Token |
|
Token ID | 198 |
Feature activation | -2.70e-02 |
Pos logit contributions | |
erella | +0.094 |
inburgh | +0.080 |
achus | +0.079 |
lime | +0.073 |
cised | +0.068 |
Neg logit contributions | |
advertisement | -0.092 |
IMAGES | -0.085 |
� | -0.082 |
� | -0.079 |
PDATE | -0.079 |
Token | pur |
Token ID | 14225 |
Feature activation | -6.79e-01 |
Pos logit contributions | |
erella | +0.070 |
inburgh | +0.061 |
achus | +0.060 |
lime | +0.057 |
acted | +0.054 |
Neg logit contributions | |
advertisement | -0.050 |
IMAGES | -0.046 |
� | -0.044 |
PDATE | -0.043 |
� | -0.042 |
Token | ak |
Token ID | 461 |
Feature activation | +3.82e-01 |
Pos logit contributions | |
erella | +1.463 |
inburgh | +1.245 |
achus | +1.232 |
lime | +1.133 |
acted | +1.055 |
Neg logit contributions | |
advertisement | -1.567 |
IMAGES | -1.462 |
� | -1.402 |
PDATE | -1.383 |
� | -1.358 |
Token | ushi |
Token ID | 17731 |
Feature activation | -3.82e-01 |
Pos logit contributions | |
advertisement | +0.819 |
IMAGES | +0.759 |
� | +0.730 |
� | +0.708 |
PDATE | +0.703 |
Neg logit contributions | |
erella | -0.891 |
inburgh | -0.758 |
achus | -0.751 |
lime | -0.691 |
cised | -0.650 |
Token | Profile |
Token ID | 13118 |
Feature activation | +3.85e-01 |
Pos logit contributions | |
erella | +0.858 |
inburgh | +0.729 |
achus | +0.719 |
lime | +0.664 |
acted | +0.617 |
Neg logit contributions | |
advertisement | -0.847 |
IMAGES | -0.785 |
� | -0.754 |
PDATE | -0.739 |
� | -0.733 |
Token | Joined |
Token ID | 16709 |
Feature activation | +2.30e+00 |
Pos logit contributions | |
advertisement | +0.792 |
IMAGES | +0.730 |
� | +0.701 |
� | +0.679 |
PDATE | +0.676 |
Neg logit contributions | |
erella | -0.933 |
inburgh | -0.800 |
achus | -0.794 |
lime | -0.733 |
cised | -0.690 |
Token | August |
Token ID | 2932 |
Feature activation | -8.85e-02 |
Pos logit contributions | |
advertisement | +5.161 |
IMAGES | +4.872 |
� | +4.624 |
OFF | +4.518 |
� | +4.486 |
Neg logit contributions | |
erella | -5.056 |
inburgh | -4.300 |
achus | -4.227 |
lime | -3.819 |
cised | -3.619 |
Token | 2012 |
Token ID | 2321 |
Feature activation | -1.61e-01 |
Pos logit contributions | |
erella | +0.215 |
inburgh | +0.184 |
achus | +0.182 |
lime | +0.169 |
acted | +0.159 |
Neg logit contributions | |
advertisement | -0.181 |
IMAGES | -0.166 |
� | -0.160 |
PDATE | -0.157 |
� | -0.155 |
Token | United |
Token ID | 1578 |
Feature activation | +6.61e-01 |
Pos logit contributions | |
erella | +0.342 |
inburgh | +0.286 |
achus | +0.284 |
lime | +0.259 |
cised | +0.241 |
Neg logit contributions | |
advertisement | -0.378 |
IMAGES | -0.352 |
� | -0.339 |
� | -0.330 |
PDATE | -0.330 |
Token | States |
Token ID | 1829 |
Feature activation | -2.58e-01 |
Pos logit contributions | |
advertisement | +1.634 |
IMAGES | +1.554 |
� | +1.495 |
� | +1.459 |
tti | +1.449 |
Neg logit contributions | |
erella | -1.309 |
inburgh | -1.072 |
achus | -1.065 |
lime | -0.963 |
cised | -0.886 |
Token | 3 |
Token ID | 513 |
Feature activation | +1.15e-01 |
Pos logit contributions | |
erella | +0.518 |
inburgh | +0.428 |
achus | +0.424 |
lime | +0.383 |
cised | +0.354 |
Neg logit contributions | |
advertisement | -0.635 |
IMAGES | -0.591 |
� | -0.572 |
PDATE | -0.562 |
� | -0.558 |
Token |
|
Token ID | 198 |
Feature activation | -7.78e-02 |
Pos logit contributions | |
erella | +0.032 |
inburgh | +0.027 |
achus | +0.027 |
lime | +0.025 |
cised | +0.023 |
Neg logit contributions | |
advertisement | -0.032 |
IMAGES | -0.030 |
� | -0.029 |
PDATE | -0.028 |
� | -0.028 |
Token | heart |
Token ID | 11499 |
Feature activation | -4.31e-01 |
Pos logit contributions | |
erella | +0.196 |
inburgh | +0.169 |
achus | +0.168 |
lime | +0.157 |
cised | +0.147 |
Neg logit contributions | |
advertisement | -0.151 |
IMAGES | -0.141 |
� | -0.133 |
PDATE | -0.131 |
� | -0.129 |
Token | l |
Token ID | 75 |
Feature activation | -4.90e-01 |
Pos logit contributions | |
erella | +1.003 |
inburgh | +0.850 |
achus | +0.844 |
lime | +0.784 |
acted | +0.730 |
Neg logit contributions | |
advertisement | -0.927 |
IMAGES | -0.867 |
� | -0.822 |
PDATE | -0.809 |
� | -0.801 |
Token | xp |
Token ID | 42372 |
Feature activation | -3.68e-01 |
Pos logit contributions | |
erella | +1.087 |
inburgh | +0.921 |
achus | +0.915 |
lime | +0.836 |
illary | +0.779 |
Neg logit contributions | |
advertisement | -1.104 |
IMAGES | -1.030 |
� | -0.988 |
PDATE | -0.968 |
� | -0.961 |
Token | Profile |
Token ID | 13118 |
Feature activation | +3.98e-01 |
Pos logit contributions | |
erella | +0.850 |
inburgh | +0.721 |
achus | +0.718 |
lime | +0.660 |
acted | +0.619 |
Neg logit contributions | |
advertisement | -0.797 |
IMAGES | -0.748 |
� | -0.708 |
PDATE | -0.699 |
� | -0.691 |
Token | Joined |
Token ID | 16709 |
Feature activation | +2.18e+00 |
Pos logit contributions | |
advertisement | +0.794 |
IMAGES | +0.732 |
� | +0.700 |
� | +0.678 |
PDATE | +0.674 |
Neg logit contributions | |
erella | -0.986 |
inburgh | -0.848 |
achus | -0.841 |
lime | -0.779 |
cised | -0.735 |
Token | September |
Token ID | 2693 |
Feature activation | +1.07e-02 |
Pos logit contributions | |
advertisement | +4.857 |
IMAGES | +4.599 |
� | +4.353 |
OFF | +4.274 |
� | +4.213 |
Neg logit contributions | |
erella | -4.830 |
inburgh | -4.127 |
achus | -4.064 |
lime | -3.655 |
cised | -3.458 |
Token | 2010 |
Token ID | 3050 |
Feature activation | -6.68e-03 |
Pos logit contributions | |
advertisement | +0.021 |
IMAGES | +0.020 |
� | +0.019 |
PDATE | +0.018 |
� | +0.018 |
Neg logit contributions | |
erella | -0.026 |
inburgh | -0.023 |
achus | -0.022 |
lime | -0.021 |
cised | -0.020 |
Token | United |
Token ID | 1578 |
Feature activation | +6.33e-01 |
Pos logit contributions | |
erella | +0.014 |
inburgh | +0.012 |
achus | +0.012 |
lime | +0.011 |
cised | +0.010 |
Neg logit contributions | |
advertisement | -0.016 |
IMAGES | -0.015 |
� | -0.014 |
� | -0.014 |
PDATE | -0.014 |
Token | States |
Token ID | 1829 |
Feature activation | -3.20e-01 |
Pos logit contributions | |
advertisement | +1.547 |
IMAGES | +1.481 |
� | +1.419 |
� | +1.386 |
tti | +1.375 |
Neg logit contributions | |
erella | -1.259 |
achus | -1.032 |
inburgh | -1.032 |
lime | -0.936 |
cised | -0.857 |
Token | 10 |
Token ID | 838 |
Feature activation | -1.09e-01 |
Pos logit contributions | |
erella | +0.644 |
inburgh | +0.531 |
achus | +0.525 |
lime | +0.476 |
cised | +0.440 |
Neg logit contributions | |
advertisement | -0.787 |
IMAGES | -0.733 |
� | -0.711 |
PDATE | -0.699 |
� | -0.693 |
Token | of |
Token ID | 286 |
Feature activation | -8.17e-02 |
Pos logit contributions | |
erella | +1.006 |
inburgh | +0.864 |
achus | +0.859 |
lime | +0.803 |
cised | +0.753 |
Neg logit contributions | |
advertisement | -0.803 |
IMAGES | -0.748 |
� | -0.706 |
PDATE | -0.692 |
� | -0.684 |
Token | Dal |
Token ID | 12348 |
Feature activation | -4.46e-01 |
Pos logit contributions | |
erella | +0.193 |
inburgh | +0.165 |
achus | +0.164 |
lime | +0.151 |
cised | +0.142 |
Neg logit contributions | |
advertisement | -0.171 |
IMAGES | -0.158 |
� | -0.152 |
PDATE | -0.148 |
� | -0.147 |
Token | K |
Token ID | 42 |
Feature activation | +3.99e-02 |
Pos logit contributions | |
erella | +0.960 |
inburgh | +0.811 |
achus | +0.798 |
lime | +0.729 |
cised | +0.677 |
Neg logit contributions | |
advertisement | -1.036 |
IMAGES | -0.965 |
� | -0.932 |
PDATE | -0.919 |
� | -0.897 |
Token | Profile |
Token ID | 13118 |
Feature activation | +5.34e-01 |
Pos logit contributions | |
advertisement | +0.085 |
IMAGES | +0.079 |
� | +0.076 |
PDATE | +0.074 |
� | +0.073 |
Neg logit contributions | |
erella | -0.094 |
inburgh | -0.080 |
achus | -0.080 |
lime | -0.074 |
illary | -0.069 |
Token | Blog |
Token ID | 14001 |
Feature activation | +5.04e-01 |
Pos logit contributions | |
advertisement | +0.997 |
IMAGES | +0.916 |
� | +0.873 |
� | +0.842 |
PDATE | +0.836 |
Neg logit contributions | |
erella | -1.391 |
inburgh | -1.204 |
achus | -1.195 |
lime | -1.112 |
cised | -1.052 |
Token | Joined |
Token ID | 16709 |
Feature activation | +2.10e+00 |
Pos logit contributions | |
advertisement | +1.026 |
IMAGES | +0.953 |
� | +0.910 |
� | +0.890 |
OFF | +0.876 |
Neg logit contributions | |
erella | -1.227 |
inburgh | -1.047 |
achus | -1.041 |
lime | -0.964 |
illary | -0.905 |
Token | June |
Token ID | 2795 |
Feature activation | -2.52e-02 |
Pos logit contributions | |
advertisement | +4.635 |
IMAGES | +4.384 |
� | +4.167 |
OFF | +4.076 |
� | +4.024 |
Neg logit contributions | |
erella | -4.678 |
inburgh | -3.984 |
achus | -3.923 |
lime | -3.552 |
cised | -3.362 |
Token | 2012 |
Token ID | 2321 |
Feature activation | -2.81e-01 |
Pos logit contributions | |
erella | +0.063 |
inburgh | +0.054 |
achus | +0.054 |
lime | +0.050 |
cised | +0.047 |
Neg logit contributions | |
advertisement | -0.049 |
IMAGES | -0.045 |
� | -0.044 |
� | -0.042 |
PDATE | -0.042 |
Token | Canada |
Token ID | 3340 |
Feature activation | +8.58e-02 |
Pos logit contributions | |
erella | +0.596 |
inburgh | +0.498 |
achus | +0.493 |
lime | +0.449 |
cised | +0.418 |
Neg logit contributions | |
advertisement | -0.662 |
IMAGES | -0.616 |
� | -0.594 |
PDATE | -0.580 |
� | -0.579 |
Token | 16 |
Token ID | 1467 |
Feature activation | -2.96e-01 |
Pos logit contributions | |
advertisement | +0.208 |
IMAGES | +0.195 |
� | +0.188 |
� | +0.183 |
PDATE | +0.182 |
Neg logit contributions | |
erella | -0.176 |
inburgh | -0.146 |
achus | -0.145 |
lime | -0.131 |
cised | -0.122 |
Token | 10 |
Token ID | 940 |
Feature activation | -3.46e-01 |
Pos logit contributions | |
erella | +0.757 |
inburgh | +0.653 |
achus | +0.647 |
lime | +0.602 |
cised | +0.569 |
Neg logit contributions | |
advertisement | -0.568 |
IMAGES | -0.521 |
� | -0.498 |
PDATE | -0.482 |
� | -0.482 |
Token | J |
Token ID | 41 |
Feature activation | -3.15e-01 |
Pos logit contributions | |
advertisement | +0.092 |
IMAGES | +0.086 |
� | +0.081 |
PDATE | +0.079 |
� | +0.079 |
Neg logit contributions | |
erella | -0.115 |
inburgh | -0.099 |
achus | -0.098 |
lime | -0.091 |
cised | -0.085 |
Token | aaaa |
Token ID | 24794 |
Feature activation | +5.14e-01 |
Pos logit contributions | |
erella | +0.755 |
inburgh | +0.645 |
achus | +0.639 |
lime | +0.592 |
acted | +0.554 |
Neg logit contributions | |
advertisement | -0.655 |
IMAGES | -0.613 |
� | -0.581 |
PDATE | -0.568 |
� | -0.565 |
Token | asper |
Token ID | 32981 |
Feature activation | +2.79e-01 |
Pos logit contributions | |
advertisement | +1.019 |
IMAGES | +0.941 |
� | +0.897 |
� | +0.869 |
PDATE | +0.867 |
Neg logit contributions | |
erella | -1.283 |
inburgh | -1.103 |
achus | -1.095 |
lime | -1.015 |
cised | -0.956 |
Token | Profile |
Token ID | 13118 |
Feature activation | +4.73e-01 |
Pos logit contributions | |
advertisement | +0.619 |
IMAGES | +0.577 |
� | +0.552 |
PDATE | +0.537 |
� | +0.537 |
Neg logit contributions | |
erella | -0.629 |
inburgh | -0.532 |
achus | -0.527 |
lime | -0.486 |
cised | -0.453 |
Token | Blog |
Token ID | 14001 |
Feature activation | +3.74e-01 |
Pos logit contributions | |
advertisement | +0.932 |
IMAGES | +0.860 |
� | +0.822 |
� | +0.795 |
PDATE | +0.790 |
Neg logit contributions | |
erella | -1.187 |
inburgh | -1.022 |
achus | -1.013 |
lime | -0.940 |
cised | -0.887 |
Token | Joined |
Token ID | 16709 |
Feature activation | +2.10e+00 |
Pos logit contributions | |
advertisement | +0.766 |
IMAGES | +0.711 |
� | +0.680 |
� | +0.663 |
OFF | +0.652 |
Neg logit contributions | |
erella | -0.909 |
inburgh | -0.778 |
achus | -0.771 |
lime | -0.714 |
illary | -0.670 |
Token | April |
Token ID | 3035 |
Feature activation | -4.00e-01 |
Pos logit contributions | |
advertisement | +4.663 |
IMAGES | +4.414 |
� | +4.200 |
OFF | +4.105 |
� | +4.051 |
Neg logit contributions | |
erella | -4.649 |
inburgh | -3.962 |
achus | -3.890 |
lime | -3.517 |
cised | -3.329 |
Token | 2012 |
Token ID | 2321 |
Feature activation | -2.42e-01 |
Pos logit contributions | |
erella | +0.988 |
inburgh | +0.844 |
achus | +0.838 |
lime | +0.779 |
acted | +0.734 |
Neg logit contributions | |
advertisement | -0.804 |
IMAGES | -0.736 |
� | -0.707 |
PDATE | -0.694 |
� | -0.685 |
Token | United |
Token ID | 1578 |
Feature activation | +6.34e-01 |
Pos logit contributions | |
erella | +0.518 |
inburgh | +0.433 |
achus | +0.429 |
lime | +0.392 |
cised | +0.364 |
Neg logit contributions | |
advertisement | -0.565 |
IMAGES | -0.526 |
� | -0.507 |
PDATE | -0.495 |
� | -0.494 |
Token | States |
Token ID | 1829 |
Feature activation | -2.90e-01 |
Pos logit contributions | |
advertisement | +1.515 |
IMAGES | +1.449 |
� | +1.388 |
� | +1.358 |
tti | +1.345 |
Neg logit contributions | |
erella | -1.293 |
achus | -1.066 |
inburgh | -1.065 |
lime | -0.971 |
cised | -0.890 |
Token | 5 |
Token ID | 642 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
erella | +0.586 |
inburgh | +0.484 |
achus | +0.479 |
lime | +0.433 |
cised | +0.401 |
Neg logit contributions | |
advertisement | -0.713 |
IMAGES | -0.665 |
� | -0.644 |
PDATE | -0.632 |
� | -0.628 |
Token |
|
Token ID | 198 |
Feature activation | -1.07e-01 |
Pos logit contributions | |
erella | +0.218 |
inburgh | +0.183 |
achus | +0.182 |
lime | +0.167 |
acted | +0.156 |
Neg logit contributions | |
advertisement | -0.219 |
IMAGES | -0.204 |
� | -0.197 |
PDATE | -0.194 |
� | -0.192 |
Token | re |
Token ID | 260 |
Feature activation | -2.55e-01 |
Pos logit contributions | |
erella | +0.269 |
inburgh | +0.232 |
achus | +0.230 |
lime | +0.215 |
cised | +0.202 |
Neg logit contributions | |
advertisement | -0.210 |
IMAGES | -0.196 |
� | -0.185 |
PDATE | -0.182 |
� | -0.180 |
Token | i |
Token ID | 72 |
Feature activation | +3.67e-01 |
Pos logit contributions | |
erella | +0.580 |
inburgh | +0.493 |
achus | +0.488 |
lime | +0.450 |
acted | +0.421 |
Neg logit contributions | |
advertisement | -0.558 |
IMAGES | -0.522 |
� | -0.499 |
PDATE | -0.486 |
� | -0.484 |
Token | Profile |
Token ID | 13118 |
Feature activation | +3.17e-01 |
Pos logit contributions | |
advertisement | +0.752 |
IMAGES | +0.695 |
� | +0.666 |
� | +0.645 |
PDATE | +0.642 |
Neg logit contributions | |
erella | -0.888 |
inburgh | -0.761 |
achus | -0.755 |
lime | -0.697 |
cised | -0.657 |
Token | Blog |
Token ID | 14001 |
Feature activation | +3.14e-01 |
Pos logit contributions | |
advertisement | +0.634 |
IMAGES | +0.584 |
� | +0.559 |
� | +0.541 |
PDATE | +0.538 |
Neg logit contributions | |
erella | -0.785 |
inburgh | -0.675 |
achus | -0.670 |
lime | -0.620 |
cised | -0.585 |
Token | Joined |
Token ID | 16709 |
Feature activation | +2.09e+00 |
Pos logit contributions | |
advertisement | +0.644 |
IMAGES | +0.597 |
� | +0.571 |
� | +0.558 |
OFF | +0.547 |
Neg logit contributions | |
erella | -0.761 |
inburgh | -0.649 |
achus | -0.645 |
lime | -0.597 |
illary | -0.560 |
Token | October |
Token ID | 3267 |
Feature activation | -5.22e-02 |
Pos logit contributions | |
advertisement | +4.711 |
IMAGES | +4.441 |
� | +4.238 |
OFF | +4.122 |
� | +4.095 |
Neg logit contributions | |
erella | -4.583 |
inburgh | -3.893 |
achus | -3.814 |
lime | -3.467 |
cised | -3.266 |
Token | 2002 |
Token ID | 6244 |
Feature activation | +5.16e-02 |
Pos logit contributions | |
erella | +0.130 |
inburgh | +0.112 |
achus | +0.112 |
lime | +0.103 |
cised | +0.098 |
Neg logit contributions | |
advertisement | -0.103 |
IMAGES | -0.095 |
� | -0.091 |
� | -0.088 |
PDATE | -0.087 |
Token | United |
Token ID | 1578 |
Feature activation | +5.84e-01 |
Pos logit contributions | |
advertisement | +0.119 |
IMAGES | +0.111 |
� | +0.107 |
� | +0.104 |
PDATE | +0.103 |
Neg logit contributions | |
erella | -0.112 |
inburgh | -0.095 |
achus | -0.094 |
lime | -0.086 |
cised | -0.080 |
Token | States |
Token ID | 1829 |
Feature activation | -2.53e-01 |
Pos logit contributions | |
advertisement | +1.409 |
IMAGES | +1.342 |
� | +1.288 |
� | +1.259 |
tti | +1.248 |
Neg logit contributions | |
erella | -1.183 |
inburgh | -0.975 |
achus | -0.971 |
lime | -0.885 |
cised | -0.812 |
Token | 3 |
Token ID | 513 |
Feature activation | +1.11e-01 |
Pos logit contributions | |
erella | +0.515 |
inburgh | +0.426 |
achus | +0.421 |
lime | +0.382 |
cised | +0.353 |
Neg logit contributions | |
advertisement | -0.618 |
IMAGES | -0.576 |
� | -0.558 |
PDATE | -0.547 |
� | -0.544 |
Token |
|
Token ID | 198 |
Feature activation | +5.98e-02 |
Pos logit contributions | |
advertisement | +0.121 |
IMAGES | +0.113 |
� | +0.108 |
� | +0.105 |
PDATE | +0.104 |
Neg logit contributions | |
erella | -0.125 |
inburgh | -0.106 |
achus | -0.105 |
lime | -0.096 |
cised | -0.090 |
Token | Dark |
Token ID | 17367 |
Feature activation | -3.43e-01 |
Pos logit contributions | |
advertisement | +0.118 |
IMAGES | +0.110 |
� | +0.104 |
PDATE | +0.101 |
� | +0.101 |
Neg logit contributions | |
erella | -0.149 |
inburgh | -0.128 |
achus | -0.127 |
lime | -0.118 |
cised | -0.111 |
Token | horse |
Token ID | 30527 |
Feature activation | -5.44e-01 |
Pos logit contributions | |
erella | +0.839 |
inburgh | +0.718 |
achus | +0.710 |
lime | +0.664 |
cised | +0.620 |
Neg logit contributions | |
advertisement | -0.699 |
IMAGES | -0.649 |
� | -0.612 |
PDATE | -0.601 |
� | -0.597 |
Token | Profile |
Token ID | 13118 |
Feature activation | +3.31e-01 |
Pos logit contributions | |
erella | +1.293 |
inburgh | +1.102 |
achus | +1.095 |
lime | +1.024 |
cised | +0.953 |
Neg logit contributions | |
advertisement | -1.135 |
IMAGES | -1.057 |
� | -0.998 |
PDATE | -0.989 |
� | -0.973 |
Token | Blog |
Token ID | 14001 |
Feature activation | +3.90e-01 |
Pos logit contributions | |
advertisement | +0.641 |
IMAGES | +0.590 |
� | +0.564 |
� | +0.545 |
PDATE | +0.541 |
Neg logit contributions | |
erella | -0.842 |
inburgh | -0.726 |
achus | -0.721 |
lime | -0.669 |
cised | -0.632 |
Token | Joined |
Token ID | 16709 |
Feature activation | +2.04e+00 |
Pos logit contributions | |
advertisement | +0.803 |
IMAGES | +0.745 |
� | +0.712 |
� | +0.694 |
OFF | +0.684 |
Neg logit contributions | |
erella | -0.943 |
inburgh | -0.806 |
achus | -0.799 |
lime | -0.739 |
illary | -0.694 |
Token | December |
Token ID | 3426 |
Feature activation | -1.35e-01 |
Pos logit contributions | |
advertisement | +4.560 |
IMAGES | +4.307 |
� | +4.098 |
OFF | +4.002 |
� | +3.956 |
Neg logit contributions | |
erella | -4.508 |
inburgh | -3.829 |
achus | -3.764 |
lime | -3.409 |
cised | -3.222 |
Token | 2011 |
Token ID | 2813 |
Feature activation | -2.51e-01 |
Pos logit contributions | |
erella | +0.337 |
inburgh | +0.290 |
achus | +0.287 |
lime | +0.267 |
cised | +0.251 |
Neg logit contributions | |
advertisement | -0.267 |
IMAGES | -0.246 |
� | -0.235 |
� | -0.228 |
PDATE | -0.228 |
Token | United |
Token ID | 1578 |
Feature activation | +5.04e-01 |
Pos logit contributions | |
erella | +0.528 |
inburgh | +0.440 |
achus | +0.436 |
lime | +0.396 |
cised | +0.368 |
Neg logit contributions | |
advertisement | -0.596 |
IMAGES | -0.556 |
� | -0.536 |
PDATE | -0.522 |
� | -0.522 |
Token | States |
Token ID | 1829 |
Feature activation | -3.71e-01 |
Pos logit contributions | |
advertisement | +1.240 |
IMAGES | +1.182 |
� | +1.136 |
� | +1.110 |
tti | +1.101 |
Neg logit contributions | |
erella | -0.997 |
inburgh | -0.817 |
achus | -0.816 |
lime | -0.742 |
illary | -0.676 |
Token | 228 |
Token ID | 29041 |
Feature activation | +7.98e-02 |
Pos logit contributions | |
erella | +0.747 |
inburgh | +0.616 |
achus | +0.609 |
lime | +0.552 |
cised | +0.510 |
Neg logit contributions | |
advertisement | -0.912 |
IMAGES | -0.850 |
� | -0.825 |
PDATE | -0.811 |
� | -0.804 |
Token | TB |
Token ID | 22737 |
Feature activation | -4.77e-01 |
Pos logit contributions | |
erella | +0.134 |
inburgh | +0.116 |
achus | +0.115 |
lime | +0.107 |
cised | +0.100 |
Neg logit contributions | |
advertisement | -0.107 |
IMAGES | -0.100 |
� | -0.094 |
PDATE | -0.093 |
� | -0.092 |
Token | one |
Token ID | 505 |
Feature activation | +2.51e-01 |
Pos logit contributions | |
erella | +1.187 |
inburgh | +1.015 |
achus | +1.007 |
lime | +0.944 |
illary | +0.884 |
Neg logit contributions | |
advertisement | -0.950 |
IMAGES | -0.880 |
� | -0.833 |
PDATE | -0.813 |
� | -0.811 |
Token | - |
Token ID | 12 |
Feature activation | -2.37e-01 |
Pos logit contributions | |
advertisement | +0.529 |
IMAGES | +0.490 |
� | +0.470 |
� | +0.456 |
PDATE | +0.454 |
Neg logit contributions | |
erella | -0.594 |
inburgh | -0.507 |
achus | -0.503 |
lime | -0.464 |
cised | -0.436 |
Token | Profile |
Token ID | 13118 |
Feature activation | +1.34e-01 |
Pos logit contributions | |
erella | +0.588 |
inburgh | +0.507 |
achus | +0.503 |
lime | +0.468 |
cised | +0.439 |
Neg logit contributions | |
advertisement | -0.468 |
IMAGES | -0.435 |
� | -0.413 |
PDATE | -0.402 |
� | -0.400 |
Token | Blog |
Token ID | 14001 |
Feature activation | +3.23e-01 |
Pos logit contributions | |
advertisement | +0.262 |
IMAGES | +0.241 |
� | +0.230 |
� | +0.223 |
PDATE | +0.222 |
Neg logit contributions | |
erella | -0.337 |
inburgh | -0.291 |
achus | -0.289 |
lime | -0.268 |
cised | -0.253 |
Token | Joined |
Token ID | 16709 |
Feature activation | +1.97e+00 |
Pos logit contributions | |
advertisement | +0.669 |
IMAGES | +0.621 |
� | +0.594 |
� | +0.579 |
OFF | +0.569 |
Neg logit contributions | |
erella | -0.778 |
inburgh | -0.664 |
achus | -0.658 |
lime | -0.608 |
illary | -0.571 |
Token | November |
Token ID | 3389 |
Feature activation | -1.47e-01 |
Pos logit contributions | |
advertisement | +4.407 |
IMAGES | +4.176 |
� | +3.971 |
OFF | +3.875 |
� | +3.836 |
Neg logit contributions | |
erella | -4.358 |
inburgh | -3.703 |
achus | -3.647 |
lime | -3.308 |
cised | -3.120 |
Token | 2010 |
Token ID | 3050 |
Feature activation | -1.18e-01 |
Pos logit contributions | |
erella | +0.360 |
inburgh | +0.308 |
achus | +0.306 |
lime | +0.283 |
cised | +0.266 |
Neg logit contributions | |
advertisement | -0.298 |
IMAGES | -0.274 |
� | -0.263 |
PDATE | -0.255 |
� | -0.255 |
Token | United |
Token ID | 1578 |
Feature activation | +5.77e-01 |
Pos logit contributions | |
erella | +0.251 |
inburgh | +0.210 |
achus | +0.208 |
lime | +0.189 |
cised | +0.176 |
Neg logit contributions | |
advertisement | -0.276 |
IMAGES | -0.257 |
� | -0.247 |
� | -0.241 |
PDATE | -0.241 |
Token | States |
Token ID | 1829 |
Feature activation | -2.79e-01 |
Pos logit contributions | |
advertisement | +1.401 |
IMAGES | +1.337 |
� | +1.283 |
� | +1.254 |
tti | +1.245 |
Neg logit contributions | |
erella | -1.152 |
inburgh | -0.948 |
achus | -0.947 |
lime | -0.862 |
cised | -0.788 |
Token | 21 |
Token ID | 2310 |
Feature activation | -1.91e-01 |
Pos logit contributions | |
erella | +0.563 |
inburgh | +0.464 |
achus | +0.459 |
lime | +0.416 |
cised | +0.384 |
Neg logit contributions | |
advertisement | -0.688 |
IMAGES | -0.642 |
� | -0.622 |
PDATE | -0.611 |
� | -0.606 |
Token | K |
Token ID | 42 |
Feature activation | -2.01e-01 |
Pos logit contributions | |
advertisement | +0.197 |
IMAGES | +0.182 |
� | +0.173 |
PDATE | +0.168 |
� | +0.167 |
Neg logit contributions | |
erella | -0.247 |
inburgh | -0.213 |
achus | -0.211 |
lime | -0.197 |
cised | -0.185 |
Token | ier |
Token ID | 959 |
Feature activation | -1.14e-01 |
Pos logit contributions | |
erella | +0.461 |
inburgh | +0.391 |
achus | +0.390 |
lime | +0.360 |
illary | +0.335 |
Neg logit contributions | |
advertisement | -0.437 |
IMAGES | -0.409 |
� | -0.390 |
PDATE | -0.381 |
� | -0.378 |
Token | an |
Token ID | 272 |
Feature activation | -2.29e-01 |
Pos logit contributions | |
erella | +0.269 |
inburgh | +0.230 |
achus | +0.228 |
lime | +0.211 |
cised | +0.198 |
Neg logit contributions | |
advertisement | -0.240 |
IMAGES | -0.223 |
� | -0.213 |
PDATE | -0.208 |
� | -0.206 |
Token | Profile |
Token ID | 13118 |
Feature activation | +2.54e-01 |
Pos logit contributions | |
erella | +0.503 |
inburgh | +0.422 |
achus | +0.421 |
lime | +0.387 |
acted | +0.359 |
Neg logit contributions | |
advertisement | -0.520 |
IMAGES | -0.486 |
� | -0.461 |
PDATE | -0.454 |
� | -0.451 |
Token | Blog |
Token ID | 14001 |
Feature activation | +3.07e-01 |
Pos logit contributions | |
advertisement | +0.479 |
IMAGES | +0.439 |
� | +0.419 |
� | +0.405 |
PDATE | +0.403 |
Neg logit contributions | |
erella | -0.656 |
inburgh | -0.568 |
achus | -0.564 |
lime | -0.524 |
cised | -0.496 |
Token | Joined |
Token ID | 16709 |
Feature activation | +1.95e+00 |
Pos logit contributions | |
