-0.016
--0.016
 really-0.016
 away-0.016
Token reached
Token ID4251
Feature activation+2.10e-03

Pos logit contributions
icycle+0.056
aker+0.055
run+0.053
itten+0.053
imm+0.053
Neg logit contributions
 they-0.045
 narrative-0.045
 museums-0.042
 maybe-0.042
 volcan-0.039
Token the
Token ID262
Feature activation+1.35e-03

Pos logit contributions
run+0.004
icycle+0.004
aker+0.004
imm+0.004
itten+0.004
Neg logit contributions
 narrative-0.002
 museums-0.002
 islands-0.002
 volcan-0.002
 they-0.002
Token sand
Token ID6450
Feature activation+5.70e-03

Pos logit contributions
run+0.003
imm+0.003
ow+0.003
itten+0.003
aker+0.003
Neg logit contributions
 narrative-0.001
 museums-0.001
 islands-0.001
 volcan-0.001
 Mercury-0.001
Tokenpit
Token ID15544
Feature activation+5.67e-02

Pos logit contributions
run+0.009
itten+0.008
chie+0.008
eting+0.008
icycle+0.008
Neg logit contributions
 narrative-0.008
 islands-0.008
 they-0.008
 volcan-0.008
 museums-0.008
Token,
Token ID11
Feature activation+1.08e-02

Pos logit contributions
run+0.094
icycle+0.089
aker+0.088
chie+0.087
itten+0.087
Neg logit contributions
 they-0.078
 maybe-0.072
 narrative-0.072
 volcan-0.070
 museums-0.070
Token Jessie
Token ID43425
Feature activation+2.16e-03

Pos logit contributions
run+0.022
!+0.021
 really+0.021
-+0.021
 away+0.020
Neg logit contributions
 narrative-0.012
 museums-0.009
 islands-0.009
 volcan-0.009
 Mercury-0.008
Token could
Token ID714
Feature activation-1.40e-02

Pos logit contributions
run+0.003
ay+0.003
chie+0.003
eting+0.003
 leash+0.003
Neg logit contributions
 they-0.003
 islands-0.003
lan-0.003
 narrative-0.003
 volcan-0.003
Token see
Token ID766
Feature activation-1.44e-02

Pos logit contributions
 narrative+0.017
 museums+0.015
 maybe+0.015
 volcan+0.015
 islands+0.014
Neg logit contributions
run-0.025
icycle-0.024
--0.024
itten-0.024
ay-0.024
Token that
Token ID326
Feature activation+9.09e-03

Pos logit contributions
 narrative+0.021
 museums+0.019
 islands+0.018
 volcan+0.018
 maybe+0.017
Neg logit contributions
run-0.023
icycle-0.022
ay-0.022
imm-0.022
itten-0.022
Token friends
Token ID2460
Feature activation+2.17e-02

Pos logit contributions
run+0.002
icycle+0.002
chie+0.002
imm+0.002
aker+0.002
Neg logit contributions
 narrative-0.002
 volcan-0.001
 museums-0.001
 they-0.001
lan-0.001
Token.
Token ID13
Feature activation-1.28e-03

Pos logit contributions
chie+0.033
icycle+0.033
imm+0.032
run+0.031
ay+0.030
Neg logit contributions
 they-0.033
 narrative-0.032
 maybe-0.030
 sometimes-0.030
 museums-0.030
Token But
Token ID887
Feature activation-3.41e-03

Pos logit contributions
 narrative+0.002
 museums+0.002
 volcan+0.001
 maybe+0.001
 Mercury+0.001
Neg logit contributions
run-0.002
icycle-0.002
--0.002
itten-0.002
imm-0.002
Token when
Token ID618
Feature activation-1.68e-03

Pos logit contributions
 narrative+0.006
 islands+0.005
 museums+0.005
 volcan+0.005
lan+0.005
Neg logit contributions
 away-0.005
--0.005
run-0.005
 too-0.005
 really-0.005
Token her
Token ID607
Feature activation+1.92e-02

Pos logit contributions
 narrative+0.003
 islands+0.002
 volcan+0.002
adies+0.002
 Mercury+0.002
Neg logit contributions
!-0.003
 away-0.003
 really-0.003
 too-0.002
 very-0.002
Token friends
Token ID2460
Feature activation+5.66e-02

Pos logit contributions
run+0.027
icycle+0.025
-+0.025
!+0.024
ow+0.024
Neg logit contributions
 narrative-0.033
 museums-0.028
 volcan-0.028
 islands-0.027
 Mercury-0.027
Token arrived
Token ID5284
Feature activation+1.03e-02

Pos logit contributions
run+0.091
icycle+0.085
imm+0.084
aker+0.083
itten+0.082
Neg logit contributions
 narrative-0.085
 museums-0.075
 they-0.070
 maybe-0.070
 volcan-0.070
Token,
Token ID11
Feature activation+1.86e-02

Pos logit contributions
run+0.012
icycle+0.011
-+0.011
itten+0.011
ow+0.011
Neg logit contributions
 narrative-0.020
 museums-0.018
 volcan-0.017
 islands-0.017
 Mercury-0.017
Token they
Token ID484
Feature activation+2.70e-03

Pos logit contributions
run+0.032
 really+0.032
 away+0.032
!+0.032
-+0.031
Neg logit contributions
 narrative-0.025
 islands-0.020
 volcan-0.020
adies-0.020
 museums-0.020
Token hugged
Token ID41130
Feature activation+6.58e-04

Pos logit contributions
run+0.004
icycle+0.004
ay+0.004
itten+0.004
imm+0.004
Neg logit contributions
 narrative-0.004
 they-0.004
 museums-0.003
 islands-0.003
 maybe-0.003
Token her
Token ID607
Feature activation-6.25e-04

Pos logit contributions
icycle+0.001
run+0.001
itten+0.001
imm+0.001
chie+0.001
Neg logit contributions
 narrative-0.001
 museums-0.001
 they-0.001
 maybe-0.001
 volcan-0.001
Token should
Token ID815
Feature activation-1.04e-03

Pos logit contributions
run+0.006
icycle+0.005
-+0.005
itten+0.005
ay+0.005
Neg logit contributions
 narrative-0.006
 museums-0.005
 volcan-0.005
 islands-0.005
 Mercury-0.005
Token not
Token ID407
Feature activation+1.70e-03

Pos logit contributions
 narrative+0.002
 museums+0.002
 volcan+0.002
 islands+0.002
 Mercury+0.002
Neg logit contributions
run-0.001
!-0.001
 really-0.001
 away-0.001
--0.001
Token tease
Token ID34691
Feature activation+1.99e-02

Pos logit contributions
run+0.003
icycle+0.002
imm+0.002
-+0.002
itten+0.002
Neg logit contributions
 narrative-0.003
 museums-0.002
 volcan-0.002
 islands-0.002
 Mercury-0.002
Token the
Token ID262
Feature activation+1.31e-02

Pos logit contributions
run+0.033
icycle+0.033
itten+0.032
imm+0.031
ay+0.030
Neg logit contributions
 narrative-0.028
 museums-0.025
 volcan-0.024
 islands-0.023
 maybe-0.022
Token ost
Token ID23619
Feature activation+3.54e-03

Pos logit contributions
run+0.028
icycle+0.027
imm+0.026
ay+0.025
ow+0.025
Neg logit contributions
 narrative-0.014
 volcan-0.012
 museums-0.011
 islands-0.010
lan-0.010
Tokenric
Token ID1173
Feature activation+5.65e-02

Pos logit contributions
run+0.005
icycle+0.004
 away+0.004
 really+0.004
 package+0.004
Neg logit contributions
 narrative-0.006
 volcan-0.006
 museums-0.005
 islands-0.005
lan-0.005
Tokenhes
Token ID956
Feature activation+1.98e-02

Pos logit contributions
run+0.098
icycle+0.097
itten+0.091
 package+0.090
chie+0.089
Neg logit contributions
 narrative-0.077
 volcan-0.065
 museums-0.062
ighters-0.061
adies-0.060
Token.
Token ID13
Feature activation-6.38e-03

Pos logit contributions
icycle+0.033
imm+0.029
run+0.029
itten+0.027
ippy+0.027
Neg logit contributions
 narrative-0.032
 they-0.030
 maybe-0.028
 sometimes-0.028
 volcan-0.027
Token They
Token ID1119
Feature activation+2.55e-03

Pos logit contributions
 narrative+0.011
 museums+0.009
 islands+0.009
 Mercury+0.009
 deserts+0.008
Neg logit contributions
run-0.009
ow-0.009
ay-0.008
--0.008
itter-0.008
Token are
Token ID389
Feature activation+3.28e-03

Pos logit contributions
run+0.004
icycle+0.004
itten+0.004
-+0.004
ay+0.004
Neg logit contributions
 narrative-0.004
 museums-0.003
 volcan-0.003
 islands-0.003
 maybe-0.003
Token not
Token ID407
Feature activation-2.61e-02

Pos logit contributions
icycle+0.005
run+0.005
ay+0.005
itten+0.005
imm+0.005
Neg logit contributions
 narrative-0.005
 volcan-0.004
 museums-0.004
 islands-0.004
 maybe-0.004
Token tried
Token ID3088
Feature activation+1.86e-02

Pos logit contributions
icycle+0.061
run+0.060
ooter+0.058
chie+0.057
ay+0.056
Neg logit contributions
 narrative-0.075
 maybe-0.065
 they-0.065
 museums-0.065
 volcan-0.061
Token to
Token ID284
Feature activation+7.10e-03

Pos logit contributions
run+0.020
-+0.018
icycle+0.018
!+0.017
 away+0.017
Neg logit contributions
 narrative-0.039
 museums-0.034
 volcan-0.033
 islands-0.033
 Mercury-0.032
Token open
Token ID1280
Feature activation+2.08e-02

Pos logit contributions
run+0.012
icycle+0.012
ay+0.011
imm+0.011
-+0.011
Neg logit contributions
 narrative-0.010
 museums-0.008
 volcan-0.008
 islands-0.008
 maybe-0.008
Token the
Token ID262
Feature activation+9.80e-03

Pos logit contributions
run+0.038
icycle+0.037
imm+0.036
aker+0.036
itten+0.036
Neg logit contributions
 narrative-0.024
 museums-0.022
 volcan-0.021
 islands-0.020
 maybe-0.020
Token new
Token ID649
Feature activation+6.01e-03

Pos logit contributions
run+0.013
imm+0.013
ow+0.012
icycle+0.012
chie+0.011
Neg logit contributions
 narrative-0.017
 museums-0.015
 Mercury-0.015
 tul-0.014
 volcan-0.014
Token bottle
Token ID9294
Feature activation+5.64e-02

Pos logit contributions
run+0.007
imm+0.007
icycle+0.007
ow+0.006
chie+0.006
Neg logit contributions
 narrative-0.011
 museums-0.010
 volcan-0.010
 Mercury-0.010
 islands-0.010
Token,
Token ID11
Feature activation+2.64e-02

Pos logit contributions
icycle+0.101
run+0.100
aker+0.097
chie+0.094
eting+0.093
Neg logit contributions
 they-0.066
 narrative-0.063
 maybe-0.063
 volcan-0.054
 museums-0.053
Token she
Token ID673
Feature activation+6.83e-03

Pos logit contributions
run+0.040
 really+0.039
!+0.039
-+0.038
 away+0.038
Neg logit contributions
 narrative-0.042
 islands-0.035
 museums-0.034
 volcan-0.034
adies-0.033
Token accidentally
Token ID14716
Feature activation+9.06e-03

Pos logit contributions
run+0.011
ay+0.011
icycle+0.010
imm+0.010
chie+0.010
Neg logit contributions
 narrative-0.010
 museums-0.008
 islands-0.008
 maybe-0.008
 they-0.008
Token tore
Token ID26197
Feature activation+1.79e-02

Pos logit contributions
ay+0.014
icycle+0.014
run+0.013
imm+0.013
chie+0.013
Neg logit contributions
 narrative-0.012
 they-0.012
 maybe-0.012
 volcan-0.011
 museums-0.011
Token the
Token ID262
Feature activation+5.94e-03

Pos logit contributions
run+0.033
icycle+0.033
imm+0.032
itten+0.031
ow+0.031
Neg logit contributions
 narrative-0.020
 volcan-0.018
 museums-0.017
 islands-0.016
 they-0.016
Token mom
Token ID1995
Feature activation+1.47e-02

Pos logit contributions
 very+0.007
-+0.007
 too+0.007
 really+0.007
ow+0.007
Neg logit contributions
 narrative-0.011
 Mercury-0.009
 museums-0.009
adies-0.009
ooks-0.009
Token and
Token ID290
Feature activation+1.43e-02

Pos logit contributions
run+0.012
-+0.012
!+0.011
 away+0.011
ow+0.010
Neg logit contributions
 narrative-0.034
 museums-0.030
 islands-0.030
 volcan-0.030
 Mercury-0.029
Token to
Token ID284
Feature activation+5.38e-03

Pos logit contributions
run+0.015
-+0.014
 away+0.014
ow+0.013
!+0.013
Neg logit contributions
 narrative-0.029
 museums-0.026
 volcan-0.026
 islands-0.026
 Mercury-0.025
Token the
Token ID262
Feature activation+1.57e-02

Pos logit contributions
run+0.008
ow+0.008
 away+0.008
 really+0.008
-+0.007
Neg logit contributions
 narrative-0.009
 museums-0.008
 Mercury-0.007
 islands-0.007
 volcan-0.007
Token ost
Token ID23619
Feature activation+1.43e-03

Pos logit contributions
run+0.030
icycle+0.028
-+0.028
!+0.027
ow+0.027
Neg logit contributions
 narrative-0.019
 museums-0.016
 volcan-0.015
 islands-0.015
 Mercury-0.014
Tokenric
Token ID1173
Feature activation+5.64e-02

Pos logit contributions
run+0.002
icycle+0.002
 away+0.002
 really+0.002
ay+0.002
Neg logit contributions
 narrative-0.002
 volcan-0.002
 islands-0.002
 museums-0.002
lan-0.002
Tokenhes
Token ID956
Feature activation+1.38e-02

Pos logit contributions
run+0.098
icycle+0.095
itten+0.091
ay+0.090
!+0.089
Neg logit contributions
 narrative-0.077
 volcan-0.065
 museums-0.063
 islands-0.061
ighters-0.061
Token.
Token ID13
Feature activation-2.80e-03

Pos logit contributions
icycle+0.022
run+0.021
imm+0.020
itten+0.019
chie+0.019
Neg logit contributions
 narrative-0.022
 they-0.021
 maybe-0.019
 volcan-0.019
 God-0.018
Token They
Token ID1119
Feature activation+2.93e-03

Pos logit contributions
 narrative+0.004
adies+0.004
 museums+0.004
 islands+0.004
 Mercury+0.003
Neg logit contributions
run-0.004
 really-0.004
 very-0.004
--0.004
!-0.004
Token learned
Token ID4499
Feature activation-1.19e-03

Pos logit contributions
run+0.004
icycle+0.004
-+0.004
ay+0.004
itten+0.004
Neg logit contributions
 narrative-0.005
 museums-0.004
 volcan-0.004
 islands-0.004
 Mercury-0.004
Token their
Token ID511
Feature activation+5.64e-03

Pos logit contributions
 narrative+0.002
 museums+0.002
 islands+0.002
adies+0.002
 volcan+0.002
Neg logit contributions
--0.002
run-0.002
 really-0.002
 very-0.002
 away-0.002
Token she
Token ID673
Feature activation+5.13e-02

Pos logit contributions
run+0.012
icycle+0.011
-+0.011
!+0.011
ow+0.011
Neg logit contributions
 narrative-0.009
 museums-0.008
 volcan-0.007
 islands-0.007
 Mercury-0.007
Token tried
Token ID3088
Feature activation+2.10e-02

Pos logit contributions
run+0.072
icycle+0.070
 leash+0.068
eting+0.068
ooter+0.067
Neg logit contributions
 narrative-0.085
 museums-0.071
 volcan-0.070
 maybe-0.069
 islands-0.067
Token to
Token ID284
Feature activation+1.43e-02

Pos logit contributions
run+0.027
icycle+0.024
-+0.024
ow+0.023
itten+0.023
Neg logit contributions
 narrative-0.039
 museums-0.034
 volcan-0.034
 islands-0.033
 Mercury-0.032
Token imagine
Token ID5967
Feature activation+1.75e-02

Pos logit contributions
run+0.025
icycle+0.024
itten+0.023
-+0.023
itter+0.023
Neg logit contributions
 narrative-0.019
 museums-0.017
 volcan-0.016
 islands-0.016
 maybe-0.015
Token a
Token ID257
Feature activation+1.94e-02

Pos logit contributions
run+0.027
icycle+0.026
itten+0.025
ay+0.024
aker+0.024
Neg logit contributions
 narrative-0.027
 museums-0.025
 volcan-0.024
 maybe-0.024
 they-0.023
Token drink
Token ID4144
Feature activation+5.63e-02

Pos logit contributions
ow+0.031
run+0.030
imm+0.030
icycle+0.030
ippy+0.029
Neg logit contributions
 volcan-0.028
 narrative-0.028
lan-0.027
 they-0.026
 museums-0.026
Token to
Token ID284
Feature activation+1.30e-02

Pos logit contributions
icycle+0.107
run+0.103
chie+0.095
ippy+0.093
ooter+0.092
Neg logit contributions
 they-0.072
 narrative-0.065
 maybe-0.061
 volcan-0.057
adays-0.056
Token go
Token ID467
Feature activation+2.42e-02

Pos logit contributions
run+0.022
icycle+0.022
itten+0.021
ay+0.020
itter+0.020
Neg logit contributions
 narrative-0.018
 museums-0.016
 volcan-0.016
 islands-0.015
 maybe-0.015
Token along
Token ID1863
Feature activation+1.72e-02

Pos logit contributions
run+0.036
icycle+0.034
itten+0.033
imm+0.033
eting+0.032
Neg logit contributions
 narrative-0.037
 they-0.037
 museums-0.035
 islands-0.032
 maybe-0.032
Token with
Token ID351
Feature activation+1.48e-02

Pos logit contributions
run+0.028
icycle+0.027
itten+0.027
ooter+0.025
imm+0.025
Neg logit contributions
 narrative-0.025
 museums-0.024
 volcan-0.023
 they-0.023
 islands-0.023
Token it
Token ID340
Feature activation+8.58e-03

Pos logit contributions
icycle+0.025
run+0.025
imm+0.023
itten+0.023
ooter+0.023
Neg logit contributions
 narrative-0.020
 volcan-0.019
 museums-0.018
 islands-0.017
 maybe-0.017
Token a
Token ID257
Feature activation+9.06e-03

Pos logit contributions
icycle+0.027
itten+0.026
run+0.026
ay+0.025
ooter+0.025
Neg logit contributions
 narrative-0.023
 they-0.021
 volcan-0.021
 museums-0.020
 maybe-0.020
Token secret
Token ID3200
Feature activation+9.69e-03

Pos logit contributions
run+0.014
ow+0.013
imm+0.013
icycle+0.012
atter+0.012
Neg logit contributions
 narrative-0.015
 volcan-0.014
 museums-0.013
 islands-0.013
 Mercury-0.013
Token code
Token ID2438
Feature activation+3.94e-04

Pos logit contributions
run+0.013
icycle+0.013
itten+0.012
bo+0.012
ow+0.012
Neg logit contributions
 narrative-0.017
 islands-0.015
 volcan-0.015
 museums-0.015
 they-0.015
Token.
Token ID13
Feature activation+6.61e-03

Pos logit contributions
icycle+0.001
itten+0.001
run+0.001
ippy+0.001
chie+0.001
Neg logit contributions
 they-0.001
 narrative-0.001
 maybe-0.001
 museums-0.001
 islands-0.001
Token If
Token ID1002
Feature activation-7.02e-03

Pos logit contributions
itten+0.013
icycle+0.013
-+0.013
aker+0.013
chie+0.013
Neg logit contributions
 narrative-0.007
 museums-0.006
 maybe-0.006
 volcan-0.006
 Mercury-0.005
Token she
Token ID673
Feature activation+5.62e-02

Pos logit contributions
 narrative+0.010
 islands+0.008
 museums+0.008
 volcan+0.008
lan+0.008
Neg logit contributions
run-0.012
!-0.011
--0.011
 really-0.011
 away-0.011
Token broke
Token ID6265
Feature activation+1.45e-02

Pos logit contributions
run+0.085
icycle+0.084
ay+0.081
itten+0.080
itter+0.080
Neg logit contributions
 narrative-0.083
 maybe-0.072
 museums-0.069
 volcan-0.068
 they-0.067
Token the
Token ID262
Feature activation+5.42e-05

Pos logit contributions
run+0.022
icycle+0.020
itten+0.020
imm+0.020
ay+0.020
Neg logit contributions
 narrative-0.023
 museums-0.020
 volcan-0.020
 islands-0.019
 Mercury-0.019
Token code
Token ID2438
Feature activation+2.12e-02

Pos logit contributions
run+0.000
icycle+0.000
imm+0.000
ow+0.000
itten+0.000
Neg logit contributions
 narrative-0.000
 museums-0.000
 volcan-0.000
 islands-0.000
 Mercury-0.000
Token,
Token ID11
Feature activation+1.26e-02

Pos logit contributions
icycle+0.034
aker+0.034
run+0.033
itten+0.033
ippy+0.033
Neg logit contributions
 they-0.029
 narrative-0.028
 maybe-0.028
 museums-0.026
 volcan-0.024
Token she
Token ID673
Feature activation+8.77e-03

Pos logit contributions
run+0.020
!+0.019
-+0.019
 away+0.019
 really+0.019
Neg logit contributions
 narrative-0.019
 museums-0.016
 islands-0.015
 volcan-0.015
lan-0.015
Token Mrs
Token ID9074
Feature activation+1.48e-02

