Token | v |
Token ID | 85 |
Feature activation | +1.78e-01 |
Pos logit contributions | |
phe | +2.698 |
ū | +2.590 |
aug | +2.555 |
Ur | +2.540 |
kus | +2.399 |
Neg logit contributions | |
Howell | -4.539 |
Birch | -4.016 |
Clarkson | -4.012 |
JJ | -3.871 |
Hendricks | -3.714 |
Token | 2 |
Token ID | 17 |
Feature activation | +2.36e-01 |
Pos logit contributions | |
phe | +1.127 |
ū | +1.080 |
aug | +1.060 |
Ur | +1.056 |
kus | +1.002 |
Neg logit contributions | |
Howell | -1.810 |
Birch | -1.601 |
Clarkson | -1.596 |
JJ | -1.543 |
Hendricks | -1.476 |
Token | . |
Token ID | 13 |
Feature activation | +5.51e-01 |
Pos logit contributions | |
phe | +1.668 |
ū | +1.609 |
aug | +1.587 |
Ur | +1.586 |
UG | +1.506 |
Neg logit contributions | |
Howell | -2.218 |
Birch | -1.939 |
Clarkson | -1.934 |
JJ | -1.861 |
Hendricks | -1.774 |
Token | im |
Token ID | 320 |
Feature activation | +8.52e-02 |
Pos logit contributions | |
phe | +4.313 |
Ur | +4.268 |
ū | +4.246 |
aug | +4.213 |
kus | +4.028 |
Neg logit contributions | |
Howell | -4.658 |
Clarkson | -4.104 |
Birch | -4.020 |
JJ | -3.807 |
Hendricks | -3.674 |
Token | read |
Token ID | 961 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
phe | +0.654 |
ū | +0.632 |
aug | +0.622 |
Ur | +0.622 |
kus | +0.595 |
Neg logit contributions | |
Howell | -0.751 |
Birch | -0.650 |
Clarkson | -0.649 |
JJ | -0.623 |
Hendricks | -0.591 |
Token | ( |
Token ID | 7 |
Feature activation | +1.03e+00 |
Pos logit contributions | |
phe | +0.851 |
ū | +0.823 |
aug | +0.810 |
Ur | +0.808 |
kus | +0.776 |
Neg logit contributions | |
Howell | -0.908 |
Birch | -0.783 |
Clarkson | -0.781 |
JJ | -0.749 |
Hendricks | -0.708 |
Token | image |
Token ID | 9060 |
Feature activation | +2.54e-01 |
Pos logit contributions | |
aug | +4.220 |
phe | +4.178 |
ū | +4.091 |
Ur | +3.811 |
kus | +3.639 |
Neg logit contributions | |
Howell | -12.474 |
Clarkson | -11.484 |
Birch | -11.351 |
JJ | -10.868 |
Hendricks | -10.685 |
Token | Path |
Token ID | 15235 |
Feature activation | +3.70e-01 |
Pos logit contributions | |
phe | +1.477 |
ū | +1.410 |
Ur | +1.384 |
aug | +1.383 |
kus | +1.299 |
Neg logit contributions | |
Howell | -2.712 |
Birch | -2.412 |
Clarkson | -2.407 |
JJ | -2.330 |
Hendricks | -2.235 |
Token | ) |
Token ID | 8 |
Feature activation | +2.99e-01 |
Pos logit contributions | |
phe | +2.546 |
ū | +2.455 |
Ur | +2.431 |
aug | +2.426 |
UG | +2.305 |
Neg logit contributions | |
Howell | -3.534 |
Birch | -3.094 |
Clarkson | -3.091 |
JJ | -2.971 |
Hendricks | -2.837 |
Token | image |
Token ID | 2939 |
Feature activation | +2.52e-02 |
Pos logit contributions | |
phe | +1.812 |
ū | +1.766 |
aug | +1.743 |
Ur | +1.733 |
UG | +1.634 |
Neg logit contributions | |
Howell | -3.065 |
Clarkson | -2.730 |
Birch | -2.710 |
JJ | -2.610 |
Hendricks | -2.524 |
Token | = |
Token ID | 796 |
Feature activation | +2.02e-01 |
Pos logit contributions | |
phe | +0.268 |
ū | +0.261 |
aug | +0.258 |
Ur | +0.255 |
kus | +0.250 |
Neg logit contributions | |
Howell | -0.148 |
Birch | -0.118 |
Clarkson | -0.117 |
JJ | -0.110 |
Hendricks | -0.101 |
Token | 2 |
Token ID | 17 |
Feature activation | +1.22e-01 |
Pos logit contributions | |
Howell | +1.940 |
Birch | +1.705 |
Clarkson | +1.696 |
JJ | +1.645 |
Dalton | +1.558 |
Neg logit contributions | |
phe | -1.470 |
ū | -1.416 |
aug | -1.374 |
anamo | -1.350 |
Ur | -1.348 |
Token | . |
Token ID | 13 |
Feature activation | +6.57e-01 |
Pos logit contributions | |
phe | +0.897 |
ū | +0.864 |
aug | +0.848 |
Ur | +0.846 |
kus | +0.811 |
Neg logit contributions | |
Howell | -1.111 |
Birch | -0.969 |
Clarkson | -0.966 |
JJ | -0.930 |
Hendricks | -0.884 |
Token | M |
Token ID | 44 |
Feature activation | +6.72e-01 |
Pos logit contributions | |
phe | +3.597 |
Ur | +3.528 |
UG | +3.455 |
aug | +3.401 |
ū | +3.389 |
Neg logit contributions | |
Howell | -6.965 |
Clarkson | -6.203 |
Birch | -6.078 |
JJ | -5.861 |
Hendricks | -5.715 |
Token | OR |
Token ID | 1581 |
Feature activation | +4.79e-02 |
Pos logit contributions | |
phe | +2.510 |
Ur | +2.446 |
ū | +2.391 |
UG | +2.385 |
aug | +2.298 |
Neg logit contributions | |
Howell | -8.414 |
Birch | -7.589 |
Clarkson | -7.537 |
JJ | -7.342 |
Hendricks | -7.178 |
Token | PH |
Token ID | 11909 |
Feature activation | +1.24e-01 |
Pos logit contributions | |
phe | +0.290 |
ū | +0.282 |
aug | +0.276 |
Ur | +0.274 |
kus | +0.262 |
Neg logit contributions | |
Howell | -0.495 |
Birch | -0.440 |
Clarkson | -0.439 |
JJ | -0.424 |
Hendricks | -0.408 |
Token | _ |
Token ID | 62 |
Feature activation | +9.12e-01 |
Pos logit contributions | |
phe | +0.873 |
ū | +0.838 |
aug | +0.825 |
Ur | +0.820 |
kus | +0.784 |
Neg logit contributions | |
Howell | -1.176 |
Birch | -1.030 |
Clarkson | -1.028 |
JJ | -0.992 |
Hendricks | -0.944 |
Token | RECT |
Token ID | 23988 |
Feature activation | +6.29e-02 |
Pos logit contributions | |
UG | +4.682 |
phe | +4.606 |
ū | +4.548 |
Ur | +4.529 |
aug | +4.401 |
Neg logit contributions | |
Howell | -9.906 |
Birch | -8.700 |
Clarkson | -8.600 |
JJ | -8.382 |
Gad | -8.185 |
Token | , |
Token ID | 11 |
Feature activation | +3.91e-01 |
Pos logit contributions | |
phe | +0.443 |
ū | +0.423 |
Ur | +0.418 |
aug | +0.414 |
UG | +0.406 |
Neg logit contributions | |
Howell | -0.599 |
Birch | -0.525 |
Clarkson | -0.523 |
JJ | -0.500 |
Hendricks | -0.481 |
Token | ( |
Token ID | 357 |
Feature activation | -7.14e-02 |
Pos logit contributions | |
phe | +2.699 |
ū | +2.581 |
Ur | +2.544 |
aug | +2.533 |
UG | +2.405 |
Neg logit contributions | |
Howell | -3.686 |
Clarkson | -3.199 |
Birch | -3.149 |
JJ | -3.026 |
Hendricks | -2.954 |
Token | 21 |
Token ID | 2481 |
Feature activation | +5.11e-02 |
Pos logit contributions | |
Howell | +0.707 |
Birch | +0.622 |
Clarkson | +0.620 |
JJ | +0.598 |
Hendricks | +0.572 |
Neg logit contributions | |
phe | -0.474 |
ū | -0.455 |
aug | -0.448 |
Ur | -0.447 |
kus | -0.424 |
Token | , |
Token ID | 11 |
Feature activation | +2.26e-01 |
Pos logit contributions | |
phe | +0.367 |
ū | +0.354 |
aug | +0.348 |
Ur | +0.347 |
UG | +0.333 |
Neg logit contributions | |
Howell | -0.479 |
Clarkson | -0.417 |
Birch | -0.417 |
JJ | -0.399 |
Hendricks | -0.381 |
Token | has |
Token ID | 468 |
Feature activation | +1.30e-01 |
Pos logit contributions | |
Howell | +1.123 |
Birch | +0.967 |
Clarkson | +0.963 |
JJ | +0.923 |
Hendricks | +0.874 |
Neg logit contributions | |
phe | -1.060 |
ū | -1.027 |
aug | -1.013 |
Ur | -1.009 |
kus | -0.966 |
Token | been |
Token ID | 587 |
Feature activation | +2.21e-01 |
Pos logit contributions | |
phe | +0.896 |
ū | +0.879 |
aug | +0.864 |
Ur | +0.860 |
UG | +0.817 |
Neg logit contributions | |
Howell | -1.226 |
Birch | -1.080 |
Clarkson | -1.066 |
JJ | -1.026 |
Hendricks | -0.983 |
Token | the |
Token ID | 262 |
Feature activation | +3.51e-01 |
Pos logit contributions | |
phe | +1.477 |
ū | +1.441 |
aug | +1.425 |
Ur | +1.410 |
UG | +1.346 |
Neg logit contributions | |
Howell | -2.134 |
Clarkson | -1.872 |
Birch | -1.872 |
JJ | -1.795 |
Hendricks | -1.718 |
Token | greatest |
Token ID | 6000 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
phe | +2.459 |
ū | +2.458 |
aug | +2.398 |
Ur | +2.348 |
kus | +2.252 |
Neg logit contributions | |
Howell | -3.269 |
Birch | -2.805 |
Clarkson | -2.797 |
JJ | -2.701 |
Hendricks | -2.560 |
Token | honor |
Token ID | 7522 |
Feature activation | +4.08e-01 |
Pos logit contributions | |
phe | +0.756 |
ū | +0.735 |
aug | +0.724 |
Ur | +0.716 |
UG | +0.681 |
Neg logit contributions | |
Howell | -1.092 |
Birch | -0.963 |
Clarkson | -0.948 |
JJ | -0.919 |
Hendricks | -0.874 |
Token | of |
Token ID | 286 |
Feature activation | +9.00e-01 |
Pos logit contributions | |
phe | +2.786 |
ū | +2.707 |
aug | +2.643 |
Ur | +2.600 |
UG | +2.502 |
Neg logit contributions | |
Howell | -3.855 |
Birch | -3.401 |
Clarkson | -3.363 |
JJ | -3.267 |
Hendricks | -3.114 |
Token | my |
Token ID | 616 |
Feature activation | +6.20e-01 |
Pos logit contributions | |
phe | +5.393 |
aug | +5.299 |
ū | +5.249 |
Ur | +5.122 |
� | +4.902 |
Neg logit contributions | |
Howell | -8.631 |
Birch | -7.607 |
Clarkson | -7.550 |
JJ | -7.280 |
Hendricks | -6.918 |
Token | life |
Token ID | 1204 |
Feature activation | +5.23e-01 |
Pos logit contributions | |
phe | +3.691 |
ū | +3.609 |
aug | +3.576 |
Ur | +3.486 |
ogo | +3.311 |
Neg logit contributions | |
Howell | -6.315 |
Birch | -5.583 |
Clarkson | -5.565 |
JJ | -5.403 |
Hendricks | -5.159 |
Token | , |
Token ID | 11 |
Feature activation | +2.68e-01 |
Pos logit contributions | |
phe | +3.224 |
ū | +3.145 |
aug | +3.059 |
Ur | +3.001 |
ura | +2.867 |
Neg logit contributions | |
Howell | -5.252 |
Birch | -4.647 |
Clarkson | -4.571 |
JJ | -4.427 |
Hendricks | -4.225 |
Token | and |
Token ID | 290 |
Feature activation | +1.97e-01 |
Pos logit contributions | |
phe | +2.012 |
ū | +1.947 |
aug | +1.940 |
Ur | +1.917 |
UG | +1.827 |
Neg logit contributions | |
Howell | -2.369 |
Birch | -2.052 |
Clarkson | -2.040 |
JJ | -1.959 |
Hendricks | -1.858 |
Token | I |
Token ID | 314 |
Feature activation | +1.98e-01 |
Pos logit contributions | |
phe | +1.478 |
aug | +1.431 |
ū | +1.430 |
Ur | +1.409 |
UG | +1.344 |
Neg logit contributions | |
Howell | -1.742 |
Birch | -1.511 |
Clarkson | -1.500 |
JJ | -1.433 |
Hendricks | -1.368 |
Token | , |
Token ID | 11 |
Feature activation | -2.21e-01 |
Pos logit contributions | |
Howell | +2.728 |
Birch | +2.370 |
Clarkson | +2.363 |
JJ | +2.271 |
Hendricks | +2.157 |
Neg logit contributions | |
phe | -2.289 |
ū | -2.210 |
aug | -2.176 |
Ur | -2.174 |
kus | -2.077 |
Token | detected |
Token ID | 12326 |
Feature activation | +2.75e-01 |
Pos logit contributions | |
Howell | +1.787 |
Birch | +1.528 |
Clarkson | +1.523 |
JJ | +1.454 |
Hendricks | +1.372 |
Neg logit contributions | |
phe | -1.859 |
ū | -1.802 |
aug | -1.779 |
Ur | -1.779 |
UG | -1.706 |
Token | as |
Token ID | 355 |
Feature activation | +4.11e-01 |
Pos logit contributions | |
phe | +2.003 |
ū | +1.948 |
Ur | +1.920 |
aug | +1.913 |
UG | +1.837 |
Neg logit contributions | |
Howell | -2.493 |
Birch | -2.163 |
Clarkson | -2.161 |
JJ | -2.070 |
Hendricks | -1.971 |
Token | EL |
Token ID | 17852 |
Feature activation | +2.42e-01 |
Pos logit contributions | |
phe | +3.126 |
ū | +3.037 |
aug | +3.021 |
Ur | +3.011 |
UG | +2.938 |
Neg logit contributions | |
Howell | -3.512 |
Clarkson | -2.984 |
Birch | -2.980 |
JJ | -2.788 |
Hendricks | -2.710 |
Token | F |
Token ID | 37 |
Feature activation | +3.34e-01 |
Pos logit contributions | |
phe | +1.387 |
ū | +1.315 |
Ur | +1.306 |
aug | +1.295 |
UG | +1.239 |
Neg logit contributions | |
Howell | -2.605 |
Birch | -2.312 |
Clarkson | -2.310 |
JJ | -2.230 |
Hendricks | -2.139 |
Token | _ |
Token ID | 62 |
Feature activation | +8.62e-01 |
Pos logit contributions | |
phe | +2.144 |
ū | +2.072 |
Ur | +2.052 |
aug | +2.036 |
UG | +1.977 |
Neg logit contributions | |
Howell | -3.345 |
Birch | -2.938 |
Clarkson | -2.925 |
JJ | -2.822 |
Hendricks | -2.703 |
Token | UM |
Token ID | 5883 |
Feature activation | +5.90e-02 |
Pos logit contributions | |
phe | +4.567 |
Ur | +4.476 |
UG | +4.474 |
ū | +4.413 |
aug | +4.327 |
Neg logit contributions | |
Howell | -9.367 |
Birch | -8.303 |
Clarkson | -8.206 |
JJ | -7.971 |
Hendricks | -7.642 |
Token | RE |
Token ID | 2200 |
Feature activation | +7.51e-02 |
Pos logit contributions | |
phe | +0.290 |
Ur | +0.277 |
ū | +0.275 |
aug | +0.269 |
UG | +0.260 |
Neg logit contributions | |
Howell | -0.684 |
Birch | -0.615 |
Clarkson | -0.610 |
JJ | -0.592 |
Hendricks | -0.569 |
Token | ON |
Token ID | 1340 |
Feature activation | -2.60e-01 |
Pos logit contributions | |
phe | +0.398 |
ū | +0.380 |
Ur | +0.374 |
aug | +0.371 |
UG | +0.357 |
Neg logit contributions | |
Howell | -0.843 |
Birch | -0.754 |
Clarkson | -0.748 |
JJ | -0.728 |
Hendricks | -0.698 |
Token | . |
Token ID | 13 |
Feature activation | +3.51e-01 |
Pos logit contributions | |
Howell | +2.438 |
Birch | +2.131 |
Clarkson | +2.125 |
JJ | +2.047 |
Hendricks | +1.950 |
Neg logit contributions | |
phe | -1.846 |
ū | -1.779 |
aug | -1.751 |
Ur | -1.749 |
kus | -1.666 |
Token | B |
Token ID | 33 |
Feature activation | +1.83e-01 |
Pos logit contributions | |
phe | +1.979 |
Ur | +1.892 |
ū | +1.885 |
aug | +1.864 |
UG | +1.836 |
Neg logit contributions | |
Howell | -3.806 |
Birch | -3.377 |
Clarkson | -3.358 |
JJ | -3.228 |
Hendricks | -3.108 |
Token | – |
Token ID | 784 |
Feature activation | +4.37e-02 |
Pos logit contributions | |
phe | +0.305 |
ū | +0.295 |
Ur | +0.289 |
aug | +0.287 |
UG | +0.276 |
Neg logit contributions | |
Howell | -0.372 |
Birch | -0.322 |
Clarkson | -0.322 |
JJ | -0.309 |
Hendricks | -0.294 |
Token | whereabouts |
Token ID | 33035 |
Feature activation | -7.97e-02 |
Pos logit contributions | |
phe | +0.335 |
ū | +0.326 |
aug | +0.321 |
Ur | +0.320 |
kus | +0.306 |
Neg logit contributions | |
Howell | -0.381 |
Birch | -0.330 |
Clarkson | -0.329 |
JJ | -0.315 |
Hendricks | -0.300 |
Token | unknown |
Token ID | 6439 |
Feature activation | +1.49e-01 |
Pos logit contributions | |
Howell | +0.733 |
Birch | +0.640 |
Clarkson | +0.638 |
JJ | +0.613 |
Hendricks | +0.584 |
Neg logit contributions | |
phe | -0.581 |
ū | -0.561 |
Ur | -0.552 |
aug | -0.551 |
kus | -0.525 |
Token | ) |
Token ID | 8 |
Feature activation | +3.13e-01 |
Pos logit contributions | |
phe | +1.070 |
ū | +1.042 |
Ur | +1.019 |
aug | +1.012 |
UG | +0.973 |
Neg logit contributions | |
Howell | -1.378 |
Birch | -1.202 |
Clarkson | -1.196 |
JJ | -1.144 |
Hendricks | -1.100 |
Token | TIM |
Token ID | 31742 |
Feature activation | +2.36e-01 |
Pos logit contributions | |
phe | +2.234 |
ū | +2.191 |
Ur | +2.166 |
aug | +2.139 |
UG | +2.077 |
Neg logit contributions | |
Howell | -2.825 |
Birch | -2.467 |
Clarkson | -2.464 |
JJ | -2.325 |
Hendricks | -2.260 |
Token | EL |
Token ID | 3698 |
Feature activation | +8.32e-01 |
Pos logit contributions | |
phe | +1.547 |
Ur | +1.510 |
ū | +1.503 |
aug | +1.462 |
UG | +1.445 |
Neg logit contributions | |
Howell | -2.272 |
Birch | -2.008 |
Clarkson | -1.993 |
JJ | -1.915 |
Hendricks | -1.846 |
Token | INE |
Token ID | 8881 |
Feature activation | +5.25e-02 |
Pos logit contributions | |
UG | +3.550 |
phe | +3.473 |
Ur | +3.342 |
ū | +3.321 |
� | +3.238 |
Neg logit contributions | |
Howell | -9.672 |
Birch | -8.704 |
Clarkson | -8.616 |
JJ | -8.218 |
Hendricks | -8.161 |
Token | : |
Token ID | 25 |
Feature activation | +1.30e-01 |
Pos logit contributions | |
phe | +0.377 |
ū | +0.367 |
aug | +0.361 |
Ur | +0.361 |
UG | +0.346 |
Neg logit contributions | |
Howell | -0.484 |
Birch | -0.422 |
Clarkson | -0.421 |
JJ | -0.404 |
Hendricks | -0.386 |
Token | 04 |
Token ID | 8702 |
Feature activation | +2.33e-01 |
Pos logit contributions | |
phe | +0.970 |
ū | +0.945 |
aug | +0.932 |
Ur | +0.928 |
UG | +0.889 |
Neg logit contributions | |
Howell | -1.162 |
Birch | -1.010 |
Clarkson | -1.006 |
JJ | -0.967 |
Hendricks | -0.918 |
Token | 30 |
Token ID | 1270 |
Feature activation | +9.72e-02 |
Pos logit contributions | |
phe | +1.500 |
ū | +1.469 |
aug | +1.453 |
Ur | +1.451 |
UG | +1.390 |
Neg logit contributions | |
Howell | -2.295 |
Birch | -2.020 |
Clarkson | -2.016 |
JJ | -1.942 |
Hendricks | -1.866 |
Token | z |
Token ID | 89 |
Feature activation | -2.04e-01 |
Pos logit contributions | |
phe | +0.688 |
ū | +0.670 |
aug | +0.660 |
Ur | +0.659 |
UG | +0.631 |
Neg logit contributions | |
Howell | -0.906 |
Birch | -0.793 |
Clarkson | -0.791 |
JJ | -0.758 |
Hendricks | -0.725 |
Token | 2 |
Token ID | 362 |
Feature activation | +1.44e-01 |
Pos logit contributions | |
phe | +1.785 |
Ur | +1.733 |
ū | +1.711 |
aug | +1.709 |
UG | +1.637 |
Neg logit contributions | |
Howell | -2.639 |
Birch | -2.316 |
Clarkson | -2.299 |
JJ | -2.197 |
Hendricks | -2.141 |
Token | . |
Token ID | 13 |
Feature activation | +3.23e-01 |
Pos logit contributions | |
phe | +1.025 |
ū | +0.989 |
aug | +0.974 |
Ur | +0.972 |
kus | +0.925 |
Neg logit contributions | |
Howell | -1.346 |
Birch | -1.176 |
Clarkson | -1.172 |
JJ | -1.128 |
Hendricks | -1.076 |
Token | [ |
Token ID | 685 |
Feature activation | +6.91e-02 |
Pos logit contributions | |
phe | +1.786 |
ū | +1.693 |
Ur | +1.684 |
aug | +1.671 |
UG | +1.562 |
Neg logit contributions | |
Howell | -3.476 |
Birch | -3.089 |
Clarkson | -3.085 |
JJ | -2.950 |
Hendricks | -2.879 |
Token | color |
Token ID | 8043 |
Feature activation | +5.04e-01 |
Pos logit contributions | |
phe | +0.316 |
ū | +0.296 |
aug | +0.290 |
Ur | +0.290 |
UG | +0.266 |
Neg logit contributions | |
Howell | -0.831 |
Clarkson | -0.747 |
Birch | -0.745 |
JJ | -0.727 |
Hendricks | -0.703 |
Token | = |
Token ID | 28 |
Feature activation | +2.32e-01 |
Pos logit contributions | |
phe | +4.031 |
ū | +3.915 |
Ur | +3.880 |
aug | +3.851 |
UG | +3.690 |
Neg logit contributions | |
Howell | -4.249 |
Birch | -3.645 |
Clarkson | -3.626 |
JJ | -3.457 |
Hendricks | -3.297 |
Token | red |
Token ID | 445 |
Feature activation | +8.30e-01 |
Pos logit contributions | |
phe | +1.517 |
ū | +1.440 |
aug | +1.437 |
Ur | +1.423 |
kus | +1.349 |
Neg logit contributions | |
Howell | -2.323 |
Clarkson | -2.045 |
Birch | -2.043 |
JJ | -1.970 |
Hendricks | -1.892 |
Token | ][ |
Token ID | 7131 |
Feature activation | +4.45e-01 |
Pos logit contributions | |
phe | +1.824 |
ū | +1.596 |
aug | +1.508 |
]] | +1.480 |
Ur | +1.449 |
Neg logit contributions | |
Howell | -11.418 |
Clarkson | -10.344 |
Birch | -10.329 |
JJ | -10.101 |
Hendricks | -9.889 |
Token | b |
Token ID | 65 |
Feature activation | +3.29e-01 |
Pos logit contributions | |
phe | +2.603 |
aug | +2.453 |
ū | +2.442 |
Ur | +2.421 |
ura | +2.280 |
Neg logit contributions | |
Howell | -4.732 |
Clarkson | -4.157 |
Birch | -4.157 |
JJ | -4.050 |
Hendricks | -3.914 |
Token | ] |
Token ID | 60 |
Feature activation | +7.72e-02 |
Pos logit contributions | |
phe | +1.182 |
ū | +1.099 |
aug | +1.065 |
Ur | +1.062 |
kus | +0.953 |
Neg logit contributions | |
Howell | -4.234 |
Birch | -3.848 |
Clarkson | -3.839 |
JJ | -3.739 |
Hendricks | -3.617 |
Token | U |
Token ID | 52 |
Feature activation | +8.62e-02 |
Pos logit contributions | |
phe | +0.448 |
Ur | +0.427 |
ū | +0.424 |
aug | +0.419 |
UG | +0.397 |
Neg logit contributions | |
Howell | -0.831 |
Birch | -0.738 |
Clarkson | -0.736 |
JJ | -0.709 |
Hendricks | -0.683 |
Token | f |
Token ID | 69 |
Feature activation | +3.29e-01 |
Pos logit contributions | |
phe | +0.553 |
ū | +0.531 |
Ur | +0.522 |
aug | +0.520 |
UG | +0.500 |
Neg logit contributions | |
Howell | -0.869 |
Birch | -0.769 |
Clarkson | -0.769 |
JJ | -0.739 |
Hendricks | -0.708 |
Token | . |
Token ID | 13 |
Feature activation | -2.59e-02 |
Pos logit contributions | |
phe | +1.292 |
ū | +1.245 |
aug | +1.224 |
Ur | +1.223 |
kus | +1.167 |
Neg logit contributions | |
Howell | -1.658 |
Birch | -1.448 |
Clarkson | -1.444 |
JJ | -1.390 |
Hendricks | -1.323 |
Token | Heat |
Token ID | 12308 |
Feature activation | +3.65e-01 |
Pos logit contributions | |
Howell | +0.130 |
Clarkson | +0.100 |
Birch | +0.099 |
JJ | +0.093 |
Hendricks | +0.080 |
Neg logit contributions | |
phe | -0.292 |
ū | -0.287 |
aug | -0.286 |
Ur | -0.286 |
UG | -0.275 |
Token | about |
Token ID | 546 |
Feature activation | +3.09e-01 |
Pos logit contributions | |
phe | +3.161 |
ū | +3.095 |
aug | +3.062 |
Ur | +3.033 |
kus | +2.937 |
Neg logit contributions | |
Howell | -2.808 |
Clarkson | -2.380 |
Birch | -2.378 |
JJ | -2.244 |
Hendricks | -2.133 |
Token | an |
Token ID | 281 |
Feature activation | +3.00e-01 |
Pos logit contributions | |
phe | +2.164 |
ū | +2.105 |
aug | +2.079 |
Ur | +2.072 |
kus | +1.966 |
Neg logit contributions | |
Howell | -2.893 |
Birch | -2.532 |
Clarkson | -2.512 |
JJ | -2.399 |
Hendricks | -2.292 |
Token | inch |
Token ID | 11111 |
Feature activation | +3.27e-01 |
Pos logit contributions | |
phe | +1.981 |
ū | +1.919 |
aug | +1.889 |
Ur | +1.885 |
UG | +1.779 |
Neg logit contributions | |
Howell | -2.908 |
Clarkson | -2.575 |
Birch | -2.563 |
JJ | -2.455 |
Hendricks | -2.338 |
Token | of |
Token ID | 286 |
Feature activation | +8.16e-01 |
Pos logit contributions | |
phe | +2.433 |
ū | +2.352 |
Ur | +2.320 |
aug | +2.314 |
UG | +2.210 |
Neg logit contributions | |
Howell | -2.943 |
Birch | -2.558 |
Clarkson | -2.547 |
JJ | -2.444 |
Hendricks | -2.326 |
Token | oil |
Token ID | 3056 |
Feature activation | +1.34e-01 |
Pos logit contributions | |
phe | +4.434 |
ū | +4.298 |
aug | +4.252 |
Ur | +4.191 |
� | +4.051 |
Neg logit contributions | |
Howell | -8.396 |
Birch | -7.434 |
Clarkson | -7.356 |
JJ | -7.032 |
Hendricks | -6.811 |
Token | in |
Token ID | 287 |
Feature activation | +6.13e-01 |
Pos logit contributions | |
phe | +1.067 |
ū | +1.032 |
aug | +1.017 |
Ur | +1.016 |
kus | +0.973 |
Neg logit contributions | |
Howell | -1.148 |
Birch | -0.990 |
Clarkson | -0.987 |
JJ | -0.946 |
Hendricks | -0.896 |
Token | a |
Token ID | 257 |
Feature activation | +7.19e-01 |
Pos logit contributions | |
phe | +3.709 |
aug | +3.559 |
ū | +3.546 |
Ur | +3.494 |
kus | +3.335 |
Neg logit contributions | |
Howell | -6.202 |
Birch | -5.475 |
Clarkson | -5.435 |
JJ | -5.220 |
Hendricks | -4.979 |
Token | skillet |
Token ID | 41306 |
Feature activation | +2.23e-01 |
Pos logit contributions | |
phe | +4.079 |
ū | +3.879 |
Ur | +3.855 |
aug | +3.841 |
kus | +3.695 |
Neg logit contributions | |
Howell | -7.424 |
Birch | -6.672 |
Clarkson | -6.548 |
JJ | -6.373 |
Davies | -6.068 |
Token | over |
Token ID | 625 |
Feature activation | +6.89e-01 |
Pos logit contributions | |
phe | +1.722 |
ū | +1.663 |
aug | +1.639 |
Ur | +1.630 |
kus | +1.558 |
Neg logit contributions | |
Howell | -1.957 |
Birch | -1.696 |
Clarkson | -1.686 |
JJ | -1.618 |
Hendricks | -1.524 |
Token | v |
Token ID | 85 |
Feature activation | +1.09e-02 |
Pos logit contributions | |
phe | +0.921 |
ū | +0.885 |
aug | +0.870 |
Ur | +0.866 |
kus | +0.819 |
Neg logit contributions | |
Howell | -1.522 |
Clarkson | -1.350 |
Birch | -1.348 |
JJ | -1.300 |
Hendricks | -1.248 |
Token | 2 |
Token ID | 17 |
Feature activation | +2.56e-01 |
Pos logit contributions | |
phe | +0.071 |
ū | +0.068 |
aug | +0.066 |
Ur | +0.066 |
kus | +0.063 |
Neg logit contributions | |
Howell | -0.108 |
Birch | -0.096 |
Clarkson | -0.095 |
JJ | -0.092 |
Hendricks | -0.088 |
Token | . |
Token ID | 13 |
Feature activation | +2.50e-01 |
Pos logit contributions | |
phe | +1.807 |
ū | +1.745 |
aug | +1.718 |
Ur | +1.718 |
kus | +1.633 |
Neg logit contributions | |
Howell | -2.404 |
Birch | -2.101 |
Clarkson | -2.099 |
JJ | -2.019 |
Hendricks | -1.923 |
Token | im |
Token ID | 320 |
Feature activation | -3.04e-01 |
Pos logit contributions | |
phe | +1.645 |
Ur | +1.590 |
aug | +1.570 |
ū | +1.569 |
kus | +1.501 |
Neg logit contributions | |
Howell | -2.481 |
Clarkson | -2.186 |
Birch | -2.184 |
JJ | -2.085 |
Hendricks | -2.009 |
Token | read |
Token ID | 961 |
Feature activation | -2.10e-01 |
Pos logit contributions | |
Howell | +3.025 |
Birch | +2.662 |
Clarkson | +2.657 |
JJ | +2.568 |
Hendricks | +2.453 |
Neg logit contributions | |
phe | -1.982 |
ū | -1.904 |
aug | -1.870 |
Ur | -1.863 |
UG | -1.765 |
Token | ( |
Token ID | 7 |
Feature activation | +8.01e-01 |
Pos logit contributions | |
Howell | +1.949 |
Birch | +1.703 |
Clarkson | +1.699 |
JJ | +1.633 |
Hendricks | +1.556 |
Neg logit contributions | |
phe | -1.520 |
