Token | . |
Token ID | 13 |
Feature activation | -1.30e-01 |
Pos logit contributions | |
Crit | +0.641 |
Nit | +0.598 |
Poverty | +0.573 |
Crit | +0.570 |
Ps | +0.533 |
Neg logit contributions | |
Kislyak | -0.465 |
================================================================= | -0.458 |
IMAGES | -0.446 |
advertising | -0.427 |
zinski | -0.425 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +1.21e-01 |
Pos logit contributions | |
Crit | +0.318 |
Nit | +0.289 |
Poverty | +0.280 |
Crit | +0.276 |
Adapt | +0.252 |
Neg logit contributions | |
Kislyak | -0.362 |
================================================================= | -0.356 |
IMAGES | -0.349 |
advertising | -0.340 |
zinski | -0.339 |
Token | Abstract |
Token ID | 23839 |
Feature activation | +2.01e-01 |
Pos logit contributions | |
Kislyak | +0.357 |
================================================================= | +0.356 |
IMAGES | +0.346 |
advertising | +0.333 |
zinski | +0.333 |
Neg logit contributions | |
Crit | -0.276 |
Nit | -0.251 |
Poverty | -0.234 |
Crit | -0.231 |
Adapt | -0.214 |
Token |
|
Token ID | 198 |
Feature activation | +1.28e-01 |
Pos logit contributions | |
Kislyak | +0.553 |
================================================================= | +0.550 |
IMAGES | +0.532 |
advertising | +0.520 |
zinski | +0.515 |
Neg logit contributions | |
Crit | -0.500 |
Nit | -0.458 |
Poverty | -0.438 |
Crit | -0.431 |
Adapt | -0.396 |
Token |
|
Token ID | 198 |
Feature activation | +1.88e-01 |
Pos logit contributions | |
Kislyak | +0.357 |
================================================================= | +0.356 |
IMAGES | +0.344 |
advertising | +0.335 |
zinski | +0.331 |
Neg logit contributions | |
Crit | -0.314 |
Nit | -0.288 |
Poverty | -0.274 |
Crit | -0.269 |
corruption | -0.249 |
Token | Background |
Token ID | 21756 |
Feature activation | +1.39e+00 |
Pos logit contributions | |
Kislyak | +0.572 |
================================================================= | +0.570 |
IMAGES | +0.554 |
advertising | +0.538 |
zinski | +0.535 |
Neg logit contributions | |
Crit | -0.415 |
Nit | -0.376 |
Poverty | -0.352 |
Crit | -0.346 |
Adapt | -0.318 |
Token | Depression |
Token ID | 22483 |
Feature activation | +5.05e-01 |
Pos logit contributions | |
================================================================= | +3.624 |
Kislyak | +3.619 |
IMAGES | +3.473 |
advertising | +3.364 |
DEA | +3.337 |
Neg logit contributions | |
Crit | -3.616 |
Nit | -3.327 |
Poverty | -3.112 |
Crit | -3.064 |
Lumpur | -2.932 |
Token | presents |
Token ID | 10969 |
Feature activation | +2.49e-01 |
Pos logit contributions | |
================================================================= | +1.361 |
Kislyak | +1.358 |
IMAGES | +1.310 |
advertising | +1.268 |
zinski | +1.265 |
Neg logit contributions | |
Crit | -1.269 |
Nit | -1.169 |
Poverty | -1.107 |
Crit | -1.086 |
Adapt | -1.015 |
Token | a |
Token ID | 257 |
Feature activation | +4.60e-01 |
Pos logit contributions | |
Kislyak | +0.640 |
================================================================= | +0.638 |
IMAGES | +0.615 |
advertising | +0.597 |
zinski | +0.591 |
Neg logit contributions | |
Crit | -0.662 |
Nit | -0.610 |
Poverty | -0.582 |
Crit | -0.574 |
corruption | -0.534 |
Token | significant |
Token ID | 2383 |
Feature activation | +3.78e-01 |
Pos logit contributions | |
Kislyak | +1.279 |
================================================================= | +1.276 |
IMAGES | +1.232 |
advertising | +1.199 |
zinski | +1.189 |
Neg logit contributions | |
Crit | -1.125 |
Nit | -1.030 |
Poverty | -0.978 |
Crit | -0.961 |
corruption | -0.889 |
Token | burden |
Token ID | 10538 |
Feature activation | +3.96e-01 |
Pos logit contributions | |
Kislyak | +1.022 |
================================================================= | +1.019 |
IMAGES | +0.984 |
advertising | +0.957 |
zinski | +0.947 |
Neg logit contributions | |
Crit | -0.957 |
Nit | -0.878 |
Poverty | -0.836 |
Crit | -0.823 |
corruption | -0.763 |
Token | travel |
Token ID | 3067 |
Feature activation | +4.69e-02 |
Pos logit contributions | |
Crit | +0.300 |
Nit | +0.273 |
Poverty | +0.260 |
Crit | +0.256 |
Lumpur | +0.233 |
Neg logit contributions | |
Kislyak | -0.377 |
================================================================= | -0.376 |
IMAGES | -0.365 |
advertising | -0.357 |
Accessory | -0.353 |
Token | back |
Token ID | 736 |
Feature activation | -3.70e-01 |
Pos logit contributions | |
Kislyak | +0.118 |
================================================================= | +0.118 |
IMAGES | +0.114 |
advertising | +0.111 |
zinski | +0.110 |
Neg logit contributions | |
Crit | -0.127 |
Nit | -0.116 |
Poverty | -0.112 |
Crit | -0.111 |
corruption | -0.102 |
Token | to |
Token ID | 284 |
Feature activation | -7.50e-02 |
Pos logit contributions | |
Crit | +0.999 |
Nit | +0.919 |
Poverty | +0.885 |
Crit | +0.879 |
Lumpur | +0.814 |
Neg logit contributions | |
Kislyak | -0.926 |
================================================================= | -0.922 |
IMAGES | -0.895 |
advertising | -0.864 |
zinski | -0.860 |
Token | Beijing |
Token ID | 11618 |
Feature activation | +5.37e-02 |
Pos logit contributions | |
Crit | +0.197 |
Nit | +0.181 |
Poverty | +0.174 |
Crit | +0.172 |
Lumpur | +0.159 |
Neg logit contributions | |
Kislyak | -0.194 |
================================================================= | -0.194 |
IMAGES | -0.188 |
advertising | -0.182 |
zinski | -0.181 |
Token | by |
Token ID | 416 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
Kislyak | +0.131 |
================================================================= | +0.130 |
IMAGES | +0.126 |
advertising | +0.122 |
zinski | +0.121 |
Neg logit contributions | |
Crit | -0.150 |
Nit | -0.138 |
Poverty | -0.133 |
Crit | -0.131 |
Lumpur | -0.122 |
Token | train |
Token ID | 4512 |
Feature activation | +1.22e+00 |
Pos logit contributions | |
Crit | +0.035 |
Nit | +0.032 |
Poverty | +0.031 |
Crit | +0.031 |
Lumpur | +0.028 |
Neg logit contributions | |
================================================================= | -0.038 |
Kislyak | -0.037 |
IMAGES | -0.037 |
advertising | -0.036 |
zinski | -0.035 |
Token | . |
Token ID | 13 |
Feature activation | -3.40e-02 |
Pos logit contributions | |
Kislyak | +2.795 |
================================================================= | +2.791 |
IMAGES | +2.665 |
advertising | +2.593 |
Accessory | +2.537 |
Neg logit contributions | |
Crit | -3.531 |
Nit | -3.285 |
Poverty | -3.146 |
Crit | -3.098 |
Adapt | -2.943 |
Token | Rather |
Token ID | 11317 |
Feature activation | -4.69e-02 |
Pos logit contributions | |
Crit | +0.083 |
Nit | +0.075 |
Poverty | +0.073 |
Crit | +0.072 |
Adapt | +0.065 |
Neg logit contributions | |
Kislyak | -0.094 |
================================================================= | -0.094 |
IMAGES | -0.091 |
advertising | -0.089 |
zinski | -0.089 |
Token | , |
Token ID | 11 |
Feature activation | -1.02e-01 |
Pos logit contributions | |
Crit | +0.132 |
Nit | +0.122 |
Poverty | +0.117 |
Crit | +0.115 |
corruption | +0.108 |
Neg logit contributions | |
Kislyak | -0.113 |
================================================================= | -0.113 |
IMAGES | -0.108 |
advertising | -0.105 |
zinski | -0.104 |
Token | Americans |
Token ID | 3399 |
Feature activation | +1.66e-01 |
Pos logit contributions | |
Crit | +0.272 |
Nit | +0.248 |
Crit | +0.239 |
Poverty | +0.239 |
Ps | +0.219 |
Neg logit contributions | |
================================================================= | -0.260 |
Kislyak | -0.258 |
IMAGES | -0.252 |
advertising | -0.245 |
zinski | -0.243 |
Token | must |
Token ID | 1276 |
Feature activation | -4.04e-03 |
Pos logit contributions | |
Kislyak | +0.507 |
================================================================= | +0.506 |
IMAGES | +0.491 |
advertising | +0.480 |
zinski | +0.475 |
Neg logit contributions | |
Crit | -0.360 |
Nit | -0.324 |
Poverty | -0.307 |
Crit | -0.302 |
corruption | -0.274 |
Token | lock |
Token ID | 5354 |
Feature activation | +7.46e-03 |
Pos logit contributions | |
Crit | +1.126 |
Nit | +1.048 |
Poverty | +0.998 |
Crit | +0.990 |
Adapt | +0.922 |
Neg logit contributions | |
================================================================= | -1.021 |
Kislyak | -1.011 |
IMAGES | -0.995 |
advertising | -0.947 |
Accessory | -0.945 |
Token | device |
Token ID | 3335 |
Feature activation | -2.97e-01 |
Pos logit contributions | |
Kislyak | +0.019 |
================================================================= | +0.019 |
IMAGES | +0.019 |
advertising | +0.018 |
zinski | +0.018 |
Neg logit contributions | |
Crit | -0.020 |
Nit | -0.018 |
Poverty | -0.017 |
Crit | -0.017 |
Ps | -0.016 |
Token |
|
Token ID | 198 |
Feature activation | -4.09e-01 |
Pos logit contributions | |
Crit | +0.749 |
Nit | +0.686 |
Poverty | +0.657 |
Crit | +0.651 |
Ps | +0.595 |
Neg logit contributions | |
Kislyak | -0.800 |
================================================================= | -0.796 |
IMAGES | -0.769 |
advertising | -0.756 |
zinski | -0.749 |
Token |
|
Token ID | 198 |
Feature activation | -2.48e-01 |
Pos logit contributions | |
Crit | +0.956 |
Nit | +0.865 |
Poverty | +0.819 |
Crit | +0.808 |
Adapt | +0.742 |
Neg logit contributions | |
Kislyak | -1.189 |
================================================================= | -1.178 |
IMAGES | -1.144 |
advertising | -1.109 |
zinski | -1.107 |
Token | A |
Token ID | 32 |
Feature activation | +2.28e-01 |
Pos logit contributions | |
Crit | +0.592 |
Nit | +0.538 |
Poverty | +0.500 |
Crit | +0.493 |
Adapt | +0.467 |
Neg logit contributions | |
Kislyak | -0.714 |
================================================================= | -0.708 |
IMAGES | -0.683 |
zinski | -0.663 |
advertising | -0.656 |
Token | Kirk |
Token ID | 13855 |
Feature activation | +1.05e+00 |
Pos logit contributions | |
Kislyak | +0.601 |
================================================================= | +0.595 |
IMAGES | +0.574 |
advertising | +0.559 |
Accessory | +0.557 |
Neg logit contributions | |
Crit | -0.593 |
Nit | -0.546 |
Poverty | -0.523 |
Crit | -0.518 |
corruption | -0.477 |
Token | land |
Token ID | 1044 |
Feature activation | +1.58e-02 |
Pos logit contributions | |
Kislyak | +2.581 |
================================================================= | +2.577 |
IMAGES | +2.468 |
advertising | +2.389 |
zinski | +2.370 |
Neg logit contributions | |
Crit | -2.896 |
Nit | -2.672 |
Poverty | -2.549 |
Crit | -2.507 |
corruption | -2.355 |
Token | man |
Token ID | 582 |
Feature activation | +3.10e-01 |
Pos logit contributions | |
Kislyak | +0.048 |
================================================================= | +0.048 |
IMAGES | +0.047 |
advertising | +0.046 |
zinski | +0.045 |
Neg logit contributions | |
Crit | -0.034 |
Nit | -0.031 |
Poverty | -0.029 |
Crit | -0.029 |
Lumpur | -0.026 |
Token | , |
Token ID | 11 |
Feature activation | -4.50e-02 |
Pos logit contributions | |
Kislyak | +0.808 |
================================================================= | +0.805 |
IMAGES | +0.776 |
advertising | +0.754 |
zinski | +0.747 |
Neg logit contributions | |
Crit | -0.812 |
Nit | -0.748 |
Poverty | -0.714 |
Crit | -0.704 |
corruption | -0.653 |
Token | convicted |
Token ID | 9118 |
Feature activation | -9.35e-02 |
Pos logit contributions | |
Crit | +0.119 |
Nit | +0.109 |
Poverty | +0.105 |
Crit | +0.104 |
corruption | +0.096 |
Neg logit contributions | |
================================================================= | -0.116 |
Kislyak | -0.116 |
IMAGES | -0.112 |
advertising | -0.109 |
Accessory | -0.108 |
Token | of |
Token ID | 286 |
Feature activation | -4.87e-01 |
Pos logit contributions | |
Crit | +0.253 |
Nit | +0.234 |
Poverty | +0.223 |
Crit | +0.219 |
Adapt | +0.205 |
Neg logit contributions | |
Kislyak | -0.236 |
================================================================= | -0.235 |
IMAGES | -0.226 |
advertising | -0.219 |
zinski | -0.217 |
Token | All |
Token ID | 1439 |
Feature activation | -1.54e-02 |
Pos logit contributions | |
Kislyak | +0.911 |
================================================================= | +0.904 |
IMAGES | +0.878 |
advertising | +0.864 |
zinski | +0.849 |
Neg logit contributions | |
Crit | -0.871 |
Nit | -0.797 |
Poverty | -0.767 |
Crit | -0.762 |
Ps | -0.692 |
Token | treatments |
Token ID | 13820 |
Feature activation | -7.89e-02 |
Pos logit contributions | |
Crit | +0.048 |
Nit | +0.044 |
Crit | +0.043 |
Poverty | +0.042 |
Ps | +0.040 |
Neg logit contributions | |
Kislyak | -0.033 |
================================================================= | -0.032 |
IMAGES | -0.031 |
advertising | -0.031 |
Accessory | -0.030 |
Token | were |
Token ID | 547 |
Feature activation | +3.35e-01 |
Pos logit contributions | |
Crit | +0.207 |
Nit | +0.188 |
Poverty | +0.182 |
Crit | +0.181 |
Ps | +0.166 |
Neg logit contributions | |
Kislyak | -0.207 |
================================================================= | -0.203 |
IMAGES | -0.200 |
advertising | -0.194 |
Accessory | -0.193 |
Token | self |
Token ID | 2116 |
Feature activation | -3.21e-01 |
Pos logit contributions | |
Kislyak | +0.920 |
================================================================= | +0.914 |
IMAGES | +0.892 |
advertising | +0.868 |
zinski | +0.858 |
Neg logit contributions | |
Crit | -0.837 |
Nit | -0.760 |
Poverty | -0.728 |
Crit | -0.726 |
Adapt | -0.662 |
Token | - |
Token ID | 12 |
Feature activation | +2.91e-01 |
Pos logit contributions | |
Crit | +0.849 |
Nit | +0.774 |
Poverty | +0.741 |
Crit | +0.738 |
Adapt | +0.678 |
Neg logit contributions | |
Kislyak | -0.837 |
================================================================= | -0.828 |
IMAGES | -0.807 |
advertising | -0.776 |
zinski | -0.772 |
Token | administ |
Token ID | 39081 |
Feature activation | +1.04e+00 |
Pos logit contributions | |
Kislyak | +0.581 |
================================================================= | +0.575 |
IMAGES | +0.553 |
Accessory | +0.523 |
advertising | +0.522 |
Neg logit contributions | |
Crit | -0.951 |
Nit | -0.886 |
Poverty | -0.850 |
Crit | -0.846 |
corruption | -0.800 |
Token | ered |
Token ID | 1068 |
Feature activation | +1.45e-01 |
Pos logit contributions | |
Kislyak | +1.995 |
================================================================= | +1.985 |
IMAGES | +1.890 |
advertising | +1.816 |
zinski | +1.790 |
Neg logit contributions | |
Crit | -3.433 |
Nit | -3.217 |
Poverty | -3.102 |
Crit | -3.068 |
corruption | -2.901 |
Token | by |
Token ID | 416 |
Feature activation | +8.09e-02 |
Pos logit contributions | |
Kislyak | +0.361 |
================================================================= | +0.357 |
IMAGES | +0.348 |
advertising | +0.339 |
Accessory | +0.335 |
Neg logit contributions | |
Crit | -0.395 |
Nit | -0.363 |
Crit | -0.348 |
Poverty | -0.348 |
Ps | -0.321 |
Token | the |
Token ID | 262 |
Feature activation | +1.29e-01 |
Pos logit contributions | |
Kislyak | +0.213 |
================================================================= | +0.212 |
IMAGES | +0.208 |
advertising | +0.203 |
Accessory | +0.200 |
Neg logit contributions | |
Crit | -0.207 |
Nit | -0.191 |
Poverty | -0.183 |
Crit | -0.182 |
Lumpur | -0.167 |
Token | patients |
Token ID | 3871 |
Feature activation | -3.72e-02 |
Pos logit contributions | |
Kislyak | +0.334 |
================================================================= | +0.332 |
IMAGES | +0.322 |
advertising | +0.316 |
Accessory | +0.311 |
Neg logit contributions | |
Crit | -0.340 |
Nit | -0.311 |
Poverty | -0.300 |
Crit | -0.299 |
Lumpur | -0.274 |
Token | at |
Token ID | 379 |
Feature activation | +8.90e-02 |
Pos logit contributions | |
Crit | +0.106 |
Nit | +0.097 |
Poverty | +0.094 |
Crit | +0.093 |
Ps | +0.086 |
Neg logit contributions | |
Kislyak | -0.089 |
================================================================= | -0.088 |
IMAGES | -0.086 |
advertising | -0.083 |
Accessory | -0.082 |
Token | Toronto |
Token ID | 6586 |
Feature activation | +3.12e-01 |
Pos logit contributions | |
Kislyak | +0.362 |
================================================================= | +0.360 |
IMAGES | +0.348 |
advertising | +0.341 |
zinski | +0.337 |
Neg logit contributions | |
Crit | -0.303 |
Nit | -0.274 |
Poverty | -0.263 |
Crit | -0.260 |
Ps | -0.236 |
Token | . |
Token ID | 13 |
Feature activation | +3.48e-02 |
Pos logit contributions | |
Kislyak | +0.768 |
================================================================= | +0.763 |
IMAGES | +0.736 |
advertising | +0.715 |
zinski | +0.707 |
Neg logit contributions | |
Crit | -0.866 |
Nit | -0.799 |
Poverty | -0.768 |
Crit | -0.757 |
corruption | -0.705 |
Token | But |
Token ID | 887 |
Feature activation | -2.49e-01 |
Pos logit contributions | |
Kislyak | +0.095 |
================================================================= | +0.095 |
IMAGES | +0.092 |
advertising | +0.090 |
zinski | +0.089 |
Neg logit contributions | |
Crit | -0.086 |
Nit | -0.079 |
Poverty | -0.076 |
Crit | -0.075 |
Adapt | -0.068 |
Token | the |
Token ID | 262 |
Feature activation | +7.15e-02 |
Pos logit contributions | |
Crit | +0.626 |
Nit | +0.569 |
Poverty | +0.549 |
Crit | +0.546 |
corruption | +0.499 |
Neg logit contributions | |
Kislyak | -0.671 |
================================================================= | -0.669 |
IMAGES | -0.650 |
advertising | -0.634 |
zinski | -0.629 |
Token | court |
Token ID | 2184 |
Feature activation | +3.78e-01 |
Pos logit contributions | |
Kislyak | +0.188 |
================================================================= | +0.187 |
IMAGES | +0.181 |
advertising | +0.177 |
zinski | +0.175 |
Neg logit contributions | |
Crit | -0.186 |
Nit | -0.170 |
Poverty | -0.164 |
Crit | -0.162 |
corruption | -0.149 |
Token | filings |
Token ID | 28058 |
Feature activation | +1.03e+00 |
Pos logit contributions | |
Kislyak | +1.091 |
================================================================= | +1.087 |
IMAGES | +1.052 |
advertising | +1.025 |
zinski | +1.016 |
Neg logit contributions | |
Crit | -0.884 |
Nit | -0.805 |
Poverty | -0.763 |
Crit | -0.751 |
corruption | -0.690 |
Token | make |
Token ID | 787 |
Feature activation | -1.21e-01 |
Pos logit contributions | |
Kislyak | +2.822 |
================================================================= | +2.810 |
IMAGES | +2.697 |
zinski | +2.605 |
advertising | +2.600 |
Neg logit contributions | |
Crit | -2.550 |
Nit | -2.347 |
Poverty | -2.213 |
Crit | -2.159 |
corruption | -2.053 |
Token | it |
Token ID | 340 |
Feature activation | -7.15e-02 |
Pos logit contributions | |
Crit | +0.300 |
Nit | +0.275 |
Poverty | +0.261 |
Crit | +0.258 |
corruption | +0.237 |
Neg logit contributions | |
Kislyak | -0.334 |
================================================================= | -0.333 |
IMAGES | -0.323 |
advertising | -0.313 |
zinski | -0.312 |
Token | very |
Token ID | 845 |
Feature activation | +6.47e-02 |
Pos logit contributions | |
Crit | +0.175 |
Nit | +0.160 |
Poverty | +0.152 |
Crit | +0.151 |
Ps | +0.139 |
Neg logit contributions | |
Kislyak | -0.198 |
================================================================= | -0.198 |
IMAGES | -0.191 |
advertising | -0.186 |
zinski | -0.184 |
Token | clear |
Token ID | 1598 |
Feature activation | +5.32e-01 |
Pos logit contributions | |
Kislyak | +0.180 |
================================================================= | +0.179 |
IMAGES | +0.173 |
advertising | +0.169 |
zinski | +0.167 |
Neg logit contributions | |
Crit | -0.158 |
Nit | -0.145 |
Poverty | -0.138 |
Crit | -0.136 |
Ps | -0.125 |
Token | the |
Token ID | 262 |
Feature activation | +1.65e-01 |
Pos logit contributions | |
Kislyak | +1.376 |
================================================================= | +1.370 |
IMAGES | +1.320 |
advertising | +1.282 |
zinski | +1.269 |
Neg logit contributions | |
Crit | -1.405 |
Nit | -1.295 |
Poverty | -1.235 |
Crit | -1.216 |
Adapt | -1.134 |
Token | happens |
Token ID | 4325 |
Feature activation | -3.76e-02 |
Pos logit contributions | |
Kislyak | +0.030 |
================================================================= | +0.030 |
IMAGES | +0.029 |
advertising | +0.028 |
zinski | +0.028 |
Neg logit contributions | |
Crit | -0.028 |
Nit | -0.026 |
Poverty | -0.025 |
Crit | -0.024 |
Adapt | -0.022 |
Token | , |
Token ID | 11 |
Feature activation | +1.59e-01 |
Pos logit contributions | |
Crit | +0.107 |
Nit | +0.099 |
Poverty | +0.095 |
Crit | +0.094 |
corruption | +0.088 |
Neg logit contributions | |
Kislyak | -0.089 |
================================================================= | -0.089 |
IMAGES | -0.086 |
advertising | -0.083 |
zinski | -0.082 |
Token | its |
Token ID | 663 |
Feature activation | -7.75e-02 |
Pos logit contributions | |
Kislyak | +0.425 |
================================================================= | +0.424 |
IMAGES | +0.409 |
advertising | +0.399 |
zinski | +0.394 |
Neg logit contributions | |
Crit | -0.404 |
Nit | -0.371 |
Poverty | -0.354 |
Crit | -0.349 |
corruption | -0.323 |
Token | top |
Token ID | 1353 |
Feature activation | +6.52e-01 |
Pos logit contributions | |
Crit | +0.201 |
Nit | +0.185 |
Poverty | +0.177 |
Crit | +0.175 |
Adapt | +0.162 |
Neg logit contributions | |
Kislyak | -0.203 |
================================================================= | -0.202 |
IMAGES | -0.196 |
advertising | -0.191 |
zinski | -0.188 |
Token | ( |
Token ID | 357 |
Feature activation | +1.92e-01 |
Pos logit contributions | |
Kislyak | +1.653 |
================================================================= | +1.647 |
IMAGES | +1.584 |
advertising | +1.538 |
zinski | +1.525 |
Neg logit contributions | |
Crit | -1.755 |
Nit | -1.619 |
Poverty | -1.547 |
Crit | -1.526 |
Adapt | -1.420 |
Token | above |
Token ID | 29370 |
Feature activation | +1.03e+00 |
Pos logit contributions | |
Kislyak | +0.458 |
================================================================= | +0.456 |
IMAGES | +0.440 |
advertising | +0.422 |
zinski | +0.419 |
Neg logit contributions | |
Crit | -0.548 |
Nit | -0.508 |
Poverty | -0.484 |
Crit | -0.478 |
corruption | -0.450 |
Token | water |
Token ID | 1660 |
Feature activation | -3.79e-01 |
Pos logit contributions | |
Kislyak | +1.936 |
================================================================= | +1.915 |
IMAGES | +1.826 |
advertising | +1.745 |
zinski | +1.713 |
Neg logit contributions | |
Crit | -3.417 |
Nit | -3.203 |
Poverty | -3.063 |
Crit | -3.035 |
Adapt | -2.910 |
Token | ) |
Token ID | 8 |
Feature activation | -1.95e-02 |
Pos logit contributions | |
Crit | +1.225 |
Nit | +1.147 |
Poverty | +1.107 |
Crit | +1.094 |
Ps | +1.030 |
Neg logit contributions | |
Kislyak | -0.758 |
================================================================= | -0.750 |
IMAGES | -0.717 |
advertising | -0.695 |
Accessory | -0.685 |
Token | falls |
Token ID | 8953 |
Feature activation | +1.19e-01 |
Pos logit contributions | |
Crit | +0.051 |
Nit | +0.047 |
Poverty | +0.045 |
Crit | +0.044 |
corruption | +0.041 |
Neg logit contributions | |
Kislyak | -0.051 |
================================================================= | -0.051 |
IMAGES | -0.049 |
advertising | -0.048 |
zinski | -0.047 |
Token | and |
Token ID | 290 |
Feature activation | -1.46e-02 |
Pos logit contributions | |
Kislyak | +0.294 |
================================================================= | +0.292 |
IMAGES | +0.281 |
advertising | +0.275 |
zinski | +0.271 |
Neg logit contributions | |
Crit | -0.328 |
Nit | -0.302 |
Poverty | -0.290 |
Crit | -0.287 |
corruption | -0.266 |
Token | pushes |
Token ID | 20070 |
Feature activation | +2.18e-03 |
Pos logit contributions | |
Crit | +0.039 |
Nit | +0.035 |
Poverty | +0.034 |
Crit | +0.034 |
corruption | +0.031 |
Neg logit contributions | |
Kislyak | -0.038 |
================================================================= | -0.037 |
IMAGES | -0.036 |
advertising | -0.035 |
zinski | -0.035 |
Token | react |
Token ID | 6324 |
Feature activation | +4.12e-01 |
Pos logit contributions | |
Crit | +0.151 |
Nit | +0.132 |
Poverty | +0.123 |
Crit | +0.121 |
corruption | +0.107 |
Neg logit contributions | |
Kislyak | -0.298 |
================================================================= | -0.296 |
IMAGES | -0.288 |
advertising | -0.283 |
zinski | -0.281 |
Token | , |
Token ID | 11 |
Feature activation | -4.76e-03 |
Pos logit contributions | |
Kislyak | +1.059 |
================================================================= | +1.055 |
IMAGES | +1.017 |
advertising | +0.988 |
zinski | +0.977 |
Neg logit contributions | |
Crit | -1.092 |
Nit | -1.007 |
Poverty | -0.964 |
Crit | -0.946 |
corruption | -0.883 |
Token | the |
Token ID | 262 |
Feature activation | -3.43e-02 |
Pos logit contributions | |
Crit | +0.013 |
Nit | +0.012 |
Poverty | +0.011 |
Crit | +0.011 |
corruption | +0.010 |
Neg logit contributions | |
Kislyak | -0.012 |
================================================================= | -0.012 |
IMAGES | -0.012 |
advertising | -0.011 |
zinski | -0.011 |
Token | body |
Token ID | 1767 |
Feature activation | -3.71e-01 |
Pos logit contributions | |
Crit | +0.084 |
Nit | +0.076 |
Poverty | +0.072 |
Crit | +0.072 |
corruption | +0.066 |
Neg logit contributions | |
Kislyak | -0.096 |
================================================================= | -0.095 |
IMAGES | -0.092 |
advertising | -0.090 |
zinski | -0.090 |
Token | will |
Token ID | 481 |
Feature activation | +3.32e-01 |
Pos logit contributions | |
Crit | +0.757 |
Nit | +0.677 |
Poverty | +0.634 |
Crit | +0.627 |
corruption | +0.566 |
Neg logit contributions | |
Kislyak | -1.185 |
================================================================= | -1.182 |
IMAGES | -1.147 |
advertising | -1.120 |
zinski | -1.114 |
Token | literally |
Token ID | 7360 |
Feature activation | +1.01e+00 |
Pos logit contributions | |
Kislyak | +1.010 |
================================================================= | +1.009 |
IMAGES | +0.977 |
advertising | +0.952 |
zinski | +0.944 |
Neg logit contributions | |
Crit | -0.723 |
Nit | -0.656 |
Poverty | -0.618 |
Crit | -0.604 |
Adapt | -0.553 |
Token | push |
Token ID | 4574 |
Feature activation | +4.44e-02 |
Pos logit contributions | |
================================================================= | +3.120 |
Kislyak | +3.092 |
IMAGES | +3.018 |
advertising | +2.913 |
DEA | +2.895 |
Neg logit contributions | |
Crit | -2.122 |
Nit | -1.951 |
Poverty | -1.809 |
Crit | -1.745 |
Lumpur | -1.634 |
Token | it |
Token ID | 340 |
Feature activation | -4.53e-02 |
Pos logit contributions | |
Kislyak | +0.119 |
================================================================= | +0.118 |
IMAGES | +0.114 |
advertising | +0.111 |
zinski | +0.111 |
Neg logit contributions | |
Crit | -0.114 |
Nit | -0.104 |
Poverty | -0.099 |
Crit | -0.099 |
corruption | -0.091 |
Token | out |
Token ID | 503 |
Feature activation | +1.09e-01 |
Pos logit contributions | |
Crit | +0.125 |
Nit | +0.115 |
Poverty | +0.110 |
Crit | +0.110 |
crit | +0.101 |
Neg logit contributions | |
Kislyak | -0.112 |
================================================================= | -0.111 |
IMAGES | -0.107 |
zinski | -0.104 |
advertising | -0.104 |
Token | , |
Token ID | 11 |
Feature activation | -1.88e-01 |
Pos logit contributions | |
Kislyak | +0.272 |
================================================================= | +0.270 |
IMAGES | +0.261 |
advertising | +0.253 |
zinski | +0.251 |
Neg logit contributions | |
Crit | -0.298 |
Nit | -0.274 |
Poverty | -0.262 |
Crit | -0.259 |
corruption | -0.242 |
Token | causing |
Token ID | 6666 |
Feature activation | +4.19e-02 |
Pos logit contributions | |
Crit | +0.508 |
Nit | +0.468 |
Poverty | +0.447 |
Crit | +0.444 |
corruption | +0.411 |
Neg logit contributions | |
Kislyak | -0.477 |
================================================================= | -0.475 |
IMAGES | -0.458 |
advertising | -0.444 |
zinski | -0.441 |
Token | 13 |
Token ID | 1511 |
Feature activation | +1.39e-01 |
Pos logit contributions | |
Kislyak | +0.085 |
================================================================= | +0.085 |
IMAGES | +0.082 |
advertising | +0.079 |
Accessory | +0.078 |
Neg logit contributions | |
Crit | -0.111 |
Nit | -0.102 |
Poverty | -0.099 |
Crit | -0.098 |
Adapt | -0.091 |
Token | , |
Token ID | 11 |
Feature activation | +8.19e-03 |
Pos logit contributions | |
Kislyak | +0.319 |
================================================================= | +0.319 |
IMAGES | +0.305 |
advertising | +0.296 |
zinski | +0.292 |
Neg logit contributions | |
Crit | -0.405 |
Nit | -0.377 |
Poverty | -0.361 |
Crit | -0.356 |
corruption | -0.335 |
Token | 2010 |
Token ID | 3050 |
Feature activation | -2.58e-02 |
Pos logit contributions | |
Kislyak | +0.020 |
================================================================= | +0.020 |
IMAGES | +0.020 |
advertising | +0.019 |
zinski | +0.019 |
Neg logit contributions | |
Crit | -0.022 |
Nit | -0.021 |
Poverty | -0.020 |
Crit | -0.019 |
corruption | -0.018 |
Token |
|
Token ID | 198 |
Feature activation | +1.77e-02 |
Pos logit contributions | |
Crit | +0.064 |
Nit | +0.058 |
Poverty | +0.055 |
Crit | +0.054 |
corruption | +0.051 |
Neg logit contributions | |
Kislyak | -0.071 |
================================================================= | -0.071 |
IMAGES | -0.069 |
advertising | -0.067 |
zinski | -0.066 |
Token |
|
Token ID | 198 |
Feature activation | +6.35e-03 |
Pos logit contributions | |
Kislyak | +0.049 |
================================================================= | +0.049 |
IMAGES | +0.047 |
advertising | +0.046 |
zinski | +0.046 |
Neg logit contributions | |
Crit | -0.043 |
Nit | -0.040 |
Poverty | -0.038 |
Crit | -0.037 |
corruption | -0.034 |
Token | Subject |
Token ID | 19776 |
Feature activation | +1.00e+00 |
Pos logit contributions | |
Kislyak | +0.018 |