Pos logit contributions
icycle+0.035
ay+0.034
imm+0.033
acked+0.033
atter+0.032
Neg logit contributions
 maybe-0.020
 they-0.019
 God-0.019
 museums-0.018
 volcan-0.018
Token.
Token ID13
Feature activation+3.21e-02

Pos logit contributions
run+0.009
!+0.008
-+0.008
 away+0.007
ow+0.007
Neg logit contributions
 narrative-0.037
 museums-0.033
 volcan-0.033
 islands-0.033
 Mercury-0.032
Token Smith
Token ID4176
Feature activation-9.88e-04

Pos logit contributions
run+0.046
icycle+0.042
-+0.042
!+0.041
ow+0.041
Neg logit contributions
 narrative-0.054
 museums-0.047
 volcan-0.046
 islands-0.045
 Mercury-0.044
Token,
Token ID11
Feature activation+1.30e-02

Pos logit contributions
 narrative+0.001
 they+0.001
 maybe+0.001
 volcan+0.001
 museums+0.001
Neg logit contributions
icycle-0.002
ay-0.001
run-0.001
imm-0.001
chie-0.001
Token if
Token ID611
Feature activation+4.56e-03

Pos logit contributions
icycle+0.024
imm+0.022
ay+0.022
chie+0.022
atter+0.021
Neg logit contributions
 maybe-0.017
 they-0.017
 narrative-0.016
 God-0.016
 volcan-0.015
Token she
Token ID673
Feature activation+5.61e-02

Pos logit contributions
!+0.007
 too+0.007
-+0.007
 really+0.007
 away+0.007
Neg logit contributions
 narrative-0.007
ooks-0.006
 islands-0.006
ighters-0.005
 volcan-0.005
Token could
Token ID714
Feature activation+8.62e-04

Pos logit contributions
run+0.079
icycle+0.075
ay+0.073
 away+0.073
itten+0.072
Neg logit contributions
 narrative-0.092
 museums-0.078
 volcan-0.078
 maybe-0.077
 islands-0.075
Token borrow
Token ID8804
Feature activation+1.62e-02

Pos logit contributions
run+0.001
icycle+0.001
itten+0.001
imm+0.001
-+0.001
Neg logit contributions
 narrative-0.001
 museums-0.001
 volcan-0.001
 maybe-0.001
 islands-0.001
Token it
Token ID340
Feature activation+2.68e-02

Pos logit contributions
icycle+0.029
run+0.028
imm+0.028
itten+0.027
aker+0.027
Neg logit contributions
 narrative-0.020
 museums-0.019
 volcan-0.018
 maybe-0.018
 they-0.017
Token.
Token ID13
Feature activation+3.79e-03

Pos logit contributions
icycle+0.048
itten+0.046
aker+0.045
chie+0.045
run+0.044
Neg logit contributions
 they-0.035
 maybe-0.035
 narrative-0.031
 museums-0.030
 sometimes-0.030
Token But
Token ID887
Feature activation-2.80e-03

Pos logit contributions
run+0.006
icycle+0.006
-+0.006
itten+0.006
ow+0.006
Neg logit contributions
 narrative-0.006
 museums-0.005
 volcan-0.005
 islands-0.005
 Mercury-0.005
Token on
Token ID319
Feature activation+4.07e-03

Pos logit contributions
 narrative+0.001
 they+0.001
 museums+0.001
 volcan+0.001
 islands+0.001
Neg logit contributions
icycle-0.001
run-0.001
ooter-0.001
itten-0.001
itter-0.001
Token the
Token ID262
Feature activation+1.62e-03

Pos logit contributions
run+0.008
icycle+0.007
itter+0.007
itten+0.007
imm+0.007
Neg logit contributions
 narrative-0.005
 museums-0.005
 volcan-0.005
 islands-0.004
 planets-0.004
Token horizon
Token ID17810
Feature activation+4.77e-03

Pos logit contributions
run+0.003
icycle+0.002
-+0.002
ow+0.002
itten+0.002
Neg logit contributions
 narrative-0.002
 museums-0.002
 volcan-0.002
 islands-0.002
 Mercury-0.002
Token.
Token ID13
Feature activation+8.62e-03

Pos logit contributions
run+0.006
icycle+0.006
chie+0.006
itten+0.006
ippy+0.006
Neg logit contributions
 narrative-0.008
 they-0.008
 museums-0.007
 volcan-0.007
 islands-0.007
Token Before
Token ID7413
Feature activation+1.02e-02

Pos logit contributions
run+0.013
-+0.012
icycle+0.012
 away+0.012
imm+0.012
Neg logit contributions
 narrative-0.013
 museums-0.012
 volcan-0.012
 Mercury-0.011
 islands-0.011
Token she
Token ID673
Feature activation+5.59e-02

Pos logit contributions
run+0.017
!+0.017
-+0.016
 really+0.016
 away+0.016
Neg logit contributions
 narrative-0.015
 museums-0.012
 volcan-0.012
 islands-0.012
 Mercury-0.011
Token could
Token ID714
Feature activation+1.63e-04

Pos logit contributions
icycle+0.092
run+0.086
ay+0.085
itter+0.084
 leash+0.083
Neg logit contributions
 narrative-0.079
 maybe-0.071
 volcan-0.064
 they-0.064
 museums-0.063
Token investigate
Token ID9161
Feature activation+6.72e-04

Pos logit contributions
run+0.000
icycle+0.000
itten+0.000
-+0.000
ay+0.000
Neg logit contributions
 narrative-0.000
 museums-0.000
 volcan-0.000
 islands-0.000
 maybe-0.000
Token,
Token ID11
Feature activation+1.90e-02

Pos logit contributions
run+0.001
icycle+0.001
itten+0.001
ay+0.001
imm+0.001
Neg logit contributions
 narrative-0.001
 museums-0.001
 volcan-0.001
 islands-0.001
 maybe-0.001
Token the
Token ID262
Feature activation-6.47e-03

Pos logit contributions
run+0.032
-+0.030
!+0.030
icycle+0.030
 away+0.030
Neg logit contributions
 narrative-0.027
 museums-0.022
 volcan-0.022
 islands-0.021
 Mercury-0.021
Token fairy
Token ID25607
Feature activation-3.54e-03

Pos logit contributions
 narrative+0.011
 museums+0.010
 volcan+0.009
 islands+0.009
 Mercury+0.009
Neg logit contributions
run-0.009
icycle-0.008
--0.008
ow-0.008
!-0.008
Token
Token ID220
Feature activation-4.64e-04

Pos logit contributions
run+0.003
icycle+0.003
-+0.003
itten+0.003
chie+0.003
Neg logit contributions
 narrative-0.003
 museums-0.002
 maybe-0.002
 volcan-0.002
 Mercury-0.002
Token
Token ID198
Feature activation+9.41e-03

Pos logit contributions
 narrative+0.001
 maybe+0.001
 museums+0.001
ooks+0.001
 they+0.001
Neg logit contributions
 away-0.001
--0.001
chie-0.001
icycle-0.001
run-0.001
Token
Token ID198
Feature activation+1.05e-02

Pos logit contributions
run+0.014
icycle+0.013
-+0.013
ow+0.013
 away+0.013
Neg logit contributions
 narrative-0.015
 museums-0.013
 volcan-0.012
 maybe-0.012
 islands-0.012
TokenJust
Token ID5703
Feature activation-3.93e-03

Pos logit contributions
run+0.019
icycle+0.018
ow+0.018
-+0.018
!+0.017
Neg logit contributions
 narrative-0.014
 museums-0.011
 volcan-0.011
 islands-0.010
 Mercury-0.010
Token when
Token ID618
Feature activation+2.58e-03

Pos logit contributions
 narrative+0.005
 museums+0.004
 volcan+0.004
 islands+0.004
 Mercury+0.004
Neg logit contributions
run-0.007
icycle-0.007
--0.006
!-0.006
ow-0.006
Token she
Token ID673
Feature activation+5.59e-02

Pos logit contributions
run+0.004
!+0.004
 away+0.004
 really+0.004
-+0.004
Neg logit contributions
 narrative-0.004
 islands-0.003
 museums-0.003
 volcan-0.003
 Mercury-0.003
Token was
Token ID373
Feature activation+3.13e-02

Pos logit contributions
run+0.076
icycle+0.076
eting+0.075
 leash+0.075
chie+0.074
Neg logit contributions
 narrative-0.087
 maybe-0.077
 museums-0.074
 volcan-0.073
 they-0.071
Token about
Token ID546
Feature activation-2.11e-03

Pos logit contributions
icycle+0.050
chie+0.049
ay+0.049
imm+0.049
run+0.049
Neg logit contributions
 narrative-0.042
 museums-0.038
 maybe-0.038
 volcan-0.037
adies-0.036
Token to
Token ID284
Feature activation+8.29e-03

Pos logit contributions
 narrative+0.005
 museums+0.004
 volcan+0.004
 islands+0.004
 Mercury+0.004
Neg logit contributions
run-0.002
--0.002
icycle-0.002
!-0.002
ow-0.002
Token give
Token ID1577
Feature activation+2.46e-02

Pos logit contributions
run+0.012
icycle+0.012
-+0.011
itten+0.011
ay+0.011
Neg logit contributions
 narrative-0.014
 museums-0.012
 volcan-0.012
 islands-0.011
 maybe-0.011
Token up
Token ID510
Feature activation+2.80e-02

Pos logit contributions
run+0.030
icycle+0.028
chie+0.027
itten+0.027
imm+0.027
Neg logit contributions
 narrative-0.044
 museums-0.041
 volcan-0.040
 maybe-0.040
 islands-0.038
Token the
Token ID262
Feature activation+2.21e-04

Pos logit contributions
run+0.011
-+0.010
!+0.010
 away+0.010
ow+0.010
Neg logit contributions
 narrative-0.014
 museums-0.012
 volcan-0.011
 islands-0.011
 Mercury-0.011
Token door
Token ID3420
Feature activation+8.69e-03

Pos logit contributions
run+0.000
imm+0.000
itten+0.000
icycle+0.000
ow+0.000
Neg logit contributions
 narrative-0.000
 museums-0.000
 volcan-0.000
 islands-0.000
 Mercury-0.000
Token.
Token ID13
Feature activation+6.30e-03

Pos logit contributions
icycle+0.011
imm+0.011
aker+0.011
itten+0.011
chie+0.011
Neg logit contributions
 they-0.013
 maybe-0.013
 volcan-0.012
 narrative-0.012
 museums-0.012
Token But
Token ID887
Feature activation+6.60e-03

Pos logit contributions
run+0.009
icycle+0.008
-+0.008
imm+0.008
itten+0.008
Neg logit contributions
 narrative-0.010
 museums-0.010
 volcan-0.009
 maybe-0.009
 Mercury-0.009
Token when
Token ID618
Feature activation+1.00e-02

Pos logit contributions
run+0.010
-+0.009
!+0.009
 away+0.009
ow+0.009
Neg logit contributions
 narrative-0.011
 museums-0.009
 islands-0.009
 volcan-0.009
lan-0.009
Token she
Token ID673
Feature activation+5.59e-02

Pos logit contributions
run+0.017
!+0.017
 away+0.016
-+0.016
 really+0.016
Neg logit contributions
 narrative-0.014
 islands-0.012
 volcan-0.012
 museums-0.012
lan-0.011
Token got
Token ID1392
Feature activation+2.92e-02

Pos logit contributions
icycle+0.076
run+0.076
eting+0.074
aker+0.074
chie+0.073
Neg logit contributions
 narrative-0.082
 maybe-0.075
 they-0.074
 museums-0.073
 volcan-0.069
Token there
Token ID612
Feature activation+3.08e-02

Pos logit contributions
run+0.032
-+0.030
 away+0.030
ow+0.029
!+0.029
Neg logit contributions
 narrative-0.060
 museums-0.052
 volcan-0.051
 islands-0.051
 Mercury-0.049
Token,
Token ID11
Feature activation+2.28e-02

Pos logit contributions
run+0.047
icycle+0.045
aker+0.043
itten+0.042
eting+0.042
Neg logit contributions
 narrative-0.048
 museums-0.043
 they-0.042
 maybe-0.042
 volcan-0.041
Token it
Token ID340
Feature activation+7.61e-03

Pos logit contributions
run+0.038
!+0.036
-+0.036
 really+0.035
 away+0.035
Neg logit contributions
 narrative-0.034
 islands-0.028
 museums-0.027
 volcan-0.027
lan-0.027
Token was
Token ID373
Feature activation+6.51e-03

Pos logit contributions
run+0.011
ay+0.011
icycle+0.011
itten+0.010
ooter+0.010
Neg logit contributions
 narrative-0.012
 they-0.010
 volcan-0.010
 maybe-0.010
 museums-0.010
Token to
Token ID284
Feature activation+7.40e-03

Pos logit contributions
 narrative+0.008
 museums+0.008
 volcan+0.007
 islands+0.007
 Mercury+0.007
Neg logit contributions
run-0.004
--0.004
icycle-0.004
!-0.004
 away-0.003
Token him
Token ID683
Feature activation+5.24e-04

Pos logit contributions
run+0.011
icycle+0.011
-+0.010
itten+0.010
imm+0.010
Neg logit contributions
 narrative-0.012
 museums-0.010
 volcan-0.010
 islands-0.010
 maybe-0.009
Token and
Token ID290
Feature activation+1.70e-02

Pos logit contributions
chie+0.001
ay+0.001
imm+0.001
icycle+0.001
aker+0.001
Neg logit contributions
 they-0.001
 maybe-0.001
 sometimes-0.000
 islands-0.000
 volcan-0.000
Token asked
Token ID1965
Feature activation-9.27e-03

Pos logit contributions
ay+0.030
icycle+0.029
itten+0.028
run+0.027
chie+0.027
Neg logit contributions
 they-0.024
 maybe-0.021
 museums-0.019
 narrative-0.019
 volcan-0.018
Token if
Token ID611
Feature activation+5.22e-03

Pos logit contributions
 narrative+0.016
 islands+0.013
 museums+0.013
 volcan+0.013
 Mercury+0.013
Neg logit contributions
run-0.013
--0.013
!-0.013
 away-0.012
 really-0.012
Token she
Token ID673
Feature activation+5.55e-02

Pos logit contributions
!+0.007
-+0.007
 too+0.006
 away+0.006
 really+0.006
Neg logit contributions
 narrative-0.010
ooks-0.008
 islands-0.008
 volcan-0.008
 museums-0.008
Token could
Token ID714
Feature activation+3.73e-03

Pos logit contributions
run+0.075
icycle+0.071
 away+0.069
itten+0.068
imm+0.068
Neg logit contributions
 narrative-0.094
 museums-0.083
 volcan-0.082
 maybe-0.080
 islands-0.080
Token play
Token ID711
Feature activation+2.44e-02

Pos logit contributions
run+0.007
icycle+0.007
imm+0.007
aker+0.006
itten+0.006
Neg logit contributions
 narrative-0.004
 maybe-0.004
 museums-0.004
 they-0.004
 volcan-0.003
Token too
Token ID1165
Feature activation+1.35e-02

Pos logit contributions
run+0.036
itten+0.035
icycle+0.035
chie+0.034
itter+0.034
Neg logit contributions
 narrative-0.038
 museums-0.034
 maybe-0.034
 they-0.032
 islands-0.032
Token.
Token ID13
Feature activation+1.46e-03

Pos logit contributions
run+0.014
imm+0.013
icycle+0.013
itten+0.012
chie+0.012
Neg logit contributions
 narrative-0.027
 museums-0.025
 volcan-0.024
 islands-0.024
 maybe-0.024
Token The
Token ID383
Feature activation-2.98e-03

Pos logit contributions
run+0.003
icycle+0.003
itten+0.003
imm+0.003
chie+0.003
Neg logit contributions
 narrative-0.001
 they-0.001
 museums-0.001
 maybe-0.001
 volcan-0.001
Token began
Token ID2540
Feature activation+8.90e-03

Pos logit contributions
run+0.013
icycle+0.012
-+0.011
ay+0.011
itten+0.011
Neg logit contributions
 narrative-0.016
 museums-0.014
 volcan-0.013
 islands-0.013
 Mercury-0.013
Token to
Token ID284
Feature activation-7.06e-03

Pos logit contributions
ay+0.011
icycle+0.011
run+0.011
imm+0.011
itten+0.010
Neg logit contributions
 narrative-0.015
 museums-0.014
 volcan-0.014
 islands-0.014
lan-0.014
Token think
Token ID892
Feature activation-3.10e-03

Pos logit contributions
 narrative+0.011
 museums+0.010
 volcan+0.010
 islands+0.010
 Mercury+0.009
Neg logit contributions
run-0.011
icycle-0.010
ay-0.010
--0.010
itten-0.009
Token.
Token ID13
Feature activation+6.23e-03

Pos logit contributions
 narrative+0.005
 museums+0.005
 volcan+0.005
 maybe+0.004
 islands+0.004
Neg logit contributions
icycle-0.004
run-0.004
imm-0.004
itten-0.004
ow-0.004
Token Before
Token ID7413
Feature activation+6.58e-03

Pos logit contributions
run+0.010
-+0.010
 away+0.010
imm+0.010
icycle+0.010
Neg logit contributions
 narrative-0.008
 museums-0.008
 volcan-0.007
 maybe-0.007
 Mercury-0.007
Token she
Token ID673
Feature activation+5.55e-02

Pos logit contributions
run+0.012
icycle+0.012
-+0.011
ow+0.011
 away+0.011
Neg logit contributions
 narrative-0.008
 museums-0.007
 volcan-0.007
 islands-0.006
 Mercury-0.006
Token knew
Token ID2993
Feature activation+2.29e-02

Pos logit contributions
icycle+0.088
run+0.083
ay+0.081
chie+0.080
 leash+0.080
Neg logit contributions
 narrative-0.081
 maybe-0.075
 volcan-0.069
 they-0.068
 museums-0.067
Token it
Token ID340
Feature activation+3.79e-02

Pos logit contributions
run+0.033
!+0.031
-+0.031
 away+0.031
ow+0.030
Neg logit contributions
 narrative-0.039
 museums-0.033
 islands-0.032
 volcan-0.032
 Mercury-0.031
Token,
Token ID11
Feature activation+1.74e-02

Pos logit contributions
icycle+0.062
aker+0.060
run+0.058
itten+0.057
chie+0.057
Neg logit contributions
 maybe-0.053
 narrative-0.052
 they-0.051
 museums-0.048
 sometimes-0.046
Token she
Token ID673
Feature activation+1.15e-02

Pos logit contributions
run+0.031
-+0.029
icycle+0.029
!+0.028
 away+0.028
Neg logit contributions
 narrative-0.024
 museums-0.020
 volcan-0.019
 islands-0.019
 Mercury-0.018
Token had
Token ID550
Feature activation-1.37e-03

Pos logit contributions
ay+0.019
icycle+0.019
run+0.018
chie+0.018
itten+0.018
Neg logit contributions
 narrative-0.015
 maybe-0.014
 they-0.013
 islands-0.012
 museums-0.012
Token pinch
Token ID25394
Feature activation+1.81e-02

Pos logit contributions
run+0.015
icycle+0.014
ay+0.014
itten+0.013
imm+0.013
Neg logit contributions
 narrative-0.013
 museums-0.012
 volcan-0.012
 islands-0.011
 maybe-0.011
Token it
Token ID340
Feature activation+2.79e-02

Pos logit contributions
run+0.032
icycle+0.032
imm+0.031
itten+0.030
chie+0.030
Neg logit contributions
 narrative-0.023
 museums-0.020
 volcan-0.019
 maybe-0.019
 islands-0.018
Token.
Token ID13
Feature activation+1.54e-03

Pos logit contributions
icycle+0.053
aker+0.046
itten+0.045
chie+0.045
ippy+0.043
Neg logit contributions
 maybe-0.038
 sometimes-0.036
 they-0.036
 if-0.029
 volcan-0.028
Token But
Token ID887
Feature activation+2.44e-03

Pos logit contributions
aker+0.003
run+0.003
imm+0.003
icycle+0.003
chie+0.003
Neg logit contributions
 maybe-0.001
 narrative-0.001
 museums-0.001
 they-0.001
 volcan-0.001
Token when
Token ID618
Feature activation+1.30e-03

Pos logit contributions
run+0.004
icycle+0.004
ay+0.004
imm+0.004
itten+0.004
Neg logit contributions
 narrative-0.003
 museums-0.003
 volcan-0.003
 maybe-0.003
 islands-0.002
Token she
Token ID673
Feature activation+5.55e-02

Pos logit contributions
run+0.002
icycle+0.002
-+0.002
!+0.002
ow+0.002
Neg logit contributions
 narrative-0.002
 museums-0.001
 volcan-0.001
 islands-0.001
 Mercury-0.001
Token did
Token ID750
Feature activation+8.95e-03

Pos logit contributions
icycle+0.078
run+0.075
ay+0.074
eting+0.073
chie+0.073
Neg logit contributions
 narrative-0.087
 maybe-0.076
 volcan-0.072
 they-0.072
 museums-0.072
Token,
Token ID11
Feature activation+1.93e-02

Pos logit contributions
run+0.013
icycle+0.013
ay+0.012
imm+0.012
itten+0.012
Neg logit contributions
 narrative-0.014
 museums-0.013
 volcan-0.013
 maybe-0.012
 they-0.012
Token the
Token ID262
Feature activation-8.83e-03

Pos logit contributions
run+0.038
icycle+0.037
imm+0.036
ay+0.036
itten+0.036
Neg logit contributions
 narrative-0.022
 museums-0.019
 maybe-0.019
 volcan-0.017
 they-0.017
Token bug
Token ID5434
Feature activation+3.40e-03