ū | -1.466 |
aug | -1.444 |
Ur | -1.441 |
kus | -1.376 |
Token | image |
Token ID | 9060 |
Feature activation | -4.39e-02 |
Pos logit contributions | |
phe | +5.144 |
aug | +4.879 |
ū | +4.737 |
Ur | +4.737 |
kus | +4.594 |
Neg logit contributions | |
Howell | -7.899 |
Clarkson | -7.031 |
Birch | -6.803 |
JJ | -6.637 |
Hendricks | -6.370 |
Token | Path |
Token ID | 15235 |
Feature activation | +7.42e-02 |
Pos logit contributions | |
Howell | +0.472 |
Clarkson | +0.419 |
Birch | +0.418 |
JJ | +0.405 |
Hendricks | +0.388 |
Neg logit contributions | |
phe | -0.254 |
ū | -0.242 |
Ur | -0.240 |
aug | -0.237 |
kus | -0.224 |
Token | ) |
Token ID | 8 |
Feature activation | +2.43e-01 |
Pos logit contributions | |
phe | +0.510 |
ū | +0.492 |
aug | +0.485 |
Ur | +0.485 |
kus | +0.459 |
Neg logit contributions | |
Howell | -0.712 |
Clarkson | -0.628 |
Birch | -0.623 |
JJ | -0.598 |
Hendricks | -0.573 |
Token | image |
Token ID | 2939 |
Feature activation | -2.03e-01 |
Pos logit contributions | |
phe | +1.730 |
ū | +1.679 |
aug | +1.673 |
Ur | +1.663 |
kus | +1.587 |
Neg logit contributions | |
Howell | -2.225 |
Clarkson | -1.942 |
Birch | -1.928 |
JJ | -1.842 |
Hendricks | -1.777 |
Token | = |
Token ID | 796 |
Feature activation | +3.30e-01 |
Pos logit contributions | |
Howell | +1.929 |
Birch | +1.691 |
Clarkson | +1.686 |
JJ | +1.624 |
Hendricks | +1.548 |
Neg logit contributions | |
phe | -1.419 |
ū | -1.367 |
aug | -1.344 |
Ur | -1.343 |
kus | -1.278 |
Token | 34 |
Token ID | 2682 |
Feature activation | -2.03e-01 |
Pos logit contributions | |
Howell | +1.562 |
Birch | +1.367 |
Clarkson | +1.363 |
JJ | +1.314 |
Hendricks | +1.252 |
Neg logit contributions | |
phe | -1.165 |
ū | -1.123 |
aug | -1.104 |
Ur | -1.102 |
kus | -1.051 |
Token | lb |
Token ID | 18360 |
Feature activation | +3.02e-01 |
Pos logit contributions | |
Howell | +1.951 |
Birch | +1.712 |
Clarkson | +1.708 |
JJ | +1.648 |
Hendricks | +1.571 |
Neg logit contributions | |
phe | -1.402 |
ū | -1.350 |
aug | -1.329 |
Ur | -1.326 |
kus | -1.259 |
Token | .- |
Token ID | 7874 |
Feature activation | +3.34e-01 |
Pos logit contributions | |
phe | +2.036 |
ū | +1.964 |
Ur | +1.931 |
aug | +1.921 |
kus | +1.827 |
Neg logit contributions | |
Howell | -2.939 |
Birch | -2.582 |
Clarkson | -2.564 |
JJ | -2.471 |
Hendricks | -2.361 |
Token | ft |
Token ID | 701 |
Feature activation | +1.18e-01 |
Pos logit contributions | |
phe | +2.009 |
ū | +1.920 |
aug | +1.899 |
Ur | +1.892 |
kus | +1.786 |
Neg logit contributions | |
Howell | -3.502 |
Birch | -3.121 |
Clarkson | -3.088 |
JJ | -2.983 |
Hendricks | -2.855 |
Token | . |
Token ID | 13 |
Feature activation | +6.28e-01 |
Pos logit contributions | |
phe | +0.809 |
ū | +0.782 |
Ur | +0.766 |
aug | +0.765 |
kus | +0.727 |
Neg logit contributions | |
Howell | -1.136 |
Birch | -0.998 |
Clarkson | -0.989 |
JJ | -0.955 |
Hendricks | -0.910 |
Token | of |
Token ID | 286 |
Feature activation | +7.87e-01 |
Pos logit contributions | |
phe | +4.176 |
ū | +3.952 |
Ur | +3.892 |
aug | +3.862 |
UG | +3.745 |
Neg logit contributions | |
Howell | -5.902 |
Birch | -5.178 |
Clarkson | -5.040 |
JJ | -4.875 |
Hendricks | -4.560 |
Token | torque |
Token ID | 26415 |
Feature activation | +5.56e-01 |
Pos logit contributions | |
phe | +3.195 |
ū | +3.063 |
aug | +3.012 |
Ur | +2.979 |
ura | +2.861 |
Neg logit contributions | |
Howell | -9.044 |
Birch | -8.095 |
Clarkson | -8.026 |
JJ | -7.746 |
Gad | -7.467 |
Token | . |
Token ID | 13 |
Feature activation | +2.64e-01 |
Pos logit contributions | |
phe | +3.469 |
ū | +3.334 |
aug | +3.207 |
Ur | +3.198 |
UG | +3.076 |
Neg logit contributions | |
Howell | -5.450 |
Birch | -4.840 |
Clarkson | -4.690 |
JJ | -4.622 |
Hendricks | -4.405 |
Token |
|
Token ID | 198 |
Feature activation | +1.10e-01 |
Pos logit contributions | |
phe | +1.902 |
aug | +1.823 |
Ur | +1.822 |
ū | +1.821 |
UG | +1.735 |
Neg logit contributions | |
Howell | -2.384 |
Birch | -2.056 |
Clarkson | -2.023 |
JJ | -1.962 |
Dalton | -1.848 |
Token |
|
Token ID | 198 |
Feature activation | +1.55e-01 |
Pos logit contributions | |
phe | +0.912 |
ū | +0.883 |
aug | +0.869 |
Ur | +0.868 |
kus | +0.835 |
Neg logit contributions | |
Howell | -0.910 |
Birch | -0.781 |
Clarkson | -0.779 |
JJ | -0.745 |
Hendricks | -0.704 |
Token | The |
Token ID | 464 |
Feature activation | +2.26e-01 |
Pos logit contributions | |
phe | +0.896 |
Ur | +0.878 |
aug | +0.861 |
ū | +0.855 |
UG | +0.821 |
Neg logit contributions | |
Howell | -1.661 |
Birch | -1.478 |
Clarkson | -1.458 |
JJ | -1.425 |
Hendricks | -1.354 |
Token | s |
Token ID | 82 |
Feature activation | -2.46e-01 |
Pos logit contributions | |
Howell | +1.410 |
Birch | +1.255 |
Clarkson | +1.252 |
JJ | +1.212 |
Hendricks | +1.162 |
Neg logit contributions | |
phe | -0.774 |
ū | -0.738 |
aug | -0.726 |
Ur | -0.724 |
kus | -0.679 |
Token | Format |
Token ID | 26227 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
Howell | +2.045 |
Clarkson | +1.748 |
Birch | +1.748 |
JJ | +1.653 |
Hendricks | +1.597 |
Neg logit contributions | |
phe | -1.967 |
ū | -1.914 |
aug | -1.865 |
anamo | -1.833 |
kus | -1.832 |
Token | . |
Token ID | 13 |
Feature activation | +7.29e-02 |
Pos logit contributions | |
Howell | +1.215 |
Birch | +1.054 |
Clarkson | +1.050 |
JJ | +1.010 |
Hendricks | +0.956 |
Neg logit contributions | |
phe | -1.073 |
ū | -1.035 |
aug | -1.010 |
Ur | -0.994 |
kus | -0.984 |
Token | Al |
Token ID | 2348 |
Feature activation | -1.33e-02 |
Pos logit contributions | |
phe | +0.442 |
Ur | +0.426 |
ū | +0.422 |
aug | +0.418 |
UG | +0.393 |
Neg logit contributions | |
Howell | -0.762 |
Birch | -0.675 |
Clarkson | -0.675 |
JJ | -0.654 |
Hendricks | -0.626 |
Token | ignment |
Token ID | 16747 |
Feature activation | -4.61e-02 |
Pos logit contributions | |
Howell | +0.116 |
Clarkson | +0.101 |
Birch | +0.100 |
JJ | +0.096 |
Hendricks | +0.091 |
Neg logit contributions | |
phe | -0.102 |
ū | -0.099 |
Ur | -0.098 |
aug | -0.097 |
UG | -0.093 |
Token | = |
Token ID | 796 |
Feature activation | +7.83e-01 |
Pos logit contributions | |
Howell | +0.327 |
Birch | +0.275 |
Clarkson | +0.274 |
JJ | +0.259 |
Hendricks | +0.242 |
Neg logit contributions | |
phe | -0.431 |
ū | -0.418 |
Ur | -0.418 |
aug | -0.413 |
UG | -0.400 |
Token | [ |
Token ID | 685 |
Feature activation | +4.26e-02 |
Pos logit contributions | |
phe | +3.605 |
ū | +3.305 |
Ur | +3.281 |
aug | +3.280 |
ura | +3.054 |
Neg logit contributions | |
Howell | -8.510 |
Clarkson | -7.774 |
Birch | -7.607 |
JJ | -7.361 |
Hendricks | -7.170 |
Token | system |
Token ID | 10057 |
Feature activation | -1.43e-02 |
Pos logit contributions | |
phe | +0.316 |
ū | +0.304 |
aug | +0.301 |
Ur | +0.298 |
kus | +0.285 |
Neg logit contributions | |
Howell | -0.388 |
Clarkson | -0.340 |
Birch | -0.338 |
JJ | -0.325 |
Hendricks | -0.308 |
Token | . |
Token ID | 13 |
Feature activation | -6.53e-03 |
Pos logit contributions | |
Howell | +0.124 |
Birch | +0.108 |
Clarkson | +0.108 |
JJ | +0.104 |
Hendricks | +0.098 |
Neg logit contributions | |
phe | -0.111 |
ū | -0.107 |
aug | -0.104 |
Ur | -0.104 |
anamo | -0.102 |
Token | draw |
Token ID | 19334 |
Feature activation | -9.31e-02 |
Pos logit contributions | |
Howell | +0.062 |
Birch | +0.054 |
Clarkson | +0.054 |
JJ | +0.052 |
Hendricks | +0.050 |
Neg logit contributions | |
phe | -0.046 |
ū | -0.044 |
Ur | -0.043 |
aug | -0.043 |
kus | -0.041 |
Token | ing |
Token ID | 278 |
Feature activation | +1.72e-02 |
Pos logit contributions | |
Howell | +0.722 |
Birch | +0.618 |
Clarkson | +0.611 |
JJ | +0.588 |
Dalton | +0.546 |
Neg logit contributions | |
phe | -0.801 |
ū | -0.773 |
Ur | -0.754 |
aug | -0.748 |
anamo | -0.743 |
Token | . |
Token ID | 13 |
Feature activation | +2.10e-01 |
Pos logit contributions | |
phe | +1.011 |
ū | +0.973 |
aug | +0.958 |
Ur | +0.956 |
kus | +0.913 |
Neg logit contributions | |
Howell | -1.279 |
Birch | -1.116 |
Clarkson | -1.113 |
JJ | -1.072 |
Hendricks | -1.020 |
Token | 5 |
Token ID | 20 |
Feature activation | +1.45e-01 |
Pos logit contributions | |
phe | +0.928 |
ū | +0.874 |
aug | +0.849 |
Ur | +0.847 |
kus | +0.782 |
Neg logit contributions | |
Howell | -2.528 |
Birch | -2.282 |
Clarkson | -2.278 |
JJ | -2.214 |
Hendricks | -2.135 |
Token | - |
Token ID | 12 |
Feature activation | +1.21e-02 |
Pos logit contributions | |
phe | +0.873 |
ū | +0.843 |
Ur | +0.831 |
aug | +0.829 |
UG | +0.780 |
Neg logit contributions | |
Howell | -1.497 |
Birch | -1.334 |
Clarkson | -1.321 |
JJ | -1.273 |
Hendricks | -1.223 |
Token | liter |
Token ID | 17201 |
Feature activation | +5.47e-01 |
Pos logit contributions | |
phe | +0.096 |
ū | +0.093 |
aug | +0.092 |
Ur | +0.092 |
kus | +0.088 |
Neg logit contributions | |
Howell | -0.104 |
Birch | -0.090 |
Clarkson | -0.090 |
JJ | -0.086 |
Hendricks | -0.081 |
Token | Eco |
Token ID | 38719 |
Feature activation | +5.79e-01 |
Pos logit contributions | |
phe | +2.957 |
aug | +2.850 |
ū | +2.831 |
Ur | +2.803 |
UG | +2.675 |
Neg logit contributions | |
Howell | -5.705 |
Birch | -5.077 |
Clarkson | -5.047 |
JJ | -4.844 |
Hendricks | -4.673 |
Token | Boost |
Token ID | 45686 |
Feature activation | +7.82e-01 |
Pos logit contributions | |
phe | +1.728 |
Ur | +1.596 |
ū | +1.588 |
aug | +1.586 |
UG | +1.382 |
Neg logit contributions | |
Howell | -7.711 |
Birch | -7.015 |
Clarkson | -6.991 |
JJ | -6.834 |
Hendricks | -6.580 |
Token | proved |
Token ID | 8302 |
Feature activation | -4.86e-02 |
Pos logit contributions | |
phe | +4.735 |
ū | +4.621 |
Ur | +4.581 |
aug | +4.559 |
ura | +4.350 |
Neg logit contributions | |
Howell | -7.512 |
Birch | -6.709 |
Clarkson | -6.517 |
JJ | -6.242 |
Hendricks | -6.054 |
Token | its |
Token ID | 663 |
Feature activation | +2.29e-01 |
Pos logit contributions | |
Howell | +0.423 |
Birch | +0.366 |
Clarkson | +0.361 |
JJ | +0.349 |
Hendricks | +0.329 |
Neg logit contributions | |
phe | -0.375 |
ū | -0.363 |
aug | -0.359 |
Ur | -0.359 |
kus | -0.344 |
Token | met |
Token ID | 1138 |
Feature activation | +4.68e-02 |
Pos logit contributions | |
phe | +1.588 |
ū | +1.551 |
Ur | +1.541 |
aug | +1.523 |
UG | +1.444 |
Neg logit contributions | |
Howell | -2.154 |
Birch | -1.876 |
Clarkson | -1.861 |
JJ | -1.784 |
Hendricks | -1.697 |
Token | tle |
Token ID | 7100 |
Feature activation | +3.83e-01 |
Pos logit contributions | |
phe | +0.213 |
ū | +0.200 |
aug | +0.197 |
Ur | +0.194 |
kus | +0.179 |
Neg logit contributions | |
Howell | -0.560 |
Birch | -0.503 |
Clarkson | -0.502 |
JJ | -0.488 |
Hendricks | -0.472 |
Token | off |
Token ID | 572 |
Feature activation | -1.87e-01 |
Pos logit contributions | |
phe | +2.689 |
ū | +2.626 |
aug | +2.607 |
Ur | +2.579 |
UG | +2.483 |
Neg logit contributions | |
Howell | -3.480 |
Birch | -3.049 |
Clarkson | -2.986 |
JJ | -2.895 |
Hendricks | -2.762 |
Token | Element |
Token ID | 20180 |
Feature activation | -2.31e-01 |
Pos logit contributions | |
Howell | +0.838 |
Birch | +0.706 |
Clarkson | +0.703 |
JJ | +0.666 |
Hendricks | +0.627 |
Neg logit contributions | |
phe | -1.034 |
ū | -1.008 |
aug | -0.992 |
Ur | -0.981 |
kus | -0.960 |
Token | ( |
Token ID | 7 |
Feature activation | +3.47e-01 |
Pos logit contributions | |
Howell | +1.808 |
Clarkson | +1.547 |
Birch | +1.536 |
JJ | +1.473 |
Hendricks | +1.377 |
Neg logit contributions | |
phe | -1.954 |
ū | -1.910 |
aug | -1.863 |
Ur | -1.850 |
anamo | -1.826 |
Token | cv |
Token ID | 33967 |
Feature activation | -7.38e-02 |
Pos logit contributions | |
phe | +1.773 |
ū | +1.675 |
aug | +1.660 |
Ur | +1.634 |
kus | +1.538 |
Neg logit contributions | |
Howell | -3.958 |
Clarkson | -3.540 |
Birch | -3.540 |
JJ | -3.431 |
Hendricks | -3.312 |
Token | 2 |
Token ID | 17 |
Feature activation | +3.56e-02 |
Pos logit contributions | |
Howell | +0.703 |
Birch | +0.620 |
Clarkson | +0.617 |
JJ | +0.599 |
Hendricks | +0.567 |
Neg logit contributions | |
phe | -0.508 |
ū | -0.488 |
aug | -0.475 |
Ur | -0.465 |
anamo | -0.461 |
Token | . |
Token ID | 13 |
Feature activation | +5.35e-01 |
Pos logit contributions | |
phe | +0.266 |
ū | +0.257 |
aug | +0.251 |
Ur | +0.250 |
kus | +0.241 |
Neg logit contributions | |
Howell | -0.320 |
Birch | -0.279 |
Clarkson | -0.278 |
JJ | -0.268 |
Hendricks | -0.254 |
Token | M |
Token ID | 44 |
Feature activation | +7.78e-01 |
Pos logit contributions | |
phe | +2.979 |
Ur | +2.963 |
UG | +2.854 |
aug | +2.838 |
ū | +2.802 |
Neg logit contributions | |
Howell | -5.692 |
Clarkson | -5.137 |
Birch | -5.050 |
JJ | -4.854 |
Hendricks | -4.658 |
Token | OR |
Token ID | 1581 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
Ur | +2.652 |
phe | +2.622 |
UG | +2.610 |
ū | +2.523 |
aug | +2.428 |
Neg logit contributions | |
Howell | -9.938 |
Birch | -8.979 |
Clarkson | -8.914 |
JJ | -8.627 |
Hendricks | -8.488 |
Token | PH |
Token ID | 11909 |
Feature activation | +8.69e-02 |
Pos logit contributions | |
phe | +0.628 |
ū | +0.608 |
aug | +0.594 |
Ur | +0.592 |
kus | +0.563 |
Neg logit contributions | |
Howell | -1.081 |
Birch | -0.961 |
Clarkson | -0.958 |
JJ | -0.926 |
Hendricks | -0.891 |
Token | _ |
Token ID | 62 |
Feature activation | +6.74e-01 |
Pos logit contributions | |
phe | +0.624 |
ū | +0.598 |
aug | +0.588 |
Ur | +0.583 |
kus | +0.561 |
Neg logit contributions | |
Howell | -0.810 |
Birch | -0.708 |
Clarkson | -0.707 |
JJ | -0.683 |
Hendricks | -0.649 |
Token | RECT |
Token ID | 23988 |
Feature activation | +1.38e-01 |
Pos logit contributions | |
phe | +5.185 |
Ur | +5.140 |
UG | +5.125 |
ū | +5.103 |
aug | +5.038 |
Neg logit contributions | |
Howell | -5.744 |
Birch | -4.940 |
Clarkson | -4.915 |
JJ | -4.673 |
Hendricks | -4.451 |
Token | , |
Token ID | 11 |
Feature activation | +4.05e-01 |
Pos logit contributions | |
phe | +1.037 |
ū | +1.002 |
aug | +0.986 |
Ur | +0.985 |
kus | +0.941 |
Neg logit contributions | |
Howell | -1.233 |
Birch | -1.071 |
Clarkson | -1.068 |
JJ | -1.027 |
Hendricks | -0.975 |
Token | ) |
Token ID | 1267 |
Feature activation | +1.28e-01 |
Pos logit contributions | |
phe | +0.391 |
ū | +0.382 |
Ur | +0.377 |
aug | +0.374 |
UG | +0.363 |
Neg logit contributions | |
Howell | -0.374 |
Birch | -0.319 |
Clarkson | -0.318 |
JJ | -0.303 |
Hendricks | -0.288 |
Token | = |
Token ID | 796 |
Feature activation | +2.35e-01 |
Pos logit contributions | |
phe | +1.226 |
ū | +1.204 |
Ur | +1.202 |
aug | +1.178 |
UG | +1.167 |
Neg logit contributions | |
Howell | -0.871 |
Clarkson | -0.724 |
Birch | -0.719 |
JJ | -0.671 |
Hendricks | -0.637 |
Token | '. |
Token ID | 45302 |
Feature activation | -7.13e-02 |
Pos logit contributions | |
phe | +1.683 |
ū | +1.628 |
aug | +1.615 |
Ur | +1.602 |
UG | +1.520 |
Neg logit contributions | |
Howell | -2.138 |
Birch | -1.839 |
Clarkson | -1.808 |
JJ | -1.770 |
Hendricks | -1.701 |
Token | cl |
Token ID | 565 |
Feature activation | -5.75e-01 |
Pos logit contributions | |
Howell | +0.696 |
Birch | +0.609 |
Clarkson | +0.606 |
JJ | +0.585 |
Hendricks | +0.559 |
Neg logit contributions | |
phe | -0.486 |
ū | -0.467 |
aug | -0.463 |
Ur | -0.457 |
kus | -0.438 |
Token | js |
Token ID | 8457 |
Feature activation | +3.72e-01 |
Pos logit contributions | |
Howell | +5.162 |
Clarkson | +4.523 |
Birch | +4.501 |
JJ | +4.394 |
Hendricks | +4.137 |
Neg logit contributions | |
phe | -4.144 |
ū | -3.983 |
Ur | -3.930 |
aug | -3.898 |
UG | -3.752 |
Token | ' |
Token ID | 6 |
Feature activation | +7.73e-01 |
Pos logit contributions | |
phe | +1.825 |
ū | +1.752 |
Ur | +1.738 |
aug | +1.703 |
UG | +1.606 |
Neg logit contributions | |
Howell | -4.264 |
Birch | -3.824 |
Clarkson | -3.775 |
JJ | -3.670 |
Hendricks | -3.554 |
Token | OR |
Token ID | 6375 |
Feature activation | +2.09e-01 |
Pos logit contributions | |
phe | +4.913 |
Ur | +4.880 |
ū | +4.773 |
UG | +4.700 |
]] | +4.605 |
Neg logit contributions | |
Howell | -7.088 |
Birch | -6.105 |
Clarkson | -6.055 |
Hendricks | -5.722 |
JJ | -5.708 |
Token | RIGHT |
Token ID | 33621 |
Feature activation | +2.40e-02 |
Pos logit contributions | |
phe | +1.710 |
ū | +1.674 |
Ur | +1.648 |
aug | +1.629 |
UG | +1.601 |
Neg logit contributions | |
Howell | -1.675 |
Birch | -1.417 |
Clarkson | -1.393 |
JJ | -1.326 |
Hendricks | -1.291 |
Token | ( |
Token ID | 357 |
Feature activation | +3.22e-01 |
Pos logit contributions | |
phe | +0.190 |
ū | +0.183 |
aug | +0.181 |
Ur | +0.179 |
kus | +0.173 |
Neg logit contributions | |
Howell | -0.207 |
Birch | -0.179 |
Clarkson | -0.179 |
JJ | -0.172 |
Hendricks | -0.162 |
Token | path |
Token ID | 3108 |
Feature activation | +2.97e-02 |
Pos logit contributions | |
phe | +1.799 |
ū | +1.734 |
Ur | +1.702 |
aug | +1.676 |
UG | +1.598 |
Neg logit contributions | |
Howell | -3.435 |
Clarkson | -3.009 |
Birch | -2.994 |
JJ | -2.899 |
Hendricks | -2.823 |
Token | , |
Token ID | 837 |
Feature activation | +2.87e-01 |
Pos logit contributions | |
phe | +0.255 |
ū | +0.247 |
aug | +0.244 |
Ur | +0.242 |
kus | +0.234 |
Neg logit contributions | |
Howell | -0.235 |
Birch | -0.201 |
Clarkson | -0.200 |
JJ | -0.192 |
Hendricks | -0.180 |
Token | - |
Token ID | 12 |
Feature activation | +1.41e-01 |
Pos logit contributions | |
phe | +0.398 |
ū | +0.384 |
aug | +0.377 |
Ur | +0.375 |
UG | +0.358 |
Neg logit contributions | |
Howell | -0.587 |
Birch | -0.516 |
Clarkson | -0.514 |
JJ | -0.494 |
Hendricks | -0.474 |
Token | to |
Token ID | 1462 |
Feature activation | -7.88e-02 |
Pos logit contributions | |
phe | +0.851 |
ū | +0.803 |
aug | +0.795 |
Ur | +0.790 |
UG | +0.751 |
Neg logit contributions | |
Howell | -1.491 |
Clarkson | -1.324 |
Birch | -1.322 |
JJ | -1.276 |
Hendricks | -1.226 |
Token | - |
Token ID | 12 |
Feature activation | +5.04e-02 |
Pos logit contributions | |
Howell | +0.758 |
Birch | +0.664 |
Clarkson | +0.663 |
JJ | +0.642 |
Hendricks | +0.611 |
Neg logit contributions | |
phe | -0.537 |
ū | -0.518 |
aug | -0.508 |
Ur | -0.508 |
UG | -0.483 |
Token | female |
Token ID | 24724 |
Feature activation | -2.01e-02 |
Pos logit contributions | |
phe | +0.209 |
ū | +0.193 |
aug | +0.190 |
Ur | +0.188 |
kus | +0.174 |
Neg logit contributions | |
Howell | -0.626 |
Clarkson | -0.571 |
Birch | -0.570 |
JJ | -0.553 |
Hendricks | -0.532 |
Token | ( |
Token ID | 357 |
Feature activation | +4.68e-01 |
Pos logit contributions | |
Howell | +0.182 |
Birch | +0.159 |
Clarkson | +0.158 |
JJ | +0.151 |
Hendricks | +0.144 |
Neg logit contributions | |
phe | -0.148 |
ū | -0.143 |
aug | -0.140 |
Ur | -0.140 |
UG | -0.134 |
Token | M |
Token ID | 44 |
Feature activation | +7.69e-01 |
Pos logit contributions | |
phe | +2.366 |
aug | +2.232 |
ū | +2.221 |
UG | +2.162 |
Ur | +2.142 |
Neg logit contributions | |
Howell | -5.283 |
Clarkson | -4.806 |
Birch | -4.776 |
JJ | -4.468 |
Dalton | -4.429 |
Token | TF |
Token ID | 10234 |
Feature activation | +5.62e-02 |
Pos logit contributions | |
phe | +2.352 |
ū | +2.230 |
Ur | +2.216 |
UG | +2.189 |
aug | +2.099 |
Neg logit contributions | |
Howell | -10.134 |
Birch | -9.253 |
Clarkson | -9.128 |
JJ | -8.727 |
Hendricks | -8.650 |
Token | ) |
Token ID | 8 |
Feature activation | -3.91e-01 |
Pos logit contributions | |
phe | +0.386 |
ū | +0.375 |
Ur | +0.368 |
aug | +0.366 |
UG | +0.352 |
Neg logit contributions | |
Howell | -0.538 |
Birch | -0.472 |
Clarkson | -0.471 |
JJ | -0.452 |
Hendricks | -0.432 |
Token | transgender |
Token ID | 10637 |
Feature activation | -2.17e-01 |
Pos logit contributions | |
Howell | +3.384 |
Birch | +2.921 |
Clarkson | +2.912 |
JJ | +2.804 |
Hendricks | +2.655 |
Neg logit contributions | |
phe | -3.054 |
ū | -2.954 |
aug | -2.910 |
Ur | -2.909 |
kus | -2.790 |
Token | people |
Token ID | 661 |
Feature activation | +6.26e-02 |
Pos logit contributions | |
Howell | +1.818 |
Birch | +1.563 |
Clarkson | +1.559 |
JJ | +1.489 |
Hendricks | +1.408 |
Neg logit contributions | |
phe | -1.765 |
ū | -1.709 |
aug | -1.685 |
Ur | -1.680 |
UG | -1.613 |
Token | are |
Token ID | 389 |
Feature activation | +3.62e-02 |
Pos logit contributions | |
phe | +0.453 |
ū | +0.435 |
aug | +0.428 |
Ur | +0.425 |
kus | +0.408 |
Neg logit contributions | |
Howell | -0.580 |
Birch | -0.505 |
Clarkson | -0.504 |
JJ | -0.484 |
Hendricks | -0.463 |
Token | 102 |
Token ID | 15377 |
Feature activation | +1.64e-01 |
Pos logit contributions | |
phe | +0.804 |
ū | +0.782 |
Ur | +0.774 |
aug | +0.774 |
UG | +0.745 |
Neg logit contributions | |
Howell | -0.886 |
Birch | -0.762 |
Clarkson | -0.757 |
JJ | -0.728 |
Hendricks | -0.692 |
Token | ] |
Token ID | 60 |
Feature activation | +1.81e-02 |
Pos logit contributions | |
phe | +0.546 |
ū | +0.511 |
aug | +0.494 |
Ur | +0.486 |
UG | +0.433 |
Neg logit contributions | |
Howell | -2.146 |
Birch | -1.950 |
Clarkson | -1.949 |
JJ | -1.901 |
Hendricks | -1.840 |
Token | For |
Token ID | 1114 |
Feature activation | +3.02e-01 |
Pos logit contributions | |
phe | +0.130 |
ū | +0.126 |
Ur | +0.124 |
aug | +0.124 |
UG | +0.118 |
Neg logit contributions | |
Howell | -0.166 |
Birch | -0.145 |
Clarkson | -0.144 |
JJ | -0.139 |
Hendricks | -0.132 |
Token | the |
Token ID | 262 |
Feature activation | +2.27e-01 |
Pos logit contributions | |
phe | +2.049 |
ū | +1.994 |
aug | +1.945 |
Ur | +1.910 |
kus | +1.879 |
Neg logit contributions | |
Howell | -2.860 |
Birch | -2.499 |
Clarkson | -2.462 |
JJ | -2.401 |
Hendricks | -2.289 |
Token | 2014 |
Token ID | 1946 |
Feature activation | +4.56e-01 |
Pos logit contributions | |
phe | +1.649 |
ū | +1.606 |
aug | +1.565 |
Ur | +1.553 |
kus | +1.516 |
Neg logit contributions | |
Howell | -2.052 |
Birch | -1.775 |
Clarkson | -1.755 |
JJ | -1.696 |
Hendricks | -1.605 |
Token | – |
Token ID | 1906 |
Feature activation | +7.69e-01 |
Pos logit contributions | |
phe | +2.905 |
ū | +2.863 |
aug | +2.808 |
Ur | +2.751 |
UG | +2.712 |
Neg logit contributions | |
Howell | -4.401 |
Birch | -3.823 |
Clarkson | -3.766 |
JJ | -3.688 |
Hendricks | -3.549 |
Token | 15 |
Token ID | 1314 |
Feature activation | +5.12e-01 |
Pos logit contributions | |
aug | +3.284 |
phe | +3.219 |
Ur | +3.119 |
ū | +2.996 |
kus | +2.812 |
Neg logit contributions | |
Howell | -8.957 |
Birch | -8.176 |
Clarkson | -7.991 |
JJ | -7.892 |
Hendricks | -7.676 |
Token | Atlantic |
Token ID | 10596 |
Feature activation | +2.08e-01 |
Pos logit contributions | |
phe | +3.378 |
ū | +3.338 |
aug | +3.229 |
Ur | +3.166 |
UG | +3.127 |
Neg logit contributions | |
Howell | -4.822 |
Birch | -4.183 |
Clarkson | -4.101 |
JJ | -4.013 |
Hendricks | -3.860 |
Token | Coast |
Token ID | 8545 |
Feature activation | +2.16e-01 |
Pos logit contributions | |
phe | +1.732 |
ū | +1.693 |
aug | +1.685 |
Ur | +1.658 |
UG | +1.599 |
Neg logit contributions | |
Howell | -1.665 |
Clarkson | -1.403 |
Birch | -1.396 |
JJ | -1.355 |
Hendricks | -1.277 |
Token | Conference |
Token ID | 8785 |
Feature activation | +5.90e-01 |
Pos logit contributions | |
phe | +1.584 |
ū | +1.540 |
aug | +1.517 |
Ur | +1.502 |
UG | +1.445 |
Neg logit contributions | |
Howell | -1.955 |
Birch | -1.698 |
Clarkson | -1.675 |
JJ | -1.634 |
Hendricks | -1.547 |
Token | men |
Token ID | 1450 |
Feature activation | +4.44e-01 |
Pos logit contributions | |
phe | +4.016 |
ū | +3.941 |
aug | +3.830 |
Ur | +3.775 |
kus | +3.656 |
Neg logit contributions | |
Howell | -5.449 |
Birch | -4.708 |
Clarkson | -4.632 |
JJ | -4.578 |
Hendricks | -4.327 |
Token | YR |
Token ID | 38162 |
Feature activation | +6.76e-02 |
Pos logit contributions | |
phe | +1.089 |