================================================================= | +0.018 |
IMAGES | +0.018 |
advertising | +0.017 |
zinski | +0.017 |
Neg logit contributions | |
Crit | -0.015 |
Nit | -0.014 |
Poverty | -0.013 |
Crit | -0.013 |
Adapt | -0.012 |
Token | : |
Token ID | 25 |
Feature activation | +1.79e-01 |
Pos logit contributions | |
================================================================= | +1.770 |
Kislyak | +1.743 |
IMAGES | +1.627 |
advertising | +1.607 |
zinski | +1.560 |
Neg logit contributions | |
Crit | -3.435 |
Nit | -3.252 |
Poverty | -3.118 |
Crit | -3.073 |
Lumpur | -2.975 |
Token | boys |
Token ID | 6510 |
Feature activation | +2.92e-01 |
Pos logit contributions | |
Kislyak | +0.476 |
================================================================= | +0.473 |
IMAGES | +0.458 |
advertising | +0.446 |
zinski | +0.442 |
Neg logit contributions | |
Crit | -0.462 |
Nit | -0.423 |
Poverty | -0.406 |
Crit | -0.401 |
Adapt | -0.369 |
Token | beware |
Token ID | 40600 |
Feature activation | -2.03e-01 |
Pos logit contributions | |
Kislyak | +0.805 |
================================================================= | +0.803 |
IMAGES | +0.776 |
advertising | +0.755 |
zinski | +0.748 |
Neg logit contributions | |
Crit | -0.722 |
Nit | -0.661 |
Poverty | -0.628 |
Crit | -0.619 |
corruption | -0.572 |
Token | This |
Token ID | 770 |
Feature activation | +4.32e-01 |
Pos logit contributions | |
Crit | +0.544 |
Nit | +0.500 |
Poverty | +0.483 |
Crit | +0.479 |
Adapt | +0.439 |
Neg logit contributions | |
Kislyak | -0.518 |
================================================================= | -0.507 |
IMAGES | -0.498 |
Accessory | -0.485 |
advertising | -0.482 |
Token | film |
Token ID | 2646 |
Feature activation | +2.73e-01 |
Pos logit contributions | |
Kislyak | +1.164 |
================================================================= | +1.161 |
IMAGES | +1.121 |
advertising | +1.090 |
zinski | +1.079 |
Neg logit contributions | |
Crit | -1.092 |
Nit | -1.003 |
Poverty | -0.955 |
Crit | -0.939 |
corruption | -0.872 |
Token | . |
Token ID | 13 |
Feature activation | -6.21e-02 |
Pos logit contributions | |
Kislyak | +1.129 |
================================================================= | +1.124 |
IMAGES | +1.077 |
advertising | +1.041 |
zinski | +1.028 |
Neg logit contributions | |
Crit | -1.548 |
Nit | -1.441 |
Poverty | -1.385 |
Crit | -1.368 |
corruption | -1.285 |
Token | 25 |
Token ID | 1679 |
Feature activation | +1.51e-01 |
Pos logit contributions | |
Crit | +0.151 |
Nit | +0.138 |
Poverty | +0.133 |
Crit | +0.131 |
Adapt | +0.119 |
Neg logit contributions | |
Kislyak | -0.173 |
================================================================= | -0.173 |
IMAGES | -0.166 |
advertising | -0.163 |
zinski | -0.162 |
Token | Journal |
Token ID | 4913 |
Feature activation | +5.52e-03 |
Pos logit contributions | |
Kislyak | +0.368 |
================================================================= | +0.367 |
IMAGES | +0.352 |
advertising | +0.343 |
zinski | +0.339 |
Neg logit contributions | |
Crit | -0.420 |
Nit | -0.388 |
Poverty | -0.372 |
Crit | -0.368 |
corruption | -0.342 |
Token | Square |
Token ID | 9276 |
Feature activation | +8.20e-02 |
Pos logit contributions | |
Kislyak | +0.017 |
================================================================= | +0.017 |
IMAGES | +0.017 |
advertising | +0.017 |
zinski | +0.016 |
Neg logit contributions | |
Crit | -0.011 |
Nit | -0.010 |
Poverty | -0.010 |
Crit | -0.010 |
Ps | -0.009 |
Token | is |
Token ID | 318 |
Feature activation | -6.43e-02 |
Pos logit contributions | |
Kislyak | +0.226 |
================================================================= | +0.226 |
IMAGES | +0.217 |
advertising | +0.213 |
zinski | +0.211 |
Neg logit contributions | |
Crit | -0.202 |
Nit | -0.185 |
Poverty | -0.176 |
Crit | -0.174 |
Adapt | -0.160 |
Token | literally |
Token ID | 7360 |
Feature activation | +9.92e-01 |
Pos logit contributions | |
Crit | +0.169 |
Nit | +0.155 |
Poverty | +0.149 |
Crit | +0.148 |
Ps | +0.135 |
Neg logit contributions | |
================================================================= | -0.167 |
Kislyak | -0.167 |
IMAGES | -0.162 |
advertising | -0.157 |
zinski | -0.156 |
Token | next |
Token ID | 1306 |
Feature activation | -2.01e-02 |
Pos logit contributions | |
Kislyak | +2.539 |
================================================================= | +2.526 |
IMAGES | +2.439 |
advertising | +2.358 |
Accessory | +2.329 |
Neg logit contributions | |
Crit | -2.639 |
Nit | -2.438 |
Poverty | -2.312 |
Crit | -2.278 |
corruption | -2.138 |
Token | door |
Token ID | 3420 |
Feature activation | -1.75e-02 |
Pos logit contributions | |
Crit | +0.057 |
Nit | +0.053 |
Poverty | +0.050 |
Crit | +0.050 |
corruption | +0.047 |
Neg logit contributions | |
Kislyak | -0.048 |
================================================================= | -0.048 |
IMAGES | -0.046 |
advertising | -0.045 |
Accessory | -0.044 |
Token | to |
Token ID | 284 |
Feature activation | +9.65e-02 |
Pos logit contributions | |
Crit | +0.051 |
Nit | +0.047 |
Poverty | +0.045 |
Crit | +0.045 |
corruption | +0.042 |
Neg logit contributions | |
Kislyak | -0.041 |
================================================================= | -0.040 |
IMAGES | -0.039 |
advertising | -0.038 |
zinski | -0.037 |
Token | the |
Token ID | 262 |
Feature activation | -1.16e-01 |
Pos logit contributions | |
Kislyak | +0.257 |
================================================================= | +0.257 |
IMAGES | +0.247 |
advertising | +0.242 |
zinski | +0.240 |
Neg logit contributions | |
Crit | -0.248 |
Nit | -0.226 |
Poverty | -0.219 |
Crit | -0.216 |
Lumpur | -0.197 |
Token | PATH |
Token ID | 46490 |
Feature activation | +1.66e-01 |
Pos logit contributions | |
Crit | +0.310 |
Nit | +0.285 |
Poverty | +0.277 |
Crit | +0.275 |
Lumpur | +0.250 |
Neg logit contributions | |
================================================================= | -0.295 |
Kislyak | -0.295 |
IMAGES | -0.281 |
advertising | -0.278 |
zinski | -0.274 |
Token | When |
Token ID | 1649 |
Feature activation | +8.76e-02 |
Pos logit contributions | |
Kislyak | +0.076 |
================================================================= | +0.076 |
IMAGES | +0.074 |
advertising | +0.072 |
zinski | +0.071 |
Neg logit contributions | |
Crit | -0.067 |
Nit | -0.061 |
Poverty | -0.058 |
Crit | -0.058 |
Adapt | -0.052 |
Token | I |
Token ID | 314 |
Feature activation | +1.29e-01 |
Pos logit contributions | |
================================================================= | +0.227 |
Kislyak | +0.226 |
IMAGES | +0.220 |
advertising | +0.214 |
zinski | +0.212 |
Neg logit contributions | |
Crit | -0.230 |
Nit | -0.211 |
Poverty | -0.202 |
Crit | -0.201 |
Ps | -0.185 |
Token | have |
Token ID | 423 |
Feature activation | +2.45e-03 |
Pos logit contributions | |
Kislyak | +0.359 |
================================================================= | +0.358 |
IMAGES | +0.347 |
advertising | +0.338 |
Accessory | +0.335 |
Neg logit contributions | |
Crit | -0.315 |
Nit | -0.287 |
Poverty | -0.274 |
Crit | -0.270 |
corruption | -0.248 |
Token | taken |
Token ID | 2077 |
Feature activation | +1.20e-01 |
Pos logit contributions | |
================================================================= | +0.007 |
Kislyak | +0.007 |
IMAGES | +0.007 |
advertising | +0.006 |
zinski | +0.006 |
Neg logit contributions | |
Crit | -0.006 |
Nit | -0.005 |
Poverty | -0.005 |
Crit | -0.005 |
crit | -0.005 |
Token | the |
Token ID | 262 |
Feature activation | -7.01e-02 |
Pos logit contributions | |
================================================================= | +0.309 |
Kislyak | +0.307 |
IMAGES | +0.298 |
advertising | +0.292 |
zinski | +0.290 |
Neg logit contributions | |
Crit | -0.314 |
Nit | -0.287 |
Poverty | -0.277 |
Crit | -0.276 |
Ps | -0.253 |
Token | train |
Token ID | 4512 |
Feature activation | +9.91e-01 |
Pos logit contributions | |
Crit | +0.183 |
Nit | +0.167 |
Crit | +0.161 |
Poverty | +0.161 |
Lumpur | +0.147 |
Neg logit contributions | |
================================================================= | -0.182 |
Kislyak | -0.182 |
IMAGES | -0.176 |
advertising | -0.174 |
Accessory | -0.170 |
Token | from |
Token ID | 422 |
Feature activation | +9.61e-02 |
Pos logit contributions | |
Kislyak | +2.471 |
================================================================= | +2.459 |
IMAGES | +2.365 |
advertising | +2.288 |
zinski | +2.260 |
Neg logit contributions | |
Crit | -2.704 |
Nit | -2.503 |
Poverty | -2.389 |
Crit | -2.346 |
Adapt | -2.206 |
Token | Pyongyang |
Token ID | 21614 |
Feature activation | -3.51e-01 |
Pos logit contributions | |
Kislyak | +0.242 |
================================================================= | +0.242 |
IMAGES | +0.234 |
advertising | +0.227 |
zinski | +0.225 |
Neg logit contributions | |
Crit | -0.259 |
Nit | -0.238 |
Poverty | -0.230 |
Crit | -0.227 |
Lumpur | -0.210 |
Token | to |
Token ID | 284 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
Crit | +0.955 |
Nit | +0.879 |
Poverty | +0.846 |
Crit | +0.841 |
Lumpur | +0.790 |
Neg logit contributions | |
================================================================= | -0.873 |
Kislyak | -0.871 |
IMAGES | -0.848 |
advertising | -0.822 |
zinski | -0.811 |
Token | Beijing |
Token ID | 11618 |
Feature activation | +1.21e-01 |
Pos logit contributions | |
Crit | +0.366 |
Nit | +0.334 |
Poverty | +0.324 |
Crit | +0.322 |
Lumpur | +0.296 |
Neg logit contributions | |
================================================================= | -0.361 |
Kislyak | -0.360 |
IMAGES | -0.351 |
advertising | -0.341 |
zinski | -0.338 |
Token | , |
Token ID | 11 |
Feature activation | +1.34e-01 |
Pos logit contributions | |
Kislyak | +0.301 |
================================================================= | +0.301 |
IMAGES | +0.290 |
advertising | +0.282 |
zinski | +0.278 |
Neg logit contributions | |
Crit | -0.331 |
Nit | -0.305 |
Poverty | -0.292 |
Crit | -0.289 |
Lumpur | -0.270 |
Token |
|
Token ID | 198 |
Feature activation | +2.56e-01 |
Pos logit contributions | |
Kislyak | +0.864 |
================================================================= | +0.862 |
IMAGES | +0.828 |
advertising | +0.813 |
zinski | +0.804 |
Neg logit contributions | |
Crit | -0.806 |
Nit | -0.742 |
Poverty | -0.709 |
Crit | -0.695 |
Lumpur | -0.655 |
Token | � |
Token ID | 447 |
Feature activation | +3.09e-01 |
Pos logit contributions | |
Kislyak | +0.783 |
================================================================= | +0.781 |
IMAGES | +0.755 |
advertising | +0.746 |
Accessory | +0.736 |
Neg logit contributions | |
Crit | -0.548 |
Nit | -0.497 |
Poverty | -0.476 |
Crit | -0.465 |
Lumpur | -0.425 |
Token | � |
Token ID | 250 |
Feature activation | -8.32e-02 |
Pos logit contributions | |
Kislyak | +0.757 |
================================================================= | +0.754 |
IMAGES | +0.722 |
advertising | +0.704 |
zinski | +0.694 |
Neg logit contributions | |
Crit | -0.858 |
Nit | -0.797 |
Poverty | -0.763 |
Crit | -0.750 |
corruption | -0.703 |
Token | Mark |
Token ID | 9704 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
Crit | +0.175 |
Nit | +0.156 |
Poverty | +0.146 |
Crit | +0.144 |
corruption | +0.131 |
Neg logit contributions | |
Kislyak | -0.263 |
================================================================= | -0.262 |
IMAGES | -0.255 |
zinski | -0.246 |
advertising | -0.246 |
Token | was |
Token ID | 373 |
Feature activation | +4.20e-03 |
Pos logit contributions | |
Kislyak | +0.292 |
================================================================= | +0.291 |
IMAGES | +0.281 |
advertising | +0.272 |
zinski | +0.269 |
Neg logit contributions | |
Crit | -0.303 |
Nit | -0.279 |
Poverty | -0.266 |
Crit | -0.263 |
corruption | -0.244 |
Token | literally |
Token ID | 7360 |
Feature activation | +9.88e-01 |
Pos logit contributions | |
================================================================= | +0.011 |
Kislyak | +0.011 |
IMAGES | +0.011 |
advertising | +0.010 |
Accessory | +0.010 |
Neg logit contributions | |
Crit | -0.011 |
Nit | -0.010 |
Crit | -0.009 |
Poverty | -0.009 |
Lumpur | -0.009 |
Token | terrified |
Token ID | 22144 |
Feature activation | +7.78e-02 |
Pos logit contributions | |
Kislyak | +2.592 |
================================================================= | +2.583 |
IMAGES | +2.488 |
advertising | +2.413 |
zinski | +2.388 |
Neg logit contributions | |
Crit | -2.565 |
Nit | -2.369 |
Poverty | -2.254 |
Crit | -2.213 |
corruption | -2.071 |
Token | and |
Token ID | 290 |
Feature activation | +8.73e-02 |
Pos logit contributions | |
Kislyak | +0.193 |
================================================================= | +0.192 |
IMAGES | +0.186 |
advertising | +0.180 |
zinski | +0.179 |
Neg logit contributions | |
Crit | -0.213 |
Nit | -0.196 |
Poverty | -0.188 |
Crit | -0.187 |
corruption | -0.173 |
Token | as |
Token ID | 355 |
Feature activation | +1.93e-01 |
Pos logit contributions | |
================================================================= | +0.236 |
Kislyak | +0.236 |
IMAGES | +0.230 |
advertising | +0.224 |
zinski | +0.222 |
Neg logit contributions | |
Crit | -0.220 |
Nit | -0.199 |
Crit | -0.191 |
Poverty | -0.190 |
Ps | -0.174 |
Token | a |
Token ID | 257 |
Feature activation | -6.79e-02 |
Pos logit contributions | |
Kislyak | +0.538 |
================================================================= | +0.535 |
IMAGES | +0.519 |
advertising | +0.507 |
zinski | +0.501 |
Neg logit contributions | |
Crit | -0.471 |
Nit | -0.427 |
Poverty | -0.410 |
Crit | -0.405 |
Ps | -0.371 |
Token | result |
Token ID | 1255 |
Feature activation | +2.90e-01 |
Pos logit contributions | |
Crit | +0.183 |
Nit | +0.169 |
Poverty | +0.162 |
Crit | +0.162 |
Lumpur | +0.149 |
Neg logit contributions | |
Kislyak | -0.171 |
================================================================= | -0.169 |
IMAGES | -0.164 |
advertising | -0.160 |
zinski | -0.159 |
Token | the |
Token ID | 262 |
Feature activation | +4.82e-03 |
Pos logit contributions | |
Crit | +0.025 |
Nit | +0.023 |
Poverty | +0.022 |
Crit | +0.021 |
Adapt | +0.020 |
Neg logit contributions | |
Kislyak | -0.027 |
================================================================= | -0.027 |
IMAGES | -0.026 |
advertising | -0.025 |
zinski | -0.025 |
Token | ankles |
Token ID | 42415 |
Feature activation | +2.83e-01 |
Pos logit contributions | |
Kislyak | +0.013 |
================================================================= | +0.013 |
IMAGES | +0.013 |
advertising | +0.012 |
zinski | +0.012 |
Neg logit contributions | |
Crit | -0.012 |
Nit | -0.011 |
Poverty | -0.011 |
Crit | -0.011 |
corruption | -0.010 |
Token | and |
Token ID | 290 |
Feature activation | +4.45e-02 |
Pos logit contributions | |
Kislyak | +0.707 |
================================================================= | +0.703 |
IMAGES | +0.679 |
advertising | +0.659 |
zinski | +0.652 |
Neg logit contributions | |
Crit | -0.772 |
Nit | -0.713 |
Poverty | -0.681 |
Crit | -0.673 |
corruption | -0.627 |
Token | climb |
Token ID | 12080 |
Feature activation | -2.10e-02 |
Pos logit contributions | |
Kislyak | +0.123 |
================================================================= | +0.122 |
IMAGES | +0.119 |
advertising | +0.117 |
zinski | +0.115 |
Neg logit contributions | |
Crit | -0.109 |
Nit | -0.099 |
Poverty | -0.095 |
Crit | -0.094 |
corruption | -0.086 |
Token | on |
Token ID | 319 |
Feature activation | +4.62e-01 |
Pos logit contributions | |
Crit | +0.056 |
Nit | +0.052 |
Poverty | +0.050 |
Crit | +0.049 |
Ps | +0.045 |
Neg logit contributions | |
Kislyak | -0.053 |
================================================================= | -0.053 |
IMAGES | -0.051 |
advertising | -0.050 |
zinski | -0.049 |
Token | top |
Token ID | 1353 |
Feature activation | +9.82e-01 |
Pos logit contributions | |
Kislyak | +1.177 |
================================================================= | +1.172 |
IMAGES | +1.131 |
advertising | +1.098 |
zinski | +1.087 |
Neg logit contributions | |
Crit | -1.237 |
Nit | -1.140 |
Poverty | -1.089 |
Crit | -1.075 |
corruption | -1.000 |
Token | of |
Token ID | 286 |
Feature activation | +1.84e-01 |
Pos logit contributions | |
================================================================= | +2.257 |
Kislyak | +2.241 |
IMAGES | +2.118 |
advertising | +2.070 |
zinski | +2.033 |
Neg logit contributions | |
Crit | -2.844 |
Nit | -2.652 |
Poverty | -2.564 |
Crit | -2.490 |
Adapt | -2.380 |
Token | them |
Token ID | 606 |
Feature activation | -8.00e-02 |
Pos logit contributions | |
Kislyak | +0.521 |
================================================================= | +0.518 |
IMAGES | +0.502 |
advertising | +0.489 |
zinski | +0.486 |
Neg logit contributions | |
Crit | -0.442 |
Nit | -0.403 |
Poverty | -0.383 |
Crit | -0.379 |
corruption | -0.347 |
Token | . |
Token ID | 13 |
Feature activation | +1.31e-01 |
Pos logit contributions | |
Crit | +0.232 |
Nit | +0.215 |
Poverty | +0.206 |
Crit | +0.204 |
corruption | +0.191 |
Neg logit contributions | |
Kislyak | -0.186 |
================================================================= | -0.184 |
IMAGES | -0.177 |
advertising | -0.172 |
zinski | -0.171 |
Token | That |
Token ID | 1320 |
Feature activation | +3.44e-01 |
Pos logit contributions | |
Kislyak | +0.366 |
================================================================= | +0.363 |
IMAGES | +0.352 |
advertising | +0.343 |
zinski | +0.342 |
Neg logit contributions | |
Crit | -0.320 |
Nit | -0.293 |
Poverty | -0.280 |
Crit | -0.276 |
Adapt | -0.253 |
Token | � |
Token ID | 447 |
Feature activation | -7.12e-02 |
Pos logit contributions | |
Kislyak | +0.858 |
================================================================= | +0.855 |
IMAGES | +0.823 |
advertising | +0.799 |
zinski | +0.790 |
Neg logit contributions | |
Crit | -0.939 |
Nit | -0.867 |
Poverty | -0.829 |
Crit | -0.818 |
corruption | -0.763 |
Token | up |
Token ID | 510 |
Feature activation | +2.53e-01 |
Pos logit contributions | |
Kislyak | +0.687 |
================================================================= | +0.683 |
IMAGES | +0.661 |
advertising | +0.642 |
zinski | +0.638 |
Neg logit contributions | |
Crit | -0.663 |
Nit | -0.608 |
Poverty | -0.582 |
Crit | -0.575 |
corruption | -0.530 |
Token | for |
Token ID | 329 |
Feature activation | +3.37e-01 |
Pos logit contributions | |
Kislyak | +0.677 |
================================================================= | +0.676 |
IMAGES | +0.652 |
advertising | +0.634 |
zinski | +0.628 |
Neg logit contributions | |
Crit | -0.646 |
Nit | -0.594 |
Poverty | -0.565 |
Crit | -0.556 |
corruption | -0.516 |
Token | BRE |
Token ID | 29377 |
Feature activation | -1.93e-01 |
Pos logit contributions | |
Kislyak | +0.720 |
================================================================= | +0.713 |
IMAGES | +0.681 |
advertising | +0.659 |
zinski | +0.651 |
Neg logit contributions | |
Crit | -1.048 |
Nit | -0.975 |
Poverty | -0.946 |
Crit | -0.936 |
corruption | -0.874 |
Token | AKING |
Token ID | 43602 |
Feature activation | -4.31e-01 |
Pos logit contributions | |
Crit | +0.499 |
Nit | +0.453 |
Crit | +0.437 |
Poverty | +0.436 |
Ps | +0.411 |
Neg logit contributions | |
Kislyak | -0.513 |
================================================================= | -0.510 |
IMAGES | -0.501 |
Accessory | -0.482 |
advertising | -0.473 |
Token | NEWS |
Token ID | 28840 |
Feature activation | +3.49e-01 |
Pos logit contributions | |
Crit | +1.450 |
Nit | +1.357 |
Poverty | +1.340 |
Crit | +1.329 |
Ps | +1.242 |
Neg logit contributions | |
Kislyak | -0.812 |
================================================================= | -0.803 |
Accessory | -0.763 |
zinski | -0.749 |
IMAGES | -0.718 |
Token | Emails |
Token ID | 39617 |
Feature activation | +9.70e-01 |
Pos logit contributions | |
Kislyak | +0.620 |
================================================================= | +0.605 |
IMAGES | +0.566 |
zinski | +0.554 |
advertising | +0.552 |
Neg logit contributions | |
Crit | -1.217 |
Nit | -1.132 |
Poverty | -1.110 |
Crit | -1.104 |
Ps | -1.040 |
Token | privacy |
Token ID | 6782 |
Feature activation | +3.07e-01 |
Pos logit contributions | |
Kislyak | +2.114 |
================================================================= | +2.105 |
IMAGES | +2.017 |
advertising | +1.946 |
zinski | +1.925 |
Neg logit contributions | |
Crit | -2.960 |
Nit | -2.758 |
Poverty | -2.651 |
Crit | -2.619 |
corruption | -2.463 |
Token | policy |
Token ID | 2450 |
Feature activation | +4.16e-01 |
Pos logit contributions | |
Kislyak | +0.792 |
================================================================= | +0.787 |
IMAGES | +0.755 |
advertising | +0.735 |
zinski | +0.723 |
Neg logit contributions | |
Crit | -0.816 |
Nit | -0.750 |
Poverty | -0.717 |
Crit | -0.704 |
Adapt | -0.663 |
Token | Thanks |
Token ID | 6930 |
Feature activation | +4.98e-01 |
Pos logit contributions | |
Kislyak | +1.037 |
================================================================= | +1.027 |
IMAGES | +1.001 |
advertising | +0.969 |
zinski | +0.967 |
Neg logit contributions | |
Crit | -1.139 |
Nit | -1.054 |
Poverty | -1.008 |
Crit | -1.001 |
Ps | -0.926 |
Token | for |
Token ID | 329 |
Feature activation | -1.48e-01 |
Pos logit contributions | |
Kislyak | +1.212 |
================================================================= | +1.205 |
IMAGES | +1.150 |
advertising | +1.124 |
DEA | +1.105 |
Neg logit contributions | |
Crit | -1.392 |
Nit | -1.290 |
Poverty | -1.234 |
Crit | -1.211 |
corruption | -1.141 |
Token | subscribing |
Token ID | 18412 |
Feature activation | -2.91e-01 |
Pos logit contributions | |
Crit | +0.436 |
Nit | +0.405 |
Poverty | +0.390 |
Crit | +0.385 |
corruption | +0.360 |
Neg logit contributions | |
================================================================= | -0.339 |
Kislyak | -0.339 |
IMAGES | -0.327 |
advertising | -0.315 |
zinski | -0.311 |
Token | eling |
Token ID | 10809 |
Feature activation | -2.14e-01 |
Pos logit contributions | |
Kislyak | +0.633 |
================================================================= | +0.629 |
IMAGES | +0.604 |
advertising | +0.577 |
zinski | +0.573 |
Neg logit contributions | |
Crit | -0.938 |
Nit | -0.873 |
Poverty | -0.842 |
Crit | -0.832 |
corruption | -0.785 |
Token | forums |
Token ID | 14216 |
Feature activation | -8.17e-02 |
Pos logit contributions | |
Crit | +0.590 |
Nit | +0.542 |
Crit | +0.527 |
Poverty | +0.519 |
crit | +0.493 |
Neg logit contributions | |
Kislyak | -0.530 |
================================================================= | -0.520 |
IMAGES | -0.501 |
zinski | -0.498 |
advertising | -0.484 |
Token | where |
Token ID | 810 |
Feature activation | -1.94e-01 |
Pos logit contributions | |
Crit | +0.216 |
Nit | +0.198 |
Crit | +0.190 |
Poverty | +0.189 |
crit | +0.176 |
Neg logit contributions | |
Kislyak | -0.211 |
================================================================= | -0.210 |
IMAGES | -0.204 |
zinski | -0.198 |
advertising | -0.196 |
Token | users |
Token ID | 2985 |
Feature activation | -2.27e-01 |
Pos logit contributions | |
Crit | +0.553 |
Nit | +0.506 |
Crit | +0.491 |
Poverty | +0.486 |
crit | +0.458 |
Neg logit contributions | |
Kislyak | -0.457 |
================================================================= | -0.457 |
IMAGES | -0.433 |
zinski | -0.430 |
advertising | -0.424 |
Token | could |
Token ID | 714 |
Feature activation | +2.21e-01 |
Pos logit contributions | |
Crit | +0.486 |
Nit | +0.434 |
Crit | +0.412 |
Poverty | +0.411 |
crit | +0.372 |
Neg logit contributions | |
Kislyak | -0.699 |
================================================================= | -0.694 |
IMAGES | -0.677 |
advertising | -0.663 |
Accessory | -0.659 |
Token | literally |
Token ID | 7360 |
Feature activation | +9.62e-01 |
Pos logit contributions | |
Kislyak | +0.623 |
================================================================= | +0.619 |
IMAGES | +0.599 |
advertising | +0.588 |
zinski | +0.580 |
Neg logit contributions | |
Crit | -0.533 |
Nit | -0.485 |
Poverty | -0.460 |
Crit | -0.459 |
corruption | -0.417 |
Token | let |
Token ID | 1309 |
Feature activation | -4.47e-02 |
Pos logit contributions | |
================================================================= | +2.671 |
Kislyak | +2.662 |
IMAGES | +2.592 |
advertising | +2.478 |
zinski | +2.463 |
Neg logit contributions | |
Crit | -2.342 |
Nit | -2.156 |
Poverty | -2.037 |
Crit | -1.983 |
Adapt | -1.862 |
Token | it |
Token ID | 340 |
Feature activation | -7.93e-02 |
Pos logit contributions | |
Crit | +0.120 |
Nit | +0.110 |
Poverty | +0.106 |
Crit | +0.106 |
crit | +0.098 |
Neg logit contributions | |
Kislyak | -0.112 |
================================================================= | -0.112 |
IMAGES | -0.107 |
zinski | -0.106 |
advertising | -0.105 |
Token | all |
Token ID | 477 |
Feature activation | +7.30e-02 |
Pos logit contributions | |
Crit | +0.173 |
Nit | +0.156 |
Poverty | +0.148 |
Crit | +0.146 |
corruption | +0.132 |
Neg logit contributions | |
Kislyak | -0.241 |
================================================================= | -0.240 |
IMAGES | -0.231 |
zinski | -0.227 |
advertising | -0.227 |
Token | hang |
Token ID | 8181 |
Feature activation | -6.87e-02 |
Pos logit contributions | |
Kislyak | +0.211 |
================================================================= | +0.210 |
IMAGES | +0.204 |
advertising | +0.198 |
zinski | +0.197 |
Neg logit contributions | |
Crit | -0.171 |
Nit | -0.156 |
Poverty | -0.148 |
Crit | -0.145 |
corruption | -0.133 |
Token | out |
Token ID | 503 |
Feature activation | -1.49e-01 |
Pos logit contributions | |
Crit | +0.157 |
Nit | +0.142 |
Poverty | +0.136 |
Crit | +0.134 |
corruption | +0.122 |
Neg logit contributions | |
Kislyak | -0.202 |
================================================================= | -0.200 |
IMAGES | -0.193 |
zinski | -0.191 |
advertising | -0.190 |
Token | us |
Token ID | 385 |
Feature activation | +3.22e-01 |
Pos logit contributions | |
Kislyak | +0.899 |
================================================================= | +0.890 |
IMAGES | +0.861 |
advertising | +0.825 |
zinski | +0.812 |
Neg logit contributions | |
Crit | -1.421 |
Nit | -1.325 |
Crit | -1.268 |
Poverty | -1.265 |
corruption | -1.190 |
Token | nerve |
Token ID | 16384 |
Feature activation | +2.03e-01 |
Pos logit contributions | |
Kislyak | +0.907 |
================================================================= | +0.907 |
IMAGES | +0.896 |
advertising | +0.863 |
zinski | +0.855 |
Neg logit contributions | |
Crit | -0.778 |
Nit | -0.714 |
Crit | -0.668 |
Poverty | -0.668 |
Adapt | -0.609 |
Token | stimulation |
Token ID | 20087 |
Feature activation | +2.94e-01 |
Pos logit contributions | |
Kislyak | +0.462 |
================================================================= | +0.460 |
IMAGES | +0.450 |
advertising | +0.428 |
Accessory | +0.424 |
Neg logit contributions | |
Crit | -0.600 |
Nit | -0.561 |
Poverty | -0.534 |
Crit | -0.533 |
Adapt | -0.497 |
Token | ( |
Token ID | 357 |
Feature activation | +1.16e-01 |
Pos logit contributions | |
Kislyak | +0.733 |
================================================================= | +0.729 |
IMAGES | +0.704 |
advertising | +0.682 |
zinski | +0.674 |
Neg logit contributions | |
Crit | -0.807 |
Nit | -0.745 |
Poverty | -0.712 |
Crit | -0.705 |
corruption | -0.655 |
Token | V |
Token ID | 53 |
Feature activation | +3.31e-01 |
Pos logit contributions | |
Kislyak | +0.309 |
================================================================= | +0.308 |
IMAGES | +0.297 |
advertising | +0.286 |
zinski | +0.285 |
Neg logit contributions | |
Crit | -0.302 |
Nit | -0.277 |
Poverty | -0.261 |
Crit | -0.258 |
corruption | -0.242 |
Token | NS |
Token ID | 8035 |
Feature activation | +9.61e-01 |
Pos logit contributions | |
Kislyak | +1.138 |
================================================================= | +1.135 |
IMAGES | +1.104 |
advertising | +1.081 |
zinski | +1.073 |
Neg logit contributions | |
Crit | -0.595 |
Nit | -0.526 |
Poverty | -0.489 |
Crit | -0.478 |
corruption | -0.425 |
Token | ), |
Token ID | 828 |
Feature activation | +1.53e-01 |
Pos logit contributions | |
Kislyak | +1.402 |
================================================================= | +1.396 |
IMAGES | +1.301 |
advertising | +1.238 |
zinski | +1.219 |
Neg logit contributions | |
Crit | -3.601 |
Nit | -3.408 |
Poverty | -3.303 |
Crit | -3.262 |
corruption | -3.115 |
Token | an |
Token ID | 281 |
Feature activation | -1.26e-02 |
Pos logit contributions | |
Kislyak | +0.388 |
================================================================= | +0.386 |
IMAGES | +0.373 |
advertising | +0.361 |
zinski | +0.358 |
Neg logit contributions | |
Crit | -0.410 |
Nit | -0.379 |
Poverty | -0.362 |
Crit | -0.358 |
Ps | -0.332 |
Token | FDA |
Token ID | 14672 |
Feature activation | +2.39e-01 |
Pos logit contributions | |
Crit | +0.032 |
Nit | +0.029 |
Crit | +0.028 |
Poverty | +0.028 |
Adapt | +0.026 |
Neg logit contributions | |
Kislyak | -0.034 |
================================================================= | -0.034 |
IMAGES | -0.033 |
zinski | -0.032 |
advertising | -0.032 |
Token | - |
Token ID | 12 |
Feature activation | +3.63e-01 |
Pos logit contributions | |
Kislyak | +0.676 |
================================================================= | +0.672 |
IMAGES | +0.652 |
advertising | +0.633 |
zinski | +0.628 |
Neg logit contributions | |
Crit | -0.579 |
Nit | -0.528 |
Poverty | -0.501 |
Crit | -0.496 |
Adapt | -0.454 |
Token | approved |
Token ID | 29137 |
Feature activation | +4.05e-01 |
Pos logit contributions | |
Kislyak | +0.913 |
================================================================= | +0.908 |
IMAGES | +0.877 |
advertising | +0.844 |
zinski | +0.839 |