Pos logit contributions
 narrative+0.015
 museums+0.012
 volcan+0.012
 Mercury+0.012
lan+0.012
Neg logit contributions
run-0.013
icycle-0.012
imm-0.012
ow-0.012
--0.012
Token flew
Token ID13112
Feature activation+1.18e-02

Pos logit contributions
ay+0.006
icycle+0.006
run+0.006
chie+0.005
ippy+0.005
Neg logit contributions
 narrative-0.005
 they-0.004
 maybe-0.004
 volcan-0.004
 sometimes-0.004
Token her
Token ID607
Feature activation-1.04e-02

Pos logit contributions
run+0.032
icycle+0.032
itten+0.031
 leash+0.030
itter+0.030
Neg logit contributions
 narrative-0.026
 museums-0.022
 they-0.022
 volcan-0.022
 planets-0.021
Token bright
Token ID6016
Feature activation-1.43e-02

Pos logit contributions
 narrative+0.010
 planets+0.010
 museums+0.009
 moons+0.009
 islands+0.009
Neg logit contributions
run-0.020
itten-0.020
icycle-0.020
ow-0.020
umb-0.019
Token clothes
Token ID8242
Feature activation+1.44e-02

Pos logit contributions
 narrative+0.019
 islands+0.017
 planets+0.017
 moons+0.017
 museums+0.017
Neg logit contributions
run-0.024
icycle-0.023
ow-0.023
itten-0.023
itter-0.022
Token.
Token ID13
Feature activation+6.66e-04

Pos logit contributions
run+0.020
icycle+0.018
chie+0.018
itten+0.018
ippy+0.018
Neg logit contributions
 they-0.024
 narrative-0.023
 maybe-0.020
lan-0.020
 museums-0.020
Token Before
Token ID7413
Feature activation+4.30e-03

Pos logit contributions
run+0.001
icycle+0.001
-+0.001
!+0.001
ow+0.001
Neg logit contributions
 narrative-0.001
 museums-0.001
 volcan-0.001
 islands-0.001
 Mercury-0.001
Token she
Token ID673
Feature activation+5.54e-02

Pos logit contributions
run+0.008
-+0.007
!+0.007
icycle+0.007
ow+0.007
Neg logit contributions
 narrative-0.006
 museums-0.005
 volcan-0.004
 islands-0.004
 Mercury-0.004
Token knew
Token ID2993
Feature activation+2.26e-02

Pos logit contributions
icycle+0.080
run+0.079
itten+0.075
 leash+0.074
chie+0.074
Neg logit contributions
 narrative-0.090
 maybe-0.078
 museums-0.075
 volcan-0.073
 they-0.073
Token it
Token ID340
Feature activation+4.11e-02

Pos logit contributions
run+0.033
!+0.031
-+0.031
 away+0.031
 really+0.030
Neg logit contributions
 narrative-0.038
 museums-0.032
 volcan-0.031
 islands-0.031
 Mercury-0.030
Token,
Token ID11
Feature activation+1.62e-02

Pos logit contributions
aker+0.069
run+0.068
icycle+0.068
itten+0.068
chie+0.066
Neg logit contributions
 narrative-0.056
 maybe-0.054
 they-0.053
 museums-0.052
 sometimes-0.047
Token her
Token ID607
Feature activation+1.42e-02

Pos logit contributions
run+0.026
 really+0.025
!+0.025
-+0.025
 away+0.024
Neg logit contributions
 narrative-0.024
 volcan-0.021
 islands-0.020
 museums-0.020
adies-0.019
Token closet
Token ID21615
Feature activation+2.02e-02

Pos logit contributions
run+0.027
imm+0.026
aker+0.026
it+0.025
umb+0.025
Neg logit contributions
 narrative-0.017
 museums-0.015
 planets-0.014
 unic-0.014
 maybe-0.013
Token wrap
Token ID14441
Feature activation+4.14e-03

Pos logit contributions
 narrative+0.016
 museums+0.013
 volcan+0.013
 islands+0.013
ighters+0.012
Neg logit contributions
run-0.014
icycle-0.013
--0.013
imm-0.013
itten-0.013
Token some
Token ID617
Feature activation-1.75e-02

Pos logit contributions
run+0.008
icycle+0.007
itten+0.007
imm+0.007
ow+0.007
Neg logit contributions
 narrative-0.005
 museums-0.004
 volcan-0.004
 islands-0.004
lan-0.004
Token tape
Token ID9154
Feature activation-6.51e-03

Pos logit contributions
 narrative+0.026
 museums+0.022
 islands+0.021
 volcan+0.021
 Mercury+0.020
Neg logit contributions
run-0.028
icycle-0.027
ow-0.027
imm-0.026
itten-0.026
Token around
Token ID1088
Feature activation+1.92e-03

Pos logit contributions
 narrative+0.010
 they+0.009
lan+0.009
 islands+0.009
 volcan+0.009
Neg logit contributions
run-0.010
itten-0.009
icycle-0.009
chie-0.009
imm-0.009
Token the
Token ID262
Feature activation-1.56e-02

Pos logit contributions
run+0.003
icycle+0.003
itten+0.003
ow+0.003
imm+0.003
Neg logit contributions
 narrative-0.003
 museums-0.002
 volcan-0.002
 islands-0.002
 Mercury-0.002
Token prope
Token ID44408
Feature activation-7.44e-02

Pos logit contributions
 narrative+0.027
 museums+0.023
 islands+0.023
 volcan+0.023
 Mercury+0.021
Neg logit contributions
run-0.023
ow-0.021
imm-0.020
icycle-0.020
itten-0.019
Tokenller
Token ID6051
Feature activation-1.37e-02

Pos logit contributions
 narrative+0.157
 museums+0.140
 volcan+0.137
 islands+0.136
 Mercury+0.133
Neg logit contributions
run-0.075
icycle-0.066
--0.066
!-0.064
ow-0.064
Token and
Token ID290
Feature activation-4.91e-03

Pos logit contributions
 narrative+0.022
 volcan+0.020
 they+0.020
 museums+0.020
lan+0.020
Neg logit contributions
run-0.019
icycle-0.018
imm-0.017
chie-0.017
itten-0.017
Token attach
Token ID10199
Feature activation+1.19e-03

Pos logit contributions
 narrative+0.009
 museums+0.008
 volcan+0.008
adies+0.007
 islands+0.007
Neg logit contributions
run-0.006
 away-0.006
!-0.006
--0.006
 really-0.006
Token it
Token ID340
Feature activation+5.99e-03

Pos logit contributions
run+0.002
icycle+0.002
itten+0.002
imm+0.002
ivering+0.002
Neg logit contributions
 narrative-0.001
 volcan-0.001
 islands-0.001
 museums-0.001
lan-0.001
Token to
Token ID284
Feature activation+3.66e-03

Pos logit contributions
icycle+0.009
itten+0.009
run+0.009
ivering+0.008
aker+0.008
Neg logit contributions
 narrative-0.008
 they-0.008
lan-0.008
 volcan-0.008
 museums-0.008
Token,
Token ID11
Feature activation-2.89e-02

Pos logit contributions
 narrative+0.090
 museums+0.079
 Mercury+0.076
 volcan+0.076
 islands+0.076
Neg logit contributions
!-0.043
run-0.042
--0.040
 really-0.038
 away-0.038
Token fifteen
Token ID17280
Feature activation-5.63e-02

Pos logit contributions
 narrative+0.040
 museums+0.033
 islands+0.032
 Mercury+0.031
 volcan+0.030
Neg logit contributions
run-0.049
!-0.048
--0.047
 really-0.047
 away-0.046
Token,
Token ID11
Feature activation-3.28e-02

Pos logit contributions
 narrative+0.106
 museums+0.095
 volcan+0.095
 islands+0.092
adies+0.088
Neg logit contributions
run-0.069
icycle-0.064
ow-0.061
imm-0.060
itten-0.059
Token sixteen
Token ID27913
Feature activation-5.27e-02

Pos logit contributions
 narrative+0.043
 museums+0.036
 islands+0.033
 volcan+0.033
 Mercury+0.032
Neg logit contributions
run-0.059
--0.056
!-0.056
 away-0.055
 really-0.055
Token,
Token ID11
Feature activation-3.24e-02

Pos logit contributions
 narrative+0.105
 museums+0.093
 volcan+0.091
 islands+0.091
 Mercury+0.089
Neg logit contributions
run-0.058
--0.053
!-0.053
icycle-0.052
 away-0.051
Token seventeen
Token ID38741
Feature activation-6.80e-02

Pos logit contributions
 narrative+0.042
 museums+0.035
 islands+0.033
 volcan+0.033
 Mercury+0.031
Neg logit contributions
run-0.059
--0.056
!-0.055
 away-0.055
 really-0.055
Token,
Token ID11
Feature activation-3.30e-02

Pos logit contributions
 narrative+0.126
 museums+0.110
 volcan+0.109
 islands+0.107
 Mercury+0.102
Neg logit contributions
run-0.089
icycle-0.079
ow-0.076
itten-0.075
 away-0.075
Token eighteen
Token ID29095
Feature activation-5.64e-02

Pos logit contributions
 narrative+0.041
 museums+0.034
 islands+0.031
 volcan+0.031
 Mercury+0.030
Neg logit contributions
run-0.062
--0.059
!-0.059
 away-0.058
 really-0.057
Token,
Token ID11
Feature activation-3.00e-02

Pos logit contributions
 narrative+0.104
 museums+0.093
 volcan+0.092
 islands+0.090
 Mercury+0.086
Neg logit contributions
run-0.072
icycle-0.066
ow-0.063
 away-0.062
itten-0.062
Token nineteen
Token ID43063
Feature activation-2.62e-02

Pos logit contributions
 narrative+0.038
 museums+0.031
 islands+0.029
 volcan+0.029
 Mercury+0.028
Neg logit contributions
run-0.056
--0.053
!-0.053
 away-0.052
 really-0.052
Token,
Token ID11
Feature activation-2.74e-02

Pos logit contributions
 narrative+0.048
 museums+0.042
 volcan+0.042
 islands+0.041
 maybe+0.040
Neg logit contributions
run-0.034
icycle-0.033
itten-0.031
ow-0.030
imm-0.030
Token had
Token ID550
Feature activation-2.02e-02

Pos logit contributions
 they+0.008
 narrative+0.007
 maybe+0.007
 sometimes+0.007
 volcan+0.007
Neg logit contributions
run-0.012
itter-0.012
eting-0.012
itten-0.011
ow-0.011
Token been
Token ID587
Feature activation-1.30e-02

Pos logit contributions
 narrative+0.028
 museums+0.026
 volcan+0.026
 islands+0.025
 planets+0.024
Neg logit contributions
run-0.033
imm-0.031
icycle-0.031
itter-0.031
itten-0.031
Token made
Token ID925
Feature activation+6.99e-03

Pos logit contributions
 they+0.017
lan+0.016
 islands+0.016
 volcan+0.016
 pirates+0.016
Neg logit contributions
itten-0.019
icycle-0.019
ay-0.018
run-0.018
itter-0.017
Token many
Token ID867
Feature activation-3.46e-02

Pos logit contributions
icycle+0.013
itten+0.012
oodle+0.012
run+0.012
 leash+0.011
Neg logit contributions
 they-0.008
 Mercury-0.007
adies-0.007
 maybe-0.007
 museums-0.007
Token years
Token ID812
Feature activation-8.21e-03

Pos logit contributions
 narrative+0.050
 museums+0.045
 islands+0.043
 volcan+0.042
 planets+0.041
Neg logit contributions
run-0.059
ow-0.059
itter-0.055
itten-0.054
atter-0.054
Token ago
Token ID2084
Feature activation-6.79e-02

Pos logit contributions
 narrative+0.013
 museums+0.012
lan+0.012
 volcan+0.012
 islands+0.012
Neg logit contributions
run-0.011
icycle-0.011
ow-0.011
imm-0.010
itten-0.010
Token,
Token ID11
Feature activation+9.97e-03

Pos logit contributions
 narrative+0.118
 volcan+0.113
 museums+0.111
 they+0.109
 islands+0.108
Neg logit contributions
run-0.081
chie-0.077
icycle-0.074
ay-0.073
imm-0.073
Token but
Token ID475
Feature activation+9.11e-03

Pos logit contributions
run+0.015
icycle+0.014
ay+0.014
ow+0.014
 away+0.014
Neg logit contributions
 narrative-0.015
 museums-0.014
 volcan-0.014
 islands-0.013
 maybe-0.013
Token it
Token ID340
Feature activation+9.98e-03

Pos logit contributions
run+0.015
icycle+0.015
itten+0.014
ay+0.014
ow+0.014
Neg logit contributions
 narrative-0.013
 museums-0.011
 volcan-0.011
 islands-0.011
 maybe-0.011
Token still
Token ID991
Feature activation+1.47e-02

Pos logit contributions
run+0.018
ay+0.017
eting+0.017
itter+0.016
itten+0.016
Neg logit contributions
 narrative-0.012
 they-0.011
 sometimes-0.011
 maybe-0.011
 volcan-0.011
Token looked
Token ID3114
Feature activation-1.19e-02

Pos logit contributions
ippy+0.030
ay+0.029
icycle+0.028
chie+0.027
itten+0.027
Neg logit contributions
 they-0.015
 maybe-0.014
 planets-0.013
 sometimes-0.013
 moons-0.013
Token,
Token ID11
Feature activation-1.36e-02

Pos logit contributions
 narrative+0.010
 they+0.010
 volcan+0.009
 islands+0.009
 maybe+0.009
Neg logit contributions
icycle-0.011
run-0.011
itten-0.010
aker-0.010
ippy-0.009
Token the
Token ID262
Feature activation-7.50e-03

Pos logit contributions
 narrative+0.029
 museums+0.025
adies+0.024
lan+0.024
ooks+0.024
Neg logit contributions
run-0.014
 away-0.014
 too-0.013
 very-0.013
!-0.013
Token nose
Token ID9686
Feature activation-6.65e-04

Pos logit contributions
 narrative+0.012
 museums+0.010
 volcan+0.010
 islands+0.010
 Mercury+0.009
Neg logit contributions
run-0.012
icycle-0.011
imm-0.011
ow-0.011
--0.011
Token,
Token ID11
Feature activation-1.35e-02

Pos logit contributions
 they+0.001
 narrative+0.001
 volcan+0.001
 maybe+0.001
 islands+0.001
Neg logit contributions
icycle-0.001
run-0.001
chie-0.001
 package-0.001
itten-0.001
Token the
Token ID262
Feature activation-8.03e-03

Pos logit contributions
 narrative+0.030
 museums+0.025
adies+0.025
lan+0.025
ooks+0.025
Neg logit contributions
run-0.013
 away-0.012
!-0.012
--0.012
 too-0.012
Token prope
Token ID44408
Feature activation-6.77e-02

Pos logit contributions
 narrative+0.013
 museums+0.011
 volcan+0.011
 islands+0.011
 planets+0.010
Neg logit contributions
run-0.013
icycle-0.012
imm-0.012
ow-0.012
--0.011
Tokenller
Token ID6051
Feature activation-2.20e-02

Pos logit contributions
 narrative+0.126
 islands+0.112
 museums+0.112
 volcan+0.112
lan+0.108
Neg logit contributions
run-0.082
icycle-0.070
 really-0.070
 away-0.070
ow-0.070
Token.
Token ID13
Feature activation-3.20e-03

Pos logit contributions
 narrative+0.034
 they+0.033
 volcan+0.032
 maybe+0.031
 islands+0.031
Neg logit contributions
icycle-0.032
run-0.032
chie-0.029
itten-0.028
imm-0.028
Token They
Token ID1119
Feature activation+1.58e-04

Pos logit contributions
 narrative+0.005
 museums+0.004
 islands+0.004
adies+0.004
 volcan+0.004
Neg logit contributions
run-0.006
ay-0.005
 really-0.005
!-0.005
 too-0.005
Token work
Token ID670
Feature activation-1.34e-02

Pos logit contributions
run+0.000
!+0.000
 away+0.000
 really+0.000
-+0.000
Neg logit contributions
 narrative-0.000
 museums-0.000
 volcan-0.000
 islands-0.000
 Mercury-0.000
Token hard
Token ID1327
Feature activation-6.82e-03

Pos logit contributions
 narrative+0.022
 museums+0.019
 volcan+0.019
 islands+0.018
 Mercury+0.018
Neg logit contributions
run-0.019
icycle-0.019
itten-0.019
imm-0.018
ow-0.018
Token,
Token ID11
Feature activation-2.89e-02

Pos logit contributions
 narrative+0.089
 museums+0.078
 islands+0.076
 Mercury+0.075
 volcan+0.075
Neg logit contributions
!-0.045
run-0.045
--0.043
 really-0.042
 away-0.041
Token fifteen
Token ID17280
Feature activation-5.73e-02

Pos logit contributions
 narrative+0.043
 museums+0.035
 islands+0.034
 Mercury+0.033
ooks+0.032
Neg logit contributions
!-0.046
run-0.046
 really-0.046
--0.045
 too-0.045
Token,
Token ID11
Feature activation-3.28e-02

Pos logit contributions
 narrative+0.107
 museums+0.096
 volcan+0.096
 islands+0.093
 maybe+0.088
Neg logit contributions
run-0.072
icycle-0.066
ow-0.064
itten-0.063
imm-0.062
Token sixteen
Token ID27913
Feature activation-5.31e-02

Pos logit contributions
 narrative+0.047
 museums+0.040
 islands+0.038
 Mercury+0.036
 volcan+0.036
Neg logit contributions
run-0.054
!-0.053
--0.053
 really-0.052
ay-0.051
Token,
Token ID11
Feature activation-3.26e-02

Pos logit contributions
 narrative+0.104
 museums+0.092
 volcan+0.090
 islands+0.089
 Mercury+0.087
Neg logit contributions
run-0.061
icycle-0.055
--0.055
ow-0.053
!-0.053
Token seventeen
Token ID38741
Feature activation-6.77e-02

Pos logit contributions
 narrative+0.046
 museums+0.039
 islands+0.036
 Mercury+0.035
 volcan+0.035
Neg logit contributions
run-0.055
--0.053
!-0.053
 really-0.053
ay-0.052
Token,
Token ID11
Feature activation-3.31e-02

Pos logit contributions
 narrative+0.126
 museums+0.111
 volcan+0.109
 islands+0.107
 Mercury+0.102
Neg logit contributions
run-0.087
icycle-0.078
ow-0.075
itten-0.075
 away-0.074
Token eighteen
Token ID29095
Feature activation-5.66e-02

Pos logit contributions
 narrative+0.045
 museums+0.038
 islands+0.035
 Mercury+0.034
 volcan+0.034
Neg logit contributions
run-0.057
--0.056
!-0.056
 really-0.055
 away-0.054
Token,
Token ID11
Feature activation-2.99e-02

Pos logit contributions
 narrative+0.106
 museums+0.095
 volcan+0.094
 islands+0.092
 Mercury+0.088
Neg logit contributions
run-0.071
icycle-0.064
ow-0.062
itten-0.061
 away-0.061
Token nineteen
Token ID43063
Feature activation-2.68e-02

Pos logit contributions
 narrative+0.040
 museums+0.033
 islands+0.031
 Mercury+0.030
 volcan+0.030
Neg logit contributions
run-0.053
!-0.051
--0.051
 really-0.050
 away-0.049
Token,
Token ID11
Feature activation-2.50e-02

Pos logit contributions
 narrative+0.050
 museums+0.044
 volcan+0.044
 islands+0.043
 maybe+0.042
Neg logit contributions
run-0.034
icycle-0.032
itten-0.031
ow-0.030
imm-0.029
Token ac
Token ID936
Feature activation-2.46e-02

Pos logit contributions
 narrative+0.021
 maybe+0.020
 unic+0.019
 tul+0.019
 sometimes+0.019
Neg logit contributions
run-0.021
ow-0.021
eting-0.021
itter-0.020
icycle-0.020
Tokenorn
Token ID1211
Feature activation-5.60e-03

Pos logit contributions
 narrative+0.039
lan+0.031
 maybe+0.030
 museums+0.030
adies+0.030
Neg logit contributions
run-0.035
eting-0.034
chie-0.032
ow-0.031
itter-0.031
Token a
Token ID257
Feature activation-7.61e-04

Pos logit contributions
 narrative+0.008
 maybe+0.008
 they+0.007
 volcan+0.007
 sometimes+0.007
Neg logit contributions
run-0.008
icycle-0.008
ay-0.008
chie-0.008
itten-0.008
Token few
Token ID1178
Feature activation-2.75e-02

Pos logit contributions
 narrative+0.001
 tul+0.001
lan+0.001
 volcan+0.001
 unic+0.001
Neg logit contributions
ow-0.001
run-0.001
imm-0.001
icycle-0.001
acked-0.001
Token days
Token ID1528
Feature activation-1.89e-02

Pos logit contributions
 narrative+0.043
 volcan+0.037
 museums+0.037
 islands+0.037
lan+0.035
Neg logit contributions
icycle-0.040
run-0.040
ow-0.039
imm-0.038
itter-0.038
Token ago
Token ID2084
Feature activation-6.74e-02

Pos logit contributions
 narrative+0.031
 museums+0.028
 islands+0.027
lan+0.027
 volcan+0.026
Neg logit contributions
run-0.025
icycle-0.024
imm-0.024
itten-0.023
ay-0.023
Token so
Token ID523
Feature activation-1.54e-02

Pos logit contributions
 narrative+0.114
 volcan+0.102
adies+0.102
 museums+0.101
 they+0.101
Neg logit contributions
run-0.084
icycle-0.080
imm-0.079
chie-0.078
itten-0.077
Token I
Token ID314
Feature activation-6.30e-03

Pos logit contributions
 narrative+0.021
 museums+0.018
 maybe+0.017
 volcan+0.017
lan+0.017
Neg logit contributions
run-0.026
icycle-0.025
imm-0.025
ow-0.024
itten-0.024
Token could
Token ID714
Feature activation-1.75e-02

