Token | . |
Token ID | 13 |
Feature activation | +5.57e+00 |
Pos logit contributions | |
atography | +0.286 |
IDA | +0.273 |
Ru | +0.271 |
irable | +0.263 |
wrongdoing | +0.261 |
Neg logit contributions | |
weather | -0.338 |
pite | -0.288 |
ritical | -0.276 |
usage | -0.274 |
Hurricane | -0.273 |
Token |
|
Token ID | 198 |
Feature activation | +7.98e+00 |
Pos logit contributions | |
atography | +3.218 |
IDA | +3.063 |
Ru | +3.036 |
kaya | +2.983 |
wrongdoing | +2.957 |
Neg logit contributions | |
weather | -4.741 |
pite | -4.144 |
ritical | -3.973 |
usage | -3.966 |
Hurricane | -3.836 |
Token |
|
Token ID | 198 |
Feature activation | +5.13e+00 |
Pos logit contributions | |
atography | +3.763 |
Ru | +3.748 |
IDA | +3.710 |
kaya | +3.556 |
wrongdoing | +3.547 |
Neg logit contributions | |
weather | -7.294 |
pite | -6.575 |
ritical | -6.504 |
usage | -6.301 |
reconc | -6.184 |
Token | ( |
Token ID | 7 |
Feature activation | +7.43e+00 |
Pos logit contributions | |
atography | +3.052 |
IDA | +2.950 |
Ru | +2.918 |
kaya | +2.830 |
wrongdoing | +2.823 |
Neg logit contributions | |
weather | -4.313 |
pite | -3.763 |
ritical | -3.634 |
usage | -3.597 |
Hurricane | -3.534 |
Token | Reporting |
Token ID | 42159 |
Feature activation | +9.43e+00 |
Pos logit contributions | |
atography | +4.460 |
IDA | +4.350 |
Ru | +4.295 |
kaya | +4.134 |
wrongdoing | +4.102 |
Neg logit contributions | |
weather | -6.121 |
pite | -5.377 |
ritical | -5.133 |
usage | -5.110 |
Hurricane | -5.033 |
Token | by |
Token ID | 416 |
Feature activation | +2.25e+01 |
Pos logit contributions | |
atography | +4.487 |
IDA | +4.261 |
Ru | +4.260 |
wrongdoing | +4.195 |
Marina | +4.171 |
Neg logit contributions | |
weather | -8.856 |
pite | -7.887 |
usage | -7.568 |
ritical | -7.563 |
Hurricane | -7.242 |
Token | Z |
Token ID | 1168 |
Feature activation | +1.43e+01 |
Pos logit contributions | |
Marina | +12.513 |
atography | +11.363 |
Transparency | +11.318 |
Ru | +11.102 |
IDA | +10.729 |
Neg logit contributions | |
weather | -15.563 |
pite | -13.876 |
ritical | -13.578 |
usage | -13.374 |
ricanes | -12.403 |
Token | ey |
Token ID | 2959 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
atography | +9.189 |
IDA | +9.013 |
Ru | +8.776 |
kaya | +8.649 |
vana | +8.558 |
Neg logit contributions | |
weather | -9.914 |
pite | -8.198 |
Hurricane | -8.159 |
usage | -8.042 |
ritical | -7.963 |
Token | ad |
Token ID | 324 |
Feature activation | +1.64e+00 |
Pos logit contributions | |
atography | +3.521 |
IDA | +3.371 |
Ru | +3.319 |
wrongdoing | +3.257 |
kaya | +3.246 |
Neg logit contributions | |
weather | -4.766 |
pite | -4.141 |
ritical | -3.989 |
usage | -3.967 |
Hurricane | -3.867 |
Token | Mas |
Token ID | 11066 |
Feature activation | +2.85e+00 |
Pos logit contributions | |
atography | +1.052 |
IDA | +1.013 |
Ru | +0.989 |
wrongdoing | +0.982 |
irable | +0.963 |
Neg logit contributions | |
weather | -1.326 |
pite | -1.149 |
ritical | -1.104 |
usage | -1.089 |
Hurricane | -1.073 |
Token | ro |
Token ID | 305 |
Feature activation | +4.66e+00 |
Pos logit contributions | |
atography | +1.821 |
IDA | +1.758 |
Ru | +1.717 |
wrongdoing | +1.716 |
kaya | +1.660 |
Neg logit contributions | |
weather | -2.319 |
pite | -2.020 |
ritical | -1.938 |
usage | -1.927 |
Hurricane | -1.855 |
Token | ro |
Token ID | 305 |
Feature activation | +4.66e+00 |
Pos logit contributions | |
atography | +1.821 |
IDA | +1.758 |
Ru | +1.717 |
wrongdoing | +1.716 |
kaya | +1.660 |
Neg logit contributions | |
weather | -2.319 |
pite | -2.020 |
ritical | -1.938 |
usage | -1.927 |
Hurricane | -1.855 |
Token | or |
Token ID | 273 |
Feature activation | +2.60e+00 |
Pos logit contributions | |
atography | +3.417 |
IDA | +3.287 |
Ru | +3.229 |
wrongdoing | +3.205 |
kaya | +3.166 |
Neg logit contributions | |
weather | -3.363 |
pite | -2.834 |
ritical | -2.717 |
usage | -2.699 |
Hurricane | -2.600 |
Token | Khan |
Token ID | 11356 |
Feature activation | +6.02e+00 |
Pos logit contributions | |
atography | +1.719 |
IDA | +1.647 |
Ru | +1.599 |
wrongdoing | +1.596 |
irable | +1.583 |
Neg logit contributions | |
weather | -2.078 |
pite | -1.775 |
ritical | -1.703 |
usage | -1.690 |
Hurricane | -1.657 |
Token | ; |
Token ID | 26 |
Feature activation | +1.88e+01 |
Pos logit contributions | |
atography | +3.423 |
IDA | +3.264 |
Ru | +3.212 |
wrongdoing | +3.148 |
kaya | +3.140 |
Neg logit contributions | |
weather | -5.261 |
pite | -4.609 |
ritical | -4.449 |
usage | -4.425 |
Hurricane | -4.319 |
Token | Writing |
Token ID | 22183 |
Feature activation | +1.49e+01 |
Pos logit contributions | |
atography | +7.033 |
Marina | +6.492 |
IDA | +6.393 |
enrich | +6.260 |
wrongdoing | +6.230 |
Neg logit contributions | |
weather | -17.582 |
pite | -15.976 |
ritical | -14.886 |
usage | -14.608 |
Hurricane | -14.532 |
Token | by |
Token ID | 416 |
Feature activation | +2.23e+01 |
Pos logit contributions | |
Marina | +5.525 |
atography | +5.517 |
wrongdoing | +5.157 |
Ru | +5.099 |
Transparency | +4.889 |
Neg logit contributions | |
weather | -14.839 |
pite | -13.391 |
ritical | -12.964 |
usage | -12.892 |
reconc | -12.143 |
Token | Z |
Token ID | 1168 |
Feature activation | +1.75e+01 |
Pos logit contributions | |
Marina | +12.496 |
atography | +11.332 |
Transparency | +11.271 |
Ru | +11.071 |
IDA | +10.663 |
Neg logit contributions | |
weather | -15.564 |
pite | -13.967 |
ritical | -13.639 |
usage | -13.427 |
ricanes | -12.338 |
Token | eb |
Token ID | 1765 |
Feature activation | +4.84e+00 |
Pos logit contributions | |
atography | +10.864 |
IDA | +10.452 |
kaya | +10.375 |
Ru | +10.188 |
Marina | +10.001 |
Neg logit contributions | |
weather | -11.857 |
usage | -9.749 |
Hurricane | -9.684 |
pite | -9.604 |
ritical | -9.369 |
Token | a |
Token ID | 64 |
Feature activation | +2.10e+00 |
Pos logit contributions | |
atography | +2.990 |
IDA | +2.864 |
Ru | +2.819 |
wrongdoing | +2.770 |
kaya | +2.756 |
Neg logit contributions | |
weather | -3.998 |
pite | -3.470 |
ritical | -3.343 |
usage | -3.324 |
Hurricane | -3.239 |
Token | Sidd |
Token ID | 44487 |
Feature activation | +4.44e+00 |
Pos logit contributions | |
atography | +1.321 |
IDA | +1.268 |
Ru | +1.241 |
wrongdoing | +1.240 |
irable | +1.212 |
Neg logit contributions | |
weather | -1.715 |
pite | -1.478 |
ritical | -1.425 |
usage | -1.419 |
Hurricane | -1.392 |
Token | iqu |
Token ID | 1557 |
Feature activation | -1.49e+00 |
Pos logit contributions | |
atography | +2.918 |
IDA | +2.821 |
Ru | +2.765 |
wrongdoing | +2.725 |
kaya | +2.701 |
Neg logit contributions | |
weather | -3.502 |
pite | -3.019 |
ritical | -2.890 |
usage | -2.881 |
Hurricane | -2.794 |
Token | 28 |
Token ID | 2579 |
Feature activation | +9.49e+00 |
Pos logit contributions | |
atography | +3.226 |
IDA | +3.139 |
Ru | +3.097 |
kaya | +3.052 |
wrongdoing | +3.046 |
Neg logit contributions | |
weather | -4.042 |
pite | -3.519 |
ritical | -3.441 |
usage | -3.374 |
Hurricane | -3.226 |
Token | , |
Token ID | 11 |
Feature activation | +1.07e+01 |
Pos logit contributions | |
atography | +5.325 |
IDA | +5.281 |
Ru | +5.170 |
wrongdoing | +5.140 |
kaya | +5.089 |
Neg logit contributions | |
weather | -7.829 |
ritical | -6.780 |
pite | -6.779 |
usage | -6.595 |
Hurricane | -6.187 |
Token | 2014 |
Token ID | 1946 |
Feature activation | +5.34e+00 |
Pos logit contributions | |
atography | +5.118 |
IDA | +4.977 |
Ru | +4.843 |
kaya | +4.789 |
wrongdoing | +4.755 |
Neg logit contributions | |
weather | -9.881 |
pite | -8.583 |
ritical | -8.518 |
usage | -8.453 |
Hurricane | -8.059 |
Token | . |
Token ID | 13 |
Feature activation | +1.42e+01 |
Pos logit contributions | |
atography | +3.393 |
IDA | +3.286 |
Ru | +3.230 |
wrongdoing | +3.179 |
kaya | +3.146 |
Neg logit contributions | |
weather | -4.255 |
pite | -3.691 |
ritical | -3.568 |
usage | -3.514 |
Hurricane | -3.398 |
Token | REUTERS |
Token ID | 15862 |
Feature activation | +1.41e+01 |
Pos logit contributions | |
atography | +5.676 |
IDA | +5.617 |
Ru | +5.585 |
Marina | +5.484 |
Transparency | +5.304 |
Neg logit contributions | |
weather | -13.510 |
pite | -12.127 |
ritical | -12.016 |
usage | -11.851 |
sled | -11.152 |
Token | / |
Token ID | 14 |
Feature activation | +2.23e+01 |
Pos logit contributions | |
IDA | +6.255 |
Ru | +6.228 |
Marina | +5.881 |
atography | +5.797 |
wrongdoing | +5.684 |
Neg logit contributions | |
weather | -12.711 |
pite | -11.603 |
ritical | -11.575 |
usage | -11.154 |
reconc | -10.979 |
Token | Michael |
Token ID | 13256 |
Feature activation | +3.16e+00 |
Pos logit contributions | |
Ru | +10.362 |
IDA | +9.564 |
Step | +9.051 |
Marina | +8.917 |
Sus | +8.602 |
Neg logit contributions | |
weather | -18.574 |
pite | -16.965 |
ritical | -16.520 |
reconc | -16.390 |
usage | -16.171 |
Token | K |
Token ID | 509 |
Feature activation | -6.53e-01 |
Pos logit contributions | |
atography | +1.866 |
IDA | +1.773 |
Ru | +1.747 |
wrongdoing | +1.722 |
kaya | +1.705 |
Neg logit contributions | |
weather | -2.716 |
pite | -2.373 |
ritical | -2.276 |
usage | -2.264 |
Hurricane | -2.222 |
Token | lim |
Token ID | 2475 |
Feature activation | +5.04e-01 |
Pos logit contributions | |
weather | +0.369 |
pite | +0.302 |
ritical | +0.282 |
usage | +0.274 |
Hurricane | +0.263 |
Neg logit contributions | |
atography | -0.573 |
IDA | -0.555 |
Ru | -0.555 |
wrongdoing | -0.551 |
kaya | -0.549 |
Token | enty |
Token ID | 3787 |
Feature activation | +4.66e+00 |
Pos logit contributions | |
IDA | +0.328 |
atography | +0.328 |
Ru | +0.322 |
wrongdoing | +0.320 |
enrich | +0.309 |
Neg logit contributions | |
weather | -0.395 |
pite | -0.343 |
usage | -0.322 |
ritical | -0.319 |
Hurricane | -0.306 |
Token | ev |
Token ID | 1990 |
Feature activation | +1.09e+01 |
Pos logit contributions | |
atography | +2.059 |
IDA | +1.937 |
Ru | +1.899 |
wrongdoing | +1.842 |
kaya | +1.837 |
Neg logit contributions | |
weather | -4.662 |
pite | -4.150 |
ritical | -4.030 |
usage | -4.017 |
Hurricane | -3.935 |
Token | Sidd |
Token ID | 44487 |
Feature activation | +4.44e+00 |
Pos logit contributions | |
atography | +1.321 |
IDA | +1.268 |
Ru | +1.241 |
wrongdoing | +1.240 |
irable | +1.212 |
Neg logit contributions | |
weather | -1.715 |
pite | -1.478 |
ritical | -1.425 |
usage | -1.419 |
Hurricane | -1.392 |
Token | iqu |
Token ID | 1557 |
Feature activation | -1.49e+00 |
Pos logit contributions | |
atography | +2.918 |
IDA | +2.821 |
Ru | +2.765 |
wrongdoing | +2.725 |
kaya | +2.701 |
Neg logit contributions | |
weather | -3.502 |
pite | -3.019 |
ritical | -2.890 |
usage | -2.881 |
Hurricane | -2.794 |
Token | i |
Token ID | 72 |
Feature activation | +5.51e+00 |
Pos logit contributions | |
weather | +1.485 |
pite | +1.312 |
ritical | +1.287 |
usage | +1.250 |
Hurricane | +1.220 |
Neg logit contributions | |
atography | -0.670 |
IDA | -0.642 |
Ru | -0.639 |
wrongdoing | -0.634 |
Transparency | -0.614 |
Token | ; |
Token ID | 26 |
Feature activation | +1.92e+01 |
Pos logit contributions | |
atography | +3.148 |
IDA | +3.008 |
Ru | +2.950 |
wrongdoing | +2.897 |
kaya | +2.875 |
Neg logit contributions | |
weather | -4.824 |
pite | -4.215 |
ritical | -4.070 |
usage | -4.051 |
Hurricane | -3.954 |
Token | Editing |
Token ID | 39883 |
Feature activation | +1.28e+01 |
Pos logit contributions | |
atography | +6.323 |
Marina | +6.002 |
IDA | +5.726 |
enrich | +5.544 |
wrongdoing | +5.438 |
Neg logit contributions | |
weather | -18.821 |
pite | -17.398 |
ritical | -16.270 |
usage | -15.941 |
awks | -15.923 |
Token | by |
Token ID | 416 |
Feature activation | +2.06e+01 |
Pos logit contributions | |
atography | +4.865 |
Marina | +4.702 |
Ru | +4.613 |
IDA | +4.544 |
wrongdoing | +4.524 |
Neg logit contributions | |
weather | -12.678 |
pite | -11.510 |
ritical | -11.242 |
usage | -11.116 |
reconc | -10.588 |
Token | Mark |
Token ID | 2940 |
Feature activation | +4.82e+00 |
Pos logit contributions | |
Marina | +10.608 |
atography | +9.767 |
Transparency | +9.714 |
Ru | +9.442 |
IDA | +9.226 |
Neg logit contributions | |
weather | -15.919 |
pite | -14.729 |
ritical | -14.295 |
usage | -14.127 |
reconc | -13.186 |
Token | Potter |
Token ID | 14179 |
Feature activation | -2.22e-01 |
Pos logit contributions | |
atography | +2.720 |
IDA | +2.590 |
Ru | +2.556 |
wrongdoing | +2.502 |
kaya | +2.495 |
Neg logit contributions | |
weather | -4.179 |
pite | -3.685 |
usage | -3.568 |
ritical | -3.563 |
Hurricane | -3.398 |
Token | ) |
Token ID | 8 |
Feature activation | +8.34e+00 |
Pos logit contributions | |
weather | +0.173 |
pite | +0.143 |
ritical | +0.137 |
usage | +0.136 |
Hurricane | +0.133 |
Neg logit contributions | |
atography | -0.152 |
IDA | -0.150 |
Ru | -0.145 |
wrongdoing | -0.144 |
vana | -0.141 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.64e+00 |
Pos logit contributions | |
atography | +2.541 |
IDA | +2.260 |
Ru | +2.163 |
Marina | +2.095 |
wrongdoing | +2.043 |
Neg logit contributions | |
weather | -9.209 |
pite | -8.426 |
ritical | -8.105 |
usage | -8.081 |
Hurricane | -7.798 |
Token | The |
Token ID | 464 |
Feature activation | -2.09e+00 |
Pos logit contributions | |
atography | +1.632 |
IDA | +1.543 |
wrongdoing | +1.513 |
Ru | +1.513 |
kaya | +1.497 |
Neg logit contributions | |
weather | -2.197 |
pite | -1.898 |
usage | -1.830 |
ritical | -1.825 |
Hurricane | -1.767 |
Token | and |
Token ID | 290 |
Feature activation | +1.61e+00 |
Pos logit contributions | |
atography | +2.679 |
IDA | +2.575 |
Ru | +2.542 |
wrongdoing | +2.498 |
kaya | +2.491 |
Neg logit contributions | |
weather | -3.156 |
pite | -2.711 |
ritical | -2.610 |
usage | -2.598 |
Hurricane | -2.525 |
Token | grid |
Token ID | 10706 |
Feature activation | +5.26e+00 |
Pos logit contributions | |
atography | +1.200 |
IDA | +1.167 |
Ru | +1.144 |
irable | +1.124 |
kaya | +1.122 |
Neg logit contributions | |
weather | -1.127 |
pite | -0.944 |
ritical | -0.906 |
usage | -0.901 |
Hurricane | -0.897 |
Token | cell |
Token ID | 2685 |
Feature activation | +5.01e+00 |
Pos logit contributions | |
atography | +3.486 |
IDA | +3.344 |
Ru | +3.299 |
wrongdoing | +3.242 |
kaya | +3.228 |
Neg logit contributions | |
weather | -4.098 |
pite | -3.527 |
ritical | -3.384 |
usage | -3.365 |
Hurricane | -3.270 |
Token | , |
Token ID | 11 |
Feature activation | +5.19e+00 |
Pos logit contributions | |
atography | +3.345 |
IDA | +3.214 |
Ru | +3.168 |
wrongdoing | +3.116 |
kaya | +3.104 |
Neg logit contributions | |
weather | -3.890 |
pite | -3.344 |
ritical | -3.211 |
usage | -3.192 |
Hurricane | -3.104 |
Token | as |
Token ID | 355 |
Feature activation | +1.34e+01 |
Pos logit contributions | |
atography | +3.439 |
IDA | +3.305 |
Ru | +3.257 |
wrongdoing | +3.203 |
kaya | +3.191 |
Neg logit contributions | |
weather | -4.045 |
pite | -3.481 |
ritical | -3.343 |
usage | -3.323 |
Hurricane | -3.230 |
Token | shown |
Token ID | 3402 |
Feature activation | +2.03e+01 |
Pos logit contributions | |
atography | +7.221 |
IDA | +7.139 |
Ru | +6.954 |
Transparency | +6.953 |
kaya | +6.901 |
Neg logit contributions | |
weather | -10.895 |
pite | -9.489 |
ritical | -9.220 |
usage | -9.054 |
Hurricane | -8.420 |
Token | in |
Token ID | 287 |
Feature activation | +1.41e+01 |
Pos logit contributions | |
atography | +8.078 |
→ | +7.937 |
enrich | +7.835 |
wrongdoing | +7.793 |
furthe | +7.751 |
Neg logit contributions | |
weather | -17.148 |
pite | -15.351 |
ritical | -15.134 |
usage | -14.612 |
okingly | -13.690 |
Token | Figure |
Token ID | 11291 |
Feature activation | -7.27e-01 |
Pos logit contributions | |
atography | +6.794 |
Transparency | +6.458 |
IDA | +6.391 |
wrongdoing | +6.277 |
Marina | +6.179 |
Neg logit contributions | |
weather | -12.272 |
pite | -10.956 |
ritical | -10.574 |
usage | -10.432 |
okingly | -9.761 |
Token | 3 |
Token ID | 513 |
Feature activation | +5.24e+00 |
Pos logit contributions | |
weather | +0.496 |
pite | +0.415 |
usage | +0.400 |
ritical | +0.392 |
Hurricane | +0.391 |
Neg logit contributions | |
atography | -0.549 |
IDA | -0.533 |
Ru | -0.527 |
wrongdoing | -0.515 |
kaya | -0.514 |
Token | . |
Token ID | 13 |
Feature activation | +4.12e+00 |
Pos logit contributions | |
atography | +3.411 |
IDA | +3.276 |
Ru | +3.227 |
wrongdoing | +3.171 |
kaya | +3.166 |
Neg logit contributions | |
weather | -4.145 |
pite | -3.574 |
ritical | -3.437 |
usage | -3.413 |
Hurricane | -3.320 |
Token |
|
Token ID | 198 |
Feature activation | -4.79e-01 |
Pos logit contributions | |
atography | +2.821 |
IDA | +2.695 |
Ru | +2.656 |
wrongdoing | +2.630 |
kaya | +2.611 |
Neg logit contributions | |
weather | -3.155 |
pite | -2.696 |
ritical | -2.589 |
usage | -2.580 |
Hurricane | -2.529 |
Token | 26 |
Token ID | 2608 |
Feature activation | +1.10e+01 |
Pos logit contributions | |
IDA | +5.719 |
Marina | +5.667 |
Ru | +5.666 |
atography | +5.603 |
wrongdoing | +5.559 |
Neg logit contributions | |
weather | -8.822 |
ritical | -8.012 |
pite | -7.893 |
usage | -7.613 |
reconc | -7.345 |
Token | , |
Token ID | 11 |
Feature activation | +9.28e+00 |
Pos logit contributions | |
IDA | +5.508 |
atography | +5.507 |
Ru | +5.463 |
wrongdoing | +5.392 |
Marina | +5.380 |
Neg logit contributions | |
weather | -9.355 |
ritical | -8.428 |
pite | -8.262 |
usage | -8.073 |
reconc | -7.738 |
Token | 2014 |
Token ID | 1946 |
Feature activation | +7.50e+00 |
Pos logit contributions | |
atography | +4.802 |
IDA | +4.630 |
Ru | +4.573 |
wrongdoing | +4.486 |
kaya | +4.409 |
Neg logit contributions | |
weather | -8.291 |
pite | -7.177 |
ritical | -7.168 |
usage | -7.038 |
Hurricane | -6.674 |
Token | . |
Token ID | 13 |
Feature activation | +1.19e+01 |
Pos logit contributions | |
atography | +4.131 |
IDA | +4.104 |
Ru | +4.002 |
wrongdoing | +3.961 |
Marina | +3.882 |
Neg logit contributions | |
weather | -6.315 |
pite | -5.574 |
ritical | -5.523 |
usage | -5.302 |
Hurricane | -5.027 |
Token | REUTERS |
Token ID | 15862 |
Feature activation | +1.39e+01 |
Pos logit contributions | |
atography | +4.704 |
Ru | +4.498 |
IDA | +4.497 |
Marina | +4.317 |
Transparency | +4.115 |
Neg logit contributions | |
weather | -11.825 |
pite | -10.603 |
ritical | -10.598 |
usage | -10.379 |
sled | -9.937 |
Token | / |
Token ID | 14 |
Feature activation | +1.98e+01 |
Pos logit contributions | |
IDA | +5.990 |
Ru | +5.977 |
Marina | +5.532 |
atography | +5.480 |
]) | +5.431 |
Neg logit contributions | |
weather | -12.556 |
ritical | -11.610 |
pite | -11.440 |
ortunately | -11.262 |
usage | -11.107 |
Token | En |
Token ID | 4834 |
Feature activation | +8.10e+00 |
Pos logit contributions | |
Ru | +10.266 |
IDA | +9.715 |
Step | +9.338 |
Sus | +8.920 |
atography | +8.866 |
Neg logit contributions | |
weather | -15.759 |
pite | -13.926 |
ritical | -13.893 |
sled | -13.740 |
reconc | -13.532 |
Token | rique |
Token ID | 33865 |
Feature activation | +3.84e+00 |
Pos logit contributions | |
atography | +2.675 |
IDA | +2.521 |
Ru | +2.471 |
kaya | +2.404 |
wrongdoing | +2.290 |
Neg logit contributions | |
weather | -8.771 |
pite | -7.931 |
usage | -7.763 |
ritical | -7.746 |
Hurricane | -7.535 |
Token | de |
Token ID | 390 |
Feature activation | +2.88e+00 |
Pos logit contributions | |
atography | +2.472 |
IDA | +2.387 |
Ru | +2.352 |
wrongdoing | +2.306 |
kaya | +2.295 |
Neg logit contributions | |
weather | -3.054 |
pite | -2.635 |
ritical | -2.551 |
usage | -2.529 |
Hurricane | -2.438 |
Token | la |
Token ID | 8591 |
Feature activation | -7.30e-01 |
Pos logit contributions | |
atography | +1.730 |
IDA | +1.652 |
Ru | +1.624 |
wrongdoing | +1.595 |
kaya | +1.588 |
Neg logit contributions | |
weather | -2.433 |
pite | -2.122 |
ritical | -2.043 |
usage | -2.031 |
Hurricane | -1.985 |
Token | Os |
Token ID | 8834 |
Feature activation | +5.44e+00 |
Pos logit contributions | |
weather | +0.559 |
pite | +0.490 |
ritical | +0.464 |
Hurricane | +0.455 |
Pend | +0.447 |
Neg logit contributions | |
atography | -0.494 |
IDA | -0.470 |
vana | -0.456 |
wrongdoing | -0.453 |
irable | -0.449 |
Token | better |
Token ID | 1365 |
Feature activation | -4.39e+00 |
Pos logit contributions | |
atography | +0.104 |
IDA | +0.100 |
Ru | +0.098 |
kaya | +0.095 |
vana | +0.094 |
Neg logit contributions | |
weather | -0.133 |
pite | -0.115 |
usage | -0.110 |
Hurricane | -0.110 |
ritical | -0.110 |
Token | . |
Token ID | 13 |
Feature activation | +6.53e+00 |
Pos logit contributions | |
weather | +2.955 |
pite | +2.448 |
usage | +2.348 |
Hurricane | +2.335 |
timed | +2.263 |
Neg logit contributions | |
atography | -3.284 |
Ru | -3.172 |
IDA | -3.152 |
vana | -3.027 |
kaya | -3.015 |
Token | Having |
Token ID | 11136 |
Feature activation | +3.87e-01 |
Pos logit contributions | |
atography | +4.359 |
IDA | +4.233 |
Ru | +4.171 |
wrongdoing | +4.084 |
kaya | +4.065 |
Neg logit contributions | |
weather | -4.986 |
pite | -4.290 |
ritical | -4.147 |
usage | -4.082 |
Hurricane | -3.937 |
Token | done |
Token ID | 1760 |
Feature activation | +8.22e+00 |
Pos logit contributions | |
atography | +0.273 |
IDA | +0.262 |
Ru | +0.259 |
kaya | +0.258 |
wrongdoing | +0.252 |
Neg logit contributions | |
weather | -0.286 |
pite | -0.243 |
usage | -0.231 |
Hurricane | -0.230 |
ritical | -0.229 |
Token | all |
Token ID | 477 |
Feature activation | +1.25e+01 |
Pos logit contributions | |
atography | +5.073 |
IDA | +4.874 |
Ru | +4.768 |
wrongdoing | +4.741 |
kaya | +4.655 |
Neg logit contributions | |
weather | -6.669 |
pite | -5.771 |
ritical | -5.611 |
usage | -5.549 |
Hurricane | -5.331 |
Token | that |
Token ID | 326 |
Feature activation | +1.97e+01 |
Pos logit contributions | |
atography | +6.767 |
wrongdoing | +6.540 |
IDA | +6.469 |
Ru | +6.190 |
Marina | +6.177 |
Neg logit contributions | |
weather | -10.402 |
pite | -9.228 |
ritical | -9.081 |
usage | -8.921 |
Hurricane | -8.383 |
Token | , |
Token ID | 11 |
Feature activation | +7.64e+00 |
Pos logit contributions | |
atography | +11.227 |
wrongdoing | +10.907 |
IDA | +10.621 |
enrich | +10.600 |
Marina | +10.527 |
Neg logit contributions | |
weather | -14.004 |
ritical | -11.937 |
pite | -11.826 |
usage | -11.481 |
awks | -10.729 |
Token | what |
Token ID | 644 |
Feature activation | +1.35e+01 |
Pos logit contributions | |
atography | +5.081 |
IDA | +4.917 |
Ru | +4.832 |
wrongdoing | +4.776 |
kaya | +4.709 |
Neg logit contributions | |
weather | -5.812 |
pite | -4.971 |
ritical | -4.833 |
usage | -4.727 |
Hurricane | -4.567 |
Token | else |
Token ID | 2073 |
Feature activation | +9.31e+00 |
Pos logit contributions | |
atography | +7.777 |
IDA | +7.716 |
wrongdoing | +7.663 |
Ru | +7.498 |
Marina | +7.427 |
Neg logit contributions | |
weather | -10.431 |
pite | -9.066 |
ritical | -8.886 |
usage | -8.565 |
Hurricane | -8.186 |
Token | can |
Token ID | 460 |
Feature activation | -4.45e-01 |
Pos logit contributions | |
atography | +6.017 |
IDA | +5.773 |
wrongdoing | +5.726 |
Ru | +5.673 |
kaya | +5.605 |
Neg logit contributions | |
weather | -7.170 |
pite | -6.171 |
ritical | -6.007 |
usage | -5.892 |
Hurricane | -5.681 |
Token | be |
Token ID | 307 |
Feature activation | +6.54e+00 |
Pos logit contributions | |
weather | +0.345 |
pite | +0.296 |
usage | +0.284 |
ritical | +0.283 |
Hurricane | +0.282 |
Neg logit contributions | |
atography | -0.298 |
Ru | -0.278 |
IDA | -0.278 |
kaya | -0.274 |
vana | -0.271 |
Token | , |
Token ID | 11 |
Feature activation | +9.26e+00 |
Pos logit contributions | |
atography | +5.997 |
IDA | +5.635 |
Ru | +5.626 |
wrongdoing | +5.594 |
Marina | +5.428 |
Neg logit contributions | |
weather | -8.340 |
pite | -7.265 |
usage | -7.019 |
ritical | -6.957 |
Hurricane | -6.717 |
Token | the |
Token ID | 262 |
Feature activation | +6.55e+00 |
Pos logit contributions | |
atography | +6.053 |
wrongdoing | +5.816 |
Ru | +5.793 |
IDA | +5.750 |
kaya | +5.594 |
Neg logit contributions | |
weather | -7.072 |
pite | -6.000 |
ritical | -5.775 |
usage | -5.738 |
Hurricane | -5.527 |
Token | excitement |
Token ID | 14067 |
Feature activation | +1.09e+01 |
Pos logit contributions | |
atography | +4.296 |
IDA | +4.094 |
Ru | +4.074 |
wrongdoing | +4.051 |
kaya | +3.979 |
Neg logit contributions | |
weather | -5.116 |
pite | -4.382 |
ritical | -4.218 |
usage | -4.192 |
Hurricane | -4.060 |
Token | and |
Token ID | 290 |
Feature activation | +9.26e+00 |
Pos logit contributions | |
atography | +6.333 |
IDA | +5.960 |
Ru | +5.954 |
wrongdoing | +5.863 |
kaya | +5.759 |
Neg logit contributions | |
weather | -8.849 |
pite | -7.717 |
usage | -7.433 |
ritical | -7.407 |
Hurricane | -7.119 |
Token | the |
Token ID | 262 |
Feature activation | +5.71e+00 |
Pos logit contributions | |
atography | +5.617 |
Ru | +5.353 |
IDA | +5.347 |
wrongdoing | +5.328 |
kaya | +5.203 |
Neg logit contributions | |
weather | -7.537 |
pite | -6.487 |
ritical | -6.280 |
usage | -6.261 |
Hurricane | -6.029 |
Token | immersion |
Token ID | 36375 |
Feature activation | +1.96e+01 |
