Token | games |
Token ID | 1830 |
Feature activation | +9.75e-02 |
Pos logit contributions | |
oons | +0.045 |
ries | +0.042 |
flour | +0.042 |
cakes | +0.038 |
ingredients | +0.038 |
Neg logit contributions | |
heel | -0.065 |
badly | -0.062 |
boo | -0.062 |
former | -0.062 |
sing | -0.061 |
Token | of |
Token ID | 286 |
Feature activation | +8.29e-02 |
Pos logit contributions | |
flour | +0.150 |
oons | +0.135 |
ingredients | +0.133 |
pans | +0.131 |
tables | +0.130 |
Neg logit contributions | |
badly | -0.177 |
sing | -0.175 |
heel | -0.175 |
recovery | -0.171 |
hero | -0.170 |
Token | all |
Token ID | 477 |
Feature activation | +1.19e-01 |
Pos logit contributions | |
flour | +0.121 |
oons | +0.114 |
ingredients | +0.108 |
ries | +0.108 |
cakes | +0.108 |
Neg logit contributions | |
heel | -0.149 |
badly | -0.146 |
recovery | -0.145 |
boo | -0.145 |
sing | -0.145 |
Token | shapes |
Token ID | 15268 |
Feature activation | +2.18e-01 |
Pos logit contributions | |
flour | +0.198 |
oons | +0.177 |
pans | +0.172 |
tables | +0.172 |
pots | +0.170 |
Neg logit contributions | |
badly | -0.187 |
promise | -0.179 |
roar | -0.179 |
hero | -0.177 |
forever | -0.175 |
Token | and |
Token ID | 290 |
Feature activation | +2.46e-02 |
Pos logit contributions | |
flour | +0.361 |
tables | +0.330 |
pans | +0.320 |
ports | +0.318 |
ingredients | +0.314 |
Neg logit contributions | |
badly | -0.367 |
sing | -0.341 |
hero | -0.336 |
heel | -0.336 |
forever | -0.328 |
Token | sizes |
Token ID | 10620 |
Feature activation | +2.38e-01 |
Pos logit contributions | |
oons | +0.031 |
flour | +0.030 |
ries | +0.029 |
cakes | +0.026 |
ingredients | +0.026 |
Neg logit contributions | |
heel | -0.047 |
former | -0.047 |
roar | -0.046 |
sing | -0.046 |
boo | -0.046 |
Token | ! |
Token ID | 0 |
Feature activation | +2.16e-02 |
Pos logit contributions | |
flour | +0.389 |
pans | +0.355 |
tables | +0.354 |
ingredients | +0.337 |
pots | +0.336 |
Neg logit contributions | |
badly | -0.412 |
heel | -0.384 |
sing | -0.383 |
hero | -0.376 |
recovery | -0.373 |
Token | Jack |
Token ID | 3619 |
Feature activation | +7.43e-02 |
Pos logit contributions | |
flour | +0.037 |
oons | +0.033 |
ingredients | +0.032 |
ries | +0.032 |
pots | +0.032 |
Neg logit contributions | |
sing | -0.039 |
symp | -0.037 |
recovery | -0.037 |
heel | -0.037 |
spine | -0.036 |
Token | couldn |
Token ID | 3521 |
Feature activation | -9.30e-03 |
Pos logit contributions | |
flour | +0.091 |
ingredients | +0.081 |
tables | +0.080 |
oons | +0.079 |
pots | +0.078 |
Neg logit contributions | |
sing | -0.167 |
recovery | -0.163 |
heel | -0.161 |
badly | -0.160 |
boo | -0.159 |
Token | 't |
Token ID | 470 |
Feature activation | +1.07e-02 |
Pos logit contributions | |
sing | +0.017 |
recovery | +0.016 |
heel | +0.016 |
spine | +0.016 |
boo | +0.016 |
Neg logit contributions | |
flour | -0.015 |
oons | -0.014 |
tables | -0.014 |
ingredients | -0.014 |
pots | -0.013 |
Token | believe |
Token ID | 1975 |
Feature activation | +3.16e-04 |
Pos logit contributions | |
tables | +0.019 |
flour | +0.018 |
pots | +0.018 |
brushes | +0.017 |
chairs | +0.017 |
Neg logit contributions | |
symp | -0.015 |
agging | -0.015 |
retri | -0.014 |
OO | -0.014 |
berth | -0.014 |
Token | . |
Token ID | 13 |
Feature activation | -1.51e-02 |
Pos logit contributions | |
flour | +0.126 |
ingredients | +0.111 |
oons | +0.110 |
tables | +0.109 |
pots | +0.105 |
Neg logit contributions | |
sing | -0.170 |
badly | -0.168 |
heel | -0.166 |
recovery | -0.163 |
boo | -0.161 |
Token | There |
Token ID | 1318 |
Feature activation | +5.08e-02 |
Pos logit contributions | |
sing | +0.021 |
recovery | +0.020 |
heel | +0.020 |
badly | +0.020 |
boo | +0.020 |
Neg logit contributions | |
flour | -0.031 |
ingredients | -0.029 |
tables | -0.029 |
oons | -0.029 |
pots | -0.029 |
Token | were |
Token ID | 547 |
Feature activation | +2.33e-02 |
Pos logit contributions | |
flour | +0.061 |
ingredients | +0.053 |
oons | +0.053 |
pans | +0.052 |
tables | +0.052 |
Neg logit contributions | |
sing | -0.112 |
badly | -0.110 |
recovery | -0.109 |
heel | -0.109 |
symp | -0.108 |
Token | paintings |
Token ID | 21641 |
Feature activation | +1.57e-01 |
Pos logit contributions | |
oons | +0.031 |
flour | +0.031 |
ries | +0.030 |
ingredients | +0.029 |
pans | +0.028 |
Neg logit contributions | |
heel | -0.043 |
badly | -0.043 |
sing | -0.043 |
hero | -0.042 |
former | -0.042 |
Token | and |
Token ID | 290 |
Feature activation | +2.47e-02 |
Pos logit contributions | |
flour | +0.267 |
pans | +0.262 |
ports | +0.257 |
oons | +0.254 |
ingredients | +0.249 |
Neg logit contributions | |
badly | -0.225 |
hero | -0.210 |
heel | -0.206 |
agging | -0.206 |
roar | -0.205 |
Token | sculptures |
Token ID | 42645 |
Feature activation | +2.35e-01 |
Pos logit contributions | |
oons | +0.031 |
flour | +0.028 |
ries | +0.027 |
ingredients | +0.026 |
cakes | +0.025 |
Neg logit contributions | |
roar | -0.046 |
former | -0.045 |
wish | -0.044 |
heel | -0.043 |
villains | -0.043 |
Token | and |
Token ID | 290 |
Feature activation | +4.25e-02 |
Pos logit contributions | |
flour | +0.398 |
pans | +0.381 |
ingredients | +0.362 |
ports | +0.362 |
oons | +0.356 |
Neg logit contributions | |
badly | -0.374 |
roar | -0.346 |
hero | -0.343 |
sing | -0.340 |
recovery | -0.337 |
Token | other |
Token ID | 584 |
Feature activation | +3.67e-02 |
Pos logit contributions | |
oons | +0.056 |
flour | +0.052 |
ries | +0.049 |
cakes | +0.047 |
ingredients | +0.047 |
Neg logit contributions | |
roar | -0.081 |
former | -0.079 |
wish | -0.077 |
sing | -0.077 |
heel | -0.077 |
Token | fun |
Token ID | 1257 |
Feature activation | +1.21e-02 |
Pos logit contributions | |
flour | +0.047 |
oons | +0.044 |
ingredients | +0.040 |
cakes | +0.039 |
ries | +0.039 |
Neg logit contributions | |
hero | -0.073 |
sing | -0.072 |
heel | -0.072 |
promise | -0.071 |
badly | -0.071 |
Token | items |
Token ID | 3709 |
Feature activation | +1.64e-01 |
Pos logit contributions | |
flour | +0.017 |
oons | +0.015 |
pans | +0.014 |
ingredients | +0.014 |
tables | +0.014 |
Neg logit contributions | |
badly | -0.023 |
sing | -0.022 |
recovery | -0.022 |
heel | -0.022 |
former | -0.022 |
Token | . |
Token ID | 13 |
Feature activation | -1.22e-03 |
Pos logit contributions | |
flour | +0.255 |
tables | +0.224 |
oons | +0.221 |
ingredients | +0.221 |
pans | +0.216 |
Neg logit contributions | |
badly | -0.310 |
sing | -0.300 |
heel | -0.294 |
recovery | -0.292 |
hero | -0.291 |
Token | creatures |
Token ID | 8109 |
Feature activation | +1.43e-01 |
Pos logit contributions | |
flour | +0.057 |
tables | +0.051 |
oons | +0.051 |
ingredients | +0.050 |
ries | +0.047 |
Neg logit contributions | |
sing | -0.080 |
heel | -0.079 |
badly | -0.077 |
recovery | -0.076 |
promise | -0.076 |
Token | of |
Token ID | 286 |
Feature activation | +9.65e-02 |
Pos logit contributions | |
flour | +0.191 |
ingredients | +0.170 |
tables | +0.168 |
oons | +0.167 |
pots | +0.164 |
Neg logit contributions | |
sing | -0.304 |
recovery | -0.295 |
heel | -0.294 |
badly | -0.292 |
boo | -0.290 |
Token | all |
Token ID | 477 |
Feature activation | +1.09e-01 |
Pos logit contributions | |
oons | +0.112 |
flour | +0.112 |
cakes | +0.111 |
ups | +0.105 |
ingredients | +0.104 |
Neg logit contributions | |
recovery | -0.208 |
heel | -0.207 |
badly | -0.204 |
boo | -0.202 |
sing | -0.202 |
Token | shapes |
Token ID | 15268 |
Feature activation | +2.09e-01 |
Pos logit contributions | |
flour | +0.147 |
tables | +0.133 |
oons | +0.130 |
cakes | +0.128 |
pots | +0.126 |
Neg logit contributions | |
badly | -0.217 |
sing | -0.215 |
recovery | -0.214 |
promise | -0.214 |
heel | -0.212 |
Token | and |
Token ID | 290 |
Feature activation | +1.98e-02 |
Pos logit contributions | |
flour | +0.314 |
tables | +0.285 |
ingredients | +0.279 |
oons | +0.271 |
pans | +0.269 |
Neg logit contributions | |
badly | -0.401 |
sing | -0.394 |
heel | -0.382 |
recovery | -0.377 |
symp | -0.371 |
Token | sizes |
Token ID | 10620 |
Feature activation | +2.34e-01 |
Pos logit contributions | |
flour | +0.025 |
oons | +0.025 |
cakes | +0.024 |
ries | +0.023 |
ingredients | +0.022 |
Neg logit contributions | |
sing | -0.042 |
heel | -0.041 |
boo | -0.040 |
tweet | -0.039 |
roar | -0.039 |
Token | . |
Token ID | 13 |
Feature activation | +1.15e-03 |
Pos logit contributions | |
flour | +0.360 |
tables | +0.330 |
pans | +0.325 |
ingredients | +0.319 |
pots | +0.312 |
Neg logit contributions | |
badly | -0.434 |
sing | -0.422 |
heel | -0.417 |
recovery | -0.411 |
hero | -0.398 |
Token | Everyone |
Token ID | 11075 |
Feature activation | +1.61e-01 |
Pos logit contributions | |
flour | +0.002 |
oons | +0.002 |
ingredients | +0.002 |
pots | +0.002 |
pans | +0.002 |
Neg logit contributions | |
sing | -0.002 |
recovery | -0.002 |
symp | -0.002 |
heel | -0.002 |
badly | -0.002 |
Token | cheered |
Token ID | 34767 |
Feature activation | +1.09e-01 |
Pos logit contributions | |
flour | +0.233 |
ingredients | +0.207 |
tables | +0.206 |
pots | +0.201 |
oons | +0.201 |
Neg logit contributions | |
sing | -0.323 |
badly | -0.317 |
heel | -0.315 |
boo | -0.313 |
symp | -0.312 |
Token | and |
Token ID | 290 |
Feature activation | +6.49e-03 |
Pos logit contributions | |
flour | +0.169 |
ingredients | +0.153 |
tables | +0.152 |
oons | +0.151 |
pots | +0.149 |
Neg logit contributions | |
sing | -0.212 |
recovery | -0.205 |
heel | -0.204 |
badly | -0.202 |
boo | -0.201 |
Token | smiled |
Token ID | 13541 |
Feature activation | +5.92e-02 |
Pos logit contributions | |
flour | +0.012 |
oons | +0.011 |
ingredients | +0.011 |
tables | +0.010 |
pots | +0.010 |
Neg logit contributions | |
sing | -0.011 |
heel | -0.011 |
boo | -0.010 |
recovery | -0.010 |
symp | -0.010 |
Token | things |
Token ID | 1243 |
Feature activation | +1.27e-01 |
Pos logit contributions | |
oons | +0.027 |
cakes | +0.026 |
flour | +0.024 |
pans | +0.023 |
ries | +0.022 |
Neg logit contributions | |
badly | -0.028 |
heel | -0.028 |
recovery | -0.027 |
boo | -0.026 |
promise | -0.025 |
Token | of |
Token ID | 286 |
Feature activation | +9.70e-02 |
Pos logit contributions | |
flour | +0.216 |
oons | +0.205 |
tables | +0.201 |
ingredients | +0.192 |
pans | +0.192 |
Neg logit contributions | |
badly | -0.219 |
sing | -0.212 |
heel | -0.208 |
promise | -0.207 |
recovery | -0.204 |
Token | all |
Token ID | 477 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
flour | +0.138 |
oons | +0.130 |
cakes | +0.128 |
pans | +0.125 |
ingredients | +0.125 |
Neg logit contributions | |
recovery | -0.187 |
badly | -0.186 |
heel | -0.185 |
sing | -0.185 |
promise | -0.179 |
Token | shapes |
Token ID | 15268 |
Feature activation | +2.06e-01 |
Pos logit contributions | |
flour | +0.184 |
tables | +0.170 |
oons | +0.168 |
cakes | +0.165 |
pans | +0.164 |
Neg logit contributions | |
sing | -0.179 |
promise | -0.176 |
recovery | -0.175 |
badly | -0.175 |
heel | -0.173 |
Token | and |
Token ID | 290 |
Feature activation | +1.39e-02 |
Pos logit contributions | |
flour | +0.374 |
tables | +0.349 |
pans | +0.347 |
ports | +0.335 |
ingredients | +0.334 |
Neg logit contributions | |
badly | -0.313 |
forever | -0.297 |
sing | -0.295 |
heel | -0.291 |
tweet | -0.288 |
Token | sizes |
Token ID | 10620 |
Feature activation | +2.33e-01 |
Pos logit contributions | |
oons | +0.017 |
cakes | +0.017 |
flour | +0.016 |
ries | +0.015 |
ingredients | +0.013 |
Neg logit contributions | |
heel | -0.028 |
former | -0.027 |
sing | -0.027 |
roar | -0.027 |
wish | -0.026 |
Token | . |
Token ID | 13 |
Feature activation | +7.33e-05 |
Pos logit contributions | |
flour | +0.390 |
pans | +0.369 |
tables | +0.363 |
ingredients | +0.345 |
ports | +0.342 |
Neg logit contributions | |
badly | -0.391 |
heel | -0.368 |
sing | -0.361 |
recovery | -0.360 |
forever | -0.360 |
Token | When |
Token ID | 1649 |
Feature activation | +7.13e-02 |
Pos logit contributions | |
flour | +0.000 |
ingredients | +0.000 |
oons | +0.000 |
tables | +0.000 |
pots | +0.000 |
Neg logit contributions | |
sing | -0.000 |
recovery | -0.000 |
heel | -0.000 |
symp | -0.000 |
boo | -0.000 |
Token | her |
Token ID | 607 |
Feature activation | -3.61e-02 |
Pos logit contributions | |
oons | +0.098 |
flour | +0.097 |
ingredients | +0.093 |
pans | +0.087 |
cakes | +0.087 |
Neg logit contributions | |
sing | -0.139 |
badly | -0.138 |
recovery | -0.137 |
symp | -0.137 |
heel | -0.135 |
Token | pink |
Token ID | 11398 |
Feature activation | +3.30e-03 |
Pos logit contributions | |
sing | +0.066 |
recovery | +0.063 |
badly | +0.063 |
symp | +0.062 |
heel | +0.061 |
Neg logit contributions | |
flour | -0.061 |
ingredients | -0.056 |
tables | -0.055 |
pots | -0.055 |
oons | -0.053 |
Token | sack |
Token ID | 23704 |
Feature activation | +1.70e-02 |
Pos logit contributions | |
flour | +0.006 |
oons | +0.006 |
ingredients | +0.006 |
tables | +0.006 |
pots | +0.006 |
Neg logit contributions | |
sing | -0.005 |
heel | -0.005 |
recovery | -0.005 |
boo | -0.005 |
badly | -0.005 |
Token | flowers |
Token ID | 12734 |
Feature activation | +1.37e-01 |
Pos logit contributions | |
former | +0.002 |
heel | +0.002 |
sing | +0.002 |
tweet | +0.002 |
wish | +0.002 |
Neg logit contributions | |
oons | -0.001 |
flour | -0.001 |
ries | -0.001 |
cakes | -0.001 |
ingredients | -0.001 |
Token | of |
Token ID | 286 |
Feature activation | +9.33e-02 |
Pos logit contributions | |
flour | +0.207 |
oons | +0.186 |
tables | +0.185 |
ingredients | +0.182 |
pans | +0.177 |
Neg logit contributions | |
sing | -0.259 |
badly | -0.258 |
recovery | -0.252 |
heel | -0.251 |
symp | -0.249 |
Token | all |
Token ID | 477 |
Feature activation | +1.11e-01 |
Pos logit contributions | |
oons | +0.124 |
cakes | +0.116 |
ups | +0.114 |
flour | +0.113 |
ingredients | +0.107 |
Neg logit contributions | |
heel | -0.189 |
recovery | -0.188 |
badly | -0.187 |
tweet | -0.186 |
former | -0.184 |
Token | shapes |
Token ID | 15268 |
Feature activation | +2.11e-01 |
Pos logit contributions | |
flour | +0.166 |
oons | +0.153 |
tables | +0.148 |
pans | +0.144 |
cakes | +0.141 |
Neg logit contributions | |
badly | -0.207 |
promise | -0.205 |
symp | -0.201 |
sing | -0.200 |
recovery | -0.200 |
Token | and |
Token ID | 290 |
Feature activation | +1.94e-02 |
Pos logit contributions | |
flour | +0.338 |
tables | +0.312 |
pans | +0.299 |
ingredients | +0.295 |
oons | +0.294 |
Neg logit contributions | |
badly | -0.374 |
sing | -0.356 |
heel | -0.344 |
recovery | -0.344 |
symp | -0.342 |
Token | sizes |
Token ID | 10620 |
Feature activation | +2.30e-01 |
Pos logit contributions | |
oons | +0.025 |
flour | +0.023 |
cakes | +0.021 |
ries | +0.021 |
ingredients | +0.020 |
Neg logit contributions | |
former | -0.041 |
sing | -0.040 |
heel | -0.040 |
tweet | -0.040 |
roar | -0.038 |
Token | . |
Token ID | 13 |
Feature activation | -1.97e-03 |
Pos logit contributions | |
flour | +0.365 |
pans | +0.338 |
tables | +0.335 |
ingredients | +0.317 |
pots | +0.316 |
Neg logit contributions | |
badly | -0.410 |
sing | -0.388 |
recovery | -0.387 |
heel | -0.385 |
promise | -0.378 |
Token | It |
Token ID | 632 |
Feature activation | +1.41e-02 |
Pos logit contributions | |
sing | +0.003 |
recovery | +0.003 |
symp | +0.003 |
heel | +0.003 |
former | +0.003 |
Neg logit contributions | |
flour | -0.004 |
oons | -0.003 |
ingredients | -0.003 |
pots | -0.003 |
tables | -0.003 |
Token | was |
Token ID | 373 |
Feature activation | +1.34e-02 |
Pos logit contributions | |
flour | +0.023 |
oons | +0.020 |
ries | +0.020 |
tables | +0.020 |
ingredients | +0.019 |
Neg logit contributions | |
symp | -0.025 |
sing | -0.025 |
tweet | -0.025 |
promise | -0.025 |
heel | -0.025 |
Token | the |
Token ID | 262 |
Feature activation | +5.28e-02 |
Pos logit contributions | |
oons | +0.024 |
flour | +0.022 |
cakes | +0.022 |
ries | +0.021 |
pans | +0.021 |
Neg logit contributions | |
sing | -0.023 |
badly | -0.022 |
former | -0.022 |
recovery | -0.021 |
heel | -0.021 |
Token | most |
Token ID | 749 |
Feature activation | +4.27e-02 |
Pos logit contributions | |
flour | +0.086 |
oons | +0.080 |
tables | +0.079 |
ingredients | +0.078 |
pots | +0.077 |
Neg logit contributions | |
sing | -0.097 |
recovery | -0.093 |
heel | -0.093 |
badly | -0.091 |
boo | -0.091 |
Token | life |
Token ID | 1204 |
Feature activation | -7.31e-03 |
Pos logit contributions | |
sing | +0.045 |
heel | +0.043 |
former | +0.043 |
wish | +0.042 |
boo | +0.042 |
Neg logit contributions | |
flour | -0.032 |
oons | -0.031 |
ries | -0.030 |
ingredients | -0.029 |
cakes | -0.028 |
Token | . |
Token ID | 13 |
Feature activation | -1.63e-03 |
Pos logit contributions | |
heel | +0.013 |
hero | +0.013 |
recovery | +0.013 |
badly | +0.013 |
promise | +0.012 |
Neg logit contributions | |
flour | -0.011 |
oons | -0.010 |
tables | -0.010 |
ries | -0.010 |
pans | -0.010 |
Token | Fish |
Token ID | 13388 |
Feature activation | +7.90e-02 |
Pos logit contributions | |
sing | +0.003 |
badly | +0.003 |
promise | +0.003 |
recovery | +0.003 |
heel | +0.003 |
Neg logit contributions | |
flour | -0.003 |
tables | -0.003 |
ingredients | -0.003 |
pots | -0.003 |
oons | -0.003 |
Token | of |
Token ID | 286 |
Feature activation | +6.77e-02 |
Pos logit contributions | |
flour | +0.110 |
ingredients | +0.098 |
tables | +0.097 |
oons | +0.094 |
pots | +0.094 |
Neg logit contributions | |
sing | -0.164 |
badly | -0.158 |
heel | -0.158 |
symp | -0.158 |
boo | -0.157 |
Token | all |
Token ID | 477 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
flour | +0.077 |
oons | +0.073 |
cakes | +0.072 |
ingredients | +0.072 |
ups | +0.068 |
Neg logit contributions | |
recovery | -0.152 |
heel | -0.149 |
hero | -0.148 |
sing | -0.148 |
badly | -0.147 |
Token | sizes |
Token ID | 10620 |
Feature activation | +2.28e-01 |
Pos logit contributions | |
flour | +0.128 |
ingredients | +0.113 |
tables | +0.111 |
oons | +0.109 |
pots | +0.109 |
Neg logit contributions | |
sing | -0.224 |
recovery | -0.220 |
heel | -0.219 |
boo | -0.217 |
badly | -0.217 |
Token | were |
Token ID | 547 |
Feature activation | +3.88e-02 |
Pos logit contributions | |
flour | +0.352 |
tables | +0.323 |
ingredients | +0.316 |
pans | +0.310 |
pots | +0.307 |
Neg logit contributions | |
badly | -0.425 |
sing | -0.423 |
heel | -0.410 |
promise | -0.407 |
recovery | -0.405 |
Token | everywhere |
Token ID | 8347 |
Feature activation | +5.14e-02 |
Pos logit contributions | |
flour | +0.075 |
ingredients | +0.069 |
oons | +0.069 |
tables | +0.069 |
pots | +0.068 |
Neg logit contributions | |
sing | -0.060 |
recovery | -0.057 |
heel | -0.057 |
badly | -0.057 |
boo | -0.056 |
Token | , |
Token ID | 11 |
Feature activation | +4.90e-02 |
Pos logit contributions | |
flour | +0.069 |
ingredients | +0.061 |
tables | +0.061 |
pots | +0.060 |
oons | +0.060 |
Neg logit contributions | |
sing | -0.109 |
recovery | -0.107 |
heel | -0.106 |
boo | -0.105 |
symp | -0.104 |
Token | swimming |
Token ID | 14899 |
Feature activation | -2.92e-02 |
Pos logit contributions | |
flour | +0.075 |
tables | +0.068 |
ingredients | +0.068 |
pots | +0.066 |
oons | +0.065 |
Neg logit contributions | |
sing | -0.095 |
recovery | -0.093 |
heel | -0.092 |
badly | -0.091 |
boo | -0.091 |
Token | around |
Token ID | 1088 |
Feature activation | +5.60e-02 |
Pos logit contributions | |
sing | +0.053 |
badly | +0.052 |
heel | +0.051 |
symp | +0.050 |
hero | +0.050 |
Neg logit contributions | |
flour | -0.048 |
oons | -0.043 |
ingredients | -0.043 |
ries | -0.043 |
tables | -0.042 |
Token | decorations |
Token ID | 39011 |
Feature activation | +1.43e-01 |
Pos logit contributions | |
flour | +0.026 |
oons | +0.023 |
ingredients | +0.023 |
tables | +0.022 |
pots | +0.022 |
Neg logit contributions | |
sing | -0.048 |
recovery | -0.046 |
badly | -0.046 |
heel | -0.046 |
symp | -0.046 |
Token | of |
Token ID | 286 |
Feature activation | +7.72e-02 |
Pos logit contributions | |
flour | +0.229 |
oons | +0.209 |
ingredients | +0.204 |
tables | +0.204 |
pots | +0.198 |
Neg logit contributions | |
sing | -0.253 |
badly | -0.247 |
heel | -0.242 |
spine | -0.241 |
hero | -0.240 |
Token | all |
Token ID | 477 |
Feature activation | +9.57e-02 |
Pos logit contributions | |
cakes | +0.099 |
oons | +0.098 |
ups | +0.097 |
flour | +0.095 |
ingredients | +0.086 |
Neg logit contributions | |
wish | -0.142 |
recovery | -0.142 |
sing | -0.141 |
roar | -0.141 |
badly | -0.141 |
Token | shapes |
Token ID | 15268 |
Feature activation | +1.95e-01 |
Pos logit contributions | |
flour | +0.134 |
oons | +0.122 |
cakes | +0.120 |
tables | +0.119 |
pots | +0.118 |
Neg logit contributions | |
promise | -0.171 |
recovery | -0.170 |
badly | -0.169 |
sing | -0.168 |
heel | -0.168 |
Token | and |
Token ID | 290 |
Feature activation | +9.77e-03 |
Pos logit contributions | |
flour | +0.305 |
tables | +0.273 |
ingredients | +0.272 |
oons | +0.268 |
pots | +0.261 |
Neg logit contributions | |
sing | -0.356 |
badly | -0.348 |
heel | -0.341 |
hero | -0.336 |
recovery | -0.336 |
Token | sizes |
Token ID | 10620 |
Feature activation | +2.26e-01 |
Pos logit contributions | |
flour | +0.011 |
oons | +0.011 |
cakes | +0.010 |
ries | +0.010 |
ingredients | +0.009 |
Neg logit contributions | |
sing | -0.021 |
former | -0.020 |
heel | -0.020 |
roar | -0.020 |
boo | -0.020 |
Token | . |
Token ID | 13 |
Feature activation | -2.93e-03 |
Pos logit contributions | |
flour | +0.336 |
tables | +0.300 |
ingredients | +0.297 |
pans | +0.292 |
pots | +0.290 |
Neg logit contributions | |
sing | -0.427 |
badly | -0.426 |
heel | -0.417 |
recovery | -0.413 |
hero | -0.408 |
Token | He |
Token ID | 679 |
Feature activation | -5.80e-02 |
Pos logit contributions | |
sing | +0.004 |
recovery | +0.004 |
heel | +0.004 |
symp | +0.004 |
badly | +0.004 |
Neg logit contributions | |
flour | -0.006 |
oons | -0.005 |
ingredients | -0.005 |
tables | -0.005 |
pots | -0.005 |
Token | noticed |
Token ID | 6810 |
Feature activation | +6.71e-02 |
Pos logit contributions | |
sing | +0.124 |
recovery | +0.121 |
heel | +0.120 |
badly | +0.119 |
boo | +0.118 |
Neg logit contributions | |
flour | -0.077 |
ingredients | -0.069 |
tables | -0.069 |
oons | -0.068 |
pots | -0.067 |
Token | many |
Token ID | 867 |
Feature activation | +3.47e-02 |
Pos logit contributions | |
oons | +0.102 |
flour | +0.100 |
pans | +0.094 |
ingredients | +0.092 |
cakes | +0.090 |
Neg logit contributions | |
badly | -0.121 |
sing | -0.117 |
hero | -0.115 |
recovery | -0.115 |
heel | -0.114 |
Token | were |
Token ID | 547 |
Feature activation | -9.67e-03 |
Pos logit contributions | |
cakes | +0.052 |
flour | +0.051 |
oons | +0.047 |
ries | +0.045 |
ingredients | +0.045 |
Neg logit contributions | |
wish | -0.058 |
badly | -0.055 |
promise | -0.055 |
hero | -0.054 |
fears | -0.053 |
Token | paint |
Token ID | 7521 |
Feature activation | +1.21e-01 |
Pos logit contributions | |
oons | +0.010 |
ries | +0.010 |
flour | +0.009 |
cakes | +0.009 |
ingredients | +0.009 |
Neg logit contributions | |
heel | -0.011 |
boo | -0.010 |
badly | -0.010 |
hero | -0.010 |
sword | -0.010 |
Token | br |
Token ID | 1671 |
Feature activation | +1.76e-02 |
Pos logit contributions | |
flour | +0.194 |
oons | +0.191 |
ingredients | +0.188 |
pans | +0.180 |
tables | +0.179 |
Neg logit contributions | |
badly | -0.177 |
heel | -0.174 |
horn | -0.171 |
hero | -0.163 |
symp | -0.162 |
Token | ushes |
Token ID | 17237 |
Feature activation | +1.50e-01 |
Pos logit contributions | |
flour | +0.030 |
tables | +0.029 |
ingredients | +0.027 |
cakes | +0.027 |
pots | +0.027 |
Neg logit contributions | |
sing | -0.028 |
badly | -0.027 |
promise | -0.027 |
ighed | -0.027 |
symp | -0.027 |
Token | of |
Token ID | 286 |
Feature activation | +6.97e-02 |
Pos logit contributions | |
flour | +0.249 |
pans | +0.231 |
oons | +0.228 |
ingredients | +0.226 |
tables | +0.221 |
Neg logit contributions | |
badly | -0.258 |
hero | -0.242 |
sing | -0.239 |
recovery | -0.237 |
forever | -0.234 |
Token | all |
Token ID | 477 |
Feature activation | +1.18e-01 |
Pos logit contributions | |
oons | +0.084 |
cakes | +0.081 |
flour | +0.080 |
ingredients | +0.074 |
ups | +0.072 |
Neg logit contributions | |
heel | -0.144 |
boo | -0.142 |
sing | -0.138 |
badly | -0.138 |
recovery | -0.138 |
Token | sizes |
Token ID | 10620 |
Feature activation | +2.25e-01 |