advertisement | +0.642 |
IMAGES | +0.595 |
� | +0.570 |
� | +0.554 |
tti | +0.545 |
Neg logit contributions | |
erella | -0.733 |
inburgh | -0.626 |
achus | -0.620 |
lime | -0.573 |
cised | -0.537 |
Token | August |
Token ID | 2932 |
Feature activation | -4.58e-02 |
Pos logit contributions | |
advertisement | +4.359 |
IMAGES | +4.118 |
� | +3.927 |
OFF | +3.828 |
� | +3.789 |
Neg logit contributions | |
erella | -4.320 |
inburgh | -3.665 |
achus | -3.614 |
lime | -3.273 |
cised | -3.094 |
Token | 2011 |
Token ID | 2813 |
Feature activation | -1.67e-01 |
Pos logit contributions | |
erella | +0.111 |
inburgh | +0.095 |
achus | +0.094 |
lime | +0.087 |
acted | +0.082 |
Neg logit contributions | |
advertisement | -0.094 |
IMAGES | -0.086 |
� | -0.083 |
PDATE | -0.081 |
� | -0.080 |
Token | United |
Token ID | 1578 |
Feature activation | +5.72e-01 |
Pos logit contributions | |
erella | +0.354 |
inburgh | +0.296 |
achus | +0.293 |
lime | +0.267 |
cised | +0.248 |
Neg logit contributions | |
advertisement | -0.392 |
IMAGES | -0.365 |
� | -0.352 |
PDATE | -0.343 |
� | -0.343 |
Token | States |
Token ID | 1829 |
Feature activation | -2.63e-01 |
Pos logit contributions | |
advertisement | +1.409 |
IMAGES | +1.346 |
� | +1.295 |
� | +1.265 |
tti | +1.254 |
Neg logit contributions | |
erella | -1.127 |
inburgh | -0.921 |
achus | -0.919 |
lime | -0.833 |
cised | -0.760 |
Token | 6 |
Token ID | 718 |
Feature activation | -1.14e-02 |
Pos logit contributions | |
erella | +0.532 |
inburgh | +0.439 |
achus | +0.434 |
lime | +0.393 |
cised | +0.364 |
Neg logit contributions | |
advertisement | -0.647 |
IMAGES | -0.604 |
� | -0.585 |
PDATE | -0.573 |
� | -0.570 |
Token |
|
Token ID | 198 |
Feature activation | +1.27e-01 |
Pos logit contributions | |
erella | +1.593 |
inburgh | +1.364 |
achus | +1.336 |
lime | +1.230 |
illary | +1.159 |
Neg logit contributions | |
advertisement | -1.716 |
IMAGES | -1.577 |
� | -1.523 |
PDATE | -1.510 |
� | -1.485 |
Token |
|
Token ID | 198 |
Feature activation | +1.18e-01 |
Pos logit contributions | |
advertisement | +0.281 |
IMAGES | +0.261 |
� | +0.251 |
� | +0.244 |
PDATE | +0.243 |
Neg logit contributions | |
erella | -0.289 |
inburgh | -0.245 |
achus | -0.243 |
lime | -0.223 |
cised | -0.209 |
Token | Ko |
Token ID | 48735 |
Feature activation | -5.26e-01 |
Pos logit contributions | |
advertisement | +0.231 |
IMAGES | +0.214 |
� | +0.204 |
PDATE | +0.198 |
� | +0.197 |
Neg logit contributions | |
erella | -0.296 |
inburgh | -0.255 |
achus | -0.253 |
lime | -0.236 |
cised | -0.222 |
Token | Do |
Token ID | 5211 |
Feature activation | +2.75e-01 |
Pos logit contributions | |
erella | +1.283 |
inburgh | +1.089 |
achus | +1.087 |
lime | +1.009 |
illary | +0.943 |
Neg logit contributions | |
advertisement | -1.078 |
IMAGES | -1.008 |
� | -0.951 |
PDATE | -0.931 |
� | -0.915 |
Token | Profile |
Token ID | 13118 |
Feature activation | +5.35e-01 |
Pos logit contributions | |
advertisement | +0.545 |
IMAGES | +0.502 |
� | +0.480 |
� | +0.464 |
PDATE | +0.463 |
Neg logit contributions | |
erella | -0.686 |
inburgh | -0.590 |
achus | -0.585 |
lime | -0.543 |
cised | -0.511 |
Token | Joined |
Token ID | 16709 |
Feature activation | +1.95e+00 |
Pos logit contributions | |
advertisement | +1.003 |
IMAGES | +0.924 |
� | +0.880 |
� | +0.849 |
PDATE | +0.843 |
Neg logit contributions | |
erella | -1.389 |
inburgh | -1.201 |
achus | -1.191 |
lime | -1.109 |
cised | -1.049 |
Token | December |
Token ID | 3426 |
Feature activation | +7.76e-02 |
Pos logit contributions | |
advertisement | +4.375 |
IMAGES | +4.148 |
� | +3.931 |
OFF | +3.837 |
� | +3.814 |
Neg logit contributions | |
erella | -4.275 |
inburgh | -3.633 |
achus | -3.583 |
lime | -3.229 |
cised | -3.061 |
Token | 2010 |
Token ID | 3050 |
Feature activation | +1.99e-01 |
Pos logit contributions | |
advertisement | +0.155 |
IMAGES | +0.143 |
� | +0.137 |
� | +0.133 |
PDATE | +0.133 |
Neg logit contributions | |
erella | -0.192 |
inburgh | -0.165 |
achus | -0.163 |
lime | -0.151 |
cised | -0.143 |
Token | United |
Token ID | 1578 |
Feature activation | +6.39e-01 |
Pos logit contributions | |
advertisement | +0.464 |
IMAGES | +0.434 |
� | +0.418 |
� | +0.406 |
PDATE | +0.403 |
Neg logit contributions | |
erella | -0.425 |
inburgh | -0.357 |
achus | -0.353 |
lime | -0.322 |
cised | -0.300 |
Token | States |
Token ID | 1829 |
Feature activation | -1.87e-01 |
Pos logit contributions | |
advertisement | +1.601 |
IMAGES | +1.530 |
� | +1.468 |
� | +1.428 |
tti | +1.424 |
Neg logit contributions | |
erella | -1.239 |
inburgh | -1.026 |
achus | -1.015 |
lime | -0.913 |
cised | -0.840 |
Token | 6 |
Token ID | 718 |
Feature activation | -9.28e-02 |
Pos logit contributions | |
erella | +0.382 |
inburgh | +0.316 |
achus | +0.313 |
lime | +0.284 |
cised | +0.263 |
Neg logit contributions | |
advertisement | -0.455 |
IMAGES | -0.425 |
� | -0.411 |
PDATE | -0.403 |
� | -0.401 |
Token | ada |
Token ID | 4763 |
Feature activation | +5.18e-02 |
Pos logit contributions | |
erella | +0.363 |
inburgh | +0.306 |
achus | +0.304 |
lime | +0.282 |
illary | +0.263 |
Neg logit contributions | |
advertisement | -0.351 |
IMAGES | -0.328 |
� | -0.313 |
PDATE | -0.306 |
� | -0.304 |
Token | ver |
Token ID | 332 |
Feature activation | -5.03e-01 |
Pos logit contributions | |
advertisement | +0.107 |
IMAGES | +0.099 |
� | +0.095 |
PDATE | +0.092 |
� | +0.092 |
Neg logit contributions | |
erella | -0.125 |
inburgh | -0.107 |
achus | -0.106 |
lime | -0.098 |
cised | -0.092 |
Token | BB |
Token ID | 15199 |
Feature activation | -2.70e-01 |
Pos logit contributions | |
erella | +1.149 |
inburgh | +0.978 |
achus | +0.963 |
lime | +0.900 |
cised | +0.835 |
Neg logit contributions | |
advertisement | -1.098 |
IMAGES | -1.022 |
� | -0.973 |
PDATE | -0.955 |
� | -0.949 |
Token | Profile |
Token ID | 13118 |
Feature activation | +3.49e-01 |
Pos logit contributions | |
erella | +0.626 |
inburgh | +0.531 |
achus | +0.524 |
lime | +0.486 |
illary | +0.455 |
Neg logit contributions | |
advertisement | -0.583 |
IMAGES | -0.546 |
� | -0.515 |
PDATE | -0.506 |
� | -0.502 |
Token | Blog |
Token ID | 14001 |
Feature activation | +4.27e-01 |
Pos logit contributions | |
advertisement | +0.662 |
IMAGES | +0.608 |
� | +0.580 |
� | +0.560 |
PDATE | +0.557 |
Neg logit contributions | |
erella | -0.901 |
inburgh | -0.780 |
achus | -0.774 |
lime | -0.719 |
cised | -0.680 |
Token | Joined |
Token ID | 16709 |
Feature activation | +1.90e+00 |
Pos logit contributions | |
advertisement | +0.883 |
IMAGES | +0.819 |
� | +0.784 |
� | +0.763 |
OFF | +0.752 |
Neg logit contributions | |
erella | -1.030 |
inburgh | -0.880 |
achus | -0.873 |
lime | -0.807 |
illary | -0.757 |
Token | June |
Token ID | 2795 |
Feature activation | +4.37e-02 |
Pos logit contributions | |
advertisement | +4.243 |
IMAGES | +4.029 |
� | +3.838 |
OFF | +3.743 |
� | +3.705 |
Neg logit contributions | |
erella | -4.198 |
inburgh | -3.563 |
achus | -3.513 |
lime | -3.183 |
cised | -3.003 |
Token | 2009 |
Token ID | 3717 |
Feature activation | +6.14e-02 |
Pos logit contributions | |
advertisement | +0.086 |
IMAGES | +0.079 |
� | +0.076 |
� | +0.074 |
PDATE | +0.073 |
Neg logit contributions | |
erella | -0.109 |
inburgh | -0.094 |
achus | -0.093 |
lime | -0.086 |
cised | -0.082 |
Token | Germany |
Token ID | 4486 |
Feature activation | -3.99e-01 |
Pos logit contributions | |
advertisement | +0.141 |
IMAGES | +0.132 |
� | +0.127 |
� | +0.123 |
PDATE | +0.123 |
Neg logit contributions | |
erella | -0.134 |
inburgh | -0.112 |
achus | -0.111 |
lime | -0.102 |
cised | -0.095 |
Token | 247 |
Token ID | 30179 |
Feature activation | +2.32e-01 |
Pos logit contributions | |
erella | +0.804 |
inburgh | +0.662 |
achus | +0.654 |
lime | +0.593 |
cised | +0.548 |
Neg logit contributions | |
advertisement | -0.983 |
IMAGES | -0.916 |
� | -0.888 |
PDATE | -0.874 |
� | -0.866 |
Token | 34 |
Token ID | 2682 |
Feature activation | -4.69e-01 |
Pos logit contributions | |
advertisement | +0.430 |
IMAGES | +0.395 |
� | +0.376 |
� | +0.363 |
PDATE | +0.360 |
Neg logit contributions | |
erella | -0.607 |
inburgh | -0.527 |
achus | -0.523 |
lime | -0.486 |
cised | -0.461 |
Token |
|
Token ID | 198 |
Feature activation | -9.15e-02 |
Pos logit contributions | |
erella | +0.162 |
inburgh | +0.136 |
achus | +0.135 |
lime | +0.124 |
acted | +0.116 |
Neg logit contributions | |
advertisement | -0.162 |
IMAGES | -0.151 |
� | -0.145 |
PDATE | -0.143 |
� | -0.142 |
Token | Big |
Token ID | 12804 |
Feature activation | -3.49e-01 |
Pos logit contributions | |
erella | +0.233 |
inburgh | +0.202 |
achus | +0.200 |
lime | +0.188 |
acted | +0.177 |
Neg logit contributions | |
advertisement | -0.174 |
IMAGES | -0.162 |
� | -0.153 |
PDATE | -0.151 |
� | -0.148 |
Token | Fan |
Token ID | 22480 |
Feature activation | -3.02e-02 |
Pos logit contributions | |
erella | +0.816 |
inburgh | +0.694 |
achus | +0.687 |
lime | +0.637 |
cised | +0.593 |
Neg logit contributions | |
advertisement | -0.749 |
IMAGES | -0.693 |
� | -0.659 |
� | -0.647 |
PDATE | -0.643 |
Token | Profile |
Token ID | 13118 |
Feature activation | +1.64e-01 |
Pos logit contributions | |
erella | +0.071 |
inburgh | +0.060 |
achus | +0.060 |
lime | +0.055 |
cised | +0.052 |
Neg logit contributions | |
advertisement | -0.064 |
IMAGES | -0.060 |
� | -0.057 |
PDATE | -0.055 |
� | -0.055 |
Token | Blog |
Token ID | 14001 |
Feature activation | +1.68e-01 |
Pos logit contributions | |
advertisement | +0.320 |
IMAGES | +0.294 |
� | +0.281 |
� | +0.272 |
PDATE | +0.271 |
Neg logit contributions | |
erella | -0.415 |
inburgh | -0.358 |
achus | -0.355 |
lime | -0.329 |
cised | -0.311 |
Token | Joined |
Token ID | 16709 |
Feature activation | +1.73e+00 |
Pos logit contributions | |
advertisement | +0.358 |
IMAGES | +0.331 |
� | +0.318 |
� | +0.308 |
PDATE | +0.306 |
Neg logit contributions | |
erella | -0.397 |
inburgh | -0.338 |
achus | -0.335 |
lime | -0.309 |
cised | -0.290 |
Token | December |
Token ID | 3426 |
Feature activation | +1.91e-02 |
Pos logit contributions | |
advertisement | +3.948 |
IMAGES | +3.728 |
� | +3.561 |
OFF | +3.460 |
� | +3.437 |
Neg logit contributions | |
erella | -3.744 |
inburgh | -3.166 |
achus | -3.118 |
lime | -2.828 |
cised | -2.662 |
Token | 2010 |
Token ID | 3050 |
Feature activation | +3.16e-02 |
Pos logit contributions | |
advertisement | +0.038 |
IMAGES | +0.035 |
� | +0.034 |
� | +0.033 |
PDATE | +0.033 |
Neg logit contributions | |
erella | -0.047 |
inburgh | -0.041 |
achus | -0.040 |
lime | -0.037 |
cised | -0.035 |
Token | TL |
Token ID | 24811 |
Feature activation | -2.83e-01 |
Pos logit contributions | |
advertisement | +0.074 |
IMAGES | +0.069 |
� | +0.066 |
� | +0.065 |
PDATE | +0.064 |
Neg logit contributions | |
erella | -0.068 |
inburgh | -0.057 |
achus | -0.056 |
lime | -0.051 |
cised | -0.048 |
Token | AD |
Token ID | 2885 |
Feature activation | -4.49e-02 |
Pos logit contributions | |
erella | +0.682 |
inburgh | +0.582 |
achus | +0.574 |
lime | +0.530 |
acted | +0.504 |
Neg logit contributions | |
advertisement | -0.585 |
IMAGES | -0.534 |
� | -0.514 |
PDATE | -0.504 |
� | -0.500 |
Token | T |
Token ID | 51 |
Feature activation | -2.66e-01 |
Pos logit contributions | |
erella | +0.137 |
achus | +0.121 |
inburgh | +0.121 |
lime | +0.113 |
illary | +0.108 |
Neg logit contributions | |
advertisement | -0.064 |
IMAGES | -0.057 |
� | -0.054 |
� | -0.051 |
PDATE | -0.050 |
Token | into |
Token ID | 656 |
Feature activation | +5.68e-02 |
Pos logit contributions | |
erella | +0.295 |
inburgh | +0.254 |
achus | +0.252 |
lime | +0.234 |
cised | +0.220 |
Neg logit contributions | |
advertisement | -0.236 |
IMAGES | -0.216 |
� | -0.208 |
PDATE | -0.202 |
� | -0.202 |
Token | them |
Token ID | 606 |
Feature activation | +9.49e-02 |
Pos logit contributions | |
advertisement | +0.128 |
IMAGES | +0.119 |
� | +0.115 |
� | +0.111 |
PDATE | +0.111 |
Neg logit contributions | |
erella | -0.126 |
inburgh | -0.107 |
achus | -0.105 |
lime | -0.097 |
cised | -0.090 |
Token | ". |
Token ID | 1911 |
Feature activation | +2.31e-01 |
Pos logit contributions | |
advertisement | +0.211 |
IMAGES | +0.197 |
� | +0.189 |
� | +0.184 |
PDATE | +0.182 |
Neg logit contributions | |
erella | -0.214 |
inburgh | -0.180 |
achus | -0.179 |
lime | -0.164 |
cised | -0.154 |
Token |
|
Token ID | 198 |
Feature activation | -5.36e-02 |
Pos logit contributions | |
advertisement | +0.518 |
IMAGES | +0.481 |
� | +0.463 |
PDATE | +0.454 |
� | +0.451 |
Neg logit contributions | |
erella | -0.517 |
inburgh | -0.435 |
achus | -0.432 |
lime | -0.396 |
cised | -0.370 |
Token |
|
Token ID | 198 |
Feature activation | -6.61e-02 |
Pos logit contributions | |
erella | +0.119 |
inburgh | +0.100 |
achus | +0.099 |
lime | +0.092 |
acted | +0.085 |
Neg logit contributions | |
advertisement | -0.120 |
IMAGES | -0.112 |
� | -0.108 |
PDATE | -0.106 |
� | -0.105 |
Token | Topics |
Token ID | 25902 |
Feature activation | +1.66e+00 |
Pos logit contributions | |
erella | +0.154 |
inburgh | +0.131 |
achus | +0.130 |
lime | +0.121 |
acted | +0.113 |
Neg logit contributions | |
advertisement | -0.140 |
IMAGES | -0.131 |
� | -0.125 |
PDATE | -0.124 |
� | -0.122 |
Token | : |
Token ID | 25 |
Feature activation | +3.06e-02 |
Pos logit contributions | |
advertisement | +3.354 |
IMAGES | +3.132 |
� | +2.852 |
OFF | +2.817 |
tti | +2.768 |
Neg logit contributions | |
erella | -4.099 |
inburgh | -3.549 |
achus | -3.540 |
lime | -3.203 |
izabeth | -3.035 |
Token | marine |
Token ID | 16050 |
Feature activation | -4.02e-01 |
Pos logit contributions | |
advertisement | +0.066 |
IMAGES | +0.061 |
� | +0.059 |
PDATE | +0.058 |
� | +0.058 |
Neg logit contributions | |
erella | -0.070 |
inburgh | -0.060 |
achus | -0.059 |
lime | -0.055 |
cised | -0.051 |
Token | - |
Token ID | 12 |
Feature activation | -8.75e-02 |
Pos logit contributions | |
erella | +0.846 |
inburgh | +0.707 |
achus | +0.702 |
lime | +0.645 |
acted | +0.598 |
Neg logit contributions | |
advertisement | -0.951 |
IMAGES | -0.882 |
� | -0.858 |
PDATE | -0.842 |
� | -0.831 |
Token | biology |
Token ID | 43592 |
Feature activation | +1.61e-01 |
Pos logit contributions | |
erella | +0.192 |
inburgh | +0.161 |
achus | +0.160 |
lime | +0.147 |
cised | +0.136 |
Neg logit contributions | |
advertisement | -0.200 |
IMAGES | -0.186 |
� | -0.180 |
� | -0.175 |
PDATE | -0.174 |
Token | , |
Token ID | 11 |
Feature activation | +4.92e-02 |
Pos logit contributions | |
advertisement | +0.371 |
IMAGES | +0.346 |
� | +0.333 |
� | +0.323 |
tti | +0.320 |
Neg logit contributions | |
erella | -0.350 |
inburgh | -0.294 |
achus | -0.291 |
lime | -0.265 |
cised | -0.248 |
Token | / |
Token ID | 14 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
advertisement | +0.344 |
IMAGES | +0.308 |
� | +0.299 |
PDATE | +0.290 |
� | +0.289 |
Neg logit contributions | |
erella | -0.499 |
inburgh | -0.433 |
achus | -0.428 |
lime | -0.399 |
cised | -0.379 |
Token | File |
Token ID | 8979 |
Feature activation | +4.16e-01 |
Pos logit contributions | |
advertisement | +0.213 |
IMAGES | +0.195 |
� | +0.186 |
PDATE | +0.180 |
� | +0.180 |
Neg logit contributions | |
erella | -0.295 |
inburgh | -0.255 |
achus | -0.250 |
lime | -0.236 |
illary | -0.222 |
Token | photo |
Token ID | 4590 |
Feature activation | +4.36e-01 |
Pos logit contributions | |
advertisement | +0.823 |
IMAGES | +0.742 |
� | +0.720 |
� | +0.697 |
PDATE | +0.696 |
Neg logit contributions | |
erella | -1.047 |
inburgh | -0.901 |
achus | -0.890 |
lime | -0.832 |
cised | -0.784 |
Token | TP |
Token ID | 24525 |
Feature activation | +4.38e-01 |
Pos logit contributions | |
advertisement | +0.720 |
IMAGES | +0.647 |
� | +0.615 |
� | +0.591 |
PDATE | +0.588 |
Neg logit contributions | |
erella | -1.231 |
inburgh | -1.082 |
achus | -1.073 |
lime | -1.004 |
cised | -0.957 |
Token | X |
Token ID | 55 |
Feature activation | +4.49e-01 |
Pos logit contributions | |
advertisement | +1.012 |
IMAGES | +0.925 |
� | +0.896 |
PDATE | +0.873 |
� | +0.870 |
Neg logit contributions | |
erella | -0.961 |
inburgh | -0.804 |
achus | -0.798 |
lime | -0.737 |
illary | -0.685 |
Token | IMAGES |
Token ID | 45325 |
Feature activation | +1.63e+00 |
Pos logit contributions | |
advertisement | +0.153 |
� | +0.042 |
PDATE | +0.034 |
tti | +0.018 |
� | +0.006 |
Neg logit contributions | |
erella | -1.851 |
inburgh | -1.707 |
achus | -1.702 |
lime | -1.619 |
illary | -1.588 |
Token | OF |
Token ID | 3963 |
Feature activation | -1.44e-02 |
Pos logit contributions | |
advertisement | +2.960 |
IMAGES | +2.814 |
� | +2.604 |
OFF | +2.531 |
� | +2.456 |
Neg logit contributions | |
erella | -4.275 |
inburgh | -3.762 |
achus | -3.747 |
lime | -3.451 |
cised | -3.275 |
Token | THE |
Token ID | 3336 |
Feature activation | -2.29e-01 |
Pos logit contributions | |
erella | +0.033 |
inburgh | +0.028 |
achus | +0.028 |
lime | +0.026 |
cised | +0.024 |
Neg logit contributions | |
advertisement | -0.031 |
IMAGES | -0.029 |
� | -0.028 |
� | -0.027 |
PDATE | -0.027 |
Token | DAY |
Token ID | 24644 |
Feature activation | +5.48e-01 |
Pos logit contributions | |
erella | +0.499 |
inburgh | +0.416 |
achus | +0.410 |
lime | +0.376 |
acted | +0.354 |
Neg logit contributions | |
advertisement | -0.520 |
IMAGES | -0.468 |
� | -0.460 |
PDATE | -0.458 |
� | -0.452 |
Token | ) |
Token ID | 8 |
Feature activation | -1.45e-01 |
Pos logit contributions | |
advertisement | +1.059 |
IMAGES | +0.957 |
� | +0.924 |
� | +0.898 |
PDATE | +0.895 |
Neg logit contributions | |
erella | -1.399 |
inburgh | -1.205 |
achus | -1.193 |
lime | -1.113 |
cised | -1.049 |
Token | Chris |
Token ID | 5180 |
Feature activation | -1.14e-01 |
Pos logit contributions | |
erella | +0.351 |
inburgh | +0.299 |
achus | +0.298 |
lime | +0.275 |
illary | +0.259 |
Neg logit contributions | |
advertisement | -0.297 |
IMAGES | -0.270 |
� | -0.261 |
PDATE | -0.257 |
� | -0.253 |
Token | . |
Token ID | 13 |
Feature activation | +6.36e-02 |
Pos logit contributions | |
advertisement | +0.740 |
IMAGES | +0.685 |
� | +0.660 |
PDATE | +0.641 |
� | +0.641 |
Neg logit contributions | |
erella | -0.792 |
inburgh | -0.673 |
achus | -0.667 |
lime | -0.615 |
cised | -0.576 |
Token | [ |
Token ID | 685 |
Feature activation | +7.03e-01 |
Pos logit contributions | |
advertisement | +0.167 |
IMAGES | +0.156 |
� | +0.152 |
� | +0.148 |
PDATE | +0.148 |
Neg logit contributions | |
erella | -0.118 |
inburgh | -0.096 |
achus | -0.095 |
lime | -0.085 |
cised | -0.077 |
Token | color |
Token ID | 8043 |
Feature activation | -2.04e-01 |
Pos logit contributions | |
advertisement | +1.416 |
IMAGES | +1.304 |
� | +1.248 |
� | +1.210 |
PDATE | +1.199 |
Neg logit contributions | |
erella | -1.729 |
inburgh | -1.486 |
achus | -1.475 |
lime | -1.360 |
cised | -1.285 |
Token | = |
Token ID | 28 |
Feature activation | +4.14e-01 |
Pos logit contributions | |
erella | +0.563 |
inburgh | +0.491 |
achus | +0.487 |
lime | +0.459 |
acted | +0.437 |
Neg logit contributions | |
advertisement | -0.349 |
IMAGES | -0.319 |
� | -0.304 |
PDATE | -0.298 |
� | -0.291 |
Token | red |
Token ID | 445 |
Feature activation | -2.83e-01 |
Pos logit contributions | |
advertisement | +1.194 |
IMAGES | +1.129 |
� | +1.097 |
� | +1.073 |
PDATE | +1.068 |
Neg logit contributions | |
erella | -0.658 |
inburgh | -0.514 |
achus | -0.507 |
lime | -0.440 |
cised | -0.396 |
Token | ][ |
Token ID | 7131 |
Feature activation | +1.57e+00 |
Pos logit contributions | |
erella | +0.647 |
inburgh | +0.550 |
achus | +0.541 |
lime | +0.500 |
acted | +0.473 |
Neg logit contributions | |
advertisement | -0.622 |
IMAGES | -0.577 |
� | -0.554 |
� | -0.538 |
PDATE | -0.537 |
Token | b |
Token ID | 65 |
Feature activation | -1.64e-01 |
Pos logit contributions | |
advertisement | +3.084 |
IMAGES | +2.785 |
� | +2.678 |
� | +2.596 |
tti | +2.584 |
Neg logit contributions | |
erella | -3.935 |
achus | -3.410 |
inburgh | -3.388 |
lime | -3.080 |
cised | -2.930 |
Token | ] |
Token ID | 60 |
Feature activation | +1.19e-01 |
Pos logit contributions | |
erella | +0.324 |
inburgh | +0.264 |
achus | +0.262 |
lime | +0.239 |
acted | +0.222 |
Neg logit contributions | |
advertisement | -0.408 |
IMAGES | -0.383 |
� | -0.370 |
PDATE | -0.369 |
� | -0.362 |
Token | U |
Token ID | 52 |
Feature activation | -2.34e-02 |
Pos logit contributions | |
advertisement | +0.228 |
IMAGES | +0.210 |
� | +0.200 |
� | +0.194 |
PDATE | +0.193 |
Neg logit contributions | |
erella | -0.306 |
inburgh | -0.264 |
achus | -0.262 |
lime | -0.244 |
cised | -0.230 |
Token | f |
Token ID | 69 |
Feature activation | -2.65e-01 |
Pos logit contributions | |
erella | +0.055 |
inburgh | +0.047 |
achus | +0.046 |
lime | +0.043 |
cised | +0.040 |
Neg logit contributions | |
advertisement | -0.050 |
IMAGES | -0.046 |
� | -0.044 |
� | -0.043 |
PDATE | -0.043 |
Token | ou |
Token ID | 280 |
Feature activation | -7.08e-01 |
Pos logit contributions | |
erella | +0.651 |
inburgh | +0.558 |
achus | +0.554 |
lime | +0.516 |
illary | +0.487 |
Neg logit contributions | |
advertisement | -0.541 |
IMAGES | -0.500 |
� | -0.474 |
PDATE | -0.469 |
� | -0.460 |
Token | 45 |
Token ID | 2231 |
Feature activation | -8.24e-02 |
Pos logit contributions | |
erella | +0.220 |
inburgh | +0.190 |
achus | +0.189 |
lime | +0.175 |
cised | +0.165 |
Neg logit contributions | |
advertisement | -0.167 |
IMAGES | -0.154 |
� | -0.147 |
� | -0.142 |
PDATE | -0.141 |
Token | 366 |
Token ID | 44856 |
Feature activation | -1.45e-01 |
Pos logit contributions | |
erella | +0.166 |
inburgh | +0.138 |
achus | +0.136 |
lime | +0.124 |
cised | +0.114 |
Neg logit contributions | |
advertisement | -0.202 |
IMAGES | -0.189 |
� | -0.183 |
� | -0.178 |
PDATE | -0.177 |
Token | . |
Token ID | 13 |
Feature activation | -1.79e-01 |
Pos logit contributions | |
erella | +0.327 |
inburgh | +0.277 |
achus | +0.274 |
lime | +0.252 |
cised | +0.236 |
Neg logit contributions | |
advertisement | -0.320 |
IMAGES | -0.297 |
� | -0.286 |
� | -0.278 |
PDATE | -0.276 |
Token |
|
Token ID | 198 |
Feature activation | -9.14e-02 |
Pos logit contributions | |
erella | +0.422 |
inburgh | +0.359 |
achus | +0.356 |
lime | +0.327 |
cised | +0.307 |
Neg logit contributions | |
advertisement | -0.380 |
IMAGES | -0.349 |
� | -0.334 |
� | -0.326 |
PDATE | -0.325 |
Token |
|
Token ID | 198 |
Feature activation | -8.85e-02 |
Pos logit contributions | |
erella | +0.204 |
inburgh | +0.172 |
achus | +0.170 |
lime | +0.157 |
acted | +0.146 |
Neg logit contributions | |
advertisement | -0.204 |
IMAGES | -0.191 |
� | -0.183 |
PDATE | -0.181 |
� | -0.179 |
Token | Topics |
Token ID | 25902 |
Feature activation | +1.56e+00 |
Pos logit contributions | |
erella | +0.170 |
inburgh | +0.139 |
achus | +0.137 |
lime | +0.125 |
acted | +0.115 |
Neg logit contributions | |
advertisement | -0.225 |
IMAGES | -0.213 |
� | -0.204 |
PDATE | -0.202 |
� | -0.200 |
Token | : |
Token ID | 25 |
Feature activation | +7.71e-03 |
Pos logit contributions | |
advertisement | +3.208 |
IMAGES | +2.991 |
� | +2.751 |
OFF | +2.709 |
tti | +2.669 |
Neg logit contributions | |
erella | -3.805 |
inburgh | -3.279 |
achus | -3.258 |
lime | -2.963 |
izabeth | -2.788 |
Token | community |
Token ID | 2055 |
Feature activation | -1.09e-01 |
Pos logit contributions | |
advertisement | +0.018 |
IMAGES | +0.016 |
� | +0.016 |
PDATE | +0.015 |
� | +0.015 |
Neg logit contributions | |
erella | -0.017 |
inburgh | -0.014 |
achus | -0.014 |
lime | -0.013 |
cised | -0.012 |
Token | - |
Token ID | 12 |
Feature activation | -1.41e-01 |
Pos logit contributions | |
erella | +0.224 |
inburgh | +0.188 |
achus | +0.185 |
lime | +0.169 |
acted | +0.158 |
Neg logit contributions | |
advertisement | -0.261 |
IMAGES | -0.242 |
PDATE | -0.236 |
� | -0.235 |
� | -0.228 |
Token | and |
Token ID | 392 |
Feature activation | -3.13e-02 |
Pos logit contributions | |
erella | +0.386 |
inburgh | +0.338 |
achus | +0.333 |
lime | +0.318 |
acted | +0.300 |
Neg logit contributions | |
advertisement | -0.241 |
IMAGES | -0.224 |
PDATE | -0.213 |
� | -0.208 |
� | -0.204 |
Token | - |
Token ID | 12 |
Feature activation | +6.05e-02 |
Pos logit contributions | |
erella | +0.070 |
inburgh | +0.059 |
achus | +0.059 |
lime | +0.054 |
cised | +0.050 |
Neg logit contributions | |
advertisement | -0.070 |
IMAGES | -0.065 |
� | -0.063 |
� | -0.061 |
tti | -0.060 |
Token | . |