ū | +1.074 |
Ur | +1.049 |
aug | +1.026 |
UG | +1.006 |
Neg logit contributions | |
Howell | -1.753 |
Birch | -1.538 |
Clarkson | -1.534 |
JJ | -1.491 |
Hendricks | -1.427 |
Token | DS |
Token ID | 17400 |
Feature activation | -2.53e-01 |
Pos logit contributions | |
phe | +0.471 |
ū | +0.454 |
Ur | +0.444 |
aug | +0.435 |
UG | +0.430 |
Neg logit contributions | |
Howell | -0.646 |
Clarkson | -0.563 |
Birch | -0.561 |
JJ | -0.538 |
Hendricks | -0.518 |
Token | cent |
Token ID | 1247 |
Feature activation | -1.25e-01 |
Pos logit contributions | |
Howell | +2.292 |
Birch | +1.997 |
Clarkson | +1.989 |
JJ | +1.916 |
Hendricks | +1.819 |
Neg logit contributions | |
phe | -1.871 |
ū | -1.805 |
aug | -1.779 |
Ur | -1.775 |
kus | -1.696 |
Token | ering |
Token ID | 1586 |
Feature activation | -6.01e-02 |
Pos logit contributions | |
Howell | +1.144 |
Birch | +0.995 |
Clarkson | +0.990 |
JJ | +0.953 |
Hendricks | +0.910 |
Neg logit contributions | |
phe | -0.916 |
ū | -0.889 |
aug | -0.872 |
Ur | -0.870 |
kus | -0.830 |
Token |
|
Token ID | 198 |
Feature activation | +4.92e-01 |
Pos logit contributions | |
Howell | +0.528 |
Birch | +0.454 |
Clarkson | +0.453 |
JJ | +0.434 |
Hendricks | +0.413 |
Neg logit contributions | |
phe | -0.465 |
ū | -0.450 |
Ur | -0.441 |
aug | -0.440 |
UG | -0.423 |
Token |
|
Token ID | 198 |
Feature activation | +7.66e-01 |
Pos logit contributions | |
phe | +3.614 |
ū | +3.501 |
aug | +3.492 |
Ur | +3.482 |
UG | +3.307 |
Neg logit contributions | |
Howell | -4.424 |
Birch | -3.838 |
Clarkson | -3.801 |
JJ | -3.661 |
Hendricks | -3.478 |
Token | The |
Token ID | 464 |
Feature activation | +1.96e-01 |
Pos logit contributions | |
Ur | +3.381 |
phe | +3.371 |
UG | +3.161 |
aug | +3.150 |
ū | +3.109 |
Neg logit contributions | |
Howell | -9.024 |
Birch | -8.117 |
Clarkson | -8.071 |
JJ | -7.810 |
Hendricks | -7.536 |
Token | Garg |
Token ID | 39013 |
Feature activation | -5.05e-02 |
Pos logit contributions | |
phe | +1.285 |
ū | +1.239 |
Ur | +1.208 |
aug | +1.201 |
UG | +1.158 |
Neg logit contributions | |
Howell | -1.930 |
Clarkson | -1.697 |
Birch | -1.693 |
JJ | -1.614 |
Hendricks | -1.556 |
Token | ant |
Token ID | 415 |
Feature activation | -4.70e-02 |
Pos logit contributions | |
Howell | +0.451 |
Birch | +0.392 |
Clarkson | +0.389 |
JJ | +0.374 |
Hendricks | +0.356 |
Neg logit contributions | |
phe | -0.379 |
ū | -0.362 |
Ur | -0.361 |
aug | -0.356 |
kus | -0.346 |
Token | ua |
Token ID | 6413 |
Feature activation | -1.41e-01 |
Pos logit contributions | |
Howell | +0.421 |
Birch | +0.365 |
Clarkson | +0.363 |
JJ | +0.351 |
Hendricks | +0.333 |
Neg logit contributions | |
phe | -0.352 |
ū | -0.337 |
Ur | -0.335 |
aug | -0.333 |
kus | -0.319 |
Token | is |
Token ID | 318 |
Feature activation | +1.12e-01 |
Pos logit contributions | |
Howell | +1.233 |
Birch | +1.067 |
Clarkson | +1.064 |
JJ | +1.019 |
Hendricks | +0.969 |
Neg logit contributions | |
phe | -1.090 |
ū | -1.055 |
aug | -1.037 |
Ur | -1.037 |
kus | -0.992 |
Token | Mysteries |
Token ID | 28298 |
Feature activation | +4.60e-01 |
Pos logit contributions | |
phe | +0.620 |
ū | +0.598 |
Ur | +0.590 |
aug | +0.590 |
kus | +0.568 |
Neg logit contributions | |
Howell | -0.558 |
Birch | -0.475 |
Clarkson | -0.471 |
JJ | -0.449 |
Hendricks | -0.423 |
Token | , |
Token ID | 11 |
Feature activation | +4.34e-01 |
Pos logit contributions | |
phe | +3.222 |
ū | +3.134 |
Ur | +3.071 |
aug | +3.057 |
UG | +2.917 |
Neg logit contributions | |
Howell | -4.302 |
Birch | -3.759 |
Clarkson | -3.747 |
JJ | -3.596 |
Hendricks | -3.427 |
Token | comic |
Token ID | 9048 |
Feature activation | +2.23e-01 |
Pos logit contributions | |
phe | +3.150 |
ū | +3.110 |
aug | +3.026 |
Ur | +2.996 |
UG | +2.894 |
Neg logit contributions | |
Howell | -3.920 |
Birch | -3.359 |
Clarkson | -3.354 |
JJ | -3.191 |
Hendricks | -3.077 |
Token | book |
Token ID | 1492 |
Feature activation | +9.03e-02 |
Pos logit contributions | |
phe | +1.933 |
ū | +1.881 |
aug | +1.856 |
Ur | +1.853 |
kus | +1.783 |
Neg logit contributions | |
Howell | -1.735 |
Birch | -1.469 |
Clarkson | -1.466 |
JJ | -1.395 |
Hendricks | -1.316 |
Token | short |
Token ID | 1790 |
Feature activation | +4.31e-01 |
Pos logit contributions | |
phe | +0.646 |
ū | +0.621 |
Ur | +0.610 |
aug | +0.609 |
kus | +0.584 |
Neg logit contributions | |
Howell | -0.846 |
Birch | -0.744 |
Clarkson | -0.741 |
JJ | -0.711 |
Hendricks | -0.679 |
Token | stories |
Token ID | 3923 |
Feature activation | +7.66e-01 |
Pos logit contributions | |
phe | +3.475 |
ū | +3.442 |
Ur | +3.397 |
aug | +3.360 |
kus | +3.238 |
Neg logit contributions | |
Howell | -3.553 |
Clarkson | -3.042 |
Birch | -3.032 |
JJ | -2.860 |
Hendricks | -2.721 |
Token | , |
Token ID | 11 |
Feature activation | +2.62e-01 |
Pos logit contributions | |
phe | +4.841 |
ū | +4.723 |
Ur | +4.657 |
aug | +4.626 |
UG | +4.403 |
Neg logit contributions | |
Howell | -7.437 |
Birch | -6.543 |
Clarkson | -6.530 |
JJ | -6.258 |
Hendricks | -5.981 |
Token | prose |
Token ID | 27149 |
Feature activation | +6.17e-01 |
Pos logit contributions | |
phe | +1.622 |
ū | +1.569 |
Ur | +1.520 |
aug | +1.518 |
kus | +1.431 |
Neg logit contributions | |
Howell | -2.683 |
Birch | -2.377 |
Clarkson | -2.366 |
JJ | -2.281 |
Hendricks | -2.183 |
Token | , |
Token ID | 11 |
Feature activation | -6.37e-02 |
Pos logit contributions | |
phe | +4.074 |
ū | +3.959 |
aug | +3.939 |
Ur | +3.935 |
UG | +3.727 |
Neg logit contributions | |
Howell | -5.899 |
Clarkson | -5.176 |
Birch | -5.166 |
JJ | -4.934 |
Hendricks | -4.722 |
Token | and |
Token ID | 290 |
Feature activation | +9.10e-02 |
Pos logit contributions | |
Howell | +0.512 |
Birch | +0.440 |
Clarkson | +0.436 |
JJ | +0.418 |
Hendricks | +0.395 |
Neg logit contributions | |
phe | -0.541 |
ū | -0.521 |
aug | -0.514 |
Ur | -0.512 |
kus | -0.496 |
Token | non |
Token ID | 1729 |
Feature activation | +5.20e-01 |
Pos logit contributions | |
phe | +0.699 |
ū | +0.674 |
aug | +0.665 |
Ur | +0.664 |
kus | +0.635 |
Neg logit contributions | |
Howell | -0.803 |
Birch | -0.696 |
Clarkson | -0.693 |
JJ | -0.667 |
Hendricks | -0.632 |
Token | with |
Token ID | 351 |
Feature activation | -3.83e-02 |
Pos logit contributions | |
Howell | +0.352 |
Birch | +0.300 |
Clarkson | +0.298 |
JJ | +0.284 |
Hendricks | +0.269 |
Neg logit contributions | |
phe | -0.380 |
ū | -0.368 |
aug | -0.363 |
Ur | -0.363 |
kus | -0.348 |
Token | bad |
Token ID | 2089 |
Feature activation | +2.38e-02 |
Pos logit contributions | |
Howell | +0.335 |
Birch | +0.290 |
Clarkson | +0.289 |
JJ | +0.277 |
Hendricks | +0.263 |
Neg logit contributions | |
phe | -0.296 |
ū | -0.285 |
aug | -0.282 |
Ur | -0.281 |
kus | -0.270 |
Token | manners |
Token ID | 33558 |
Feature activation | -1.46e-01 |
Pos logit contributions | |
phe | +0.179 |
ū | +0.172 |
Ur | +0.170 |
aug | +0.170 |
kus | +0.163 |
Neg logit contributions | |
Howell | -0.213 |
Clarkson | -0.186 |
Birch | -0.186 |
JJ | -0.177 |
Hendricks | -0.169 |
Token | . |
Token ID | 13 |
Feature activation | +2.67e-02 |
Pos logit contributions | |
Howell | +1.321 |
Birch | +1.150 |
Clarkson | +1.146 |
JJ | +1.101 |
Hendricks | +1.047 |
Neg logit contributions | |
phe | -1.081 |
ū | -1.044 |
aug | -1.028 |
Ur | -1.027 |
kus | -0.979 |
Token | - |
Token ID | 532 |
Feature activation | -1.27e-01 |
Pos logit contributions | |
phe | +0.187 |
ū | +0.180 |
Ur | +0.178 |
aug | +0.178 |
UG | +0.169 |
Neg logit contributions | |
Howell | -0.253 |
Birch | -0.221 |
Clarkson | -0.221 |
JJ | -0.212 |
Hendricks | -0.203 |
Token | January |
Token ID | 3269 |
Feature activation | +7.61e-01 |
Pos logit contributions | |
Howell | +1.060 |
Clarkson | +0.916 |
Birch | +0.910 |
JJ | +0.876 |
Hendricks | +0.834 |
Neg logit contributions | |
phe | -1.033 |
ū | -0.997 |
Ur | -0.985 |
aug | -0.977 |
anamo | -0.945 |
Token | 4 |
Token ID | 604 |
Feature activation | +2.89e-01 |
Pos logit contributions | |
ū | +4.520 |
phe | +4.475 |
Ur | +4.389 |
aug | +4.387 |
UG | +4.191 |
Neg logit contributions | |
Howell | -7.571 |
Clarkson | -6.696 |
Birch | -6.678 |
JJ | -6.435 |
Hendricks | -6.216 |
Token | , |
Token ID | 11 |
Feature activation | +2.60e-01 |
Pos logit contributions | |
phe | +2.114 |
ū | +2.043 |
aug | +2.019 |
Ur | +2.016 |
UG | +1.921 |
Neg logit contributions | |
Howell | -2.648 |
Birch | -2.310 |
Clarkson | -2.300 |
JJ | -2.209 |
Hendricks | -2.102 |
Token | 2010 |
Token ID | 3050 |
Feature activation | +5.86e-01 |
Pos logit contributions | |
phe | +1.816 |
ū | +1.750 |
aug | +1.723 |
Ur | +1.721 |
UG | +1.640 |
Neg logit contributions | |
Howell | -2.430 |
Birch | -2.136 |
Clarkson | -2.135 |
JJ | -2.051 |
Hendricks | -1.951 |
Token | User |
Token ID | 12982 |
Feature activation | -2.34e-02 |
Pos logit contributions | |
phe | +2.063 |
Ur | +2.051 |
ū | +2.003 |
aug | +1.970 |
UG | +1.894 |
Neg logit contributions | |
Howell | -7.536 |
Birch | -6.817 |
Clarkson | -6.758 |
JJ | -6.615 |
Hendricks | -6.448 |
Token | reviews |
Token ID | 8088 |
Feature activation | +3.51e-01 |
Pos logit contributions | |
Howell | +0.183 |
Birch | +0.156 |
Clarkson | +0.156 |
JJ | +0.149 |
Hendricks | +0.140 |
Neg logit contributions | |
phe | -0.204 |
ū | -0.196 |
Ur | -0.194 |
aug | -0.194 |
kus | -0.187 |
Token | ' |
Token ID | 6 |
Feature activation | -6.94e-02 |
Pos logit contributions | |
Howell | +1.874 |
Birch | +1.653 |
Clarkson | +1.648 |
JJ | +1.592 |
Hendricks | +1.525 |
Neg logit contributions | |
phe | -1.206 |
ū | -1.159 |
aug | -1.138 |
Ur | -1.136 |
kus | -1.077 |
Token | " |
Token ID | 366 |
Feature activation | +5.26e-02 |
Pos logit contributions | |
Howell | +0.559 |
Birch | +0.475 |
Clarkson | +0.473 |
JJ | +0.451 |
Hendricks | +0.429 |
Neg logit contributions | |
phe | -0.585 |
ū | -0.565 |
aug | -0.559 |
Ur | -0.557 |
kus | -0.537 |
Token | I |
Token ID | 40 |
Feature activation | -1.59e-02 |
Pos logit contributions | |
phe | +0.301 |
Ur | +0.288 |
ū | +0.285 |
aug | +0.281 |
UG | +0.267 |
Neg logit contributions | |
Howell | -0.570 |
Birch | -0.506 |
Clarkson | -0.504 |
JJ | -0.489 |
Hendricks | -0.471 |
Token | 'll |
Token ID | 1183 |
Feature activation | +2.12e-01 |
Pos logit contributions | |
Howell | +0.144 |
Birch | +0.125 |
Clarkson | +0.125 |
JJ | +0.120 |
Hendricks | +0.115 |
Neg logit contributions | |
phe | -0.118 |
ū | -0.113 |
aug | -0.111 |
Ur | -0.111 |
kus | -0.107 |
Token | Feel |
Token ID | 18571 |
Feature activation | -2.47e-01 |
Pos logit contributions | |
phe | +0.969 |
ū | +0.916 |
aug | +0.893 |
Ur | +0.877 |
kus | +0.806 |
Neg logit contributions | |
Howell | -2.506 |
Clarkson | -2.255 |
Birch | -2.248 |
JJ | -2.174 |
Hendricks | -2.134 |
Token | a |
Token ID | 257 |
Feature activation | +7.52e-01 |
Pos logit contributions | |
Howell | +2.324 |
Birch | +2.034 |
Clarkson | +2.028 |
JJ | +1.954 |
Hendricks | +1.861 |
Neg logit contributions | |
phe | -1.746 |
ū | -1.682 |
aug | -1.655 |
Ur | -1.653 |
kus | -1.574 |
Token | Whole |
Token ID | 23431 |
Feature activation | +2.17e-01 |
Pos logit contributions | |
phe | +4.644 |
ū | +4.367 |
aug | +4.346 |
Ur | +4.234 |
UG | +4.146 |
Neg logit contributions | |
Howell | -7.320 |
Birch | -6.484 |
Clarkson | -6.362 |
JJ | -6.253 |
Hendricks | -5.934 |
Token | Lot |
Token ID | 15099 |
Feature activation | +4.34e-02 |
Pos logit contributions | |
phe | +1.392 |
ū | +1.336 |
aug | +1.312 |
Ur | +1.310 |
kus | +1.241 |
Neg logit contributions | |
Howell | -2.193 |
Birch | -1.937 |
Clarkson | -1.932 |
JJ | -1.867 |
Hendricks | -1.785 |
Token | Better |
Token ID | 11625 |
Feature activation | -1.25e-01 |
Pos logit contributions | |
phe | +0.308 |
ū | +0.297 |
aug | +0.292 |
Ur | +0.291 |
UG | +0.278 |
Neg logit contributions | |
Howell | -0.407 |
Birch | -0.355 |
Clarkson | -0.354 |
JJ | -0.340 |
Hendricks | -0.326 |
Token | " |
Token ID | 1 |
Feature activation | -7.15e-02 |
Pos logit contributions | |
Howell | +1.124 |
Birch | +0.976 |
Clarkson | +0.973 |
JJ | +0.934 |
Hendricks | +0.888 |
Neg logit contributions | |
phe | -0.945 |
ū | -0.913 |
aug | -0.899 |
Ur | -0.897 |
kus | -0.857 |
Token | ( |
Token ID | 357 |
Feature activation | -4.20e-03 |
Pos logit contributions | |
Howell | +0.632 |
Birch | +0.548 |
Clarkson | +0.545 |
JJ | +0.523 |
Hendricks | +0.498 |
Neg logit contributions | |
phe | -0.546 |
ū | -0.530 |
Ur | -0.521 |
aug | -0.520 |
kus | -0.497 |
Token | the |
Token ID | 262 |
Feature activation | +7.44e-02 |
Pos logit contributions | |
Howell | +0.985 |
Birch | +0.861 |
Clarkson | +0.857 |
JJ | +0.823 |
Hendricks | +0.788 |
Neg logit contributions | |
phe | -0.754 |
ū | -0.732 |
aug | -0.724 |
Ur | -0.717 |
UG | -0.687 |
Token | Democratic |
Token ID | 4390 |
Feature activation | -1.28e-01 |
Pos logit contributions | |
phe | +0.546 |
ū | +0.534 |
aug | +0.529 |
Ur | +0.523 |
UG | +0.504 |
Neg logit contributions | |
Howell | -0.659 |
Birch | -0.574 |
Clarkson | -0.573 |
JJ | -0.548 |
Hendricks | -0.525 |
Token | National |
Token ID | 2351 |
Feature activation | -8.63e-02 |
Pos logit contributions | |
Howell | +0.932 |
Clarkson | +0.782 |
Birch | +0.781 |
JJ | +0.742 |
Hendricks | +0.693 |
Neg logit contributions | |
phe | -1.168 |
ū | -1.140 |
aug | -1.128 |
Ur | -1.124 |
kus | -1.085 |
Token | Committee |
Token ID | 4606 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
Howell | +0.879 |
Birch | +0.777 |
Clarkson | +0.774 |
JJ | +0.750 |
Hendricks | +0.716 |
Neg logit contributions | |
phe | -0.553 |
ū | -0.525 |
Ur | -0.518 |
aug | -0.512 |
UG | -0.490 |
Token | leak |
Token ID | 13044 |
Feature activation | +2.36e-01 |
Pos logit contributions | |
phe | +0.773 |
ū | +0.758 |
aug | +0.752 |
Ur | +0.748 |
UG | +0.714 |
Neg logit contributions | |
Howell | -0.886 |
Clarkson | -0.773 |
Birch | -0.772 |
JJ | -0.739 |
Hendricks | -0.704 |
Token |
|
Token ID | 198 |
Feature activation | +7.50e-01 |
Pos logit contributions | |
phe | +1.724 |
ū | +1.694 |
aug | +1.675 |
Ur | +1.670 |
UG | +1.585 |
Neg logit contributions | |
Howell | -2.099 |
Birch | -1.834 |
Clarkson | -1.829 |
JJ | -1.744 |
Hendricks | -1.690 |
Token |
|
Token ID | 198 |
Feature activation | +5.03e-01 |
Pos logit contributions | |
phe | +5.116 |
Ur | +5.052 |
aug | +4.997 |
ū | +4.985 |
UG | +4.740 |
Neg logit contributions | |
Howell | -6.891 |
Birch | -5.988 |
Clarkson | -5.910 |
JJ | -5.710 |
Hendricks | -5.462 |
Token | R |
Token ID | 49 |
Feature activation | -1.21e-01 |
Pos logit contributions | |
Ur | +2.736 |
phe | +2.644 |
aug | +2.622 |
ū | +2.545 |
UG | +2.535 |
Neg logit contributions | |
Howell | -5.522 |
Birch | -4.948 |
Clarkson | -4.918 |
JJ | -4.784 |
Hendricks | -4.605 |
Token | obby |
Token ID | 11369 |
Feature activation | +1.92e-01 |
Pos logit contributions | |
Howell | +1.241 |
Clarkson | +1.100 |
Birch | +1.099 |
JJ | +1.061 |
Hendricks | +1.020 |
Neg logit contributions | |
phe | -0.764 |
ū | -0.738 |
aug | -0.723 |
Ur | -0.718 |
UG | -0.685 |
Token | M |
Token ID | 337 |
Feature activation | -2.07e-01 |
Pos logit contributions | |
phe | +1.523 |
ū | +1.469 |
aug | +1.444 |
Ur | +1.431 |
kus | +1.377 |
Neg logit contributions | |
Howell | -1.603 |
Clarkson | -1.369 |
Birch | -1.362 |
JJ | -1.327 |
Hendricks | -1.226 |
Token | ook |
Token ID | 566 |
Feature activation | +5.03e-02 |
Pos logit contributions | |
Howell | +1.596 |
Birch | +1.361 |
Clarkson | +1.346 |
JJ | +1.270 |
Hendricks | +1.198 |
Neg logit contributions | |
phe | -1.809 |
ū | -1.734 |
Ur | -1.728 |
aug | -1.704 |
kus | -1.654 |
Token | this |
Token ID | 428 |
Feature activation | -2.39e-01 |
Pos logit contributions | |
Howell | +3.588 |
Birch | +3.145 |
Clarkson | +3.136 |
JJ | +3.023 |
Hendricks | +2.867 |
Neg logit contributions | |
phe | -2.816 |
ū | -2.710 |
Ur | -2.673 |
aug | -2.667 |
kus | -2.539 |
Token | webpage |
Token ID | 35699 |
Feature activation | +1.95e-01 |
Pos logit contributions | |
Howell | +2.031 |
Birch | +1.750 |
Clarkson | +1.744 |
JJ | +1.672 |
Hendricks | +1.583 |
Neg logit contributions | |
phe | -1.909 |
ū | -1.848 |
aug | -1.821 |
Ur | -1.819 |
kus | -1.743 |
Token | , |
Token ID | 11 |
Feature activation | +1.22e-01 |
Pos logit contributions | |
phe | +1.406 |
ū | +1.376 |
aug | +1.346 |
Ur | +1.345 |
ura | +1.275 |
Neg logit contributions | |
Howell | -1.791 |
Birch | -1.551 |
Clarkson | -1.530 |
JJ | -1.489 |
Hendricks | -1.416 |
Token | in |
Token ID | 287 |
Feature activation | -2.67e-01 |
Pos logit contributions | |
phe | +0.906 |
ū | +0.877 |
aug | +0.864 |
Ur | +0.864 |
kus | +0.824 |
Neg logit contributions | |
Howell | -1.091 |
Birch | -0.946 |
Clarkson | -0.939 |
JJ | -0.906 |
Hendricks | -0.859 |
Token | whole |
Token ID | 2187 |
Feature activation | -1.65e-01 |
Pos logit contributions | |
Howell | +2.414 |
Birch | +2.100 |
Clarkson | +2.092 |
JJ | +2.010 |
Hendricks | +1.911 |
Neg logit contributions | |
phe | -1.992 |
ū | -1.924 |
aug | -1.901 |
Ur | -1.894 |
kus | -1.810 |
Token | or |
Token ID | 393 |
Feature activation | -9.60e-01 |
Pos logit contributions | |
Howell | +1.526 |
Birch | +1.329 |
Clarkson | +1.311 |
JJ | +1.268 |
Hendricks | +1.219 |
Neg logit contributions | |
phe | -1.203 |
ū | -1.168 |
Ur | -1.149 |
aug | -1.146 |
kus | -1.092 |
Token | in |
Token ID | 287 |
Feature activation | -8.36e-01 |
Pos logit contributions | |
Howell | +7.115 |
Birch | +6.118 |
Clarkson | +6.067 |
JJ | +5.813 |
Hendricks | +5.495 |
Neg logit contributions | |
phe | -8.360 |
aug | -7.976 |
ū | -7.951 |
Ur | -7.877 |
anamo | -7.679 |
Token | part |
Token ID | 636 |
Feature activation | -6.60e-01 |
Pos logit contributions | |
Howell | +3.841 |
Birch | +2.903 |
Clarkson | +2.886 |
JJ | +2.761 |
Hendricks | +2.410 |
Neg logit contributions | |
phe | -9.685 |
ū | -9.400 |
aug | -9.268 |
Ur | -9.252 |
anamo | -9.214 |
Token | , |
Token ID | 11 |
Feature activation | -6.99e-02 |
Pos logit contributions | |
Howell | +5.532 |
Birch | +4.815 |
Clarkson | +4.812 |
JJ | +4.583 |
Hendricks | +4.372 |
Neg logit contributions | |
phe | -5.200 |
ū | -4.982 |
Ur | -4.947 |
aug | -4.939 |
kus | -4.756 |
Token | is |
Token ID | 318 |
Feature activation | -4.20e-01 |
Pos logit contributions | |
Howell | +0.630 |
Birch | +0.543 |
Clarkson | +0.540 |
JJ | +0.522 |
Hendricks | +0.495 |
Neg logit contributions | |
phe | -0.522 |
ū | -0.508 |
aug | -0.499 |
Ur | -0.499 |
kus | -0.475 |
Token | strictly |
Token ID | 14084 |
Feature activation | -1.43e-02 |
Pos logit contributions | |
Howell | +3.411 |
Clarkson | +2.930 |
Birch | +2.921 |
JJ | +2.792 |
Hendricks | +2.640 |
Neg logit contributions | |
phe | -3.514 |
ū | -3.381 |
aug | -3.330 |
Ur | -3.327 |
kus | -3.191 |
Token | any |
Token ID | 597 |
Feature activation | -2.72e-01 |
Pos logit contributions | |
Howell | +1.364 |
Birch | +1.191 |
Clarkson | +1.189 |
JJ | +1.141 |
Hendricks | +1.094 |
Neg logit contributions | |
phe | -1.033 |
ū | -0.996 |
aug | -0.978 |
Ur | -0.977 |
kus | -0.932 |
Token | antenna |
Token ID | 20509 |
Feature activation | -1.60e-01 |
Pos logit contributions | |
Howell | +2.403 |
Birch | +2.084 |
Clarkson | +2.078 |
JJ | +1.992 |
Hendricks | +1.894 |
Neg logit contributions | |
phe | -2.083 |
ū | -2.012 |
Ur | -1.981 |
aug | -1.979 |
kus | -1.895 |
Token | , |
Token ID | 11 |
Feature activation | -6.71e-02 |
Pos logit contributions | |
Howell | +1.423 |
Birch | +1.231 |
Clarkson | +1.230 |
JJ | +1.176 |
Hendricks | +1.123 |
Neg logit contributions | |
phe | -1.216 |
ū | -1.175 |
Ur | -1.161 |
aug | -1.160 |
UG | -1.106 |
Token | both |
Token ID | 1111 |
Feature activation | -1.06e-01 |
Pos logit contributions | |
Howell | +0.572 |
Birch | +0.493 |
Clarkson | +0.492 |
JJ | +0.469 |
Hendricks | +0.447 |
Neg logit contributions | |
phe | -0.532 |
ū | -0.512 |
Ur | -0.505 |
aug | -0.505 |
kus | -0.487 |
Token | external |
Token ID | 7097 |
Feature activation | -2.99e-01 |
Pos logit contributions | |
Howell | +0.930 |
Birch | +0.805 |
Clarkson | +0.803 |
JJ | +0.769 |
Hendricks | +0.732 |
Neg logit contributions | |
phe | -0.811 |
ū | -0.784 |
Ur | -0.772 |
aug | -0.771 |
kus | -0.739 |
Token | and |
Token ID | 290 |
Feature activation | -9.42e-01 |
Pos logit contributions | |
Howell | +2.576 |
Birch | +2.222 |
Clarkson | +2.217 |
JJ | +2.121 |
Hendricks | +2.016 |
Neg logit contributions | |
phe | -2.361 |
ū | -2.284 |
Ur | -2.250 |
aug | -2.249 |
kus | -2.153 |
Token | internal |
Token ID | 5387 |
Feature activation | -4.97e-01 |
Pos logit contributions | |
Howell | +7.531 |
Birch | +6.432 |
Clarkson | +6.300 |
JJ | +6.099 |
Gad | +5.778 |
Neg logit contributions | |
phe | -7.790 |
aug | -7.502 |
ū | -7.430 |
Ur | -7.417 |
anamo | -7.103 |
Token | , |
Token ID | 11 |
Feature activation | -2.74e-01 |
Pos logit contributions | |
Howell | +4.403 |
Birch | +3.827 |
Clarkson | +3.811 |
JJ | +3.674 |
Hendricks | +3.466 |
Neg logit contributions | |
phe | -3.767 |
ū | -3.642 |
aug | -3.595 |
Ur | -3.576 |
kus | -3.426 |
Token | and |
Token ID | 290 |
Feature activation | -1.72e-01 |
Pos logit contributions | |
Howell | +2.329 |
Birch | +2.005 |
Clarkson | +1.998 |
JJ | +1.915 |
Hendricks | +1.815 |
Neg logit contributions | |
phe | -2.190 |
ū | -2.120 |
aug | -2.091 |
Ur | -2.088 |
kus | -2.001 |
Token | note |
Token ID | 3465 |
Feature activation | +6.90e-02 |
Pos logit contributions | |
Howell | +1.516 |
Birch | +1.312 |
Clarkson | +1.307 |
JJ | +1.254 |
Hendricks | +1.194 |
Neg logit contributions | |
phe | -1.327 |
ū | -1.284 |
aug | -1.264 |
Ur | -1.263 |
kus | -1.212 |
Token | if |
Token ID | 611 |
Feature activation | -1.79e-01 |
Pos logit contributions | |
phe | +0.485 |
ū | +0.470 |
aug | +0.463 |
Ur | +0.461 |
UG | +0.443 |
Neg logit contributions | |
Howell | -0.644 |
Clarkson | -0.562 |
Birch | -0.562 |
JJ | -0.536 |
Hendricks | -0.518 |
Token | el |
Token ID | 417 |
Feature activation | -5.04e-01 |
Pos logit contributions | |
Howell | +4.758 |
Birch | +4.107 |
Clarkson | +4.099 |
JJ | +3.939 |
Hendricks | +3.738 |
Neg logit contributions | |
phe | -3.979 |
ū | -3.826 |
Ur | -3.783 |
aug | -3.763 |
UG | -3.619 |
Token | , |
Token ID | 11 |
Feature activation | -3.54e-01 |
Pos logit contributions | |
Howell | +4.471 |
Birch | +3.884 |
Clarkson | +3.850 |
JJ | +3.723 |
Hendricks | +3.529 |
Neg logit contributions | |
phe | -3.776 |
ū | -3.646 |
Ur | -3.619 |
aug | -3.617 |
UG | -3.445 |
Token | and |
Token ID | 290 |
Feature activation | -5.35e-01 |
Pos logit contributions | |
Howell | +3.067 |
Birch | +2.650 |
Clarkson | +2.641 |
JJ | +2.539 |
Hendricks | +2.403 |
Neg logit contributions | |
phe | -2.773 |
ū | -2.679 |
aug | -2.648 |
Ur | -2.639 |
kus | -2.531 |
Token | Schiff |
Token ID | 41665 |
Feature activation | -4.01e-01 |
Pos logit contributions | |
Howell | +4.589 |
Birch | +3.942 |
Clarkson | +3.937 |
JJ | +3.782 |
Hendricks | +3.604 |
Neg logit contributions | |
phe | -4.199 |
ū | -4.086 |
aug | -4.047 |
Ur | -4.030 |
UG | -3.862 |
Token | man |
Token ID | 805 |
Feature activation | -6.34e-01 |
Pos logit contributions | |
Howell | +3.256 |
Birch | +2.784 |
Clarkson | +2.783 |
JJ | +2.722 |
Hendricks | +2.529 |
Neg logit contributions | |
phe | -3.201 |
Ur | -3.111 |
ū | -3.092 |
aug | -3.022 |
UG | -2.997 |
Token | , |
Token ID | 11 |
Feature activation | -8.62e-01 |
Pos logit contributions | |
Howell | +5.373 |
Birch | +4.656 |
Clarkson | +4.627 |
JJ | +4.522 |
Hendricks | +4.194 |