Pos logit contributions | |
flour | +0.205 |
cakes | +0.199 |
tables | +0.196 |
oons | +0.189 |
liquids | +0.187 |
Neg logit contributions | |
badly | -0.160 |
promise | -0.158 |
forever | -0.153 |
hero | -0.151 |
heart | -0.149 |
Token | . |
Token ID | 13 |
Feature activation | -6.00e-03 |
Pos logit contributions | |
flour | +0.397 |
tables | +0.374 |
pans | +0.370 |
ingredients | +0.358 |
pots | +0.346 |
Neg logit contributions | |
badly | -0.356 |
agging | -0.327 |
hero | -0.317 |
heel | -0.315 |
sing | -0.313 |
Token |
|
Token ID | 198 |
Feature activation | -3.02e-02 |
Pos logit contributions | |
sing | +0.011 |
recovery | +0.010 |
heel | +0.010 |
badly | +0.010 |
boo | +0.010 |
Neg logit contributions | |
flour | -0.010 |
ingredients | -0.009 |
tables | -0.009 |
oons | -0.009 |
pots | -0.009 |
Token |
|
Token ID | 198 |
Feature activation | -3.84e-02 |
Pos logit contributions | |
sing | +0.054 |
recovery | +0.053 |
badly | +0.053 |
heel | +0.052 |
symp | +0.052 |
Neg logit contributions | |
flour | -0.050 |
ingredients | -0.046 |
tables | -0.045 |
pots | -0.044 |
oons | -0.044 |
Token | The |
Token ID | 464 |
Feature activation | +2.64e-02 |
Pos logit contributions | |
sing | +0.075 |
boo | +0.071 |
spine | +0.070 |
promise | +0.068 |
recovery | +0.068 |
Neg logit contributions | |
flour | -0.064 |
tables | -0.060 |
oons | -0.059 |
pots | -0.055 |
ingredients | -0.054 |
Token | painter |
Token ID | 34537 |
Feature activation | +5.06e-02 |
Pos logit contributions | |
flour | +0.044 |
oons | +0.043 |
ries | +0.041 |
tables | +0.041 |
ingredients | +0.040 |
Neg logit contributions | |
sing | -0.044 |
boo | -0.043 |
hero | -0.043 |
promise | -0.043 |
heel | -0.043 |
Token | , |
Token ID | 11 |
Feature activation | +6.99e-02 |
Pos logit contributions | |
flour | +0.198 |
oons | +0.187 |
tables | +0.187 |
pans | +0.181 |
pots | +0.177 |
Neg logit contributions | |
badly | -0.201 |
promise | -0.200 |
heel | -0.199 |
tweet | -0.198 |
sing | -0.196 |
Token | glitter |
Token ID | 31133 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
flour | +0.091 |
oons | +0.087 |
ingredients | +0.085 |
ries | +0.080 |
tables | +0.080 |
Neg logit contributions | |
sing | -0.139 |
heel | -0.138 |
boo | -0.136 |
badly | -0.135 |
recovery | -0.134 |
Token | , |
Token ID | 11 |
Feature activation | +5.20e-02 |
Pos logit contributions | |
tables | +0.179 |
flour | +0.174 |
pans | +0.165 |
cakes | +0.164 |
ups | +0.164 |
Neg logit contributions | |
badly | -0.137 |
heel | -0.125 |
tweet | -0.124 |
forever | -0.123 |
ute | -0.123 |
Token | and |
Token ID | 290 |
Feature activation | +9.11e-03 |
Pos logit contributions | |
oons | +0.070 |
flour | +0.070 |
ingredients | +0.067 |
cakes | +0.064 |
tables | +0.064 |
Neg logit contributions | |
sing | -0.099 |
heel | -0.097 |
boo | -0.096 |
badly | -0.094 |
symp | -0.094 |
Token | other |
Token ID | 584 |
Feature activation | +3.44e-02 |
Pos logit contributions | |
flour | +0.012 |
oons | +0.012 |
cakes | +0.011 |
ingredients | +0.011 |
tables | +0.011 |
Neg logit contributions | |
sing | -0.018 |
heel | -0.018 |
boo | -0.018 |
badly | -0.017 |
recovery | -0.017 |
Token | materials |
Token ID | 5696 |
Feature activation | +2.25e-01 |
Pos logit contributions | |
oons | +0.038 |
flour | +0.038 |
cakes | +0.037 |
ries | +0.032 |
pans | +0.032 |
Neg logit contributions | |
promise | -0.071 |
heel | -0.070 |
boo | -0.070 |
wish | -0.069 |
recovery | -0.068 |
Token | to |
Token ID | 284 |
Feature activation | +1.17e-03 |
Pos logit contributions | |
flour | +0.332 |
tables | +0.305 |
ingredients | +0.294 |
oons | +0.294 |
pans | +0.286 |
Neg logit contributions | |
badly | -0.419 |
sing | -0.417 |
promise | -0.409 |
recovery | -0.407 |
heel | -0.405 |
Token | decor |
Token ID | 11705 |
Feature activation | +5.53e-02 |
Pos logit contributions | |
flour | +0.002 |
ingredients | +0.002 |
oons | +0.002 |
tables | +0.002 |
pots | +0.002 |
Neg logit contributions | |
sing | -0.002 |
recovery | -0.002 |
heel | -0.002 |
boo | -0.002 |
promise | -0.002 |
Token | ate |
Token ID | 378 |
Feature activation | +1.30e-01 |
Pos logit contributions | |
flour | +0.092 |
tables | +0.090 |
cakes | +0.087 |
cans | +0.086 |
pots | +0.086 |
Neg logit contributions | |
former | -0.076 |
ute | -0.071 |
badly | -0.070 |
symp | -0.069 |
oop | -0.067 |
Token | his |
Token ID | 465 |
Feature activation | -1.33e-01 |
Pos logit contributions | |
flour | +0.175 |
tables | +0.160 |
oons | +0.156 |
cakes | +0.155 |
ingredients | +0.154 |
Neg logit contributions | |
badly | -0.242 |
promise | -0.224 |
heel | -0.223 |
symp | -0.220 |
sing | -0.220 |
Token | pole |
Token ID | 16825 |
Feature activation | -4.31e-02 |
Pos logit contributions | |
sing | +0.260 |
recovery | +0.252 |
heel | +0.250 |
badly | +0.249 |
boo | +0.247 |
Neg logit contributions | |
flour | -0.202 |
ingredients | -0.183 |
tables | -0.180 |
oons | -0.178 |
pots | -0.177 |
Token | all |
Token ID | 477 |
Feature activation | +9.35e-02 |
Pos logit contributions | |
flour | +0.065 |
oons | +0.065 |
cakes | +0.063 |
ingredients | +0.059 |
ups | +0.059 |
Neg logit contributions | |
heel | -0.095 |
badly | -0.092 |
sing | -0.092 |
recovery | -0.092 |
boo | -0.091 |
Token | colours |
Token ID | 18915 |
Feature activation | +1.70e-01 |
Pos logit contributions | |
flour | +0.128 |
tables | +0.126 |
oons | +0.125 |
liquids | +0.124 |
ports | +0.122 |
Neg logit contributions | |
badly | -0.156 |
promise | -0.154 |
recovery | -0.151 |
hero | -0.148 |
heel | -0.146 |
Token | , |
Token ID | 11 |
Feature activation | +7.25e-02 |
Pos logit contributions | |
flour | +0.299 |
pans | +0.288 |
tables | +0.283 |
ports | +0.281 |
oons | +0.271 |
Neg logit contributions | |
badly | -0.255 |
heel | -0.228 |
sing | -0.228 |
roar | -0.222 |
hero | -0.221 |
Token | shapes |
Token ID | 15268 |
Feature activation | +2.00e-01 |
Pos logit contributions | |
flour | +0.098 |
oons | +0.092 |
ingredients | +0.089 |
tables | +0.086 |
pots | +0.085 |
Neg logit contributions | |
sing | -0.148 |
heel | -0.144 |
recovery | -0.142 |
boo | -0.141 |
badly | -0.140 |
Token | and |
Token ID | 290 |
Feature activation | +2.39e-02 |
Pos logit contributions | |
flour | +0.342 |
tables | +0.318 |
pans | +0.317 |
ingredients | +0.310 |
oons | +0.306 |
Neg logit contributions | |
sing | -0.329 |
badly | -0.323 |
heel | -0.316 |
recovery | -0.306 |
hero | -0.304 |
Token | sizes |
Token ID | 10620 |
Feature activation | +2.25e-01 |
Pos logit contributions | |
oons | +0.030 |
cakes | +0.027 |
ries | +0.026 |
flour | +0.026 |
ups | +0.025 |
Neg logit contributions | |
heel | -0.049 |
sing | -0.047 |
boo | -0.047 |
former | -0.046 |
roar | -0.046 |
Token | . |
Token ID | 13 |
Feature activation | +5.51e-03 |
Pos logit contributions | |
flour | +0.349 |
pans | +0.318 |
tables | +0.318 |
ingredients | +0.308 |
oons | +0.300 |
Neg logit contributions | |
badly | -0.408 |
sing | -0.402 |
heel | -0.395 |
recovery | -0.390 |
hero | -0.383 |
Token | Some |
Token ID | 2773 |
Feature activation | +1.83e-01 |
Pos logit contributions | |
flour | +0.010 |
tables | +0.009 |
ingredients | +0.009 |
oons | +0.009 |
pots | +0.009 |
Neg logit contributions | |
sing | -0.010 |
recovery | -0.009 |
heel | -0.009 |
badly | -0.009 |
boo | -0.009 |
Token | were |
Token ID | 547 |
Feature activation | +4.88e-02 |
Pos logit contributions | |
flour | +0.248 |
oons | +0.230 |
pans | +0.225 |
pots | +0.220 |
ingredients | +0.218 |
Neg logit contributions | |
heel | -0.348 |
badly | -0.348 |
promise | -0.335 |
recovery | -0.331 |
wish | -0.330 |
Token | spark |
Token ID | 9009 |
Feature activation | +4.52e-02 |
Pos logit contributions | |
flour | +0.079 |
oons | +0.076 |
ingredients | +0.073 |
tables | +0.073 |
ries | +0.071 |
Neg logit contributions | |
sing | -0.087 |
heel | -0.084 |
badly | -0.083 |
boo | -0.081 |
recovery | -0.080 |
Token | ly |
Token ID | 306 |
Feature activation | +3.81e-02 |
Pos logit contributions | |
flour | +0.074 |
tables | +0.067 |
pots | +0.064 |
pans | +0.063 |
ingredients | +0.063 |
Neg logit contributions | |
badly | -0.079 |
boo | -0.076 |
heel | -0.076 |
hero | -0.075 |
symp | -0.075 |
Token | and |
Token ID | 290 |
Feature activation | +1.30e-02 |
Pos logit contributions | |
flour | +0.263 |
ingredients | +0.245 |
oons | +0.245 |
pans | +0.239 |
cakes | +0.237 |
Neg logit contributions | |
badly | -0.272 |
sing | -0.265 |
heel | -0.258 |
spine | -0.252 |
recovery | -0.246 |
Token | boxes |
Token ID | 10559 |
Feature activation | +1.60e-01 |
Pos logit contributions | |
oons | +0.015 |
ries | +0.014 |
cakes | +0.014 |
flour | +0.014 |
ingredients | +0.013 |
Neg logit contributions | |
heel | -0.028 |
sing | -0.027 |
badly | -0.027 |
boo | -0.027 |
recovery | -0.026 |
Token | and |
Token ID | 290 |
Feature activation | +2.40e-02 |
Pos logit contributions | |
flour | +0.263 |
ingredients | +0.241 |
oons | +0.240 |
pans | +0.234 |
tables | +0.229 |
Neg logit contributions | |
badly | -0.286 |
sing | -0.282 |
spine | -0.267 |
heel | -0.264 |
promise | -0.263 |
Token | pots |
Token ID | 32195 |
Feature activation | +1.93e-01 |
Pos logit contributions | |
oons | +0.028 |
cakes | +0.026 |
ries | +0.026 |
flour | +0.025 |
ingredients | +0.023 |
Neg logit contributions | |
heel | -0.051 |
boo | -0.049 |
sing | -0.049 |
badly | -0.048 |
hero | -0.047 |
Token | and |
Token ID | 290 |
Feature activation | -1.37e-02 |
Pos logit contributions | |
flour | +0.336 |
oons | +0.314 |
ingredients | +0.313 |
pans | +0.311 |
tables | +0.307 |
Neg logit contributions | |
badly | -0.305 |
sing | -0.285 |
heel | -0.273 |
symp | -0.269 |
spine | -0.268 |
Token | pans |
Token ID | 36209 |
Feature activation | +2.25e-01 |
Pos logit contributions | |
heel | +0.031 |
boo | +0.030 |
sing | +0.029 |
promise | +0.029 |
badly | +0.029 |
Neg logit contributions | |
oons | -0.014 |
ries | -0.013 |
flour | -0.013 |
cakes | -0.012 |
ingredients | -0.012 |
Token | . |
Token ID | 13 |
Feature activation | +2.23e-03 |
Pos logit contributions | |
flour | +0.346 |
oons | +0.322 |
ingredients | +0.319 |
tables | +0.311 |
cakes | +0.305 |
Neg logit contributions | |
badly | -0.395 |
sing | -0.390 |
symp | -0.373 |
heel | -0.365 |
promise | -0.363 |
Token | There |
Token ID | 1318 |
Feature activation | +6.91e-02 |
Pos logit contributions | |
flour | +0.003 |
oons | +0.003 |
ingredients | +0.003 |
ries | +0.003 |
tables | +0.003 |
Neg logit contributions | |
symp | -0.004 |
sing | -0.004 |
recovery | -0.004 |
heel | -0.004 |
badly | -0.004 |
Token | were |
Token ID | 547 |
Feature activation | +3.83e-02 |
Pos logit contributions | |
flour | +0.092 |
pans | +0.090 |
ingredients | +0.086 |
oons | +0.084 |
tables | +0.082 |
Neg logit contributions | |
badly | -0.132 |
symp | -0.130 |
promise | -0.129 |
sing | -0.129 |
heel | -0.127 |
Token | also |
Token ID | 635 |
Feature activation | -2.12e-02 |
Pos logit contributions | |
flour | +0.039 |
oons | +0.038 |
ries | +0.036 |
ingredients | +0.036 |
cakes | +0.035 |
Neg logit contributions | |
sing | -0.088 |
badly | -0.085 |
heel | -0.085 |
recovery | -0.084 |
boo | -0.083 |
Token | many |
Token ID | 867 |
Feature activation | +5.46e-02 |
Pos logit contributions | |
sing | +0.043 |
badly | +0.042 |
heel | +0.042 |
recovery | +0.041 |
former | +0.041 |
Neg logit contributions | |
flour | -0.027 |
oons | -0.027 |
ries | -0.026 |
cakes | -0.025 |
ingredients | -0.025 |
Token | wood |
Token ID | 4898 |
Feature activation | +2.84e-02 |
Pos logit contributions | |
flour | +0.170 |
oons | +0.154 |
tables | +0.153 |
ingredients | +0.150 |
pots | +0.147 |
Neg logit contributions | |
sing | -0.295 |
recovery | -0.291 |
heel | -0.289 |
boo | -0.288 |
badly | -0.287 |
Token | and |
Token ID | 290 |
Feature activation | -8.86e-03 |
Pos logit contributions | |
flour | +0.043 |
tables | +0.040 |
oons | +0.040 |
ingredients | +0.039 |
pots | +0.038 |
Neg logit contributions | |
sing | -0.053 |
badly | -0.052 |
recovery | -0.051 |
heel | -0.051 |
boo | -0.050 |
Token | paint |
Token ID | 7521 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
boo | +0.015 |
heel | +0.015 |
sing | +0.015 |
badly | +0.014 |
former | +0.014 |
Neg logit contributions | |
oons | -0.012 |
flour | -0.011 |
cakes | -0.011 |
ries | -0.011 |
ports | -0.011 |
Token | and |
Token ID | 290 |
Feature activation | +1.42e-02 |
Pos logit contributions | |
flour | +0.171 |
oons | +0.167 |
pans | +0.162 |
ingredients | +0.161 |
tables | +0.160 |
Neg logit contributions | |
badly | -0.187 |
heel | -0.171 |
symp | -0.170 |
recovery | -0.168 |
sing | -0.167 |
Token | other |
Token ID | 584 |
Feature activation | +2.67e-02 |
Pos logit contributions | |
flour | +0.019 |
oons | +0.019 |
ries | +0.018 |
ingredients | +0.018 |
cakes | +0.017 |
Neg logit contributions | |
boo | -0.026 |
sing | -0.026 |
heel | -0.026 |
former | -0.025 |
recovery | -0.025 |
Token | materials |
Token ID | 5696 |
Feature activation | +2.24e-01 |
Pos logit contributions | |
flour | +0.030 |
cakes | +0.028 |
oons | +0.028 |
ries | +0.026 |
ingredients | +0.026 |
Neg logit contributions | |
recovery | -0.056 |
sing | -0.056 |
heel | -0.056 |
badly | -0.056 |
boo | -0.056 |
Token | . |
Token ID | 13 |
Feature activation | -1.31e-02 |
Pos logit contributions | |
flour | +0.339 |
tables | +0.312 |
ingredients | +0.304 |
oons | +0.303 |
pans | +0.297 |
Neg logit contributions | |
badly | -0.418 |
sing | -0.412 |
recovery | -0.404 |
heel | -0.400 |
promise | -0.394 |
Token | With |
Token ID | 2080 |
Feature activation | +6.90e-02 |
Pos logit contributions | |
sing | +0.021 |
recovery | +0.021 |
heel | +0.020 |
boo | +0.020 |
badly | +0.020 |
Neg logit contributions | |
flour | -0.024 |
tables | -0.022 |
ingredients | -0.022 |
oons | -0.022 |
pots | -0.022 |
Token | a |
Token ID | 257 |
Feature activation | +4.24e-02 |
Pos logit contributions | |
flour | +0.102 |
oons | +0.098 |
cakes | +0.096 |
ingredients | +0.096 |
tables | +0.094 |
Neg logit contributions | |
sing | -0.129 |
recovery | -0.126 |
badly | -0.125 |
boo | -0.124 |
heel | -0.124 |
Token | lot |
Token ID | 1256 |
Feature activation | -4.54e-02 |
Pos logit contributions | |
flour | +0.072 |
ingredients | +0.065 |
pots | +0.063 |
tables | +0.062 |
pans | +0.061 |
Neg logit contributions | |
sing | -0.078 |
recovery | -0.076 |
badly | -0.076 |
symp | -0.075 |
heel | -0.074 |
Token | of |
Token ID | 286 |
Feature activation | +9.84e-03 |
Pos logit contributions | |
recovery | +0.077 |
badly | +0.074 |
hero | +0.073 |
promise | +0.072 |
aring | +0.069 |
Neg logit contributions | |
tables | -0.069 |
ries | -0.069 |
flour | -0.069 |
pans | -0.066 |
ports | -0.066 |
Token | potatoes |
Token ID | 18821 |
Feature activation | +6.59e-02 |
Pos logit contributions | |
sing | +0.041 |
badly | +0.041 |
heel | +0.041 |
recovery | +0.041 |
boo | +0.040 |
Neg logit contributions | |
flour | -0.033 |
oons | -0.033 |
ingredients | -0.030 |
cakes | -0.030 |
tables | -0.030 |
Token | of |
Token ID | 286 |
Feature activation | +3.75e-02 |
Pos logit contributions | |
flour | +0.100 |
oons | +0.091 |
tables | +0.089 |
ingredients | +0.088 |
pots | +0.087 |
Neg logit contributions | |
sing | -0.122 |
heel | -0.120 |
badly | -0.119 |
recovery | -0.118 |
promise | -0.116 |
Token | all |
Token ID | 477 |
Feature activation | +8.72e-02 |
Pos logit contributions | |
cakes | +0.052 |
oons | +0.051 |
ups | +0.051 |
flour | +0.048 |
ries | +0.047 |
Neg logit contributions | |
heel | -0.069 |
recovery | -0.069 |
sing | -0.069 |
boo | -0.067 |
badly | -0.067 |
Token | shapes |
Token ID | 15268 |
Feature activation | +1.94e-01 |
Pos logit contributions | |
flour | +0.114 |
tables | +0.107 |
oons | +0.104 |
pots | +0.101 |
cakes | +0.100 |
Neg logit contributions | |
badly | -0.170 |
promise | -0.169 |
sing | -0.169 |
recovery | -0.167 |
heel | -0.164 |
Token | and |
Token ID | 290 |
Feature activation | +1.09e-02 |
Pos logit contributions | |
flour | +0.321 |
tables | +0.297 |
pans | +0.284 |
oons | +0.284 |
ingredients | +0.282 |
Neg logit contributions | |
badly | -0.326 |
sing | -0.325 |
heel | -0.307 |
recovery | -0.302 |
symp | -0.299 |
Token | sizes |
Token ID | 10620 |
Feature activation | +2.24e-01 |
Pos logit contributions | |
oons | +0.014 |
cakes | +0.013 |
ries | +0.013 |
flour | +0.012 |
ups | +0.012 |
Neg logit contributions | |
heel | -0.021 |
former | -0.020 |
sing | -0.020 |
boo | -0.020 |
roar | -0.020 |
Token | . |
Token ID | 13 |
Feature activation | -3.52e-03 |
Pos logit contributions | |
flour | +0.348 |
tables | +0.321 |
pans | +0.315 |
pots | +0.307 |
ingredients | +0.304 |
Neg logit contributions | |
badly | -0.400 |
sing | -0.388 |
heel | -0.381 |
recovery | -0.378 |
promise | -0.372 |
Token | And |
Token ID | 843 |
Feature activation | -3.23e-02 |
Pos logit contributions | |
sing | +0.006 |
recovery | +0.006 |
heel | +0.006 |
badly | +0.006 |
boo | +0.006 |
Neg logit contributions | |
flour | -0.006 |
ingredients | -0.006 |
tables | -0.006 |
oons | -0.006 |
pots | -0.006 |
Token | the |
Token ID | 262 |
Feature activation | +9.10e-02 |
Pos logit contributions | |
sing | +0.069 |
heel | +0.067 |
recovery | +0.067 |
badly | +0.067 |
symp | +0.066 |
Neg logit contributions | |
flour | -0.043 |
oons | -0.038 |
ingredients | -0.038 |
tables | -0.038 |
pots | -0.037 |
Token | tall |
Token ID | 7331 |
Feature activation | +2.08e-02 |
Pos logit contributions | |
flour | +0.140 |
oons | +0.131 |
ries | +0.126 |
tables | +0.126 |
ingredients | +0.124 |
Neg logit contributions | |
sing | -0.171 |
heel | -0.169 |
recovery | -0.168 |
hero | -0.166 |
boo | -0.165 |
Token | potato |
Token ID | 21219 |
Feature activation | -1.17e-02 |
Pos logit contributions | |
flour | +0.037 |
oons | +0.034 |
ingredients | +0.034 |
tables | +0.033 |
pots | +0.033 |
Neg logit contributions | |
sing | -0.034 |
heel | -0.034 |
recovery | -0.034 |
symp | -0.033 |
boo | -0.033 |
Token | all |
Token ID | 477 |
Feature activation | +8.00e-02 |
Pos logit contributions | |
flour | +0.050 |
oons | +0.046 |
ingredients | +0.046 |
tables | +0.046 |
pots | +0.045 |
Neg logit contributions | |
sing | -0.038 |
heel | -0.036 |
badly | -0.036 |
recovery | -0.035 |
boo | -0.035 |
Token | colors |
Token ID | 7577 |
Feature activation | +1.66e-01 |
Pos logit contributions | |
flour | +0.154 |
ingredients | +0.144 |
pots | +0.141 |
tables | +0.141 |
oons | +0.134 |
Neg logit contributions | |
sing | -0.123 |
recovery | -0.121 |
symp | -0.119 |
heel | -0.119 |
former | -0.115 |
Token | , |
Token ID | 11 |
Feature activation | +6.76e-02 |
Pos logit contributions | |
flour | +0.264 |
tables | +0.247 |
ingredients | +0.240 |
pans | +0.238 |
oons | +0.236 |
Neg logit contributions | |
sing | -0.306 |
badly | -0.306 |
heel | -0.289 |
recovery | -0.285 |
symp | -0.285 |
Token | shapes |
Token ID | 15268 |
Feature activation | +1.94e-01 |
Pos logit contributions | |
flour | +0.103 |
oons | +0.098 |
ingredients | +0.094 |
tables | +0.092 |
pots | +0.090 |
Neg logit contributions | |
sing | -0.128 |
heel | -0.123 |
recovery | -0.122 |
boo | -0.122 |
badly | -0.121 |
Token | and |
Token ID | 290 |
Feature activation | +1.92e-02 |
Pos logit contributions | |
flour | +0.311 |
tables | +0.289 |
ingredients | +0.282 |
pans | +0.278 |
oons | +0.278 |
Neg logit contributions | |
sing | -0.354 |
badly | -0.343 |
heel | -0.335 |
recovery | -0.329 |
tweet | -0.325 |
Token | sizes |
Token ID | 10620 |
Feature activation | +2.24e-01 |
Pos logit contributions | |
oons | +0.025 |
flour | +0.023 |
ries | +0.022 |
ingredients | +0.021 |
cakes | +0.021 |
Neg logit contributions | |
sing | -0.039 |
heel | -0.038 |
boo | -0.038 |
roar | -0.038 |
tweet | -0.037 |
Token | . |
Token ID | 13 |
Feature activation | +6.53e-03 |
Pos logit contributions | |
flour | +0.333 |
tables | +0.306 |
pans | +0.299 |
ingredients | +0.297 |
oons | +0.288 |
Neg logit contributions | |
badly | -0.422 |
sing | -0.422 |
heel | -0.409 |
recovery | -0.404 |
promise | -0.397 |
Token | They |
Token ID | 1119 |
Feature activation | +3.32e-02 |
Pos logit contributions | |
oons | +0.012 |
flour | +0.012 |
ries | +0.011 |
ingredients | +0.011 |
pots | +0.011 |
Neg logit contributions | |
sing | -0.010 |
symp | -0.009 |
berth | -0.009 |
bye | -0.009 |
spine | -0.009 |
Token | ran |
Token ID | 4966 |
Feature activation | +2.14e-02 |
Pos logit contributions | |
flour | +0.050 |
ingredients | +0.046 |
oons | +0.045 |
tables | +0.043 |
ries | +0.043 |
Neg logit contributions | |
sing | -0.064 |
heel | -0.062 |
boo | -0.061 |
badly | -0.061 |
promise | -0.060 |
Token | around |
Token ID | 1088 |
Feature activation | +5.62e-02 |
Pos logit contributions | |
flour | +0.030 |
ingredients | +0.027 |
oons | +0.027 |
tables | +0.026 |
pots | +0.026 |
Neg logit contributions | |
sing | -0.044 |
heel | -0.043 |
recovery | -0.042 |
badly | -0.042 |
boo | -0.042 |
Token | the |
Token ID | 262 |
Feature activation | +4.21e-02 |
Pos logit contributions | |
flour | +0.061 |
tables | +0.053 |
ingredients | +0.053 |
pots | +0.052 |
oons | +0.050 |
Neg logit contributions | |
sing | -0.134 |
recovery | -0.131 |
heel | -0.131 |
badly | -0.130 |
boo | -0.129 |
Token | and |
Token ID | 290 |
Feature activation | -1.57e-02 |
Pos logit contributions | |
flour | +0.050 |
ingredients | +0.045 |
tables | +0.045 |
pots | +0.044 |
oons | +0.043 |
Neg logit contributions | |
sing | -0.066 |
recovery | -0.064 |
heel | -0.063 |
badly | -0.063 |
symp | -0.063 |
Token | all |
Token ID | 477 |
Feature activation | +1.25e-01 |
Pos logit contributions | |
sing | +0.033 |
heel | +0.033 |
badly | +0.032 |
recovery | +0.032 |
boo | +0.032 |
Neg logit contributions | |
flour | -0.021 |
oons | -0.019 |
ingredients | -0.019 |
tables | -0.018 |
pots | -0.018 |
Token | different |
Token ID | 1180 |
Feature activation | +5.09e-02 |
Pos logit contributions | |
flour | +0.149 |
ingredients | +0.129 |
oons | +0.128 |
tables | +0.125 |
pots | +0.124 |
Neg logit contributions | |
sing | -0.277 |
recovery | -0.277 |
badly | -0.274 |
heel | -0.273 |
hero | -0.272 |
Token | shapes |
Token ID | 15268 |
Feature activation | +2.06e-01 |
Pos logit contributions | |
flour | +0.070 |
oons | +0.064 |
cakes | +0.062 |
pans | +0.059 |
tables | +0.057 |
Neg logit contributions | |
sing | -0.097 |
badly | -0.096 |
former | -0.096 |
recovery | -0.095 |
hero | -0.095 |
Token | and |
Token ID | 290 |
Feature activation | +7.14e-03 |
Pos logit contributions | |
flour | +0.318 |
tables | +0.287 |
ingredients | +0.279 |
oons | +0.276 |
pans | +0.275 |
Neg logit contributions | |
badly | -0.380 |
sing | -0.378 |
recovery | -0.366 |
heel | -0.365 |
hero | -0.365 |
Token | sizes |
Token ID | 10620 |
Feature activation | +2.24e-01 |
Pos logit contributions | |
oons | +0.009 |
flour | +0.009 |
ries | +0.008 |
cakes | +0.008 |
ingredients | +0.008 |
Neg logit contributions | |
sing | -0.015 |
heel | -0.015 |
boo | -0.014 |
recovery | -0.014 |
former | -0.014 |
Token | . |
Token ID | 13 |
Feature activation | -6.41e-03 |
Pos logit contributions | |
flour | +0.330 |
tables | +0.296 |
ingredients | +0.290 |
pans | +0.287 |
pots | +0.284 |
Neg logit contributions | |
badly | -0.429 |
sing | -0.425 |
recovery | -0.419 |
heel | -0.417 |
hero | -0.411 |
Token | Every |
Token ID | 3887 |
Feature activation | +6.52e-02 |
Pos logit contributions | |
sing | +0.013 |
heel | +0.012 |
badly | +0.012 |
boo | +0.012 |
promise | +0.012 |
Neg logit contributions | |
flour | -0.010 |
tables | -0.009 |
ingredients | -0.009 |
oons | -0.009 |
pots | -0.009 |
Token | day |
Token ID | 1110 |
Feature activation | -1.36e-02 |
Pos logit contributions | |
flour | +0.099 |
ingredients | +0.091 |
pots | +0.086 |
tables | +0.085 |
oons | +0.083 |
Neg logit contributions | |
sing | -0.131 |
heel | -0.126 |
spine | -0.126 |
boo | -0.125 |
symp | -0.124 |
Token | the |
Token ID | 262 |
Feature activation | +7.20e-02 |
Pos logit contributions | |
hero | +0.026 |
heel | +0.026 |
badly | +0.026 |
recovery | +0.026 |
sing | +0.025 |
Neg logit contributions | |
flour | -0.020 |
oons | -0.019 |
ingredients | -0.018 |
ries | -0.018 |
pots | -0.018 |
Token | bear |
Token ID | 6842 |
Feature activation | -2.65e-02 |
Pos logit contributions | |
flour | +0.123 |
ingredients | +0.112 |
tables | +0.112 |
pots | +0.110 |
oons | +0.108 |
Neg logit contributions | |
sing | -0.130 |