Token ID | 13 |
Feature activation | -1.09e-01 |
Pos logit contributions | |
advertisement | +0.654 |
IMAGES | +0.606 |
� | +0.582 |
� | +0.565 |
PDATE | +0.564 |
Neg logit contributions | |
erella | -0.710 |
inburgh | -0.604 |
achus | -0.599 |
lime | -0.551 |
cised | -0.517 |
Token | [ |
Token ID | 685 |
Feature activation | +6.34e-01 |
Pos logit contributions | |
erella | +0.194 |
inburgh | +0.157 |
achus | +0.155 |
lime | +0.138 |
cised | +0.126 |
Neg logit contributions | |
advertisement | -0.292 |
IMAGES | -0.273 |
� | -0.266 |
� | -0.260 |
PDATE | -0.259 |
Token | color |
Token ID | 8043 |
Feature activation | -2.29e-01 |
Pos logit contributions | |
advertisement | +1.276 |
IMAGES | +1.176 |
� | +1.126 |
� | +1.091 |
PDATE | +1.081 |
Neg logit contributions | |
erella | -1.562 |
inburgh | -1.342 |
achus | -1.331 |
lime | -1.229 |
cised | -1.161 |
Token | = |
Token ID | 28 |
Feature activation | +3.50e-01 |
Pos logit contributions | |
erella | +0.626 |
inburgh | +0.544 |
achus | +0.541 |
lime | +0.510 |
acted | +0.486 |
Neg logit contributions | |
advertisement | -0.394 |
IMAGES | -0.361 |
� | -0.343 |
PDATE | -0.342 |
� | -0.330 |
Token | red |
Token ID | 445 |
Feature activation | -3.34e-01 |
Pos logit contributions | |
advertisement | +1.033 |
IMAGES | +0.977 |
� | +0.950 |
� | +0.931 |
PDATE | +0.926 |
Neg logit contributions | |
erella | -0.536 |
inburgh | -0.414 |
achus | -0.408 |
lime | -0.351 |
cised | -0.314 |
Token | ][ |
Token ID | 7131 |
Feature activation | +1.56e+00 |
Pos logit contributions | |
erella | +0.765 |
inburgh | +0.652 |
achus | +0.642 |
lime | +0.593 |
acted | +0.558 |
Neg logit contributions | |
advertisement | -0.729 |
IMAGES | -0.676 |
� | -0.649 |
� | -0.630 |
PDATE | -0.628 |
Token | b |
Token ID | 65 |
Feature activation | -2.83e-01 |
Pos logit contributions | |
advertisement | +3.054 |
IMAGES | +2.752 |
� | +2.634 |
tti | +2.564 |
� | +2.552 |
Neg logit contributions | |
erella | -3.899 |
inburgh | -3.376 |
achus | -3.369 |
lime | -3.049 |
cised | -2.900 |
Token | ] |
Token ID | 60 |
Feature activation | +5.29e-02 |
Pos logit contributions | |
erella | +0.548 |
inburgh | +0.443 |
achus | +0.440 |
lime | +0.403 |
acted | +0.377 |
Neg logit contributions | |
advertisement | -0.713 |
IMAGES | -0.674 |
PDATE | -0.660 |
� | -0.651 |
� | -0.637 |
Token | Mega |
Token ID | 43471 |
Feature activation | -2.15e-01 |
Pos logit contributions | |
advertisement | +0.100 |
IMAGES | +0.092 |
� | +0.088 |
PDATE | +0.085 |
� | +0.085 |
Neg logit contributions | |
erella | -0.137 |
inburgh | -0.118 |
achus | -0.117 |
lime | -0.109 |
acted | -0.103 |
Token | Pony |
Token ID | 30377 |
Feature activation | -1.05e-01 |
Pos logit contributions | |
erella | +0.479 |
inburgh | +0.403 |
achus | +0.402 |
lime | +0.367 |
cised | +0.342 |
Neg logit contributions | |
advertisement | -0.484 |
IMAGES | -0.451 |
� | -0.434 |
PDATE | -0.419 |
� | -0.419 |
Token | [/ |
Token ID | 13412 |
Feature activation | -3.52e-01 |
Pos logit contributions | |
erella | +0.196 |
inburgh | +0.161 |
achus | +0.160 |
lime | +0.143 |
cised | +0.130 |
Neg logit contributions | |
advertisement | -0.273 |
IMAGES | -0.257 |
� | -0.248 |
PDATE | -0.243 |
� | -0.241 |
Token | News |
Token ID | 3000 |
Feature activation | -3.05e-01 |
Pos logit contributions | |
advertisement | +0.127 |
IMAGES | +0.117 |
� | +0.113 |
PDATE | +0.109 |
� | +0.109 |
Neg logit contributions | |
erella | -0.140 |
inburgh | -0.120 |
achus | -0.118 |
lime | -0.109 |
cised | -0.102 |
Token | 24 |
Token ID | 1987 |
Feature activation | -1.73e-01 |
Pos logit contributions | |
erella | +0.689 |
inburgh | +0.586 |
achus | +0.580 |
lime | +0.531 |
cised | +0.497 |
Neg logit contributions | |
advertisement | -0.673 |
IMAGES | -0.623 |
� | -0.599 |
PDATE | -0.583 |
� | -0.581 |
Token | . |
Token ID | 13 |
Feature activation | -2.11e-01 |
Pos logit contributions | |
erella | +0.399 |
inburgh | +0.339 |
achus | +0.336 |
lime | +0.309 |
cised | +0.290 |
Neg logit contributions | |
advertisement | -0.375 |
IMAGES | -0.348 |
� | -0.334 |
� | -0.324 |
PDATE | -0.323 |
Token |
|
Token ID | 198 |
Feature activation | -2.66e-02 |
Pos logit contributions | |
erella | +0.510 |
inburgh | +0.437 |
achus | +0.433 |
lime | +0.400 |
acted | +0.376 |
Neg logit contributions | |
advertisement | -0.434 |
IMAGES | -0.398 |
� | -0.382 |
� | -0.370 |
PDATE | -0.370 |
Token |
|
Token ID | 198 |
Feature activation | -3.47e-02 |
Pos logit contributions | |
erella | +0.059 |
inburgh | +0.050 |
achus | +0.050 |
lime | +0.046 |
cised | +0.042 |
Neg logit contributions | |
advertisement | -0.059 |
IMAGES | -0.055 |
� | -0.053 |
PDATE | -0.052 |
� | -0.052 |
Token | Topics |
Token ID | 25902 |
Feature activation | +1.50e+00 |
Pos logit contributions | |
erella | +0.061 |
inburgh | +0.049 |
achus | +0.049 |
lime | +0.044 |
acted | +0.040 |
Neg logit contributions | |
advertisement | -0.093 |
IMAGES | -0.088 |
� | -0.085 |
PDATE | -0.084 |
� | -0.084 |
Token | : |
Token ID | 25 |
Feature activation | -3.33e-02 |
Pos logit contributions | |
advertisement | +3.070 |
IMAGES | +2.859 |
� | +2.641 |
OFF | +2.594 |
tti | +2.556 |
Neg logit contributions | |
erella | -3.636 |
inburgh | -3.126 |
achus | -3.114 |
lime | -2.834 |
izabeth | -2.670 |
Token | beef |
Token ID | 12023 |
Feature activation | -2.54e-01 |
Pos logit contributions | |
erella | +0.076 |
inburgh | +0.065 |
achus | +0.064 |
lime | +0.059 |
acted | +0.055 |
Neg logit contributions | |
advertisement | -0.073 |
IMAGES | -0.067 |
� | -0.065 |
PDATE | -0.063 |
� | -0.063 |
Token | - |
Token ID | 12 |
Feature activation | -2.73e-01 |
Pos logit contributions | |
erella | +0.520 |
inburgh | +0.441 |
achus | +0.434 |
lime | +0.395 |
acted | +0.366 |
Neg logit contributions | |
advertisement | -0.611 |
IMAGES | -0.569 |
� | -0.546 |
PDATE | -0.541 |
� | -0.534 |
Token | c |
Token ID | 66 |
Feature activation | -2.02e-01 |
Pos logit contributions | |
erella | +0.694 |
inburgh | +0.601 |
achus | +0.594 |
lime | +0.556 |
acted | +0.523 |
Neg logit contributions | |
advertisement | -0.525 |
IMAGES | -0.485 |
� | -0.462 |
PDATE | -0.451 |
� | -0.448 |
Token | attle |
Token ID | 1999 |
Feature activation | +1.05e-01 |
Pos logit contributions | |
erella | +0.424 |
inburgh | +0.354 |
achus | +0.351 |
lime | +0.318 |
illary | +0.296 |
Neg logit contributions | |
advertisement | -0.481 |
IMAGES | -0.450 |
� | -0.434 |
PDATE | -0.425 |
� | -0.423 |
Token | , |
Token ID | 11 |
Feature activation | +2.44e-02 |
Pos logit contributions | |
advertisement | +0.061 |
IMAGES | +0.057 |
� | +0.055 |
� | +0.053 |
PDATE | +0.053 |
Neg logit contributions | |
erella | -0.062 |
inburgh | -0.052 |
achus | -0.052 |
lime | -0.047 |
cised | -0.044 |
Token | Clinton |
Token ID | 2605 |
Feature activation | -1.13e-01 |
Pos logit contributions | |
advertisement | +0.052 |
IMAGES | +0.048 |
� | +0.046 |
PDATE | +0.045 |
� | +0.045 |
Neg logit contributions | |
erella | -0.058 |
inburgh | -0.049 |
achus | -0.049 |
lime | -0.045 |
cised | -0.042 |
Token | Foundation |
Token ID | 5693 |
Feature activation | -2.00e-01 |
Pos logit contributions | |
erella | +0.242 |
inburgh | +0.202 |
achus | +0.201 |
lime | +0.182 |
cised | +0.170 |
Neg logit contributions | |
advertisement | -0.263 |
IMAGES | -0.246 |
� | -0.237 |
PDATE | -0.231 |
� | -0.230 |
Token | President |
Token ID | 1992 |
Feature activation | -9.21e-02 |
Pos logit contributions | |
erella | +0.466 |
inburgh | +0.394 |
achus | +0.392 |
lime | +0.361 |
illary | +0.340 |
Neg logit contributions | |
advertisement | -0.424 |
IMAGES | -0.394 |
� | -0.380 |
PDATE | -0.376 |
� | -0.369 |
Token | Donna |
Token ID | 29216 |
Feature activation | -1.09e-01 |
Pos logit contributions | |
erella | +0.231 |
inburgh | +0.198 |
achus | +0.196 |
lime | +0.183 |
acted | +0.174 |
Neg logit contributions | |
advertisement | -0.179 |
IMAGES | -0.164 |
PDATE | -0.159 |
� | -0.158 |
� | -0.153 |
Token | Shal |
Token ID | 47974 |
Feature activation | +1.45e+00 |
Pos logit contributions | |
erella | +0.292 |
inburgh | +0.252 |
achus | +0.249 |
lime | +0.235 |
acted | +0.226 |
Neg logit contributions | |
advertisement | -0.198 |
IMAGES | -0.179 |
� | -0.170 |
PDATE | -0.170 |
� | -0.162 |
Token | ala |
Token ID | 6081 |
Feature activation | -2.88e-01 |
Pos logit contributions | |
advertisement | +4.076 |
IMAGES | +3.809 |
� | +3.751 |
tti | +3.680 |
� | +3.649 |
Neg logit contributions | |
erella | -2.390 |
inburgh | -1.913 |
achus | -1.883 |
lime | -1.631 |
cised | -1.481 |
Token | summed |
Token ID | 32794 |
Feature activation | +4.04e-01 |
Pos logit contributions | |
erella | +0.607 |
inburgh | +0.505 |
achus | +0.501 |
lime | +0.458 |
acted | +0.427 |
Neg logit contributions | |
advertisement | -0.680 |
IMAGES | -0.638 |
� | -0.612 |
PDATE | -0.603 |
� | -0.596 |
Token | up |
Token ID | 510 |
Feature activation | -4.16e-02 |
Pos logit contributions | |
advertisement | +0.888 |
IMAGES | +0.823 |
� | +0.791 |
� | +0.769 |
PDATE | +0.765 |
Neg logit contributions | |
erella | -0.918 |
inburgh | -0.780 |
achus | -0.773 |
lime | -0.709 |
cised | -0.663 |
Token | the |
Token ID | 262 |
Feature activation | +1.53e-02 |
Pos logit contributions | |
erella | +0.087 |
inburgh | +0.073 |
achus | +0.072 |
lime | +0.066 |
acted | +0.061 |
Neg logit contributions | |
advertisement | -0.098 |
IMAGES | -0.092 |
� | -0.089 |
PDATE | -0.087 |
� | -0.086 |
Token | complicated |
Token ID | 8253 |
Feature activation | +7.85e-02 |
Pos logit contributions | |
advertisement | +0.032 |
IMAGES | +0.030 |
� | +0.029 |
PDATE | +0.028 |
� | +0.028 |
Neg logit contributions | |
erella | -0.036 |
inburgh | -0.031 |
achus | -0.030 |
lime | -0.028 |
acted | -0.026 |
Token | uk |
Token ID | 2724 |
Feature activation | +4.89e-01 |
Pos logit contributions | |
erella | +0.151 |
inburgh | +0.132 |
achus | +0.131 |
lime | +0.122 |
cised | +0.116 |
Neg logit contributions | |
advertisement | -0.099 |
IMAGES | -0.091 |
� | -0.086 |
� | -0.083 |
PDATE | -0.082 |
Token | / |
Token ID | 14 |
Feature activation | +4.37e-01 |
Pos logit contributions | |
advertisement | +1.010 |
IMAGES | +0.931 |
� | +0.892 |
� | +0.859 |
tti | +0.853 |
Neg logit contributions | |
erella | -1.181 |
inburgh | -1.016 |
achus | -1.009 |
lime | -0.928 |
cised | -0.876 |
Token | in |
Token ID | 259 |
Feature activation | +1.02e-03 |
Pos logit contributions | |
advertisement | +0.860 |
IMAGES | +0.797 |
� | +0.762 |
PDATE | +0.745 |
� | +0.739 |
Neg logit contributions | |
erella | -1.092 |
inburgh | -0.943 |
achus | -0.930 |
lime | -0.865 |
acted | -0.815 |
Token | coming |
Token ID | 4976 |
Feature activation | +1.91e-01 |
Pos logit contributions | |
advertisement | +0.002 |
IMAGES | +0.002 |
� | +0.002 |
� | +0.002 |
PDATE | +0.002 |
Neg logit contributions | |
erella | -0.002 |
inburgh | -0.002 |
achus | -0.002 |
lime | -0.002 |
cised | -0.001 |
Token | / |
Token ID | 14 |
Feature activation | +1.92e-01 |
Pos logit contributions | |
advertisement | +0.417 |
IMAGES | +0.388 |
PDATE | +0.373 |
� | +0.373 |
� | +0.363 |
Neg logit contributions | |
erella | -0.431 |
inburgh | -0.369 |
achus | -0.361 |
lime | -0.335 |
acted | -0.313 |
Token | article |
Token ID | 20205 |
Feature activation | +1.45e+00 |
Pos logit contributions | |
advertisement | +0.295 |
IMAGES | +0.267 |
� | +0.253 |
PDATE | +0.250 |
� | +0.243 |
Neg logit contributions | |
erella | -0.559 |
inburgh | -0.496 |
achus | -0.490 |
lime | -0.460 |
acted | -0.440 |
Token | 35 |
Token ID | 2327 |
Feature activation | -1.49e-02 |
Pos logit contributions | |
advertisement | +2.831 |
IMAGES | +2.534 |
� | +2.417 |
� | +2.342 |
OFF | +2.311 |
Neg logit contributions | |
erella | -3.659 |
inburgh | -3.196 |
achus | -3.174 |
lime | -2.949 |
acted | -2.789 |
Token | 79 |
Token ID | 3720 |
Feature activation | +6.52e-03 |
Pos logit contributions | |
erella | +0.034 |
inburgh | +0.029 |
achus | +0.029 |
lime | +0.026 |
cised | +0.025 |
Neg logit contributions | |
advertisement | -0.033 |
IMAGES | -0.030 |
� | -0.029 |
� | -0.028 |
PDATE | -0.028 |
Token | 23 |
Token ID | 1954 |
Feature activation | -2.89e-01 |
Pos logit contributions | |
advertisement | +0.013 |
IMAGES | +0.012 |
� | +0.011 |
PDATE | +0.011 |
� | +0.011 |
Neg logit contributions | |
erella | -0.017 |
inburgh | -0.014 |
achus | -0.014 |
lime | -0.013 |
cised | -0.012 |
Token | 12 |
Token ID | 1065 |
Feature activation | -2.04e-01 |
Pos logit contributions | |
erella | +0.725 |
inburgh | +0.620 |
achus | +0.615 |
lime | +0.576 |
acted | +0.542 |
Neg logit contributions | |
advertisement | -0.564 |
IMAGES | -0.522 |
� | -0.497 |
PDATE | -0.496 |
� | -0.484 |
Token | . |
Token ID | 13 |
Feature activation | +9.80e-02 |
Pos logit contributions | |
erella | +0.461 |
inburgh | +0.389 |
achus | +0.385 |
lime | +0.356 |
acted | +0.333 |
Neg logit contributions | |
advertisement | -0.447 |
IMAGES | -0.420 |
� | -0.404 |
PDATE | -0.403 |
� | -0.393 |
Token | all |
Token ID | 477 |
Feature activation | -2.67e-02 |
Pos logit contributions | |
advertisement | +0.423 |
IMAGES | +0.391 |
� | +0.374 |
PDATE | +0.367 |
� | +0.363 |
Neg logit contributions | |
erella | -0.447 |
inburgh | -0.380 |
achus | -0.376 |
lime | -0.347 |
acted | -0.326 |
Token | the |
Token ID | 262 |
Feature activation | +3.80e-01 |
Pos logit contributions | |
erella | +0.060 |
inburgh | +0.051 |
achus | +0.050 |
lime | +0.047 |
illary | +0.044 |
Neg logit contributions | |
advertisement | -0.059 |
IMAGES | -0.055 |
� | -0.053 |
PDATE | -0.052 |
� | -0.051 |
Token | evidence |
Token ID | 2370 |
Feature activation | +1.46e-01 |
Pos logit contributions | |
advertisement | +0.796 |
IMAGES | +0.729 |
� | +0.703 |
PDATE | +0.689 |
� | +0.681 |
Neg logit contributions | |
erella | -0.906 |
inburgh | -0.773 |
achus | -0.763 |
lime | -0.710 |
acted | -0.672 |
Token | that |
Token ID | 326 |
Feature activation | +1.17e-01 |
Pos logit contributions | |
advertisement | +0.313 |
IMAGES | +0.291 |
� | +0.277 |
PDATE | +0.270 |
� | +0.267 |
Neg logit contributions | |
erella | -0.340 |
inburgh | -0.289 |
achus | -0.286 |
lime | -0.265 |
acted | -0.250 |
Token | the |
Token ID | 262 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
advertisement | +0.269 |
IMAGES | +0.248 |
� | +0.240 |
PDATE | +0.236 |
� | +0.232 |
Neg logit contributions | |
erella | -0.255 |
inburgh | -0.214 |
achus | -0.212 |
lime | -0.194 |
illary | -0.181 |
Token | Commonwealth |
Token ID | 17165 |
Feature activation | +1.43e+00 |
Pos logit contributions | |
advertisement | +0.228 |
IMAGES | +0.209 |
� | +0.201 |
PDATE | +0.197 |
� | +0.195 |
Neg logit contributions | |
erella | -0.249 |
inburgh | -0.212 |
achus | -0.208 |
lime | -0.193 |
acted | -0.184 |
Token | possesses |
Token ID | 22194 |
Feature activation | +3.40e-01 |
Pos logit contributions | |
advertisement | +3.050 |
IMAGES | +2.864 |
� | +2.730 |
� | +2.670 |
OFF | +2.635 |
Neg logit contributions | |
erella | -3.308 |
inburgh | -2.815 |
achus | -2.784 |
lime | -2.564 |
cised | -2.412 |
Token | so |
Token ID | 523 |
Feature activation | +2.62e-02 |
Pos logit contributions | |
advertisement | +0.725 |
IMAGES | +0.678 |
� | +0.649 |
PDATE | +0.634 |
� | +0.625 |
Neg logit contributions | |
erella | -0.796 |
inburgh | -0.678 |
achus | -0.670 |
lime | -0.619 |
illary | -0.585 |
Token | that |
Token ID | 326 |
Feature activation | -5.34e-01 |
Pos logit contributions | |
advertisement | +0.063 |
IMAGES | +0.058 |
� | +0.056 |
PDATE | +0.055 |
� | +0.054 |
Neg logit contributions | |
erella | -0.056 |
inburgh | -0.046 |
achus | -0.045 |
lime | -0.041 |
acted | -0.039 |
Token | they |
Token ID | 484 |
Feature activation | -1.10e-01 |
Pos logit contributions | |
erella | +1.069 |
inburgh | +0.874 |
achus | +0.866 |
lime | +0.784 |
illary | +0.737 |
Neg logit contributions | |
advertisement | -1.322 |
IMAGES | -1.227 |
� | -1.178 |
PDATE | -1.178 |
� | -1.142 |
Token | can |
Token ID | 460 |
Feature activation | -1.02e-01 |
Pos logit contributions | |
erella | +0.193 |
inburgh | +0.152 |
achus | +0.151 |
lime | +0.135 |
illary | +0.123 |
Neg logit contributions | |
advertisement | -0.303 |
IMAGES | -0.284 |
� | -0.276 |
PDATE | -0.272 |
� | -0.268 |
Token | , |
Token ID | 11 |
Feature activation | -1.72e-02 |
Pos logit contributions | |
erella | +0.670 |
inburgh | +0.563 |
achus | +0.562 |
lime | +0.517 |
acted | +0.482 |
Neg logit contributions | |
advertisement | -0.668 |
IMAGES | -0.623 |
� | -0.592 |
PDATE | -0.581 |
� | -0.579 |
Token | and |
Token ID | 290 |
Feature activation | -1.47e-01 |
Pos logit contributions | |
erella | +0.036 |
inburgh | +0.029 |
achus | +0.029 |
lime | +0.027 |
acted | +0.025 |
Neg logit contributions | |
advertisement | -0.041 |
IMAGES | -0.039 |
� | -0.037 |
PDATE | -0.036 |
� | -0.036 |
Token | then |
Token ID | 788 |
Feature activation | +7.21e-02 |
Pos logit contributions | |
erella | +0.334 |
inburgh | +0.280 |
achus | +0.280 |
lime | +0.257 |
acted | +0.241 |
Neg logit contributions | |
advertisement | -0.323 |
IMAGES | -0.300 |
� | -0.289 |
PDATE | -0.281 |
� | -0.280 |
Token | you |
Token ID | 345 |
Feature activation | -2.53e-01 |
Pos logit contributions | |
advertisement | +0.157 |
IMAGES | +0.146 |
� | +0.141 |
PDATE | +0.137 |
� | +0.137 |
Neg logit contributions | |
erella | -0.165 |
inburgh | -0.139 |
achus | -0.139 |
lime | -0.128 |
acted | -0.119 |
Token | Special |
Token ID | 6093 |
Feature activation | -3.77e-01 |
Pos logit contributions | |
erella | +0.558 |
inburgh | +0.467 |
achus | +0.464 |
lime | +0.427 |
acted | +0.396 |
Neg logit contributions | |
advertisement | -0.577 |
IMAGES | -0.539 |
� | -0.516 |
� | -0.503 |
PDATE | -0.502 |
Token | Summon |
Token ID | 19515 |
Feature activation | -1.59e+00 |
Pos logit contributions | |
erella | +0.816 |
achus | +0.678 |
inburgh | +0.678 |
lime | +0.628 |
acted | +0.581 |
Neg logit contributions | |
advertisement | -0.874 |
IMAGES | -0.812 |
� | -0.779 |
PDATE | -0.762 |
� | -0.755 |
Token | an |
Token ID | 281 |
Feature activation | -6.63e-02 |
Pos logit contributions | |
erella | +3.160 |
inburgh | +2.536 |
achus | +2.493 |
lime | +2.363 |
ウ | +2.240 |
Neg logit contributions | |
advertisement | -3.828 |
IMAGES | -3.659 |
� | -3.424 |
PDATE | -3.413 |
� | -3.364 |
Token | ATK |
Token ID | 24526 |
Feature activation | -6.66e-01 |
Pos logit contributions | |
erella | +0.155 |
inburgh | +0.129 |
achus | +0.127 |
lime | +0.120 |
acted | +0.113 |
Neg logit contributions | |
advertisement | -0.142 |
IMAGES | -0.132 |
� | -0.127 |
PDATE | -0.123 |
� | -0.123 |
Token | 0 |
Token ID | 657 |
Feature activation | -5.88e-02 |
Pos logit contributions | |
erella | +1.494 |
inburgh | +1.242 |
achus | +1.225 |
lime | +1.148 |
illary | +1.091 |
Neg logit contributions | |
advertisement | -1.488 |
IMAGES | -1.384 |
� | -1.308 |
� | -1.272 |
PDATE | -1.266 |
Token | / |
Token ID | 14 |
Feature activation | -4.22e-02 |
Pos logit contributions | |
erella | +0.127 |
inburgh | +0.105 |
achus | +0.105 |
lime | +0.097 |
acted | +0.092 |
Neg logit contributions | |
advertisement | -0.136 |
IMAGES | -0.127 |
� | -0.121 |
� | -0.118 |
PDATE | -0.117 |
Token | DEF |
Token ID | 32988 |
Feature activation | -7.73e-01 |
Pos logit contributions | |
erella | +0.117 |
inburgh | +0.101 |
achus | +0.100 |
lime | +0.095 |
ウ | +0.091 |
Neg logit contributions | |
advertisement | -0.071 |
IMAGES | -0.066 |
� | -0.060 |
PDATE | -0.060 |
� | -0.060 |
Token | a |
Token ID | 257 |
Feature activation | -4.66e-01 |
Pos logit contributions | |
erella | +0.600 |
inburgh | +0.501 |
achus | +0.495 |
lime | +0.457 |
ウ | +0.429 |
Neg logit contributions | |
advertisement | -0.640 |
IMAGES | -0.601 |
� | -0.569 |
PDATE | -0.566 |
� | -0.557 |
Token | commonly |
Token ID | 8811 |
Feature activation | -3.44e-01 |
Pos logit contributions | |
erella | +1.121 |
achus | +0.943 |
inburgh | +0.942 |
lime | +0.876 |
acted | +0.828 |
Neg logit contributions | |
advertisement | -0.963 |
IMAGES | -0.889 |
� | -0.835 |
PDATE | -0.834 |
� | -0.820 |
Token | used |
Token ID | 973 |
Feature activation | +4.17e-02 |
Pos logit contributions | |
erella | +0.837 |
inburgh | +0.705 |
achus | +0.699 |
lime | +0.657 |
acted | +0.628 |
Neg logit contributions | |
advertisement | -0.704 |
IMAGES | -0.660 |
� | -0.613 |
PDATE | -0.610 |
� | -0.596 |
Token | effect |
Token ID | 1245 |
Feature activation | -9.94e-01 |
Pos logit contributions | |
advertisement | +0.086 |
IMAGES | +0.081 |
� | +0.076 |
PDATE | +0.075 |
� | +0.074 |
Neg logit contributions | |
erella | -0.100 |
inburgh | -0.085 |
achus | -0.084 |
lime | -0.079 |
ウ | -0.074 |
Token | new |
Token ID | 649 |
Feature activation | -1.43e-01 |
Pos logit contributions | |
erella | +2.125 |
inburgh | +1.789 |
achus | +1.770 |
lime | +1.638 |
illary | +1.543 |
Neg logit contributions | |
advertisement | -2.262 |
IMAGES | -2.149 |
PDATE | -2.009 |
� | -2.003 |
� | -1.990 |
Token | life |
Token ID | 1204 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
erella | +0.353 |
inburgh | +0.300 |
achus | +0.298 |
lime | +0.280 |
illary | +0.262 |
Neg logit contributions | |
advertisement | -0.289 |
IMAGES | -0.266 |
� | -0.253 |
PDATE | -0.249 |
� | -0.245 |
Token | , |
Token ID | 11 |
Feature activation | +1.73e-01 |
Pos logit contributions | |
erella | +2.869 |
inburgh | +2.424 |
achus | +2.417 |
lime | +2.265 |
illary | +2.082 |
Neg logit contributions | |
advertisement | -3.131 |
IMAGES | -2.899 |
PDATE | -2.698 |
� | -2.697 |
� | -2.650 |
Token | as |
Token ID | 355 |
Feature activation | -5.58e-03 |
Pos logit contributions | |
advertisement | +0.373 |
IMAGES | +0.350 |
� | +0.333 |
PDATE | +0.327 |
� | +0.324 |
Neg logit contributions | |
erella | -0.398 |
inburgh | -0.336 |
achus | -0.335 |
lime | -0.312 |
acted | -0.288 |
Token | an |
Token ID | 281 |
Feature activation | +9.10e-02 |
Pos logit contributions | |
erella | +0.014 |
inburgh | +0.012 |
achus | +0.012 |
lime | +0.011 |
acted | +0.010 |
Neg logit contributions | |
advertisement | -0.011 |
IMAGES | -0.010 |
� | -0.010 |
PDATE | -0.009 |
� | -0.009 |
Token | alternative |
Token ID | 5559 |
Feature activation | -1.78e-01 |
Pos logit contributions | |
advertisement | +0.185 |
IMAGES | +0.172 |
� | +0.163 |
� | +0.158 |
PDATE | +0.158 |
Neg logit contributions | |
erella | -0.222 |
inburgh | -0.190 |
achus | -0.187 |
lime | -0.174 |
acted | -0.164 |
Token | to |
Token ID | 284 |
Feature activation | -1.84e-01 |
Pos logit contributions | |
erella | +0.407 |
inburgh | +0.345 |
achus | +0.341 |
lime | +0.317 |
ウ | +0.296 |
Neg logit contributions | |
advertisement | -0.387 |
IMAGES | -0.365 |
� | -0.342 |
PDATE | -0.337 |
� | -0.336 |
Token | you |
Token ID | 345 |
Feature activation | -1.18e-01 |
Pos logit contributions | |
advertisement | +0.072 |
IMAGES | +0.068 |
� | +0.065 |
PDATE | +0.063 |
� | +0.063 |
Neg logit contributions | |
erella | -0.068 |
inburgh | -0.057 |
achus | -0.057 |
lime | -0.052 |
acted | -0.049 |
Token | cannot |
Token ID | 2314 |
Feature activation | -3.69e-03 |
Pos logit contributions | |
erella | +0.244 |
inburgh | +0.200 |
achus | +0.199 |
lime | +0.182 |
acted | +0.169 |
Neg logit contributions | |
advertisement | -0.284 |
IMAGES | -0.266 |
� | -0.255 |
PDATE | -0.251 |
� | -0.248 |
Token | Normal |
Token ID | 14435 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
erella | +0.010 |
achus | +0.008 |
inburgh | +0.008 |
lime | +0.008 |
acted | +0.007 |
Neg logit contributions | |
advertisement | -0.007 |
IMAGES | -0.007 |
� | -0.006 |
PDATE | -0.006 |
� | -0.006 |
Token | or |
Token ID | 393 |
Feature activation | -4.99e-02 |
Pos logit contributions | |
advertisement | +0.268 |
IMAGES | +0.249 |
� | +0.237 |
PDATE | +0.232 |
� | +0.231 |
Neg logit contributions | |
erella | -0.246 |
inburgh | -0.206 |
achus | -0.204 |
lime | -0.188 |
acted | -0.175 |
Token | Special |
Token ID | 6093 |
Feature activation | -2.54e-01 |
Pos logit contributions | |
erella | +0.132 |
achus | +0.114 |
inburgh | +0.113 |
lime | +0.107 |
acted | +0.101 |
Neg logit contributions | |
advertisement | -0.092 |
IMAGES | -0.083 |
� | -0.079 |
� | -0.076 |
PDATE | -0.076 |
Token | Summon |
Token ID | 19515 |
Feature activation | -1.34e+00 |
Pos logit contributions | |
erella | +0.530 |
inburgh | +0.438 |
achus | +0.436 |
lime | +0.404 |
acted | +0.373 |
Neg logit contributions | |