Neg logit contributions | |
Crit | -0.994 |
Nit | -0.917 |
Poverty | -0.873 |
Crit | -0.865 |
Adapt | -0.806 |
Token | of |
Token ID | 286 |
Feature activation | +1.66e-01 |
Pos logit contributions | |
Crit | +0.647 |
Nit | +0.596 |
Poverty | +0.572 |
Crit | +0.566 |
corruption | +0.525 |
Neg logit contributions | |
Kislyak | -0.609 |
================================================================= | -0.606 |
IMAGES | -0.585 |
advertising | -0.569 |
zinski | -0.559 |
Token | England |
Token ID | 4492 |
Feature activation | +3.13e-01 |
Pos logit contributions | |
Kislyak | +0.471 |
================================================================= | +0.470 |
IMAGES | +0.454 |
advertising | +0.443 |
zinski | +0.438 |
Neg logit contributions | |
Crit | -0.394 |
Nit | -0.360 |
Poverty | -0.343 |
Crit | -0.337 |
corruption | -0.310 |
Token | has |
Token ID | 468 |
Feature activation | +1.57e-01 |
Pos logit contributions | |
Kislyak | +0.859 |
================================================================= | +0.858 |
IMAGES | +0.828 |
advertising | +0.804 |
zinski | +0.798 |
Neg logit contributions | |
Crit | -0.776 |
Nit | -0.712 |
Poverty | -0.675 |
Crit | -0.665 |
Adapt | -0.618 |
Token | held |
Token ID | 2714 |
Feature activation | -7.88e-02 |
Pos logit contributions | |
Kislyak | +0.432 |
================================================================= | +0.431 |
IMAGES | +0.417 |
advertising | +0.406 |
zinski | +0.402 |
Neg logit contributions | |
Crit | -0.389 |
Nit | -0.355 |
Poverty | -0.338 |
Crit | -0.334 |
corruption | -0.308 |
Token | UK |
Token ID | 3482 |
Feature activation | -3.40e-02 |
Pos logit contributions | |
Crit | +0.206 |
Nit | +0.188 |
Poverty | +0.181 |
Crit | +0.180 |
corruption | +0.165 |
Neg logit contributions | |
Kislyak | -0.205 |
================================================================= | -0.205 |
IMAGES | -0.199 |
advertising | -0.193 |
zinski | -0.192 |
Token | interest |
Token ID | 1393 |
Feature activation | +9.60e-01 |
Pos logit contributions | |
Crit | +0.091 |
Nit | +0.084 |
Poverty | +0.080 |
Crit | +0.079 |
corruption | +0.073 |
Neg logit contributions | |
Kislyak | -0.087 |
================================================================= | -0.087 |
IMAGES | -0.084 |
advertising | -0.081 |
zinski | -0.080 |
Token | rates |
Token ID | 3965 |
Feature activation | +1.46e-01 |
Pos logit contributions | |
Kislyak | +1.672 |
================================================================= | +1.669 |
IMAGES | +1.577 |
advertising | +1.506 |
zinski | +1.487 |
Neg logit contributions | |
Crit | -3.271 |
Nit | -3.085 |
Poverty | -2.992 |
Crit | -2.927 |
Adapt | -2.820 |
Token | at |
Token ID | 379 |
Feature activation | +1.38e-02 |
Pos logit contributions | |
Kislyak | +0.385 |
================================================================= | +0.384 |
IMAGES | +0.372 |
advertising | +0.361 |
zinski | +0.357 |
Neg logit contributions | |
Crit | -0.380 |
Nit | -0.349 |
Poverty | -0.333 |
Crit | -0.329 |
corruption | -0.304 |
Token | a |
Token ID | 257 |
Feature activation | -3.32e-03 |
Pos logit contributions | |
Kislyak | +0.034 |
================================================================= | +0.034 |
IMAGES | +0.032 |
advertising | +0.031 |
zinski | +0.031 |
Neg logit contributions | |
Crit | -0.039 |
Nit | -0.036 |
Poverty | -0.034 |
Crit | -0.034 |
corruption | -0.031 |
Token | record |
Token ID | 1700 |
Feature activation | -2.46e-01 |
Pos logit contributions | |
Crit | +0.009 |
Nit | +0.008 |
Poverty | +0.008 |
Crit | +0.008 |
Adapt | +0.007 |
Neg logit contributions | |
Kislyak | -0.009 |
================================================================= | -0.009 |
IMAGES | -0.008 |
advertising | -0.008 |
Accessory | -0.008 |
Token | low |
Token ID | 1877 |
Feature activation | -1.27e-01 |
Pos logit contributions | |
Crit | +0.615 |
Nit | +0.566 |
Poverty | +0.540 |
Crit | +0.537 |
crit | +0.487 |
Neg logit contributions | |
================================================================= | -0.669 |
Kislyak | -0.668 |
IMAGES | -0.653 |
advertising | -0.632 |
Accessory | -0.623 |
Token | of |
Token ID | 286 |
Feature activation | +7.52e-02 |
Pos logit contributions | |
Kislyak | +0.950 |
================================================================= | +0.945 |
IMAGES | +0.914 |
advertising | +0.885 |
zinski | +0.877 |
Neg logit contributions | |
Crit | -0.998 |
Nit | -0.919 |
Poverty | -0.879 |
Crit | -0.870 |
corruption | -0.806 |
Token | medical |
Token ID | 3315 |
Feature activation | +6.05e-02 |
Pos logit contributions | |
Kislyak | +0.138 |
================================================================= | +0.137 |
IMAGES | +0.131 |
advertising | +0.125 |
Accessory | +0.123 |
Neg logit contributions | |
Crit | -0.256 |
Nit | -0.240 |
Poverty | -0.233 |
Crit | -0.231 |
corruption | -0.218 |
Token | thought |
Token ID | 1807 |
Feature activation | +1.76e-01 |
Pos logit contributions | |
Kislyak | +0.153 |
================================================================= | +0.152 |
IMAGES | +0.147 |
advertising | +0.142 |
zinski | +0.141 |
Neg logit contributions | |
Crit | -0.164 |
Nit | -0.151 |
Poverty | -0.144 |
Crit | -0.143 |
corruption | -0.133 |
Token | focused |
Token ID | 5670 |
Feature activation | +1.87e-01 |
Pos logit contributions | |
Kislyak | +0.461 |
================================================================= | +0.460 |
IMAGES | +0.445 |
advertising | +0.432 |
zinski | +0.427 |
Neg logit contributions | |
Crit | -0.459 |
Nit | -0.421 |
Poverty | -0.403 |
Crit | -0.399 |
Adapt | -0.367 |
Token | on |
Token ID | 319 |
Feature activation | +1.21e-01 |
Pos logit contributions | |
Kislyak | +0.496 |
================================================================= | +0.494 |
IMAGES | +0.478 |
advertising | +0.464 |
zinski | +0.459 |
Neg logit contributions | |
Crit | -0.485 |
Nit | -0.444 |
Poverty | -0.425 |
Crit | -0.420 |
Adapt | -0.388 |
Token | literally |
Token ID | 7360 |
Feature activation | +9.56e-01 |
Pos logit contributions | |
Kislyak | +0.316 |
================================================================= | +0.314 |
IMAGES | +0.304 |
advertising | +0.295 |
zinski | +0.293 |
Neg logit contributions | |
Crit | -0.318 |
Nit | -0.292 |
Poverty | -0.280 |
Crit | -0.276 |
corruption | -0.256 |
Token | removing |
Token ID | 10829 |
Feature activation | +2.87e-01 |
Pos logit contributions | |
================================================================= | +2.572 |
Kislyak | +2.567 |
IMAGES | +2.480 |
advertising | +2.409 |
Accessory | +2.376 |
Neg logit contributions | |
Crit | -2.411 |
Nit | -2.218 |
Poverty | -2.106 |
Crit | -2.072 |
Adapt | -1.926 |
Token | a |
Token ID | 257 |
Feature activation | +1.88e-02 |
Pos logit contributions | |
Kislyak | +0.759 |
================================================================= | +0.757 |
IMAGES | +0.730 |
advertising | +0.710 |
zinski | +0.703 |
Neg logit contributions | |
Crit | -0.741 |
Nit | -0.681 |
Poverty | -0.650 |
Crit | -0.640 |
corruption | -0.594 |
Token | disease |
Token ID | 4369 |
Feature activation | -1.55e-01 |
Pos logit contributions | |
Kislyak | +0.054 |
================================================================= | +0.054 |
IMAGES | +0.052 |
advertising | +0.051 |
zinski | +0.050 |
Neg logit contributions | |
Crit | -0.045 |
Nit | -0.041 |
Poverty | -0.039 |
Crit | -0.038 |
corruption | -0.035 |
Token | from |
Token ID | 422 |
Feature activation | +2.11e-01 |
Pos logit contributions | |
Crit | +0.407 |
Nit | +0.374 |
Poverty | +0.357 |
Crit | +0.353 |
corruption | +0.327 |
Neg logit contributions | |
Kislyak | -0.403 |
================================================================= | -0.400 |
IMAGES | -0.386 |
advertising | -0.375 |
zinski | -0.371 |
Token | the |
Token ID | 262 |
Feature activation | +1.44e-01 |
Pos logit contributions | |
Kislyak | +0.557 |
================================================================= | +0.554 |
IMAGES | +0.535 |
advertising | +0.520 |
zinski | +0.516 |
Neg logit contributions | |
Crit | -0.548 |
Nit | -0.503 |
Poverty | -0.481 |
Crit | -0.474 |
corruption | -0.439 |
Token | made |
Token ID | 9727 |
Feature activation | +1.21e-01 |
Pos logit contributions | |
Kislyak | +0.376 |
================================================================= | +0.372 |
IMAGES | +0.362 |
advertising | +0.352 |
Accessory | +0.349 |
Neg logit contributions | |
Crit | -0.319 |
Nit | -0.289 |
Poverty | -0.276 |
Crit | -0.273 |
corruption | -0.251 |
Token | ath |
Token ID | 776 |
Feature activation | -1.28e-01 |
Pos logit contributions | |
Kislyak | +0.296 |
================================================================= | +0.290 |
IMAGES | +0.281 |
advertising | +0.272 |
zinski | +0.270 |
Neg logit contributions | |
Crit | -0.338 |
Nit | -0.311 |
Poverty | -0.300 |
Crit | -0.296 |
corruption | -0.277 |
Token | ) |
Token ID | 8 |
Feature activation | -8.96e-02 |
Pos logit contributions | |
Crit | +0.344 |
Nit | +0.316 |
Poverty | +0.302 |
Crit | +0.300 |
corruption | +0.279 |
Neg logit contributions | |
Kislyak | -0.332 |
================================================================= | -0.323 |
IMAGES | -0.314 |
advertising | -0.306 |
Accessory | -0.304 |
Token |
|
Token ID | 198 |
Feature activation | -2.85e-05 |
Pos logit contributions | |
Crit | +0.221 |
Nit | +0.201 |
Poverty | +0.192 |
Crit | +0.190 |
corruption | +0.174 |
Neg logit contributions | |
Kislyak | -0.250 |
================================================================= | -0.245 |
IMAGES | -0.240 |
advertising | -0.234 |
zinski | -0.231 |
Token |
|
Token ID | 198 |
Feature activation | +6.04e-02 |
Pos logit contributions | |
Crit | +0.000 |
Nit | +0.000 |
Poverty | +0.000 |
Crit | +0.000 |
corruption | +0.000 |
Neg logit contributions | |
Kislyak | -0.000 |
================================================================= | -0.000 |
IMAGES | -0.000 |
advertising | -0.000 |
zinski | -0.000 |
Token | Craft |
Token ID | 14467 |
Feature activation | +9.51e-01 |
Pos logit contributions | |
Kislyak | +0.184 |
================================================================= | +0.182 |
IMAGES | +0.178 |
advertising | +0.171 |
zinski | +0.171 |
Neg logit contributions | |
Crit | -0.135 |
Nit | -0.123 |
Poverty | -0.113 |
Crit | -0.112 |
Adapt | -0.105 |
Token | thousands |
Token ID | 4138 |
Feature activation | +3.36e-01 |
Pos logit contributions | |
Kislyak | +2.419 |
================================================================= | +2.414 |
IMAGES | +2.325 |
advertising | +2.257 |
zinski | +2.234 |
Neg logit contributions | |
Crit | -2.540 |
Nit | -2.344 |
Poverty | -2.236 |
Crit | -2.202 |
corruption | -2.056 |
Token | of |
Token ID | 286 |
Feature activation | +5.17e-03 |
Pos logit contributions | |
================================================================= | +0.795 |
Kislyak | +0.793 |
IMAGES | +0.763 |
advertising | +0.736 |
zinski | +0.730 |
Neg logit contributions | |
Crit | -0.959 |
Nit | -0.891 |
Poverty | -0.856 |
Crit | -0.836 |
Adapt | -0.789 |
Token | objects |
Token ID | 5563 |
Feature activation | +7.03e-02 |
Pos logit contributions | |
Kislyak | +0.015 |
================================================================= | +0.014 |
IMAGES | +0.014 |
advertising | +0.014 |
zinski | +0.014 |
Neg logit contributions | |
Crit | -0.013 |
Nit | -0.011 |
Poverty | -0.011 |
Crit | -0.011 |
Ps | -0.010 |
Token | - |
Token ID | 532 |
Feature activation | -4.90e-02 |
Pos logit contributions | |
Kislyak | +0.184 |
================================================================= | +0.180 |
IMAGES | +0.175 |
advertising | +0.171 |
zinski | +0.170 |
Neg logit contributions | |
Crit | -0.185 |
Nit | -0.171 |
Poverty | -0.163 |
Crit | -0.162 |
Ps | -0.149 |
Token | building |
Token ID | 2615 |
Feature activation | +7.50e-02 |
Pos logit contributions | |
Crit | +0.126 |
Nit | +0.116 |
Poverty | +0.112 |
Crit | +0.111 |
Adapt | +0.101 |
Neg logit contributions | |
Kislyak | -0.131 |
================================================================= | -0.129 |
IMAGES | -0.124 |
advertising | -0.122 |
zinski | -0.121 |
Token | of |
Token ID | 286 |
Feature activation | +1.09e-01 |
Pos logit contributions | |
Kislyak | +0.821 |
================================================================= | +0.816 |
IMAGES | +0.790 |
advertising | +0.766 |
Accessory | +0.759 |
Neg logit contributions | |
Crit | -0.856 |
Nit | -0.788 |
Poverty | -0.756 |
Crit | -0.748 |
corruption | -0.692 |
Token | the |
Token ID | 262 |
Feature activation | +1.95e-01 |
Pos logit contributions | |
Kislyak | +0.289 |
================================================================= | +0.289 |
IMAGES | +0.280 |
advertising | +0.274 |
zinski | +0.272 |
Neg logit contributions | |
Crit | -0.276 |
Nit | -0.253 |
Poverty | -0.244 |
Crit | -0.242 |
corruption | -0.222 |
Token | Greeks |
Token ID | 25059 |
Feature activation | +1.52e-01 |
Pos logit contributions | |
Kislyak | +0.501 |
================================================================= | +0.498 |
IMAGES | +0.484 |
advertising | +0.474 |
zinski | +0.466 |
Neg logit contributions | |
Crit | -0.517 |
Nit | -0.475 |
Poverty | -0.457 |
Crit | -0.454 |
corruption | -0.418 |
Token | and |
Token ID | 290 |
Feature activation | +7.55e-02 |
Pos logit contributions | |
Kislyak | +0.381 |
================================================================= | +0.379 |
IMAGES | +0.368 |
advertising | +0.356 |
zinski | +0.353 |
Neg logit contributions | |
Crit | -0.413 |
Nit | -0.381 |
Poverty | -0.365 |
Crit | -0.361 |
corruption | -0.335 |
Token | the |
Token ID | 262 |
Feature activation | +5.14e-02 |
Pos logit contributions | |
Kislyak | +0.199 |
================================================================= | +0.199 |
IMAGES | +0.194 |
advertising | +0.189 |
Accessory | +0.187 |
Neg logit contributions | |
Crit | -0.193 |
Nit | -0.176 |
Poverty | -0.170 |
Crit | -0.169 |
Ps | -0.155 |
Token | Pers |
Token ID | 9467 |
Feature activation | +9.48e-01 |
Pos logit contributions | |
Kislyak | +0.136 |
================================================================= | +0.135 |
IMAGES | +0.132 |
advertising | +0.129 |
Accessory | +0.127 |
Neg logit contributions | |
Crit | -0.132 |
Nit | -0.121 |
Poverty | -0.117 |
Crit | -0.116 |
corruption | -0.106 |
Token | ians |
Token ID | 1547 |
Feature activation | +2.10e-01 |
Pos logit contributions | |
Kislyak | +2.248 |
================================================================= | +2.237 |
IMAGES | +2.144 |
advertising | +2.087 |
zinski | +2.064 |
Neg logit contributions | |
Crit | -2.707 |
Nit | -2.509 |
Poverty | -2.409 |
Crit | -2.371 |
Lumpur | -2.224 |
Token | , |
Token ID | 11 |
Feature activation | +8.81e-02 |
Pos logit contributions | |
Kislyak | +0.527 |
================================================================= | +0.524 |
IMAGES | +0.509 |
advertising | +0.493 |
Accessory | +0.488 |
Neg logit contributions | |
Crit | -0.568 |
Nit | -0.523 |
Poverty | -0.504 |
Crit | -0.499 |
corruption | -0.462 |
Token | in |
Token ID | 287 |
Feature activation | +1.66e-02 |
Pos logit contributions | |
Kislyak | +0.229 |
================================================================= | +0.228 |
IMAGES | +0.222 |
advertising | +0.216 |
zinski | +0.214 |
Neg logit contributions | |
Crit | -0.229 |
Nit | -0.210 |
Poverty | -0.204 |
Crit | -0.202 |
corruption | -0.185 |
Token | the |
Token ID | 262 |
Feature activation | +1.16e-01 |
Pos logit contributions | |
Kislyak | +0.044 |
================================================================= | +0.044 |
IMAGES | +0.042 |
advertising | +0.041 |
zinski | +0.041 |
Neg logit contributions | |
Crit | -0.043 |
Nit | -0.039 |
Poverty | -0.038 |
Crit | -0.037 |
corruption | -0.034 |
Token | second |
Token ID | 1218 |
Feature activation | +1.48e-01 |
Pos logit contributions | |
Kislyak | +0.285 |
================================================================= | +0.284 |
IMAGES | +0.276 |
advertising | +0.271 |
zinski | +0.265 |
Neg logit contributions | |
Crit | -0.316 |
Nit | -0.293 |
Poverty | -0.283 |
Crit | -0.281 |
corruption | -0.259 |
Token | deduct |
Token ID | 18777 |
Feature activation | +3.04e-01 |
Pos logit contributions | |
================================================================= | +0.595 |
Kislyak | +0.594 |
IMAGES | +0.576 |
advertising | +0.558 |
zinski | +0.554 |
Neg logit contributions | |
Crit | -0.505 |
Nit | -0.460 |
Poverty | -0.440 |
Crit | -0.434 |
Ps | -0.396 |
Token | ibles |
Token ID | 18764 |
Feature activation | +1.32e-01 |
Pos logit contributions | |
Kislyak | +0.768 |
================================================================= | +0.764 |
IMAGES | +0.739 |
advertising | +0.711 |
Accessory | +0.706 |
Neg logit contributions | |
Crit | -0.823 |
Nit | -0.758 |
Poverty | -0.729 |
Crit | -0.717 |
corruption | -0.666 |
Token | and |
Token ID | 290 |
Feature activation | +1.37e-01 |
Pos logit contributions | |
Kislyak | +0.336 |
================================================================= | +0.334 |
IMAGES | +0.323 |
advertising | +0.312 |
zinski | +0.310 |
Neg logit contributions | |
Crit | -0.358 |
Nit | -0.329 |
Poverty | -0.316 |
Crit | -0.312 |
Ps | -0.289 |
Token | out |
Token ID | 503 |
Feature activation | -1.23e-01 |
Pos logit contributions | |
Kislyak | +0.365 |
================================================================= | +0.365 |
IMAGES | +0.352 |
advertising | +0.340 |
zinski | +0.339 |
Neg logit contributions | |
Crit | -0.355 |
Nit | -0.324 |
Poverty | -0.311 |
Crit | -0.307 |
Ps | -0.284 |
Token | - |
Token ID | 12 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
Crit | +0.319 |
Nit | +0.290 |
Poverty | +0.277 |
Crit | +0.275 |
Ps | +0.255 |
Neg logit contributions | |
Kislyak | -0.325 |
================================================================= | -0.324 |
IMAGES | -0.315 |
zinski | -0.301 |
advertising | -0.300 |
Token | of |
Token ID | 1659 |
Feature activation | +9.44e-01 |
Pos logit contributions | |
Kislyak | +0.311 |
================================================================= | +0.310 |
IMAGES | +0.301 |
zinski | +0.290 |
advertising | +0.289 |
Neg logit contributions | |
Crit | -0.236 |
Nit | -0.213 |
Poverty | -0.200 |
Crit | -0.198 |
Ps | -0.182 |
Token | - |
Token ID | 12 |
Feature activation | +2.31e-01 |
Pos logit contributions | |
Kislyak | +2.262 |
================================================================= | +2.261 |
IMAGES | +2.176 |
advertising | +2.121 |
Accessory | +2.084 |
Neg logit contributions | |
Crit | -2.631 |
Nit | -2.445 |
Poverty | -2.308 |
Crit | -2.294 |
Adapt | -2.168 |
Token | pocket |
Token ID | 31991 |
Feature activation | +2.59e-01 |
Pos logit contributions | |
Kislyak | +0.348 |
================================================================= | +0.346 |
IMAGES | +0.325 |
advertising | +0.302 |
zinski | +0.301 |
Neg logit contributions | |
Crit | -0.863 |
Nit | -0.813 |
Poverty | -0.789 |
Crit | -0.780 |
corruption | -0.744 |
Token | costs |
Token ID | 3484 |
Feature activation | +6.41e-02 |
Pos logit contributions | |
Kislyak | +0.742 |
================================================================= | +0.741 |
IMAGES | +0.723 |
advertising | +0.698 |
zinski | +0.696 |
Neg logit contributions | |
Crit | -0.613 |
Nit | -0.558 |
Poverty | -0.536 |
Crit | -0.525 |
Ps | -0.481 |
Token | , |
Token ID | 11 |
Feature activation | -6.25e-02 |
Pos logit contributions | |
Kislyak | +0.158 |
================================================================= | +0.158 |
IMAGES | +0.152 |
advertising | +0.147 |
zinski | +0.146 |
Neg logit contributions | |
Crit | -0.177 |
Nit | -0.163 |
Poverty | -0.157 |
Crit | -0.155 |
Ps | -0.144 |
Token | but |
Token ID | 475 |
Feature activation | -9.45e-02 |
Pos logit contributions | |
Crit | +0.162 |
Nit | +0.149 |
Poverty | +0.143 |
Crit | +0.141 |
Ps | +0.130 |
Neg logit contributions | |
Kislyak | -0.165 |
================================================================= | -0.164 |
IMAGES | -0.159 |
advertising | -0.154 |
Accessory | -0.152 |
Token | er |
Token ID | 263 |
Feature activation | -4.54e-01 |
Pos logit contributions | |
Crit | +1.963 |
Nit | +1.784 |
Poverty | +1.736 |
Crit | +1.725 |
corruption | +1.615 |
Neg logit contributions | |
Kislyak | -1.463 |
================================================================= | -1.459 |
IMAGES | -1.415 |
Accessory | -1.335 |
DEA | -1.330 |
Token | magnet |
Token ID | 19972 |
Feature activation | +2.04e-01 |
Pos logit contributions | |
Crit | +1.149 |
Nit | +1.046 |
Poverty | +1.007 |
Crit | +0.997 |
Adapt | +0.922 |
Neg logit contributions | |
Kislyak | -1.224 |
================================================================= | -1.210 |
IMAGES | -1.177 |
advertising | -1.142 |
zinski | -1.134 |
Token | ? |
Token ID | 30 |
Feature activation | -1.47e-01 |
Pos logit contributions | |
Kislyak | +0.382 |
================================================================= | +0.381 |
IMAGES | +0.360 |
advertising | +0.346 |
zinski | +0.342 |
Neg logit contributions | |
Crit | -0.683 |
Nit | -0.642 |
Poverty | -0.619 |
Crit | -0.611 |
Adapt | -0.580 |
Token |
|
Token ID | 198 |
Feature activation | -2.92e-01 |
Pos logit contributions | |
Crit | +0.370 |
Nit | +0.340 |
Poverty | +0.323 |
Crit | +0.318 |
corruption | +0.295 |
Neg logit contributions | |
Kislyak | -0.398 |
================================================================= | -0.397 |
IMAGES | -0.383 |
advertising | -0.372 |
zinski | -0.369 |
Token |
|
Token ID | 198 |
Feature activation | -2.06e-01 |
Pos logit contributions | |
Crit | +0.693 |
Nit | +0.630 |
Poverty | +0.597 |
Crit | +0.590 |
Adapt | +0.541 |
Neg logit contributions | |
Kislyak | -0.833 |
================================================================= | -0.829 |
IMAGES | -0.805 |
advertising | -0.781 |
zinski | -0.775 |
Token | Reviewer |
Token ID | 35407 |
Feature activation | -1.66e+00 |
Pos logit contributions | |
Crit | +0.376 |
Nit | +0.332 |
Poverty | +0.300 |
Crit | +0.297 |
Ps | +0.268 |
Neg logit contributions | |
Kislyak | -0.709 |
================================================================= | -0.705 |
IMAGES | -0.692 |
advertising | -0.666 |
zinski | -0.666 |
Token | : |
Token ID | 25 |
Feature activation | -2.04e-01 |
Pos logit contributions | |
Crit | +4.747 |
Nit | +4.325 |
Crit | +4.183 |
Poverty | +4.123 |
Ps | +3.862 |
Neg logit contributions | |
conclud | -3.848 |
Kislyak | -3.804 |
ategory | -3.765 |
================================================================= | -3.725 |
IMAGES | -3.715 |
Token | hey |
Token ID | 17207 |
Feature activation | -9.11e-02 |
Pos logit contributions | |
Crit | +0.510 |
Nit | +0.469 |
Poverty | +0.448 |
Crit | +0.442 |
Adapt | +0.405 |
Neg logit contributions | |
Kislyak | -0.554 |
================================================================= | -0.550 |
IMAGES | -0.533 |
advertising | -0.521 |
Accessory | -0.515 |
Token | _ |
Token ID | 62 |
Feature activation | -2.93e-01 |
Pos logit contributions | |
Crit | +0.236 |
Nit | +0.217 |
Poverty | +0.207 |
Crit | +0.204 |
corruption | +0.190 |
Neg logit contributions | |
Kislyak | -0.241 |
================================================================= | -0.239 |
IMAGES | -0.232 |
advertising | -0.224 |
zinski | -0.222 |
Token | ler |
Token ID | 1754 |
Feature activation | +1.09e-01 |
Pos logit contributions | |
Crit | +0.767 |
Nit | +0.706 |
Poverty | +0.669 |
Crit | +0.660 |
corruption | +0.620 |
Neg logit contributions | |
Kislyak | -0.766 |
================================================================= | -0.761 |
IMAGES | -0.738 |
advertising | -0.709 |
Accessory | -0.705 |
Token | oy |
Token ID | 726 |
Feature activation | +3.17e-01 |
Pos logit contributions | |
Kislyak | +0.279 |
================================================================= | +0.277 |
IMAGES | +0.269 |
advertising | +0.260 |
Accessory | +0.258 |
Neg logit contributions | |
Crit | -0.293 |
Nit | -0.270 |
Poverty | -0.258 |
Crit | -0.255 |
corruption | -0.237 |
Token | was |
Token ID | 373 |
Feature activation | +1.49e-01 |
Pos logit contributions | |
Crit | +0.139 |
Nit | +0.128 |
Poverty | +0.123 |
Crit | +0.121 |
Adapt | +0.112 |
Neg logit contributions | |
Kislyak | -0.137 |
================================================================= | -0.135 |
IMAGES | -0.131 |
advertising | -0.126 |
Accessory | -0.126 |
Token | about |
Token ID | 546 |
Feature activation | +1.53e-01 |
Pos logit contributions | |
Kislyak | +0.402 |
================================================================= | +0.398 |
IMAGES | +0.389 |
advertising | +0.376 |
zinski | +0.374 |
Neg logit contributions | |
Crit | -0.379 |
Nit | -0.347 |
Poverty | -0.332 |
Crit | -0.329 |
Adapt | -0.301 |
Token | homosexuality |
Token ID | 19822 |
Feature activation | +2.81e-01 |
Pos logit contributions | |
Kislyak | +0.391 |
================================================================= | +0.389 |
IMAGES | +0.379 |
advertising | +0.367 |
zinski | +0.365 |
Neg logit contributions | |
Crit | -0.405 |
Nit | -0.371 |
Poverty | -0.361 |
Crit | -0.357 |
corruption | -0.327 |
Token |
|
Token ID | 198 |
Feature activation | +2.73e-02 |
Pos logit contributions | |
Kislyak | +0.701 |
================================================================= | +0.694 |
IMAGES | +0.673 |
advertising | +0.650 |
zinski | +0.646 |
Neg logit contributions | |
Crit | -0.771 |
Nit | -0.710 |
Poverty | -0.683 |
Crit | -0.673 |
corruption | -0.625 |
Token |
|
Token ID | 198 |
Feature activation | +2.03e-02 |
Pos logit contributions | |
Kislyak | +0.075 |
================================================================= | +0.075 |
IMAGES | +0.073 |
advertising | +0.071 |
zinski | +0.070 |
Neg logit contributions | |
Crit | -0.067 |
Nit | -0.061 |
Poverty | -0.059 |
Crit | -0.058 |
Lumpur | -0.053 |
Token | Reviewer |
Token ID | 35407 |
Feature activation | -1.64e+00 |
Pos logit contributions | |
Kislyak | +0.059 |
================================================================= | +0.058 |
IMAGES | +0.057 |
zinski | +0.055 |
advertising | +0.054 |
Neg logit contributions | |
Crit | -0.048 |
Nit | -0.044 |
Poverty | -0.041 |
Crit | -0.040 |
Adapt | -0.038 |
Token | : |
Token ID | 25 |
Feature activation | +8.07e-02 |
Pos logit contributions | |
Crit | +4.657 |
Crit | +4.246 |
Nit | +4.207 |
Poverty | +4.182 |
crit | +3.875 |
Neg logit contributions | |
Kislyak | -3.695 |
================================================================= | -3.609 |
ategory | -3.529 |
IMAGES | -3.523 |
advertising | -3.520 |
Token | k |
Token ID | 479 |
Feature activation | +1.34e-01 |
Pos logit contributions | |
Kislyak | +0.221 |
================================================================= | +0.220 |
IMAGES | +0.213 |
advertising | +0.207 |
zinski | +0.205 |
Neg logit contributions | |
Crit | -0.201 |
Nit | -0.184 |
Poverty | -0.175 |
Crit | -0.172 |
corruption | -0.159 |
Token | imp |
Token ID | 11011 |
Feature activation | +2.09e-01 |
Pos logit contributions | |
Kislyak | +0.339 |
================================================================= | +0.336 |
IMAGES | +0.326 |
advertising | +0.314 |
Accessory | +0.312 |
Neg logit contributions | |
Crit | -0.361 |
Nit | -0.334 |
Poverty | -0.318 |
Crit | -0.314 |
corruption | -0.293 |
Token | unk |
Token ID | 2954 |
Feature activation | +3.10e-01 |
Pos logit contributions | |
Kislyak | +0.532 |
================================================================= | +0.528 |
IMAGES | +0.510 |
advertising | +0.494 |
Accessory | +0.489 |
Neg logit contributions | |
Crit | -0.561 |
Nit | -0.518 |
Poverty | -0.494 |
Crit | -0.488 |
corruption | -0.454 |
Token | rock |
Token ID | 10823 |
Feature activation | +8.24e-02 |
Pos logit contributions | |
================================================================= | +0.782 |
Kislyak | +0.782 |
IMAGES | +0.751 |
advertising | +0.732 |
zinski | +0.724 |
Neg logit contributions | |
Crit | -0.836 |
Nit | -0.772 |
Poverty | -0.739 |
Crit | -0.727 |
Adapt | -0.679 |
Token | iff |
Token ID | 733 |
Feature activation | -2.35e-01 |
Pos logit contributions | |
Crit | +0.766 |
Nit | +0.699 |
Poverty | +0.656 |
Crit | +0.653 |
Ps | +0.602 |
Neg logit contributions | |
Kislyak | -0.867 |
================================================================= | -0.858 |
IMAGES | -0.834 |
DEA | -0.799 |
Accessory | -0.799 |
Token | Tra |
Token ID | 15721 |
Feature activation | -3.55e-01 |
Pos logit contributions | |
Crit | +0.556 |
Nit | +0.501 |
Poverty | +0.470 |
Crit | +0.465 |
Adapt | +0.431 |
Neg logit contributions | |
Kislyak | -0.684 |
================================================================= | -0.675 |
IMAGES | -0.663 |
zinski | -0.637 |
advertising | -0.636 |
Token | x |
Token ID | 87 |
Feature activation | +5.71e-02 |
Pos logit contributions | |
Crit | +0.821 |
Nit | +0.744 |
Poverty | +0.703 |
Crit | +0.693 |
corruption | +0.638 |
Neg logit contributions | |
Kislyak | -1.037 |
================================================================= | -1.034 |
IMAGES | -1.006 |
advertising | -0.973 |
Accessory | -0.969 |
Token |
|
Token ID | 198 |
Feature activation | -3.54e-01 |
Pos logit contributions | |
Kislyak | +0.160 |
================================================================= | +0.158 |
IMAGES | +0.154 |
advertising | +0.150 |
Accessory | +0.148 |
Neg logit contributions | |
Crit | -0.139 |
Nit | -0.127 |
Poverty | -0.121 |
Crit | -0.119 |
Adapt | -0.110 |
Token |
|
Token ID | 198 |
Feature activation | -2.32e-01 |
Pos logit contributions | |
Crit | +0.835 |
Nit | +0.758 |
Poverty | +0.718 |
Crit | +0.711 |
Adapt | +0.649 |
Neg logit contributions | |
Kislyak | -1.019 |
================================================================= | -1.013 |
IMAGES | -0.986 |
advertising | -0.955 |
Accessory | -0.947 |
Token | Reviewer |
