Token | i |
Token ID | 72 |
Feature activation | -1.50e+00 |
Pos logit contributions | |
ringe | +3.072 |
ritten | +3.003 |
itness | +2.774 |
kefeller | +2.760 |
ickson | +2.687 |
Neg logit contributions | |
REDACTED | -1.970 |
Capitol | -1.725 |
Clown | -1.725 |
Certain | -1.701 |
Franch | -1.642 |
Token | Dar |
Token ID | 7491 |
Feature activation | -1.84e+00 |
Pos logit contributions | |
ringe | +2.424 |
ritten | +2.362 |
itness | +2.181 |
kefeller | +2.158 |
ickson | +2.118 |
Neg logit contributions | |
REDACTED | -1.625 |
Clown | -1.440 |
Capitol | -1.433 |
Certain | -1.409 |
Franch | -1.371 |
Token | uss |
Token ID | 1046 |
Feature activation | +9.39e-01 |
Pos logit contributions | |
ringe | +2.691 |
ritten | +2.625 |
itness | +2.400 |
kefeller | +2.387 |
ickson | +2.310 |
Neg logit contributions | |
REDACTED | -2.249 |
Clown | -2.032 |
Capitol | -2.015 |
Certain | -2.011 |
Franch | -1.952 |
Token | alam |
Token ID | 44949 |
Feature activation | -6.31e-01 |
Pos logit contributions | |
REDACTED | +0.907 |
Capitol | +0.812 |
Clown | +0.790 |
Certain | +0.780 |
Franch | +0.770 |
Neg logit contributions | |
ringe | -1.582 |
ritten | -1.570 |
kefeller | -1.485 |
itness | -1.432 |
ickson | -1.384 |
Token | ( |
Token ID | 357 |
Feature activation | +1.05e+00 |
Pos logit contributions | |
ringe | +1.013 |
ritten | +0.986 |
itness | +0.913 |
kefeller | +0.901 |
ickson | +0.882 |
Neg logit contributions | |
REDACTED | -0.688 |
Clown | -0.612 |
Capitol | -0.611 |
Certain | -0.600 |
Franch | -0.584 |
Token | Aut |
Token ID | 16541 |
Feature activation | +5.29e+00 |
Pos logit contributions | |
REDACTED | +1.322 |
Capitol | +1.159 |
Clown | +1.153 |
Certain | +1.146 |
Franch | +1.115 |
Neg logit contributions | |
ringe | -1.502 |
ritten | -1.466 |
itness | -1.342 |
kefeller | -1.339 |
ickson | -1.302 |
Token | or |
Token ID | 273 |
Feature activation | -1.71e-01 |
Pos logit contributions | |
REDACTED | +6.285 |
Capitol | +6.010 |
Certain | +5.869 |
Clown | +5.860 |
Franch | +5.756 |
Neg logit contributions | |
ringe | -6.879 |
ritten | -6.703 |
itness | -6.101 |
kefeller | -6.097 |
ickson | -5.776 |
Token | iti |
Token ID | 8846 |
Feature activation | +8.77e-01 |
Pos logit contributions | |
ringe | +0.240 |
ritten | +0.232 |
itness | +0.213 |
kefeller | +0.211 |
ickson | +0.205 |
Neg logit contributions | |
REDACTED | -0.215 |
Clown | -0.197 |
Capitol | -0.197 |
Certain | -0.194 |
Franch | -0.190 |
Token | Mon |
Token ID | 2892 |
Feature activation | +1.85e+00 |
Pos logit contributions | |
REDACTED | +1.024 |
Capitol | +0.933 |
Clown | +0.932 |
Certain | +0.918 |
Franch | +0.899 |
Neg logit contributions | |
ringe | -1.317 |
ritten | -1.282 |
itness | -1.177 |
kefeller | -1.171 |
ickson | -1.146 |
Token | et |
Token ID | 316 |
Feature activation | -1.19e-01 |
Pos logit contributions | |
REDACTED | +2.236 |
Capitol | +2.085 |
Clown | +2.061 |
Certain | +2.053 |
Franch | +2.020 |
Neg logit contributions | |
ringe | -2.608 |
ritten | -2.519 |
itness | -2.286 |
kefeller | -2.283 |
ickson | -2.240 |
Token | ari |
Token ID | 2743 |
Feature activation | +9.75e-01 |
Pos logit contributions | |
ringe | +0.177 |
ritten | +0.172 |
itness | +0.157 |
kefeller | +0.156 |
ickson | +0.153 |
Neg logit contributions | |
REDACTED | -0.140 |
Clown | -0.127 |
Capitol | -0.127 |
Certain | -0.125 |
Franch | -0.122 |
Token | drama |
Token ID | 10512 |
Feature activation | +9.77e-01 |
Pos logit contributions | |
ringe | +1.410 |
ritten | +1.370 |
itness | +1.248 |
kefeller | +1.233 |
ickson | +1.215 |
Neg logit contributions | |
REDACTED | -1.264 |
Clown | -1.138 |
Capitol | -1.125 |
Certain | -1.115 |
Franch | -1.092 |
Token | parody |
Token ID | 29695 |
Feature activation | +2.09e+00 |
Pos logit contributions | |
REDACTED | +1.256 |
Clown | +1.154 |
Capitol | +1.138 |
Certain | +1.124 |
Franch | +1.094 |
Neg logit contributions | |