advertisement | -0.610 |
IMAGES | -0.568 |
� | -0.543 |
PDATE | -0.530 |
� | -0.525 |
Token | monsters |
Token ID | 10963 |
Feature activation | -3.68e-01 |
Pos logit contributions | |
erella | +2.729 |
inburgh | +2.212 |
achus | +2.208 |
lime | +2.055 |
ウ | +1.957 |
Neg logit contributions | |
advertisement | -3.155 |
IMAGES | -2.995 |
� | -2.804 |
PDATE | -2.796 |
� | -2.782 |
Token | , |
Token ID | 11 |
Feature activation | +3.14e-02 |
Pos logit contributions | |
erella | +0.786 |
inburgh | +0.656 |
achus | +0.652 |
lime | +0.602 |
acted | +0.563 |
Neg logit contributions | |
advertisement | -0.854 |
IMAGES | -0.806 |
� | -0.766 |
� | -0.749 |
PDATE | -0.747 |
Token | except |
Token ID | 2845 |
Feature activation | +5.99e-02 |
Pos logit contributions | |
advertisement | +0.067 |
IMAGES | +0.062 |
� | +0.060 |
PDATE | +0.058 |
� | +0.058 |
Neg logit contributions | |
erella | -0.073 |
inburgh | -0.062 |
achus | -0.062 |
lime | -0.058 |
acted | -0.054 |
Token | for |
Token ID | 329 |
Feature activation | +6.44e-02 |
Pos logit contributions | |
advertisement | +0.135 |
IMAGES | +0.127 |
� | +0.121 |
PDATE | +0.119 |
� | +0.118 |
Neg logit contributions | |
erella | -0.133 |
achus | -0.110 |
inburgh | -0.110 |
lime | -0.101 |
acted | -0.095 |
Token | � |
Token ID | 564 |
Feature activation | +3.11e-02 |
Pos logit contributions | |
advertisement | +0.135 |
IMAGES | +0.126 |
� | +0.120 |
PDATE | +0.118 |
� | +0.117 |
Neg logit contributions | |
erella | -0.152 |
achus | -0.128 |
inburgh | -0.127 |
lime | -0.119 |
acted | -0.112 |
Token | between |
Token ID | 1022 |
Feature activation | -6.39e-02 |
Pos logit contributions | |
advertisement | +0.185 |
IMAGES | +0.173 |
� | +0.167 |
PDATE | +0.162 |
� | +0.161 |
Neg logit contributions | |
erella | -0.172 |
inburgh | -0.144 |
achus | -0.143 |
lime | -0.130 |
illary | -0.121 |
Token | the |
Token ID | 262 |
Feature activation | -1.74e-01 |
Pos logit contributions | |
erella | +0.146 |
inburgh | +0.124 |
achus | +0.123 |
lime | +0.113 |
acted | +0.107 |
Neg logit contributions | |
advertisement | -0.140 |
IMAGES | -0.129 |
� | -0.125 |
PDATE | -0.122 |
� | -0.121 |
Token | two |
Token ID | 734 |
Feature activation | +6.48e-02 |
Pos logit contributions | |
erella | +0.388 |
inburgh | +0.328 |
achus | +0.327 |
lime | +0.302 |
acted | +0.283 |
Neg logit contributions | |
advertisement | -0.388 |
IMAGES | -0.356 |
� | -0.345 |
PDATE | -0.339 |
� | -0.334 |
Token | countries |
Token ID | 2678 |
Feature activation | -5.24e-01 |
Pos logit contributions | |
advertisement | +0.145 |
IMAGES | +0.134 |
� | +0.130 |
PDATE | +0.127 |
� | +0.126 |
Neg logit contributions | |
erella | -0.144 |
inburgh | -0.122 |
achus | -0.121 |
lime | -0.112 |
acted | -0.104 |
Token | broke |
Token ID | 6265 |
Feature activation | -1.44e-01 |
Pos logit contributions | |
erella | +1.083 |
inburgh | +0.897 |
achus | +0.894 |
lime | +0.815 |
illary | +0.759 |
Neg logit contributions | |
advertisement | -1.254 |
IMAGES | -1.172 |
� | -1.142 |
PDATE | -1.133 |
� | -1.098 |
Token | down |
Token ID | 866 |
Feature activation | -1.33e+00 |
Pos logit contributions | |
erella | +0.363 |
inburgh | +0.312 |
achus | +0.310 |
lime | +0.287 |
cised | +0.273 |
Neg logit contributions | |
advertisement | -0.282 |
IMAGES | -0.259 |
� | -0.248 |
PDATE | -0.241 |
� | -0.240 |
Token | in |
Token ID | 287 |
Feature activation | -1.07e-01 |
Pos logit contributions | |
erella | +3.017 |
achus | +2.584 |
inburgh | +2.574 |
lime | +2.334 |
cised | +2.259 |
Neg logit contributions | |
advertisement | -2.833 |
IMAGES | -2.648 |
� | -2.565 |
PDATE | -2.546 |
� | -2.471 |
Token | 2010 |
Token ID | 3050 |
Feature activation | -5.56e-02 |
Pos logit contributions | |
erella | +0.235 |
inburgh | +0.198 |
achus | +0.197 |
lime | +0.180 |
cised | +0.171 |
Neg logit contributions | |
advertisement | -0.241 |
IMAGES | -0.224 |
� | -0.217 |
PDATE | -0.212 |
� | -0.211 |
Token | after |
Token ID | 706 |
Feature activation | -1.92e-01 |
Pos logit contributions | |
erella | +0.128 |
inburgh | +0.108 |
achus | +0.108 |
lime | +0.099 |
cised | +0.093 |
Neg logit contributions | |
advertisement | -0.121 |
IMAGES | -0.113 |
� | -0.108 |
PDATE | -0.105 |
� | -0.105 |
Token | an |
Token ID | 281 |
Feature activation | +3.16e-02 |
Pos logit contributions | |
erella | +0.429 |
inburgh | +0.363 |
achus | +0.361 |
lime | +0.331 |
cised | +0.313 |
Neg logit contributions | |
advertisement | -0.426 |
IMAGES | -0.396 |
� | -0.385 |
PDATE | -0.375 |
� | -0.369 |
Token | Israeli |
Token ID | 6085 |
Feature activation | +1.61e-01 |
Pos logit contributions | |
advertisement | +0.065 |
IMAGES | +0.060 |
� | +0.058 |
PDATE | +0.056 |
� | +0.056 |
Neg logit contributions | |
erella | -0.076 |
inburgh | -0.065 |
achus | -0.065 |
lime | -0.060 |
cised | -0.057 |
Token | claiming |
Token ID | 8512 |
Feature activation | +1.91e-01 |
Pos logit contributions | |
advertisement | +0.698 |
IMAGES | +0.639 |
� | +0.607 |
� | +0.582 |
PDATE | +0.582 |
Neg logit contributions | |
erella | -1.103 |
inburgh | -0.963 |
achus | -0.960 |
lime | -0.895 |
acted | -0.850 |
Token | that |
Token ID | 326 |
Feature activation | +2.36e-01 |
Pos logit contributions | |
advertisement | +0.424 |
IMAGES | +0.394 |
� | +0.380 |
PDATE | +0.369 |
� | +0.368 |
Neg logit contributions | |
erella | -0.432 |
inburgh | -0.365 |
achus | -0.361 |
lime | -0.332 |
acted | -0.312 |
Token | he |
Token ID | 339 |
Feature activation | +3.20e-01 |
Pos logit contributions | |
advertisement | +0.516 |
IMAGES | +0.478 |
� | +0.461 |
PDATE | +0.448 |
� | +0.447 |
Neg logit contributions | |
erella | -0.541 |
inburgh | -0.460 |
achus | -0.456 |
lime | -0.420 |
cised | -0.394 |
Token | was |
Token ID | 373 |
Feature activation | +1.71e-01 |
Pos logit contributions | |
advertisement | +0.738 |
IMAGES | +0.688 |
� | +0.663 |
� | +0.645 |
PDATE | +0.640 |
Neg logit contributions | |
erella | -0.693 |
inburgh | -0.582 |
achus | -0.576 |
lime | -0.526 |
cised | -0.490 |
Token | t |
Token ID | 256 |
Feature activation | -4.80e-02 |
Pos logit contributions | |
advertisement | +0.383 |
IMAGES | +0.357 |
� | +0.343 |
PDATE | +0.335 |
� | +0.331 |
Neg logit contributions | |
erella | -0.381 |
inburgh | -0.322 |
achus | -0.318 |
lime | -0.293 |
cised | -0.274 |
Token | aunted |
Token ID | 20227 |
Feature activation | -1.33e+00 |
Pos logit contributions | |
erella | +0.091 |
inburgh | +0.074 |
achus | +0.074 |
lime | +0.065 |
acted | +0.062 |
Neg logit contributions | |
advertisement | -0.125 |
IMAGES | -0.118 |
� | -0.114 |
PDATE | -0.112 |
� | -0.111 |
Token | and |
Token ID | 290 |
Feature activation | -1.42e-01 |
Pos logit contributions | |
erella | +2.511 |
inburgh | +2.111 |
achus | +2.056 |
lime | +1.865 |
acted | +1.731 |
Neg logit contributions | |
advertisement | -3.305 |
IMAGES | -3.111 |
� | -3.021 |
PDATE | -3.008 |
� | -2.944 |
Token | falsely |
Token ID | 24566 |
Feature activation | -2.74e-01 |
Pos logit contributions | |
erella | +0.309 |
inburgh | +0.258 |
achus | +0.256 |
lime | +0.234 |
acted | +0.220 |
Neg logit contributions | |
advertisement | -0.328 |
IMAGES | -0.305 |
� | -0.294 |
PDATE | -0.291 |
� | -0.285 |
Token | arrested |
Token ID | 5169 |
Feature activation | -1.27e-01 |
Pos logit contributions | |
erella | +0.644 |
inburgh | +0.548 |
achus | +0.543 |
lime | +0.501 |
acted | +0.474 |
Neg logit contributions | |
advertisement | -0.580 |
IMAGES | -0.539 |
� | -0.516 |
PDATE | -0.509 |
� | -0.499 |
Token | by |
Token ID | 416 |
Feature activation | +5.57e-02 |
Pos logit contributions | |
erella | +0.259 |
inburgh | +0.216 |
achus | +0.214 |
lime | +0.194 |
cised | +0.180 |
Neg logit contributions | |
advertisement | -0.307 |
IMAGES | -0.287 |
� | -0.278 |
PDATE | -0.271 |
� | -0.271 |
Token | police |
Token ID | 1644 |
Feature activation | -3.61e-01 |
Pos logit contributions | |
advertisement | +0.122 |
IMAGES | +0.113 |
� | +0.109 |
PDATE | +0.107 |
� | +0.106 |
Neg logit contributions | |
erella | -0.126 |
inburgh | -0.107 |
achus | -0.106 |
lime | -0.098 |
acted | -0.092 |
Token | park |
Token ID | 3952 |
Feature activation | -4.15e-01 |
Pos logit contributions | |
erella | +0.120 |
inburgh | +0.100 |
achus | +0.099 |
lime | +0.090 |
acted | +0.084 |
Neg logit contributions | |
advertisement | -0.134 |
IMAGES | -0.125 |
� | -0.121 |
� | -0.118 |
PDATE | -0.117 |
Token | anymore |
Token ID | 7471 |
Feature activation | +3.23e-01 |
Pos logit contributions | |
erella | +0.925 |
inburgh | +0.781 |
achus | +0.768 |
lime | +0.706 |
illary | +0.663 |
Neg logit contributions | |
advertisement | -0.930 |
IMAGES | -0.873 |
� | -0.839 |
PDATE | -0.822 |
� | -0.812 |
Token | ." |
Token ID | 526 |
Feature activation | -4.70e-02 |
Pos logit contributions | |
advertisement | +0.699 |
IMAGES | +0.649 |
� | +0.623 |
� | +0.605 |
PDATE | +0.604 |
Neg logit contributions | |
erella | -0.749 |
inburgh | -0.637 |
achus | -0.630 |
lime | -0.581 |
cised | -0.543 |
Token |
|
Token ID | 198 |
Feature activation | +5.06e-02 |
Pos logit contributions | |
erella | +0.104 |
inburgh | +0.087 |
achus | +0.086 |
lime | +0.079 |
acted | +0.074 |
Neg logit contributions | |
advertisement | -0.106 |
IMAGES | -0.099 |
� | -0.095 |
PDATE | -0.095 |
� | -0.093 |
Token |
|
Token ID | 198 |
Feature activation | +5.66e-02 |
Pos logit contributions | |
advertisement | +0.112 |
IMAGES | +0.104 |
� | +0.100 |
� | +0.097 |
PDATE | +0.097 |
Neg logit contributions | |
erella | -0.115 |
inburgh | -0.097 |
achus | -0.096 |
lime | -0.088 |
cised | -0.083 |
Token | CLOSE |
Token ID | 32737 |
Feature activation | -1.29e+00 |
Pos logit contributions | |
advertisement | +0.109 |
IMAGES | +0.102 |
� | +0.097 |
PDATE | +0.095 |
� | +0.094 |
Neg logit contributions | |
erella | -0.142 |
inburgh | -0.122 |
achus | -0.121 |
lime | -0.113 |
acted | -0.107 |
Token | New |
Token ID | 968 |
Feature activation | -4.88e-03 |
Pos logit contributions | |
erella | +2.956 |
inburgh | +2.480 |
achus | +2.471 |
lime | +2.291 |
ウ | +2.171 |
Neg logit contributions | |
advertisement | -2.666 |
PDATE | -2.430 |
� | -2.367 |
� | -2.336 |
IMAGES | -2.328 |
Token | Castle |
Token ID | 11312 |
Feature activation | -6.15e-02 |
Pos logit contributions | |
erella | +0.007 |
inburgh | +0.005 |
achus | +0.005 |
lime | +0.005 |
cised | +0.004 |
Neg logit contributions | |
advertisement | -0.015 |
IMAGES | -0.014 |
� | -0.014 |
� | -0.013 |
PDATE | -0.013 |
Token | police |
Token ID | 1644 |
Feature activation | -2.43e-01 |
Pos logit contributions | |
erella | +0.143 |
inburgh | +0.122 |
achus | +0.122 |
lime | +0.111 |
illary | +0.104 |
Neg logit contributions | |
advertisement | -0.132 |
IMAGES | -0.121 |
� | -0.117 |
PDATE | -0.114 |
� | -0.113 |
Token | enforce |
Token ID | 4605 |
Feature activation | -6.04e-02 |
Pos logit contributions | |
erella | +0.534 |
inburgh | +0.451 |
achus | +0.449 |
lime | +0.411 |
acted | +0.384 |
Neg logit contributions | |
advertisement | -0.551 |
IMAGES | -0.508 |
� | -0.497 |
PDATE | -0.485 |
� | -0.476 |
Token | dawn |
Token ID | 17577 |
Feature activation | -7.22e-02 |
Pos logit contributions | |
erella | +0.136 |
inburgh | +0.115 |
achus | +0.114 |
lime | +0.105 |
cised | +0.099 |
Neg logit contributions | |
advertisement | -0.134 |
IMAGES | -0.124 |
� | -0.120 |
PDATE | -0.116 |
� | -0.116 |
Token | minutes |
Token ID | 2431 |
Feature activation | -4.58e-01 |
Pos logit contributions | |
advertisement | +0.130 |
IMAGES | +0.123 |
� | +0.119 |
PDATE | +0.117 |
� | +0.117 |
Neg logit contributions | |
erella | -0.066 |
inburgh | -0.050 |
achus | -0.050 |
lime | -0.043 |
cised | -0.038 |
Token | later |
Token ID | 1568 |
Feature activation | -4.99e-01 |
Pos logit contributions | |
erella | +1.008 |
inburgh | +0.848 |
achus | +0.846 |
lime | +0.778 |
illary | +0.732 |
Neg logit contributions | |
advertisement | -1.037 |
IMAGES | -0.956 |
� | -0.922 |
PDATE | -0.905 |
tti | -0.885 |
Token | . |
Token ID | 13 |
Feature activation | +3.37e-02 |
Pos logit contributions | |
erella | +1.085 |
inburgh | +0.916 |
achus | +0.909 |
lime | +0.830 |
acted | +0.772 |
Neg logit contributions | |
advertisement | -1.134 |
IMAGES | -1.065 |
� | -1.014 |
PDATE | -1.000 |
� | -0.982 |
Token | Players |
Token ID | 13094 |
Feature activation | +8.78e-02 |
Pos logit contributions | |
advertisement | +0.073 |
IMAGES | +0.067 |
� | +0.065 |
PDATE | +0.064 |
� | +0.063 |
Neg logit contributions | |
erella | -0.078 |
inburgh | -0.066 |
achus | -0.065 |
lime | -0.060 |
cised | -0.056 |
Token | s |
Token ID | 264 |
Feature activation | -1.27e-01 |
Pos logit contributions | |
advertisement | +0.198 |
IMAGES | +0.185 |
� | +0.178 |
PDATE | +0.173 |
� | +0.172 |
Neg logit contributions | |
erella | -0.195 |
inburgh | -0.164 |
achus | -0.163 |
lime | -0.149 |
cised | -0.139 |
Token | ifted |
Token ID | 21715 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
erella | +0.269 |
inburgh | +0.225 |
achus | +0.223 |
lime | +0.203 |
acted | +0.190 |
Neg logit contributions | |
advertisement | -0.301 |
IMAGES | -0.282 |
� | -0.270 |
PDATE | -0.264 |
� | -0.263 |
Token | over |
Token ID | 625 |
Feature activation | -2.19e-01 |
Pos logit contributions | |
erella | +3.134 |
achus | +2.702 |
inburgh | +2.699 |
lime | +2.556 |
folk | +2.352 |
Neg logit contributions | |
advertisement | -2.430 |
IMAGES | -2.190 |
PDATE | -2.135 |
� | -2.134 |
� | -2.060 |
Token | what |
Token ID | 644 |
Feature activation | -1.07e-02 |
Pos logit contributions | |
erella | +0.465 |
inburgh | +0.390 |
achus | +0.388 |
lime | +0.359 |
acted | +0.332 |
Neg logit contributions | |
advertisement | -0.509 |
IMAGES | -0.473 |
� | -0.459 |
PDATE | -0.451 |
� | -0.446 |
Token | � |
Token ID | 447 |
Feature activation | -8.32e-02 |
Pos logit contributions | |
erella | +0.024 |
inburgh | +0.020 |
achus | +0.020 |
lime | +0.018 |
acted | +0.017 |
Neg logit contributions | |
advertisement | -0.024 |
IMAGES | -0.023 |
� | -0.022 |
PDATE | -0.021 |
� | -0.021 |
Token | � |
Token ID | 247 |
Feature activation | -2.49e-01 |
Pos logit contributions | |
erella | +0.230 |
inburgh | +0.201 |
achus | +0.200 |
lime | +0.188 |
acted | +0.179 |
Neg logit contributions | |
advertisement | -0.140 |
IMAGES | -0.129 |
� | -0.122 |
PDATE | -0.117 |
tti | -0.114 |
Token | s |
Token ID | 82 |
Feature activation | -1.57e-01 |
Pos logit contributions | |
erella | +0.604 |
inburgh | +0.519 |
achus | +0.514 |
lime | +0.482 |
acted | +0.452 |
Neg logit contributions | |
advertisement | -0.503 |
IMAGES | -0.467 |
� | -0.447 |
PDATE | -0.438 |
� | -0.430 |
Token | moving |
Token ID | 3867 |
Feature activation | -1.41e-01 |
Pos logit contributions | |
advertisement | +0.371 |
IMAGES | +0.344 |
� | +0.333 |
� | +0.323 |
PDATE | +0.322 |
Neg logit contributions | |
erella | -0.361 |
inburgh | -0.303 |
achus | -0.302 |
lime | -0.276 |
cised | -0.257 |
Token | from |
Token ID | 422 |
Feature activation | +5.14e-02 |
Pos logit contributions | |
erella | +0.310 |
achus | +0.262 |
inburgh | +0.260 |
lime | +0.237 |
acted | +0.221 |
Neg logit contributions | |
advertisement | -0.323 |
IMAGES | -0.294 |
� | -0.287 |
PDATE | -0.280 |
� | -0.280 |
Token | Android |
Token ID | 5565 |
Feature activation | -5.96e-01 |
Pos logit contributions | |
advertisement | +0.111 |
IMAGES | +0.103 |
� | +0.099 |
PDATE | +0.096 |
� | +0.096 |
Neg logit contributions | |
erella | -0.118 |
inburgh | -0.100 |
achus | -0.100 |
lime | -0.092 |
acted | -0.086 |
Token | to |
Token ID | 284 |
Feature activation | +1.77e-01 |
Pos logit contributions | |
erella | +1.280 |
achus | +1.075 |
inburgh | +1.068 |
lime | +0.994 |
acted | +0.924 |
Neg logit contributions | |
advertisement | -1.357 |
IMAGES | -1.249 |
� | -1.220 |
PDATE | -1.179 |
� | -1.172 |
Token | T |
Token ID | 309 |
Feature activation | -3.07e-01 |
Pos logit contributions | |
advertisement | +0.377 |
IMAGES | +0.348 |
� | +0.335 |
� | +0.324 |
PDATE | +0.324 |
Neg logit contributions | |
erella | -0.416 |
inburgh | -0.354 |
achus | -0.352 |
lime | -0.326 |
cised | -0.305 |
Token | izen |
Token ID | 33977 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
erella | +0.674 |
inburgh | +0.565 |
achus | +0.565 |
lime | +0.520 |
acted | +0.481 |
Neg logit contributions | |
advertisement | -0.700 |
IMAGES | -0.651 |
� | -0.627 |
� | -0.611 |
PDATE | -0.610 |
Token | , |
Token ID | 11 |
Feature activation | -1.10e-01 |
Pos logit contributions | |
erella | +2.657 |
achus | +2.229 |
inburgh | +2.209 |
lime | +2.033 |
illary | +1.891 |
Neg logit contributions | |
advertisement | -2.908 |
IMAGES | -2.670 |
� | -2.602 |
PDATE | -2.530 |
� | -2.508 |
Token | or |
Token ID | 393 |
Feature activation | -9.92e-03 |
Pos logit contributions | |
erella | +0.245 |
inburgh | +0.206 |
achus | +0.206 |
lime | +0.189 |
cised | +0.176 |
Neg logit contributions | |
advertisement | -0.248 |
IMAGES | -0.229 |
� | -0.221 |
PDATE | -0.215 |
� | -0.215 |
Token | from |
Token ID | 422 |
Feature activation | +3.19e-01 |
Pos logit contributions | |
erella | +0.023 |
inburgh | +0.019 |
achus | +0.019 |
lime | +0.018 |
cised | +0.016 |
Neg logit contributions | |
advertisement | -0.022 |
IMAGES | -0.020 |
� | -0.019 |
PDATE | -0.019 |
� | -0.019 |
Token | T |
Token ID | 309 |
Feature activation | +6.06e-03 |
Pos logit contributions | |
advertisement | +0.643 |
IMAGES | +0.592 |
� | +0.566 |
PDATE | +0.550 |
� | +0.549 |
Neg logit contributions | |
erella | -0.787 |
inburgh | -0.674 |
achus | -0.672 |
lime | -0.625 |
acted | -0.584 |
Token | izen |
Token ID | 33977 |
Feature activation | -7.84e-01 |
Pos logit contributions | |
advertisement | +0.014 |
IMAGES | +0.013 |
� | +0.012 |
� | +0.012 |
PDATE | +0.012 |
Neg logit contributions | |
erella | -0.014 |
inburgh | -0.011 |
achus | -0.011 |
lime | -0.010 |
illary | -0.010 |
Token | use |
Token ID | 779 |
Feature activation | -3.80e-01 |
Pos logit contributions | |
erella | +0.181 |
inburgh | +0.152 |
achus | +0.152 |
lime | +0.140 |
acted | +0.132 |
Neg logit contributions | |
advertisement | -0.164 |
IMAGES | -0.153 |
� | -0.146 |
PDATE | -0.144 |
� | -0.141 |
Token | it |
Token ID | 340 |
Feature activation | -1.05e-01 |
Pos logit contributions | |
erella | +0.798 |
inburgh | +0.654 |
achus | +0.647 |
lime | +0.601 |
ウ | +0.558 |
Neg logit contributions | |
advertisement | -0.892 |
IMAGES | -0.848 |
� | -0.800 |
PDATE | -0.790 |
� | -0.787 |
Token | to |
Token ID | 284 |
Feature activation | -9.06e-02 |
Pos logit contributions | |
erella | +0.243 |
inburgh | +0.206 |
achus | +0.206 |
lime | +0.188 |
acted | +0.177 |
Neg logit contributions | |
advertisement | -0.228 |
IMAGES | -0.213 |
� | -0.202 |
� | -0.198 |
PDATE | -0.197 |
Token | protect |
Token ID | 1805 |
Feature activation | -7.44e-01 |
Pos logit contributions | |
erella | +0.221 |
achus | +0.189 |
inburgh | +0.189 |
lime | +0.177 |
illary | +0.166 |
Neg logit contributions | |
advertisement | -0.183 |
IMAGES | -0.168 |
� | -0.159 |
PDATE | -0.158 |
� | -0.156 |
Token | your |
Token ID | 534 |
Feature activation | -2.50e-01 |
Pos logit contributions | |
erella | +1.615 |
inburgh | +1.346 |
achus | +1.338 |
lime | +1.230 |
ウ | +1.160 |
Neg logit contributions | |
advertisement | -1.696 |
IMAGES | -1.593 |
� | -1.502 |
PDATE | -1.486 |
� | -1.478 |
Token | Life |
Token ID | 5155 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
erella | +0.574 |
inburgh | +0.485 |
achus | +0.482 |
lime | +0.445 |
ウ | +0.420 |
Neg logit contributions | |
advertisement | -0.542 |
IMAGES | -0.505 |
� | -0.477 |
PDATE | -0.471 |
� | -0.465 |
Token | Points |
Token ID | 11045 |
Feature activation | -1.33e-01 |
Pos logit contributions | |
erella | +2.793 |
achus | +2.356 |
inburgh | +2.337 |
lime | +2.203 |
folk | +2.057 |
Neg logit contributions | |
advertisement | -2.831 |
IMAGES | -2.550 |
� | -2.412 |
PDATE | -2.373 |
� | -2.325 |
Token | ! |
Token ID | 0 |
Feature activation | +9.89e-02 |
Pos logit contributions | |
erella | +0.298 |
inburgh | +0.253 |
achus | +0.251 |
lime | +0.231 |
illary | +0.215 |
Neg logit contributions | |
advertisement | -0.297 |
IMAGES | -0.276 |
� | -0.263 |
� | -0.257 |
PDATE | -0.256 |
Token | It |
Token ID | 632 |
Feature activation | -4.92e-01 |
Pos logit contributions | |
advertisement | +0.210 |
IMAGES | +0.195 |
� | +0.187 |
PDATE | +0.184 |
� | +0.182 |
Neg logit contributions | |
erella | -0.231 |
inburgh | -0.196 |
achus | -0.195 |
lime | -0.180 |
illary | -0.169 |
Token | � |
Token ID | 447 |
Feature activation | -4.13e-01 |
Pos logit contributions | |
erella | +1.117 |
inburgh | +0.946 |
achus | +0.932 |
lime | +0.861 |
illary | +0.808 |
Neg logit contributions | |
advertisement | -1.082 |
IMAGES | -1.009 |
� | -0.963 |
PDATE | -0.959 |
� | -0.941 |
Token | � |
Token ID | 247 |
Feature activation | -1.77e-01 |
Pos logit contributions | |
erella | +0.923 |
inburgh | +0.776 |
achus | +0.769 |
lime | +0.712 |
acted | +0.674 |
Neg logit contributions | |
advertisement | -0.899 |
IMAGES | -0.858 |
� | -0.818 |
PDATE | -0.816 |
OFF | -0.774 |
Token | m |
Token ID | 76 |
Feature activation | -3.73e-01 |
Pos logit contributions | |
erella | +1.266 |
inburgh | +1.073 |
achus | +1.059 |
lime | +0.998 |
acted | +0.922 |
Neg logit contributions | |
advertisement | -1.230 |
IMAGES | -1.162 |
� | -1.100 |
PDATE | -1.094 |
� | -1.067 |
Token | ew |
Token ID | 413 |
Feature activation | -2.72e-02 |
Pos logit contributions | |
erella | +0.855 |
inburgh | +0.724 |
achus | +0.722 |
lime | +0.661 |
acted | +0.619 |
Neg logit contributions | |
advertisement | -0.818 |
IMAGES | -0.764 |
� | -0.730 |
PDATE | -0.716 |
� | -0.711 |
Token | 2 |
Token ID | 17 |
Feature activation | -2.52e-01 |
Pos logit contributions | |
erella | +0.064 |
inburgh | +0.054 |
achus | +0.054 |
lime | +0.050 |
cised | +0.046 |
Neg logit contributions | |
advertisement | -0.058 |
IMAGES | -0.054 |
� | -0.052 |
PDATE | -0.050 |
� | -0.050 |
Token | king |
Token ID | 3364 |
Feature activation | -4.25e-01 |
Pos logit contributions | |
erella | +0.571 |
inburgh | +0.484 |
achus | +0.479 |
lime | +0.442 |
acted | +0.413 |
Neg logit contributions | |
advertisement | -0.555 |
IMAGES | -0.521 |
� | -0.498 |
PDATE | -0.487 |
� | -0.484 |
Token | > |
Token ID | 29 |
Feature activation | -5.73e-01 |
Pos logit contributions | |
erella | +0.771 |
inburgh | +0.626 |
achus | +0.612 |
lime | +0.556 |
acted | +0.506 |
Neg logit contributions | |
advertisement | -1.125 |
IMAGES | -1.071 |
� | -1.030 |
PDATE | -1.015 |
� | -1.004 |
Token | OM |
Token ID | 32468 |
Feature activation | -1.24e+00 |
Pos logit contributions | |
erella | +1.231 |
inburgh | +1.026 |
achus | +1.019 |
lime | +0.943 |
illary | +0.867 |
Neg logit contributions | |
advertisement | -1.332 |
IMAGES | -1.227 |
� | -1.189 |
PDATE | -1.165 |
� | -1.151 |
Token | FG |
Token ID | 30386 |
Feature activation | -5.73e-01 |
Pos logit contributions | |
erella | +3.029 |
inburgh | +2.629 |
achus | +2.629 |
lime | +2.407 |
illary | +2.285 |
Neg logit contributions | |
advertisement | -2.440 |
IMAGES | -2.223 |
� | -2.120 |
PDATE | -2.063 |
� | -2.058 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -1.74e-01 |
Pos logit contributions | |
erella | +1.274 |
inburgh | +1.076 |
achus | +1.068 |
lime | +0.990 |
cised | +0.918 |
Neg logit contributions | |
advertisement | -1.278 |
IMAGES | -1.175 |
� | -1.133 |
� | -1.105 |
PDATE | -1.102 |
Token | By |
Token ID | 3886 |
Feature activation | +7.56e-03 |
Pos logit contributions | |
erella | +0.443 |
inburgh | +0.382 |
achus | +0.378 |
lime | +0.357 |
acted | +0.336 |
Neg logit contributions | |
advertisement | -0.325 |
IMAGES | -0.308 |
� | -0.290 |
PDATE | -0.289 |
� | -0.281 |
Token | Andrew |
Token ID | 6858 |
Feature activation | +2.86e-01 |
Pos logit contributions | |
advertisement | +0.015 |
IMAGES | +0.014 |
� | +0.013 |
PDATE | +0.013 |
� | +0.013 |
Neg logit contributions | |
erella | -0.019 |
inburgh | -0.016 |
achus | -0.016 |
lime | -0.015 |
acted | -0.014 |
Token | Benson |
Token ID | 33940 |
Feature activation | +2.01e-01 |
Pos logit contributions | |
advertisement | +0.565 |
IMAGES | +0.521 |
� | +0.497 |
PDATE | +0.486 |
� | +0.482 |
Neg logit contributions | |
erella | -0.718 |
inburgh | -0.617 |
achus | -0.611 |
lime | -0.566 |
cised | -0.533 |
Token | eventually |
Token ID | 4191 |
Feature activation | -1.93e-01 |
Pos logit contributions | |
erella | +0.106 |