Pos logit contributions | |
atography | +3.680 |
IDA | +3.518 |
Ru | +3.481 |
wrongdoing | +3.439 |
kaya | +3.405 |
Neg logit contributions | |
weather | -4.547 |
pite | -3.914 |
ritical | -3.769 |
usage | -3.749 |
Hurricane | -3.638 |
Token | – |
Token ID | 784 |
Feature activation | +7.73e+00 |
Pos logit contributions | |
atography | +10.880 |
enrich | +10.513 |
iness | +10.394 |
Ru | +10.361 |
wrongdoing | +10.352 |
Neg logit contributions | |
weather | -13.982 |
pite | -12.009 |
usage | -11.535 |
ritical | -11.380 |
okingly | -10.733 |
Token | it |
Token ID | 340 |
Feature activation | +5.32e+00 |
Pos logit contributions | |
atography | +5.280 |
IDA | +5.096 |
Ru | +5.035 |
wrongdoing | +4.965 |
kaya | +4.924 |
Neg logit contributions | |
weather | -5.769 |
pite | -4.923 |
ritical | -4.748 |
usage | -4.692 |
Hurricane | -4.525 |
Token | is |
Token ID | 318 |
Feature activation | +1.33e+01 |
Pos logit contributions | |
atography | +3.930 |
IDA | +3.769 |
Ru | +3.741 |
wrongdoing | +3.693 |
kaya | +3.685 |
Neg logit contributions | |
weather | -3.709 |
pite | -3.146 |
ritical | -3.005 |
usage | -2.990 |
Hurricane | -2.877 |
Token | your |
Token ID | 534 |
Feature activation | +3.75e+00 |
Pos logit contributions | |
atography | +7.567 |
enrich | +7.386 |
wrongdoing | +7.322 |
Ru | +7.219 |
IDA | +7.189 |
Neg logit contributions | |
weather | -10.355 |
pite | -9.065 |
ritical | -8.822 |
usage | -8.693 |
awks | -8.080 |
Token | job |
Token ID | 1693 |
Feature activation | -4.67e+00 |
Pos logit contributions | |
atography | +2.338 |
IDA | +2.240 |
Ru | +2.206 |
wrongdoing | +2.165 |
kaya | +2.158 |
Neg logit contributions | |
weather | -3.074 |
pite | -2.665 |
ritical | -2.566 |
usage | -2.553 |
Hurricane | -2.488 |
Token | room |
Token ID | 2119 |
Feature activation | +5.57e+00 |
Pos logit contributions | |
atography | +2.145 |
IDA | +2.030 |
Ru | +1.988 |
wrongdoing | +1.957 |
kaya | +1.942 |
Neg logit contributions | |
weather | -3.550 |
pite | -3.117 |
ritical | -3.016 |
usage | -2.997 |
Hurricane | -2.936 |
Token | maze |
Token ID | 31237 |
Feature activation | -1.48e+00 |
Pos logit contributions | |
atography | +3.156 |
IDA | +3.011 |
Ru | +2.961 |
wrongdoing | +2.902 |
kaya | +2.890 |
Neg logit contributions | |
weather | -4.870 |
pite | -4.264 |
ritical | -4.114 |
usage | -4.096 |
Hurricane | -3.996 |
Token | was |
Token ID | 373 |
Feature activation | +3.64e+00 |
Pos logit contributions | |
weather | +1.013 |
pite | +0.850 |
ritical | +0.830 |
usage | +0.811 |
Hurricane | +0.799 |
Neg logit contributions | |
atography | -1.118 |
IDA | -1.087 |
Ru | -1.060 |
wrongdoing | -1.044 |
irable | -1.039 |
Token | designed |
Token ID | 3562 |
Feature activation | -2.22e-01 |
Pos logit contributions | |
atography | +2.736 |
IDA | +2.645 |
Ru | +2.607 |
wrongdoing | +2.568 |
kaya | +2.559 |
Neg logit contributions | |
weather | -2.528 |
pite | -2.130 |
ritical | -2.032 |
usage | -2.019 |
Hurricane | -1.961 |
Token | as |
Token ID | 355 |
Feature activation | +5.82e+00 |
Pos logit contributions | |
weather | +0.167 |
pite | +0.144 |
ritical | +0.138 |
usage | +0.136 |
Hurricane | +0.134 |
Neg logit contributions | |
atography | -0.153 |
IDA | -0.149 |
Ru | -0.147 |
wrongdoing | -0.143 |
kaya | -0.142 |
Token | shown |
Token ID | 3402 |
Feature activation | +1.94e+01 |
Pos logit contributions | |
atography | +4.040 |
IDA | +3.887 |
Ru | +3.841 |
wrongdoing | +3.779 |
kaya | +3.761 |
Neg logit contributions | |
weather | -4.331 |
pite | -3.690 |
ritical | -3.542 |
usage | -3.522 |
Hurricane | -3.397 |
Token | in |
Token ID | 287 |
Feature activation | +1.62e+01 |
Pos logit contributions | |
atography | +8.071 |
Marina | +7.453 |
→ | +7.441 |
kaya | +7.432 |
enrich | +7.430 |
Neg logit contributions | |
weather | -16.661 |
pite | -14.782 |
ritical | -14.335 |
usage | -14.202 |
Hurricane | -13.298 |
Token | Figure |
Token ID | 11291 |
Feature activation | +4.94e+00 |
Pos logit contributions | |
atography | +7.517 |
IDA | +7.144 |
Transparency | +6.953 |
Marina | +6.949 |
wrongdoing | +6.899 |
Neg logit contributions | |
weather | -14.107 |
pite | -12.438 |
ritical | -12.144 |
usage | -11.802 |
okingly | -11.107 |
Token | 1 |
Token ID | 352 |
Feature activation | +5.45e+00 |
Pos logit contributions | |
atography | +2.875 |
IDA | +2.751 |
Ru | +2.700 |
wrongdoing | +2.653 |
kaya | +2.642 |
Neg logit contributions | |
weather | -4.246 |
pite | -3.707 |
ritical | -3.578 |
usage | -3.553 |
Hurricane | -3.465 |
Token | . |
Token ID | 13 |
Feature activation | +3.69e+00 |
Pos logit contributions | |
atography | +3.485 |
IDA | +3.347 |
Ru | +3.296 |
wrongdoing | +3.236 |
kaya | +3.230 |
Neg logit contributions | |
weather | -4.372 |
pite | -3.777 |
ritical | -3.636 |
usage | -3.610 |
Hurricane | -3.515 |
Token | The |
Token ID | 383 |
Feature activation | +6.56e+00 |
Pos logit contributions | |
atography | +2.510 |
IDA | +2.400 |
Ru | +2.358 |
wrongdoing | +2.333 |
kaya | +2.321 |
Neg logit contributions | |
weather | -2.836 |
pite | -2.434 |
ritical | -2.329 |
usage | -2.320 |
Hurricane | -2.280 |
Token | , |
Token ID | 11 |
Feature activation | +1.03e+01 |
Pos logit contributions | |
atography | +5.060 |
IDA | +4.869 |
wrongdoing | +4.760 |
Ru | +4.741 |
kaya | +4.683 |
Neg logit contributions | |
weather | -6.861 |
pite | -6.010 |
ritical | -5.852 |
usage | -5.801 |
Hurricane | -5.509 |
Token | standing |
Token ID | 5055 |
Feature activation | +1.13e+01 |
Pos logit contributions | |
atography | +5.770 |
IDA | +5.617 |
wrongdoing | +5.511 |
Ru | +5.447 |
Marina | +5.406 |
Neg logit contributions | |
weather | -8.493 |
pite | -7.439 |
ritical | -7.346 |
usage | -7.236 |
Hurricane | -6.695 |
Token | 5 |
Token ID | 642 |
Feature activation | +9.07e+00 |
Pos logit contributions | |
atography | +6.459 |
wrongdoing | +6.196 |
Marina | +6.189 |
IDA | +6.165 |
Ru | +6.111 |
Neg logit contributions | |
weather | -8.966 |
pite | -7.929 |
ritical | -7.809 |
usage | -7.707 |
okingly | -6.955 |
Token | ' |
Token ID | 6 |
Feature activation | +1.04e+01 |
Pos logit contributions | |
atography | +5.984 |
Ru | +5.763 |
IDA | +5.716 |
wrongdoing | +5.678 |
kaya | +5.646 |
Neg logit contributions | |
weather | -6.719 |
pite | -5.790 |
ritical | -5.611 |
usage | -5.542 |
Hurricane | -5.346 |
Token | 7 |
Token ID | 22 |
Feature activation | +1.54e+01 |
Pos logit contributions | |
atography | +7.107 |
IDA | +7.027 |
Ru | +6.935 |
wrongdoing | +6.770 |
kaya | +6.725 |
Neg logit contributions | |
weather | -7.238 |
pite | -6.263 |
ritical | -6.086 |
usage | -5.999 |
Hurricane | -5.701 |
Token | " |
Token ID | 1 |
Feature activation | +1.92e+01 |
Pos logit contributions | |
atography | +8.377 |
IDA | +8.003 |
wrongdoing | +7.941 |
Ru | +7.875 |
Marina | +7.850 |
Neg logit contributions | |
weather | -12.174 |
pite | -10.882 |
ritical | -10.476 |
usage | -10.330 |
Hurricane | -9.719 |
Token | with |
Token ID | 351 |
Feature activation | +1.08e+01 |
Pos logit contributions | |
atography | +8.238 |
IDA | +8.177 |
Marina | +7.957 |
wrongdoing | +7.939 |
Ru | +7.788 |
Neg logit contributions | |
weather | -15.660 |
pite | -14.508 |
ritical | -14.211 |
usage | -13.969 |
reconc | -13.157 |
Token | dark |
Token ID | 3223 |
Feature activation | +7.62e+00 |
Pos logit contributions | |
atography | +6.282 |
wrongdoing | +6.085 |
IDA | +5.974 |
Ru | +5.916 |
enrich | +5.855 |
Neg logit contributions | |
weather | -8.689 |
pite | -7.542 |
ritical | -7.469 |
usage | -7.404 |
Hurricane | -6.730 |
Token | hair |
Token ID | 4190 |
Feature activation | +4.21e+00 |
Pos logit contributions | |
atography | +3.764 |
IDA | +3.561 |
wrongdoing | +3.538 |
Ru | +3.498 |
kaya | +3.454 |
Neg logit contributions | |
weather | -6.956 |
pite | -6.208 |
ritical | -6.056 |
usage | -6.018 |
Hurricane | -5.740 |
Token | , |
Token ID | 11 |
Feature activation | +1.11e+01 |
Pos logit contributions | |
atography | +2.780 |
IDA | +2.671 |
Ru | +2.631 |
wrongdoing | +2.586 |
kaya | +2.578 |
Neg logit contributions | |
weather | -3.305 |
pite | -2.845 |
ritical | -2.731 |
usage | -2.717 |
Hurricane | -2.645 |
Token | a |
Token ID | 257 |
Feature activation | +5.77e+00 |
Pos logit contributions | |
atography | +6.223 |
IDA | +5.956 |
wrongdoing | +5.952 |
Ru | +5.830 |
Marina | +5.813 |
Neg logit contributions | |
weather | -9.127 |
pite | -7.976 |
ritical | -7.848 |
usage | -7.779 |
Hurricane | -7.144 |
Token | eb |
Token ID | 1765 |
Feature activation | +4.84e+00 |
Pos logit contributions | |
atography | +10.864 |
IDA | +10.452 |
kaya | +10.375 |
Ru | +10.188 |
Marina | +10.001 |
Neg logit contributions | |
weather | -11.857 |
usage | -9.749 |
Hurricane | -9.684 |
pite | -9.604 |
ritical | -9.369 |
Token | a |
Token ID | 64 |
Feature activation | +2.10e+00 |
Pos logit contributions | |
atography | +2.990 |
IDA | +2.864 |
Ru | +2.819 |
wrongdoing | +2.770 |
kaya | +2.756 |
Neg logit contributions | |
weather | -3.998 |
pite | -3.470 |
ritical | -3.343 |
usage | -3.324 |
Hurricane | -3.239 |
Token | Sidd |
Token ID | 44487 |
Feature activation | +4.44e+00 |
Pos logit contributions | |
atography | +1.321 |
IDA | +1.268 |
Ru | +1.241 |
wrongdoing | +1.240 |
irable | +1.212 |
Neg logit contributions | |
weather | -1.715 |
pite | -1.478 |
ritical | -1.425 |
usage | -1.419 |
Hurricane | -1.392 |
Token | iqu |
Token ID | 1557 |
Feature activation | -1.49e+00 |
Pos logit contributions | |
atography | +2.918 |
IDA | +2.821 |
Ru | +2.765 |
wrongdoing | +2.725 |
kaya | +2.701 |
Neg logit contributions | |
weather | -3.502 |
pite | -3.019 |
ritical | -2.890 |
usage | -2.881 |
Hurricane | -2.794 |
Token | i |
Token ID | 72 |
Feature activation | +5.51e+00 |
Pos logit contributions | |
weather | +1.485 |
pite | +1.312 |
ritical | +1.287 |
usage | +1.250 |
Hurricane | +1.220 |
Neg logit contributions | |
atography | -0.670 |
IDA | -0.642 |
Ru | -0.639 |
wrongdoing | -0.634 |
Transparency | -0.614 |
Token | ; |
Token ID | 26 |
Feature activation | +1.92e+01 |
Pos logit contributions | |
atography | +3.148 |
IDA | +3.008 |
Ru | +2.950 |
wrongdoing | +2.897 |
kaya | +2.875 |
Neg logit contributions | |
weather | -4.824 |
pite | -4.215 |
ritical | -4.070 |
usage | -4.051 |
Hurricane | -3.954 |
Token | Editing |
Token ID | 39883 |
Feature activation | +1.28e+01 |
Pos logit contributions | |
atography | +6.323 |
Marina | +6.002 |
IDA | +5.726 |
enrich | +5.544 |
wrongdoing | +5.438 |
Neg logit contributions | |
weather | -18.821 |
pite | -17.398 |
ritical | -16.270 |
usage | -15.941 |
awks | -15.923 |
Token | by |
Token ID | 416 |
Feature activation | +2.06e+01 |
Pos logit contributions | |
atography | +4.865 |
Marina | +4.702 |
Ru | +4.613 |
IDA | +4.544 |
wrongdoing | +4.524 |
Neg logit contributions | |
weather | -12.678 |
pite | -11.510 |
ritical | -11.242 |
usage | -11.116 |
reconc | -10.588 |
Token | Mark |
Token ID | 2940 |
Feature activation | +4.82e+00 |
Pos logit contributions | |
Marina | +10.608 |
atography | +9.767 |
Transparency | +9.714 |
Ru | +9.442 |
IDA | +9.226 |
Neg logit contributions | |
weather | -15.919 |
pite | -14.729 |
ritical | -14.295 |
usage | -14.127 |
reconc | -13.186 |
Token | Potter |
Token ID | 14179 |
Feature activation | -2.22e-01 |
Pos logit contributions | |
atography | +2.720 |
IDA | +2.590 |
Ru | +2.556 |
wrongdoing | +2.502 |
kaya | +2.495 |
Neg logit contributions | |
weather | -4.179 |
pite | -3.685 |
usage | -3.568 |
ritical | -3.563 |
Hurricane | -3.398 |
Token | ) |
Token ID | 8 |
Feature activation | +8.34e+00 |
Pos logit contributions | |
weather | +0.173 |
pite | +0.143 |
ritical | +0.137 |
usage | +0.136 |
Hurricane | +0.133 |
Neg logit contributions | |
atography | -0.152 |
IDA | -0.150 |
Ru | -0.145 |
wrongdoing | -0.144 |
vana | -0.141 |
Token | 2010 |
Token ID | 3050 |
Feature activation | -7.13e-01 |
Pos logit contributions | |
atography | +1.679 |
IDA | +1.605 |
Ru | +1.593 |
irable | +1.552 |
kaya | +1.545 |
Neg logit contributions | |
weather | -1.990 |
pite | -1.717 |
ritical | -1.623 |
usage | -1.617 |
Hurricane | -1.600 |
Token | . |
Token ID | 13 |
Feature activation | -8.01e-02 |
Pos logit contributions | |
weather | +0.519 |
pite | +0.436 |
Hurricane | +0.411 |
ritical | +0.409 |
usage | +0.407 |
Neg logit contributions | |
atography | -0.519 |
IDA | -0.503 |
Ru | -0.497 |
irable | -0.486 |
kaya | -0.485 |
Token | P |
Token ID | 350 |
Feature activation | +1.30e+00 |
Pos logit contributions | |
weather | +0.060 |
pite | +0.050 |
Hurricane | +0.049 |
usage | +0.048 |
ritical | +0.047 |
Neg logit contributions | |
atography | -0.056 |
IDA | -0.053 |
irable | -0.053 |
enrich | -0.053 |
Ru | -0.052 |
Token | eder |
Token ID | 5702 |
Feature activation | +8.65e+00 |
Pos logit contributions | |
atography | +0.929 |
IDA | +0.885 |
Ru | +0.881 |
wrongdoing | +0.870 |
kaya | +0.858 |
Neg logit contributions | |
weather | -0.959 |
pite | -0.816 |
usage | -0.784 |
ritical | -0.784 |
Hurricane | -0.740 |
Token | had |
Token ID | 550 |
Feature activation | +3.24e+00 |
Pos logit contributions | |
atography | +5.437 |
IDA | +5.209 |
wrongdoing | +5.128 |
Ru | +5.101 |
Marina | +5.012 |
Neg logit contributions | |
weather | -6.806 |
pite | -5.926 |
ritical | -5.774 |
usage | -5.726 |
Hurricane | -5.430 |
Token | already |
Token ID | 1541 |
Feature activation | +1.91e+01 |
Pos logit contributions | |
atography | +2.245 |
IDA | +2.156 |
Ru | +2.134 |
kaya | +2.097 |
irable | +2.079 |
Neg logit contributions | |
weather | -2.437 |
pite | -2.082 |
ritical | -1.990 |
usage | -1.988 |
Hurricane | -1.948 |
Token | had |
Token ID | 550 |
Feature activation | +3.34e+00 |
Pos logit contributions | |
wrongdoing | +9.590 |
atography | +9.287 |
enrich | +9.106 |
furthe | +8.963 |
Marina | +8.815 |
Neg logit contributions | |
weather | -14.903 |
ritical | -13.451 |
pite | -13.264 |
usage | -12.883 |
awks | -12.222 |
Token | two |
Token ID | 734 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
atography | +2.562 |
IDA | +2.467 |
Ru | +2.443 |
kaya | +2.403 |
wrongdoing | +2.388 |
Neg logit contributions | |
weather | -2.263 |
pite | -1.901 |
usage | -1.808 |
ritical | -1.806 |
Hurricane | -1.757 |
Token | prior |
Token ID | 3161 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
weather | +0.852 |
pite | +0.752 |
usage | +0.728 |
Hurricane | +0.720 |
ritical | +0.713 |
Neg logit contributions | |
atography | -0.618 |
Ru | -0.586 |
IDA | -0.574 |
irable | -0.573 |
kaya | -0.567 |
Token | DUI |
Token ID | 39414 |
Feature activation | +4.69e+00 |
Pos logit contributions | |
weather | +0.947 |
pite | +0.829 |
usage | +0.807 |
Hurricane | +0.800 |
ritical | +0.789 |
Neg logit contributions | |
atography | -0.663 |
Ru | -0.629 |
IDA | -0.620 |
kaya | -0.615 |
irable | -0.609 |
Token | incidents |
Token ID | 10207 |
Feature activation | +1.31e+00 |
Pos logit contributions | |
atography | +2.407 |
IDA | +2.313 |
wrongdoing | +2.245 |
Ru | +2.230 |
kaya | +2.169 |
Neg logit contributions | |
weather | -4.333 |
pite | -3.854 |
ritical | -3.702 |
usage | -3.662 |
Hurricane | -3.583 |
Token | through |
Token ID | 832 |
Feature activation | +6.37e+00 |
Pos logit contributions | |
atography | +4.157 |
IDA | +3.971 |
Ru | +3.936 |
wrongdoing | +3.888 |
kaya | +3.798 |
Neg logit contributions | |
weather | -6.439 |
pite | -5.624 |
ritical | -5.485 |
usage | -5.431 |
Hurricane | -5.205 |
Token | the |
Token ID | 262 |
Feature activation | +5.57e+00 |
Pos logit contributions | |
atography | +4.062 |
IDA | +3.900 |
Ru | +3.860 |
wrongdoing | +3.826 |
kaya | +3.745 |
Neg logit contributions | |
weather | -5.079 |
pite | -4.394 |
ritical | -4.253 |
usage | -4.208 |
Hurricane | -4.037 |
Token | rubbish |
Token ID | 35022 |
Feature activation | +4.38e+00 |
Pos logit contributions | |
atography | +3.469 |
IDA | +3.322 |
Ru | +3.281 |
wrongdoing | +3.231 |
kaya | +3.198 |
Neg logit contributions | |
weather | -4.560 |
pite | -3.955 |
ritical | -3.816 |
usage | -3.789 |
Hurricane | -3.676 |
Token | to |
Token ID | 284 |
Feature activation | +6.86e+00 |
Pos logit contributions | |
atography | +2.865 |
IDA | +2.753 |
Ru | +2.707 |
wrongdoing | +2.660 |
kaya | +2.656 |
Neg logit contributions | |
weather | -3.465 |
pite | -2.986 |
ritical | -2.866 |
usage | -2.854 |
Hurricane | -2.782 |
Token | remove |
Token ID | 4781 |
Feature activation | +1.01e+01 |
Pos logit contributions | |
atography | +4.606 |
IDA | +4.362 |
Ru | +4.362 |
wrongdoing | +4.319 |
enrich | +4.274 |
Neg logit contributions | |
weather | -5.190 |
pite | -4.475 |
ritical | -4.307 |
usage | -4.297 |
Hurricane | -4.077 |
Token | any |
Token ID | 597 |
Feature activation | +1.91e+01 |
Pos logit contributions | |
atography | +6.490 |
wrongdoing | +6.355 |
Ru | +6.212 |
IDA | +6.199 |
irable | +6.076 |
Neg logit contributions | |
weather | -7.678 |
pite | -6.634 |
ritical | -6.507 |
usage | -6.382 |
Hurricane | -5.910 |
Token | hazardous |
Token ID | 26082 |
Feature activation | +3.99e+00 |
Pos logit contributions | |
wrongdoing | +12.036 |
atography | +10.984 |
Marina | +10.903 |
enrich | +10.859 |
Ru | +10.856 |
Neg logit contributions | |
weather | -12.932 |
ritical | -11.341 |
pite | -11.203 |
usage | -10.826 |
States | -9.756 |
Token | or |
Token ID | 393 |
Feature activation | +2.90e+00 |
Pos logit contributions | |
atography | +2.180 |
IDA | +2.078 |
Ru | +2.035 |
kaya | +1.989 |
wrongdoing | +1.982 |
Neg logit contributions | |
weather | -3.586 |
pite | -3.149 |
ritical | -3.034 |
usage | -3.024 |
Hurricane | -2.964 |
Token | toxic |
Token ID | 11422 |
Feature activation | +9.87e+00 |
Pos logit contributions | |
atography | +2.013 |
IDA | +1.953 |
Ru | +1.911 |
kaya | +1.870 |
irable | +1.851 |
Neg logit contributions | |
weather | -2.191 |
pite | -1.875 |
usage | -1.789 |
ritical | -1.784 |
Hurricane | -1.753 |
Token | material |
Token ID | 2587 |
Feature activation | +1.02e+01 |
Pos logit contributions | |
atography | +4.947 |
wrongdoing | +4.650 |
Ru | +4.633 |
IDA | +4.608 |
kaya | +4.447 |
Neg logit contributions | |
weather | -8.964 |
pite | -7.911 |
ritical | -7.790 |
usage | -7.689 |
Hurricane | -7.438 |
Token |
|
Token ID | 198 |
Feature activation | +1.67e+00 |
Pos logit contributions | |
atography | +5.452 |
wrongdoing | +5.206 |
Ru | +5.205 |
IDA | +5.173 |
Marina | +5.100 |
Neg logit contributions | |
weather | -8.796 |
pite | -7.710 |
ritical | -7.591 |
usage | -7.362 |
Hurricane | -7.052 |
Token | said |
Token ID | 531 |
Feature activation | +8.82e-01 |
Pos logit contributions | |
atography | +1.169 |
IDA | +1.114 |
Ru | +1.097 |
wrongdoing | +1.079 |
kaya | +1.070 |
Neg logit contributions | |
weather | -1.701 |
pite | -1.498 |
ritical | -1.440 |
usage | -1.438 |
Hurricane | -1.385 |
Token | . |
Token ID | 13 |
Feature activation | +1.29e+00 |
Pos logit contributions | |
atography | +0.576 |
IDA | +0.553 |
Ru | +0.547 |
kaya | +0.533 |
wrongdoing | +0.533 |
Neg logit contributions | |
weather | -0.701 |
pite | -0.601 |
ritical | -0.576 |
usage | -0.571 |
Hurricane | -0.566 |
Token |
|
Token ID | 198 |
Feature activation | -3.15e-02 |
Pos logit contributions | |
atography | +0.781 |
IDA | +0.746 |
Ru | +0.735 |
wrongdoing | +0.721 |
kaya | +0.717 |
Neg logit contributions | |
weather | -1.075 |
pite | -0.936 |
ritical | -0.901 |
usage | -0.897 |
Hurricane | -0.870 |
Token |
|
Token ID | 198 |
Feature activation | -2.90e+00 |
Pos logit contributions | |
weather | +0.025 |
pite | +0.021 |
usage | +0.021 |
Hurricane | +0.020 |
ritical | +0.020 |
Neg logit contributions | |
atography | -0.021 |
IDA | -0.020 |
Ru | -0.019 |
wrongdoing | -0.019 |
irable | -0.019 |
Token | Both |
Token ID | 10265 |
Feature activation | +1.03e+01 |
Pos logit contributions | |
weather | +2.038 |
pite | +1.674 |
usage | +1.636 |
ritical | +1.587 |
States | +1.523 |
Neg logit contributions | |
atography | -2.176 |
irable | -2.055 |
enrich | -2.050 |
wrongdoing | -2.038 |
kaya | -2.007 |
Token | suspects |
Token ID | 11826 |
Feature activation | +1.89e+01 |
Pos logit contributions | |
atography | +5.855 |
wrongdoing | +5.779 |
IDA | +5.771 |
Ru | +5.644 |
Marina | +5.502 |
Neg logit contributions | |
weather | -8.406 |
pite | -7.418 |
ritical | -7.098 |
usage | -7.095 |
Hurricane | -6.689 |
Token | have |
Token ID | 423 |
Feature activation | +1.41e+01 |
Pos logit contributions | |
atography | +9.651 |
wrongdoing | +9.511 |
IDA | +9.446 |
Marina | +9.251 |
kaya | +9.224 |
Neg logit contributions | |
weather | -14.033 |
pite | -12.611 |
ritical | -12.355 |
usage | -12.145 |
Hurricane | -11.129 |
Token | been |
Token ID | 587 |
Feature activation | +1.09e+01 |
Pos logit contributions | |
wrongdoing | +7.849 |
atography | +7.709 |
IDA | +7.689 |
Marina | +7.466 |
enrich | +7.327 |
Neg logit contributions | |
weather | -11.312 |
pite | -10.032 |
ritical | -9.814 |
usage | -9.616 |
awks | -8.709 |
Token | arrested |
Token ID | 5169 |
Feature activation | +9.20e+00 |
Pos logit contributions | |
atography | +5.808 |
IDA | +5.708 |
wrongdoing | +5.535 |
Ru | +5.453 |
kaya | +5.326 |
Neg logit contributions | |
weather | -9.426 |
pite | -8.419 |
ritical | -8.168 |
usage | -8.125 |
Hurricane | -7.640 |
Token | and |
Token ID | 290 |
Feature activation | +9.32e+00 |
Pos logit contributions | |
atography | +5.603 |
IDA | +5.440 |
wrongdoing | +5.407 |
Ru | +5.276 |
Marina | +5.240 |
Neg logit contributions | |
weather | -7.307 |
pite | -6.425 |
ritical | -6.261 |
usage | -6.162 |
Hurricane | -5.836 |
Token | convicted |
Token ID | 9118 |
Feature activation | +1.22e+01 |
Pos logit contributions | |
atography | +5.713 |
IDA | +5.626 |
wrongdoing | +5.513 |
Ru | +5.481 |
Marina | +5.335 |
Neg logit contributions | |
weather | -7.323 |
pite | -6.392 |
ritical | -6.220 |
usage | -6.168 |
Hurricane | -5.799 |
Token |
|
Token ID | 198 |
Feature activation | +2.48e+00 |
Pos logit contributions | |
atography | +2.198 |
IDA | +2.094 |
Ru | +2.062 |
wrongdoing | +2.036 |
irable | +2.030 |
Neg logit contributions | |
weather | -2.580 |
pite | -2.208 |
usage | -2.106 |
ritical | -2.101 |
Hurricane | -2.088 |
Token |
|
Token ID | 198 |
Feature activation | +3.23e+00 |
Pos logit contributions | |
atography | +1.500 |
IDA | +1.431 |
Ru | +1.405 |
wrongdoing | +1.383 |
kaya | +1.377 |
Neg logit contributions | |
weather | -2.081 |
pite | -1.808 |
ritical | -1.737 |
usage | -1.735 |
Hurricane | -1.694 |
Token | The |
Token ID | 464 |
Feature activation | +6.75e+00 |
Pos logit contributions | |
atography | +2.069 |
IDA | +1.951 |
wrongdoing | +1.918 |
Ru | +1.912 |
kaya | +1.897 |
Neg logit contributions | |
weather | -2.634 |
pite | -2.264 |
usage | -2.177 |
ritical | -2.160 |
Hurricane | -2.103 |
Token | three |
Token ID | 1115 |
Feature activation | +8.50e+00 |
Pos logit contributions | |
atography | +4.384 |
IDA | +4.200 |
Ru | +4.151 |
wrongdoing | +4.103 |
kaya | +4.066 |
Neg logit contributions | |
weather | -5.326 |
pite | -4.592 |
ritical | -4.426 |
usage | -4.395 |
Hurricane | -4.254 |
Token | officers |
Token ID | 3790 |
Feature activation | +1.62e+01 |
Pos logit contributions | |
atography | +5.377 |
IDA | +5.210 |
wrongdoing | +5.139 |
Ru | +5.119 |
kaya | +5.062 |
Neg logit contributions | |
weather | -6.683 |
pite | -5.756 |
ritical | -5.595 |
usage | -5.550 |
Hurricane | -5.330 |
Token | who |
Token ID | 508 |
Feature activation | +1.89e+01 |
Pos logit contributions | |
atography | +9.048 |
wrongdoing | +9.031 |
IDA | +8.818 |
Marina | +8.655 |
kaya | +8.645 |
Neg logit contributions | |
weather | -12.020 |
pite | -10.766 |
ritical | -10.663 |
usage | -10.180 |
Hurricane | -9.529 |
Token | conducted |
Token ID | 5952 |
Feature activation | +1.16e+01 |
Pos logit contributions | |
atography | +10.382 |
wrongdoing | +10.225 |
enrich | +9.884 |
furthe | +9.869 |
IDA | +9.808 |
Neg logit contributions | |
weather | -13.541 |
ritical | -12.221 |
pite | -12.083 |
usage | -11.895 |
States | -10.534 |
Token | the |
Token ID | 262 |
Feature activation | +1.11e+01 |
Pos logit contributions | |
atography | +6.492 |
wrongdoing | +6.244 |
IDA | +6.135 |
Marina | +5.942 |
Ru | +5.886 |
Neg logit contributions | |
weather | -9.543 |
pite | -8.575 |
ritical | -8.368 |
usage | -8.164 |
Hurricane | -7.559 |
Token | raid |
Token ID | 9513 |
Feature activation | +1.39e+01 |
Pos logit contributions | |
atography | +5.490 |
wrongdoing | +5.234 |
IDA | +5.030 |
Ru | +4.971 |
Marina | +4.911 |
Neg logit contributions | |
weather | -10.057 |
pite | -8.981 |
ritical | -8.741 |
usage | -8.600 |
Hurricane | -8.158 |
Token | are |
Token ID | 389 |
Feature activation | +1.80e+01 |
Pos logit contributions | |
atography | +8.401 |
wrongdoing | +8.152 |
kaya | +7.934 |
IDA | +7.913 |
Marina | +7.867 |
Neg logit contributions | |
weather | -10.409 |
pite | -9.267 |
ritical | -9.008 |
usage | -8.683 |
Hurricane | -8.234 |
Token | no |
Token ID | 645 |
Feature activation | +6.38e+00 |