badly | -0.124 |
recovery | -0.124 |
heel | -0.123 |
symp | -0.123 |
Token | toys |
Token ID | 14958 |
Feature activation | +8.38e-02 |
Pos logit contributions | |
oons | +0.051 |
ries | +0.048 |
cakes | +0.045 |
flour | +0.045 |
liquids | +0.044 |
Neg logit contributions | |
heel | -0.051 |
badly | -0.051 |
boo | -0.050 |
sing | -0.049 |
recovery | -0.047 |
Token | of |
Token ID | 286 |
Feature activation | +6.04e-02 |
Pos logit contributions | |
flour | +0.139 |
oons | +0.132 |
tables | +0.128 |
pans | +0.127 |
ingredients | +0.127 |
Neg logit contributions | |
badly | -0.146 |
heel | -0.144 |
sing | -0.142 |
hero | -0.137 |
recovery | -0.137 |
Token | all |
Token ID | 477 |
Feature activation | +8.35e-02 |
Pos logit contributions | |
flour | +0.089 |
oons | +0.084 |
cakes | +0.083 |
ingredients | +0.082 |
tables | +0.080 |
Neg logit contributions | |
heel | -0.112 |
boo | -0.110 |
sing | -0.110 |
recovery | -0.109 |
badly | -0.108 |
Token | shapes |
Token ID | 15268 |
Feature activation | +1.92e-01 |
Pos logit contributions | |
flour | +0.160 |
tables | +0.154 |
oons | +0.149 |
pans | +0.148 |
cakes | +0.146 |
Neg logit contributions | |
badly | -0.117 |
sing | -0.116 |
heel | -0.115 |
promise | -0.114 |
recovery | -0.113 |
Token | and |
Token ID | 290 |
Feature activation | +8.71e-03 |
Pos logit contributions | |
flour | +0.356 |
tables | +0.349 |
pans | +0.341 |
ports | +0.333 |
oons | +0.324 |
Neg logit contributions | |
badly | -0.284 |
sing | -0.258 |
forever | -0.256 |
heel | -0.252 |
tweet | -0.250 |
Token | sizes |
Token ID | 10620 |
Feature activation | +2.23e-01 |
Pos logit contributions | |
oons | +0.011 |
cakes | +0.010 |
flour | +0.010 |
ries | +0.010 |
ports | +0.009 |
Neg logit contributions | |
heel | -0.017 |
boo | -0.016 |
roar | -0.016 |
tune | -0.016 |
personality | -0.016 |
Token | . |
Token ID | 13 |
Feature activation | +2.94e-03 |
Pos logit contributions | |
flour | +0.388 |
pans | +0.382 |
tables | +0.371 |
ports | +0.345 |
ingredients | +0.343 |
Neg logit contributions | |
badly | -0.361 |
agging | -0.334 |
heel | -0.328 |
forever | -0.325 |
recovery | -0.321 |
Token | Her |
Token ID | 2332 |
Feature activation | -7.95e-02 |
Pos logit contributions | |
flour | +0.005 |
ingredients | +0.005 |
pans | +0.005 |
tables | +0.005 |
oons | +0.005 |
Neg logit contributions | |
sing | -0.005 |
symp | -0.005 |
recovery | -0.005 |
heel | -0.005 |
promise | -0.004 |
Token | favorite |
Token ID | 4004 |
Feature activation | +1.57e-02 |
Pos logit contributions | |
sing | +0.122 |
recovery | +0.122 |
badly | +0.120 |
symp | +0.119 |
heel | +0.118 |
Neg logit contributions | |
flour | -0.153 |
ingredients | -0.143 |
pots | -0.140 |
tables | -0.140 |
oons | -0.137 |
Token | were |
Token ID | 547 |
Feature activation | +2.51e-02 |
Pos logit contributions | |
flour | +0.024 |
oons | +0.022 |
tables | +0.022 |
pans | +0.021 |
ries | +0.021 |
Neg logit contributions | |
sing | -0.029 |
heel | -0.029 |
boo | -0.028 |
recovery | -0.028 |
badly | -0.028 |
Token | two |
Token ID | 734 |
Feature activation | +1.33e-01 |
Pos logit contributions | |
oons | +0.041 |
ports | +0.038 |
flour | +0.037 |
ingredients | +0.037 |
liquids | +0.037 |
Neg logit contributions | |
badly | -0.039 |
heel | -0.038 |
boo | -0.037 |
sing | -0.037 |
hero | -0.037 |
Token | boxes |
Token ID | 10559 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
oons | +0.013 |
flour | +0.012 |
cakes | +0.012 |
ups | +0.012 |
ingredients | +0.011 |
Neg logit contributions | |
heel | -0.022 |
boo | -0.021 |
hero | -0.021 |
badly | -0.021 |
recovery | -0.020 |
Token | , |
Token ID | 11 |
Feature activation | +5.18e-02 |
Pos logit contributions | |
flour | +0.187 |
ingredients | +0.177 |
oons | +0.173 |
pans | +0.169 |
tables | +0.168 |
Neg logit contributions | |
badly | -0.198 |
sing | -0.190 |
symp | -0.185 |
heel | -0.183 |
promise | -0.183 |
Token | colorful |
Token ID | 20239 |
Feature activation | +3.98e-02 |
Pos logit contributions | |
oons | +0.078 |
ries | +0.069 |
ups | +0.068 |
ingredients | +0.068 |
cakes | +0.068 |
Neg logit contributions | |
heel | -0.090 |
badly | -0.089 |
boo | -0.086 |
sing | -0.084 |
symp | -0.083 |
Token | toys |
Token ID | 14958 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
oons | +0.042 |
flour | +0.041 |
cakes | +0.041 |
ups | +0.040 |
hes | +0.038 |
Neg logit contributions | |
heel | -0.076 |
badly | -0.072 |
hero | -0.071 |
horn | -0.070 |
boo | -0.069 |
Token | and |
Token ID | 290 |
Feature activation | +2.25e-03 |
Pos logit contributions | |
flour | +0.189 |
oons | +0.178 |
ingredients | +0.176 |
pans | +0.175 |
tables | +0.168 |
Neg logit contributions | |
badly | -0.168 |
heel | -0.158 |
sing | -0.152 |
hero | -0.148 |
promise | -0.148 |
Token | walls |
Token ID | 7714 |
Feature activation | +2.22e-01 |
Pos logit contributions | |
oons | +0.004 |
ries | +0.004 |
flour | +0.003 |
ingredients | +0.003 |
cakes | +0.003 |
Neg logit contributions | |
heel | -0.004 |
badly | -0.004 |
boo | -0.004 |
wish | -0.003 |
roar | -0.003 |
Token | full |
Token ID | 1336 |
Feature activation | +5.23e-02 |
Pos logit contributions | |
flour | +0.356 |
pans | +0.334 |
ingredients | +0.329 |
tables | +0.325 |
ports | +0.323 |
Neg logit contributions | |
badly | -0.384 |
sing | -0.357 |
agging | -0.343 |
symp | -0.341 |
spine | -0.339 |
Token | of |
Token ID | 286 |
Feature activation | +4.80e-02 |
Pos logit contributions | |
oons | +0.068 |
flour | +0.066 |
pans | +0.065 |
ries | +0.063 |
ingredients | +0.062 |
Neg logit contributions | |
recovery | -0.096 |
badly | -0.095 |
heel | -0.095 |
hero | -0.091 |
sing | -0.090 |
Token | books |
Token ID | 3835 |
Feature activation | +7.40e-02 |
Pos logit contributions | |
oons | +0.065 |
cakes | +0.062 |
ups | +0.062 |
flour | +0.061 |
ries | +0.058 |
Neg logit contributions | |
badly | -0.086 |
boo | -0.086 |
heel | -0.086 |
recovery | -0.084 |
wish | -0.082 |
Token | . |
Token ID | 13 |
Feature activation | -7.08e-03 |
Pos logit contributions | |
flour | +0.122 |
ingredients | +0.113 |
tables | +0.109 |
pans | +0.108 |
pots | +0.107 |
Neg logit contributions | |
badly | -0.132 |
sing | -0.118 |
promise | -0.115 |
symp | -0.113 |
forever | -0.112 |
Token | He |
Token ID | 679 |
Feature activation | -5.60e-02 |
Pos logit contributions | |
sing | +0.010 |
recovery | +0.009 |
heel | +0.009 |
badly | +0.009 |
boo | +0.009 |
Neg logit contributions | |
flour | -0.015 |
ingredients | -0.014 |
tables | -0.014 |
oons | -0.014 |
pots | -0.013 |
Token | gave |
Token ID | 2921 |
Feature activation | +4.79e-02 |
Pos logit contributions | |
flour | +0.152 |
ingredients | +0.135 |
oons | +0.133 |
tables | +0.133 |
pots | +0.132 |
Neg logit contributions | |
sing | -0.220 |
recovery | -0.217 |
heel | -0.216 |
badly | -0.215 |
symp | -0.214 |
Token | him |
Token ID | 683 |
Feature activation | -6.59e-02 |
Pos logit contributions | |
flour | +0.097 |
oons | +0.091 |
tables | +0.089 |
ingredients | +0.088 |
pots | +0.088 |
Neg logit contributions | |
heel | -0.064 |
badly | -0.064 |
sing | -0.064 |
recovery | -0.064 |
boo | -0.063 |
Token | some |
Token ID | 617 |
Feature activation | +1.33e-01 |
Pos logit contributions | |
sing | +0.139 |
heel | +0.139 |
recovery | +0.138 |
badly | +0.138 |
boo | +0.136 |
Neg logit contributions | |
flour | -0.084 |
oons | -0.075 |
tables | -0.075 |
ingredients | -0.074 |
pots | -0.073 |
Token | parts |
Token ID | 3354 |
Feature activation | +1.18e-01 |
Pos logit contributions | |
flour | +0.188 |
tables | +0.171 |
oons | +0.170 |
ingredients | +0.168 |
pots | +0.166 |
Neg logit contributions | |
sing | -0.260 |
recovery | -0.259 |
heel | -0.256 |
badly | -0.254 |
boo | -0.253 |
Token | and |
Token ID | 290 |
Feature activation | +4.70e-03 |
Pos logit contributions | |
flour | +0.166 |
tables | +0.153 |
oons | +0.147 |
ingredients | +0.147 |
pots | +0.145 |
Neg logit contributions | |
sing | -0.233 |
badly | -0.231 |
recovery | -0.227 |
heel | -0.226 |
promise | -0.222 |
Token | materials |
Token ID | 5696 |
Feature activation | +2.22e-01 |
Pos logit contributions | |
flour | +0.007 |
oons | +0.006 |
ries | +0.006 |
ingredients | +0.006 |
tables | +0.006 |
Neg logit contributions | |
heel | -0.009 |
boo | -0.009 |
sing | -0.009 |
recovery | -0.009 |
badly | -0.009 |
Token | to |
Token ID | 284 |
Feature activation | +1.78e-03 |
Pos logit contributions | |
flour | +0.352 |
tables | +0.325 |
oons | +0.313 |
ingredients | +0.311 |
pots | +0.310 |
Neg logit contributions | |
badly | -0.394 |
sing | -0.383 |
recovery | -0.382 |
heel | -0.375 |
promise | -0.373 |
Token | make |
Token ID | 787 |
Feature activation | +7.88e-02 |
Pos logit contributions | |
flour | +0.003 |
ingredients | +0.003 |
oons | +0.003 |
tables | +0.003 |
pots | +0.003 |
Neg logit contributions | |
sing | -0.003 |
recovery | -0.003 |
heel | -0.003 |
badly | -0.003 |
boo | -0.003 |
Token | the |
Token ID | 262 |
Feature activation | +2.25e-02 |
Pos logit contributions | |
flour | +0.107 |
oons | +0.099 |
tables | +0.096 |
cakes | +0.095 |
ingredients | +0.095 |
Neg logit contributions | |
badly | -0.151 |
recovery | -0.150 |
sing | -0.149 |
heel | -0.148 |
boo | -0.147 |
Token | car |
Token ID | 1097 |
Feature activation | -7.48e-02 |
Pos logit contributions | |
flour | +0.034 |
ingredients | +0.031 |
tables | +0.030 |
pots | +0.030 |
oons | +0.029 |
Neg logit contributions | |
sing | -0.045 |
recovery | -0.043 |
badly | -0.043 |
symp | -0.043 |
heel | -0.043 |
Token | . |
Token ID | 13 |
Feature activation | -1.49e-02 |
Pos logit contributions | |
sing | +0.128 |
recovery | +0.124 |
heel | +0.122 |
symp | +0.121 |
boo | +0.121 |
Neg logit contributions | |
flour | -0.132 |
ingredients | -0.121 |
tables | -0.120 |
pots | -0.118 |
oons | -0.118 |
Token | colors |
Token ID | 7577 |
Feature activation | +1.37e-01 |
Pos logit contributions | |
badly | +0.019 |
heel | +0.019 |
former | +0.019 |
boo | +0.018 |
wish | +0.018 |
Neg logit contributions | |
cakes | -0.012 |
oons | -0.011 |
flour | -0.010 |
pans | -0.010 |
hes | -0.010 |
Token | , |
Token ID | 11 |
Feature activation | +4.92e-02 |
Pos logit contributions | |
pans | +0.247 |
flour | +0.242 |
tables | +0.237 |
ports | +0.235 |
ingredients | +0.231 |
Neg logit contributions | |
badly | -0.204 |
heel | -0.194 |
sing | -0.186 |
recovery | -0.182 |
wing | -0.177 |
Token | shapes |
Token ID | 15268 |
Feature activation | +1.84e-01 |
Pos logit contributions | |
flour | +0.069 |
oons | +0.067 |
ingredients | +0.064 |
tables | +0.063 |
pots | +0.061 |
Neg logit contributions | |
sing | -0.097 |
heel | -0.094 |
badly | -0.093 |
boo | -0.092 |
recovery | -0.091 |
Token | , |
Token ID | 11 |
Feature activation | +7.12e-02 |
Pos logit contributions | |
flour | +0.314 |
pans | +0.295 |
tables | +0.293 |
ingredients | +0.291 |
oons | +0.282 |
Neg logit contributions | |
sing | -0.303 |
badly | -0.301 |
heel | -0.296 |
recovery | -0.290 |
tweet | -0.285 |
Token | and |
Token ID | 290 |
Feature activation | +1.67e-02 |
Pos logit contributions | |
oons | +0.105 |
flour | +0.101 |
ingredients | +0.097 |
ups | +0.095 |
cakes | +0.094 |
Neg logit contributions | |
sing | -0.132 |
heel | -0.128 |
boo | -0.126 |
tweet | -0.123 |
badly | -0.121 |
Token | sizes |
Token ID | 10620 |
Feature activation | +2.21e-01 |
Pos logit contributions | |
oons | +0.021 |
flour | +0.019 |
cakes | +0.019 |
ries | +0.018 |
ingredients | +0.018 |
Neg logit contributions | |
boo | -0.034 |
heel | -0.034 |
sing | -0.034 |
badly | -0.033 |
recovery | -0.032 |
Token | . |
Token ID | 13 |
Feature activation | +1.11e-03 |
Pos logit contributions | |
flour | +0.356 |
pans | +0.333 |
tables | +0.332 |
ingredients | +0.322 |
pots | +0.307 |
Neg logit contributions | |
badly | -0.388 |
heel | -0.373 |
sing | -0.370 |
recovery | -0.367 |
promise | -0.365 |
Token | But |
Token ID | 887 |
Feature activation | +5.43e-03 |
Pos logit contributions | |
flour | +0.002 |
ingredients | +0.002 |
tables | +0.002 |
oons | +0.002 |
pots | +0.002 |
Neg logit contributions | |
sing | -0.002 |
recovery | -0.002 |
heel | -0.002 |
badly | -0.002 |
symp | -0.002 |
Token | the |
Token ID | 262 |
Feature activation | +8.21e-02 |
Pos logit contributions | |
flour | +0.006 |
pots | +0.005 |
ingredients | +0.005 |
tables | +0.005 |
ries | +0.004 |
Neg logit contributions | |
recovery | -0.013 |
heel | -0.013 |
sing | -0.013 |
symp | -0.013 |
boo | -0.013 |
Token | cube |
Token ID | 23441 |
Feature activation | +3.25e-02 |
Pos logit contributions | |
flour | +0.153 |
oons | +0.143 |
ingredients | +0.141 |
tables | +0.140 |
pots | +0.138 |
Neg logit contributions | |
sing | -0.130 |
heel | -0.125 |
boo | -0.125 |
recovery | -0.124 |
badly | -0.122 |
Token | loved |
Token ID | 6151 |
Feature activation | +1.37e-02 |
Pos logit contributions | |
flour | +0.041 |
ingredients | +0.036 |
tables | +0.036 |
oons | +0.035 |
pots | +0.035 |
Neg logit contributions | |
sing | -0.072 |
recovery | -0.070 |
heel | -0.070 |
badly | -0.069 |
boo | -0.069 |
Token | . |
Token ID | 13 |
Feature activation | -6.19e-03 |
Pos logit contributions | |
flour | +0.056 |
ingredients | +0.049 |
tables | +0.048 |
oons | +0.048 |
pots | +0.047 |
Neg logit contributions | |
sing | -0.104 |
recovery | -0.101 |
heel | -0.101 |
badly | -0.100 |
symp | -0.099 |
Token | All |
Token ID | 1439 |
Feature activation | +1.16e-01 |
Pos logit contributions | |
sing | +0.007 |
heel | +0.006 |
recovery | +0.006 |
badly | +0.006 |
boo | +0.006 |
Neg logit contributions | |
flour | -0.015 |
oons | -0.014 |
pots | -0.014 |
ingredients | -0.014 |
ries | -0.013 |
Token | different |
Token ID | 1180 |
Feature activation | +3.83e-02 |
Pos logit contributions | |
flour | +0.177 |
oons | +0.165 |
tables | +0.163 |
pots | +0.157 |
cakes | +0.154 |
Neg logit contributions | |
badly | -0.200 |
heel | -0.200 |
hero | -0.198 |
recovery | -0.193 |
boo | -0.191 |
Token | shapes |
Token ID | 15268 |
Feature activation | +1.91e-01 |
Pos logit contributions | |
cakes | +0.053 |
flour | +0.050 |
oons | +0.046 |
pans | +0.045 |
tables | +0.044 |
Neg logit contributions | |
heel | -0.067 |
badly | -0.066 |
symp | -0.066 |
former | -0.065 |
fears | -0.065 |
Token | and |
Token ID | 290 |
Feature activation | +1.12e-02 |
Pos logit contributions | |
flour | +0.315 |
tables | +0.294 |
ingredients | +0.284 |
pans | +0.281 |
oons | +0.277 |
Neg logit contributions | |
badly | -0.321 |
sing | -0.318 |
heel | -0.308 |
tweet | -0.303 |
roar | -0.301 |
Token | sizes |
Token ID | 10620 |
Feature activation | +2.21e-01 |
Pos logit contributions | |
oons | +0.014 |
cakes | +0.013 |
ries | +0.012 |
flour | +0.012 |
ports | +0.012 |
Neg logit contributions | |
roar | -0.021 |
sing | -0.021 |
heel | -0.021 |
former | -0.021 |
boo | -0.021 |
Token | . |
Token ID | 13 |
Feature activation | +7.78e-03 |
Pos logit contributions | |
flour | +0.343 |
tables | +0.321 |
pans | +0.312 |
ingredients | +0.305 |
pots | +0.299 |
Neg logit contributions | |
badly | -0.396 |
sing | -0.389 |
heel | -0.382 |
recovery | -0.371 |
promise | -0.370 |
Token | He |
Token ID | 679 |
Feature activation | -5.07e-02 |
Pos logit contributions | |
flour | +0.016 |
oons | +0.015 |
pots | +0.015 |
pans | +0.015 |
ingredients | +0.015 |
Neg logit contributions | |
sing | -0.010 |
recovery | -0.009 |
symp | -0.009 |
badly | -0.009 |
berth | -0.009 |
Token | could |
Token ID | 714 |
Feature activation | -5.39e-03 |
Pos logit contributions | |
sing | +0.098 |
heel | +0.096 |
boo | +0.095 |
symp | +0.094 |
badly | +0.093 |
Neg logit contributions | |
flour | -0.076 |
ingredients | -0.068 |
tables | -0.068 |
oons | -0.068 |
ries | -0.066 |
Token | hardly |
Token ID | 8941 |
Feature activation | -9.23e-03 |
Pos logit contributions | |
recovery | +0.010 |
sing | +0.010 |
symp | +0.010 |
badly | +0.010 |
heel | +0.010 |
Neg logit contributions | |
flour | -0.008 |
pots | -0.008 |
tables | -0.008 |
ingredients | -0.007 |
oons | -0.007 |
Token | believe |
Token ID | 1975 |
Feature activation | +1.26e-02 |
Pos logit contributions | |
symp | +0.012 |
badly | +0.012 |
recovery | +0.012 |
agging | +0.012 |
heel | +0.011 |
Neg logit contributions | |
flour | -0.018 |
tables | -0.018 |
pots | -0.017 |
ingredients | -0.017 |
oons | -0.017 |
Token | determined |
Token ID | 5295 |
Feature activation | -6.18e-02 |
Pos logit contributions | |
flour | +0.003 |
ingredients | +0.002 |
tables | +0.002 |
pots | +0.002 |
chairs | +0.002 |
Neg logit contributions | |
sing | -0.002 |
spine | -0.002 |
recovery | -0.002 |
symp | -0.002 |
former | -0.002 |
Token | to |
Token ID | 284 |
Feature activation | -4.36e-02 |
Pos logit contributions | |
sing | +0.131 |
recovery | +0.126 |
heel | +0.125 |
boo | +0.124 |
symp | +0.124 |
Neg logit contributions | |
flour | -0.085 |
ingredients | -0.076 |
tables | -0.075 |
oons | -0.074 |
pots | -0.073 |
Token | keep |
Token ID | 1394 |
Feature activation | -2.44e-02 |
Pos logit contributions | |
sing | +0.072 |
symp | +0.070 |
spine | +0.067 |
recovery | +0.065 |
badly | +0.065 |
Neg logit contributions | |
flour | -0.077 |
tables | -0.071 |
ingredients | -0.070 |
pots | -0.069 |
oons | -0.068 |
Token | the |
Token ID | 262 |
Feature activation | -3.17e-03 |
Pos logit contributions | |
sing | +0.053 |
heel | +0.052 |
recovery | +0.052 |
badly | +0.051 |
boo | +0.051 |
Neg logit contributions | |
flour | -0.032 |
oons | -0.028 |
ingredients | -0.028 |
tables | -0.028 |
pots | -0.027 |
Token | fierce |
Token ID | 14800 |
Feature activation | -1.28e-01 |
Pos logit contributions | |
sing | +0.005 |
symp | +0.005 |
badly | +0.005 |
recovery | +0.005 |
spine | +0.004 |
Neg logit contributions | |
flour | -0.006 |
ingredients | -0.006 |
pots | -0.006 |
tables | -0.006 |
pans | -0.005 |
Token | fl |
Token ID | 781 |
Feature activation | -2.37e-01 |
Pos logit contributions | |
sing | +0.194 |
symp | +0.188 |
recovery | +0.184 |
former | +0.184 |
badly | +0.183 |
Neg logit contributions | |
flour | -0.244 |
ingredients | -0.226 |
tables | -0.224 |
pots | -0.224 |
oons | -0.216 |
Token | ute |
Token ID | 1133 |
Feature activation | -1.55e-01 |
Pos logit contributions | |
bru | +0.233 |
boo | +0.231 |
symp | +0.224 |
sing | +0.214 |
appetite | +0.211 |
Neg logit contributions | |
ups | -0.526 |
oons | -0.525 |
flour | -0.523 |
ingredients | -0.496 |
pots | -0.492 |
Token | for |
Token ID | 329 |
Feature activation | -1.03e-02 |
Pos logit contributions | |
sing | +0.327 |
boo | +0.315 |
recovery | +0.315 |
symp | +0.313 |
former | +0.307 |
Neg logit contributions | |
flour | -0.202 |
ingredients | -0.185 |
pots | -0.178 |
tables | -0.176 |
oons | -0.166 |
Token | herself |
Token ID | 5223 |
Feature activation | -2.49e-02 |
Pos logit contributions | |
sing | +0.023 |
boo | +0.022 |
heel | +0.021 |
recovery | +0.021 |
symp | +0.021 |
Neg logit contributions | |
flour | -0.013 |
tables | -0.012 |
ingredients | -0.012 |
pots | -0.011 |
oons | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | -2.34e-02 |
Pos logit contributions | |
sing | +0.058 |
recovery | +0.056 |
boo | +0.055 |
heel | +0.055 |
symp | +0.055 |
Neg logit contributions | |
flour | -0.028 |
ingredients | -0.025 |
tables | -0.025 |
oons | -0.024 |
pots | -0.024 |
Token | After |
Token ID | 2293 |
Feature activation | +2.60e-02 |
Pos logit contributions | |
sing | +0.041 |
recovery | +0.039 |
heel | +0.039 |
badly | +0.038 |
boo | +0.038 |
Neg logit contributions | |
flour | -0.041 |
ingredients | -0.038 |
tables | -0.037 |
oons | -0.037 |
pots | -0.036 |
Token | was |
Token ID | 373 |
Feature activation | -1.15e-02 |
Pos logit contributions | |
flour | +0.039 |
ingredients | +0.035 |
tables | +0.035 |
oons | +0.034 |
pots | +0.033 |
Neg logit contributions | |
sing | -0.077 |
recovery | -0.074 |
spine | -0.072 |
badly | -0.072 |
heel | -0.072 |
Token | able |
Token ID | 1498 |
Feature activation | +3.12e-02 |
Pos logit contributions | |
sing | +0.019 |
recovery | +0.019 |
spine | +0.019 |
former | +0.018 |
heel | +0.018 |
Neg logit contributions | |
flour | -0.020 |
ingredients | -0.018 |
tables | -0.018 |
pots | -0.017 |
oons | -0.017 |
Token | to |
Token ID | 284 |
Feature activation | -5.07e-02 |
Pos logit contributions | |
flour | +0.039 |
ingredients | +0.036 |
tables | +0.034 |
pots | +0.034 |
oons | +0.033 |
Neg logit contributions | |
sing | -0.070 |
recovery | -0.067 |
heel | -0.066 |
symp | -0.066 |
boo | -0.066 |
Token | keep |
Token ID | 1394 |
Feature activation | -2.54e-02 |
Pos logit contributions | |
sing | +0.093 |
spine | +0.089 |
symp | +0.088 |
OO | +0.083 |
badly | +0.082 |
Neg logit contributions | |
flour | -0.079 |
tables | -0.072 |
pots | -0.070 |
ingredients | -0.069 |
brushes | -0.067 |
Token | her |
Token ID | 607 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
sing | +0.056 |
heel | +0.055 |
recovery | +0.055 |
badly | +0.055 |
boo | +0.055 |
Neg logit contributions | |
flour | -0.032 |
oons | -0.028 |
ingredients | -0.028 |
tables | -0.028 |
pots | -0.027 |
Token | fl |
Token ID | 781 |
Feature activation | -2.37e-01 |
Pos logit contributions | |
sing | +0.206 |
recovery | +0.195 |
symp | +0.193 |
badly | +0.193 |
former | +0.193 |
Neg logit contributions | |
flour | -0.263 |
ingredients | -0.249 |
pots | -0.244 |
tables | -0.236 |
pans | -0.234 |
Token | ute |
Token ID | 1133 |
Feature activation | -1.56e-01 |
Pos logit contributions | |
bru | +0.244 |
boo | +0.242 |
sing | +0.229 |
appetite | +0.225 |
recovery | +0.221 |
Neg logit contributions | |
ups | -0.510 |
flour | -0.496 |
oons | -0.495 |
ingredients | -0.471 |
pans | -0.462 |
Token | . |
Token ID | 13 |
Feature activation | -2.15e-02 |
Pos logit contributions | |
sing | +0.307 |
boo | +0.291 |
recovery | +0.289 |
symp | +0.285 |
bru | +0.283 |
Neg logit contributions | |
flour | -0.228 |
ingredients | -0.212 |
pots | -0.205 |
tables | -0.202 |
pans | -0.193 |
Token | She |
Token ID | 1375 |
Feature activation | -5.38e-02 |
Pos logit contributions | |
symp | +0.033 |
sing | +0.032 |
recovery | +0.032 |
heel | +0.031 |
boo | +0.030 |
Neg logit contributions | |
flour | -0.042 |
ingredients | -0.038 |
pots | -0.038 |
oons | -0.038 |
ries | -0.038 |
Token | was |
Token ID | 373 |
Feature activation | -1.82e-03 |
Pos logit contributions | |
sing | +0.123 |
recovery | +0.119 |
spine | +0.117 |
heel | +0.115 |
badly | +0.115 |
Neg logit contributions | |
flour | -0.065 |
ingredients | -0.058 |
tables | -0.057 |
oons | -0.056 |
pots | -0.055 |
Token | so |
Token ID | 523 |
Feature activation | +1.04e-02 |
Pos logit contributions | |
sing | +0.003 |
spine | +0.003 |
recovery | +0.003 |
symp | +0.003 |
former | +0.003 |
Neg logit contributions | |
flour | -0.003 |
ingredients | -0.003 |
tables | -0.003 |
pots | -0.002 |
ries | -0.002 |
Token | going |
Token ID | 1016 |
Feature activation | -8.70e-02 |
Pos logit contributions | |
flour | +0.003 |
ingredients | +0.003 |
tables | +0.003 |
pots | +0.002 |
oons | +0.002 |
Neg logit contributions | |
sing | -0.003 |
recovery | -0.003 |
heel | -0.003 |
symp | -0.003 |
badly | -0.003 |
Token | to |
Token ID | 284 |
Feature activation | -5.12e-02 |
Pos logit contributions | |
sing | +0.189 |
recovery | +0.184 |
heel | +0.181 |
boo | +0.181 |
symp | +0.180 |
Neg logit contributions | |
flour | -0.114 |
ingredients | -0.103 |
tables | -0.102 |
pots | -0.099 |
oons | -0.098 |
Token | enjoy |
Token ID | 2883 |
Feature activation | -2.51e-02 |
Pos logit contributions | |
sing | +0.106 |
symp | +0.103 |
badly | +0.102 |
uer | +0.101 |
spine | +0.101 |
Neg logit contributions | |
flour | -0.074 |
tables | -0.071 |
pots | -0.064 |
chairs | -0.063 |
ingredients | -0.063 |
Token | her |
Token ID | 607 |
Feature activation | -1.26e-01 |
Pos logit contributions | |
sing | +0.061 |
recovery | +0.059 |
heel | +0.058 |
boo | +0.058 |
badly | +0.058 |
Neg logit contributions | |
flour | -0.027 |
tables | -0.023 |
ingredients | -0.023 |
pots | -0.022 |
oons | -0.022 |
Token | first |
Token ID | 717 |
Feature activation | -7.14e-02 |
Pos logit contributions | |
sing | +0.175 |
aring | +0.162 |
symp | +0.158 |
ur | +0.154 |
recovery | +0.153 |
Neg logit contributions | |
ingredients | -0.257 |
flour | -0.257 |
tables | -0.250 |
chairs | -0.249 |
plates | -0.245 |
Token | gymn |
Token ID | 38581 |
Feature activation | -2.36e-01 |