Pos logit contributions
 narrative+0.008
 maybe+0.007
 museums+0.007
ighters+0.007
 they+0.007
Neg logit contributions
icycle-0.011
run-0.010
itten-0.010
ippy-0.010
 leash-0.010
Token have
Token ID423
Feature activation-2.26e-03

Pos logit contributions
 narrative+0.023
 museums+0.019
 volcan+0.019
 maybe+0.019
 islands+0.018
Neg logit contributions
run-0.031
icycle-0.030
ay-0.029
itten-0.028
imm-0.028
Token some
Token ID617
Feature activation-3.19e-02

Pos logit contributions
 narrative+0.003
 museums+0.003
 volcan+0.003
 islands+0.003
 maybe+0.003
Neg logit contributions
run-0.004
icycle-0.004
itter-0.004
ay-0.004
itten-0.004
Token I
Token ID314
Feature activation-4.77e-03

Pos logit contributions
run+0.005
ow+0.005
icycle+0.005
ay+0.005
!+0.005
Neg logit contributions
 narrative-0.006
 museums-0.005
 islands-0.005
 volcan-0.005
 Mercury-0.005
Token built
Token ID3170
Feature activation+8.62e-03

Pos logit contributions
 narrative+0.007
 maybe+0.006
 museums+0.005
 volcan+0.005
 they+0.005
Neg logit contributions
icycle-0.008
run-0.008
itten-0.007
ay-0.007
 leash-0.007
Token it
Token ID340
Feature activation+8.65e-05

Pos logit contributions
run+0.013
icycle+0.012
itten+0.012
itter+0.012
imm+0.012
Neg logit contributions
 narrative-0.013
 volcan-0.012
 museums-0.012
 islands-0.012
 Mercury-0.011
Token many
Token ID867
Feature activation-3.40e-02

Pos logit contributions
run+0.000
icycle+0.000
itten+0.000
ay+0.000
aker+0.000
Neg logit contributions
 narrative-0.000
 museums-0.000
 volcan-0.000
 islands-0.000
 Mercury-0.000
Token years
Token ID812
Feature activation-1.18e-02

Pos logit contributions
 narrative+0.056
 museums+0.049
 volcan+0.048
 islands+0.047
 Mercury+0.045
Neg logit contributions
run-0.051
icycle-0.047
ow-0.047
itten-0.046
itter-0.045
Token ago
Token ID2084
Feature activation-6.71e-02

Pos logit contributions
 narrative+0.021
 museums+0.019
 volcan+0.019
 islands+0.019
lan+0.019
Neg logit contributions
run-0.015
icycle-0.014
itten-0.014
imm-0.014
ow-0.014
Token and
Token ID290
Feature activation-6.33e-03

Pos logit contributions
 narrative+0.110
 volcan+0.104
 museums+0.101
 islands+0.100
 they+0.098
Neg logit contributions
run-0.088
icycle-0.082
itten-0.081
imm-0.080
chie-0.080
Token I
Token ID314
Feature activation+1.44e-03

Pos logit contributions
 narrative+0.009
 museums+0.008
 volcan+0.008
 maybe+0.008
 islands+0.008
Neg logit contributions
run-0.010
icycle-0.010
itten-0.010
imm-0.009
ow-0.009
Token am
Token ID716
Feature activation+9.74e-03

Pos logit contributions
icycle+0.002
run+0.002
itten+0.002
ippy+0.002
 leash+0.002
Neg logit contributions
 narrative-0.002
 maybe-0.002
 volcan-0.002
 museums-0.002
 they-0.002
Token very
Token ID845
Feature activation+2.82e-02

Pos logit contributions
icycle+0.014
itten+0.013
run+0.013
ay+0.013
imm+0.013
Neg logit contributions
 volcan-0.015
 maybe-0.015
 God-0.014
 narrative-0.014
 they-0.014
Token proud
Token ID6613
Feature activation+8.23e-03

Pos logit contributions
icycle+0.039
run+0.039
itten+0.038
ay+0.037
imm+0.037
Neg logit contributions
 narrative-0.045
 volcan-0.041
 museums-0.041
lan-0.040
 maybe-0.040
Token.
Token ID13
Feature activation-5.91e-04

Pos logit contributions
icycle+0.012
itten+0.012
itter+0.012
chie+0.012
aker+0.011
Neg logit contributions
 they-0.012
 islands-0.012
 narrative-0.011
lan-0.011
 volcan-0.011
Token The
Token ID383
Feature activation-4.72e-03

Pos logit contributions
 narrative+0.001
 museums+0.001
 volcan+0.001
 maybe+0.001
 Mercury+0.001
Neg logit contributions
run-0.001
itten-0.001
icycle-0.001
--0.001
imm-0.001
Token helicopter
Token ID15392
Feature activation-6.58e-03

Pos logit contributions
 narrative+0.010
 museums+0.009
 volcan+0.008
 islands+0.008
lan+0.008
Neg logit contributions
run-0.005
--0.005
icycle-0.005
 away-0.004
!-0.004
Token had
Token ID550
Feature activation-1.82e-02

Pos logit contributions
 narrative+0.010
 they+0.009
 volcan+0.009
 islands+0.009
 museums+0.009
Neg logit contributions
run-0.010
ay-0.009
icycle-0.009
itten-0.009
chie-0.009
Token big
Token ID1263
Feature activation-1.89e-02

Pos logit contributions
 volcan+0.024
 narrative+0.023
 islands+0.023
 museums+0.023
 Mercury+0.022
Neg logit contributions
run-0.030
icycle-0.029
imm-0.029
itten-0.029
itter-0.029
Token prope
Token ID44408
Feature activation-6.63e-02

Pos logit contributions
 narrative+0.039
 museums+0.034
 volcan+0.033
 islands+0.033
lan+0.033
Neg logit contributions
run-0.020
--0.019
!-0.018
icycle-0.018
 away-0.018
Tokenllers
Token ID13802
Feature activation-1.71e-02

Pos logit contributions
 narrative+0.130
 museums+0.114
 volcan+0.112
 islands+0.110
 Mercury+0.108
Neg logit contributions
run-0.078
--0.071
icycle-0.071
!-0.069
ow-0.068
Token that
Token ID326
Feature activation-6.55e-03

Pos logit contributions
 they+0.023
 islands+0.019
 volcan+0.019
 sometimes+0.019
 maybe+0.018
Neg logit contributions
chie-0.030
icycle-0.029
run-0.028
aker-0.027
itter-0.027
Token spun
Token ID26843
Feature activation+6.16e-03

Pos logit contributions
 narrative+0.011
 volcan+0.010
 museums+0.010
 maybe+0.009
 islands+0.009
Neg logit contributions
run-0.010
icycle-0.009
itten-0.009
chie-0.009
ippy-0.009
Token around
Token ID1088
Feature activation-1.84e-03

Pos logit contributions
itten+0.009
run+0.009
icycle+0.008
ippy+0.008
itter+0.008
Neg logit contributions
 planets-0.010
 narrative-0.010
 islands-0.010
 volcan-0.009
 museums-0.009
Token and
Token ID290
Feature activation+2.77e-04

Pos logit contributions
 narrative+0.003
 volcan+0.003
 museums+0.003
 islands+0.003
 Mercury+0.003
Neg logit contributions
run-0.002
icycle-0.002
itten-0.002
imm-0.002
ow-0.002
Token touch
Token ID3638
Feature activation-7.17e-03

Pos logit contributions
 narrative+0.006
 museums+0.005
 islands+0.005
 volcan+0.005
 Mercury+0.005
Neg logit contributions
run-0.007
icycle-0.006
--0.006
itten-0.006
imm-0.006
Token the
Token ID262
Feature activation-7.02e-05

Pos logit contributions
 narrative+0.010
 museums+0.008
 volcan+0.008
 islands+0.008
 Mercury+0.008
Neg logit contributions
run-0.012
icycle-0.012
itten-0.011
imm-0.011
chie-0.011
Token wings
Token ID12098
Feature activation-1.60e-02

Pos logit contributions
 narrative+0.000
 museums+0.000
 volcan+0.000
 islands+0.000
 Mercury+0.000
Neg logit contributions
run-0.000
icycle-0.000
ow-0.000
--0.000
imm-0.000
Token and
Token ID290
Feature activation+1.73e-03

Pos logit contributions
 narrative+0.025
 they+0.023
 islands+0.022
 volcan+0.022
 maybe+0.022
Neg logit contributions
icycle-0.024
run-0.023
chie-0.022
imm-0.021
itten-0.021
Token the
Token ID262
Feature activation-6.43e-03

Pos logit contributions
 away+0.002
 really+0.002
!+0.002
run+0.002
ow+0.002
Neg logit contributions
 narrative-0.003
ooks-0.003
 museums-0.003
adies-0.003
 volcan-0.003
Token prope
Token ID44408
Feature activation-6.57e-02

Pos logit contributions
 narrative+0.010
 museums+0.008
 volcan+0.008
 islands+0.008
 Mercury+0.008
Neg logit contributions
run-0.011
icycle-0.010
--0.010
ow-0.010
imm-0.009
Tokenller
Token ID6051
Feature activation-1.76e-02

Pos logit contributions
 narrative+0.140
 museums+0.125
 volcan+0.123
 islands+0.121
 Mercury+0.119
Neg logit contributions
run-0.065
icycle-0.057
--0.057
!-0.056
ow-0.055
Token and
Token ID290
Feature activation+9.97e-03

Pos logit contributions
 narrative+0.028
 islands+0.025
 museums+0.025
 volcan+0.025
 they+0.024
Neg logit contributions
run-0.026
icycle-0.026
imm-0.024
chie-0.024
itten-0.024
Token the
Token ID262
Feature activation-9.35e-03

Pos logit contributions
 away+0.013
run+0.013
!+0.012
 really+0.012
ow+0.012
Neg logit contributions
 narrative-0.018
ooks-0.016
 museums-0.015
 volcan-0.015
adies-0.015
Token wheels
Token ID13666
Feature activation-1.44e-02

Pos logit contributions
 narrative+0.014
 museums+0.012
 islands+0.012
 volcan+0.012
 Mercury+0.011
Neg logit contributions
run-0.016
icycle-0.014
ow-0.014
imm-0.014
--0.014
Token.
Token ID13
Feature activation-9.33e-03

Pos logit contributions
 narrative+0.024
 they+0.024
 islands+0.022
 maybe+0.022
 museums+0.021
Neg logit contributions
icycle-0.019
run-0.019
imm-0.017
chie-0.017
itten-0.017
Token,
Token ID11
Feature activation-2.23e-02

Pos logit contributions
 narrative+0.071
 museums+0.060
 volcan+0.060
 islands+0.060
adies+0.058
Neg logit contributions
run-0.063
icycle-0.059
itter-0.056
ooter-0.055
eting-0.055
Token twenty
Token ID8208
Feature activation-6.07e-02

Pos logit contributions
 narrative+0.032
 museums+0.027
 volcan+0.027
 islands+0.026
 Mercury+0.025
Neg logit contributions
run-0.037
--0.034
icycle-0.034
!-0.034
 away-0.034
Token-
Token ID12
Feature activation-1.81e-02

Pos logit contributions
 narrative+0.116
 volcan+0.106
 museums+0.104
 islands+0.101
lan+0.101
Neg logit contributions
run-0.071
icycle-0.067
ow-0.061
eting-0.060
 terribly-0.059
Tokenfour
Token ID14337
Feature activation-2.26e-02

Pos logit contributions
 narrative+0.025
 museums+0.021
 volcan+0.020
 islands+0.020
lan+0.019
Neg logit contributions
run-0.031
icycle-0.030
--0.029
ow-0.029
!-0.029
Token,
Token ID11
Feature activation-2.81e-02

Pos logit contributions
 narrative+0.039
 volcan+0.033
 museums+0.033
 islands+0.032
adies+0.031
Neg logit contributions
run-0.032
icycle-0.031
itten-0.029
ooter-0.029
imm-0.028
Token twenty
Token ID8208
Feature activation-6.53e-02

Pos logit contributions
 narrative+0.044
 museums+0.038
 islands+0.036
ighters+0.036
 volcan+0.035
Neg logit contributions
run-0.043
!-0.041
--0.040
 too-0.040
 really-0.040
Token-
Token ID12
Feature activation-1.79e-02

Pos logit contributions
 narrative+0.158
 volcan+0.144
 museums+0.142
lan+0.139
 islands+0.139
Neg logit contributions
run-0.045
icycle-0.040
ow-0.033
 away-0.033
eting-0.033
Tokenfive
Token ID13261
Feature activation-4.88e-02

Pos logit contributions
 narrative+0.027
 museums+0.023
 volcan+0.022
 islands+0.022
 Mercury+0.021
Neg logit contributions
run-0.029
--0.027
icycle-0.027
!-0.026
ow-0.026
Token,
Token ID11
Feature activation-2.15e-02

Pos logit contributions
 narrative+0.077
 volcan+0.069
 islands+0.068
 museums+0.067
lan+0.065
Neg logit contributions
icycle-0.073
run-0.072
itten-0.070
ooter-0.067
 leash-0.066
Token twenty
Token ID8208
Feature activation-6.33e-02

Pos logit contributions
 narrative+0.034
 museums+0.030
 islands+0.028
 volcan+0.028
ighters+0.027
Neg logit contributions
run-0.032
!-0.031
--0.031
 really-0.030
 too-0.030
Token-
Token ID12
Feature activation-2.05e-02

Pos logit contributions
 narrative+0.136
 volcan+0.124
 museums+0.121
 islands+0.119
lan+0.118
Neg logit contributions
run-0.060
icycle-0.055
ow-0.049
eting-0.049
 away-0.048
Token wing
Token ID8539
Feature activation-2.21e-02

Pos logit contributions
 narrative+0.022
 museums+0.019
 volcan+0.018
 islands+0.018
 planets+0.018
Neg logit contributions
run-0.022
imm-0.020
ow-0.020
icycle-0.020
itter-0.019
Token came
Token ID1625
Feature activation-1.37e-02

Pos logit contributions
 they+0.024
 narrative+0.024
 sometimes+0.023
 maybe+0.022
 bridges+0.020
Neg logit contributions
run-0.039
chie-0.039
ippy-0.037
icycle-0.037
eting-0.036
Token off
Token ID572
Feature activation-1.85e-03

Pos logit contributions
 narrative+0.022
 volcan+0.021
 islands+0.021
 museums+0.021
 Mercury+0.021
Neg logit contributions
run-0.017
icycle-0.016
aker-0.016
itten-0.016
imm-0.016
Token and
Token ID290
Feature activation+7.64e-03

Pos logit contributions
 they+0.003
 volcan+0.002
 narrative+0.002
 museums+0.002
 islands+0.002
Neg logit contributions
icycle-0.003
itten-0.003
imm-0.003
chie-0.003
run-0.003
Token the
Token ID262
Feature activation-1.79e-02

Pos logit contributions
run+0.011
icycle+0.010
ay+0.010
itten+0.010
-+0.010
Neg logit contributions
 narrative-0.013
 museums-0.011
 volcan-0.011
 islands-0.011
 Mercury-0.011
Token prope
Token ID44408
Feature activation-6.51e-02

Pos logit contributions
 narrative+0.032
 museums+0.028
 volcan+0.027
 islands+0.027
 Mercury+0.026
Neg logit contributions
run-0.025
icycle-0.022
ow-0.022
imm-0.022
--0.022
Tokenller
Token ID6051
Feature activation-8.87e-03

Pos logit contributions
 narrative+0.119
 museums+0.106
 islands+0.106
 volcan+0.105
lan+0.104
Neg logit contributions
run-0.080
ow-0.068
 away-0.067
 really-0.067
icycle-0.067
Token fell
Token ID3214
Feature activation-4.34e-03

Pos logit contributions
 narrative+0.012
 they+0.011
 volcan+0.011
 maybe+0.011
 museums+0.010
Neg logit contributions
run-0.015
icycle-0.014
chie-0.014
ippy-0.014
itten-0.014
Token down
Token ID866
Feature activation+8.92e-03

Pos logit contributions
 they+0.006
 sometimes+0.006
 bridges+0.006
 narrative+0.006
 volcan+0.006
Neg logit contributions
icycle-0.006
aker-0.006
run-0.006
itten-0.006
imm-0.006
Token.
Token ID13
Feature activation+1.05e-03

Pos logit contributions
run+0.011
itten+0.010
icycle+0.010
imm+0.010
ooter+0.010
Neg logit contributions
 narrative-0.016
 volcan-0.015
 museums-0.015
 they-0.015
 islands-0.014
Token Sam
Token ID3409
Feature activation-2.00e-03

Pos logit contributions
run+0.002
icycle+0.002
 away+0.002
chie+0.002
-+0.002
Neg logit contributions
 narrative-0.001
 volcan-0.001
 museums-0.001
 maybe-0.001
 Mercury-0.001
Token books
Token ID3835
Feature activation-9.28e-03

Pos logit contributions
 narrative+0.022
adies+0.019
ooks+0.018
lan+0.018
ighters+0.017
Neg logit contributions
 really-0.015
 very-0.015
 too-0.013
run-0.013
 so-0.013
Token,
Token ID11
Feature activation-1.55e-02

Pos logit contributions
 they+0.015
 narrative+0.014
 museums+0.013
 maybe+0.013
 sometimes+0.013
Neg logit contributions
chie-0.014
run-0.014
itten-0.013
icycle-0.013
itter-0.012
Token clothes
Token ID8242
Feature activation-1.17e-02

Pos logit contributions
 narrative+0.033
lan+0.027
ooks+0.027
adies+0.027
 museums+0.026
Neg logit contributions
 really-0.017
 too-0.017
 very-0.017
!-0.016
run-0.016
Token from
Token ID422
Feature activation-2.27e-02

Pos logit contributions
 they+0.018
 narrative+0.017
 museums+0.016
 islands+0.016
 maybe+0.015
Neg logit contributions
run-0.018
chie-0.017
icycle-0.017
itten-0.017
imm-0.016
Token long
Token ID890
Feature activation-4.44e-02

Pos logit contributions
 narrative+0.033
 museums+0.030
 volcan+0.029
 islands+0.029
 Mercury+0.027
Neg logit contributions
run-0.037
icycle-0.035
itten-0.035
imm-0.034
ow-0.034
Token ago
Token ID2084
Feature activation-6.49e-02

Pos logit contributions
 narrative+0.079
 museums+0.069
 volcan+0.068
 islands+0.067
 Mercury+0.065
Neg logit contributions
run-0.059
--0.054
icycle-0.054
!-0.053
 away-0.052
Token,
Token ID11
Feature activation-6.49e-03

Pos logit contributions
 narrative+0.120
 volcan+0.111
 museums+0.110
 islands+0.109
 they+0.107
Neg logit contributions
run-0.076
chie-0.069
icycle-0.069
itten-0.068
imm-0.068
Token and
Token ID290
Feature activation-1.11e-02

Pos logit contributions
 narrative+0.014
adies+0.012
lan+0.012
 museums+0.012
ooks+0.012
Neg logit contributions
 really-0.006
 very-0.006
--0.006
 too-0.006
!-0.006
Token special
Token ID2041
Feature activation+4.33e-03

Pos logit contributions
 narrative+0.020
adies+0.017
 museums+0.016
ooks+0.016
 volcan+0.016
Neg logit contributions
run-0.015
 really-0.014
 very-0.014
--0.014
 away-0.014
Token family
Token ID1641
Feature activation+5.66e-03

Pos logit contributions
run+0.007
itten+0.007
imm+0.007
icycle+0.007
ow+0.007
Neg logit contributions
 narrative-0.006
 museums-0.005
 islands-0.005
 volcan-0.005
 planets-0.005
Token keeps
Token ID7622
Feature activation-2.94e-03

Pos logit contributions
itten+0.008
imm+0.008
chie+0.008
icycle+0.008
ivered+0.008
Neg logit contributions
 they-0.009
lan-0.008
 museums-0.008
 islands-0.008
 volcan-0.008
Token,
Token ID11
Feature activation-1.55e-02

Pos logit contributions
 narrative+0.061
 museums+0.055
 volcan+0.054
 islands+0.053
 Mercury+0.051
Neg logit contributions
run-0.050
icycle-0.048
ow-0.045
itten-0.045
imm-0.045
Token three
Token ID1115
Feature activation-5.73e-02

Pos logit contributions
 narrative+0.037
 conclusions+0.031
ighters+0.030
ooks+0.030
 museums+0.030
Neg logit contributions
 really-0.014
 very-0.014
--0.013
!-0.013
 too-0.013
Token,
Token ID11
Feature activation-1.27e-02

Pos logit contributions
 volcan+0.093
 narrative+0.092
 museums+0.091
 islands+0.088
 they+0.086
Neg logit contributions
icycle-0.076
run-0.073
ooter-0.070
ow-0.069
itten-0.069
Token four
Token ID1440
Feature activation-4.37e-02

Pos logit contributions
 narrative+0.029
ighters+0.024
 conclusions+0.024
ooks+0.023
 museums+0.023
Neg logit contributions
 really-0.013
!-0.012
--0.012
 very-0.012
 too-0.012
Token,
Token ID11
Feature activation-1.40e-02

Pos logit contributions
 narrative+0.072
 volcan+0.068
 museums+0.068
 islands+0.066
 maybe+0.062
Neg logit contributions
run-0.058
icycle-0.057
imm-0.054
ow-0.053
 away-0.053
Token five
Token ID1936
Feature activation-6.45e-02

Pos logit contributions
 narrative+0.030
 museums+0.025
ighters+0.025
ooks+0.024
 islands+0.024
Neg logit contributions
--0.015
 really-0.015
!-0.014
 very-0.014
run-0.014
Token,
Token ID11
Feature activation-1.71e-02