Pos logit contributions | |
Marina | +10.576 |
atography | +10.432 |
wrongdoing | +10.214 |
Ru | +10.047 |
IDA | +10.045 |
Neg logit contributions | |
weather | -12.429 |
ritical | -11.309 |
pite | -11.106 |
usage | -10.533 |
okingly | -9.517 |
Token | ad |
Token ID | 324 |
Feature activation | +1.64e+00 |
Pos logit contributions | |
atography | +3.521 |
IDA | +3.371 |
Ru | +3.319 |
wrongdoing | +3.257 |
kaya | +3.246 |
Neg logit contributions | |
weather | -4.766 |
pite | -4.141 |
ritical | -3.989 |
usage | -3.967 |
Hurricane | -3.867 |
Token | Mas |
Token ID | 11066 |
Feature activation | +2.85e+00 |
Pos logit contributions | |
atography | +1.052 |
IDA | +1.013 |
Ru | +0.989 |
wrongdoing | +0.982 |
irable | +0.963 |
Neg logit contributions | |
weather | -1.326 |
pite | -1.149 |
ritical | -1.104 |
usage | -1.089 |
Hurricane | -1.073 |
Token | ro |
Token ID | 305 |
Feature activation | +4.66e+00 |
Pos logit contributions | |
atography | +1.821 |
IDA | +1.758 |
Ru | +1.717 |
wrongdoing | +1.716 |
kaya | +1.660 |
Neg logit contributions | |
weather | -2.319 |
pite | -2.020 |
ritical | -1.938 |
usage | -1.927 |
Hurricane | -1.855 |
Token | or |
Token ID | 273 |
Feature activation | +2.60e+00 |
Pos logit contributions | |
atography | +3.417 |
IDA | +3.287 |
Ru | +3.229 |
wrongdoing | +3.205 |
kaya | +3.166 |
Neg logit contributions | |
weather | -3.363 |
pite | -2.834 |
ritical | -2.717 |
usage | -2.699 |
Hurricane | -2.600 |
Token | Khan |
Token ID | 11356 |
Feature activation | +6.02e+00 |
Pos logit contributions | |
atography | +1.719 |
IDA | +1.647 |
Ru | +1.599 |
wrongdoing | +1.596 |
irable | +1.583 |
Neg logit contributions | |
weather | -2.078 |
pite | -1.775 |
ritical | -1.703 |
usage | -1.690 |
Hurricane | -1.657 |
Token | ; |
Token ID | 26 |
Feature activation | +1.88e+01 |
Pos logit contributions | |
atography | +3.423 |
IDA | +3.264 |
Ru | +3.212 |
wrongdoing | +3.148 |
kaya | +3.140 |
Neg logit contributions | |
weather | -5.261 |
pite | -4.609 |
ritical | -4.449 |
usage | -4.425 |
Hurricane | -4.319 |
Token | Writing |
Token ID | 22183 |
Feature activation | +1.49e+01 |
Pos logit contributions | |
atography | +7.033 |
Marina | +6.492 |
IDA | +6.393 |
enrich | +6.260 |
wrongdoing | +6.230 |
Neg logit contributions | |
weather | -17.582 |
pite | -15.976 |
ritical | -14.886 |
usage | -14.608 |
Hurricane | -14.532 |
Token | by |
Token ID | 416 |
Feature activation | +2.23e+01 |
Pos logit contributions | |
Marina | +5.525 |
atography | +5.517 |
wrongdoing | +5.157 |
Ru | +5.099 |
Transparency | +4.889 |
Neg logit contributions | |
weather | -14.839 |
pite | -13.391 |
ritical | -12.964 |
usage | -12.892 |
reconc | -12.143 |
Token | Z |
Token ID | 1168 |
Feature activation | +1.75e+01 |
Pos logit contributions | |
Marina | +12.496 |
atography | +11.332 |
Transparency | +11.271 |
Ru | +11.071 |
IDA | +10.663 |
Neg logit contributions | |
weather | -15.564 |
pite | -13.967 |
ritical | -13.639 |
usage | -13.427 |
ricanes | -12.338 |
Token | eb |
Token ID | 1765 |
Feature activation | +4.84e+00 |
Pos logit contributions | |
atography | +10.864 |
IDA | +10.452 |
kaya | +10.375 |
Ru | +10.188 |
Marina | +10.001 |
Neg logit contributions | |
weather | -11.857 |
usage | -9.749 |
Hurricane | -9.684 |
pite | -9.604 |
ritical | -9.369 |
Token | a |
Token ID | 64 |
Feature activation | +2.10e+00 |
Pos logit contributions | |
atography | +2.990 |
IDA | +2.864 |
Ru | +2.819 |
wrongdoing | +2.770 |
kaya | +2.756 |
Neg logit contributions | |
weather | -3.998 |
pite | -3.470 |
ritical | -3.343 |
usage | -3.324 |
Hurricane | -3.239 |
Token | the |
Token ID | 262 |
Feature activation | -2.77e+00 |
Pos logit contributions | |
weather | +2.567 |
pite | +2.193 |
usage | +2.121 |
Hurricane | +2.098 |
ritical | +2.071 |
Neg logit contributions | |
atography | -2.582 |
IDA | -2.466 |
Ru | -2.427 |
irable | -2.376 |
kaya | -2.356 |
Token | skills |
Token ID | 4678 |
Feature activation | +4.62e+00 |
Pos logit contributions | |
weather | +1.835 |
pite | +1.554 |
Hurricane | +1.488 |
ritical | +1.482 |
usage | +1.480 |
Neg logit contributions | |
atography | -2.058 |
IDA | -2.029 |
Ru | -2.006 |
arthed | -1.954 |
kaya | -1.952 |
Token | they |
Token ID | 484 |
Feature activation | +8.58e+00 |
Pos logit contributions | |
atography | +3.249 |
IDA | +3.130 |
Ru | +3.086 |
wrongdoing | +3.039 |
kaya | +3.027 |
Neg logit contributions | |
weather | -3.425 |
pite | -2.920 |
ritical | -2.797 |
usage | -2.782 |
Hurricane | -2.706 |
Token | need |
Token ID | 761 |
Feature activation | -1.60e+00 |
Pos logit contributions | |
atography | +5.530 |
IDA | +5.251 |
Ru | +5.227 |
kaya | +5.177 |
wrongdoing | +5.080 |
Neg logit contributions | |
weather | -6.664 |
pite | -5.820 |
ritical | -5.597 |
usage | -5.495 |
Hurricane | -5.372 |
Token | to |
Token ID | 284 |
Feature activation | +3.49e+00 |
Pos logit contributions | |
weather | +1.146 |
pite | +0.951 |
usage | +0.916 |
ritical | +0.908 |
Hurricane | +0.901 |
Neg logit contributions | |
atography | -1.134 |
IDA | -1.111 |
Ru | -1.101 |
wrongdoing | -1.078 |
irable | -1.065 |
Token | enter |
Token ID | 3802 |
Feature activation | +1.88e+01 |
Pos logit contributions | |
atography | +2.374 |
IDA | +2.283 |
Ru | +2.249 |
wrongdoing | +2.204 |
kaya | +2.191 |
Neg logit contributions | |
weather | -2.666 |
pite | -2.288 |
ritical | -2.184 |
usage | -2.181 |
Hurricane | -2.136 |
Token | the |
Token ID | 262 |
Feature activation | +9.78e+00 |
Pos logit contributions | |
atography | +9.348 |
IDA | +9.322 |
wrongdoing | +9.076 |
enrich | +9.061 |
vana | +9.035 |
Neg logit contributions | |
weather | -14.587 |
pite | -12.806 |
ritical | -12.639 |
usage | -12.219 |
Hurricane | -11.481 |
Token | predomin |
Token ID | 15960 |
Feature activation | -1.76e+00 |
Pos logit contributions | |
atography | +5.670 |
IDA | +5.400 |
wrongdoing | +5.343 |
Ru | +5.311 |
irable | +5.250 |
Neg logit contributions | |
weather | -8.144 |
pite | -7.017 |
ritical | -6.814 |
usage | -6.800 |
Hurricane | -6.498 |
Token | ately |
Token ID | 1286 |
Feature activation | -3.23e-01 |
Pos logit contributions | |
weather | +0.912 |
pite | +0.721 |
ritical | +0.660 |
usage | +0.659 |
Hurricane | +0.622 |
Neg logit contributions | |
atography | -1.615 |
IDA | -1.598 |
Ru | -1.578 |
wrongdoing | -1.568 |
kaya | -1.546 |
Token | male |
Token ID | 4257 |
Feature activation | +8.59e+00 |
Pos logit contributions | |
weather | +0.269 |
pite | +0.231 |
ritical | +0.225 |
usage | +0.223 |
Hurricane | +0.222 |
Neg logit contributions | |
atography | -0.200 |
IDA | -0.194 |
Ru | -0.187 |
kaya | -0.181 |
wrongdoing | -0.180 |
Token | - |
Token ID | 12 |
Feature activation | +4.88e+00 |
Pos logit contributions | |
atography | +5.178 |
IDA | +4.960 |
wrongdoing | +4.869 |
Ru | +4.831 |
kaya | +4.792 |
Neg logit contributions | |
weather | -7.020 |
pite | -6.100 |
ritical | -5.889 |
usage | -5.832 |
Hurricane | -5.626 |
Token | Photo |
Token ID | 6191 |
Feature activation | +1.54e+01 |
Pos logit contributions | |
IDA | +7.445 |
Ru | +7.369 |
atography | +6.954 |
Marina | +6.745 |
Step | +6.713 |
Neg logit contributions | |
weather | -12.202 |
pite | -11.029 |
ritical | -10.784 |
usage | -10.261 |
Hurricane | -10.061 |
Token | 11 |
Token ID | 1157 |
Feature activation | +7.08e+00 |
Pos logit contributions | |
atography | +6.706 |
Marina | +6.339 |
Ru | +6.258 |
IDA | +6.121 |
→ | +5.969 |
Neg logit contributions | |
weather | -13.896 |
pite | -12.482 |
ritical | -11.979 |
usage | -11.594 |
Hurricane | -11.193 |
Token | : |
Token ID | 25 |
Feature activation | +1.12e+01 |
Pos logit contributions | |
atography | +4.425 |
IDA | +4.250 |
Ru | +4.194 |
wrongdoing | +4.099 |
kaya | +4.095 |
Neg logit contributions | |
weather | -5.717 |
pite | -4.963 |
ritical | -4.794 |
usage | -4.746 |
Hurricane | -4.619 |
Token | File |
Token ID | 9220 |
Feature activation | +4.57e+00 |
Pos logit contributions | |
atography | +6.967 |
IDA | +6.847 |
Ru | +6.725 |
Marina | +6.660 |
kaya | +6.520 |
Neg logit contributions | |
weather | -8.713 |
pite | -7.542 |
ritical | -7.350 |
usage | -7.226 |
Hurricane | -6.788 |
Token | photo |
Token ID | 4590 |
Feature activation | +7.91e+00 |
Pos logit contributions | |
atography | +2.695 |
IDA | +2.574 |
Ru | +2.526 |
wrongdoing | +2.488 |
kaya | +2.482 |
Neg logit contributions | |
weather | -3.926 |
pite | -3.422 |
ritical | -3.293 |
usage | -3.282 |
Hurricane | -3.210 |
Token | by |
Token ID | 416 |
Feature activation | +1.87e+01 |
Pos logit contributions | |
atography | +4.470 |
IDA | +4.277 |
Ru | +4.180 |
wrongdoing | +4.079 |
kaya | +4.069 |
Neg logit contributions | |
weather | -6.815 |
pite | -5.947 |
ritical | -5.787 |
usage | -5.698 |
Hurricane | -5.554 |
Token | J |
Token ID | 449 |
Feature activation | +8.12e+00 |
Pos logit contributions | |
Marina | +11.107 |
IDA | +10.901 |
atography | +10.715 |
Transparency | +10.466 |
Ru | +10.240 |
Neg logit contributions | |
weather | -12.916 |
pite | -11.626 |
ritical | -11.348 |
usage | -10.882 |
ricanes | -10.115 |
Token | iji |
Token ID | 20770 |
Feature activation | +1.92e+00 |
Pos logit contributions | |
atography | +5.256 |
IDA | +5.138 |
Ru | +4.976 |
kaya | +4.856 |
wrongdoing | +4.832 |
Neg logit contributions | |
weather | -6.279 |
pite | -5.446 |
ritical | -5.226 |
usage | -5.152 |
Hurricane | -5.098 |
Token | Press |
Token ID | 4332 |
Feature activation | +7.48e+00 |
Pos logit contributions | |
atography | +1.219 |
IDA | +1.160 |
Ru | +1.136 |
wrongdoing | +1.125 |
irable | +1.115 |
Neg logit contributions | |
weather | -1.563 |
pite | -1.340 |
usage | -1.294 |
ritical | -1.287 |
Hurricane | -1.275 |
Token | , |
Token ID | 11 |
Feature activation | +8.40e+00 |
Pos logit contributions | |
atography | +4.614 |
IDA | +4.468 |
Ru | +4.383 |
kaya | +4.344 |
wrongdoing | +4.295 |
Neg logit contributions | |
weather | -5.994 |
pite | -5.222 |
ritical | -5.064 |
usage | -4.967 |
Hurricane | -4.867 |
Token | AFP |
Token ID | 8916 |
Feature activation | +8.00e+00 |
Pos logit contributions | |
atography | +5.347 |
IDA | +5.212 |
Ru | +5.104 |
wrongdoing | +4.998 |
kaya | +4.988 |
Neg logit contributions | |
weather | -6.603 |
pite | -5.707 |
ritical | -5.527 |
usage | -5.418 |
Hurricane | -5.231 |
Token | session |
Token ID | 6246 |
Feature activation | +6.75e-01 |
Pos logit contributions | |
weather | +1.664 |
pite | +1.409 |
usage | +1.357 |
ritical | +1.340 |
Hurricane | +1.338 |
Neg logit contributions | |
atography | -1.457 |
IDA | -1.382 |
irable | -1.357 |
Ru | -1.350 |
wrongdoing | -1.317 |
Token | but |
Token ID | 475 |
Feature activation | +4.38e+00 |
Pos logit contributions | |
atography | +0.457 |
IDA | +0.435 |
Ru | +0.434 |
irable | +0.423 |
wrongdoing | +0.421 |
Neg logit contributions | |
weather | -0.523 |
pite | -0.444 |
usage | -0.428 |
ritical | -0.422 |
Hurricane | -0.420 |
Token | there |
Token ID | 612 |
Feature activation | +4.42e+00 |
Pos logit contributions | |
atography | +3.006 |
IDA | +2.889 |
Ru | +2.849 |
wrongdoing | +2.802 |
kaya | +2.795 |
Neg logit contributions | |
weather | -3.320 |
pite | -2.842 |
ritical | -2.722 |
usage | -2.711 |
Hurricane | -2.636 |
Token | is |
Token ID | 318 |
Feature activation | +1.42e+01 |
Pos logit contributions | |
atography | +2.901 |
IDA | +2.786 |
Ru | +2.744 |
wrongdoing | +2.698 |
kaya | +2.689 |
Neg logit contributions | |
weather | -3.476 |
pite | -2.993 |
ritical | -2.876 |
usage | -2.861 |
Hurricane | -2.783 |
Token | so |
Token ID | 523 |
Feature activation | +1.22e+01 |
Pos logit contributions | |
atography | +8.711 |
IDA | +8.629 |
Ru | +8.495 |
wrongdoing | +8.437 |
kaya | +8.201 |
Neg logit contributions | |
weather | -10.380 |
pite | -9.007 |
ritical | -8.708 |
usage | -8.437 |
Hurricane | -8.133 |
Token | much |
Token ID | 881 |
Feature activation | +1.86e+01 |
Pos logit contributions | |
atography | +5.953 |
IDA | +5.617 |
Ru | +5.563 |
wrongdoing | +5.427 |
kaya | +5.373 |
Neg logit contributions | |
weather | -11.030 |
pite | -9.631 |
ritical | -9.394 |
usage | -9.354 |
Hurricane | -9.105 |
Token | more |
Token ID | 517 |
Feature activation | +1.46e+01 |
Pos logit contributions | |
wrongdoing | +10.118 |
atography | +9.821 |
IDA | +9.629 |
iness | +9.533 |
enrich | +9.476 |
Neg logit contributions | |
weather | -14.143 |
pite | -12.079 |
ritical | -11.891 |
usage | -11.516 |
okingly | -10.998 |
Token | to |
Token ID | 284 |
Feature activation | +3.87e+00 |
Pos logit contributions | |
atography | +8.287 |
wrongdoing | +7.977 |
IDA | +7.950 |
enrich | +7.658 |
Ru | +7.658 |
Neg logit contributions | |
weather | -11.709 |
pite | -10.054 |
ritical | -9.817 |
usage | -9.532 |
Hurricane | -9.156 |
Token | it |
Token ID | 340 |
Feature activation | +4.22e+00 |
Pos logit contributions | |
atography | +2.993 |
IDA | +2.877 |
Ru | +2.841 |
wrongdoing | +2.798 |
kaya | +2.796 |
Neg logit contributions | |
weather | -2.606 |
pite | -2.183 |
ritical | -2.077 |
usage | -2.072 |
Hurricane | -2.016 |
Token | than |
Token ID | 621 |
Feature activation | +9.61e+00 |
Pos logit contributions | |
atography | +2.798 |
IDA | +2.686 |
Ru | +2.648 |
wrongdoing | +2.600 |
kaya | +2.590 |
Neg logit contributions | |
weather | -3.311 |
pite | -2.848 |
ritical | -2.730 |
usage | -2.725 |
Hurricane | -2.654 |
Token | that |
Token ID | 326 |
Feature activation | +5.18e+00 |
Pos logit contributions | |
atography | +5.681 |
IDA | +5.443 |
wrongdoing | +5.421 |
Ru | +5.388 |
kaya | +5.241 |
Neg logit contributions | |
weather | -7.934 |
pite | -6.824 |
ritical | -6.636 |
usage | -6.558 |
Hurricane | -6.273 |
Token | I |
Token ID | 314 |
Feature activation | +2.08e+00 |
Pos logit contributions | |
atography | +4.223 |
IDA | +4.064 |
Ru | +4.011 |
wrongdoing | +3.945 |
kaya | +3.934 |
Neg logit contributions | |
weather | -4.531 |
pite | -3.870 |
ritical | -3.709 |
usage | -3.689 |
Hurricane | -3.583 |
Token | add |
Token ID | 751 |
Feature activation | +2.47e+00 |
Pos logit contributions | |
atography | +1.519 |
IDA | +1.473 |
Ru | +1.457 |
wrongdoing | +1.427 |
kaya | +1.413 |
Neg logit contributions | |
weather | -1.484 |
pite | -1.255 |
usage | -1.207 |
ritical | -1.196 |
Hurricane | -1.170 |
Token | the |
Token ID | 262 |
Feature activation | +1.96e+00 |
Pos logit contributions | |
atography | +1.800 |
IDA | +1.729 |
Ru | +1.727 |
kaya | +1.696 |
irable | +1.688 |
Neg logit contributions | |
weather | -1.787 |
pite | -1.503 |
usage | -1.442 |
ritical | -1.439 |
Hurricane | -1.403 |
Token | comparison |
Token ID | 7208 |
Feature activation | -8.30e-01 |
Pos logit contributions | |
atography | +1.359 |
IDA | +1.322 |
Ru | +1.308 |
irable | +1.279 |
kaya | +1.275 |
Neg logit contributions | |
weather | -1.472 |
pite | -1.242 |
usage | -1.211 |
ritical | -1.204 |
Hurricane | -1.168 |
Token | code |
Token ID | 2438 |
Feature activation | +1.01e+00 |
Pos logit contributions | |
weather | +0.636 |
pite | +0.540 |
usage | +0.530 |
ritical | +0.530 |
Hurricane | +0.514 |
Neg logit contributions | |
atography | -0.558 |
IDA | -0.532 |
Ru | -0.529 |
irable | -0.524 |
kaya | -0.520 |
Token | like |
Token ID | 588 |
Feature activation | +1.86e+01 |
Pos logit contributions | |
atography | +0.696 |
IDA | +0.677 |
Ru | +0.671 |
irable | +0.654 |
wrongdoing | +0.653 |
Neg logit contributions | |
weather | -0.756 |
pite | -0.647 |
usage | -0.630 |
ritical | -0.623 |
Hurricane | -0.606 |
Token | in |
Token ID | 287 |
Feature activation | +1.79e+01 |
Pos logit contributions | |
atography | +9.566 |
IDA | +9.348 |
Transparency | +9.153 |
wrongdoing | +8.938 |
Marina | +8.876 |
Neg logit contributions | |
weather | -14.490 |
pite | -12.839 |
ritical | -12.263 |
usage | -11.873 |
awks | -11.411 |
Token | the |
Token ID | 262 |
Feature activation | +1.33e+01 |
Pos logit contributions | |
atography | +9.843 |
IDA | +9.520 |
Transparency | +9.384 |
kaya | +9.153 |
Ru | +9.121 |
Neg logit contributions | |
weather | -13.304 |
pite | -11.707 |
ritical | -11.291 |
usage | -10.842 |
Hurricane | -10.537 |
Token | above |
Token ID | 2029 |
Feature activation | +8.63e+00 |
Pos logit contributions | |
atography | +7.527 |
IDA | +7.283 |
wrongdoing | +7.094 |
Transparency | +7.069 |
Ru | +7.011 |
Neg logit contributions | |
weather | -10.565 |
pite | -9.330 |
ritical | -8.872 |
usage | -8.585 |
Hurricane | -8.360 |
Token | code |
Token ID | 2438 |
Feature activation | +5.82e+00 |
Pos logit contributions | |
atography | +5.522 |
IDA | +5.345 |
Ru | +5.217 |
wrongdoing | +5.190 |
kaya | +5.132 |
Neg logit contributions | |
weather | -6.748 |
pite | -5.844 |
ritical | -5.582 |
usage | -5.489 |
Hurricane | -5.338 |
Token | snippet |
Token ID | 39442 |
Feature activation | +7.42e+00 |
Pos logit contributions | |
atography | +3.985 |
IDA | +3.834 |
Ru | +3.781 |
wrongdoing | +3.720 |
kaya | +3.707 |
Neg logit contributions | |
weather | -4.416 |
pite | -3.782 |
ritical | -3.627 |
usage | -3.608 |
Hurricane | -3.504 |
Token | Example |
Token ID | 16281 |
Feature activation | -4.16e+00 |
Pos logit contributions | |
weather | +2.496 |
usage | +2.041 |
pite | +2.038 |
ritical | +1.856 |
Hurricane | +1.844 |
Neg logit contributions | |
atography | -2.949 |
enrich | -2.752 |
wrongdoing | -2.724 |
irable | -2.680 |
enriched | -2.642 |
Token | : |
Token ID | 25 |
Feature activation | -1.14e+01 |
Pos logit contributions | |
weather | +3.025 |
pite | +2.507 |
Hurricane | +2.462 |
usage | +2.452 |
Pend | +2.354 |
Neg logit contributions | |
atography | -2.991 |
irable | -2.735 |
enrich | -2.724 |
IDA | -2.704 |
vana | -2.689 |
Token | Because |
Token ID | 4362 |
Feature activation | -6.87e+00 |
Pos logit contributions | |
weather | +3.003 |
Hurricane | +2.332 |
pite | +1.896 |
Pend | +1.803 |
Welsh | +1.701 |
Neg logit contributions | |
atography | -12.008 |
irable | -11.863 |
enrich | -11.852 |
challeng | -11.527 |
exhib | -11.506 |
Token | of |
Token ID | 286 |
Feature activation | -1.12e+01 |
Pos logit contributions | |
weather | +4.054 |
pite | +3.306 |
Hurricane | +3.207 |
usage | +3.008 |
Pend | +2.909 |
Neg logit contributions | |
atography | -5.793 |
irable | -5.275 |
IDA | -5.274 |
challeng | -5.258 |
enrich | -5.239 |
Token | a |
Token ID | 257 |
Feature activation | -1.14e+01 |
Pos logit contributions | |
weather | +4.859 |
Hurricane | +3.798 |
pite | +3.668 |
timed | +3.519 |
misplaced | +3.408 |
Neg logit contributions | |
atography | -10.094 |
�� | -9.777 |
challeng | -9.666 |
vana | -9.648 |
IDA | -9.519 |
Token | failed |
Token ID | 4054 |
Feature activation | -2.29e+01 |
Pos logit contributions | |
weather | +5.878 |
misplaced | +5.266 |
timed | +5.145 |
Hurricane | +4.871 |
pite | +4.812 |
Neg logit contributions | |
atography | -9.503 |
�� | -8.853 |
vana | -8.793 |
Ru | -8.685 |
IDA | -8.678 |
Token | will |
Token ID | 481 |
Feature activation | -4.68e+00 |
Pos logit contributions | |
weather | +5.205 |
misplaced | +4.694 |
timed | +4.677 |
Hurricane | +4.048 |
sled | +3.893 |
Neg logit contributions | |
�� | -21.891 |
ThumbnailImage | -20.982 |
atography | -20.804 |
IDA | -20.770 |
arthed | -19.905 |
Token | save |
Token ID | 3613 |
Feature activation | -2.94e+00 |
Pos logit contributions | |
weather | +2.960 |
pite | +2.429 |
Hurricane | +2.327 |
usage | +2.321 |
ritical | +2.311 |
Neg logit contributions | |
atography | -3.701 |
irable | -3.478 |
IDA | -3.474 |
Ru | -3.465 |
kaya | -3.427 |
Token | , |
Token ID | 11 |
Feature activation | -8.63e+00 |
Pos logit contributions | |
weather | +1.865 |
pite | +1.484 |
Hurricane | +1.435 |
usage | +1.406 |
Pend | +1.329 |
Neg logit contributions | |
atography | -2.440 |
IDA | -2.280 |
irable | -2.247 |
vana | -2.240 |
enrich | -2.230 |
Token | your |
Token ID | 534 |
Feature activation | -3.34e+00 |
Pos logit contributions | |
weather | +4.101 |
Hurricane | +3.237 |
pite | +3.003 |
usage | +2.952 |
Pend | +2.950 |
Neg logit contributions | |
atography | -8.077 |
vana | -7.497 |
IDA | -7.453 |
irable | -7.419 |
�� | -7.197 |
Token | character |
Token ID | 2095 |
Feature activation | -7.25e+00 |
Pos logit contributions | |
weather | +2.301 |
pite | +1.950 |
Hurricane | +1.835 |
ritical | +1.833 |
usage | +1.815 |
Neg logit contributions | |
atography | -2.541 |
IDA | -2.379 |
Ru | -2.293 |
irable | -2.292 |
vana | -2.257 |
Token | several |
Token ID | 1811 |
Feature activation | -6.22e-01 |
Pos logit contributions | |
weather | +0.825 |
pite | +0.707 |
usage | +0.674 |
ritical | +0.673 |
Hurricane | +0.670 |
Neg logit contributions | |
atography | -0.755 |
IDA | -0.719 |
Ru | -0.715 |
kaya | -0.694 |
irable | -0.691 |
Token | people |
Token ID | 661 |
Feature activation | -1.31e+00 |
Pos logit contributions | |
weather | +0.482 |
pite | +0.415 |
usage | +0.397 |
ritical | +0.396 |
Hurricane | +0.390 |
Neg logit contributions | |
atography | -0.414 |
Ru | -0.396 |
IDA | -0.395 |
kaya | -0.385 |
irable | -0.381 |
Token | seemed |
Token ID | 3947 |
Feature activation | -2.48e+00 |
Pos logit contributions | |
weather | +0.997 |
pite | +0.851 |
usage | +0.810 |
ritical | +0.802 |
Hurricane | +0.794 |
Neg logit contributions | |
atography | -0.882 |
IDA | -0.857 |
Ru | -0.856 |
kaya | -0.821 |
irable | -0.816 |
Token | to |
Token ID | 284 |
Feature activation | -1.82e-01 |
Pos logit contributions | |
weather | +1.829 |
pite | +1.579 |
usage | +1.483 |
ritical | +1.472 |
Hurricane | +1.462 |
Neg logit contributions | |
atography | -1.686 |
IDA | -1.656 |
Ru | -1.627 |
kaya | -1.617 |
irable | -1.612 |
Token | have |
Token ID | 423 |
Feature activation | -4.79e+00 |
Pos logit contributions | |
weather | +0.120 |
pite | +0.100 |
ritical | +0.094 |
usage | +0.094 |
Hurricane | +0.092 |
Neg logit contributions | |
atography | -0.143 |
IDA | -0.141 |
Ru | -0.138 |
irable | -0.134 |
kaya | -0.133 |
Token | bad |
Token ID | 2089 |
Feature activation | -2.26e+01 |
Pos logit contributions | |
weather | +3.302 |
pite | +2.764 |
Hurricane | +2.700 |
usage | +2.689 |
misplaced | +2.677 |
Neg logit contributions | |
atography | -3.438 |
Ru | -3.293 |
IDA | -3.269 |
kaya | -3.246 |
irable | -3.229 |
Token | copies |
Token ID | 9088 |
Feature activation | -1.20e+01 |
Pos logit contributions | |
weather | +12.670 |
usage | +10.829 |
timed | +10.750 |
misplaced | +10.469 |
pite | +10.319 |
Neg logit contributions | |
Ru | -14.032 |
arthed | -13.060 |
kaya | -12.916 |
irable | -12.909 |
atography | -12.798 |
Token | . |
Token ID | 13 |
Feature activation | -8.86e-02 |
Pos logit contributions | |
weather | +7.201 |
pite | +6.071 |
misplaced | +5.868 |
timed | +5.665 |
Hurricane | +5.607 |
Neg logit contributions | |
Ru | -8.818 |
atography | -8.815 |
vana | -8.334 |
IDA | -8.311 |
arthed | -8.294 |
Token |
|
Token ID | 198 |
Feature activation | +2.97e+00 |
Pos logit contributions | |
weather | +0.064 |
pite | +0.054 |
Hurricane | +0.052 |
usage | +0.051 |
ritical | +0.050 |
Neg logit contributions | |
atography | -0.064 |
IDA | -0.061 |
irable | -0.060 |
wrongdoing | -0.060 |
kaya | -0.060 |
Token |
|
Token ID | 198 |
Feature activation | +1.03e+00 |
Pos logit contributions | |
atography | +1.866 |
IDA | +1.753 |
wrongdoing | +1.735 |
kaya | +1.722 |
Ru | +1.720 |
Neg logit contributions | |
weather | -2.455 |
pite | -2.129 |
usage | -2.040 |
ritical | -2.029 |
Hurricane | -1.984 |
Token | Of |
Token ID | 5189 |
Feature activation | +5.42e-01 |
Pos logit contributions | |
atography | +0.647 |
wrongdoing | +0.601 |
IDA | +0.598 |
kaya | +0.594 |
irable | +0.592 |
Neg logit contributions | |
weather | -0.860 |
pite | -0.742 |
usage | -0.717 |
ritical | -0.704 |
Hurricane | -0.686 |
Token | loader |
Token ID | 29356 |
Feature activation | -7.96e+00 |
Pos logit contributions | |
weather | +2.542 |
pite | +2.014 |
usage | +1.996 |
ritical | +1.980 |
Hurricane | +1.863 |
Neg logit contributions | |
atography | -3.419 |
IDA | -3.360 |
Ru | -3.295 |
wrongdoing | -3.209 |
kaya | -3.202 |
Token | in |
Token ID | 287 |
Feature activation | -7.62e-01 |
Pos logit contributions | |
weather | +5.044 |
pite | +4.141 |
usage | +4.111 |
Hurricane | +4.109 |
sled | +3.979 |
Neg logit contributions | |
atography | -5.986 |
Ru | -5.791 |
IDA | -5.723 |
arthed | -5.649 |
irable | -5.581 |
Token | this |
Token ID | 428 |
Feature activation | -1.41e+00 |
Pos logit contributions | |
weather | +0.548 |
pite | +0.466 |
usage | +0.453 |
ritical | +0.444 |
Hurricane | +0.441 |
Neg logit contributions | |
atography | -0.552 |
Ru | -0.528 |
IDA | -0.527 |
kaya | -0.511 |
wrongdoing | -0.509 |
Token | watch |
Token ID | 2342 |
Feature activation | -2.10e+00 |
Pos logit contributions | |
weather | +0.981 |
pite | +0.833 |
usage | +0.802 |
Hurricane | +0.800 |
ritical | +0.783 |
Neg logit contributions | |
atography | -1.046 |
Ru | -1.000 |
IDA | -0.987 |
kaya | -0.970 |
irable | -0.963 |
Token | is |
Token ID | 318 |
Feature activation | -6.99e+00 |