Pos logit contributions | |
sing | +0.125 |
symp | +0.118 |
former | +0.116 |
tweet | +0.115 |
utt | +0.114 |
Neg logit contributions | |
flour | -0.113 |
ingredients | -0.112 |
tables | -0.110 |
chairs | -0.106 |
ones | -0.103 |
Token | astics |
Token ID | 24232 |
Feature activation | -1.26e-01 |
Pos logit contributions | |
tweet | +0.321 |
boo | +0.314 |
sing | +0.311 |
bru | +0.302 |
spine | +0.295 |
Neg logit contributions | |
flour | -0.466 |
ups | -0.461 |
ingredients | -0.452 |
tables | -0.437 |
oons | -0.428 |
Token | lesson |
Token ID | 11483 |
Feature activation | -4.16e-02 |
Pos logit contributions | |
sing | +0.213 |
recovery | +0.204 |
heel | +0.203 |
badly | +0.201 |
boo | +0.201 |
Neg logit contributions | |
flour | -0.224 |
ingredients | -0.207 |
tables | -0.205 |
oons | -0.203 |
pots | -0.201 |
Token | . |
Token ID | 13 |
Feature activation | -1.28e-02 |
Pos logit contributions | |
sing | +0.080 |
badly | +0.080 |
recovery | +0.079 |
heel | +0.078 |
symp | +0.077 |
Neg logit contributions | |
flour | -0.064 |
ingredients | -0.057 |
oons | -0.056 |
pots | -0.056 |
tables | -0.056 |
Token | |
Token ID | 220 |
Feature activation | +2.98e-02 |
Pos logit contributions | |
sing | +0.018 |
symp | +0.018 |
heel | +0.018 |
recovery | +0.017 |
badly | +0.017 |
Neg logit contributions | |
flour | -0.026 |
oons | -0.024 |
pans | -0.024 |
pots | -0.024 |
ingredients | -0.023 |
Token |
|
Token ID | 198 |
Feature activation | -3.14e-02 |
Pos logit contributions | |
flour | +0.055 |
pots | +0.051 |
ingredients | +0.051 |
tables | +0.051 |
pans | +0.050 |
Neg logit contributions | |
badly | -0.046 |
symp | -0.045 |
sing | -0.045 |
recovery | -0.045 |
heel | -0.044 |
Token | hard |
Token ID | 1327 |
Feature activation | -3.36e-02 |
Pos logit contributions | |
sing | +0.111 |
spine | +0.110 |
badly | +0.108 |
recovery | +0.106 |
tweet | +0.106 |
Neg logit contributions | |
flour | -0.114 |
ingredients | -0.105 |
pots | -0.102 |
tables | -0.100 |
plates | -0.097 |
Token | work |
Token ID | 670 |
Feature activation | +1.42e-02 |
Pos logit contributions | |
sing | +0.059 |
recovery | +0.056 |
badly | +0.056 |
heel | +0.056 |
symp | +0.056 |
Neg logit contributions | |
flour | -0.058 |
ingredients | -0.054 |
oons | -0.053 |
tables | -0.053 |
pots | -0.053 |
Token | had |
Token ID | 550 |
Feature activation | +1.16e-02 |
Pos logit contributions | |
flour | +0.021 |
tables | +0.019 |
pots | +0.018 |
oons | +0.018 |
ingredients | +0.018 |
Neg logit contributions | |
badly | -0.028 |
sing | -0.028 |
recovery | -0.028 |
promise | -0.027 |
heel | -0.027 |
Token | made |
Token ID | 925 |
Feature activation | +3.89e-02 |
Pos logit contributions | |
flour | +0.018 |
oons | +0.018 |
tables | +0.017 |
pans | +0.017 |
ries | +0.016 |
Neg logit contributions | |
badly | -0.020 |
sing | -0.020 |
recovery | -0.020 |
promise | -0.020 |
heel | -0.020 |
Token | the |
Token ID | 262 |
Feature activation | -6.95e-03 |
Pos logit contributions | |
flour | +0.051 |
oons | +0.048 |
tables | +0.047 |
pans | +0.046 |
cakes | +0.045 |
Neg logit contributions | |
sing | -0.079 |
recovery | -0.079 |
badly | -0.079 |
boo | -0.077 |
heel | -0.076 |
Token | fl |
Token ID | 781 |
Feature activation | -2.36e-01 |
Pos logit contributions | |
sing | +0.013 |
badly | +0.013 |
recovery | +0.012 |
symp | +0.012 |
heel | +0.012 |
Neg logit contributions | |
flour | -0.012 |
ingredients | -0.011 |
pots | -0.010 |
tables | -0.010 |
oons | -0.010 |
Token | ute |
Token ID | 1133 |
Feature activation | -1.55e-01 |
Pos logit contributions | |
boo | +0.179 |
bru | +0.175 |
symp | +0.175 |
sing | +0.173 |
recovery | +0.165 |
Neg logit contributions | |
flour | -0.605 |
oons | -0.591 |
ingredients | -0.579 |
ups | -0.577 |
pots | -0.564 |
Token | beautiful |
Token ID | 4950 |
Feature activation | -5.38e-02 |
Pos logit contributions | |
sing | +0.302 |
recovery | +0.292 |
heel | +0.291 |
boo | +0.288 |
symp | +0.288 |
Neg logit contributions | |
flour | -0.235 |
ingredients | -0.214 |
tables | -0.210 |
oons | -0.208 |
pots | -0.207 |
Token | . |
Token ID | 13 |
Feature activation | -3.47e-02 |
Pos logit contributions | |
sing | +0.101 |
badly | +0.100 |
heel | +0.099 |
recovery | +0.098 |
hero | +0.096 |
Neg logit contributions | |
flour | -0.083 |
oons | -0.076 |
tables | -0.074 |
ingredients | -0.073 |
pots | -0.072 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +4.86e-02 |
Pos logit contributions | |
sing | +0.063 |
recovery | +0.061 |
heel | +0.060 |
badly | +0.059 |
symp | +0.059 |
Neg logit contributions | |
flour | -0.058 |
ingredients | -0.053 |
tables | -0.052 |
oons | -0.052 |
pots | -0.052 |
Token | Once |
Token ID | 7454 |
Feature activation | +2.44e-02 |
Pos logit contributions | |
flour | +0.091 |
pans | +0.084 |
ingredients | +0.084 |
pots | +0.082 |
brushes | +0.081 |
Neg logit contributions | |
heel | -0.069 |
symp | -0.069 |
recovery | -0.069 |
hero | -0.068 |
badly | -0.067 |
Token | music |
Token ID | 2647 |
Feature activation | -5.56e-02 |
Pos logit contributions | |
flour | +0.037 |
ingredients | +0.035 |
pots | +0.034 |
colors | +0.033 |
tables | +0.032 |
Neg logit contributions | |
sing | -0.042 |
badly | -0.041 |
symp | -0.041 |
uer | -0.040 |
ur | -0.040 |
Token | she |
Token ID | 673 |
Feature activation | -6.14e-02 |
Pos logit contributions | |
sing | +0.128 |
symp | +0.122 |
boo | +0.121 |
former | +0.121 |
recovery | +0.121 |
Neg logit contributions | |
flour | -0.064 |
ingredients | -0.058 |
tables | -0.056 |
pots | -0.054 |
oons | -0.053 |
Token | played |
Token ID | 2826 |
Feature activation | -6.44e-02 |
Pos logit contributions | |
sing | +0.136 |
recovery | +0.131 |
symp | +0.129 |
badly | +0.129 |
heel | +0.128 |
Neg logit contributions | |
flour | -0.076 |
ingredients | -0.069 |
tables | -0.067 |
pots | -0.066 |
oons | -0.066 |
Token | on |
Token ID | 319 |
Feature activation | -7.97e-02 |
Pos logit contributions | |
sing | +0.159 |
spine | +0.152 |
boo | +0.149 |
promise | +0.148 |
former | +0.148 |
Neg logit contributions | |
flour | -0.063 |
tables | -0.059 |
pots | -0.058 |
ingredients | -0.057 |
oons | -0.053 |
Token | her |
Token ID | 607 |
Feature activation | -1.41e-01 |
Pos logit contributions | |
sing | +0.193 |
recovery | +0.187 |
heel | +0.187 |
badly | +0.185 |
boo | +0.185 |
Neg logit contributions | |
flour | -0.084 |
ingredients | -0.072 |
tables | -0.072 |
oons | -0.071 |
pots | -0.069 |
Token | fl |
Token ID | 781 |
Feature activation | -2.34e-01 |
Pos logit contributions | |
utt | +0.241 |
OO | +0.239 |
ur | +0.238 |
berth | +0.238 |
tweet | +0.237 |
Neg logit contributions | |
chairs | -0.215 |
tables | -0.214 |
flour | -0.207 |
pots | -0.203 |
plates | -0.199 |
Token | ute |
Token ID | 1133 |
Feature activation | -1.60e-01 |
Pos logit contributions | |
boo | +0.287 |
bru | +0.284 |
tweet | +0.269 |
symp | +0.267 |
sing | +0.263 |
Neg logit contributions | |
oons | -0.485 |
ups | -0.478 |
flour | -0.429 |
ingredients | -0.423 |
cakes | -0.420 |
Token | that |
Token ID | 326 |
Feature activation | +1.42e-02 |
Pos logit contributions | |
sing | +0.407 |
boo | +0.397 |
symp | +0.391 |
recovery | +0.381 |
heel | +0.381 |
Neg logit contributions | |
flour | -0.128 |
ingredients | -0.118 |
tables | -0.115 |
pots | -0.110 |
oons | -0.104 |
Token | night |
Token ID | 1755 |
Feature activation | -5.68e-02 |
Pos logit contributions | |
flour | +0.021 |
ingredients | +0.019 |
tables | +0.019 |
oons | +0.019 |
pots | +0.019 |
Neg logit contributions | |
sing | -0.028 |
recovery | -0.027 |
heel | -0.027 |
badly | -0.027 |
boo | -0.027 |
Token | . |
Token ID | 13 |
Feature activation | -3.09e-02 |
Pos logit contributions | |
sing | +0.128 |
recovery | +0.124 |
heel | +0.124 |
boo | +0.123 |
badly | +0.123 |
Neg logit contributions | |
flour | -0.068 |
ingredients | -0.060 |
tables | -0.060 |
oons | -0.059 |
pots | -0.058 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +4.64e-02 |
Pos logit contributions | |
sing | +0.067 |
recovery | +0.064 |
heel | +0.064 |
badly | +0.063 |
boo | +0.063 |
Neg logit contributions | |
flour | -0.040 |
tables | -0.036 |
ingredients | -0.036 |
oons | -0.035 |
pots | -0.034 |
Token | a |
Token ID | 257 |
Feature activation | +3.52e-03 |
Pos logit contributions | |
badly | +0.010 |
sing | +0.010 |
heel | +0.010 |
recovery | +0.010 |
boo | +0.010 |
Neg logit contributions | |
flour | -0.009 |
oons | -0.009 |
tables | -0.008 |
ingredients | -0.008 |
pots | -0.008 |
Token | bad |
Token ID | 2089 |
Feature activation | -1.35e-01 |
Pos logit contributions | |
flour | +0.006 |
ingredients | +0.005 |
pots | +0.005 |
tables | +0.004 |
plates | +0.004 |
Neg logit contributions | |
amb | -0.007 |
symp | -0.007 |
utt | -0.007 |
agging | -0.007 |
sing | -0.007 |
Token | case |
Token ID | 1339 |
Feature activation | -1.16e-01 |
Pos logit contributions | |
symp | +0.200 |
utt | +0.192 |
ighed | +0.192 |
amb | +0.189 |
shadow | +0.187 |
Neg logit contributions | |
flour | -0.276 |
ingredients | -0.259 |
pots | -0.252 |
tables | -0.236 |
oons | -0.232 |
Token | of |
Token ID | 286 |
Feature activation | -7.73e-03 |
Pos logit contributions | |
sing | +0.249 |
recovery | +0.242 |
badly | +0.241 |
heel | +0.240 |
boo | +0.237 |
Neg logit contributions | |
flour | -0.153 |
ingredients | -0.136 |
tables | -0.134 |
oons | -0.134 |
pots | -0.131 |
Token | the |
Token ID | 262 |
Feature activation | +6.94e-03 |
Pos logit contributions | |
sing | +0.016 |
symp | +0.016 |
spine | +0.015 |
uer | +0.015 |
berth | +0.015 |
Neg logit contributions | |
flour | -0.012 |
ingredients | -0.010 |
tables | -0.010 |
pots | -0.010 |
oons | -0.009 |
Token | flu |
Token ID | 6562 |
Feature activation | -2.34e-01 |
Pos logit contributions | |
flour | +0.011 |
ingredients | +0.010 |
tables | +0.010 |
pots | +0.010 |
chairs | +0.009 |
Neg logit contributions | |
sing | -0.013 |
symp | -0.013 |
ree | -0.013 |
berth | -0.013 |
amb | -0.013 |
Token | . |
Token ID | 13 |
Feature activation | -2.83e-02 |
Pos logit contributions | |
sing | +0.370 |
recovery | +0.366 |
badly | +0.362 |
heel | +0.353 |
promise | +0.349 |
Neg logit contributions | |
flour | -0.441 |
tables | -0.406 |
ingredients | -0.406 |
pots | -0.403 |
pans | -0.391 |
Token | Lily |
Token ID | 20037 |
Feature activation | +5.52e-02 |
Pos logit contributions | |
sing | +0.053 |
boo | +0.053 |
recovery | +0.053 |
badly | +0.052 |
symp | +0.052 |
Neg logit contributions | |
flour | -0.043 |
ingredients | -0.039 |
oons | -0.039 |
ries | -0.039 |
pots | -0.038 |
Token | had |
Token ID | 550 |
Feature activation | +2.46e-02 |
Pos logit contributions | |
flour | +0.068 |
ingredients | +0.061 |
tables | +0.060 |
oons | +0.059 |
pots | +0.058 |
Neg logit contributions | |
sing | -0.123 |
recovery | -0.120 |
heel | -0.118 |
symp | -0.116 |
spine | -0.116 |
Token | to |
Token ID | 284 |
Feature activation | -5.86e-02 |
Pos logit contributions | |
flour | +0.038 |
ingredients | +0.035 |
tables | +0.034 |
oons | +0.034 |
pots | +0.034 |
Neg logit contributions | |
sing | -0.047 |
recovery | -0.046 |
heel | -0.045 |
badly | -0.045 |
boo | -0.045 |
Token | stay |
Token ID | 2652 |
Feature activation | -8.12e-02 |
Pos logit contributions | |
sing | +0.105 |
spine | +0.098 |
uer | +0.097 |
symp | +0.096 |
OO | +0.092 |
Neg logit contributions | |
flour | -0.103 |
tables | -0.091 |
pots | -0.089 |
ingredients | -0.087 |
plates | -0.085 |
Token | and |
Token ID | 290 |
Feature activation | -4.21e-02 |
Pos logit contributions | |
sing | +0.148 |
recovery | +0.144 |
heel | +0.142 |
boo | +0.142 |
symp | +0.141 |
Neg logit contributions | |
flour | -0.080 |
ingredients | -0.072 |
tables | -0.070 |
oons | -0.069 |
pots | -0.069 |
Token | he |
Token ID | 339 |
Feature activation | -3.24e-02 |
Pos logit contributions | |
sing | +0.075 |
recovery | +0.072 |
heel | +0.072 |
badly | +0.071 |
boo | +0.071 |
Neg logit contributions | |
flour | -0.071 |
ingredients | -0.065 |
tables | -0.065 |
oons | -0.064 |
pots | -0.064 |
Token | kept |
Token ID | 4030 |
Feature activation | -1.76e-03 |
Pos logit contributions | |
sing | +0.071 |
recovery | +0.069 |
symp | +0.068 |
spine | +0.067 |
badly | +0.067 |
Neg logit contributions | |
flour | -0.040 |
tables | -0.036 |
ingredients | -0.036 |
oons | -0.035 |
pots | -0.035 |
Token | playing |
Token ID | 2712 |
Feature activation | -5.77e-02 |
Pos logit contributions | |
spine | +0.004 |
sing | +0.003 |
recovery | +0.003 |
former | +0.003 |
uer | +0.003 |
Neg logit contributions | |
flour | -0.002 |
ingredients | -0.002 |
pots | -0.002 |
the | -0.002 |
tables | -0.002 |
Token | his |
Token ID | 465 |
Feature activation | -1.47e-01 |
Pos logit contributions | |
sing | +0.145 |
spine | +0.139 |
former | +0.136 |
boo | +0.135 |
promise | +0.133 |
Neg logit contributions | |
tables | -0.058 |
flour | -0.054 |
pots | -0.050 |
chairs | -0.050 |
ingredients | -0.047 |
Token | fl |
Token ID | 781 |
Feature activation | -2.33e-01 |
Pos logit contributions | |
symp | +0.273 |
badly | +0.271 |
berth | +0.263 |
aring | +0.261 |
ur | +0.261 |
Neg logit contributions | |
ingredients | -0.219 |
flour | -0.219 |
chairs | -0.218 |
colors | -0.215 |
flowers | -0.213 |
Token | ute |
Token ID | 1133 |
Feature activation | -1.61e-01 |
Pos logit contributions | |
boo | +0.267 |
bru | +0.257 |
symp | +0.241 |
weep | +0.220 |
recovery | +0.219 |
Neg logit contributions | |
oons | -0.507 |
ups | -0.497 |
flour | -0.473 |
ingredients | -0.460 |
ons | -0.448 |
Token | as |
Token ID | 355 |
Feature activation | +7.22e-02 |
Pos logit contributions | |
sing | +0.406 |
boo | +0.406 |
symp | +0.397 |
bru | +0.396 |
recovery | +0.392 |
Neg logit contributions | |
flour | -0.120 |
tables | -0.111 |
ingredients | -0.110 |
pots | -0.108 |
chairs | -0.100 |
Token | they |
Token ID | 484 |
Feature activation | +2.59e-02 |
Pos logit contributions | |
flour | +0.134 |
ingredients | +0.126 |
tables | +0.122 |
oons | +0.122 |
pots | +0.121 |
Neg logit contributions | |
sing | -0.115 |
badly | -0.114 |
recovery | -0.111 |
heel | -0.109 |
hero | -0.109 |
Token | danced |
Token ID | 39480 |
Feature activation | +2.12e-02 |
Pos logit contributions | |
flour | +0.037 |
tables | +0.035 |
pots | +0.034 |
oons | +0.034 |
ingredients | +0.033 |
Neg logit contributions | |
sing | -0.052 |
recovery | -0.050 |
badly | -0.049 |
boo | -0.049 |
spine | -0.048 |
Token | . |
Token ID | 13 |
Feature activation | -2.65e-02 |
Pos logit contributions | |
flour | +0.021 |
tables | +0.019 |
ingredients | +0.018 |
pots | +0.018 |
oons | +0.018 |
Neg logit contributions | |
sing | -0.052 |
boo | -0.051 |
recovery | -0.050 |
heel | -0.050 |
symp | -0.049 |
Token | who |
Token ID | 508 |
Feature activation | +2.54e-02 |
Pos logit contributions | |
flour | +0.022 |
pots | +0.019 |
tables | +0.019 |
ingredients | +0.019 |
oons | +0.018 |
Neg logit contributions | |
sing | -0.040 |
recovery | -0.038 |
former | -0.038 |
heel | -0.038 |
spine | -0.038 |
Token | had |
Token ID | 550 |
Feature activation | +3.01e-02 |
Pos logit contributions | |
flour | +0.033 |
ingredients | +0.029 |
tables | +0.029 |
oons | +0.028 |
pots | +0.028 |
Neg logit contributions | |
sing | -0.055 |
badly | -0.054 |
heel | -0.054 |
boo | -0.054 |
recovery | -0.054 |
Token | never |
Token ID | 1239 |
Feature activation | -5.59e-02 |
Pos logit contributions | |
flour | +0.038 |
ingredients | +0.034 |
oons | +0.034 |
tables | +0.034 |
pots | +0.033 |
Neg logit contributions | |
sing | -0.066 |
heel | -0.065 |
recovery | -0.065 |
badly | -0.064 |
boo | -0.064 |
Token | heard |
Token ID | 2982 |
Feature activation | +6.73e-02 |
Pos logit contributions | |
sing | +0.099 |
recovery | +0.096 |
heel | +0.096 |
boo | +0.094 |
badly | +0.094 |
Neg logit contributions | |
flour | -0.095 |
oons | -0.087 |
ingredients | -0.086 |
tables | -0.086 |
pots | -0.085 |
Token | a |
Token ID | 257 |
Feature activation | +2.23e-02 |
Pos logit contributions | |
flour | +0.099 |
ingredients | +0.090 |
oons | +0.089 |
tables | +0.087 |
pots | +0.087 |
Neg logit contributions | |
sing | -0.132 |
recovery | -0.131 |
heel | -0.130 |
badly | -0.129 |
boo | -0.129 |
Token | fl |
Token ID | 781 |
Feature activation | -2.32e-01 |
Pos logit contributions | |
flour | +0.037 |
tables | +0.034 |
pots | +0.033 |
chairs | +0.032 |
pans | +0.032 |
Neg logit contributions | |
ighed | -0.040 |
ur | -0.040 |
amb | -0.039 |
berth | -0.039 |
sing | -0.038 |
Token | ute |
Token ID | 1133 |
Feature activation | -1.56e-01 |
Pos logit contributions | |
symp | +0.222 |
bru | +0.221 |
boo | +0.220 |
former | +0.219 |
sing | +0.216 |
Neg logit contributions | |
flour | -0.500 |
oons | -0.494 |
ups | -0.485 |
pots | -0.473 |
ingredients | -0.465 |
Token | before |
Token ID | 878 |
Feature activation | +9.76e-02 |
Pos logit contributions | |
ighed | +0.386 |
former | +0.384 |
oof | +0.380 |
sing | +0.380 |
symp | +0.375 |
Neg logit contributions | |
flour | -0.120 |
tables | -0.114 |
pots | -0.112 |
cans | -0.096 |
ingredients | -0.095 |
Token | . |
Token ID | 13 |
Feature activation | -8.66e-03 |
Pos logit contributions | |
flour | +0.086 |
tables | +0.079 |
pots | +0.073 |
ingredients | +0.070 |
oons | +0.067 |
Neg logit contributions | |
sing | -0.249 |
promise | -0.244 |
former | -0.241 |
uer | -0.240 |
recovery | -0.240 |
Token | He |
Token ID | 679 |
Feature activation | -5.70e-02 |
Pos logit contributions | |
heel | +0.014 |
sing | +0.014 |
boo | +0.014 |
recovery | +0.014 |
symp | +0.014 |
Neg logit contributions | |
flour | -0.016 |
ingredients | -0.015 |
oons | -0.014 |
pots | -0.014 |
ries | -0.014 |
Token | asked |
Token ID | 1965 |
Feature activation | -5.67e-03 |
Pos logit contributions | |
sing | +0.134 |
recovery | +0.131 |
spine | +0.129 |
heel | +0.127 |
symp | +0.126 |
Neg logit contributions | |
flour | -0.063 |
ingredients | -0.055 |
tables | -0.055 |
oons | -0.054 |
pots | -0.053 |
Token | bending |
Token ID | 30808 |
Feature activation | +3.67e-02 |
Pos logit contributions | |
sing | +0.068 |
heel | +0.066 |
badly | +0.066 |
boo | +0.065 |
recovery | +0.064 |
Neg logit contributions | |
flour | -0.060 |
oons | -0.057 |
ingredients | -0.054 |
tables | -0.054 |
ries | -0.052 |
Token | her |
Token ID | 607 |
Feature activation | -8.63e-02 |
Pos logit contributions | |
flour | +0.056 |
oons | +0.051 |
ingredients | +0.050 |
cakes | +0.049 |
tables | +0.048 |
Neg logit contributions | |
badly | -0.069 |
heel | -0.068 |
sing | -0.067 |
symp | -0.066 |
boo | -0.065 |
Token | body |
Token ID | 1767 |
Feature activation | -1.43e-02 |
Pos logit contributions | |
sing | +0.154 |
recovery | +0.150 |
heel | +0.147 |
badly | +0.147 |
symp | +0.146 |
Neg logit contributions | |
flour | -0.147 |
ingredients | -0.135 |
tables | -0.133 |
oons | -0.132 |
pots | -0.131 |
Token | like |
Token ID | 588 |
Feature activation | +2.06e-02 |
Pos logit contributions | |
sing | +0.030 |
badly | +0.029 |
heel | +0.029 |
recovery | +0.029 |
boo | +0.029 |
Neg logit contributions | |
flour | -0.019 |
ingredients | -0.017 |
tables | -0.017 |
oons | -0.017 |
pots | -0.017 |
Token | a |
Token ID | 257 |
Feature activation | +2.56e-02 |
Pos logit contributions | |
flour | +0.026 |
oons | +0.024 |
ingredients | +0.023 |
tables | +0.023 |
pots | +0.022 |
Neg logit contributions | |
sing | -0.044 |
recovery | -0.043 |
heel | -0.043 |
badly | -0.043 |
boo | -0.043 |
Token | gymn |
Token ID | 38581 |
Feature activation | -2.32e-01 |
Pos logit contributions | |
flour | +0.044 |
ingredients | +0.040 |
tables | +0.040 |
pots | +0.039 |
chairs | +0.037 |
Neg logit contributions | |
sing | -0.046 |
symp | -0.045 |
recovery | -0.043 |
spine | -0.042 |
berth | -0.042 |
Token | ast |
Token ID | 459 |
Feature activation | -3.56e-02 |
Pos logit contributions | |
tweet | +0.268 |
recovery | +0.262 |
sing | +0.257 |
hero | +0.256 |
badly | +0.255 |
Neg logit contributions | |
flour | -0.509 |
ingredients | -0.482 |
tables | -0.479 |
pots | -0.473 |
oons | -0.471 |
Token | . |
Token ID | 13 |
Feature activation | -1.19e-02 |
Pos logit contributions | |
sing | +0.068 |
recovery | +0.067 |
heel | +0.067 |
boo | +0.067 |
badly | +0.066 |
Neg logit contributions | |
flour | -0.053 |
ingredients | -0.047 |
tables | -0.047 |
pots | -0.047 |
oons | -0.046 |
Token | Lily |
Token ID | 20037 |
Feature activation | +6.51e-02 |
Pos logit contributions | |
sing | +0.025 |
recovery | +0.024 |
heel | +0.024 |
badly | +0.024 |
boo | +0.024 |
Neg logit contributions | |
flour | -0.016 |
tables | -0.015 |
ingredients | -0.015 |
oons | -0.014 |
pots | -0.014 |
Token | tried |
Token ID | 3088 |
Feature activation | +1.63e-02 |
Pos logit contributions | |
flour | +0.090 |
ingredients | +0.081 |
tables | +0.080 |
oons | +0.079 |
pots | +0.078 |
Neg logit contributions | |
sing | -0.136 |
recovery | -0.132 |
heel | -0.131 |
badly | -0.130 |
boo | -0.129 |
Token | her |
Token ID | 607 |
Feature activation | -9.12e-02 |
Pos logit contributions | |
flour | +0.022 |
ingredients | +0.020 |
tables | +0.020 |
pots | +0.019 |
oons | +0.019 |
Neg logit contributions | |
sing | -0.035 |
recovery | -0.034 |
heel | -0.033 |
symp | -0.033 |
badly | -0.033 |
Token | and |
Token ID | 290 |
Feature activation | -1.60e-02 |
Pos logit contributions | |
sing | +0.228 |
recovery | +0.220 |
heel | +0.219 |
badly | +0.219 |
boo | +0.217 |
Neg logit contributions | |
flour | -0.183 |
ingredients | -0.166 |
tables | -0.164 |
oons | -0.163 |
pots | -0.161 |
Token | I |
Token ID | 314 |
Feature activation | +1.68e-02 |
Pos logit contributions | |
sing | +0.029 |
symp | +0.028 |
recovery | +0.028 |
spine | +0.027 |
badly | +0.027 |
Neg logit contributions | |
flour | -0.027 |
tables | -0.025 |
ingredients | -0.024 |
pots | -0.024 |
oons | -0.023 |
Token | can |
Token ID | 460 |
Feature activation | -2.29e-03 |
Pos logit contributions | |
flour | +0.018 |
ingredients | +0.017 |
oons | +0.016 |
pots | +0.016 |
brushes | +0.015 |
Neg logit contributions | |
sing | -0.040 |
symp | -0.038 |
spine | -0.038 |
recovery | -0.038 |
agging | -0.037 |
Token | have |
Token ID | 423 |
Feature activation | -9.42e-03 |
Pos logit contributions | |
sing | +0.004 |
symp | +0.004 |
spine | +0.004 |
recovery | +0.004 |
former | +0.003 |
Neg logit contributions | |
flour | -0.004 |
tables | -0.004 |
oons | -0.004 |
pots | -0.004 |
ingredients | -0.004 |
Token | your |
Token ID | 534 |
Feature activation | -1.50e-01 |
Pos logit contributions | |
sing | +0.019 |
recovery | +0.019 |
heel | +0.019 |
badly | +0.019 |
boo | +0.019 |
Neg logit contributions | |
flour | -0.013 |
ingredients | -0.012 |
oons | -0.012 |
tables | -0.012 |
pots | -0.012 |
Token | sc |
Token ID | 629 |
Feature activation | -2.32e-01 |
Pos logit contributions | |
symp | +0.265 |
sing | +0.263 |
recovery | +0.250 |
ree | +0.242 |
spine | +0.241 |
Neg logit contributions | |
flour | -0.256 |
ingredients | -0.242 |
pots | -0.229 |
tables | -0.225 |
cans | -0.220 |
Token | ooter |
Token ID | 25141 |
Feature activation | -1.16e-01 |
Pos logit contributions | |
symp | +0.260 |
sing | +0.259 |
recovery | +0.254 |
boo | +0.229 |
bru | +0.223 |
Neg logit contributions | |
ups | -0.514 |
flour | -0.501 |
cakes | -0.493 |
brushes | -0.493 |
ingredients | -0.489 |
Token | . |
Token ID | 13 |
Feature activation | -4.25e-02 |
Pos logit contributions | |
sing | +0.240 |
recovery | +0.233 |
heel | +0.231 |
badly | +0.230 |
boo | +0.229 |
Neg logit contributions | |
flour | -0.161 |
ingredients | -0.144 |
tables | -0.143 |
oons | -0.142 |
pots | -0.140 |
Token | Then |
Token ID | 3244 |
Feature activation | +4.66e-02 |
Pos logit contributions | |
sing | +0.073 |
recovery | +0.071 |
badly | +0.070 |
symp | +0.070 |
spine | +0.068 |
Neg logit contributions | |
flour | -0.074 |
ingredients | -0.065 |
tables | -0.065 |
pots | -0.065 |
cans | -0.063 |
Token | we |
Token ID | 356 |
Feature activation | +3.24e-02 |
Pos logit contributions | |
oons | +0.089 |
flour | +0.088 |
ingredients | +0.084 |
cakes | +0.083 |
tables | +0.083 |
Neg logit contributions | |
promise | -0.073 |
heel | -0.071 |
badly | -0.069 |
boo | -0.068 |
recovery | -0.068 |
Token | can |
Token ID | 460 |
Feature activation | +6.54e-03 |
Pos logit contributions | |
flour | +0.038 |
ingredients | +0.032 |
oons | +0.032 |
pots | +0.032 |
tables | +0.032 |
Neg logit contributions | |