inburgh | +0.089 |
achus | +0.089 |
lime | +0.081 |
cised | +0.076 |
Neg logit contributions | |
advertisement | -0.100 |
IMAGES | -0.093 |
� | -0.089 |
PDATE | -0.089 |
� | -0.086 |
Token | hook |
Token ID | 8011 |
Feature activation | -1.19e-01 |
Pos logit contributions | |
erella | +0.442 |
achus | +0.371 |
inburgh | +0.370 |
lime | +0.341 |
izabeth | +0.316 |
Neg logit contributions | |
advertisement | -0.425 |
IMAGES | -0.398 |
PDATE | -0.381 |
� | -0.379 |
� | -0.368 |
Token | up |
Token ID | 510 |
Feature activation | -3.78e-01 |
Pos logit contributions | |
erella | +0.258 |
inburgh | +0.216 |
achus | +0.215 |
lime | +0.196 |
acted | +0.181 |
Neg logit contributions | |
advertisement | -0.276 |
IMAGES | -0.258 |
� | -0.247 |
PDATE | -0.242 |
� | -0.239 |
Token | with |
Token ID | 351 |
Feature activation | -1.56e-02 |
Pos logit contributions | |
erella | +0.815 |
inburgh | +0.684 |
achus | +0.678 |
lime | +0.619 |
acted | +0.576 |
Neg logit contributions | |
advertisement | -0.877 |
IMAGES | -0.821 |
� | -0.790 |
PDATE | -0.770 |
� | -0.766 |
Token | her |
Token ID | 607 |
Feature activation | -2.40e-01 |
Pos logit contributions | |
erella | +0.035 |
achus | +0.030 |
inburgh | +0.030 |
lime | +0.028 |
acted | +0.026 |
Neg logit contributions | |
advertisement | -0.034 |
IMAGES | -0.032 |
� | -0.031 |
PDATE | -0.030 |
� | -0.030 |
Token | adul |
Token ID | 19046 |
Feature activation | -1.23e+00 |
Pos logit contributions | |
erella | +0.548 |
inburgh | +0.465 |
achus | +0.458 |
lime | +0.426 |
acted | +0.396 |
Neg logit contributions | |
advertisement | -0.530 |
IMAGES | -0.494 |
� | -0.471 |
PDATE | -0.459 |
� | -0.458 |
Token | ter |
Token ID | 353 |
Feature activation | -5.85e-01 |
Pos logit contributions | |
erella | +2.555 |
achus | +2.149 |
inburgh | +2.118 |
lime | +1.938 |
cised | +1.837 |
Neg logit contributions | |
advertisement | -2.971 |
IMAGES | -2.773 |
PDATE | -2.652 |
� | -2.631 |
� | -2.565 |
Token | ous |
Token ID | 516 |
Feature activation | -1.40e-01 |
Pos logit contributions | |
erella | +1.472 |
inburgh | +1.268 |
achus | +1.264 |
lime | +1.165 |
cised | +1.120 |
Neg logit contributions | |
advertisement | -1.149 |
IMAGES | -1.060 |
� | -1.007 |
PDATE | -0.990 |
� | -0.970 |
Token | l |
Token ID | 75 |
Feature activation | -4.37e-01 |
Pos logit contributions | |
erella | +0.299 |
inburgh | +0.251 |
achus | +0.248 |
lime | +0.228 |
cised | +0.210 |
Neg logit contributions | |
advertisement | -0.331 |
IMAGES | -0.308 |
� | -0.297 |
� | -0.289 |
PDATE | -0.289 |
Token | over |
Token ID | 2502 |
Feature activation | -1.40e-01 |
Pos logit contributions | |
erella | +0.733 |
achus | +0.585 |
inburgh | +0.585 |
lime | +0.508 |
illary | +0.461 |
Neg logit contributions | |
advertisement | -1.225 |
IMAGES | -1.159 |
� | -1.124 |
PDATE | -1.103 |
� | -1.100 |
Token | ( |
Token ID | 357 |
Feature activation | -1.98e-01 |
Pos logit contributions | |
erella | +0.315 |
inburgh | +0.266 |
achus | +0.264 |
lime | +0.243 |
acted | +0.226 |
Neg logit contributions | |
advertisement | -0.315 |
IMAGES | -0.293 |
� | -0.282 |
PDATE | -0.275 |
� | -0.273 |
Token | conflict |
Token ID | 5358 |
Feature activation | -2.92e-01 |
Pos logit contributions | |
erella | +0.180 |
inburgh | +0.153 |
achus | +0.152 |
lime | +0.139 |
acted | +0.131 |
Neg logit contributions | |
advertisement | -0.169 |
IMAGES | -0.157 |
� | -0.150 |
PDATE | -0.146 |
� | -0.146 |
Token | scored |
Token ID | 7781 |
Feature activation | +9.20e-02 |
Pos logit contributions | |
erella | +0.641 |
inburgh | +0.540 |
achus | +0.537 |
lime | +0.492 |
illary | +0.459 |
Neg logit contributions | |
advertisement | -0.661 |
IMAGES | -0.623 |
� | -0.593 |
PDATE | -0.581 |
� | -0.577 |
Token | into |
Token ID | 656 |
Feature activation | -1.06e-01 |
Pos logit contributions | |
advertisement | +0.213 |
IMAGES | +0.199 |
� | +0.192 |
� | +0.186 |
PDATE | +0.186 |
Neg logit contributions | |
erella | -0.198 |
inburgh | -0.167 |
achus | -0.165 |
lime | -0.151 |
cised | -0.141 |
Token | the |
Token ID | 262 |
Feature activation | -3.98e-02 |
Pos logit contributions | |
erella | +0.235 |
inburgh | +0.199 |
achus | +0.198 |
lime | +0.181 |
cised | +0.169 |
Neg logit contributions | |
advertisement | -0.238 |
IMAGES | -0.222 |
� | -0.213 |
� | -0.207 |
PDATE | -0.207 |
Token | very |
Token ID | 845 |
Feature activation | +1.57e-01 |
Pos logit contributions | |
erella | +0.092 |
inburgh | +0.078 |
achus | +0.078 |
lime | +0.071 |
cised | +0.067 |
Neg logit contributions | |
advertisement | -0.086 |
IMAGES | -0.080 |
� | -0.077 |
PDATE | -0.075 |
� | -0.074 |
Token | continental |
Token ID | 29843 |
Feature activation | -1.22e+00 |
Pos logit contributions | |
advertisement | +0.337 |
IMAGES | +0.312 |
� | +0.300 |
� | +0.291 |
PDATE | +0.291 |
Neg logit contributions | |
erella | -0.367 |
inburgh | -0.313 |
achus | -0.310 |
lime | -0.286 |
cised | -0.267 |
Token | land |
Token ID | 1956 |
Feature activation | -6.04e-01 |
Pos logit contributions | |
erella | +2.556 |
inburgh | +2.163 |
achus | +2.157 |
lime | +1.973 |
kingdom | +1.864 |
Neg logit contributions | |
advertisement | -2.795 |
IMAGES | -2.613 |
PDATE | -2.551 |
� | -2.525 |
� | -2.451 |
Token | mass |
Token ID | 22208 |
Feature activation | -5.11e-01 |
Pos logit contributions | |
erella | +1.293 |
inburgh | +1.099 |
achus | +1.081 |
lime | +0.992 |
illary | +0.927 |
Neg logit contributions | |
advertisement | -1.405 |
IMAGES | -1.305 |
� | -1.258 |
PDATE | -1.239 |
� | -1.214 |
Token | itself |
Token ID | 2346 |
Feature activation | +2.68e-01 |
Pos logit contributions | |
erella | +1.107 |
inburgh | +0.944 |
achus | +0.940 |
lime | +0.846 |
acted | +0.790 |
Neg logit contributions | |
advertisement | -1.164 |
IMAGES | -1.088 |
� | -1.040 |
PDATE | -1.036 |
� | -1.021 |
Token | . |
Token ID | 13 |
Feature activation | +3.28e-02 |
Pos logit contributions | |
advertisement | +0.588 |
IMAGES | +0.547 |
� | +0.525 |
� | +0.510 |
PDATE | +0.509 |
Neg logit contributions | |
erella | -0.610 |
inburgh | -0.518 |
achus | -0.512 |
lime | -0.471 |
cised | -0.441 |
Token | But |
Token ID | 887 |
Feature activation | +1.10e-01 |
Pos logit contributions | |
advertisement | +0.067 |
IMAGES | +0.061 |
� | +0.059 |
PDATE | +0.058 |
� | +0.057 |
Neg logit contributions | |
erella | -0.080 |
inburgh | -0.068 |
achus | -0.068 |
lime | -0.063 |
illary | -0.059 |
Token |
|
Token ID | 198 |
Feature activation | -1.91e-01 |
Pos logit contributions | |
advertisement | +0.091 |
IMAGES | +0.085 |
� | +0.082 |
PDATE | +0.079 |
� | +0.079 |
Neg logit contributions | |
erella | -0.095 |
inburgh | -0.081 |
achus | -0.080 |
lime | -0.074 |
illary | -0.069 |
Token |
|
Token ID | 198 |
Feature activation | -2.12e-01 |
Pos logit contributions | |
erella | +0.421 |
inburgh | +0.352 |
achus | +0.349 |
lime | +0.322 |
acted | +0.301 |
Neg logit contributions | |
advertisement | -0.432 |
IMAGES | -0.405 |
PDATE | -0.390 |
� | -0.389 |
� | -0.381 |
Token | Col |
Token ID | 5216 |
Feature activation | +2.60e-02 |
Pos logit contributions | |
erella | +0.545 |
inburgh | +0.470 |
achus | +0.465 |
lime | +0.440 |
acted | +0.415 |
Neg logit contributions | |
advertisement | -0.391 |
IMAGES | -0.370 |
PDATE | -0.347 |
� | -0.347 |
� | -0.338 |
Token | on |
Token ID | 261 |
Feature activation | +3.10e-01 |
Pos logit contributions | |
advertisement | +0.063 |
IMAGES | +0.059 |
� | +0.057 |
PDATE | +0.057 |
� | +0.055 |
Neg logit contributions | |
erella | -0.054 |
inburgh | -0.045 |
achus | -0.044 |
lime | -0.040 |
illary | -0.037 |
Token | ize |
Token ID | 1096 |
Feature activation | -1.48e-01 |
Pos logit contributions | |
advertisement | +0.609 |
IMAGES | +0.561 |
� | +0.536 |
� | +0.517 |
PDATE | +0.517 |
Neg logit contributions | |
erella | -0.780 |
inburgh | -0.671 |
achus | -0.667 |
lime | -0.618 |
cised | -0.584 |
Token | unch |
Token ID | 16258 |
Feature activation | -1.20e+00 |
Pos logit contributions | |
erella | +0.319 |
inburgh | +0.268 |
achus | +0.265 |
lime | +0.243 |
acted | +0.227 |
Neg logit contributions | |
advertisement | -0.341 |
IMAGES | -0.316 |
� | -0.307 |
PDATE | -0.299 |
� | -0.298 |
Token | arted |
Token ID | 19112 |
Feature activation | -5.44e-01 |
Pos logit contributions | |
erella | +1.952 |
inburgh | +1.532 |
achus | +1.505 |
lime | +1.302 |
acted | +1.232 |
Neg logit contributions | |
advertisement | -3.422 |
IMAGES | -3.270 |
PDATE | -3.264 |
� | -3.205 |
� | -3.132 |
Token | planets |
Token ID | 14705 |
Feature activation | +1.32e-01 |
Pos logit contributions | |
erella | +1.074 |
inburgh | +0.888 |
achus | +0.883 |
lime | +0.800 |
acted | +0.740 |
Neg logit contributions | |
advertisement | -1.352 |
IMAGES | -1.252 |
� | -1.228 |
PDATE | -1.206 |
� | -1.191 |
Token | and |
Token ID | 290 |
Feature activation | +6.21e-02 |
Pos logit contributions | |
advertisement | +0.294 |
IMAGES | +0.274 |
� | +0.263 |
� | +0.256 |
PDATE | +0.255 |
Neg logit contributions | |
erella | -0.297 |
inburgh | -0.251 |
achus | -0.249 |
lime | -0.229 |
cised | -0.214 |
Token | collect |
Token ID | 2824 |
Feature activation | -4.01e-01 |
Pos logit contributions | |
advertisement | +0.139 |
IMAGES | +0.129 |
� | +0.125 |
PDATE | +0.122 |
� | +0.121 |
Neg logit contributions | |
erella | -0.139 |
achus | -0.117 |
inburgh | -0.117 |
lime | -0.107 |
acted | -0.099 |
Token | gifts |
Token ID | 13201 |
Feature activation | -8.25e-03 |
Pos logit contributions | |
erella | +0.899 |
achus | +0.754 |
inburgh | +0.752 |
lime | +0.693 |
acted | +0.647 |
Neg logit contributions | |
advertisement | -0.891 |
IMAGES | -0.826 |
� | -0.798 |
PDATE | -0.784 |
� | -0.776 |
Token | , |
Token ID | 11 |
Feature activation | +9.67e-03 |
Pos logit contributions | |
advertisement | +0.044 |
IMAGES | +0.041 |
� | +0.039 |
PDATE | +0.038 |
� | +0.038 |
Neg logit contributions | |
erella | -0.046 |
inburgh | -0.039 |
achus | -0.039 |
lime | -0.036 |
cised | -0.033 |
Token | and |
Token ID | 290 |
Feature activation | -1.51e-02 |
Pos logit contributions | |
advertisement | +0.021 |
IMAGES | +0.019 |
� | +0.019 |
PDATE | +0.018 |
� | +0.018 |
Neg logit contributions | |
erella | -0.022 |
inburgh | -0.019 |
achus | -0.019 |
lime | -0.017 |
cised | -0.016 |
Token | could |
Token ID | 714 |
Feature activation | -5.18e-02 |
Pos logit contributions | |
erella | +0.034 |
inburgh | +0.029 |
achus | +0.029 |
lime | +0.026 |
cised | +0.024 |
Neg logit contributions | |
advertisement | -0.034 |
IMAGES | -0.031 |
� | -0.030 |
PDATE | -0.029 |
� | -0.029 |
Token | upset |
Token ID | 9247 |
Feature activation | -5.55e-01 |
Pos logit contributions | |
erella | +0.122 |
inburgh | +0.103 |
achus | +0.102 |
lime | +0.094 |
cised | +0.089 |
Neg logit contributions | |
advertisement | -0.111 |
IMAGES | -0.102 |
� | -0.099 |
PDATE | -0.098 |
� | -0.095 |
Token | Na |
Token ID | 11013 |
Feature activation | +2.93e-01 |
Pos logit contributions | |
erella | +1.298 |
inburgh | +1.107 |
achus | +1.095 |
lime | +1.016 |
acted | +0.951 |
Neg logit contributions | |
advertisement | -1.169 |
IMAGES | -1.076 |
� | -1.034 |
PDATE | -1.034 |
� | -1.013 |
Token | ` |
Token ID | 63 |
Feature activation | -1.18e+00 |
Pos logit contributions | |
advertisement | +0.709 |
IMAGES | +0.662 |
� | +0.637 |
PDATE | +0.621 |
tti | +0.617 |
Neg logit contributions | |
erella | -0.605 |
inburgh | -0.504 |
achus | -0.499 |
lime | -0.458 |
illary | -0.420 |
Token | Vi |
Token ID | 38432 |
Feature activation | -2.61e-01 |
Pos logit contributions | |
erella | +2.504 |
inburgh | +2.118 |
achus | +2.109 |
lime | +1.958 |
cised | +1.811 |
Neg logit contributions | |
advertisement | -2.727 |
IMAGES | -2.564 |
� | -2.453 |
PDATE | -2.392 |
� | -2.357 |
Token | in |
Token ID | 287 |
Feature activation | -1.14e-01 |
Pos logit contributions | |
erella | +0.578 |
inburgh | +0.489 |
achus | +0.487 |
lime | +0.445 |
cised | +0.411 |
Neg logit contributions | |
advertisement | -0.586 |
IMAGES | -0.541 |
� | -0.523 |
PDATE | -0.514 |
� | -0.510 |
Token | some |
Token ID | 617 |
Feature activation | -2.49e-01 |
Pos logit contributions | |
erella | +0.243 |
inburgh | +0.205 |
achus | +0.203 |
lime | +0.186 |
cised | +0.172 |
Neg logit contributions | |
advertisement | -0.266 |
IMAGES | -0.246 |
� | -0.238 |
PDATE | -0.233 |
� | -0.230 |
Token | universe |
Token ID | 6881 |
Feature activation | +2.27e-01 |
Pos logit contributions | |
erella | +0.608 |
inburgh | +0.525 |
achus | +0.519 |
lime | +0.487 |
illary | +0.459 |
Neg logit contributions | |
advertisement | -0.501 |
IMAGES | -0.458 |
� | -0.447 |
PDATE | -0.437 |
� | -0.425 |
Token | , |
Token ID | 11 |
Feature activation | +2.40e-02 |
Pos logit contributions | |
advertisement | +0.507 |
IMAGES | +0.472 |
� | +0.455 |
PDATE | +0.442 |
� | +0.440 |
Neg logit contributions | |
erella | -0.508 |
inburgh | -0.429 |
achus | -0.425 |
lime | -0.390 |
illary | -0.365 |
Token | negotiators |
Token ID | 45373 |
Feature activation | -1.75e-01 |
Pos logit contributions | |
advertisement | +0.156 |
IMAGES | +0.145 |
� | +0.140 |
PDATE | +0.136 |
� | +0.135 |
Neg logit contributions | |
erella | -0.151 |
inburgh | -0.127 |
achus | -0.127 |
lime | -0.116 |
acted | -0.108 |
Token | and |
Token ID | 290 |
Feature activation | +2.13e-02 |
Pos logit contributions | |
erella | +0.391 |
inburgh | +0.330 |
achus | +0.328 |
lime | +0.302 |
acted | +0.283 |
Neg logit contributions | |
advertisement | -0.396 |
IMAGES | -0.368 |
� | -0.354 |
PDATE | -0.344 |
� | -0.340 |
Token | Taliban |
Token ID | 15471 |
Feature activation | -3.47e-01 |
Pos logit contributions | |
advertisement | +0.045 |
IMAGES | +0.041 |
� | +0.040 |
PDATE | +0.039 |
� | +0.039 |
Neg logit contributions | |
erella | -0.050 |
inburgh | -0.043 |
achus | -0.043 |
lime | -0.039 |
cised | -0.037 |
Token | militants |
Token ID | 13162 |
Feature activation | +9.40e-02 |
Pos logit contributions | |
erella | +0.776 |
achus | +0.658 |
inburgh | +0.656 |
lime | +0.607 |
cised | +0.565 |
Neg logit contributions | |
advertisement | -0.774 |
IMAGES | -0.711 |
� | -0.692 |
PDATE | -0.674 |
� | -0.661 |
Token | broke |
Token ID | 6265 |
Feature activation | -1.22e-01 |
Pos logit contributions | |
advertisement | +0.213 |
IMAGES | +0.197 |
� | +0.190 |
PDATE | +0.185 |
� | +0.184 |
Neg logit contributions | |
erella | -0.208 |
inburgh | -0.176 |
achus | -0.175 |
lime | -0.160 |
cised | -0.149 |
Token | down |
Token ID | 866 |
Feature activation | -1.18e+00 |
Pos logit contributions | |
erella | +0.315 |
inburgh | +0.271 |
achus | +0.270 |
lime | +0.251 |
cised | +0.238 |
Neg logit contributions | |
advertisement | -0.232 |
IMAGES | -0.212 |
� | -0.203 |
PDATE | -0.196 |
� | -0.196 |
Token | earlier |
Token ID | 2961 |
Feature activation | +3.81e-01 |
Pos logit contributions | |
erella | +2.602 |
inburgh | +2.226 |
achus | +2.220 |
lime | +2.009 |
cised | +1.925 |
Neg logit contributions | |
advertisement | -2.601 |
IMAGES | -2.396 |
� | -2.336 |
PDATE | -2.317 |
� | -2.259 |
Token | this |
Token ID | 428 |
Feature activation | +5.90e-02 |
Pos logit contributions | |
advertisement | +0.866 |
IMAGES | +0.805 |
� | +0.776 |
� | +0.755 |
PDATE | +0.747 |
Neg logit contributions | |
erella | -0.838 |
inburgh | -0.706 |
achus | -0.698 |
lime | -0.639 |
cised | -0.595 |
Token | week |
Token ID | 1285 |
Feature activation | -1.58e-01 |
Pos logit contributions | |
advertisement | +0.161 |
IMAGES | +0.152 |
� | +0.147 |
� | +0.143 |
PDATE | +0.143 |
Neg logit contributions | |
erella | -0.103 |
inburgh | -0.083 |
achus | -0.082 |
lime | -0.073 |
cised | -0.066 |
Token | . |
Token ID | 13 |
Feature activation | +1.54e-01 |
Pos logit contributions | |
erella | +0.358 |
inburgh | +0.304 |
achus | +0.303 |
lime | +0.277 |
cised | +0.261 |
Neg logit contributions | |
advertisement | -0.349 |
IMAGES | -0.324 |
� | -0.311 |
PDATE | -0.305 |
� | -0.301 |
Token |
|
Token ID | 198 |
Feature activation | +3.37e-02 |
Pos logit contributions | |
advertisement | +0.341 |
IMAGES | +0.317 |
� | +0.305 |
� | +0.296 |
PDATE | +0.296 |
Neg logit contributions | |
erella | -0.349 |
inburgh | -0.295 |
achus | -0.292 |
lime | -0.268 |
cised | -0.251 |
Token | Gl |
Token ID | 2671 |
Feature activation | +2.33e-01 |
Pos logit contributions | |
erella | +0.251 |
inburgh | +0.212 |
achus | +0.208 |
lime | +0.193 |
illary | +0.178 |
Neg logit contributions | |
advertisement | -0.279 |
IMAGES | -0.262 |
� | -0.251 |
PDATE | -0.248 |
� | -0.244 |
Token | ü |
Token ID | 9116 |
Feature activation | -3.14e-01 |
Pos logit contributions | |
advertisement | +0.455 |
IMAGES | +0.420 |
� | +0.401 |
� | +0.388 |
PDATE | +0.387 |
Neg logit contributions | |
erella | -0.587 |
inburgh | -0.506 |
achus | -0.502 |
lime | -0.466 |
cised | -0.440 |
Token | ck |
Token ID | 694 |
Feature activation | -6.96e-02 |
Pos logit contributions | |
erella | +0.819 |
inburgh | +0.709 |
achus | +0.704 |
lime | +0.655 |
cised | +0.620 |
Neg logit contributions | |
advertisement | -0.587 |
IMAGES | -0.539 |
� | -0.514 |
PDATE | -0.498 |
� | -0.496 |
Token | is |
Token ID | 318 |
Feature activation | -7.18e-02 |
Pos logit contributions | |
erella | +0.158 |
inburgh | +0.134 |
achus | +0.133 |
lime | +0.122 |
illary | +0.114 |
Neg logit contributions | |
advertisement | -0.154 |
IMAGES | -0.143 |
� | -0.136 |
PDATE | -0.134 |
� | -0.133 |
Token | the |
Token ID | 262 |
Feature activation | -1.31e-01 |
Pos logit contributions | |
erella | +0.165 |
inburgh | +0.141 |
achus | +0.140 |
lime | +0.129 |
acted | +0.120 |
Neg logit contributions | |
advertisement | -0.156 |
IMAGES | -0.146 |
� | -0.138 |
PDATE | -0.135 |
� | -0.134 |
Token | closest |
Token ID | 11706 |
Feature activation | -1.17e+00 |
Pos logit contributions | |
erella | +0.327 |
inburgh | +0.282 |
achus | +0.279 |
lime | +0.259 |
acted | +0.245 |
Neg logit contributions | |
advertisement | -0.258 |
IMAGES | -0.238 |
� | -0.226 |
PDATE | -0.221 |
� | -0.218 |
Token | to |
Token ID | 284 |
Feature activation | -5.36e-01 |
Pos logit contributions | |
erella | +2.566 |
inburgh | +2.173 |
achus | +2.156 |
lime | +1.967 |
folk | +1.837 |
Neg logit contributions | |
advertisement | -2.652 |
IMAGES | -2.495 |
� | -2.373 |
PDATE | -2.367 |
� | -2.287 |
Token | being |
Token ID | 852 |
Feature activation | -1.42e-02 |
Pos logit contributions | |
erella | +1.206 |
inburgh | +1.024 |
achus | +1.017 |
lime | +0.931 |
acted | +0.873 |
Neg logit contributions | |
advertisement | -1.178 |
IMAGES | -1.109 |
� | -1.057 |
PDATE | -1.043 |
� | -1.019 |
Token | his |
Token ID | 465 |
Feature activation | -1.77e-02 |
Pos logit contributions | |
erella | +0.032 |
inburgh | +0.028 |
achus | +0.027 |
lime | +0.025 |
acted | +0.023 |
Neg logit contributions | |
advertisement | -0.031 |
IMAGES | -0.029 |
� | -0.028 |
PDATE | -0.027 |
� | -0.027 |
Token | secret |
Token ID | 3200 |
Feature activation | -4.79e-01 |
Pos logit contributions | |
erella | +0.041 |
inburgh | +0.035 |
achus | +0.034 |
lime | +0.032 |
acted | +0.029 |
Neg logit contributions | |
advertisement | -0.039 |
IMAGES | -0.036 |
� | -0.034 |
PDATE | -0.034 |
� | -0.033 |
Token | myth |
Token ID | 7918 |
Feature activation | -4.98e-02 |
Pos logit contributions | |
erella | +1.035 |
inburgh | +0.862 |
achus | +0.855 |
lime | +0.784 |
illary | +0.727 |
Neg logit contributions | |
advertisement | -1.110 |
IMAGES | -1.046 |
� | -0.997 |
PDATE | -0.983 |
� | -0.973 |
Token | to |
Token ID | 284 |
Feature activation | -9.80e-02 |
Pos logit contributions | |
erella | +0.900 |
inburgh | +0.760 |
achus | +0.757 |
lime | +0.698 |
acted | +0.651 |
Neg logit contributions | |
advertisement | -0.872 |
IMAGES | -0.825 |
� | -0.781 |
PDATE | -0.769 |
� | -0.754 |
Token | figure |
Token ID | 3785 |
Feature activation | +2.13e-01 |
Pos logit contributions | |
erella | +0.220 |
achus | +0.185 |
inburgh | +0.185 |
lime | +0.170 |
cised | +0.158 |
Neg logit contributions | |
advertisement | -0.219 |
IMAGES | -0.203 |
� | -0.197 |
PDATE | -0.194 |
� | -0.189 |
Token | out |
Token ID | 503 |
Feature activation | +7.14e-02 |
Pos logit contributions | |
advertisement | +0.358 |
IMAGES | +0.324 |
� | +0.308 |
� | +0.296 |
PDATE | +0.296 |
Neg logit contributions | |
erella | -0.595 |
inburgh | -0.521 |
achus | -0.518 |
lime | -0.485 |
cised | -0.460 |
Token | how |
Token ID | 703 |
Feature activation | -1.63e-01 |
Pos logit contributions | |
advertisement | +0.165 |
IMAGES | +0.154 |
� | +0.148 |
PDATE | +0.145 |
� | +0.144 |
Neg logit contributions | |
erella | -0.154 |
inburgh | -0.130 |
achus | -0.129 |
lime | -0.118 |
illary | -0.109 |
Token | to |
Token ID | 284 |
Feature activation | -2.48e-01 |
Pos logit contributions | |
erella | +0.372 |
inburgh | +0.313 |
achus | +0.311 |
lime | +0.291 |
illary | +0.268 |
Neg logit contributions | |
advertisement | -0.362 |
IMAGES | -0.334 |
� | -0.319 |
PDATE | -0.310 |
� | -0.308 |
Token | fix |
Token ID | 4259 |
Feature activation | -1.16e+00 |
Pos logit contributions | |
erella | +0.522 |
inburgh | +0.433 |
achus | +0.431 |
lime | +0.395 |
illary | +0.367 |
Neg logit contributions | |
advertisement | -0.591 |
IMAGES | -0.548 |
� | -0.533 |
PDATE | -0.526 |
� | -0.516 |
Token | the |
Token ID | 262 |
Feature activation | -1.17e-01 |
Pos logit contributions | |
erella | +2.507 |
inburgh | +2.096 |
achus | +2.080 |
lime | +1.947 |
illary | +1.806 |
Neg logit contributions | |
advertisement | -2.645 |
IMAGES | -2.425 |
� | -2.365 |
PDATE | -2.313 |
� | -2.254 |
Token | walk |
Token ID | 2513 |
Feature activation | -3.51e-01 |
Pos logit contributions | |
erella | +0.287 |
inburgh | +0.246 |
achus | +0.245 |
lime | +0.228 |
acted | +0.214 |
Neg logit contributions | |
advertisement | -0.234 |
IMAGES | -0.214 |
� | -0.207 |
PDATE | -0.201 |
� | -0.197 |
Token | ways |
Token ID | 1322 |
Feature activation | +1.49e-01 |
Pos logit contributions | |
erella | +0.784 |
inburgh | +0.660 |
achus | +0.648 |
lime | +0.606 |
illary | +0.566 |
Neg logit contributions | |
advertisement | -0.785 |
IMAGES | -0.731 |
� | -0.700 |
PDATE | -0.699 |
� | -0.676 |
Token | without |
Token ID | 1231 |
Feature activation | -3.56e-01 |
Pos logit contributions | |
advertisement | +0.334 |
IMAGES | +0.312 |
� | +0.301 |
PDATE | +0.294 |
� | +0.291 |
Neg logit contributions | |
erella | -0.332 |
inburgh | -0.282 |
achus | -0.279 |
lime | -0.257 |
acted | -0.239 |
Token | diver |
Token ID | 12312 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
erella | +0.863 |
inburgh | +0.740 |
achus | +0.738 |
lime | +0.688 |
illary | +0.641 |
Neg logit contributions | |
advertisement | -0.723 |
IMAGES | -0.667 |
� | -0.641 |
PDATE | -0.633 |
� | -0.619 |
Token | n |
Token ID | 77 |
Feature activation | -3.20e-01 |
Pos logit contributions | |
advertisement | +0.345 |
IMAGES | +0.315 |
� | +0.300 |
PDATE | +0.291 |
� | +0.288 |
Neg logit contributions | |
erella | -0.525 |
inburgh | -0.458 |
achus | -0.455 |
lime | -0.425 |
cised | -0.403 |
Token | has |
Token ID | 468 |
Feature activation | -2.53e-01 |
Pos logit contributions | |
erella | +0.701 |
inburgh | +0.591 |
achus | +0.589 |
lime | +0.535 |
acted | +0.502 |
Neg logit contributions | |
advertisement | -0.729 |
IMAGES | -0.678 |
� | -0.652 |
PDATE | -0.639 |
� | -0.630 |
Token | contributed |
Token ID | 8639 |
Feature activation | -1.50e-01 |
Pos logit contributions | |
erella | +0.575 |
inburgh | +0.488 |
achus | +0.485 |
lime | +0.444 |
acted | +0.416 |
Neg logit contributions | |
advertisement | -0.558 |
IMAGES | -0.521 |
� | -0.498 |
PDATE | -0.487 |
� | -0.481 |
Token | one |
Token ID | 530 |
Feature activation | +9.54e-03 |
Pos logit contributions | |
erella | +0.332 |
inburgh | +0.281 |
achus | +0.280 |
lime | +0.256 |
acted | +0.239 |
Neg logit contributions | |
advertisement | -0.337 |
IMAGES | -0.314 |
� | -0.303 |
PDATE | -0.296 |
� | -0.293 |
Token | mill |
Token ID | 3939 |
Feature activation | -8.39e-01 |
Pos logit contributions | |
advertisement | +0.022 |
IMAGES | +0.021 |
� | +0.020 |
PDATE | +0.019 |
� | +0.019 |
Neg logit contributions | |
erella | -0.020 |
inburgh | -0.017 |
achus | -0.017 |
lime | -0.016 |
illary | -0.015 |
Token | imeter |
Token ID | 16912 |
Feature activation | -1.16e+00 |
Pos logit contributions | |
erella | +1.385 |
inburgh | +1.089 |
achus | +1.068 |
lime | +0.968 |
illary | +0.872 |
Neg logit contributions | |
advertisement | -2.360 |