ringe | -1.353 |
ritten | -1.322 |
kefeller | -1.200 |
itness | -1.192 |
ickson | -1.158 |
Token | sk |
Token ID | 1341 |
Feature activation | +9.08e-01 |
Pos logit contributions | |
REDACTED | +2.574 |
Clown | +2.403 |
Capitol | +2.367 |
Certain | +2.342 |
Franch | +2.272 |
Neg logit contributions | |
ringe | -2.966 |
ritten | -2.895 |
kefeller | -2.679 |
itness | -2.642 |
ickson | -2.569 |
Token | it |
Token ID | 270 |
Feature activation | +9.60e-01 |
Pos logit contributions | |
REDACTED | +1.122 |
Clown | +1.034 |
Certain | +1.025 |
Capitol | +1.016 |
Franch | +0.983 |
Neg logit contributions | |
ringe | -1.250 |
ritten | -1.217 |
kefeller | -1.140 |
itness | -1.105 |
ickson | -1.073 |
Token | " |
Token ID | 366 |
Feature activation | -4.15e-01 |
Pos logit contributions | |
REDACTED | +1.185 |
Clown | +1.096 |
Capitol | +1.076 |
Certain | +1.066 |
Franch | +1.036 |
Neg logit contributions | |
ringe | -1.377 |
ritten | -1.338 |
kefeller | -1.231 |
itness | -1.225 |
ickson | -1.190 |
Token | Veter |
Token ID | 45182 |
Feature activation | +5.22e+00 |
Pos logit contributions | |
ringe | +0.626 |
ritten | +0.610 |
itness | +0.559 |
kefeller | +0.555 |
ickson | +0.545 |
Neg logit contributions | |
REDACTED | -0.491 |
Clown | -0.438 |
Capitol | -0.434 |
Certain | -0.429 |
Franch | -0.417 |
Token | in |
Token ID | 259 |
Feature activation | -4.48e-01 |
Pos logit contributions | |
REDACTED | +5.578 |
Clown | +5.231 |
Certain | +5.218 |
Capitol | +5.209 |
atorium | +4.950 |
Neg logit contributions | |
ringe | -7.713 |
ritten | -7.533 |
kefeller | -6.967 |
itness | -6.787 |
ickson | -6.689 |
Token | arian |
Token ID | 3699 |
Feature activation | +4.75e-02 |
Pos logit contributions | |
ringe | +0.635 |
ritten | +0.622 |
kefeller | +0.562 |
itness | +0.562 |
ickson | +0.547 |
Neg logit contributions | |
REDACTED | -0.567 |
Clown | -0.515 |
Capitol | -0.509 |
Certain | -0.503 |
Franch | -0.491 |
Token | 's |
Token ID | 338 |
Feature activation | +4.67e-01 |
Pos logit contributions | |
REDACTED | +0.056 |
Clown | +0.051 |
Capitol | +0.050 |
Certain | +0.050 |
Franch | +0.048 |
Neg logit contributions | |
ringe | -0.071 |
ritten | -0.070 |
itness | -0.063 |
kefeller | -0.063 |
ickson | -0.062 |
Token | Hospital |
Token ID | 9256 |
Feature activation | -6.92e-01 |
Pos logit contributions | |
REDACTED | +0.489 |
Clown | +0.450 |
Capitol | +0.443 |
Certain | +0.436 |
Franch | +0.424 |
Neg logit contributions | |
ringe | -0.757 |
ritten | -0.739 |
kefeller | -0.682 |
itness | -0.679 |
ickson | -0.667 |
Token | ", |
Token ID | 1600 |
Feature activation | -4.75e-02 |
Pos logit contributions | |
ringe | +1.000 |
ritten | +0.976 |
itness | +0.888 |
kefeller | +0.885 |
ickson | +0.866 |
Neg logit contributions | |
REDACTED | -0.853 |
Clown | -0.779 |
Capitol | -0.768 |
Certain | -0.758 |
Franch | -0.741 |
Token | S |
Token ID | 311 |
Feature activation | -5.18e-01 |
Pos logit contributions | |
REDACTED | +0.057 |
Clown | +0.051 |
Capitol | +0.050 |
Certain | +0.050 |
Franch | +0.049 |
Neg logit contributions | |
ringe | -0.086 |
ritten | -0.084 |
itness | -0.077 |
kefeller | -0.077 |
ickson | -0.075 |
Token | Mr |
Token ID | 1770 |
Feature activation | +3.70e-01 |
Pos logit contributions | |
ringe | +0.701 |
ritten | +0.684 |
itness | +0.619 |
kefeller | +0.614 |
ickson | +0.599 |
Neg logit contributions | |
REDACTED | -0.683 |
Clown | -0.618 |
Capitol | -0.615 |
Certain | -0.610 |
Franch | -0.596 |
Token | á |
Token ID | 6557 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
REDACTED | +0.474 |
Clown | +0.432 |
Capitol | +0.428 |
Certain | +0.424 |
Franch | +0.415 |
Neg logit contributions | |
ringe | -0.507 |
ritten | -0.501 |
kefeller | -0.452 |
itness | -0.452 |
ickson | -0.435 |
Token | z |
Token ID | 89 |
Feature activation | -1.12e-01 |
Pos logit contributions | |
REDACTED | +1.471 |