Token ID | 35407 |
Feature activation | -1.63e+00 |
Pos logit contributions | |
Crit | +0.422 |
Nit | +0.372 |
Poverty | +0.335 |
Crit | +0.331 |
Ps | +0.301 |
Neg logit contributions | |
Kislyak | -0.803 |
================================================================= | -0.797 |
IMAGES | -0.783 |
advertising | -0.753 |
zinski | -0.753 |
Token | : |
Token ID | 25 |
Feature activation | +6.46e-02 |
Pos logit contributions | |
Crit | +4.691 |
Nit | +4.275 |
Crit | +4.125 |
Poverty | +4.068 |
Ps | +3.826 |
Neg logit contributions | |
conclud | -3.745 |
Kislyak | -3.741 |
================================================================= | -3.663 |
IMAGES | -3.645 |
ategory | -3.644 |
Token | V |
Token ID | 569 |
Feature activation | +2.08e-01 |
Pos logit contributions | |
Kislyak | +0.178 |
================================================================= | +0.177 |
IMAGES | +0.172 |
advertising | +0.167 |
zinski | +0.165 |
Neg logit contributions | |
Crit | -0.160 |
Nit | -0.146 |
Poverty | -0.139 |
Crit | -0.137 |
corruption | -0.127 |
Token | the |
Token ID | 1169 |
Feature activation | +1.47e-01 |
Pos logit contributions | |
Kislyak | +0.552 |
================================================================= | +0.550 |
IMAGES | +0.531 |
advertising | +0.516 |
zinski | +0.511 |
Neg logit contributions | |
Crit | -0.535 |
Nit | -0.492 |
Poverty | -0.469 |
Crit | -0.462 |
corruption | -0.428 |
Token | Happy |
Token ID | 25082 |
Feature activation | -5.36e-01 |
Pos logit contributions | |
Kislyak | +0.408 |
================================================================= | +0.406 |
IMAGES | +0.393 |
advertising | +0.381 |
zinski | +0.377 |
Neg logit contributions | |
Crit | -0.362 |
Nit | -0.331 |
Poverty | -0.313 |
Crit | -0.309 |
Adapt | -0.286 |
Token | L |
Token ID | 43 |
Feature activation | -3.18e-01 |
Pos logit contributions | |
Crit | +1.293 |
Nit | +1.175 |
Poverty | +1.112 |
Crit | +1.101 |
Adapt | +1.014 |
Neg logit contributions | |
Kislyak | -1.520 |
================================================================= | -1.510 |
IMAGES | -1.468 |
advertising | -1.421 |
zinski | -1.410 |
Token | , |
Token ID | 11 |
Feature activation | -7.21e-04 |
Pos logit contributions | |
Crit | +0.663 |
Nit | +0.612 |
Poverty | +0.585 |
Crit | +0.578 |
Adapt | +0.538 |
Neg logit contributions | |
Kislyak | -0.588 |
================================================================= | -0.581 |
IMAGES | -0.563 |
advertising | -0.542 |
Accessory | -0.540 |
Token | I |
Token ID | 314 |
Feature activation | -1.22e-01 |
Pos logit contributions | |
Crit | +0.002 |
Nit | +0.002 |
Poverty | +0.002 |
Crit | +0.002 |
corruption | +0.002 |
Neg logit contributions | |
================================================================= | -0.002 |
Kislyak | -0.002 |
IMAGES | -0.002 |
advertising | -0.002 |
zinski | -0.002 |
Token | think |
Token ID | 892 |
Feature activation | -2.89e-01 |
Pos logit contributions | |
Crit | +0.280 |
Nit | +0.255 |
Poverty | +0.241 |
Crit | +0.238 |
corruption | +0.218 |
Neg logit contributions | |
Kislyak | -0.356 |
================================================================= | -0.355 |
IMAGES | -0.344 |
advertising | -0.335 |
zinski | -0.332 |
Token |
|
Token ID | 198 |
Feature activation | -3.49e-01 |
Pos logit contributions | |
Crit | +0.701 |
Nit | +0.639 |
Poverty | +0.610 |
Crit | +0.603 |
Adapt | +0.551 |
Neg logit contributions | |
Kislyak | -0.809 |
================================================================= | -0.802 |
IMAGES | -0.778 |
advertising | -0.760 |
Accessory | -0.755 |
Token |
|
Token ID | 198 |
Feature activation | -2.50e-01 |
Pos logit contributions | |
Crit | +0.827 |
Nit | +0.751 |
Poverty | +0.712 |
Crit | +0.704 |
Adapt | +0.644 |
Neg logit contributions | |
Kislyak | -1.002 |
================================================================= | -0.997 |
IMAGES | -0.969 |
advertising | -0.940 |
zinski | -0.932 |
Token | Reviewer |
Token ID | 35407 |
Feature activation | -1.63e+00 |
Pos logit contributions | |
Crit | +0.458 |
Nit | +0.404 |
Poverty | +0.366 |
Crit | +0.363 |
Ps | +0.328 |
Neg logit contributions | |
Kislyak | -0.858 |
================================================================= | -0.853 |
IMAGES | -0.837 |
advertising | -0.809 |
zinski | -0.807 |
Token | : |
Token ID | 25 |
Feature activation | -1.02e-02 |
Pos logit contributions | |
Crit | +4.685 |
Nit | +4.273 |
Crit | +4.124 |
Poverty | +4.065 |
Ps | +3.816 |
Neg logit contributions | |
conclud | -3.756 |
Kislyak | -3.739 |
ategory | -3.675 |
================================================================= | -3.663 |
IMAGES | -3.641 |
Token | B |
Token ID | 347 |
Feature activation | +1.69e-01 |
Pos logit contributions | |
Crit | +0.025 |
Nit | +0.023 |
Poverty | +0.022 |
Crit | +0.022 |
Adapt | +0.020 |
Neg logit contributions | |
Kislyak | -0.028 |
================================================================= | -0.028 |
IMAGES | -0.027 |
advertising | -0.026 |
zinski | -0.026 |
Token | iz |
Token ID | 528 |
Feature activation | -2.13e-01 |
Pos logit contributions | |
Kislyak | +0.432 |
================================================================= | +0.429 |
IMAGES | +0.415 |
advertising | +0.402 |
zinski | +0.397 |
Neg logit contributions | |
Crit | -0.455 |
Nit | -0.420 |
Poverty | -0.400 |
Crit | -0.396 |
Ps | -0.368 |
Token | r |
Token ID | 81 |
Feature activation | +2.00e-02 |
Pos logit contributions | |
Crit | +0.562 |
Nit | +0.518 |
Poverty | +0.488 |
Crit | +0.485 |
Adapt | +0.451 |
Neg logit contributions | |
Kislyak | -0.555 |
================================================================= | -0.552 |
IMAGES | -0.537 |
advertising | -0.516 |
Accessory | -0.510 |
Token | Pix |
Token ID | 21642 |
Feature activation | +1.96e-01 |
Pos logit contributions | |
Kislyak | +0.053 |
================================================================= | +0.053 |
IMAGES | +0.051 |
advertising | +0.050 |
zinski | +0.049 |
Neg logit contributions | |
Crit | -0.052 |
Nit | -0.047 |
Poverty | -0.045 |
Crit | -0.044 |
corruption | -0.041 |
Token | 2010 |
Token ID | 3050 |
Feature activation | -3.21e-01 |
Pos logit contributions | |
Crit | +0.902 |
Nit | +0.833 |
Poverty | +0.811 |
Crit | +0.797 |
Adapt | +0.734 |
Neg logit contributions | |
Kislyak | -0.822 |
================================================================= | -0.819 |
IMAGES | -0.783 |
advertising | -0.760 |
zinski | -0.758 |
Token | boys |
Token ID | 13202 |
Feature activation | -1.87e-01 |
Pos logit contributions | |
Crit | +1.095 |
Nit | +1.014 |
Crit | +0.956 |
Poverty | +0.946 |
corruption | +0.935 |
Neg logit contributions | |
Kislyak | -0.608 |
================================================================= | -0.593 |
IMAGES | -0.570 |
Accessory | -0.538 |
DEA | -0.538 |
Token | beware |
Token ID | 40600 |
Feature activation | -3.21e-01 |
Pos logit contributions | |
Crit | +0.478 |
Nit | +0.439 |
Poverty | +0.419 |
Crit | +0.413 |
corruption | +0.384 |
Neg logit contributions | |
Kislyak | -0.500 |
================================================================= | -0.496 |
IMAGES | -0.478 |
zinski | -0.464 |
advertising | -0.463 |
Token |
|
Token ID | 198 |
Feature activation | -1.32e-01 |
Pos logit contributions | |
Crit | +0.772 |
Nit | +0.702 |
Poverty | +0.669 |
Crit | +0.662 |
Adapt | +0.606 |
Neg logit contributions | |
Kislyak | -0.910 |
================================================================= | -0.898 |
IMAGES | -0.875 |
advertising | -0.850 |
Accessory | -0.847 |
Token |
|
Token ID | 198 |
Feature activation | -7.91e-02 |
Pos logit contributions | |
Crit | +0.319 |
Nit | +0.292 |
Poverty | +0.277 |
Crit | +0.273 |
Adapt | +0.251 |
Neg logit contributions | |
Kislyak | -0.373 |
================================================================= | -0.371 |
IMAGES | -0.360 |
advertising | -0.350 |
zinski | -0.346 |
Token | Reviewer |
Token ID | 35407 |
Feature activation | -1.62e+00 |
Pos logit contributions | |
Crit | +0.145 |
Nit | +0.128 |
Poverty | +0.116 |
Crit | +0.115 |
Ps | +0.104 |
Neg logit contributions | |
Kislyak | -0.272 |
================================================================= | -0.270 |
IMAGES | -0.265 |
advertising | -0.256 |
zinski | -0.255 |
Token | : |
Token ID | 25 |
Feature activation | -2.55e-03 |
Pos logit contributions | |
Crit | +4.664 |
Nit | +4.253 |
Crit | +4.106 |
Poverty | +4.053 |
Ps | +3.806 |
Neg logit contributions | |
conclud | -3.717 |
Kislyak | -3.703 |
================================================================= | -3.648 |
IMAGES | -3.632 |
ategory | -3.616 |
Token | j |
Token ID | 474 |
Feature activation | +1.73e-01 |
Pos logit contributions | |
Crit | +0.006 |
Nit | +0.006 |
Poverty | +0.006 |
Crit | +0.005 |
Adapt | +0.005 |
Neg logit contributions | |
Kislyak | -0.007 |
================================================================= | -0.007 |
IMAGES | -0.007 |
advertising | -0.007 |
zinski | -0.006 |
Token | wt |
Token ID | 46569 |
Feature activation | -2.57e-01 |
Pos logit contributions | |
Kislyak | +0.443 |
================================================================= | +0.440 |
IMAGES | +0.427 |
advertising | +0.410 |
Accessory | +0.408 |
Neg logit contributions | |
Crit | -0.466 |
Nit | -0.430 |
Poverty | -0.408 |
Crit | -0.404 |
corruption | -0.377 |
Token | 0001 |
Token ID | 18005 |
Feature activation | +3.03e-01 |
Pos logit contributions | |
Crit | +0.664 |
Nit | +0.611 |
Poverty | +0.580 |
Crit | +0.574 |
corruption | +0.534 |
Neg logit contributions | |
Kislyak | -0.679 |
================================================================= | -0.673 |
IMAGES | -0.651 |
advertising | -0.630 |
Accessory | -0.626 |
Token | - |
Token ID | 532 |
Feature activation | -1.51e-01 |
Pos logit contributions | |
================================================================= | +0.740 |
Kislyak | +0.734 |
IMAGES | +0.704 |
advertising | +0.690 |
zinski | +0.682 |
Neg logit contributions | |
Crit | -0.837 |
Nit | -0.775 |
Poverty | -0.744 |
Crit | -0.729 |
Adapt | -0.689 |
Token | null |
Token ID | 9242 |
Feature activation | +3.60e-01 |
Pos logit contributions | |
Kislyak | +0.208 |
================================================================= | +0.205 |
IMAGES | +0.200 |
advertising | +0.194 |
zinski | +0.192 |
Neg logit contributions | |
Crit | -0.189 |
Nit | -0.175 |
Poverty | -0.165 |
Crit | -0.164 |
Adapt | -0.152 |
Token | atomic |
Token ID | 17226 |
Feature activation | +3.47e-01 |
Pos logit contributions | |
Kislyak | +0.513 |
================================================================= | +0.502 |
IMAGES | +0.479 |
advertising | +0.446 |
zinski | +0.440 |
Neg logit contributions | |
Crit | -1.376 |
Nit | -1.307 |
Poverty | -1.258 |
Crit | -1.253 |
Adapt | -1.196 |
Token | pointer |
Token ID | 17562 |
Feature activation | +1.54e-01 |
Pos logit contributions | |
Kislyak | +0.935 |
================================================================= | +0.913 |
IMAGES | +0.896 |
advertising | +0.877 |
zinski | +0.872 |
Neg logit contributions | |
Crit | -0.887 |
Nit | -0.821 |
Crit | -0.773 |
Poverty | -0.773 |
Adapt | -0.716 |
Token | . |
Token ID | 13 |
Feature activation | -1.96e-01 |
Pos logit contributions | |
Kislyak | +0.390 |
================================================================= | +0.379 |
IMAGES | +0.370 |
advertising | +0.362 |
zinski | +0.361 |
Neg logit contributions | |
Crit | -0.417 |
Nit | -0.389 |
Crit | -0.368 |
Poverty | -0.367 |
Adapt | -0.341 |
Token | pub |
Token ID | 2240 |
Feature activation | -8.21e-01 |
Pos logit contributions | |
Crit | +0.500 |
Nit | +0.462 |
Crit | +0.438 |
Poverty | +0.437 |
Adapt | +0.402 |
Neg logit contributions | |
Kislyak | -0.524 |
================================================================= | -0.512 |
IMAGES | -0.500 |
advertising | -0.489 |
zinski | -0.486 |
Token | fn |
Token ID | 24714 |
Feature activation | -1.52e+00 |
Pos logit contributions | |
Crit | +1.966 |
Nit | +1.811 |
Crit | +1.709 |
Poverty | +1.697 |
Adapt | +1.566 |
Neg logit contributions | |
Kislyak | -2.307 |
================================================================= | -2.233 |
IMAGES | -2.189 |
advertising | -2.188 |
zinski | -2.155 |
Token | null |
Token ID | 9242 |
Feature activation | +3.53e-01 |
Pos logit contributions | |
Crit | +3.241 |
Nit | +3.022 |
Crit | +2.878 |
Poverty | +2.795 |
crit | +2.651 |
Neg logit contributions | |
Kislyak | -4.535 |
advertising | -4.358 |
================================================================= | -4.348 |
Accessory | -4.253 |
IMAGES | -4.234 |
Token | () |
Token ID | 7499 |
Feature activation | -7.75e-03 |
Pos logit contributions | |
================================================================= | +0.842 |
Kislyak | +0.841 |
IMAGES | +0.805 |
advertising | +0.781 |
Accessory | +0.773 |
Neg logit contributions | |
Crit | -0.998 |
Nit | -0.923 |
Poverty | -0.889 |
Crit | -0.876 |
Lumpur | -0.818 |
Token | -> |
Token ID | 4613 |
Feature activation | +4.27e-01 |
Pos logit contributions | |
Crit | +0.024 |
Nit | +0.023 |
Poverty | +0.022 |
Crit | +0.021 |
corruption | +0.020 |
Neg logit contributions | |
Kislyak | -0.016 |
================================================================= | -0.016 |
IMAGES | -0.016 |
advertising | -0.015 |
zinski | -0.015 |
Token | Atomic |
Token ID | 28976 |
Feature activation | +3.39e-01 |
Pos logit contributions | |
Kislyak | +1.116 |
================================================================= | +1.104 |
IMAGES | +1.066 |
advertising | +1.048 |
zinski | +1.033 |
Neg logit contributions | |
Crit | -1.118 |
Nit | -1.033 |
Poverty | -0.987 |
Crit | -0.981 |
Adapt | -0.902 |
Token | < |
Token ID | 1279 |
Feature activation | -4.19e-01 |
Pos logit contributions | |
Kislyak | +0.656 |
================================================================= | +0.653 |
IMAGES | +0.622 |
advertising | +0.598 |
zinski | +0.589 |
Neg logit contributions | |
Crit | -1.115 |
Nit | -1.044 |
Poverty | -1.006 |
Crit | -0.996 |
corruption | -0.941 |
Token | { |
Token ID | 1391 |
Feature activation | -4.88e-01 |
Pos logit contributions | |
Kislyak | +1.183 |
================================================================= | +1.179 |
IMAGES | +1.137 |
advertising | +1.105 |
zinski | +1.093 |
Neg logit contributions | |
Crit | -1.190 |
Nit | -1.095 |
Poverty | -1.046 |
Crit | -1.031 |
corruption | -0.958 |
Token | Board |
Token ID | 5926 |
Feature activation | +2.80e-01 |
Pos logit contributions | |
Crit | +1.178 |
Nit | +1.070 |
Poverty | +1.016 |
Crit | +1.014 |
Adapt | +0.925 |
Neg logit contributions | |
Kislyak | -1.379 |
================================================================= | -1.357 |
IMAGES | -1.320 |
advertising | -1.300 |
zinski | -1.289 |
Token | { |
Token ID | 1391 |
Feature activation | -5.54e-01 |
Pos logit contributions | |
================================================================= | +0.727 |
Kislyak | +0.726 |
IMAGES | +0.698 |
advertising | +0.679 |
Accessory | +0.671 |
Neg logit contributions | |
Crit | -0.734 |
Nit | -0.677 |
Poverty | -0.648 |
Crit | -0.637 |
corruption | -0.592 |
Token | turn |
Token ID | 1210 |
Feature activation | -1.28e-01 |
Pos logit contributions | |
Crit | +1.357 |
Nit | +1.233 |
Crit | +1.172 |
Poverty | +1.171 |
Adapt | +1.072 |
Neg logit contributions | |
Kislyak | -1.541 |
================================================================= | -1.516 |
IMAGES | -1.475 |
advertising | -1.451 |
zinski | -1.440 |
Token | : |
Token ID | 1058 |
Feature activation | -4.40e-01 |
Pos logit contributions | |
Crit | +0.391 |
Nit | +0.364 |
Poverty | +0.350 |
Crit | +0.346 |
corruption | +0.325 |
Neg logit contributions | |
Kislyak | -0.281 |
================================================================= | -0.279 |
IMAGES | -0.268 |
advertising | -0.258 |
zinski | -0.256 |
Token | self |
Token ID | 2116 |
Feature activation | -1.51e+00 |
Pos logit contributions | |
Crit | +1.062 |
Nit | +0.965 |
Poverty | +0.923 |
Crit | +0.917 |
Adapt | +0.835 |
Neg logit contributions | |
Kislyak | -1.237 |
================================================================= | -1.228 |
IMAGES | -1.190 |
advertising | -1.171 |
Accessory | -1.157 |
Token | . |
Token ID | 764 |
Feature activation | -5.40e-02 |
Pos logit contributions | |
Crit | +3.999 |
Nit | +3.641 |
Crit | +3.454 |
Poverty | +3.436 |
corruption | +3.210 |
Neg logit contributions | |
Kislyak | -3.821 |
================================================================= | -3.747 |
IMAGES | -3.675 |
advertising | -3.595 |
zinski | -3.577 |
Token | clone |
Token ID | 21018 |
Feature activation | -2.22e-02 |
Pos logit contributions | |
Crit | +0.142 |
Nit | +0.130 |
Poverty | +0.124 |
Crit | +0.123 |
corruption | +0.114 |
Neg logit contributions | |
Kislyak | -0.141 |
================================================================= | -0.140 |
IMAGES | -0.135 |
advertising | -0.131 |
zinski | -0.130 |
Token | (), |
Token ID | 29994 |
Feature activation | -4.62e-02 |
Pos logit contributions | |
Crit | +0.062 |
Nit | +0.058 |
Poverty | +0.055 |
Crit | +0.055 |
corruption | +0.051 |
Neg logit contributions | |
Kislyak | -0.054 |
================================================================= | -0.053 |
IMAGES | -0.051 |
advertising | -0.050 |
zinski | -0.049 |
Token | pieces |
Token ID | 5207 |
Feature activation | +3.41e-01 |
Pos logit contributions | |
Crit | +0.110 |
Nit | +0.101 |
Poverty | +0.096 |
Crit | +0.094 |
corruption | +0.087 |
Neg logit contributions | |
Kislyak | -0.131 |
================================================================= | -0.130 |
IMAGES | -0.126 |
advertising | -0.123 |
zinski | -0.122 |
Token | : |
Token ID | 1058 |
Feature activation | -2.36e-01 |
Pos logit contributions | |
================================================================= | +0.633 |
Kislyak | +0.623 |
IMAGES | +0.594 |
advertising | +0.573 |
Accessory | +0.563 |
Neg logit contributions | |
Crit | -1.141 |
Nit | -1.076 |
Poverty | -1.043 |
Crit | -1.026 |
Lumpur | -0.976 |
Token | bad |
Token ID | 2089 |
Feature activation | -1.56e-01 |
Pos logit contributions | |
Crit | +0.191 |
Nit | +0.167 |
Poverty | +0.154 |
Crit | +0.150 |
corruption | +0.131 |
Neg logit contributions | |
Kislyak | -0.422 |
================================================================= | -0.421 |
IMAGES | -0.410 |
advertising | -0.402 |
zinski | -0.399 |
Token | idea |
Token ID | 2126 |
Feature activation | -2.74e-01 |
Pos logit contributions | |
Crit | +0.402 |
Nit | +0.371 |
Poverty | +0.356 |
Crit | +0.351 |
corruption | +0.325 |
Neg logit contributions | |
Kislyak | -0.414 |
================================================================= | -0.412 |
IMAGES | -0.399 |
zinski | -0.385 |
advertising | -0.384 |
Token | . |
Token ID | 13 |
Feature activation | -1.97e-02 |
Pos logit contributions | |
Crit | +0.864 |
Nit | +0.806 |
Poverty | +0.775 |
Crit | +0.771 |
Ps | +0.721 |
Neg logit contributions | |
Kislyak | -0.569 |
================================================================= | -0.559 |
IMAGES | -0.546 |
Accessory | -0.519 |
zinski | -0.516 |
Token |
|
Token ID | 198 |
Feature activation | -1.60e-01 |
Pos logit contributions | |
Crit | +0.051 |
Nit | +0.047 |
Poverty | +0.045 |
Crit | +0.044 |
corruption | +0.041 |
Neg logit contributions | |
================================================================= | -0.052 |
Kislyak | -0.052 |
IMAGES | -0.050 |
advertising | -0.048 |
zinski | -0.048 |
Token |
|
Token ID | 198 |
Feature activation | -1.50e-01 |
Pos logit contributions | |
Crit | +0.387 |
Nit | +0.353 |
Poverty | +0.335 |
Crit | +0.330 |
corruption | +0.304 |
Neg logit contributions | |
Kislyak | -0.452 |
================================================================= | -0.450 |
IMAGES | -0.436 |
advertising | -0.424 |
zinski | -0.420 |
Token | Reviewer |
Token ID | 35407 |
Feature activation | -1.48e+00 |
Pos logit contributions | |
Crit | +0.273 |
Nit | +0.241 |
Poverty | +0.215 |
Crit | +0.213 |
Adapt | +0.194 |
Neg logit contributions | |
Kislyak | -0.519 |
================================================================= | -0.516 |
IMAGES | -0.508 |
zinski | -0.487 |
advertising | -0.487 |
Token | : |
Token ID | 25 |
Feature activation | -3.45e-02 |
Pos logit contributions | |
Crit | +4.310 |
Nit | +3.939 |
Crit | +3.809 |
Poverty | +3.765 |
Ps | +3.520 |
Neg logit contributions | |
Kislyak | -3.341 |
conclud | -3.290 |
================================================================= | -3.275 |
IMAGES | -3.248 |
ategory | -3.241 |
Token | Mike |
Token ID | 4995 |
Feature activation | +1.94e-01 |
Pos logit contributions | |
Crit | +0.086 |
Nit | +0.079 |
Poverty | +0.075 |
Crit | +0.074 |
Adapt | +0.068 |
Neg logit contributions | |
Kislyak | -0.094 |
================================================================= | -0.094 |
IMAGES | -0.091 |
advertising | -0.089 |
zinski | -0.088 |
Token | M |
Token ID | 44 |
Feature activation | -4.16e-01 |
Pos logit contributions | |
Kislyak | +0.560 |
================================================================= | +0.558 |
IMAGES | +0.540 |
advertising | +0.526 |
zinski | +0.521 |
Neg logit contributions | |
Crit | -0.456 |
Nit | -0.416 |
Poverty | -0.394 |
Crit | -0.388 |
corruption | -0.357 |
Token | anners |
Token ID | 15672 |
Feature activation | -1.44e-01 |
Pos logit contributions | |
Crit | +0.820 |
Nit | +0.730 |
Poverty | +0.683 |
Crit | +0.671 |
corruption | +0.606 |
Neg logit contributions | |
Kislyak | -1.359 |
================================================================= | -1.354 |
IMAGES | -1.315 |
advertising | -1.281 |
Accessory | -1.273 |
Token | - |
Token ID | 532 |
Feature activation | -1.91e-01 |
Pos logit contributions | |
Crit | +0.379 |
Nit | +0.350 |
Poverty | +0.334 |
Crit | +0.329 |
corruption | +0.306 |
Neg logit contributions | |
Kislyak | -0.370 |
================================================================= | -0.370 |
IMAGES | -0.355 |
advertising | -0.346 |
zinski | -0.343 |
Token | < |
Token ID | 1279 |
Feature activation | -2.78e-01 |
Pos logit contributions | |
Crit | +0.186 |
Nit | +0.173 |
Poverty | +0.165 |
Crit | +0.164 |
Adapt | +0.155 |
Neg logit contributions | |
Kislyak | -0.131 |
================================================================= | -0.129 |
IMAGES | -0.125 |
advertising | -0.120 |
zinski | -0.120 |
Token | T |
Token ID | 309 |
Feature activation | -1.31e-01 |
Pos logit contributions | |
Crit | +0.725 |
Nit | +0.668 |
Crit | +0.646 |
Poverty | +0.645 |
Adapt | +0.586 |
Neg logit contributions | |
Kislyak | -0.719 |
================================================================= | -0.712 |
IMAGES | -0.688 |
advertising | -0.683 |
zinski | -0.675 |
Token | > |
Token ID | 1875 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
Crit | +0.401 |
Nit | +0.371 |
Poverty | +0.357 |
Crit | +0.357 |
Ps | +0.333 |
Neg logit contributions | |
Kislyak | -0.282 |
================================================================= | -0.281 |
IMAGES | -0.268 |
advertising | -0.264 |
zinski | -0.258 |
Token | { |
Token ID | 1391 |
Feature activation | -3.40e-01 |
Pos logit contributions | |
Kislyak | +0.283 |
================================================================= | +0.281 |
IMAGES | +0.273 |
advertising | +0.268 |
zinski | +0.266 |
Neg logit contributions | |
Crit | -0.262 |
Nit | -0.239 |
Poverty | -0.229 |
Crit | -0.228 |
Adapt | -0.208 |
Token | pub |
Token ID | 2240 |
Feature activation | -5.74e-01 |
Pos logit contributions | |
Crit | +0.831 |
Nit | +0.759 |
Poverty | +0.722 |
Crit | +0.715 |
corruption | +0.656 |
Neg logit contributions | |
Kislyak | -0.949 |
================================================================= | -0.940 |
IMAGES | -0.911 |
advertising | -0.891 |
zinski | -0.882 |
Token | fn |
Token ID | 24714 |
Feature activation | -1.45e+00 |
Pos logit contributions | |
Crit | +1.393 |
Nit | +1.270 |
Crit | +1.210 |
Poverty | +1.207 |
Adapt | +1.113 |
Neg logit contributions | |
Kislyak | -1.602 |
================================================================= | -1.566 |
advertising | -1.523 |
IMAGES | -1.522 |
zinski | -1.502 |
Token | new |
Token ID | 649 |
Feature activation | -8.55e-02 |
Pos logit contributions | |
Crit | +3.541 |
Nit | +3.247 |
Crit | +3.172 |
Poverty | +3.151 |
crit | +2.975 |
Neg logit contributions | |
Kislyak | -3.818 |
================================================================= | -3.711 |
advertising | -3.672 |
Accessory | -3.604 |
IMAGES | -3.581 |
Token | ( |
Token ID | 357 |
Feature activation | -2.35e-01 |
Pos logit contributions | |
Crit | +0.236 |
Nit | +0.218 |
Poverty | +0.209 |
Crit | +0.206 |
Adapt | +0.192 |
Neg logit contributions | |
Kislyak | -0.211 |
================================================================= | -0.210 |
IMAGES | -0.202 |
advertising | -0.196 |
zinski | -0.194 |
Token | t |
Token ID | 256 |
Feature activation | +1.45e-01 |
Pos logit contributions | |
Crit | +0.589 |
Nit | +0.541 |
Poverty | +0.513 |
Crit | +0.509 |
Ps | +0.471 |
Neg logit contributions | |
Kislyak | -0.638 |
================================================================= | -0.632 |
IMAGES | -0.611 |
advertising | -0.598 |
Accessory | -0.591 |
Token | : |
Token ID | 1058 |
Feature activation | -4.84e-01 |
Pos logit contributions | |
Kislyak | +0.301 |
================================================================= | +0.295 |
IMAGES | +0.282 |
advertising | +0.277 |
zinski | +0.272 |
Neg logit contributions | |
Crit | -0.462 |
Nit | -0.430 |
Poverty | -0.415 |
Crit | -0.412 |
Ps | -0.387 |
Token | T |
Token ID | 309 |
Feature activation | -1.08e-01 |
Pos logit contributions | |
Crit | +1.189 |
Nit | +1.091 |
Poverty | +1.061 |
Crit | +1.058 |
Adapt | +0.957 |
Neg logit contributions | |
Kislyak | -1.310 |
================================================================= | -1.289 |
advertising | -1.262 |
IMAGES | -1.250 |
zinski | -1.247 |
Token | are |
Token ID | 389 |
Feature activation | -1.37e-01 |
Pos logit contributions | |
Crit | +0.486 |
Nit | +0.433 |
Poverty | +0.408 |
Crit | +0.404 |
corruption | +0.360 |
Neg logit contributions | |
Kislyak | -0.790 |
================================================================= | -0.781 |
IMAGES | -0.764 |
advertising | -0.744 |
zinski | -0.742 |
Token | Here |
Token ID | 3423 |
Feature activation | +1.45e-01 |
Pos logit contributions | |
Crit | +0.315 |
Nit | +0.286 |
Poverty | +0.272 |
Crit | +0.268 |
Adapt | +0.245 |
Neg logit contributions | |
Kislyak | -0.400 |
================================================================= | -0.396 |
IMAGES | -0.385 |
advertising | -0.375 |
zinski | -0.373 |
Token | .... |
Token ID | 1106 |
Feature activation | -4.33e-02 |
Pos logit contributions | |
Kislyak | +0.471 |
================================================================= | +0.470 |
IMAGES | +0.457 |
advertising | +0.445 |
zinski | +0.443 |
Neg logit contributions | |
Crit | -0.285 |
Nit | -0.255 |
Poverty | -0.239 |
Crit | -0.234 |
corruption | -0.211 |
Token |
|
Token ID | 198 |
Feature activation | -3.48e-01 |
Pos logit contributions | |
Crit | +0.111 |
Nit | +0.102 |
Poverty | +0.097 |
Crit | +0.096 |
corruption | +0.089 |
Neg logit contributions | |
================================================================= | -0.115 |
Kislyak | -0.115 |
IMAGES | -0.111 |
advertising | -0.107 |
zinski | -0.107 |
Token |
|
Token ID | 198 |
Feature activation | -1.93e-01 |
Pos logit contributions | |
Crit | +0.827 |
Nit | +0.753 |
Poverty | +0.714 |
Crit | +0.705 |
Adapt | +0.646 |
Neg logit contributions | |
Kislyak | -0.993 |
================================================================= | -0.988 |
IMAGES | -0.960 |
advertising | -0.931 |
zinski | -0.923 |
Token | Reviewer |
Token ID | 35407 |
Feature activation | -1.44e+00 |
Pos logit contributions | |
Crit | +0.353 |
Nit | +0.311 |
Poverty | +0.283 |
Crit | +0.281 |
Adapt | +0.252 |
Neg logit contributions | |
Kislyak | -0.661 |
================================================================= | -0.658 |
IMAGES | -0.645 |
advertising | -0.623 |
zinski | -0.622 |
Token | : |
Token ID | 25 |
Feature activation | -4.21e-02 |
Pos logit contributions | |
Crit | +4.207 |
Nit | +3.843 |
Crit | +3.716 |
Poverty | +3.676 |
Ps | +3.435 |
Neg logit contributions | |
Kislyak | -3.260 |
conclud | -3.220 |
================================================================= | -3.187 |
ategory | -3.161 |
IMAGES | -3.157 |
Token | SAS |
Token ID | 35516 |
Feature activation | +3.19e-01 |
Pos logit contributions | |
Crit | +0.106 |
Nit | +0.097 |
Poverty | +0.093 |
Crit | +0.091 |
Adapt | +0.084 |
Neg logit contributions | |
Kislyak | -0.114 |
================================================================= | -0.113 |
IMAGES | -0.110 |
advertising | -0.107 |
Accessory | -0.106 |
Token | K |
Token ID | 42 |
Feature activation | -2.59e-02 |
Pos logit contributions | |
Kislyak | +0.873 |
================================================================= | +0.871 |
IMAGES | +0.840 |
advertising | +0.818 |
zinski | +0.811 |
Neg logit contributions | |
Crit | -0.795 |
Nit | -0.728 |
Poverty | -0.694 |
Crit | -0.682 |
corruption | -0.632 |
Token | ATCH |
Token ID | 11417 |
Feature activation | -2.79e-01 |
Pos logit contributions | |
Crit | +0.065 |
Nit | +0.060 |
Poverty | +0.057 |
Crit | +0.056 |
corruption | +0.052 |
Neg logit contributions | |
Kislyak | -0.070 |
================================================================= | -0.070 |
IMAGES | -0.068 |
advertising | -0.066 |
zinski | -0.065 |
Token | EW |
Token ID | 6217 |
Feature activation | -4.23e-02 |
Pos logit contributions | |
Crit | +0.706 |
Nit | +0.648 |
Poverty | +0.616 |
Crit | +0.608 |
Adapt | +0.562 |
Neg logit contributions | |
Kislyak | -0.755 |
================================================================= | -0.747 |