Pos logit contributions | |
weather | +1.488 |
pite | +1.214 |
ritical | +1.184 |
usage | +1.175 |
Hurricane | +1.138 |
Neg logit contributions | |
atography | -1.568 |
IDA | -1.536 |
Ru | -1.501 |
wrongdoing | -1.473 |
kaya | -1.465 |
Token | poorly |
Token ID | 13455 |
Feature activation | -2.23e+01 |
Pos logit contributions | |
weather | +4.761 |
pite | +4.050 |
Hurricane | +3.908 |
timed | +3.866 |
usage | +3.865 |
Neg logit contributions | |
atography | -4.942 |
Ru | -4.735 |
vana | -4.664 |
arthed | -4.632 |
IDA | -4.595 |
Token | coded |
Token ID | 30817 |
Feature activation | -3.07e+00 |
Pos logit contributions | |
weather | +11.794 |
timed | +11.167 |
pite | +9.534 |
usage | +9.368 |
Hurricane | +9.321 |
Neg logit contributions | |
atography | -14.256 |
Ru | -14.144 |
kaya | -13.532 |
NetMessage | -13.527 |
vana | -13.304 |
Token | The |
Token ID | 383 |
Feature activation | -1.79e+00 |
Pos logit contributions | |
weather | +2.184 |
pite | +1.887 |
usage | +1.789 |
ritical | +1.781 |
Hurricane | +1.755 |
Neg logit contributions | |
atography | -2.211 |
IDA | -2.112 |
Ru | -2.097 |
kaya | -2.035 |
wrongdoing | -2.034 |
Token | boot |
Token ID | 6297 |
Feature activation | -4.54e+00 |
Pos logit contributions | |
weather | +1.199 |
pite | +0.996 |
usage | +0.957 |
ritical | +0.956 |
Hurricane | +0.945 |
Neg logit contributions | |
atography | -1.384 |
IDA | -1.330 |
Ru | -1.330 |
irable | -1.287 |
kaya | -1.285 |
Token | loader |
Token ID | 29356 |
Feature activation | -4.06e+00 |
Pos logit contributions | |
weather | +2.241 |
pite | +1.651 |
ritical | +1.631 |
usage | +1.628 |
Hurricane | +1.508 |
Neg logit contributions | |
atography | -4.272 |
IDA | -4.174 |
Ru | -4.129 |
kaya | -4.040 |
wrongdoing | -4.039 |
Token | res |
Token ID | 581 |
Feature activation | -7.28e+00 |
Pos logit contributions | |
weather | +2.817 |
pite | +2.410 |
usage | +2.361 |
Hurricane | +2.295 |
ritical | +2.264 |
Neg logit contributions | |
atography | -2.945 |
IDA | -2.819 |
Ru | -2.819 |
vana | -2.736 |
irable | -2.722 |
Token | simple |
Token ID | 2829 |
Feature activation | +6.62e-01 |
Pos logit contributions | |
weather | +2.426 |
pite | +2.124 |
Hurricane | +2.028 |
usage | +1.983 |
ritical | +1.971 |
Neg logit contributions | |
atography | -2.207 |
IDA | -2.080 |
Ru | -2.036 |
kaya | -2.027 |
irable | -1.978 |
Token | to |
Token ID | 284 |
Feature activation | -8.20e+00 |
Pos logit contributions | |
atography | +0.454 |
IDA | +0.443 |
Ru | +0.434 |
kaya | +0.426 |
irable | +0.420 |
Neg logit contributions | |
weather | -0.506 |
pite | -0.426 |
usage | -0.408 |
ritical | -0.405 |
Hurricane | -0.404 |
Token | mod |
Token ID | 953 |
Feature activation | -2.37e+00 |
Pos logit contributions | |
weather | +5.045 |
pite | +4.305 |
Hurricane | +4.299 |
Pend | +4.126 |
timed | +4.125 |
Neg logit contributions | |
IDA | -6.078 |
atography | -5.999 |
Ru | -5.834 |
arthed | -5.638 |
kaya | -5.593 |
Token | , |
Token ID | 11 |
Feature activation | -2.80e+00 |
Pos logit contributions | |
weather | +1.749 |
pite | +1.462 |
Hurricane | +1.391 |
usage | +1.381 |
ritical | +1.359 |
Neg logit contributions | |
atography | -1.697 |
IDA | -1.605 |
Ru | -1.586 |
wrongdoing | -1.565 |
kaya | -1.549 |
Token | and |
Token ID | 290 |
Feature activation | -1.11e+01 |
Pos logit contributions | |
weather | +2.064 |
pite | +1.766 |
Hurricane | +1.682 |
usage | +1.667 |
ritical | +1.643 |
Neg logit contributions | |
atography | -1.944 |
IDA | -1.890 |
Ru | -1.844 |
irable | -1.838 |
kaya | -1.828 |
Token | partly |
Token ID | 11476 |
Feature activation | -2.23e+01 |
Pos logit contributions | |
weather | +6.600 |
pite | +5.835 |
Hurricane | +5.576 |
usage | +5.227 |
timed | +5.204 |
Neg logit contributions | |
atography | -8.117 |
IDA | -7.852 |
kaya | -7.600 |
irable | -7.593 |
Ru | -7.569 |
Token | due |
Token ID | 2233 |
Feature activation | -9.21e+00 |
Pos logit contributions | |
weather | +10.530 |
pite | +9.463 |
Hurricane | +9.190 |
timed | +8.724 |
misplaced | +8.572 |
Neg logit contributions | |
atography | -15.163 |
IDA | -14.613 |
Ru | -14.167 |
kaya | -13.998 |
irable | -13.912 |
Token | to |
Token ID | 284 |
Feature activation | -9.32e+00 |
Pos logit contributions | |
weather | +6.331 |
pite | +5.436 |
Hurricane | +5.204 |
ritical | +5.034 |
usage | +4.897 |
Neg logit contributions | |
atography | -6.459 |
IDA | -6.176 |
kaya | -6.139 |
Ru | -6.124 |
irable | -6.090 |
Token | bits |
Token ID | 10340 |
Feature activation | -1.41e+00 |
Pos logit contributions | |
weather | +5.844 |
pite | +5.164 |
Hurricane | +5.066 |
Pend | +4.742 |
usage | +4.625 |
Neg logit contributions | |
atography | -6.859 |
Ru | -6.391 |
kaya | -6.388 |
IDA | -6.345 |
irable | -6.310 |
Token | of |
Token ID | 286 |
Feature activation | +2.25e+00 |
Pos logit contributions | |
weather | +1.060 |
pite | +0.912 |
usage | +0.868 |
ritical | +0.867 |
Hurricane | +0.846 |
Neg logit contributions | |
atography | -0.978 |
IDA | -0.937 |
Ru | -0.928 |
wrongdoing | -0.907 |
kaya | -0.897 |
Token | it |
Token ID | 340 |
Feature activation | -2.33e+00 |
Pos logit contributions | |
atography | +1.540 |
IDA | +1.482 |
Ru | +1.464 |
kaya | +1.431 |
irable | +1.414 |
Neg logit contributions | |
weather | -1.718 |
pite | -1.485 |
ritical | -1.419 |
usage | -1.412 |
Hurricane | -1.391 |
Token | reon |
Token ID | 21833 |
Feature activation | -9.01e+00 |
Pos logit contributions | |
atography | +1.522 |
IDA | +1.467 |
Ru | +1.443 |
wrongdoing | +1.426 |
kaya | +1.420 |
Neg logit contributions | |
weather | -1.601 |
pite | -1.366 |
ritical | -1.312 |
usage | -1.299 |
Hurricane | -1.255 |
Token | in |
Token ID | 287 |
Feature activation | -5.34e+00 |
Pos logit contributions | |
weather | +6.018 |
Hurricane | +5.072 |
pite | +5.043 |
usage | +4.834 |
ritical | +4.689 |
Neg logit contributions | |
atography | -6.443 |
IDA | -5.957 |
Ru | -5.952 |
irable | -5.914 |
kaya | -5.882 |
Token | 2011 |
Token ID | 2813 |
Feature activation | -2.35e+00 |
Pos logit contributions | |
weather | +3.557 |
pite | +3.142 |
Hurricane | +3.022 |
ritical | +2.851 |
usage | +2.841 |
Neg logit contributions | |
atography | -4.004 |
IDA | -3.748 |
Ru | -3.729 |
irable | -3.687 |
kaya | -3.648 |
Token | , |
Token ID | 11 |
Feature activation | +1.46e+00 |
Pos logit contributions | |
weather | +1.680 |
pite | +1.421 |
usage | +1.334 |
ritical | +1.324 |
Hurricane | +1.323 |
Neg logit contributions | |
atography | -1.721 |
IDA | -1.647 |
Ru | -1.643 |
wrongdoing | -1.597 |
irable | -1.595 |
Token | bas |
Token ID | 1615 |
Feature activation | -2.93e+00 |
Pos logit contributions | |
atography | +1.006 |
IDA | +0.961 |
Ru | +0.956 |
kaya | +0.937 |
irable | +0.930 |
Neg logit contributions | |
weather | -1.118 |
pite | -0.962 |
usage | -0.919 |
ritical | -0.913 |
Hurricane | -0.901 |
Token | ing |
Token ID | 278 |
Feature activation | -2.19e+01 |
Pos logit contributions | |
weather | +2.316 |
pite | +2.062 |
usage | +1.957 |
ritical | +1.950 |
Hurricane | +1.879 |
Neg logit contributions | |
atography | -1.852 |
IDA | -1.808 |
Ru | -1.777 |
wrongdoing | -1.756 |
kaya | -1.729 |
Token | it |
Token ID | 340 |
Feature activation | -1.48e+01 |
Pos logit contributions | |
weather | +11.701 |
Hurricane | +10.195 |
Pend | +9.813 |
usage | +9.613 |
pite | +9.610 |
Neg logit contributions | |
kaya | -13.157 |
Ru | -12.932 |
atography | -12.777 |
irable | -12.674 |
IDA | -12.423 |
Token | off |
Token ID | 572 |
Feature activation | -2.68e+00 |
Pos logit contributions | |
weather | +7.648 |
pite | +5.998 |
Hurricane | +5.954 |
usage | +5.891 |
Pend | +5.695 |
Neg logit contributions | |
atography | -11.106 |
Ru | -11.105 |
IDA | -11.072 |
kaya | -10.802 |
irable | -10.800 |
Token | three |
Token ID | 1115 |
Feature activation | -6.96e+00 |
Pos logit contributions | |
weather | +1.942 |
pite | +1.658 |
Hurricane | +1.586 |
usage | +1.575 |
ritical | +1.574 |
Neg logit contributions | |
atography | -1.888 |
Ru | -1.809 |
kaya | -1.801 |
IDA | -1.790 |
wrongdoing | -1.756 |
Token | existing |
Token ID | 4683 |
Feature activation | -8.53e+00 |
Pos logit contributions | |
weather | +4.813 |
pite | +4.096 |
Hurricane | +4.038 |
usage | +3.818 |
ritical | +3.818 |
Neg logit contributions | |
atography | -4.934 |
Ru | -4.708 |
kaya | -4.606 |
irable | -4.556 |
vana | -4.517 |
Token | root |
Token ID | 6808 |
Feature activation | -1.95e+00 |
Pos logit contributions | |
weather | +5.677 |
pite | +4.698 |
Hurricane | +4.675 |
Pend | +4.556 |
usage | +4.430 |
Neg logit contributions | |
atography | -6.112 |
Ru | -5.895 |
kaya | -5.818 |
vana | -5.675 |
irable | -5.669 |
Token | reflect |
Token ID | 4079 |
Feature activation | +2.67e+00 |
Pos logit contributions | |
weather | +2.635 |
pite | +2.212 |
ritical | +2.106 |
Hurricane | +2.036 |
usage | +2.021 |
Neg logit contributions | |
atography | -3.025 |
IDA | -3.009 |
Ru | -2.897 |
arthed | -2.807 |
wrongdoing | -2.794 |
Token | true |
Token ID | 2081 |
Feature activation | -3.28e-01 |
Pos logit contributions | |
atography | +1.815 |
IDA | +1.745 |
Ru | +1.720 |
kaya | +1.688 |
wrongdoing | +1.674 |
Neg logit contributions | |
weather | -2.044 |
pite | -1.753 |
ritical | -1.678 |
usage | -1.671 |
Hurricane | -1.636 |
Token | disagreement |
Token ID | 25800 |
Feature activation | -8.28e-01 |
Pos logit contributions | |
weather | +0.249 |
pite | +0.213 |
ritical | +0.204 |
usage | +0.202 |
Hurricane | +0.200 |
Neg logit contributions | |
atography | -0.226 |
IDA | -0.217 |
Ru | -0.215 |
kaya | -0.211 |
irable | -0.208 |
Token | ; |
Token ID | 26 |
Feature activation | -2.43e+00 |
Pos logit contributions | |
weather | +0.607 |
pite | +0.516 |
ritical | +0.493 |
usage | +0.490 |
Hurricane | +0.479 |
Neg logit contributions | |
atography | -0.586 |
IDA | -0.572 |
Ru | -0.558 |
kaya | -0.549 |
irable | -0.548 |
Token | they |
Token ID | 484 |
Feature activation | -1.03e+01 |
Pos logit contributions | |
weather | +1.790 |
pite | +1.515 |
Hurricane | +1.437 |
ritical | +1.428 |
usage | +1.421 |
Neg logit contributions | |
atography | -1.721 |
IDA | -1.647 |
Ru | -1.631 |
irable | -1.610 |
kaya | -1.597 |
Token | may |
Token ID | 743 |
Feature activation | -2.10e+01 |
Pos logit contributions | |
weather | +6.130 |
pite | +5.032 |
ritical | +4.689 |
Hurricane | +4.593 |
timed | +4.591 |
Neg logit contributions | |
IDA | -8.047 |
atography | -8.027 |
arthed | -7.818 |
Ru | -7.627 |
Transparency | -7.275 |
Token | be |
Token ID | 307 |
Feature activation | -1.10e+01 |
Pos logit contributions | |
weather | +10.841 |
pite | +8.881 |
cope | +8.655 |
Hurricane | +8.542 |
mistakenly | +8.425 |
Neg logit contributions | |
IDA | -14.424 |
arthed | -14.279 |
atography | -13.996 |
Ru | -13.322 |
MF | -12.940 |
Token | explain |
Token ID | 4727 |
Feature activation | +2.78e+00 |
Pos logit contributions | |
weather | +7.384 |
pite | +6.372 |
misplaced | +6.168 |
Hurricane | +6.166 |
timed | +6.006 |
Neg logit contributions | |
atography | -7.656 |
IDA | -7.150 |
Ru | -7.055 |
irable | -6.960 |
vana | -6.947 |
Token | able |
Token ID | 540 |
Feature activation | -2.50e+00 |
Pos logit contributions | |
atography | +1.762 |
IDA | +1.689 |
Ru | +1.661 |
kaya | +1.633 |
wrongdoing | +1.626 |
Neg logit contributions | |
weather | -2.269 |
pite | -1.948 |
ritical | -1.884 |
usage | -1.871 |
Hurricane | -1.821 |
Token | as |
Token ID | 355 |
Feature activation | -7.24e+00 |
Pos logit contributions | |
weather | +1.968 |
pite | +1.716 |
Hurricane | +1.611 |
usage | +1.597 |
ritical | +1.590 |
Neg logit contributions | |
atography | -1.668 |
IDA | -1.582 |
Ru | -1.542 |
kaya | -1.503 |
irable | -1.496 |
Token | actually |
Token ID | 1682 |
Feature activation | -3.50e+00 |
Pos logit contributions | |
weather | +5.059 |
pite | +4.401 |
Hurricane | +4.190 |
ritical | +4.028 |
usage | +4.020 |
Neg logit contributions | |
atography | -5.139 |
IDA | -4.936 |
Ru | -4.851 |
irable | -4.699 |
kaya | -4.688 |
Token | millennial |
Token ID | 46451 |
Feature activation | +9.48e-01 |
Pos logit contributions | |
weather | +5.182 |
pite | +4.436 |
ritical | +4.400 |
Hurricane | +4.394 |
usage | +4.246 |
Neg logit contributions | |
atography | -5.397 |
IDA | -5.328 |
kaya | -5.229 |
arthed | -5.184 |
Ru | -5.167 |
Token | vote |
Token ID | 3015 |
Feature activation | -5.89e+00 |
Pos logit contributions | |
atography | +0.646 |
IDA | +0.615 |
Ru | +0.610 |
kaya | +0.601 |
irable | +0.593 |
Neg logit contributions | |
weather | -0.725 |
pite | -0.621 |
ritical | -0.605 |
usage | -0.595 |
Hurricane | -0.590 |
Token | , |
Token ID | 11 |
Feature activation | -3.34e+00 |
Pos logit contributions | |
weather | +3.902 |
pite | +3.190 |
Hurricane | +3.043 |
ritical | +3.018 |
usage | +2.997 |
Neg logit contributions | |
atography | -4.415 |
IDA | -4.317 |
Ru | -4.206 |
irable | -4.172 |
arthed | -4.146 |
Token | and |
Token ID | 290 |
Feature activation | -3.05e+00 |
Pos logit contributions | |
weather | +2.301 |
pite | +1.915 |
Hurricane | +1.867 |
usage | +1.842 |
ritical | +1.829 |
Neg logit contributions | |
atography | -2.450 |
IDA | -2.366 |
Ru | -2.333 |
irable | -2.332 |
kaya | -2.320 |
Token | with |
Token ID | 351 |
Feature activation | -4.57e+00 |
Pos logit contributions | |
weather | +2.121 |
pite | +1.788 |
Hurricane | +1.732 |
ritical | +1.697 |
usage | +1.693 |
Neg logit contributions | |
atography | -2.267 |
IDA | -2.126 |
Ru | -2.106 |
kaya | -2.104 |
irable | -2.098 |
Token | good |
Token ID | 922 |
Feature activation | -2.08e+01 |
Pos logit contributions | |
weather | +3.051 |
pite | +2.569 |
Hurricane | +2.569 |
ritical | +2.480 |
usage | +2.439 |
Neg logit contributions | |
atography | -3.421 |
IDA | -3.222 |
Ru | -3.191 |
kaya | -3.188 |
irable | -3.165 |
Token | reason |
Token ID | 1738 |
Feature activation | -3.50e+00 |
Pos logit contributions | |
weather | +5.883 |
reasons | +4.429 |
Hurricane | +3.778 |
timed | +3.424 |
misplaced | +3.403 |
Neg logit contributions | |
IDA | -19.239 |
arthed | -19.162 |
atography | -18.939 |
Ru | -18.863 |
kaya | -18.574 |
Token | . |
Token ID | 13 |
Feature activation | -2.99e+00 |
Pos logit contributions | |
weather | +2.456 |
pite | +2.009 |
ritical | +1.941 |
Hurricane | +1.932 |
usage | +1.927 |
Neg logit contributions | |
atography | -2.554 |
IDA | -2.488 |
Ru | -2.407 |
irable | -2.407 |
kaya | -2.359 |
Token | This |
Token ID | 770 |
Feature activation | -3.04e-01 |
Pos logit contributions | |
weather | +2.062 |
Hurricane | +1.751 |
pite | +1.705 |
Pend | +1.642 |
usage | +1.640 |
Neg logit contributions | |
atography | -2.220 |
enrich | -2.089 |
IDA | -2.081 |
irable | -2.065 |
wrongdoing | -2.062 |
Token | year |
Token ID | 614 |
Feature activation | -7.22e+00 |
Pos logit contributions | |
weather | +0.217 |
pite | +0.185 |
ritical | +0.177 |
usage | +0.176 |
Hurricane | +0.176 |
Neg logit contributions | |
atography | -0.224 |
IDA | -0.213 |
Ru | -0.209 |
irable | -0.205 |
kaya | -0.205 |
Token | , |
Token ID | 11 |
Feature activation | -1.74e+00 |
Pos logit contributions | |
weather | +4.414 |
Hurricane | +3.640 |
pite | +3.534 |
ritical | +3.308 |
usage | +3.285 |
Neg logit contributions | |
atography | -5.777 |
IDA | -5.441 |
irable | -5.417 |
kaya | -5.302 |
Ru | -5.275 |
Token | times |
Token ID | 1661 |
Feature activation | -6.55e-01 |
Pos logit contributions | |
weather | +1.769 |
pite | +1.200 |
Hurricane | +1.026 |
usage | +0.978 |
Pend | +0.941 |
Neg logit contributions | |
atography | -7.144 |
IDA | -6.944 |
Ru | -6.919 |
irable | -6.869 |
kaya | -6.770 |
Token | to |
Token ID | 284 |
Feature activation | -9.43e+00 |
Pos logit contributions | |
weather | +0.477 |
pite | +0.403 |
usage | +0.385 |
Hurricane | +0.379 |
ritical | +0.375 |
Neg logit contributions | |
atography | -0.468 |
IDA | -0.461 |
Ru | -0.452 |
irable | -0.443 |
wrongdoing | -0.434 |
Token | get |
Token ID | 651 |
Feature activation | -5.54e-01 |
Pos logit contributions | |
weather | +5.681 |
Hurricane | +4.725 |
pite | +4.582 |
Welsh | +4.555 |
Pend | +4.490 |
Neg logit contributions | |
atography | -7.062 |
IDA | -6.957 |
arthed | -6.805 |
Ru | -6.800 |
irable | -6.698 |
Token | people |
Token ID | 661 |
Feature activation | -7.90e+00 |
Pos logit contributions | |
weather | +0.367 |
pite | +0.309 |
usage | +0.297 |
ritical | +0.293 |
Hurricane | +0.292 |
Neg logit contributions | |
atography | -0.430 |
IDA | -0.420 |
Ru | -0.410 |
kaya | -0.407 |
irable | -0.405 |
Token | to |
Token ID | 284 |
Feature activation | -7.45e+00 |
Pos logit contributions | |
weather | +5.034 |
usage | +4.157 |
pite | +4.077 |
Hurricane | +4.047 |
Welsh | +3.948 |
Neg logit contributions | |
IDA | -5.908 |
atography | -5.855 |
Ru | -5.779 |
irable | -5.717 |
kaya | -5.636 |
Token | call |
Token ID | 869 |
Feature activation | -2.05e+01 |
Pos logit contributions | |
weather | +3.816 |
pite | +2.952 |
Hurricane | +2.945 |
usage | +2.932 |
Welsh | +2.770 |
Neg logit contributions | |
IDA | -6.503 |
atography | -6.494 |
arthed | -6.338 |
Ru | -6.326 |
irable | -6.206 |
Token | it |
Token ID | 340 |
Feature activation | -7.38e+00 |
Pos logit contributions | |
weather | +9.409 |
Hurricane | +8.162 |
pite | +7.801 |
Pend | +7.763 |
usage | +7.294 |
Neg logit contributions | |
atography | -14.416 |
IDA | -13.997 |
kaya | -13.400 |
Ru | -13.341 |
irable | -13.304 |
Token | J |
Token ID | 449 |
Feature activation | -3.09e+00 |
Pos logit contributions | |
weather | +5.404 |
Hurricane | +4.635 |
pite | +4.589 |
Pend | +4.492 |
usage | +4.348 |
Neg logit contributions | |
atography | -4.848 |
IDA | -4.769 |
Ru | -4.599 |
irable | -4.539 |
kaya | -4.473 |
Token | oust |
Token ID | 23968 |
Feature activation | -7.21e-01 |
Pos logit contributions | |
weather | +1.866 |
pite | +1.539 |
usage | +1.431 |
ritical | +1.416 |
Hurricane | +1.328 |
Neg logit contributions | |
atography | -2.575 |
IDA | -2.541 |
Ru | -2.511 |
wrongdoing | -2.493 |
kaya | -2.454 |
Token | . |
Token ID | 13 |
Feature activation | -1.13e+00 |
Pos logit contributions | |
weather | +0.533 |
pite | +0.452 |
usage | +0.429 |
ritical | +0.429 |
Hurricane | +0.422 |
Neg logit contributions | |
atography | -0.516 |
IDA | -0.501 |
Ru | -0.489 |
kaya | -0.480 |
wrongdoing | -0.479 |
Token | Then |
Token ID | 3244 |
Feature activation | -2.91e+00 |
Pos logit contributions | |
weather | +0.767 |
pite | +0.642 |
Hurricane | +0.633 |
usage | +0.615 |
Pend | +0.611 |
Neg logit contributions | |
atography | -0.851 |
IDA | -0.817 |
irable | -0.804 |
enrich | -0.797 |
wrongdoing | -0.797 |
Token | Psych |
Token ID | 31923 |
Feature activation | +4.16e+00 |
Pos logit contributions | |
weather | +0.787 |
pite | +0.658 |
ritical | +0.638 |
usage | +0.636 |
Hurricane | +0.631 |
Neg logit contributions | |
atography | -0.684 |
IDA | -0.652 |
Ru | -0.646 |
kaya | -0.640 |
irable | -0.634 |
Token | o |
Token ID | 78 |
Feature activation | -6.50e+00 |
Pos logit contributions | |
atography | +1.968 |
IDA | +1.861 |
Ru | +1.820 |
wrongdoing | +1.778 |
kaya | +1.769 |
Neg logit contributions | |
weather | -4.019 |
pite | -3.568 |
ritical | -3.458 |
usage | -3.446 |
Hurricane | -3.373 |
Token | , |
Token ID | 11 |
Feature activation | -5.08e+00 |
Pos logit contributions | |
weather | +4.486 |
pite | +3.641 |
usage | +3.562 |
ritical | +3.524 |
Hurricane | +3.460 |
Neg logit contributions | |
atography | -4.590 |
IDA | -4.538 |
Ru | -4.423 |
irable | -4.410 |
kaya | -4.384 |
Token | and |
Token ID | 290 |
Feature activation | -8.73e+00 |
Pos logit contributions | |
weather | +3.867 |
pite | +3.309 |
usage | +3.161 |
Hurricane | +3.147 |
ritical | +3.106 |
Neg logit contributions | |
atography | -3.240 |
IDA | -3.181 |
irable | -3.142 |
Ru | -3.079 |
kaya | -3.058 |
Token | with |
Token ID | 351 |
Feature activation | -2.20e+00 |
Pos logit contributions | |
weather | +5.981 |
pite | +5.070 |
ritical | +4.819 |
Hurricane | +4.800 |
usage | +4.797 |
Neg logit contributions | |
atography | -5.953 |
IDA | -5.686 |
kaya | -5.671 |
arthed | -5.659 |
Ru | -5.601 |
Token | good |
Token ID | 922 |
Feature activation | -2.04e+01 |
Pos logit contributions | |
weather | +1.612 |
pite | +1.371 |
ritical | +1.322 |
usage | +1.302 |
Hurricane | +1.286 |
Neg logit contributions | |
atography | -1.519 |
IDA | -1.472 |
Ru | -1.467 |
kaya | -1.450 |
wrongdoing | -1.426 |
Token | reason |
Token ID | 1738 |
Feature activation | -5.90e+00 |
Pos logit contributions | |
weather | +7.118 |
pite | +4.694 |
reasons | +4.546 |
ritical | +4.287 |
timed | +4.124 |
Neg logit contributions | |
IDA | -18.022 |
arthed | -17.649 |
Ru | -17.565 |
atography | -17.154 |
kaya | -17.012 |
Token | : |
Token ID | 25 |
Feature activation | -4.64e+00 |
Pos logit contributions | |
weather | +4.057 |
pite | +3.325 |
ritical | +3.218 |
usage | +3.209 |
Hurricane | +3.099 |
Neg logit contributions | |
atography | -4.228 |
IDA | -4.169 |
irable | -4.016 |
Ru | -4.014 |
wrongdoing | -4.008 |
Token | it |
Token ID | 340 |
Feature activation | -3.20e+00 |
Pos logit contributions | |
weather | +3.154 |
pite | +2.632 |
usage | +2.538 |
Hurricane | +2.511 |
ritical | +2.497 |
Neg logit contributions | |
atography | -3.304 |
IDA | -3.261 |
kaya | -3.197 |
Ru | -3.194 |
irable | -3.179 |
Token | 's |
Token ID | 338 |
Feature activation | +1.97e+00 |
Pos logit contributions | |
weather | +2.239 |
pite | +1.817 |
ritical | +1.751 |
usage | +1.718 |
Hurricane | +1.697 |
Neg logit contributions | |
IDA | -2.354 |
atography | -2.313 |
wrongdoing | -2.261 |
Ru | -2.252 |
irable | -2.202 |
Token | one |
Token ID | 530 |
Feature activation | +9.36e+00 |
Pos logit contributions | |
atography | +1.400 |
IDA | +1.356 |
Ru | +1.333 |
wrongdoing | +1.323 |
kaya | +1.320 |
Neg logit contributions | |
weather | -1.451 |
pite | -1.222 |
ritical | -1.179 |
usage | -1.175 |
Hurricane | -1.141 |
Token | to |
Token ID | 284 |
Feature activation | -4.71e+00 |
Pos logit contributions | |
weather | +0.327 |
pite | +0.276 |
Hurricane | +0.264 |
ritical | +0.262 |
usage | +0.258 |
Neg logit contributions | |
atography | -0.298 |
IDA | -0.284 |
Ru | -0.284 |
irable | -0.280 |
kaya | -0.276 |
Token | do |
Token ID | 466 |
Feature activation | +2.86e+00 |
Pos logit contributions | |
weather | +2.716 |
Hurricane | +2.125 |
pite | +2.106 |
ritical | +1.989 |
sled | +1.948 |
Neg logit contributions | |
atography | -3.970 |
IDA | -3.899 |
Ru | -3.819 |
wrongdoing | -3.747 |
irable | -3.732 |
Token | if |
Token ID | 611 |
Feature activation | -4.82e+00 |
Pos logit contributions | |
atography | +1.902 |
IDA | +1.820 |
Ru | +1.804 |
wrongdoing | +1.761 |
irable | +1.758 |
Neg logit contributions | |
weather | -2.262 |
pite | -1.933 |
ritical | -1.844 |
usage | -1.840 |
Hurricane | -1.818 |
Token | the |
Token ID | 262 |
Feature activation | -1.08e+01 |
Pos logit contributions | |
weather | +3.106 |
pite | +2.494 |
Hurricane | +2.443 |
usage | +2.361 |
ritical | +2.327 |
Neg logit contributions | |
atography | -3.848 |
Ru | -3.592 |
irable | -3.581 |
kaya | -3.564 |
arthed | -3.543 |
Token | season |
Token ID | 1622 |
Feature activation | -1.74e+01 |
Pos logit contributions | |
weather | +6.767 |
pite | +5.687 |
Hurricane | +5.528 |
usage | +5.218 |
sled | +5.209 |
Neg logit contributions | |
atography | -7.890 |
IDA | -7.843 |
Ru | -7.619 |
arthed | -7.542 |
irable | -7.418 |
Token | was |
Token ID | 373 |
Feature activation | -2.03e+01 |
Pos logit contributions | |
weather | +9.721 |
sled | +8.449 |
timed | +7.981 |
pite | +7.930 |
Hurricane | +7.880 |
Neg logit contributions | |
atography | -11.315 |
IDA | -11.256 |
irable | -10.996 |
Ru | -10.841 |
rance | -10.695 |
Token | to |
Token ID | 284 |
Feature activation | -1.19e+01 |
Pos logit contributions | |
weather | +9.314 |
timed | +8.338 |
sled | +7.852 |
Hurricane | +7.658 |
misplaced | +7.512 |
Neg logit contributions | |
vana | -14.152 |
atography | -14.053 |
irable | -13.924 |
Ru | -13.664 |
IDA | -13.578 |
Token | be |
Token ID | 307 |
Feature activation | -1.18e+01 |
Pos logit contributions | |
weather | +7.031 |
pite | +5.607 |
sled | +5.581 |
Hurricane | +5.574 |
cope | +5.273 |
Neg logit contributions | |
atography | -8.907 |
IDA | -8.723 |
arthed | -8.330 |
Ru | -8.312 |
Transparency | -8.269 |
Token | a |
Token ID | 257 |
Feature activation | -7.86e+00 |
Pos logit contributions | |
weather | +7.741 |
Hurricane | +6.475 |
timed | +6.394 |
sled | +6.229 |
misplaced | +6.119 |
Neg logit contributions | |
atography | -8.188 |
Ru | -7.450 |
vana | -7.439 |
wrongdoing | -7.420 |
kaya | -7.292 |
Token | success |
Token ID | 1943 |
Feature activation | -2.82e+00 |
Pos logit contributions | |
weather | +5.551 |
pite | +4.753 |
Hurricane | +4.669 |
ritical | +4.659 |
usage | +4.645 |
Neg logit contributions | |
atography | -5.485 |
Ru | -4.994 |
wrongdoing | -4.909 |
Transparency | -4.848 |
IDA | -4.840 |
Token | – |