Neg logit contributions | |
phe | -4.904 |
Ur | -4.733 |
ū | -4.709 |
aug | -4.703 |
UG | -4.526 |
Token | Outside |
Token ID | 22151 |
Feature activation | -2.56e-01 |
Pos logit contributions | |
Howell | +6.720 |
Birch | +5.714 |
Clarkson | +5.685 |
JJ | +5.626 |
Hendricks | +5.122 |
Neg logit contributions | |
phe | -7.216 |
ū | -6.970 |
aug | -6.951 |
Ur | -6.881 |
kus | -6.702 |
Token | the |
Token ID | 262 |
Feature activation | -2.31e-01 |
Pos logit contributions | |
Howell | +2.321 |
Birch | +2.020 |
Clarkson | +2.012 |
JJ | +1.937 |
Hendricks | +1.839 |
Neg logit contributions | |
phe | -1.906 |
ū | -1.839 |
aug | -1.813 |
Ur | -1.810 |
kus | -1.729 |
Token | Bible |
Token ID | 9111 |
Feature activation | -5.52e-01 |
Pos logit contributions | |
Howell | +1.868 |
Birch | +1.604 |
Clarkson | +1.580 |
JJ | +1.519 |
Hendricks | +1.435 |
Neg logit contributions | |
phe | -1.957 |
ū | -1.883 |
aug | -1.873 |
Ur | -1.864 |
kus | -1.795 |
Token | , |
Token ID | 11 |
Feature activation | -3.64e-01 |
Pos logit contributions | |
Howell | +5.085 |
Birch | +4.444 |
Clarkson | +4.419 |
JJ | +4.280 |
Hendricks | +4.065 |
Neg logit contributions | |
phe | -3.931 |
ū | -3.759 |
aug | -3.740 |
Ur | -3.725 |
kus | -3.561 |
Token | 3 |
Token ID | 513 |
Feature activation | -3.32e-01 |
Pos logit contributions | |
Howell | +3.198 |
Birch | +2.769 |
Clarkson | +2.760 |
JJ | +2.656 |
Hendricks | +2.515 |
Neg logit contributions | |
phe | -2.801 |
ū | -2.708 |
aug | -2.670 |
Ur | -2.664 |
kus | -2.562 |
Token |
|
Token ID | 198 |
Feature activation | +2.23e-02 |
Pos logit contributions | |
Howell | +0.437 |
Clarkson | +0.374 |
Birch | +0.373 |
JJ | +0.356 |
Hendricks | +0.336 |
Neg logit contributions | |
phe | -0.480 |
ū | -0.463 |
aug | -0.453 |
Ur | -0.452 |
anamo | -0.442 |
Token | Hall |
Token ID | 34194 |
Feature activation | -8.19e-02 |
Pos logit contributions | |
phe | +0.126 |
ū | +0.120 |
Ur | +0.119 |
aug | +0.118 |
UG | +0.110 |
Neg logit contributions | |
Howell | -0.243 |
Birch | -0.217 |
Clarkson | -0.215 |
JJ | -0.210 |
Hendricks | -0.201 |
Token | ion |
Token ID | 295 |
Feature activation | -2.72e-02 |
Pos logit contributions | |
Howell | +0.799 |
Clarkson | +0.700 |
Birch | +0.699 |
JJ | +0.674 |
Hendricks | +0.642 |
Neg logit contributions | |
phe | -0.555 |
ū | -0.536 |
aug | -0.529 |
Ur | -0.520 |
kus | -0.497 |
Token | , |
Token ID | 11 |
Feature activation | +8.62e-02 |
Pos logit contributions | |
Howell | +0.248 |
Birch | +0.216 |
Clarkson | +0.216 |
JJ | +0.208 |
Hendricks | +0.197 |
Neg logit contributions | |
phe | -0.199 |
ū | -0.192 |
aug | -0.189 |
Ur | -0.187 |
kus | -0.180 |
Token | Richard |
Token ID | 6219 |
Feature activation | -7.31e-01 |
Pos logit contributions | |
phe | +0.712 |
ū | +0.690 |
Ur | +0.684 |
aug | +0.682 |
UG | +0.653 |
Neg logit contributions | |
Howell | -0.689 |
Birch | -0.589 |
Clarkson | -0.587 |
JJ | -0.556 |
Hendricks | -0.528 |
Token | P |
Token ID | 350 |
Feature activation | -8.51e-01 |
Pos logit contributions | |
Howell | +6.485 |
Birch | +5.647 |
Clarkson | +5.640 |
JJ | +5.393 |
Hendricks | +5.157 |
Neg logit contributions | |
phe | -5.375 |
ū | -5.214 |
Ur | -5.194 |
aug | -5.131 |
anamo | -4.991 |
Token | . |
Token ID | 13 |
Feature activation | -3.22e-01 |
Pos logit contributions | |
Howell | +7.580 |
Clarkson | +6.630 |
Birch | +6.591 |
JJ | +6.329 |
Hendricks | +6.013 |
Neg logit contributions | |
phe | -5.959 |
Ur | -5.713 |
ū | -5.677 |
aug | -5.605 |
kus | -5.475 |
Token | Design |
Token ID | 8495 |
Feature activation | +5.60e-02 |
Pos logit contributions | |
Howell | +2.793 |
Birch | +2.415 |
Clarkson | +2.407 |
JJ | +2.310 |
Hendricks | +2.190 |
Neg logit contributions | |
phe | -2.520 |
ū | -2.435 |
aug | -2.400 |
Ur | -2.397 |
kus | -2.297 |
Token | ers |
Token ID | 364 |
Feature activation | +1.86e-01 |
Pos logit contributions | |
phe | +0.420 |
ū | +0.411 |
aug | +0.403 |
Ur | +0.401 |
kus | +0.384 |
Neg logit contributions | |
Howell | -0.502 |
Clarkson | -0.434 |
Birch | -0.433 |
JJ | -0.417 |
Hendricks | -0.395 |
Token | and |
Token ID | 290 |
Feature activation | -1.21e-01 |
Pos logit contributions | |
phe | +1.349 |
ū | +1.319 |
Ur | +1.282 |
aug | +1.282 |
kus | +1.226 |
Neg logit contributions | |
Howell | -1.694 |
Clarkson | -1.473 |
Birch | -1.473 |
JJ | -1.417 |
Hendricks | -1.347 |
Token | Test |
Token ID | 6208 |
Feature activation | +9.93e-03 |
Pos logit contributions | |
Howell | +1.087 |
Birch | +0.944 |
Clarkson | +0.942 |
JJ | +0.905 |
Hendricks | +0.860 |
Neg logit contributions | |
phe | -0.903 |
ū | -0.875 |
aug | -0.861 |
Ur | -0.859 |
kus | -0.822 |
Token | webpage |
Token ID | 35699 |
Feature activation | +1.95e-01 |
Pos logit contributions | |
Howell | +2.031 |
Birch | +1.750 |
Clarkson | +1.744 |
JJ | +1.672 |
Hendricks | +1.583 |
Neg logit contributions | |
phe | -1.909 |
ū | -1.848 |
aug | -1.821 |
Ur | -1.819 |
kus | -1.743 |
Token | , |
Token ID | 11 |
Feature activation | +1.22e-01 |
Pos logit contributions | |
phe | +1.406 |
ū | +1.376 |
aug | +1.346 |
Ur | +1.345 |
ura | +1.275 |
Neg logit contributions | |
Howell | -1.791 |
Birch | -1.551 |
Clarkson | -1.530 |
JJ | -1.489 |
Hendricks | -1.416 |
Token | in |
Token ID | 287 |
Feature activation | -2.67e-01 |
Pos logit contributions | |
phe | +0.906 |
ū | +0.877 |
aug | +0.864 |
Ur | +0.864 |
kus | +0.824 |
Neg logit contributions | |
Howell | -1.091 |
Birch | -0.946 |
Clarkson | -0.939 |
JJ | -0.906 |
Hendricks | -0.859 |
Token | whole |
Token ID | 2187 |
Feature activation | -1.65e-01 |
Pos logit contributions | |
Howell | +2.414 |
Birch | +2.100 |
Clarkson | +2.092 |
JJ | +2.010 |
Hendricks | +1.911 |
Neg logit contributions | |
phe | -1.992 |
ū | -1.924 |
aug | -1.901 |
Ur | -1.894 |
kus | -1.810 |
Token | or |
Token ID | 393 |
Feature activation | -9.60e-01 |
Pos logit contributions | |
Howell | +1.526 |
Birch | +1.329 |
Clarkson | +1.311 |
JJ | +1.268 |
Hendricks | +1.219 |
Neg logit contributions | |
phe | -1.203 |
ū | -1.168 |
Ur | -1.149 |
aug | -1.146 |
kus | -1.092 |
Token | in |
Token ID | 287 |
Feature activation | -8.36e-01 |
Pos logit contributions | |
Howell | +7.115 |
Birch | +6.118 |
Clarkson | +6.067 |
JJ | +5.813 |
Hendricks | +5.495 |
Neg logit contributions | |
phe | -8.360 |
aug | -7.976 |
ū | -7.951 |
Ur | -7.877 |
anamo | -7.679 |
Token | part |
Token ID | 636 |
Feature activation | -6.60e-01 |
Pos logit contributions | |
Howell | +3.841 |
Birch | +2.903 |
Clarkson | +2.886 |
JJ | +2.761 |
Hendricks | +2.410 |
Neg logit contributions | |
phe | -9.685 |
ū | -9.400 |
aug | -9.268 |
Ur | -9.252 |
anamo | -9.214 |
Token | , |
Token ID | 11 |
Feature activation | -6.99e-02 |
Pos logit contributions | |
Howell | +5.532 |
Birch | +4.815 |
Clarkson | +4.812 |
JJ | +4.583 |
Hendricks | +4.372 |
Neg logit contributions | |
phe | -5.200 |
ū | -4.982 |
Ur | -4.947 |
aug | -4.939 |
kus | -4.756 |
Token | is |
Token ID | 318 |
Feature activation | -4.20e-01 |
Pos logit contributions | |
Howell | +0.630 |
Birch | +0.543 |
Clarkson | +0.540 |
JJ | +0.522 |
Hendricks | +0.495 |
Neg logit contributions | |
phe | -0.522 |
ū | -0.508 |
aug | -0.499 |
Ur | -0.499 |
kus | -0.475 |
Token | strictly |
Token ID | 14084 |
Feature activation | -1.43e-02 |
Pos logit contributions | |
Howell | +3.411 |
Clarkson | +2.930 |
Birch | +2.921 |
JJ | +2.792 |
Hendricks | +2.640 |
Neg logit contributions | |
phe | -3.514 |
ū | -3.381 |
aug | -3.330 |
Ur | -3.327 |
kus | -3.191 |
Token | prohibited |
Token ID | 12244 |
Feature activation | +1.33e-01 |
Pos logit contributions | |
Howell | +0.132 |
Birch | +0.116 |
Clarkson | +0.115 |
JJ | +0.111 |
Hendricks | +0.105 |
Neg logit contributions | |
phe | -0.103 |
ū | -0.100 |
Ur | -0.099 |
aug | -0.099 |
UG | -0.094 |
Token | Follow |
Token ID | 7155 |
Feature activation | -2.30e-01 |
Pos logit contributions | |
Howell | +0.072 |
Birch | +0.064 |
Clarkson | +0.064 |
JJ | +0.062 |
Hendricks | +0.059 |
Neg logit contributions | |
phe | -0.041 |
ū | -0.040 |
aug | -0.039 |
Ur | -0.039 |
UG | -0.037 |
Token | Justin |
Token ID | 10799 |
Feature activation | +2.67e-01 |
Pos logit contributions | |
Howell | +1.918 |
Birch | +1.649 |
Clarkson | +1.641 |
JJ | +1.579 |
Hendricks | +1.486 |
Neg logit contributions | |
phe | -1.876 |
ū | -1.814 |
aug | -1.788 |
Ur | -1.785 |
kus | -1.714 |
Token | on |
Token ID | 319 |
Feature activation | -1.45e-01 |
Pos logit contributions | |
phe | +1.814 |
Ur | +1.726 |
aug | +1.726 |
ū | +1.721 |
kus | +1.636 |
Neg logit contributions | |
Howell | -2.503 |
Birch | -2.222 |
Clarkson | -2.220 |
JJ | -2.128 |
Hendricks | -2.012 |
Token | Twitter |
Token ID | 3009 |
Feature activation | -9.56e-02 |
Pos logit contributions | |
Howell | +1.260 |
Birch | +1.090 |
Clarkson | +1.087 |
JJ | +1.043 |
Hendricks | +0.989 |
Neg logit contributions | |
phe | -1.127 |
ū | -1.089 |
aug | -1.073 |
Ur | -1.072 |
kus | -1.026 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -9.91e-02 |
Pos logit contributions | |
Howell | +0.927 |
Birch | +0.814 |
Clarkson | +0.814 |
JJ | +0.783 |
Hendricks | +0.749 |
Neg logit contributions | |
phe | -0.647 |
ū | -0.623 |
Ur | -0.614 |
aug | -0.614 |
UG | -0.583 |
Token | by |
Token ID | 1525 |
Feature activation | -8.35e-01 |
Pos logit contributions | |
Howell | +1.061 |
Birch | +0.944 |
Clarkson | +0.941 |
JJ | +0.911 |
Hendricks | +0.874 |
Neg logit contributions | |
phe | -0.577 |
ū | -0.551 |
Ur | -0.543 |
aug | -0.542 |
UG | -0.509 |
Token | Julia |
Token ID | 22300 |
Feature activation | -2.78e-03 |
Pos logit contributions | |
Howell | +7.072 |
Birch | +6.106 |
Clarkson | +6.092 |
JJ | +5.929 |
Dalton | +5.510 |
Neg logit contributions | |
phe | -6.493 |
aug | -6.166 |
ū | -6.163 |
Ur | -6.086 |
kus | -5.921 |
Token | Py |
Token ID | 9485 |
Feature activation | +4.64e-02 |
Pos logit contributions | |
Howell | +0.023 |
Birch | +0.020 |
Clarkson | +0.020 |
JJ | +0.019 |
Hendricks | +0.018 |
Neg logit contributions | |
phe | -0.022 |
ū | -0.022 |
aug | -0.021 |
Ur | -0.021 |
UG | -0.020 |
Token | per |
Token ID | 525 |
Feature activation | +1.26e-01 |
Pos logit contributions | |
phe | +0.276 |
ū | +0.269 |
aug | +0.260 |
Ur | +0.258 |
kus | +0.250 |
Neg logit contributions | |
Howell | -0.490 |
Birch | -0.440 |
Clarkson | -0.437 |
JJ | -0.422 |
Hendricks | -0.403 |
Token |
|
Token ID | 198 |
Feature activation | -4.69e-01 |
Pos logit contributions | |
phe | +0.897 |
ū | +0.875 |
aug | +0.868 |
Ur | +0.850 |
kus | +0.830 |
Neg logit contributions | |
Howell | -1.155 |
Birch | -1.014 |
Clarkson | -1.011 |
JJ | -0.967 |
Hendricks | -0.927 |
Token |
|
Token ID | 198 |
Feature activation | -4.79e-01 |
Pos logit contributions | |
Howell | +3.272 |
Clarkson | +2.766 |
Birch | +2.735 |
JJ | +2.621 |
Hendricks | +2.431 |
Neg logit contributions | |
phe | -4.254 |
ū | -4.106 |
anamo | -4.096 |
aug | -3.988 |
Ur | -3.944 |
Token | Frank |
Token ID | 17439 |
Feature activation | +1.76e-01 |
Pos logit contributions | |
Howell | +0.066 |
Birch | +0.059 |
Clarkson | +0.059 |
JJ | +0.057 |
Hendricks | +0.055 |
Neg logit contributions | |
phe | -0.030 |
ū | -0.029 |
Ur | -0.029 |
aug | -0.028 |
UG | -0.027 |
Token | iv |
Token ID | 452 |
Feature activation | +2.87e-01 |
Pos logit contributions | |
phe | +1.617 |
ū | +1.609 |
aug | +1.589 |
Ur | +1.585 |
kus | +1.541 |
Neg logit contributions | |
Howell | -1.271 |
Birch | -1.060 |
Clarkson | -1.055 |
JJ | -0.993 |
Hendricks | -0.948 |
Token | sk |
Token ID | 8135 |
Feature activation | +3.93e-01 |
Pos logit contributions | |
phe | +1.055 |
ū | +0.998 |
Ur | +0.958 |
aug | +0.955 |
kus | +0.868 |
Neg logit contributions | |
Howell | -3.655 |
Birch | -3.332 |
Clarkson | -3.320 |
JJ | -3.234 |
Hendricks | -3.141 |
Token | . |
Token ID | 13 |
Feature activation | -7.17e-02 |
Pos logit contributions | |
phe | +2.820 |
ū | +2.739 |
aug | +2.704 |
Ur | +2.697 |
ogo | +2.574 |
Neg logit contributions | |
Howell | -3.587 |
Birch | -3.157 |
Clarkson | -3.112 |
JJ | -2.950 |
Hendricks | -2.876 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -2.37e-01 |
Pos logit contributions | |
Howell | +0.655 |
Birch | +0.572 |
Clarkson | +0.570 |
JJ | +0.547 |
Hendricks | +0.520 |
Neg logit contributions | |
phe | -0.521 |
ū | -0.503 |
Ur | -0.500 |
aug | -0.497 |
UG | -0.474 |
Token | [ |
Token ID | 58 |
Feature activation | -8.31e-01 |
Pos logit contributions | |
Howell | +2.524 |
Birch | +2.246 |
Clarkson | +2.242 |
JJ | +2.170 |
Hendricks | +2.082 |
Neg logit contributions | |
phe | -1.367 |
ū | -1.307 |
aug | -1.278 |
Ur | -1.273 |
kus | -1.204 |
Token | Ex |
Token ID | 3109 |
Feature activation | +1.31e-01 |
Pos logit contributions | |
Howell | +8.589 |
Birch | +7.692 |
Clarkson | +7.651 |
JJ | +7.422 |
Hendricks | +7.106 |
Neg logit contributions | |
phe | -4.654 |
ū | -4.446 |
aug | -4.248 |
Ur | -4.228 |
kus | -4.136 |
Token | clusive |
Token ID | 5731 |
Feature activation | +9.51e-02 |
Pos logit contributions | |
phe | +0.935 |
aug | +0.904 |
ū | +0.900 |
Ur | +0.895 |
kus | +0.848 |
Neg logit contributions | |
Howell | -1.224 |
Birch | -1.062 |
Clarkson | -1.059 |
JJ | -1.020 |
Hendricks | -0.978 |
Token | ] |
Token ID | 60 |
Feature activation | -1.21e-01 |
Pos logit contributions | |
phe | +0.505 |
ū | +0.480 |
aug | +0.473 |
Ur | +0.471 |
UG | +0.441 |
Neg logit contributions | |
Howell | -1.059 |
Birch | -0.945 |
Clarkson | -0.943 |
JJ | -0.915 |
Hendricks | -0.880 |
Token | Best |
Token ID | 6705 |
Feature activation | -1.41e-01 |
Pos logit contributions | |
Howell | +1.084 |
Birch | +0.939 |
Clarkson | +0.936 |
JJ | +0.897 |
Hendricks | +0.854 |
Neg logit contributions | |
phe | -0.916 |
ū | -0.882 |
Ur | -0.875 |
aug | -0.871 |
UG | -0.834 |
Token | Buy |
Token ID | 11763 |
Feature activation | -7.07e-04 |
Pos logit contributions | |
Howell | +1.280 |
Birch | +1.114 |
Clarkson | +1.110 |
JJ | +1.067 |
Hendricks | +1.017 |
Neg logit contributions | |
phe | -1.051 |
ū | -1.012 |
aug | -0.997 |
Ur | -0.994 |
kus | -0.953 |
Token | and |
Token ID | 290 |
Feature activation | -6.07e-01 |
Pos logit contributions | |
Howell | +1.632 |
Birch | +1.414 |
Clarkson | +1.410 |
JJ | +1.350 |
Hendricks | +1.285 |
Neg logit contributions | |
phe | -1.411 |
ū | -1.368 |
Ur | -1.345 |
aug | -1.344 |
kus | -1.287 |
Token | metal |
Token ID | 6147 |
Feature activation | +1.34e-01 |
Pos logit contributions | |
Howell | +5.606 |
Birch | +4.910 |
Clarkson | +4.882 |
JJ | +4.732 |
Hendricks | +4.465 |
Neg logit contributions | |
phe | -4.348 |
ū | -4.210 |
aug | -4.149 |
Ur | -4.116 |
UG | -3.928 |
Token | , |
Token ID | 11 |
Feature activation | -3.20e-02 |
Pos logit contributions | |
phe | +0.992 |
ū | +0.965 |
Ur | +0.947 |
aug | +0.945 |
UG | +0.910 |
Neg logit contributions | |
Howell | -1.210 |
Birch | -1.050 |
Clarkson | -1.050 |
JJ | -0.995 |
Hendricks | -0.959 |
Token | sing |
Token ID | 1702 |
Feature activation | +4.30e-02 |
Pos logit contributions | |
Howell | +0.282 |
Birch | +0.244 |
Clarkson | +0.243 |
JJ | +0.233 |
Hendricks | +0.222 |
Neg logit contributions | |
phe | -0.244 |
ū | -0.236 |
aug | -0.232 |
Ur | -0.232 |
kus | -0.223 |
Token | ly |
Token ID | 306 |
Feature activation | +3.39e-02 |
Pos logit contributions | |
phe | +0.228 |
ū | +0.218 |
aug | +0.213 |
Ur | +0.212 |
kus | +0.200 |
Neg logit contributions | |
Howell | -0.479 |
Birch | -0.432 |
Clarkson | -0.427 |
JJ | -0.415 |
Hendricks | -0.400 |
Token | or |
Token ID | 393 |
Feature activation | -8.18e-01 |
Pos logit contributions | |
phe | +0.248 |
ū | +0.240 |
aug | +0.236 |
Ur | +0.230 |
kus | +0.227 |
Neg logit contributions | |
Howell | -0.313 |
Birch | -0.273 |
Clarkson | -0.270 |
JJ | -0.260 |
Hendricks | -0.249 |
Token | in |
Token ID | 287 |
Feature activation | -3.59e-01 |
Pos logit contributions | |
Howell | +6.592 |
Clarkson | +5.613 |
Birch | +5.591 |
JJ | +5.454 |
Hendricks | +5.103 |
Neg logit contributions | |
phe | -6.629 |
ū | -6.435 |
aug | -6.408 |
Ur | -6.360 |
UG | -6.078 |
Token | combination |
Token ID | 6087 |
Feature activation | -2.40e-01 |
Pos logit contributions | |
Howell | +3.115 |
Birch | +2.690 |
Clarkson | +2.686 |
JJ | +2.583 |
Hendricks | +2.442 |
Neg logit contributions | |
phe | -2.794 |
ū | -2.701 |
aug | -2.658 |
Ur | -2.658 |
kus | -2.542 |
Token | , |
Token ID | 11 |
Feature activation | -4.47e-02 |
Pos logit contributions | |
Howell | +2.204 |
Birch | +1.924 |
Clarkson | +1.919 |
JJ | +1.846 |
Hendricks | +1.757 |
Neg logit contributions | |
phe | -1.755 |
ū | -1.690 |
aug | -1.666 |
Ur | -1.662 |
kus | -1.586 |
Token | every |
Token ID | 790 |
Feature activation | -3.15e-01 |
Pos logit contributions | |
Howell | +0.381 |
Birch | +0.327 |
Clarkson | +0.326 |
JJ | +0.311 |
Hendricks | +0.297 |
Neg logit contributions | |
phe | -0.355 |
ū | -0.344 |
aug | -0.339 |
Ur | -0.337 |
kus | -0.324 |
Token | Monday |
Token ID | 3321 |
Feature activation | -1.84e-01 |
Pos logit contributions | |
Howell | +2.872 |
Birch | +2.508 |
Clarkson | +2.497 |
JJ | +2.405 |
Hendricks | +2.285 |
Neg logit contributions | |
phe | -2.318 |
ū | -2.233 |
aug | -2.204 |
Ur | -2.193 |
kus | -2.093 |
Token | � |
Token ID | 447 |
Feature activation | +5.47e-02 |
Pos logit contributions | |
Howell | +1.161 |
Birch | +1.015 |
Clarkson | +1.012 |
JJ | +0.974 |
Hendricks | +0.930 |
Neg logit contributions | |
phe | -0.872 |
ū | -0.842 |
aug | -0.826 |
Ur | -0.826 |
UG | -0.784 |
Token | � |
Token ID | 247 |
Feature activation | -1.15e-02 |
Pos logit contributions | |
phe | +0.243 |
ū | +0.237 |
Ur | +0.226 |
aug | +0.225 |
ogo | +0.206 |
Neg logit contributions | |
Howell | -0.655 |
Birch | -0.593 |
Clarkson | -0.589 |
JJ | -0.569 |
Hendricks | -0.554 |
Token | ? |
Token ID | 30 |
Feature activation | -2.47e-01 |
Pos logit contributions | |
Howell | +0.108 |
Birch | +0.095 |
Clarkson | +0.094 |
JJ | +0.091 |
Hendricks | +0.087 |
Neg logit contributions | |
phe | -0.082 |
ū | -0.079 |
aug | -0.078 |
Ur | -0.078 |
UG | -0.074 |
Token |
|
Token ID | 198 |
Feature activation | -8.67e-02 |
Pos logit contributions | |
Howell | +2.110 |
Birch | +1.819 |
Clarkson | +1.814 |
JJ | +1.738 |
Hendricks | +1.647 |
Neg logit contributions | |
phe | -1.962 |
ū | -1.898 |
aug | -1.871 |
Ur | -1.869 |
kus | -1.790 |
Token |
|
Token ID | 198 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
Howell | +0.673 |
Clarkson | +0.575 |
Birch | +0.573 |
JJ | +0.548 |
Hendricks | +0.515 |
Neg logit contributions | |
phe | -0.753 |
ū | -0.728 |
Ur | -0.712 |
aug | -0.711 |
anamo | -0.699 |
Token | and |
Token ID | 392 |
Feature activation | -8.13e-01 |
Pos logit contributions | |
Howell | +1.435 |
Birch | +1.271 |
Clarkson | +1.268 |
JJ | +1.226 |
Hendricks | +1.175 |
Neg logit contributions | |
phe | -0.859 |
ū | -0.823 |
aug | -0.808 |
Ur | -0.807 |
kus | -0.762 |
Token | how |
Token ID | 703 |
Feature activation | -4.07e-01 |
Pos logit contributions | |
Howell | +5.732 |
Birch | +4.787 |
Clarkson | +4.695 |
JJ | +4.623 |
Hendricks | +4.226 |
Neg logit contributions | |
phe | -7.368 |
ū | -7.146 |
aug | -7.106 |
Ur | -7.094 |
anamo | -6.885 |
Token | does |
Token ID | 857 |
Feature activation | -5.58e-01 |
Pos logit contributions | |
Howell | +3.739 |
Clarkson | +3.256 |
Birch | +3.246 |
JJ | +3.146 |
Hendricks | +2.978 |
Neg logit contributions | |
phe | -2.966 |
ū | -2.841 |
Ur | -2.807 |
aug | -2.796 |
kus | -2.684 |
Token | it |
Token ID | 340 |
Feature activation | -4.15e-01 |
Pos logit contributions | |
Howell | +3.903 |
Birch | +3.304 |
Clarkson | +3.231 |
JJ | +3.133 |
Hendricks | +2.850 |
Neg logit contributions | |
phe | -5.274 |
ū | -4.980 |
anamo | -4.948 |
aug | -4.943 |
Ur | -4.924 |
Token | look |
Token ID | 804 |
Feature activation | -5.29e-01 |
Pos logit contributions | |
Howell | +2.813 |
Clarkson | +2.334 |
Birch | +2.303 |
JJ | +2.243 |
Hendricks | +2.004 |
Neg logit contributions | |
phe | -3.988 |
ū | -3.860 |
Ur | -3.841 |
aug | -3.807 |
anamo | -3.783 |
Token | from |
Token ID | 422 |
Feature activation | -4.05e-01 |
Pos logit contributions | |
Howell | +4.064 |
Clarkson | +3.449 |
Birch | +3.441 |
JJ | +3.269 |
Hendricks | +3.078 |
Neg logit contributions | |
phe | -4.586 |
ū | -4.418 |
aug | -4.404 |
Ur | -4.356 |
anamo | -4.335 |
Token | the |
Token ID | 262 |
Feature activation | -2.46e-01 |
Pos logit contributions | |
Howell | +1.198 |
Birch | +1.039 |
Clarkson | +1.032 |
JJ | +0.993 |
Hendricks | +0.940 |
Neg logit contributions | |
phe | -1.094 |
ū | -1.056 |
Ur | -1.041 |
aug | -1.036 |
UG | -0.993 |
Token | flow |
Token ID | 5202 |
Feature activation | -2.83e-01 |
Pos logit contributions | |
Howell | +2.051 |
Birch | +1.762 |
Clarkson | +1.755 |
JJ | +1.682 |
Hendricks | +1.590 |
Neg logit contributions | |
phe | -2.002 |
ū | -1.939 |
aug | -1.911 |
Ur | -1.910 |
kus | -1.831 |
Token | of |
Token ID | 286 |
Feature activation | -4.81e-01 |
Pos logit contributions | |
Howell | +2.473 |
Birch | +2.141 |
Clarkson | +2.135 |
JJ | +2.050 |
Hendricks | +1.944 |
Neg logit contributions | |
phe | -2.191 |
ū | -2.120 |
aug | -2.094 |
Ur | -2.086 |
kus | -2.000 |
Token | on |
Token ID | 319 |
Feature activation | -3.99e-01 |
Pos logit contributions | |
Howell | +3.874 |
Clarkson | +3.310 |
Birch | +3.301 |
JJ | +3.163 |
Hendricks | +2.986 |
Neg logit contributions | |
phe | -4.061 |
ū | -3.908 |
aug | -3.881 |
Ur | -3.853 |
kus | -3.740 |
Token | coming |
Token ID | 4976 |
Feature activation | -4.08e-01 |
Pos logit contributions | |
Howell | +3.292 |
Birch | +2.829 |
Clarkson | +2.814 |
JJ | +2.687 |
Hendricks | +2.547 |
Neg logit contributions | |
phe | -3.271 |
ū | -3.168 |
Ur | -3.121 |
aug | -3.120 |
kus | -3.003 |
Token | cars |
Token ID | 5006 |
Feature activation | -8.06e-01 |
Pos logit contributions | |
Howell | +3.057 |
Birch | +2.571 |
Clarkson | +2.570 |
JJ | +2.451 |
Hendricks | +2.294 |
Neg logit contributions | |
phe | -3.686 |
ū | -3.558 |
aug | -3.523 |
Ur | -3.510 |
kus | -3.398 |
Token | and |
Token ID | 290 |
Feature activation | -5.02e-01 |
Pos logit contributions | |
Howell | +6.294 |
Clarkson | +5.416 |
Birch | +5.334 |
JJ | +5.178 |
Hendricks | +4.888 |
Neg logit contributions | |
phe | -6.650 |
aug | -6.412 |
ū | -6.398 |
Ur | -6.350 |
kus | -6.250 |
Token | pedestrians |
Token ID | 26735 |
Feature activation | -4.86e-01 |
Pos logit contributions | |
Howell | +4.369 |
Clarkson | +3.764 |
Birch | +3.764 |
JJ | +3.612 |
Hendricks | +3.425 |
Neg logit contributions | |
phe | -3.876 |
ū | -3.754 |
aug | -3.727 |
Ur | -3.684 |
kus | -3.548 |
Token | to |
Token ID | 284 |
Feature activation | -3.68e-01 |
Pos logit contributions | |
Howell | +4.186 |
Clarkson | +3.607 |
Birch | +3.601 |
JJ | +3.472 |
Hendricks | +3.279 |
Neg logit contributions | |
phe | -3.776 |
ū | -3.643 |
aug | -3.612 |
Ur | -3.603 |
kus | -3.493 |
Token | make |
Token ID | 787 |
Feature activation | -4.70e-02 |
Pos logit contributions | |
Howell | +3.227 |
Birch | +2.791 |
Clarkson | +2.786 |
JJ | +2.674 |
Hendricks | +2.534 |
Neg logit contributions | |
phe | -2.847 |
ū | -2.748 |
aug | -2.711 |
Ur | -2.702 |
kus | -2.586 |
Token | a |
Token ID | 257 |
Feature activation | +2.48e-02 |
Pos logit contributions | |
Howell | +0.394 |
Birch | +0.342 |
Clarkson | +0.338 |
JJ | +0.325 |
Hendricks | +0.308 |
Neg logit contributions | |
phe | -0.376 |
ū | -0.366 |
Ur | -0.360 |
aug | -0.359 |