IMAGES | -2.249 |
� | -2.158 |
PDATE | -2.145 |
� | -2.095 |
Token | to |
Token ID | 284 |
Feature activation | +4.10e-02 |
Pos logit contributions | |
erella | +2.318 |
achus | +1.945 |
inburgh | +1.920 |
lime | +1.763 |
illary | +1.625 |
Neg logit contributions | |
advertisement | -2.803 |
IMAGES | -2.661 |
� | -2.528 |
PDATE | -2.513 |
� | -2.467 |
Token | sea |
Token ID | 5417 |
Feature activation | -3.25e-01 |
Pos logit contributions | |
advertisement | +0.096 |
IMAGES | +0.089 |
� | +0.086 |
� | +0.083 |
PDATE | +0.083 |
Neg logit contributions | |
erella | -0.088 |
inburgh | -0.074 |
achus | -0.073 |
lime | -0.067 |
cised | -0.062 |
Token | level |
Token ID | 1241 |
Feature activation | -4.28e-01 |
Pos logit contributions | |
erella | +0.800 |
inburgh | +0.689 |
achus | +0.682 |
lime | +0.634 |
illary | +0.597 |
Neg logit contributions | |
advertisement | -0.655 |
IMAGES | -0.603 |
� | -0.576 |
PDATE | -0.556 |
� | -0.555 |
Token | rise |
Token ID | 4485 |
Feature activation | -2.20e-01 |
Pos logit contributions | |
erella | +0.981 |
inburgh | +0.840 |
achus | +0.833 |
lime | +0.763 |
illary | +0.715 |
Neg logit contributions | |
advertisement | -0.927 |
IMAGES | -0.862 |
� | -0.826 |
PDATE | -0.803 |
� | -0.802 |
Token | between |
Token ID | 1022 |
Feature activation | -2.46e-01 |
Pos logit contributions | |
erella | +0.490 |
inburgh | +0.416 |
achus | +0.412 |
lime | +0.376 |
acted | +0.352 |
Neg logit contributions | |
advertisement | -0.492 |
IMAGES | -0.459 |
� | -0.442 |
PDATE | -0.430 |
� | -0.429 |
Token | For |
Token ID | 1890 |
Feature activation | +2.60e-01 |
Pos logit contributions | |
advertisement | +0.103 |
IMAGES | +0.095 |
� | +0.091 |
PDATE | +0.088 |
� | +0.088 |
Neg logit contributions | |
erella | -0.139 |
inburgh | -0.120 |
achus | -0.119 |
lime | -0.111 |
acted | -0.105 |
Token | comparison |
Token ID | 7208 |
Feature activation | -7.94e-02 |
Pos logit contributions | |
advertisement | +0.573 |
IMAGES | +0.533 |
� | +0.511 |
� | +0.498 |
PDATE | +0.493 |
Neg logit contributions | |
erella | -0.592 |
inburgh | -0.501 |
achus | -0.497 |
lime | -0.457 |
cised | -0.427 |
Token | � |
Token ID | 447 |
Feature activation | +1.16e-01 |
Pos logit contributions | |
erella | +0.175 |
inburgh | +0.147 |
achus | +0.146 |
lime | +0.133 |
cised | +0.125 |
Neg logit contributions | |
advertisement | -0.181 |
IMAGES | -0.168 |
� | -0.162 |
� | -0.158 |
PDATE | -0.157 |
Token | � |
Token ID | 247 |
Feature activation | +1.30e-02 |
Pos logit contributions | |
advertisement | +0.261 |
IMAGES | +0.245 |
� | +0.235 |
PDATE | +0.230 |
tti | +0.225 |
Neg logit contributions | |
erella | -0.257 |
inburgh | -0.216 |
achus | -0.214 |
lime | -0.197 |
acted | -0.184 |
Token | s |
Token ID | 82 |
Feature activation | -1.11e-01 |
Pos logit contributions | |
advertisement | +0.029 |
IMAGES | +0.027 |
� | +0.026 |
PDATE | +0.025 |
� | +0.025 |
Neg logit contributions | |
erella | -0.030 |
inburgh | -0.025 |
achus | -0.025 |
lime | -0.023 |
cised | -0.021 |
Token | sake |
Token ID | 11060 |
Feature activation | -1.16e+00 |
Pos logit contributions | |
erella | +0.246 |
inburgh | +0.207 |
achus | +0.205 |
lime | +0.188 |
cised | +0.175 |
Neg logit contributions | |
advertisement | -0.252 |
IMAGES | -0.235 |
� | -0.226 |
� | -0.220 |
PDATE | -0.219 |
Token | , |
Token ID | 11 |
Feature activation | -3.37e-01 |
Pos logit contributions | |
erella | +2.380 |
inburgh | +1.999 |
achus | +1.957 |
lime | +1.783 |
illary | +1.659 |
Neg logit contributions | |
advertisement | -2.769 |
IMAGES | -2.594 |
� | -2.469 |
PDATE | -2.416 |
� | -2.405 |
Token | American |
Token ID | 1605 |
Feature activation | -3.12e-02 |
Pos logit contributions | |
erella | +0.776 |
inburgh | +0.660 |
achus | +0.655 |
lime | +0.603 |
acted | +0.567 |
Neg logit contributions | |
advertisement | -0.728 |
IMAGES | -0.675 |
� | -0.652 |
PDATE | -0.633 |
� | -0.629 |
Token | Horror |
Token ID | 22732 |
Feature activation | +3.68e-01 |
Pos logit contributions | |
erella | +0.076 |
inburgh | +0.065 |
achus | +0.064 |
lime | +0.060 |
acted | +0.056 |
Neg logit contributions | |
advertisement | -0.064 |
IMAGES | -0.059 |
� | -0.057 |
PDATE | -0.055 |
� | -0.055 |
Token | Story |
Token ID | 8362 |
Feature activation | -3.24e-01 |
Pos logit contributions | |
advertisement | +0.954 |
IMAGES | +0.897 |
� | +0.868 |
� | +0.847 |
PDATE | +0.843 |
Neg logit contributions | |
erella | -0.693 |
inburgh | -0.566 |
achus | -0.559 |
lime | -0.502 |
cised | -0.461 |
Token | : |
Token ID | 25 |
Feature activation | +6.10e-02 |
Pos logit contributions | |
erella | +0.712 |
inburgh | +0.600 |
achus | +0.598 |
lime | +0.545 |
illary | +0.509 |
Neg logit contributions | |
advertisement | -0.729 |
IMAGES | -0.680 |
� | -0.655 |
PDATE | -0.645 |
� | -0.635 |
Token | - |
Token ID | 12 |
Feature activation | -1.98e-01 |
Pos logit contributions | |
erella | +0.880 |
inburgh | +0.724 |
achus | +0.724 |
lime | +0.659 |
acted | +0.625 |
Neg logit contributions | |
advertisement | -0.965 |
IMAGES | -0.913 |
� | -0.873 |
PDATE | -0.859 |
� | -0.852 |
Token | known |
Token ID | 4002 |
Feature activation | -6.04e-02 |
Pos logit contributions | |
erella | +0.433 |
inburgh | +0.361 |
achus | +0.360 |
lime | +0.331 |
acted | +0.311 |
Neg logit contributions | |
advertisement | -0.451 |
IMAGES | -0.427 |
� | -0.410 |
PDATE | -0.403 |
� | -0.399 |
Token | type |
Token ID | 2099 |
Feature activation | -1.31e-01 |
Pos logit contributions | |
erella | +0.132 |
inburgh | +0.110 |
achus | +0.109 |
lime | +0.100 |
acted | +0.093 |
Neg logit contributions | |
advertisement | -0.139 |
IMAGES | -0.130 |
� | -0.125 |
PDATE | -0.123 |
� | -0.121 |
Token | of |
Token ID | 286 |
Feature activation | -1.11e-01 |
Pos logit contributions | |
erella | +0.273 |
inburgh | +0.226 |
achus | +0.224 |
lime | +0.205 |
acted | +0.189 |
Neg logit contributions | |
advertisement | -0.316 |
IMAGES | -0.296 |
� | -0.285 |
PDATE | -0.279 |
� | -0.277 |
Token | ign |
Token ID | 3627 |
Feature activation | -1.97e-01 |
Pos logit contributions | |
erella | +0.229 |
inburgh | +0.187 |
achus | +0.187 |
lime | +0.171 |
acted | +0.158 |
Neg logit contributions | |
advertisement | -0.267 |
IMAGES | -0.250 |
� | -0.242 |
PDATE | -0.236 |
� | -0.235 |
Token | eous |
Token ID | 15303 |
Feature activation | -1.15e+00 |
Pos logit contributions | |
erella | +0.247 |
inburgh | +0.175 |
achus | +0.172 |
lime | +0.144 |
acted | +0.122 |
Neg logit contributions | |
advertisement | -0.637 |
IMAGES | -0.610 |
� | -0.593 |
PDATE | -0.589 |
� | -0.581 |
Token | rock |
Token ID | 3881 |
Feature activation | -2.89e-01 |
Pos logit contributions | |
erella | +2.275 |
inburgh | +1.874 |
achus | +1.859 |
lime | +1.741 |
izabeth | +1.592 |
Neg logit contributions | |
advertisement | -2.823 |
IMAGES | -2.649 |
PDATE | -2.548 |
� | -2.536 |
tti | -2.462 |
Token | found |
Token ID | 1043 |
Feature activation | +2.41e-01 |
Pos logit contributions | |
erella | +0.613 |
inburgh | +0.511 |
achus | +0.506 |
lime | +0.463 |
acted | +0.430 |
Neg logit contributions | |
advertisement | -0.682 |
IMAGES | -0.638 |
� | -0.615 |
PDATE | -0.604 |
� | -0.598 |
Token | in |
Token ID | 287 |
Feature activation | +3.45e-02 |
Pos logit contributions | |
advertisement | +0.496 |
IMAGES | +0.459 |
� | +0.440 |
� | +0.426 |
PDATE | +0.423 |
Neg logit contributions | |
erella | -0.582 |
inburgh | -0.498 |
achus | -0.494 |
lime | -0.457 |
cised | -0.430 |
Token | many |
Token ID | 867 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
advertisement | +0.083 |
IMAGES | +0.077 |
� | +0.075 |
PDATE | +0.073 |
� | +0.072 |
Neg logit contributions | |
erella | -0.072 |
achus | -0.060 |
inburgh | -0.060 |
lime | -0.054 |
acted | -0.050 |
Token | volcanic |
Token ID | 31513 |
Feature activation | -6.49e-01 |
Pos logit contributions | |
advertisement | +0.235 |
IMAGES | +0.218 |
� | +0.212 |
PDATE | +0.206 |
� | +0.205 |
Neg logit contributions | |
erella | -0.222 |
inburgh | -0.185 |
achus | -0.184 |
lime | -0.169 |
illary | -0.157 |
Token | / |
Token ID | 14 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
advertisement | +0.344 |
IMAGES | +0.308 |
� | +0.299 |
PDATE | +0.290 |
� | +0.289 |
Neg logit contributions | |
erella | -0.499 |
inburgh | -0.433 |
achus | -0.428 |
lime | -0.399 |
cised | -0.379 |
Token | File |
Token ID | 8979 |
Feature activation | +4.16e-01 |
Pos logit contributions | |
advertisement | +0.213 |
IMAGES | +0.195 |
� | +0.186 |
PDATE | +0.180 |
� | +0.180 |
Neg logit contributions | |
erella | -0.295 |
inburgh | -0.255 |
achus | -0.250 |
lime | -0.236 |
illary | -0.222 |
Token | photo |
Token ID | 4590 |
Feature activation | +4.36e-01 |
Pos logit contributions | |
advertisement | +0.823 |
IMAGES | +0.742 |
� | +0.720 |
� | +0.697 |
PDATE | +0.696 |
Neg logit contributions | |
erella | -1.047 |
inburgh | -0.901 |
achus | -0.890 |
lime | -0.832 |
cised | -0.784 |
Token | TP |
Token ID | 24525 |
Feature activation | +4.38e-01 |
Pos logit contributions | |
advertisement | +0.720 |
IMAGES | +0.647 |
� | +0.615 |
� | +0.591 |
PDATE | +0.588 |
Neg logit contributions | |
erella | -1.231 |
inburgh | -1.082 |
achus | -1.073 |
lime | -1.004 |
cised | -0.957 |
Token | X |
Token ID | 55 |
Feature activation | +4.49e-01 |
Pos logit contributions | |
advertisement | +1.012 |
IMAGES | +0.925 |
� | +0.896 |
PDATE | +0.873 |
� | +0.870 |
Neg logit contributions | |
erella | -0.961 |
inburgh | -0.804 |
achus | -0.798 |
lime | -0.737 |
illary | -0.685 |
Token | IMAGES |
Token ID | 45325 |
Feature activation | +1.63e+00 |
Pos logit contributions | |
advertisement | +0.153 |
� | +0.042 |
PDATE | +0.034 |
tti | +0.018 |
� | +0.006 |
Neg logit contributions | |
erella | -1.851 |
inburgh | -1.707 |
achus | -1.702 |
lime | -1.619 |
illary | -1.588 |
Token | OF |
Token ID | 3963 |
Feature activation | -1.44e-02 |
Pos logit contributions | |
advertisement | +2.960 |
IMAGES | +2.814 |
� | +2.604 |
OFF | +2.531 |
� | +2.456 |
Neg logit contributions | |
erella | -4.275 |
inburgh | -3.762 |
achus | -3.747 |
lime | -3.451 |
cised | -3.275 |
Token | THE |
Token ID | 3336 |
Feature activation | -2.29e-01 |
Pos logit contributions | |
erella | +0.033 |
inburgh | +0.028 |
achus | +0.028 |
lime | +0.026 |
cised | +0.024 |
Neg logit contributions | |
advertisement | -0.031 |
IMAGES | -0.029 |
� | -0.028 |
� | -0.027 |
PDATE | -0.027 |
Token | DAY |
Token ID | 24644 |
Feature activation | +5.48e-01 |
Pos logit contributions | |
erella | +0.499 |
inburgh | +0.416 |
achus | +0.410 |
lime | +0.376 |
acted | +0.354 |
Neg logit contributions | |
advertisement | -0.520 |
IMAGES | -0.468 |
� | -0.460 |
PDATE | -0.458 |
� | -0.452 |
Token | ) |
Token ID | 8 |
Feature activation | -1.45e-01 |
Pos logit contributions | |
advertisement | +1.059 |
IMAGES | +0.957 |
� | +0.924 |
� | +0.898 |
PDATE | +0.895 |
Neg logit contributions | |
erella | -1.399 |
inburgh | -1.205 |
achus | -1.193 |
lime | -1.113 |
cised | -1.049 |
Token | Chris |
Token ID | 5180 |
Feature activation | -1.14e-01 |
Pos logit contributions | |
erella | +0.351 |
inburgh | +0.299 |
achus | +0.298 |
lime | +0.275 |
illary | +0.259 |
Neg logit contributions | |
advertisement | -0.297 |
IMAGES | -0.270 |
� | -0.261 |
PDATE | -0.257 |
� | -0.253 |
Token | 55 |
Token ID | 2816 |
Feature activation | -6.79e-01 |
Pos logit contributions | |
erella | +0.163 |
inburgh | +0.140 |
achus | +0.139 |
lime | +0.131 |
acted | +0.123 |
Neg logit contributions | |
advertisement | -0.121 |
IMAGES | -0.113 |
PDATE | -0.107 |
� | -0.107 |
� | -0.104 |
Token | Recommend |
Token ID | 30361 |
Feature activation | +1.24e+00 |
Pos logit contributions | |
erella | +1.921 |
inburgh | +1.673 |
achus | +1.670 |
lime | +1.582 |
illary | +1.504 |
Neg logit contributions | |
advertisement | -1.102 |
IMAGES | -0.996 |
PDATE | -0.952 |
� | -0.948 |
tti | -0.903 |
Token |
|
Token ID | 198 |
Feature activation | +3.64e-01 |
Pos logit contributions | |
advertisement | +2.589 |
IMAGES | +2.379 |
� | +2.270 |
� | +2.192 |
OFF | +2.192 |
Neg logit contributions | |
erella | -2.945 |
inburgh | -2.568 |
achus | -2.518 |
lime | -2.298 |
cised | -2.184 |
Token |
|
Token ID | 198 |
Feature activation | +3.18e-01 |
Pos logit contributions | |
advertisement | +0.788 |
IMAGES | +0.729 |
� | +0.699 |
� | +0.676 |
tti | +0.671 |
Neg logit contributions | |
erella | -0.842 |
inburgh | -0.719 |
achus | -0.712 |
lime | -0.651 |
cised | -0.613 |
Token | Recommend |
Token ID | 41248 |
Feature activation | +1.22e+00 |
Pos logit contributions | |
advertisement | +0.402 |
IMAGES | +0.355 |
� | +0.327 |
PDATE | +0.314 |
� | +0.310 |
Neg logit contributions | |
erella | -1.018 |
inburgh | -0.907 |
achus | -0.901 |
lime | -0.856 |
illary | -0.817 |
Token | Share |
Token ID | 8734 |
Feature activation | +1.34e+00 |
Pos logit contributions | |
advertisement | +1.692 |
IMAGES | +1.504 |
� | +1.410 |
� | +1.357 |
OFF | +1.321 |
Neg logit contributions | |
erella | -3.725 |
inburgh | -3.332 |
achus | -3.287 |
lime | -3.087 |
cised | -2.973 |
Token | this |
Token ID | 428 |
Feature activation | -2.27e-01 |
Pos logit contributions | |
advertisement | +3.022 |
IMAGES | +2.815 |
� | +2.677 |
� | +2.615 |
OFF | +2.603 |
Neg logit contributions | |
erella | -2.952 |
inburgh | -2.490 |
achus | -2.463 |
lime | -2.250 |
cised | -2.109 |
Token | comment |
Token ID | 2912 |
Feature activation | -3.21e-01 |
Pos logit contributions | |
erella | +0.519 |
inburgh | +0.440 |
achus | +0.435 |
lime | +0.403 |
acted | +0.378 |
Neg logit contributions | |
advertisement | -0.493 |
IMAGES | -0.461 |
� | -0.442 |
PDATE | -0.439 |
� | -0.433 |
Token | on |
Token ID | 319 |
Feature activation | -1.31e-01 |
Pos logit contributions | |
erella | +0.709 |
inburgh | +0.594 |
achus | +0.589 |
lime | +0.541 |
acted | +0.505 |
Neg logit contributions | |
advertisement | -0.725 |
IMAGES | -0.677 |
� | -0.654 |
PDATE | -0.649 |
� | -0.638 |
Token | Facebook |
Token ID | 3203 |
Feature activation | +2.81e-01 |
Pos logit contributions | |
erella | +0.300 |
inburgh | +0.255 |
achus | +0.253 |
lime | +0.233 |
acted | +0.219 |
Neg logit contributions | |
advertisement | -0.282 |
IMAGES | -0.262 |
PDATE | -0.254 |
� | -0.253 |
� | -0.246 |
Token | Share |
Token ID | 8734 |
Feature activation | +1.09e+00 |
Pos logit contributions | |
advertisement | +0.415 |
IMAGES | +0.373 |
� | +0.349 |
� | +0.334 |
PDATE | +0.327 |
Neg logit contributions | |
erella | -0.845 |
inburgh | -0.747 |
achus | -0.742 |
lime | -0.697 |
cised | -0.667 |
Token | of |
Token ID | 286 |
Feature activation | -8.17e-02 |
Pos logit contributions | |
erella | +1.006 |
inburgh | +0.864 |
achus | +0.859 |
lime | +0.803 |
cised | +0.753 |
Neg logit contributions | |
advertisement | -0.803 |
IMAGES | -0.748 |
� | -0.706 |
PDATE | -0.692 |
� | -0.684 |
Token | Dal |
Token ID | 12348 |
Feature activation | -4.46e-01 |
Pos logit contributions | |
erella | +0.193 |
inburgh | +0.165 |
achus | +0.164 |
lime | +0.151 |
cised | +0.142 |
Neg logit contributions | |
advertisement | -0.171 |
IMAGES | -0.158 |
� | -0.152 |
PDATE | -0.148 |
� | -0.147 |
Token | K |
Token ID | 42 |
Feature activation | +3.99e-02 |
Pos logit contributions | |
erella | +0.960 |
inburgh | +0.811 |
achus | +0.798 |
lime | +0.729 |
cised | +0.677 |
Neg logit contributions | |
advertisement | -1.036 |
IMAGES | -0.965 |
� | -0.932 |
PDATE | -0.919 |
� | -0.897 |
Token | Profile |
Token ID | 13118 |
Feature activation | +5.34e-01 |
Pos logit contributions | |
advertisement | +0.085 |
IMAGES | +0.079 |
� | +0.076 |
PDATE | +0.074 |
� | +0.073 |
Neg logit contributions | |
erella | -0.094 |
inburgh | -0.080 |
achus | -0.080 |
lime | -0.074 |
illary | -0.069 |
Token | Blog |
Token ID | 14001 |
Feature activation | +5.04e-01 |
Pos logit contributions | |
advertisement | +0.997 |
IMAGES | +0.916 |
� | +0.873 |
� | +0.842 |
PDATE | +0.836 |
Neg logit contributions | |
erella | -1.391 |
inburgh | -1.204 |
achus | -1.195 |
lime | -1.112 |
cised | -1.052 |
Token | Joined |
Token ID | 16709 |
Feature activation | +2.10e+00 |
Pos logit contributions | |
advertisement | +1.026 |
IMAGES | +0.953 |
� | +0.910 |
� | +0.890 |
OFF | +0.876 |
Neg logit contributions | |
erella | -1.227 |
inburgh | -1.047 |
achus | -1.041 |
lime | -0.964 |
illary | -0.905 |
Token | June |
Token ID | 2795 |
Feature activation | -2.52e-02 |
Pos logit contributions | |
advertisement | +4.635 |
IMAGES | +4.384 |
� | +4.167 |
OFF | +4.076 |
� | +4.024 |
Neg logit contributions | |
erella | -4.678 |
inburgh | -3.984 |
achus | -3.923 |
lime | -3.552 |
cised | -3.362 |
Token | 2012 |
Token ID | 2321 |
Feature activation | -2.81e-01 |
Pos logit contributions | |
erella | +0.063 |
inburgh | +0.054 |
achus | +0.054 |
lime | +0.050 |
cised | +0.047 |
Neg logit contributions | |
advertisement | -0.049 |
IMAGES | -0.045 |
� | -0.044 |
� | -0.042 |
PDATE | -0.042 |
Token | Canada |
Token ID | 3340 |
Feature activation | +8.58e-02 |
Pos logit contributions | |
erella | +0.596 |
inburgh | +0.498 |
achus | +0.493 |
lime | +0.449 |
cised | +0.418 |
Neg logit contributions | |
advertisement | -0.662 |
IMAGES | -0.616 |
� | -0.594 |
PDATE | -0.580 |
� | -0.579 |
Token | 16 |
Token ID | 1467 |
Feature activation | -2.96e-01 |
Pos logit contributions | |
advertisement | +0.208 |
IMAGES | +0.195 |
� | +0.188 |
� | +0.183 |
PDATE | +0.182 |
Neg logit contributions | |
erella | -0.176 |
inburgh | -0.146 |
achus | -0.145 |
lime | -0.131 |
cised | -0.122 |
Token | 10 |
Token ID | 940 |
Feature activation | -3.46e-01 |
Pos logit contributions | |
erella | +0.757 |
inburgh | +0.653 |
achus | +0.647 |
lime | +0.602 |
cised | +0.569 |
Neg logit contributions | |
advertisement | -0.568 |
IMAGES | -0.521 |
� | -0.498 |
PDATE | -0.482 |
� | -0.482 |
Token | K |
Token ID | 42 |
Feature activation | -2.01e-01 |
Pos logit contributions | |
advertisement | +0.197 |
IMAGES | +0.182 |
� | +0.173 |
PDATE | +0.168 |
� | +0.167 |
Neg logit contributions | |
erella | -0.247 |
inburgh | -0.213 |
achus | -0.211 |
lime | -0.197 |
cised | -0.185 |
Token | ier |
Token ID | 959 |
Feature activation | -1.14e-01 |
Pos logit contributions | |
erella | +0.461 |
inburgh | +0.391 |
achus | +0.390 |
lime | +0.360 |
illary | +0.335 |
Neg logit contributions | |
advertisement | -0.437 |
IMAGES | -0.409 |
� | -0.390 |
PDATE | -0.381 |
� | -0.378 |
Token | an |
Token ID | 272 |
Feature activation | -2.29e-01 |
Pos logit contributions | |
erella | +0.269 |
inburgh | +0.230 |
achus | +0.228 |
lime | +0.211 |
cised | +0.198 |
Neg logit contributions | |
advertisement | -0.240 |
IMAGES | -0.223 |
� | -0.213 |
PDATE | -0.208 |
� | -0.206 |
Token | Profile |
Token ID | 13118 |
Feature activation | +2.54e-01 |
Pos logit contributions | |
erella | +0.503 |
inburgh | +0.422 |
achus | +0.421 |
lime | +0.387 |
acted | +0.359 |
Neg logit contributions | |
advertisement | -0.520 |
IMAGES | -0.486 |
� | -0.461 |
PDATE | -0.454 |
� | -0.451 |
Token | Blog |
Token ID | 14001 |
Feature activation | +3.07e-01 |
Pos logit contributions | |
advertisement | +0.479 |
IMAGES | +0.439 |
� | +0.419 |
� | +0.405 |
PDATE | +0.403 |
Neg logit contributions | |
erella | -0.656 |
inburgh | -0.568 |
achus | -0.564 |
lime | -0.524 |
cised | -0.496 |
Token | Joined |
Token ID | 16709 |
Feature activation | +1.95e+00 |
Pos logit contributions | |
advertisement | +0.642 |
IMAGES | +0.595 |
� | +0.570 |
� | +0.554 |
tti | +0.545 |
Neg logit contributions | |
erella | -0.733 |
inburgh | -0.626 |
achus | -0.620 |
lime | -0.573 |
cised | -0.537 |
Token | August |
Token ID | 2932 |
Feature activation | -4.58e-02 |
Pos logit contributions | |
advertisement | +4.359 |
IMAGES | +4.118 |
� | +3.927 |
OFF | +3.828 |
� | +3.789 |
Neg logit contributions | |
erella | -4.320 |
inburgh | -3.665 |
achus | -3.614 |
lime | -3.273 |
cised | -3.094 |
Token | 2011 |
Token ID | 2813 |
Feature activation | -1.67e-01 |
Pos logit contributions | |
erella | +0.111 |
inburgh | +0.095 |
achus | +0.094 |
lime | +0.087 |
acted | +0.082 |
Neg logit contributions | |
advertisement | -0.094 |
IMAGES | -0.086 |
� | -0.083 |
PDATE | -0.081 |
� | -0.080 |
Token | United |
Token ID | 1578 |
Feature activation | +5.72e-01 |
Pos logit contributions | |
erella | +0.354 |
inburgh | +0.296 |
achus | +0.293 |
lime | +0.267 |
cised | +0.248 |
Neg logit contributions | |
advertisement | -0.392 |
IMAGES | -0.365 |
� | -0.352 |
PDATE | -0.343 |
� | -0.343 |
Token | States |
Token ID | 1829 |
Feature activation | -2.63e-01 |
Pos logit contributions | |
advertisement | +1.409 |
IMAGES | +1.346 |
� | +1.295 |
� | +1.265 |
tti | +1.254 |
Neg logit contributions | |
erella | -1.127 |
inburgh | -0.921 |
achus | -0.919 |
lime | -0.833 |
cised | -0.760 |
Token | 6 |
Token ID | 718 |
Feature activation | -1.14e-02 |
Pos logit contributions | |
erella | +0.532 |
inburgh | +0.439 |
achus | +0.434 |
lime | +0.393 |
cised | +0.364 |
Neg logit contributions | |
advertisement | -0.647 |
IMAGES | -0.604 |
� | -0.585 |
PDATE | -0.573 |
� | -0.570 |
Token | , |
Token ID | 11 |
Feature activation | +2.44e-02 |
Pos logit contributions | |
advertisement | +0.061 |
IMAGES | +0.057 |
� | +0.055 |
� | +0.053 |
PDATE | +0.053 |
Neg logit contributions | |
erella | -0.062 |
inburgh | -0.052 |
achus | -0.052 |
lime | -0.047 |
cised | -0.044 |
Token | Clinton |
Token ID | 2605 |
Feature activation | -1.13e-01 |
Pos logit contributions | |
advertisement | +0.052 |
IMAGES | +0.048 |
� | +0.046 |
PDATE | +0.045 |
� | +0.045 |
Neg logit contributions | |
erella | -0.058 |
inburgh | -0.049 |
achus | -0.049 |
lime | -0.045 |
cised | -0.042 |
Token | Foundation |
Token ID | 5693 |
Feature activation | -2.00e-01 |
Pos logit contributions | |
erella | +0.242 |
inburgh | +0.202 |
achus | +0.201 |
lime | +0.182 |
cised | +0.170 |
Neg logit contributions | |
advertisement | -0.263 |
IMAGES | -0.246 |
� | -0.237 |
PDATE | -0.231 |
� | -0.230 |
Token | President |
Token ID | 1992 |
Feature activation | -9.21e-02 |
Pos logit contributions | |
erella | +0.466 |
inburgh | +0.394 |
achus | +0.392 |
lime | +0.361 |
illary | +0.340 |
Neg logit contributions | |
advertisement | -0.424 |
IMAGES | -0.394 |
� | -0.380 |
PDATE | -0.376 |
� | -0.369 |
Token | Donna |
Token ID | 29216 |
Feature activation | -1.09e-01 |
Pos logit contributions | |
erella | +0.231 |
inburgh | +0.198 |
achus | +0.196 |
lime | +0.183 |
acted | +0.174 |
Neg logit contributions | |
advertisement | -0.179 |
IMAGES | -0.164 |
PDATE | -0.159 |
� | -0.158 |
� | -0.153 |
Token | Shal |
Token ID | 47974 |
Feature activation | +1.45e+00 |
Pos logit contributions | |
erella | +0.292 |
inburgh | +0.252 |
achus | +0.249 |
lime | +0.235 |
acted | +0.226 |
Neg logit contributions | |
advertisement | -0.198 |
IMAGES | -0.179 |
� | -0.170 |
PDATE | -0.170 |
� | -0.162 |
Token | ala |
Token ID | 6081 |
Feature activation | -2.88e-01 |
Pos logit contributions | |
advertisement | +4.076 |
IMAGES | +3.809 |
� | +3.751 |
tti | +3.680 |
� | +3.649 |
Neg logit contributions | |
erella | -2.390 |
inburgh | -1.913 |
achus | -1.883 |
lime | -1.631 |
cised | -1.481 |
Token | summed |
Token ID | 32794 |
Feature activation | +4.04e-01 |
Pos logit contributions | |
erella | +0.607 |
inburgh | +0.505 |
achus | +0.501 |
lime | +0.458 |
acted | +0.427 |
Neg logit contributions | |
advertisement | -0.680 |
IMAGES | -0.638 |
� | -0.612 |
PDATE | -0.603 |
� | -0.596 |
Token | up |
Token ID | 510 |
Feature activation | -4.16e-02 |
Pos logit contributions | |
advertisement | +0.888 |
IMAGES | +0.823 |
� | +0.791 |
� | +0.769 |
PDATE | +0.765 |
Neg logit contributions | |
erella | -0.918 |
inburgh | -0.780 |
achus | -0.773 |
lime | -0.709 |
cised | -0.663 |
Token | the |
Token ID | 262 |
Feature activation | +1.53e-02 |
Pos logit contributions | |
erella | +0.087 |
inburgh | +0.073 |
achus | +0.072 |
lime | +0.066 |
acted | +0.061 |
Neg logit contributions | |
advertisement | -0.098 |
IMAGES | -0.092 |
� | -0.089 |
PDATE | -0.087 |
� | -0.086 |
Token | complicated |
Token ID | 8253 |
Feature activation | +7.85e-02 |
Pos logit contributions | |
advertisement | +0.032 |
IMAGES | +0.030 |
� | +0.029 |
PDATE | +0.028 |
� | +0.028 |
Neg logit contributions | |
erella | -0.036 |
inburgh | -0.031 |
achus | -0.030 |
lime | -0.028 |
acted | -0.026 |
Token | were |
Token ID | 547 |
Feature activation | -1.48e-01 |