Pos logit contributions
 narrative+0.116
 museums+0.106
 volcan+0.106
 islands+0.104
 maybe+0.098
Neg logit contributions
run-0.079
icycle-0.077
ay-0.072
imm-0.072
ow-0.071
Token six
Token ID2237
Feature activation-5.82e-02

Pos logit contributions
 narrative+0.037
 museums+0.031
ighters+0.031
ooks+0.030
 conclusions+0.030
Neg logit contributions
--0.018
 really-0.017
!-0.017
 very-0.017
run-0.016
Token,
Token ID11
Feature activation-2.49e-02

Pos logit contributions
 narrative+0.104
 museums+0.095
 volcan+0.094
 islands+0.092
 Mercury+0.087
Neg logit contributions
run-0.073
icycle-0.070
imm-0.065
ay-0.065
ow-0.065
Token seven
Token ID3598
Feature activation-3.51e-02

Pos logit contributions
 narrative+0.051
 museums+0.045
ighters+0.043
 volcan+0.043
 islands+0.043
Neg logit contributions
--0.027
run-0.026
 really-0.026
!-0.025
 away-0.025
Token,
Token ID11
Feature activation-1.52e-02

Pos logit contributions
 narrative+0.066
 museums+0.059
 volcan+0.058
 islands+0.057
 Mercury+0.055
Neg logit contributions
run-0.043
icycle-0.040
 away-0.038
--0.038
itten-0.038
Token "
Token ID366
Feature activation-9.89e-03

Pos logit contributions
 narrative+0.013
 museums+0.012
 volcan+0.012
 maybe+0.011
 they+0.011
Neg logit contributions
run-0.013
icycle-0.012
 away-0.012
imm-0.012
--0.012
TokenIt
Token ID1026
Feature activation+3.99e-03

Pos logit contributions
 narrative+0.019
 museums+0.016
 volcan+0.015
 islands+0.015
adays+0.015
Neg logit contributions
run-0.012
 really-0.012
it-0.012
icycle-0.011
--0.011
Token was
Token ID373
Feature activation+2.56e-03

Pos logit contributions
run+0.005
icycle+0.004
itten+0.004
-+0.004
ay+0.004
Neg logit contributions
 narrative-0.008
 museums-0.007
 volcan-0.007
 islands-0.007
 Mercury-0.006
Token hidden
Token ID7104
Feature activation-3.17e-03

Pos logit contributions
icycle+0.004
run+0.004
itten+0.004
atter+0.004
itter+0.004
Neg logit contributions
 narrative-0.004
 volcan-0.004
 maybe-0.004
 islands-0.004
 museums-0.004
Token long
Token ID890
Feature activation-1.34e-02

Pos logit contributions
 narrative+0.006
 museums+0.005
 volcan+0.005
 islands+0.005
 Mercury+0.005
Neg logit contributions
run-0.004
--0.003
icycle-0.003
!-0.003
 away-0.003
Token ago
Token ID2084
Feature activation-6.44e-02

Pos logit contributions
 narrative+0.028
 museums+0.024
 volcan+0.024
 islands+0.023
 Mercury+0.023
Neg logit contributions
run-0.014
--0.013
!-0.013
 away-0.013
icycle-0.012
Token,
Token ID11
Feature activation+3.51e-03

Pos logit contributions
 narrative+0.113
 volcan+0.105
 museums+0.104
adies+0.101
 islands+0.101
Neg logit contributions
run-0.080
icycle-0.073
imm-0.073
itten-0.071
chie-0.071
Token and
Token ID290
Feature activation-5.38e-03

Pos logit contributions
run+0.004
!+0.004
-+0.004
 really+0.004
 away+0.003
Neg logit contributions
 narrative-0.007
 museums-0.006
 islands-0.006
 volcan-0.006
lan-0.006
Token only
Token ID691
Feature activation+8.08e-03

Pos logit contributions
 narrative+0.009
 museums+0.008
 volcan+0.007
 islands+0.007
adies+0.007
Neg logit contributions
run-0.008
!-0.007
 really-0.007
--0.007
 away-0.007
Token seems
Token ID2331
Feature activation+7.02e-03

Pos logit contributions
run+0.010
-+0.009
!+0.009
 away+0.009
 really+0.009
Neg logit contributions
 narrative-0.015
 museums-0.013
 volcan-0.013
 islands-0.013
 Mercury-0.012
Token fancy
Token ID14996
Feature activation-1.63e-03

Pos logit contributions
icycle+0.010
itten+0.009
imm+0.009
ay+0.009
run+0.009
Neg logit contributions
 narrative-0.012
 volcan-0.012
 islands-0.011
 museums-0.011
 maybe-0.011
Token light
Token ID1657
Feature activation-9.07e-03

Pos logit contributions
run+0.020
icycle+0.018
itten+0.018
ow+0.017
 away+0.017
Neg logit contributions
 narrative-0.030
 museums-0.027
 volcan-0.027
 islands-0.026
lan-0.025
Token,
Token ID11
Feature activation-1.47e-03

Pos logit contributions
 narrative+0.014
 volcan+0.013
 they+0.013
 museums+0.013
lan+0.013
Neg logit contributions
run-0.014
icycle-0.013
imm-0.012
itten-0.012
ippy-0.012
Token Lily
Token ID20037
Feature activation-1.54e-02

Pos logit contributions
 narrative+0.002
 museums+0.002
 islands+0.002
 volcan+0.002
lan+0.002
Neg logit contributions
run-0.002
!-0.002
--0.002
 really-0.002
ay-0.002
Token.
Token ID13
Feature activation-6.98e-03

Pos logit contributions
 narrative+0.027
 museums+0.024
lan+0.024
 volcan+0.024
 they+0.023
Neg logit contributions
run-0.021
icycle-0.018
chie-0.018
imm-0.017
ow-0.017
Token Only
Token ID5514
Feature activation-1.97e-02

Pos logit contributions
 narrative+0.014
 islands+0.013
 museums+0.012
 volcan+0.012
 farms+0.011
Neg logit contributions
ay-0.008
run-0.008
ow-0.007
!-0.007
eting-0.007
Token twenty
Token ID8208
Feature activation-6.42e-02

Pos logit contributions
 narrative+0.034
 museums+0.029
 volcan+0.028
 islands+0.028
 Mercury+0.027
Neg logit contributions
run-0.028
--0.026
!-0.025
icycle-0.025
 away-0.025
Token pounds
Token ID8059
Feature activation-2.90e-02

Pos logit contributions
 narrative+0.114
 volcan+0.103
 museums+0.103
 islands+0.101
lan+0.096
Neg logit contributions
run-0.085
icycle-0.079
ow-0.076
imm-0.075
itten-0.075
Token,"
Token ID553
Feature activation+5.84e-04

Pos logit contributions
 narrative+0.044
 they+0.041
 maybe+0.039
 museums+0.039
 volcan+0.038
Neg logit contributions
run-0.046
imm-0.045
icycle-0.044
itter-0.042
eting-0.041
Token Mom
Token ID11254
Feature activation+2.00e-02

Pos logit contributions
run+0.001
-+0.001
!+0.001
 really+0.001
 away+0.001
Neg logit contributions
 narrative-0.001
 islands-0.001
 museums-0.001
 volcan-0.001
adies-0.001
Token said
Token ID531
Feature activation+3.97e-03

Pos logit contributions
chie+0.030
run+0.029
icycle+0.029
 leash+0.029
ay+0.028
Neg logit contributions
 they-0.027
 narrative-0.027
 maybe-0.026
lan-0.025
 God-0.025
Token.
Token ID13
Feature activation-1.45e-02

Pos logit contributions
icycle+0.004
run+0.004
ay+0.003
itten+0.003
imm+0.003
Neg logit contributions
 narrative-0.008
 maybe-0.007
 volcan-0.007
 they-0.007
 museums-0.007
Token Thank
Token ID6952
Feature activation-2.41e-04

Pos logit contributions
 narrative+0.001
 islands+0.001
 museums+0.001
 crabs+0.001
 bridges+0.001
Neg logit contributions
run-0.000
 too-0.000
 really-0.000
ay-0.000
ow-0.000
Token you
Token ID345
Feature activation-2.65e-02

Pos logit contributions
 narrative+0.000
 museums+0.000
 volcan+0.000
 islands+0.000
 Mercury+0.000
Neg logit contributions
--0.000
!-0.000
ow-0.000
run-0.000
 away-0.000
Token,
Token ID11
Feature activation-6.33e-03

Pos logit contributions
 narrative+0.056
 museums+0.050
 volcan+0.049
 islands+0.048
 Mercury+0.048
Neg logit contributions
!-0.026
run-0.026
--0.026
 away-0.024
 really-0.024
Token God
Token ID1793
Feature activation-3.83e-02

Pos logit contributions
 narrative+0.010
 museums+0.009
 islands+0.008
 volcan+0.008
 Mercury+0.008
Neg logit contributions
run-0.010
!-0.009
--0.009
 really-0.009
ow-0.009
Token.
Token ID13
Feature activation+3.34e-04

Pos logit contributions
 narrative+0.060
lan+0.053
 they+0.052
 God+0.052
 islands+0.051
Neg logit contributions
run-0.059
icycle-0.059
imm-0.056
itten-0.053
chie-0.052
Token Amen
Token ID34717
Feature activation-6.41e-02

Pos logit contributions
run+0.000
ay+0.000
ow+0.000
 too+0.000
 really+0.000
Neg logit contributions
 narrative-0.001
 islands-0.001
 museums-0.001
 bridges-0.001
 volcan-0.001
Token."
Token ID526
Feature activation-5.26e-03

Pos logit contributions
 narrative+0.129
 museums+0.116
 islands+0.113
 volcan+0.113
 Mercury+0.111
Neg logit contributions
run-0.070
--0.067
ow-0.065
!-0.064
ay-0.062
Token
Token ID198
Feature activation-6.87e-03

Pos logit contributions
 narrative+0.012
 islands+0.010
 museums+0.010
adies+0.010
ooks+0.010
Neg logit contributions
!-0.004
 really-0.004
 very-0.004
 too-0.004
ow-0.004
Token
Token ID198
Feature activation-2.20e-03

Pos logit contributions
 narrative+0.013
 museums+0.012
 volcan+0.011
 islands+0.011
 Mercury+0.011
Neg logit contributions
run-0.008
--0.008
!-0.008
icycle-0.007
 really-0.007
TokenHe
Token ID1544
Feature activation+1.86e-03

Pos logit contributions
 narrative+0.003
 museums+0.003
 volcan+0.003
 islands+0.003
 Mercury+0.003
Neg logit contributions
run-0.004
--0.003
icycle-0.003
!-0.003
 really-0.003
Token opened
Token ID4721
Feature activation-4.22e-03

Pos logit contributions
run+0.002
-+0.002
icycle+0.002
!+0.002
 away+0.002
Neg logit contributions
 narrative-0.003
 museums-0.003
 volcan-0.003
 islands-0.003
 Mercury-0.003
Token,
Token ID11
Feature activation-2.86e-02

Pos logit contributions
 narrative+0.042
 museums+0.035
 volcan+0.034
 islands+0.034
 maybe+0.033
Neg logit contributions
run-0.035
icycle-0.033
itten-0.032
imm-0.031
ooter-0.031
Token forty
Token ID16571
Feature activation-6.38e-02

Pos logit contributions
 narrative+0.047
 museums+0.040
 islands+0.039
ighters+0.039
 volcan+0.039
Neg logit contributions
run-0.042
!-0.040
--0.040
 really-0.040
 too-0.039
Token-
Token ID12
Feature activation-1.89e-02

Pos logit contributions
 narrative+0.162
 volcan+0.147
lan+0.146
 museums+0.145
 islands+0.143
Neg logit contributions
run-0.033
icycle-0.029
eting-0.024
 away-0.022
ow-0.021
Tokenfive
Token ID13261
Feature activation-5.25e-02

Pos logit contributions
 narrative+0.026
 museums+0.022
 volcan+0.022
 islands+0.021
lan+0.020
Neg logit contributions
run-0.032
icycle-0.030
--0.030
!-0.029
 away-0.029
Token,
Token ID11
Feature activation-2.29e-02

Pos logit contributions
 narrative+0.080
 islands+0.071
 museums+0.071
 volcan+0.070
 they+0.066
Neg logit contributions
icycle-0.080
run-0.079
itten-0.078
ooter-0.074
imm-0.073
Token forty
Token ID16571
Feature activation-6.40e-02

Pos logit contributions
 narrative+0.038
 museums+0.032
 islands+0.031
 volcan+0.031
ighters+0.031
Neg logit contributions
run-0.033
!-0.032
--0.032
 really-0.031
 too-0.031
Token-
Token ID12
Feature activation-2.21e-02

Pos logit contributions
 narrative+0.160
 volcan+0.146
 museums+0.145
lan+0.144
 islands+0.143
Neg logit contributions
run-0.035
icycle-0.030
eting-0.026
 away-0.024
ow-0.023
Tokensix
Token ID19412
Feature activation-3.76e-02

Pos logit contributions
 narrative+0.030
 museums+0.025
 volcan+0.024
 islands+0.024
lan+0.023
Neg logit contributions
run-0.039
icycle-0.036
--0.036
!-0.036
 away-0.035
Token,
Token ID11
Feature activation-1.87e-02

Pos logit contributions
 narrative+0.060
 museums+0.052
 volcan+0.052
 islands+0.051
 maybe+0.050
Neg logit contributions
run-0.058
icycle-0.053
aker-0.052
imm-0.052
itten-0.051
Token forty
Token ID16571
Feature activation-5.51e-02

Pos logit contributions
 narrative+0.028
 museums+0.024
 volcan+0.023
 islands+0.023
 Mercury+0.022
Neg logit contributions
run-0.030
--0.028
!-0.028
icycle-0.028
 really-0.028
Token-
Token ID12
Feature activation-1.25e-02

Pos logit contributions
 narrative+0.128
 volcan+0.119
 museums+0.117
lan+0.117
 islands+0.115
Neg logit contributions
run-0.038
icycle-0.032
eting-0.030
ainted-0.027
 away-0.027
Token,
Token ID11
Feature activation-2.33e-02

Pos logit contributions
 narrative+0.066
 museums+0.057
 islands+0.054
 volcan+0.054
 maybe+0.053
Neg logit contributions
run-0.070
icycle-0.066
eting-0.064
itter-0.064
 package-0.064
Token forty
Token ID16571
Feature activation-5.66e-02

Pos logit contributions
 narrative+0.034
 museums+0.029
 volcan+0.028
 islands+0.028
 Mercury+0.027
Neg logit contributions
run-0.038
--0.036
icycle-0.035
!-0.035
 away-0.035
Token-
Token ID12
Feature activation-1.68e-02

Pos logit contributions
 narrative+0.131
 volcan+0.121
 museums+0.120
lan+0.118
 islands+0.118
Neg logit contributions
run-0.040
icycle-0.034
eting-0.033
ow-0.030
ainted-0.029
Tokenfour
Token ID14337
Feature activation-2.42e-02

Pos logit contributions
 narrative+0.020
 museums+0.016
 volcan+0.016
 islands+0.015
lan+0.015
Neg logit contributions
run-0.032
icycle-0.030
--0.029
ow-0.029
 away-0.029
Token,
Token ID11
Feature activation-2.86e-02

Pos logit contributions
 narrative+0.042
 museums+0.035
 volcan+0.034
 islands+0.034
 maybe+0.033
Neg logit contributions
run-0.035
icycle-0.033
itten-0.032
imm-0.031
ooter-0.031
Token forty
Token ID16571
Feature activation-6.38e-02

Pos logit contributions
 narrative+0.047
 museums+0.040
 islands+0.039
ighters+0.039
 volcan+0.039
Neg logit contributions
run-0.042
!-0.040
--0.040
 really-0.040
 too-0.039
Token-
Token ID12
Feature activation-1.89e-02

Pos logit contributions
 narrative+0.162
 volcan+0.147
lan+0.146
 museums+0.145
 islands+0.143
Neg logit contributions
run-0.033
icycle-0.029
eting-0.024
 away-0.022
ow-0.021
Tokenfive
Token ID13261
Feature activation-5.25e-02

Pos logit contributions
 narrative+0.026
 museums+0.022
 volcan+0.022
 islands+0.021
lan+0.020
Neg logit contributions
run-0.032
icycle-0.030
--0.030
!-0.029
 away-0.029
Token,
Token ID11
Feature activation-2.29e-02

Pos logit contributions
 narrative+0.080
 islands+0.071
 museums+0.071
 volcan+0.070
 they+0.066
Neg logit contributions
icycle-0.080
run-0.079
itten-0.078
ooter-0.074
imm-0.073
Token forty
Token ID16571
Feature activation-6.40e-02

Pos logit contributions
 narrative+0.038
 museums+0.032
 islands+0.031
 volcan+0.031
ighters+0.031
Neg logit contributions
run-0.033
!-0.032
--0.032
 really-0.031
 too-0.031
Token-
Token ID12
Feature activation-2.21e-02

Pos logit contributions
 narrative+0.160
 volcan+0.146
 museums+0.145
lan+0.144
 islands+0.143
Neg logit contributions
run-0.035
icycle-0.030
eting-0.026
 away-0.024
ow-0.023
Token her
Token ID607
Feature activation-3.17e-03

Pos logit contributions
 narrative+0.012
 museums+0.012
 volcan+0.011
 God+0.011
 islands+0.011
Neg logit contributions
icycle-0.018
run-0.017
itten-0.017
imm-0.017
atter-0.016
Token birthday
Token ID10955
Feature activation-4.84e-03

Pos logit contributions
 narrative+0.005
 maybe+0.005
 they+0.005
lan+0.004
 museums+0.004
Neg logit contributions
imm-0.005
acked-0.004
aker-0.004
icycle-0.004
umb-0.004
Token a
Token ID257
Feature activation+7.28e-03

Pos logit contributions
 they+0.007
lan+0.007
 narrative+0.007
 museums+0.006
 someday+0.006
Neg logit contributions
chie-0.007
run-0.007
ivered-0.007
imm-0.007
icycle-0.007
Token few
Token ID1178
Feature activation-1.97e-02

Pos logit contributions
imm+0.012
ow+0.012
run+0.012
atter+0.011
ivered+0.011
Neg logit contributions
 narrative-0.011
 they-0.010
lan-0.010
 museums-0.009
 volcan-0.009
Token days
Token ID1528
Feature activation-1.70e-02

Pos logit contributions
 narrative+0.018
 museums+0.013
 islands+0.012
 planets+0.012
 volcan+0.011
Neg logit contributions
imm-0.043
ow-0.043
itter-0.043
icycle-0.043
atter-0.042
Token ago
Token ID2084
Feature activation-6.38e-02

Pos logit contributions
 narrative+0.029
 they+0.028
lan+0.027
 museums+0.027
 islands+0.027
Neg logit contributions
run-0.020
imm-0.019
icycle-0.019
chie-0.018
itten-0.017
Token.
Token ID13
Feature activation-8.18e-03

Pos logit contributions
 narrative+0.116
 they+0.107
 volcan+0.105
lan+0.105
 museums+0.104
Neg logit contributions
run-0.073
icycle-0.068
chie-0.066
imm-0.065
ay-0.064
Token She
Token ID1375
Feature activation-7.21e-03

Pos logit contributions
 narrative+0.014
 museums+0.012
 volcan+0.011
 islands+0.011
 Mercury+0.011
Neg logit contributions
run-0.012
icycle-0.011
--0.011
!-0.011
ay-0.011
Token liked
Token ID8288
Feature activation-1.20e-02

Pos logit contributions
 narrative+0.010
 they+0.009
 maybe+0.009
 museums+0.009
 islands+0.009
Neg logit contributions
run-0.011
chie-0.011
itten-0.011
ow-0.011
eting-0.010
Token to
Token ID284
Feature activation-6.01e-03

Pos logit contributions
 narrative+0.016
 museums+0.016
 they+0.015
 volcan+0.015
 islands+0.014
Neg logit contributions
imm-0.019
run-0.019
itten-0.019
aker-0.019
icycle-0.019
Token practice
Token ID3357
Feature activation-1.39e-02

Pos logit contributions
 narrative+0.008
 museums+0.007
 maybe+0.006
 islands+0.006
 volcan+0.006
Neg logit contributions
run-0.011
itten-0.010
icycle-0.010
itter-0.010
imm-0.010
Token,
Token ID11
Feature activation-3.31e-02

Pos logit contributions
 narrative+0.126
 museums+0.111
 volcan+0.109
 islands+0.107
 Mercury+0.102
Neg logit contributions
run-0.087
icycle-0.078
ow-0.075
itten-0.075
 away-0.074
Token eighteen
Token ID29095
Feature activation-5.66e-02

Pos logit contributions
 narrative+0.045
 museums+0.038
 islands+0.035
 Mercury+0.034
 volcan+0.034
Neg logit contributions
run-0.057
--0.056
!-0.056
 really-0.055
 away-0.054
Token,
Token ID11
Feature activation-2.99e-02

Pos logit contributions
 narrative+0.106
 museums+0.095
 volcan+0.094
 islands+0.092
 Mercury+0.088
Neg logit contributions
run-0.071
icycle-0.064
ow-0.062
itten-0.061
 away-0.061
Token nineteen
Token ID43063
Feature activation-2.68e-02

Pos logit contributions
 narrative+0.040
 museums+0.033
 islands+0.031
 Mercury+0.030
 volcan+0.030
Neg logit contributions
run-0.053
!-0.051
--0.051
 really-0.050
 away-0.049
Token,
Token ID11
Feature activation-2.50e-02

Pos logit contributions
 narrative+0.050
 museums+0.044
 volcan+0.044
 islands+0.043
 maybe+0.042
Neg logit contributions
run-0.034
icycle-0.032
itten-0.031
ow-0.030
imm-0.029
Token twenty
Token ID8208
Feature activation-6.35e-02