kus | -0.343 |
Token |
|
Token ID | 198 |
Feature activation | -2.24e-01 |
Pos logit contributions | |
Howell | +1.811 |
Birch | +1.562 |
Clarkson | +1.559 |
JJ | +1.492 |
Hendricks | +1.417 |
Neg logit contributions | |
phe | -1.648 |
ū | -1.595 |
aug | -1.572 |
Ur | -1.571 |
kus | -1.502 |
Token |
|
Token ID | 198 |
Feature activation | -2.72e-01 |
Pos logit contributions | |
Howell | +1.672 |
Clarkson | +1.426 |
Birch | +1.415 |
JJ | +1.355 |
Hendricks | +1.266 |
Neg logit contributions | |
phe | -2.005 |
ū | -1.942 |
anamo | -1.896 |
Ur | -1.889 |
aug | -1.888 |
Token | ( |
Token ID | 7 |
Feature activation | -5.88e-01 |
Pos logit contributions | |
Howell | +2.837 |
Birch | +2.522 |
Clarkson | +2.512 |
JJ | +2.434 |
Hendricks | +2.328 |
Neg logit contributions | |
phe | -1.623 |
ū | -1.554 |
aug | -1.516 |
Ur | -1.513 |
kus | -1.437 |
Token | i |
Token ID | 72 |
Feature activation | -4.07e-01 |
Pos logit contributions | |
Howell | +5.367 |
Birch | +4.692 |
Clarkson | +4.690 |
JJ | +4.528 |
Hendricks | +4.294 |
Neg logit contributions | |
phe | -4.152 |
ū | -4.001 |
aug | -3.906 |
Ur | -3.886 |
kus | -3.741 |
Token | . |
Token ID | 13 |
Feature activation | -1.45e-01 |
Pos logit contributions | |
Howell | +3.254 |
Clarkson | +2.807 |
Birch | +2.805 |
JJ | +2.693 |
Hendricks | +2.545 |
Neg logit contributions | |
phe | -3.362 |
ū | -3.181 |
anamo | -3.175 |
aug | -3.114 |
Ur | -3.101 |
Token | e |
Token ID | 68 |
Feature activation | -7.82e-01 |
Pos logit contributions | |
Howell | +1.232 |
Clarkson | +1.062 |
Birch | +1.057 |
JJ | +1.012 |
Hendricks | +0.967 |
Neg logit contributions | |
phe | -1.127 |
ū | -1.090 |
Ur | -1.076 |
aug | -1.066 |
anamo | -1.051 |
Token | . |
Token ID | 13 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
Howell | +5.770 |
Birch | +5.005 |
Clarkson | +4.974 |
JJ | +4.696 |
Hendricks | +4.471 |
Neg logit contributions | |
phe | -6.557 |
anamo | -6.387 |
ū | -6.301 |
Ur | -6.199 |
aug | -6.158 |
Token | is |
Token ID | 318 |
Feature activation | -4.53e-01 |
Pos logit contributions | |
Howell | +1.229 |
Clarkson | +1.061 |
Birch | +1.058 |
JJ | +1.016 |
Hendricks | +0.967 |
Neg logit contributions | |
phe | -1.070 |
ū | -1.041 |
Ur | -1.020 |
aug | -1.019 |
kus | -0.971 |
Token | the |
Token ID | 262 |
Feature activation | -3.04e-01 |
Pos logit contributions | |
Howell | +3.588 |
Birch | +3.101 |
Clarkson | +3.052 |
JJ | +2.969 |
Hendricks | +2.754 |
Neg logit contributions | |
phe | -3.824 |
aug | -3.610 |
Ur | -3.602 |
ū | -3.592 |
anamo | -3.577 |
Token | syntax |
Token ID | 15582 |
Feature activation | +5.30e-02 |
Pos logit contributions | |
Howell | +2.425 |
Birch | +2.113 |
Clarkson | +2.084 |
JJ | +2.010 |
Hendricks | +1.893 |
Neg logit contributions | |
phe | -2.571 |
aug | -2.460 |
Ur | -2.449 |
ū | -2.443 |
kus | -2.372 |
Token | just |
Token ID | 655 |
Feature activation | +8.72e-03 |
Pos logit contributions | |
phe | +0.423 |
ū | +0.405 |
Ur | +0.401 |
aug | +0.401 |
UG | +0.383 |
Neg logit contributions | |
Howell | -0.454 |
Birch | -0.393 |
Clarkson | -0.391 |
JJ | -0.376 |
Hendricks | -0.355 |
Token | to |
Token ID | 284 |
Feature activation | -7.28e-03 |
Pos logit contributions | |
phe | +1.192 |
ū | +1.160 |
Ur | +1.139 |
aug | +1.135 |
UG | +1.093 |
Neg logit contributions | |
Howell | -1.388 |
Birch | -1.211 |
Clarkson | -1.201 |
JJ | -1.148 |
Hendricks | -1.099 |
Token | behave |
Token ID | 17438 |
Feature activation | +2.87e-01 |
Pos logit contributions | |
Howell | +0.064 |
Birch | +0.055 |
Clarkson | +0.055 |
JJ | +0.052 |
Hendricks | +0.050 |
Neg logit contributions | |
phe | -0.056 |
ū | -0.054 |
Ur | -0.054 |
aug | -0.053 |
kus | -0.052 |
Token | properly |
Token ID | 6105 |
Feature activation | +1.93e-01 |
Pos logit contributions | |
phe | +2.183 |
ū | +2.137 |
Ur | +2.112 |
aug | +2.092 |
UG | +2.028 |
Neg logit contributions | |
Howell | -2.452 |
Birch | -2.142 |
Clarkson | -2.102 |
JJ | -2.027 |
Hendricks | -1.941 |
Token | . |
Token ID | 13 |
Feature activation | -6.43e-02 |
Pos logit contributions | |
phe | +1.396 |
ū | +1.360 |
aug | +1.330 |
Ur | +1.330 |
UG | +1.282 |
Neg logit contributions | |
Howell | -1.760 |
Birch | -1.536 |
Clarkson | -1.524 |
JJ | -1.467 |
Hendricks | -1.402 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -1.85e-01 |
Pos logit contributions | |
Howell | +0.590 |
Birch | +0.514 |
Clarkson | +0.513 |
JJ | +0.493 |
Hendricks | +0.469 |
Neg logit contributions | |
phe | -0.469 |
ū | -0.453 |
Ur | -0.446 |
aug | -0.446 |
UG | -0.425 |
Token | by |
Token ID | 1525 |
Feature activation | -7.81e-01 |
Pos logit contributions | |
Howell | +1.981 |
Birch | +1.763 |
Clarkson | +1.758 |
JJ | +1.702 |
Hendricks | +1.633 |
Neg logit contributions | |
phe | -1.075 |
ū | -1.027 |
aug | -1.007 |
Ur | -1.006 |
kus | -0.946 |
Token | Erik |
Token ID | 22722 |
Feature activation | +1.57e-01 |
Pos logit contributions | |
Howell | +6.610 |
Birch | +5.697 |
Clarkson | +5.677 |
JJ | +5.524 |
Hendricks | +5.144 |
Neg logit contributions | |
phe | -6.131 |
ū | -5.841 |
aug | -5.829 |
Ur | -5.761 |
kus | -5.585 |
Token | Alt |
Token ID | 12344 |
Feature activation | +8.86e-02 |
Pos logit contributions | |
phe | +1.217 |
ū | +1.179 |
aug | +1.161 |
Ur | +1.157 |
kus | +1.109 |
Neg logit contributions | |
Howell | -1.337 |
Clarkson | -1.156 |
Birch | -1.149 |
JJ | -1.113 |
Hendricks | -1.040 |
Token | ieri |
Token ID | 29864 |
Feature activation | +3.10e-01 |
Pos logit contributions | |
phe | +0.889 |
ū | +0.874 |
aug | +0.865 |
ura | +0.843 |
kus | +0.842 |
Neg logit contributions | |
Howell | -0.579 |
Birch | -0.467 |
Clarkson | -0.464 |
JJ | -0.439 |
Hendricks | -0.400 |
Token | , |
Token ID | 11 |
Feature activation | -1.27e-01 |
Pos logit contributions | |
phe | +2.193 |
ū | +2.185 |
aug | +2.138 |
Ur | +2.094 |
UG | +2.043 |
Neg logit contributions | |
Howell | -2.810 |
Birch | -2.445 |
Clarkson | -2.419 |
JJ | -2.324 |
Hendricks | -2.226 |
Token | NOR |
Token ID | 25273 |
Feature activation | -1.23e-01 |
Pos logit contributions | |
Howell | +1.116 |
Birch | +0.965 |
Clarkson | +0.962 |
JJ | +0.922 |
Hendricks | +0.878 |
Neg logit contributions | |
phe | -0.971 |
ū | -0.938 |
aug | -0.925 |
Ur | -0.924 |
kus | -0.884 |
Token | by |
Token ID | 416 |
Feature activation | -2.97e-01 |
Pos logit contributions | |
phe | +0.038 |
ū | +0.037 |
aug | +0.037 |
Ur | +0.037 |
UG | +0.035 |
Neg logit contributions | |
Howell | -0.044 |
Birch | -0.038 |
Clarkson | -0.038 |
JJ | -0.037 |
Hendricks | -0.035 |
Token | | |
Token ID | 930 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
Howell | +2.575 |
Birch | +2.227 |
Clarkson | +2.218 |
JJ | +2.132 |
Hendricks | +2.017 |
Neg logit contributions | |
phe | -2.326 |
ū | -2.247 |
aug | -2.215 |
Ur | -2.212 |
kus | -2.119 |
Token | in |
Token ID | 287 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
Howell | +1.560 |
Birch | +1.349 |
Clarkson | +1.344 |
JJ | +1.289 |
Hendricks | +1.222 |
Neg logit contributions | |
phe | -1.405 |
ū | -1.361 |
aug | -1.342 |
Ur | -1.340 |
UG | -1.281 |
Token |
|
Token ID | 198 |
Feature activation | -2.17e-01 |
Pos logit contributions | |
Howell | +1.636 |
Birch | +1.424 |
Clarkson | +1.420 |
JJ | +1.365 |
Hendricks | +1.298 |
Neg logit contributions | |
phe | -1.339 |
ū | -1.292 |
aug | -1.272 |
Ur | -1.271 |
kus | -1.213 |
Token |
|
Token ID | 198 |
Feature activation | -2.84e-01 |
Pos logit contributions | |
Howell | +1.636 |
Clarkson | +1.393 |
Birch | +1.381 |
JJ | +1.325 |
Hendricks | +1.244 |
Neg logit contributions | |
phe | -1.921 |
ū | -1.860 |
aug | -1.812 |
anamo | -1.805 |
Ur | -1.797 |
Token | | |
Token ID | 91 |
Feature activation | -7.81e-01 |
Pos logit contributions | |
Howell | +3.001 |
Clarkson | +2.666 |
Birch | +2.665 |
JJ | +2.582 |
Hendricks | +2.473 |
Neg logit contributions | |
phe | -1.652 |
ū | -1.580 |
aug | -1.539 |
Ur | -1.523 |
kus | -1.456 |
Token | in |
Token ID | 287 |
Feature activation | -2.65e-01 |
Pos logit contributions | |
Howell | +7.029 |
Birch | +6.114 |
Clarkson | +6.113 |
JJ | +5.926 |
Hendricks | +5.562 |
Neg logit contributions | |
phe | -5.642 |
ū | -5.337 |
Ur | -5.306 |
aug | -5.251 |
kus | -5.100 |
Token | babys |
Token ID | 46711 |
Feature activation | -9.89e-02 |
Pos logit contributions | |
Howell | +2.401 |
Birch | +2.090 |
Clarkson | +2.083 |
JJ | +2.004 |
Hendricks | +1.905 |
Neg logit contributions | |
phe | -1.964 |
ū | -1.894 |
aug | -1.864 |
Ur | -1.863 |
kus | -1.780 |
Token | itting |
Token ID | 2535 |
Feature activation | -3.07e-01 |
Pos logit contributions | |
Howell | +1.005 |
Birch | +0.890 |
Clarkson | +0.885 |
JJ | +0.856 |
Hendricks | +0.825 |
Neg logit contributions | |
phe | -0.629 |
ū | -0.607 |
aug | -0.595 |
Ur | -0.592 |
UG | -0.561 |
Token |
|
Token ID | 198 |
Feature activation | -1.21e-01 |
Pos logit contributions | |
Howell | +2.857 |
Birch | +2.496 |
Clarkson | +2.489 |
JJ | +2.397 |
Hendricks | +2.279 |
Neg logit contributions | |
phe | -2.210 |
ū | -2.127 |
aug | -2.095 |
Ur | -2.092 |
kus | -1.993 |
Token |
|
Token ID | 198 |
Feature activation | -6.03e-02 |
Pos logit contributions | |
Howell | +0.937 |
Clarkson | +0.798 |
Birch | +0.795 |
JJ | +0.762 |
Hendricks | +0.715 |
Neg logit contributions | |
phe | -1.061 |
ū | -1.025 |
Ur | -1.002 |
aug | -1.001 |
anamo | -0.988 |
Token | ` |
Token ID | 63 |
Feature activation | -1.25e-01 |
Pos logit contributions | |
Howell | +1.278 |
Birch | +1.122 |
Clarkson | +1.118 |
JJ | +1.073 |
Hendricks | +1.030 |
Neg logit contributions | |
phe | -0.906 |
ū | -0.871 |
Ur | -0.863 |
aug | -0.857 |
UG | -0.814 |
Token | represent |
Token ID | 2380 |
Feature activation | -2.05e-01 |
Pos logit contributions | |
Howell | +1.104 |
Birch | +0.960 |
Clarkson | +0.956 |
JJ | +0.913 |
Hendricks | +0.871 |
Neg logit contributions | |
phe | -0.956 |
ū | -0.924 |
aug | -0.913 |
Ur | -0.912 |
UG | -0.869 |
Token | able |
Token ID | 540 |
Feature activation | -1.68e-01 |
Pos logit contributions | |
Howell | +1.847 |
Birch | +1.606 |
Clarkson | +1.601 |
JJ | +1.542 |
Hendricks | +1.462 |
Neg logit contributions | |
phe | -1.532 |
ū | -1.479 |
aug | -1.455 |
Ur | -1.453 |
kus | -1.389 |
Token |
|
Token ID | 198 |
Feature activation | -9.50e-02 |
Pos logit contributions | |
Howell | +1.529 |
Birch | +1.332 |
Clarkson | +1.328 |
JJ | +1.277 |
Hendricks | +1.215 |
Neg logit contributions | |
phe | -1.235 |
ū | -1.192 |
aug | -1.174 |
Ur | -1.172 |
kus | -1.119 |
Token |
|
Token ID | 198 |
Feature activation | -6.80e-03 |
Pos logit contributions | |
Howell | +0.731 |
Clarkson | +0.624 |
Birch | +0.622 |
JJ | +0.593 |
Hendricks | +0.557 |
Neg logit contributions | |
phe | -0.834 |
ū | -0.804 |
aug | -0.787 |
Ur | -0.784 |
anamo | -0.777 |
Token | :( |
Token ID | 37498 |
Feature activation | -7.76e-01 |
Pos logit contributions | |
Howell | +0.073 |
Birch | +0.065 |
Clarkson | +0.065 |
JJ | +0.063 |
Hendricks | +0.061 |
Neg logit contributions | |
phe | -0.039 |
ū | -0.037 |
Ur | -0.037 |
aug | -0.037 |
UG | -0.034 |
Token |
|
Token ID | 198 |
Feature activation | -2.21e-01 |
Pos logit contributions | |
Howell | +7.730 |
Birch | +6.862 |
Clarkson | +6.821 |
JJ | +6.663 |
Hendricks | +6.287 |
Neg logit contributions | |
phe | -4.757 |
ū | -4.546 |
aug | -4.382 |
Ur | -4.339 |
kus | -4.271 |
Token |
|
Token ID | 198 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
Howell | +1.635 |
Clarkson | +1.391 |
Birch | +1.384 |
JJ | +1.320 |
Hendricks | +1.234 |
Neg logit contributions | |
phe | -1.984 |
ū | -1.910 |
anamo | -1.872 |
aug | -1.863 |
Ur | -1.858 |
Token | I |
Token ID | 40 |
Feature activation | -3.56e-03 |
Pos logit contributions | |
Howell | +1.401 |
Birch | +1.239 |
Clarkson | +1.235 |
JJ | +1.193 |
Hendricks | +1.142 |
Neg logit contributions | |
phe | -0.877 |
ū | -0.841 |
aug | -0.826 |
Ur | -0.825 |
kus | -0.780 |
Token | � |
Token ID | 447 |
Feature activation | -1.78e-01 |
Pos logit contributions | |
Howell | +0.033 |
Clarkson | +0.028 |
Birch | +0.028 |
JJ | +0.027 |
Hendricks | +0.026 |
Neg logit contributions | |
phe | -0.026 |
ū | -0.025 |
Ur | -0.025 |
aug | -0.025 |
UG | -0.024 |
Token | � |
Token ID | 247 |
Feature activation | -3.16e-01 |
Pos logit contributions | |
Howell | +2.205 |
Birch | +1.999 |
Clarkson | +1.984 |
JJ | +1.943 |
Hendricks | +1.870 |
Neg logit contributions | |
phe | -0.738 |
ū | -0.667 |
aug | -0.661 |
Ur | -0.652 |
kus | -0.609 |
Token | o |
Token ID | 78 |
Feature activation | +1.38e-01 |
Pos logit contributions | |
phe | +0.487 |
ū | +0.471 |
aug | +0.458 |
Ur | +0.454 |
kus | +0.428 |
Neg logit contributions | |
Howell | -0.888 |
Birch | -0.792 |
Clarkson | -0.792 |
JJ | -0.763 |
Hendricks | -0.735 |
Token |
|
Token ID | 198 |
Feature activation | -1.26e-01 |
Pos logit contributions | |
phe | +1.028 |
ū | +1.004 |
Ur | +0.978 |
aug | +0.977 |
kus | +0.935 |
Neg logit contributions | |
Howell | -1.244 |
Birch | -1.090 |
Clarkson | -1.085 |
JJ | -1.036 |
Hendricks | -0.993 |
Token |
|
Token ID | 198 |
Feature activation | -9.68e-02 |
Pos logit contributions | |
Howell | +0.977 |
Clarkson | +0.831 |
Birch | +0.829 |
JJ | +0.792 |
Hendricks | +0.745 |
Neg logit contributions | |
phe | -1.104 |
ū | -1.062 |
aug | -1.043 |
Ur | -1.039 |
anamo | -1.031 |
Token | ko |
Token ID | 7204 |
Feature activation | +3.04e-04 |
Pos logit contributions | |
Howell | +1.030 |
Birch | +0.917 |
Clarkson | +0.914 |
JJ | +0.885 |
Hendricks | +0.849 |
Neg logit contributions | |
phe | -0.566 |
ū | -0.541 |
aug | -0.530 |
Ur | -0.530 |
kus | -0.499 |
Token | Kal |
Token ID | 12612 |
Feature activation | +5.52e-03 |
Pos logit contributions | |
phe | +0.002 |
ū | +0.002 |
aug | +0.002 |
Ur | +0.002 |
kus | +0.002 |
Neg logit contributions | |
Howell | -0.003 |
Birch | -0.003 |
Clarkson | -0.002 |
JJ | -0.002 |
Hendricks | -0.002 |
Token | an |
Token ID | 272 |
Feature activation | -7.69e-01 |
Pos logit contributions | |
phe | +0.024 |
ū | +0.022 |
aug | +0.022 |
Ur | +0.021 |
kus | +0.020 |
Neg logit contributions | |
Howell | -0.068 |
Birch | -0.061 |
Clarkson | -0.061 |
JJ | -0.059 |
Hendricks | -0.057 |
Token | ial |
Token ID | 498 |
Feature activation | -3.43e-01 |
Pos logit contributions | |
Howell | +6.600 |
Birch | +5.713 |
Clarkson | +5.603 |
JJ | +5.557 |
Hendricks | +5.112 |
Neg logit contributions | |
phe | -5.830 |
aug | -5.456 |
Ur | -5.415 |
ū | -5.381 |
UG | -5.294 |
Token | i |
Token ID | 72 |
Feature activation | -2.89e-01 |
Pos logit contributions | |
Howell | +3.099 |
Birch | +2.692 |
Clarkson | +2.673 |
JJ | +2.589 |
Hendricks | +2.449 |
Neg logit contributions | |
phe | -2.545 |
ū | -2.439 |
aug | -2.414 |
Ur | -2.413 |
UG | -2.311 |
Token | � |
Token ID | 134 |
Feature activation | -6.05e-02 |
Pos logit contributions | |
Howell | +3.126 |
Birch | +2.785 |
Clarkson | +2.775 |
JJ | +2.695 |
Hendricks | +2.583 |
Neg logit contributions | |
phe | -1.632 |
ū | -1.540 |
aug | -1.517 |
Ur | -1.516 |
kus | -1.425 |
Token | � |
Token ID | 119 |
Feature activation | -2.58e-01 |
Pos logit contributions | |
Howell | +0.516 |
Birch | +0.449 |
Clarkson | +0.442 |
JJ | +0.426 |
Hendricks | +0.407 |
Neg logit contributions | |
phe | -0.481 |
aug | -0.455 |
ū | -0.454 |
Ur | -0.449 |
anamo | -0.441 |
Token | il |
Token ID | 346 |
Feature activation | +2.39e-01 |
Pos logit contributions | |
Howell | +2.924 |
Birch | +2.621 |
Clarkson | +2.608 |
JJ | +2.536 |
Hendricks | +2.437 |
Neg logit contributions | |
phe | -1.323 |
ū | -1.251 |
aug | -1.226 |
Ur | -1.224 |
kus | -1.143 |
Token | d |
Token ID | 67 |
Feature activation | -2.76e-01 |
Pos logit contributions | |
Howell | +1.281 |
Birch | +1.106 |
Clarkson | +1.102 |
JJ | +1.056 |
Hendricks | +1.003 |
Neg logit contributions | |
phe | -1.170 |
ū | -1.131 |
aug | -1.115 |
Ur | -1.114 |
kus | -1.067 |
Token | up |
Token ID | 510 |
Feature activation | +3.12e-02 |
Pos logit contributions | |
Howell | +2.469 |
Birch | +2.143 |
Clarkson | +2.136 |
JJ | +2.052 |
Hendricks | +1.950 |
Neg logit contributions | |
phe | -2.082 |
ū | -2.013 |
aug | -1.981 |
Ur | -1.979 |
kus | -1.891 |
Token | , |
Token ID | 11 |
Feature activation | -1.56e-02 |
Pos logit contributions | |
phe | +0.228 |
ū | +0.222 |
Ur | +0.217 |
aug | +0.216 |
kus | +0.208 |
Neg logit contributions | |
Howell | -0.286 |
Birch | -0.248 |
Clarkson | -0.247 |
JJ | -0.237 |
Hendricks | -0.226 |
Token | aren |
Token ID | 3588 |
Feature activation | -2.16e-01 |
Pos logit contributions | |
Howell | +0.138 |
Birch | +0.119 |
Clarkson | +0.119 |
JJ | +0.113 |
Hendricks | +0.108 |
Neg logit contributions | |
phe | -0.118 |
ū | -0.114 |
Ur | -0.113 |
aug | -0.112 |
UG | -0.108 |
Token | 't |
Token ID | 470 |
Feature activation | -5.67e-01 |
Pos logit contributions | |
Howell | +1.654 |
Clarkson | +1.406 |
Birch | +1.403 |
JJ | +1.346 |
Hendricks | +1.263 |
Neg logit contributions | |
phe | -1.876 |
ū | -1.818 |
Ur | -1.777 |
aug | -1.761 |
anamo | -1.751 |
Token | you |
Token ID | 345 |
Feature activation | -7.64e-01 |
Pos logit contributions | |
Howell | +5.146 |
Birch | +4.500 |
Clarkson | +4.475 |
JJ | +4.337 |
Hendricks | +4.092 |
Neg logit contributions | |
phe | -4.084 |
ū | -3.983 |
Ur | -3.931 |
aug | -3.906 |
UG | -3.721 |
Token | ? |
Token ID | 30 |
Feature activation | -5.28e-02 |
Pos logit contributions | |
Howell | +6.574 |
Birch | +5.728 |
Clarkson | +5.652 |
JJ | +5.524 |
Hendricks | +5.161 |
Neg logit contributions | |
phe | -5.823 |
ū | -5.604 |
aug | -5.546 |
Ur | -5.495 |
anamo | -5.283 |
Token | Let |
Token ID | 3914 |
Feature activation | -1.89e-01 |
Pos logit contributions | |
Howell | +0.508 |
Birch | +0.444 |
Clarkson | +0.443 |
JJ | +0.426 |
Hendricks | +0.408 |
Neg logit contributions | |
phe | -0.362 |
ū | -0.348 |
Ur | -0.344 |
aug | -0.343 |
UG | -0.327 |
Token | 's |
Token ID | 338 |
Feature activation | -2.46e-01 |
Pos logit contributions | |
Howell | +1.627 |
Birch | +1.404 |
Clarkson | +1.395 |
JJ | +1.335 |
Hendricks | +1.273 |
Neg logit contributions | |
phe | -1.489 |
ū | -1.443 |
aug | -1.418 |
Ur | -1.413 |
kus | -1.360 |
Token | try |
Token ID | 1949 |
Feature activation | -1.24e-01 |
Pos logit contributions | |
Howell | +2.212 |
Birch | +1.925 |
Clarkson | +1.919 |
JJ | +1.847 |
Hendricks | +1.753 |
Neg logit contributions | |
phe | -1.836 |
ū | -1.772 |
aug | -1.744 |
Ur | -1.742 |
kus | -1.662 |
Token | some |
Token ID | 617 |
Feature activation | -1.17e-02 |
Pos logit contributions | |
Howell | +1.091 |
Birch | +0.941 |
Clarkson | +0.939 |
JJ | +0.898 |
Hendricks | +0.854 |
Neg logit contributions | |
phe | -0.951 |
ū | -0.923 |
aug | -0.912 |
Ur | -0.907 |
UG | -0.871 |
Token | YouTube |
Token ID | 7444 |
Feature activation | +1.84e-01 |
Pos logit contributions | |
phe | +1.068 |
ū | +1.039 |
Ur | +1.020 |
aug | +1.020 |
UG | +0.983 |
Neg logit contributions | |
Howell | -1.188 |
Clarkson | -1.022 |
Birch | -1.019 |
JJ | -0.971 |
Hendricks | -0.933 |
Token | screen |
Token ID | 3159 |
Feature activation | +1.94e-01 |
Pos logit contributions | |
phe | +0.877 |
ū | +0.856 |
Ur | +0.832 |
aug | +0.830 |
UG | +0.773 |
Neg logit contributions | |
Howell | -2.119 |
Birch | -1.899 |
Clarkson | -1.895 |
JJ | -1.840 |
Hendricks | -1.780 |
Token | grab |
Token ID | 5552 |
Feature activation | +1.38e-01 |
Pos logit contributions | |
phe | +1.003 |
ū | +0.981 |
aug | +0.958 |
Ur | +0.941 |
UG | +0.881 |
Neg logit contributions | |
Howell | -2.183 |
Birch | -1.955 |
Clarkson | -1.924 |
JJ | -1.869 |
Hendricks | -1.811 |
Token | ] |
Token ID | 60 |
Feature activation | -2.83e-01 |
Pos logit contributions | |
phe | +0.704 |
ū | +0.678 |
aug | +0.666 |
Ur | +0.661 |
UG | +0.620 |
Neg logit contributions | |
Howell | -1.562 |
Birch | -1.394 |
Clarkson | -1.392 |
JJ | -1.349 |
Hendricks | -1.301 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -1.51e-01 |
Pos logit contributions | |
Howell | +2.707 |
Birch | +2.377 |
Clarkson | +2.368 |
JJ | +2.284 |
Hendricks | +2.178 |
Neg logit contributions | |
phe | -1.952 |
ū | -1.872 |
aug | -1.841 |
Ur | -1.838 |
kus | -1.757 |
Token | ( |
Token ID | 7 |
Feature activation | -7.63e-01 |
Pos logit contributions | |
Howell | +1.615 |
Birch | +1.437 |
Clarkson | +1.433 |
JJ | +1.388 |
Hendricks | +1.331 |
Neg logit contributions | |
phe | -0.878 |
ū | -0.839 |
aug | -0.823 |
Ur | -0.823 |
kus | -0.773 |
Token | Updated |
Token ID | 17354 |
Feature activation | -2.42e-01 |
Pos logit contributions | |
Howell | +7.964 |
Birch | +7.085 |
Clarkson | +7.081 |
JJ | +6.852 |
Hendricks | +6.562 |
Neg logit contributions | |
phe | -4.397 |
ū | -4.158 |
aug | -4.027 |
Ur | -3.963 |
kus | -3.851 |
Token | below |
Token ID | 2174 |
Feature activation | +6.71e-02 |
Pos logit contributions | |
Howell | +2.127 |
Birch | +1.841 |
Clarkson | +1.837 |
JJ | +1.763 |
Hendricks | +1.670 |
Neg logit contributions | |
phe | -1.864 |
ū | -1.798 |
aug | -1.771 |
Ur | -1.769 |
kus | -1.694 |
Token | ) |
Token ID | 8 |
Feature activation | -2.76e-01 |
Pos logit contributions | |
phe | +0.467 |
ū | +0.451 |
aug | +0.444 |
Ur | +0.444 |
UG | +0.424 |
Neg logit contributions | |
Howell | -0.635 |
Birch | -0.556 |
Clarkson | -0.553 |
JJ | -0.532 |
Hendricks | -0.509 |
Token |
|
Token ID | 198 |
Feature activation | -5.49e-01 |
Pos logit contributions | |
Howell | +2.454 |
Birch | +2.129 |
Clarkson | +2.122 |
JJ | +2.037 |
Hendricks | +1.935 |
Neg logit contributions | |
phe | -2.095 |
ū | -2.025 |
Ur | -1.996 |
aug | -1.995 |
UG | -1.903 |
Token |
|
Token ID | 198 |
Feature activation | -4.25e-01 |
Pos logit contributions | |
Howell | +3.734 |
Clarkson | +3.165 |
Birch | +3.114 |
JJ | +2.978 |
Hendricks | +2.781 |
Neg logit contributions | |
phe | -5.083 |
anamo | -4.941 |
ū | -4.898 |
aug | -4.760 |
trak | -4.727 |
Token | 333 |
Token ID | 23460 |
Feature activation | -3.64e-01 |
Pos logit contributions | |
Howell | +2.152 |
Birch | +1.662 |
Clarkson | +1.597 |
JJ | +1.497 |
Gad | +1.415 |
Neg logit contributions | |
phe | -4.339 |
ū | -4.335 |
Ur | -4.212 |
aug | -4.160 |
trak | -4.048 |
Token | 68 |
Token ID | 8257 |
Feature activation | -2.15e-01 |
Pos logit contributions | |
Howell | +2.871 |
Birch | +2.499 |
Clarkson | +2.368 |
JJ | +2.356 |
Gad | +2.210 |
Neg logit contributions | |
phe | -3.082 |
ū | -2.998 |
Ur | -2.969 |
aug | -2.946 |
kus | -2.836 |
Token | 46 |
Token ID | 3510 |
Feature activation | -2.79e-01 |
Pos logit contributions | |
Howell | +2.137 |
Birch | +1.888 |
Clarkson | +1.877 |
JJ | +1.803 |
Hendricks | +1.734 |
Neg logit contributions | |
phe | -1.415 |
ū | -1.362 |
Ur | -1.320 |
aug | -1.316 |
kus | -1.255 |
Token | . |
Token ID | 13 |
Feature activation | +6.19e-02 |
Pos logit contributions | |
Howell | +2.479 |
Birch | +2.172 |
Clarkson | +2.153 |
JJ | +2.062 |
Hendricks | +1.952 |
Neg logit contributions | |
phe | -2.105 |
ū | -2.042 |
Ur | -2.008 |
aug | -1.994 |
kus | -1.911 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -1.30e-01 |
Pos logit contributions | |
phe | +0.450 |
ū | +0.434 |
aug | +0.428 |
Ur | +0.427 |
kus | +0.407 |
Neg logit contributions | |
Howell | -0.570 |
Birch | -0.497 |
Clarkson | -0.495 |
JJ | -0.478 |
Hendricks | -0.454 |
Token | ( |
Token ID | 7 |
Feature activation | -7.61e-01 |
Pos logit contributions | |
Howell | +1.395 |
Birch | +1.242 |
Clarkson | +1.238 |
JJ | +1.198 |
Hendricks | +1.150 |
Neg logit contributions | |