Token ID | 784 |
Feature activation | -3.32e-01 |
Pos logit contributions | |
weather | +2.045 |
pite | +1.689 |
usage | +1.603 |
ritical | +1.597 |
Hurricane | +1.592 |
Neg logit contributions | |
atography | -2.048 |
IDA | -1.966 |
Ru | -1.934 |
irable | -1.905 |
wrongdoing | -1.898 |
Token |
|
Token ID | 198 |
Feature activation | -5.16e+00 |
Pos logit contributions | |
weather | +0.129 |
pite | +0.111 |
ritical | +0.106 |
Hurricane | +0.106 |
usage | +0.106 |
Neg logit contributions | |
atography | -0.096 |
IDA | -0.092 |
Ru | -0.090 |
kaya | -0.089 |
wrongdoing | -0.089 |
Token |
|
Token ID | 198 |
Feature activation | -2.90e+00 |
Pos logit contributions | |
weather | +3.560 |
pite | +2.861 |
Hurricane | +2.802 |
usage | +2.762 |
Pend | +2.627 |
Neg logit contributions | |
atography | -3.910 |
vana | -3.633 |
IDA | -3.632 |
enrich | -3.597 |
irable | -3.569 |
Token | Hung |
Token ID | 39505 |
Feature activation | -3.86e-01 |
Pos logit contributions | |
weather | +2.254 |
pite | +1.884 |
usage | +1.832 |
ritical | +1.763 |
Hurricane | +1.745 |
Neg logit contributions | |
atography | -1.976 |
wrongdoing | -1.850 |
irable | -1.839 |
enrich | -1.826 |
kaya | -1.818 |
Token | ry |
Token ID | 563 |
Feature activation | -4.02e+00 |
Pos logit contributions | |
weather | +0.348 |
pite | +0.301 |
ritical | +0.292 |
Hurricane | +0.286 |
usage | +0.286 |
Neg logit contributions | |
atography | -0.214 |
IDA | -0.203 |
Ru | -0.197 |
wrongdoing | -0.195 |
irable | -0.192 |
Token | and |
Token ID | 290 |
Feature activation | -2.66e+00 |
Pos logit contributions | |
weather | +2.944 |
pite | +2.497 |
Hurricane | +2.427 |
ritical | +2.298 |
usage | +2.288 |
Neg logit contributions | |
atography | -2.895 |
IDA | -2.746 |
kaya | -2.696 |
irable | -2.666 |
Ru | -2.648 |
Token | dis |
Token ID | 595 |
Feature activation | -2.01e+01 |
Pos logit contributions | |
weather | +1.797 |
pite | +1.537 |
Hurricane | +1.459 |
ritical | +1.415 |
usage | +1.404 |
Neg logit contributions | |
atography | -2.038 |
IDA | -1.945 |
irable | -1.932 |
Ru | -1.889 |
kaya | -1.876 |
Token | he |
Token ID | 258 |
Feature activation | -5.41e+00 |
Pos logit contributions | |
weather | +10.767 |
pite | +9.830 |
usage | +8.570 |
heart | +8.305 |
ritical | +7.885 |
Neg logit contributions | |
IDA | -14.814 |
atography | -14.632 |
Transparency | -13.985 |
wrongdoing | -13.802 |
�� | -13.557 |
Token | ve |
Token ID | 303 |
Feature activation | +1.78e+00 |
Pos logit contributions | |
weather | +4.162 |
pite | +3.471 |
ritical | +3.326 |
usage | +3.321 |
Hurricane | +3.273 |
Neg logit contributions | |
atography | -3.573 |
IDA | -3.512 |
Ru | -3.380 |
kaya | -3.358 |
wrongdoing | -3.322 |
Token | lled |
Token ID | 3353 |
Feature activation | +3.39e+00 |
Pos logit contributions | |
atography | +1.576 |
IDA | +1.516 |
Ru | +1.506 |
wrongdoing | +1.489 |
kaya | +1.486 |
Neg logit contributions | |
weather | -1.023 |
pite | -0.807 |
usage | -0.762 |
ritical | -0.760 |
Hurricane | -0.732 |
Token |
|
Token ID | 198 |
Feature activation | -5.08e-02 |
Pos logit contributions | |
atography | +2.301 |
IDA | +2.213 |
Ru | +2.178 |
kaya | +2.145 |
wrongdoing | +2.131 |
Neg logit contributions | |
weather | -2.611 |
pite | -2.239 |
ritical | -2.134 |
usage | -2.123 |
Hurricane | -2.095 |
Token |
|
Token ID | 198 |
Feature activation | -1.54e+00 |
Pos logit contributions | |
weather | +0.040 |
pite | +0.034 |
usage | +0.033 |
Hurricane | +0.032 |
ritical | +0.032 |
Neg logit contributions | |
atography | -0.034 |
IDA | -0.032 |
wrongdoing | -0.031 |
Ru | -0.031 |
irable | -0.031 |
Token | Minister |
Token ID | 4139 |
Feature activation | -2.74e+00 |
Pos logit contributions | |
atography | +1.886 |
IDA | +1.817 |
Ru | +1.793 |
wrongdoing | +1.766 |
kaya | +1.761 |
Neg logit contributions | |
weather | -1.636 |
pite | -1.366 |
ritical | -1.300 |
usage | -1.294 |
Hurricane | -1.254 |
Token | for |
Token ID | 329 |
Feature activation | -5.98e+00 |
Pos logit contributions | |
weather | +1.929 |
pite | +1.609 |
Hurricane | +1.516 |
usage | +1.512 |
ritical | +1.507 |
Neg logit contributions | |
atography | -2.029 |
IDA | -1.955 |
irable | -1.922 |
Ru | -1.919 |
kaya | -1.891 |
Token | having |
Token ID | 1719 |
Feature activation | +4.20e+00 |
Pos logit contributions | |
weather | +3.970 |
pite | +3.320 |
Hurricane | +3.303 |
usage | +3.134 |
sled | +3.132 |
Neg logit contributions | |
atography | -4.491 |
IDA | -4.292 |
irable | -4.237 |
Ru | -4.217 |
kaya | -4.126 |
Token | � |
Token ID | 564 |
Feature activation | +1.07e+00 |
Pos logit contributions | |
atography | +2.866 |
IDA | +2.757 |
Ru | +2.717 |
wrongdoing | +2.664 |
kaya | +2.663 |
Neg logit contributions | |
weather | -3.208 |
pite | -2.745 |
ritical | -2.636 |
usage | -2.616 |
Hurricane | -2.556 |
Token | � |
Token ID | 250 |
Feature activation | +9.00e-01 |
Pos logit contributions | |
atography | +0.694 |
IDA | +0.658 |
wrongdoing | +0.640 |
Ru | +0.638 |
kaya | +0.628 |
Neg logit contributions | |
weather | -0.853 |
pite | -0.742 |
usage | -0.708 |
ritical | -0.703 |
Hurricane | -0.692 |
Token | un |
Token ID | 403 |
Feature activation | -1.96e+01 |
Pos logit contributions | |
atography | +0.588 |
IDA | +0.564 |
wrongdoing | +0.546 |
Ru | +0.545 |
irable | +0.538 |
Neg logit contributions | |
weather | -0.727 |
pite | -0.619 |
usage | -0.598 |
ritical | -0.595 |
Hurricane | -0.569 |
Token | le |
Token ID | 293 |
Feature activation | -9.57e+00 |
Pos logit contributions | |
weather | +12.992 |
pite | +11.030 |
usage | +10.552 |
ritical | +10.503 |
foreseen | +9.865 |
Neg logit contributions | |
IDA | -11.108 |
atography | -10.866 |
Marina | -10.366 |
wrongdoing | -10.208 |
Ru | -10.202 |
Token | ashed |
Token ID | 5263 |
Feature activation | -1.20e+00 |
Pos logit contributions | |
weather | +5.483 |
pite | +4.577 |
ritical | +3.976 |
Hurricane | +3.949 |
usage | +3.923 |
Neg logit contributions | |
IDA | -7.614 |
atography | -7.609 |
Ru | -7.170 |
wrongdoing | -7.103 |
kaya | -7.064 |
Token | a |
Token ID | 257 |
Feature activation | -9.43e-01 |
Pos logit contributions | |
weather | +0.838 |
pite | +0.713 |
ritical | +0.682 |
Hurricane | +0.680 |
usage | +0.675 |
Neg logit contributions | |
atography | -0.886 |
IDA | -0.831 |
Ru | -0.823 |
kaya | -0.819 |
irable | -0.812 |
Token | national |
Token ID | 2260 |
Feature activation | -7.89e+00 |
Pos logit contributions | |
weather | +0.669 |
pite | +0.577 |
ritical | +0.558 |
Hurricane | +0.554 |
usage | +0.546 |
Neg logit contributions | |
atography | -0.692 |
IDA | -0.647 |
Ru | -0.643 |
irable | -0.632 |
kaya | -0.631 |
Token | unity |
Token ID | 14111 |
Feature activation | +5.55e-01 |
Pos logit contributions | |
weather | +5.207 |
Hurricane | +4.492 |
pite | +4.398 |
ritical | +4.347 |
misplaced | +4.181 |
Neg logit contributions | |
atography | -5.716 |
IDA | -5.283 |
irable | -5.278 |
arthed | -5.257 |
vana | -5.215 |
Token | / |
Token ID | 14 |
Feature activation | +7.97e+00 |
Pos logit contributions | |
atography | +0.780 |
IDA | +0.737 |
Ru | +0.725 |
wrongdoing | +0.718 |
kaya | +0.715 |
Neg logit contributions | |
weather | -0.924 |
pite | -0.792 |
usage | -0.757 |
ritical | -0.751 |
Hurricane | -0.743 |
Token | 205 |
Token ID | 21261 |
Feature activation | -4.69e+00 |
Pos logit contributions | |
atography | +4.604 |
IDA | +4.520 |
Ru | +4.444 |
wrongdoing | +4.264 |
kaya | +4.250 |
Neg logit contributions | |
weather | -6.627 |
pite | -5.808 |
ritical | -5.681 |
usage | -5.528 |
Hurricane | -5.405 |
Token | 3 |
Token ID | 513 |
Feature activation | +6.05e-01 |
Pos logit contributions | |
weather | +3.028 |
pite | +2.426 |
Hurricane | +2.402 |
usage | +2.340 |
Welsh | +2.247 |
Neg logit contributions | |
atography | -3.710 |
kaya | -3.474 |
IDA | -3.456 |
enrich | -3.449 |
Ru | -3.438 |
Token | * |
Token ID | 9 |
Feature activation | +1.85e+00 |
Pos logit contributions | |
atography | +0.391 |
IDA | +0.367 |
Ru | +0.361 |
wrongdoing | +0.358 |
kaya | +0.354 |
Neg logit contributions | |
weather | -0.489 |
pite | -0.422 |
usage | -0.404 |
ritical | -0.400 |
Hurricane | -0.395 |
Token | 0 |
Token ID | 657 |
Feature activation | -3.26e+00 |
Pos logit contributions | |
atography | +1.014 |
IDA | +0.967 |
Ru | +0.950 |
wrongdoing | +0.930 |
kaya | +0.925 |
Neg logit contributions | |
weather | -1.652 |
pite | -1.453 |
ritical | -1.403 |
usage | -1.395 |
Hurricane | -1.359 |
Token | . |
Token ID | 13 |
Feature activation | -1.96e+01 |
Pos logit contributions | |
weather | +2.193 |
pite | +1.804 |
usage | +1.731 |
Hurricane | +1.726 |
ritical | +1.635 |
Neg logit contributions | |
atography | -2.548 |
IDA | -2.354 |
enrich | -2.328 |
irable | -2.326 |
wrongdoing | -2.317 |
Token | 85 |
Token ID | 5332 |
Feature activation | -4.62e+00 |
Pos logit contributions | |
weather | +9.014 |
States | +6.911 |
pite | +6.900 |
712 | +6.800 |
usage | +6.523 |
Neg logit contributions | |
enrich | -14.531 |
enriched | -14.037 |
atography | -13.979 |
wrongdoing | -13.722 |
�� | -13.683 |
Token | 17 |
Token ID | 1558 |
Feature activation | -3.60e+00 |
Pos logit contributions | |
weather | +3.195 |
pite | +2.707 |
Hurricane | +2.638 |
usage | +2.515 |
Pend | +2.481 |
Neg logit contributions | |
atography | -3.410 |
enrich | -3.207 |
vana | -3.182 |
irable | -3.143 |
wrongdoing | -3.119 |
Token | 4 |
Token ID | 604 |
Feature activation | +2.90e-01 |
Pos logit contributions | |
weather | +2.270 |
pite | +1.827 |
Hurricane | +1.778 |
usage | +1.736 |
ritical | +1.687 |
Neg logit contributions | |
atography | -2.899 |
IDA | -2.806 |
kaya | -2.737 |
wrongdoing | -2.720 |
Ru | -2.714 |
Token | * |
Token ID | 9 |
Feature activation | -6.21e+00 |
Pos logit contributions | |
atography | +0.201 |
IDA | +0.186 |
irable | +0.183 |
Ru | +0.182 |
kaya | +0.182 |
Neg logit contributions | |
weather | -0.223 |
pite | -0.191 |
Hurricane | -0.181 |
usage | -0.181 |
ritical | -0.177 |
Token | 3 |
Token ID | 513 |
Feature activation | -6.52e-01 |
Pos logit contributions | |
weather | +3.711 |
Hurricane | +2.919 |
pite | +2.885 |
usage | +2.762 |
Pend | +2.663 |
Neg logit contributions | |
atography | -5.050 |
IDA | -4.840 |
kaya | -4.764 |
Ru | -4.738 |
irable | -4.725 |
Token | / |
Token ID | 14 |
Feature activation | +4.84e+00 |
Pos logit contributions | |
atography | +1.988 |
IDA | +1.913 |
Ru | +1.885 |
wrongdoing | +1.847 |
kaya | +1.840 |
Neg logit contributions | |
weather | -2.622 |
pite | -2.277 |
ritical | -2.200 |
usage | -2.180 |
Hurricane | -2.119 |
Token | 226 |
Token ID | 24909 |
Feature activation | -1.97e+00 |
Pos logit contributions | |
atography | +3.041 |
IDA | +2.922 |
Ru | +2.878 |
wrongdoing | +2.820 |
kaya | +2.810 |
Neg logit contributions | |
weather | -3.936 |
pite | -3.410 |
ritical | -3.288 |
usage | -3.262 |
Hurricane | -3.179 |
Token | 4 |
Token ID | 604 |
Feature activation | +2.07e+00 |
Pos logit contributions | |
weather | +1.367 |
pite | +1.124 |
Hurricane | +1.095 |
usage | +1.075 |
ritical | +1.047 |
Neg logit contributions | |
atography | -1.490 |
IDA | -1.392 |
kaya | -1.389 |
Ru | -1.388 |
vana | -1.374 |
Token | * |
Token ID | 9 |
Feature activation | +3.34e+00 |
Pos logit contributions | |
atography | +1.290 |
IDA | +1.224 |
Ru | +1.204 |
wrongdoing | +1.187 |
kaya | +1.179 |
Neg logit contributions | |
weather | -1.706 |
pite | -1.478 |
usage | -1.416 |
ritical | -1.415 |
Hurricane | -1.382 |
Token | 0 |
Token ID | 657 |
Feature activation | -2.03e+00 |
Pos logit contributions | |
atography | +1.768 |
IDA | +1.688 |
Ru | +1.652 |
wrongdoing | +1.618 |
kaya | +1.605 |
Neg logit contributions | |
weather | -3.032 |
pite | -2.679 |
ritical | -2.588 |
usage | -2.571 |
Hurricane | -2.490 |
Token | . |
Token ID | 13 |
Feature activation | -1.96e+01 |
Pos logit contributions | |
weather | +1.417 |
pite | +1.178 |
usage | +1.127 |
Hurricane | +1.116 |
ritical | +1.082 |
Neg logit contributions | |
atography | -1.540 |
IDA | -1.436 |
wrongdoing | -1.409 |
Ru | -1.408 |
irable | -1.407 |
Token | 9 |
Token ID | 24 |
Feature activation | +1.30e+00 |
Pos logit contributions | |
weather | +8.362 |
pite | +6.547 |
States | +5.867 |
usage | +5.843 |
pring | +5.710 |
Neg logit contributions | |
enrich | -15.162 |
exhib | -14.857 |
enriched | -14.666 |
NetMessage | -14.577 |
wrongdoing | -14.484 |
Token | 242 |
Token ID | 27877 |
Feature activation | -5.30e-01 |
Pos logit contributions | |
atography | +0.795 |
IDA | +0.758 |
Ru | +0.748 |
wrongdoing | +0.745 |
kaya | +0.729 |
Neg logit contributions | |
weather | -1.093 |
pite | -0.944 |
usage | -0.902 |
ritical | -0.898 |
Hurricane | -0.879 |
Token | 4 |
Token ID | 604 |
Feature activation | +1.81e+00 |
Pos logit contributions | |
weather | +0.405 |
pite | +0.346 |
ritical | +0.332 |
usage | +0.330 |
Hurricane | +0.323 |
Neg logit contributions | |
atography | -0.360 |
IDA | -0.346 |
Ru | -0.341 |
wrongdoing | -0.336 |
kaya | -0.334 |
Token | * |
Token ID | 9 |
Feature activation | -2.38e+00 |
Pos logit contributions | |
atography | +1.138 |
IDA | +1.077 |
Ru | +1.058 |
wrongdoing | +1.046 |
kaya | +1.037 |
Neg logit contributions | |
weather | -1.492 |
pite | -1.289 |
usage | -1.234 |
ritical | -1.231 |
Hurricane | -1.205 |
Token | 4 |
Token ID | 604 |
Feature activation | +1.01e+00 |
Pos logit contributions | |
weather | +1.540 |
Hurricane | +1.252 |
pite | +1.233 |
usage | +1.177 |
Pend | +1.155 |
Neg logit contributions | |
atography | -1.919 |
IDA | -1.802 |
Ru | -1.776 |
irable | -1.766 |
vana | -1.764 |
Token | system |
Token ID | 1080 |
Feature activation | -5.23e+00 |
Pos logit contributions | |
weather | +1.562 |
pite | +1.140 |
Hurricane | +1.103 |
ritical | +1.062 |
sled | +1.022 |
Neg logit contributions | |
IDA | -2.806 |
atography | -2.804 |
Ru | -2.746 |
irable | -2.734 |
kaya | -2.705 |
Token | appear |
Token ID | 1656 |
Feature activation | +3.25e+00 |
Pos logit contributions | |
weather | +3.566 |
pite | +2.823 |
Hurricane | +2.794 |
ritical | +2.677 |
usage | +2.655 |
Neg logit contributions | |
atography | -3.850 |
irable | -3.803 |
IDA | -3.791 |
Ru | -3.759 |
arthed | -3.665 |
Token | to |
Token ID | 284 |
Feature activation | -5.77e+00 |
Pos logit contributions | |
atography | +2.146 |
IDA | +2.061 |
Ru | +2.029 |
wrongdoing | +1.989 |
kaya | +1.987 |
Neg logit contributions | |
weather | -2.564 |
pite | -2.201 |
ritical | -2.114 |
usage | -2.094 |
Hurricane | -2.059 |
Token | be |
Token ID | 307 |
Feature activation | -6.89e+00 |
Pos logit contributions | |
weather | +3.703 |
pite | +2.982 |
Hurricane | +2.932 |
Pend | +2.751 |
ritical | +2.745 |
Neg logit contributions | |
IDA | -4.500 |
atography | -4.427 |
Ru | -4.296 |
irable | -4.233 |
arthed | -4.199 |
Token | moving |
Token ID | 3867 |
Feature activation | -3.35e+00 |
Pos logit contributions | |
weather | +4.462 |
Hurricane | +3.603 |
pite | +3.533 |
timed | +3.532 |
Pend | +3.415 |
Neg logit contributions | |
atography | -5.189 |
IDA | -5.086 |
irable | -5.049 |
Ru | -4.948 |
vana | -4.899 |
Token | at |
Token ID | 379 |
Feature activation | -1.96e+01 |
Pos logit contributions | |
weather | +2.470 |
pite | +2.102 |
Hurricane | +2.040 |
usage | +1.958 |
sled | +1.953 |
Neg logit contributions | |
atography | -2.277 |
IDA | -2.230 |
Ru | -2.189 |
irable | -2.176 |
vana | -2.109 |
Token | 200 |
Token ID | 939 |
Feature activation | -4.20e+00 |
Pos logit contributions | |
weather | +10.146 |
timed | +8.867 |
Hurricane | +8.834 |
pite | +8.215 |
Pend | +8.118 |
Neg logit contributions | |
IDA | -12.984 |
atography | -12.944 |
arthed | -12.698 |
kaya | -12.374 |
Ru | -12.359 |
Token | kilometers |
Token ID | 18212 |
Feature activation | -1.81e+00 |
Pos logit contributions | |
weather | +2.641 |
pite | +2.144 |
ritical | +2.056 |
Hurricane | +2.009 |
usage | +1.946 |
Neg logit contributions | |
atography | -3.329 |
IDA | -3.209 |
Ru | -3.176 |
irable | -3.173 |
wrongdoing | -3.135 |
Token | per |
Token ID | 583 |
Feature activation | +6.29e+00 |
Pos logit contributions | |
weather | +1.412 |
pite | +1.229 |
ritical | +1.180 |
Hurricane | +1.175 |
usage | +1.146 |
Neg logit contributions | |
atography | -1.194 |
IDA | -1.112 |
Ru | -1.109 |
kaya | -1.096 |
irable | -1.086 |
Token | second |
Token ID | 1218 |
Feature activation | -9.56e-01 |
Pos logit contributions | |
atography | +3.671 |
IDA | +3.531 |
Ru | +3.453 |
kaya | +3.401 |
wrongdoing | +3.400 |
Neg logit contributions | |
weather | -5.357 |
pite | -4.672 |
ritical | -4.510 |
usage | -4.496 |
Hurricane | -4.353 |
Token | , |
Token ID | 11 |
Feature activation | -1.42e-01 |
Pos logit contributions | |
weather | +0.640 |
pite | +0.533 |
ritical | +0.509 |
Hurricane | +0.498 |
usage | +0.494 |
Neg logit contributions | |
atography | -0.749 |
IDA | -0.717 |
Ru | -0.710 |
kaya | -0.695 |
wrongdoing | -0.695 |
Token | . |
Token ID | 13 |
Feature activation | -1.38e+00 |
Pos logit contributions | |
weather | +1.559 |
pite | +1.356 |
Hurricane | +1.273 |
usage | +1.238 |
sled | +1.214 |
Neg logit contributions | |
atography | -1.509 |
IDA | -1.462 |
Ru | -1.440 |
irable | -1.406 |
wrongdoing | -1.401 |
Token | Good |
Token ID | 4599 |
Feature activation | -1.49e+01 |
Pos logit contributions | |
weather | +1.011 |
pite | +0.853 |
Hurricane | +0.819 |
usage | +0.804 |
ritical | +0.797 |
Neg logit contributions | |
atography | -0.964 |
IDA | -0.928 |
kaya | -0.910 |
irable | -0.909 |
Ru | -0.908 |
Token | ? |
Token ID | 30 |
Feature activation | -1.11e+01 |
Pos logit contributions | |
weather | +8.521 |
pite | +6.397 |
Hurricane | +6.230 |
timed | +6.165 |
ritical | +6.021 |
Neg logit contributions | |
IDA | -10.940 |
atography | -10.884 |
Ru | -10.596 |
kaya | -10.545 |
arthed | -10.422 |
Token | Why |
Token ID | 4162 |
Feature activation | -1.41e+01 |
Pos logit contributions | |
weather | +4.285 |
Hurricane | +3.208 |
usage | +2.829 |
Welsh | +2.719 |
pite | +2.683 |
Neg logit contributions | |
atography | -10.475 |
irable | -10.305 |
enrich | -10.300 |
arthed | -10.158 |
vana | -10.088 |
Token | good |
Token ID | 922 |
Feature activation | -9.70e+00 |
Pos logit contributions | |
weather | +7.276 |
Hurricane | +5.568 |
timed | +5.547 |
Welsh | +5.462 |
misplaced | +5.375 |
Neg logit contributions | |
atography | -11.848 |
IDA | -11.643 |
arthed | -11.084 |
�� | -10.963 |
Ru | -10.946 |
Token | ? |
Token ID | 30 |
Feature activation | -1.95e+01 |
Pos logit contributions | |
weather | +5.197 |
usage | +3.922 |
pite | +3.914 |
Hurricane | +3.792 |
ritical | +3.472 |
Neg logit contributions | |
atography | -8.522 |
IDA | -7.874 |
arthed | -7.755 |
irable | -7.740 |
kaya | -7.684 |
Token | We |
Token ID | 775 |
Feature activation | -8.97e+00 |
Pos logit contributions | |
weather | +6.101 |
Hurricane | +5.063 |
Pend | +4.148 |
Welsh | +4.147 |
Listen | +3.839 |
Neg logit contributions | |
�� | -18.603 |
atography | -18.570 |
enrich | -18.556 |
vana | -18.449 |
challeng | -18.434 |
Token | need |
Token ID | 761 |
Feature activation | -5.12e+00 |
Pos logit contributions | |
weather | +4.711 |
pite | +3.613 |
Hurricane | +3.549 |
misplaced | +3.458 |
usage | +3.440 |
Neg logit contributions | |
atography | -7.880 |
IDA | -7.503 |
arthed | -7.441 |
vana | -7.282 |
kaya | -7.269 |
Token | something |
Token ID | 1223 |
Feature activation | -3.94e+00 |
Pos logit contributions | |
weather | +2.440 |
pite | +1.835 |
Hurricane | +1.702 |
timed | +1.658 |
ritical | +1.649 |
Neg logit contributions | |
atography | -4.850 |
IDA | -4.683 |
irable | -4.633 |
kaya | -4.576 |
Ru | -4.550 |
Token | to |
Token ID | 284 |
Feature activation | -8.98e+00 |
Pos logit contributions | |
weather | +2.718 |
pite | +2.260 |
Hurricane | +2.172 |
usage | +2.157 |
ritical | +2.145 |
Neg logit contributions | |
atography | -2.932 |
IDA | -2.767 |
irable | -2.716 |
Ru | -2.714 |
vana | -2.696 |
Token | raise |
Token ID | 5298 |
Feature activation | -1.45e+01 |
Pos logit contributions | |
weather | +4.903 |
pite | +4.002 |
usage | +3.715 |
ritical | +3.706 |
Hurricane | +3.701 |
Neg logit contributions | |
IDA | -7.632 |
atography | -7.504 |
Ru | -7.058 |
vana | -7.039 |
Transparency | -6.961 |
Token | Likewise |
Token ID | 22660 |
Feature activation | -1.28e+00 |
Pos logit contributions | |
weather | +2.055 |
pite | +1.768 |
Hurricane | +1.740 |
usage | +1.665 |
ritical | +1.654 |
Neg logit contributions | |
atography | -2.020 |
IDA | -1.885 |
kaya | -1.869 |
irable | -1.866 |
enrich | -1.850 |
Token | , |
Token ID | 11 |
Feature activation | +7.25e-02 |
Pos logit contributions | |
weather | +0.918 |
pite | +0.784 |
ritical | +0.743 |
Hurricane | +0.738 |
usage | +0.736 |
Neg logit contributions | |
atography | -0.935 |
IDA | -0.897 |
Ru | -0.886 |
kaya | -0.874 |
wrongdoing | -0.867 |
Token | IGN |
Token ID | 28730 |
Feature activation | -1.17e+00 |
Pos logit contributions | |
atography | +0.050 |
IDA | +0.048 |
Ru | +0.047 |
kaya | +0.047 |
irable | +0.046 |
Neg logit contributions | |
weather | -0.055 |
pite | -0.047 |
Hurricane | -0.045 |
ritical | -0.045 |
usage | -0.045 |
Token | gave |
Token ID | 2921 |
Feature activation | -5.55e+00 |
Pos logit contributions | |
weather | +0.935 |
pite | +0.792 |
ritical | +0.772 |
Hurricane | +0.765 |
usage | +0.759 |
Neg logit contributions | |
atography | -0.757 |
IDA | -0.707 |
Ru | -0.703 |
kaya | -0.694 |
wrongdoing | -0.694 |
Token | the |
Token ID | 262 |
Feature activation | -4.69e+00 |
Pos logit contributions | |
weather | +3.552 |
pite | +3.047 |
Hurricane | +2.930 |
ritical | +2.870 |
usage | +2.820 |
Neg logit contributions | |
atography | -4.225 |
IDA | -4.027 |
kaya | -4.021 |
Ru | -4.019 |
irable | -3.959 |
Token | game |
Token ID | 983 |
Feature activation | -1.94e+01 |
Pos logit contributions | |
weather | +3.095 |
pite | +2.608 |
Hurricane | +2.533 |
ritical | +2.472 |
usage | +2.463 |
Neg logit contributions | |
atography | -3.570 |
IDA | -3.459 |
kaya | -3.397 |
Ru | -3.341 |
irable | -3.300 |
Token | a |
Token ID | 257 |
Feature activation | -1.33e+01 |
Pos logit contributions | |
weather | +9.205 |
Hurricane | +7.819 |
usage | +7.332 |
ritical | +7.246 |
pite | +7.058 |
Neg logit contributions | |
atography | -13.380 |
kaya | -13.307 |
Ru | -13.067 |
irable | -12.814 |
vana | -12.697 |
Token | ' |
Token ID | 705 |
Feature activation | -4.38e+00 |
Pos logit contributions | |
weather | +6.993 |
Hurricane | +5.969 |
pite | +5.841 |
ritical | +5.797 |
usage | +5.572 |
Neg logit contributions | |
atography | -9.832 |
Ru | -9.673 |
kaya | -9.426 |
IDA | -9.406 |
arthed | -9.379 |
Token | Pain |
Token ID | 38490 |
Feature activation | +4.82e-01 |
Pos logit contributions | |
weather | +3.542 |
pite | +2.975 |
usage | +2.962 |
ritical | +2.879 |
Hurricane | +2.814 |
Neg logit contributions | |
atography | -2.816 |
IDA | -2.569 |
wrongdoing | -2.550 |
kaya | -2.541 |
irable | -2.524 |
Token | ful |
Token ID | 913 |
Feature activation | -5.25e+00 |
Pos logit contributions | |
atography | +0.304 |
IDA | +0.290 |
Ru | +0.282 |
wrongdoing | +0.278 |
kaya | +0.275 |
Neg logit contributions | |
weather | -0.398 |
pite | -0.347 |
usage | -0.333 |
ritical | -0.332 |
Hurricane | -0.326 |
Token | ' |
Token ID | 6 |
Feature activation | -1.18e+01 |
Pos logit contributions | |
weather | +3.069 |
pite | +2.510 |
usage | +2.406 |
Hurricane | +2.403 |
ritical | +2.334 |
Neg logit contributions | |
atography | -4.320 |
Ru | -4.183 |
IDA | -4.156 |
kaya | -4.078 |
wrongdoing | -4.005 |
Token | , |
Token ID | 11 |
Feature activation | -2.11e-02 |
Pos logit contributions | |
atography | +0.969 |
IDA | +0.938 |
Ru | +0.931 |
irable | +0.903 |
kaya | +0.898 |
Neg logit contributions | |
weather | -1.023 |
pite | -0.884 |
ritical | -0.834 |
usage | -0.831 |
Hurricane | -0.825 |
Token | and |
Token ID | 290 |
Feature activation | -2.53e-01 |
Pos logit contributions | |
weather | +0.015 |
pite | +0.013 |
Hurricane | +0.012 |
ritical | +0.012 |
usage | +0.012 |
Neg logit contributions | |
atography | -0.016 |
IDA | -0.015 |
Ru | -0.015 |
irable | -0.015 |
kaya | -0.015 |
Token | our |
Token ID | 674 |
Feature activation | -2.67e+00 |
Pos logit contributions | |
weather | +0.188 |
pite | +0.162 |
Hurricane | +0.154 |
ritical | +0.154 |
usage | +0.151 |
Neg logit contributions | |
atography | -0.178 |
IDA | -0.170 |
Ru | -0.168 |
kaya | -0.165 |
irable | -0.165 |
Token | abilities |
Token ID | 7883 |
Feature activation | -5.59e+00 |
Pos logit contributions | |
weather | +1.937 |
pite | +1.696 |
Hurricane | +1.627 |
ritical | +1.601 |
usage | +1.582 |
Neg logit contributions | |
atography | -1.871 |
IDA | -1.816 |
Ru | -1.792 |
irable | -1.738 |
kaya | -1.729 |
Token | were |
Token ID | 547 |
Feature activation | -6.36e+00 |
Pos logit contributions | |
weather | +3.902 |
pite | +3.339 |
usage | +3.188 |
ritical | +3.186 |
timed | +3.116 |
Neg logit contributions | |
IDA | -3.919 |
atography | -3.905 |