Capitol | +1.334 |
Clown | +1.334 |
Certain | +1.313 |
Franch | +1.288 |
Neg logit contributions | |
ringe | -1.530 |
ritten | -1.500 |
kefeller | -1.365 |
itness | -1.360 |
ickson | -1.322 |
Token | ( |
Token ID | 357 |
Feature activation | +9.39e-01 |
Pos logit contributions | |
ringe | +0.153 |
ritten | +0.148 |
itness | +0.134 |
kefeller | +0.134 |
ickson | +0.130 |
Neg logit contributions | |
REDACTED | -0.145 |
Clown | -0.133 |
Capitol | -0.131 |
Certain | -0.130 |
Franch | -0.128 |
Token | Mich |
Token ID | 11180 |
Feature activation | +5.17e+00 |
Pos logit contributions | |
REDACTED | +1.172 |
Clown | +1.041 |
Capitol | +1.032 |
Certain | +1.019 |
Franch | +0.995 |
Neg logit contributions | |
ringe | -1.360 |
ritten | -1.325 |
itness | -1.207 |
kefeller | -1.202 |
ickson | -1.173 |
Token | l |
Token ID | 75 |
Feature activation | +7.30e-02 |
Pos logit contributions | |
REDACTED | +5.674 |
Clown | +5.247 |
Capitol | +5.180 |
Certain | +5.121 |
Franch | +4.968 |
Neg logit contributions | |
ringe | -7.380 |
ritten | -7.110 |
kefeller | -6.697 |
itness | -6.667 |
ickson | -6.289 |
Token | í |
Token ID | 8836 |
Feature activation | +1.25e-01 |
Pos logit contributions | |
REDACTED | +0.088 |
Clown | +0.080 |
Capitol | +0.079 |
Certain | +0.079 |
Franch | +0.076 |
Neg logit contributions | |
ringe | -0.107 |
ritten | -0.104 |
itness | -0.095 |
kefeller | -0.095 |
ickson | -0.092 |
Token | k |
Token ID | 74 |
Feature activation | +6.22e-01 |
Pos logit contributions | |
REDACTED | +0.152 |
Clown | +0.137 |
Capitol | +0.136 |
Certain | +0.135 |
Franch | +0.132 |
Neg logit contributions | |
ringe | -0.181 |
ritten | -0.177 |
itness | -0.162 |
kefeller | -0.161 |
ickson | -0.158 |
Token | 41 |
Token ID | 6073 |
Feature activation | +5.23e-01 |
Pos logit contributions | |
REDACTED | +0.528 |
Clown | +0.461 |
Capitol | +0.453 |
Certain | +0.447 |
Franch | +0.426 |
Neg logit contributions | |
ringe | -1.129 |
ritten | -1.108 |
kefeller | -1.031 |
itness | -1.029 |
ickson | -1.008 |
Token | ), |
Token ID | 828 |
Feature activation | +1.59e-01 |
Pos logit contributions | |
REDACTED | +0.645 |
Clown | +0.584 |
Capitol | +0.578 |
Certain | +0.572 |
Franch | +0.558 |
Neg logit contributions | |
ringe | -0.756 |
ritten | -0.737 |
itness | -0.669 |
kefeller | -0.668 |
ickson | -0.654 |
Token | s |
Token ID | 82 |
Feature activation | -4.50e-01 |
Pos logit contributions | |
ringe | +0.038 |
ritten | +0.036 |
itness | +0.032 |
kefeller | +0.032 |
ickson | +0.031 |
Neg logit contributions | |
REDACTED | -0.049 |
Capitol | -0.046 |
Clown | -0.045 |
Certain | -0.045 |
Franch | -0.044 |
Token | two |
Token ID | 734 |
Feature activation | +2.03e+00 |
Pos logit contributions | |
ringe | +0.681 |
ritten | +0.664 |
itness | +0.607 |
kefeller | +0.602 |
ickson | +0.593 |
Neg logit contributions | |
REDACTED | -0.528 |
Clown | -0.474 |
Capitol | -0.472 |
Certain | -0.464 |
Franch | -0.452 |
Token | federal |
Token ID | 2717 |
Feature activation | +3.44e-01 |
Pos logit contributions | |
REDACTED | +2.237 |
Capitol | +2.138 |
Clown | +2.089 |
Certain | +2.006 |
Franch | +1.982 |
Neg logit contributions | |
ringe | -3.123 |
ritten | -3.042 |
itness | -2.801 |
kefeller | -2.759 |
ickson | -2.734 |
Token | PACs |
Token ID | 43838 |
Feature activation | +9.64e-01 |
Pos logit contributions | |
REDACTED | +0.398 |
Capitol | +0.365 |
Clown | +0.362 |
Certain | +0.354 |
Franch | +0.347 |
Neg logit contributions | |
ringe | -0.520 |
ritten | -0.509 |
itness | -0.465 |
kefeller | -0.461 |
ickson | -0.454 |
Token | — |
Token ID | 960 |
Feature activation | -6.43e-01 |
Pos logit contributions | |
REDACTED | +1.235 |
Capitol | +1.124 |
Clown | +1.123 |
Certain | +1.109 |
Franch | +1.080 |
Neg logit contributions | |
ringe | -1.348 |
ritten | -1.307 |
itness | -1.190 |
kefeller | -1.181 |
ickson | -1.161 |
Token | Liber |
Token ID | 31199 |