phe | -0.760 |
ū | -0.726 |
Ur | -0.713 |
aug | -0.713 |
UG | -0.669 |
Token | FIL |
Token ID | 46700 |
Feature activation | -3.28e-01 |
Pos logit contributions | |
Howell | +7.927 |
Birch | +7.046 |
Clarkson | +7.038 |
JJ | +6.814 |
Hendricks | +6.525 |
Neg logit contributions | |
phe | -4.396 |
ū | -4.161 |
aug | -4.023 |
Ur | -3.961 |
kus | -3.856 |
Token | ES |
Token ID | 1546 |
Feature activation | +1.94e-01 |
Pos logit contributions | |
Howell | +3.327 |
Birch | +2.942 |
Clarkson | +2.934 |
JJ | +2.834 |
Hendricks | +2.712 |
Neg logit contributions | |
phe | -2.077 |
ū | -1.992 |
aug | -1.955 |
Ur | -1.953 |
kus | -1.848 |
Token | ) |
Token ID | 8 |
Feature activation | +7.43e-02 |
Pos logit contributions | |
phe | +1.439 |
ū | +1.399 |
Ur | +1.373 |
aug | +1.371 |
UG | +1.345 |
Neg logit contributions | |
Howell | -1.731 |
Birch | -1.507 |
Clarkson | -1.495 |
JJ | -1.411 |
Hendricks | -1.358 |
Token | US |
Token ID | 2937 |
Feature activation | -2.79e-01 |
Pos logit contributions | |
phe | +0.560 |
ū | +0.543 |
Ur | +0.537 |
aug | +0.532 |
UG | +0.515 |
Neg logit contributions | |
Howell | -0.658 |
Birch | -0.570 |
Clarkson | -0.568 |
JJ | -0.542 |
Hendricks | -0.517 |
Token | Treasury |
Token ID | 13419 |
Feature activation | -2.05e-01 |
Pos logit contributions | |
Howell | +2.640 |
Birch | +2.312 |
Clarkson | +2.305 |
JJ | +2.219 |
Hendricks | +2.116 |
Neg logit contributions | |
phe | -1.960 |
ū | -1.892 |
aug | -1.860 |
Ur | -1.858 |
UG | -1.771 |
Token | hard |
Token ID | 1327 |
Feature activation | +2.32e-02 |
Pos logit contributions | |
Howell | +2.184 |
Birch | +1.893 |
Clarkson | +1.883 |
JJ | +1.809 |
Hendricks | +1.719 |
Neg logit contributions | |
phe | -1.872 |
ū | -1.811 |
aug | -1.783 |
Ur | -1.782 |
kus | -1.703 |
Token | work |
Token ID | 670 |
Feature activation | -1.06e-01 |
Pos logit contributions | |
phe | +0.150 |
ū | +0.146 |
Ur | +0.140 |
aug | +0.140 |
kus | +0.137 |
Neg logit contributions | |
Howell | -0.235 |
Birch | -0.206 |
Clarkson | -0.205 |
JJ | -0.194 |
Hendricks | -0.190 |
Token | , |
Token ID | 11 |
Feature activation | -3.00e-01 |
Pos logit contributions | |
Howell | +0.961 |
Birch | +0.836 |
Clarkson | +0.828 |
JJ | +0.794 |
Hendricks | +0.758 |
Neg logit contributions | |
phe | -0.794 |
ū | -0.771 |
aug | -0.758 |
Ur | -0.757 |
kus | -0.724 |
Token | leadership |
Token ID | 5531 |
Feature activation | -1.25e-01 |
Pos logit contributions | |
Howell | +2.563 |
Birch | +2.208 |
Clarkson | +2.197 |
JJ | +2.105 |
Hendricks | +1.996 |
Neg logit contributions | |
phe | -2.381 |
ū | -2.305 |
aug | -2.272 |
Ur | -2.269 |
kus | -2.175 |
Token | and |
Token ID | 290 |
Feature activation | -1.60e-01 |
Pos logit contributions | |
Howell | +1.126 |
Birch | +0.979 |
Clarkson | +0.969 |
JJ | +0.929 |
Hendricks | +0.890 |
Neg logit contributions | |
phe | -0.927 |
ū | -0.901 |
Ur | -0.888 |
aug | -0.888 |
UG | -0.846 |
Token | ( |
Token ID | 357 |
Feature activation | -7.60e-01 |
Pos logit contributions | |
Howell | +1.414 |
Birch | +1.223 |
Clarkson | +1.216 |
JJ | +1.165 |
Hendricks | +1.110 |
Neg logit contributions | |
phe | -1.218 |
ū | -1.179 |
aug | -1.162 |
Ur | -1.161 |
kus | -1.112 |
Token | are |
Token ID | 533 |
Feature activation | -2.38e-01 |
Pos logit contributions | |
Howell | +7.601 |
Clarkson | +6.813 |
Birch | +6.751 |
JJ | +6.607 |
Hendricks | +6.271 |
Neg logit contributions | |
phe | -4.478 |
ū | -4.319 |
aug | -4.153 |
Ur | -4.153 |
anamo | -3.967 |
Token | ) |
Token ID | 8 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
Howell | +2.217 |
Birch | +1.937 |
Clarkson | +1.929 |
JJ | +1.855 |
Hendricks | +1.767 |
Neg logit contributions | |
phe | -1.718 |
ū | -1.655 |
aug | -1.633 |
Ur | -1.627 |
kus | -1.555 |
Token | great |
Token ID | 1049 |
Feature activation | +4.00e-02 |
Pos logit contributions | |
Howell | +0.889 |
Birch | +0.772 |
Clarkson | +0.762 |
JJ | +0.729 |
Hendricks | +0.696 |
Neg logit contributions | |
phe | -0.774 |
ū | -0.749 |
aug | -0.742 |
Ur | -0.738 |
kus | -0.711 |
Token | characters |
Token ID | 3435 |
Feature activation | +1.35e-01 |
Pos logit contributions | |
phe | +0.294 |
ū | +0.288 |
aug | +0.283 |
Ur | +0.282 |
kus | +0.271 |
Neg logit contributions | |
Howell | -0.361 |
Birch | -0.313 |
Clarkson | -0.310 |
JJ | -0.297 |
Hendricks | -0.284 |
Token | ," |
Token ID | 553 |
Feature activation | -3.95e-03 |
Pos logit contributions | |
phe | +0.966 |
ū | +0.946 |
aug | +0.933 |
Ur | +0.919 |
UG | +0.878 |
Neg logit contributions | |
Howell | -1.240 |
Birch | -1.085 |
Clarkson | -1.069 |
JJ | -1.016 |
Hendricks | -0.982 |
Token | type |
Token ID | 4906 |
Feature activation | +1.54e-01 |
Pos logit contributions | |
Howell | +0.565 |
Clarkson | +0.500 |
Birch | +0.497 |
JJ | +0.474 |
Hendricks | +0.451 |
Neg logit contributions | |
phe | -0.440 |
ū | -0.426 |
Ur | -0.419 |
aug | -0.414 |
anamo | -0.398 |
Token | ) |
Token ID | 8 |
Feature activation | -4.79e-02 |
Pos logit contributions | |
phe | +1.142 |
ū | +1.101 |
aug | +1.083 |
Ur | +1.082 |
kus | +1.034 |
Neg logit contributions | |
Howell | -1.402 |
Birch | -1.222 |
Clarkson | -1.218 |
JJ | -1.171 |
Hendricks | -1.114 |
Token | @ |
Token ID | 2488 |
Feature activation | -2.49e-01 |
Pos logit contributions | |
Howell | +0.401 |
Birch | +0.347 |
Clarkson | +0.343 |
JJ | +0.329 |
Hendricks | +0.313 |
Neg logit contributions | |
phe | -0.384 |
ū | -0.373 |
Ur | -0.368 |
aug | -0.367 |
UG | -0.353 |
Token | 144 |
Token ID | 20224 |
Feature activation | +6.48e-02 |
Pos logit contributions | |
Howell | +2.018 |
Birch | +1.726 |
Clarkson | +1.721 |
JJ | +1.645 |
Hendricks | +1.553 |
Neg logit contributions | |
phe | -2.090 |
ū | -2.029 |
aug | -2.000 |
Ur | -1.997 |
kus | -1.916 |
Token | Hz |
Token ID | 7399 |
Feature activation | -3.68e-01 |
Pos logit contributions | |
phe | +0.326 |
ū | +0.310 |
aug | +0.302 |
Ur | +0.301 |
kus | +0.282 |
Neg logit contributions | |
Howell | -0.741 |
Birch | -0.664 |
Clarkson | -0.663 |
JJ | -0.641 |
Hendricks | -0.620 |
Token | ( |
Token ID | 357 |
Feature activation | -7.57e-01 |
Pos logit contributions | |
Howell | +3.269 |
Birch | +2.836 |
Clarkson | +2.826 |
JJ | +2.716 |
Hendricks | +2.580 |
Neg logit contributions | |
phe | -2.801 |
ū | -2.707 |
aug | -2.665 |
Ur | -2.662 |
kus | -2.544 |
Token | OD |
Token ID | 3727 |
Feature activation | -2.01e-01 |
Pos logit contributions | |
Howell | +7.261 |
Clarkson | +6.472 |
Birch | +6.451 |
JJ | +6.185 |
Hendricks | +6.073 |
Neg logit contributions | |
phe | -4.629 |
anamo | -4.490 |
ū | -4.470 |
aug | -4.345 |
kus | -4.171 |
Token | = |
Token ID | 796 |
Feature activation | +7.29e-02 |
Pos logit contributions | |
Howell | +2.164 |
Birch | +1.929 |
Clarkson | +1.924 |
JJ | +1.862 |
Hendricks | +1.791 |
Neg logit contributions | |
phe | -1.160 |
ū | -1.104 |
Ur | -1.086 |
aug | -1.081 |
UG | -1.022 |
Token | Normal |
Token ID | 14435 |
Feature activation | -3.27e-02 |
Pos logit contributions | |
phe | +0.570 |
ū | +0.555 |
Ur | +0.547 |
aug | +0.543 |
UG | +0.527 |
Neg logit contributions | |
Howell | -0.616 |
Birch | -0.537 |
Clarkson | -0.533 |
JJ | -0.510 |
Hendricks | -0.484 |
Token | ) |
Token ID | 8 |
Feature activation | -1.98e-01 |
Pos logit contributions | |
Howell | +0.307 |
Birch | +0.270 |
Clarkson | +0.269 |
JJ | +0.260 |
Hendricks | +0.250 |
Neg logit contributions | |
phe | -0.233 |
ū | -0.221 |
aug | -0.217 |
Ur | -0.217 |
UG | -0.207 |
Token | The |
Token ID | 383 |
Feature activation | +4.92e-02 |
Pos logit contributions | |
Howell | +1.722 |
Birch | +1.489 |
Clarkson | +1.478 |
JJ | +1.424 |
Hendricks | +1.353 |
Neg logit contributions | |
phe | -1.537 |
ū | -1.493 |
Ur | -1.470 |
aug | -1.468 |
UG | -1.412 |
Token | the |
Token ID | 262 |
Feature activation | +7.91e-02 |
Pos logit contributions | |
phe | +1.262 |
ū | +1.232 |
Ur | +1.213 |
aug | +1.208 |
kus | +1.159 |
Neg logit contributions | |
Howell | -1.630 |
Birch | -1.416 |
Clarkson | -1.405 |
JJ | -1.357 |
Hendricks | -1.304 |
Token | subjects |
Token ID | 7481 |
Feature activation | -2.29e-02 |
Pos logit contributions | |
phe | +0.604 |
ū | +0.589 |
Ur | +0.580 |
aug | +0.575 |
kus | +0.554 |
Neg logit contributions | |
Howell | -0.693 |
Birch | -0.598 |
Clarkson | -0.596 |
JJ | -0.571 |
Hendricks | -0.546 |
Token | featured |
Token ID | 8096 |
Feature activation | -7.02e-02 |
Pos logit contributions | |
Howell | +0.210 |
Birch | +0.183 |
Clarkson | +0.182 |
JJ | +0.176 |
Hendricks | +0.167 |
Neg logit contributions | |
phe | -0.167 |
ū | -0.161 |
Ur | -0.158 |
aug | -0.158 |
kus | -0.151 |
Token | in |
Token ID | 287 |
Feature activation | +9.75e-02 |
Pos logit contributions | |
Howell | +0.618 |
Birch | +0.535 |
Clarkson | +0.532 |
JJ | +0.510 |
Hendricks | +0.485 |
Neg logit contributions | |
phe | -0.540 |
ū | -0.524 |
Ur | -0.515 |
aug | -0.514 |
kus | -0.492 |
Token | the |
Token ID | 262 |
Feature activation | +1.74e-01 |
Pos logit contributions | |
phe | +0.669 |
ū | +0.650 |
aug | +0.641 |
Ur | +0.634 |
kus | +0.606 |
Neg logit contributions | |
Howell | -0.931 |
Birch | -0.814 |
Clarkson | -0.806 |
JJ | -0.781 |
Hendricks | -0.749 |
Token | images |
Token ID | 4263 |
Feature activation | +5.38e-01 |
Pos logit contributions | |
phe | +1.254 |
ū | +1.226 |
Ur | +1.203 |
aug | +1.197 |
kus | +1.146 |
Neg logit contributions | |
Howell | -1.588 |
Birch | -1.379 |
Clarkson | -1.367 |
JJ | -1.318 |
Hendricks | -1.263 |
Token | intended |
Token ID | 5292 |
Feature activation | -1.61e-01 |
Pos logit contributions | |
phe | +3.464 |
ū | +3.324 |
Ur | +3.222 |
aug | +3.217 |
ura | +3.108 |
Neg logit contributions | |
Howell | -5.269 |
Birch | -4.645 |
Clarkson | -4.544 |
JJ | -4.384 |
Hendricks | -4.290 |
Token | them |
Token ID | 606 |
Feature activation | +7.13e-02 |
Pos logit contributions | |
Howell | +1.348 |
Birch | +1.161 |
Clarkson | +1.152 |
JJ | +1.104 |
Hendricks | +1.046 |
Neg logit contributions | |
phe | -1.300 |
ū | -1.265 |
aug | -1.246 |
Ur | -1.244 |
kus | -1.194 |
Token | to |
Token ID | 284 |
Feature activation | -5.22e-02 |
Pos logit contributions | |
phe | +0.549 |
ū | +0.533 |
aug | +0.526 |
Ur | +0.520 |
kus | +0.500 |
Neg logit contributions | |
Howell | -0.624 |
Birch | -0.546 |
Clarkson | -0.539 |
JJ | -0.521 |
Hendricks | -0.494 |
Token | remain |
Token ID | 3520 |
Feature activation | +1.87e-01 |
Pos logit contributions | |
Howell | +0.510 |
Birch | +0.450 |
Clarkson | +0.446 |
JJ | +0.430 |
Hendricks | +0.412 |
Neg logit contributions | |
phe | -0.350 |
ū | -0.338 |
aug | -0.332 |
Ur | -0.331 |
kus | -0.317 |
Token | private |
Token ID | 2839 |
Feature activation | -3.18e-03 |
Pos logit contributions | |
phe | +1.445 |
ū | +1.415 |
Ur | +1.400 |
aug | +1.391 |
kus | +1.329 |
Neg logit contributions | |
Howell | -1.600 |
Birch | -1.392 |
Clarkson | -1.374 |
JJ | -1.316 |
Hendricks | -1.262 |
Token | had |
Token ID | 550 |
Feature activation | +1.03e-01 |
Pos logit contributions | |
phe | +0.778 |
ū | +0.759 |
aug | +0.741 |
Ur | +0.739 |
kus | +0.705 |
Neg logit contributions | |
Howell | -0.906 |
Birch | -0.780 |
Clarkson | -0.780 |
JJ | -0.753 |
Hendricks | -0.710 |
Token | not |
Token ID | 407 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
phe | +0.763 |
ū | +0.745 |
Ur | +0.731 |
aug | +0.729 |
kus | +0.696 |
Neg logit contributions | |
Howell | -0.918 |
Birch | -0.797 |
Clarkson | -0.795 |
JJ | -0.764 |
Hendricks | -0.723 |
Token | been |
Token ID | 587 |
Feature activation | +2.21e-01 |
Pos logit contributions | |
phe | +0.848 |
ū | +0.825 |
aug | +0.806 |
Ur | +0.804 |
kus | +0.769 |
Neg logit contributions | |
Howell | -1.026 |
Birch | -0.893 |
Clarkson | -0.886 |
JJ | -0.853 |
Hendricks | -0.813 |
Token | captured |
Token ID | 7907 |
Feature activation | +1.43e-01 |
Pos logit contributions | |
phe | +1.612 |
ū | +1.562 |
Ur | +1.535 |
aug | +1.529 |
kus | +1.464 |
Neg logit contributions | |
Howell | -2.002 |
Birch | -1.746 |
Clarkson | -1.728 |
JJ | -1.664 |
Hendricks | -1.581 |
Token | but |
Token ID | 475 |
Feature activation | +3.66e-01 |
Pos logit contributions | |
phe | +1.086 |
ū | +1.059 |
Ur | +1.035 |
aug | +1.035 |
kus | +0.989 |
Neg logit contributions | |
Howell | -1.265 |
Birch | -1.098 |
Clarkson | -1.090 |
JJ | -1.047 |
Hendricks | -0.995 |
Token | rather |
Token ID | 2138 |
Feature activation | +5.84e-01 |
Pos logit contributions | |
phe | +2.764 |
ū | +2.694 |
aug | +2.649 |
Ur | +2.627 |
kus | +2.507 |
Neg logit contributions | |
Howell | -3.162 |
Birch | -2.752 |
Clarkson | -2.738 |
JJ | -2.645 |
Hendricks | -2.465 |
Token | walked |
Token ID | 6807 |
Feature activation | +1.79e-01 |
Pos logit contributions | |
phe | +4.201 |
ū | +4.113 |
aug | +4.009 |
Ur | +3.977 |
ogo | +3.790 |
Neg logit contributions | |
Howell | -5.075 |
Birch | -4.455 |
Clarkson | -4.425 |
JJ | -4.286 |
Hendricks | -3.976 |
Token | away |
Token ID | 1497 |
Feature activation | +2.93e-01 |
Pos logit contributions | |
phe | +1.371 |
ū | +1.342 |
Ur | +1.304 |
aug | +1.294 |
UG | +1.243 |
Neg logit contributions | |
Howell | -1.557 |
Birch | -1.349 |
Clarkson | -1.339 |
JJ | -1.289 |
Hendricks | -1.216 |
Token | from |
Token ID | 422 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
phe | +2.103 |
ū | +2.057 |
Ur | +2.018 |
aug | +2.016 |
UG | +1.908 |
Neg logit contributions | |
Howell | -2.676 |
Birch | -2.333 |
Clarkson | -2.308 |
JJ | -2.233 |
Hendricks | -2.113 |
Token | his |
Token ID | 465 |
Feature activation | -5.03e-02 |
Pos logit contributions | |
Howell | +1.150 |
Birch | +1.009 |
Clarkson | +1.006 |
JJ | +0.969 |
Hendricks | +0.924 |
Neg logit contributions | |
phe | -0.828 |
ū | -0.797 |
aug | -0.783 |
Ur | -0.783 |
kus | -0.744 |
Token | unit |
Token ID | 4326 |
Feature activation | +1.19e-01 |
Pos logit contributions | |
Howell | +0.446 |
Birch | +0.386 |
Clarkson | +0.385 |
JJ | +0.370 |
Hendricks | +0.349 |
Neg logit contributions | |
phe | -0.382 |
ū | -0.371 |
Ur | -0.364 |
aug | -0.363 |
kus | -0.349 |
Token | be |
Token ID | 1350 |
Feature activation | +2.01e-01 |
Pos logit contributions | |
Howell | +0.470 |
Birch | +0.418 |
Clarkson | +0.417 |
JJ | +0.403 |
Hendricks | +0.386 |
Neg logit contributions | |
phe | -0.266 |
ū | -0.253 |
aug | -0.249 |
Ur | -0.246 |
kus | -0.236 |
Token | ek |
Token ID | 988 |
Feature activation | +4.27e-02 |
Pos logit contributions | |
phe | +1.343 |
aug | +1.299 |
ū | +1.298 |
Ur | +1.254 |
kus | +1.224 |
Neg logit contributions | |
Howell | -1.940 |
Birch | -1.715 |
Clarkson | -1.694 |
JJ | -1.635 |
Hendricks | -1.575 |
Token | . |
Token ID | 13 |
Feature activation | +3.11e-01 |
Pos logit contributions | |
phe | +0.300 |
ū | +0.291 |
aug | +0.286 |
Ur | +0.283 |
kus | +0.273 |
Neg logit contributions | |
Howell | -0.401 |
Birch | -0.352 |
Clarkson | -0.351 |
JJ | -0.337 |
Hendricks | -0.323 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -1.19e-01 |
Pos logit contributions | |
ū | +1.829 |
phe | +1.822 |
Ur | +1.805 |
aug | +1.772 |
UG | +1.697 |
Neg logit contributions | |
Howell | -3.143 |
Birch | -2.774 |
Clarkson | -2.742 |
JJ | -2.676 |
Hendricks | -2.574 |
Token | Share |
Token ID | 11649 |
Feature activation | -6.27e-01 |
Pos logit contributions | |
Howell | +1.272 |
Birch | +1.132 |
Clarkson | +1.129 |
JJ | +1.093 |
Hendricks | +1.048 |
Neg logit contributions | |
phe | -0.694 |
ū | -0.663 |
Ur | -0.651 |
aug | -0.650 |
UG | -0.611 |
Token | . |
Token ID | 13 |
Feature activation | +5.84e-01 |
Pos logit contributions | |
Howell | +5.426 |
Birch | +4.735 |
Clarkson | +4.733 |
JJ | +4.523 |
Hendricks | +4.296 |
Neg logit contributions | |
phe | -4.738 |
ū | -4.582 |
Ur | -4.541 |
aug | -4.522 |
kus | -4.334 |
Token | We |
Token ID | 775 |
Feature activation | -5.18e-03 |
Pos logit contributions | |
phe | +4.232 |
Ur | +4.106 |
aug | +4.069 |
ū | +4.039 |
UG | +3.910 |
Neg logit contributions | |
Howell | -4.965 |
Birch | -4.262 |
Clarkson | -4.250 |
JJ | -3.994 |
Hendricks | -3.912 |
Token | found |
Token ID | 1043 |
Feature activation | +1.51e-01 |
Pos logit contributions | |
Howell | +0.048 |
Birch | +0.042 |
Clarkson | +0.042 |
JJ | +0.040 |
Hendricks | +0.039 |
Neg logit contributions | |
phe | -0.037 |
ū | -0.036 |
aug | -0.035 |
Ur | -0.035 |
kus | -0.034 |
Token | a |
Token ID | 257 |
Feature activation | -1.75e-02 |
Pos logit contributions | |
phe | +1.204 |
ū | +1.171 |
aug | +1.158 |
Ur | +1.155 |
UG | +1.109 |
Neg logit contributions | |
Howell | -1.252 |
Birch | -1.064 |
Clarkson | -1.064 |
JJ | -1.008 |
Hendricks | -0.969 |
Token | home |
Token ID | 1363 |
Feature activation | +1.93e-01 |
Pos logit contributions | |
Howell | +0.156 |
Birch | +0.135 |
Clarkson | +0.134 |
JJ | +0.129 |
Hendricks | +0.123 |
Neg logit contributions | |
phe | -0.133 |
ū | -0.128 |
Ur | -0.126 |
aug | -0.126 |
kus | -0.121 |
Token | for |
Token ID | 329 |
Feature activation | -1.51e-01 |
Pos logit contributions | |
phe | +1.445 |
ū | +1.415 |
aug | +1.403 |
Ur | +1.392 |
kus | +1.341 |
Neg logit contributions | |
Howell | -1.692 |
Clarkson | -1.463 |
Birch | -1.460 |
JJ | -1.385 |
Hendricks | -1.336 |
Token | ] |
Token ID | 60 |
Feature activation | +4.23e-01 |
Pos logit contributions | |
phe | +1.214 |
Ur | +1.093 |
ū | +1.055 |
aug | +1.043 |
]] | +0.959 |
Neg logit contributions | |
Howell | -5.055 |
Clarkson | -4.631 |
Birch | -4.602 |
JJ | -4.474 |
Hendricks | -4.332 |
Token | :: |
Token ID | 3712 |
Feature activation | +5.10e-01 |
Pos logit contributions | |
phe | +3.027 |
ū | +2.905 |
Ur | +2.893 |
aug | +2.862 |
UG | +2.737 |
Neg logit contributions | |
Howell | -3.916 |
Birch | -3.433 |
Clarkson | -3.400 |
JJ | -3.276 |
Hendricks | -3.114 |
Token | From |
Token ID | 4863 |
Feature activation | -4.81e-02 |
Pos logit contributions | |
phe | +2.733 |
Ur | +2.706 |
aug | +2.500 |
ū | +2.497 |
UG | +2.488 |
Neg logit contributions | |
Howell | -5.463 |
Clarkson | -4.925 |
Birch | -4.855 |
JJ | -4.787 |
Davies | -4.552 |
Token | Ar |
Token ID | 3163 |
Feature activation | +4.78e-02 |
Pos logit contributions | |
Howell | +0.513 |
Clarkson | +0.457 |
Birch | +0.456 |
JJ | +0.441 |
Hendricks | +0.422 |
Neg logit contributions | |
phe | -0.281 |
Ur | -0.269 |
ū | -0.267 |
aug | -0.264 |
UG | -0.250 |
Token | gb |
Token ID | 22296 |
Feature activation | +3.62e-02 |
Pos logit contributions | |
phe | +0.378 |
ū | +0.365 |
Ur | +0.362 |
aug | +0.358 |
UG | +0.344 |
Neg logit contributions | |
Howell | -0.405 |
Birch | -0.350 |
Clarkson | -0.347 |
JJ | -0.336 |
Hendricks | -0.316 |
Token | ($ |
Token ID | 16763 |
Feature activation | +5.38e-01 |
Pos logit contributions | |
phe | +0.288 |
ū | +0.275 |
Ur | +0.273 |
aug | +0.271 |
UG | +0.260 |
Neg logit contributions | |
Howell | -0.313 |
Birch | -0.270 |
Clarkson | -0.269 |
JJ | -0.257 |
Hendricks | -0.243 |
Token | FC |
Token ID | 4851 |
Feature activation | +4.55e-01 |
Pos logit contributions | |
phe | +3.183 |
Ur | +2.985 |
aug | +2.950 |
ū | +2.915 |
UG | +2.857 |
Neg logit contributions | |
Howell | -5.652 |
Clarkson | -5.014 |
Birch | -4.953 |
JJ | -4.782 |
Dalton | -4.624 |
Token | 1 |
Token ID | 16 |
Feature activation | +5.12e-01 |
Pos logit contributions | |
phe | +2.198 |
Ur | +2.065 |
ū | +2.053 |
aug | +2.030 |
UG | +1.915 |
Neg logit contributions | |
Howell | -5.281 |
Clarkson | -4.760 |
Birch | -4.733 |
JJ | -4.580 |
Hendricks | -4.408 |
Token | [ |
Token ID | 58 |
Feature activation | +3.14e-01 |
Pos logit contributions | |
phe | +2.848 |
ū | +2.758 |
aug | +2.688 |
Ur | +2.665 |
UG | +2.546 |
Neg logit contributions | |
Howell | -5.454 |
Clarkson | -4.861 |
Birch | -4.826 |
JJ | -4.682 |
Hendricks | -4.505 |
Token | 0 |
Token ID | 15 |
Feature activation | +3.65e-01 |
Pos logit contributions | |
phe | +1.662 |
Ur | +1.565 |
aug | +1.556 |
ū | +1.546 |
UG | +1.438 |
Neg logit contributions | |
Howell | -3.532 |
Clarkson | -3.151 |
Birch | -3.146 |
JJ | -3.033 |
Dalton | -2.911 |
Token | ], |
Token ID | 4357 |
Feature activation | +2.91e-01 |
Pos logit contributions | |
phe | +1.417 |
ū | +1.351 |
aug | +1.321 |
Ur | +1.310 |
kus | +1.198 |
Neg logit contributions | |
Howell | -4.543 |
Clarkson | -4.120 |
Birch | -4.119 |
JJ | -3.989 |
Hendricks | -3.854 |
Token | the |
Token ID | 262 |
Feature activation | +3.32e-01 |
Pos logit contributions | |
phe | +2.330 |
ū | +2.255 |
aug | +2.230 |
Ur | +2.213 |
kus | +2.116 |
Neg logit contributions | |
Howell | -3.199 |
Birch | -2.802 |
Clarkson | -2.795 |
JJ | -2.672 |
Hendricks | -2.575 |
Token | CG |
Token ID | 29925 |
Feature activation | +9.44e-02 |
Pos logit contributions | |
phe | +2.396 |
ū | +2.327 |
aug | +2.293 |
Ur | +2.288 |
UG | +2.170 |
Neg logit contributions | |
Howell | -3.010 |
Birch | -2.607 |
Clarkson | -2.602 |
JJ | -2.495 |
Hendricks | -2.395 |
Token | PR |
Token ID | 4805 |
Feature activation | +2.16e-02 |
Pos logit contributions | |
phe | +0.578 |
ū | +0.556 |
Ur | +0.553 |
aug | +0.543 |
UG | +0.516 |
Neg logit contributions | |
Howell | -0.980 |
Birch | -0.870 |
Clarkson | -0.867 |
JJ | -0.840 |
Hendricks | -0.805 |
Token | OG |
Token ID | 7730 |
Feature activation | +1.38e-01 |
Pos logit contributions | |
phe | +0.123 |
ū | +0.118 |
Ur | +0.117 |
aug | +0.116 |
UG | +0.110 |
Neg logit contributions | |
Howell | -0.233 |
Birch | -0.209 |
Clarkson | -0.207 |
JJ | -0.200 |
Hendricks | -0.192 |
Token | RAM |
Token ID | 24115 |
Feature activation | +3.56e-02 |
Pos logit contributions | |
phe | +0.984 |
Ur | +0.947 |
ū | +0.947 |
aug | +0.920 |
UG | +0.896 |
Neg logit contributions | |
Howell | -1.288 |
Birch | -1.128 |
Clarkson | -1.119 |
JJ | -1.074 |
Hendricks | -1.018 |
Token | / |
Token ID | 14 |
Feature activation | +5.42e-01 |
Pos logit contributions | |
phe | +0.246 |
ū | +0.236 |
Ur | +0.235 |
aug | +0.232 |
UG | +0.223 |
Neg logit contributions | |
Howell | -0.342 |
Birch | -0.299 |
Clarkson | -0.297 |
JJ | -0.287 |
Hendricks | -0.274 |
Token | END |
Token ID | 23578 |
Feature activation | +3.70e-01 |
Pos logit contributions | |
phe | +3.390 |
aug | +3.222 |
ū | +3.214 |
Ur | +3.214 |
UG | +3.038 |
Neg logit contributions | |
Howell | -5.473 |
Birch | -4.858 |
Clarkson | -4.835 |
JJ | -4.660 |
Hendricks | -4.481 |
Token | CG |
Token ID | 39816 |
Feature activation | -1.75e-02 |
Pos logit contributions | |
phe | +2.371 |
ū | +2.311 |
Ur | +2.296 |
aug | +2.269 |
UG | +2.226 |
Neg logit contributions | |
Howell | -3.659 |
Birch | -3.220 |
Clarkson | -3.192 |
JJ | -3.061 |
Hendricks | -2.966 |
Token | section |
Token ID | 2665 |
Feature activation | -5.95e-02 |
Pos logit contributions | |
Howell | +0.173 |
Birch | +0.153 |
Clarkson | +0.151 |
JJ | +0.146 |
Hendricks | +0.140 |
Neg logit contributions | |
phe | -0.116 |
ū | -0.112 |
Ur | -0.111 |
aug | -0.111 |
UG | -0.106 |
Token | ! |
Token ID | 0 |
Feature activation | -1.91e-01 |
Pos logit contributions | |
Howell | +0.536 |
Birch | +0.465 |
Clarkson | +0.463 |
JJ | +0.445 |
Hendricks | +0.424 |
Neg logit contributions | |
phe | -0.448 |
ū | -0.433 |
aug | -0.425 |
Ur | -0.425 |
UG | -0.405 |
Token | ): |
Token ID | 2599 |
Feature activation | -2.56e-02 |
Pos logit contributions | |
Howell | +1.562 |
Birch | +1.338 |
Clarkson | +1.332 |
JJ | +1.274 |
Hendricks | +1.204 |
Neg logit contributions | |
phe | -1.590 |
ū | -1.542 |
aug | -1.520 |
Ur | -1.519 |
kus | -1.457 |
Token | ," |
Token ID | 553 |
Feature activation | -4.34e-02 |