sing | -0.078 |
recovery | -0.074 |
spine | -0.073 |
symp | -0.072 |
heel | -0.072 |
Token | and |
Token ID | 290 |
Feature activation | -2.80e-02 |
Pos logit contributions | |
sing | +0.204 |
recovery | +0.199 |
heel | +0.196 |
boo | +0.195 |
symp | +0.195 |
Neg logit contributions | |
flour | -0.125 |
ingredients | -0.111 |
tables | -0.110 |
oons | -0.109 |
pots | -0.107 |
Token | Lily |
Token ID | 20037 |
Feature activation | +3.86e-02 |
Pos logit contributions | |
heel | +0.052 |
sing | +0.052 |
badly | +0.051 |
boo | +0.050 |
hero | +0.049 |
Neg logit contributions | |
flour | -0.044 |
ingredients | -0.041 |
oons | -0.041 |
tables | -0.039 |
ries | -0.039 |
Token | had |
Token ID | 550 |
Feature activation | +7.14e-03 |
Pos logit contributions | |
flour | +0.059 |
ingredients | +0.053 |
tables | +0.052 |
oons | +0.052 |
pots | +0.051 |
Neg logit contributions | |
sing | -0.075 |
heel | -0.074 |
badly | -0.073 |
boo | -0.072 |
recovery | -0.071 |
Token | a |
Token ID | 257 |
Feature activation | +3.12e-03 |
Pos logit contributions | |
flour | +0.010 |
oons | +0.010 |
ingredients | +0.009 |
tables | +0.009 |
ries | +0.009 |
Neg logit contributions | |
heel | -0.014 |
badly | -0.014 |
sing | -0.014 |
recovery | -0.014 |
boo | -0.013 |
Token | white |
Token ID | 2330 |
Feature activation | -7.16e-02 |
Pos logit contributions | |
flour | +0.004 |
brushes | +0.003 |
pots | +0.003 |
ingredients | +0.003 |
pans | +0.003 |
Neg logit contributions | |
recovery | -0.007 |
symp | -0.007 |
aring | -0.007 |
onel | -0.007 |
sing | -0.007 |
Token | sc |
Token ID | 629 |
Feature activation | -2.31e-01 |
Pos logit contributions | |
former | +0.129 |
symp | +0.128 |
sing | +0.127 |
recovery | +0.124 |
aring | +0.121 |
Neg logit contributions | |
flour | -0.116 |
ones | -0.106 |
pots | -0.104 |
tables | -0.103 |
brushes | -0.102 |
Token | ooter |
Token ID | 25141 |
Feature activation | -1.05e-01 |
Pos logit contributions | |
recovery | +0.249 |
symp | +0.246 |
former | +0.234 |
sing | +0.234 |
boo | +0.230 |
Neg logit contributions | |
flour | -0.519 |
ups | -0.511 |
brushes | -0.508 |
cakes | -0.503 |
oons | -0.489 |
Token | . |
Token ID | 13 |
Feature activation | -2.97e-02 |
Pos logit contributions | |
sing | +0.250 |
recovery | +0.244 |
symp | +0.240 |
heel | +0.240 |
boo | +0.240 |
Neg logit contributions | |
flour | -0.115 |
ingredients | -0.099 |
tables | -0.099 |
oons | -0.097 |
pots | -0.096 |
Token | They |
Token ID | 1119 |
Feature activation | +1.82e-02 |
Pos logit contributions | |
sing | +0.050 |
recovery | +0.049 |
heel | +0.048 |
badly | +0.048 |
boo | +0.048 |
Neg logit contributions | |
flour | -0.053 |
ingredients | -0.049 |
tables | -0.048 |
oons | -0.048 |
pots | -0.048 |
Token | raced |
Token ID | 32897 |
Feature activation | +3.24e-02 |
Pos logit contributions | |
flour | +0.028 |
ingredients | +0.026 |
oons | +0.025 |
tables | +0.025 |
pots | +0.024 |
Neg logit contributions | |
sing | -0.035 |
heel | -0.034 |
boo | -0.034 |
badly | -0.034 |
symp | -0.034 |
Token | each |
Token ID | 1123 |
Feature activation | +4.52e-02 |
Pos logit contributions | |
flour | +0.043 |
tables | +0.038 |
ingredients | +0.038 |
oons | +0.037 |
pots | +0.037 |
Neg logit contributions | |
sing | -0.070 |
recovery | -0.068 |
heel | -0.067 |
boo | -0.067 |
badly | -0.067 |
Token | , |
Token ID | 11 |
Feature activation | +6.15e-02 |
Pos logit contributions | |
badly | +0.035 |
sing | +0.033 |
bye | +0.030 |
tweet | +0.029 |
heel | +0.029 |
Neg logit contributions | |
flour | -0.045 |
liquids | -0.044 |
ups | -0.043 |
ingredients | -0.043 |
pans | -0.043 |
Token | she |
Token ID | 673 |
Feature activation | -1.15e-02 |
Pos logit contributions | |
oons | +0.111 |
ups | +0.111 |
liquids | +0.108 |
flour | +0.107 |
ingredients | +0.104 |
Neg logit contributions | |
badly | -0.097 |
sing | -0.096 |
tweet | -0.087 |
bye | -0.083 |
heel | -0.082 |
Token | saw |
Token ID | 2497 |
Feature activation | +3.97e-02 |
Pos logit contributions | |
boo | +0.021 |
heel | +0.020 |
badly | +0.019 |
promise | +0.019 |
sing | +0.019 |
Neg logit contributions | |
flour | -0.019 |
ries | -0.018 |
cakes | -0.018 |
ingredients | -0.018 |
oons | -0.017 |
Token | a |
Token ID | 257 |
Feature activation | +2.04e-02 |
Pos logit contributions | |
oons | +0.077 |
pans | +0.071 |
liquids | +0.070 |
ups | +0.069 |
glers | +0.069 |
Neg logit contributions | |
badly | -0.059 |
boo | -0.059 |
heel | -0.055 |
knee | -0.054 |
wing | -0.052 |
Token | beautiful |
Token ID | 4950 |
Feature activation | -1.00e-01 |
Pos logit contributions | |
flour | +0.028 |
pots | +0.023 |
tables | +0.023 |
ingredients | +0.022 |
pans | +0.022 |
Neg logit contributions | |
sing | -0.044 |
recovery | -0.044 |
former | -0.043 |
spine | -0.043 |
symp | -0.043 |
Token | fl |
Token ID | 781 |
Feature activation | -2.31e-01 |
Pos logit contributions | |
sing | +0.203 |
recovery | +0.198 |
symp | +0.197 |
former | +0.194 |
spine | +0.192 |
Neg logit contributions | |
flour | -0.150 |
ingredients | -0.129 |
tables | -0.127 |
pots | -0.127 |
chairs | -0.117 |
Token | ute |
Token ID | 1133 |
Feature activation | -1.44e-01 |
Pos logit contributions | |
former | +0.254 |
symp | +0.233 |
recovery | +0.230 |
istence | +0.229 |
boo | +0.225 |
Neg logit contributions | |
flour | -0.492 |
oons | -0.468 |
pots | -0.456 |
ingredients | -0.450 |
cakes | -0.438 |
Token | in |
Token ID | 287 |
Feature activation | -3.02e-02 |
Pos logit contributions | |
sing | +0.325 |
recovery | +0.319 |
boo | +0.314 |
heel | +0.314 |
former | +0.314 |
Neg logit contributions | |
flour | -0.172 |
ingredients | -0.150 |
tables | -0.149 |
pots | -0.148 |
oons | -0.143 |
Token | the |
Token ID | 262 |
Feature activation | +2.60e-02 |
Pos logit contributions | |
heel | +0.056 |
recovery | +0.056 |
boo | +0.054 |
symp | +0.054 |
badly | +0.053 |
Neg logit contributions | |
oons | -0.046 |
cakes | -0.043 |
flour | -0.041 |
ingredients | -0.041 |
ups | -0.041 |
Token | window |
Token ID | 4324 |
Feature activation | -4.54e-02 |
Pos logit contributions | |
flour | +0.038 |
ingredients | +0.034 |
tables | +0.034 |
pots | +0.033 |
oons | +0.033 |
Neg logit contributions | |
sing | -0.053 |
recovery | -0.051 |
badly | -0.050 |
heel | -0.050 |
symp | -0.050 |
Token | of |
Token ID | 286 |
Feature activation | +3.70e-02 |
Pos logit contributions | |
sing | +0.099 |
recovery | +0.096 |
heel | +0.095 |
badly | +0.095 |
symp | +0.094 |
Neg logit contributions | |
flour | -0.059 |
ingredients | -0.052 |
oons | -0.052 |
tables | -0.052 |
pots | -0.050 |
Token | found |
Token ID | 1043 |
Feature activation | +5.57e-02 |
Pos logit contributions | |
sing | +0.104 |
recovery | +0.101 |
heel | +0.100 |
badly | +0.099 |
symp | +0.099 |
Neg logit contributions | |
flour | -0.067 |
ingredients | -0.060 |
tables | -0.059 |
oons | -0.059 |
pots | -0.058 |
Token | she |
Token ID | 673 |
Feature activation | -4.60e-02 |
Pos logit contributions | |
oons | +0.079 |
flour | +0.078 |
tables | +0.073 |
pans | +0.072 |
cakes | +0.071 |
Neg logit contributions | |
heel | -0.109 |
badly | -0.106 |
boo | -0.105 |
recovery | -0.104 |
hero | -0.102 |
Token | could |
Token ID | 714 |
Feature activation | -5.06e-02 |
Pos logit contributions | |
sing | +0.104 |
recovery | +0.101 |
heel | +0.099 |
symp | +0.099 |
spine | +0.099 |
Neg logit contributions | |
flour | -0.055 |
ingredients | -0.049 |
tables | -0.048 |
pots | -0.047 |
oons | -0.047 |
Token | make |
Token ID | 787 |
Feature activation | +4.21e-02 |
Pos logit contributions | |
sing | +0.089 |
symp | +0.088 |
spine | +0.088 |
OO | +0.085 |
recovery | +0.084 |
Neg logit contributions | |
flour | -0.076 |
tables | -0.072 |
brushes | -0.071 |
pots | -0.069 |
oons | -0.068 |
Token | the |
Token ID | 262 |
Feature activation | +7.47e-03 |
Pos logit contributions | |
flour | +0.055 |
ingredients | +0.049 |
tables | +0.048 |
oons | +0.048 |
pots | +0.047 |
Neg logit contributions | |
sing | -0.091 |
recovery | -0.088 |
heel | -0.088 |
badly | -0.087 |
boo | -0.087 |
Token | fl |
Token ID | 781 |
Feature activation | -2.31e-01 |
Pos logit contributions | |
flour | +0.012 |
ingredients | +0.011 |
tables | +0.010 |
pots | +0.010 |
chairs | +0.010 |
Neg logit contributions | |
sing | -0.014 |
badly | -0.014 |
ur | -0.014 |
recovery | -0.014 |
symp | -0.014 |
Token | ute |
Token ID | 1133 |
Feature activation | -1.55e-01 |
Pos logit contributions | |
bru | +0.207 |
boo | +0.206 |
symp | +0.195 |
sing | +0.191 |
recovery | +0.188 |
Neg logit contributions | |
flour | -0.533 |
ups | -0.533 |
oons | -0.527 |
ingredients | -0.508 |
cakes | -0.502 |
Token | sing |
Token ID | 1702 |
Feature activation | -3.60e-02 |
Pos logit contributions | |
sing | +0.281 |
recovery | +0.272 |
symp | +0.270 |
spine | +0.268 |
boo | +0.267 |
Neg logit contributions | |
flour | -0.247 |
ingredients | -0.230 |
pots | -0.223 |
tables | -0.222 |
oons | -0.213 |
Token | like |
Token ID | 588 |
Feature activation | +2.62e-03 |
Pos logit contributions | |
sing | +0.076 |
recovery | +0.073 |
heel | +0.073 |
boo | +0.073 |
symp | +0.072 |
Neg logit contributions | |
flour | -0.049 |
ingredients | -0.044 |
tables | -0.044 |
oons | -0.043 |
pots | -0.043 |
Token | a |
Token ID | 257 |
Feature activation | +1.83e-02 |
Pos logit contributions | |
flour | +0.003 |
ingredients | +0.003 |
oons | +0.003 |
tables | +0.003 |
pots | +0.003 |
Neg logit contributions | |
sing | -0.006 |
recovery | -0.006 |
heel | -0.006 |
badly | -0.006 |
boo | -0.006 |
Token | bird |
Token ID | 6512 |
Feature activation | -8.92e-02 |
Pos logit contributions | |
flour | +0.030 |
ingredients | +0.027 |
tables | +0.026 |
chairs | +0.025 |
pots | +0.025 |
Neg logit contributions | |
sing | -0.032 |
symp | -0.032 |
spine | -0.031 |
ighed | -0.031 |
recovery | -0.031 |
Token | drum |
Token ID | 13026 |
Feature activation | -8.68e-02 |
Pos logit contributions | |
flour | +0.016 |
pots | +0.014 |
ingredients | +0.014 |
brushes | +0.013 |
pans | +0.013 |
Neg logit contributions | |
symp | -0.030 |
recovery | -0.029 |
utt | -0.028 |
ur | -0.028 |
aring | -0.028 |
Token | and |
Token ID | 290 |
Feature activation | -9.23e-03 |
Pos logit contributions | |
sing | +0.196 |
symp | +0.192 |
recovery | +0.190 |
promise | +0.188 |
boo | +0.187 |
Neg logit contributions | |
flour | -0.102 |
tables | -0.090 |
pots | -0.090 |
ingredients | -0.089 |
oons | -0.088 |
Token | Mia |
Token ID | 32189 |
Feature activation | +4.46e-02 |
Pos logit contributions | |
heel | +0.017 |
sing | +0.017 |
boo | +0.016 |
badly | +0.016 |
hero | +0.016 |
Neg logit contributions | |
flour | -0.015 |
ingredients | -0.014 |
oons | -0.013 |
ries | -0.013 |
cakes | -0.013 |
Token | has |
Token ID | 468 |
Feature activation | +1.72e-02 |
Pos logit contributions | |
flour | +0.073 |
ingredients | +0.067 |
tables | +0.066 |
oons | +0.065 |
pots | +0.064 |
Neg logit contributions | |
sing | -0.081 |
heel | -0.079 |
boo | -0.078 |
recovery | -0.078 |
badly | -0.077 |
Token | a |
Token ID | 257 |
Feature activation | +7.10e-03 |
Pos logit contributions | |
flour | +0.024 |
oons | +0.022 |
ingredients | +0.022 |
ries | +0.022 |
tables | +0.021 |
Neg logit contributions | |
sing | -0.035 |
heel | -0.034 |
badly | -0.033 |
boo | -0.033 |
recovery | -0.033 |
Token | fl |
Token ID | 781 |
Feature activation | -2.31e-01 |
Pos logit contributions | |
flour | +0.011 |
pots | +0.010 |
ingredients | +0.010 |
brushes | +0.009 |
pans | +0.009 |
Neg logit contributions | |
symp | -0.014 |
recovery | -0.014 |
utt | -0.013 |
ur | -0.013 |
sing | -0.013 |
Token | ute |
Token ID | 1133 |
Feature activation | -1.57e-01 |
Pos logit contributions | |
recovery | +0.219 |
symp | +0.212 |
former | +0.209 |
bru | +0.202 |
boo | +0.199 |
Neg logit contributions | |
flour | -0.531 |
oons | -0.526 |
pots | -0.507 |
brushes | -0.499 |
ingredients | -0.497 |
Token | . |
Token ID | 13 |
Feature activation | -1.63e-02 |
Pos logit contributions | |
sing | +0.379 |
symp | +0.372 |
recovery | +0.371 |
boo | +0.367 |
heel | +0.364 |
Neg logit contributions | |
flour | -0.158 |
pots | -0.137 |
ingredients | -0.136 |
tables | -0.133 |
oons | -0.131 |
Token | They |
Token ID | 1119 |
Feature activation | +1.85e-02 |
Pos logit contributions | |
sing | +0.032 |
recovery | +0.031 |
heel | +0.031 |
symp | +0.031 |
boo | +0.030 |
Neg logit contributions | |
flour | -0.024 |
ingredients | -0.022 |
tables | -0.022 |
oons | -0.021 |
pots | -0.021 |
Token | make |
Token ID | 787 |
Feature activation | +7.60e-02 |
Pos logit contributions | |
flour | +0.033 |
tables | +0.030 |
pots | +0.030 |
oons | +0.030 |
ingredients | +0.030 |
Neg logit contributions | |
sing | -0.031 |
recovery | -0.031 |
badly | -0.030 |
symp | -0.029 |
heel | -0.029 |
Token | different |
Token ID | 1180 |
Feature activation | -3.52e-02 |
Pos logit contributions | |
flour | +0.113 |
ingredients | +0.101 |
tables | +0.101 |
oons | +0.100 |
pots | +0.099 |
Neg logit contributions | |
sing | -0.151 |
recovery | -0.147 |
heel | -0.145 |
badly | -0.144 |
symp | -0.144 |
Token | , |
Token ID | 11 |
Feature activation | +1.67e-02 |
Pos logit contributions | |
flour | +0.019 |
ingredients | +0.018 |
oons | +0.018 |
pots | +0.017 |
tables | +0.017 |
Neg logit contributions | |
heel | -0.024 |
badly | -0.023 |
sing | -0.023 |
boo | -0.023 |
recovery | -0.023 |
Token | I |
Token ID | 314 |
Feature activation | +1.55e-02 |
Pos logit contributions | |
flour | +0.032 |
tables | +0.029 |
pots | +0.028 |
ingredients | +0.028 |
pans | +0.028 |
Neg logit contributions | |
sing | -0.027 |
former | -0.026 |
recovery | -0.026 |
badly | -0.025 |
symp | -0.025 |
Token | cannot |
Token ID | 2314 |
Feature activation | -5.77e-02 |
Pos logit contributions | |
flour | +0.026 |
ingredients | +0.023 |
oons | +0.023 |
pots | +0.023 |
tables | +0.022 |
Neg logit contributions | |
sing | -0.028 |
recovery | -0.027 |
spine | -0.027 |
symp | -0.027 |
heel | -0.026 |
Token | play |
Token ID | 711 |
Feature activation | -9.79e-02 |
Pos logit contributions | |
symp | +0.096 |
sing | +0.094 |
former | +0.094 |
spine | +0.093 |
recovery | +0.089 |
Neg logit contributions | |
flour | -0.098 |
pots | -0.091 |
oons | -0.089 |
tables | -0.087 |
ingredients | -0.085 |
Token | your |
Token ID | 534 |
Feature activation | -1.41e-01 |
Pos logit contributions | |
sing | +0.235 |
spine | +0.232 |
former | +0.224 |
symp | +0.222 |
berth | +0.218 |
Neg logit contributions | |
flour | -0.108 |
tables | -0.096 |
pots | -0.096 |
cans | -0.090 |
ingredients | -0.089 |
Token | fl |
Token ID | 781 |
Feature activation | -2.30e-01 |
Pos logit contributions | |
symp | +0.239 |
badly | +0.232 |
ur | +0.230 |
berth | +0.229 |
sing | +0.226 |
Neg logit contributions | |
flour | -0.237 |
pots | -0.229 |
flowers | -0.223 |
chairs | -0.222 |
tables | -0.220 |
Token | ute |
Token ID | 1133 |
Feature activation | -1.55e-01 |
Pos logit contributions | |
symp | +0.248 |
bru | +0.243 |
boo | +0.239 |
sing | +0.229 |
recovery | +0.224 |
Neg logit contributions | |
oons | -0.496 |
ups | -0.489 |
flour | -0.483 |
ingredients | -0.458 |
pots | -0.457 |
Token | . |
Token ID | 13 |
Feature activation | -4.68e-02 |
Pos logit contributions | |
sing | +0.370 |
symp | +0.364 |
recovery | +0.361 |
boo | +0.356 |
spine | +0.355 |
Neg logit contributions | |
flour | -0.153 |
ingredients | -0.134 |
pots | -0.133 |
tables | -0.131 |
oons | -0.127 |
Token | It |
Token ID | 632 |
Feature activation | -5.74e-03 |
Pos logit contributions | |
badly | +0.095 |
sing | +0.095 |
former | +0.095 |
recovery | +0.094 |
symp | +0.094 |
Neg logit contributions | |
flour | -0.059 |
pots | -0.052 |
brushes | -0.051 |
tables | -0.049 |
cans | -0.048 |
Token | is |
Token ID | 318 |
Feature activation | -2.37e-02 |
Pos logit contributions | |
sing | +0.014 |
recovery | +0.014 |
spine | +0.014 |
symp | +0.014 |
badly | +0.014 |
Neg logit contributions | |
flour | -0.006 |
ingredients | -0.005 |
pots | -0.005 |
oons | -0.005 |
tables | -0.005 |
Token | too |
Token ID | 1165 |
Feature activation | -2.41e-02 |
Pos logit contributions | |
recovery | +0.042 |
symp | +0.041 |
sing | +0.040 |
uer | +0.040 |
spine | +0.040 |
Neg logit contributions | |
flour | -0.042 |
pots | -0.037 |
ingredients | -0.037 |
tables | -0.036 |
pans | -0.034 |
Token | . |
Token ID | 13 |
Feature activation | +1.37e-03 |
Pos logit contributions | |
sing | +0.287 |
recovery | +0.276 |
boo | +0.274 |
heel | +0.273 |
symp | +0.273 |
Neg logit contributions | |
flour | -0.138 |
ingredients | -0.120 |
tables | -0.120 |
pots | -0.118 |
oons | -0.117 |
Token | She |
Token ID | 1375 |
Feature activation | -4.96e-02 |
Pos logit contributions | |
flour | +0.002 |
ingredients | +0.002 |
oons | +0.002 |
ries | +0.002 |
tables | +0.002 |
Neg logit contributions | |
recovery | -0.002 |
sing | -0.002 |
heel | -0.002 |
boo | -0.002 |
symp | -0.002 |
Token | had |
Token ID | 550 |
Feature activation | +3.13e-02 |
Pos logit contributions | |
sing | +0.091 |
heel | +0.090 |
boo | +0.090 |
badly | +0.087 |
recovery | +0.086 |
Neg logit contributions | |
flour | -0.080 |
ingredients | -0.074 |
ries | -0.073 |
tables | -0.071 |
cakes | -0.070 |
Token | a |
Token ID | 257 |
Feature activation | +1.90e-02 |
Pos logit contributions | |
oons | +0.051 |
ries | +0.050 |
cakes | +0.047 |
flour | +0.047 |
pans | +0.046 |
Neg logit contributions | |
heel | -0.056 |
badly | -0.054 |
boo | -0.054 |
sing | -0.052 |
recovery | -0.051 |
Token | small |
Token ID | 1402 |
Feature activation | -4.37e-02 |
Pos logit contributions | |
flour | +0.022 |
pots | +0.018 |
ingredients | +0.017 |
brushes | +0.016 |
cups | +0.016 |
Neg logit contributions | |
aring | -0.046 |
utt | -0.046 |
symp | -0.046 |
sing | -0.045 |
spine | -0.045 |
Token | fl |
Token ID | 781 |
Feature activation | -2.29e-01 |
Pos logit contributions | |
symp | +0.088 |
aring | +0.087 |
utt | +0.087 |
shadow | +0.086 |
sing | +0.085 |
Neg logit contributions | |
flour | -0.066 |
pots | -0.059 |
ingredients | -0.057 |
cups | -0.055 |
plates | -0.054 |
Token | ute |
Token ID | 1133 |
Feature activation | -1.48e-01 |
Pos logit contributions | |
recovery | +0.232 |
former | +0.230 |
boo | +0.227 |
symp | +0.224 |
bru | +0.217 |
Neg logit contributions | |
flour | -0.505 |
oons | -0.489 |
pots | -0.473 |
ingredients | -0.467 |
brushes | -0.465 |
Token | that |
Token ID | 326 |
Feature activation | +1.18e-02 |
Pos logit contributions | |
sing | +0.299 |
heel | +0.294 |
recovery | +0.290 |
badly | +0.290 |
boo | +0.286 |
Neg logit contributions | |
flour | -0.215 |
ingredients | -0.193 |
oons | -0.193 |
tables | -0.192 |
pots | -0.184 |
Token | made |
Token ID | 925 |
Feature activation | +5.22e-02 |
Pos logit contributions | |
flour | +0.017 |
ingredients | +0.016 |
oons | +0.016 |
ries | +0.016 |
tables | +0.015 |
Neg logit contributions | |
heel | -0.023 |
boo | -0.023 |
sing | -0.022 |
badly | -0.022 |
recovery | -0.022 |
Token | pretty |
Token ID | 2495 |
Feature activation | -5.77e-02 |
Pos logit contributions | |
flour | +0.074 |
ingredients | +0.067 |
oons | +0.067 |
tables | +0.066 |
pans | +0.065 |
Neg logit contributions | |
sing | -0.101 |
recovery | -0.101 |
heel | -0.100 |
boo | -0.100 |
badly | -0.099 |
Token | sounds |
Token ID | 5238 |
Feature activation | +8.89e-03 |
Pos logit contributions | |
sing | +0.115 |
heel | +0.115 |
badly | +0.113 |
recovery | +0.111 |
boo | +0.111 |
Neg logit contributions | |
flour | -0.083 |
ingredients | -0.075 |
oons | -0.074 |
tables | -0.072 |
pots | -0.071 |
Token | to |
Token ID | 284 |
Feature activation | -5.46e-02 |
Pos logit contributions | |
heel | +0.091 |
boo | +0.090 |
sing | +0.089 |
recovery | +0.089 |
badly | +0.089 |
Neg logit contributions | |
flour | -0.084 |
oons | -0.080 |
ingredients | -0.077 |
tables | -0.077 |
pots | -0.074 |
Token | hear |
Token ID | 3285 |
Feature activation | +3.47e-02 |
Pos logit contributions | |
sing | +0.080 |
symp | +0.078 |
recovery | +0.075 |
spine | +0.074 |
badly | +0.074 |
Neg logit contributions | |
flour | -0.109 |
tables | -0.101 |
pots | -0.100 |
ingredients | -0.098 |
oons | -0.096 |
Token | me |
Token ID | 502 |
Feature activation | -1.13e-01 |
Pos logit contributions | |
flour | +0.053 |
ingredients | +0.049 |
oons | +0.049 |
tables | +0.048 |
pots | +0.047 |
Neg logit contributions | |
sing | -0.066 |
recovery | -0.064 |
heel | -0.064 |
boo | -0.064 |
badly | -0.064 |
Token | play |
Token ID | 711 |
Feature activation | -1.04e-01 |
Pos logit contributions | |
recovery | +0.275 |
sing | +0.275 |
symp | +0.274 |
spine | +0.272 |
former | +0.272 |
Neg logit contributions | |
flour | -0.102 |
tables | -0.087 |
pots | -0.086 |
oons | -0.085 |
ries | -0.084 |
Token | my |
Token ID | 616 |
Feature activation | -1.17e-01 |
Pos logit contributions | |
sing | +0.245 |
spine | +0.238 |
former | +0.230 |
symp | +0.229 |
recovery | +0.227 |
Neg logit contributions | |
flour | -0.122 |
tables | -0.109 |
pots | -0.106 |
oons | -0.104 |
ingredients | -0.103 |
Token | fl |
Token ID | 781 |
Feature activation | -2.29e-01 |
Pos logit contributions | |
badly | +0.188 |
symp | +0.188 |
recovery | +0.187 |
spine | +0.185 |
sing | +0.184 |
Neg logit contributions | |
flour | -0.209 |
ingredients | -0.193 |
tables | -0.193 |
pots | -0.192 |
chairs | -0.191 |
Token | ute |
Token ID | 1133 |
Feature activation | -1.54e-01 |
Pos logit contributions | |
bru | +0.229 |
symp | +0.229 |
boo | +0.228 |
recovery | +0.212 |
sing | +0.206 |
Neg logit contributions | |
oons | -0.521 |
flour | -0.507 |
ups | -0.498 |
cakes | -0.485 |
ingredients | -0.479 |
Token | ?" |
Token ID | 1701 |
Feature activation | +6.64e-02 |
Pos logit contributions | |
symp | +0.398 |
sing | +0.396 |
recovery | +0.390 |
spine | +0.390 |
former | +0.382 |
Neg logit contributions | |
flour | -0.123 |
pots | -0.106 |
ingredients | -0.100 |
tables | -0.100 |
oons | -0.095 |
Token |
|
Token ID | 198 |
Feature activation | -4.84e-02 |
Pos logit contributions | |
flour | +0.096 |
tables | +0.086 |
ingredients | +0.085 |
oons | +0.084 |
pots | +0.083 |
Neg logit contributions | |
sing | -0.136 |
recovery | -0.132 |
heel | -0.131 |
badly | -0.130 |
boo | -0.129 |
Token |
|
Token ID | 198 |
Feature activation | -5.05e-02 |
Pos logit contributions | |
recovery | +0.079 |
sing | +0.079 |
badly | +0.079 |
heel | +0.079 |
symp | +0.078 |
Neg logit contributions | |
flour | -0.086 |
ingredients | -0.082 |
pots | -0.078 |
tables | -0.077 |
ries | -0.077 |
Token | The |
Token ID | 464 |
Feature activation | +2.11e-02 |
Pos logit contributions | |
sing | +0.099 |
recovery | +0.096 |
heel | +0.095 |
boo | +0.094 |
badly | +0.094 |
Neg logit contributions | |
flour | -0.077 |
tables | -0.070 |
ingredients | -0.070 |
oons | -0.069 |
pots | -0.068 |
Token | . |
Token ID | 13 |
Feature activation | -2.16e-03 |
Pos logit contributions | |
sing | +0.281 |
recovery | +0.273 |
heel | +0.271 |
boo | +0.270 |
symp | +0.269 |
Neg logit contributions | |
flour | -0.143 |
ingredients | -0.125 |
tables | -0.125 |
pots | -0.123 |
oons | -0.122 |
Token | She |
Token ID | 1375 |
Feature activation | -4.61e-02 |
Pos logit contributions | |
sing | +0.004 |
recovery | +0.004 |
heel | +0.004 |
boo | +0.004 |
badly | +0.004 |
Neg logit contributions | |
flour | -0.004 |
ingredients | -0.003 |
tables | -0.003 |
oons | -0.003 |
pots | -0.003 |
Token | has |
Token ID | 468 |
Feature activation | +2.47e-02 |
Pos logit contributions | |
sing | +0.082 |
heel | +0.081 |
boo | +0.081 |
recovery | +0.079 |
promise | +0.078 |
Neg logit contributions | |
flour | -0.077 |
ingredients | -0.071 |
tables | -0.069 |
ries | -0.069 |
oons | -0.068 |
Token | a |
Token ID | 257 |
Feature activation | +1.46e-02 |
Pos logit contributions | |
oons | +0.040 |
ries | +0.039 |
flour | +0.038 |
cakes | +0.037 |
ingredients | +0.036 |
Neg logit contributions | |
heel | -0.045 |
badly | -0.042 |
boo | -0.042 |
sing | -0.041 |
recovery | -0.041 |
Token | pink |
Token ID | 11398 |
Feature activation | -7.14e-02 |
Pos logit contributions | |
flour | +0.015 |
pots | +0.012 |
brushes | +0.012 |
ingredients | +0.011 |
cups | +0.010 |
Neg logit contributions | |
symp | -0.037 |
utt | -0.037 |
aring | -0.037 |
recovery | -0.036 |
sing | -0.036 |
Token | fl |
Token ID | 781 |
Feature activation | -2.29e-01 |
Pos logit contributions | |