Feature activation | +5.11e+00 |
Pos logit contributions | |
ringe | +0.955 |
ritten | +0.930 |
itness | +0.850 |
kefeller | +0.842 |
ickson | +0.829 |
Neg logit contributions | |
REDACTED | -0.774 |
Clown | -0.692 |
Capitol | -0.686 |
Certain | -0.678 |
Franch | -0.662 |
Token | ty |
Token ID | 774 |
Feature activation | -1.27e+00 |
Pos logit contributions | |
REDACTED | +6.311 |
Capitol | +6.007 |
Clown | +6.004 |
Certain | +5.892 |
Franch | +5.674 |
Neg logit contributions | |
ringe | -6.755 |
ritten | -6.682 |
kefeller | -6.016 |
itness | -6.004 |
ickson | -5.824 |
Token | 2 |
Token ID | 362 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
ringe | +1.998 |
ritten | +1.946 |
itness | +1.789 |
kefeller | +1.772 |
ickson | +1.746 |
Neg logit contributions | |
REDACTED | -1.419 |
Clown | -1.254 |
Capitol | -1.238 |
Certain | -1.226 |
Franch | -1.195 |
Token | . |
Token ID | 13 |
Feature activation | +1.03e+00 |
Pos logit contributions | |
REDACTED | +1.495 |
Capitol | +1.390 |
Clown | +1.390 |
Certain | +1.363 |
Franch | +1.337 |
Neg logit contributions | |
ringe | -1.607 |
ritten | -1.574 |
itness | -1.423 |
kefeller | -1.400 |
ickson | -1.388 |
Token | 0 |
Token ID | 15 |
Feature activation | -9.83e-01 |
Pos logit contributions | |
REDACTED | +1.000 |
Clown | +0.878 |
Capitol | +0.875 |
Certain | +0.862 |
Franch | +0.828 |
Neg logit contributions | |
ringe | -1.753 |
ritten | -1.722 |
itness | -1.585 |
kefeller | -1.574 |
ickson | -1.550 |
Token | and |
Token ID | 290 |
Feature activation | +2.11e-01 |
Pos logit contributions | |
ringe | +1.303 |
ritten | +1.264 |
itness | +1.142 |
kefeller | +1.129 |
ickson | +1.109 |
Neg logit contributions | |
REDACTED | -1.338 |
Clown | -1.217 |
Capitol | -1.208 |
Certain | -1.196 |
Franch | -1.170 |
Token | ap |
Token ID | 2471 |
Feature activation | +2.03e+00 |
Pos logit contributions | |
REDACTED | +2.744 |
Capitol | +2.500 |
Clown | +2.479 |
Certain | +2.441 |
Franch | +2.360 |
Neg logit contributions | |
ringe | -3.319 |
ritten | -3.244 |
kefeller | -2.970 |
itness | -2.933 |
ickson | -2.840 |
Token | = |
Token ID | 796 |
Feature activation | +9.22e-02 |
Pos logit contributions | |
REDACTED | +2.538 |
Capitol | +2.277 |
Certain | +2.233 |
Clown | +2.219 |
Franch | +2.147 |
Neg logit contributions | |
ringe | -2.895 |
ritten | -2.837 |
kefeller | -2.656 |
itness | -2.571 |
ickson | -2.493 |
Token | arg |
Token ID | 1822 |
Feature activation | +3.56e+00 |
Pos logit contributions | |
REDACTED | +0.080 |
Capitol | +0.069 |
Clown | +0.069 |
Certain | +0.067 |
Franch | +0.063 |
Neg logit contributions | |
ringe | -0.168 |
ritten | -0.165 |
kefeller | -0.154 |
itness | -0.153 |
ickson | -0.149 |
Token | parse |
Token ID | 29572 |
Feature activation | +9.60e-01 |
Pos logit contributions | |
REDACTED | +2.406 |
Certain | +1.852 |
Capitol | +1.822 |
Clown | +1.745 |
Posts | +1.686 |
Neg logit contributions | |
ringe | -6.973 |
ritten | -6.759 |
kefeller | -6.407 |
itness | -6.251 |
ickson | -6.126 |
Token | . |
Token ID | 13 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
REDACTED | +1.233 |
Capitol | +1.109 |
Clown | +1.095 |
Certain | +1.095 |
Franch | +1.050 |
Neg logit contributions | |
ringe | -1.360 |
ritten | -1.322 |
kefeller | -1.226 |
itness | -1.208 |
ickson | -1.159 |
Token | Arg |
Token ID | 28100 |
Feature activation | +5.10e+00 |
Pos logit contributions | |
REDACTED | +1.281 |
Clown | +1.086 |
Capitol | +1.080 |
Certain | +1.077 |
Franch | +1.001 |
Neg logit contributions | |
ringe | -2.318 |
ritten | -2.265 |
kefeller | -2.149 |
itness | -2.105 |
ickson | -2.045 |
Token | ument |
Token ID | 1713 |
Feature activation | +1.13e+00 |
Pos logit contributions | |
REDACTED | +3.164 |
Capitol | +2.509 |
igation | +2.377 |
Certain | +2.368 |
Clown | +2.316 |
Neg logit contributions | |
ritten | -9.921 |
ringe | -9.892 |
kefeller | -9.684 |