Pos logit contributions | |
phe | +0.843 |
ū | +0.820 |
aug | +0.808 |
Ur | +0.803 |
UG | +0.766 |
Neg logit contributions | |
Howell | -1.101 |
Birch | -0.964 |
Clarkson | -0.956 |
JJ | -0.918 |
Hendricks | -0.878 |
Token | he |
Token ID | 339 |
Feature activation | +1.83e-01 |
Pos logit contributions | |
Howell | +0.366 |
Birch | +0.316 |
Clarkson | +0.313 |
JJ | +0.303 |
Hendricks | +0.284 |
Neg logit contributions | |
phe | -0.345 |
ū | -0.332 |
aug | -0.329 |
Ur | -0.327 |
kus | -0.315 |
Token | said |
Token ID | 531 |
Feature activation | -5.91e-02 |
Pos logit contributions | |
phe | +1.329 |
aug | +1.291 |
ū | +1.290 |
Ur | +1.248 |
kus | +1.221 |
Neg logit contributions | |
Howell | -1.649 |
Birch | -1.448 |
Clarkson | -1.418 |
JJ | -1.377 |
Hendricks | -1.305 |
Token | . |
Token ID | 13 |
Feature activation | +1.48e-01 |
Pos logit contributions | |
Howell | +0.564 |
Birch | +0.496 |
Clarkson | +0.492 |
JJ | +0.475 |
Hendricks | +0.453 |
Neg logit contributions | |
phe | -0.409 |
ū | -0.395 |
aug | -0.390 |
Ur | -0.385 |
UG | -0.369 |
Token |
|
Token ID | 198 |
Feature activation | -1.58e-02 |
Pos logit contributions | |
phe | +1.200 |
ū | +1.160 |
aug | +1.148 |
Ur | +1.138 |
kus | +1.094 |
Neg logit contributions | |
Howell | -1.221 |
Birch | -1.049 |
Clarkson | -1.037 |
JJ | -1.004 |
Hendricks | -0.939 |
Token |
|
Token ID | 198 |
Feature activation | +5.26e-02 |
Pos logit contributions | |
Howell | +0.126 |
Clarkson | +0.108 |
Birch | +0.108 |
JJ | +0.103 |
Hendricks | +0.097 |
Neg logit contributions | |
phe | -0.135 |
ū | -0.130 |
aug | -0.128 |
Ur | -0.128 |
anamo | -0.124 |
Token | " |
Token ID | 1 |
Feature activation | -6.63e-02 |
Pos logit contributions | |
phe | +0.310 |
ū | +0.296 |
Ur | +0.296 |
aug | +0.293 |
UG | +0.277 |
Neg logit contributions | |
Howell | -0.560 |
Birch | -0.499 |
Clarkson | -0.495 |
JJ | -0.482 |
Hendricks | -0.460 |
Token | We |
Token ID | 1135 |
Feature activation | +4.00e-02 |
Pos logit contributions | |
Howell | +0.726 |
Birch | +0.649 |
Clarkson | +0.647 |
JJ | +0.627 |
Hendricks | +0.602 |
Neg logit contributions | |
phe | -0.366 |
ū | -0.349 |
Ur | -0.343 |
aug | -0.343 |
UG | -0.321 |
Token | 're |
Token ID | 821 |
Feature activation | +8.11e-02 |
Pos logit contributions | |
phe | +0.275 |
ū | +0.265 |
aug | +0.260 |
Ur | +0.259 |
kus | +0.248 |
Neg logit contributions | |
Howell | -0.384 |
Birch | -0.337 |
Clarkson | -0.336 |
JJ | -0.323 |
Hendricks | -0.310 |
Token | a |
Token ID | 257 |
Feature activation | +6.91e-02 |
Pos logit contributions | |
phe | +0.651 |
ū | +0.628 |
aug | +0.626 |
Ur | +0.624 |
kus | +0.600 |
Neg logit contributions | |
Howell | -0.673 |
Birch | -0.581 |
Clarkson | -0.575 |
JJ | -0.547 |
Hendricks | -0.522 |
Token | little |
Token ID | 1310 |
Feature activation | +2.29e-01 |
Pos logit contributions | |
phe | +0.480 |
ū | +0.459 |
aug | +0.451 |
Ur | +0.450 |
kus | +0.435 |
Neg logit contributions | |
Howell | -0.655 |
Birch | -0.575 |
Clarkson | -0.570 |
JJ | -0.550 |
Hendricks | -0.526 |
Token | , |
Token ID | 11 |
Feature activation | +2.20e-01 |
Pos logit contributions | |
Howell | +0.167 |
Birch | +0.145 |
Clarkson | +0.145 |
JJ | +0.139 |
Hendricks | +0.132 |
Neg logit contributions | |
phe | -0.137 |
ū | -0.133 |
aug | -0.131 |
Ur | -0.131 |
kus | -0.125 |
Token | an |
Token ID | 281 |
Feature activation | +1.52e-02 |
Pos logit contributions | |
phe | +1.608 |
ū | +1.559 |
Ur | +1.540 |
aug | +1.535 |
UG | +1.463 |
Neg logit contributions | |
Howell | -1.991 |
Birch | -1.740 |
Clarkson | -1.734 |
JJ | -1.665 |
Hendricks | -1.600 |
Token | anti |
Token ID | 3098 |
Feature activation | +1.15e-01 |
Pos logit contributions | |
phe | +0.114 |
ū | +0.111 |
Ur | +0.109 |
aug | +0.109 |
kus | +0.104 |
Neg logit contributions | |
Howell | -0.136 |
Birch | -0.119 |
Clarkson | -0.118 |
JJ | -0.113 |
Hendricks | -0.108 |
Token | - |
Token ID | 12 |
Feature activation | +2.36e-01 |
Pos logit contributions | |
phe | +0.702 |
ū | +0.684 |
Ur | +0.673 |
aug | +0.672 |
kus | +0.629 |
Neg logit contributions | |
Howell | -1.194 |
Birch | -1.060 |
Clarkson | -1.051 |
JJ | -1.023 |
Hendricks | -0.982 |
Token | fascist |
Token ID | 46928 |
Feature activation | +5.38e-02 |
Pos logit contributions | |
phe | +1.345 |
Ur | +1.308 |
aug | +1.295 |
ū | +1.294 |
UG | +1.225 |
Neg logit contributions | |
Howell | -2.551 |
Birch | -2.282 |
Clarkson | -2.252 |
JJ | -2.202 |
Hendricks | -2.130 |
Token | activist |
Token ID | 11276 |
Feature activation | +1.22e-01 |
Pos logit contributions | |
phe | +0.384 |
ū | +0.375 |
Ur | +0.368 |
aug | +0.366 |
UG | +0.350 |
Neg logit contributions | |
Howell | -0.497 |
Birch | -0.435 |
Clarkson | -0.433 |
JJ | -0.418 |
Hendricks | -0.399 |
Token | , |
Token ID | 11 |
Feature activation | +1.30e-01 |
Pos logit contributions | |
phe | +0.910 |
ū | +0.887 |
Ur | +0.868 |
aug | +0.865 |
kus | +0.826 |
Neg logit contributions | |
Howell | -1.091 |
Birch | -0.945 |
Clarkson | -0.944 |
JJ | -0.909 |
Hendricks | -0.865 |
Token | was |
Token ID | 373 |
Feature activation | -2.44e-02 |
Pos logit contributions | |
phe | +1.009 |
ū | +0.981 |
Ur | +0.967 |
aug | +0.963 |
UG | +0.919 |
Neg logit contributions | |
Howell | -1.125 |
Birch | -0.971 |
Clarkson | -0.970 |
JJ | -0.933 |
Hendricks | -0.883 |
Token | released |
Token ID | 2716 |
Feature activation | +1.44e-01 |
Pos logit contributions | |
Howell | +0.243 |
Birch | +0.215 |
Clarkson | +0.214 |
JJ | +0.206 |
Hendricks | +0.197 |
Neg logit contributions | |
phe | -0.159 |
ū | -0.153 |
Ur | -0.150 |
aug | -0.150 |
kus | -0.142 |
Token | after |
Token ID | 706 |
Feature activation | +1.69e-01 |
Pos logit contributions | |
phe | +1.080 |
ū | +1.045 |
Ur | +1.026 |
aug | +1.025 |
UG | +0.983 |
Neg logit contributions | |
Howell | -1.293 |
Birch | -1.124 |
Clarkson | -1.116 |
JJ | -1.070 |
Hendricks | -1.020 |
Token | three |
Token ID | 1115 |
Feature activation | -2.43e-02 |
Pos logit contributions | |
phe | +1.265 |
ū | +1.228 |
Ur | +1.210 |
aug | +1.206 |
UG | +1.161 |
Neg logit contributions | |
Howell | -1.506 |
Clarkson | -1.307 |
Birch | -1.305 |
JJ | -1.253 |
Hendricks | -1.186 |
Token | stop |
Token ID | 2245 |
Feature activation | +5.48e-02 |
Pos logit contributions | |
Howell | +0.055 |
Birch | +0.048 |
Clarkson | +0.048 |
JJ | +0.046 |
Hendricks | +0.044 |
Neg logit contributions | |
phe | -0.046 |
ū | -0.045 |
Ur | -0.044 |
aug | -0.044 |
kus | -0.042 |
Token | Laura |
Token ID | 16753 |
Feature activation | -1.24e-03 |
Pos logit contributions | |
phe | +0.387 |
ū | +0.375 |
Ur | +0.368 |
aug | +0.368 |
UG | +0.351 |
Neg logit contributions | |
Howell | -0.513 |
Clarkson | -0.448 |
Birch | -0.448 |
JJ | -0.429 |
Hendricks | -0.409 |
Token | from |
Token ID | 422 |
Feature activation | -2.34e-02 |
Pos logit contributions | |
Howell | +0.012 |
Birch | +0.011 |
Clarkson | +0.010 |
JJ | +0.010 |
Hendricks | +0.010 |
Neg logit contributions | |
phe | -0.009 |
ū | -0.008 |
aug | -0.008 |
Ur | -0.008 |
UG | -0.008 |
Token | overwhelming |
Token ID | 9721 |
Feature activation | +1.47e-02 |
Pos logit contributions | |
Howell | +0.194 |
Birch | +0.167 |
Clarkson | +0.166 |
JJ | +0.158 |
Hendricks | +0.149 |
Neg logit contributions | |
phe | -0.190 |
ū | -0.186 |
aug | -0.183 |
Ur | -0.182 |
kus | -0.175 |
Token | them |
Token ID | 606 |
Feature activation | +2.11e-01 |
Pos logit contributions | |
phe | +0.098 |
ū | +0.095 |
aug | +0.093 |
Ur | +0.093 |
kus | +0.088 |
Neg logit contributions | |
Howell | -0.144 |
Birch | -0.127 |
Clarkson | -0.127 |
JJ | -0.122 |
Hendricks | -0.116 |
Token | quickly |
Token ID | 2952 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
phe | +1.433 |
ū | +1.393 |
aug | +1.367 |
Ur | +1.359 |
UG | +1.286 |
Neg logit contributions | |
Howell | -2.035 |
Birch | -1.783 |
Clarkson | -1.779 |
JJ | -1.704 |
Hendricks | -1.633 |
Token | . |
Token ID | 13 |
Feature activation | +7.70e-03 |
Pos logit contributions | |
phe | +0.790 |
ū | +0.770 |
aug | +0.752 |
Ur | +0.750 |
UG | +0.714 |
Neg logit contributions | |
Howell | -1.078 |
Birch | -0.946 |
Clarkson | -0.942 |
JJ | -0.908 |
Hendricks | -0.862 |
Token |
|
Token ID | 198 |
Feature activation | +5.89e-02 |
Pos logit contributions | |
phe | +0.056 |
ū | +0.055 |
Ur | +0.054 |
aug | +0.054 |
UG | +0.051 |
Neg logit contributions | |
Howell | -0.070 |
Birch | -0.061 |
Clarkson | -0.061 |
JJ | -0.058 |
Hendricks | -0.055 |
Token |
|
Token ID | 198 |
Feature activation | -9.92e-02 |
Pos logit contributions | |
phe | +0.497 |
ū | +0.481 |
aug | +0.472 |
Ur | +0.471 |
kus | +0.455 |
Neg logit contributions | |
Howell | -0.475 |
Clarkson | -0.407 |
Birch | -0.407 |
JJ | -0.388 |
Hendricks | -0.366 |
Token | Laura |
Token ID | 43687 |
Feature activation | +6.49e-02 |
Pos logit contributions | |
Howell | +1.059 |
Birch | +0.943 |
Clarkson | +0.940 |
JJ | +0.910 |
Hendricks | +0.872 |
Neg logit contributions | |
phe | -0.579 |
ū | -0.553 |
Ur | -0.544 |
aug | -0.543 |
kus | -0.510 |
Token | is |
Token ID | 318 |
Feature activation | -2.07e-02 |
Pos logit contributions | |
phe | +0.443 |
ū | +0.429 |
Ur | +0.422 |
aug | +0.421 |
kus | +0.397 |
Neg logit contributions | |
Howell | -0.623 |
Birch | -0.547 |
Clarkson | -0.545 |
JJ | -0.528 |
Hendricks | -0.501 |
Token | , |
Token ID | 11 |
Feature activation | -5.54e-02 |
Pos logit contributions | |
Howell | +2.313 |
Birch | +2.003 |
Clarkson | +1.998 |
JJ | +1.920 |
Hendricks | +1.818 |
Neg logit contributions | |
phe | -2.023 |
ū | -1.956 |
aug | -1.929 |
Ur | -1.924 |
kus | -1.841 |
Token | young |
Token ID | 1862 |
Feature activation | -5.28e-02 |
Pos logit contributions | |
Howell | +0.476 |
Birch | +0.413 |
Clarkson | +0.411 |
JJ | +0.393 |
Hendricks | +0.373 |
Neg logit contributions | |
phe | -0.434 |
ū | -0.420 |
Ur | -0.414 |
aug | -0.413 |
kus | -0.396 |
Token | human |
Token ID | 1692 |
Feature activation | -6.20e-03 |
Pos logit contributions | |
Howell | +0.481 |
Birch | +0.421 |
Clarkson | +0.418 |
JJ | +0.402 |
Hendricks | +0.384 |
Neg logit contributions | |
phe | -0.388 |
ū | -0.375 |
Ur | -0.369 |
aug | -0.368 |
kus | -0.352 |
Token | , |
Token ID | 11 |
Feature activation | +4.32e-03 |
Pos logit contributions | |
Howell | +0.055 |
Birch | +0.048 |
Clarkson | +0.047 |
JJ | +0.045 |
Hendricks | +0.043 |
Neg logit contributions | |
phe | -0.047 |
ū | -0.046 |
Ur | -0.045 |
aug | -0.045 |
kus | -0.043 |
Token | motor |
Token ID | 5584 |
Feature activation | +3.73e-02 |
Pos logit contributions | |
phe | +0.034 |
ū | +0.032 |
Ur | +0.032 |
aug | +0.032 |
kus | +0.030 |
Neg logit contributions | |
Howell | -0.037 |
Birch | -0.033 |
Clarkson | -0.032 |
JJ | -0.031 |
Hendricks | -0.029 |
Token | ized |
Token ID | 1143 |
Feature activation | +4.90e-02 |
Pos logit contributions | |
phe | +0.260 |
ū | +0.250 |
aug | +0.247 |
Ur | +0.246 |
kus | +0.235 |
Neg logit contributions | |
Howell | -0.353 |
Birch | -0.312 |
Clarkson | -0.310 |
JJ | -0.298 |
Hendricks | -0.286 |
Token | wheelchair |
Token ID | 27181 |
Feature activation | -1.65e-01 |
Pos logit contributions | |
phe | +0.368 |
ū | +0.356 |
Ur | +0.350 |
aug | +0.349 |
kus | +0.334 |
Neg logit contributions | |
Howell | -0.434 |
Birch | -0.380 |
Clarkson | -0.376 |
JJ | -0.361 |
Hendricks | -0.345 |
Token | ( |
Token ID | 357 |
Feature activation | -1.73e-02 |
Pos logit contributions | |
Howell | +1.471 |
Birch | +1.280 |
Clarkson | +1.274 |
JJ | +1.223 |
Hendricks | +1.163 |
Neg logit contributions | |
phe | -1.247 |
ū | -1.205 |
Ur | -1.185 |
aug | -1.184 |
UG | -1.130 |
Token | Figure |
Token ID | 11337 |
Feature activation | -1.05e-01 |
Pos logit contributions | |
Howell | +0.178 |
Birch | +0.158 |
Clarkson | +0.157 |
JJ | +0.152 |
Hendricks | +0.146 |
Neg logit contributions | |
phe | -0.108 |
ū | -0.103 |
Ur | -0.102 |
aug | -0.101 |
UG | -0.096 |
Token | 2 |
Token ID | 362 |
Feature activation | -3.04e-02 |
Pos logit contributions | |
Howell | +0.909 |
Birch | +0.785 |
Clarkson | +0.784 |
JJ | +0.749 |
Hendricks | +0.712 |
Neg logit contributions | |
phe | -0.828 |
ū | -0.802 |
aug | -0.788 |
Ur | -0.788 |
kus | -0.755 |
Token | A |
Token ID | 32 |
Feature activation | -5.86e-02 |
Pos logit contributions | |
Howell | +0.296 |
Birch | +0.260 |
Clarkson | +0.260 |
JJ | +0.250 |
Hendricks | +0.240 |
Neg logit contributions | |
phe | -0.205 |
ū | -0.198 |
aug | -0.194 |
Ur | -0.193 |
UG | -0.184 |
Token | book |
Token ID | 1492 |
Feature activation | -1.47e-01 |
Pos logit contributions | |
phe | +0.420 |
ū | +0.408 |
Ur | +0.404 |
aug | +0.402 |
kus | +0.385 |
Neg logit contributions | |
Howell | -0.494 |
Birch | -0.428 |
Clarkson | -0.426 |
JJ | -0.410 |
Hendricks | -0.389 |
Token | sales |
Token ID | 4200 |
Feature activation | +1.98e-01 |
Pos logit contributions | |
Howell | +1.328 |
Birch | +1.152 |
Clarkson | +1.144 |
JJ | +1.103 |
Hendricks | +1.046 |
Neg logit contributions | |
phe | -1.110 |
ū | -1.072 |
Ur | -1.057 |
aug | -1.055 |
UG | -1.007 |
Token | , |
Token ID | 11 |
Feature activation | +1.21e-03 |
Pos logit contributions | |
phe | +1.424 |
ū | +1.391 |
Ur | +1.356 |
aug | +1.352 |
UG | +1.290 |
Neg logit contributions | |
Howell | -1.826 |
Birch | -1.595 |
Clarkson | -1.578 |
JJ | -1.526 |
Hendricks | -1.435 |
Token | rental |
Token ID | 14447 |
Feature activation | +1.98e-02 |
Pos logit contributions | |
phe | +0.009 |
ū | +0.009 |
Ur | +0.009 |
aug | +0.009 |
UG | +0.009 |
Neg logit contributions | |
Howell | -0.010 |
Birch | -0.009 |
Clarkson | -0.009 |
JJ | -0.009 |
Hendricks | -0.008 |
Token | , |
Token ID | 11 |
Feature activation | -1.35e-01 |
Pos logit contributions | |
phe | +0.144 |
ū | +0.139 |
Ur | +0.137 |
aug | +0.136 |
kus | +0.132 |
Neg logit contributions | |
Howell | -0.182 |
Birch | -0.157 |
Clarkson | -0.156 |
JJ | -0.151 |
Hendricks | -0.143 |
Token | or |
Token ID | 393 |
Feature activation | +5.52e-02 |
Pos logit contributions | |
Howell | +1.186 |
Birch | +1.030 |
Clarkson | +1.018 |
JJ | +0.986 |
Hendricks | +0.929 |
Neg logit contributions | |
phe | -1.038 |
ū | -1.002 |
Ur | -0.986 |
aug | -0.984 |
UG | -0.945 |
Token | lending |
Token ID | 20983 |
Feature activation | +2.47e-01 |
Pos logit contributions | |
phe | +0.434 |
ū | +0.418 |
Ur | +0.413 |
aug | +0.411 |
UG | +0.396 |
Neg logit contributions | |
Howell | -0.473 |
Birch | -0.410 |
Clarkson | -0.403 |
JJ | -0.391 |
Hendricks | -0.367 |
Token | records |
Token ID | 4406 |
Feature activation | +2.05e-02 |
Pos logit contributions | |
phe | +1.771 |
ū | +1.713 |
Ur | +1.690 |
aug | +1.669 |
kus | +1.611 |
Neg logit contributions | |
Howell | -2.286 |
Clarkson | -1.971 |
Birch | -1.970 |
JJ | -1.896 |
Hendricks | -1.810 |
Token | . |
Token ID | 13 |
Feature activation | -5.97e-02 |
Pos logit contributions | |
phe | +0.147 |
ū | +0.142 |
Ur | +0.139 |
aug | +0.139 |
UG | +0.132 |
Neg logit contributions | |
Howell | -0.192 |
Birch | -0.167 |
Clarkson | -0.167 |
JJ | -0.161 |
Hendricks | -0.153 |
Token |
|
Token ID | 198 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
Howell | +0.542 |
Birch | +0.471 |
Clarkson | +0.470 |
JJ | +0.452 |
Hendricks | +0.431 |
Neg logit contributions | |
phe | -0.440 |
ū | -0.424 |
Ur | -0.420 |
aug | -0.418 |
UG | -0.399 |
Token |
|
Token ID | 198 |
Feature activation | +4.14e-02 |
Pos logit contributions | |
Howell | +1.050 |
Clarkson | +0.895 |
Birch | +0.891 |
JJ | +0.849 |
Hendricks | +0.798 |
Neg logit contributions | |
phe | -1.228 |
ū | -1.184 |
aug | -1.159 |
anamo | -1.153 |
Ur | -1.152 |
Token | to |
Token ID | 284 |
Feature activation | -4.16e-02 |
Pos logit contributions | |
Howell | +0.190 |
Birch | +0.165 |
Clarkson | +0.164 |
JJ | +0.158 |
Hendricks | +0.150 |
Neg logit contributions | |
phe | -0.168 |
ū | -0.163 |
aug | -0.160 |
Ur | -0.159 |
UG | -0.152 |
Token | federal |
Token ID | 2717 |
Feature activation | -1.86e-01 |
Pos logit contributions | |
Howell | +0.372 |
Birch | +0.323 |
Clarkson | +0.323 |
JJ | +0.310 |
Hendricks | +0.295 |
Neg logit contributions | |
phe | -0.313 |
ū | -0.302 |
aug | -0.298 |
Ur | -0.297 |
kus | -0.284 |
Token | public |
Token ID | 1171 |
Feature activation | -7.77e-02 |
Pos logit contributions | |
Howell | +1.657 |
Birch | +1.439 |
Clarkson | +1.434 |
JJ | +1.378 |
Hendricks | +1.309 |
Neg logit contributions | |
phe | -1.407 |
ū | -1.359 |
aug | -1.338 |
Ur | -1.337 |
kus | -1.278 |
Token | lands |
Token ID | 8604 |
Feature activation | -5.02e-02 |
Pos logit contributions | |
Howell | +0.700 |
Birch | +0.609 |
Clarkson | +0.608 |
JJ | +0.583 |
Hendricks | +0.555 |
Neg logit contributions | |
phe | -0.582 |
ū | -0.561 |
aug | -0.553 |
Ur | -0.553 |
UG | -0.527 |
Token | , |
Token ID | 11 |
Feature activation | +3.21e-04 |
Pos logit contributions | |
Howell | +0.453 |
Birch | +0.394 |
Clarkson | +0.392 |
JJ | +0.377 |
Hendricks | +0.359 |
Neg logit contributions | |
phe | -0.376 |
ū | -0.364 |
aug | -0.358 |
Ur | -0.357 |
UG | -0.341 |
Token | and |
Token ID | 290 |
Feature activation | -6.95e-02 |
Pos logit contributions | |
phe | +0.003 |
ū | +0.002 |
Ur | +0.002 |
aug | +0.002 |
UG | +0.002 |
Neg logit contributions | |
Howell | -0.003 |
Birch | -0.002 |
Clarkson | -0.002 |
JJ | -0.002 |
Hendricks | -0.002 |
Token | then |
Token ID | 788 |
Feature activation | -9.07e-02 |
Pos logit contributions | |
Howell | +0.595 |
Birch | +0.513 |
Clarkson | +0.512 |
JJ | +0.491 |
Hendricks | +0.465 |
Neg logit contributions | |
phe | -0.550 |
ū | -0.532 |
Ur | -0.524 |
aug | -0.524 |
kus | -0.502 |
Token | figure |
Token ID | 3785 |
Feature activation | -1.24e-01 |
Pos logit contributions | |
Howell | +0.799 |
Birch | +0.692 |
Clarkson | +0.690 |
JJ | +0.663 |
Hendricks | +0.629 |
Neg logit contributions | |
phe | -0.695 |
ū | -0.672 |
Ur | -0.662 |
aug | -0.662 |
kus | -0.633 |
Token | out |
Token ID | 503 |
Feature activation | -1.97e-01 |
Pos logit contributions | |
Howell | +0.934 |
Birch | +0.787 |
Clarkson | +0.783 |
JJ | +0.746 |
Hendricks | +0.705 |
Neg logit contributions | |
phe | -1.113 |
ū | -1.086 |
aug | -1.072 |
Ur | -1.067 |
UG | -1.028 |
Token | how |
Token ID | 703 |
Feature activation | -3.59e-01 |
Pos logit contributions | |
Howell | +1.531 |
Birch | +1.300 |
Clarkson | +1.295 |
JJ | +1.236 |
Hendricks | +1.161 |
Neg logit contributions | |
phe | -1.716 |
ū | -1.664 |
aug | -1.643 |
Ur | -1.641 |
kus | -1.579 |
Token | to |
Token ID | 284 |
Feature activation | -1.76e-02 |
Pos logit contributions | |
Howell | +3.000 |
Birch | +2.583 |
Clarkson | +2.567 |
JJ | +2.465 |
Hendricks | +2.320 |
Neg logit contributions | |
phe | -2.921 |
ū | -2.804 |
aug | -2.767 |
Ur | -2.767 |
kus | -2.666 |
Token | foundation |
Token ID | 8489 |
Feature activation | -6.53e-02 |
Pos logit contributions | |
Howell | +0.490 |
Birch | +0.418 |
Clarkson | +0.416 |
JJ | +0.398 |
Hendricks | +0.375 |
Neg logit contributions | |
phe | -0.508 |
ū | -0.494 |
aug | -0.486 |
Ur | -0.485 |
kus | -0.466 |
Token | of |
Token ID | 286 |
Feature activation | -2.07e-01 |
Pos logit contributions | |
Howell | +0.577 |
Clarkson | +0.499 |
Birch | +0.498 |
JJ | +0.479 |
Hendricks | +0.455 |
Neg logit contributions | |
phe | -0.499 |
ū | -0.485 |
aug | -0.475 |
Ur | -0.475 |
UG | -0.454 |
Token | everything |
Token ID | 2279 |
Feature activation | +7.96e-02 |
Pos logit contributions | |
Howell | +1.844 |
Birch | +1.601 |
Clarkson | +1.594 |
JJ | +1.536 |
Hendricks | +1.457 |
Neg logit contributions | |
phe | -1.569 |
ū | -1.513 |
aug | -1.490 |
Ur | -1.488 |
kus | -1.424 |
Token | and |
Token ID | 290 |
Feature activation | -2.63e-01 |
Pos logit contributions | |
phe | +0.583 |
ū | +0.568 |
aug | +0.559 |
Ur | +0.558 |
UG | +0.533 |
Neg logit contributions | |
Howell | -0.723 |
Birch | -0.627 |
Clarkson | -0.626 |
JJ | -0.599 |
Hendricks | -0.570 |
Token | Vietnamese |
Token ID | 23618 |
Feature activation | -3.11e-01 |
Pos logit contributions | |
Howell | +2.235 |
Birch | +1.927 |
Clarkson | +1.918 |
JJ | +1.848 |
Hendricks | +1.744 |
Neg logit contributions | |
phe | -2.108 |
ū | -2.037 |
aug | -2.007 |
Ur | -2.003 |
kus | -1.920 |
Token | people |
Token ID | 661 |
Feature activation | -1.32e-01 |
Pos logit contributions | |
Howell | +2.610 |
Birch | +2.243 |
Clarkson | +2.230 |
JJ | +2.146 |
Hendricks | +2.029 |
Neg logit contributions | |
phe | -2.512 |
ū | -2.429 |
aug | -2.394 |
Ur | -2.390 |
kus | -2.298 |
Token | work |
Token ID | 670 |
Feature activation | -5.23e-02 |
Pos logit contributions | |
Howell | +1.216 |
Birch | +1.060 |
Clarkson | +1.056 |
JJ | +1.016 |
Hendricks | +0.967 |
Neg logit contributions | |
phe | -0.969 |
ū | -0.935 |
aug | -0.920 |
Ur | -0.918 |
kus | -0.877 |
Token | hard |
Token ID | 1327 |
Feature activation | +3.83e-01 |
Pos logit contributions | |
Howell | +0.441 |
Birch | +0.380 |
Clarkson | +0.378 |
JJ | +0.362 |
Hendricks | +0.343 |
Neg logit contributions | |
phe | -0.417 |
ū | -0.407 |
aug | -0.400 |
Ur | -0.400 |
kus | -0.385 |
Token | — |
Token ID | 851 |
Feature activation | +1.03e-02 |
Pos logit contributions | |
phe | +2.746 |
ū | +2.721 |
aug | +2.670 |
Ur | +2.608 |
UG | +2.534 |
Neg logit contributions | |
Howell | -3.464 |
Birch | -3.024 |
Clarkson | -3.007 |
JJ | -2.853 |
Hendricks | -2.757 |
Token | I |
Token ID | 314 |
Feature activation | +6.98e-02 |
Pos logit contributions | |
phe | +0.078 |
ū | +0.076 |
aug | +0.075 |
Ur | +0.075 |
kus | +0.072 |
Neg logit contributions | |
Howell | -0.091 |
Birch | -0.079 |
Clarkson | -0.079 |
JJ | -0.075 |
Hendricks | -0.072 |
Token | � |
Token ID | 447 |
Feature activation | +9.92e-02 |
Pos logit contributions | |
phe | +0.492 |
ū | +0.476 |
aug | +0.463 |
Ur | +0.463 |
kus | +0.446 |
Neg logit contributions | |
Howell | -0.661 |
Birch | -0.580 |
Clarkson | -0.579 |
JJ | -0.556 |
Hendricks | -0.532 |
Token | for |
Token ID | 329 |
Feature activation | -7.90e-02 |
Pos logit contributions | |
Howell | +0.481 |
Birch | +0.414 |
Clarkson | +0.412 |
JJ | +0.395 |
Hendricks | +0.374 |
Neg logit contributions | |
phe | -0.456 |
ū | -0.444 |
Ur | -0.437 |
aug | -0.437 |
UG | -0.418 |
Token | women |
Token ID | 1466 |
Feature activation | +1.46e-01 |
Pos logit contributions | |
Howell | +0.753 |
Birch | +0.661 |
Clarkson | +0.658 |
JJ | +0.633 |
Hendricks | +0.604 |
Neg logit contributions | |
phe | -0.548 |
ū | -0.529 |
Ur | -0.521 |
aug | -0.520 |
UG | -0.495 |
Token | that |
Token ID | 326 |
Feature activation | -4.43e-01 |
Pos logit contributions | |
phe | +1.107 |
ū | +1.083 |
Ur | +1.066 |
aug | +1.059 |
UG | +1.016 |
Neg logit contributions | |
Howell | -1.301 |
Birch | -1.125 |
Clarkson | -1.117 |
JJ | -1.071 |
Hendricks | -1.016 |
Token | would |
Token ID | 561 |
Feature activation | -2.64e-01 |
Pos logit contributions | |
Howell | +3.605 |
Clarkson | +3.077 |
Birch | +3.072 |
JJ | +2.934 |
Hendricks | +2.773 |
Neg logit contributions | |
phe | -3.688 |
ū | -3.535 |
aug | -3.512 |
Ur | -3.488 |
UG | -3.361 |
Token | basically |
Token ID | 6209 |
Feature activation | -1.49e-02 |
Pos logit contributions | |
Howell | +2.481 |
Birch | +2.170 |
Clarkson | +2.166 |
JJ | +2.087 |
Hendricks | +1.986 |
Neg logit contributions | |
phe | -1.876 |
ū | -1.798 |
aug | -1.773 |
Ur | -1.771 |
UG | -1.684 |
Token | consist |
Token ID | 3473 |
Feature activation | +2.91e-02 |
Pos logit contributions | |
Howell | +0.132 |
Birch | +0.115 |