Pos logit contributions | |
erella | +0.119 |
inburgh | +0.099 |
achus | +0.098 |
lime | +0.089 |
cised | +0.083 |
Neg logit contributions | |
advertisement | -0.130 |
IMAGES | -0.121 |
� | -0.117 |
PDATE | -0.114 |
� | -0.114 |
Token | able |
Token ID | 1498 |
Feature activation | +1.83e-01 |
Pos logit contributions | |
erella | +0.355 |
inburgh | +0.302 |
achus | +0.298 |
lime | +0.275 |
acted | +0.263 |
Neg logit contributions | |
advertisement | -0.307 |
IMAGES | -0.286 |
� | -0.272 |
PDATE | -0.269 |
� | -0.260 |
Token | to |
Token ID | 284 |
Feature activation | -3.24e-02 |
Pos logit contributions | |
advertisement | +0.379 |
IMAGES | +0.350 |
� | +0.336 |
� | +0.325 |
PDATE | +0.324 |
Neg logit contributions | |
erella | -0.439 |
inburgh | -0.376 |
achus | -0.372 |
lime | -0.344 |
cised | -0.323 |
Token | discriminate |
Token ID | 28433 |
Feature activation | -1.11e-01 |
Pos logit contributions | |
erella | +0.074 |
inburgh | +0.062 |
achus | +0.062 |
lime | +0.056 |
cised | +0.053 |
Neg logit contributions | |
advertisement | -0.072 |
IMAGES | -0.066 |
� | -0.064 |
PDATE | -0.063 |
� | -0.062 |
Token | between |
Token ID | 1022 |
Feature activation | +2.37e-01 |
Pos logit contributions | |
erella | +0.238 |
inburgh | +0.200 |
achus | +0.199 |
lime | +0.183 |
illary | +0.169 |
Neg logit contributions | |
advertisement | -0.256 |
IMAGES | -0.238 |
� | -0.232 |
PDATE | -0.227 |
� | -0.224 |
Token | 3 |
Token ID | 513 |
Feature activation | +3.94e-01 |
Pos logit contributions | |
advertisement | +0.530 |
IMAGES | +0.491 |
� | +0.478 |
PDATE | +0.466 |
� | +0.461 |
Neg logit contributions | |
erella | -0.530 |
inburgh | -0.448 |
achus | -0.443 |
lime | -0.408 |
acted | -0.381 |
Token | distinct |
Token ID | 7310 |
Feature activation | +4.21e-01 |
Pos logit contributions | |
advertisement | +0.809 |
IMAGES | +0.745 |
� | +0.720 |
PDATE | +0.696 |
� | +0.693 |
Neg logit contributions | |
erella | -0.955 |
inburgh | -0.820 |
achus | -0.809 |
lime | -0.752 |
illary | -0.708 |
Token | spatial |
Token ID | 21739 |
Feature activation | +1.47e-02 |
Pos logit contributions | |
advertisement | +0.932 |
IMAGES | +0.864 |
� | +0.836 |
PDATE | +0.813 |
� | +0.810 |
Neg logit contributions | |
erella | -0.954 |
inburgh | -0.808 |
achus | -0.799 |
lime | -0.737 |
illary | -0.687 |
Token | navigation |
Token ID | 16408 |
Feature activation | +1.64e-01 |
Pos logit contributions | |
advertisement | +0.034 |
IMAGES | +0.031 |
� | +0.030 |
PDATE | +0.030 |
tti | +0.029 |
Neg logit contributions | |
erella | -0.032 |
inburgh | -0.027 |
achus | -0.027 |
lime | -0.024 |
illary | -0.023 |
Token | strategies |
Token ID | 10064 |
Feature activation | +3.98e-01 |
Pos logit contributions | |
advertisement | +0.321 |
IMAGES | +0.295 |
� | +0.285 |
PDATE | +0.280 |
� | +0.276 |
Neg logit contributions | |
erella | -0.409 |
inburgh | -0.353 |
achus | -0.349 |
lime | -0.325 |
illary | -0.309 |
Token | , |
Token ID | 11 |
Feature activation | +1.64e-01 |
Pos logit contributions | |
advertisement | +0.892 |
IMAGES | +0.832 |
� | +0.801 |
� | +0.777 |
PDATE | +0.777 |
Neg logit contributions | |
erella | -0.886 |
inburgh | -0.748 |
achus | -0.742 |
lime | -0.680 |
cised | -0.635 |
Token | prisoner |
Token ID | 17234 |
Feature activation | +6.15e-01 |
Pos logit contributions | |
erella | +0.604 |
inburgh | +0.517 |
achus | +0.513 |
lime | +0.473 |
illary | +0.445 |
Neg logit contributions | |
advertisement | -0.567 |
IMAGES | -0.524 |
� | -0.499 |
PDATE | -0.490 |
� | -0.485 |
Token | , |
Token ID | 11 |
Feature activation | +9.31e-02 |
Pos logit contributions | |
advertisement | +1.392 |
IMAGES | +1.298 |
� | +1.246 |
� | +1.210 |
PDATE | +1.210 |
Neg logit contributions | |
erella | -1.362 |
inburgh | -1.149 |
achus | -1.139 |
lime | -1.043 |
cised | -0.972 |
Token | has |
Token ID | 468 |
Feature activation | +1.70e-01 |
Pos logit contributions | |
advertisement | +0.215 |
IMAGES | +0.200 |
� | +0.192 |
PDATE | +0.188 |
� | +0.187 |
Neg logit contributions | |
erella | -0.202 |
inburgh | -0.170 |
achus | -0.169 |
lime | -0.154 |
acted | -0.144 |
Token | not |
Token ID | 407 |
Feature activation | +1.12e-01 |
Pos logit contributions | |
advertisement | +0.368 |
IMAGES | +0.343 |
� | +0.327 |
PDATE | +0.322 |
� | +0.317 |
Neg logit contributions | |
erella | -0.392 |
inburgh | -0.334 |
achus | -0.330 |
lime | -0.304 |
acted | -0.286 |
Token | been |
Token ID | 587 |
Feature activation | +1.40e-01 |
Pos logit contributions | |
advertisement | +0.243 |
IMAGES | +0.226 |
PDATE | +0.216 |
� | +0.216 |
� | +0.210 |
Neg logit contributions | |
erella | -0.256 |
inburgh | -0.217 |
achus | -0.215 |
lime | -0.199 |
acted | -0.189 |
Token | released |
Token ID | 2716 |
Feature activation | +3.76e-01 |
Pos logit contributions | |
advertisement | +0.280 |
IMAGES | +0.260 |
� | +0.247 |
PDATE | +0.246 |
� | +0.238 |
Neg logit contributions | |
erella | -0.343 |
inburgh | -0.296 |
achus | -0.293 |
lime | -0.272 |
acted | -0.257 |
Token | from |
Token ID | 422 |
Feature activation | -2.52e-02 |
Pos logit contributions | |
advertisement | +0.822 |
IMAGES | +0.763 |
� | +0.733 |
� | +0.713 |
PDATE | +0.711 |
Neg logit contributions | |
erella | -0.859 |
inburgh | -0.731 |
achus | -0.725 |
lime | -0.665 |
cised | -0.623 |
Token | isolation |
Token ID | 15133 |
Feature activation | +1.22e-01 |
Pos logit contributions | |
erella | +0.070 |
achus | +0.062 |
inburgh | +0.061 |
lime | +0.057 |
illary | +0.054 |
Neg logit contributions | |
advertisement | -0.043 |
IMAGES | -0.039 |
� | -0.037 |
PDATE | -0.036 |
� | -0.035 |
Token | prison |
Token ID | 3770 |
Feature activation | +4.96e-01 |
Pos logit contributions | |
advertisement | +0.268 |
IMAGES | +0.250 |
� | +0.239 |
PDATE | +0.234 |
� | +0.232 |
Neg logit contributions | |
erella | -0.276 |
inburgh | -0.236 |
achus | -0.234 |
lime | -0.215 |
illary | -0.201 |
Token | after |
Token ID | 706 |
Feature activation | +6.09e-02 |
Pos logit contributions | |
advertisement | +1.063 |
IMAGES | +0.986 |
� | +0.946 |
� | +0.917 |
PDATE | +0.916 |
Neg logit contributions | |
erella | -1.157 |
inburgh | -0.988 |
achus | -0.980 |
lime | -0.901 |
cised | -0.846 |
Token | he |
Token ID | 339 |
Feature activation | +2.52e-01 |
Pos logit contributions | |
advertisement | +0.129 |
IMAGES | +0.120 |
� | +0.115 |
PDATE | +0.112 |
� | +0.111 |
Neg logit contributions | |
erella | -0.143 |
inburgh | -0.123 |
achus | -0.122 |
lime | -0.112 |
acted | -0.105 |
Token | . |
Token ID | 13 |
Feature activation | +1.99e-01 |
Pos logit contributions | |
advertisement | +0.722 |
IMAGES | +0.672 |
� | +0.646 |
� | +0.628 |
PDATE | +0.626 |
Neg logit contributions | |
erella | -0.719 |
inburgh | -0.606 |
achus | -0.601 |
lime | -0.553 |
cised | -0.515 |
Token | I |
Token ID | 314 |
Feature activation | +3.09e-01 |
Pos logit contributions | |
advertisement | +0.413 |
IMAGES | +0.381 |
� | +0.366 |
PDATE | +0.357 |
� | +0.355 |
Neg logit contributions | |
erella | -0.477 |
inburgh | -0.405 |
achus | -0.404 |
lime | -0.373 |
cised | -0.351 |
Token | am |
Token ID | 716 |
Feature activation | +1.88e-01 |
Pos logit contributions | |
advertisement | +0.726 |
IMAGES | +0.673 |
� | +0.650 |
PDATE | +0.638 |
� | +0.631 |
Neg logit contributions | |
erella | -0.665 |
inburgh | -0.552 |
achus | -0.549 |
lime | -0.501 |
cised | -0.468 |
Token | going |
Token ID | 1016 |
Feature activation | -3.86e-02 |
Pos logit contributions | |
advertisement | +0.414 |
IMAGES | +0.383 |
� | +0.368 |
PDATE | +0.363 |
� | +0.355 |
Neg logit contributions | |
erella | -0.428 |
inburgh | -0.364 |
achus | -0.359 |
lime | -0.330 |
cised | -0.311 |
Token | to |
Token ID | 284 |
Feature activation | +2.29e-01 |
Pos logit contributions | |
erella | +0.092 |
inburgh | +0.078 |
achus | +0.077 |
lime | +0.072 |
cised | +0.067 |
Neg logit contributions | |
advertisement | -0.081 |
IMAGES | -0.076 |
� | -0.072 |
� | -0.070 |
PDATE | -0.069 |
Token | show |
Token ID | 905 |
Feature activation | +5.19e-01 |
Pos logit contributions | |
advertisement | +0.497 |
IMAGES | +0.459 |
� | +0.443 |
PDATE | +0.436 |
� | +0.428 |
Neg logit contributions | |
erella | -0.529 |
inburgh | -0.448 |
achus | -0.445 |
lime | -0.408 |
cised | -0.385 |
Token | you |
Token ID | 345 |
Feature activation | +1.35e-01 |
Pos logit contributions | |
advertisement | +1.166 |
IMAGES | +1.084 |
� | +1.045 |
� | +1.014 |
PDATE | +1.013 |
Neg logit contributions | |
erella | -1.161 |
inburgh | -0.980 |
achus | -0.970 |
lime | -0.890 |
cised | -0.831 |
Token | a |
Token ID | 257 |
Feature activation | +1.55e-02 |
Pos logit contributions | |
advertisement | +0.311 |
IMAGES | +0.288 |
� | +0.280 |
PDATE | +0.273 |
� | +0.272 |
Neg logit contributions | |
erella | -0.293 |
inburgh | -0.245 |
achus | -0.242 |
lime | -0.223 |
illary | -0.207 |
Token | way |
Token ID | 835 |
Feature activation | +3.48e-01 |
Pos logit contributions | |
advertisement | +0.032 |
IMAGES | +0.030 |
� | +0.029 |
PDATE | +0.028 |
� | +0.028 |
Neg logit contributions | |
erella | -0.037 |
inburgh | -0.032 |
achus | -0.031 |
lime | -0.029 |
cised | -0.027 |
Token | to |
Token ID | 284 |
Feature activation | -1.13e-01 |
Pos logit contributions | |
advertisement | +0.742 |
IMAGES | +0.688 |
� | +0.660 |
� | +0.641 |
PDATE | +0.639 |
Neg logit contributions | |
erella | -0.816 |
inburgh | -0.694 |
achus | -0.688 |
lime | -0.635 |
cised | -0.595 |
Token | fix |
Token ID | 4259 |
Feature activation | -2.89e-01 |
Pos logit contributions | |
erella | +0.253 |
inburgh | +0.211 |
achus | +0.211 |
lime | +0.194 |
cised | +0.181 |
Neg logit contributions | |
advertisement | -0.253 |
IMAGES | -0.232 |
� | -0.226 |
PDATE | -0.224 |
� | -0.219 |
Token | and |
Token ID | 290 |
Feature activation | -1.80e-02 |
Pos logit contributions | |
advertisement | +0.104 |
IMAGES | +0.097 |
� | +0.093 |
PDATE | +0.090 |
� | +0.090 |
Neg logit contributions | |
erella | -0.106 |
inburgh | -0.089 |
achus | -0.088 |
lime | -0.081 |
illary | -0.076 |
Token | CEO |
Token ID | 6123 |
Feature activation | +6.90e-02 |
Pos logit contributions | |
erella | +0.044 |
inburgh | +0.038 |
achus | +0.037 |
lime | +0.035 |
acted | +0.033 |
Neg logit contributions | |
advertisement | -0.036 |
IMAGES | -0.034 |
� | -0.032 |
PDATE | -0.032 |
� | -0.031 |
Token | of |
Token ID | 286 |
Feature activation | +5.38e-02 |
Pos logit contributions | |
advertisement | +0.150 |
IMAGES | +0.139 |
� | +0.134 |
PDATE | +0.131 |
� | +0.130 |
Neg logit contributions | |
erella | -0.159 |
inburgh | -0.134 |
achus | -0.133 |
lime | -0.122 |
illary | -0.115 |
Token | the |
Token ID | 262 |
Feature activation | -4.88e-03 |
Pos logit contributions | |
advertisement | +0.117 |
IMAGES | +0.108 |
� | +0.104 |
PDATE | +0.101 |
� | +0.101 |
Neg logit contributions | |
erella | -0.125 |
inburgh | -0.105 |
achus | -0.105 |
lime | -0.096 |
illary | -0.090 |
Token | Climate |
Token ID | 13963 |
Feature activation | +7.07e-01 |
Pos logit contributions | |
erella | +0.011 |
inburgh | +0.010 |
achus | +0.010 |
lime | +0.009 |
illary | +0.008 |
Neg logit contributions | |
advertisement | -0.010 |
IMAGES | -0.010 |
� | -0.009 |
PDATE | -0.009 |
� | -0.009 |
Token | Works |
Token ID | 23044 |
Feature activation | +5.94e-01 |
Pos logit contributions | |
advertisement | +1.498 |
IMAGES | +1.380 |
� | +1.333 |
� | +1.286 |
tti | +1.276 |
Neg logit contributions | |
erella | -1.664 |
inburgh | -1.424 |
achus | -1.407 |
lime | -1.298 |
cised | -1.226 |
Token | Foundation |
Token ID | 5693 |
Feature activation | +8.76e-02 |
Pos logit contributions | |
advertisement | +1.231 |
IMAGES | +1.142 |
� | +1.093 |
PDATE | +1.062 |
� | +1.061 |
Neg logit contributions | |
erella | -1.424 |
inburgh | -1.220 |
achus | -1.209 |
lime | -1.118 |
cised | -1.047 |
Token | 2008 |
Token ID | 3648 |
Feature activation | -9.39e-02 |
Pos logit contributions | |
advertisement | +0.171 |
IMAGES | +0.157 |
� | +0.150 |
PDATE | +0.147 |
� | +0.145 |
Neg logit contributions | |
erella | -0.221 |
inburgh | -0.191 |
achus | -0.189 |
lime | -0.175 |
illary | -0.166 |
Token | - |
Token ID | 12 |
Feature activation | +5.70e-02 |
Pos logit contributions | |
erella | +0.201 |
inburgh | +0.169 |
achus | +0.165 |
lime | +0.153 |
acted | +0.142 |
Neg logit contributions | |
advertisement | -0.219 |
IMAGES | -0.204 |
� | -0.196 |
PDATE | -0.193 |
� | -0.192 |
Token | 2011 |
Token ID | 9804 |
Feature activation | +1.42e-01 |
Pos logit contributions | |
advertisement | +0.075 |
IMAGES | +0.067 |
� | +0.062 |
PDATE | +0.060 |
� | +0.059 |
Neg logit contributions | |
erella | -0.180 |
inburgh | -0.159 |
achus | -0.158 |
lime | -0.150 |
acted | -0.144 |
Token | ), |
Token ID | 828 |
Feature activation | +1.38e-01 |
Pos logit contributions | |
advertisement | +0.335 |
IMAGES | +0.312 |
� | +0.300 |
PDATE | +0.294 |
� | +0.292 |
Neg logit contributions | |
erella | -0.301 |
inburgh | -0.253 |
achus | -0.248 |
lime | -0.228 |
illary | -0.212 |
Token | upcoming |
Token ID | 7865 |
Feature activation | +3.00e-02 |
Pos logit contributions | |
advertisement | +0.043 |
IMAGES | +0.040 |
� | +0.038 |
PDATE | +0.037 |
� | +0.037 |
Neg logit contributions | |
erella | -0.049 |
inburgh | -0.042 |
achus | -0.041 |
lime | -0.038 |
acted | -0.036 |
Token | revival |
Token ID | 26624 |
Feature activation | +3.23e-02 |
Pos logit contributions | |
advertisement | +0.062 |
IMAGES | +0.057 |
� | +0.055 |
PDATE | +0.053 |
� | +0.053 |
Neg logit contributions | |
erella | -0.073 |
inburgh | -0.062 |
achus | -0.061 |
lime | -0.057 |
acted | -0.054 |
Token | of |
Token ID | 286 |
Feature activation | +2.09e-01 |
Pos logit contributions | |
advertisement | +0.078 |
IMAGES | +0.074 |
� | +0.071 |
PDATE | +0.069 |
� | +0.069 |
Neg logit contributions | |
erella | -0.066 |
inburgh | -0.055 |
achus | -0.054 |
lime | -0.049 |
acted | -0.046 |
Token | a |
Token ID | 257 |
Feature activation | +1.39e-01 |
Pos logit contributions | |
advertisement | +0.449 |
IMAGES | +0.414 |
� | +0.397 |
PDATE | +0.388 |
� | +0.385 |
Neg logit contributions | |
erella | -0.490 |
inburgh | -0.416 |
achus | -0.412 |
lime | -0.381 |
acted | -0.358 |
Token | historic |
Token ID | 9566 |
Feature activation | -3.38e-03 |
Pos logit contributions | |
advertisement | +0.293 |
IMAGES | +0.270 |
� | +0.257 |
PDATE | +0.254 |
� | +0.250 |
Neg logit contributions | |
erella | -0.332 |
inburgh | -0.282 |
achus | -0.278 |
lime | -0.258 |
acted | -0.244 |
Token | Broadway |
Token ID | 20617 |
Feature activation | +5.27e-01 |
Pos logit contributions | |
erella | +0.008 |
inburgh | +0.007 |
achus | +0.006 |
lime | +0.006 |
acted | +0.006 |
Neg logit contributions | |
advertisement | -0.007 |
IMAGES | -0.007 |
� | -0.007 |
PDATE | -0.006 |
� | -0.006 |
Token | musical |
Token ID | 10530 |
Feature activation | +2.82e-01 |
Pos logit contributions | |
advertisement | +1.221 |
IMAGES | +1.141 |
� | +1.093 |
PDATE | +1.071 |
� | +1.062 |
Neg logit contributions | |
erella | -1.146 |
inburgh | -0.961 |
achus | -0.945 |
lime | -0.866 |
acted | -0.807 |
Token | has |
Token ID | 468 |
Feature activation | +7.48e-02 |
Pos logit contributions | |
advertisement | +0.644 |
IMAGES | +0.599 |
� | +0.577 |
PDATE | +0.561 |
� | +0.561 |
Neg logit contributions | |
erella | -0.619 |
inburgh | -0.521 |
achus | -0.515 |
lime | -0.472 |
cised | -0.441 |
Token | gone |
Token ID | 3750 |
Feature activation | -3.67e-01 |
Pos logit contributions | |
advertisement | +0.171 |
IMAGES | +0.160 |
� | +0.153 |
PDATE | +0.151 |
� | +0.149 |
Neg logit contributions | |
erella | -0.164 |
inburgh | -0.138 |
achus | -0.136 |
lime | -0.125 |
acted | -0.117 |
Token | missing |
Token ID | 4814 |
Feature activation | -3.00e-01 |
Pos logit contributions | |
erella | +0.844 |
inburgh | +0.720 |
achus | +0.706 |
lime | +0.652 |
acted | +0.615 |
Neg logit contributions | |
advertisement | -0.803 |
IMAGES | -0.732 |
� | -0.715 |
PDATE | -0.698 |
� | -0.688 |
Token | . |
Token ID | 13 |
Feature activation | +5.27e-02 |
Pos logit contributions | |
erella | +0.693 |
inburgh | +0.590 |
achus | +0.585 |
lime | +0.539 |
acted | +0.507 |
Neg logit contributions | |
advertisement | -0.649 |
IMAGES | -0.601 |
� | -0.578 |
PDATE | -0.565 |
� | -0.560 |
Token | I |
Token ID | 314 |
Feature activation | -4.72e-02 |
Pos logit contributions | |
erella | +0.253 |
inburgh | +0.213 |
achus | +0.210 |
lime | +0.193 |
acted | +0.181 |
Neg logit contributions | |
advertisement | -0.261 |
IMAGES | -0.244 |
� | -0.235 |
PDATE | -0.230 |
� | -0.227 |
Token | � |
Token ID | 447 |
Feature activation | -2.27e-01 |
Pos logit contributions | |
erella | +0.114 |
inburgh | +0.096 |
achus | +0.096 |
lime | +0.089 |
cised | +0.084 |
Neg logit contributions | |
advertisement | -0.098 |
IMAGES | -0.091 |
PDATE | -0.086 |
� | -0.086 |
� | -0.083 |
Token | � |
Token ID | 247 |
Feature activation | -7.24e-02 |
Pos logit contributions | |
erella | +0.516 |
inburgh | +0.434 |
achus | +0.430 |
lime | +0.400 |
acted | +0.377 |
Neg logit contributions | |
advertisement | -0.489 |
IMAGES | -0.461 |
� | -0.439 |
PDATE | -0.432 |
tti | -0.421 |
Token | ve |
Token ID | 303 |
Feature activation | +2.37e-02 |
Pos logit contributions | |
erella | +0.155 |
inburgh | +0.130 |
achus | +0.127 |
lime | +0.119 |
cised | +0.111 |
Neg logit contributions | |
advertisement | -0.168 |
IMAGES | -0.159 |
� | -0.151 |
PDATE | -0.150 |
� | -0.145 |
Token | had |
Token ID | 550 |
Feature activation | -1.87e-02 |
Pos logit contributions | |
advertisement | +0.051 |
IMAGES | +0.048 |
� | +0.046 |
PDATE | +0.045 |
� | +0.044 |
Neg logit contributions | |
erella | -0.054 |
inburgh | -0.046 |
achus | -0.045 |
lime | -0.042 |
acted | -0.040 |
Token | the |
Token ID | 262 |
Feature activation | +5.86e-02 |
Pos logit contributions | |
erella | +0.042 |
inburgh | +0.036 |
achus | +0.035 |
lime | +0.033 |
acted | +0.031 |
Neg logit contributions | |
advertisement | -0.041 |
IMAGES | -0.038 |
� | -0.037 |
PDATE | -0.036 |
� | -0.036 |
Token | experience |
Token ID | 1998 |
Feature activation | -3.25e-01 |
Pos logit contributions | |
advertisement | +0.126 |
IMAGES | +0.118 |
� | +0.112 |
PDATE | +0.110 |
� | +0.109 |
Neg logit contributions | |
erella | -0.135 |
inburgh | -0.115 |
achus | -0.114 |
lime | -0.104 |
cised | -0.098 |
Token | of |
Token ID | 286 |
Feature activation | -1.10e-01 |
Pos logit contributions | |
erella | +0.679 |
inburgh | +0.570 |
achus | +0.559 |
lime | +0.510 |
illary | +0.475 |
Neg logit contributions | |
advertisement | -0.759 |
IMAGES | -0.727 |
� | -0.686 |
PDATE | -0.682 |
� | -0.672 |
Token | calling |
Token ID | 4585 |
Feature activation | +1.23e-01 |
Pos logit contributions | |
erella | +0.260 |
inburgh | +0.223 |
achus | +0.220 |
lime | +0.202 |
acted | +0.192 |
Neg logit contributions | |
advertisement | -0.228 |
IMAGES | -0.215 |
� | -0.203 |
PDATE | -0.203 |
� | -0.198 |
Token | agents |
Token ID | 6554 |
Feature activation | -2.98e-01 |
Pos logit contributions | |
advertisement | +0.270 |
IMAGES | +0.250 |
� | +0.241 |
PDATE | +0.237 |
� | +0.236 |
Neg logit contributions | |
erella | -0.282 |
inburgh | -0.239 |
achus | -0.235 |
lime | -0.217 |
acted | -0.203 |
Token | and |
Token ID | 290 |
Feature activation | +9.94e-02 |
Pos logit contributions | |
erella | +0.660 |
inburgh | +0.560 |
achus | +0.554 |
lime | +0.507 |
acted | +0.474 |
Neg logit contributions | |
advertisement | -0.666 |
IMAGES | -0.626 |
� | -0.597 |
PDATE | -0.590 |
� | -0.584 |
Token | be |
Token ID | 307 |
Feature activation | -5.08e-02 |
Pos logit contributions | |
advertisement | +0.196 |
IMAGES | +0.183 |
� | +0.176 |
PDATE | +0.173 |
� | +0.171 |
Neg logit contributions | |
erella | -0.195 |
inburgh | -0.164 |
achus | -0.162 |
lime | -0.148 |
cised | -0.139 |
Token | ruled |
Token ID | 8879 |
Feature activation | -7.14e-02 |
Pos logit contributions | |
erella | +0.119 |
inburgh | +0.101 |
achus | +0.100 |
lime | +0.092 |
cised | +0.086 |
Neg logit contributions | |
advertisement | -0.109 |
IMAGES | -0.101 |
� | -0.097 |
� | -0.094 |
PDATE | -0.094 |
Token | out |
Token ID | 503 |
Feature activation | +6.97e-01 |
Pos logit contributions | |
erella | +0.200 |
inburgh | +0.175 |
achus | +0.174 |
lime | +0.163 |
acted | +0.155 |
Neg logit contributions | |
advertisement | -0.120 |
IMAGES | -0.109 |
� | -0.102 |
� | -0.099 |
PDATE | -0.098 |
Token | . |
Token ID | 13 |
Feature activation | +4.92e-03 |
Pos logit contributions | |
advertisement | +1.522 |
IMAGES | +1.411 |
� | +1.359 |
� | +1.317 |
PDATE | +1.308 |
Neg logit contributions | |
erella | -1.599 |
inburgh | -1.356 |
achus | -1.344 |
lime | -1.236 |
cised | -1.160 |
Token | � |
Token ID | 447 |
Feature activation | +3.91e-02 |
Pos logit contributions | |
advertisement | +0.009 |
IMAGES | +0.009 |
� | +0.008 |
PDATE | +0.008 |
� | +0.008 |
Neg logit contributions | |
erella | -0.013 |
inburgh | -0.011 |
achus | -0.011 |
lime | -0.010 |
illary | -0.010 |
Token | � |
Token ID | 251 |
Feature activation | +1.25e-01 |
Pos logit contributions | |
advertisement | +0.054 |
IMAGES | +0.049 |
� | +0.045 |
PDATE | +0.043 |
tti | +0.042 |
Neg logit contributions | |
erella | -0.121 |
inburgh | -0.107 |
achus | -0.106 |
lime | -0.100 |
acted | -0.096 |
Token |
|
Token ID | 198 |
Feature activation | +3.14e-02 |
Pos logit contributions | |
advertisement | +0.280 |
IMAGES | +0.259 |
� | +0.248 |
� | +0.242 |
PDATE | +0.242 |
Neg logit contributions | |
erella | -0.282 |
inburgh | -0.238 |
achus | -0.236 |
lime | -0.216 |
illary | -0.202 |
Token |
|
Token ID | 198 |
Feature activation | +2.04e-02 |
Pos logit contributions | |
advertisement | +0.070 |
IMAGES | +0.066 |
� | +0.063 |
PDATE | +0.063 |
� | +0.062 |
Neg logit contributions | |
erella | -0.070 |
inburgh | -0.059 |
achus | -0.058 |
lime | -0.054 |
acted | -0.050 |
Token | A |
Token ID | 32 |
Feature activation | +1.29e-01 |
Pos logit contributions | |
advertisement | +0.036 |
IMAGES | +0.034 |
� | +0.031 |
PDATE | +0.031 |
� | +0.031 |
Neg logit contributions | |
erella | -0.055 |
inburgh | -0.048 |
achus | -0.047 |
lime | -0.044 |
acted | -0.042 |
Token | key |
Token ID | 1994 |
Feature activation | +3.68e-01 |
Pos logit contributions | |
advertisement | +0.271 |
IMAGES | +0.249 |
� | +0.240 |
PDATE | +0.233 |
� | +0.232 |
Neg logit contributions | |
erella | -0.305 |
inburgh | -0.261 |
achus | -0.260 |
lime | -0.238 |
cised | -0.224 |
Token | warning |
Token ID | 6509 |
Feature activation | +9.50e-02 |
Pos logit contributions | |
advertisement | +0.707 |
IMAGES | +0.653 |
� | +0.622 |
PDATE | +0.604 |
� | +0.599 |
Neg logit contributions | |
erella | -0.937 |
inburgh | -0.810 |
achus | -0.804 |
lime | -0.747 |
illary | -0.704 |
Token | , |
Token ID | 11 |
Feature activation | +3.12e-02 |
Pos logit contributions | |
erella | +0.372 |
inburgh | +0.312 |
achus | +0.309 |
lime | +0.282 |
cised | +0.263 |
Neg logit contributions | |
advertisement | -0.401 |
IMAGES | -0.374 |
� | -0.360 |
� | -0.351 |
PDATE | -0.349 |
Token | it |
Token ID | 340 |
Feature activation | -5.53e-02 |
Pos logit contributions | |
advertisement | +0.073 |
IMAGES | +0.068 |
� | +0.066 |
PDATE | +0.065 |
� | +0.064 |
Neg logit contributions | |
erella | -0.066 |
inburgh | -0.055 |
achus | -0.055 |
lime | -0.050 |
acted | -0.046 |
Token | is |
Token ID | 318 |
Feature activation | +1.34e-01 |
Pos logit contributions | |
erella | +0.122 |
inburgh | +0.103 |
achus | +0.102 |
lime | +0.094 |
cised | +0.087 |
Neg logit contributions | |
advertisement | -0.125 |
IMAGES | -0.116 |
� | -0.112 |
PDATE | -0.110 |
� | -0.109 |
Token | something |
Token ID | 1223 |
Feature activation | -3.46e-01 |
Pos logit contributions | |
advertisement | +0.296 |
IMAGES | +0.275 |
� | +0.265 |
� | +0.257 |
PDATE | +0.257 |
Neg logit contributions | |
erella | -0.304 |
inburgh | -0.257 |
achus | -0.255 |
lime | -0.234 |
cised | -0.219 |
Token | the |
Token ID | 262 |
Feature activation | +7.78e-02 |
Pos logit contributions | |
erella | +0.709 |
inburgh | +0.587 |
achus | +0.582 |
lime | +0.534 |
acted | +0.491 |
Neg logit contributions | |
advertisement | -0.832 |
IMAGES | -0.778 |
� | -0.756 |
PDATE | -0.744 |
� | -0.736 |
Token | player |
Token ID | 2137 |
Feature activation | +2.43e-01 |
Pos logit contributions | |
advertisement | +0.181 |
IMAGES | +0.168 |
� | +0.163 |
PDATE | +0.160 |