itness | -9.192 |
lished | -9.071 |
Token | Parser |
Token ID | 46677 |
Feature activation | +3.31e+00 |
Pos logit contributions | |
REDACTED | +1.319 |
Clown | +1.155 |
Capitol | +1.153 |
Certain | +1.132 |
Franch | +1.085 |
Neg logit contributions | |
ringe | -1.733 |
ritten | -1.685 |
kefeller | -1.555 |
itness | -1.540 |
ickson | -1.491 |
Token | () |
Token ID | 3419 |
Feature activation | +5.05e-01 |
Pos logit contributions | |
REDACTED | +2.806 |
Capitol | +2.334 |
Clown | +2.271 |
Certain | +2.251 |
Assass | +2.082 |
Neg logit contributions | |
ringe | -6.129 |
ritten | -6.007 |
kefeller | -5.668 |
itness | -5.613 |
ickson | -5.420 |
Token | ap |
Token ID | 2471 |
Feature activation | +1.87e+00 |
Pos logit contributions | |
REDACTED | +0.615 |
Capitol | +0.561 |
Clown | +0.551 |
Certain | +0.548 |
Franch | +0.524 |
Neg logit contributions | |
ringe | -0.741 |
ritten | -0.722 |
kefeller | -0.668 |
itness | -0.656 |
ickson | -0.635 |
Token | . |
Token ID | 13 |
Feature activation | +1.36e+00 |
Pos logit contributions | |
REDACTED | +2.322 |
Capitol | +2.090 |
Certain | +2.044 |
Clown | +2.035 |
Franch | +1.945 |
Neg logit contributions | |
ringe | -2.714 |
ritten | -2.628 |
kefeller | -2.490 |
itness | -2.417 |
ickson | -2.323 |
Token | History |
Token ID | 7443 |
Feature activation | +7.45e-01 |
Pos logit contributions | |
REDACTED | +1.200 |
Clown | +1.133 |
Capitol | +1.102 |
Certain | +1.080 |
Assass | +0.993 |
Neg logit contributions | |
ringe | -3.005 |
ritten | -2.967 |
kefeller | -2.744 |
itness | -2.723 |
ickson | -2.692 |
Token | ( |
Token ID | 357 |
Feature activation | +1.06e+00 |
Pos logit contributions | |
REDACTED | +0.899 |
Clown | +0.824 |
Capitol | +0.821 |
Certain | +0.800 |
Franch | +0.783 |
Neg logit contributions | |
ringe | -1.088 |
ritten | -1.068 |
itness | -0.971 |
kefeller | -0.971 |
ickson | -0.944 |
Token | which |
Token ID | 4758 |
Feature activation | +1.59e+00 |
Pos logit contributions | |
REDACTED | +1.379 |
Capitol | +1.230 |
Clown | +1.227 |
Certain | +1.215 |
Franch | +1.172 |
Neg logit contributions | |
ringe | -1.461 |
ritten | -1.429 |
itness | -1.294 |
kefeller | -1.290 |
ickson | -1.257 |
Token | houses |
Token ID | 7777 |
Feature activation | +4.58e-02 |
Pos logit contributions | |
REDACTED | +1.869 |
Capitol | +1.720 |
Clown | +1.714 |
Certain | +1.656 |
Franch | +1.603 |
Neg logit contributions | |
ringe | -2.358 |
ritten | -2.307 |
kefeller | -2.102 |
itness | -2.098 |
ickson | -2.016 |
Token | the |
Token ID | 262 |
Feature activation | +1.60e-01 |
Pos logit contributions | |
REDACTED | +0.055 |
Clown | +0.050 |
Capitol | +0.050 |
Certain | +0.049 |
Franch | +0.047 |
Neg logit contributions | |
ringe | -0.068 |
ritten | -0.066 |
itness | -0.060 |
kefeller | -0.060 |
ickson | -0.058 |
Token | Prot |
Token ID | 5038 |
Feature activation | +5.06e+00 |
Pos logit contributions | |
REDACTED | +0.188 |
Capitol | +0.172 |
Clown | +0.172 |
Certain | +0.168 |
Franch | +0.163 |
Neg logit contributions | |
ringe | -0.241 |
ritten | -0.236 |
itness | -0.216 |
kefeller | -0.213 |
ickson | -0.209 |
Token | oc |
Token ID | 420 |
Feature activation | -5.78e-01 |
Pos logit contributions | |
REDACTED | +6.174 |
Clown | +5.810 |
Certain | +5.747 |
Capitol | +5.730 |
inct | +5.548 |
Neg logit contributions | |
ringe | -6.600 |
ritten | -6.408 |
kefeller | -5.917 |
itness | -5.721 |
ickson | -5.455 |
Token | er |
Token ID | 263 |
Feature activation | -2.54e-01 |
Pos logit contributions | |
ringe | +0.948 |
ritten | +0.924 |
itness | +0.855 |
kefeller | +0.844 |
ickson | +0.835 |
Neg logit contributions | |
REDACTED | -0.598 |
Clown | -0.531 |
Capitol | -0.527 |
Certain | -0.518 |
Franch | -0.500 |
Token | at |
Token ID | 265 |
Feature activation | -5.26e-01 |
Pos logit contributions | |
ringe | +0.392 |
ritten | +0.383 |
itness | +0.353 |
kefeller | +0.350 |