aring | +0.144 |
utt | +0.140 |
former | +0.138 |
symp | +0.137 |
shadow | +0.136 |
Neg logit contributions | |
flour | -0.101 |
ones | -0.093 |
pots | -0.092 |
flowers | -0.092 |
chairs | -0.092 |
Token | ute |
Token ID | 1133 |
Feature activation | -1.46e-01 |
Pos logit contributions | |
former | +0.242 |
recovery | +0.234 |
symp | +0.230 |
boo | +0.227 |
bru | +0.224 |
Neg logit contributions | |
flour | -0.496 |
oons | -0.488 |
pots | -0.467 |
brushes | -0.466 |
ingredients | -0.459 |
Token | that |
Token ID | 326 |
Feature activation | +1.96e-02 |
Pos logit contributions | |
sing | +0.293 |
heel | +0.288 |
badly | +0.283 |
recovery | +0.283 |
boo | +0.277 |
Neg logit contributions | |
flour | -0.214 |
ingredients | -0.192 |
oons | -0.192 |
tables | -0.190 |
pots | -0.182 |
Token | makes |
Token ID | 1838 |
Feature activation | +4.72e-02 |
Pos logit contributions | |
flour | +0.027 |
ingredients | +0.025 |
tables | +0.024 |
ries | +0.024 |
oons | +0.024 |
Neg logit contributions | |
heel | -0.039 |
boo | -0.039 |
sing | -0.038 |
recovery | -0.038 |
hero | -0.037 |
Token | pretty |
Token ID | 2495 |
Feature activation | -5.28e-02 |
Pos logit contributions | |
flour | +0.062 |
ingredients | +0.055 |
tables | +0.054 |
oons | +0.054 |
pots | +0.053 |
Neg logit contributions | |
sing | -0.101 |
recovery | -0.099 |
heel | -0.098 |
boo | -0.097 |
badly | -0.097 |
Token | sounds |
Token ID | 5238 |
Feature activation | +1.28e-02 |
Pos logit contributions | |
sing | +0.107 |
heel | +0.106 |
badly | +0.105 |
symp | +0.104 |
recovery | +0.103 |
Neg logit contributions | |
flour | -0.075 |
ingredients | -0.067 |
oons | -0.066 |
tables | -0.064 |
pots | -0.064 |
Token | garden |
Token ID | 11376 |
Feature activation | -4.45e-02 |
Pos logit contributions | |
flour | +0.047 |
ingredients | +0.043 |
tables | +0.043 |
oons | +0.042 |
pots | +0.042 |
Neg logit contributions | |
sing | -0.062 |
recovery | -0.059 |
heel | -0.059 |
badly | -0.059 |
symp | -0.058 |
Token | and |
Token ID | 290 |
Feature activation | -6.20e-03 |
Pos logit contributions | |
sing | +0.091 |
recovery | +0.088 |
heel | +0.087 |
badly | +0.087 |
spine | +0.085 |
Neg logit contributions | |
flour | -0.063 |
oons | -0.059 |
pans | -0.056 |
ingredients | -0.055 |
cakes | -0.055 |
Token | found |
Token ID | 1043 |
Feature activation | +4.83e-02 |
Pos logit contributions | |
sing | +0.011 |
heel | +0.011 |
boo | +0.011 |
badly | +0.010 |
tweet | +0.010 |
Neg logit contributions | |
flour | -0.010 |
oons | -0.010 |
ingredients | -0.009 |
cakes | -0.009 |
ries | -0.009 |
Token | a |
Token ID | 257 |
Feature activation | +2.11e-02 |
Pos logit contributions | |
oons | +0.091 |
pans | +0.083 |
cakes | +0.081 |
tables | +0.080 |
glers | +0.080 |
Neg logit contributions | |
badly | -0.074 |
heel | -0.074 |
boo | -0.073 |
recovery | -0.070 |
knee | -0.069 |
Token | beautiful |
Token ID | 4950 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
flour | +0.032 |
ingredients | +0.028 |
pots | +0.028 |
tables | +0.027 |
pans | +0.026 |
Neg logit contributions | |
symp | -0.042 |
sing | -0.042 |
recovery | -0.042 |
spine | -0.041 |
former | -0.040 |
Token | fl |
Token ID | 781 |
Feature activation | -2.29e-01 |
Pos logit contributions | |
sing | +0.190 |
recovery | +0.183 |
symp | +0.183 |
spine | +0.177 |
badly | +0.176 |
Neg logit contributions | |
flour | -0.164 |
ingredients | -0.147 |
tables | -0.144 |
pots | -0.143 |
oons | -0.138 |
Token | ute |
Token ID | 1133 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
boo | +0.220 |
former | +0.210 |
recovery | +0.207 |
symp | +0.204 |
bru | +0.202 |
Neg logit contributions | |
flour | -0.530 |
oons | -0.514 |
pots | -0.499 |
ingredients | -0.498 |
cakes | -0.485 |
Token | . |
Token ID | 13 |
Feature activation | -1.20e-02 |
Pos logit contributions | |
sing | +0.318 |
recovery | +0.310 |
heel | +0.308 |
boo | +0.306 |
symp | +0.305 |
Neg logit contributions | |
flour | -0.163 |
ingredients | -0.143 |
tables | -0.141 |
oons | -0.139 |
pots | -0.139 |
Token | It |
Token ID | 632 |
Feature activation | +4.25e-03 |
Pos logit contributions | |
symp | +0.015 |
heel | +0.015 |
boo | +0.015 |
sing | +0.015 |
berth | +0.014 |
Neg logit contributions | |
flour | -0.025 |
ries | -0.024 |
glers | -0.023 |
ingredients | -0.023 |
ups | -0.023 |
Token | had |
Token ID | 550 |
Feature activation | +3.71e-02 |
Pos logit contributions | |
flour | +0.006 |
ingredients | +0.005 |
oons | +0.005 |
tables | +0.005 |
ries | +0.005 |
Neg logit contributions | |
heel | -0.009 |
sing | -0.009 |
boo | -0.009 |
badly | -0.009 |
symp | -0.008 |
Token | a |
Token ID | 257 |
Feature activation | +2.37e-02 |
Pos logit contributions | |
flour | +0.054 |
oons | +0.052 |
ingredients | +0.049 |
ries | +0.049 |
tables | +0.048 |
Neg logit contributions | |
heel | -0.072 |
sing | -0.072 |
boo | -0.070 |
badly | -0.070 |
recovery | -0.069 |
Token | glad |
Token ID | 9675 |
Feature activation | -3.77e-02 |
Pos logit contributions | |
sing | +0.006 |
recovery | +0.005 |
spine | +0.005 |
heel | +0.005 |
symp | +0.005 |
Neg logit contributions | |
flour | -0.006 |
ingredients | -0.006 |
tables | -0.006 |
pots | -0.005 |
oons | -0.005 |
Token | she |
Token ID | 673 |
Feature activation | -2.77e-02 |
Pos logit contributions | |
sing | +0.083 |
recovery | +0.081 |
heel | +0.081 |
badly | +0.080 |
boo | +0.080 |
Neg logit contributions | |
flour | -0.048 |
ingredients | -0.042 |
oons | -0.042 |
tables | -0.042 |
pots | -0.041 |
Token | had |
Token ID | 550 |
Feature activation | +1.44e-02 |
Pos logit contributions | |
sing | +0.058 |
heel | +0.057 |
recovery | +0.057 |
badly | +0.057 |
boo | +0.056 |
Neg logit contributions | |
flour | -0.039 |
oons | -0.034 |
ingredients | -0.033 |
tables | -0.033 |
pots | -0.033 |
Token | taken |
Token ID | 2077 |
Feature activation | -3.37e-02 |
Pos logit contributions | |
flour | +0.024 |
oons | +0.023 |
tables | +0.022 |
pots | +0.021 |
ries | +0.021 |
Neg logit contributions | |
heel | -0.026 |
badly | -0.026 |
recovery | -0.025 |
boo | -0.025 |
promise | -0.024 |
Token | the |
Token ID | 262 |
Feature activation | +5.28e-03 |
Pos logit contributions | |
heel | +0.058 |
badly | +0.058 |
recovery | +0.057 |
boo | +0.056 |
hero | +0.055 |
Neg logit contributions | |
oons | -0.057 |
flour | -0.057 |
pots | -0.052 |
pans | -0.052 |
tables | -0.052 |
Token | fl |
Token ID | 781 |
Feature activation | -2.29e-01 |
Pos logit contributions | |
flour | +0.011 |
ingredients | +0.010 |
pots | +0.009 |
tables | +0.009 |
pans | +0.009 |
Neg logit contributions | |
sing | -0.008 |
symp | -0.008 |
badly | -0.008 |
recovery | -0.007 |
former | -0.007 |
Token | ute |
Token ID | 1133 |
Feature activation | -1.42e-01 |
Pos logit contributions | |
sing | +0.205 |
symp | +0.198 |
bru | +0.193 |
boo | +0.192 |
appetite | +0.177 |
Neg logit contributions | |
flour | -0.545 |
oons | -0.533 |
ups | -0.530 |
ingredients | -0.529 |
brushes | -0.507 |
Token | home |
Token ID | 1363 |
Feature activation | -8.01e-02 |
Pos logit contributions | |
sing | +0.302 |
boo | +0.290 |
recovery | +0.289 |
symp | +0.288 |
heel | +0.284 |
Neg logit contributions | |
flour | -0.189 |
ingredients | -0.174 |
tables | -0.168 |
pots | -0.165 |
oons | -0.160 |
Token | , |
Token ID | 11 |
Feature activation | +3.12e-02 |
Pos logit contributions | |
sing | +0.165 |
recovery | +0.161 |
badly | +0.160 |
heel | +0.160 |
symp | +0.157 |
Neg logit contributions | |
flour | -0.113 |
ingredients | -0.100 |
oons | -0.100 |
tables | -0.100 |
pots | -0.098 |
Token | so |
Token ID | 523 |
Feature activation | +1.66e-02 |
Pos logit contributions | |
flour | +0.043 |
ingredients | +0.040 |
tables | +0.038 |
pots | +0.037 |
chairs | +0.035 |
Neg logit contributions | |
sing | -0.064 |
recovery | -0.061 |
symp | -0.061 |
spine | -0.060 |
heel | -0.059 |
Token | she |
Token ID | 673 |
Feature activation | -2.45e-02 |
Pos logit contributions | |
flour | +0.019 |
ingredients | +0.017 |
tables | +0.017 |
oons | +0.017 |
pots | +0.016 |
Neg logit contributions | |
sing | -0.038 |
recovery | -0.037 |
heel | -0.037 |
badly | -0.037 |
boo | -0.037 |
Token | ice |
Token ID | 4771 |
Feature activation | -9.29e-02 |
Pos logit contributions | |
flour | +0.073 |
ingredients | +0.067 |
oons | +0.067 |
tables | +0.066 |
pots | +0.065 |
Neg logit contributions | |
sing | -0.086 |
heel | -0.084 |
recovery | -0.084 |
boo | -0.083 |
badly | -0.083 |
Token | cream |
Token ID | 8566 |
Feature activation | +3.75e-03 |
Pos logit contributions | |
badly | +0.190 |
sing | +0.187 |
promise | +0.187 |
heel | +0.185 |
recovery | +0.181 |
Neg logit contributions | |
flour | -0.120 |
tables | -0.111 |
pots | -0.106 |
liquids | -0.106 |
ingredients | -0.105 |
Token | truck |
Token ID | 7779 |
Feature activation | -4.55e-03 |
Pos logit contributions | |
flour | +0.006 |
oons | +0.006 |
tables | +0.006 |
ingredients | +0.006 |
pots | +0.006 |
Neg logit contributions | |
sing | -0.006 |
heel | -0.006 |
recovery | -0.006 |
promise | -0.006 |
badly | -0.006 |
Token | to |
Token ID | 284 |
Feature activation | -3.39e-02 |
Pos logit contributions | |
sing | +0.009 |
heel | +0.009 |
recovery | +0.009 |
badly | +0.009 |
boo | +0.008 |
Neg logit contributions | |
flour | -0.007 |
oons | -0.006 |
ingredients | -0.006 |
tables | -0.006 |
pots | -0.006 |
Token | buy |
Token ID | 2822 |
Feature activation | +4.20e-02 |
Pos logit contributions | |
sing | +0.051 |
recovery | +0.050 |
heel | +0.050 |
boo | +0.049 |
badly | +0.049 |
Neg logit contributions | |
flour | -0.066 |
ingredients | -0.061 |
oons | -0.061 |
tables | -0.060 |
pots | -0.060 |
Token | some |
Token ID | 617 |
Feature activation | +1.23e-01 |
Pos logit contributions | |
oons | +0.066 |
tables | +0.065 |
pans | +0.064 |
flour | +0.063 |
pots | +0.063 |
Neg logit contributions | |
badly | -0.073 |
heel | -0.065 |
retri | -0.063 |
recovery | -0.063 |
boo | -0.063 |
Token | . |
Token ID | 13 |
Feature activation | -1.08e-02 |
Pos logit contributions | |
tables | +0.215 |
flour | +0.212 |
oons | +0.210 |
pots | +0.207 |
pans | +0.201 |
Neg logit contributions | |
badly | -0.184 |
promise | -0.180 |
heel | -0.174 |
recovery | -0.171 |
boo | -0.170 |
Token | She |
Token ID | 1375 |
Feature activation | -4.52e-02 |
Pos logit contributions | |
sing | +0.017 |
heel | +0.017 |
recovery | +0.017 |
boo | +0.017 |
badly | +0.017 |
Neg logit contributions | |
flour | -0.020 |
oons | -0.018 |
ingredients | -0.018 |
pots | -0.018 |
tables | -0.018 |
Token | asked |
Token ID | 1965 |
Feature activation | +1.10e-03 |
Pos logit contributions | |
boo | +0.087 |
badly | +0.085 |
sing | +0.085 |
heel | +0.084 |
symp | +0.083 |
Neg logit contributions | |
flour | -0.065 |
oons | -0.060 |
ingredients | -0.060 |
ries | -0.058 |
pots | -0.058 |
Token | the |
Token ID | 262 |
Feature activation | +6.24e-02 |
Pos logit contributions | |
flour | +0.002 |
oons | +0.002 |
ingredients | +0.002 |
pans | +0.001 |
tables | +0.001 |
Neg logit contributions | |
badly | -0.002 |
recovery | -0.002 |
heel | -0.002 |
boo | -0.002 |
symp | -0.002 |
Token | man |
Token ID | 582 |
Feature activation | -7.72e-03 |
Pos logit contributions | |
flour | +0.109 |
oons | +0.104 |
ingredients | +0.102 |
tables | +0.101 |
pots | +0.101 |
Neg logit contributions | |
sing | -0.102 |
heel | -0.100 |
recovery | -0.100 |
boo | -0.099 |
promise | -0.097 |
Token |
|
Token ID | 198 |
Feature activation | -3.98e-02 |
Pos logit contributions | |
sing | +0.087 |
spine | +0.082 |
heel | +0.081 |
boo | +0.081 |
recovery | +0.081 |
Neg logit contributions | |
flour | -0.044 |
oons | -0.039 |
tables | -0.037 |
ingredients | -0.037 |
cakes | -0.035 |
Token | He |
Token ID | 1544 |
Feature activation | -6.30e-02 |
Pos logit contributions | |
sing | +0.099 |
spine | +0.095 |
boo | +0.090 |
heel | +0.088 |
injury | +0.088 |
Neg logit contributions | |
flour | -0.047 |
tables | -0.041 |
oons | -0.041 |
ups | -0.041 |
ingredients | -0.039 |
Token | grabbed |
Token ID | 13176 |
Feature activation | +9.63e-03 |
Pos logit contributions | |
sing | +0.135 |
recovery | +0.131 |
heel | +0.129 |
badly | +0.128 |
boo | +0.128 |
Neg logit contributions | |
flour | -0.085 |
ingredients | -0.076 |
tables | -0.075 |
oons | -0.074 |
pots | -0.073 |
Token | some |
Token ID | 617 |
Feature activation | +1.29e-01 |
Pos logit contributions | |
ups | +0.013 |
oons | +0.013 |
flour | +0.013 |
cakes | +0.013 |
tables | +0.012 |
Neg logit contributions | |
boo | -0.018 |
hero | -0.018 |
sing | -0.018 |
heel | -0.018 |
recovery | -0.018 |
Token | cleaning |
Token ID | 12724 |
Feature activation | +2.30e-02 |
Pos logit contributions | |
cakes | +0.155 |
ups | +0.152 |
flour | +0.152 |
oons | +0.151 |
tables | +0.150 |
Neg logit contributions | |
recovery | -0.253 |
promise | -0.252 |
badly | -0.251 |
boo | -0.250 |
heel | -0.245 |
Token | supplies |
Token ID | 9416 |
Feature activation | +1.23e-01 |
Pos logit contributions | |
cakes | +0.034 |
flour | +0.033 |
ups | +0.033 |
glers | +0.033 |
oons | +0.032 |
Neg logit contributions | |
badly | -0.035 |
knee | -0.034 |
ankle | -0.033 |
heel | -0.033 |
symp | -0.033 |
Token | and |
Token ID | 290 |
Feature activation | +5.02e-03 |
Pos logit contributions | |
flour | +0.203 |
oons | +0.187 |
tables | +0.187 |
ingredients | +0.181 |
pots | +0.181 |
Neg logit contributions | |
badly | -0.214 |
sing | -0.209 |
symp | -0.204 |
promise | -0.203 |
recovery | -0.203 |
Token | began |
Token ID | 2540 |
Feature activation | +1.51e-02 |
Pos logit contributions | |
oons | +0.007 |
flour | +0.007 |
ries | +0.007 |
cakes | +0.007 |
pots | +0.007 |
Neg logit contributions | |
boo | -0.009 |
heel | -0.008 |
sing | -0.008 |
badly | -0.008 |
recovery | -0.008 |
Token | to |
Token ID | 284 |
Feature activation | -3.57e-02 |
Pos logit contributions | |
flour | +0.022 |
oons | +0.020 |
ingredients | +0.020 |
tables | +0.020 |
pots | +0.020 |
Neg logit contributions | |
sing | -0.029 |
recovery | -0.029 |
heel | -0.028 |
badly | -0.028 |
boo | -0.028 |
Token | scrub |
Token ID | 27268 |
Feature activation | +3.08e-02 |
Pos logit contributions | |
sing | +0.063 |
recovery | +0.061 |
badly | +0.061 |
heel | +0.060 |
symp | +0.060 |
Neg logit contributions | |
flour | -0.062 |
tables | -0.056 |
ingredients | -0.056 |
pots | -0.055 |
oons | -0.053 |
Token | the |
Token ID | 262 |
Feature activation | +1.78e-02 |
Pos logit contributions | |
cakes | +0.059 |
flour | +0.058 |
ingredients | +0.053 |
oons | +0.052 |
pots | +0.052 |
Neg logit contributions | |
symp | -0.044 |
heel | -0.043 |
badly | -0.043 |
boo | -0.043 |
recovery | -0.041 |
Token | at |
Token ID | 379 |
Feature activation | -1.03e-01 |
Pos logit contributions | |
badly | +0.100 |
sing | +0.100 |
recovery | +0.098 |
heel | +0.097 |
symp | +0.096 |
Neg logit contributions | |
flour | -0.081 |
oons | -0.075 |
ingredients | -0.075 |
tables | -0.073 |
pots | -0.072 |
Token | the |
Token ID | 262 |
Feature activation | +2.44e-02 |
Pos logit contributions | |
sing | +0.222 |
heel | +0.219 |
recovery | +0.218 |
boo | +0.216 |
badly | +0.215 |
Neg logit contributions | |
flour | -0.131 |
oons | -0.120 |
ingredients | -0.118 |
cakes | -0.114 |
tables | -0.113 |
Token | water |
Token ID | 1660 |
Feature activation | +2.06e-02 |
Pos logit contributions | |
flour | +0.036 |
oons | +0.034 |
ries | +0.033 |
ingredients | +0.033 |
tables | +0.032 |
Neg logit contributions | |
sing | -0.048 |
heel | -0.046 |
recovery | -0.046 |
boo | -0.045 |
spine | -0.044 |
Token | hose |
Token ID | 31489 |
Feature activation | +6.26e-02 |
Pos logit contributions | |
flour | +0.031 |
oons | +0.029 |
ingredients | +0.028 |
tables | +0.028 |
ries | +0.027 |
Neg logit contributions | |
sing | -0.039 |
heel | -0.038 |
recovery | -0.038 |
badly | -0.037 |
spine | -0.037 |
Token | and |
Token ID | 290 |
Feature activation | +2.02e-02 |
Pos logit contributions | |
flour | +0.094 |
oons | +0.087 |
ingredients | +0.085 |
tables | +0.084 |
pans | +0.082 |
Neg logit contributions | |
sing | -0.120 |
heel | -0.116 |
badly | -0.116 |
recovery | -0.114 |
symp | -0.113 |
Token | turned |
Token ID | 2900 |
Feature activation | +1.23e-01 |
Pos logit contributions | |
flour | +0.031 |
oons | +0.031 |
ingredients | +0.028 |
cakes | +0.028 |
ries | +0.028 |
Neg logit contributions | |
sing | -0.038 |
heel | -0.037 |
boo | -0.037 |
badly | -0.034 |
promise | -0.034 |
Token | the |
Token ID | 262 |
Feature activation | +5.96e-02 |
Pos logit contributions | |
flour | +0.174 |
oons | +0.168 |
cakes | +0.163 |
tables | +0.162 |
ingredients | +0.161 |
Neg logit contributions | |
badly | -0.234 |
boo | -0.229 |
heel | -0.228 |
symp | -0.228 |
sing | -0.220 |
Token | handle |
Token ID | 5412 |
Feature activation | +3.77e-02 |
Pos logit contributions | |
oons | +0.073 |
ries | +0.072 |
flour | +0.072 |
cakes | +0.068 |
tables | +0.067 |
Neg logit contributions | |
sing | -0.126 |
promise | -0.121 |
boo | -0.120 |
heel | -0.120 |
spine | -0.119 |
Token | . |
Token ID | 13 |
Feature activation | -5.35e-03 |
Pos logit contributions | |
flour | +0.053 |
ingredients | +0.048 |
tables | +0.048 |
oons | +0.047 |
pots | +0.046 |
Neg logit contributions | |
sing | -0.076 |
badly | -0.074 |
heel | -0.072 |
symp | -0.072 |
recovery | -0.070 |
Token | The |
Token ID | 383 |
Feature activation | +5.35e-02 |
Pos logit contributions | |
sing | +0.008 |
recovery | +0.008 |
heel | +0.008 |
badly | +0.008 |
boo | +0.008 |
Neg logit contributions | |
flour | -0.010 |
ingredients | -0.010 |
tables | -0.010 |
oons | -0.009 |
pots | -0.009 |
Token | pot |
Token ID | 1787 |
Feature activation | +5.43e-02 |
Pos logit contributions | |
flour | +0.060 |
oons | +0.058 |
ries | +0.056 |
tables | +0.054 |
cakes | +0.053 |
Neg logit contributions | |
sing | -0.122 |
heel | -0.117 |
boo | -0.116 |
promise | -0.116 |
recovery | -0.115 |
Token | They |
Token ID | 1119 |
Feature activation | +1.23e-02 |
Pos logit contributions | |
sing | +0.053 |
recovery | +0.051 |
heel | +0.051 |
badly | +0.050 |
symp | +0.050 |
Neg logit contributions | |
flour | -0.042 |
ingredients | -0.039 |
oons | -0.038 |
tables | -0.037 |
pots | -0.037 |
Token | pretended |
Token ID | 36766 |
Feature activation | +9.87e-02 |
Pos logit contributions | |
flour | +0.019 |
ingredients | +0.018 |
oons | +0.018 |
ries | +0.017 |
tables | +0.017 |
Neg logit contributions | |
sing | -0.023 |
heel | -0.023 |
boo | -0.022 |
promise | -0.022 |
symp | -0.022 |
Token | to |
Token ID | 284 |
Feature activation | -1.89e-02 |
Pos logit contributions | |
flour | +0.154 |
oons | +0.150 |
pans | +0.143 |
ingredients | +0.142 |
tables | +0.135 |
Neg logit contributions | |
sing | -0.180 |
badly | -0.177 |
heel | -0.170 |
recovery | -0.168 |
boo | -0.165 |
Token | play |
Token ID | 711 |
Feature activation | -6.98e-02 |
Pos logit contributions | |
symp | +0.029 |
agging | +0.029 |
sing | +0.028 |
aring | +0.027 |
badly | +0.027 |
Neg logit contributions | |
flour | -0.035 |
tables | -0.034 |
pots | -0.034 |
cans | -0.032 |
chairs | -0.032 |
Token | with |
Token ID | 351 |
Feature activation | +1.32e-02 |
Pos logit contributions | |
sing | +0.146 |
heel | +0.139 |
recovery | +0.139 |
boo | +0.139 |
symp | +0.138 |
Neg logit contributions | |
flour | -0.099 |
tables | -0.091 |
pots | -0.088 |
ingredients | -0.088 |
oons | -0.087 |
Token | some |
Token ID | 617 |
Feature activation | +1.29e-01 |
Pos logit contributions | |
cakes | +0.017 |
oons | +0.017 |
ingredients | +0.015 |
ups | +0.015 |
flour | +0.015 |
Neg logit contributions | |
badly | -0.026 |
sing | -0.026 |
heel | -0.025 |
boo | -0.025 |
recovery | -0.025 |
Token | dolls |
Token ID | 36062 |
Feature activation | +6.87e-02 |
Pos logit contributions | |
flour | +0.146 |
oons | +0.130 |
ingredients | +0.129 |
tables | +0.125 |
pots | +0.122 |
Neg logit contributions | |
sing | -0.285 |
recovery | -0.284 |
heel | -0.284 |
boo | -0.280 |
badly | -0.279 |
Token | and |
Token ID | 290 |
Feature activation | +7.21e-03 |
Pos logit contributions | |
flour | +0.112 |
ingredients | +0.102 |
oons | +0.101 |
tables | +0.098 |
pans | +0.096 |
Neg logit contributions | |
sing | -0.125 |
heel | -0.122 |
badly | -0.121 |
recovery | -0.119 |
hero | -0.119 |
Token | cars |
Token ID | 5006 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
flour | +0.010 |
oons | +0.009 |
ingredients | +0.009 |
ries | +0.008 |
cakes | +0.008 |
Neg logit contributions | |
sing | -0.015 |
heel | -0.015 |
boo | -0.015 |
recovery | -0.014 |
promise | -0.014 |
Token | . |
Token ID | 13 |
Feature activation | -5.50e-03 |
Pos logit contributions | |
flour | +0.156 |
ingredients | +0.139 |
tables | +0.136 |
oons | +0.136 |
pots | +0.133 |
Neg logit contributions | |
sing | -0.237 |
recovery | -0.230 |
heel | -0.229 |
badly | -0.228 |
boo | -0.226 |
Token | They |
Token ID | 1119 |
Feature activation | +2.50e-02 |
Pos logit contributions | |
sing | +0.010 |
recovery | +0.010 |
heel | +0.010 |
symp | +0.010 |
former | +0.010 |
Neg logit contributions | |
flour | -0.009 |
ingredients | -0.008 |
oons | -0.008 |
ries | -0.008 |
pans | -0.008 |
Token | big |
Token ID | 1263 |
Feature activation | +3.58e-02 |
Pos logit contributions | |
sing | +0.117 |
recovery | +0.113 |
heel | +0.112 |
badly | +0.110 |
boo | +0.110 |
Neg logit contributions | |
flour | -0.116 |
ingredients | -0.106 |
tables | -0.105 |
oons | -0.104 |
pots | -0.103 |
Token | and |
Token ID | 290 |
Feature activation | -6.15e-02 |
Pos logit contributions | |
flour | +0.068 |
oons | +0.065 |
cakes | +0.062 |
ingredients | +0.061 |
ries | +0.061 |
Neg logit contributions | |
hero | -0.053 |
badly | -0.053 |
heel | -0.052 |
sing | -0.050 |
boo | -0.050 |
Token | had |
Token ID | 550 |
Feature activation | +3.30e-02 |
Pos logit contributions | |
badly | +0.096 |
sing | +0.095 |
boo | +0.092 |
heel | +0.091 |
promise | +0.090 |
Neg logit contributions | |
oons | -0.110 |
flour | -0.106 |
ingredients | -0.105 |
ries | -0.104 |
ups | -0.101 |
Token | lots |
Token ID | 6041 |
Feature activation | +8.51e-02 |
Pos logit contributions | |
oons | +0.051 |
ries | +0.050 |
flour | +0.046 |
pans | +0.045 |
liquids | +0.044 |
Neg logit contributions | |
badly | -0.055 |
heel | -0.055 |
hero | -0.055 |
recovery | -0.053 |
roar | -0.053 |
Token | of |
Token ID | 286 |
Feature activation | +1.02e-02 |
Pos logit contributions | |
flour | +0.133 |
ports | +0.128 |
pans | +0.128 |
pots | +0.126 |
ries | +0.123 |
Neg logit contributions | |
badly | -0.140 |
recovery | -0.137 |
heel | -0.132 |
hero | -0.131 |
agging | -0.123 |
Token | tables |
Token ID | 8893 |
Feature activation | +1.52e-01 |
Pos logit contributions | |
ries | +0.015 |
flour | +0.015 |
cakes | +0.014 |
oons | +0.014 |
ups | +0.014 |
Neg logit contributions | |
wish | -0.017 |
badly | -0.016 |
roar | -0.016 |
promise | -0.016 |
heel | -0.016 |
Token | and |
Token ID | 290 |
Feature activation | -7.00e-04 |
Pos logit contributions | |
flour | +0.244 |
ingredients | +0.225 |
oons | +0.220 |
pans | +0.212 |
tables | +0.208 |
Neg logit contributions | |
badly | -0.276 |
sing | -0.275 |
spine | -0.262 |
heel | -0.261 |
recovery | -0.257 |
Token | chairs |
Token ID | 18791 |
Feature activation | +1.53e-01 |
Pos logit contributions | |
badly | +0.001 |
heel | +0.001 |
former | +0.001 |
roar | +0.001 |
boo | +0.001 |
Neg logit contributions | |
oons | -0.001 |
ries | -0.001 |
ures | -0.001 |
flour | -0.001 |
ingredients | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | +1.03e-02 |
Pos logit contributions | |
flour | +0.231 |
ingredients | +0.209 |
oons | +0.201 |
pans | +0.200 |
tables | +0.195 |
Neg logit contributions | |
sing | -0.295 |
badly | -0.291 |
recovery | -0.279 |
spine | -0.278 |
heel | -0.277 |
Token |
|
Token ID | 198 |
Feature activation | -2.03e-02 |
Pos logit contributions | |
flour | +0.020 |
ries | +0.020 |
oons | +0.019 |
pans | +0.019 |
ingredients | +0.019 |
Neg logit contributions | |
symp | -0.014 |
heel | -0.013 |
boo | -0.013 |
badly | -0.013 |
sing | -0.013 |
Token |
|
Token ID | 198 |
Feature activation | -2.88e-02 |
Pos logit contributions | |
badly | +0.029 |
recovery | +0.027 |
symp | +0.027 |
promise | +0.027 |
sing | +0.026 |
Neg logit contributions | |
flour | -0.040 |
ingredients | -0.038 |
ries | -0.038 |
tables | -0.038 |