Ru | -3.831 |
irable | -3.675 |
arthed | -3.649 |
Token | at |
Token ID | 379 |
Feature activation | -1.94e+01 |
Pos logit contributions | |
weather | +4.036 |
pite | +3.514 |
misplaced | +3.374 |
timed | +3.368 |
Hurricane | +3.344 |
Neg logit contributions | |
atography | -4.738 |
IDA | -4.612 |
Ru | -4.520 |
irable | -4.499 |
vana | -4.418 |
Token | a |
Token ID | 257 |
Feature activation | -1.17e+01 |
Pos logit contributions | |
weather | +9.983 |
pite | +9.006 |
Hurricane | +8.930 |
usage | +8.488 |
timed | +8.406 |
Neg logit contributions | |
atography | -13.616 |
IDA | -12.546 |
arthed | -12.187 |
Ru | -12.079 |
kaya | -12.038 |
Token | certain |
Token ID | 1728 |
Feature activation | -8.48e+00 |
Pos logit contributions | |
weather | +6.906 |
pite | +6.337 |
Hurricane | +6.266 |
sled | +6.126 |
misplaced | +6.058 |
Neg logit contributions | |
atography | -8.574 |
arthed | -7.862 |
IDA | -7.851 |
Ru | -7.806 |
kaya | -7.633 |
Token | point |
Token ID | 966 |
Feature activation | -2.21e+00 |
Pos logit contributions | |
weather | +5.223 |
pite | +4.671 |
Hurricane | +4.443 |
ritical | +4.346 |
timed | +4.328 |
Neg logit contributions | |
atography | -6.412 |
IDA | -6.069 |
Ru | -5.998 |
arthed | -5.926 |
kaya | -5.808 |
Token | - |
Token ID | 532 |
Feature activation | +1.02e+00 |
Pos logit contributions | |
weather | +1.562 |
pite | +1.348 |
Hurricane | +1.276 |
usage | +1.274 |
ritical | +1.266 |
Neg logit contributions | |
atography | -1.600 |
IDA | -1.560 |
Ru | -1.520 |
irable | -1.488 |
kaya | -1.468 |
Token | we |
Token ID | 356 |
Feature activation | +4.83e+00 |
Pos logit contributions | |
atography | +0.740 |
IDA | +0.710 |
Ru | +0.698 |
irable | +0.687 |
kaya | +0.685 |
Neg logit contributions | |
weather | -0.737 |
pite | -0.630 |
ritical | -0.595 |
usage | -0.592 |
Hurricane | -0.589 |
Token | month |
Token ID | 1227 |
Feature activation | -6.52e-01 |
Pos logit contributions | |
atography | +4.178 |
IDA | +4.002 |
Ru | +3.968 |
wrongdoing | +3.938 |
kaya | +3.884 |
Neg logit contributions | |
weather | -5.920 |
pite | -5.154 |
ritical | -4.976 |
usage | -4.904 |
Hurricane | -4.772 |
Token | , |
Token ID | 11 |
Feature activation | +1.67e+00 |
Pos logit contributions | |
weather | +0.468 |
pite | +0.395 |
usage | +0.386 |
ritical | +0.375 |
Hurricane | +0.371 |
Neg logit contributions | |
atography | -0.483 |
IDA | -0.464 |
Ru | -0.456 |
kaya | -0.445 |
irable | -0.441 |
Token | or |
Token ID | 393 |
Feature activation | +2.43e+00 |
Pos logit contributions | |
atography | +1.208 |
IDA | +1.162 |
Ru | +1.142 |
kaya | +1.124 |
irable | +1.114 |
Neg logit contributions | |
weather | -1.211 |
pite | -1.030 |
usage | -0.989 |
ritical | -0.984 |
Hurricane | -0.968 |
Token | as |
Token ID | 355 |
Feature activation | +1.60e+00 |
Pos logit contributions | |
atography | +1.720 |
IDA | +1.648 |
Ru | +1.623 |
kaya | +1.593 |
wrongdoing | +1.585 |
Neg logit contributions | |
weather | -1.794 |
pite | -1.534 |
usage | -1.477 |
ritical | -1.469 |
Hurricane | -1.437 |
Token | the |
Token ID | 262 |
Feature activation | -1.48e-01 |
Pos logit contributions | |
atography | +1.079 |
IDA | +1.034 |
Ru | +1.011 |
kaya | +1.005 |
wrongdoing | +0.998 |
Neg logit contributions | |
weather | -1.232 |
pite | -1.071 |
usage | -1.029 |
ritical | -1.023 |
Hurricane | -1.011 |
Token | report |
Token ID | 989 |
Feature activation | -1.93e+01 |
Pos logit contributions | |
weather | +0.103 |
pite | +0.086 |
usage | +0.083 |
ritical | +0.082 |
Hurricane | +0.081 |
Neg logit contributions | |
atography | -0.113 |
IDA | -0.108 |
Ru | -0.106 |
kaya | -0.105 |
irable | -0.103 |
Token | says |
Token ID | 1139 |
Feature activation | -7.52e+00 |
Pos logit contributions | |
weather | +9.681 |
timed | +8.053 |
warns | +7.951 |
Hurricane | +7.852 |
refers | +7.569 |
Neg logit contributions | |
atography | -13.837 |
IDA | -13.690 |
�� | -13.559 |
Ru | -13.187 |
irable | -13.108 |
Token | , |
Token ID | 11 |
Feature activation | +1.90e+00 |
Pos logit contributions | |
weather | +4.602 |
pite | +3.923 |
usage | +3.915 |
Hurricane | +3.764 |
ritical | +3.604 |
Neg logit contributions | |
atography | -5.913 |
IDA | -5.617 |
Ru | -5.453 |
kaya | -5.378 |
arthed | -5.300 |
Token | the |
Token ID | 262 |
Feature activation | +3.18e+00 |
Pos logit contributions | |
atography | +1.354 |
IDA | +1.298 |
Ru | +1.269 |
kaya | +1.252 |
wrongdoing | +1.242 |
Neg logit contributions | |
weather | -1.404 |
pite | -1.206 |
usage | -1.161 |
ritical | -1.150 |
Hurricane | -1.126 |
Token | equivalent |
Token ID | 7548 |
Feature activation | +5.79e+00 |
Pos logit contributions | |
atography | +2.210 |
IDA | +2.132 |
Ru | +2.093 |
kaya | +2.055 |
wrongdoing | +2.041 |
Neg logit contributions | |
weather | -2.379 |
pite | -2.049 |
usage | -1.968 |
ritical | -1.959 |
Hurricane | -1.909 |
Token | of |
Token ID | 286 |
Feature activation | +3.25e+00 |
Pos logit contributions | |
atography | +3.288 |
IDA | +3.195 |
wrongdoing | +3.173 |
Ru | +3.165 |
kaya | +3.060 |
Neg logit contributions | |
weather | -4.914 |
pite | -4.253 |
ritical | -4.123 |
usage | -3.955 |
Hurricane | -3.863 |
Token | regional |
Token ID | 7915 |
Feature activation | -1.93e+00 |
Pos logit contributions | |
atography | +2.003 |
IDA | +1.933 |
Ru | +1.898 |
kaya | +1.855 |
wrongdoing | +1.848 |
Neg logit contributions | |
weather | -2.184 |
pite | -1.860 |
ritical | -1.783 |
usage | -1.765 |
Hurricane | -1.751 |
Token | powers |
Token ID | 5635 |
Feature activation | +2.67e+00 |
Pos logit contributions | |
weather | +1.315 |
pite | +1.092 |
Hurricane | +1.047 |
ritical | +1.040 |
usage | +1.017 |
Neg logit contributions | |
atography | -1.464 |
IDA | -1.433 |
Ru | -1.390 |
irable | -1.374 |
vana | -1.362 |
Token | ; |
Token ID | 26 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
atography | +1.846 |
IDA | +1.780 |
Ru | +1.749 |
wrongdoing | +1.714 |
irable | +1.705 |
Neg logit contributions | |
weather | -2.034 |
pite | -1.734 |
ritical | -1.659 |
usage | -1.646 |
Hurricane | -1.619 |
Token | but |
Token ID | 475 |
Feature activation | -3.99e+00 |
Pos logit contributions | |
atography | +0.879 |
IDA | +0.849 |
Ru | +0.824 |
irable | +0.810 |
wrongdoing | +0.802 |
Neg logit contributions | |
weather | -1.052 |
pite | -0.907 |
ritical | -0.863 |
usage | -0.860 |
Hurricane | -0.859 |
Token | at |
Token ID | 379 |
Feature activation | -6.07e+00 |
Pos logit contributions | |
weather | +2.895 |
pite | +2.513 |
Hurricane | +2.445 |
usage | +2.331 |
ritical | +2.301 |
Neg logit contributions | |
atography | -2.850 |
IDA | -2.729 |
irable | -2.636 |
Ru | -2.623 |
kaya | -2.533 |
Token | other |
Token ID | 584 |
Feature activation | -1.93e+01 |
Pos logit contributions | |
weather | +4.100 |
pite | +3.540 |
Hurricane | +3.437 |
ritical | +3.371 |
usage | +3.311 |
Neg logit contributions | |
IDA | -4.360 |
atography | -4.283 |
Ru | -4.062 |
irable | -4.033 |
kaya | -3.968 |
Token | times |
Token ID | 1661 |
Feature activation | -1.66e+01 |
Pos logit contributions | |
weather | +9.666 |
pite | +7.926 |
ritical | +7.867 |
times | +7.834 |
Hurricane | +7.425 |
Neg logit contributions | |
atography | -14.020 |
IDA | -13.761 |
Ru | -13.078 |
irable | -13.018 |
arthed | -12.801 |
Token | it |
Token ID | 340 |
Feature activation | -2.72e+00 |
Pos logit contributions | |
weather | +8.734 |
pite | +7.313 |
Hurricane | +7.271 |
Pend | +6.719 |
ritical | +6.703 |
Neg logit contributions | |
atography | -12.148 |
IDA | -11.614 |
irable | -11.545 |
Ru | -11.050 |
MF | -10.701 |
Token | was |
Token ID | 373 |
Feature activation | -2.71e+00 |
Pos logit contributions | |
weather | +2.041 |
pite | +1.683 |
ritical | +1.637 |
Hurricane | +1.623 |
usage | +1.573 |
Neg logit contributions | |
IDA | -1.897 |
atography | -1.885 |
Ru | -1.794 |
wrongdoing | -1.754 |
irable | -1.742 |
Token | the |
Token ID | 262 |
Feature activation | -2.48e+00 |
Pos logit contributions | |
weather | +1.921 |
pite | +1.592 |
ritical | +1.573 |
Hurricane | +1.553 |
usage | +1.529 |
Neg logit contributions | |
atography | -1.958 |
IDA | -1.907 |
irable | -1.834 |
Ru | -1.826 |
wrongdoing | -1.803 |
Token | regional |
Token ID | 7915 |
Feature activation | -3.19e+00 |
Pos logit contributions | |
weather | +1.852 |
pite | +1.577 |
ritical | +1.544 |
Hurricane | +1.522 |
usage | +1.473 |
Neg logit contributions | |
atography | -1.712 |
IDA | -1.680 |
Ru | -1.606 |
irable | -1.582 |
kaya | -1.561 |
Token | , |
Token ID | 11 |
Feature activation | +8.87e+00 |
Pos logit contributions | |
atography | +2.467 |
IDA | +2.379 |
Ru | +2.348 |
wrongdoing | +2.311 |
kaya | +2.305 |
Neg logit contributions | |
weather | -2.366 |
pite | -1.999 |
ritical | -1.909 |
usage | -1.898 |
Hurricane | -1.842 |
Token | PD |
Token ID | 14340 |
Feature activation | +9.08e+00 |
Pos logit contributions | |
atography | +5.482 |
IDA | +5.320 |
Ru | +5.202 |
wrongdoing | +5.202 |
kaya | +5.082 |
Neg logit contributions | |
weather | -7.124 |
pite | -6.154 |
ritical | -5.963 |
usage | -5.914 |
Hurricane | -5.616 |
Token | V |
Token ID | 53 |
Feature activation | +7.12e+00 |
Pos logit contributions | |
Ru | +4.600 |
IDA | +4.569 |
wrongdoing | +4.333 |
atography | +4.322 |
kaya | +4.154 |
Neg logit contributions | |
weather | -8.119 |
pite | -7.197 |
ritical | -6.948 |
usage | -6.870 |
Hurricane | -6.699 |
Token | SA |
Token ID | 4090 |
Feature activation | +1.37e+01 |
Pos logit contributions | |
IDA | +3.038 |
Ru | +2.952 |
atography | +2.929 |
wrongdoing | +2.738 |
kaya | +2.681 |
Neg logit contributions | |
weather | -7.038 |
pite | -6.348 |
ritical | -6.190 |
usage | -6.064 |
Hurricane | -5.966 |
Token | has |
Token ID | 468 |
Feature activation | +1.06e+01 |
Pos logit contributions | |
atography | +8.286 |
wrongdoing | +8.113 |
IDA | +8.080 |
Ru | +8.051 |
enrich | +7.906 |
Neg logit contributions | |
weather | -10.172 |
pite | -9.118 |
ritical | -8.586 |
usage | -8.509 |
Hurricane | -7.748 |
Token | not |
Token ID | 407 |
Feature activation | +1.13e+01 |
Pos logit contributions | |
atography | +6.725 |
IDA | +6.432 |
wrongdoing | +6.419 |
Ru | +6.416 |
enrich | +6.325 |
Neg logit contributions | |
weather | -8.105 |
pite | -7.085 |
ritical | -6.750 |
usage | -6.724 |
Hurricane | -6.277 |
Token | thri |
Token ID | 19531 |
Feature activation | +4.58e+00 |
Pos logit contributions | |
atography | +7.205 |
wrongdoing | +6.787 |
kaya | +6.767 |
IDA | +6.752 |
Ru | +6.746 |
Neg logit contributions | |
weather | -8.580 |
pite | -7.420 |
usage | -7.242 |
ritical | -7.166 |
Hurricane | -6.725 |
Token | ved |
Token ID | 1079 |
Feature activation | +2.49e+00 |
Pos logit contributions | |
atography | +0.812 |
IDA | +0.694 |
Ru | +0.673 |
kaya | +0.628 |
wrongdoing | +0.609 |
Neg logit contributions | |
weather | -5.746 |
pite | -5.251 |
ritical | -5.130 |
usage | -5.123 |
Hurricane | -5.044 |
Token | . |
Token ID | 13 |
Feature activation | +1.72e+00 |
Pos logit contributions | |
atography | +1.757 |
IDA | +1.690 |
Ru | +1.664 |
kaya | +1.621 |
irable | +1.621 |
Neg logit contributions | |
weather | -1.873 |
pite | -1.589 |
usage | -1.510 |
ritical | -1.509 |
Hurricane | -1.497 |
Token | Oil |
Token ID | 11474 |
Feature activation | +5.86e-01 |
Pos logit contributions | |
atography | +1.303 |
IDA | +1.230 |
irable | +1.207 |
wrongdoing | +1.205 |
kaya | +1.205 |
Neg logit contributions | |
weather | -1.196 |
pite | -1.006 |
Hurricane | -0.978 |
usage | -0.962 |
ritical | -0.954 |
Token | exports |
Token ID | 15319 |
Feature activation | -4.50e+00 |
Pos logit contributions | |
atography | +0.429 |
IDA | +0.406 |
irable | +0.400 |
Ru | +0.398 |
wrongdoing | +0.393 |
Neg logit contributions | |
weather | -0.422 |
pite | -0.356 |
usage | -0.345 |
ritical | -0.344 |
Hurricane | -0.336 |
Token | drugs |
Token ID | 5010 |
Feature activation | -1.35e-01 |
Pos logit contributions | |
atography | +2.652 |
IDA | +2.542 |
Ru | +2.505 |
kaya | +2.458 |
wrongdoing | +2.445 |
Neg logit contributions | |
weather | -3.315 |
pite | -2.872 |
ritical | -2.754 |
usage | -2.744 |
Hurricane | -2.698 |
Token | " |
Token ID | 366 |
Feature activation | +5.25e+00 |
Pos logit contributions | |
weather | +0.098 |
pite | +0.085 |
usage | +0.080 |
Hurricane | +0.080 |
ritical | +0.079 |
Neg logit contributions | |
atography | -0.097 |
IDA | -0.094 |
Ru | -0.093 |
irable | -0.091 |
kaya | -0.089 |
Token | five |
Token ID | 13261 |
Feature activation | +4.46e-01 |
Pos logit contributions | |
atography | +3.320 |
IDA | +3.181 |
Ru | +3.147 |
wrongdoing | +3.080 |
kaya | +3.076 |
Neg logit contributions | |
weather | -4.227 |
pite | -3.663 |
ritical | -3.532 |
usage | -3.499 |
Hurricane | -3.428 |
Token | or |
Token ID | 393 |
Feature activation | -5.09e+00 |
Pos logit contributions | |
atography | +0.391 |
IDA | +0.378 |
Ru | +0.375 |
irable | +0.368 |
kaya | +0.366 |
Neg logit contributions | |
weather | -0.252 |
pite | -0.206 |
ritical | -0.195 |
usage | -0.190 |
Hurricane | -0.189 |
Token | six |
Token ID | 2237 |
Feature activation | +5.99e+00 |
Pos logit contributions | |
weather | +2.621 |
pite | +1.985 |
ritical | +1.876 |
Hurricane | +1.871 |
timed | +1.831 |
Neg logit contributions | |
atography | -4.644 |
irable | -4.513 |
IDA | -4.503 |
Ru | -4.458 |
Transparency | -4.343 |
Token | times |
Token ID | 1661 |
Feature activation | +1.19e+01 |
Pos logit contributions | |
atography | +6.274 |
IDA | +6.121 |
Ru | +6.062 |
wrongdoing | +6.024 |
kaya | +5.994 |
Neg logit contributions | |
weather | -2.350 |
pite | -1.688 |
ritical | -1.531 |
usage | -1.517 |
Hurricane | -1.387 |
Token | " |
Token ID | 1 |
Feature activation | +9.27e+00 |
Pos logit contributions | |
atography | +7.225 |
wrongdoing | +6.932 |
IDA | +6.931 |
Ru | +6.778 |
kaya | +6.757 |
Neg logit contributions | |
weather | -9.277 |
pite | -8.023 |
ritical | -7.854 |
usage | -7.714 |
Hurricane | -7.217 |
Token | from |
Token ID | 422 |
Feature activation | -1.10e+00 |
Pos logit contributions | |
atography | +5.658 |
IDA | +5.393 |
wrongdoing | +5.339 |
Ru | +5.321 |
kaya | +5.216 |
Neg logit contributions | |
weather | -7.451 |
pite | -6.462 |
ritical | -6.296 |
usage | -6.210 |
Hurricane | -5.933 |
Token | Jones |
Token ID | 5437 |
Feature activation | -3.02e+00 |
Pos logit contributions | |
weather | +0.749 |
pite | +0.632 |
Hurricane | +0.626 |
ritical | +0.595 |
usage | +0.590 |
Neg logit contributions | |
atography | -0.841 |
IDA | -0.805 |
Ru | -0.796 |
kaya | -0.794 |
irable | -0.781 |
Token | , |
Token ID | 11 |
Feature activation | +4.66e+00 |
Pos logit contributions | |
weather | +2.167 |
pite | +1.798 |
Hurricane | +1.728 |
usage | +1.687 |
ritical | +1.676 |
Neg logit contributions | |
atography | -2.200 |
IDA | -2.114 |
Ru | -2.087 |
irable | -2.066 |
kaya | -1.994 |
Token | but |
Token ID | 475 |
Feature activation | +4.50e+00 |
Pos logit contributions | |
atography | +3.068 |
IDA | +2.945 |
Ru | +2.901 |
kaya | +2.845 |
wrongdoing | +2.845 |
Neg logit contributions | |
weather | -3.673 |
pite | -3.163 |
ritical | -3.033 |
usage | -3.020 |
Hurricane | -2.952 |
Token | Photo |
Token ID | 5555 |
Feature activation | +7.10e+00 |
Pos logit contributions | |
atography | +3.516 |
IDA | +3.375 |
Ru | +3.323 |
wrongdoing | +3.272 |
kaya | +3.258 |
Neg logit contributions | |
weather | -4.152 |
pite | -3.573 |
ritical | -3.431 |
usage | -3.410 |
Hurricane | -3.324 |
Token | by |
Token ID | 416 |
Feature activation | +1.18e+01 |
Pos logit contributions | |
atography | +4.085 |
IDA | +3.903 |
Ru | +3.861 |
wrongdoing | +3.757 |
kaya | +3.740 |
Neg logit contributions | |
weather | -6.108 |
pite | -5.324 |
ritical | -5.128 |
usage | -5.124 |
Hurricane | -4.974 |
Token | : |
Token ID | 25 |
Feature activation | +1.04e+01 |
Pos logit contributions | |
atography | +7.291 |
IDA | +7.053 |
Marina | +7.040 |
Ru | +6.962 |
kaya | +6.860 |
Neg logit contributions | |
weather | -9.022 |
pite | -7.867 |
ritical | -7.649 |
usage | -7.587 |
Hurricane | -7.063 |
Token | NO |
Token ID | 8005 |
Feature activation | +6.88e+00 |
Pos logit contributions | |
atography | +6.578 |
IDA | +6.404 |
Ru | +6.307 |
Marina | +6.269 |
kaya | +6.184 |
Neg logit contributions | |
weather | -7.958 |
pite | -6.876 |
ritical | -6.643 |
usage | -6.634 |
Hurricane | -6.260 |
Token | VA |
Token ID | 11731 |
Feature activation | +6.63e+00 |
Pos logit contributions | |
atography | +3.930 |
IDA | +3.851 |
Ru | +3.735 |
wrongdoing | +3.642 |
kaya | +3.638 |
Neg logit contributions | |
weather | -5.910 |
pite | -5.190 |
ritical | -5.008 |
usage | -4.984 |
Hurricane | -4.832 |
Token | . |
Token ID | 13 |
Feature activation | +1.20e+01 |
Pos logit contributions | |
atography | +4.226 |
IDA | +4.097 |
Ru | +4.021 |
kaya | +3.937 |
wrongdoing | +3.932 |
Neg logit contributions | |
weather | -5.293 |
pite | -4.588 |
ritical | -4.413 |
usage | -4.392 |
Hurricane | -4.244 |
Token | TV |
Token ID | 6849 |
Feature activation | +4.79e+00 |
Pos logit contributions | |
atography | +7.169 |
Ru | +7.141 |
IDA | +7.084 |
kaya | +6.781 |
Marina | +6.735 |
Neg logit contributions | |
weather | -9.480 |
pite | -8.292 |
ritical | -8.015 |
usage | -7.995 |
Hurricane | -7.623 |
Token |
|
Token ID | 198 |
Feature activation | +3.03e+00 |
Pos logit contributions | |
atography | +3.053 |
IDA | +2.925 |
Ru | +2.881 |
wrongdoing | +2.833 |
kaya | +2.820 |
Neg logit contributions | |
weather | -3.866 |
pite | -3.342 |
ritical | -3.215 |
usage | -3.196 |
Hurricane | -3.117 |
Token |
|
Token ID | 198 |
Feature activation | +4.39e+00 |
Pos logit contributions | |
atography | +1.892 |
IDA | +1.774 |
wrongdoing | +1.744 |
Ru | +1.734 |
kaya | +1.722 |
Neg logit contributions | |
weather | -2.514 |
pite | -2.171 |
usage | -2.085 |
ritical | -2.078 |
Hurricane | -2.035 |
Token | On |
Token ID | 2202 |
Feature activation | +7.94e+00 |
Pos logit contributions | |
atography | +2.654 |
IDA | +2.539 |
Ru | +2.498 |
wrongdoing | +2.453 |
kaya | +2.443 |
Neg logit contributions | |
weather | -3.686 |
pite | -3.207 |
ritical | -3.091 |
usage | -3.074 |
Hurricane | -2.997 |
Token | a |
Token ID | 257 |
Feature activation | +1.05e+00 |
Pos logit contributions | |
atography | +4.808 |
IDA | +4.632 |
Ru | +4.605 |
wrongdoing | +4.541 |
kaya | +4.528 |
Neg logit contributions | |
weather | -6.449 |
pite | -5.619 |
ritical | -5.435 |
usage | -5.386 |
Hurricane | -5.163 |
Token | think |
Token ID | 892 |
Feature activation | +8.16e-01 |
Pos logit contributions | |
weather | +2.484 |
pite | +1.971 |
Hurricane | +1.845 |
usage | +1.839 |
ritical | +1.833 |
Neg logit contributions | |
IDA | -3.673 |
atography | -3.619 |
Ru | -3.522 |
arthed | -3.473 |
irable | -3.455 |
Token | , |
Token ID | 11 |
Feature activation | +4.68e+00 |
Pos logit contributions | |
atography | +0.587 |
IDA | +0.569 |
Ru | +0.556 |
kaya | +0.543 |
irable | +0.540 |
Neg logit contributions | |
weather | -0.596 |
pite | -0.508 |
ritical | -0.482 |
usage | -0.477 |
Hurricane | -0.474 |
Token | the |
Token ID | 262 |
Feature activation | +2.96e-01 |
Pos logit contributions | |
atography | +3.481 |
IDA | +3.362 |
Ru | +3.311 |
kaya | +3.259 |
wrongdoing | +3.255 |
Neg logit contributions | |
weather | -3.291 |
pite | -2.776 |
ritical | -2.653 |
usage | -2.638 |
Hurricane | -2.566 |
Token | celebrities |
Token ID | 21305 |
Feature activation | +4.47e+00 |
Pos logit contributions | |
atography | +0.198 |
IDA | +0.194 |
Ru | +0.187 |
kaya | +0.186 |
irable | +0.182 |
Neg logit contributions | |
weather | -0.230 |
pite | -0.197 |
usage | -0.188 |
ritical | -0.187 |
Hurricane | -0.187 |
Token | in |
Token ID | 287 |
Feature activation | +5.35e+00 |
Pos logit contributions | |
atography | +3.148 |
IDA | +3.034 |
Ru | +2.989 |
wrongdoing | +2.932 |
kaya | +2.929 |
Neg logit contributions | |
weather | -3.319 |
pite | -2.828 |
ritical | -2.706 |
usage | -2.692 |
Hurricane | -2.620 |
Token | attendance |
Token ID | 14858 |
Feature activation | +1.14e+01 |
Pos logit contributions | |
atography | +3.863 |
IDA | +3.717 |
Ru | +3.677 |
wrongdoing | +3.629 |
kaya | +3.605 |
Neg logit contributions | |
weather | -3.842 |
pite | -3.263 |
ritical | -3.122 |
usage | -3.093 |
Hurricane | -2.995 |
Token | quite |
Token ID | 2407 |
Feature activation | -4.89e-02 |
Pos logit contributions | |
atography | +7.085 |
wrongdoing | +6.737 |
enrich | +6.642 |
IDA | +6.553 |
Ru | +6.552 |
Neg logit contributions | |
weather | -8.732 |
pite | -7.606 |
ritical | -7.509 |
usage | -7.372 |
Hurricane | -6.955 |
Token | enjoyed |
Token ID | 8359 |
Feature activation | -8.49e-01 |
Pos logit contributions | |
weather | +0.033 |
pite | +0.028 |
ritical | +0.027 |
usage | +0.026 |
Hurricane | +0.026 |
Neg logit contributions | |
atography | -0.037 |
IDA | -0.037 |
Ru | -0.035 |
kaya | -0.035 |
irable | -0.035 |
Token | his |
Token ID | 465 |
Feature activation | +5.34e+00 |
Pos logit contributions | |
weather | +0.555 |
pite | +0.459 |
Hurricane | +0.442 |
usage | +0.436 |
ritical | +0.430 |
Neg logit contributions | |
atography | -0.676 |
IDA | -0.653 |
Ru | -0.634 |
kaya | -0.630 |
irable | -0.626 |
Token | qu |
Token ID | 627 |
Feature activation | -2.50e-01 |
Pos logit contributions | |
atography | +3.416 |
IDA | +3.266 |
Ru | +3.227 |
wrongdoing | +3.190 |
kaya | +3.162 |
Neg logit contributions | |
weather | -4.279 |
pite | -3.691 |
ritical | -3.553 |
usage | -3.526 |
Hurricane | -3.423 |
Token | ips |
Token ID | 2419 |
Feature activation | -4.15e-01 |
Pos logit contributions | |
weather | +0.181 |
pite | +0.156 |
ritical | +0.151 |
usage | +0.148 |
Hurricane | +0.138 |
Neg logit contributions | |
atography | -0.178 |
IDA | -0.177 |
wrongdoing | -0.174 |
Ru | -0.172 |
kaya | -0.168 |
Token | The |
Token ID | 464 |
Feature activation | +4.10e+00 |
Pos logit contributions | |
atography | +0.228 |
irable | +0.210 |
wrongdoing | +0.210 |
IDA | +0.209 |
enrich | +0.208 |
Neg logit contributions | |
weather | -0.290 |
pite | -0.250 |
usage | -0.242 |
ritical | -0.235 |
Hurricane | -0.230 |
Token | Bottom |
Token ID | 24530 |
Feature activation | -3.04e+00 |
Pos logit contributions | |
atography | +2.834 |
IDA | +2.732 |
Ru | +2.687 |
kaya | +2.633 |
wrongdoing | +2.622 |
Neg logit contributions | |
weather | -3.103 |
pite | -2.661 |
ritical | -2.550 |
usage | -2.540 |
Hurricane | -2.473 |
Token | Line |
Token ID | 6910 |
Feature activation | +2.76e+00 |
Pos logit contributions | |
weather | +1.637 |
pite | +1.329 |
Hurricane | +1.254 |
usage | +1.239 |
ritical | +1.210 |
Neg logit contributions | |
atography | -2.698 |
IDA | -2.690 |
Ru | -2.599 |
wrongdoing | -2.570 |
irable | -2.534 |
Token |
|
Token ID | 198 |
Feature activation | +2.44e+00 |
Pos logit contributions | |
atography | +1.653 |
IDA | +1.581 |
Ru | +1.556 |
wrongdoing | +1.527 |
kaya | +1.520 |
Neg logit contributions | |
weather | -2.331 |
pite | -2.030 |
ritical | -1.957 |
usage | -1.947 |
Hurricane | -1.899 |
Token |
|
Token ID | 198 |
Feature activation | +4.34e+00 |
Pos logit contributions | |
atography | +1.472 |
IDA | +1.404 |
Ru | +1.379 |
wrongdoing | +1.357 |
kaya | +1.351 |
Neg logit contributions | |
weather | -2.054 |
pite | -1.785 |
ritical | -1.716 |
usage | -1.714 |
Hurricane | -1.672 |
Token | Any |
Token ID | 7149 |
Feature activation | +1.14e+01 |
Pos logit contributions | |
atography | +2.682 |
IDA | +2.548 |
Ru | +2.499 |
wrongdoing | +2.480 |
kaya | +2.466 |
Neg logit contributions | |
weather | -3.607 |
pite | -3.128 |
usage | -3.006 |
ritical | -3.000 |
Hurricane | -2.916 |
Token | objective |
Token ID | 9432 |
Feature activation | +4.02e+00 |
Pos logit contributions | |
atography | +7.264 |
wrongdoing | +7.204 |
IDA | +7.078 |
Ru | +6.973 |
kaya | +6.945 |
Neg logit contributions | |
weather | -8.464 |
pite | -7.230 |
ritical | -7.011 |
usage | -6.907 |
Hurricane | -6.669 |
Token | observer |
Token ID | 22890 |
Feature activation | +8.51e+00 |
Pos logit contributions | |
atography | +2.670 |
IDA | +2.563 |
Ru | +2.526 |
kaya | +2.468 |
wrongdoing | +2.467 |
Neg logit contributions | |
weather | -3.149 |
pite | -2.707 |
ritical | -2.603 |
usage | -2.589 |
Hurricane | -2.524 |
Token | would |
Token ID | 561 |
Feature activation | +9.94e+00 |
Pos logit contributions | |
atography | +5.444 |
wrongdoing | +5.142 |
IDA | +5.116 |
kaya | +5.108 |
Ru | +5.072 |
Neg logit contributions | |
weather | -6.563 |
pite | -5.667 |
ritical | -5.499 |
usage | -5.469 |
Hurricane | -5.269 |
Token | have |
Token ID | 423 |
Feature activation | +1.02e+01 |
Pos logit contributions | |
atography | +6.921 |
kaya | +6.520 |
enrich | +6.486 |
vana | +6.404 |
wrongdoing | +6.396 |
Neg logit contributions | |
weather | -7.076 |
pite | -6.066 |
ritical | -5.834 |
usage | -5.823 |
Hurricane | -5.555 |
Token | to |
Token ID | 284 |
Feature activation | +5.55e-01 |
Pos logit contributions | |
atography | +6.105 |
wrongdoing | +5.854 |