� | +0.158 |
Neg logit contributions | |
erella | -0.167 |
inburgh | -0.140 |
achus | -0.139 |
lime | -0.127 |
cised | -0.118 |
Token | should |
Token ID | 815 |
Feature activation | +8.10e-02 |
Pos logit contributions | |
advertisement | +0.556 |
IMAGES | +0.518 |
� | +0.499 |
� | +0.485 |
PDATE | +0.484 |
Neg logit contributions | |
erella | -0.530 |
inburgh | -0.445 |
achus | -0.441 |
lime | -0.403 |
cised | -0.376 |
Token | do |
Token ID | 466 |
Feature activation | -2.70e-01 |
Pos logit contributions | |
advertisement | +0.179 |
IMAGES | +0.166 |
� | +0.161 |
PDATE | +0.158 |
� | +0.156 |
Neg logit contributions | |
erella | -0.184 |
inburgh | -0.155 |
achus | -0.154 |
lime | -0.141 |
cised | -0.132 |
Token | . |
Token ID | 13 |
Feature activation | +1.36e-01 |
Pos logit contributions | |
erella | +0.569 |
inburgh | +0.476 |
achus | +0.470 |
lime | +0.427 |
acted | +0.399 |
Neg logit contributions | |
advertisement | -0.639 |
IMAGES | -0.594 |
� | -0.578 |
PDATE | -0.568 |
� | -0.565 |
Token | Personally |
Token ID | 31342 |
Feature activation | -2.98e-01 |
Pos logit contributions | |
advertisement | +0.277 |
IMAGES | +0.255 |
� | +0.246 |
PDATE | +0.241 |
� | +0.239 |
Neg logit contributions | |
erella | -0.329 |
inburgh | -0.281 |
achus | -0.279 |
lime | -0.258 |
acted | -0.243 |
Token | , |
Token ID | 11 |
Feature activation | +2.02e-01 |
Pos logit contributions | |
erella | +0.645 |
inburgh | +0.542 |
achus | +0.536 |
lime | +0.491 |
cised | +0.457 |
Neg logit contributions | |
advertisement | -0.689 |
IMAGES | -0.642 |
� | -0.617 |
� | -0.601 |
PDATE | -0.601 |
Token | - |
Token ID | 12 |
Feature activation | -3.64e-01 |
Pos logit contributions | |
advertisement | +0.057 |
IMAGES | +0.053 |
� | +0.050 |
� | +0.049 |
PDATE | +0.049 |
Neg logit contributions | |
erella | -0.059 |
inburgh | -0.050 |
achus | -0.049 |
lime | -0.046 |
acted | -0.043 |
Token | Rest |
Token ID | 8324 |
Feature activation | +9.25e-02 |
Pos logit contributions | |
erella | +0.902 |
inburgh | +0.770 |
achus | +0.766 |
lime | +0.714 |
acted | +0.671 |
Neg logit contributions | |
advertisement | -0.722 |
IMAGES | -0.666 |
� | -0.635 |
PDATE | -0.628 |
� | -0.618 |
Token | less |
Token ID | 1203 |
Feature activation | +4.85e-01 |
Pos logit contributions | |
advertisement | +0.183 |
IMAGES | +0.169 |
� | +0.162 |
� | +0.156 |
PDATE | +0.156 |
Neg logit contributions | |
erella | -0.231 |
inburgh | -0.199 |
achus | -0.197 |
lime | -0.183 |
cised | -0.172 |
Token | , |
Token ID | 11 |
Feature activation | +8.92e-02 |
Pos logit contributions | |
advertisement | +1.095 |
IMAGES | +1.019 |
� | +0.985 |
� | +0.955 |
PDATE | +0.948 |
Neg logit contributions | |
erella | -1.072 |
inburgh | -0.905 |
achus | -0.895 |
lime | -0.816 |
cised | -0.765 |
Token | Quick |
Token ID | 12029 |
Feature activation | +1.37e-01 |
Pos logit contributions | |
advertisement | +0.179 |
IMAGES | +0.165 |
� | +0.158 |
PDATE | +0.154 |
� | +0.153 |
Neg logit contributions | |
erella | -0.220 |
inburgh | -0.189 |
achus | -0.188 |
lime | -0.174 |
cised | -0.164 |
Token | to |
Token ID | 284 |
Feature activation | -7.44e-02 |
Pos logit contributions | |
advertisement | +0.306 |
IMAGES | +0.285 |
� | +0.273 |
� | +0.266 |
PDATE | +0.265 |
Neg logit contributions | |
erella | -0.307 |
inburgh | -0.259 |
achus | -0.257 |
lime | -0.236 |
cised | -0.220 |
Token | anger |
Token ID | 8993 |
Feature activation | +1.68e-01 |
Pos logit contributions | |
erella | +0.174 |
inburgh | +0.148 |
achus | +0.148 |
lime | +0.136 |
cised | +0.127 |
Neg logit contributions | |
advertisement | -0.158 |
IMAGES | -0.147 |
� | -0.141 |
PDATE | -0.137 |
� | -0.137 |
Token | , |
Token ID | 11 |
Feature activation | +4.68e-03 |
Pos logit contributions | |
advertisement | +0.388 |
IMAGES | +0.361 |
� | +0.347 |
� | +0.337 |
PDATE | +0.337 |
Neg logit contributions | |
erella | -0.365 |
inburgh | -0.307 |
achus | -0.304 |
lime | -0.278 |
cised | -0.259 |
Token | Unt |
Token ID | 26970 |
Feature activation | -3.09e-01 |
Pos logit contributions | |
advertisement | +0.010 |
IMAGES | +0.009 |
� | +0.008 |
PDATE | +0.008 |
� | +0.008 |
Neg logit contributions | |
erella | -0.011 |
inburgh | -0.010 |
achus | -0.010 |
lime | -0.009 |
cised | -0.008 |
Token | rust |
Token ID | 11469 |
Feature activation | -6.26e-01 |
Pos logit contributions | |
erella | +0.693 |
inburgh | +0.583 |
achus | +0.580 |
lime | +0.536 |
acted | +0.503 |
Neg logit contributions | |
advertisement | -0.689 |
IMAGES | -0.650 |
� | -0.628 |
PDATE | -0.619 |
� | -0.605 |
Token | worthy |
Token ID | 18275 |
Feature activation | +1.09e-01 |
Pos logit contributions | |
erella | +1.307 |
inburgh | +1.090 |
achus | +1.076 |
lime | +0.988 |
acted | +0.916 |
Neg logit contributions | |
advertisement | -1.486 |
IMAGES | -1.394 |
� | -1.342 |
PDATE | -1.315 |
� | -1.303 |
Token | Federal |
Token ID | 5618 |
Feature activation | +9.52e-02 |
Pos logit contributions | |
erella | +0.246 |
inburgh | +0.210 |
achus | +0.209 |
lime | +0.194 |
cised | +0.182 |
Neg logit contributions | |
advertisement | -0.207 |
IMAGES | -0.193 |
� | -0.186 |
PDATE | -0.181 |
� | -0.179 |
Token | Election |
Token ID | 14219 |
Feature activation | +2.43e-01 |
Pos logit contributions | |
advertisement | +0.229 |
IMAGES | +0.215 |
� | +0.207 |
PDATE | +0.202 |
� | +0.201 |
Neg logit contributions | |
erella | -0.197 |
inburgh | -0.163 |
achus | -0.162 |
lime | -0.147 |
cised | -0.136 |
Token | Commission |
Token ID | 4513 |
Feature activation | +6.49e-02 |
Pos logit contributions | |
advertisement | +0.568 |
IMAGES | +0.532 |
� | +0.512 |
� | +0.498 |
PDATE | +0.496 |
Neg logit contributions | |
erella | -0.519 |
inburgh | -0.435 |
achus | -0.432 |
lime | -0.393 |
cised | -0.366 |
Token | complaints |
Token ID | 9687 |
Feature activation | -1.73e-01 |
Pos logit contributions | |
advertisement | +0.139 |
IMAGES | +0.130 |
� | +0.126 |
PDATE | +0.121 |
� | +0.121 |
Neg logit contributions | |
erella | -0.150 |
inburgh | -0.127 |
achus | -0.126 |
lime | -0.117 |
illary | -0.109 |
Token | against |
Token ID | 1028 |
Feature activation | -1.12e-01 |
Pos logit contributions | |
erella | +0.408 |
inburgh | +0.348 |
achus | +0.347 |
lime | +0.319 |
acted | +0.300 |
Neg logit contributions | |
advertisement | -0.363 |
IMAGES | -0.340 |
� | -0.325 |
PDATE | -0.316 |
� | -0.314 |
Token | the |
Token ID | 262 |
Feature activation | -1.16e-01 |
Pos logit contributions | |
erella | +0.238 |
inburgh | +0.200 |
achus | +0.199 |
lime | +0.180 |
illary | +0.169 |
Neg logit contributions | |
advertisement | -0.259 |
IMAGES | -0.243 |
� | -0.235 |
PDATE | -0.230 |
� | -0.226 |
Token | Trump |
Token ID | 1301 |
Feature activation | -8.43e-02 |
Pos logit contributions | |
erella | +0.250 |
inburgh | +0.212 |
achus | +0.210 |
lime | +0.191 |
cised | +0.180 |
Neg logit contributions | |
advertisement | -0.265 |
IMAGES | -0.248 |
� | -0.240 |
PDATE | -0.235 |
� | -0.231 |
Token | campaign |
Token ID | 1923 |
Feature activation | -2.52e-01 |
Pos logit contributions | |
erella | +0.221 |
inburgh | +0.192 |
achus | +0.191 |
lime | +0.177 |
illary | +0.168 |
Neg logit contributions | |
advertisement | -0.155 |
IMAGES | -0.142 |
� | -0.135 |
PDATE | -0.132 |
� | -0.130 |
Token | for |
Token ID | 329 |
Feature activation | -5.19e-02 |
Pos logit contributions | |
erella | +0.554 |
inburgh | +0.470 |
achus | +0.469 |
lime | +0.425 |
illary | +0.402 |
Neg logit contributions | |
advertisement | -0.563 |
IMAGES | -0.528 |
� | -0.508 |
PDATE | -0.497 |
� | -0.493 |
Token | solic |
Token ID | 25806 |
Feature activation | -1.56e-02 |
Pos logit contributions | |
erella | +0.123 |
inburgh | +0.105 |
achus | +0.104 |
lime | +0.096 |
acted | +0.091 |
Neg logit contributions | |
advertisement | -0.108 |
IMAGES | -0.101 |
� | -0.097 |
PDATE | -0.094 |
� | -0.093 |
Token | iting |
Token ID | 1780 |
Feature activation | -2.08e-01 |
Pos logit contributions | |
erella | +0.037 |
inburgh | +0.032 |
achus | +0.031 |
lime | +0.029 |
illary | +0.027 |
Neg logit contributions | |
advertisement | -0.033 |
IMAGES | -0.031 |
� | -0.030 |
PDATE | -0.029 |
� | -0.028 |
Token | only |
Token ID | 691 |
Feature activation | -2.56e-01 |
Pos logit contributions | |
advertisement | +0.193 |
IMAGES | +0.179 |
� | +0.171 |
PDATE | +0.167 |
� | +0.166 |
Neg logit contributions | |
erella | -0.221 |
inburgh | -0.190 |
achus | -0.188 |
lime | -0.173 |
acted | -0.163 |
Token | rarely |
Token ID | 8365 |
Feature activation | -3.65e-01 |
Pos logit contributions | |
erella | +0.587 |
inburgh | +0.500 |
achus | +0.499 |
lime | +0.455 |
acted | +0.426 |
Neg logit contributions | |
advertisement | -0.550 |
IMAGES | -0.511 |
� | -0.495 |
PDATE | -0.484 |
� | -0.478 |
Token | . |
Token ID | 13 |
Feature activation | +6.30e-02 |
Pos logit contributions | |
erella | +0.821 |
inburgh | +0.694 |
achus | +0.686 |
lime | +0.629 |
acted | +0.589 |
Neg logit contributions | |
advertisement | -0.813 |
IMAGES | -0.757 |
� | -0.726 |
PDATE | -0.714 |
� | -0.707 |
Token |
|
Token ID | 198 |
Feature activation | +3.48e-02 |
Pos logit contributions | |
advertisement | +0.139 |
IMAGES | +0.129 |
� | +0.124 |
PDATE | +0.123 |
� | +0.122 |
Neg logit contributions | |
erella | -0.142 |
inburgh | -0.119 |
achus | -0.118 |
lime | -0.108 |
illary | -0.102 |
Token |
|
Token ID | 198 |
Feature activation | +2.42e-02 |
Pos logit contributions | |
advertisement | +0.077 |
IMAGES | +0.072 |
� | +0.069 |
PDATE | +0.068 |
� | +0.068 |
Neg logit contributions | |
erella | -0.078 |
inburgh | -0.066 |
achus | -0.065 |
lime | -0.060 |
cised | -0.056 |
Token | Advertisement |
Token ID | 4723 |
Feature activation | -8.44e-01 |
Pos logit contributions | |
advertisement | +0.051 |
IMAGES | +0.048 |
PDATE | +0.046 |
� | +0.045 |
� | +0.045 |
Neg logit contributions | |
erella | -0.056 |
inburgh | -0.048 |
achus | -0.047 |
lime | -0.044 |
acted | -0.041 |
Token | Continue |
Token ID | 10054 |
Feature activation | +4.82e-02 |
Pos logit contributions | |
erella | +1.895 |
achus | +1.609 |
inburgh | +1.603 |
lime | +1.459 |
illary | +1.392 |
Neg logit contributions | |
advertisement | -1.827 |
IMAGES | -1.691 |
PDATE | -1.653 |
� | -1.648 |
� | -1.603 |
Token | reading |
Token ID | 3555 |
Feature activation | -3.28e-01 |
Pos logit contributions | |
advertisement | +0.078 |
IMAGES | +0.071 |
� | +0.067 |
PDATE | +0.065 |
� | +0.064 |
Neg logit contributions | |
erella | -0.137 |
inburgh | -0.121 |
achus | -0.120 |
lime | -0.113 |
illary | -0.107 |
Token | the |
Token ID | 262 |
Feature activation | +1.24e-01 |
Pos logit contributions | |
erella | +0.660 |
inburgh | +0.541 |
achus | +0.536 |
lime | +0.491 |
acted | +0.455 |
Neg logit contributions | |
advertisement | -0.799 |
IMAGES | -0.754 |
� | -0.728 |
PDATE | -0.725 |
� | -0.714 |
Token | main |
Token ID | 1388 |
Feature activation | -3.41e-02 |
Pos logit contributions | |
advertisement | +0.207 |
IMAGES | +0.191 |
� | +0.180 |
PDATE | +0.178 |
� | +0.174 |
Neg logit contributions | |
erella | -0.348 |
inburgh | -0.303 |
achus | -0.301 |
lime | -0.282 |
acted | -0.269 |
Token | story |
Token ID | 1621 |
Feature activation | -6.88e-02 |
Pos logit contributions | |
erella | +0.071 |
inburgh | +0.059 |
achus | +0.059 |
lime | +0.054 |
acted | +0.050 |
Neg logit contributions | |
advertisement | -0.080 |
IMAGES | -0.076 |
� | -0.073 |
PDATE | -0.073 |
� | -0.072 |
Token | JR |
Token ID | 44817 |
Feature activation | -5.14e-01 |
Pos logit contributions | |
erella | +0.023 |
inburgh | +0.020 |
achus | +0.020 |
lime | +0.018 |
acted | +0.017 |
Neg logit contributions | |
advertisement | -0.019 |
IMAGES | -0.018 |
� | -0.017 |
PDATE | -0.016 |
� | -0.016 |
Token | ) |
Token ID | 8 |
Feature activation | +4.22e-02 |
Pos logit contributions | |
erella | +1.406 |
inburgh | +1.240 |
achus | +1.221 |
lime | +1.152 |
acted | +1.084 |
Neg logit contributions | |
advertisement | -0.887 |
IMAGES | -0.806 |
� | -0.753 |
PDATE | -0.735 |
� | -0.729 |
Token | April |
Token ID | 3035 |
Feature activation | -2.51e-01 |
Pos logit contributions | |
advertisement | +0.105 |
IMAGES | +0.098 |
� | +0.096 |
PDATE | +0.094 |
� | +0.094 |
Neg logit contributions | |
erella | -0.083 |
inburgh | -0.068 |
achus | -0.067 |
lime | -0.061 |
acted | -0.056 |
Token | 14 |
Token ID | 1478 |
Feature activation | -2.27e-01 |
Pos logit contributions | |
erella | +0.484 |
inburgh | +0.393 |
achus | +0.388 |
lime | +0.355 |
acted | +0.327 |
Neg logit contributions | |
advertisement | -0.633 |
IMAGES | -0.594 |
PDATE | -0.579 |
� | -0.578 |
� | -0.564 |
Token | , |
Token ID | 11 |
Feature activation | +1.33e-01 |
Pos logit contributions | |
erella | +0.469 |
inburgh | +0.388 |
achus | +0.385 |
lime | +0.353 |
acted | +0.327 |
Neg logit contributions | |
advertisement | -0.544 |
IMAGES | -0.508 |
PDATE | -0.491 |
� | -0.490 |
� | -0.479 |
Token | 2017 |
Token ID | 2177 |
Feature activation | -6.95e-01 |
Pos logit contributions | |
advertisement | +0.305 |
IMAGES | +0.282 |
� | +0.272 |
PDATE | +0.268 |
� | +0.266 |
Neg logit contributions | |
erella | -0.292 |
inburgh | -0.246 |
achus | -0.243 |
lime | -0.223 |
cised | -0.208 |
Token |
|
Token ID | 198 |
Feature activation | -8.58e-02 |
Pos logit contributions | |
erella | +1.488 |
inburgh | +1.249 |
achus | +1.230 |
lime | +1.130 |
illary | +1.065 |
Neg logit contributions | |
advertisement | -1.597 |
IMAGES | -1.458 |
� | -1.423 |
PDATE | -1.398 |
� | -1.385 |
Token |
|
Token ID | 198 |
Feature activation | +3.76e-02 |
Pos logit contributions | |
erella | +0.192 |
inburgh | +0.161 |
achus | +0.160 |
lime | +0.147 |
acted | +0.137 |
Neg logit contributions | |
advertisement | -0.192 |
IMAGES | -0.179 |
� | -0.172 |
PDATE | -0.169 |
� | -0.167 |
Token | Michael |
Token ID | 13256 |
Feature activation | +2.66e-01 |
Pos logit contributions | |
advertisement | +0.072 |
IMAGES | +0.067 |
� | +0.063 |
PDATE | +0.062 |
� | +0.062 |
Neg logit contributions | |
erella | -0.096 |
inburgh | -0.083 |
achus | -0.082 |
lime | -0.076 |
acted | -0.072 |
Token | G |
Token ID | 402 |
Feature activation | +4.25e-01 |
Pos logit contributions | |
advertisement | +0.530 |
IMAGES | +0.488 |
� | +0.467 |
PDATE | +0.453 |
� | +0.452 |
Neg logit contributions | |
erella | -0.661 |
inburgh | -0.569 |
achus | -0.564 |
lime | -0.523 |
acted | -0.494 |
Token | . |
Token ID | 13 |
Feature activation | +7.45e-02 |
Pos logit contributions | |
advertisement | +0.974 |
IMAGES | +0.911 |
� | +0.878 |
PDATE | +0.856 |
� | +0.852 |
Neg logit contributions | |
erella | -0.931 |
inburgh | -0.786 |
achus | -0.779 |
lime | -0.709 |
acted | -0.664 |
Token | that |
Token ID | 326 |
Feature activation | -1.23e-01 |
Pos logit contributions | |
erella | +0.426 |
inburgh | +0.358 |
achus | +0.355 |
lime | +0.327 |
illary | +0.306 |
Neg logit contributions | |
advertisement | -0.433 |
IMAGES | -0.404 |
� | -0.388 |
PDATE | -0.380 |
� | -0.376 |
Token | never |
Token ID | 1239 |
Feature activation | -5.53e-02 |
Pos logit contributions | |
erella | +0.287 |
inburgh | +0.243 |
achus | +0.242 |
lime | +0.223 |
illary | +0.209 |
Neg logit contributions | |
advertisement | -0.263 |
IMAGES | -0.246 |
� | -0.234 |
PDATE | -0.230 |
� | -0.227 |
Token | sleeps |
Token ID | 44263 |
Feature activation | +4.90e-01 |
Pos logit contributions | |
erella | +0.132 |
inburgh | +0.112 |
achus | +0.111 |
lime | +0.103 |
acted | +0.096 |
Neg logit contributions | |
advertisement | -0.116 |
IMAGES | -0.108 |
� | -0.104 |
PDATE | -0.102 |
� | -0.101 |
Token | and |
Token ID | 290 |
Feature activation | -4.14e-02 |
Pos logit contributions | |
advertisement | +1.121 |
IMAGES | +1.046 |
� | +1.002 |
� | +0.976 |
PDATE | +0.976 |
Neg logit contributions | |
erella | -1.076 |
inburgh | -0.906 |
achus | -0.897 |
lime | -0.822 |
cised | -0.764 |
Token | the |
Token ID | 262 |
Feature activation | -1.17e-01 |
Pos logit contributions | |
erella | +0.097 |
inburgh | +0.083 |
achus | +0.083 |
lime | +0.076 |
acted | +0.072 |
Neg logit contributions | |
advertisement | -0.087 |
IMAGES | -0.081 |
� | -0.077 |
PDATE | -0.076 |
� | -0.075 |
Token | moral |
Token ID | 6573 |
Feature activation | -6.68e-01 |
Pos logit contributions | |
erella | +0.271 |
inburgh | +0.230 |
achus | +0.229 |
lime | +0.210 |
acted | +0.197 |
Neg logit contributions | |
advertisement | -0.252 |
IMAGES | -0.236 |
� | -0.225 |
PDATE | -0.220 |
� | -0.218 |
Token | decay |
Token ID | 22119 |
Feature activation | +1.63e-01 |
Pos logit contributions | |
erella | +1.542 |
inburgh | +1.315 |
achus | +1.295 |
lime | +1.201 |
illary | +1.120 |
Neg logit contributions | |
advertisement | -1.429 |
IMAGES | -1.334 |
� | -1.274 |
PDATE | -1.264 |
� | -1.236 |
Token | of |
Token ID | 286 |
Feature activation | +8.03e-02 |
Pos logit contributions | |
advertisement | +0.376 |
IMAGES | +0.353 |
� | +0.338 |
PDATE | +0.332 |
� | +0.328 |
Neg logit contributions | |
erella | -0.353 |
inburgh | -0.297 |
achus | -0.294 |
lime | -0.269 |
illary | -0.250 |
Token | a |
Token ID | 257 |
Feature activation | +4.71e-02 |
Pos logit contributions | |
advertisement | +0.177 |
IMAGES | +0.165 |
� | +0.158 |
PDATE | +0.154 |
� | +0.152 |
Neg logit contributions | |
erella | -0.182 |
inburgh | -0.154 |
achus | -0.153 |
lime | -0.140 |
cised | -0.132 |
Token | des |
Token ID | 748 |
Feature activation | +1.69e-02 |
Pos logit contributions | |
advertisement | +0.108 |
IMAGES | +0.101 |
� | +0.097 |
PDATE | +0.095 |
� | +0.094 |
Neg logit contributions | |
erella | -0.103 |
inburgh | -0.086 |
achus | -0.086 |
lime | -0.078 |
cised | -0.073 |
Token | ens |
Token ID | 641 |
Feature activation | +1.80e-01 |
Pos logit contributions | |
advertisement | +0.038 |
IMAGES | +0.035 |
� | +0.034 |
� | +0.033 |
PDATE | +0.033 |
Neg logit contributions | |
erella | -0.038 |
inburgh | -0.032 |
achus | -0.032 |
lime | -0.029 |
cised | -0.027 |
Token | of |
Token ID | 286 |
Feature activation | -3.01e-02 |
Pos logit contributions | |
erella | +0.318 |
inburgh | +0.263 |
achus | +0.263 |
lime | +0.239 |
illary | +0.221 |
Neg logit contributions | |
advertisement | -0.382 |
IMAGES | -0.361 |
� | -0.346 |
PDATE | -0.338 |
� | -0.336 |
Token | Nikki |
Token ID | 34108 |
Feature activation | +7.61e-02 |
Pos logit contributions | |
erella | +0.068 |
inburgh | +0.058 |
achus | +0.058 |
lime | +0.053 |
cised | +0.050 |
Neg logit contributions | |
advertisement | -0.066 |
IMAGES | -0.061 |
� | -0.059 |
PDATE | -0.058 |
� | -0.057 |
Token | J |
Token ID | 449 |
Feature activation | -7.30e-02 |
Pos logit contributions | |
advertisement | +0.154 |
IMAGES | +0.142 |
� | +0.136 |
PDATE | +0.133 |
� | +0.130 |
Neg logit contributions | |
erella | -0.187 |
inburgh | -0.161 |
achus | -0.161 |
lime | -0.148 |
acted | -0.139 |
Token | oly |
Token ID | 3366 |
Feature activation | +1.67e-01 |
Pos logit contributions | |
erella | +0.167 |
inburgh | +0.141 |
achus | +0.141 |
lime | +0.128 |
acted | +0.121 |
Neg logit contributions | |
advertisement | -0.160 |
IMAGES | -0.150 |
� | -0.144 |
PDATE | -0.140 |
� | -0.139 |
Token | burned |
Token ID | 11544 |
Feature activation | +1.17e-01 |
Pos logit contributions | |
advertisement | +0.375 |
IMAGES | +0.349 |
� | +0.336 |
� | +0.326 |
PDATE | +0.325 |
Neg logit contributions | |
erella | -0.373 |
inburgh | -0.316 |
achus | -0.313 |
lime | -0.287 |
cised | -0.268 |
Token | down |
Token ID | 866 |
Feature activation | -6.24e-01 |
Pos logit contributions | |
advertisement | +0.232 |
IMAGES | +0.214 |
� | +0.205 |
PDATE | +0.198 |
� | +0.198 |
Neg logit contributions | |
erella | -0.289 |
inburgh | -0.250 |
achus | -0.248 |
lime | -0.230 |
cised | -0.217 |
Token | in |
Token ID | 287 |
Feature activation | +5.59e-02 |
Pos logit contributions | |
erella | +1.286 |
inburgh | +1.087 |
achus | +1.078 |
lime | +0.980 |
cised | +0.918 |
Neg logit contributions | |
advertisement | -1.477 |
IMAGES | -1.385 |
� | -1.334 |
PDATE | -1.308 |
� | -1.296 |
Token | 2017 |
Token ID | 2177 |
Feature activation | -3.88e-01 |
Pos logit contributions | |
advertisement | +0.123 |
IMAGES | +0.114 |
� | +0.110 |
PDATE | +0.107 |
� | +0.106 |
Neg logit contributions | |
erella | -0.127 |
inburgh | -0.108 |
achus | -0.107 |
lime | -0.098 |
cised | -0.092 |
Token | , |
Token ID | 11 |
Feature activation | -4.99e-02 |
Pos logit contributions | |
erella | +0.801 |
inburgh | +0.674 |
achus | +0.665 |
lime | +0.608 |
acted | +0.564 |
Neg logit contributions | |
advertisement | -0.923 |
IMAGES | -0.866 |
� | -0.836 |
PDATE | -0.813 |
� | -0.810 |
Token | killing |
Token ID | 5170 |
Feature activation | +1.81e-01 |
Pos logit contributions | |
erella | +0.110 |
inburgh | +0.093 |
achus | +0.092 |
lime | +0.084 |
acted | +0.079 |
Neg logit contributions | |
advertisement | -0.113 |
IMAGES | -0.105 |
� | -0.101 |
PDATE | -0.099 |
� | -0.098 |
Token | five |
Token ID | 1936 |
Feature activation | -1.52e-01 |
Pos logit contributions | |
advertisement | +0.419 |
IMAGES | +0.390 |
� | +0.376 |
PDATE | +0.370 |
� | +0.363 |
Neg logit contributions | |
erella | -0.387 |
inburgh | -0.328 |
achus | -0.324 |
lime | -0.297 |
cised | -0.276 |
Token | . |
Token ID | 13 |
Feature activation | -1.66e-01 |
Pos logit contributions | |
erella | +1.476 |
inburgh | +1.256 |
achus | +1.236 |
lime | +1.133 |
illary | +1.059 |
Neg logit contributions | |
advertisement | -1.528 |
IMAGES | -1.422 |
� | -1.361 |
PDATE | -1.336 |
� | -1.330 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -7.88e-02 |
Pos logit contributions | |
erella | +0.394 |
inburgh | +0.338 |
achus | +0.335 |
lime | +0.309 |
illary | +0.292 |
Neg logit contributions | |
advertisement | -0.345 |
IMAGES | -0.319 |
� | -0.305 |
� | -0.298 |
PDATE | -0.298 |
Token | C |
Token ID | 34 |
Feature activation | +1.84e-01 |
Pos logit contributions | |
erella | +0.201 |
inburgh | +0.173 |
achus | +0.172 |
lime | +0.162 |
acted | +0.152 |
Neg logit contributions | |
advertisement | -0.148 |
IMAGES | -0.140 |
� | -0.132 |
PDATE | -0.131 |
� | -0.128 |
Token | ort |
Token ID | 419 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
advertisement | +0.395 |
IMAGES | +0.368 |
� | +0.352 |
PDATE | +0.342 |
� | +0.341 |
Neg logit contributions | |
erella | -0.432 |
inburgh | -0.367 |
achus | -0.365 |
lime | -0.336 |
illary | -0.314 |
Token | ic |
Token ID | 291 |
Feature activation | -1.56e-01 |
Pos logit contributions | |
erella | +0.315 |
inburgh | +0.266 |
achus | +0.264 |
lime | +0.242 |
illary | +0.227 |
Neg logit contributions | |
advertisement | -0.304 |
IMAGES | -0.284 |
� | -0.272 |
PDATE | -0.266 |
� | -0.264 |
Token | otrop |
Token ID | 34248 |
Feature activation | -6.89e-01 |
Pos logit contributions | |
erella | +0.369 |
inburgh | +0.311 |
achus | +0.310 |
lime | +0.286 |
illary | +0.273 |
Neg logit contributions | |
advertisement | -0.331 |
IMAGES | -0.307 |
� | -0.295 |
PDATE | -0.290 |
� | -0.286 |
Token | in |
Token ID | 259 |
Feature activation | -1.48e-01 |
Pos logit contributions | |
erella | +1.962 |
inburgh | +1.705 |
achus | +1.702 |
lime | +1.587 |
illary | +1.528 |
Neg logit contributions | |
advertisement | -1.128 |
IMAGES | -1.039 |
� | -0.976 |
PDATE | -0.959 |
� | -0.932 |
Token | - |
Token ID | 12 |
Feature activation | -1.90e-01 |
Pos logit contributions | |
erella | +0.329 |
inburgh | +0.278 |
achus | +0.278 |
lime | +0.250 |
illary | +0.239 |
Neg logit contributions | |
advertisement | -0.330 |
IMAGES | -0.309 |
� | -0.294 |
PDATE | -0.290 |
� | -0.288 |
Token | re |
Token ID | 260 |
Feature activation | -1.11e-01 |
Pos logit contributions | |
erella | +0.377 |
inburgh | +0.313 |
achus | +0.310 |
lime | +0.280 |
acted | +0.261 |
Neg logit contributions | |
advertisement | -0.469 |
IMAGES | -0.444 |
� | -0.424 |
PDATE | -0.417 |
� | -0.415 |
Token | leasing |
Token ID | 48764 |
Feature activation | +1.43e-01 |
Pos logit contributions | |
erella | +0.287 |
inburgh | +0.254 |
achus | +0.252 |
lime | +0.232 |
acted | +0.226 |
Neg logit contributions | |
advertisement | -0.202 |
IMAGES | -0.193 |
� | -0.179 |
PDATE | -0.175 |
tti | -0.171 |
Token | factor |
Token ID | 5766 |
Feature activation | +1.96e-01 |
Pos logit contributions | |
advertisement | +0.254 |
IMAGES | +0.233 |
� | +0.217 |
PDATE | +0.212 |
� | +0.209 |
Neg logit contributions | |
erella | -0.388 |
inburgh | -0.339 |
achus | -0.335 |
lime | -0.313 |
illary | -0.299 |