ickson | +0.343 |
Neg logit contributions | |
REDACTED | -0.286 |
Capitol | -0.255 |
Clown | -0.255 |
Certain | -0.252 |
Franch | -0.243 |
Token | ops |
Token ID | 2840 |
Feature activation | +1.50e+00 |
Pos logit contributions | |
ringe | +0.527 |
ritten | +0.505 |
itness | +0.440 |
kefeller | +0.440 |
ickson | +0.419 |
Neg logit contributions | |
REDACTED | -0.880 |
Clown | -0.821 |
Capitol | -0.818 |
Certain | -0.810 |
Franch | -0.791 |
Token | eggs |
Token ID | 9653 |
Feature activation | -3.11e-01 |
Pos logit contributions | |
REDACTED | +1.853 |
Clown | +1.697 |
Capitol | +1.696 |
Certain | +1.646 |
Franch | +1.614 |
Neg logit contributions | |
ringe | -2.162 |
ritten | -2.104 |
itness | -1.924 |
kefeller | -1.913 |
ickson | -1.877 |
Token | Jeff |
Token ID | 19139 |
Feature activation | -1.55e+00 |
Pos logit contributions | |
REDACTED | +1.380 |
Clown | +1.245 |
Capitol | +1.241 |
Certain | +1.212 |
Franch | +1.180 |
Neg logit contributions | |
ringe | -2.143 |
ritten | -2.078 |
kefeller | -1.917 |
itness | -1.917 |
ickson | -1.867 |
Token | Flake |
Token ID | 39727 |
Feature activation | +2.53e-01 |
Pos logit contributions | |
ringe | +2.856 |
ritten | +2.797 |
itness | +2.607 |
kefeller | +2.584 |
ickson | +2.555 |
Neg logit contributions | |
REDACTED | -1.311 |
Clown | -1.105 |
Capitol | -1.083 |
Certain | -1.068 |
Franch | -1.025 |
Token | ( |
Token ID | 7 |
Feature activation | +6.27e-01 |
Pos logit contributions | |
REDACTED | +0.319 |
Clown | +0.290 |
Capitol | +0.288 |
Certain | +0.283 |
Franch | +0.277 |
Neg logit contributions | |
ringe | -0.361 |
ritten | -0.351 |
itness | -0.320 |
kefeller | -0.318 |
ickson | -0.311 |
Token | R |
Token ID | 49 |
Feature activation | +1.55e+00 |
Pos logit contributions | |
REDACTED | +0.750 |
Clown | +0.671 |
Capitol | +0.670 |
Certain | +0.656 |
Franch | +0.637 |
Neg logit contributions | |
ringe | -0.935 |
ritten | -0.916 |
kefeller | -0.836 |
itness | -0.835 |
ickson | -0.811 |
Token | - |
Token ID | 12 |
Feature activation | +2.49e-01 |
Pos logit contributions | |
REDACTED | +2.138 |
Clown | +1.934 |
Capitol | +1.911 |
Certain | +1.900 |
Franch | +1.862 |
Neg logit contributions | |
ringe | -2.031 |
ritten | -1.969 |
itness | -1.781 |
kefeller | -1.758 |
ickson | -1.730 |
Token | Ari |
Token ID | 26529 |
Feature activation | +4.96e+00 |
Pos logit contributions | |
REDACTED | +0.306 |
Clown | +0.275 |
Capitol | +0.274 |
Certain | +0.270 |
Franch | +0.263 |
Neg logit contributions | |
ringe | -0.364 |
ritten | -0.354 |
itness | -0.322 |
kefeller | -0.322 |
ickson | -0.313 |
Token | z |
Token ID | 89 |
Feature activation | +4.92e-02 |
Pos logit contributions | |
REDACTED | +4.643 |
Capitol | +4.518 |
Clown | +4.213 |
Certain | +4.206 |
Franch | +3.960 |
Neg logit contributions | |
ringe | -8.012 |
ritten | -7.979 |
itness | -7.413 |
kefeller | -7.347 |
ickson | -7.040 |
Token | .) |
Token ID | 2014 |
Feature activation | +1.24e+00 |
Pos logit contributions | |
REDACTED | +0.066 |
Clown | +0.059 |
Capitol | +0.058 |
Certain | +0.058 |
Franch | +0.057 |
Neg logit contributions | |
ringe | -0.067 |
ritten | -0.065 |
itness | -0.059 |
kefeller | -0.058 |
ickson | -0.057 |
Token | after |
Token ID | 706 |
Feature activation | -2.69e-01 |
Pos logit contributions | |
REDACTED | +1.498 |
Capitol | +1.386 |
Clown | +1.375 |
Certain | +1.331 |
Franch | +1.305 |
Neg logit contributions | |
ringe | -1.811 |
ritten | -1.763 |
itness | -1.605 |
kefeller | -1.604 |
ickson | -1.569 |
Token | he |
Token ID | 339 |
Feature activation | +9.36e-01 |
Pos logit contributions | |
ringe | +0.403 |
ritten | +0.393 |
itness | +0.359 |
kefeller | +0.358 |
ickson | +0.351 |
Neg logit contributions | |
REDACTED | -0.320 |
Clown | -0.289 |
Capitol | -0.289 |
Certain | -0.282 |
Franch | -0.275 |
Token | was |
Token ID | 373 |
Feature activation | -6.88e-01 |