Pos logit contributions
 narrative+0.041
 museums+0.033
 islands+0.032
 Mercury+0.031
adies+0.031
Neg logit contributions
!-0.038
run-0.037
 really-0.036
 too-0.036
--0.036
Token.
Token ID13
Feature activation+7.94e-03

Pos logit contributions
 narrative+0.125
 volcan+0.114
 museums+0.112
 islands+0.108
lan+0.106
Neg logit contributions
run-0.071
icycle-0.067
ow-0.064
itten-0.063
 away-0.062
Token They
Token ID1119
Feature activation+2.93e-03

Pos logit contributions
run+0.011
ay+0.010
!+0.010
 really+0.010
icycle+0.010
Neg logit contributions
 narrative-0.014
 islands-0.012
 museums-0.012
 volcan-0.011
adies-0.011
Token are
Token ID389
Feature activation+6.04e-03

Pos logit contributions
run+0.005
icycle+0.005
-+0.005
itten+0.005
imm+0.005
Neg logit contributions
 narrative-0.004
 museums-0.003
 volcan-0.003
 islands-0.003
 Mercury-0.003
Token happy
Token ID3772
Feature activation-5.82e-03

Pos logit contributions
run+0.007
icycle+0.007
itten+0.007
imm+0.006
aker+0.006
Neg logit contributions
 narrative-0.012
 volcan-0.011
 museums-0.011
 maybe-0.010
 they-0.010
Token.
Token ID13
Feature activation-3.29e-03

Pos logit contributions
 narrative+0.014
 museums+0.012
 volcan+0.012
 islands+0.012
 Mercury+0.012
Neg logit contributions
!-0.004
run-0.004
--0.004
 really-0.004
 away-0.004
TokenThe
Token ID464
Feature activation-3.03e-03

Pos logit contributions
run+0.011
icycle+0.010
-+0.010
!+0.010
ow+0.010
Neg logit contributions
 narrative-0.010
 museums-0.008
 volcan-0.008
 islands-0.008
 Mercury-0.007
Token little
Token ID1310
Feature activation-1.45e-02

Pos logit contributions
 narrative+0.005
 museums+0.005
 volcan+0.004
 islands+0.004
 Mercury+0.004
Neg logit contributions
run-0.004
imm-0.004
ow-0.004
icycle-0.004
itten-0.004
Token girl
Token ID2576
Feature activation+3.52e-03

Pos logit contributions
 narrative+0.026
 museums+0.022
 volcan+0.022
 islands+0.021
 Mercury+0.021
Neg logit contributions
run-0.020
icycle-0.018
--0.018
ow-0.018
!-0.018
Token's
Token ID338
Feature activation+5.85e-03

Pos logit contributions
run+0.005
ay+0.005
chie+0.005
icycle+0.005
itten+0.005
Neg logit contributions
 narrative-0.005
 they-0.005
 museums-0.004
 islands-0.004
 maybe-0.004
Token wish
Token ID4601
Feature activation-2.21e-02

Pos logit contributions
imm+0.011
aker+0.010
run+0.010
chie+0.010
itten+0.010
Neg logit contributions
 narrative-0.007
 museums-0.006
 maybe-0.006
 Mercury-0.006
 God-0.006
Token was
Token ID373
Feature activation+3.06e-02

Pos logit contributions
 they+0.030
 narrative+0.030
 maybe+0.027
 museums+0.026
 moons+0.025
Neg logit contributions
icycle-0.035
run-0.035
ippy-0.034
chie-0.034
aker-0.033
Token to
Token ID284
Feature activation+5.79e-03

Pos logit contributions
icycle+0.054
itten+0.054
ippy+0.052
run+0.051
aker+0.050
Neg logit contributions
 narrative-0.040
 they-0.039
 museums-0.037
 maybe-0.037
 God-0.034
Token play
Token ID711
Feature activation+2.36e-02

Pos logit contributions
run+0.010
icycle+0.010
itten+0.010
aker+0.010
imm+0.010
Neg logit contributions
 narrative-0.008
 museums-0.007
 maybe-0.006
 God-0.006
 islands-0.006
Token like
Token ID588
Feature activation+2.25e-02

Pos logit contributions
itten+0.040
run+0.039
icycle+0.038
chie+0.036
imm+0.036
Neg logit contributions
 narrative-0.033
 maybe-0.030
 museums-0.030
 they-0.028
 knights-0.027
Token the
Token ID262
Feature activation+3.12e-04

Pos logit contributions
run+0.035
icycle+0.034
imm+0.033
itten+0.033
ay+0.033
Neg logit contributions
 narrative-0.033
 museums-0.030
 volcan-0.029
 islands-0.028
 Mercury-0.028
Token best
Token ID1266
Feature activation+2.03e-02

Pos logit contributions
run+0.001
ow+0.000
imm+0.000
itten+0.000
aker+0.000
Neg logit contributions
 narrative-0.001
 museums-0.000
 Mercury-0.000
 planets-0.000
 islands-0.000
Token.
Token ID13
Feature activation-1.60e-04

Pos logit contributions
 they+0.044
lan+0.042
 islands+0.041
 narrative+0.041
 farms+0.040
Neg logit contributions
run-0.030
icycle-0.027
itten-0.027
 leash-0.027
ippy-0.027
Token Not
Token ID1892
Feature activation-1.41e-02

Pos logit contributions
 narrative+0.000
 museums+0.000
 volcan+0.000
 islands+0.000
 Mercury+0.000
Neg logit contributions
run-0.000
icycle-0.000
ay-0.000
!-0.000
--0.000
Token wanting
Token ID10291
Feature activation-8.84e-04

Pos logit contributions
 narrative+0.021
 museums+0.019
 volcan+0.019
 islands+0.018
 maybe+0.018
Neg logit contributions
icycle-0.022
run-0.021
ay-0.020
itten-0.020
imm-0.020
Token her
Token ID607
Feature activation+1.22e-02

Pos logit contributions
 narrative+0.002
 museums+0.001
 volcan+0.001
 islands+0.001
 maybe+0.001
Neg logit contributions
icycle-0.001
run-0.001
imm-0.001
itten-0.001
ay-0.001
Token furry
Token ID40857
Feature activation+1.92e-02

Pos logit contributions
imm+0.018
ippy+0.018
aker+0.017
acked+0.017
icycle+0.017
Neg logit contributions
 volcan-0.017
 narrative-0.017
 they-0.017
 museums-0.017
 maybe-0.017
Token friend
Token ID1545
Feature activation+3.04e-02

Pos logit contributions
imm+0.025
run+0.025
ocked+0.025
it+0.024
icycle+0.024
Neg logit contributions
 volcan-0.029
 narrative-0.028
 boats-0.026
 museums-0.026
 islands-0.026
Token to
Token ID284
Feature activation+7.72e-03

Pos logit contributions
icycle+0.052
ippy+0.048
imm+0.046
aker+0.046
chie+0.046
Neg logit contributions
 they-0.042
 maybe-0.040
 sometimes-0.036
 volcan-0.035
 museums-0.034
Token get
Token ID651
Feature activation+2.34e-02

Pos logit contributions
run+0.013
icycle+0.012
imm+0.011
itter+0.011
itten+0.011
Neg logit contributions
 narrative-0.011
 museums-0.010
 volcan-0.010
 islands-0.010
 maybe-0.009
Token hurt
Token ID5938
Feature activation+1.32e-02

Pos logit contributions
icycle+0.036
ay+0.034
imm+0.033
ippy+0.033
itten+0.033
Neg logit contributions
 they-0.036
 narrative-0.032
 volcan-0.032
lan-0.032
 islands-0.031
Token,
Token ID11
Feature activation+1.26e-02

Pos logit contributions
icycle+0.018
run+0.017
itten+0.016
imm+0.016
ooter+0.015
Neg logit contributions
 narrative-0.023
 museums-0.020
 volcan-0.020
 they-0.020
 maybe-0.020
Token she
Token ID673
Feature activation-6.84e-04

Pos logit contributions
icycle+0.020
run+0.019
imm+0.018
ay+0.018
itten+0.018
Neg logit contributions
 narrative-0.019
 museums-0.017
 volcan-0.016
 maybe-0.016
 they-0.016
Token came
Token ID1625
Feature activation-1.41e-02

Pos logit contributions
run+0.018
icycle+0.017
chie+0.016
itten+0.016
ay+0.016
Neg logit contributions
 narrative-0.021
 museums-0.019
 volcan-0.018
 islands-0.018
 Mercury-0.018
Token and
Token ID290
Feature activation+3.82e-03

Pos logit contributions
 museums+0.021
 islands+0.021
 narrative+0.021
 farms+0.021
 volcan+0.020
Neg logit contributions
icycle-0.019
run-0.019
imm-0.019
itten-0.018
ay-0.018
Token asked
Token ID1965
Feature activation-1.84e-02

Pos logit contributions
run+0.004
 really+0.004
-+0.004
!+0.004
 very+0.003
Neg logit contributions
 narrative-0.008
ooks-0.007
 islands-0.007
 volcan-0.007
 museums-0.007
Token them
Token ID606
Feature activation+1.83e-02

Pos logit contributions
 narrative+0.030
 museums+0.024
 islands+0.024
 volcan+0.024
 Mercury+0.023
Neg logit contributions
run-0.028
--0.027
!-0.027
 away-0.026
 really-0.026
Token what
Token ID644
Feature activation+8.11e-04

Pos logit contributions
run+0.024
-+0.022
!+0.022
ow+0.021
 really+0.021
Neg logit contributions
 narrative-0.034
 museums-0.029
 volcan-0.028
 islands-0.028
 Mercury-0.027
Token they
Token ID484
Feature activation+3.01e-02

Pos logit contributions
run+0.001
!+0.001
-+0.001
 really+0.001
 very+0.001
Neg logit contributions
 narrative-0.001
 islands-0.001
adies-0.001
 museums-0.001
ooks-0.001
Token wanted
Token ID2227
Feature activation-1.01e-02

Pos logit contributions
run+0.050
icycle+0.046
itten+0.045
-+0.044
ay+0.044
Neg logit contributions
 narrative-0.046
 museums-0.040
 volcan-0.038
 islands-0.038
 maybe-0.036
Token.
Token ID13
Feature activation-1.15e-02

Pos logit contributions
 narrative+0.018
 volcan+0.018
 museums+0.017
 they+0.017
 islands+0.017
Neg logit contributions
icycle-0.013
run-0.012
itten-0.011
chie-0.011
imm-0.011
Token
Token ID198
Feature activation+3.74e-04

Pos logit contributions
 narrative+0.017
 museums+0.014
 volcan+0.014
 islands+0.013
 Mercury+0.013
Neg logit contributions
run-0.019
icycle-0.018
--0.018
!-0.018
 away-0.018
Token
Token ID198
Feature activation-1.22e-03

Pos logit contributions
run+0.001
icycle+0.000
-+0.000
ow+0.000
itten+0.000
Neg logit contributions
 narrative-0.001
 museums-0.001
 volcan-0.001
 islands-0.001
 Mercury-0.001
Token"
Token ID1
Feature activation+6.90e-03

Pos logit contributions
 narrative+0.002
 museums+0.002
 volcan+0.002
 islands+0.002
 Mercury+0.002
Neg logit contributions
run-0.001
--0.001
icycle-0.001
!-0.001
 away-0.001
Token
Token ID198
Feature activation+7.44e-03

Pos logit contributions
run+0.011
icycle+0.011
chie+0.010
-+0.010
 away+0.010
Neg logit contributions
 narrative-0.012
 museums-0.010
 volcan-0.010
 islands-0.010
 maybe-0.010
TokenWhen
Token ID2215
Feature activation-1.08e-02

Pos logit contributions
run+0.012
ow+0.012
itten+0.011
imm+0.011
-+0.011
Neg logit contributions
 narrative-0.011
 museums-0.009
 islands-0.009
 volcan-0.009
 Mercury-0.009
Token he
Token ID339
Feature activation+3.00e-02

Pos logit contributions
 narrative+0.019
 volcan+0.017
adies+0.017
ooks+0.016
 museums+0.016
Neg logit contributions
!-0.014
 really-0.014
 away-0.014
 very-0.014
 too-0.014
Token thought
Token ID1807
Feature activation-1.28e-02

Pos logit contributions
run+0.040
icycle+0.039
itten+0.038
itter+0.037
ay+0.037
Neg logit contributions
 narrative-0.048
 museums-0.045
 maybe-0.044
 they-0.044
 volcan-0.043
Token he
Token ID339
Feature activation+3.35e-02

Pos logit contributions
 narrative+0.025
 museums+0.022
 volcan+0.021
 islands+0.021
 Mercury+0.020
Neg logit contributions
run-0.015
--0.014
!-0.014
icycle-0.014
 away-0.013
Token was
Token ID373
Feature activation+3.10e-02

Pos logit contributions
run+0.046
icycle+0.041
itten+0.041
-+0.040
ow+0.040
Neg logit contributions
 narrative-0.058
 museums-0.051
 volcan-0.050
 islands-0.049
 Mercury-0.048
Token almost
Token ID2048
Feature activation+4.09e-02

Pos logit contributions
run+0.045
ay+0.044
itten+0.044
icycle+0.043
imm+0.042
Neg logit contributions
 narrative-0.049
 museums-0.046
 volcan-0.044
 maybe-0.044
adies-0.043
Token done
Token ID1760
Feature activation+2.00e-02

Pos logit contributions
run+0.046
icycle+0.044
itten+0.043
-+0.042
ay+0.041
Neg logit contributions
 narrative-0.080
 museums-0.072
 volcan-0.070
 islands-0.070
adies-0.068
Token,
Token ID11
Feature activation+1.52e-02

Pos logit contributions
run+0.026
-+0.024
!+0.024
icycle+0.024
 away+0.023
Neg logit contributions
 narrative-0.036
 museums-0.031
 volcan-0.031
 islands-0.031
 Mercury-0.030
Token a
Token ID257
Feature activation+3.18e-03

Pos logit contributions
run+0.024
!+0.023
-+0.023
 really+0.023
 away+0.023
Neg logit contributions
 narrative-0.023
 museums-0.019
 volcan-0.019
 islands-0.019
adies-0.018
Token fox
Token ID21831
Feature activation-1.16e-03

Pos logit contributions
imm+0.005
run+0.005
ow+0.005
icycle+0.004
ash+0.004
Neg logit contributions
 narrative-0.005
 museums-0.005
 volcan-0.004
 maybe-0.004
 they-0.004
Token didn
Token ID1422
Feature activation+1.08e-02

Pos logit contributions
run+0.012
ay+0.011
chie+0.011
imm+0.010
icycle+0.010
Neg logit contributions
 narrative-0.008
 they-0.008
 maybe-0.008
 sometimes-0.007
 museums-0.007
Token't
Token ID470
Feature activation-3.85e-03

Pos logit contributions
run+0.014
icycle+0.013
-+0.013
itten+0.013
 away+0.013
Neg logit contributions
 narrative-0.019
 museums-0.017
 volcan-0.016
 islands-0.016
 Mercury-0.016
Token want
Token ID765
Feature activation+7.92e-03

Pos logit contributions
 narrative+0.006
 museums+0.005
 volcan+0.005
 islands+0.005
 maybe+0.004
Neg logit contributions
run-0.006
icycle-0.006
itten-0.006
--0.006
ay-0.006
Token to
Token ID284
Feature activation+6.30e-03

Pos logit contributions
run+0.009
icycle+0.008
itten+0.008
imm+0.008
-+0.008
Neg logit contributions
 narrative-0.015
 museums-0.014
 volcan-0.014
 islands-0.013
 Mercury-0.013
Token be
Token ID307
Feature activation+2.15e-02

Pos logit contributions
run+0.009
!+0.009
-+0.009
 away+0.009
ow+0.009
Neg logit contributions
 narrative-0.010
 museums-0.009
 volcan-0.009
 islands-0.009
 Mercury-0.009
Token friends
Token ID2460
Feature activation+3.15e-02

Pos logit contributions
icycle+0.035
itten+0.034
ay+0.033
imm+0.030
 package+0.029
Neg logit contributions
 narrative-0.031
 they-0.030
 knights-0.030
 museums-0.029
 pirates-0.029
Token with
Token ID351
Feature activation+6.27e-03

Pos logit contributions
chie+0.053
icycle+0.052
imm+0.052
itten+0.052
run+0.051
Neg logit contributions
 narrative-0.044
 they-0.037
 maybe-0.037
 museums-0.037
 sometimes-0.035
Token Tim
Token ID5045
Feature activation+1.11e-05

Pos logit contributions
icycle+0.011
imm+0.011
itten+0.011
run+0.011
chie+0.011
Neg logit contributions
 narrative-0.007
 volcan-0.007
 museums-0.007
 maybe-0.007
 they-0.007
Tokenmy
Token ID1820
Feature activation-1.49e-03

Pos logit contributions
chie+0.000
icycle+0.000
run+0.000
ay+0.000
imm+0.000
Neg logit contributions
 narrative-0.000
 sometimes-0.000
 museums-0.000
 maybe-0.000
 they-0.000
Token anymore
Token ID7471
Feature activation-1.37e-02

Pos logit contributions
 narrative+0.002
 sometimes+0.002
 they+0.002
 maybe+0.002
 museums+0.002
Neg logit contributions
chie-0.002
imm-0.002
icycle-0.002
itten-0.002
run-0.002
Token.
Token ID13
Feature activation-1.08e-03

Pos logit contributions
 narrative+0.026
 they+0.024
 maybe+0.024
 museums+0.024
 volcan+0.023
Neg logit contributions
chie-0.016
icycle-0.016
run-0.015
imm-0.015
ay-0.015
Token delicious
Token ID12625
Feature activation+1.31e-02

Pos logit contributions
icycle+0.008
run+0.007
ooter+0.007
itten+0.007
ow+0.007
Neg logit contributions
 they-0.007
 narrative-0.007
 maybe-0.006
 volcan-0.006
 museums-0.006
Token!
Token ID0
Feature activation+3.83e-03

Pos logit contributions
run+0.015
chie+0.015
icycle+0.015
ooter+0.013
ow+0.013
Neg logit contributions
 narrative-0.025
 they-0.024
 volcan-0.024
 museums-0.023
 islands-0.023
Token
Token ID220
Feature activation-5.05e-03

Pos logit contributions
run+0.006
-+0.006
icycle+0.006
 away+0.006
itten+0.006
Neg logit contributions
 narrative-0.005
 museums-0.005
 volcan-0.005
 islands-0.004
 Mercury-0.004
Token
Token ID198
Feature activation+4.66e-03

Pos logit contributions
 narrative+0.007
 museums+0.006
 volcan+0.006
 maybe+0.006
ooks+0.006
Neg logit contributions
run-0.008
 away-0.008
itten-0.008
chie-0.008
--0.008
Token
Token ID198
Feature activation+7.65e-03

Pos logit contributions
run+0.006
icycle+0.006
-+0.006
!+0.006
ay+0.005
Neg logit contributions
 narrative-0.008
 museums-0.007
 volcan-0.007
 islands-0.007
 Mercury-0.007
TokenHe
Token ID1544
Feature activation-4.62e-03

Pos logit contributions
run+0.013
icycle+0.012
 really+0.012
-+0.012
!+0.012
Neg logit contributions
 narrative-0.011
 museums-0.009
 volcan-0.009
 islands-0.009
 Mercury-0.009
Token enjoyed
Token ID8359
Feature activation-7.47e-03

Pos logit contributions
 narrative+0.007
 islands+0.006
 they+0.006
 museums+0.006
 volcan+0.006
Neg logit contributions
run-0.007
icycle-0.007
ay-0.007
itten-0.007
chie-0.006
Token the
Token ID262
Feature activation-1.68e-02

Pos logit contributions
 narrative+0.013
 museums+0.011
 volcan+0.010
 islands+0.010
 Mercury+0.010
Neg logit contributions
run-0.011
--0.010
!-0.010
 really-0.010
icycle-0.010
Token tea
Token ID8887
Feature activation+8.62e-03

Pos logit contributions
 narrative+0.019
 museums+0.015
 volcan+0.014
 Mercury+0.014
 islands+0.014
Neg logit contributions
run-0.035
icycle-0.032
chie-0.032
ow-0.031
itten-0.031
Token and
Token ID290
Feature activation+1.15e-03

Pos logit contributions
run+0.015
chie+0.014
icycle+0.013
itten+0.013
ow+0.012
Neg logit contributions
 narrative-0.012
 they-0.012
 museums-0.011
 maybe-0.011
 islands-0.011
Token the
Token ID262
Feature activation-5.53e-03

Pos logit contributions
run+0.002
-+0.001
!+0.001
 really+0.001
 away+0.001
Neg logit contributions
 narrative-0.002
 museums-0.002
 volcan-0.002
adies-0.002
 Mercury-0.002
Token Tom
Token ID4186
Feature activation+2.12e-03

Pos logit contributions
 narrative+0.029
 museums+0.025
 volcan+0.024
 maybe+0.024
 islands+0.023
Neg logit contributions
run-0.032
icycle-0.031
--0.030
imm-0.030
ay-0.030
Token asks
Token ID7893
Feature activation-1.02e-02

Pos logit contributions
chie+0.004
run+0.004
ay+0.004
imm+0.003
icycle+0.003
Neg logit contributions
 narrative-0.003
 maybe-0.002
 museums-0.002
 they-0.002
 God-0.002
Token Mia
Token ID32189
Feature activation+2.80e-02

Pos logit contributions
 narrative+0.019
 maybe+0.017
 they+0.017
 volcan+0.016
 museums+0.016
Neg logit contributions
icycle-0.013
imm-0.011
itten-0.011
chie-0.011
ay-0.011
Token.
Token ID13
Feature activation+7.65e-03