pans | -0.037 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.00e-01 |
Pos logit contributions | |
flour | +0.181 |
tables | +0.162 |
oons | +0.161 |
ingredients | +0.161 |
pots | +0.156 |
Neg logit contributions | |
sing | -0.259 |
badly | -0.251 |
recovery | -0.250 |
heel | -0.250 |
boo | -0.244 |
Token | |
Token ID | 220 |
Feature activation | +4.18e-02 |
Pos logit contributions | |
flour | +0.132 |
ingredients | +0.117 |
tables | +0.117 |
oons | +0.114 |
pots | +0.113 |
Neg logit contributions | |
sing | -0.217 |
recovery | -0.211 |
heel | -0.208 |
boo | -0.206 |
badly | -0.206 |
Token |
|
Token ID | 198 |
Feature activation | -1.92e-02 |
Pos logit contributions | |
flour | +0.070 |
tables | +0.064 |
ingredients | +0.064 |
pots | +0.063 |
oons | +0.061 |
Neg logit contributions | |
sing | -0.073 |
badly | -0.071 |
symp | -0.071 |
recovery | -0.071 |
heel | -0.071 |
Token |
|
Token ID | 198 |
Feature activation | -2.41e-02 |
Pos logit contributions | |
sing | +0.044 |
spine | +0.042 |
recovery | +0.042 |
boo | +0.042 |
heel | +0.041 |
Neg logit contributions | |
flour | -0.024 |
oons | -0.021 |
tables | -0.020 |
ingredients | -0.020 |
pots | -0.019 |
Token | Suddenly |
Token ID | 38582 |
Feature activation | +9.49e-02 |
Pos logit contributions | |
sing | +0.055 |
spine | +0.052 |
boo | +0.050 |
injury | +0.050 |
recovery | +0.049 |
Neg logit contributions | |
flour | -0.032 |
tables | -0.029 |
oons | -0.028 |
ups | -0.027 |
ingredients | -0.027 |
Token | , |
Token ID | 11 |
Feature activation | +6.52e-02 |
Pos logit contributions | |
flour | +0.126 |
ingredients | +0.113 |
oons | +0.113 |
tables | +0.111 |
pots | +0.109 |
Neg logit contributions | |
sing | -0.200 |
heel | -0.195 |
recovery | -0.194 |
badly | -0.194 |
boo | -0.192 |
Token | they |
Token ID | 484 |
Feature activation | +6.11e-02 |
Pos logit contributions | |
flour | +0.081 |
oons | +0.074 |
ingredients | +0.073 |
tables | +0.070 |
pots | +0.070 |
Neg logit contributions | |
sing | -0.142 |
heel | -0.139 |
badly | -0.139 |
recovery | -0.137 |
boo | -0.137 |
Token | heard |
Token ID | 2982 |
Feature activation | +9.51e-02 |
Pos logit contributions | |
flour | +0.092 |
ingredients | +0.085 |
oons | +0.084 |
cakes | +0.078 |
tables | +0.077 |
Neg logit contributions | |
sing | -0.122 |
symp | -0.118 |
heel | -0.118 |
badly | -0.116 |
tweet | -0.116 |
Token | a |
Token ID | 257 |
Feature activation | +4.23e-02 |
Pos logit contributions | |
flour | +0.123 |
ingredients | +0.111 |
oons | +0.111 |
tables | +0.107 |
pots | +0.106 |
Neg logit contributions | |
sing | -0.204 |
recovery | -0.199 |
heel | -0.197 |
badly | -0.197 |
boo | -0.195 |
Token | loud |
Token ID | 7812 |
Feature activation | -8.35e-02 |
Pos logit contributions | |
flour | +0.068 |
pans | +0.065 |
tables | +0.063 |
pots | +0.061 |
chairs | +0.059 |
Neg logit contributions | |
amb | -0.080 |
former | -0.077 |
spine | -0.076 |
recovery | -0.074 |
agging | -0.074 |
Token | noise |
Token ID | 7838 |
Feature activation | -2.20e-02 |
Pos logit contributions | |
former | +0.167 |
recovery | +0.146 |
spine | +0.145 |
uer | +0.145 |
prisoner | +0.145 |
Neg logit contributions | |
flour | -0.129 |
tables | -0.125 |
pots | -0.121 |
pans | -0.116 |
chairs | -0.113 |
Token | and |
Token ID | 290 |
Feature activation | +1.07e-02 |
Pos logit contributions | |
flour | +0.056 |
oons | +0.053 |
ingredients | +0.051 |
pots | +0.050 |
tables | +0.049 |
Neg logit contributions | |
badly | -0.066 |
sing | -0.065 |
recovery | -0.063 |
heel | -0.063 |
symp | -0.061 |
Token | months |
Token ID | 1933 |
Feature activation | +1.37e-01 |
Pos logit contributions | |
flour | +0.016 |
oons | +0.014 |
ingredients | +0.014 |
tables | +0.014 |
pots | +0.014 |
Neg logit contributions | |
sing | -0.021 |
heel | -0.021 |
recovery | -0.021 |
badly | -0.020 |
boo | -0.020 |
Token | . |
Token ID | 13 |
Feature activation | -7.20e-03 |
Pos logit contributions | |
flour | +0.204 |
ingredients | +0.183 |
oons | +0.180 |
pots | +0.178 |
tables | +0.175 |
Neg logit contributions | |
recovery | -0.263 |
badly | -0.261 |
sing | -0.257 |
heel | -0.255 |
symp | -0.253 |
Token | Fl |
Token ID | 1610 |
Feature activation | -3.93e-02 |
Pos logit contributions | |
sing | +0.014 |
boo | +0.013 |
recovery | +0.013 |
heel | +0.013 |
promise | +0.013 |
Neg logit contributions | |
flour | -0.011 |
tables | -0.010 |
ingredients | -0.010 |
oons | -0.010 |
pots | -0.009 |
Token | uffy |
Token ID | 15352 |
Feature activation | -3.41e-03 |
Pos logit contributions | |
recovery | +0.066 |
heel | +0.065 |
berth | +0.064 |
tweet | +0.064 |
spine | +0.064 |
Neg logit contributions | |
flour | -0.064 |
oons | -0.063 |
ingredients | -0.058 |
ries | -0.057 |
pots | -0.055 |
Token | thought |
Token ID | 1807 |
Feature activation | +3.37e-02 |
Pos logit contributions | |
sing | +0.008 |
recovery | +0.007 |
heel | +0.007 |
badly | +0.007 |
symp | +0.007 |
Neg logit contributions | |
flour | -0.004 |
ingredients | -0.004 |
tables | -0.004 |
oons | -0.004 |
pots | -0.004 |
Token | it |
Token ID | 340 |
Feature activation | +5.78e-02 |
Pos logit contributions | |
flour | +0.043 |
tables | +0.038 |
ingredients | +0.038 |
oons | +0.037 |
pots | +0.037 |
Neg logit contributions | |
sing | -0.076 |
recovery | -0.072 |
heel | -0.072 |
symp | -0.071 |
former | -0.071 |
Token | was |
Token ID | 373 |
Feature activation | -9.20e-04 |
Pos logit contributions | |
flour | +0.069 |
tables | +0.060 |
ingredients | +0.060 |
oons | +0.060 |
pots | +0.058 |
Neg logit contributions | |
sing | -0.131 |
heel | -0.127 |
recovery | -0.127 |
badly | -0.127 |
symp | -0.126 |
Token | very |
Token ID | 845 |
Feature activation | -8.27e-02 |
Pos logit contributions | |
sing | +0.002 |
badly | +0.002 |
heel | +0.002 |
boo | +0.001 |
recovery | +0.001 |
Neg logit contributions | |
flour | -0.002 |
oons | -0.002 |
tables | -0.001 |
ingredients | -0.001 |
ries | -0.001 |
Token | interesting |
Token ID | 3499 |
Feature activation | -1.56e-02 |
Pos logit contributions | |
sing | +0.124 |
badly | +0.120 |
heel | +0.119 |
recovery | +0.119 |
boo | +0.117 |
Neg logit contributions | |
flour | -0.162 |
ingredients | -0.151 |
tables | -0.150 |
oons | -0.149 |
pots | -0.147 |
Token | and |
Token ID | 290 |
Feature activation | -5.16e-02 |
Pos logit contributions | |
badly | +0.028 |
sing | +0.027 |
heel | +0.027 |
recovery | +0.027 |
hero | +0.027 |
Neg logit contributions | |
flour | -0.025 |
oons | -0.023 |
tables | -0.023 |
pots | -0.022 |
pans | -0.022 |
Token | a |
Token ID | 257 |
Feature activation | +2.29e-02 |
Pos logit contributions | |
oons | +0.080 |
pans | +0.076 |
liquids | +0.073 |
ingredients | +0.072 |
flour | +0.069 |
Neg logit contributions | |
badly | -0.060 |
boo | -0.060 |
heel | -0.059 |
bye | -0.057 |
sing | -0.056 |
Token | thick |
Token ID | 6546 |
Feature activation | -6.68e-02 |
Pos logit contributions | |
flour | +0.032 |
tables | +0.028 |
ingredients | +0.027 |
pots | +0.027 |
chairs | +0.026 |
Neg logit contributions | |
sing | -0.048 |
symp | -0.048 |
recovery | -0.048 |
former | -0.047 |
heel | -0.046 |
Token | book |
Token ID | 1492 |
Feature activation | -7.92e-02 |
Pos logit contributions | |
sing | +0.122 |
former | +0.119 |
symp | +0.119 |
recovery | +0.118 |
badly | +0.116 |
Neg logit contributions | |
flour | -0.112 |
tables | -0.103 |
ingredients | -0.100 |
pots | -0.099 |
oons | -0.096 |
Token | on |
Token ID | 319 |
Feature activation | -5.93e-02 |
Pos logit contributions | |
sing | +0.147 |
badly | +0.143 |
heel | +0.141 |
recovery | +0.140 |
symp | +0.138 |
Neg logit contributions | |
flour | -0.127 |
ingredients | -0.117 |
oons | -0.114 |
tables | -0.114 |
pots | -0.113 |
Token | the |
Token ID | 262 |
Feature activation | +1.13e-02 |
Pos logit contributions | |
heel | +0.123 |
badly | +0.123 |
recovery | +0.120 |
sing | +0.118 |
boo | +0.116 |
Neg logit contributions | |
ingredients | -0.074 |
oons | -0.073 |
cakes | -0.071 |
ups | -0.071 |
flour | -0.069 |
Token | top |
Token ID | 1353 |
Feature activation | +3.30e-02 |
Pos logit contributions | |
flour | +0.016 |
oons | +0.015 |
ingredients | +0.015 |
ries | +0.015 |
pots | +0.014 |
Neg logit contributions | |
sing | -0.023 |
heel | -0.022 |
recovery | -0.022 |
spine | -0.022 |
boo | -0.021 |
Token | shelf |
Token ID | 18316 |
Feature activation | +4.20e-02 |
Pos logit contributions | |
flour | +0.048 |
ingredients | +0.045 |
pans | +0.043 |
oons | +0.043 |
pots | +0.043 |
Neg logit contributions | |
sing | -0.062 |
recovery | -0.061 |
badly | -0.061 |
spine | -0.061 |
heel | -0.060 |
Token | . |
Token ID | 13 |
Feature activation | -1.19e-02 |
Pos logit contributions | |
flour | +0.059 |
ingredients | +0.053 |
oons | +0.053 |
pans | +0.052 |
tables | +0.052 |
Neg logit contributions | |
sing | -0.087 |
badly | -0.084 |
spine | -0.082 |
recovery | -0.082 |
heel | -0.081 |
Token | She |
Token ID | 1375 |
Feature activation | -4.43e-02 |
Pos logit contributions | |
heel | +0.014 |
bye | +0.014 |
ree | +0.013 |
sing | +0.013 |
ighed | +0.013 |
Neg logit contributions | |
flour | -0.024 |
ingredients | -0.024 |
pans | -0.023 |
pots | -0.023 |
liquids | -0.022 |
Token | wanted |
Token ID | 2227 |
Feature activation | -8.68e-05 |
Pos logit contributions | |
sing | +0.086 |
heel | +0.085 |
boo | +0.085 |
badly | +0.084 |
symp | +0.082 |
Neg logit contributions | |
flour | -0.067 |
ingredients | -0.060 |
oons | -0.059 |
tables | -0.059 |
pots | -0.059 |
Token | to |
Token ID | 284 |
Feature activation | -4.97e-02 |
Pos logit contributions | |
heel | +0.000 |
badly | +0.000 |
boo | +0.000 |
sing | +0.000 |
roar | +0.000 |
Neg logit contributions | |
oons | -0.000 |
flour | -0.000 |
pans | -0.000 |
glers | -0.000 |
cakes | -0.000 |
Token | so |
Token ID | 523 |
Feature activation | +1.84e-02 |
Pos logit contributions | |
flour | +0.046 |
ingredients | +0.042 |
tables | +0.041 |
oons | +0.041 |
pots | +0.041 |
Neg logit contributions | |
sing | -0.055 |
recovery | -0.053 |
heel | -0.053 |
badly | -0.052 |
boo | -0.052 |
Token | they |
Token ID | 484 |
Feature activation | +4.19e-02 |
Pos logit contributions | |
flour | +0.019 |
oons | +0.019 |
ingredients | +0.018 |
tables | +0.017 |
pots | +0.017 |
Neg logit contributions | |
badly | -0.044 |
sing | -0.044 |
heel | -0.043 |
recovery | -0.042 |
former | -0.041 |
Token | buried |
Token ID | 11694 |
Feature activation | +3.53e-02 |
Pos logit contributions | |
flour | +0.058 |
ingredients | +0.052 |
oons | +0.051 |
tables | +0.051 |
pots | +0.050 |
Neg logit contributions | |
sing | -0.087 |
heel | -0.084 |
recovery | -0.084 |
badly | -0.084 |
boo | -0.084 |
Token | it |
Token ID | 340 |
Feature activation | +7.11e-03 |
Pos logit contributions | |
oons | +0.055 |
cakes | +0.052 |
ups | +0.052 |
flour | +0.050 |
tables | +0.048 |
Neg logit contributions | |
badly | -0.064 |
sing | -0.061 |
symp | -0.058 |
promise | -0.057 |
heel | -0.056 |
Token | in |
Token ID | 287 |
Feature activation | -3.19e-02 |
Pos logit contributions | |
flour | +0.010 |
tables | +0.009 |
ingredients | +0.009 |
oons | +0.009 |
pots | +0.009 |
Neg logit contributions | |
sing | -0.014 |
badly | -0.014 |
heel | -0.014 |
recovery | -0.014 |
symp | -0.013 |
Token | the |
Token ID | 262 |
Feature activation | +2.15e-02 |
Pos logit contributions | |
badly | +0.066 |
sing | +0.064 |
recovery | +0.063 |
former | +0.063 |
promise | +0.062 |
Neg logit contributions | |
cakes | -0.044 |
oons | -0.044 |
ups | -0.042 |
boards | -0.039 |
flour | -0.039 |
Token | hole |
Token ID | 7604 |
Feature activation | -4.41e-02 |
Pos logit contributions | |
flour | +0.035 |
oons | +0.034 |
ries | +0.034 |
ingredients | +0.032 |
cakes | +0.032 |
Neg logit contributions | |
sing | -0.038 |
recovery | -0.037 |
heel | -0.037 |
hero | -0.037 |
promise | -0.037 |
Token | . |
Token ID | 13 |
Feature activation | -1.61e-02 |
Pos logit contributions | |
sing | +0.088 |
badly | +0.085 |
spine | +0.081 |
recovery | +0.080 |
heel | +0.080 |
Neg logit contributions | |
flour | -0.065 |
tables | -0.059 |
oons | -0.058 |
ingredients | -0.057 |
pans | -0.056 |
Token |
|
Token ID | 198 |
Feature activation | -4.10e-02 |
Pos logit contributions | |
sing | +0.027 |
heel | +0.026 |
symp | +0.026 |
badly | +0.026 |
recovery | +0.026 |
Neg logit contributions | |
flour | -0.029 |
oons | -0.027 |
ingredients | -0.027 |
pots | -0.026 |
tables | -0.026 |
Token |
|
Token ID | 198 |
Feature activation | -4.70e-02 |
Pos logit contributions | |
sing | +0.075 |
recovery | +0.073 |
badly | +0.072 |
heel | +0.072 |
symp | +0.072 |
Neg logit contributions | |
flour | -0.067 |
ingredients | -0.061 |
tables | -0.060 |
pots | -0.060 |
oons | -0.059 |
Token | Later |
Token ID | 18602 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
sing | +0.094 |
spine | +0.089 |
boo | +0.089 |
recovery | +0.089 |
heel | +0.087 |
Neg logit contributions | |
flour | -0.073 |
tables | -0.066 |
oons | -0.065 |
ingredients | -0.064 |
pots | -0.062 |
Token | quiet |
Token ID | 5897 |
Feature activation | -8.23e-02 |
Pos logit contributions | |
flour | +0.062 |
pots | +0.055 |
ingredients | +0.055 |
tables | +0.053 |
pans | +0.052 |
Neg logit contributions | |
symp | -0.075 |
utt | -0.071 |
ighed | -0.070 |
sing | -0.070 |
agging | -0.069 |
Token | and |
Token ID | 290 |
Feature activation | -5.70e-02 |
Pos logit contributions | |
sing | +0.152 |
recovery | +0.146 |
heel | +0.145 |
symp | +0.144 |
badly | +0.144 |
Neg logit contributions | |
flour | -0.134 |
ingredients | -0.122 |
tables | -0.121 |
oons | -0.120 |
pots | -0.119 |
Token | peaceful |
Token ID | 12309 |
Feature activation | -1.03e-01 |
Pos logit contributions | |
sing | +0.097 |
recovery | +0.094 |
heel | +0.093 |
symp | +0.093 |
badly | +0.092 |
Neg logit contributions | |
flour | -0.101 |
ingredients | -0.092 |
tables | -0.092 |
pots | -0.091 |
oons | -0.089 |
Token | place |
Token ID | 1295 |
Feature activation | -8.30e-03 |
Pos logit contributions | |
sing | +0.183 |
symp | +0.176 |
badly | +0.174 |
heel | +0.174 |
recovery | +0.174 |
Neg logit contributions | |
flour | -0.173 |
tables | -0.161 |
ingredients | -0.160 |
pots | -0.158 |
oons | -0.153 |
Token | to |
Token ID | 284 |
Feature activation | -4.98e-02 |
Pos logit contributions | |
sing | +0.016 |
badly | +0.016 |
recovery | +0.015 |
promise | +0.015 |
heel | +0.015 |
Neg logit contributions | |
flour | -0.012 |
oons | -0.011 |
tables | -0.011 |
ingredients | -0.011 |
pots | -0.011 |
Token | work |
Token ID | 670 |
Feature activation | +2.89e-02 |
Pos logit contributions | |
sing | +0.079 |
symp | +0.078 |
badly | +0.075 |
heel | +0.074 |
spine | +0.073 |
Neg logit contributions | |
flour | -0.091 |
tables | -0.088 |
pots | -0.086 |
oons | -0.081 |
ingredients | -0.081 |
Token | ." |
Token ID | 526 |
Feature activation | +4.14e-02 |
Pos logit contributions | |
flour | +0.043 |
ingredients | +0.038 |
tables | +0.038 |
ries | +0.038 |
oons | +0.038 |
Neg logit contributions | |
badly | -0.055 |
sing | -0.053 |
recovery | -0.052 |
promise | -0.052 |
hero | -0.051 |
Token |
|
Token ID | 198 |
Feature activation | -3.41e-02 |
Pos logit contributions | |
flour | +0.061 |
ingredients | +0.055 |
tables | +0.055 |
oons | +0.054 |
pots | +0.053 |
Neg logit contributions | |
sing | -0.084 |
recovery | -0.081 |
heel | -0.080 |
badly | -0.079 |
boo | -0.079 |
Token |
|
Token ID | 198 |
Feature activation | -3.95e-02 |
Pos logit contributions | |
sing | +0.065 |
recovery | +0.064 |
heel | +0.063 |
badly | +0.063 |
boo | +0.062 |
Neg logit contributions | |
flour | -0.053 |
ingredients | -0.048 |
tables | -0.048 |
oons | -0.047 |
pots | -0.047 |
Token | " |
Token ID | 1 |
Feature activation | +6.38e-02 |
Pos logit contributions | |
sing | +0.085 |
spine | +0.078 |
recovery | +0.078 |
boo | +0.077 |
heel | +0.076 |
Neg logit contributions | |
flour | -0.058 |
tables | -0.052 |
oons | -0.050 |
ingredients | -0.050 |
ups | -0.049 |
Token | Can |
Token ID | 6090 |
Feature activation | -3.31e-02 |
Pos logit contributions | |
flour | +0.109 |
ingredients | +0.099 |
tables | +0.098 |
oons | +0.097 |
pots | +0.096 |
Neg logit contributions | |
sing | -0.115 |
recovery | -0.109 |
heel | -0.109 |
badly | -0.109 |
boo | -0.108 |
Token | wanted |
Token ID | 2227 |
Feature activation | +9.17e-03 |
Pos logit contributions | |
flour | +0.035 |
oons | +0.033 |
ingredients | +0.033 |
ries | +0.031 |
tables | +0.031 |
Neg logit contributions | |
promise | -0.039 |
sing | -0.039 |
heel | -0.038 |
boo | -0.038 |
badly | -0.038 |
Token | to |
Token ID | 284 |
Feature activation | -4.39e-02 |
Pos logit contributions | |
oons | +0.020 |
pans | +0.018 |
cakes | +0.018 |
flour | +0.018 |
ries | +0.018 |
Neg logit contributions | |
badly | -0.010 |
heel | -0.010 |
boo | -0.010 |
sing | -0.010 |
roar | -0.010 |
Token | see |
Token ID | 766 |
Feature activation | +4.04e-02 |
Pos logit contributions | |
sing | +0.064 |
recovery | +0.062 |
heel | +0.061 |
badly | +0.060 |
boo | +0.060 |
Neg logit contributions | |
flour | -0.089 |
ingredients | -0.082 |
tables | -0.082 |
oons | -0.081 |
pots | -0.080 |
Token | who |
Token ID | 508 |
Feature activation | +2.58e-02 |
Pos logit contributions | |
oons | +0.087 |
cakes | +0.086 |
boards | +0.079 |
ingredients | +0.079 |
ups | +0.076 |
Neg logit contributions | |
badly | -0.052 |
promise | -0.049 |
hero | -0.046 |
heel | -0.045 |
wish | -0.044 |
Token | could |
Token ID | 714 |
Feature activation | -7.77e-03 |
Pos logit contributions | |
flour | +0.037 |
ingredients | +0.035 |
oons | +0.034 |
tables | +0.034 |
pans | +0.032 |
Neg logit contributions | |
badly | -0.050 |
promise | -0.050 |
sing | -0.050 |
hero | -0.049 |
heel | -0.049 |
Token | jump |
Token ID | 4391 |
Feature activation | -3.97e-03 |
Pos logit contributions | |
symp | +0.013 |
sing | +0.013 |
spine | +0.013 |
recovery | +0.013 |
former | +0.013 |
Neg logit contributions | |
flour | -0.013 |
tables | -0.012 |
pots | -0.012 |
oons | -0.012 |
cans | -0.011 |
Token | over |
Token ID | 625 |
Feature activation | +5.96e-02 |
Pos logit contributions | |
sing | +0.008 |
heel | +0.008 |
recovery | +0.008 |
badly | +0.008 |
boo | +0.008 |
Neg logit contributions | |
flour | -0.006 |
ingredients | -0.005 |
tables | -0.005 |
oons | -0.005 |
pots | -0.005 |
Token | the |
Token ID | 262 |
Feature activation | +3.99e-02 |
Pos logit contributions | |
oons | +0.074 |
flour | +0.071 |
cakes | +0.069 |
ingredients | +0.069 |
pots | +0.060 |
Neg logit contributions | |
badly | -0.127 |
heel | -0.121 |
hero | -0.117 |
symp | -0.115 |
recovery | -0.115 |
Token | mailbox |
Token ID | 37282 |
Feature activation | -1.55e-02 |
Pos logit contributions | |
oons | +0.084 |
ries | +0.077 |
liquids | +0.075 |
cakes | +0.074 |
flour | +0.074 |
Neg logit contributions | |
sing | -0.054 |
hero | -0.053 |
promise | -0.052 |
badly | -0.050 |
threat | -0.049 |
Token | . |
Token ID | 13 |
Feature activation | -5.72e-03 |
Pos logit contributions | |
sing | +0.034 |
recovery | +0.033 |
heel | +0.033 |
badly | +0.032 |
boo | +0.032 |
Neg logit contributions | |
flour | -0.020 |
ingredients | -0.018 |
oons | -0.018 |
tables | -0.018 |
pots | -0.017 |
Token | Tim |
Token ID | 5045 |
Feature activation | +5.23e-02 |
Pos logit contributions | |
sing | +0.009 |
heel | +0.008 |
badly | +0.008 |
boo | +0.008 |
symp | +0.008 |
Neg logit contributions | |
flour | -0.011 |
oons | -0.010 |
ingredients | -0.010 |
pans | -0.010 |
ries | -0.010 |
Token | 's |
Token ID | 338 |
Feature activation | -3.55e-02 |
Pos logit contributions | |
flour | +0.199 |
oons | +0.190 |
ingredients | +0.184 |
tables | +0.182 |
pots | +0.181 |
Neg logit contributions | |
promise | -0.177 |
boo | -0.177 |
heel | -0.176 |
badly | -0.166 |
symp | -0.162 |
Token | take |
Token ID | 1011 |
Feature activation | -1.96e-02 |
Pos logit contributions | |
sing | +0.056 |
recovery | +0.054 |
symp | +0.053 |
badly | +0.053 |
heel | +0.053 |
Neg logit contributions | |
flour | -0.068 |
tables | -0.062 |
ingredients | -0.062 |
pots | -0.061 |
oons | -0.060 |
Token | it |
Token ID | 340 |
Feature activation | +1.35e-02 |
Pos logit contributions | |
heel | +0.038 |
boo | +0.037 |
badly | +0.036 |
hero | +0.036 |
tweet | +0.036 |
Neg logit contributions | |
oons | -0.028 |
flour | -0.027 |
cakes | -0.026 |
tables | -0.025 |
pots | -0.024 |
Token | home |
Token ID | 1363 |
Feature activation | -4.70e-02 |
Pos logit contributions | |
flour | +0.027 |
ingredients | +0.025 |
oons | +0.025 |
tables | +0.025 |
pots | +0.024 |
Neg logit contributions | |
sing | -0.020 |
heel | -0.019 |
badly | -0.019 |
recovery | -0.019 |
boo | -0.018 |
Token | ." |
Token ID | 526 |
Feature activation | +4.34e-02 |
Pos logit contributions | |
badly | +0.083 |
promise | +0.082 |
sing | +0.078 |
former | +0.075 |
heel | +0.075 |
Neg logit contributions | |
flour | -0.076 |
oons | -0.071 |
ries | -0.070 |
tables | -0.069 |
ingredients | -0.069 |
Token |
|
Token ID | 198 |
Feature activation | -2.92e-02 |
Pos logit contributions | |
flour | +0.089 |
ries | +0.084 |
pots | +0.081 |
ingredients | +0.081 |
oons | +0.080 |
Neg logit contributions | |
symp | -0.056 |
heel | -0.055 |
promise | -0.054 |
sing | -0.054 |
recovery | -0.053 |
Token |
|
Token ID | 198 |
Feature activation | -3.41e-02 |
Pos logit contributions | |
symp | +0.042 |
recovery | +0.042 |
badly | +0.041 |
promise | +0.041 |
sing | +0.041 |
Neg logit contributions | |
flour | -0.056 |
ingredients | -0.053 |
ries | -0.052 |
tables | -0.052 |
pots | -0.051 |
Token | They |
Token ID | 2990 |
Feature activation | +3.36e-02 |
Pos logit contributions | |
promise | +0.052 |
symp | +0.051 |
badly | +0.050 |
heel | +0.050 |
berth | +0.049 |
Neg logit contributions | |
flour | -0.063 |
ries | -0.061 |
ingredients | -0.058 |
pots | -0.057 |
tables | -0.056 |
Token | picked |
Token ID | 6497 |
Feature activation | +3.49e-02 |
Pos logit contributions | |
flour | +0.053 |
ingredients | +0.049 |
oons | +0.048 |
ries | +0.047 |
tables | +0.046 |
Neg logit contributions | |
symp | -0.061 |
sing | -0.061 |
heel | -0.060 |
boo | -0.060 |
badly | -0.059 |
Token | up |
Token ID | 510 |
Feature activation | +1.62e-02 |
Pos logit contributions | |
oons | +0.052 |
flour | +0.050 |
cakes | +0.050 |
tables | +0.048 |
ups | +0.047 |
Neg logit contributions | |
symp | -0.062 |
boo | -0.062 |
heel | -0.062 |
sing | -0.061 |
promise | -0.060 |
Token | the |
Token ID | 262 |
Feature activation | +3.92e-02 |
Pos logit contributions | |
oons | +0.022 |
cakes | +0.019 |
flour | +0.018 |
boards | +0.018 |
ingredients | +0.018 |
Neg logit contributions | |
symp | -0.031 |
heel | -0.030 |
promise | -0.030 |
badly | -0.030 |
boo | -0.029 |
Token | would |
Token ID | 561 |
Feature activation | +3.05e-03 |
Pos logit contributions | |
sing | +0.088 |
boo | +0.087 |
heel | +0.087 |
promise | +0.086 |
recovery | +0.086 |
Neg logit contributions | |
flour | -0.070 |
ingredients | -0.063 |
tables | -0.062 |
ries | -0.061 |
pots | -0.060 |
Token | recognize |
Token ID | 7564 |
Feature activation | +2.74e-02 |
Pos logit contributions | |
flour | +0.005 |
tables | +0.005 |
pots | +0.005 |
ingredients | +0.005 |
oons | +0.005 |
Neg logit contributions | |
sing | -0.005 |
badly | -0.005 |
symp | -0.005 |
retri | -0.005 |
spine | -0.005 |
Token | the |
Token ID | 262 |
Feature activation | +4.24e-02 |
Pos logit contributions | |
oons | +0.038 |
flour | +0.036 |
pans | +0.035 |
cakes | +0.034 |
ingredients | +0.032 |
Neg logit contributions | |
recovery | -0.051 |
badly | -0.050 |
tweet | -0.050 |
hero | -0.049 |
symp | -0.049 |
Token | different |
Token ID | 1180 |
Feature activation | -4.55e-02 |
Pos logit contributions | |
flour | +0.056 |
ries | +0.055 |
oons | +0.054 |
cakes | +0.050 |
ingredients | +0.048 |
Neg logit contributions | |
sing | -0.087 |
promise | -0.086 |
heel | -0.085 |
hero | -0.085 |
recovery | -0.084 |
Token | colors |
Token ID | 7577 |
Feature activation | +1.05e-01 |
Pos logit contributions | |
promise | +0.090 |
wish | +0.088 |
heel | +0.088 |
hero | +0.088 |
former | +0.085 |
Neg logit contributions | |
flour | -0.057 |
cakes | -0.054 |
ries | -0.053 |
glers | -0.053 |
oons | -0.051 |
Token | and |
Token ID | 290 |
Feature activation | -5.94e-03 |
Pos logit contributions | |
flour | +0.169 |
tables | +0.159 |
pans | +0.156 |
oons | +0.154 |
ingredients | +0.154 |