Pos logit contributions | |
REDACTED | +1.170 |
Capitol | +1.078 |
Clown | +1.078 |
Certain | +1.038 |
Franch | +1.025 |
Neg logit contributions | |
ringe | -1.335 |
ritten | -1.290 |
kefeller | -1.188 |
itness | -1.176 |
ickson | -1.152 |
Token | ing |
Token ID | 278 |
Feature activation | -7.97e-01 |
Pos logit contributions | |
ringe | +1.091 |
ritten | +1.061 |
itness | +0.964 |
kefeller | +0.960 |
ickson | +0.934 |
Neg logit contributions | |
REDACTED | -1.061 |
Clown | -0.958 |
Capitol | -0.952 |
Certain | -0.943 |
Franch | -0.917 |
Token | . |
Token ID | 13 |
Feature activation | +9.90e-01 |
Pos logit contributions | |
ringe | +1.076 |
ritten | +1.043 |
kefeller | +0.950 |
itness | +0.949 |
ickson | +0.916 |
Neg logit contributions | |
REDACTED | -1.070 |
Clown | -0.965 |
Capitol | -0.963 |
Certain | -0.955 |
Franch | -0.924 |
Token | bit |
Token ID | 2545 |
Feature activation | +1.48e+00 |
Pos logit contributions | |
REDACTED | +1.205 |
Clown | +1.072 |
Capitol | +1.061 |
Certain | +1.056 |
Franch | +1.018 |
Neg logit contributions | |
ringe | -1.443 |
ritten | -1.410 |
itness | -1.293 |
kefeller | -1.289 |
ickson | -1.262 |
Token | map |
Token ID | 8899 |
Feature activation | -5.04e-02 |
Pos logit contributions | |
REDACTED | +1.572 |
Clown | +1.408 |
Capitol | +1.363 |
Certain | +1.360 |
Franch | +1.304 |
Neg logit contributions | |
ringe | -2.352 |
ritten | -2.307 |
kefeller | -2.143 |
itness | -2.142 |
ickson | -2.087 |
Token | - |
Token ID | 532 |
Feature activation | +2.55e-01 |
Pos logit contributions | |
ringe | +0.072 |
ritten | +0.070 |
itness | +0.064 |
kefeller | +0.064 |
ickson | +0.063 |
Neg logit contributions | |
REDACTED | -0.063 |
Clown | -0.057 |
Capitol | -0.056 |
Certain | -0.056 |
Franch | -0.054 |
Token | Arg |
Token ID | 28100 |
Feature activation | +4.91e+00 |
Pos logit contributions | |
REDACTED | +0.298 |
Clown | +0.268 |
Capitol | +0.265 |
Certain | +0.263 |
Franch | +0.255 |
Neg logit contributions | |
ringe | -0.385 |
ritten | -0.377 |
itness | -0.345 |
kefeller | -0.345 |
ickson | -0.337 |
Token | ument |
Token ID | 1713 |
Feature activation | +1.27e+00 |
Pos logit contributions | |
REDACTED | +3.144 |
Clown | +2.877 |
Capitol | +2.728 |
Certain | +2.655 |
Franch | +2.528 |
Neg logit contributions | |
ritten | -9.318 |
ringe | -9.301 |
kefeller | -8.889 |
itness | -8.762 |
ickson | -8.593 |
Token | List |
Token ID | 8053 |
Feature activation | +1.85e+00 |
Pos logit contributions | |
REDACTED | +0.855 |
Clown | +0.730 |
Capitol | +0.702 |
Certain | +0.677 |
Franch | +0.645 |
Neg logit contributions | |
ringe | -2.528 |
ritten | -2.494 |
itness | -2.332 |
kefeller | -2.330 |
ickson | -2.295 |
Token | $ |
Token ID | 720 |
Feature activation | +7.78e-03 |
Pos logit contributions | |
REDACTED | +2.107 |
Clown | +1.929 |
Capitol | +1.888 |
Certain | +1.870 |
Franch | +1.815 |
Neg logit contributions | |
ringe | -2.803 |
ritten | -2.750 |
kefeller | -2.514 |
itness | -2.507 |
ickson | -2.466 |
Token | w |
Token ID | 86 |
Feature activation | -2.17e+00 |
Pos logit contributions | |
REDACTED | +0.010 |
Clown | +0.009 |
Capitol | +0.009 |
Certain | +0.009 |
Franch | +0.009 |
Neg logit contributions | |
ringe | -0.011 |
ritten | -0.011 |
itness | -0.010 |
kefeller | -0.010 |
ickson | -0.009 |
Token | path |
Token ID | 6978 |
Feature activation | +1.62e-02 |
Pos logit contributions | |
ringe | +3.528 |
ritten | +3.442 |
itness | +3.174 |
kefeller | +3.142 |
ickson | +3.097 |
Neg logit contributions | |
REDACTED | -2.305 |
Clown | -2.032 |
Capitol | -2.008 |
Certain | -1.982 |
Franch | -1.929 |
Token | of |
Token ID | 286 |
Feature activation | -2.08e-01 |
Pos logit contributions | |
ringe | +0.890 |
ritten | +0.864 |
itness | +0.792 |
kefeller | +0.782 |
ickson | +0.772 |
Neg logit contributions | |
REDACTED | -0.738 |
Clown | -0.661 |
Capitol | -0.656 |
Certain | -0.647 |