IMAGES | -0.721 |
advertising | -0.704 |
zinski | -0.698 |
Token |
|
Token ID | 198 |
Feature activation | +2.25e-02 |
Pos logit contributions | |
Crit | +0.020 |
Nit | +0.018 |
Poverty | +0.017 |
Crit | +0.017 |
corruption | +0.015 |
Neg logit contributions | |
Kislyak | -0.022 |
================================================================= | -0.022 |
IMAGES | -0.021 |
advertising | -0.021 |
zinski | -0.021 |
Token | impl |
Token ID | 23928 |
Feature activation | -4.99e-01 |
Pos logit contributions | |
Kislyak | +0.061 |
================================================================= | +0.061 |
IMAGES | +0.059 |
advertising | +0.057 |
zinski | +0.057 |
Neg logit contributions | |
Crit | -0.057 |
Nit | -0.052 |
Poverty | -0.049 |
Crit | -0.049 |
Adapt | -0.046 |
Token | Board |
Token ID | 5926 |
Feature activation | +7.41e-01 |
Pos logit contributions | |
Crit | +1.383 |
Nit | +1.275 |
Crit | +1.230 |
Poverty | +1.229 |
Adapt | +1.132 |
Neg logit contributions | |
Kislyak | -1.234 |
================================================================= | -1.206 |
IMAGES | -1.166 |
advertising | -1.161 |
zinski | -1.140 |
Token | { |
Token ID | 1391 |
Feature activation | -3.73e-01 |
Pos logit contributions | |
================================================================= | +1.922 |
Kislyak | +1.914 |
IMAGES | +1.838 |
advertising | +1.787 |
Accessory | +1.771 |
Neg logit contributions | |
Crit | -1.935 |
Nit | -1.789 |
Poverty | -1.712 |
Crit | -1.679 |
corruption | -1.565 |
Token | pub |
Token ID | 2240 |
Feature activation | -7.00e-01 |
Pos logit contributions | |
Crit | +0.935 |
Nit | +0.856 |
Poverty | +0.813 |
Crit | +0.811 |
Adapt | +0.746 |
Neg logit contributions | |
Kislyak | -1.016 |
================================================================= | -0.999 |
IMAGES | -0.976 |
advertising | -0.956 |
zinski | -0.947 |
Token | fn |
Token ID | 24714 |
Feature activation | -1.40e+00 |
Pos logit contributions | |
Crit | +1.707 |
Nit | +1.557 |
Crit | +1.478 |
Poverty | +1.466 |
Adapt | +1.360 |
Neg logit contributions | |
Kislyak | -1.958 |
================================================================= | -1.899 |
IMAGES | -1.874 |
advertising | -1.849 |
zinski | -1.835 |
Token | copy |
Token ID | 4866 |
Feature activation | -1.08e-01 |
Pos logit contributions | |
Crit | +3.343 |
Nit | +2.992 |
Crit | +2.976 |
Poverty | +2.898 |
crit | +2.756 |
Neg logit contributions | |
Kislyak | -3.869 |
================================================================= | -3.690 |
advertising | -3.663 |
zinski | -3.623 |
Accessory | -3.617 |
Token | _ |
Token ID | 62 |
Feature activation | -3.87e-01 |
Pos logit contributions | |
Crit | +0.295 |
Nit | +0.272 |
Poverty | +0.260 |
Crit | +0.257 |
Adapt | +0.240 |
Neg logit contributions | |
Kislyak | -0.268 |
================================================================= | -0.266 |
IMAGES | -0.257 |
advertising | -0.249 |
zinski | -0.247 |
Token | board |
Token ID | 3526 |
Feature activation | +2.51e-01 |
Pos logit contributions | |
Crit | +0.979 |
Nit | +0.890 |
Poverty | +0.836 |
Crit | +0.834 |
corruption | +0.797 |
Neg logit contributions | |
Kislyak | -1.050 |
================================================================= | -1.034 |
IMAGES | -1.009 |
zinski | -0.969 |
Accessory | -0.966 |
Token | ( |
Token ID | 357 |
Feature activation | -2.10e-01 |
Pos logit contributions | |
Kislyak | +0.653 |
================================================================= | +0.646 |
IMAGES | +0.626 |
advertising | +0.609 |
zinski | +0.604 |
Neg logit contributions | |
Crit | -0.660 |
Nit | -0.606 |
Poverty | -0.578 |
Crit | -0.571 |
corruption | -0.531 |
Token | & |
Token ID | 1222 |
Feature activation | +1.23e-01 |
Pos logit contributions | |
Crit | +0.544 |
Nit | +0.498 |
Poverty | +0.475 |
Crit | +0.470 |
corruption | +0.437 |
Neg logit contributions | |
Kislyak | -0.556 |
================================================================= | -0.552 |
IMAGES | -0.533 |
advertising | -0.520 |
zinski | -0.515 |
Token | let |
Token ID | 1309 |
Feature activation | -1.01e+00 |
Pos logit contributions | |
Crit | +1.247 |
Nit | +1.132 |
Poverty | +1.079 |
Crit | +1.079 |
Adapt | +0.980 |
Neg logit contributions | |
Kislyak | -1.443 |
================================================================= | -1.421 |
IMAGES | -1.380 |
advertising | -1.364 |
zinski | -1.350 |
Token | prev |
Token ID | 8654 |
Feature activation | +2.69e-01 |
Pos logit contributions | |
Crit | +2.263 |
Nit | +2.030 |
Crit | +1.956 |
Poverty | +1.948 |
crit | +1.802 |
Neg logit contributions | |
Kislyak | -2.972 |
================================================================= | -2.908 |
advertising | -2.860 |
IMAGES | -2.816 |
zinski | -2.798 |
Token | _ |
Token ID | 62 |
Feature activation | -3.98e-01 |
Pos logit contributions | |
================================================================= | +0.662 |
Kislyak | +0.655 |
IMAGES | +0.631 |
advertising | +0.613 |
Accessory | +0.608 |
Neg logit contributions | |
Crit | -0.740 |
Nit | -0.688 |
Poverty | -0.662 |
Crit | -0.650 |
Lumpur | -0.610 |
Token | state |
Token ID | 5219 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
Crit | +1.101 |
Nit | +1.014 |
Poverty | +0.968 |
Crit | +0.960 |
corruption | +0.904 |
Neg logit contributions | |
Kislyak | -0.981 |
================================================================= | -0.972 |
IMAGES | -0.937 |
advertising | -0.906 |
zinski | -0.899 |
Token | = |
Token ID | 796 |
Feature activation | -2.42e-01 |
Pos logit contributions | |
Kislyak | +0.247 |
================================================================= | +0.243 |
IMAGES | +0.234 |
advertising | +0.228 |
zinski | +0.226 |
Neg logit contributions | |
Crit | -0.312 |
Nit | -0.289 |
Poverty | -0.277 |
Crit | -0.274 |
Ps | -0.257 |
Token | self |
Token ID | 2116 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
Crit | +0.611 |
Nit | +0.554 |
Crit | +0.536 |
Poverty | +0.534 |
Adapt | +0.488 |
Neg logit contributions | |
Kislyak | -0.651 |
================================================================= | -0.643 |
IMAGES | -0.622 |
advertising | -0.620 |
zinski | -0.612 |
Token | . |
Token ID | 764 |
Feature activation | -1.66e-01 |
Pos logit contributions | |
Crit | +3.640 |
Nit | +3.315 |
Crit | +3.165 |
Poverty | +3.160 |
corruption | +2.944 |
Neg logit contributions | |
Kislyak | -3.398 |
================================================================= | -3.318 |
IMAGES | -3.245 |
advertising | -3.175 |
zinski | -3.150 |
Token | copy |
Token ID | 30073 |
Feature activation | -3.78e-01 |
Pos logit contributions | |
Crit | +0.413 |
Nit | +0.375 |
Poverty | +0.356 |
Crit | +0.355 |
corruption | +0.332 |
Neg logit contributions | |
Kislyak | -0.458 |
================================================================= | -0.453 |
IMAGES | -0.440 |
advertising | -0.427 |
zinski | -0.424 |
Token | _ |
Token ID | 62 |
Feature activation | -4.60e-01 |
Pos logit contributions | |
Crit | +1.001 |
Nit | +0.911 |
Poverty | +0.868 |
Crit | +0.863 |
corruption | +0.811 |
Neg logit contributions | |
Kislyak | -0.985 |
================================================================= | -0.967 |
IMAGES | -0.943 |
advertising | -0.910 |
zinski | -0.909 |
Token | board |
Token ID | 3526 |
Feature activation | +2.64e-01 |
Pos logit contributions | |
Crit | +1.224 |
Nit | +1.123 |
Poverty | +1.067 |
Crit | +1.059 |
corruption | +1.000 |
Neg logit contributions | |
Kislyak | -1.185 |
================================================================= | -1.178 |
IMAGES | -1.139 |
advertising | -1.095 |
zinski | -1.091 |
Token | (); |
Token ID | 13979 |
Feature activation | -1.22e-01 |
Pos logit contributions | |
Kislyak | +0.669 |
================================================================= | +0.665 |
IMAGES | +0.642 |
advertising | +0.623 |
zinski | +0.617 |
Neg logit contributions | |
Crit | -0.714 |
Nit | -0.659 |
Poverty | -0.629 |
Crit | -0.621 |
corruption | -0.579 |
Token | < |
Token ID | 1279 |
Feature activation | -2.87e-01 |
Pos logit contributions | |
Kislyak | +0.459 |
================================================================= | +0.453 |
IMAGES | +0.435 |
advertising | +0.419 |
zinski | +0.414 |
Neg logit contributions | |
Crit | -0.738 |
Nit | -0.689 |
Poverty | -0.663 |
Crit | -0.655 |
Adapt | -0.621 |
Token | T |
Token ID | 309 |
Feature activation | -6.24e-02 |
Pos logit contributions | |
Crit | +0.765 |
Nit | +0.708 |
Crit | +0.679 |
Poverty | +0.677 |
Adapt | +0.618 |
Neg logit contributions | |
Kislyak | -0.735 |
================================================================= | -0.722 |
IMAGES | -0.700 |
advertising | -0.699 |
zinski | -0.686 |
Token | > |
Token ID | 1875 |
Feature activation | +1.31e-01 |
Pos logit contributions | |
Crit | +0.199 |
Nit | +0.186 |
Poverty | +0.179 |
Crit | +0.177 |
corruption | +0.167 |
Neg logit contributions | |
Kislyak | -0.127 |
================================================================= | -0.127 |
IMAGES | -0.121 |
advertising | -0.117 |
zinski | -0.115 |
Token | { |
Token ID | 1391 |
Feature activation | -4.88e-01 |
Pos logit contributions | |
Kislyak | +0.366 |
================================================================= | +0.361 |
IMAGES | +0.351 |
advertising | +0.345 |
zinski | +0.341 |
Neg logit contributions | |
Crit | -0.322 |
Nit | -0.294 |
Poverty | -0.281 |
Crit | -0.278 |
Adapt | -0.255 |
Token | pub |
Token ID | 2240 |
Feature activation | -6.34e-01 |
Pos logit contributions | |
Crit | +1.274 |
Nit | +1.176 |
Poverty | +1.117 |
Crit | +1.109 |
Adapt | +1.025 |
Neg logit contributions | |
Kislyak | -1.277 |
================================================================= | -1.260 |
IMAGES | -1.220 |
advertising | -1.195 |
zinski | -1.182 |
Token | fn |
Token ID | 24714 |
Feature activation | -1.31e+00 |
Pos logit contributions | |
Crit | +1.585 |
Nit | +1.461 |
Crit | +1.375 |
Poverty | +1.374 |
Adapt | +1.268 |
Neg logit contributions | |
Kislyak | -1.727 |
================================================================= | -1.688 |
IMAGES | -1.636 |
advertising | -1.630 |
zinski | -1.600 |
Token | load |
Token ID | 3440 |
Feature activation | +3.47e-01 |
Pos logit contributions | |
Crit | +3.239 |
Nit | +3.004 |
Crit | +2.911 |
Poverty | +2.862 |
crit | +2.718 |
Neg logit contributions | |
Kislyak | -3.480 |
advertising | -3.323 |
================================================================= | -3.320 |
Accessory | -3.246 |
zinski | -3.209 |
Token | < |
Token ID | 1279 |
Feature activation | -2.60e-01 |
Pos logit contributions | |
================================================================= | +0.859 |
Kislyak | +0.856 |
IMAGES | +0.824 |
advertising | +0.798 |
zinski | +0.792 |
Neg logit contributions | |
Crit | -0.950 |
Nit | -0.878 |
Poverty | -0.844 |
Crit | -0.832 |
Lumpur | -0.775 |
Token | ' |
Token ID | 705 |
Feature activation | -3.82e-03 |
Pos logit contributions | |
Crit | +0.709 |
Nit | +0.656 |
Crit | +0.632 |
Poverty | +0.630 |
Ps | +0.581 |
Neg logit contributions | |
Kislyak | -0.648 |
================================================================= | -0.643 |
advertising | -0.613 |
IMAGES | -0.607 |
zinski | -0.601 |
Token | a |
Token ID | 64 |
Feature activation | +1.72e-01 |
Pos logit contributions | |
Crit | +0.011 |
Nit | +0.010 |
Poverty | +0.010 |
Crit | +0.010 |
corruption | +0.009 |
Neg logit contributions | |
Kislyak | -0.009 |
================================================================= | -0.009 |
IMAGES | -0.009 |
advertising | -0.008 |
zinski | -0.008 |
Token | > |
Token ID | 1875 |
Feature activation | +1.89e-01 |
Pos logit contributions | |
Kislyak | +0.396 |
================================================================= | +0.393 |
IMAGES | +0.376 |
advertising | +0.368 |
zinski | +0.364 |
Neg logit contributions | |
Crit | -0.506 |
Nit | -0.471 |
Poverty | -0.450 |
Crit | -0.449 |
corruption | -0.416 |
Token | let |
Token ID | 1309 |
Feature activation | -1.05e+00 |
Pos logit contributions | |
Crit | +1.140 |
Nit | +1.029 |
Crit | +0.979 |
Poverty | +0.976 |
Ps | +0.881 |
Neg logit contributions | |
Kislyak | -1.460 |
================================================================= | -1.433 |
IMAGES | -1.398 |
advertising | -1.384 |
zinski | -1.373 |
Token | prev |
Token ID | 8654 |
Feature activation | +2.60e-01 |
Pos logit contributions | |
Crit | +1.958 |
Nit | +1.717 |
Poverty | +1.638 |
Crit | +1.638 |
crit | +1.497 |
Neg logit contributions | |
Kislyak | -3.491 |
================================================================= | -3.413 |
advertising | -3.361 |
IMAGES | -3.329 |
zinski | -3.313 |
Token | : |
Token ID | 1058 |
Feature activation | -4.65e-01 |
Pos logit contributions | |
Kislyak | +0.662 |
================================================================= | +0.661 |
IMAGES | +0.636 |
advertising | +0.619 |
zinski | +0.612 |
Neg logit contributions | |
Crit | -0.695 |
Nit | -0.642 |
Poverty | -0.614 |
Crit | -0.605 |
Lumpur | -0.563 |
Token | Board |
Token ID | 5926 |
Feature activation | +1.41e-01 |
Pos logit contributions | |
Crit | +1.348 |
Nit | +1.240 |
Crit | +1.208 |
Poverty | +1.205 |
Ps | +1.114 |
Neg logit contributions | |
Kislyak | -1.080 |
================================================================= | -1.064 |
IMAGES | -1.022 |
advertising | -1.016 |
zinski | -1.008 |
Token | = |
Token ID | 796 |
Feature activation | -3.31e-01 |
Pos logit contributions | |
Kislyak | +0.356 |
================================================================= | +0.354 |
IMAGES | +0.342 |
advertising | +0.332 |
zinski | +0.328 |
Neg logit contributions | |
Crit | -0.383 |
Nit | -0.353 |
Poverty | -0.337 |
Crit | -0.333 |
Adapt | -0.310 |
Token | self |
Token ID | 2116 |
Feature activation | -1.28e+00 |
Pos logit contributions | |
Crit | +0.696 |
Nit | +0.619 |
Crit | +0.591 |
Poverty | +0.590 |
Adapt | +0.526 |
Neg logit contributions | |
Kislyak | -1.035 |
================================================================= | -1.023 |
IMAGES | -0.995 |
advertising | -0.988 |
zinski | -0.978 |
Token | . |
Token ID | 764 |
Feature activation | -2.51e-01 |
Pos logit contributions | |
Crit | +3.504 |
Nit | +3.209 |
Poverty | +3.057 |
Crit | +3.052 |
corruption | +2.838 |
Neg logit contributions | |
Kislyak | -3.175 |
================================================================= | -3.113 |
IMAGES | -3.029 |
advertising | -2.950 |
zinski | -2.927 |
Token | prev |
Token ID | 47050 |
Feature activation | +2.64e-01 |
Pos logit contributions | |
Crit | +0.654 |
Nit | +0.596 |
Poverty | +0.566 |
Crit | +0.564 |
corruption | +0.529 |
Neg logit contributions | |
Kislyak | -0.667 |
================================================================= | -0.661 |
IMAGES | -0.639 |
advertising | -0.616 |
zinski | -0.616 |
Token | _ |
Token ID | 62 |
Feature activation | -4.32e-01 |
Pos logit contributions | |
================================================================= | +0.651 |
Kislyak | +0.646 |
IMAGES | +0.619 |
advertising | +0.605 |
Accessory | +0.592 |
Neg logit contributions | |
Crit | -0.726 |
Nit | -0.675 |
Poverty | -0.647 |
Crit | -0.635 |
Adapt | -0.595 |
Token | state |
Token ID | 5219 |
Feature activation | +9.52e-02 |
Pos logit contributions | |
Crit | +1.160 |
Nit | +1.060 |
Poverty | +1.005 |
Crit | +1.000 |
corruption | +0.956 |
Neg logit contributions | |
Kislyak | -1.103 |
================================================================= | -1.093 |
IMAGES | -1.064 |
Accessory | -1.017 |
zinski | -1.012 |
Token | . |
Token ID | 764 |
Feature activation | -2.38e-01 |
Pos logit contributions | |
Kislyak | +0.225 |
================================================================= | +0.225 |
IMAGES | +0.216 |
advertising | +0.209 |
zinski | +0.206 |
Neg logit contributions | |
Crit | -0.272 |
Nit | -0.252 |
Poverty | -0.242 |
Crit | -0.239 |
corruption | -0.223 |
Token | copy |
Token ID | 4866 |
Feature activation | -1.08e-01 |
Pos logit contributions | |
Crit | +3.343 |
Nit | +2.992 |
Crit | +2.976 |
Poverty | +2.898 |
crit | +2.756 |
Neg logit contributions | |
Kislyak | -3.869 |
================================================================= | -3.690 |
advertising | -3.663 |
zinski | -3.623 |
Accessory | -3.617 |
Token | _ |
Token ID | 62 |
Feature activation | -3.87e-01 |
Pos logit contributions | |
Crit | +0.295 |
Nit | +0.272 |
Poverty | +0.260 |
Crit | +0.257 |
Adapt | +0.240 |
Neg logit contributions | |
Kislyak | -0.268 |
================================================================= | -0.266 |
IMAGES | -0.257 |
advertising | -0.249 |
zinski | -0.247 |
Token | board |
Token ID | 3526 |
Feature activation | +2.51e-01 |
Pos logit contributions | |
Crit | +0.979 |
Nit | +0.890 |
Poverty | +0.836 |
Crit | +0.834 |
corruption | +0.797 |
Neg logit contributions | |
Kislyak | -1.050 |
================================================================= | -1.034 |
IMAGES | -1.009 |
zinski | -0.969 |
Accessory | -0.966 |
Token | ( |
Token ID | 357 |
Feature activation | -2.10e-01 |
Pos logit contributions | |
Kislyak | +0.653 |
================================================================= | +0.646 |
IMAGES | +0.626 |
advertising | +0.609 |
zinski | +0.604 |
Neg logit contributions | |
Crit | -0.660 |
Nit | -0.606 |
Poverty | -0.578 |
Crit | -0.571 |
corruption | -0.531 |
Token | & |
Token ID | 1222 |
Feature activation | +1.23e-01 |
Pos logit contributions | |
Crit | +0.544 |
Nit | +0.498 |
Poverty | +0.475 |
Crit | +0.470 |
corruption | +0.437 |
Neg logit contributions | |
Kislyak | -0.556 |
================================================================= | -0.552 |
IMAGES | -0.533 |
advertising | -0.520 |
zinski | -0.515 |
Token | self |
Token ID | 2116 |
Feature activation | -1.24e+00 |
Pos logit contributions | |
Kislyak | +0.398 |
================================================================= | +0.397 |
IMAGES | +0.385 |
advertising | +0.377 |
zinski | +0.374 |
Neg logit contributions | |
Crit | -0.245 |
Nit | -0.220 |
Poverty | -0.206 |
Crit | -0.202 |
corruption | -0.182 |
Token | ) |
Token ID | 1267 |
Feature activation | +1.20e-02 |
Pos logit contributions | |
Crit | +3.706 |
Nit | +3.391 |
Poverty | +3.264 |
Crit | +3.259 |
crit | +3.057 |
Neg logit contributions | |
Kislyak | -2.744 |
================================================================= | -2.630 |
zinski | -2.529 |
IMAGES | -2.524 |
advertising | -2.502 |
Token | -> |
Token ID | 4613 |
Feature activation | +1.89e-01 |
Pos logit contributions | |
Kislyak | +0.025 |
================================================================= | +0.025 |
IMAGES | +0.023 |
advertising | +0.023 |
zinski | +0.022 |
Neg logit contributions | |
Crit | -0.038 |
Nit | -0.036 |
Poverty | -0.034 |
Crit | -0.034 |
Adapt | -0.032 |
Token | Board |
Token ID | 5926 |
Feature activation | +4.54e-01 |
Pos logit contributions | |
Kislyak | +0.495 |
================================================================= | +0.489 |
IMAGES | +0.472 |
advertising | +0.466 |
zinski | +0.462 |
Neg logit contributions | |
Crit | -0.492 |
Nit | -0.451 |
Poverty | -0.434 |
Crit | -0.432 |
Adapt | -0.396 |
Token | { |
Token ID | 1391 |
Feature activation | -4.88e-01 |
Pos logit contributions | |
Kislyak | +1.183 |
================================================================= | +1.179 |
IMAGES | +1.137 |
advertising | +1.105 |
zinski | +1.093 |
Neg logit contributions | |
Crit | -1.190 |
Nit | -1.095 |
Poverty | -1.046 |
Crit | -1.031 |
corruption | -0.958 |
Token | Board |
Token ID | 5926 |
Feature activation | +2.80e-01 |
Pos logit contributions | |
Crit | +1.178 |
Nit | +1.070 |
Poverty | +1.016 |
Crit | +1.014 |
Adapt | +0.925 |
Neg logit contributions | |
Kislyak | -1.379 |
================================================================= | -1.357 |
IMAGES | -1.320 |
advertising | -1.300 |
zinski | -1.289 |
Token | Guard |
Token ID | 4932 |
Feature activation | +1.40e-01 |
Pos logit contributions | |
Crit | +1.684 |
Nit | +1.545 |
Poverty | +1.517 |
Crit | +1.509 |
Adapt | +1.365 |
Neg logit contributions | |
Kislyak | -1.663 |
================================================================= | -1.635 |
advertising | -1.595 |
IMAGES | -1.593 |
zinski | -1.572 |
Token | { |
Token ID | 1391 |
Feature activation | -6.47e-01 |
Pos logit contributions | |
Kislyak | +0.355 |
================================================================= | +0.353 |
IMAGES | +0.342 |
advertising | +0.331 |
zinski | +0.329 |
Neg logit contributions | |
Crit | -0.378 |
Nit | -0.349 |
Poverty | -0.332 |
Crit | -0.328 |
Adapt | -0.307 |
Token | ... |
Token ID | 2644 |
Feature activation | -4.78e-01 |
Pos logit contributions | |
Crit | +1.577 |
Nit | +1.443 |
Crit | +1.373 |
Poverty | +1.369 |
Adapt | +1.248 |
Neg logit contributions | |
Kislyak | -1.790 |
================================================================= | -1.761 |
IMAGES | -1.717 |
advertising | -1.690 |
zinski | -1.677 |
Token | } |
Token ID | 1782 |
Feature activation | +1.99e-01 |
Pos logit contributions | |
Crit | +1.041 |
Nit | +0.941 |
Crit | +0.885 |
Poverty | +0.882 |
corruption | +0.804 |
Neg logit contributions | |
Kislyak | -1.452 |
================================================================= | -1.434 |
IMAGES | -1.394 |
advertising | -1.373 |
zinski | -1.370 |
Token | pub |
Token ID | 2240 |
Feature activation | -5.20e-01 |
Pos logit contributions | |
Kislyak | +0.544 |
================================================================= | +0.542 |
IMAGES | +0.524 |
advertising | +0.512 |
zinski | +0.507 |
Neg logit contributions | |
Crit | -0.494 |
Nit | -0.452 |
Poverty | -0.430 |
Crit | -0.425 |
corruption | -0.392 |
Token | fn |
Token ID | 24714 |
Feature activation | -1.23e+00 |
Pos logit contributions | |
Crit | +1.248 |
Nit | +1.134 |
Poverty | +1.079 |
Crit | +1.071 |
Adapt | +0.985 |
Neg logit contributions | |
Kislyak | -1.471 |
================================================================= | -1.454 |
IMAGES | -1.412 |
advertising | -1.393 |
zinski | -1.374 |
Token | pin |
Token ID | 6757 |
Feature activation | +4.34e-01 |
Pos logit contributions | |
Crit | +3.168 |
Nit | +2.912 |
Crit | +2.857 |
Poverty | +2.839 |
crit | +2.646 |
Neg logit contributions | |
Kislyak | -3.128 |
================================================================= | -3.047 |
advertising | -2.982 |
IMAGES | -2.958 |
zinski | -2.935 |
Token | () |
Token ID | 7499 |
Feature activation | -6.19e-02 |
Pos logit contributions | |
Kislyak | +1.081 |
================================================================= | +1.078 |
IMAGES | +1.037 |
advertising | +1.007 |
zinski | +0.996 |
Neg logit contributions | |
Crit | -1.189 |
Nit | -1.098 |
Poverty | -1.051 |
Crit | -1.036 |
corruption | -0.966 |
Token | -> |
Token ID | 4613 |
Feature activation | +3.44e-01 |
Pos logit contributions | |
Crit | +0.173 |
Nit | +0.160 |
Poverty | +0.153 |
Crit | +0.153 |
Ps | +0.142 |
Neg logit contributions | |
Kislyak | -0.150 |
================================================================= | -0.148 |
IMAGES | -0.143 |
advertising | -0.140 |
zinski | -0.139 |
Token | Guard |
Token ID | 4932 |
Feature activation | +8.17e-02 |
Pos logit contributions | |
Kislyak | +0.902 |
================================================================= | +0.896 |
IMAGES | +0.864 |
advertising | +0.849 |
zinski | +0.837 |
Neg logit contributions | |
Crit | -0.897 |
Nit | -0.827 |
Poverty | -0.793 |
Crit | -0.786 |
Adapt | -0.723 |
Token | ; |
Token ID | 2162 |
Feature activation | -2.48e-01 |
Pos logit contributions | |
Kislyak | +0.209 |
================================================================= | +0.207 |
IMAGES | +0.201 |
advertising | +0.195 |
zinski | +0.194 |
Neg logit contributions | |
Crit | -0.221 |
Nit | -0.203 |
Poverty | -0.192 |
Crit | -0.191 |
Adapt | -0.178 |
Token | Red |
Token ID | 2297 |
Feature activation | -1.73e-01 |
Pos logit contributions | |
Crit | +0.462 |
Nit | +0.428 |
Poverty | +0.420 |
Crit | +0.418 |
Adapt | +0.383 |
Neg logit contributions | |
Kislyak | -0.333 |
================================================================= | -0.331 |
IMAGES | -0.324 |
advertising | -0.314 |
Accessory | -0.312 |
Token | Christmas |
Token ID | 6786 |
Feature activation | +2.21e-01 |
Pos logit contributions | |
Crit | +0.425 |
Nit | +0.387 |
Crit | +0.372 |
Poverty | +0.372 |
Adapt | +0.337 |
Neg logit contributions | |
Kislyak | -0.482 |
================================================================= | -0.476 |
IMAGES | -0.462 |
Accessory | -0.451 |
advertising | -0.450 |
Token | album |
Token ID | 5062 |
Feature activation | -9.25e-02 |
Pos logit contributions | |
Kislyak | +0.586 |
================================================================= | +0.582 |
IMAGES | +0.565 |
advertising | +0.550 |
Accessory | +0.545 |
Neg logit contributions | |
Crit | -0.569 |
Nit | -0.520 |
Poverty | -0.500 |
Crit | -0.497 |
Adapt | -0.457 |
Token | .[ |
Token ID | 3693 |
Feature activation | -4.73e-01 |
Pos logit contributions | |
Crit | +0.277 |
Nit | +0.257 |
Poverty | +0.248 |
Crit | +0.247 |
Ps | +0.230 |
Neg logit contributions | |
Kislyak | -0.207 |
================================================================= | -0.205 |
IMAGES | -0.197 |
advertising | -0.191 |
Accessory | -0.190 |
Token | c |
Token ID | 66 |
Feature activation | -2.53e-01 |
Pos logit contributions | |
Crit | +1.064 |
Nit | +0.966 |
Poverty | +0.892 |
Crit | +0.891 |
Adapt | +0.824 |
Neg logit contributions | |
Kislyak | -1.420 |
================================================================= | -1.406 |
IMAGES | -1.368 |
zinski | -1.324 |
Accessory | -1.319 |
Token | itation |
Token ID | 3780 |
Feature activation | -1.23e+00 |
Pos logit contributions | |
Crit | +0.493 |
Nit | +0.439 |
Poverty | +0.410 |
Crit | +0.405 |
Adapt | +0.363 |
Neg logit contributions | |
Kislyak | -0.829 |
================================================================= | -0.825 |
IMAGES | -0.803 |
Accessory | -0.785 |
advertising | -0.779 |
Token | needed |
Token ID | 2622 |
Feature activation | -3.93e-01 |
Pos logit contributions | |
Crit | +2.540 |
Nit | +2.278 |
Poverty | +2.160 |
Crit | +2.138 |
Adapt | +1.929 |
Neg logit contributions | |
Kislyak | -3.895 |
================================================================= | -3.860 |
IMAGES | -3.757 |
Accessory | -3.670 |
zinski | -3.661 |
Token | ] |
Token ID | 60 |
Feature activation | +2.60e-01 |
Pos logit contributions | |
Crit | +1.167 |
Nit | +1.084 |
Poverty | +1.041 |
Crit | +1.033 |
Adapt | +0.967 |
Neg logit contributions | |
Kislyak | -0.886 |
================================================================= | -0.876 |
IMAGES | -0.844 |
advertising | -0.814 |
zinski | -0.813 |
Token |
|
Token ID | 198 |
Feature activation | -1.75e-01 |
Pos logit contributions | |
Kislyak | +0.719 |
================================================================= | +0.717 |
IMAGES | +0.693 |
advertising | +0.674 |
zinski | +0.668 |
Neg logit contributions | |
Crit | -0.640 |
Nit | -0.586 |
Poverty | -0.557 |
Crit | -0.549 |
Adapt | -0.506 |
Token |
|
Token ID | 198 |
Feature activation | -1.34e-01 |
Pos logit contributions | |
Crit | +0.418 |
Nit | +0.381 |
Poverty | +0.361 |
Crit | +0.357 |
Adapt | +0.327 |
Neg logit contributions | |
Kislyak | -0.499 |
================================================================= | -0.497 |
IMAGES | -0.482 |
advertising | -0.467 |
zinski | -0.464 |
Token | As |
Token ID | 1722 |
Feature activation | +1.60e-01 |
Pos logit contributions | |
Crit | +0.302 |
Nit | +0.272 |
Poverty | +0.252 |
Crit | +0.249 |
Adapt | +0.233 |
Neg logit contributions | |
Kislyak | -0.406 |
================================================================= | -0.404 |
IMAGES | -0.393 |
zinski | -0.379 |
advertising | -0.377 |
Token | ) |
Token ID | 8 |
Feature activation | +3.25e-02 |
Pos logit contributions | |
Kislyak | +0.275 |
================================================================= | +0.274 |
IMAGES | +0.258 |
advertising | +0.245 |
zinski | +0.242 |
Neg logit contributions | |
Crit | -0.585 |
Nit | -0.552 |
Poverty | -0.533 |
Crit | -0.528 |
corruption | -0.501 |
Token | { |
Token ID | 1391 |
Feature activation | -8.62e-01 |
Pos logit contributions | |
================================================================= | +0.087 |
Kislyak | +0.087 |
IMAGES | +0.083 |
advertising | +0.080 |
Accessory | +0.080 |
Neg logit contributions | |
Crit | -0.083 |
Nit | -0.076 |
Poverty | -0.073 |
Crit | -0.071 |
corruption | -0.067 |
Token | // |
Token ID | 3373 |
Feature activation | +5.18e-01 |
Pos logit contributions | |
Crit | +1.888 |
Nit | +1.700 |
Poverty | +1.603 |
Crit | +1.601 |
Adapt | +1.441 |
Neg logit contributions | |
Kislyak | -2.601 |
================================================================= | -2.568 |
IMAGES | -2.511 |
advertising | -2.487 |
Accessory | -2.453 |
Token | Time |
Token ID | 3862 |
Feature activation | +1.66e-01 |
Pos logit contributions | |
Kislyak | +1.363 |
================================================================= | +1.362 |
IMAGES | +1.310 |
advertising | +1.272 |
zinski | +1.261 |
Neg logit contributions | |
Crit | -1.343 |
Nit | -1.236 |
Poverty | -1.180 |
Crit | -1.160 |
corruption | -1.079 |
Token | component |
Token ID | 7515 |
Feature activation | +9.79e-02 |
Pos logit contributions | |
Kislyak | +0.425 |
================================================================= | +0.422 |
IMAGES | +0.409 |
advertising | +0.398 |
zinski | +0.394 |
Neg logit contributions | |
Crit | -0.444 |
Nit | -0.409 |
Poverty | -0.391 |
Crit | -0.386 |
Adapt | -0.359 |
Token | const |
Token ID | 1500 |
Feature activation | -1.21e+00 |
Pos logit contributions | |
Kislyak | +0.255 |
================================================================= | +0.252 |