Pos logit contributions
icycle+0.040
ay+0.038
chie+0.037
imm+0.037
itten+0.037
Neg logit contributions
 maybe-0.045
 narrative-0.044
 they-0.044
 museums-0.041
 God-0.040
Token
Token ID198
Feature activation+5.17e-03

Pos logit contributions
icycle+0.013
run+0.012
imm+0.012
chie+0.012
itten+0.012
Neg logit contributions
 narrative-0.010
 maybe-0.009
 they-0.009
 volcan-0.009
 museums-0.009
Token
Token ID198
Feature activation+8.59e-03

Pos logit contributions
run+0.007
icycle+0.006
-+0.006
!+0.006
ow+0.006
Neg logit contributions
 narrative-0.009
 museums-0.008
 volcan-0.008
 islands-0.008
 Mercury-0.007
Token"
Token ID1
Feature activation+6.57e-03

Pos logit contributions
run+0.012
icycle+0.011
-+0.011
ow+0.010
!+0.010
Neg logit contributions
 narrative-0.015
 museums-0.013
 volcan-0.013
 islands-0.013
 Mercury-0.012
TokenLet
Token ID5756
Feature activation+4.73e-03

Pos logit contributions
run+0.010
icycle+0.009
-+0.009
!+0.009
ow+0.009
Neg logit contributions
 narrative-0.011
 museums-0.009
 volcan-0.009
 islands-0.009
 Mercury-0.009
Token's
Token ID338
Feature activation+1.58e-02

Pos logit contributions
run+0.008
icycle+0.007
imm+0.007
aker+0.007
chie+0.007
Neg logit contributions
 narrative-0.007
 they-0.007
 museums-0.006
 maybe-0.006
 volcan-0.006
Token ask
Token ID1265
Feature activation-3.47e-03

Pos logit contributions
run+0.030
imm+0.028
aker+0.027
itten+0.027
chie+0.027
Neg logit contributions
 narrative-0.019
 museums-0.017
 maybe-0.016
 they-0.015
 volcan-0.015
Token the
Token ID262
Feature activation+1.33e-02

Pos logit contributions
 narrative+0.004
 museums+0.004
 volcan+0.004
 God+0.004
 maybe+0.004
Neg logit contributions
run-0.006
icycle-0.006
imm-0.006
itten-0.006
ay-0.006
Token,
Token ID11
Feature activation+6.42e-03

Pos logit contributions
run+0.029
icycle+0.028
itten+0.027
imm+0.026
ay+0.026
Neg logit contributions
 narrative-0.034
 museums-0.030
 volcan-0.030
 maybe-0.029
 islands-0.028
Token "
Token ID366
Feature activation+9.59e-03

Pos logit contributions
icycle+0.011
run+0.011
imm+0.011
ay+0.010
ow+0.010
Neg logit contributions
 narrative-0.008
 they-0.008
 volcan-0.008
 maybe-0.008
 museums-0.007
TokenM
Token ID44
Feature activation+8.49e-03

Pos logit contributions
run+0.016
icycle+0.015
-+0.015
!+0.015
ay+0.015
Neg logit contributions
 narrative-0.013
 museums-0.011
 volcan-0.011
 islands-0.011
 Mercury-0.010
Tokenummy
Token ID13513
Feature activation+1.12e-02

Pos logit contributions
run+0.014
icycle+0.013
 away+0.013
-+0.013
ow+0.013
Neg logit contributions
 narrative-0.012
 museums-0.010
 volcan-0.010
 islands-0.010
 Mercury-0.009
Token look
Token ID804
Feature activation-2.43e-02

Pos logit contributions
run+0.015
chie+0.014
icycle+0.014
imm+0.013
itten+0.013
Neg logit contributions
 they-0.018
 museums-0.018
 narrative-0.017
 volcan-0.017
 islands-0.017
Token!
Token ID0
Feature activation+1.00e-02

Pos logit contributions
 volcan+0.042
 maybe+0.042
 museums+0.041
 narrative+0.041
 they+0.041
Neg logit contributions
icycle-0.030
itten-0.028
run-0.028
imm-0.028
ippy-0.027
Token I
Token ID314
Feature activation+1.45e-03

Pos logit contributions
run+0.015
ay+0.013
icycle+0.013
!+0.012
ow+0.012
Neg logit contributions
 narrative-0.018
 museums-0.015
 islands-0.015
 volcan-0.014
 Mercury-0.014
Token found
Token ID1043
Feature activation+3.93e-03

Pos logit contributions
icycle+0.002
run+0.002
itten+0.002
ay+0.002
imm+0.002
Neg logit contributions
 narrative-0.002
 maybe-0.002
 museums-0.002
 volcan-0.002
 they-0.002
Token a
Token ID257
Feature activation+1.01e-02

Pos logit contributions
run+0.007
icycle+0.007
imm+0.007
itter+0.007
itten+0.006
Neg logit contributions
 volcan-0.005
 narrative-0.005
 islands-0.005
 museums-0.005
lan-0.004
Token coin
Token ID10752
Feature activation-5.41e-03

Pos logit contributions
run+0.018
ow+0.018
imm+0.017
ocked+0.015
itter+0.015
Neg logit contributions
 narrative-0.014
 islands-0.013
 volcan-0.013
lan-0.012
 museums-0.012
Token".
Token ID1911
Feature activation+4.39e-03

Pos logit contributions
lan+0.008
 they+0.008
 narrative+0.007
 volcan+0.007
adies+0.007
Neg logit contributions
run-0.008
imm-0.007
chie-0.007
icycle-0.007
ivered-0.007
Token it
Token ID340
Feature activation-3.93e-03

Pos logit contributions
icycle+0.010
run+0.010
itten+0.010
imm+0.010
itter+0.010
Neg logit contributions
 narrative-0.009
 museums-0.008
 volcan-0.008
 maybe-0.008
 islands-0.008
Token.
Token ID13
Feature activation-5.12e-03

Pos logit contributions
 narrative+0.005
 they+0.005
 sometimes+0.005
 maybe+0.005
 museums+0.005
Neg logit contributions
chie-0.006
icycle-0.006
imm-0.006
run-0.006
itten-0.006
Token He
Token ID679
Feature activation-1.76e-03

Pos logit contributions
 narrative+0.011
 museums+0.010
 islands+0.010
 volcan+0.010
adies+0.010
Neg logit contributions
run-0.005
 really-0.004
!-0.004
--0.004
icycle-0.004
Token asks
Token ID7893
Feature activation-1.04e-02

Pos logit contributions
 narrative+0.003
 museums+0.003
 volcan+0.003
lan+0.002
ooks+0.002
Neg logit contributions
run-0.002
 really-0.002
--0.002
 away-0.002
!-0.002
Token his
Token ID465
Feature activation+1.20e-02

Pos logit contributions
 narrative+0.024
 museums+0.020
 volcan+0.020
 islands+0.020
 Mercury+0.020
Neg logit contributions
--0.010
run-0.009
!-0.009
 away-0.008
 really-0.008
Token mom
Token ID1995
Feature activation+4.31e-03

Pos logit contributions
run+0.020
imm+0.019
icycle+0.019
itten+0.019
chie+0.019
Neg logit contributions
 narrative-0.017
 museums-0.015
 volcan-0.014
 islands-0.014
lan-0.014
Token and
Token ID290
Feature activation+1.86e-02

Pos logit contributions
run+0.005
icycle+0.005
-+0.005
itten+0.005
ay+0.005
Neg logit contributions
 narrative-0.008
 museums-0.007
 volcan-0.007
 islands-0.007
 Mercury-0.007
Token dad
Token ID9955
Feature activation-3.06e-03

Pos logit contributions
run+0.028
icycle+0.026
-+0.026
ay+0.025
itten+0.025
Neg logit contributions
 narrative-0.030
 museums-0.026
 volcan-0.025
 islands-0.025
 Mercury-0.024
Token,
Token ID11
Feature activation+1.22e-02

Pos logit contributions
 narrative+0.006
 museums+0.005
 volcan+0.005
 islands+0.005
 Mercury+0.005
Neg logit contributions
run-0.004
--0.004
!-0.004
 away-0.003
ow-0.003
Token "
Token ID366
Feature activation+5.64e-03

Pos logit contributions
icycle+0.022
imm+0.022
ay+0.022
chie+0.021
acked+0.021
Neg logit contributions
 maybe-0.015
 they-0.015
 narrative-0.015
 God-0.013
 museums-0.013
TokenWhere
Token ID8496
Feature activation-6.61e-03

Pos logit contributions
 away+0.008
run+0.008
-+0.008
icycle+0.008
it+0.008
Neg logit contributions
 narrative-0.009
 museums-0.008
 volcan-0.008
ooks-0.008
 Mercury-0.008
Token was
Token ID373
Feature activation-4.69e-03

Pos logit contributions
run+0.020
itten+0.018
icycle+0.018
imm+0.018
ay+0.018
Neg logit contributions
 narrative-0.021
 museums-0.018
 islands-0.018
 volcan-0.017
 planets-0.017
Token so
Token ID523
Feature activation+9.56e-03

Pos logit contributions
 maybe+0.007
 narrative+0.006
 they+0.006
 God+0.006
 sometimes+0.006
Neg logit contributions
icycle-0.007
ay-0.007
itten-0.007
run-0.007
ippy-0.007
Token happy
Token ID3772
Feature activation+4.29e-03

Pos logit contributions
run+0.013
icycle+0.012
-+0.011
itten+0.011
!+0.011
Neg logit contributions
 narrative-0.017
 museums-0.015
 volcan-0.015
 islands-0.014
 Mercury-0.014
Token with
Token ID351
Feature activation+2.67e-03

Pos logit contributions
run+0.006
icycle+0.005
chie+0.005
itten+0.005
imm+0.005
Neg logit contributions
 narrative-0.007
 museums-0.007
 they-0.007
 volcan-0.007
 maybe-0.007
Token her
Token ID607
Feature activation+2.09e-03

Pos logit contributions
run+0.004
icycle+0.004
ay+0.004
itten+0.004
imm+0.004
Neg logit contributions
 narrative-0.004
 museums-0.004
 volcan-0.003
 maybe-0.003
 islands-0.003
Token computer
Token ID3644
Feature activation-6.89e-04

Pos logit contributions
imm+0.003
bo+0.003
run+0.003
itten+0.003
ippy+0.003
Neg logit contributions
 narrative-0.003
 museums-0.003
 they-0.003
 maybe-0.003
 volcan-0.003
Token.
Token ID13
Feature activation+2.33e-03

Pos logit contributions
 narrative+0.001
 they+0.001
 maybe+0.001
 sometimes+0.001
 museums+0.001
Neg logit contributions
itten-0.001
chie-0.001
aker-0.001
run-0.001
ippy-0.001
Token It
Token ID632
Feature activation-9.15e-03

Pos logit contributions
run+0.003
-+0.003
icycle+0.003
 away+0.003
itten+0.003
Neg logit contributions
 narrative-0.004
 museums-0.003
 volcan-0.003
 islands-0.003
 Mercury-0.003
Token was
Token ID373
Feature activation-4.88e-03

Pos logit contributions
 narrative+0.014
 they+0.012
 museums+0.012
 maybe+0.012
 planets+0.012
Neg logit contributions
run-0.015
itten-0.014
 leash-0.014
icycle-0.014
aker-0.014
Token her
Token ID607
Feature activation-1.95e-03

Pos logit contributions
 narrative+0.005
 maybe+0.005
 they+0.005
 museums+0.005
 God+0.005
Neg logit contributions
icycle-0.009
atter-0.009
itten-0.009
run-0.009
ooter-0.009
Token favorite
Token ID4004
Feature activation+1.66e-02

Pos logit contributions
 narrative+0.003
 maybe+0.003
 museums+0.002
 they+0.002
 volcan+0.002
Neg logit contributions
run-0.003
aker-0.003
bo-0.003
acked-0.003
imm-0.003
Token the
Token ID262
Feature activation+1.01e-02

Pos logit contributions
run+0.027
icycle+0.026
itten+0.025
-+0.025
ow+0.025
Neg logit contributions
 narrative-0.024
 museums-0.021
 volcan-0.021
 islands-0.020
 Mercury-0.019
Token caul
Token ID46087
Feature activation-2.14e-02

Pos logit contributions
run+0.020
icycle+0.018
ow+0.018
itter+0.017
ooter+0.017
Neg logit contributions
 narrative-0.013
 volcan-0.011
 museums-0.011
 islands-0.010
lan-0.010
Tokeniflower
Token ID42642
Feature activation-2.87e-03

Pos logit contributions
 narrative+0.040
 museums+0.035
 volcan+0.035
 islands+0.034
 Mercury+0.033
Neg logit contributions
run-0.027
icycle-0.024
--0.024
ow-0.023
!-0.023
Token with
Token ID351
Feature activation-5.18e-03

Pos logit contributions
 narrative+0.004
 volcan+0.004
 they+0.004
 museums+0.004
 islands+0.004
Neg logit contributions
run-0.005
icycle-0.004
chie-0.004
itten-0.004
ay-0.004
Token some
Token ID617
Feature activation-2.75e-02

Pos logit contributions
 narrative+0.007
 volcan+0.006
 museums+0.006
 islands+0.006
 Mercury+0.006
Neg logit contributions
run-0.009
icycle-0.009
imm-0.008
ooter-0.008
itter-0.008
Token spicy
Token ID26880
Feature activation-1.32e-02

Pos logit contributions
 narrative+0.039
 volcan+0.033
 museums+0.033
 islands+0.033
lan+0.031
Neg logit contributions
run-0.046
icycle-0.045
ow-0.043
imm-0.042
itter-0.042
Token sauce
Token ID10746
Feature activation+5.26e-03

Pos logit contributions
 narrative+0.014
 volcan+0.013
lan+0.012
 islands+0.012
 museums+0.012
Neg logit contributions
run-0.026
icycle-0.026
chie-0.025
ooter-0.024
ow-0.024
Token.
Token ID13
Feature activation-3.09e-03

Pos logit contributions
icycle+0.008
chie+0.008
run+0.008
itter+0.007
ooter+0.007
Neg logit contributions
 they-0.009
 narrative-0.008
lan-0.007
 volcan-0.007
 maybe-0.007
Token The
Token ID383
Feature activation-1.86e-03

Pos logit contributions
 narrative+0.005
 museums+0.004
 volcan+0.004
 maybe+0.004
 Mercury+0.004
Neg logit contributions
run-0.005
icycle-0.005
--0.005
itten-0.004
chie-0.004
Token caul
Token ID46087
Feature activation-2.71e-02

Pos logit contributions
 narrative+0.003
 volcan+0.002
 museums+0.002
 islands+0.002
lan+0.002
Neg logit contributions
run-0.004
icycle-0.003
ow-0.003
imm-0.003
itter-0.003
Tokeniflower
Token ID42642
Feature activation-6.39e-03

Pos logit contributions
 narrative+0.048
 museums+0.041
 volcan+0.041
 islands+0.040
lan+0.039
Neg logit contributions
run-0.038
ow-0.034
icycle-0.034
 away-0.032
itten-0.032
Token bathroom
Token ID12436
Feature activation+2.03e-03

Pos logit contributions
run+0.008
imm+0.008
itten+0.008
ow+0.007
icycle+0.007
Neg logit contributions
 narrative-0.009
 museums-0.008
 volcan-0.008
 islands-0.008
lan-0.007
Token.
Token ID13
Feature activation-7.84e-04

Pos logit contributions
run+0.003
itten+0.003
chie+0.003
icycle+0.003
imm+0.003
Neg logit contributions
 they-0.003
 narrative-0.003
 museums-0.003
 volcan-0.003
 islands-0.003
Token There
Token ID1318
Feature activation-2.10e-03

Pos logit contributions
 narrative+0.002
 islands+0.001
 volcan+0.001
lan+0.001
 museums+0.001
Neg logit contributions
run-0.001
!-0.001
 really-0.001
ow-0.001
ay-0.001
Token,
Token ID11
Feature activation-1.26e-02

Pos logit contributions
 narrative+0.004
 museums+0.003
 islands+0.003
 volcan+0.003
lan+0.003
Neg logit contributions
run-0.003
--0.002
!-0.002
ow-0.002
icycle-0.002
Token on
Token ID319
Feature activation-2.16e-02

Pos logit contributions
 narrative+0.021
 islands+0.017
 museums+0.016
 volcan+0.016
adies+0.016
Neg logit contributions
run-0.019
 away-0.018
 very-0.018
--0.018
 really-0.018
Token the
Token ID262
Feature activation-9.07e-03

Pos logit contributions
 narrative+0.031
 museums+0.026
 volcan+0.026
 islands+0.025
 Mercury+0.024
Neg logit contributions
run-0.037
icycle-0.035
itten-0.034
ow-0.034
--0.034
Token sink
Token ID14595
Feature activation-5.87e-03

Pos logit contributions
 narrative+0.013
 museums+0.011
 volcan+0.011
 islands+0.011
lan+0.010
Neg logit contributions
run-0.016
imm-0.015
ow-0.015
icycle-0.014
itten-0.014
Token,
Token ID11
Feature activation-6.38e-03

Pos logit contributions
 narrative+0.009
 volcan+0.009
 they+0.009
 museums+0.009
 islands+0.009
Neg logit contributions
run-0.008
itten-0.008
icycle-0.008
chie-0.008
ay-0.008
Token was
Token ID373
Feature activation-5.63e-03

Pos logit contributions
 narrative+0.013
 museums+0.011
 islands+0.011
 volcan+0.011
 conclusions+0.011
Neg logit contributions
 away-0.007
run-0.007
--0.006
 too-0.006
 very-0.006
Token a
Token ID257
Feature activation+3.08e-03

Pos logit contributions
 narrative+0.008
 volcan+0.007
 museums+0.007
 islands+0.007
 maybe+0.007
Neg logit contributions
run-0.009
icycle-0.009
chie-0.009
itten-0.009
ay-0.009
Token new
Token ID649
Feature activation+1.81e-02

Pos logit contributions
run+0.005
ow+0.004
imm+0.004
icycle+0.004
chie+0.004
Neg logit contributions
 narrative-0.005
 museums-0.005
 volcan-0.005
 islands-0.005
lan-0.004
Token lived
Token ID5615
Feature activation-8.63e-03

Pos logit contributions
 narrative+0.010
 museums+0.009
 volcan+0.009
 islands+0.009
 Mercury+0.009
Neg logit contributions
run-0.007
icycle-0.007
--0.007
!-0.007
ow-0.007
Token in
Token ID287
Feature activation+3.00e-03

Pos logit contributions
 narrative+0.017
 museums+0.015
 volcan+0.015
 islands+0.015
 Mercury+0.015
Neg logit contributions
run-0.010
icycle-0.009
--0.009
!-0.008
ow-0.008
Token a
Token ID257
Feature activation-1.09e-02

Pos logit contributions
run+0.004
!+0.003
-+0.003
 really+0.003
 away+0.003
Neg logit contributions
 narrative-0.006
 museums-0.005
 volcan-0.005
 islands-0.005
adies-0.005
Token big
Token ID1263
Feature activation+4.44e-03

Pos logit contributions
 narrative+0.020
 museums+0.018
 volcan+0.018
 islands+0.017
 Mercury+0.017
Neg logit contributions
run-0.015
ow-0.013
imm-0.013
icycle-0.013
itten-0.012
Token garden
Token ID11376
Feature activation+1.85e-03

Pos logit contributions
run+0.007
imm+0.006
icycle+0.006
ow+0.006
aker+0.006
Neg logit contributions
 narrative-0.007
 museums-0.006
 volcan-0.006
 islands-0.006
 Mercury-0.006
Token with
Token ID351
Feature activation-1.56e-02

Pos logit contributions
run+0.002
icycle+0.002
itten+0.002
ay+0.002
-+0.002
Neg logit contributions
 narrative-0.003
 museums-0.003
 volcan-0.003
 islands-0.003
 Mercury-0.003
Token many
Token ID867
Feature activation-3.62e-02

Pos logit contributions
 narrative+0.024
 museums+0.021
 volcan+0.020
 islands+0.020
 Mercury+0.019
Neg logit contributions
run-0.024
--0.023
icycle-0.022
!-0.022
ow-0.022
Token other
Token ID584
Feature activation-1.45e-02

Pos logit contributions
 narrative+0.060
 museums+0.051
 islands+0.049
 volcan+0.049
 Mercury+0.048
Neg logit contributions
--0.051
run-0.051
!-0.050
 really-0.049
 away-0.047
Token flowers
Token ID12734
Feature activation-1.47e-02

Pos logit contributions
 narrative+0.018
 museums+0.014
 islands+0.014
 volcan+0.014
 Mercury+0.013
Neg logit contributions
run-0.026
--0.026
!-0.026
 really-0.025
 away-0.025
Token.
Token ID13
Feature activation-6.24e-03

Pos logit contributions
 narrative+0.030
 volcan+0.028
 museums+0.028
 islands+0.028
lan+0.027
Neg logit contributions
run-0.015
icycle-0.014
chie-0.013
ay-0.013
itten-0.012
Token One
Token ID1881
Feature activation+7.81e-03

Pos logit contributions
 narrative+0.013
 islands+0.010
 museums+0.010
lan+0.010
 volcan+0.010
Neg logit contributions
run-0.008
ay-0.007
!-0.007
 really-0.007
ow-0.007
Token.
Token ID13
Feature activation+8.26e-03

Pos logit contributions
 narrative+0.004
 islands+0.003
 they+0.003
 volcan+0.003
 museums+0.003
Neg logit contributions
run-0.003
icycle-0.003
itten-0.003
itter-0.003
ooter-0.003
Token Jim
Token ID5395
Feature activation-7.09e-03