Clarkson | +0.114 |
JJ | +0.110 |
Hendricks | +0.104 |
Neg logit contributions | |
phe | -0.113 |
ū | -0.110 |
Ur | -0.108 |
aug | -0.108 |
kus | -0.103 |
Token | of |
Token ID | 286 |
Feature activation | +1.86e-01 |
Pos logit contributions | |
phe | +0.214 |
ū | +0.209 |
Ur | +0.203 |
aug | +0.203 |
UG | +0.195 |
Neg logit contributions | |
Howell | -0.266 |
Birch | -0.232 |
Clarkson | -0.230 |
JJ | -0.222 |
Hendricks | -0.211 |
Token | a |
Token ID | 257 |
Feature activation | +3.15e-01 |
Pos logit contributions | |
phe | +1.316 |
ū | +1.289 |
Ur | +1.254 |
aug | +1.251 |
UG | +1.197 |
Neg logit contributions | |
Howell | -1.719 |
Birch | -1.500 |
Clarkson | -1.494 |
JJ | -1.435 |
Hendricks | -1.364 |
Token | piece |
Token ID | 3704 |
Feature activation | +2.89e-01 |
Pos logit contributions | |
phe | +2.593 |
ū | +2.551 |
Ur | +2.483 |
aug | +2.465 |
UG | +2.423 |
Neg logit contributions | |
Howell | -2.493 |
Birch | -2.139 |
Clarkson | -2.128 |
JJ | -2.022 |
Hendricks | -1.888 |
Token | of |
Token ID | 286 |
Feature activation | +4.23e-01 |
Pos logit contributions | |
phe | +2.009 |
ū | +1.967 |
Ur | +1.895 |
aug | +1.888 |
UG | +1.822 |
Neg logit contributions | |
Howell | -2.708 |
Birch | -2.373 |
Clarkson | -2.361 |
JJ | -2.293 |
Hendricks | -2.181 |
Token | metal |
Token ID | 6147 |
Feature activation | +2.85e-01 |
Pos logit contributions | |
phe | +3.145 |
ū | +3.082 |
Ur | +2.999 |
aug | +2.993 |
UG | +2.867 |
Neg logit contributions | |
Howell | -3.642 |
Birch | -3.142 |
Clarkson | -3.142 |
JJ | -3.005 |
Hendricks | -2.851 |
Token | � |
Token ID | 134 |
Feature activation | -1.21e-02 |
Pos logit contributions | |
phe | +0.853 |
ū | +0.840 |
aug | +0.810 |
Ur | +0.806 |
kus | +0.752 |
Neg logit contributions | |
Howell | -1.895 |
Birch | -1.707 |
Clarkson | -1.698 |
JJ | -1.642 |
Hendricks | -1.584 |
Token | � |
Token ID | 119 |
Feature activation | -4.95e-01 |
Pos logit contributions | |
Howell | +0.100 |
Birch | +0.086 |
Clarkson | +0.086 |
JJ | +0.082 |
Hendricks | +0.078 |
Neg logit contributions | |
phe | -0.101 |
ū | -0.094 |
aug | -0.094 |
Ur | -0.094 |
anamo | -0.091 |
Token | i |
Token ID | 72 |
Feature activation | -5.62e-01 |
Pos logit contributions | |
Howell | +4.304 |
Birch | +3.705 |
Clarkson | +3.685 |
JJ | +3.587 |
Hendricks | +3.371 |
Neg logit contributions | |
phe | -3.783 |
Ur | -3.564 |
ū | -3.558 |
aug | -3.549 |
kus | -3.412 |
Token | M |
Token ID | 337 |
Feature activation | -8.78e-02 |
Pos logit contributions | |
Howell | +4.745 |
Clarkson | +4.049 |
Birch | +4.042 |
JJ | +3.906 |
Hendricks | +3.664 |
Neg logit contributions | |
phe | -4.442 |
aug | -4.249 |
ū | -4.244 |
Ur | -4.228 |
UG | -4.068 |
Token | ok |
Token ID | 482 |
Feature activation | -1.18e-01 |
Pos logit contributions | |
Howell | +1.111 |
Birch | +1.008 |
Clarkson | +1.006 |
JJ | +0.979 |
Hendricks | +0.946 |
Neg logit contributions | |
phe | -0.336 |
ū | -0.314 |
aug | -0.304 |
Ur | -0.303 |
kus | -0.275 |
Token | up |
Token ID | 929 |
Feature activation | +2.33e-02 |
Pos logit contributions | |
Howell | +1.337 |
Birch | +1.198 |
Clarkson | +1.195 |
JJ | +1.160 |
Hendricks | +1.115 |
Neg logit contributions | |
phe | -0.608 |
ū | -0.575 |
Ur | -0.564 |
aug | -0.563 |
kus | -0.525 |
Token | uni |
Token ID | 35657 |
Feature activation | +6.64e-02 |
Pos logit contributions | |
phe | +0.107 |
ū | +0.104 |
aug | +0.100 |
Ur | +0.099 |
kus | +0.092 |
Neg logit contributions | |
Howell | -0.278 |
Birch | -0.250 |
Clarkson | -0.250 |
JJ | -0.242 |
Hendricks | -0.234 |
Token | ma |
Token ID | 17266 |
Feature activation | -4.13e-03 |
Pos logit contributions | |
phe | +0.507 |
ū | +0.494 |
Ur | +0.484 |
aug | +0.483 |
kus | +0.462 |
Neg logit contributions | |
Howell | -0.584 |
Birch | -0.509 |
Clarkson | -0.508 |
JJ | -0.484 |
Hendricks | -0.462 |
Token | |
Token ID | 220 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
Howell | +0.037 |
Birch | +0.032 |
Clarkson | +0.032 |
JJ | +0.030 |
Hendricks | +0.029 |
Neg logit contributions | |
phe | -0.031 |
ū | -0.030 |
aug | -0.030 |
Ur | -0.030 |
kus | -0.029 |
Token | � |
Token ID | 134 |
Feature activation | -2.51e-01 |
Pos logit contributions | |
phe | +0.595 |
ū | +0.588 |
aug | +0.563 |
Ur | +0.561 |
kus | +0.523 |
Neg logit contributions | |
Howell | -1.283 |
Clarkson | -1.158 |
Birch | -1.155 |
JJ | -1.115 |
Hendricks | -1.076 |
Token | � |
Token ID | 119 |
Feature activation | -9.45e-02 |
Pos logit contributions | |
Howell | +2.246 |
Birch | +1.952 |
Clarkson | +1.945 |
JJ | +1.869 |
Hendricks | +1.778 |
Neg logit contributions | |
phe | -1.890 |
ū | -1.805 |
aug | -1.795 |
Ur | -1.791 |
kus | -1.711 |
Token | gave |
Token ID | 2921 |
Feature activation | +2.40e-02 |
Pos logit contributions | |
phe | +0.500 |
ū | +0.489 |
Ur | +0.476 |
aug | +0.476 |
kus | +0.455 |
Neg logit contributions | |
Howell | -0.632 |
Birch | -0.552 |
Clarkson | -0.548 |
JJ | -0.528 |
Hendricks | -0.501 |
Token | T |
Token ID | 309 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
phe | +0.145 |
ū | +0.139 |
aug | +0.137 |
Ur | +0.136 |
kus | +0.129 |
Neg logit contributions | |
Howell | -0.249 |
Clarkson | -0.221 |
Birch | -0.221 |
JJ | -0.214 |
Hendricks | -0.204 |
Token | read |
Token ID | 961 |
Feature activation | +7.30e-02 |
Pos logit contributions | |
Howell | +1.224 |
Birch | +1.062 |
Clarkson | +1.058 |
JJ | +1.017 |
Hendricks | +0.965 |
Neg logit contributions | |
phe | -1.052 |
ū | -1.015 |
aug | -1.000 |
Ur | -0.999 |
kus | -0.955 |
Token | well |
Token ID | 4053 |
Feature activation | +1.89e-01 |
Pos logit contributions | |
phe | +0.329 |
ū | +0.304 |
Ur | +0.304 |
aug | +0.291 |
UG | +0.277 |
Neg logit contributions | |
Howell | -0.876 |
Birch | -0.788 |
Clarkson | -0.783 |
JJ | -0.765 |
Hendricks | -0.738 |
Token | an |
Token ID | 281 |
Feature activation | -2.76e-02 |
Pos logit contributions | |
phe | +1.341 |
ū | +1.311 |
aug | +1.277 |
Ur | +1.277 |
UG | +1.220 |
Neg logit contributions | |
Howell | -1.766 |
Birch | -1.538 |
Clarkson | -1.532 |
JJ | -1.482 |
Hendricks | -1.408 |
Token | immediate |
Token ID | 7103 |
Feature activation | +9.56e-04 |
Pos logit contributions | |
Howell | +0.237 |
Birch | +0.204 |
Clarkson | +0.203 |
JJ | +0.193 |
Hendricks | +0.183 |
Neg logit contributions | |
phe | -0.217 |
ū | -0.212 |
Ur | -0.209 |
aug | -0.209 |
UG | -0.198 |
Token | advantage |
Token ID | 4621 |
Feature activation | +2.12e-01 |
Pos logit contributions | |
phe | +0.007 |
ū | +0.007 |
Ur | +0.007 |
aug | +0.007 |
UG | +0.007 |
Neg logit contributions | |
Howell | -0.009 |
Birch | -0.007 |
Clarkson | -0.007 |
JJ | -0.007 |
Hendricks | -0.007 |
Token | over |
Token ID | 625 |
Feature activation | -1.32e-02 |
Pos logit contributions | |
phe | +1.594 |
ū | +1.554 |
aug | +1.525 |
Ur | +1.509 |
UG | +1.444 |
Neg logit contributions | |
Howell | -1.882 |
Birch | -1.631 |
Clarkson | -1.609 |
JJ | -1.562 |
Hendricks | -1.476 |
Token | smaller |
Token ID | 4833 |
Feature activation | +1.46e-01 |
Pos logit contributions | |
Howell | +0.120 |
Clarkson | +0.105 |
Birch | +0.105 |
JJ | +0.101 |
Hendricks | +0.096 |
Neg logit contributions | |
phe | -0.095 |
ū | -0.092 |
aug | -0.091 |
Ur | -0.091 |
kus | -0.087 |
Token | corner |
Token ID | 5228 |
Feature activation | -2.09e-01 |
Pos logit contributions | |
phe | +1.141 |
ū | +1.118 |
Ur | +1.104 |
aug | +1.092 |
UG | +1.064 |
Neg logit contributions | |
Howell | -1.230 |
Clarkson | -1.061 |
Birch | -1.051 |
JJ | -0.999 |
Hendricks | -0.956 |
Token | backs |
Token ID | 10146 |
Feature activation | -5.35e-02 |
Pos logit contributions | |
Howell | +2.422 |
Birch | +2.173 |
Clarkson | +2.169 |
JJ | +2.104 |
Hendricks | +2.026 |
Neg logit contributions | |
phe | -1.032 |
ū | -0.980 |
aug | -0.957 |
Ur | -0.957 |
kus | -0.889 |
Token | > |
Token ID | 29 |
Feature activation | -2.36e-02 |
Pos logit contributions | |
Howell | +4.721 |
Clarkson | +4.098 |
Birch | +4.094 |
JJ | +3.972 |
Hendricks | +3.759 |
Neg logit contributions | |
phe | -3.766 |
ū | -3.611 |
Ur | -3.534 |
aug | -3.527 |
kus | -3.445 |
Token | |
Token ID | 7359 |
Feature activation | -7.54e-02 |
Pos logit contributions | |
Howell | +0.232 |
Birch | +0.204 |
Clarkson | +0.201 |
JJ | +0.195 |
Hendricks | +0.186 |
Neg logit contributions | |
phe | -0.158 |
ū | -0.153 |
aug | -0.151 |
Ur | -0.150 |
UG | -0.142 |
Token | li |
Token ID | 4528 |
Feature activation | -2.98e-01 |
Pos logit contributions | |
Howell | +0.782 |
Birch | +0.694 |
Clarkson | +0.690 |
JJ | +0.667 |
Hendricks | +0.640 |
Neg logit contributions | |
phe | -0.467 |
ū | -0.447 |
aug | -0.439 |
Ur | -0.438 |
kus | -0.412 |
Token | > |
Token ID | 29 |
Feature activation | -2.94e-01 |
Pos logit contributions | |
Howell | +2.645 |
Clarkson | +2.311 |
Birch | +2.290 |
JJ | +2.225 |
Hendricks | +2.106 |
Neg logit contributions | |
phe | -2.245 |
ū | -2.143 |
aug | -2.096 |
Ur | -2.081 |
kus | -2.041 |
Token | |
Token ID | 7359 |
Feature activation | -3.99e-01 |
Pos logit contributions | |
Howell | +2.788 |
Birch | +2.440 |
Clarkson | +2.423 |
JJ | +2.339 |
Hendricks | +2.229 |
Neg logit contributions | |
phe | -2.070 |
ū | -2.000 |
aug | -1.968 |
Ur | -1.966 |
UG | -1.867 |
Token | ul |
Token ID | 377 |
Feature activation | -6.86e-01 |
Pos logit contributions | |
Howell | +5.054 |
Clarkson | +4.604 |
Birch | +4.595 |
JJ | +4.486 |
Hendricks | +4.326 |
Neg logit contributions | |
phe | -1.366 |
ū | -1.254 |
Ur | -1.210 |
aug | -1.199 |
UG | -1.123 |
Token | > |
Token ID | 29 |
Feature activation | -2.59e-01 |
Pos logit contributions | |
Howell | +5.426 |
Clarkson | +4.645 |
JJ | +4.576 |
Birch | +4.550 |
Hendricks | +4.236 |
Neg logit contributions | |
phe | -5.471 |
ū | -5.294 |
aug | -5.131 |
Ur | -5.112 |
kus | -5.076 |
Token | |
Token ID | 7359 |
Feature activation | -7.29e-01 |
Pos logit contributions | |
Howell | +2.435 |
Birch | +2.126 |
Clarkson | +2.109 |
JJ | +2.035 |
Hendricks | +1.937 |
Neg logit contributions | |
phe | -1.845 |
ū | -1.786 |
Ur | -1.756 |
aug | -1.756 |
UG | -1.668 |
Token | body |
Token ID | 2618 |
Feature activation | -3.01e-01 |
Pos logit contributions | |
Howell | +6.468 |
Clarkson | +5.693 |
Birch | +5.691 |
JJ | +5.474 |
Hendricks | +5.183 |
Neg logit contributions | |
phe | -5.077 |
ū | -4.974 |
Ur | -4.827 |
aug | -4.820 |
anamo | -4.705 |
Token | > |
Token ID | 29 |
Feature activation | -1.72e-01 |
Pos logit contributions | |
Howell | +2.612 |
Clarkson | +2.291 |
Birch | +2.272 |
JJ | +2.200 |
Hendricks | +2.085 |
Neg logit contributions | |
phe | -2.294 |
ū | -2.201 |
aug | -2.139 |
Ur | -2.128 |
anamo | -2.092 |
Token |
|
Token ID | 198 |
Feature activation | -2.90e-01 |
Pos logit contributions | |
Howell | +1.620 |
Birch | +1.412 |
Clarkson | +1.399 |
JJ | +1.348 |
Hendricks | +1.285 |
Neg logit contributions | |
phe | -1.219 |
ū | -1.184 |
Ur | -1.162 |
aug | -1.161 |
UG | -1.101 |
Token | case |
Token ID | 1339 |
Feature activation | -8.14e-02 |
Pos logit contributions | |
Howell | +1.906 |
Birch | +1.656 |
Clarkson | +1.650 |
JJ | +1.587 |
Hendricks | +1.507 |
Neg logit contributions | |
phe | -1.605 |
ū | -1.549 |
aug | -1.526 |
Ur | -1.524 |
kus | -1.456 |
Token | , |
Token ID | 11 |
Feature activation | -7.00e-02 |
Pos logit contributions | |
Howell | +0.734 |
Birch | +0.637 |
Clarkson | +0.637 |
JJ | +0.609 |
Hendricks | +0.580 |
Neg logit contributions | |
phe | -0.609 |
ū | -0.588 |
Ur | -0.579 |
aug | -0.579 |
kus | -0.555 |
Token | since |
Token ID | 1201 |
Feature activation | +1.91e-01 |
Pos logit contributions | |
Howell | +0.612 |
Birch | +0.528 |
Clarkson | +0.528 |
JJ | +0.503 |
Hendricks | +0.480 |
Neg logit contributions | |
phe | -0.542 |
ū | -0.524 |
Ur | -0.517 |
aug | -0.516 |
kus | -0.497 |
Token | I |
Token ID | 314 |
Feature activation | +1.20e-01 |
Pos logit contributions | |
phe | +1.406 |
ū | +1.356 |
aug | +1.346 |
Ur | +1.338 |
kus | +1.298 |
Neg logit contributions | |
Howell | -1.709 |
Birch | -1.484 |
Clarkson | -1.479 |
JJ | -1.399 |
Hendricks | -1.348 |
Token | mm |
Token ID | 8085 |
Feature activation | -1.10e-01 |
Pos logit contributions | |
phe | +0.863 |
ū | +0.844 |
Ur | +0.821 |
aug | +0.817 |
kus | +0.797 |
Neg logit contributions | |
Howell | -1.108 |
Birch | -0.967 |
Clarkson | -0.965 |
JJ | -0.921 |
Hendricks | -0.885 |
Token | ap |
Token ID | 499 |
Feature activation | -5.02e-01 |
Pos logit contributions | |
Howell | +1.369 |
Birch | +1.239 |
Clarkson | +1.238 |
JJ | +1.202 |
Hendricks | +1.165 |
Neg logit contributions | |
phe | -0.452 |
ū | -0.424 |
aug | -0.414 |
Ur | -0.412 |
UG | -0.375 |
Token | the |
Token ID | 262 |
Feature activation | +4.41e-02 |
Pos logit contributions | |
Howell | +4.457 |
Birch | +3.880 |
Clarkson | +3.847 |
JJ | +3.731 |
Hendricks | +3.516 |
Neg logit contributions | |
phe | -3.761 |
ū | -3.623 |
aug | -3.566 |
Ur | -3.556 |
kus | -3.395 |
Token | file |
Token ID | 2393 |
Feature activation | -2.00e-01 |
Pos logit contributions | |
phe | +0.320 |
ū | +0.308 |
Ur | +0.304 |
aug | +0.303 |
kus | +0.291 |
Neg logit contributions | |
Howell | -0.405 |
Birch | -0.352 |
Clarkson | -0.352 |
JJ | -0.335 |
Hendricks | -0.323 |
Token | , |
Token ID | 11 |
Feature activation | -1.85e-01 |
Pos logit contributions | |
Howell | +1.761 |
Birch | +1.524 |
Clarkson | +1.521 |
JJ | +1.457 |
Hendricks | +1.386 |
Neg logit contributions | |
phe | -1.543 |
ū | -1.490 |
Ur | -1.470 |
aug | -1.466 |
kus | -1.405 |
Token | this |
Token ID | 428 |
Feature activation | -3.25e-01 |
Pos logit contributions | |
Howell | +1.596 |
Birch | +1.376 |
Clarkson | +1.374 |
JJ | +1.312 |
Hendricks | +1.249 |
Neg logit contributions | |
phe | -1.455 |
ū | -1.408 |
aug | -1.388 |
Ur | -1.388 |
kus | -1.330 |
Token | code |
Token ID | 2438 |
Feature activation | -2.57e-01 |
Pos logit contributions | |
Howell | +2.656 |
Birch | +2.271 |
Clarkson | +2.264 |
JJ | +2.170 |
Hendricks | +2.048 |
Neg logit contributions | |
phe | -2.703 |
ū | -2.619 |
aug | -2.583 |
Ur | -2.577 |
kus | -2.475 |
Token | FG |
Token ID | 30386 |
Feature activation | -1.34e-01 |
Pos logit contributions | |
phe | +0.319 |
ū | +0.305 |
Ur | +0.303 |
aug | +0.299 |
UG | +0.289 |
Neg logit contributions | |
Howell | -0.566 |
Birch | -0.503 |
Clarkson | -0.500 |
JJ | -0.480 |
Hendricks | -0.464 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -1.76e-01 |
Pos logit contributions | |
Howell | +1.305 |
Birch | +1.147 |
Clarkson | +1.143 |
JJ | +1.099 |
Hendricks | +1.052 |
Neg logit contributions | |
phe | -0.910 |
ū | -0.874 |
Ur | -0.861 |
aug | -0.860 |
UG | -0.818 |
Token | By |
Token ID | 3886 |
Feature activation | -5.89e-01 |
Pos logit contributions | |
Howell | +1.876 |
Birch | +1.669 |
Clarkson | +1.665 |
JJ | +1.612 |
Hendricks | +1.546 |
Neg logit contributions | |
phe | -1.025 |
ū | -0.980 |
aug | -0.960 |
Ur | -0.957 |
kus | -0.903 |
Token | Andrew |
Token ID | 6858 |
Feature activation | +1.56e-01 |
Pos logit contributions | |
Howell | +5.055 |
Birch | +4.362 |
Clarkson | +4.351 |
JJ | +4.207 |
Hendricks | +3.956 |
Neg logit contributions | |
phe | -4.590 |
ū | -4.414 |
aug | -4.362 |
Ur | -4.356 |
kus | -4.189 |
Token | Benson |
Token ID | 33940 |
Feature activation | +1.78e-01 |
Pos logit contributions | |
phe | +1.264 |
ū | +1.216 |
aug | +1.206 |
Ur | +1.204 |
kus | +1.151 |
Neg logit contributions | |
Howell | -1.252 |
Birch | -1.076 |
Clarkson | -1.073 |
JJ | -1.044 |
Hendricks | -0.975 |
Token |
|
Token ID | 198 |
Feature activation | -5.42e-01 |
Pos logit contributions | |
phe | +1.308 |
ū | +1.285 |
aug | +1.269 |
Ur | +1.249 |
UG | +1.207 |
Neg logit contributions | |
Howell | -1.586 |
Birch | -1.391 |
Clarkson | -1.374 |
JJ | -1.329 |
Hendricks | -1.262 |
Token |
|
Token ID | 198 |
Feature activation | -4.65e-01 |
Pos logit contributions | |
Howell | +3.642 |
Clarkson | +3.063 |
Birch | +3.018 |
JJ | +2.899 |
Hendricks | +2.688 |
Neg logit contributions | |
phe | -5.018 |
anamo | -4.882 |
ū | -4.834 |
aug | -4.679 |
trak | -4.673 |
Token | Veter |
Token ID | 45182 |
Feature activation | -1.46e-01 |
Pos logit contributions | |
Howell | +4.859 |
Clarkson | +4.317 |
Birch | +4.310 |
JJ | +4.174 |
Hendricks | +3.994 |
Neg logit contributions | |
phe | -2.685 |
ū | -2.574 |
aug | -2.482 |
Ur | -2.439 |
kus | -2.373 |
Token | an |
Token ID | 272 |
Feature activation | -1.66e-01 |
Pos logit contributions | |
Howell | +1.511 |
Birch | +1.338 |
Clarkson | +1.335 |
JJ | +1.291 |
Hendricks | +1.231 |
Neg logit contributions | |
phe | -0.901 |
ū | -0.872 |
aug | -0.855 |
Ur | -0.846 |
kus | -0.808 |
Token | Nick |
Token ID | 8047 |
Feature activation | -1.59e-02 |
Pos logit contributions | |
Howell | +1.498 |
Birch | +1.302 |
Clarkson | +1.297 |
JJ | +1.246 |
Hendricks | +1.187 |
Neg logit contributions | |
phe | -1.241 |
ū | -1.199 |
Ur | -1.179 |
aug | -1.178 |
kus | -1.128 |
Token | He |
Token ID | 679 |
Feature activation | +1.18e-02 |
Pos logit contributions | |
Howell | +0.139 |
Birch | +0.120 |
Clarkson | +0.120 |
JJ | +0.116 |
Hendricks | +0.109 |
Neg logit contributions | |
phe | -0.121 |
ū | -0.117 |
aug | -0.116 |
Ur | -0.116 |
kus | -0.110 |
Token | a |
Token ID | 257 |
Feature activation | -2.98e-03 |
Pos logit contributions | |
phe | +0.966 |
ū | +0.926 |
aug | +0.918 |
Ur | +0.908 |
kus | +0.872 |
Neg logit contributions | |
Howell | -1.125 |
Clarkson | -0.975 |
Birch | -0.972 |
JJ | -0.933 |
Hendricks | -0.887 |
Token | fore |
Token ID | 1674 |
Feature activation | -3.31e-02 |
Pos logit contributions | |
Howell | +0.025 |
Birch | +0.022 |
Clarkson | +0.021 |
JJ | +0.021 |
Hendricks | +0.019 |
Neg logit contributions | |
phe | -0.024 |
ū | -0.023 |
aug | -0.023 |
Ur | -0.023 |
UG | -0.022 |
Token | closed |
Token ID | 20225 |
Feature activation | +6.05e-02 |
Pos logit contributions | |
Howell | +0.322 |
Birch | +0.282 |
Clarkson | +0.279 |
JJ | +0.270 |
Hendricks | +0.259 |
Neg logit contributions | |
phe | -0.230 |
ū | -0.218 |
aug | -0.215 |
Ur | -0.210 |
kus | -0.204 |
Token | home |
Token ID | 1363 |
Feature activation | -3.98e-01 |
Pos logit contributions | |
phe | +0.521 |
ū | +0.502 |
aug | +0.495 |
Ur | +0.495 |
UG | +0.478 |
Neg logit contributions | |
Howell | -0.471 |
Birch | -0.401 |
Clarkson | -0.399 |
JJ | -0.383 |
Hendricks | -0.358 |
Token | May |
Token ID | 1737 |
Feature activation | -3.00e-02 |
Pos logit contributions | |
Howell | +3.433 |
Birch | +2.963 |
Clarkson | +2.944 |
JJ | +2.829 |
Hendricks | +2.681 |
Neg logit contributions | |
phe | -3.121 |
ū | -3.006 |
Ur | -2.967 |
aug | -2.961 |
kus | -2.848 |
Token | 7 |
Token ID | 767 |
Feature activation | -6.03e-01 |
Pos logit contributions | |
Howell | +0.248 |
Birch | +0.212 |
Clarkson | +0.212 |
JJ | +0.203 |
Hendricks | +0.191 |
Neg logit contributions | |
phe | -0.247 |
ū | -0.239 |
Ur | -0.236 |
aug | -0.236 |
kus | -0.226 |
Token | , |
Token ID | 11 |
Feature activation | -5.23e-01 |
Pos logit contributions | |
Howell | +5.122 |
Clarkson | +4.424 |
Birch | +4.403 |
JJ | +4.230 |
Hendricks | +4.010 |
Neg logit contributions | |
phe | -4.745 |
ū | -4.513 |
Ur | -4.504 |
aug | -4.464 |
kus | -4.311 |
Token | 2009 |
Token ID | 3717 |
Feature activation | -5.83e-01 |
Pos logit contributions | |
Howell | +4.339 |
Birch | +3.725 |
Clarkson | +3.718 |
JJ | +3.575 |
Hendricks | +3.360 |
Neg logit contributions | |
phe | -4.281 |
ū | -4.116 |
Ur | -4.064 |
aug | -4.050 |
kus | -3.920 |
Token | in |
Token ID | 287 |
Feature activation | -2.06e-01 |
Pos logit contributions | |
Howell | +5.084 |
Birch | +4.409 |
Clarkson | +4.399 |
JJ | +4.218 |
Hendricks | +4.011 |
Neg logit contributions | |
phe | -4.489 |
ū | -4.293 |
Ur | -4.249 |
aug | -4.243 |
kus | -4.087 |
Token | Antioch |
Token ID | 46590 |
Feature activation | -2.78e-01 |
Pos logit contributions | |
Howell | +1.796 |
Birch | +1.554 |
Clarkson | +1.549 |
JJ | +1.487 |
Hendricks | +1.410 |
Neg logit contributions | |
phe | -1.604 |
ū | -1.551 |
aug | -1.528 |
Ur | -1.526 |
kus | -1.461 |
Token | , |
Token ID | 11 |
Feature activation | -2.71e-01 |
Pos logit contributions | |
Howell | +2.487 |
Birch | +2.161 |
Clarkson | +2.153 |
JJ | +2.071 |
Hendricks | +1.966 |
Neg logit contributions | |
phe | -2.092 |
ū | -2.019 |
aug | -1.987 |
Ur | -1.987 |
kus | -1.898 |
Token | you |
Token ID | 345 |
Feature activation | -5.92e-02 |
Pos logit contributions | |
phe | +0.262 |
ū | +0.252 |
Ur | +0.250 |
aug | +0.249 |
kus | +0.239 |
Neg logit contributions | |
Howell | -0.315 |
Clarkson | -0.274 |
Birch | -0.272 |
JJ | -0.261 |
Hendricks | -0.250 |
Token | can |
Token ID | 460 |
Feature activation | -9.78e-02 |
Pos logit contributions | |
Howell | +0.511 |
Birch | +0.442 |
Clarkson | +0.439 |
JJ | +0.419 |
Hendricks | +0.402 |
Neg logit contributions | |
phe | -0.464 |
ū | -0.449 |
aug | -0.444 |
Ur | -0.443 |
kus | -0.427 |
Token | . |
Token ID | 13 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
Howell | +0.903 |
Birch | +0.787 |
Clarkson | +0.786 |
JJ | +0.754 |
Hendricks | +0.719 |
Neg logit contributions | |
phe | -0.710 |
ū | -0.685 |
aug | -0.677 |
Ur | -0.676 |
kus | -0.645 |
Token | It |
Token ID | 632 |
Feature activation | +1.23e-01 |
Pos logit contributions | |
phe | +0.663 |
ū | +0.631 |
Ur | +0.631 |
aug | +0.629 |
UG | +0.595 |
Neg logit contributions | |
Howell | -1.006 |
Clarkson | -0.887 |
Birch | -0.883 |
JJ | -0.850 |
Hendricks | -0.815 |
Token | all |
Token ID | 477 |
Feature activation | -7.79e-02 |
Pos logit contributions | |
phe | +0.859 |
ū | +0.834 |
aug | +0.824 |
Ur | +0.819 |
kus | +0.781 |
Neg logit contributions | |
Howell | -1.157 |
Birch | -1.016 |
Clarkson | -1.001 |
JJ | -0.958 |
Hendricks | -0.925 |
Token | kind |
Token ID | 1611 |
Feature activation | -5.79e-01 |
Pos logit contributions | |
Howell | +0.706 |
Birch | +0.616 |
Clarkson | +0.609 |
JJ | +0.585 |
Hendricks | +0.559 |
Neg logit contributions | |
phe | -0.578 |
ū | -0.560 |
aug | -0.553 |
Ur | -0.549 |
kus | -0.524 |
Token | of |
Token ID | 286 |
Feature activation | -2.39e-01 |
Pos logit contributions | |
Howell | +4.187 |
Clarkson | +3.588 |
Birch | +3.550 |
JJ | +3.516 |
Hendricks | +3.248 |
Neg logit contributions | |
phe | -4.965 |
anamo | -4.759 |
ū | -4.750 |
Ur | -4.744 |
aug | -4.679 |
Token | depends |
Token ID | 8338 |
Feature activation | -1.00e-01 |
Pos logit contributions | |
Howell | +2.099 |
Birch | +1.818 |
Clarkson | +1.812 |
JJ | +1.740 |
Hendricks | +1.650 |
Neg logit contributions | |
phe | -1.842 |
ū | -1.781 |
aug | -1.754 |
Ur | -1.752 |
kus | -1.676 |
Token | on |
Token ID | 319 |
Feature activation | -8.26e-02 |
Pos logit contributions | |
Howell | +0.861 |
Birch | +0.741 |
Clarkson | +0.738 |
JJ | +0.705 |
Hendricks | +0.672 |
Neg logit contributions | |
phe | -0.795 |
ū | -0.764 |
Ur | -0.755 |
aug | -0.755 |
UG | -0.718 |
Token | how |
Token ID | 703 |
Feature activation | +8.83e-02 |
Pos logit contributions | |
Howell | +0.735 |
Birch | +0.635 |
Clarkson | +0.634 |
JJ | +0.607 |
Hendricks | +0.579 |
Neg logit contributions | |
phe | -0.626 |
ū | -0.603 |
Ur | -0.596 |
aug | -0.594 |
kus | -0.566 |
Token | you |
Token ID | 345 |
Feature activation | +4.95e-02 |
Pos logit contributions | |
phe | +0.646 |
ū | +0.622 |
Ur | +0.615 |
aug | +0.614 |
kus | +0.582 |
Neg logit contributions | |
Howell | -0.808 |
Birch | -0.699 |
Clarkson | -0.698 |
JJ | -0.669 |
Hendricks | -0.640 |