IMAGES | +0.245 |
advertising | +0.240 |
zinski | +0.237 |
Neg logit contributions | |
Crit | -0.257 |
Nit | -0.237 |
Poverty | -0.225 |
Crit | -0.224 |
Adapt | -0.208 |
Token | float |
Token ID | 12178 |
Feature activation | -8.81e-01 |
Pos logit contributions | |
Crit | +2.843 |
Nit | +2.599 |
Crit | +2.481 |
Poverty | +2.463 |
Ps | +2.268 |
Neg logit contributions | |
Kislyak | -3.381 |
advertising | -3.299 |
================================================================= | -3.294 |
IMAGES | -3.242 |
Accessory | -3.201 |
Token | TA |
Token ID | 21664 |
Feature activation | +5.03e-01 |
Pos logit contributions | |
Crit | +2.290 |
Nit | +2.090 |
Poverty | +2.004 |
Crit | +1.994 |
Adapt | +1.842 |
Neg logit contributions | |
Kislyak | -2.289 |
================================================================= | -2.253 |
IMAGES | -2.216 |
advertising | -2.205 |
Accessory | -2.141 |
Token | U |
Token ID | 52 |
Feature activation | -1.79e-01 |
Pos logit contributions | |
Kislyak | +1.259 |
================================================================= | +1.253 |
IMAGES | +1.207 |
advertising | +1.172 |
zinski | +1.159 |
Neg logit contributions | |
Crit | -1.373 |
Nit | -1.268 |
Poverty | -1.212 |
Crit | -1.195 |
Adapt | -1.114 |
Token | = |
Token ID | 796 |
Feature activation | -2.50e-01 |
Pos logit contributions | |
Crit | +0.485 |
Nit | +0.446 |
Poverty | +0.424 |
Crit | +0.420 |
Adapt | +0.395 |
Neg logit contributions | |
Kislyak | -0.455 |
================================================================= | -0.445 |
IMAGES | -0.432 |
advertising | -0.424 |
zinski | -0.417 |
Token | 6 |
Token ID | 718 |
Feature activation | +3.62e-01 |
Pos logit contributions | |
Crit | +0.672 |
Nit | +0.615 |
Poverty | +0.591 |
Crit | +0.588 |
Adapt | +0.542 |
Neg logit contributions | |
Kislyak | -0.635 |
================================================================= | -0.630 |
IMAGES | -0.611 |
advertising | -0.606 |
Accessory | -0.592 |
Token | . |
Token ID | 764 |
Feature activation | -5.40e-02 |
Pos logit contributions | |
Crit | +3.999 |
Nit | +3.641 |
Crit | +3.454 |
Poverty | +3.436 |
corruption | +3.210 |
Neg logit contributions | |
Kislyak | -3.821 |
================================================================= | -3.747 |
IMAGES | -3.675 |
advertising | -3.595 |
zinski | -3.577 |
Token | clone |
Token ID | 21018 |
Feature activation | -2.22e-02 |
Pos logit contributions | |
Crit | +0.142 |
Nit | +0.130 |
Poverty | +0.124 |
Crit | +0.123 |
corruption | +0.114 |
Neg logit contributions | |
Kislyak | -0.141 |
================================================================= | -0.140 |
IMAGES | -0.135 |
advertising | -0.131 |
zinski | -0.130 |
Token | (), |
Token ID | 29994 |
Feature activation | -4.62e-02 |
Pos logit contributions | |
Crit | +0.062 |
Nit | +0.058 |
Poverty | +0.055 |
Crit | +0.055 |
corruption | +0.051 |
Neg logit contributions | |
Kislyak | -0.054 |
================================================================= | -0.053 |
IMAGES | -0.051 |
advertising | -0.050 |
zinski | -0.049 |
Token | pieces |
Token ID | 5207 |
Feature activation | +3.41e-01 |
Pos logit contributions | |
Crit | +0.110 |
Nit | +0.101 |
Poverty | +0.096 |
Crit | +0.094 |
corruption | +0.087 |
Neg logit contributions | |
Kislyak | -0.131 |
================================================================= | -0.130 |
IMAGES | -0.126 |
advertising | -0.123 |
zinski | -0.122 |
Token | : |
Token ID | 1058 |
Feature activation | -2.36e-01 |
Pos logit contributions | |
================================================================= | +0.633 |
Kislyak | +0.623 |
IMAGES | +0.594 |
advertising | +0.573 |
Accessory | +0.563 |
Neg logit contributions | |
Crit | -1.141 |
Nit | -1.076 |
Poverty | -1.043 |
Crit | -1.026 |
Lumpur | -0.976 |
Token | self |
Token ID | 2116 |
Feature activation | -1.19e+00 |
Pos logit contributions | |
Crit | +0.487 |
Nit | +0.434 |
Poverty | +0.412 |
Crit | +0.410 |
Ps | +0.367 |
Neg logit contributions | |
Kislyak | -0.747 |
================================================================= | -0.740 |
IMAGES | -0.718 |
advertising | -0.713 |
zinski | -0.705 |
Token | . |
Token ID | 764 |
Feature activation | -2.70e-02 |
Pos logit contributions | |
Crit | +3.241 |
Nit | +2.965 |
Poverty | +2.831 |
Crit | +2.822 |
corruption | +2.632 |
Neg logit contributions | |
Kislyak | -2.987 |
================================================================= | -2.936 |
IMAGES | -2.858 |
advertising | -2.790 |
zinski | -2.762 |
Token | pieces |
Token ID | 34154 |
Feature activation | +1.67e-01 |
Pos logit contributions | |
Crit | +0.076 |
Nit | +0.070 |
Poverty | +0.066 |
Crit | +0.066 |
corruption | +0.062 |
Neg logit contributions | |
Kislyak | -0.066 |
================================================================= | -0.066 |
IMAGES | -0.063 |
advertising | -0.061 |
zinski | -0.061 |
Token | . |
Token ID | 764 |
Feature activation | -4.46e-02 |
Pos logit contributions | |
Kislyak | +0.403 |
================================================================= | +0.401 |
IMAGES | +0.386 |
advertising | +0.374 |
zinski | +0.370 |
Neg logit contributions | |
Crit | -0.469 |
Nit | -0.434 |
Poverty | -0.415 |
Crit | -0.410 |
corruption | -0.383 |
Token | clone |
Token ID | 21018 |
Feature activation | +1.80e-01 |
Pos logit contributions | |
Crit | +0.129 |
Nit | +0.119 |
Poverty | +0.114 |
Crit | +0.114 |
corruption | +0.107 |
Neg logit contributions | |
Kislyak | -0.104 |
================================================================= | -0.103 |
IMAGES | -0.100 |
advertising | -0.097 |
zinski | -0.096 |
Token | (), |
Token ID | 29994 |
Feature activation | +4.95e-02 |
Pos logit contributions | |
Kislyak | +0.382 |
================================================================= | +0.379 |
IMAGES | +0.363 |
advertising | +0.350 |
zinski | +0.346 |
Neg logit contributions | |
Crit | -0.564 |
Nit | -0.525 |
Poverty | -0.504 |
Crit | -0.499 |
corruption | -0.471 |
Token | wrap |
Token ID | 37150 |
Feature activation | -3.01e-01 |
Pos logit contributions | |
Crit | +0.166 |
Nit | +0.153 |
Poverty | +0.146 |
Crit | +0.145 |
corruption | +0.139 |
Neg logit contributions | |
Kislyak | -0.122 |
================================================================= | -0.120 |
IMAGES | -0.117 |
Accessory | -0.111 |
advertising | -0.111 |
Token | ()) |
Token ID | 32865 |
Feature activation | -2.88e-02 |
Pos logit contributions | |
Crit | +0.949 |
Nit | +0.878 |
Poverty | +0.842 |
Crit | +0.838 |
corruption | +0.793 |
Neg logit contributions | |
Kislyak | -0.633 |
================================================================= | -0.622 |
IMAGES | -0.597 |
advertising | -0.571 |
Accessory | -0.570 |
Token | } |
Token ID | 1782 |
Feature activation | -7.26e-02 |
Pos logit contributions | |
Crit | +0.059 |
Nit | +0.053 |
Poverty | +0.050 |
Crit | +0.049 |
corruption | +0.044 |
Neg logit contributions | |
Kislyak | -0.092 |
================================================================= | -0.091 |
IMAGES | -0.089 |
advertising | -0.087 |
zinski | -0.086 |
Token | } |
Token ID | 1782 |
Feature activation | -2.27e-01 |
Pos logit contributions | |
Crit | +0.117 |
Nit | +0.102 |
Poverty | +0.094 |
Crit | +0.092 |
corruption | +0.080 |
Neg logit contributions | |
Kislyak | -0.262 |
================================================================= | -0.262 |
IMAGES | -0.255 |
advertising | -0.250 |
zinski | -0.248 |
Token | pub |
Token ID | 2240 |
Feature activation | -4.89e-01 |
Pos logit contributions | |
Crit | +0.487 |
Nit | +0.440 |
Poverty | +0.415 |
Crit | +0.407 |
corruption | +0.370 |
Neg logit contributions | |
Kislyak | -0.698 |
================================================================= | -0.697 |
IMAGES | -0.675 |
advertising | -0.658 |
zinski | -0.652 |
Token | fn |
Token ID | 24714 |
Feature activation | -1.18e+00 |
Pos logit contributions | |
Crit | +1.249 |
Nit | +1.143 |
Poverty | +1.088 |
Crit | +1.079 |
corruption | +0.998 |
Neg logit contributions | |
Kislyak | -1.307 |
================================================================= | -1.289 |
IMAGES | -1.250 |
advertising | -1.229 |
zinski | -1.214 |
Token | apply |
Token ID | 4174 |
Feature activation | +1.39e-01 |
Pos logit contributions | |
Crit | +2.751 |
Nit | +2.479 |
Crit | +2.434 |
Poverty | +2.405 |
crit | +2.286 |
Neg logit contributions | |
Kislyak | -3.330 |
================================================================= | -3.208 |
advertising | -3.180 |
IMAGES | -3.142 |
Accessory | -3.127 |
Token | _ |
Token ID | 62 |
Feature activation | -2.18e-01 |
Pos logit contributions | |
Kislyak | +0.368 |
================================================================= | +0.364 |
IMAGES | +0.353 |
advertising | +0.342 |
zinski | +0.339 |
Neg logit contributions | |
Crit | -0.362 |
Nit | -0.332 |
Poverty | -0.317 |
Crit | -0.313 |
corruption | -0.292 |
Token | move |
Token ID | 21084 |
Feature activation | +2.29e-02 |
Pos logit contributions | |
Crit | +0.472 |
Nit | +0.425 |
Poverty | +0.398 |
Crit | +0.394 |
corruption | +0.371 |
Neg logit contributions | |
Kislyak | -0.668 |
================================================================= | -0.665 |
IMAGES | -0.646 |
Accessory | -0.623 |
advertising | -0.623 |
Token | ( |
Token ID | 357 |
Feature activation | -8.24e-02 |
Pos logit contributions | |
Kislyak | +0.060 |
================================================================= | +0.060 |
IMAGES | +0.058 |
advertising | +0.056 |
zinski | +0.056 |
Neg logit contributions | |
Crit | -0.060 |
Nit | -0.055 |
Poverty | -0.052 |
Crit | -0.052 |
corruption | -0.048 |
Token | & |
Token ID | 1222 |
Feature activation | +1.70e-01 |
Pos logit contributions | |
Crit | +0.215 |
Nit | +0.197 |
Poverty | +0.188 |
Crit | +0.187 |
corruption | +0.173 |
Neg logit contributions | |
Kislyak | -0.217 |
================================================================= | -0.215 |
IMAGES | -0.207 |
advertising | -0.203 |
zinski | -0.201 |
Token | backs |
Token ID | 12983 |
Feature activation | +4.33e-02 |
Pos logit contributions | |
Crit | +0.357 |
Nit | +0.331 |
Crit | +0.317 |
Poverty | +0.316 |
crit | +0.296 |
Neg logit contributions | |
Kislyak | -0.306 |
================================================================= | -0.304 |
IMAGES | -0.292 |
advertising | -0.286 |
zinski | -0.284 |
Token | . |
Token ID | 13 |
Feature activation | +5.32e-02 |
Pos logit contributions | |
Kislyak | +0.095 |
================================================================= | +0.095 |
IMAGES | +0.091 |
advertising | +0.088 |
zinski | +0.087 |
Neg logit contributions | |
Crit | -0.132 |
Nit | -0.122 |
Poverty | -0.117 |
Crit | -0.116 |
corruption | -0.109 |
Token |
|
Token ID | 198 |
Feature activation | +7.84e-02 |
Pos logit contributions | |
Kislyak | +0.150 |
================================================================= | +0.149 |
IMAGES | +0.144 |
advertising | +0.141 |
zinski | +0.139 |
Neg logit contributions | |
Crit | -0.128 |
Nit | -0.117 |
Poverty | -0.112 |
Crit | -0.110 |
Adapt | -0.101 |
Token |
|
Token ID | 198 |
Feature activation | +1.98e-02 |
Pos logit contributions | |
Kislyak | +0.217 |
================================================================= | +0.216 |
IMAGES | +0.208 |
advertising | +0.203 |
zinski | +0.201 |
Neg logit contributions | |
Crit | -0.193 |
Nit | -0.177 |
Poverty | -0.168 |
Crit | -0.165 |
corruption | -0.153 |
Token | On |
Token ID | 2202 |
Feature activation | -4.14e-01 |
Pos logit contributions | |
Kislyak | +0.060 |
================================================================= | +0.060 |
IMAGES | +0.058 |
advertising | +0.056 |
zinski | +0.056 |
Neg logit contributions | |
Crit | -0.044 |
Nit | -0.040 |
Poverty | -0.037 |
Crit | -0.037 |
Adapt | -0.034 |
Token | top |
Token ID | 1353 |
Feature activation | +6.33e-01 |
Pos logit contributions | |
Crit | +1.060 |
Nit | +0.969 |
Poverty | +0.930 |
Crit | +0.928 |
Adapt | +0.850 |
Neg logit contributions | |
Kislyak | -1.094 |
================================================================= | -1.088 |
IMAGES | -1.061 |
advertising | -1.029 |
zinski | -1.019 |
Token | of |
Token ID | 286 |
Feature activation | +9.48e-02 |
Pos logit contributions | |
Kislyak | +1.534 |
================================================================= | +1.529 |
IMAGES | +1.467 |
advertising | +1.428 |
zinski | +1.408 |
Neg logit contributions | |
Crit | -1.771 |
Nit | -1.643 |
Poverty | -1.570 |
Crit | -1.543 |
corruption | -1.448 |
Token | his |
Token ID | 465 |
Feature activation | +3.45e-02 |
Pos logit contributions | |
Kislyak | +0.255 |
================================================================= | +0.254 |
IMAGES | +0.247 |
advertising | +0.239 |
zinski | +0.236 |
Neg logit contributions | |
Crit | -0.241 |
Nit | -0.221 |
Poverty | -0.210 |
Crit | -0.209 |
Adapt | -0.193 |
Token | prowess |
Token ID | 30721 |
Feature activation | +1.51e-01 |
Pos logit contributions | |
Kislyak | +0.088 |
================================================================= | +0.088 |
IMAGES | +0.086 |
advertising | +0.082 |
zinski | +0.082 |
Neg logit contributions | |
Crit | -0.092 |
Nit | -0.085 |
Poverty | -0.081 |
Crit | -0.080 |
Adapt | -0.075 |
Token | as |
Token ID | 355 |
Feature activation | +4.12e-02 |
Pos logit contributions | |
Kislyak | +0.386 |
================================================================= | +0.384 |
IMAGES | +0.372 |
advertising | +0.360 |
zinski | +0.356 |
Neg logit contributions | |
Crit | -0.402 |
Nit | -0.371 |
Poverty | -0.354 |
Crit | -0.350 |
Adapt | -0.325 |
Token | a |
Token ID | 257 |
Feature activation | -1.05e-01 |
Pos logit contributions | |
Kislyak | +0.111 |
================================================================= | +0.111 |
IMAGES | +0.107 |
advertising | +0.104 |
zinski | +0.103 |
Neg logit contributions | |
Crit | -0.104 |
Nit | -0.096 |
Poverty | -0.091 |
Crit | -0.090 |
Adapt | -0.083 |
Token | path |
Token ID | 3108 |
Feature activation | +4.14e-01 |
Pos logit contributions | |
Kislyak | +0.236 |
================================================================= | +0.232 |
IMAGES | +0.224 |
advertising | +0.219 |
zinski | +0.218 |
Neg logit contributions | |
Crit | -0.257 |
Nit | -0.236 |
Poverty | -0.226 |
Crit | -0.224 |
Adapt | -0.209 |
Token | , |
Token ID | 837 |
Feature activation | +1.55e-01 |
Pos logit contributions | |
Kislyak | +1.003 |
================================================================= | +0.991 |
IMAGES | +0.956 |
advertising | +0.928 |
zinski | +0.921 |
Neg logit contributions | |
Crit | -1.170 |
Nit | -1.080 |
Poverty | -1.034 |
Crit | -1.023 |
Adapt | -0.954 |
Token | 4 |
Token ID | 604 |
Feature activation | +4.11e-01 |
Pos logit contributions | |
Kislyak | +0.405 |
================================================================= | +0.401 |
IMAGES | +0.387 |
advertising | +0.379 |
zinski | +0.377 |
Neg logit contributions | |
Crit | -0.407 |
Nit | -0.373 |
Poverty | -0.358 |
Crit | -0.354 |
Adapt | -0.326 |
Token | ) |
Token ID | 1267 |
Feature activation | +4.54e-02 |
Pos logit contributions | |
Kislyak | +0.850 |
================================================================= | +0.836 |
IMAGES | +0.792 |
advertising | +0.782 |
zinski | +0.770 |
Neg logit contributions | |
Crit | -1.313 |
Nit | -1.219 |
Poverty | -1.174 |
Crit | -1.166 |
corruption | -1.097 |
Token | = |
Token ID | 796 |
Feature activation | +2.51e-01 |
Pos logit contributions | |
Kislyak | +0.105 |
================================================================= | +0.103 |
IMAGES | +0.099 |
advertising | +0.097 |
zinski | +0.097 |
Neg logit contributions | |
Crit | -0.133 |
Nit | -0.123 |
Poverty | -0.118 |
Crit | -0.117 |
Adapt | -0.109 |
Token | '. |
Token ID | 45302 |
Feature activation | +7.10e-01 |
Pos logit contributions | |
Kislyak | +0.612 |
================================================================= | +0.605 |
IMAGES | +0.582 |
advertising | +0.569 |
zinski | +0.565 |
Neg logit contributions | |
Crit | -0.706 |
Nit | -0.651 |
Poverty | -0.626 |
Crit | -0.619 |
Adapt | -0.576 |
Token | cl |
Token ID | 565 |
Feature activation | -1.59e-01 |
Pos logit contributions | |
Kislyak | +1.787 |
================================================================= | +1.780 |
IMAGES | +1.715 |
advertising | +1.664 |
zinski | +1.647 |
Neg logit contributions | |
Crit | -1.924 |
Nit | -1.777 |
Poverty | -1.698 |
Crit | -1.675 |
corruption | -1.560 |
Token | j |
Token ID | 73 |
Feature activation | -8.56e-02 |
Pos logit contributions | |
Crit | +0.311 |
Nit | +0.278 |
Poverty | +0.261 |
Crit | +0.255 |
Adapt | +0.230 |
Neg logit contributions | |
Kislyak | -0.521 |
================================================================= | -0.520 |
IMAGES | -0.505 |
advertising | -0.494 |
zinski | -0.490 |
Token | ' |
Token ID | 6 |
Feature activation | +5.98e-02 |
Pos logit contributions | |
Crit | +0.208 |
Nit | +0.189 |
Poverty | +0.179 |
Crit | +0.178 |
corruption | +0.163 |
Neg logit contributions | |
Kislyak | -0.241 |
================================================================= | -0.238 |
IMAGES | -0.230 |
advertising | -0.226 |
Accessory | -0.224 |
Token | OR |
Token ID | 6375 |
Feature activation | +7.83e-02 |
Pos logit contributions | |
Kislyak | +0.148 |
================================================================= | +0.145 |
IMAGES | +0.139 |
advertising | +0.139 |
zinski | +0.137 |
Neg logit contributions | |
Crit | -0.166 |
Nit | -0.152 |
Poverty | -0.147 |
Crit | -0.146 |
Adapt | -0.135 |
Token | RIGHT |
Token ID | 33621 |
Feature activation | -2.94e-01 |
Pos logit contributions | |
Kislyak | +0.199 |
================================================================= | +0.197 |
IMAGES | +0.190 |
advertising | +0.186 |
zinski | +0.184 |
Neg logit contributions | |
Crit | -0.211 |
Nit | -0.194 |
Poverty | -0.186 |
Crit | -0.184 |
Adapt | -0.171 |
Token | Maryland |
Token ID | 10769 |
Feature activation | +5.26e-01 |
Pos logit contributions | |
Kislyak | +0.514 |
================================================================= | +0.512 |
IMAGES | +0.493 |
advertising | +0.479 |
zinski | +0.474 |
Neg logit contributions | |
Crit | -0.551 |
Nit | -0.509 |
Poverty | -0.486 |
Crit | -0.480 |
corruption | -0.447 |
Token | 1 |
Token ID | 352 |
Feature activation | +1.46e-01 |
Pos logit contributions | |
Kislyak | +1.298 |
================================================================= | +1.293 |
IMAGES | +1.245 |
advertising | +1.207 |
zinski | +1.194 |
Neg logit contributions | |
Crit | -1.455 |
Nit | -1.345 |
Poverty | -1.286 |
Crit | -1.269 |
corruption | -1.185 |
Token | hour |
Token ID | 1711 |
Feature activation | +9.23e-02 |
Pos logit contributions | |
Kislyak | +0.372 |
================================================================= | +0.371 |
IMAGES | +0.358 |
advertising | +0.349 |
zinski | +0.344 |
Neg logit contributions | |
Crit | -0.388 |
Nit | -0.359 |
Poverty | -0.342 |
Crit | -0.334 |
Adapt | -0.315 |
Token | ago |
Token ID | 2084 |
Feature activation | -1.82e-01 |
Pos logit contributions | |
Kislyak | +0.190 |
================================================================= | +0.189 |
IMAGES | +0.181 |
advertising | +0.174 |
zinski | +0.172 |
Neg logit contributions | |
Crit | -0.292 |
Nit | -0.273 |
Poverty | -0.262 |
Crit | -0.259 |
Adapt | -0.245 |
Token | This |
Token ID | 770 |
Feature activation | -4.82e-02 |
Pos logit contributions | |
Crit | +0.463 |
Nit | +0.423 |
Poverty | +0.410 |
Crit | +0.404 |
Ps | +0.368 |
Neg logit contributions | |
Kislyak | -0.489 |
================================================================= | -0.485 |
IMAGES | -0.470 |
advertising | -0.457 |
zinski | -0.455 |
Token | election |
Token ID | 3071 |
Feature activation | +6.61e-01 |
Pos logit contributions | |
Crit | +0.117 |
Nit | +0.107 |
Poverty | +0.102 |
Crit | +0.100 |
corruption | +0.092 |
Neg logit contributions | |
Kislyak | -0.136 |
================================================================= | -0.135 |
IMAGES | -0.131 |
advertising | -0.127 |
zinski | -0.126 |
Token | was |
Token ID | 373 |
Feature activation | +5.50e-02 |
Pos logit contributions | |
Kislyak | +1.735 |
================================================================= | +1.730 |
IMAGES | +1.667 |
advertising | +1.619 |
zinski | +1.605 |
Neg logit contributions | |
Crit | -1.718 |
Nit | -1.581 |
Poverty | -1.505 |
Crit | -1.482 |
Adapt | -1.380 |
Token | not |
Token ID | 407 |
Feature activation | -1.98e-01 |
Pos logit contributions | |
Kislyak | +0.144 |
================================================================= | +0.143 |
IMAGES | +0.139 |
advertising | +0.135 |
zinski | +0.133 |
Neg logit contributions | |
Crit | -0.143 |
Nit | -0.132 |
Poverty | -0.126 |
Crit | -0.125 |
corruption | -0.115 |
Token | about |
Token ID | 546 |
Feature activation | -6.22e-02 |
Pos logit contributions | |
Crit | +0.578 |
Nit | +0.535 |
Poverty | +0.519 |
Crit | +0.514 |
crit | +0.478 |
Neg logit contributions | |
Kislyak | -0.452 |
================================================================= | -0.448 |
IMAGES | -0.438 |
advertising | -0.422 |
Accessory | -0.417 |
Token | policy |
Token ID | 2450 |
Feature activation | -2.57e-01 |
Pos logit contributions | |
Crit | +0.149 |
Nit | +0.135 |
Poverty | +0.131 |
Crit | +0.129 |
corruption | +0.118 |
Neg logit contributions | |
Kislyak | -0.175 |
================================================================= | -0.175 |
IMAGES | -0.170 |
advertising | -0.165 |
Accessory | -0.164 |
Token | differences |
Token ID | 5400 |
Feature activation | -8.66e-02 |
Pos logit contributions | |
Crit | +0.712 |
Nit | +0.653 |
Poverty | +0.628 |
Crit | +0.621 |
corruption | +0.578 |
Neg logit contributions | |
Kislyak | -0.634 |
================================================================= | -0.628 |
IMAGES | -0.610 |
advertising | -0.587 |
Accessory | -0.584 |
Token | it |
Token ID | 340 |
Feature activation | -5.16e-02 |
Pos logit contributions | |
Kislyak | +0.669 |
================================================================= | +0.665 |
IMAGES | +0.645 |
advertising | +0.623 |
zinski | +0.621 |
Neg logit contributions | |
Crit | -0.683 |
Nit | -0.625 |
Poverty | -0.601 |
Crit | -0.597 |
corruption | -0.548 |
Token | � |
Token ID | 447 |
Feature activation | +1.52e-01 |
Pos logit contributions | |
Crit | +0.135 |
Nit | +0.124 |
Poverty | +0.118 |
Crit | +0.117 |
corruption | +0.108 |
Neg logit contributions | |
Kislyak | -0.135 |
================================================================= | -0.134 |
IMAGES | -0.130 |
advertising | -0.125 |
zinski | -0.125 |
Token | � |
Token ID | 247 |
Feature activation | +5.39e-02 |
Pos logit contributions | |
Kislyak | +0.474 |
================================================================= | +0.471 |
IMAGES | +0.456 |
advertising | +0.447 |
zinski | +0.442 |
Neg logit contributions | |
Crit | -0.322 |
Nit | -0.292 |
Poverty | -0.274 |
Crit | -0.268 |
corruption | -0.244 |
Token | s |
Token ID | 82 |
Feature activation | -1.48e-01 |
Pos logit contributions | |
Kislyak | +0.128 |
================================================================= | +0.127 |
IMAGES | +0.123 |
advertising | +0.117 |
Accessory | +0.117 |
Neg logit contributions | |
Crit | -0.155 |
Nit | -0.143 |
Poverty | -0.137 |
Crit | -0.136 |
Ps | -0.129 |
Token | easily |
Token ID | 3538 |
Feature activation | -3.89e-02 |
Pos logit contributions | |
Crit | +0.369 |
Nit | +0.338 |
Poverty | +0.321 |
Crit | +0.320 |
corruption | +0.292 |
Neg logit contributions | |
Kislyak | -0.405 |
================================================================= | -0.402 |
IMAGES | -0.391 |
advertising | -0.379 |
zinski | -0.376 |
Token | eclips |
Token ID | 39097 |
Feature activation | +6.49e-01 |
Pos logit contributions | |
Crit | +0.098 |
Nit | +0.089 |
Poverty | +0.085 |
Crit | +0.085 |
Adapt | +0.077 |
Neg logit contributions | |
Kislyak | -0.106 |
================================================================= | -0.106 |
IMAGES | -0.103 |
advertising | -0.100 |
zinski | -0.099 |
Token | ed |
Token ID | 276 |
Feature activation | -4.59e-02 |
Pos logit contributions | |
Kislyak | +0.920 |
================================================================= | +0.916 |
IMAGES | +0.866 |
advertising | +0.801 |
Accessory | +0.795 |
Neg logit contributions | |
Crit | -2.478 |
Nit | -2.333 |
Poverty | -2.265 |
Crit | -2.249 |
corruption | -2.144 |
Token | by |
Token ID | 416 |
Feature activation | -1.03e-01 |
Pos logit contributions | |
Crit | +0.126 |
Nit | +0.115 |
Poverty | +0.112 |
Crit | +0.111 |
corruption | +0.102 |
Neg logit contributions | |
Kislyak | -0.114 |
================================================================= | -0.114 |
IMAGES | -0.110 |
zinski | -0.106 |
advertising | -0.106 |
Token | its |
Token ID | 663 |
Feature activation | +2.03e-02 |
Pos logit contributions | |
Crit | +0.271 |
Nit | +0.247 |
Crit | +0.237 |
Poverty | +0.237 |
corruption | +0.215 |
Neg logit contributions | |
Kislyak | -0.268 |
================================================================= | -0.267 |
IMAGES | -0.259 |
advertising | -0.250 |
zinski | -0.249 |
Token | successors |
Token ID | 41491 |
Feature activation | +3.96e-01 |
Pos logit contributions | |
Kislyak | +0.053 |
================================================================= | +0.053 |
IMAGES | +0.051 |
advertising | +0.050 |
zinski | +0.049 |
Neg logit contributions | |
Crit | -0.053 |
Nit | -0.049 |
Crit | -0.047 |
Poverty | -0.047 |
corruption | -0.042 |
Token | Sly |
Token ID | 31615 |
Feature activation | +6.54e-01 |
Pos logit contributions | |
Kislyak | +0.984 |
================================================================= | +0.978 |
IMAGES | +0.944 |
advertising | +0.914 |
zinski | +0.909 |
Neg logit contributions | |
Crit | -1.093 |
Nit | -1.009 |
Poverty | -0.968 |
Crit | -0.960 |
Adapt | -0.885 |
Token | British |
Token ID | 3517 |
Feature activation | -1.81e-01 |
Pos logit contributions | |
Kislyak | +0.427 |
================================================================= | +0.426 |
IMAGES | +0.411 |
advertising | +0.401 |
zinski | +0.395 |
Neg logit contributions | |
Crit | -0.436 |
Nit | -0.400 |
Poverty | -0.385 |
Crit | -0.381 |
corruption | -0.352 |
Token | commander |
Token ID | 11561 |
Feature activation | -1.12e-01 |
Pos logit contributions | |
Crit | +0.475 |
Nit | +0.436 |
Poverty | +0.420 |
Crit | +0.417 |
Ps | +0.384 |
Neg logit contributions | |
Kislyak | -0.470 |
================================================================= | -0.462 |
IMAGES | -0.450 |
advertising | -0.441 |
zinski | -0.436 |
Token | - |
Token ID | 12 |
Feature activation | -1.53e-01 |
Pos logit contributions | |
Crit | +0.299 |
Nit | +0.274 |
Crit | +0.265 |
Poverty | +0.264 |
Ps | +0.244 |
Neg logit contributions | |
Kislyak | -0.284 |
================================================================= | -0.283 |
IMAGES | -0.278 |
Accessory | -0.268 |
advertising | -0.267 |
Token | in |
Token ID | 259 |
Feature activation | +1.38e-03 |
Pos logit contributions | |
Crit | +0.503 |
Nit | +0.471 |
Poverty | +0.453 |
Crit | +0.449 |
corruption | +0.428 |
Neg logit contributions | |
Kislyak | -0.298 |
================================================================= | -0.295 |
IMAGES | -0.285 |
advertising | -0.267 |
zinski | -0.267 |
Token | - |
Token ID | 12 |
Feature activation | +1.70e-01 |
Pos logit contributions | |
Kislyak | +0.003 |
================================================================= | +0.003 |
IMAGES | +0.003 |
advertising | +0.003 |
zinski | +0.003 |
Neg logit contributions | |
Crit | -0.004 |
Nit | -0.003 |
Poverty | -0.003 |
Crit | -0.003 |
corruption | -0.003 |
Token | chief |
Token ID | 17351 |
Feature activation | +7.83e-01 |
Pos logit contributions | |
Kislyak | +0.504 |
================================================================= | +0.502 |
IMAGES | +0.487 |
advertising | +0.472 |
zinski | +0.470 |
Neg logit contributions | |
Crit | -0.388 |
Nit | -0.352 |
Poverty | -0.333 |
Crit | -0.327 |
corruption | -0.301 |
Token | , |
Token ID | 11 |
Feature activation | +2.66e-02 |
Pos logit contributions | |
Kislyak | +1.811 |
================================================================= | +1.804 |
IMAGES | +1.732 |
advertising | +1.676 |
zinski | +1.657 |
Neg logit contributions | |
Crit | -2.283 |
Nit | -2.119 |
Poverty | -2.033 |
Crit | -2.009 |
corruption | -1.881 |
Token | Sir |
Token ID | 7361 |
Feature activation | +1.09e-01 |
Pos logit contributions | |
Kislyak | +0.070 |
================================================================= | +0.070 |
IMAGES | +0.068 |
advertising | +0.066 |
Accessory | +0.066 |
Neg logit contributions | |
Crit | -0.068 |
Nit | -0.063 |
Poverty | -0.060 |
Crit | -0.060 |
Ps | -0.055 |
Token | Ian |
Token ID | 12930 |
Feature activation | +4.54e-01 |
Pos logit contributions | |
Kislyak | +0.269 |
================================================================= | +0.267 |
IMAGES | +0.260 |
Accessory | +0.255 |
advertising | +0.252 |
Neg logit contributions | |
Crit | -0.301 |
Nit | -0.276 |
Poverty | -0.270 |
Crit | -0.269 |
Lumpur | -0.246 |
Token | Hamilton |
Token ID | 11582 |
Feature activation | +1.44e-02 |
Pos logit contributions | |
Kislyak | +1.097 |
================================================================= | +1.093 |
IMAGES | +1.059 |
advertising | +1.023 |
zinski | +1.011 |
Neg logit contributions | |
Crit | -1.273 |
Nit | -1.176 |
Poverty | -1.135 |
Crit | -1.126 |
Lumpur | -1.042 |
Token | . |
Token ID | 13 |
Feature activation | +9.84e-02 |
Pos logit contributions | |
Kislyak | +0.033 |
================================================================= | +0.032 |
IMAGES | +0.032 |
advertising | +0.030 |
Accessory | +0.030 |
Neg logit contributions | |