IDA | +5.767 |
Ru | +5.737 |
kaya | +5.610 |
Neg logit contributions | |
weather | -8.155 |
pite | -7.195 |
usage | -6.808 |
ritical | -6.795 |
Hurricane | -6.423 |
Token | frequent |
Token ID | 10792 |
Feature activation | +3.26e-01 |
Pos logit contributions | |
atography | +2.213 |
IDA | +2.136 |
Ru | +2.114 |
kaya | +2.065 |
wrongdoing | +2.053 |
Neg logit contributions | |
weather | -2.271 |
pite | -1.940 |
ritical | -1.857 |
usage | -1.850 |
Hurricane | -1.804 |
Token | written |
Token ID | 3194 |
Feature activation | +4.32e+00 |
Pos logit contributions | |
atography | +0.222 |
IDA | +0.215 |
Ru | +0.213 |
irable | +0.206 |
kaya | +0.205 |
Neg logit contributions | |
weather | -0.249 |
pite | -0.215 |
usage | -0.207 |
ritical | -0.206 |
Hurricane | -0.202 |
Token | and |
Token ID | 290 |
Feature activation | -1.75e+00 |
Pos logit contributions | |
atography | +3.022 |
IDA | +2.909 |
Ru | +2.869 |
wrongdoing | +2.825 |
kaya | +2.815 |
Neg logit contributions | |
weather | -3.206 |
pite | -2.736 |
ritical | -2.622 |
usage | -2.606 |
Hurricane | -2.530 |
Token | spoken |
Token ID | 9635 |
Feature activation | +5.69e+00 |
Pos logit contributions | |
weather | +1.245 |
pite | +1.071 |
usage | +1.036 |
ritical | +1.023 |
Hurricane | +0.996 |
Neg logit contributions | |
atography | -1.258 |
IDA | -1.220 |
Ru | -1.213 |
irable | -1.183 |
kaya | -1.182 |
Token | descriptions |
Token ID | 16969 |
Feature activation | -2.19e+00 |
Pos logit contributions | |
atography | +3.784 |
IDA | +3.631 |
Ru | +3.573 |
wrongdoing | +3.561 |
kaya | +3.517 |
Neg logit contributions | |
weather | -4.390 |
pite | -3.782 |
ritical | -3.631 |
usage | -3.585 |
Hurricane | -3.490 |
Token | . |
Token ID | 13 |
Feature activation | -1.59e-01 |
Pos logit contributions | |
weather | +1.525 |
pite | +1.290 |
usage | +1.239 |
ritical | +1.235 |
Hurricane | +1.210 |
Neg logit contributions | |
atography | -1.586 |
IDA | -1.560 |
Ru | -1.555 |
kaya | -1.495 |
irable | -1.485 |
Token | As |
Token ID | 1081 |
Feature activation | +8.94e+00 |
Pos logit contributions | |
weather | +0.113 |
pite | +0.095 |
Hurricane | +0.093 |
usage | +0.091 |
ritical | +0.090 |
Neg logit contributions | |
atography | -0.117 |
IDA | -0.110 |
irable | -0.109 |
Ru | -0.109 |
wrongdoing | -0.109 |
Token | the |
Token ID | 262 |
Feature activation | +4.16e+00 |
Pos logit contributions | |
atography | +6.032 |
IDA | +5.854 |
Ru | +5.756 |
wrongdoing | +5.754 |
kaya | +5.640 |
Neg logit contributions | |
weather | -6.643 |
pite | -5.605 |
ritical | -5.445 |
usage | -5.369 |
Hurricane | -5.101 |
Token | dust |
Token ID | 8977 |
Feature activation | -1.28e+00 |
Pos logit contributions | |
atography | +3.014 |
IDA | +2.907 |
Ru | +2.869 |
wrongdoing | +2.815 |
kaya | +2.812 |
Neg logit contributions | |
weather | -2.993 |
pite | -2.540 |
ritical | -2.428 |
usage | -2.413 |
Hurricane | -2.348 |
Token | settled |
Token ID | 10282 |
Feature activation | +4.15e+00 |
Pos logit contributions | |
weather | +0.953 |
pite | +0.791 |
usage | +0.762 |
ritical | +0.758 |
Hurricane | +0.748 |
Neg logit contributions | |
atography | -0.898 |
IDA | -0.879 |
Ru | -0.849 |
wrongdoing | -0.842 |
kaya | -0.837 |
Token | and |
Token ID | 290 |
Feature activation | +1.00e+01 |
Pos logit contributions | |
atography | +2.934 |
IDA | +2.823 |
Ru | +2.785 |
wrongdoing | +2.737 |
kaya | +2.729 |
Neg logit contributions | |
weather | -3.069 |
pite | -2.618 |
ritical | -2.500 |
usage | -2.486 |
Hurricane | -2.427 |
Token | Kh |
Token ID | 5311 |
Feature activation | +4.26e+00 |
Pos logit contributions | |
atography | +5.343 |
IDA | +5.092 |
Ru | +5.075 |
wrongdoing | +5.042 |
kaya | +4.946 |
Neg logit contributions | |
weather | -6.178 |
pite | -5.351 |
ritical | -5.176 |
usage | -5.129 |
Hurricane | -4.876 |
Token | y |
Token ID | 88 |
Feature activation | +1.94e+00 |
Pos logit contributions | |
atography | +2.145 |
IDA | +2.028 |
Ru | +1.993 |
wrongdoing | +1.945 |
kaya | +1.941 |
Neg logit contributions | |
weather | -4.003 |
pite | -3.540 |
ritical | -3.428 |
usage | -3.411 |
Hurricane | -3.340 |
Token | ber |
Token ID | 527 |
Feature activation | +4.27e+00 |
Pos logit contributions | |
atography | +1.670 |
IDA | +1.619 |
Ru | +1.592 |
wrongdoing | +1.575 |
kaya | +1.564 |
Neg logit contributions | |
weather | -1.156 |
pite | -0.938 |
usage | -0.880 |
ritical | -0.878 |
Hurricane | -0.835 |
Token | district |
Token ID | 4783 |
Feature activation | +6.20e-01 |
Pos logit contributions | |
atography | +2.885 |
IDA | +2.769 |
Ru | +2.731 |
wrongdoing | +2.684 |
kaya | +2.676 |
Neg logit contributions | |
weather | -3.291 |
pite | -2.823 |
ritical | -2.713 |
usage | -2.695 |
Hurricane | -2.625 |
Token | , |
Token ID | 11 |
Feature activation | +2.53e+00 |
Pos logit contributions | |
atography | +0.441 |
IDA | +0.424 |
Ru | +0.417 |
irable | +0.408 |
wrongdoing | +0.407 |
Neg logit contributions | |
weather | -0.463 |
pite | -0.392 |
ritical | -0.373 |
usage | -0.369 |
Hurricane | -0.368 |
Token | the |
Token ID | 262 |
Feature activation | +3.64e+00 |
Pos logit contributions | |
atography | +1.807 |
IDA | +1.740 |
Ru | +1.709 |
kaya | +1.682 |
irable | +1.678 |
Neg logit contributions | |
weather | -1.868 |
pite | -1.587 |
ritical | -1.511 |
usage | -1.502 |
Hurricane | -1.487 |
Token | BBC |
Token ID | 7823 |
Feature activation | -1.46e+00 |
Pos logit contributions | |
atography | +2.447 |
IDA | +2.364 |
Ru | +2.313 |
kaya | +2.271 |
wrongdoing | +2.261 |
Neg logit contributions | |
weather | -2.823 |
pite | -2.424 |
ritical | -2.321 |
usage | -2.295 |
Hurricane | -2.264 |
Token | 's |
Token ID | 338 |
Feature activation | +5.79e+00 |
Pos logit contributions | |
weather | +0.878 |
ritical | +0.679 |
pite | +0.678 |
Hurricane | +0.649 |
usage | +0.635 |
Neg logit contributions | |
atography | -1.264 |
IDA | -1.204 |
Ru | -1.178 |
irable | -1.175 |
kaya | -1.171 |
Token | M |
Token ID | 337 |
Feature activation | -1.07e+00 |
Pos logit contributions | |
atography | +3.439 |
IDA | +3.302 |
Ru | +3.256 |
wrongdoing | +3.198 |
kaya | +3.162 |
Neg logit contributions | |
weather | -4.873 |
pite | -4.262 |
ritical | -4.103 |
usage | -4.099 |
Hurricane | -3.952 |
Token | Ily |
Token ID | 49804 |
Feature activation | +2.80e+00 |
Pos logit contributions | |
weather | +0.850 |
pite | +0.738 |
ritical | +0.717 |
usage | +0.702 |
Hurricane | +0.676 |
Neg logit contributions | |
atography | -0.681 |
IDA | -0.659 |
wrongdoing | -0.646 |
Ru | -0.636 |
kaya | -0.628 |
Token | as |
Token ID | 292 |
Feature activation | -2.84e+00 |
Pos logit contributions | |
atography | +1.723 |
IDA | +1.683 |
Ru | +1.654 |
wrongdoing | +1.639 |
irable | +1.610 |
Neg logit contributions | |
weather | -2.328 |
pite | -1.990 |
ritical | -1.936 |
usage | -1.921 |
Hurricane | -1.856 |
Token | the |
Token ID | 262 |
Feature activation | -2.74e+00 |
Pos logit contributions | |
weather | +0.709 |
pite | +0.585 |
Hurricane | +0.571 |
ritical | +0.550 |
usage | +0.548 |
Neg logit contributions | |
atography | -0.740 |
Ru | -0.704 |
IDA | -0.704 |
kaya | -0.693 |
irable | -0.683 |
Token | US |
Token ID | 1294 |
Feature activation | -7.26e-01 |
Pos logit contributions | |
weather | +1.728 |
pite | +1.386 |
Hurricane | +1.327 |
ritical | +1.308 |
usage | +1.273 |
Neg logit contributions | |
atography | -2.249 |
IDA | -2.161 |
Ru | -2.159 |
irable | -2.113 |
kaya | -2.106 |
Token | Anti |
Token ID | 12297 |
Feature activation | +1.98e+00 |
Pos logit contributions | |
weather | +0.549 |
pite | +0.458 |
ritical | +0.450 |
usage | +0.443 |
Hurricane | +0.439 |
Neg logit contributions | |
atography | -0.503 |
irable | -0.472 |
Ru | -0.472 |
kaya | -0.467 |
IDA | -0.466 |
Token | - |
Token ID | 12 |
Feature activation | -1.22e+00 |
Pos logit contributions | |
atography | +1.505 |
IDA | +1.450 |
Ru | +1.425 |
kaya | +1.408 |
wrongdoing | +1.403 |
Neg logit contributions | |
weather | -1.358 |
pite | -1.138 |
ritical | -1.090 |
usage | -1.083 |
Hurricane | -1.055 |
Token | D |
Token ID | 35 |
Feature activation | -4.04e+00 |
Pos logit contributions | |
weather | +1.149 |
pite | +0.982 |
usage | +0.975 |
ritical | +0.967 |
Hurricane | +0.942 |
Neg logit contributions | |
atography | -0.636 |
irable | -0.580 |
vana | -0.573 |
IDA | -0.571 |
kaya | -0.571 |
Token | oping |
Token ID | 15816 |
Feature activation | +2.05e+00 |
Pos logit contributions | |
weather | +2.272 |
ritical | +1.812 |
pite | +1.806 |
usage | +1.804 |
Hurricane | +1.633 |
Neg logit contributions | |
atography | -3.496 |
vana | -3.369 |
Ru | -3.333 |
kaya | -3.276 |
IDA | -3.274 |
Token | Agency |
Token ID | 7732 |
Feature activation | -2.22e+00 |
Pos logit contributions | |
atography | +1.460 |
IDA | +1.407 |
Ru | +1.394 |
kaya | +1.361 |
wrongdoing | +1.356 |
Neg logit contributions | |
weather | -1.503 |
pite | -1.273 |
ritical | -1.225 |
usage | -1.214 |
Hurricane | -1.187 |
Token | once |
Token ID | 1752 |
Feature activation | -5.07e+00 |
Pos logit contributions | |
weather | +1.436 |
pite | +1.168 |
Hurricane | +1.136 |
ritical | +1.117 |
usage | +1.107 |
Neg logit contributions | |
atography | -1.760 |
IDA | -1.686 |
Ru | -1.680 |
irable | -1.665 |
kaya | -1.659 |
Token | again |
Token ID | 757 |
Feature activation | -9.74e-01 |
Pos logit contributions | |
weather | +3.342 |
Hurricane | +2.712 |
pite | +2.671 |
Pend | +2.551 |
usage | +2.528 |
Neg logit contributions | |
atography | -3.931 |
irable | -3.738 |
Ru | -3.673 |
kaya | -3.639 |
IDA | -3.623 |
Token | after |
Token ID | 706 |
Feature activation | -2.61e+00 |
Pos logit contributions | |
weather | +0.696 |
pite | +0.582 |
Hurricane | +0.548 |
usage | +0.546 |
ritical | +0.541 |
Neg logit contributions | |
atography | -0.724 |
IDA | -0.698 |
Ru | -0.690 |
irable | -0.677 |
kaya | -0.670 |
Token | testing |
Token ID | 4856 |
Feature activation | +1.70e+00 |
Pos logit contributions | |
weather | +1.744 |
pite | +1.445 |
Hurricane | +1.437 |
usage | +1.382 |
ritical | +1.346 |
Neg logit contributions | |
atography | -2.026 |
IDA | -1.903 |
kaya | -1.900 |
irable | -1.895 |
Ru | -1.889 |
Token | but |
Token ID | 475 |
Feature activation | -1.89e+00 |
Pos logit contributions | |
atography | +0.782 |
IDA | +0.749 |
Ru | +0.737 |
irable | +0.718 |
kaya | +0.713 |
Neg logit contributions | |
weather | -0.965 |
pite | -0.825 |
usage | -0.795 |
ritical | -0.790 |
Hurricane | -0.779 |
Token | point |
Token ID | 966 |
Feature activation | -3.56e+00 |
Pos logit contributions | |
weather | +1.384 |
pite | +1.171 |
Hurricane | +1.123 |
usage | +1.119 |
ritical | +1.109 |
Neg logit contributions | |
atography | -1.353 |
IDA | -1.292 |
Ru | -1.283 |
irable | -1.253 |
kaya | -1.247 |
Token | less |
Token ID | 1342 |
Feature activation | -3.09e+00 |
Pos logit contributions | |
weather | +2.654 |
pite | +2.265 |
ritical | +2.178 |
Hurricane | +2.172 |
usage | +2.160 |
Neg logit contributions | |
atography | -2.393 |
IDA | -2.339 |
irable | -2.282 |
Ru | -2.252 |
kaya | -2.233 |
Token | to |
Token ID | 284 |
Feature activation | +3.91e+00 |
Pos logit contributions | |
weather | +2.290 |
pite | +2.009 |
Hurricane | +1.932 |
ritical | +1.892 |
usage | +1.864 |
Neg logit contributions | |
atography | -2.040 |
IDA | -1.970 |
Ru | -1.946 |
irable | -1.931 |
kaya | -1.879 |
Token | the |
Token ID | 262 |
Feature activation | +4.59e+00 |
Pos logit contributions | |
atography | +2.689 |
IDA | +2.587 |
Ru | +2.550 |
kaya | +2.500 |
wrongdoing | +2.497 |
Neg logit contributions | |
weather | -2.959 |
pite | -2.533 |
ritical | -2.429 |
usage | -2.415 |
Hurricane | -2.355 |
Token | operatives |
Token ID | 27269 |
Feature activation | +1.04e+01 |
Pos logit contributions | |
atography | +3.153 |
IDA | +3.032 |
Ru | +2.991 |
wrongdoing | +2.947 |
kaya | +2.934 |
Neg logit contributions | |
weather | -3.466 |
pite | -2.966 |
ritical | -2.845 |
usage | -2.827 |
Hurricane | -2.745 |
Token | around |
Token ID | 1088 |
Feature activation | +1.03e+01 |
Pos logit contributions | |
atography | +6.777 |
wrongdoing | +6.559 |
IDA | +6.479 |
kaya | +6.385 |
Ru | +6.344 |
Neg logit contributions | |
weather | -7.763 |
pite | -6.703 |
ritical | -6.494 |
usage | -6.428 |
Hurricane | -6.114 |
Token | Romney |
Token ID | 8431 |
Feature activation | +5.65e+00 |
Pos logit contributions | |
atography | +6.969 |
wrongdoing | +6.725 |
IDA | +6.670 |
Ru | +6.626 |
Marina | +6.534 |
Neg logit contributions | |
weather | -7.479 |
pite | -6.493 |
ritical | -6.224 |
usage | -6.127 |
Hurricane | -5.742 |
Token | � |
Token ID | 126 |
Feature activation | +2.24e+00 |
Pos logit contributions | |
atography | +3.880 |
IDA | +3.733 |
Ru | +3.681 |
wrongdoing | +3.636 |
kaya | +3.614 |
Neg logit contributions | |
weather | -4.248 |
pite | -3.642 |
ritical | -3.498 |
usage | -3.477 |
Hurricane | -3.369 |
Token | � |
Token ID | 245 |
Feature activation | +5.92e+00 |
Pos logit contributions | |
atography | +1.410 |
IDA | +1.350 |
Ru | +1.324 |
wrongdoing | +1.305 |
kaya | +1.297 |
Neg logit contributions | |
weather | -1.821 |
pite | -1.581 |
ritical | -1.516 |
usage | -1.514 |
Hurricane | -1.478 |
Token | who |
Token ID | 8727 |
Feature activation | +1.28e+01 |
Pos logit contributions | |
atography | +3.382 |
IDA | +3.233 |
Ru | +3.183 |
wrongdoing | +3.132 |
kaya | +3.112 |
Neg logit contributions | |
weather | -5.120 |
pite | -4.474 |
ritical | -4.329 |
usage | -4.299 |
Hurricane | -4.193 |
Token | ont |
Token ID | 756 |
Feature activation | +3.26e+00 |
Pos logit contributions | |
atography | +3.772 |
IDA | +3.621 |
Ru | +3.567 |
wrongdoing | +3.502 |
kaya | +3.499 |
Neg logit contributions | |
weather | -4.563 |
pite | -3.933 |
ritical | -3.785 |
usage | -3.763 |
Hurricane | -3.672 |
Token | pour |
Token ID | 12797 |
Feature activation | +8.76e+00 |
Pos logit contributions | |
atography | +2.159 |
IDA | +2.077 |
Ru | +2.035 |
kaya | +2.005 |
wrongdoing | +1.986 |
Neg logit contributions | |
weather | -2.568 |
pite | -2.214 |
ritical | -2.128 |
usage | -2.108 |
Hurricane | -2.068 |
Token | l |
Token ID | 300 |
Feature activation | -1.70e-01 |
Pos logit contributions | |
atography | +5.860 |
IDA | +5.650 |
Ru | +5.648 |
wrongdoing | +5.508 |
kaya | +5.455 |
Neg logit contributions | |
weather | -6.481 |
pite | -5.574 |
ritical | -5.322 |
usage | -5.322 |
Hurricane | -5.160 |
Token | ' |
Token ID | 6 |
Feature activation | +6.00e-01 |
Pos logit contributions | |
weather | +0.116 |
pite | +0.098 |
ritical | +0.093 |
usage | +0.092 |
Hurricane | +0.087 |
Neg logit contributions | |
atography | -0.130 |
IDA | -0.125 |
Ru | -0.124 |
kaya | -0.123 |
wrongdoing | -0.123 |
Token | inst |
Token ID | 8625 |
Feature activation | +7.21e+00 |
Pos logit contributions | |
atography | +0.392 |
IDA | +0.374 |
wrongdoing | +0.369 |
kaya | +0.366 |
Ru | +0.366 |
Neg logit contributions | |
weather | -0.482 |
pite | -0.414 |
usage | -0.407 |
ritical | -0.398 |
Hurricane | -0.376 |
Token | ant |
Token ID | 415 |
Feature activation | +3.99e+00 |
Pos logit contributions | |
atography | +4.768 |
IDA | +4.516 |
Ru | +4.490 |
irable | +4.388 |
kaya | +4.363 |
Neg logit contributions | |
weather | -5.518 |
pite | -4.729 |
usage | -4.523 |
Hurricane | -4.496 |
ritical | -4.488 |
Token | neut |
Token ID | 22190 |
Feature activation | +5.94e+00 |
Pos logit contributions | |
atography | +2.575 |
IDA | +2.468 |
Ru | +2.431 |
wrongdoing | +2.383 |
kaya | +2.382 |
Neg logit contributions | |
weather | -3.208 |
pite | -2.774 |
ritical | -2.668 |
usage | -2.650 |
Hurricane | -2.582 |
Token | re |
Token ID | 260 |
Feature activation | +2.04e+00 |
Pos logit contributions | |
atography | +3.979 |
IDA | +3.814 |
Ru | +3.774 |
wrongdoing | +3.703 |
kaya | +3.686 |
Neg logit contributions | |
weather | -4.541 |
pite | -3.912 |
ritical | -3.736 |
usage | -3.726 |
Hurricane | -3.631 |
Token | . |
Token ID | 13 |
Feature activation | +7.10e+00 |
Pos logit contributions | |
atography | +1.343 |
IDA | +1.277 |
Ru | +1.254 |
kaya | +1.240 |
wrongdoing | +1.233 |
Neg logit contributions | |
weather | -1.634 |
pite | -1.411 |
ritical | -1.353 |
usage | -1.352 |
Hurricane | -1.320 |
Token | J |
Token ID | 449 |
Feature activation | +2.43e-01 |
Pos logit contributions | |
atography | +4.147 |
IDA | +4.041 |
Ru | +4.005 |
wrongdoing | +3.853 |
kaya | +3.836 |
Neg logit contributions | |
weather | -5.985 |
pite | -5.213 |
ritical | -5.053 |
usage | -4.999 |
Hurricane | -4.827 |
Token | ' |
Token ID | 6 |
Feature activation | -1.66e+00 |
Pos logit contributions | |
atography | +0.192 |
IDA | +0.184 |
Ru | +0.182 |
wrongdoing | +0.181 |
kaya | +0.181 |
Neg logit contributions | |
weather | -0.161 |
pite | -0.133 |
ritical | -0.130 |
usage | -0.127 |
Hurricane | -0.120 |
Token | Access |
Token ID | 8798 |
Feature activation | -1.86e-01 |
Pos logit contributions | |
weather | +0.247 |
pite | +0.204 |
usage | +0.197 |
Hurricane | +0.194 |
ritical | +0.191 |
Neg logit contributions | |
atography | -0.264 |
IDA | -0.254 |
Ru | -0.247 |
wrongdoing | -0.241 |
kaya | -0.241 |
Token |
|
Token ID | 198 |
Feature activation | -3.91e+00 |
Pos logit contributions | |
weather | +0.138 |
pite | +0.117 |
usage | +0.113 |
Hurricane | +0.113 |
ritical | +0.112 |
Neg logit contributions | |
atography | -0.133 |
IDA | -0.125 |
Ru | -0.123 |
kaya | -0.122 |
irable | -0.121 |
Token |
|
Token ID | 198 |
Feature activation | -1.19e+00 |
Pos logit contributions | |
weather | +2.794 |
pite | +2.297 |
usage | +2.231 |
Hurricane | +2.201 |
ritical | +2.076 |
Neg logit contributions | |
atography | -2.897 |
IDA | -2.690 |
vana | -2.657 |
irable | -2.650 |
enrich | -2.648 |
Token | All |
Token ID | 3237 |
Feature activation | -9.25e-01 |
Pos logit contributions | |
weather | +0.988 |
pite | +0.833 |
usage | +0.819 |
ritical | +0.793 |
Hurricane | +0.763 |
Neg logit contributions | |
atography | -0.769 |
wrongdoing | -0.718 |
kaya | -0.699 |
IDA | -0.699 |
irable | -0.698 |
Token | seasons |
Token ID | 7028 |
Feature activation | -4.54e+00 |
Pos logit contributions | |
weather | +0.655 |
pite | +0.534 |
usage | +0.514 |
Hurricane | +0.511 |
ritical | +0.509 |
Neg logit contributions | |
atography | -0.699 |
IDA | -0.665 |
Ru | -0.658 |
kaya | -0.645 |
irable | -0.640 |
Token | are |
Token ID | 389 |
Feature activation | -3.11e+00 |
Pos logit contributions | |
weather | +3.196 |
pite | +2.731 |
Hurricane | +2.573 |
ritical | +2.513 |
usage | +2.511 |
Neg logit contributions | |
atography | -3.365 |
IDA | -3.145 |
Ru | -3.122 |
irable | -2.985 |
wrongdoing | -2.970 |
Token | available |
Token ID | 1695 |
Feature activation | -4.96e+00 |
Pos logit contributions | |
weather | +2.398 |
pite | +2.021 |
Hurricane | +1.941 |
usage | +1.919 |
timed | +1.911 |
Neg logit contributions | |
atography | -2.082 |
IDA | -2.011 |
Ru | -1.946 |
kaya | -1.923 |
irable | -1.908 |
Token | on |
Token ID | 319 |
Feature activation | -6.79e+00 |
Pos logit contributions | |
weather | +3.405 |
pite | +2.753 |
Hurricane | +2.667 |
usage | +2.598 |
timed | +2.578 |
Neg logit contributions | |
atography | -3.757 |
IDA | -3.582 |
Ru | -3.529 |
irable | -3.454 |
kaya | -3.447 |
Token | CBS |
Token ID | 11133 |
Feature activation | +1.81e+00 |
Pos logit contributions | |
weather | +4.343 |
Hurricane | +3.428 |
pite | +3.418 |
usage | +3.321 |
timed | +3.233 |
Neg logit contributions | |
atography | -5.368 |
IDA | -5.142 |
kaya | -5.026 |
Ru | -5.003 |
irable | -4.956 |
Token | All |
Token ID | 1439 |
Feature activation | -3.53e+00 |
Pos logit contributions | |
atography | +1.415 |
IDA | +1.355 |
Ru | +1.324 |
kaya | +1.321 |
irable | +1.312 |
Neg logit contributions | |
weather | -1.233 |
pite | -1.019 |
usage | -0.976 |
ritical | -0.975 |
Hurricane | -0.955 |
Token | Access |
Token ID | 8798 |
Feature activation | -7.04e+00 |
Pos logit contributions | |
weather | +2.531 |
pite | +2.105 |
usage | +2.022 |
ritical | +2.003 |
Hurricane | +1.983 |
Neg logit contributions | |
atography | -2.561 |
IDA | -2.488 |
Ru | -2.433 |
kaya | -2.402 |
wrongdoing | -2.395 |
Token | know |
Token ID | 760 |
Feature activation | -1.39e+00 |
Pos logit contributions | |
atography | +1.625 |
IDA | +1.575 |
Ru | +1.553 |
wrongdoing | +1.516 |
kaya | +1.515 |
Neg logit contributions | |
weather | -1.564 |
pite | -1.323 |
ritical | -1.252 |
usage | -1.250 |
Hurricane | -1.223 |
Token | what |
Token ID | 644 |
Feature activation | -1.89e+00 |
Pos logit contributions | |
weather | +0.917 |
pite | +0.750 |
Hurricane | +0.742 |
usage | +0.709 |
Pend | +0.698 |
Neg logit contributions | |
atography | -1.094 |
IDA | -1.036 |
irable | -1.031 |
kaya | -1.021 |
Ru | -1.021 |
Token | caused |
Token ID | 4073 |
Feature activation | +8.51e-01 |
Pos logit contributions | |
weather | +1.365 |
Hurricane | +1.163 |
pite | +1.163 |
usage | +1.104 |
ritical | +1.073 |
Neg logit contributions | |
atography | -1.358 |
IDA | -1.284 |
kaya | -1.277 |
irable | -1.274 |
Ru | -1.240 |
Token | the |
Token ID | 262 |
Feature activation | -1.66e+00 |
Pos logit contributions | |
atography | +0.653 |
IDA | +0.625 |
kaya | +0.611 |
Ru | +0.610 |
irable | +0.604 |
Neg logit contributions | |
weather | -0.587 |
pite | -0.486 |
Hurricane | -0.468 |
usage | -0.465 |
ritical | -0.462 |
Token | n |
Token ID | 299 |
Feature activation | -7.68e+00 |
Pos logit contributions | |
weather | +1.128 |
pite | +0.927 |
Hurricane | +0.925 |
usage | +0.890 |
ritical | +0.876 |
Neg logit contributions | |
atography | -1.284 |
IDA | -1.245 |
Ru | -1.206 |
kaya | -1.201 |
irable | -1.191 |
Token | osed |
Token ID | 1335 |
Feature activation | -5.56e+00 |
Pos logit contributions | |
weather | +5.302 |
pite | +4.379 |
usage | +4.288 |
ukes | +4.256 |
ritical | +4.226 |
Neg logit contributions | |
atography | -5.458 |
IDA | -5.299 |
wrongdoing | -5.202 |
kaya | -5.187 |
Transparency | -5.121 |
Token | ives |
Token ID | 1083 |
Feature activation | -2.27e-01 |
Pos logit contributions | |
weather | +3.710 |
pite | +3.033 |
Hurricane | +2.958 |
usage | +2.944 |
ritical | +2.849 |
Neg logit contributions | |
atography | -4.275 |
IDA | -3.999 |
Ru | -3.970 |
Transparency | -3.898 |
vana | -3.865 |
Token | , |
Token ID | 11 |
Feature activation | -4.22e+00 |
Pos logit contributions | |
weather | +0.167 |
pite | +0.141 |
Hurricane | +0.136 |
usage | +0.134 |
ritical | +0.132 |
Neg logit contributions | |
atography | -0.163 |
IDA | -0.157 |
Ru | -0.154 |
kaya | -0.150 |
irable | -0.149 |
Token | but |
Token ID | 475 |
Feature activation | -5.61e+00 |
Pos logit contributions | |
weather | +3.180 |
pite | +2.673 |
Hurricane | +2.652 |
usage | +2.561 |
ritical | +2.558 |
Neg logit contributions | |
atography | -2.862 |
IDA | -2.773 |
irable | -2.773 |
kaya | -2.684 |
Ru | -2.670 |
Token | he |
Token ID | 339 |
Feature activation | -3.67e+00 |
Pos logit contributions | |
weather | +3.742 |
Hurricane | +3.115 |
pite | +3.062 |
Pend | +2.885 |
ritical | +2.868 |
Neg logit contributions | |
atography | -4.255 |
irable | -4.079 |
IDA | -4.056 |
kaya | -3.984 |
Ru | -3.962 |
Token | and |
Token ID | 290 |
Feature activation | -1.95e+00 |
Pos logit contributions | |
weather | +2.412 |
pite | +1.924 |
Hurricane | +1.860 |
ritical | +1.836 |
timed | +1.803 |
Neg logit contributions | |
atography | -2.837 |
IDA | -2.837 |
Ru | -2.768 |
irable | -2.737 |
kaya | -2.678 |
Token | . |
Token ID | 13 |
Feature activation | +4.13e+00 |
Pos logit contributions | |
atography | +2.582 |
IDA | +2.502 |
Ru | +2.455 |
kaya | +2.404 |
wrongdoing | +2.400 |
Neg logit contributions | |
weather | -2.749 |
pite | -2.348 |
ritical | -2.240 |
usage | -2.225 |
Hurricane | -2.182 |
Token | Very |
Token ID | 9576 |
Feature activation | -2.46e+00 |
Pos logit contributions | |
atography | +3.000 |
IDA | +2.889 |
Ru | +2.847 |
wrongdoing | +2.808 |
kaya | +2.799 |
Neg logit contributions | |
weather | -2.967 |
pite | -2.515 |
ritical | -2.404 |
usage | -2.392 |
Hurricane | -2.331 |
Token | lonely |
Token ID | 21757 |
Feature activation | -1.01e+00 |
Pos logit contributions | |
weather | +1.776 |
pite | +1.559 |
Hurricane | +1.498 |
ritical | +1.489 |
usage | +1.452 |
Neg logit contributions | |
atography | -1.708 |
IDA | -1.654 |
kaya | -1.600 |
irable | -1.598 |
Ru | -1.595 |
Token | , |
Token ID | 11 |
Feature activation | -1.91e-01 |
Pos logit contributions | |
weather | +0.722 |
pite | +0.611 |
Hurricane | +0.581 |
ritical | +0.580 |
usage | +0.573 |
Neg logit contributions | |
atography | -0.731 |
IDA | -0.715 |
Ru | -0.701 |
irable | -0.689 |
kaya | -0.687 |
Token | sick |
Token ID | 6639 |
Feature activation | -6.61e-01 |
Pos logit contributions | |
weather | +0.138 |
pite | +0.117 |
Hurricane | +0.113 |
ritical | +0.111 |
usage | +0.110 |
Neg logit contributions | |
atography | -0.138 |
IDA | -0.134 |
Ru | -0.130 |
irable | -0.130 |
kaya | -0.129 |
Token | ".[ |
Token ID | 42924 |
Feature activation | -4.04e+00 |