Pos logit contributions
run+0.012
icycle+0.011
-+0.011
!+0.011
ow+0.011
Neg logit contributions
 narrative-0.014
 museums-0.012
 volcan-0.011
 islands-0.011
 Mercury-0.011
Token looked
Token ID3114
Feature activation-1.43e-02

Pos logit contributions
 islands+0.009
 narrative+0.009
 volcan+0.009
 they+0.009
 planets+0.009
Neg logit contributions
ay-0.011
run-0.011
icycle-0.011
itten-0.010
 leash-0.010
Token over
Token ID625
Feature activation-5.62e-03

Pos logit contributions
 narrative+0.026
 islands+0.025
 museums+0.024
 volcan+0.024
 maybe+0.024
Neg logit contributions
run-0.017
itten-0.017
icycle-0.016
aker-0.015
ay-0.015
Token the
Token ID262
Feature activation-6.90e-03

Pos logit contributions
 narrative+0.010
 museums+0.009
 volcan+0.008
 islands+0.008
 Mercury+0.008
Neg logit contributions
run-0.008
icycle-0.007
--0.007
ow-0.007
itten-0.007
Token horizon
Token ID17810
Feature activation-1.47e-02

Pos logit contributions
 narrative+0.010
 museums+0.009
 volcan+0.009
 islands+0.009
 Mercury+0.008
Neg logit contributions
run-0.012
ow-0.011
--0.011
icycle-0.011
imm-0.011
Token and
Token ID290
Feature activation+1.36e-02

Pos logit contributions
 narrative+0.023
 islands+0.021
 museums+0.021
 volcan+0.021
 they+0.021
Neg logit contributions
run-0.022
icycle-0.020
chie-0.020
itter-0.019
imm-0.019
Token could
Token ID714
Feature activation-1.80e-02

Pos logit contributions
run+0.017
-+0.015
icycle+0.015
!+0.015
 away+0.015
Neg logit contributions
 narrative-0.026
 museums-0.023
 volcan-0.022
 islands-0.022
 Mercury-0.021
Token see
Token ID766
Feature activation-1.06e-02

Pos logit contributions
 narrative+0.021
 museums+0.018
 maybe+0.018
 volcan+0.018
 islands+0.017
Neg logit contributions
run-0.034
icycle-0.032
ay-0.032
--0.032
itten-0.032
Token the
Token ID262
Feature activation-1.36e-03

Pos logit contributions
 narrative+0.016
 museums+0.014
 volcan+0.014
 islands+0.014
 maybe+0.013
Neg logit contributions
run-0.017
icycle-0.016
itten-0.016
--0.016
ay-0.016
Token sun
Token ID4252
Feature activation+1.28e-02

Pos logit contributions
 narrative+0.002
 museums+0.002
 volcan+0.002
 islands+0.002
 planets+0.001
Neg logit contributions
run-0.003
ow-0.002
imm-0.002
icycle-0.002
--0.002
Token happy
Token ID3772
Feature activation+1.43e-03

Pos logit contributions
icycle+0.004
itten+0.004
run+0.004
ay+0.004
ippy+0.004
Neg logit contributions
 narrative-0.005
 they-0.005
 islands-0.005
 maybe-0.004
 volcan-0.004
Token.
Token ID13
Feature activation-2.50e-03

Pos logit contributions
run+0.002
icycle+0.002
itten+0.002
ay+0.002
chie+0.002
Neg logit contributions
 narrative-0.003
 they-0.002
 islands-0.002
 museums-0.002
 volcan-0.002
Token Everyone
Token ID11075
Feature activation+2.64e-03

Pos logit contributions
 narrative+0.004
 museums+0.003
 volcan+0.003
 islands+0.003
 Mercury+0.003
Neg logit contributions
run-0.004
icycle-0.004
--0.004
!-0.004
ow-0.004
Token loved
Token ID6151
Feature activation-2.08e-02

Pos logit contributions
itten+0.004
run+0.004
icycle+0.004
 leash+0.004
ay+0.004
Neg logit contributions
 narrative-0.004
 they-0.004
 museums-0.003
 islands-0.003
 maybe-0.003
Token the
Token ID262
Feature activation-4.54e-03

Pos logit contributions
 narrative+0.029
 museums+0.026
 volcan+0.025
 islands+0.024
 Mercury+0.023
Neg logit contributions
run-0.035
icycle-0.033
itten-0.033
ay-0.032
--0.032
Token king
Token ID5822
Feature activation-2.67e-02

Pos logit contributions
 narrative+0.008
 museums+0.007
 volcan+0.006
 islands+0.006
 Mercury+0.006
Neg logit contributions
run-0.007
icycle-0.006
imm-0.006
ow-0.006
itten-0.006
Token because
Token ID780
Feature activation-4.37e-03

Pos logit contributions
 they+0.037
 narrative+0.036
 volcan+0.034
 planets+0.034
 maybe+0.033
Neg logit contributions
icycle-0.043
run-0.041
ippy-0.039
itten-0.039
chie-0.039
Token he
Token ID339
Feature activation-2.95e-03

Pos logit contributions
 narrative+0.006
 museums+0.005
 islands+0.005
 volcan+0.004
adies+0.004
Neg logit contributions
run-0.008
--0.007
!-0.007
 really-0.007
 away-0.007
Token was
Token ID373
Feature activation-1.50e-03

Pos logit contributions
 narrative+0.003
 they+0.003
 maybe+0.003
 museums+0.003
 planets+0.003
Neg logit contributions
run-0.006
itten-0.005
ooter-0.005
itter-0.005
 leash-0.005
Token so
Token ID523
Feature activation+1.12e-02

Pos logit contributions
 they+0.002
 volcan+0.002
 narrative+0.002
lan+0.002
 islands+0.002
Neg logit contributions
itten-0.002
icycle-0.002
run-0.002
ay-0.002
imm-0.002
Token compassionate
Token ID32533
Feature activation+5.48e-03

Pos logit contributions
run+0.016
icycle+0.016
itten+0.016
imm+0.015
 package+0.014
Neg logit contributions
 narrative-0.017
 volcan-0.016
lan-0.016
 museums-0.016
 they-0.016
Token are
Token ID389
Feature activation+6.27e-03

Pos logit contributions
 narrative+0.014
 museums+0.012
 volcan+0.011
 islands+0.011
 Mercury+0.011
Neg logit contributions
run-0.013
icycle-0.012
--0.012
itten-0.012
ay-0.012
Token a
Token ID257
Feature activation+2.92e-03

Pos logit contributions
run+0.009
icycle+0.009
itten+0.009
ay+0.009
imm+0.008
Neg logit contributions
 narrative-0.010
 volcan-0.009
 museums-0.009
 maybe-0.009
 islands-0.009
Token good
Token ID922
Feature activation+2.65e-02

Pos logit contributions
run+0.005
icycle+0.004
imm+0.004
ow+0.004
ay+0.004
Neg logit contributions
 narrative-0.005
 volcan-0.004
 museums-0.004
 islands-0.004
lan-0.004
Token friend
Token ID1545
Feature activation+1.43e-02

Pos logit contributions
run+0.023
icycle+0.020
-+0.018
imm+0.018
ow+0.018
Neg logit contributions
 narrative-0.060
 museums-0.054
 volcan-0.053
 islands-0.052
 Mercury-0.051
Token.
Token ID13
Feature activation+2.08e-03

Pos logit contributions
icycle+0.019
imm+0.019
run+0.018
itten+0.018
chie+0.017
Neg logit contributions
 narrative-0.026
 maybe-0.022
 they-0.022
 sometimes-0.022
 volcan-0.022
Token Can
Token ID1680
Feature activation-5.17e-02

Pos logit contributions
run+0.002
ow+0.002
ay+0.002
!+0.002
 too+0.002
Neg logit contributions
 narrative-0.004
 islands-0.004
 museums-0.004
 volcan-0.004
 farms-0.004
Token you
Token ID345
Feature activation+7.13e-04

Pos logit contributions
 narrative+0.104
 islands+0.102
ooks+0.098
 deserts+0.094
 Mercury+0.093
Neg logit contributions
 too-0.060
 me-0.055
ow-0.053
 really-0.053
!-0.052
Token forgive
Token ID20927
Feature activation-6.18e-03

Pos logit contributions
run+0.001
icycle+0.001
-+0.001
itten+0.001
ay+0.001
Neg logit contributions
 narrative-0.001
 museums-0.001
 volcan-0.001
 islands-0.001
 Mercury-0.001
Token me
Token ID502
Feature activation-7.05e-03

Pos logit contributions
 narrative+0.015
 museums+0.014
 volcan+0.013
 islands+0.013
 Mercury+0.013
Neg logit contributions
run-0.005
icycle-0.004
itten-0.004
chie-0.004
imm-0.003
Token?"
Token ID1701
Feature activation-1.71e-02

Pos logit contributions
 narrative+0.013
 museums+0.012
 volcan+0.012
 islands+0.011
 maybe+0.011
Neg logit contributions
run-0.008
icycle-0.008
itten-0.007
ay-0.007
imm-0.007
Token Tom
Token ID4186
Feature activation+1.03e-03

Pos logit contributions
 narrative+0.038
 islands+0.034
 volcan+0.033
 museums+0.033
adies+0.032
Neg logit contributions
run-0.016
!-0.016
 really-0.015
 too-0.015
--0.014
Token named
Token ID3706
Feature activation+1.92e-03

Pos logit contributions
 narrative+0.000
 museums+0.000
 islands+0.000
 volcan+0.000
 Mercury+0.000
Neg logit contributions
--0.000
run-0.000
!-0.000
ow-0.000
 really-0.000
Token Lily
Token ID20037
Feature activation-7.25e-03

Pos logit contributions
!+0.003
run+0.003
 really+0.003
 very+0.003
ow+0.003
Neg logit contributions
 narrative-0.004
 islands-0.003
 museums-0.003
 volcan-0.003
 whales-0.003
Token.
Token ID13
Feature activation-4.52e-03

Pos logit contributions
 narrative+0.016
 museums+0.015
 volcan+0.015
 islands+0.014
 Mercury+0.014
Neg logit contributions
run-0.006
icycle-0.005
itten-0.005
ay-0.005
--0.005
Token Lily
Token ID20037
Feature activation-1.30e-02

Pos logit contributions
 narrative+0.010
 islands+0.009
 whales+0.009
lan+0.009
 volcan+0.009
Neg logit contributions
run-0.004
 really-0.004
ay-0.004
 too-0.004
!-0.004
Token was
Token ID373
Feature activation-6.38e-03

Pos logit contributions
 narrative+0.028
 museums+0.025
 islands+0.025
 volcan+0.024
 Mercury+0.024
Neg logit contributions
--0.013
 really-0.013
!-0.013
run-0.012
 very-0.012
Token three
Token ID1115
Feature activation-4.83e-02

Pos logit contributions
 narrative+0.017
 museums+0.015
 islands+0.015
 volcan+0.015
lan+0.014
Neg logit contributions
 really-0.003
run-0.003
 very-0.003
ow-0.003
--0.003
Token years
Token ID812
Feature activation-2.18e-03

Pos logit contributions
 narrative+0.098
 museums+0.088
 volcan+0.087
 islands+0.085
 Mercury+0.082
Neg logit contributions
run-0.052
icycle-0.047
 away-0.044
ow-0.044
imm-0.044
Token old
Token ID1468
Feature activation-2.22e-02

Pos logit contributions
 narrative+0.005
 museums+0.004
 islands+0.004
 volcan+0.004
 Mercury+0.004
Neg logit contributions
run-0.002
--0.002
ay-0.002
!-0.002
icycle-0.001
Token and
Token ID290
Feature activation-6.46e-03

Pos logit contributions
 narrative+0.043
 museums+0.038
 volcan+0.037
 islands+0.037
lan+0.036
Neg logit contributions
run-0.027
icycle-0.024
imm-0.023
ow-0.023
itten-0.022
Token lived
Token ID5615
Feature activation-7.81e-03

Pos logit contributions
 narrative+0.011
 museums+0.009
 volcan+0.009
 maybe+0.009
 islands+0.009
Neg logit contributions
run-0.009
icycle-0.009
 away-0.008
ow-0.008
itten-0.008
Token in
Token ID287
Feature activation+5.05e-03

Pos logit contributions
 narrative+0.016
 museums+0.014
 volcan+0.014
 islands+0.014
 Mercury+0.013
Neg logit contributions
run-0.008
icycle-0.007
--0.007
!-0.007
itten-0.007
Token,"
Token ID553
Feature activation+1.00e-02

Pos logit contributions
run+0.007
icycle+0.007
chie+0.006
imm+0.006
ippy+0.006
Neg logit contributions
 narrative-0.005
 museums-0.005
 volcan-0.005
 islands-0.004
 they-0.004
Token she
Token ID673
Feature activation+2.08e-02

Pos logit contributions
run+0.016
icycle+0.014
-+0.014
!+0.014
ow+0.014
Neg logit contributions
 narrative-0.016
 museums-0.013
 volcan-0.013
 islands-0.013
 Mercury-0.012
Token said
Token ID531
Feature activation+1.05e-02

Pos logit contributions
icycle+0.031
run+0.031
ay+0.029
itten+0.028
chie+0.027
Neg logit contributions
 narrative-0.033
 maybe-0.029
 they-0.029
 volcan-0.028
 museums-0.028
Token.
Token ID13
Feature activation-6.23e-03

Pos logit contributions
icycle+0.011
run+0.010
ay+0.010
itten+0.010
imm+0.009
Neg logit contributions
 narrative-0.020
 maybe-0.019
 they-0.019
 volcan-0.019
lan-0.018
Token "
Token ID366
Feature activation-5.09e-03

Pos logit contributions
 narrative+0.009
 museums+0.008
 volcan+0.008
 maybe+0.008
 they+0.008
Neg logit contributions
run-0.009
icycle-0.009
 away-0.009
chie-0.009
imm-0.009
TokenDo
Token ID5211
Feature activation-4.10e-02

Pos logit contributions
 narrative+0.009
 museums+0.008
 volcan+0.008
 islands+0.008
adies+0.007
Neg logit contributions
run-0.007
 really-0.006
it-0.006
icycle-0.006
itten-0.006
Token you
Token ID345
Feature activation+9.35e-03

Pos logit contributions
 narrative+0.080
ooks+0.071
 islands+0.071
 museums+0.069
ighters+0.069
Neg logit contributions
 really-0.045
!-0.044
 too-0.044
ow-0.043
 very-0.043
Token want
Token ID765
Feature activation+4.81e-03

Pos logit contributions
run+0.013
!+0.012
-+0.012
 really+0.012
ow+0.012
Neg logit contributions
 narrative-0.017
 museums-0.014
 volcan-0.014
 islands-0.014
 Mercury-0.014
Token to
Token ID284
Feature activation+9.30e-03

Pos logit contributions
run+0.006
icycle+0.006
imm+0.006
itten+0.005
aker+0.005
Neg logit contributions
 narrative-0.009
 museums-0.008
 volcan-0.008
 islands-0.008
lan-0.008
Token see
Token ID766
Feature activation-4.85e-03

Pos logit contributions
run+0.014
icycle+0.013
-+0.013
itten+0.013
ay+0.013
Neg logit contributions
 narrative-0.015
 museums-0.013
 volcan-0.013
 islands-0.012
 Mercury-0.012
Token?"
Token ID1701
Feature activation-1.80e-02

Pos logit contributions
 narrative+0.006
 volcan+0.006
 museums+0.006
 islands+0.006
 maybe+0.006
Neg logit contributions
run-0.008
icycle-0.008
imm-0.007
chie-0.007
itten-0.007
Token make
Token ID787
Feature activation-4.47e-03

Pos logit contributions
 narrative+0.025
 museums+0.021
 maybe+0.021
 volcan+0.021
 islands+0.020
Neg logit contributions
run-0.031
icycle-0.031
itten-0.029
itter-0.028
imm-0.028
Token me
Token ID502
Feature activation-1.79e-02

Pos logit contributions
 narrative+0.005
 volcan+0.005
 museums+0.005
 maybe+0.005
 islands+0.004
Neg logit contributions
icycle-0.009
run-0.008
itten-0.008
suit-0.008
ooter-0.008
Token feel
Token ID1254
Feature activation-7.38e-03

Pos logit contributions
 narrative+0.021
 maybe+0.020
 museums+0.019
 volcan+0.019
 God+0.019
Neg logit contributions
icycle-0.032
run-0.031
ippy-0.030
itten-0.029
ay-0.029
Token better
Token ID1365
Feature activation-4.95e-03

Pos logit contributions
 narrative+0.010
 maybe+0.009
 volcan+0.008
 museums+0.008
 islands+0.008
Neg logit contributions
icycle-0.014
run-0.013
itten-0.013
ippy-0.013
ow-0.012
Token.
Token ID13
Feature activation-7.80e-03

Pos logit contributions
 narrative+0.009
 museums+0.008
 islands+0.007
 volcan+0.007
 maybe+0.007
Neg logit contributions
run-0.007
icycle-0.007
itten-0.006
imm-0.006
ay-0.006
Token Can
Token ID1680
Feature activation-4.77e-02

Pos logit contributions
 narrative+0.014
 islands+0.012
 museums+0.012
 volcan+0.011
lan+0.011
Neg logit contributions
run-0.012
ay-0.010
ow-0.010
 really-0.010
 too-0.009
Token you
Token ID345
Feature activation-2.55e-03

Pos logit contributions
 narrative+0.107
ooks+0.096
 islands+0.094
 museums+0.091
 nations+0.089
Neg logit contributions
 too-0.048
run-0.045
 me-0.045
 really-0.044
ow-0.043
Token use
Token ID779
Feature activation+3.07e-04

Pos logit contributions
 narrative+0.004
 museums+0.003
 volcan+0.003
 islands+0.003
 maybe+0.003
Neg logit contributions
run-0.004
icycle-0.004
itten-0.004
--0.004
imm-0.004
Token your
Token ID534
Feature activation-1.65e-02

Pos logit contributions
imm+0.001
icycle+0.001
run+0.001
itter+0.001
suit+0.001
Neg logit contributions
 narrative-0.000
 volcan-0.000
 museums-0.000
 maybe-0.000
 God-0.000
Token magical
Token ID10883
Feature activation-1.63e-02

Pos logit contributions
 narrative+0.022
 volcan+0.018
 maybe+0.018
 God+0.017
 museums+0.017
Neg logit contributions
icycle-0.029
run-0.029
imm-0.028
it-0.028
oodle-0.028
Token powers
Token ID5635
Feature activation-1.36e-02

Pos logit contributions
 narrative+0.024
 volcan+0.021
 museums+0.020
 planets+0.020
 God+0.020
Neg logit contributions
run-0.024
it-0.024
icycle-0.024
oodle-0.023
imm-0.023
Token library
Token ID5888
Feature activation-1.63e-02

Pos logit contributions
 narrative+0.016
 museums+0.014
 volcan+0.014
 islands+0.013
 Mercury+0.013
Neg logit contributions
run-0.020
icycle-0.018
itten-0.018
ow-0.018
imm-0.018
Token.
Token ID13
Feature activation-2.75e-03

Pos logit contributions
 narrative+0.032
 museums+0.030
 volcan+0.029
 islands+0.029
 they+0.029
Neg logit contributions
run-0.018
icycle-0.017
itten-0.017
imm-0.016
ay-0.016
Token She
Token ID1375
Feature activation-3.93e-03

Pos logit contributions
 narrative+0.003
 museums+0.003
 maybe+0.003
 volcan+0.003
 Mercury+0.003
Neg logit contributions
run-0.005
icycle-0.005
aker-0.005
imm-0.005
itten-0.005
Token was
Token ID373
Feature activation-1.83e-03

Pos logit contributions
 narrative+0.006
 museums+0.005
 islands+0.005
 volcan+0.005
 maybe+0.005
Neg logit contributions
run-0.006
ay-0.006
imm-0.006
itten-0.006
icycle-0.006
Token only
Token ID691
Feature activation-1.01e-04

Pos logit contributions
 maybe+0.003
 narrative+0.003
 museums+0.003
 volcan+0.003
adies+0.003
Neg logit contributions
ay-0.002
icycle-0.002
imm-0.002
run-0.002
itten-0.002
Token three
Token ID1115
Feature activation-4.89e-02

Pos logit contributions
 maybe+0.000
 museums+0.000
 volcan+0.000
 narrative+0.000
 cam+0.000
Neg logit contributions
icycle-0.000
imm-0.000
ay-0.000
itten-0.000
aker-0.000
Token years
Token ID812
Feature activation-1.04e-03

Pos logit contributions
 narrative+0.078
 museums+0.074
 volcan+0.071
 islands+0.069
 maybe+0.067
Neg logit contributions
icycle-0.067
run-0.066
imm-0.066
itten-0.063
chie-0.063
Token old
Token ID1468
Feature activation-2.74e-02

Pos logit contributions
 narrative+0.002
 museums+0.002
lan+0.002
 volcan+0.002
 islands+0.002
Neg logit contributions
run-0.001
imm-0.001
icycle-0.001
itten-0.001
ow-0.001
Token,
Token ID11
Feature activation+4.57e-03

Pos logit contributions
 narrative+0.036
 museums+0.034
 maybe+0.034
 they+0.033
 islands+0.033
Neg logit contributions
run-0.048
imm-0.045
icycle-0.045
chie-0.044
aker-0.044
Token but
Token ID475
Feature activation+3.35e-04

Pos logit contributions
icycle+0.007
run+0.007
ay+0.007
chie+0.007
imm+0.007
Neg logit contributions
 narrative-0.007
 maybe-0.006
 museums-0.006
 volcan-0.006
 they-0.006
Token she
Token ID673
Feature activation+1.58e-02

Pos logit contributions
run+0.001
 really+0.001
-+0.001
!+0.001
 away+0.001
Neg logit contributions
 narrative-0.000
 museums-0.000
 islands-0.000
 volcan-0.000
lan-0.000