Crit | -0.042 |
Nit | -0.039 |
Poverty | -0.038 |
Crit | -0.038 |
Lumpur | -0.035 |
Token | they |
Token ID | 484 |
Feature activation | +2.50e-02 |
Pos logit contributions | |
Crit | +0.525 |
Nit | +0.479 |
Poverty | +0.466 |
Crit | +0.462 |
corruption | +0.427 |
Neg logit contributions | |
================================================================= | -0.505 |
Kislyak | -0.501 |
IMAGES | -0.486 |
zinski | -0.476 |
advertising | -0.473 |
Token | mistakenly |
Token ID | 33168 |
Feature activation | -2.91e-01 |
Pos logit contributions | |
Kislyak | +0.072 |
================================================================= | +0.072 |
IMAGES | +0.070 |
advertising | +0.068 |
Accessory | +0.068 |
Neg logit contributions | |
Crit | -0.058 |
Nit | -0.053 |
Poverty | -0.051 |
Crit | -0.050 |
crit | -0.046 |
Token | detain |
Token ID | 17017 |
Feature activation | +2.79e-02 |
Pos logit contributions | |
Crit | +0.671 |
Nit | +0.603 |
Poverty | +0.586 |
Crit | +0.582 |
crit | +0.539 |
Neg logit contributions | |
================================================================= | -0.839 |
Kislyak | -0.838 |
IMAGES | -0.812 |
advertising | -0.799 |
zinski | -0.791 |
Token | somebody |
Token ID | 8276 |
Feature activation | +7.95e-02 |
Pos logit contributions | |
================================================================= | +0.075 |
Kislyak | +0.074 |
IMAGES | +0.072 |
zinski | +0.071 |
advertising | +0.070 |
Neg logit contributions | |
Crit | -0.071 |
Nit | -0.065 |
Poverty | -0.062 |
Crit | -0.062 |
corruption | -0.057 |
Token | who |
Token ID | 508 |
Feature activation | +1.59e-01 |
Pos logit contributions | |
================================================================= | +0.208 |
Kislyak | +0.208 |
IMAGES | +0.200 |
advertising | +0.196 |
zinski | +0.195 |
Neg logit contributions | |
Crit | -0.207 |
Nit | -0.189 |
Poverty | -0.182 |
Crit | -0.180 |
Lumpur | -0.166 |
Token | is |
Token ID | 318 |
Feature activation | -1.98e-02 |
Pos logit contributions | |
================================================================= | +0.430 |
Kislyak | +0.426 |
IMAGES | +0.411 |
advertising | +0.403 |
zinski | +0.401 |
Neg logit contributions | |
Crit | -0.404 |
Nit | -0.367 |
Poverty | -0.354 |
Crit | -0.350 |
Ps | -0.322 |
Token | here |
Token ID | 994 |
Feature activation | +2.24e-01 |
Pos logit contributions | |
Crit | +0.049 |
Nit | +0.044 |
Poverty | +0.043 |
Crit | +0.042 |
Ps | +0.039 |
Neg logit contributions | |
================================================================= | -0.055 |
Kislyak | -0.054 |
IMAGES | -0.053 |
advertising | -0.051 |
zinski | -0.051 |
Token | lawfully |
Token ID | 41095 |
Feature activation | -7.11e-01 |
Pos logit contributions | |
Kislyak | +0.700 |
================================================================= | +0.700 |
IMAGES | +0.678 |
advertising | +0.662 |
zinski | +0.658 |
Neg logit contributions | |
Crit | -0.470 |
Nit | -0.422 |
Poverty | -0.399 |
Crit | -0.392 |
corruption | -0.355 |
Token | or |
Token ID | 393 |
Feature activation | +9.43e-02 |
Pos logit contributions | |
Crit | +1.862 |
Nit | +1.699 |
Poverty | +1.648 |
Crit | +1.631 |
Lumpur | +1.512 |
Neg logit contributions | |
================================================================= | -1.841 |
Kislyak | -1.825 |
IMAGES | -1.772 |
zinski | -1.717 |
advertising | -1.714 |
Token | , |
Token ID | 11 |
Feature activation | +1.54e-01 |
Pos logit contributions | |
================================================================= | +0.259 |
Kislyak | +0.259 |
IMAGES | +0.250 |
advertising | +0.245 |
zinski | +0.243 |
Neg logit contributions | |
Crit | -0.233 |
Nit | -0.211 |
Poverty | -0.204 |
Crit | -0.201 |
Ps | -0.185 |
Token | in |
Token ID | 287 |
Feature activation | +2.76e-02 |
Pos logit contributions | |
Kislyak | +0.419 |
================================================================= | +0.418 |
IMAGES | +0.404 |
advertising | +0.393 |
zinski | +0.390 |
Neg logit contributions | |
Crit | -0.388 |
Nit | -0.355 |
Poverty | -0.339 |
Crit | -0.334 |
corruption | -0.309 |
Token | just |
Token ID | 655 |
Feature activation | -1.86e-01 |
Pos logit contributions | |
Crit | +0.810 |
Nit | +0.748 |
Poverty | +0.719 |
Crit | +0.717 |
Adapt | +0.670 |
Neg logit contributions | |
Kislyak | -0.598 |
================================================================= | -0.592 |
IMAGES | -0.573 |
advertising | -0.549 |
zinski | -0.547 |
Token | really |
Token ID | 1107 |
Feature activation | +2.37e-01 |
Pos logit contributions | |
Crit | +0.489 |
Nit | +0.451 |
Poverty | +0.430 |
Crit | +0.427 |
crit | +0.395 |
Neg logit contributions | |
Kislyak | -0.482 |
================================================================= | -0.480 |
IMAGES | -0.466 |
advertising | -0.449 |
zinski | -0.447 |
Token | is |
Token ID | 318 |
Feature activation | -8.29e-02 |
Pos logit contributions | |
Kislyak | +0.642 |
================================================================= | +0.639 |
IMAGES | +0.618 |
advertising | +0.600 |
zinski | +0.595 |
Neg logit contributions | |
Crit | -0.600 |
Nit | -0.552 |
Poverty | -0.524 |
Crit | -0.518 |
corruption | -0.478 |
Token | that |
Token ID | 326 |
Feature activation | -8.08e-02 |
Pos logit contributions | |
Crit | +0.207 |
Nit | +0.189 |
Poverty | +0.180 |
Crit | +0.179 |
corruption | +0.165 |
Neg logit contributions | |
Kislyak | -0.227 |
================================================================= | -0.225 |
IMAGES | -0.220 |
advertising | -0.213 |
zinski | -0.211 |
Token | popular |
Token ID | 2968 |
Feature activation | +4.77e-01 |
Pos logit contributions | |
Crit | +0.193 |
Nit | +0.175 |
Poverty | +0.167 |
Crit | +0.167 |
crit | +0.152 |
Neg logit contributions | |
Kislyak | -0.229 |
================================================================= | -0.228 |
IMAGES | -0.223 |
advertising | -0.215 |
zinski | -0.215 |
Token | , |
Token ID | 11 |
Feature activation | -1.65e-03 |
Pos logit contributions | |
Kislyak | +1.075 |
================================================================= | +1.071 |
IMAGES | +1.029 |
advertising | +0.993 |
zinski | +0.981 |
Neg logit contributions | |
Crit | -1.420 |
Nit | -1.319 |
Poverty | -1.266 |
Crit | -1.255 |
corruption | -1.176 |
Token | and |
Token ID | 290 |
Feature activation | -9.41e-03 |
Pos logit contributions | |
Crit | +0.004 |
Nit | +0.004 |
Poverty | +0.004 |
Crit | +0.004 |
Adapt | +0.003 |
Neg logit contributions | |
Kislyak | -0.004 |
================================================================= | -0.004 |
IMAGES | -0.004 |
advertising | -0.004 |
zinski | -0.004 |
Token | I |
Token ID | 314 |
Feature activation | -6.80e-02 |
Pos logit contributions | |
Crit | +0.024 |
Nit | +0.022 |
Poverty | +0.021 |
Crit | +0.021 |
corruption | +0.019 |
Neg logit contributions | |
Kislyak | -0.025 |
================================================================= | -0.025 |
IMAGES | -0.024 |
advertising | -0.024 |
zinski | -0.024 |
Token | � |
Token ID | 447 |
Feature activation | -3.96e-02 |
Pos logit contributions | |
Crit | +0.163 |
Nit | +0.149 |
Poverty | +0.141 |
Crit | +0.139 |
corruption | +0.128 |
Neg logit contributions | |
Kislyak | -0.193 |
================================================================= | -0.192 |
IMAGES | -0.187 |
advertising | -0.181 |
Accessory | -0.180 |
Token | � |
Token ID | 247 |
Feature activation | -1.32e-01 |
Pos logit contributions | |
Crit | +0.085 |
Nit | +0.077 |
Poverty | +0.073 |
Crit | +0.072 |
corruption | +0.065 |
Neg logit contributions | |
Kislyak | -0.122 |
================================================================= | -0.121 |
IMAGES | -0.117 |
advertising | -0.115 |
zinski | -0.113 |
Token | ve |
Token ID | 303 |
Feature activation | +8.43e-02 |
Pos logit contributions | |
Crit | +0.305 |
Nit | +0.277 |
Poverty | +0.262 |
Crit | +0.258 |
corruption | +0.238 |
Neg logit contributions | |
Kislyak | -0.387 |
================================================================= | -0.385 |
IMAGES | -0.376 |
advertising | -0.363 |
Accessory | -0.363 |
Token | the |
Token ID | 262 |
Feature activation | -2.40e-01 |
Pos logit contributions | |
Crit | +0.552 |
Nit | +0.504 |
Poverty | +0.482 |
Crit | +0.478 |
Ps | +0.430 |
Neg logit contributions | |
Kislyak | -0.663 |
================================================================= | -0.660 |
IMAGES | -0.645 |
advertising | -0.622 |
Accessory | -0.619 |
Token | major |
Token ID | 1688 |
Feature activation | -3.26e-02 |
Pos logit contributions | |
Crit | +0.615 |
Nit | +0.567 |
Poverty | +0.546 |
Crit | +0.541 |
crit | +0.493 |
Neg logit contributions | |
Kislyak | -0.638 |
================================================================= | -0.633 |
IMAGES | -0.620 |
advertising | -0.594 |
zinski | -0.592 |
Token | use |
Token ID | 779 |
Feature activation | -1.51e-01 |
Pos logit contributions | |
Crit | +0.084 |
Nit | +0.077 |
Poverty | +0.074 |
Crit | +0.074 |
corruption | +0.067 |
Neg logit contributions | |
Kislyak | -0.086 |
================================================================= | -0.086 |
IMAGES | -0.083 |
advertising | -0.080 |
zinski | -0.080 |
Token | of |
Token ID | 286 |
Feature activation | -2.38e-01 |
Pos logit contributions | |
Crit | +0.394 |
Nit | +0.364 |
Poverty | +0.347 |
Crit | +0.345 |
Ps | +0.317 |
Neg logit contributions | |
Kislyak | -0.394 |
================================================================= | -0.391 |
IMAGES | -0.381 |
advertising | -0.368 |
zinski | -0.364 |
Token | his |
Token ID | 465 |
Feature activation | -9.71e-02 |
Pos logit contributions | |
Crit | +0.664 |
Nit | +0.619 |
Poverty | +0.595 |
Crit | +0.592 |
crit | +0.543 |
Neg logit contributions | |
Kislyak | -0.572 |
================================================================= | -0.572 |
IMAGES | -0.555 |
advertising | -0.532 |
Accessory | -0.530 |
Token | metal |
Token ID | 6147 |
Feature activation | +1.28e-01 |
Pos logit contributions | |
Crit | +0.257 |
Nit | +0.239 |
Poverty | +0.229 |
Crit | +0.228 |
crit | +0.210 |
Neg logit contributions | |
Kislyak | -0.248 |
================================================================= | -0.245 |
IMAGES | -0.241 |
advertising | -0.231 |
Accessory | -0.230 |
Token | . |
Token ID | 13 |
Feature activation | +1.12e-01 |
Pos logit contributions | |
Kislyak | +0.299 |
================================================================= | +0.296 |
IMAGES | +0.286 |
advertising | +0.274 |
zinski | +0.274 |
Neg logit contributions | |
Crit | -0.371 |
Nit | -0.344 |
Poverty | -0.332 |
Crit | -0.328 |
corruption | -0.307 |
Token | " |
Token ID | 366 |
Feature activation | +9.12e-03 |
Pos logit contributions | |
Kislyak | +0.316 |
================================================================= | +0.315 |
IMAGES | +0.305 |
advertising | +0.297 |
zinski | +0.294 |
Neg logit contributions | |
Crit | -0.268 |
Nit | -0.244 |
Poverty | -0.233 |
Crit | -0.230 |
Adapt | -0.209 |
Token | Sh |
Token ID | 2484 |
Feature activation | -2.30e-02 |
Pos logit contributions | |
Kislyak | +0.028 |
================================================================= | +0.028 |
IMAGES | +0.027 |
advertising | +0.026 |
zinski | +0.026 |
Neg logit contributions | |
Crit | -0.020 |
Nit | -0.018 |
Poverty | -0.017 |
Crit | -0.017 |
Adapt | -0.015 |
Token | all |
Token ID | 439 |
Feature activation | +3.38e-01 |
Pos logit contributions | |
Crit | +0.062 |
Nit | +0.057 |
Poverty | +0.055 |
Crit | +0.054 |
Ps | +0.050 |
Neg logit contributions | |
Kislyak | -0.058 |
================================================================= | -0.058 |
IMAGES | -0.057 |
Accessory | -0.054 |
advertising | -0.054 |
Token | we |
Token ID | 356 |
Feature activation | -2.28e-01 |
Pos logit contributions | |
Kislyak | +0.853 |
================================================================= | +0.849 |
IMAGES | +0.819 |
advertising | +0.793 |
zinski | +0.786 |
Neg logit contributions | |
Crit | -0.916 |
Nit | -0.846 |
Poverty | -0.809 |
Crit | -0.799 |
corruption | -0.743 |
Token | a |
Token ID | 64 |
Feature activation | -1.06e-01 |
Pos logit contributions | |
Crit | +0.345 |
Nit | +0.321 |
Poverty | +0.308 |
Crit | +0.303 |
corruption | +0.285 |
Neg logit contributions | |
Kislyak | -0.284 |
================================================================= | -0.284 |
IMAGES | -0.274 |
advertising | -0.264 |
zinski | -0.261 |
Token | despite |
Token ID | 3805 |
Feature activation | -2.78e-02 |
Pos logit contributions | |
Crit | +0.293 |
Nit | +0.272 |
Poverty | +0.260 |
Crit | +0.257 |
corruption | +0.239 |
Neg logit contributions | |
Kislyak | -0.259 |
================================================================= | -0.258 |
IMAGES | -0.249 |
advertising | -0.242 |
zinski | -0.238 |
Token | the |
Token ID | 262 |
Feature activation | +7.09e-02 |
Pos logit contributions | |
Crit | +0.068 |
Nit | +0.062 |
Poverty | +0.059 |
Crit | +0.058 |
corruption | +0.054 |
Neg logit contributions | |
Kislyak | -0.077 |
================================================================= | -0.077 |
IMAGES | -0.074 |
advertising | -0.072 |
zinski | -0.072 |
Token | fact |
Token ID | 1109 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
Kislyak | +0.187 |
================================================================= | +0.186 |
IMAGES | +0.180 |
advertising | +0.175 |
zinski | +0.173 |
Neg logit contributions | |
Crit | -0.184 |
Nit | -0.169 |
Poverty | -0.162 |
Crit | -0.159 |
corruption | -0.148 |
Token | that |
Token ID | 326 |
Feature activation | -1.04e-01 |
Pos logit contributions | |
Kislyak | +0.292 |
================================================================= | +0.291 |
IMAGES | +0.280 |
advertising | +0.272 |
zinski | +0.269 |
Neg logit contributions | |
Crit | -0.303 |
Nit | -0.280 |
Poverty | -0.267 |
Crit | -0.262 |
Adapt | -0.246 |
Token | it |
Token ID | 340 |
Feature activation | -1.03e-01 |
Pos logit contributions | |
Crit | +0.249 |
Nit | +0.227 |
Poverty | +0.217 |
Crit | +0.213 |
corruption | +0.196 |
Neg logit contributions | |
Kislyak | -0.297 |
================================================================= | -0.296 |
IMAGES | -0.287 |
advertising | -0.279 |
zinski | -0.277 |
Token | is |
Token ID | 318 |
Feature activation | -6.69e-02 |
Pos logit contributions | |
Crit | +0.253 |
Nit | +0.232 |
Poverty | +0.221 |
Crit | +0.218 |
corruption | +0.200 |
Neg logit contributions | |
Kislyak | -0.283 |
================================================================= | -0.283 |
IMAGES | -0.273 |
advertising | -0.267 |
zinski | -0.264 |
Token | abundantly |
Token ID | 49244 |
Feature activation | +2.97e-01 |
Pos logit contributions | |
Crit | +0.166 |
Nit | +0.152 |
Poverty | +0.144 |
Crit | +0.142 |
corruption | +0.131 |
Neg logit contributions | |
Kislyak | -0.184 |
================================================================= | -0.184 |
IMAGES | -0.178 |
advertising | -0.173 |
zinski | -0.171 |
Token | clear |
Token ID | 1598 |
Feature activation | +4.48e-01 |
Pos logit contributions | |
Kislyak | +0.789 |
================================================================= | +0.789 |
IMAGES | +0.758 |
advertising | +0.736 |
zinski | +0.730 |
Neg logit contributions | |
Crit | -0.759 |
Nit | -0.697 |
Poverty | -0.664 |
Crit | -0.654 |
Adapt | -0.609 |
Token | that |
Token ID | 326 |
Feature activation | +3.35e-02 |
Pos logit contributions | |
Kislyak | +1.149 |
================================================================= | +1.143 |
IMAGES | +1.100 |
advertising | +1.068 |
zinski | +1.058 |
Neg logit contributions | |
Crit | -1.189 |
Nit | -1.094 |
Poverty | -1.043 |
Crit | -1.027 |
Adapt | -0.965 |
Token | V |
Token ID | 569 |
Feature activation | -2.68e-01 |
Pos logit contributions | |
Kislyak | +0.094 |
================================================================= | +0.094 |
IMAGES | +0.091 |
advertising | +0.088 |
zinski | +0.088 |
Neg logit contributions | |
Crit | -0.081 |
Nit | -0.074 |
Poverty | -0.070 |
Crit | -0.069 |
corruption | -0.064 |
Token | ucking |
Token ID | 19296 |
Feature activation | -1.25e-01 |
Pos logit contributions | |
Kislyak | +0.141 |
================================================================= | +0.140 |
IMAGES | +0.135 |
advertising | +0.130 |
Accessory | +0.130 |
Neg logit contributions | |
Crit | -0.157 |
Nit | -0.144 |
Poverty | -0.138 |
Crit | -0.137 |
corruption | -0.128 |
Token | minorities |
Token ID | 15557 |
Feature activation | -8.07e-02 |
Pos logit contributions | |
Crit | +0.319 |
Nit | +0.294 |
Poverty | +0.281 |
Crit | +0.277 |
corruption | +0.255 |
Neg logit contributions | |
Kislyak | -0.334 |
================================================================= | -0.332 |
IMAGES | -0.321 |
advertising | -0.312 |
zinski | -0.310 |
Token | . |
Token ID | 13 |
Feature activation | -7.65e-02 |
Pos logit contributions | |
Crit | +0.268 |
Nit | +0.251 |
Poverty | +0.242 |
Crit | +0.240 |
corruption | +0.226 |
Neg logit contributions | |
Kislyak | -0.154 |
================================================================= | -0.153 |
IMAGES | -0.146 |
advertising | -0.139 |
Accessory | -0.138 |
Token | � |
Token ID | 447 |
Feature activation | +1.85e-01 |
Pos logit contributions | |
Crit | +0.184 |
Nit | +0.167 |
Poverty | +0.160 |
Crit | +0.159 |
Adapt | +0.143 |
Neg logit contributions | |
Kislyak | -0.216 |
================================================================= | -0.215 |
IMAGES | -0.209 |
advertising | -0.204 |
zinski | -0.203 |
Token | � |
Token ID | 251 |
Feature activation | +2.23e-01 |
Pos logit contributions | |
Kislyak | +0.426 |
================================================================= | +0.424 |
IMAGES | +0.406 |
advertising | +0.393 |
zinski | +0.388 |
Neg logit contributions | |
Crit | -0.543 |
Nit | -0.505 |
Poverty | -0.484 |
Crit | -0.477 |
corruption | -0.449 |
Token | ) |
Token ID | 8 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
Kislyak | +0.594 |
================================================================= | +0.592 |
IMAGES | +0.571 |
advertising | +0.557 |
zinski | +0.552 |
Neg logit contributions | |
Crit | -0.572 |
Nit | -0.523 |
Poverty | -0.502 |
Crit | -0.496 |
Adapt | -0.455 |
Token |
|
Token ID | 198 |
Feature activation | +2.33e-02 |
Pos logit contributions | |
Kislyak | +0.315 |
================================================================= | +0.314 |
IMAGES | +0.304 |
advertising | +0.297 |
zinski | +0.294 |
Neg logit contributions | |
Crit | -0.241 |
Nit | -0.218 |
Poverty | -0.207 |
Crit | -0.204 |
Adapt | -0.186 |
Token |
|
Token ID | 198 |
Feature activation | +3.38e-02 |
Pos logit contributions | |
Kislyak | +0.064 |
================================================================= | +0.064 |
IMAGES | +0.062 |
advertising | +0.060 |
zinski | +0.060 |
Neg logit contributions | |
Crit | -0.057 |
Nit | -0.052 |
Poverty | -0.050 |
Crit | -0.049 |
corruption | -0.045 |
Token | None |
Token ID | 14202 |
Feature activation | -2.73e-01 |
Pos logit contributions | |
Kislyak | +0.108 |
================================================================= | +0.107 |
IMAGES | +0.104 |
advertising | +0.101 |
zinski | +0.101 |
Neg logit contributions | |
Crit | -0.070 |
Nit | -0.063 |
Poverty | -0.058 |
Crit | -0.057 |
Adapt | -0.052 |
Token | of |
Token ID | 286 |
Feature activation | -3.04e-02 |
Pos logit contributions | |
Crit | +0.772 |
Nit | +0.718 |
Poverty | +0.687 |
Crit | +0.675 |
Adapt | +0.635 |
Neg logit contributions | |
Kislyak | -0.658 |
================================================================= | -0.655 |
IMAGES | -0.627 |
advertising | -0.609 |
zinski | -0.603 |
Token | this |
Token ID | 428 |
Feature activation | -9.70e-02 |
Pos logit contributions | |
Crit | +0.074 |
Nit | +0.067 |
Poverty | +0.064 |
Crit | +0.063 |
corruption | +0.058 |
Neg logit contributions | |
Kislyak | -0.085 |
================================================================= | -0.085 |
IMAGES | -0.082 |
advertising | -0.080 |
zinski | -0.079 |
Token | , |
Token ID | 11 |
Feature activation | -4.81e-02 |
Pos logit contributions | |
Crit | +1.002 |
Nit | +0.927 |
Poverty | +0.892 |
Crit | +0.886 |
corruption | +0.822 |
Neg logit contributions | |
Kislyak | -0.820 |
================================================================= | -0.809 |
IMAGES | -0.793 |
advertising | -0.765 |
Accessory | -0.755 |
Token | I |
Token ID | 314 |
Feature activation | -9.34e-02 |
Pos logit contributions | |
Crit | +0.132 |
Nit | +0.121 |
Poverty | +0.117 |
Crit | +0.115 |
corruption | +0.107 |
Neg logit contributions | |
Kislyak | -0.120 |
================================================================= | -0.119 |
IMAGES | -0.116 |
advertising | -0.112 |
zinski | -0.111 |
Token | have |
Token ID | 423 |
Feature activation | -4.84e-02 |
Pos logit contributions | |
Crit | +0.225 |
Nit | +0.205 |
Poverty | +0.195 |
Crit | +0.192 |
Adapt | +0.177 |
Neg logit contributions | |
Kislyak | -0.264 |
================================================================= | -0.263 |
IMAGES | -0.255 |
advertising | -0.248 |
zinski | -0.245 |
Token | finally |
Token ID | 3443 |
Feature activation | -1.37e-01 |
Pos logit contributions | |
Crit | +0.123 |
Nit | +0.113 |
Poverty | +0.109 |
Crit | +0.108 |
Ps | +0.099 |
Neg logit contributions | |
Kislyak | -0.129 |
================================================================= | -0.128 |
IMAGES | -0.125 |
advertising | -0.122 |
zinski | -0.120 |
Token | started |
Token ID | 2067 |
Feature activation | -2.31e-01 |
Pos logit contributions | |
Crit | +0.351 |
Nit | +0.321 |
Poverty | +0.309 |
Crit | +0.306 |
crit | +0.282 |
Neg logit contributions | |
Kislyak | -0.362 |
================================================================= | -0.358 |
IMAGES | -0.351 |
advertising | -0.341 |
zinski | -0.337 |
Token | pursuing |
Token ID | 15461 |
Feature activation | -1.35e-01 |
Pos logit contributions | |
Crit | +0.587 |
Nit | +0.537 |
Poverty | +0.517 |
Crit | +0.513 |
crit | +0.472 |
Neg logit contributions | |
Kislyak | -0.620 |
================================================================= | -0.613 |
IMAGES | -0.602 |
advertising | -0.580 |
Accessory | -0.579 |
Token | music |
Token ID | 2647 |
Feature activation | -1.13e-01 |
Pos logit contributions | |
Crit | +0.349 |
Nit | +0.316 |
Poverty | +0.308 |
Crit | +0.306 |
crit | +0.280 |
Neg logit contributions | |
Kislyak | -0.355 |
================================================================= | -0.350 |
IMAGES | -0.343 |
Accessory | -0.334 |
advertising | -0.332 |
Token | seriously |
Token ID | 6411 |
Feature activation | +1.79e-01 |
Pos logit contributions | |
Crit | +0.309 |
Nit | +0.283 |
Poverty | +0.271 |
Crit | +0.270 |
corruption | +0.250 |
Neg logit contributions | |
Kislyak | -0.283 |
================================================================= | -0.280 |
IMAGES | -0.270 |
advertising | -0.262 |
zinski | -0.261 |
Token | , |
Token ID | 11 |
Feature activation | -5.41e-02 |
Pos logit contributions | |
Kislyak | +0.412 |
================================================================= | +0.408 |
IMAGES | +0.394 |
advertising | +0.381 |
Accessory | +0.377 |
Neg logit contributions | |
Crit | -0.526 |
Nit | -0.488 |
Poverty | -0.468 |
Crit | -0.464 |
corruption | -0.434 |
Token | I |
Token ID | 314 |
Feature activation | -8.76e-02 |
Pos logit contributions | |
Crit | +0.152 |
Nit | +0.141 |
Poverty | +0.135 |
Crit | +0.134 |
corruption | +0.124 |
Neg logit contributions | |
Kislyak | -0.131 |
================================================================= | -0.130 |
IMAGES | -0.125 |
advertising | -0.121 |
zinski | -0.120 |
Token | live |
Token ID | 2107 |
Feature activation | -4.57e-02 |
Pos logit contributions | |
Crit | +0.201 |
Nit | +0.183 |
Poverty | +0.173 |
Crit | +0.171 |
Adapt | +0.156 |
Neg logit contributions | |
Kislyak | -0.257 |
================================================================= | -0.256 |
IMAGES | -0.248 |
advertising | -0.242 |
zinski | -0.239 |
Token | ambitious |
Token ID | 14742 |
Feature activation | +8.36e-02 |
Pos logit contributions | |
Kislyak | +0.101 |
================================================================= | +0.101 |
IMAGES | +0.097 |
zinski | +0.095 |
advertising | +0.094 |
Neg logit contributions | |
Crit | -0.098 |
Nit | -0.091 |
Crit | -0.086 |
Poverty | -0.086 |
Adapt | -0.080 |
Token | 24 |
Token ID | 1987 |
Feature activation | +2.37e-01 |
Pos logit contributions | |
Kislyak | +0.241 |
================================================================= | +0.239 |
IMAGES | +0.232 |
advertising | +0.226 |
zinski | +0.225 |
Neg logit contributions | |
Crit | -0.197 |
Nit | -0.180 |
Poverty | -0.170 |
Crit | -0.169 |
Adapt | -0.154 |
Token | - |
Token ID | 12 |
Feature activation | -2.57e-01 |
Pos logit contributions | |
Kislyak | +0.730 |
================================================================= | +0.726 |
IMAGES | +0.702 |
advertising | +0.690 |
zinski | +0.683 |
Neg logit contributions | |
Crit | -0.512 |
Nit | -0.462 |
Poverty | -0.432 |
Crit | -0.430 |
Adapt | -0.390 |
Token | 75 |
Token ID | 2425 |
Feature activation | -4.53e-01 |
Pos logit contributions | |
Crit | +0.644 |
Nit | +0.594 |
Crit | +0.550 |
Poverty | +0.548 |
Adapt | +0.517 |
Neg logit contributions | |
Kislyak | -0.702 |
================================================================= | -0.697 |
IMAGES | -0.671 |
advertising | -0.653 |
zinski | -0.651 |
Token | / |
Token ID | 14 |
Feature activation | -1.81e-01 |
Pos logit contributions | |
Crit | +0.868 |
Nit | +0.771 |
Poverty | +0.713 |
Crit | +0.711 |
Ps | +0.638 |
Neg logit contributions | |
Kislyak | -1.507 |
================================================================= | -1.490 |
IMAGES | -1.438 |
advertising | -1.426 |
zinski | -1.415 |
Token | 1 |
Token ID | 16 |
Feature activation | -3.34e-01 |
Pos logit contributions | |
Crit | +0.425 |
Nit | +0.387 |
Crit | +0.361 |
Poverty | +0.360 |
Adapt | +0.339 |
Neg logit contributions | |
Kislyak | -0.523 |
================================================================= | -0.516 |
IMAGES | -0.498 |
zinski | -0.489 |
DEA | -0.484 |
Token | . |
Token ID | 13 |
Feature activation | -6.43e-02 |
Pos logit contributions | |
Crit | +1.026 |
Nit | +0.955 |
Crit | +0.918 |
Poverty | +0.912 |
Adapt | +0.865 |
Neg logit contributions | |
Kislyak | -0.713 |
================================================================= | -0.706 |
IMAGES | -0.680 |
advertising | -0.662 |
zinski | -0.659 |
Token | 7 |
Token ID | 22 |
Feature activation | -1.66e-01 |
Pos logit contributions | |
Crit | +0.142 |
Nit | +0.129 |
Poverty | +0.121 |
Crit | +0.120 |
Adapt | +0.110 |
Neg logit contributions | |
Kislyak | -0.194 |
================================================================= | -0.192 |
IMAGES | -0.187 |
zinski | -0.182 |
advertising | -0.181 |
Token | - |
Token ID | 12 |
Feature activation | -4.66e-02 |
Pos logit contributions | |
Crit | +0.409 |
Nit | +0.374 |
Crit | +0.352 |
Poverty | +0.351 |
Adapt | +0.326 |
Neg logit contributions | |
Kislyak | -0.458 |
================================================================= | -0.454 |
IMAGES | -0.439 |
advertising | -0.433 |
zinski | -0.427 |
Token | 2 |
Token ID | 17 |
Feature activation | -3.98e-02 |
Pos logit contributions | |
Crit | +0.110 |
Nit | +0.101 |
Poverty | +0.094 |
Crit | +0.094 |
Adapt | +0.087 |
Neg logit contributions | |
Kislyak | -0.134 |
================================================================= | -0.133 |
IMAGES | -0.129 |
advertising | -0.125 |
zinski | -0.124 |
Token | . |
Token ID | 13 |
Feature activation | +1.12e-01 |
Pos logit contributions | |
Crit | +0.122 |
Nit | +0.114 |
Poverty | +0.109 |
Crit | +0.109 |
Adapt | +0.102 |
Neg logit contributions | |
Kislyak | -0.086 |
================================================================= | -0.085 |
IMAGES | -0.081 |
advertising | -0.079 |
zinski | -0.078 |
Token | trade |
Token ID | 25351 |
Feature activation | +2.46e-01 |
Pos logit contributions | |
Crit | +0.239 |
Nit | +0.218 |
Poverty | +0.206 |
Crit | +0.203 |
corruption | +0.188 |
Neg logit contributions | |
Kislyak | -0.283 |
================================================================= | -0.283 |
IMAGES | -0.273 |
advertising | -0.264 |
zinski | -0.264 |
Token | clause |
Token ID | 13444 |
Feature activation | +1.45e-01 |
Pos logit contributions | |
Kislyak | +0.728 |
================================================================= | +0.726 |
IMAGES | +0.703 |
advertising | +0.685 |
zinski | +0.679 |
Neg logit contributions | |
Crit | -0.556 |
Nit | -0.505 |
Poverty | -0.478 |
Crit | -0.470 |
corruption | -0.430 |
Token | in |
Token ID | 287 |
Feature activation | +7.83e-02 |
Pos logit contributions | |
Kislyak | +0.357 |
================================================================= | +0.355 |
IMAGES | +0.342 |
advertising | +0.332 |
zinski | +0.328 |
Neg logit contributions | |
Crit | -0.401 |
Nit | -0.371 |
Poverty | -0.355 |
Crit | -0.350 |
corruption | -0.326 |
Token | his |
Token ID | 465 |
Feature activation | +4.98e-02 |
Pos logit contributions | |
================================================================= | +0.228 |
Kislyak | +0.227 |
IMAGES | +0.221 |
advertising | +0.215 |
zinski | +0.213 |
Neg logit contributions | |
Crit | -0.180 |
Nit | -0.164 |
Poverty | -0.156 |
Crit | -0.155 |
Adapt | -0.140 |
Token | contract |
Token ID | 2775 |
Feature activation | -2.70e-01 |
Pos logit contributions | |
================================================================= | +0.166 |
Kislyak | +0.165 |
IMAGES | +0.162 |
advertising | +0.157 |
zinski | +0.156 |
Neg logit contributions | |
Crit | -0.094 |
Nit | -0.084 |
Poverty | -0.079 |
Crit | -0.078 |
Adapt | -0.069 |
Token | and |
Token ID | 290 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
Crit | +0.752 |
Nit | +0.696 |