Pos logit contributions | |
weather | +0.472 |
pite | +0.396 |
ritical | +0.380 |
usage | +0.376 |
Hurricane | +0.370 |
Neg logit contributions | |
atography | -0.479 |
IDA | -0.470 |
Ru | -0.463 |
kaya | -0.452 |
irable | -0.448 |
Token | 38 |
Token ID | 2548 |
Feature activation | -7.63e-01 |
Pos logit contributions | |
weather | +2.806 |
pite | +2.262 |
usage | +2.201 |
ritical | +2.113 |
Hurricane | +2.068 |
Neg logit contributions | |
atography | -3.096 |
wrongdoing | -2.896 |
IDA | -2.884 |
Ru | -2.856 |
enrich | -2.853 |
Token | ] |
Token ID | 60 |
Feature activation | -9.12e-01 |
Pos logit contributions | |
weather | +0.736 |
pite | +0.651 |
ritical | +0.632 |
usage | +0.627 |
Hurricane | +0.617 |
Neg logit contributions | |
atography | -0.365 |
IDA | -0.347 |
Ru | -0.338 |
wrongdoing | -0.330 |
kaya | -0.329 |
Token | While |
Token ID | 2893 |
Feature activation | -1.49e-01 |
Pos logit contributions | |
weather | +0.672 |
pite | +0.576 |
Hurricane | +0.554 |
ritical | +0.545 |
usage | +0.543 |
Neg logit contributions | |
atography | -0.637 |
IDA | -0.610 |
kaya | -0.598 |
irable | -0.597 |
wrongdoing | -0.597 |
Token | Cob |
Token ID | 14828 |
Feature activation | +6.93e+00 |
Pos logit contributions | |
weather | +0.107 |
pite | +0.092 |
ritical | +0.088 |
usage | +0.087 |
Hurricane | +0.086 |
Neg logit contributions | |
atography | -0.108 |
IDA | -0.105 |
kaya | -0.103 |
Ru | -0.102 |
irable | -0.102 |
Token | ain |
Token ID | 391 |
Feature activation | +5.34e+00 |
Pos logit contributions | |
atography | +2.995 |
Ru | +2.845 |
IDA | +2.807 |
kaya | +2.732 |
vana | +2.694 |
Neg logit contributions | |
weather | -6.769 |
pite | -6.013 |
ritical | -5.894 |
usage | -5.854 |
Hurricane | -5.766 |
Token | letters |
Token ID | 7475 |
Feature activation | -5.26e+00 |
Pos logit contributions | |
weather | +0.296 |
pite | +0.255 |
ritical | +0.241 |
usage | +0.238 |
Hurricane | +0.236 |
Neg logit contributions | |
atography | -0.276 |
IDA | -0.267 |
Ru | -0.260 |
kaya | -0.258 |
irable | -0.256 |
Token | from |
Token ID | 422 |
Feature activation | +1.48e+00 |
Pos logit contributions | |
weather | +3.747 |
pite | +3.153 |
ritical | +2.962 |
usage | +2.923 |
Hurricane | +2.922 |
Neg logit contributions | |
atography | -3.760 |
IDA | -3.638 |
Ru | -3.543 |
vana | -3.475 |
irable | -3.459 |
Token | people |
Token ID | 661 |
Feature activation | +8.49e-01 |
Pos logit contributions | |
atography | +1.073 |
IDA | +1.030 |
Ru | +1.011 |
kaya | +1.001 |
wrongdoing | +0.981 |
Neg logit contributions | |
weather | -1.078 |
pite | -0.918 |
ritical | -0.870 |
usage | -0.859 |
Hurricane | -0.850 |
Token | who |
Token ID | 508 |
Feature activation | +6.42e+00 |
Pos logit contributions | |
atography | +0.593 |
IDA | +0.568 |
Ru | +0.560 |
kaya | +0.545 |
irable | +0.541 |
Neg logit contributions | |
weather | -0.645 |
pite | -0.551 |
ritical | -0.523 |
usage | -0.518 |
Hurricane | -0.513 |
Token | � |
Token ID | 447 |
Feature activation | +3.36e+00 |
Pos logit contributions | |
atography | +4.163 |
IDA | +3.980 |
Ru | +3.960 |
wrongdoing | +3.920 |
kaya | +3.893 |
Neg logit contributions | |
weather | -4.984 |
pite | -4.297 |
ritical | -4.175 |
usage | -4.147 |
Hurricane | -3.977 |
Token | � |
Token ID | 247 |
Feature activation | -5.58e-01 |
Pos logit contributions | |
atography | +1.952 |
IDA | +1.866 |
Ru | +1.836 |
wrongdoing | +1.799 |
kaya | +1.792 |
Neg logit contributions | |
weather | -2.904 |
pite | -2.537 |
ritical | -2.449 |
usage | -2.436 |
Hurricane | -2.375 |
Token | ve |
Token ID | 303 |
Feature activation | +2.52e+00 |
Pos logit contributions | |
weather | +0.443 |
pite | +0.382 |
usage | +0.364 |
ritical | +0.359 |
Hurricane | +0.349 |
Neg logit contributions | |
atography | -0.374 |
IDA | -0.357 |
Ru | -0.347 |
wrongdoing | -0.345 |
Transparency | -0.334 |
Token | known |
Token ID | 1900 |
Feature activation | +3.54e+00 |
Pos logit contributions | |
atography | +1.726 |
IDA | +1.661 |
Ru | +1.627 |
kaya | +1.594 |
wrongdoing | +1.591 |
Neg logit contributions | |
weather | -1.932 |
pite | -1.650 |
ritical | -1.569 |
usage | -1.561 |
Hurricane | -1.535 |
Token | you |
Token ID | 345 |
Feature activation | +6.94e-01 |
Pos logit contributions | |
atography | +2.271 |
IDA | +2.165 |
Ru | +2.120 |
kaya | +2.076 |
wrongdoing | +2.065 |
Neg logit contributions | |
weather | -2.886 |
pite | -2.486 |
ritical | -2.393 |
usage | -2.376 |
Hurricane | -2.334 |
Token | , |
Token ID | 11 |
Feature activation | +1.30e+00 |
Pos logit contributions | |
atography | +0.479 |
IDA | +0.462 |
Ru | +0.449 |
wrongdoing | +0.440 |
irable | +0.437 |
Neg logit contributions | |
weather | -0.537 |
pite | -0.454 |
usage | -0.432 |
ritical | -0.429 |
Hurricane | -0.423 |
Token | you |
Token ID | 345 |
Feature activation | -4.95e-01 |
Pos logit contributions | |
atography | +0.908 |
IDA | +0.876 |
Ru | +0.856 |
wrongdoing | +0.839 |
kaya | +0.839 |
Neg logit contributions | |
weather | -0.988 |
pite | -0.840 |
ritical | -0.802 |
usage | -0.796 |
Hurricane | -0.778 |
Token | of |
Token ID | 286 |
Feature activation | -8.00e-01 |
Pos logit contributions | |
atography | +3.745 |
IDA | +3.596 |
Ru | +3.535 |
wrongdoing | +3.473 |
kaya | +3.456 |
Neg logit contributions | |
weather | -5.339 |
pite | -4.663 |
ritical | -4.494 |
usage | -4.465 |
Hurricane | -4.336 |
Token | me |
Token ID | 502 |
Feature activation | +1.77e+00 |
Pos logit contributions | |
weather | +0.806 |
pite | +0.709 |
Hurricane | +0.697 |
usage | +0.696 |
ritical | +0.693 |
Neg logit contributions | |
atography | -0.351 |
IDA | -0.325 |
Ru | -0.309 |
kaya | -0.309 |
irable | -0.297 |
Token | . |
Token ID | 13 |
Feature activation | +4.98e+00 |
Pos logit contributions | |
atography | +1.210 |
IDA | +1.150 |
Ru | +1.146 |
wrongdoing | +1.118 |
irable | +1.113 |
Neg logit contributions | |
weather | -1.362 |
pite | -1.159 |
usage | -1.111 |
ritical | -1.107 |
Hurricane | -1.093 |
Token | I |
Token ID | 314 |
Feature activation | +3.56e+00 |
Pos logit contributions | |
atography | +3.082 |
IDA | +2.952 |
Ru | +2.906 |
wrongdoing | +2.854 |
kaya | +2.843 |
Neg logit contributions | |
weather | -4.111 |
pite | -3.568 |
ritical | -3.436 |
usage | -3.418 |
Hurricane | -3.330 |
Token | am |
Token ID | 716 |
Feature activation | -1.23e-01 |
Pos logit contributions | |
atography | +2.931 |
IDA | +2.837 |
Ru | +2.801 |
wrongdoing | +2.765 |
kaya | +2.750 |
Neg logit contributions | |
weather | -2.223 |
pite | -1.826 |
usage | -1.724 |
ritical | -1.722 |
Hurricane | -1.671 |
Token | ready |
Token ID | 3492 |
Feature activation | -6.23e-01 |
Pos logit contributions | |
weather | +0.081 |
pite | +0.068 |
usage | +0.065 |
Hurricane | +0.064 |
ritical | +0.063 |
Neg logit contributions | |
atography | -0.096 |
IDA | -0.092 |
Ru | -0.091 |
kaya | -0.090 |
wrongdoing | -0.089 |
Token | . |
Token ID | 13 |
Feature activation | +6.87e+00 |
Pos logit contributions | |
weather | +0.466 |
pite | +0.395 |
ritical | +0.374 |
usage | +0.372 |
Hurricane | +0.371 |
Neg logit contributions | |
atography | -0.436 |
IDA | -0.417 |
Ru | -0.416 |
irable | -0.406 |
wrongdoing | -0.404 |
Token | � |
Token ID | 564 |
Feature activation | +1.59e+00 |
Pos logit contributions | |
atography | +4.254 |
IDA | +4.119 |
Ru | +4.040 |
wrongdoing | +3.961 |
kaya | +3.948 |
Neg logit contributions | |
weather | -5.582 |
pite | -4.853 |
ritical | -4.687 |
usage | -4.632 |
Hurricane | -4.463 |
Token | � |
Token ID | 251 |
Feature activation | +1.23e+00 |
Pos logit contributions | |
atography | +1.039 |
IDA | +0.976 |
Ru | +0.958 |
wrongdoing | +0.951 |
kaya | +0.942 |
Neg logit contributions | |
weather | -1.261 |
pite | -1.098 |
usage | -1.049 |
ritical | -1.043 |
Hurricane | -1.025 |
Token | Then |
Token ID | 3244 |
Feature activation | -3.53e+00 |
Pos logit contributions | |
atography | +0.805 |
IDA | +0.756 |
wrongdoing | +0.743 |
Ru | +0.742 |
kaya | +0.741 |
Neg logit contributions | |
weather | -0.982 |
pite | -0.845 |
usage | -0.812 |
ritical | -0.801 |
Hurricane | -0.799 |
Token | , |
Token ID | 11 |
Feature activation | +1.20e+00 |
Pos logit contributions | |
weather | +2.561 |
pite | +2.226 |
Hurricane | +2.159 |
usage | +2.102 |
ritical | +2.072 |
Neg logit contributions | |
atography | -2.426 |
irable | -2.323 |
IDA | -2.302 |
kaya | -2.300 |
arthed | -2.275 |
Token | from |
Token ID | 422 |
Feature activation | +4.19e+00 |
Pos logit contributions | |
weather | +1.852 |
pite | +1.509 |
ritical | +1.465 |
Hurricane | +1.445 |
usage | +1.414 |
Neg logit contributions | |
atography | -2.175 |
IDA | -2.112 |
Ru | -2.086 |
kaya | -2.020 |
irable | -2.020 |
Token | complaining |
Token ID | 18705 |
Feature activation | +1.66e+00 |
Pos logit contributions | |
atography | +2.871 |
IDA | +2.762 |
Ru | +2.724 |
wrongdoing | +2.684 |
kaya | +2.672 |
Neg logit contributions | |
weather | -3.176 |
pite | -2.720 |
ritical | -2.610 |
usage | -2.594 |
Hurricane | -2.517 |
Token | to |
Token ID | 284 |
Feature activation | +4.54e+00 |
Pos logit contributions | |
atography | +1.197 |
IDA | +1.146 |
Ru | +1.123 |
kaya | +1.111 |
irable | +1.102 |
Neg logit contributions | |
weather | -1.219 |
pite | -1.025 |
ritical | -0.985 |
Hurricane | -0.978 |
usage | -0.977 |
Token | Congress |
Token ID | 3162 |
Feature activation | -2.70e+00 |
Pos logit contributions | |
atography | +3.004 |
IDA | +2.886 |
Ru | +2.845 |
wrongdoing | +2.800 |
kaya | +2.787 |
Neg logit contributions | |
weather | -3.543 |
pite | -3.049 |
ritical | -2.929 |
usage | -2.913 |
Hurricane | -2.829 |
Token | about |
Token ID | 546 |
Feature activation | -3.89e+00 |
Pos logit contributions | |
weather | +1.828 |
pite | +1.526 |
ritical | +1.482 |
Hurricane | +1.463 |
usage | +1.430 |
Neg logit contributions | |
atography | -2.057 |
IDA | -1.983 |
Ru | -1.926 |
irable | -1.908 |
kaya | -1.884 |
Token | poor |
Token ID | 3595 |
Feature activation | -1.24e+01 |
Pos logit contributions | |
weather | +2.557 |
pite | +2.197 |
Hurricane | +2.188 |
usage | +2.072 |
ritical | +2.026 |
Neg logit contributions | |
atography | -2.970 |
IDA | -2.839 |
kaya | -2.786 |
Ru | -2.778 |
vana | -2.739 |
Token | treatment |
Token ID | 3513 |
Feature activation | -7.94e+00 |
Pos logit contributions | |
weather | +8.826 |
Hurricane | +7.729 |
pite | +7.425 |
usage | +7.233 |
bearer | +7.226 |
Neg logit contributions | |
atography | -7.565 |
Ru | -7.510 |
IDA | -7.430 |
irable | -7.274 |
kaya | -7.226 |
Token | , |
Token ID | 11 |
Feature activation | -2.97e+00 |
Pos logit contributions | |
weather | +5.202 |
pite | +4.403 |
Hurricane | +4.344 |
usage | +4.205 |
ritical | +4.117 |
Neg logit contributions | |
atography | -5.828 |
IDA | -5.684 |
Ru | -5.612 |
kaya | -5.539 |
vana | -5.502 |
Token | according |
Token ID | 1864 |
Feature activation | +1.45e+00 |
Pos logit contributions | |
weather | +1.984 |
pite | +1.666 |
Hurricane | +1.638 |
usage | +1.573 |
ritical | +1.571 |
Neg logit contributions | |
atography | -2.245 |
IDA | -2.154 |
kaya | -2.118 |
Ru | -2.105 |
irable | -2.101 |
Token | to |
Token ID | 284 |
Feature activation | -7.88e-01 |
Pos logit contributions | |
atography | +0.942 |
IDA | +0.898 |
Ru | +0.881 |
kaya | +0.868 |
wrongdoing | +0.866 |
Neg logit contributions | |
weather | -1.164 |
pite | -1.002 |
ritical | -0.958 |
usage | -0.958 |
Hurricane | -0.939 |
Token | one |
Token ID | 530 |
Feature activation | +4.92e-01 |
Pos logit contributions | |
weather | +0.543 |
pite | +0.462 |
Hurricane | +0.441 |
ritical | +0.437 |
usage | +0.435 |
Neg logit contributions | |
atography | -0.590 |
IDA | -0.571 |
Ru | -0.559 |
kaya | -0.557 |
vana | -0.546 |
Token | that |
Token ID | 326 |
Feature activation | -4.87e+00 |
Pos logit contributions | |
atography | +0.227 |
IDA | +0.215 |
Ru | +0.211 |
kaya | +0.208 |
irable | +0.207 |
Neg logit contributions | |
weather | -0.296 |
pite | -0.255 |
ritical | -0.246 |
usage | -0.243 |
Hurricane | -0.243 |
Token | Trump |
Token ID | 1301 |
Feature activation | -7.92e+00 |
Pos logit contributions | |
weather | +3.136 |
Hurricane | +2.628 |
pite | +2.585 |
ritical | +2.454 |
usage | +2.441 |
Neg logit contributions | |
atography | -3.914 |
IDA | -3.608 |
kaya | -3.601 |
Ru | -3.585 |
irable | -3.545 |
Token | 's |
Token ID | 338 |
Feature activation | -3.09e+00 |
Pos logit contributions | |
weather | +5.000 |
Hurricane | +3.920 |
pite | +3.781 |
usage | +3.569 |
Pend | +3.525 |
Neg logit contributions | |
atography | -6.323 |
IDA | -6.024 |
irable | -6.019 |
Ru | -5.901 |
kaya | -5.822 |
Token | executive |
Token ID | 4640 |
Feature activation | -5.60e-01 |
Pos logit contributions | |
weather | +2.125 |
pite | +1.791 |
Hurricane | +1.783 |
usage | +1.741 |
ritical | +1.712 |
Neg logit contributions | |
atography | -2.324 |
IDA | -2.209 |
Ru | -2.192 |
kaya | -2.145 |
irable | -2.122 |
Token | order |
Token ID | 1502 |
Feature activation | -9.29e+00 |
Pos logit contributions | |
weather | +0.396 |
pite | +0.333 |
usage | +0.318 |
ritical | +0.317 |
Hurricane | +0.315 |
Neg logit contributions | |
atography | -0.414 |
IDA | -0.400 |
Ru | -0.395 |
kaya | -0.387 |
irable | -0.385 |
Token | was |
Token ID | 373 |
Feature activation | -1.46e+01 |
Pos logit contributions | |
weather | +5.373 |
Hurricane | +4.480 |
pite | +4.381 |
timed | +4.241 |
sled | +4.111 |
Neg logit contributions | |
atography | -7.360 |
IDA | -7.274 |
arthed | -6.980 |
irable | -6.931 |
Ru | -6.905 |
Token | not |
Token ID | 407 |
Feature activation | -8.89e+00 |
Pos logit contributions | |
weather | +8.113 |
timed | +7.530 |
misplaced | +7.418 |
Hurricane | +7.264 |
pite | +7.046 |
Neg logit contributions | |
atography | -10.707 |
IDA | -9.921 |
irable | -9.907 |
kaya | -9.837 |
Ru | -9.799 |
Token | in |
Token ID | 287 |
Feature activation | -9.83e+00 |
Pos logit contributions | |
weather | +5.349 |
Hurricane | +4.683 |
pite | +4.600 |
timed | +4.528 |
misplaced | +4.382 |
Neg logit contributions | |
atography | -6.965 |
IDA | -6.387 |
Ru | -6.370 |
kaya | -6.340 |
vana | -6.284 |
Token | fact |
Token ID | 1109 |
Feature activation | -6.74e+00 |
Pos logit contributions | |
weather | +6.632 |
pite | +5.664 |
Hurricane | +5.591 |
usage | +5.418 |
timed | +5.202 |
Neg logit contributions | |
atography | -6.834 |
IDA | -6.611 |
Ru | -6.484 |
Submit | -6.238 |
Marina | -6.195 |
Token | a |
Token ID | 257 |
Feature activation | -3.76e+00 |
Pos logit contributions | |
weather | +4.119 |
Hurricane | +3.497 |
pite | +3.488 |
timed | +3.272 |
usage | +3.224 |
Neg logit contributions | |
atography | -5.419 |
IDA | -5.052 |
kaya | -5.012 |
Ru | -4.954 |
vana | -4.935 |
Token | ban |
Token ID | 3958 |
Feature activation | -6.79e+00 |
Pos logit contributions | |
weather | +2.396 |
pite | +2.032 |
Hurricane | +1.996 |
ritical | +1.928 |
usage | +1.922 |
Neg logit contributions | |
atography | -3.008 |
Ru | -2.827 |
kaya | -2.791 |
IDA | -2.763 |
vana | -2.704 |
Token | () |
Token ID | 3419 |
Feature activation | -6.61e+00 |
Pos logit contributions | |
weather | +3.696 |
pite | +2.466 |
usage | +2.415 |
Hurricane | +2.211 |
timed | +2.052 |
Neg logit contributions | |
atography | -10.057 |
enrich | -9.885 |
�� | -9.661 |
vana | -9.580 |
irable | -9.516 |
Token | ap |
Token ID | 2471 |
Feature activation | -7.71e-01 |
Pos logit contributions | |
weather | +3.041 |
usage | +2.317 |
pite | +2.292 |
timed | +2.109 |
Hurricane | +2.103 |
Neg logit contributions | |
IDA | -6.244 |
atography | -6.202 |
Ru | -6.136 |
irable | -6.020 |
kaya | -5.902 |
Token | . |
Token ID | 13 |
Feature activation | -3.40e+00 |
Pos logit contributions | |
weather | +0.563 |
pite | +0.474 |
usage | +0.455 |
ritical | +0.447 |
Hurricane | +0.442 |
Neg logit contributions | |
atography | -0.558 |
IDA | -0.529 |
wrongdoing | -0.516 |
Ru | -0.516 |
irable | -0.510 |
Token | add |
Token ID | 2860 |
Feature activation | +4.15e-01 |
Pos logit contributions | |
weather | +2.490 |
usage | +2.066 |
pite | +2.007 |
ritical | +1.872 |
avers | +1.746 |
Neg logit contributions | |
atography | -2.497 |
wrongdoing | -2.435 |
IDA | -2.387 |
Marina | -2.370 |
Transparency | -2.325 |
Token | _ |
Token ID | 62 |
Feature activation | -4.06e+00 |
Pos logit contributions | |
atography | +0.327 |
IDA | +0.311 |
Ru | +0.306 |
wrongdoing | +0.305 |
kaya | +0.302 |
Neg logit contributions | |
weather | -0.276 |
pite | -0.230 |
usage | -0.220 |
ritical | -0.218 |
Hurricane | -0.210 |
Token | argument |
Token ID | 49140 |
Feature activation | -1.17e+01 |
Pos logit contributions | |
weather | +2.407 |
usage | +1.945 |
pite | +1.864 |
ritical | +1.752 |
Hurricane | +1.597 |
Neg logit contributions | |
atography | -3.566 |
IDA | -3.386 |
wrongdoing | -3.360 |
Transparency | -3.315 |
Ru | -3.302 |
Token | (" |
Token ID | 7203 |
Feature activation | +7.86e-01 |
Pos logit contributions | |
weather | +3.521 |
usage | +2.331 |
ritical | +2.009 |
pite | +1.773 |
Pend | +1.676 |
Neg logit contributions | |
atography | -12.279 |
enrich | -11.905 |
wrongdoing | -11.814 |
vana | -11.764 |
irable | -11.662 |
Token | - |
Token ID | 12 |
Feature activation | +3.54e+00 |
Pos logit contributions | |
atography | +0.634 |
IDA | +0.599 |
Ru | +0.589 |
wrongdoing | +0.586 |
irable | +0.581 |
Neg logit contributions | |
weather | -0.523 |
pite | -0.425 |
usage | -0.415 |
ritical | -0.402 |
Hurricane | -0.386 |
Token | i |
Token ID | 72 |
Feature activation | -2.79e+00 |
Pos logit contributions | |
atography | +1.308 |
IDA | +1.200 |
Ru | +1.168 |
wrongdoing | +1.147 |
kaya | +1.121 |
Neg logit contributions | |
weather | -3.831 |
pite | -3.438 |
usage | -3.342 |
ritical | -3.335 |
Hurricane | -3.260 |
Token | ", |
Token ID | 1600 |
Feature activation | -1.52e+00 |
Pos logit contributions | |
weather | +2.427 |
pite | +2.072 |
usage | +2.022 |
ritical | +1.967 |
Hurricane | +1.914 |
Neg logit contributions | |
atography | -1.675 |
IDA | -1.563 |
wrongdoing | -1.506 |
enrich | -1.502 |
Ru | -1.494 |
Token | " |
Token ID | 366 |
Feature activation | -2.49e+00 |
Pos logit contributions | |
weather | +1.004 |
pite | +0.825 |
usage | +0.806 |
ritical | +0.789 |
Hurricane | +0.782 |
Neg logit contributions | |
atography | -1.200 |
IDA | -1.149 |
Ru | -1.130 |
irable | -1.109 |
vana | -1.104 |
Token | this |
Token ID | 428 |
Feature activation | +2.75e+00 |
Pos logit contributions | |
atography | +0.526 |
IDA | +0.510 |
Ru | +0.494 |
kaya | +0.492 |
irable | +0.491 |
Neg logit contributions | |
weather | -0.557 |
pite | -0.478 |
ritical | -0.454 |
usage | -0.452 |
Hurricane | -0.447 |
Token | manoeuv |
Token ID | 42618 |
Feature activation | +6.84e+00 |
Pos logit contributions | |
atography | +1.908 |
IDA | +1.834 |
Ru | +1.794 |
kaya | +1.761 |
irable | +1.759 |
Neg logit contributions | |
weather | -2.073 |
pite | -1.785 |
ritical | -1.707 |
usage | -1.699 |
Hurricane | -1.667 |
Token | re |
Token ID | 260 |
Feature activation | +2.68e+00 |
Pos logit contributions | |
atography | +4.873 |
Ru | +4.658 |
IDA | +4.625 |
kaya | +4.559 |
irable | +4.534 |
Neg logit contributions | |
weather | -4.902 |
pite | -4.183 |
ritical | -3.978 |
usage | -3.959 |
Hurricane | -3.868 |
Token | failed |
Token ID | 4054 |
Feature activation | -1.32e+00 |
Pos logit contributions | |
atography | +1.734 |
IDA | +1.682 |
Ru | +1.636 |
wrongdoing | +1.599 |
irable | +1.590 |
Neg logit contributions | |
weather | -2.152 |
pite | -1.866 |
ritical | -1.788 |
usage | -1.764 |
Hurricane | -1.751 |
Token | and |
Token ID | 290 |
Feature activation | +3.58e-01 |
Pos logit contributions | |
weather | +0.955 |
pite | +0.805 |
Hurricane | +0.761 |
ritical | +0.754 |
usage | +0.746 |
Neg logit contributions | |
atography | -0.957 |
IDA | -0.941 |
Ru | -0.911 |
irable | -0.901 |
Transparency | -0.881 |
Token | dis |
Token ID | 595 |
Feature activation | -1.18e+01 |
Pos logit contributions | |
atography | +0.252 |
IDA | +0.242 |
Ru | +0.232 |
irable | +0.232 |
kaya | +0.232 |
Neg logit contributions | |
weather | -0.268 |
pite | -0.230 |
ritical | -0.219 |
Hurricane | -0.218 |
usage | -0.216 |
Token | organized |
Token ID | 30280 |
Feature activation | -1.71e+00 |
Pos logit contributions | |
weather | +7.731 |
pite | +7.320 |
usage | +6.188 |
ritical | +6.172 |
Hurricane | +5.552 |
Neg logit contributions | |
atography | -8.589 |
IDA | -8.362 |
Transparency | -8.154 |
wrongdoing | -8.008 |
Submit | -7.688 |
Token | Red |
Token ID | 2297 |
Feature activation | -7.11e-01 |
Pos logit contributions | |
weather | +1.161 |
pite | +0.988 |
ritical | +0.929 |
Hurricane | +0.925 |
usage | +0.913 |
Neg logit contributions | |
atography | -1.305 |
IDA | -1.253 |
kaya | -1.219 |
irable | -1.216 |
Ru | -1.214 |
Token | Army |
Token ID | 5407 |
Feature activation | -1.07e+00 |
Pos logit contributions | |
weather | +0.423 |
pite | +0.347 |
ritical | +0.328 |
usage | +0.321 |
Hurricane | +0.315 |
Neg logit contributions | |
atography | -0.604 |
IDA | -0.584 |
Ru | -0.574 |
wrongdoing | -0.571 |
kaya | -0.569 |
Token | units |
Token ID | 4991 |
Feature activation | -8.13e+00 |
Pos logit contributions | |
weather | +0.633 |
pite | +0.525 |
ritical | +0.488 |
Hurricane | +0.484 |
usage | +0.477 |
Neg logit contributions | |
atography | -0.909 |
IDA | -0.874 |
Ru | -0.860 |
irable | -0.860 |
kaya | -0.847 |
Token | were |
Token ID | 547 |
Feature activation | -6.33e+00 |
Pos logit contributions | |
weather | +5.642 |
pite | +4.891 |
Hurricane | +4.640 |
towed | +4.596 |
timed | +4.589 |
Neg logit contributions | |
atography | -5.502 |
IDA | -5.424 |
Ru | -5.284 |
irable | -5.160 |
arthed | -5.118 |
Token | having |
Token ID | 1719 |
Feature activation | -1.00e+00 |
Pos logit contributions | |
atography | +1.968 |
IDA | +1.899 |
Ru | +1.878 |
wrongdoing | +1.836 |
kaya | +1.834 |
Neg logit contributions | |
weather | -2.068 |
pite | -1.768 |
ritical | -1.689 |
usage | -1.685 |
Hurricane | -1.637 |
Token | the |
Token ID | 262 |
Feature activation | -1.90e-01 |
Pos logit contributions | |
weather | +0.691 |
pite | +0.589 |
usage | +0.558 |
ritical | +0.556 |
Hurricane | +0.544 |
Neg logit contributions | |
atography | -0.745 |
IDA | -0.724 |
Ru | -0.718 |
kaya | -0.703 |
wrongdoing | -0.697 |
Token | client |
Token ID | 5456 |
Feature activation | -4.33e+00 |
Pos logit contributions | |
weather | +0.137 |
pite | +0.118 |
usage | +0.112 |
ritical | +0.111 |
Hurricane | +0.109 |
Neg logit contributions | |
atography | -0.137 |
IDA | -0.133 |
Ru | -0.131 |
kaya | -0.127 |
irable | -0.126 |
Token | ask |
Token ID | 1265 |
Feature activation | -2.48e+00 |
Pos logit contributions | |
weather | +3.023 |
pite | +2.512 |
usage | +2.428 |
ritical | +2.394 |
Hurricane | +2.352 |
Neg logit contributions | |
atography | -3.125 |
Ru | -3.061 |
IDA | -3.033 |
arthed | -2.965 |
wrongdoing | -2.926 |
Token | the |
Token ID | 262 |
Feature activation | -4.41e+00 |
Pos logit contributions | |
weather | +1.741 |
pite | +1.480 |
usage | +1.430 |
ritical | +1.399 |
Hurricane | +1.395 |
Neg logit contributions | |
atography | -1.798 |
IDA | -1.710 |
Ru | -1.691 |
kaya | -1.681 |
irable | -1.665 |
Token | server |
Token ID | 4382 |
Feature activation | -1.17e+01 |
Pos logit contributions | |
weather | +2.332 |
pite | +1.879 |
usage | +1.785 |
Hurricane | +1.745 |
ritical | +1.722 |
Neg logit contributions | |
atography | -3.898 |
IDA | -3.814 |
Ru | -3.785 |
irable | -3.713 |
arthed | -3.662 |
Token | to |
Token ID | 284 |
Feature activation | -5.19e+00 |
Pos logit contributions | |
weather | +7.011 |
usage | +5.942 |
pite | +5.673 |
Hurricane | +5.664 |
ritical | +5.556 |
Neg logit contributions | |
atography | -8.405 |
Ru | -8.132 |
irable | -8.088 |
IDA | -7.993 |
arthed | -7.946 |
Token | do |
Token ID | 466 |
Feature activation | -1.87e+00 |
Pos logit contributions | |
weather | +3.365 |
pite | +2.863 |
usage | +2.770 |
Hurricane | +2.746 |
sled | +2.624 |
Neg logit contributions | |
atography | -3.906 |
Ru | -3.798 |
IDA | -3.793 |
arthed | -3.700 |
wrongdoing | -3.595 |
Token | work |
Token ID | 670 |
Feature activation | -3.17e+00 |
Pos logit contributions | |
weather | +1.264 |
pite | +1.074 |
usage | +1.028 |
ritical | +1.001 |
Hurricane | +0.990 |
Neg logit contributions | |
atography | -1.397 |
Ru | -1.347 |
IDA | -1.345 |
kaya | -1.317 |
arthed | -1.306 |
Token | for |
Token ID | 329 |
Feature activation | -1.39e+00 |
Pos logit contributions | |
weather | +2.246 |
pite | +1.925 |
usage | +1.855 |
ritical | +1.820 |
Hurricane | +1.756 |
Neg logit contributions | |
atography | -2.260 |
IDA | -2.206 |
Ru | -2.177 |
kaya | -2.107 |
arthed | -2.102 |
Token | him |
Token ID | 683 |
Feature activation | -3.90e-01 |
Pos logit contributions | |
weather | +0.859 |
pite | +0.714 |
usage | +0.678 |
ritical | +0.671 |
Hurricane | +0.663 |
Neg logit contributions | |
atography | -1.138 |
IDA | -1.099 |
Ru | -1.092 |
kaya | -1.074 |
wrongdoing | -1.057 |