Neg logit contributions | |
badly | -0.176 |
sing | -0.176 |
tweet | -0.169 |
recovery | -0.168 |
heel | -0.168 |
Token | types |
Token ID | 3858 |
Feature activation | +1.35e-01 |
Pos logit contributions | |
sing | +0.011 |
former | +0.011 |
heel | +0.011 |
wish | +0.011 |
tweet | +0.011 |
Neg logit contributions | |
oons | -0.009 |
flour | -0.008 |
ries | -0.008 |
cakes | -0.007 |
ingredients | -0.007 |
Token | of |
Token ID | 286 |
Feature activation | +3.43e-02 |
Pos logit contributions | |
flour | +0.208 |
tables | +0.188 |
pans | +0.185 |
ingredients | +0.183 |
oons | +0.182 |
Neg logit contributions | |
badly | -0.253 |
hero | -0.247 |
recovery | -0.246 |
heel | -0.234 |
tweet | -0.231 |
Token | furniture |
Token ID | 13091 |
Feature activation | +7.49e-02 |
Pos logit contributions | |
flour | +0.042 |
cakes | +0.039 |
oons | +0.039 |
ingredients | +0.038 |
ups | +0.038 |
Neg logit contributions | |
recovery | -0.069 |
boo | -0.068 |
heel | -0.067 |
sing | -0.067 |
wish | -0.066 |
Token | . |
Token ID | 13 |
Feature activation | -5.07e-03 |
Pos logit contributions | |
flour | +0.113 |
oons | +0.100 |
ingredients | +0.099 |
tables | +0.099 |
pans | +0.096 |
Neg logit contributions | |
sing | -0.141 |
recovery | -0.138 |
badly | -0.136 |
heel | -0.134 |
promise | -0.134 |
Token | She |
Token ID | 1375 |
Feature activation | -4.21e-02 |
Pos logit contributions | |
sing | +0.009 |
recovery | +0.008 |
symp | +0.008 |
heel | +0.008 |
former | +0.008 |
Neg logit contributions | |
flour | -0.009 |
oons | -0.008 |
ingredients | -0.008 |
pots | -0.008 |
tables | -0.008 |
Token | out |
Token ID | 503 |
Feature activation | +1.50e-02 |
Pos logit contributions | |
flour | +0.059 |
oons | +0.055 |
ingredients | +0.053 |
cakes | +0.052 |
tables | +0.052 |
Neg logit contributions | |
sing | -0.069 |
heel | -0.066 |
symp | -0.065 |
badly | -0.065 |
recovery | -0.065 |
Token | to |
Token ID | 284 |
Feature activation | -5.19e-02 |
Pos logit contributions | |
flour | +0.021 |
ingredients | +0.019 |
oons | +0.019 |
tables | +0.019 |
pots | +0.018 |
Neg logit contributions | |
sing | -0.031 |
recovery | -0.030 |
heel | -0.030 |
badly | -0.030 |
symp | -0.030 |
Token | take |
Token ID | 1011 |
Feature activation | -3.32e-02 |
Pos logit contributions | |
sing | +0.080 |
heel | +0.079 |
recovery | +0.078 |
boo | +0.078 |
promise | +0.076 |
Neg logit contributions | |
flour | -0.099 |
oons | -0.093 |
ingredients | -0.092 |
pots | -0.089 |
tables | -0.089 |
Token | it |
Token ID | 340 |
Feature activation | +2.75e-03 |
Pos logit contributions | |
heel | +0.064 |
badly | +0.062 |
hero | +0.062 |
promise | +0.061 |
tweet | +0.061 |
Neg logit contributions | |
oons | -0.048 |
flour | -0.046 |
cakes | -0.042 |
ingredients | -0.041 |
pans | -0.041 |
Token | , |
Token ID | 11 |
Feature activation | +3.22e-02 |
Pos logit contributions | |
flour | +0.004 |
ingredients | +0.003 |
tables | +0.003 |
pots | +0.003 |
oons | +0.003 |
Neg logit contributions | |
sing | -0.006 |
recovery | -0.006 |
boo | -0.006 |
symp | -0.006 |
heel | -0.006 |
Token | and |
Token ID | 290 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
flour | +0.048 |
tables | +0.044 |
ingredients | +0.043 |
pots | +0.042 |
pans | +0.039 |
Neg logit contributions | |
recovery | -0.064 |
sing | -0.063 |
hero | -0.061 |
badly | -0.061 |
former | -0.061 |
Token | suddenly |
Token ID | 6451 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
sing | +0.027 |
heel | +0.026 |
badly | +0.026 |
boo | +0.026 |
symp | +0.026 |
Neg logit contributions | |
flour | -0.021 |
oons | -0.020 |
ingredients | -0.019 |
pots | -0.018 |
ries | -0.018 |
Token | an |
Token ID | 281 |
Feature activation | +1.79e-02 |
Pos logit contributions | |
flour | +0.177 |
oons | +0.175 |
cakes | +0.169 |
ingredients | +0.166 |
pans | +0.161 |
Neg logit contributions | |
heel | -0.176 |
badly | -0.175 |
sing | -0.169 |
tweet | -0.168 |
symp | -0.164 |
Token | old |
Token ID | 1468 |
Feature activation | -4.76e-02 |
Pos logit contributions | |
flour | +0.028 |
tables | +0.025 |
ingredients | +0.024 |
pots | +0.024 |
oons | +0.023 |
Neg logit contributions | |
sing | -0.033 |
former | -0.033 |
recovery | -0.033 |
ighed | -0.032 |
spine | -0.032 |
Token | lady |
Token ID | 10846 |
Feature activation | -3.50e-02 |
Pos logit contributions | |
sing | +0.098 |
recovery | +0.095 |
symp | +0.095 |
former | +0.094 |
badly | +0.091 |
Neg logit contributions | |
flour | -0.068 |
ingredients | -0.061 |
tables | -0.060 |
pots | -0.059 |
cans | -0.056 |
Token | appeared |
Token ID | 4120 |
Feature activation | -1.88e-02 |
Pos logit contributions | |
sing | +0.072 |
heel | +0.070 |
badly | +0.069 |
recovery | +0.069 |
symp | +0.069 |
Neg logit contributions | |
flour | -0.050 |
ingredients | -0.044 |
oons | -0.044 |
tables | -0.043 |
pots | -0.043 |
Token | smiles |
Token ID | 21845 |
Feature activation | +5.88e-02 |
Pos logit contributions | |
boo | +0.038 |
promise | +0.034 |
heel | +0.034 |
apolog | +0.033 |
bru | +0.033 |
Neg logit contributions | |
oons | -0.030 |
flour | -0.030 |
ingredients | -0.029 |
cakes | -0.028 |
boards | -0.028 |
Token | . |
Token ID | 13 |
Feature activation | -2.52e-02 |
Pos logit contributions | |
flour | +0.047 |
pots | +0.039 |
ingredients | +0.039 |
tables | +0.038 |
oons | +0.036 |
Neg logit contributions | |
sing | -0.155 |
recovery | -0.152 |
former | -0.149 |
heel | -0.149 |
spine | -0.148 |
Token | She |
Token ID | 1375 |
Feature activation | -4.68e-02 |
Pos logit contributions | |
sing | +0.042 |
heel | +0.041 |
recovery | +0.041 |
symp | +0.041 |
boo | +0.040 |
Neg logit contributions | |
flour | -0.044 |
ingredients | -0.042 |
tables | -0.041 |
oons | -0.041 |
pots | -0.040 |
Token | makes |
Token ID | 1838 |
Feature activation | +6.40e-02 |
Pos logit contributions | |
boo | +0.088 |
promise | +0.086 |
heel | +0.085 |
sing | +0.085 |
symp | +0.084 |
Neg logit contributions | |
flour | -0.072 |
ingredients | -0.065 |
tables | -0.064 |
oons | -0.063 |
pots | -0.063 |
Token | a |
Token ID | 257 |
Feature activation | +1.12e-02 |
Pos logit contributions | |
flour | +0.077 |
ingredients | +0.068 |
tables | +0.067 |
oons | +0.067 |
pots | +0.065 |
Neg logit contributions | |
sing | -0.144 |
recovery | -0.140 |
heel | -0.140 |
boo | -0.140 |
badly | -0.139 |
Token | blue |
Token ID | 4171 |
Feature activation | -1.99e-02 |
Pos logit contributions | |
flour | +0.017 |
pots | +0.014 |
brushes | +0.014 |
cups | +0.014 |
ingredients | +0.013 |
Neg logit contributions | |
badly | -0.024 |
amb | -0.023 |
ur | -0.023 |
ez | -0.023 |
recovery | -0.023 |
Token | mark |
Token ID | 1317 |
Feature activation | -2.97e-02 |
Pos logit contributions | |
badly | +0.044 |
recovery | +0.043 |
heel | +0.043 |
former | +0.042 |
sing | +0.042 |
Neg logit contributions | |
flour | -0.026 |
ingredients | -0.023 |
pots | -0.023 |
tables | -0.021 |
cups | -0.021 |
Token | on |
Token ID | 319 |
Feature activation | -5.43e-02 |
Pos logit contributions | |
sing | +0.068 |
recovery | +0.066 |
heel | +0.066 |
former | +0.066 |
spine | +0.065 |
Neg logit contributions | |
flour | -0.036 |
ingredients | -0.031 |
oons | -0.031 |
pots | -0.030 |
tables | -0.030 |
Token | her |
Token ID | 607 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
sing | +0.120 |
heel | +0.118 |
boo | +0.117 |
symp | +0.116 |
recovery | +0.116 |
Neg logit contributions | |
flour | -0.066 |
ingredients | -0.062 |
oons | -0.062 |
cakes | -0.057 |
tables | -0.057 |
Token | paper |
Token ID | 3348 |
Feature activation | -4.79e-03 |
Pos logit contributions | |
badly | +0.227 |
recovery | +0.226 |
sing | +0.220 |
former | +0.217 |
heel | +0.216 |
Neg logit contributions | |
flour | -0.192 |
ingredients | -0.175 |
pots | -0.174 |
tables | -0.173 |
pans | -0.164 |
Token | . |
Token ID | 13 |
Feature activation | -2.55e-02 |
Pos logit contributions | |
sing | +0.012 |
recovery | +0.011 |
heel | +0.011 |
badly | +0.011 |
boo | +0.011 |
Neg logit contributions | |
flour | -0.005 |
ingredients | -0.004 |
pots | -0.004 |
tables | -0.004 |
oons | -0.004 |
Token | , |
Token ID | 11 |
Feature activation | +5.30e-02 |
Pos logit contributions | |
flour | +0.074 |
ingredients | +0.066 |
oons | +0.066 |
tables | +0.065 |
pots | +0.064 |
Neg logit contributions | |
sing | -0.112 |
recovery | -0.109 |
heel | -0.109 |
badly | -0.108 |
boo | -0.107 |
Token | he |
Token ID | 339 |
Feature activation | -2.15e-02 |
Pos logit contributions | |
flour | +0.048 |
tables | +0.040 |
ingredients | +0.039 |
pots | +0.034 |
chairs | +0.033 |
Neg logit contributions | |
sing | -0.140 |
recovery | -0.134 |
boo | -0.133 |
heel | -0.133 |
symp | -0.132 |
Token | rode |
Token ID | 22075 |
Feature activation | -2.83e-02 |
Pos logit contributions | |
sing | +0.045 |
recovery | +0.043 |
badly | +0.042 |
boo | +0.042 |
heel | +0.042 |
Neg logit contributions | |
flour | -0.030 |
ingredients | -0.027 |
tables | -0.027 |
oons | -0.026 |
pots | -0.026 |
Token | away |
Token ID | 1497 |
Feature activation | -8.95e-02 |
Pos logit contributions | |
sing | +0.053 |
recovery | +0.050 |
boo | +0.049 |
symp | +0.049 |
heel | +0.049 |
Neg logit contributions | |
flour | -0.046 |
tables | -0.043 |
ingredients | -0.043 |
pots | -0.041 |
oons | -0.041 |
Token | on |
Token ID | 319 |
Feature activation | -4.50e-02 |
Pos logit contributions | |
sing | +0.204 |
boo | +0.198 |
recovery | +0.197 |
oof | +0.195 |
symp | +0.193 |
Neg logit contributions | |
flour | -0.096 |
tables | -0.093 |
ingredients | -0.088 |
pots | -0.085 |
chairs | -0.085 |
Token | his |
Token ID | 465 |
Feature activation | -1.27e-01 |
Pos logit contributions | |
sing | +0.101 |
symp | +0.096 |
recovery | +0.095 |
boo | +0.095 |
heel | +0.095 |
Neg logit contributions | |
flour | -0.057 |
tables | -0.052 |
ingredients | -0.051 |
pots | -0.049 |
oons | -0.046 |
Token | shining |
Token ID | 22751 |
Feature activation | -4.09e-02 |
Pos logit contributions | |
aring | +0.227 |
amb | +0.226 |
utt | +0.225 |
ree | +0.224 |
ur | +0.224 |
Neg logit contributions | |
tables | -0.199 |
chairs | -0.198 |
flour | -0.192 |
boxes | -0.186 |
ingredients | -0.184 |
Token | white |
Token ID | 2330 |
Feature activation | -4.75e-02 |
Pos logit contributions | |
sing | +0.066 |
symp | +0.060 |
aring | +0.058 |
promise | +0.057 |
spine | +0.057 |
Neg logit contributions | |
flour | -0.080 |
tables | -0.074 |
ingredients | -0.074 |
pots | -0.072 |
chairs | -0.071 |
Token | horse |
Token ID | 8223 |
Feature activation | -1.17e-01 |
Pos logit contributions | |
sing | +0.077 |
utt | +0.070 |
symp | +0.070 |
badly | +0.069 |
uer | +0.069 |
Neg logit contributions | |
flour | -0.089 |
ingredients | -0.086 |
tables | -0.086 |
pots | -0.083 |
chairs | -0.083 |
Token | , |
Token ID | 11 |
Feature activation | +3.68e-02 |
Pos logit contributions | |
boo | +0.280 |
sing | +0.275 |
bru | +0.270 |
recovery | +0.266 |
former | +0.264 |
Neg logit contributions | |
flour | -0.106 |
ingredients | -0.105 |
tables | -0.105 |
chairs | -0.097 |
pots | -0.092 |
Token | and |
Token ID | 290 |
Feature activation | -1.69e-02 |
Pos logit contributions | |
flour | +0.059 |
tables | +0.056 |
ingredients | +0.054 |
pots | +0.053 |
chairs | +0.052 |
Neg logit contributions | |
recovery | -0.063 |
sing | -0.062 |
symp | -0.061 |
badly | -0.061 |
former | -0.061 |
Token | , |
Token ID | 11 |
Feature activation | +5.44e-02 |
Pos logit contributions | |
sing | +0.069 |
recovery | +0.067 |
heel | +0.067 |
boo | +0.066 |
badly | +0.066 |
Neg logit contributions | |
flour | -0.047 |
ingredients | -0.042 |
tables | -0.042 |
oons | -0.042 |
pots | -0.041 |
Token | " |
Token ID | 366 |
Feature activation | +6.27e-02 |
Pos logit contributions | |
flour | +0.053 |
tables | +0.049 |
ingredients | +0.043 |
pots | +0.043 |
pans | +0.042 |
Neg logit contributions | |
sing | -0.133 |
recovery | -0.131 |
former | -0.130 |
badly | -0.129 |
heel | -0.128 |
Token | Put |
Token ID | 11588 |
Feature activation | +3.42e-02 |
Pos logit contributions | |
flour | +0.094 |
ingredients | +0.086 |
tables | +0.083 |
oons | +0.082 |
cakes | +0.081 |
Neg logit contributions | |
sing | -0.132 |
boo | -0.121 |
heel | -0.120 |
spine | -0.120 |
promise | -0.120 |
Token | this |
Token ID | 428 |
Feature activation | -2.07e-02 |
Pos logit contributions | |
flour | +0.049 |
oons | +0.047 |
tables | +0.045 |
cakes | +0.044 |
pots | +0.044 |
Neg logit contributions | |
heel | -0.065 |
boo | -0.065 |
badly | -0.064 |
recovery | -0.063 |
promise | -0.062 |
Token | in |
Token ID | 287 |
Feature activation | -3.49e-02 |
Pos logit contributions | |
heel | +0.034 |
badly | +0.034 |
sing | +0.034 |
recovery | +0.033 |
boo | +0.033 |
Neg logit contributions | |
flour | -0.036 |
oons | -0.033 |
tables | -0.033 |
ingredients | -0.033 |
pots | -0.032 |
Token | your |
Token ID | 534 |
Feature activation | -1.31e-01 |
Pos logit contributions | |
sing | +0.073 |
recovery | +0.073 |
badly | +0.072 |
heel | +0.072 |
boo | +0.071 |
Neg logit contributions | |
flour | -0.046 |
oons | -0.041 |
ingredients | -0.041 |
tables | -0.040 |
pots | -0.040 |
Token | pocket |
Token ID | 10000 |
Feature activation | -5.86e-02 |
Pos logit contributions | |
sing | +0.233 |
tweet | +0.213 |
symp | +0.212 |
badly | +0.210 |
ree | +0.207 |
Neg logit contributions | |
flour | -0.237 |
tables | -0.220 |
ingredients | -0.220 |
pots | -0.218 |
cans | -0.207 |
Token | ." |
Token ID | 526 |
Feature activation | +3.76e-02 |
Pos logit contributions | |
badly | +0.115 |
heel | +0.113 |
promise | +0.112 |
sing | +0.112 |
recovery | +0.111 |
Neg logit contributions | |
flour | -0.084 |
oons | -0.074 |
tables | -0.074 |
ingredients | -0.074 |
pots | -0.073 |
Token |
|
Token ID | 198 |
Feature activation | -3.29e-02 |
Pos logit contributions | |
flour | +0.057 |
ingredients | +0.051 |
tables | +0.051 |
oons | +0.050 |
pots | +0.049 |
Neg logit contributions | |
sing | -0.075 |
recovery | -0.072 |
heel | -0.072 |
badly | -0.072 |
boo | -0.071 |
Token |
|
Token ID | 198 |
Feature activation | -3.69e-02 |
Pos logit contributions | |
sing | +0.063 |
recovery | +0.061 |
heel | +0.061 |
badly | +0.060 |
symp | +0.060 |
Neg logit contributions | |
flour | -0.051 |
ingredients | -0.046 |
tables | -0.046 |
oons | -0.045 |
pots | -0.045 |
Token | The |
Token ID | 464 |
Feature activation | +2.91e-02 |
Pos logit contributions | |
sing | +0.074 |
boo | +0.069 |
spine | +0.069 |
recovery | +0.069 |
heel | +0.069 |
Neg logit contributions | |
flour | -0.058 |
tables | -0.053 |
oons | -0.052 |
ingredients | -0.051 |
pots | -0.049 |
Token | the |
Token ID | 262 |
Feature activation | +2.98e-02 |
Pos logit contributions | |
recovery | +0.046 |
heel | +0.045 |
badly | +0.045 |
sing | +0.045 |
promise | +0.044 |
Neg logit contributions | |
flour | -0.027 |
oons | -0.026 |
tables | -0.024 |
ingredients | -0.024 |
cakes | -0.024 |
Token | special |
Token ID | 2041 |
Feature activation | -8.47e-02 |
Pos logit contributions | |
flour | +0.051 |
ingredients | +0.047 |
pots | +0.043 |
tables | +0.043 |
cans | +0.041 |
Neg logit contributions | |
sing | -0.055 |
symp | -0.054 |
badly | -0.053 |
spine | -0.052 |
ree | -0.052 |
Token | day |
Token ID | 1110 |
Feature activation | -8.61e-02 |
Pos logit contributions | |
sing | +0.144 |
recovery | +0.137 |
symp | +0.137 |
heel | +0.137 |
badly | +0.136 |
Neg logit contributions | |
flour | -0.151 |
ingredients | -0.140 |
tables | -0.136 |
oons | -0.135 |
pots | -0.135 |
Token | with |
Token ID | 351 |
Feature activation | +1.38e-02 |
Pos logit contributions | |
sing | +0.165 |
badly | +0.163 |
recovery | +0.162 |
heel | +0.160 |
symp | +0.160 |
Neg logit contributions | |
flour | -0.131 |
oons | -0.118 |
ingredients | -0.116 |
tables | -0.115 |
pots | -0.113 |
Token | the |
Token ID | 262 |
Feature activation | +2.78e-02 |
Pos logit contributions | |
flour | +0.018 |
oons | +0.017 |
ingredients | +0.016 |
tables | +0.016 |
pots | +0.016 |
Neg logit contributions | |
sing | -0.029 |
recovery | -0.029 |
heel | -0.029 |
badly | -0.029 |
boo | -0.028 |
Token | j |
Token ID | 474 |
Feature activation | -1.75e-01 |
Pos logit contributions | |
flour | +0.044 |
ingredients | +0.039 |
tables | +0.038 |
pots | +0.038 |
oons | +0.037 |
Neg logit contributions | |
sing | -0.054 |
recovery | -0.052 |
symp | -0.052 |
heel | -0.052 |
badly | -0.052 |
Token | olly |
Token ID | 5098 |
Feature activation | -1.04e-01 |
Pos logit contributions | |
sing | +0.268 |
recovery | +0.258 |
heel | +0.255 |
badly | +0.252 |
boo | +0.251 |
Neg logit contributions | |
flour | -0.341 |
ingredients | -0.316 |
tables | -0.314 |
oons | -0.311 |
pots | -0.309 |
Token | ducks |
Token ID | 39694 |
Feature activation | +2.58e-02 |
Pos logit contributions | |
sing | +0.180 |
recovery | +0.173 |
heel | +0.172 |
badly | +0.171 |
boo | +0.170 |
Neg logit contributions | |
flour | -0.181 |
ingredients | -0.167 |
tables | -0.165 |
oons | -0.163 |
pots | -0.162 |
Token | . |
Token ID | 13 |
Feature activation | -1.93e-02 |
Pos logit contributions | |
flour | +0.031 |
ingredients | +0.028 |
pots | +0.027 |
tables | +0.027 |
oons | +0.026 |
Neg logit contributions | |
sing | -0.058 |
boo | -0.056 |
recovery | -0.056 |
heel | -0.055 |
symp | -0.055 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +5.21e-02 |
Pos logit contributions | |
sing | +0.037 |
recovery | +0.036 |
heel | +0.036 |
badly | +0.035 |
boo | +0.035 |
Neg logit contributions | |
flour | -0.030 |
ingredients | -0.027 |
tables | -0.027 |
oons | -0.027 |
pots | -0.026 |
Token | Once |
Token ID | 7454 |
Feature activation | +2.58e-02 |
Pos logit contributions | |
flour | +0.092 |
pans | +0.085 |
ingredients | +0.084 |
pots | +0.084 |
tables | +0.082 |
Neg logit contributions | |
symp | -0.082 |
recovery | -0.081 |
heel | -0.081 |
badly | -0.080 |
hero | -0.079 |
Token | ice |
Token ID | 4771 |
Feature activation | -8.72e-02 |
Pos logit contributions | |
sing | +0.182 |
symp | +0.178 |
recovery | +0.178 |
heel | +0.175 |
tweet | +0.174 |
Neg logit contributions | |
flour | -0.155 |
ingredients | -0.140 |
pots | -0.135 |
tables | -0.132 |
oons | -0.127 |
Token | cream |
Token ID | 8566 |
Feature activation | +7.11e-03 |
Pos logit contributions | |
sing | +0.170 |
badly | +0.169 |
heel | +0.166 |
boo | +0.165 |
recovery | +0.165 |
Neg logit contributions | |
flour | -0.128 |
tables | -0.118 |
ingredients | -0.115 |
oons | -0.115 |
pots | -0.115 |
Token | cone |
Token ID | 27763 |
Feature activation | -6.97e-02 |
Pos logit contributions | |
flour | +0.012 |
ingredients | +0.011 |
tables | +0.011 |
oons | +0.011 |
pots | +0.011 |
Neg logit contributions | |
sing | -0.013 |
recovery | -0.012 |
heel | -0.012 |
badly | -0.012 |
boo | -0.012 |
Token | and |
Token ID | 290 |
Feature activation | +1.07e-02 |
Pos logit contributions | |
sing | +0.140 |
badly | +0.136 |
recovery | +0.136 |
heel | +0.136 |
boo | +0.134 |
Neg logit contributions | |
flour | -0.101 |
tables | -0.090 |
ingredients | -0.090 |
oons | -0.089 |
pots | -0.088 |
Token | played |
Token ID | 2826 |
Feature activation | -3.08e-02 |
Pos logit contributions | |
flour | +0.015 |
oons | +0.015 |
pots | +0.014 |
ingredients | +0.014 |
tables | +0.014 |
Neg logit contributions | |
boo | -0.020 |
heel | -0.020 |
badly | -0.019 |
promise | -0.019 |
knee | -0.018 |
Token | happily |
Token ID | 18177 |
Feature activation | -1.26e-01 |
Pos logit contributions | |
sing | +0.076 |
spine | +0.073 |
former | +0.070 |
symp | +0.069 |
heel | +0.068 |
Neg logit contributions | |
flour | -0.034 |
tables | -0.033 |
ingredients | -0.029 |
chairs | -0.028 |
pots | -0.028 |
Token | with |
Token ID | 351 |
Feature activation | +6.49e-03 |
Pos logit contributions | |
sing | +0.302 |
boo | +0.298 |
recovery | +0.296 |
symp | +0.291 |
former | +0.287 |
Neg logit contributions | |
flour | -0.108 |
tables | -0.105 |
ingredients | -0.104 |
cakes | -0.100 |
oons | -0.098 |
Token | her |
Token ID | 607 |
Feature activation | -1.07e-01 |
Pos logit contributions | |
oons | +0.012 |
flour | +0.012 |
pans | +0.012 |
boards | +0.012 |
cakes | +0.011 |
Neg logit contributions | |
badly | -0.010 |
recovery | -0.009 |
heel | -0.009 |
hero | -0.009 |
heart | -0.009 |
Token | fixed |
Token ID | 5969 |
Feature activation | -8.84e-03 |
Pos logit contributions | |
sing | +0.212 |
symp | +0.202 |
recovery | +0.202 |
spine | +0.201 |
tweet | +0.199 |
Neg logit contributions | |
flour | -0.168 |
ingredients | -0.152 |
tables | -0.144 |
pots | -0.144 |
ries | -0.139 |
Token | toy |
Token ID | 13373 |
Feature activation | -8.93e-02 |
Pos logit contributions | |
badly | +0.016 |
heel | +0.015 |
boo | +0.015 |
recovery | +0.015 |
promise | +0.015 |
Neg logit contributions | |
flour | -0.014 |
oons | -0.014 |
tables | -0.013 |
cakes | -0.013 |
pots | -0.013 |
Token | car |
Token ID | 1097 |
Feature activation | -5.43e-02 |
Pos logit contributions | |
sing | +0.174 |
symp | +0.166 |
recovery | +0.165 |
heel | +0.163 |
boo | +0.163 |
Neg logit contributions | |
flour | -0.140 |
ingredients | -0.127 |
tables | -0.126 |
pots | -0.123 |
oons | -0.123 |
Token | He |
Token ID | 679 |
Feature activation | -6.40e-02 |
Pos logit contributions | |
sing | +0.025 |
recovery | +0.024 |
heel | +0.024 |
badly | +0.024 |
boo | +0.023 |
Neg logit contributions | |
flour | -0.036 |
ingredients | -0.033 |
tables | -0.033 |
oons | -0.033 |
pots | -0.033 |
Token | said |
Token ID | 531 |
Feature activation | -2.76e-02 |
Pos logit contributions | |
sing | +0.146 |
recovery | +0.143 |
badly | +0.138 |
spine | +0.138 |
heel | +0.137 |
Neg logit contributions | |
flour | -0.076 |
ingredients | -0.068 |
tables | -0.067 |
oons | -0.065 |
pots | -0.064 |
Token | , |
Token ID | 11 |
Feature activation | +5.53e-02 |
Pos logit contributions | |
sing | +0.065 |
recovery | +0.062 |
spine | +0.061 |
former | +0.061 |
heel | +0.060 |
Neg logit contributions | |
flour | -0.033 |
ingredients | -0.028 |
tables | -0.028 |
oons | -0.028 |
pots | -0.028 |
Token | � |
Token ID | 6184 |
Feature activation | -7.78e-02 |
Pos logit contributions | |
flour | +0.052 |
tables | +0.045 |
pans | +0.041 |
ingredients | +0.040 |
cakes | +0.040 |
Neg logit contributions | |
sing | -0.136 |
recovery | -0.135 |
former | -0.134 |
promise | -0.134 |
badly | -0.132 |
Token | � |
Token ID | 95 |
Feature activation | -7.72e-02 |
Pos logit contributions | |
sing | +0.117 |
boo | +0.116 |
injury | +0.115 |
former | +0.114 |
weep | +0.112 |
Neg logit contributions | |
flour | -0.129 |
cakes | -0.127 |
tables | -0.126 |
oons | -0.123 |
boards | -0.120 |
Token | € |
Token ID | 26391 |
Feature activation | -1.42e-01 |
Pos logit contributions | |
sing | +0.120 |
badly | +0.119 |
boo | +0.114 |
former | +0.107 |
heel | +0.107 |
Neg logit contributions | |
flour | -0.141 |
tables | -0.136 |
ingredients | -0.134 |
cakes | -0.129 |
cans | -0.128 |
Token | � |
Token ID | 129 |
Feature activation | -6.05e-02 |
Pos logit contributions | |
sing | +0.242 |
recovery | +0.233 |
heel | +0.231 |
badly | +0.229 |
boo | +0.228 |
Neg logit contributions | |
flour | -0.249 |
ingredients | -0.229 |
tables | -0.227 |
oons | -0.225 |
pots | -0.223 |
Token | � |
Token ID | 241 |
Feature activation | -7.27e-03 |
Pos logit contributions | |
hero | +0.099 |
badly | +0.098 |
bru | +0.098 |
recovery | +0.097 |
boo | +0.095 |
Neg logit contributions | |
tables | -0.095 |
flour | -0.091 |
ers | -0.086 |
ups | -0.086 |
oons | -0.084 |
Token | Yes |
Token ID | 5297 |
Feature activation | -2.24e-02 |
Pos logit contributions | |
sing | +0.014 |
recovery | +0.013 |
boo | +0.013 |
symp | +0.013 |
heel | +0.013 |
Neg logit contributions | |
flour | -0.012 |
ingredients | -0.011 |
tables | -0.010 |
oons | -0.010 |
pots | -0.010 |
Token | , |
Token ID | 11 |
Feature activation | +4.92e-02 |
Pos logit contributions | |
sing | +0.051 |
recovery | +0.049 |
boo | +0.048 |
heel | +0.048 |
former | +0.048 |
Neg logit contributions | |
flour | -0.028 |
ingredients | -0.024 |
tables | -0.024 |
oons | -0.024 |
pots | -0.023 |
Token | I |
Token ID | 314 |
Feature activation | +3.33e-02 |
Pos logit contributions | |
flour | +0.062 |
tables | +0.053 |
ingredients | +0.053 |
pots | +0.048 |
oons | +0.048 |
Neg logit contributions | |
sing | -0.114 |
recovery | -0.106 |
boo | -0.105 |
spine | -0.105 |
symp | -0.105 |