Poverty | +0.665 |
Crit | +0.661 |
Ps | +0.612 |
Neg logit contributions | |
Kislyak | -0.654 |
================================================================= | -0.652 |
IMAGES | -0.629 |
advertising | -0.608 |
zinski | -0.603 |
Token | expected |
Token ID | 2938 |
Feature activation | +7.96e-02 |
Pos logit contributions | |
Crit | +0.331 |
Nit | +0.303 |
Poverty | +0.289 |
Crit | +0.288 |
Lumpur | +0.262 |
Neg logit contributions | |
================================================================= | -0.392 |
Kislyak | -0.387 |
IMAGES | -0.376 |
advertising | -0.369 |
Accessory | -0.364 |
Token | to |
Token ID | 284 |
Feature activation | +2.31e-02 |
Pos logit contributions | |
================================================================= | +0.195 |
Kislyak | +0.195 |
IMAGES | +0.188 |
advertising | +0.182 |
zinski | +0.180 |
Neg logit contributions | |
Crit | -0.221 |
Nit | -0.204 |
Poverty | -0.195 |
Crit | -0.194 |
Ps | -0.179 |
Token | be |
Token ID | 307 |
Feature activation | -4.40e-02 |
Pos logit contributions | |
Kislyak | +0.065 |
================================================================= | +0.065 |
IMAGES | +0.063 |
advertising | +0.062 |
Accessory | +0.061 |
Neg logit contributions | |
Crit | -0.055 |
Nit | -0.050 |
Poverty | -0.048 |
Crit | -0.047 |
Adapt | -0.043 |
Token | moved |
Token ID | 3888 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
Crit | +0.118 |
Nit | +0.108 |
Poverty | +0.104 |
Crit | +0.104 |
Lumpur | +0.096 |
Neg logit contributions | |
================================================================= | -0.112 |
Kislyak | -0.111 |
IMAGES | -0.107 |
advertising | -0.105 |
zinski | -0.104 |
Token | . |
Token ID | 13 |
Feature activation | +9.05e-03 |
Pos logit contributions | |
Crit | +0.327 |
Nit | +0.301 |
Poverty | +0.288 |
Crit | +0.286 |
Lumpur | +0.265 |
Neg logit contributions | |
Kislyak | -0.298 |
================================================================= | -0.297 |
IMAGES | -0.286 |
advertising | -0.278 |
zinski | -0.275 |
Token | Press |
Token ID | 4332 |
Feature activation | -2.69e-01 |
Pos logit contributions | |
Crit | +0.440 |
Nit | +0.412 |
Poverty | +0.397 |
Crit | +0.392 |
corruption | +0.371 |
Neg logit contributions | |
Kislyak | -0.264 |
================================================================= | -0.263 |
IMAGES | -0.251 |
advertising | -0.241 |
zinski | -0.238 |
Token | , |
Token ID | 11 |
Feature activation | -1.74e-01 |
Pos logit contributions | |
Crit | +0.776 |
Nit | +0.719 |
Poverty | +0.690 |
Crit | +0.680 |
corruption | +0.638 |
Neg logit contributions | |
Kislyak | -0.631 |
================================================================= | -0.629 |
IMAGES | -0.604 |
advertising | -0.585 |
zinski | -0.578 |
Token | 1994 |
Token ID | 9162 |
Feature activation | -3.08e-01 |
Pos logit contributions | |
Crit | +0.434 |
Nit | +0.395 |
Poverty | +0.381 |
Crit | +0.378 |
Adapt | +0.342 |
Neg logit contributions | |
Kislyak | -0.477 |
================================================================= | -0.472 |
IMAGES | -0.461 |
advertising | -0.449 |
Accessory | -0.447 |
Token | . |
Token ID | 13 |
Feature activation | +7.34e-02 |
Pos logit contributions | |
Crit | +1.018 |
Nit | +0.953 |
Poverty | +0.919 |
Crit | +0.909 |
corruption | +0.860 |
Neg logit contributions | |
Kislyak | -0.595 |
================================================================= | -0.592 |
IMAGES | -0.563 |
advertising | -0.541 |
zinski | -0.533 |
Token |
|
Token ID | 198 |
Feature activation | -1.99e-01 |
Pos logit contributions | |
================================================================= | +0.197 |
Kislyak | +0.197 |
IMAGES | +0.189 |
advertising | +0.183 |
zinski | +0.182 |
Neg logit contributions | |
Crit | -0.187 |
Nit | -0.172 |
Poverty | -0.163 |
Crit | -0.160 |
corruption | -0.150 |
Token |
|
Token ID | 198 |
Feature activation | -1.49e-01 |
Pos logit contributions | |
Crit | +0.474 |
Nit | +0.432 |
Poverty | +0.410 |
Crit | +0.404 |
Adapt | +0.371 |
Neg logit contributions | |
Kislyak | -0.568 |
================================================================= | -0.565 |
IMAGES | -0.549 |
advertising | -0.532 |
zinski | -0.528 |
Token | Ye |
Token ID | 35543 |
Feature activation | -3.19e-01 |
Pos logit contributions | |
Crit | +0.344 |
Nit | +0.312 |
Poverty | +0.287 |
Crit | +0.284 |
Adapt | +0.267 |
Neg logit contributions | |
Kislyak | -0.439 |
================================================================= | -0.438 |
IMAGES | -0.427 |
zinski | -0.409 |
advertising | -0.409 |
Token | ager |
Token ID | 3536 |
Feature activation | +3.48e-02 |
Pos logit contributions | |
Crit | +0.906 |
Nit | +0.833 |
Poverty | +0.804 |
Crit | +0.798 |
corruption | +0.749 |
Neg logit contributions | |
Kislyak | -0.755 |
================================================================= | -0.745 |
IMAGES | -0.738 |
DEA | -0.686 |
Accessory | -0.686 |
Token | , |
Token ID | 11 |
Feature activation | -4.53e-02 |
Pos logit contributions | |
Kislyak | +0.083 |
================================================================= | +0.082 |
IMAGES | +0.080 |
advertising | +0.077 |
Accessory | +0.076 |
Neg logit contributions | |
Crit | -0.099 |
Nit | -0.091 |
Poverty | -0.088 |
Crit | -0.087 |
Adapt | -0.081 |
Token | Chuck |
Token ID | 14663 |
Feature activation | -1.94e-02 |
Pos logit contributions | |
Crit | +0.115 |
Nit | +0.105 |
Poverty | +0.101 |
Crit | +0.100 |
Ps | +0.092 |
Neg logit contributions | |
Kislyak | -0.122 |
================================================================= | -0.120 |
IMAGES | -0.118 |
advertising | -0.114 |
Accessory | -0.114 |
Token | and |
Token ID | 290 |
Feature activation | -2.00e-01 |
Pos logit contributions | |
Crit | +0.059 |
Nit | +0.055 |
Poverty | +0.054 |
Crit | +0.053 |
Adapt | +0.050 |
Neg logit contributions | |
Kislyak | -0.041 |
================================================================= | -0.041 |
IMAGES | -0.040 |
advertising | -0.038 |
Accessory | -0.038 |
Token | downtown |
Token ID | 9436 |
Feature activation | +2.82e-01 |
Pos logit contributions | |
Kislyak | +0.432 |
================================================================= | +0.431 |
IMAGES | +0.415 |
advertising | +0.403 |
zinski | +0.399 |
Neg logit contributions | |
Crit | -0.445 |
Nit | -0.410 |
Poverty | -0.391 |
Crit | -0.387 |
corruption | -0.359 |
Token | news |
Token ID | 1705 |
Feature activation | +3.50e-01 |
Pos logit contributions | |
Kislyak | +0.687 |
================================================================= | +0.685 |
IMAGES | +0.658 |
advertising | +0.638 |
zinski | +0.632 |
Neg logit contributions | |
Crit | -0.789 |
Nit | -0.730 |
Poverty | -0.699 |
Crit | -0.691 |
corruption | -0.644 |
Token | , |
Token ID | 11 |
Feature activation | -1.37e-01 |
Pos logit contributions | |
Kislyak | +0.853 |
================================================================= | +0.850 |
IMAGES | +0.818 |
advertising | +0.793 |
zinski | +0.784 |
Neg logit contributions | |
Crit | -0.977 |
Nit | -0.904 |
Poverty | -0.866 |
Crit | -0.854 |
corruption | -0.798 |
Token | event |
Token ID | 1785 |
Feature activation | -1.60e-01 |
Pos logit contributions | |
Crit | +0.365 |
Nit | +0.335 |
Poverty | +0.320 |
Crit | +0.317 |
corruption | +0.293 |
Neg logit contributions | |
Kislyak | -0.354 |
================================================================= | -0.354 |
IMAGES | -0.339 |
advertising | -0.330 |
zinski | -0.329 |
Token | info |
Token ID | 7508 |
Feature activation | +1.61e-01 |
Pos logit contributions | |
Crit | +0.455 |
Nit | +0.422 |
Poverty | +0.401 |
Crit | +0.400 |
Ps | +0.373 |
Neg logit contributions | |
Kislyak | -0.381 |
================================================================= | -0.379 |
IMAGES | -0.364 |
advertising | -0.353 |
zinski | -0.351 |
Token | , |
Token ID | 11 |
Feature activation | -1.72e-01 |
Pos logit contributions | |
Kislyak | +0.399 |
================================================================= | +0.397 |
IMAGES | +0.383 |
advertising | +0.371 |
zinski | +0.367 |
Neg logit contributions | |
Crit | -0.442 |
Nit | -0.408 |
Poverty | -0.390 |
Crit | -0.385 |
corruption | -0.359 |
Token | hears |
Token ID | 23180 |
Feature activation | -4.32e-01 |
Pos logit contributions | |
Crit | +0.464 |
Nit | +0.425 |
Poverty | +0.407 |
Crit | +0.406 |
crit | +0.372 |
Neg logit contributions | |
Kislyak | -0.439 |
================================================================= | -0.437 |
IMAGES | -0.419 |
zinski | -0.407 |
advertising | -0.406 |
Token | ay |
Token ID | 323 |
Feature activation | +3.09e-01 |
Pos logit contributions | |
Crit | +1.170 |
Nit | +1.078 |
Poverty | +1.029 |
Crit | +1.022 |
corruption | +0.950 |
Neg logit contributions | |
Kislyak | -1.088 |
================================================================= | -1.082 |
IMAGES | -1.060 |
Accessory | -1.012 |
advertising | -1.007 |
Token | , |
Token ID | 11 |
Feature activation | +1.37e-01 |
Pos logit contributions | |
Kislyak | +0.776 |
================================================================= | +0.773 |
IMAGES | +0.745 |
advertising | +0.723 |
zinski | +0.717 |
Neg logit contributions | |
Crit | -0.842 |
Nit | -0.777 |
Poverty | -0.742 |
Crit | -0.734 |
corruption | -0.683 |
Token | tips |
Token ID | 9040 |
Feature activation | -4.92e-02 |
Pos logit contributions | |
Kislyak | +0.352 |
================================================================= | +0.350 |
IMAGES | +0.337 |
advertising | +0.327 |
zinski | +0.325 |
Neg logit contributions | |
Crit | -0.363 |
Nit | -0.334 |
Poverty | -0.319 |
Crit | -0.316 |
corruption | -0.293 |
Token | , |
Token ID | 11 |
Feature activation | +2.40e-01 |
Pos logit contributions | |
Crit | +0.132 |
Nit | +0.121 |
Poverty | +0.116 |
Crit | +0.115 |
corruption | +0.107 |
Neg logit contributions | |
Kislyak | -0.125 |
================================================================= | -0.125 |
IMAGES | -0.121 |
advertising | -0.116 |
Accessory | -0.116 |
Token | float |
Token ID | 12178 |
Feature activation | -3.61e-01 |
Pos logit contributions | |
Crit | +0.277 |
Nit | +0.254 |
Poverty | +0.239 |
Crit | +0.237 |
Adapt | +0.221 |
Neg logit contributions | |
Kislyak | -0.298 |
================================================================= | -0.296 |
IMAGES | -0.285 |
advertising | -0.278 |
zinski | -0.275 |
Token | ambient |
Token ID | 25237 |
Feature activation | +4.85e-01 |
Pos logit contributions | |
Crit | +0.867 |
Nit | +0.790 |
Poverty | +0.751 |
Crit | +0.743 |
Adapt | +0.680 |
Neg logit contributions | |
Kislyak | -1.023 |
================================================================= | -1.014 |
IMAGES | -0.983 |
advertising | -0.969 |
Accessory | -0.956 |
Token | D |
Token ID | 35 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
Kislyak | +1.166 |
================================================================= | +1.164 |
IMAGES | +1.117 |
advertising | +1.083 |
zinski | +1.071 |
Neg logit contributions | |
Crit | -1.359 |
Nit | -1.259 |
Poverty | -1.208 |
Crit | -1.192 |
corruption | -1.116 |
Token | ensity |
Token ID | 6377 |
Feature activation | +3.31e-02 |
Pos logit contributions | |
Kislyak | +0.242 |
================================================================= | +0.239 |
IMAGES | +0.231 |
advertising | +0.225 |
Accessory | +0.219 |
Neg logit contributions | |
Crit | -0.316 |
Nit | -0.293 |
Poverty | -0.282 |
Crit | -0.278 |
Ps | -0.261 |
Token | ) |
Token ID | 1267 |
Feature activation | +1.87e-01 |
Pos logit contributions | |
Kislyak | +0.070 |
================================================================= | +0.069 |
IMAGES | +0.067 |
advertising | +0.065 |
zinski | +0.063 |
Neg logit contributions | |
Crit | -0.104 |
Nit | -0.097 |
Poverty | -0.093 |
Crit | -0.092 |
Adapt | -0.087 |
Token | const |
Token ID | 1500 |
Feature activation | -9.98e-01 |
Pos logit contributions | |
Kislyak | +0.514 |
================================================================= | +0.508 |
IMAGES | +0.495 |
advertising | +0.482 |
zinski | +0.478 |
Neg logit contributions | |
Crit | -0.466 |
Nit | -0.426 |
Poverty | -0.405 |
Crit | -0.403 |
Adapt | -0.369 |
Token | { |
Token ID | 1391 |
Feature activation | -7.41e-01 |
Pos logit contributions | |
Crit | +2.300 |
Nit | +2.079 |
Crit | +2.008 |
Poverty | +1.960 |
Ps | +1.816 |
Neg logit contributions | |
Kislyak | -2.844 |
================================================================= | -2.781 |
IMAGES | -2.755 |
advertising | -2.751 |
zinski | -2.743 |
Token | const |
Token ID | 1500 |
Feature activation | -1.16e+00 |
Pos logit contributions | |
Crit | +1.874 |
Nit | +1.722 |
Crit | +1.634 |
Poverty | +1.625 |
Adapt | +1.498 |
Neg logit contributions | |
Kislyak | -2.004 |
================================================================= | -1.970 |
IMAGES | -1.910 |
Accessory | -1.877 |
advertising | -1.876 |
Token | float |
Token ID | 12178 |
Feature activation | -8.74e-01 |
Pos logit contributions | |
Crit | +2.758 |
Nit | +2.527 |
Crit | +2.411 |
Poverty | +2.396 |
crit | +2.191 |
Neg logit contributions | |
Kislyak | -3.272 |
================================================================= | -3.202 |
advertising | -3.142 |
Accessory | -3.103 |
IMAGES | -3.100 |
Token | f |
Token ID | 277 |
Feature activation | +1.53e-01 |
Pos logit contributions | |
Crit | +2.174 |
Nit | +1.981 |
Poverty | +1.888 |
Crit | +1.871 |
Adapt | +1.715 |
Neg logit contributions | |
Kislyak | -2.407 |
================================================================= | -2.372 |
IMAGES | -2.304 |
advertising | -2.285 |
Accessory | -2.249 |
Token | Mass |
Token ID | 20273 |
Feature activation | -1.62e-01 |
Pos logit contributions | |
Kislyak | +0.413 |
================================================================= | +0.412 |
IMAGES | +0.398 |
advertising | +0.387 |
zinski | +0.383 |
Neg logit contributions | |
Crit | -0.386 |
Nit | -0.354 |
Poverty | -0.338 |
Crit | -0.332 |
corruption | -0.308 |
Token | & |
Token ID | 1222 |
Feature activation | +1.70e-01 |
Pos logit contributions | |
Crit | +0.215 |
Nit | +0.197 |
Poverty | +0.188 |
Crit | +0.187 |
corruption | +0.173 |
Neg logit contributions | |
Kislyak | -0.217 |
================================================================= | -0.215 |
IMAGES | -0.207 |
advertising | -0.203 |
zinski | -0.201 |
Token | mut |
Token ID | 4517 |
Feature activation | -2.85e-01 |
Pos logit contributions | |
Kislyak | +0.539 |
================================================================= | +0.539 |
IMAGES | +0.523 |
advertising | +0.509 |
zinski | +0.505 |
Neg logit contributions | |
Crit | -0.350 |
Nit | -0.315 |
Poverty | -0.296 |
Crit | -0.289 |
corruption | -0.263 |
Token | self |
Token ID | 2116 |
Feature activation | -9.10e-01 |
Pos logit contributions | |
Crit | +0.535 |
Nit | +0.470 |
Crit | +0.449 |
Poverty | +0.445 |
crit | +0.396 |
Neg logit contributions | |
Kislyak | -0.951 |
================================================================= | -0.944 |
IMAGES | -0.917 |
advertising | -0.914 |
zinski | -0.903 |
Token | ) |
Token ID | 1267 |
Feature activation | -2.57e-03 |
Pos logit contributions | |
Crit | +2.916 |
Nit | +2.712 |
Poverty | +2.603 |
Crit | +2.591 |
corruption | +2.451 |
Neg logit contributions | |
Kislyak | -1.842 |
================================================================= | -1.807 |
IMAGES | -1.729 |
advertising | -1.688 |
zinski | -1.679 |
Token | { |
Token ID | 1391 |
Feature activation | -5.14e-01 |
Pos logit contributions | |
Crit | +0.008 |
Nit | +0.008 |
Poverty | +0.008 |
Crit | +0.007 |
Lumpur | +0.007 |
Neg logit contributions | |
================================================================= | -0.005 |
Kislyak | -0.005 |
IMAGES | -0.005 |
advertising | -0.005 |
zinski | -0.004 |
Token | let |
Token ID | 1309 |
Feature activation | -1.01e+00 |
Pos logit contributions | |
Crit | +1.247 |
Nit | +1.132 |
Poverty | +1.079 |
Crit | +1.079 |
Adapt | +0.980 |
Neg logit contributions | |
Kislyak | -1.443 |
================================================================= | -1.421 |
IMAGES | -1.380 |
advertising | -1.364 |
zinski | -1.350 |
Token | prev |
Token ID | 8654 |
Feature activation | +2.69e-01 |
Pos logit contributions | |
Crit | +2.263 |
Nit | +2.030 |
Crit | +1.956 |
Poverty | +1.948 |
crit | +1.802 |
Neg logit contributions | |
Kislyak | -2.972 |
================================================================= | -2.908 |
advertising | -2.860 |
IMAGES | -2.816 |
zinski | -2.798 |
Token | _ |
Token ID | 62 |
Feature activation | -3.98e-01 |
Pos logit contributions | |
================================================================= | +0.662 |
Kislyak | +0.655 |
IMAGES | +0.631 |
advertising | +0.613 |
Accessory | +0.608 |
Neg logit contributions | |
Crit | -0.740 |
Nit | -0.688 |
Poverty | -0.662 |
Crit | -0.650 |
Lumpur | -0.610 |
Token | state |
Token ID | 5219 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
Crit | +1.101 |
Nit | +1.014 |
Poverty | +0.968 |
Crit | +0.960 |
corruption | +0.904 |
Neg logit contributions | |
Kislyak | -0.981 |
================================================================= | -0.972 |
IMAGES | -0.937 |
advertising | -0.906 |
zinski | -0.899 |
Token | = |
Token ID | 796 |
Feature activation | -2.42e-01 |
Pos logit contributions | |
Kislyak | +0.247 |
================================================================= | +0.243 |
IMAGES | +0.234 |
advertising | +0.228 |
zinski | +0.226 |
Neg logit contributions | |
Crit | -0.312 |
Nit | -0.289 |
Poverty | -0.277 |
Crit | -0.274 |
Ps | -0.257 |
Token | self |
Token ID | 2116 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
Crit | +0.611 |
Nit | +0.554 |
Crit | +0.536 |
Poverty | +0.534 |
Adapt | +0.488 |
Neg logit contributions | |
Kislyak | -0.651 |
================================================================= | -0.643 |
IMAGES | -0.622 |
advertising | -0.620 |
zinski | -0.612 |
Token | response |
Token ID | 2882 |
Feature activation | +7.97e-03 |
Pos logit contributions | |
Crit | +0.834 |
Nit | +0.753 |
Poverty | +0.715 |
Crit | +0.710 |
Adapt | +0.642 |
Neg logit contributions | |
Kislyak | -1.166 |
================================================================= | -1.151 |
IMAGES | -1.125 |
advertising | -1.097 |
Accessory | -1.088 |
Token | plan |
Token ID | 1410 |
Feature activation | -4.73e-01 |
Pos logit contributions | |
Kislyak | +0.026 |
================================================================= | +0.025 |
IMAGES | +0.025 |
advertising | +0.024 |
zinski | +0.024 |
Neg logit contributions | |
Crit | -0.016 |
Nit | -0.014 |
Poverty | -0.013 |
Crit | -0.013 |
Ps | -0.012 |
Token | for |
Token ID | 329 |
Feature activation | +1.87e-02 |
Pos logit contributions | |
Crit | +1.395 |
Nit | +1.286 |
Poverty | +1.240 |
Crit | +1.236 |
Adapt | +1.150 |
Neg logit contributions | |
Kislyak | -1.070 |
================================================================= | -1.061 |
IMAGES | -1.037 |
advertising | -0.994 |
zinski | -0.984 |
Token | the |
Token ID | 262 |
Feature activation | +1.93e-02 |
Pos logit contributions | |
Kislyak | +0.050 |
================================================================= | +0.050 |
IMAGES | +0.048 |
advertising | +0.047 |
zinski | +0.046 |
Neg logit contributions | |
Crit | -0.048 |
Nit | -0.044 |
Poverty | -0.042 |
Crit | -0.042 |
corruption | -0.038 |
Token | Ch |
Token ID | 609 |
Feature activation | +1.82e-01 |
Pos logit contributions | |
================================================================= | +0.052 |
Kislyak | +0.052 |
IMAGES | +0.050 |
advertising | +0.050 |
Accessory | +0.049 |
Neg logit contributions | |
Crit | -0.049 |
Nit | -0.044 |
Poverty | -0.043 |
Crit | -0.043 |
Ps | -0.039 |
Token | uk |
Token ID | 2724 |
Feature activation | -9.12e-01 |
Pos logit contributions | |
Kislyak | +0.510 |
================================================================= | +0.508 |
IMAGES | +0.492 |
advertising | +0.479 |
zinski | +0.474 |
Neg logit contributions | |
Crit | -0.442 |
Nit | -0.404 |
Poverty | -0.384 |
Crit | -0.378 |
corruption | -0.349 |
Token | chi |
Token ID | 11072 |
Feature activation | +1.08e-01 |
Pos logit contributions | |
Crit | +2.688 |
Nit | +2.487 |
Poverty | +2.399 |
Crit | +2.370 |
Ps | +2.235 |
Neg logit contributions | |
================================================================= | -2.077 |
Kislyak | -2.074 |
IMAGES | -2.004 |
advertising | -1.941 |
Accessory | -1.912 |
Token | Sea |
Token ID | 6896 |
Feature activation | +9.52e-02 |
Pos logit contributions | |
================================================================= | +0.227 |
Kislyak | +0.227 |
IMAGES | +0.215 |
advertising | +0.207 |
zinski | +0.204 |
Neg logit contributions | |
Crit | -0.340 |
Nit | -0.318 |
Poverty | -0.306 |
Crit | -0.301 |
Adapt | -0.285 |
Token | . |
Token ID | 13 |
Feature activation | -2.13e-02 |
Pos logit contributions | |
Kislyak | +0.223 |
================================================================= | +0.222 |
IMAGES | +0.213 |
advertising | +0.206 |
zinski | +0.204 |
Neg logit contributions | |
Crit | -0.275 |
Nit | -0.255 |
Poverty | -0.244 |
Crit | -0.241 |
Adapt | -0.226 |
Token | Alaska |
Token ID | 12926 |
Feature activation | -2.44e-02 |
Pos logit contributions | |
Crit | +0.053 |
Nit | +0.048 |
Poverty | +0.046 |
Crit | +0.046 |
Adapt | +0.041 |
Neg logit contributions | |
Kislyak | -0.059 |
================================================================= | -0.058 |
IMAGES | -0.057 |
advertising | -0.055 |
zinski | -0.055 |
Token | is |
Token ID | 318 |
Feature activation | +4.00e-02 |
Pos logit contributions | |
Crit | +0.062 |
Nit | +0.057 |
Poverty | +0.054 |
Crit | +0.053 |
Adapt | +0.049 |
Neg logit contributions | |
Kislyak | -0.066 |
================================================================= | -0.065 |
IMAGES | -0.063 |
advertising | -0.062 |
zinski | -0.061 |
Token | Follow |
Token ID | 7281 |
Feature activation | +8.44e-02 |
Pos logit contributions | |
================================================================= | +0.914 |
Kislyak | +0.912 |
IMAGES | +0.875 |
advertising | +0.831 |
DEA | +0.815 |
Neg logit contributions | |
Crit | -1.544 |
Nit | -1.438 |
Poverty | -1.389 |
Crit | -1.362 |
corruption | -1.305 |
Token | him |
Token ID | 683 |
Feature activation | +2.65e-02 |
Pos logit contributions | |
Kislyak | +0.249 |
================================================================= | +0.248 |
IMAGES | +0.240 |
advertising | +0.234 |
zinski | +0.232 |
Neg logit contributions | |
Crit | -0.193 |
Nit | -0.175 |
Poverty | -0.166 |
Crit | -0.163 |
corruption | -0.149 |
Token | on |
Token ID | 319 |
Feature activation | -7.37e-02 |
Pos logit contributions | |
================================================================= | +0.069 |
Kislyak | +0.069 |
IMAGES | +0.067 |
advertising | +0.065 |
zinski | +0.064 |
Neg logit contributions | |
Crit | -0.069 |
Nit | -0.063 |
Poverty | -0.061 |
Crit | -0.060 |
corruption | -0.055 |
Token | Twitter |
Token ID | 3009 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
Crit | +0.201 |
Nit | +0.186 |
Poverty | +0.178 |
Crit | +0.175 |
corruption | +0.163 |
Neg logit contributions | |
Kislyak | -0.185 |
================================================================= | -0.184 |
IMAGES | -0.177 |
advertising | -0.172 |
zinski | -0.170 |
Token | @ |
Token ID | 2488 |
Feature activation | -8.91e-02 |
Pos logit contributions | |
Kislyak | +0.342 |
================================================================= | +0.339 |
IMAGES | +0.331 |
advertising | +0.323 |
zinski | +0.322 |
Neg logit contributions | |
Crit | -0.212 |
Nit | -0.191 |
Poverty | -0.178 |
Crit | -0.176 |
Adapt | -0.159 |
Token | jon |
Token ID | 46286 |
Feature activation | -1.08e+00 |
Pos logit contributions | |
Crit | +0.192 |
Nit | +0.175 |
Poverty | +0.160 |
Crit | +0.158 |
Adapt | +0.148 |
Neg logit contributions | |
Kislyak | -0.276 |
================================================================= | -0.276 |
IMAGES | -0.269 |
advertising | -0.258 |
Accessory | -0.257 |
Token | chest |
Token ID | 46713 |
Feature activation | -2.47e-01 |
Pos logit contributions | |
Crit | +2.767 |
Nit | +2.515 |
Poverty | +2.364 |
Crit | +2.360 |
corruption | +2.293 |
Neg logit contributions | |
Kislyak | -2.924 |
================================================================= | -2.836 |
IMAGES | -2.813 |
Accessory | -2.657 |
DEA | -2.654 |
Token | o |
Token ID | 78 |
Feature activation | +5.37e-02 |
Pos logit contributions | |
Crit | +0.735 |
Nit | +0.685 |
Poverty | +0.655 |
Crit | +0.649 |
corruption | +0.611 |
Neg logit contributions | |
Kislyak | -0.561 |
================================================================= | -0.551 |
IMAGES | -0.537 |
Accessory | -0.512 |
advertising | -0.507 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +1.28e-01 |
Pos logit contributions | |
Kislyak | +0.134 |
================================================================= | +0.133 |
IMAGES | +0.128 |
advertising | +0.124 |
zinski | +0.123 |
Neg logit contributions | |
Crit | -0.147 |
Nit | -0.136 |
Poverty | -0.130 |
Crit | -0.128 |
Adapt | -0.120 |
Token | An |
Token ID | 2025 |
Feature activation | -2.70e-01 |
Pos logit contributions | |
Kislyak | +0.381 |
================================================================= | +0.379 |
IMAGES | +0.368 |
advertising | +0.355 |
zinski | +0.355 |
Neg logit contributions | |
Crit | -0.293 |
Nit | -0.266 |
Poverty | -0.248 |
Crit | -0.245 |
Adapt | -0.227 |
Token | Open |
Token ID | 4946 |
Feature activation | +4.66e-02 |
Pos logit contributions | |
Crit | +0.714 |
Nit | +0.658 |
Crit | +0.634 |
Poverty | +0.633 |
Adapt | +0.585 |
Neg logit contributions | |
Kislyak | -0.687 |
================================================================= | -0.687 |
IMAGES | -0.664 |
zinski | -0.640 |
Accessory | -0.639 |
Token | for |
Token ID | 329 |
Feature activation | -1.27e-01 |
Pos logit contributions | |
Crit | +0.492 |
Nit | +0.453 |
Poverty | +0.435 |
Crit | +0.430 |
corruption | +0.398 |
Neg logit contributions | |
Kislyak | -0.454 |
================================================================= | -0.452 |
IMAGES | -0.437 |
advertising | -0.425 |
zinski | -0.421 |
Token | a |
Token ID | 257 |
Feature activation | +4.20e-02 |
Pos logit contributions | |
Crit | +0.324 |
Nit | +0.296 |
Poverty | +0.288 |
Crit | +0.283 |
Adapt | +0.259 |
Neg logit contributions | |
Kislyak | -0.340 |
================================================================= | -0.340 |
IMAGES | -0.328 |
advertising | -0.320 |
zinski | -0.317 |
Token | free |
Token ID | 1479 |
Feature activation | +3.78e-01 |
Pos logit contributions | |
Kislyak | +0.114 |
================================================================= | +0.114 |
IMAGES | +0.110 |
advertising | +0.108 |
Accessory | +0.107 |
Neg logit contributions | |
Crit | -0.105 |
Nit | -0.096 |
Poverty | -0.094 |
Crit | -0.092 |
Adapt | -0.083 |
Token | year |
Token ID | 614 |
Feature activation | +7.74e-02 |
Pos logit contributions | |
Kislyak | +1.010 |
================================================================= | +1.006 |
IMAGES | +0.973 |
advertising | +0.947 |
zinski | +0.937 |
Neg logit contributions | |
Crit | -0.968 |
Nit | -0.889 |
Poverty | -0.853 |
Crit | -0.841 |
Adapt | -0.774 |
Token | long |
Token ID | 6511 |
Feature activation | -3.40e-01 |
Pos logit contributions | |
Kislyak | +0.195 |
================================================================= | +0.195 |
IMAGES | +0.188 |
advertising | +0.181 |
zinski | +0.180 |
Neg logit contributions | |
Crit | -0.210 |
Nit | -0.193 |
Poverty | -0.185 |
Crit | -0.183 |
corruption | -0.170 |
Token | pass |
Token ID | 1208 |
Feature activation | -9.21e-01 |
Pos logit contributions | |
Crit | +0.830 |
Nit | +0.756 |
Poverty | +0.743 |
Crit | +0.730 |
crit | +0.662 |
Neg logit contributions | |
Kislyak | -0.938 |
================================================================= | -0.936 |
IMAGES | -0.907 |
advertising | -0.885 |
Accessory | -0.878 |
Token | to |
Token ID | 284 |
Feature activation | +4.82e-02 |
Pos logit contributions | |
Crit | +2.408 |
Nit | +2.207 |
Poverty | +2.140 |
Crit | +2.117 |
crit | +1.960 |
Neg logit contributions | |
Kislyak | -2.383 |
================================================================= | -2.338 |
IMAGES | -2.298 |
advertising | -2.185 |
zinski | -2.184 |
Token | federal |
Token ID | 2717 |
Feature activation | -7.05e-02 |
Pos logit contributions | |
Kislyak | +0.136 |
================================================================= | +0.136 |
IMAGES | +0.132 |
advertising | +0.129 |
zinski | +0.127 |
Neg logit contributions | |
Crit | -0.115 |
Nit | -0.105 |
Poverty | -0.101 |
Crit | -0.099 |
Adapt | -0.091 |
Token | public |
Token ID | 1171 |
Feature activation | +2.49e-02 |
Pos logit contributions | |
Crit | +0.180 |
Nit | +0.164 |
Poverty | +0.161 |
Crit | +0.157 |
crit | +0.144 |
Neg logit contributions | |
Kislyak | -0.189 |
================================================================= | -0.188 |
IMAGES | -0.183 |
zinski | -0.177 |
advertising | -0.177 |
Token | lands |
Token ID | 8604 |
Feature activation | -1.88e-01 |
Pos logit contributions | |
================================================================= | +0.068 |
Kislyak | +0.068 |
IMAGES | +0.066 |
zinski | +0.064 |
advertising | +0.064 |
Neg logit contributions | |
Crit | -0.062 |
Nit | -0.057 |
Poverty | -0.056 |
Crit | -0.054 |
crit | -0.050 |
Token | , |
Token ID | 11 |
Feature activation | -4.56e-02 |
Pos logit contributions | |
Crit | +0.537 |
Nit | +0.496 |
Poverty | +0.479 |
Crit | +0.473 |
corruption | +0.441 |
Neg logit contributions | |
Kislyak | -0.449 |
================================================================= | -0.444 |
IMAGES | -0.430 |
advertising | -0.415 |
zinski | -0.412 |