Franch | -0.635 |
Token | studies |
Token ID | 3640 |
Feature activation | +9.37e-02 |
Pos logit contributions | |
ringe | +0.297 |
ritten | +0.288 |
itness | +0.261 |
kefeller | +0.261 |
ickson | +0.254 |
Neg logit contributions | |
REDACTED | -0.262 |
Clown | -0.238 |
Capitol | -0.236 |
Certain | -0.233 |
Franch | -0.226 |
Token | are |
Token ID | 389 |
Feature activation | +1.64e+00 |
Pos logit contributions | |
REDACTED | +0.117 |
Clown | +0.105 |
Capitol | +0.105 |
Certain | +0.104 |
Franch | +0.100 |
Neg logit contributions | |
ringe | -0.134 |
ritten | -0.132 |
kefeller | -0.121 |
itness | -0.118 |
ickson | -0.115 |
Token | found |
Token ID | 1043 |
Feature activation | +1.12e+00 |
Pos logit contributions | |
REDACTED | +1.959 |
Capitol | +1.765 |
Clown | +1.760 |
Certain | +1.714 |
Franch | +1.630 |
Neg logit contributions | |
ringe | -2.462 |
ritten | -2.343 |
kefeller | -2.199 |
itness | -2.139 |
ickson | -2.085 |
Token | ( |
Token ID | 357 |
Feature activation | +8.16e-01 |
Pos logit contributions | |
REDACTED | +1.322 |
Clown | +1.185 |
Capitol | +1.170 |
Certain | +1.163 |
Franch | +1.115 |
Neg logit contributions | |
ringe | -1.667 |
ritten | -1.630 |
kefeller | -1.512 |
itness | -1.472 |
ickson | -1.442 |
Token | gener |
Token ID | 8612 |
Feature activation | +4.89e+00 |
Pos logit contributions | |
REDACTED | +0.956 |
Clown | +0.839 |
Capitol | +0.833 |
Certain | +0.825 |
Franch | +0.794 |
Neg logit contributions | |
ringe | -1.243 |
ritten | -1.213 |
kefeller | -1.117 |
itness | -1.109 |
ickson | -1.076 |
Token | ally |
Token ID | 453 |
Feature activation | +1.13e+00 |
Pos logit contributions | |
REDACTED | +6.287 |
Clown | +5.740 |
Capitol | +5.729 |
Certain | +5.637 |
atorium | +5.465 |
Neg logit contributions | |
ringe | -6.347 |
ritten | -6.233 |
kefeller | -5.641 |
itness | -5.617 |
ickson | -5.438 |
Token | > |
Token ID | 1875 |
Feature activation | +1.64e-01 |
Pos logit contributions | |
REDACTED | +1.357 |
Clown | +1.221 |
Capitol | +1.218 |
Certain | +1.186 |
Franch | +1.146 |
Neg logit contributions | |
ringe | -1.681 |
ritten | -1.637 |
kefeller | -1.523 |
itness | -1.501 |
ickson | -1.452 |
Token | 10 |
Token ID | 940 |
Feature activation | +4.69e-01 |
Pos logit contributions | |
REDACTED | +0.165 |
Clown | +0.143 |
Capitol | +0.141 |
Certain | +0.141 |
Franch | +0.135 |
Neg logit contributions | |
ringe | -0.276 |
ritten | -0.270 |
itness | -0.250 |
kefeller | -0.249 |
ickson | -0.244 |
Token | ). |
Token ID | 737 |
Feature activation | -2.44e-01 |
Pos logit contributions | |
REDACTED | +0.578 |
Clown | +0.523 |
Capitol | +0.513 |
Certain | +0.511 |
Franch | +0.490 |
Neg logit contributions | |
ringe | -0.678 |
ritten | -0.660 |
kefeller | -0.605 |
itness | -0.594 |
ickson | -0.580 |
Token | If |
Token ID | 1002 |
Feature activation | +9.21e-02 |
Pos logit contributions | |
ringe | +0.400 |
ritten | +0.391 |
itness | +0.359 |
kefeller | +0.359 |
ickson | +0.351 |
Neg logit contributions | |
REDACTED | -0.258 |
Clown | -0.228 |
Capitol | -0.226 |
Certain | -0.225 |
Franch | -0.217 |
Token | ap |
Token ID | 2471 |
Feature activation | +1.69e+00 |
Pos logit contributions | |
REDACTED | +2.240 |
Clown | +2.048 |
Capitol | +2.034 |
Certain | +2.010 |
Franch | +1.907 |
Neg logit contributions | |
ringe | -2.877 |
ritten | -2.808 |
kefeller | -2.603 |
itness | -2.540 |
ickson | -2.477 |
Token | = |
Token ID | 796 |
Feature activation | -3.55e-02 |
Pos logit contributions | |
REDACTED | +2.180 |
Clown | +1.976 |
Capitol | +1.972 |
Certain | +1.952 |
Franch | +1.883 |
Neg logit contributions | |
ringe | -2.279 |
ritten | -2.208 |
kefeller | -2.049 |
itness | -1.993 |
ickson | -1.937 |
Token | arg |
Token ID | 1822 |
Feature activation | +3.40e+00 |
Pos logit contributions | |
ringe | +0.056 |
ritten | +0.055 |
kefeller | +0.050 |
itness | +0.050 |
ickson | +0.049 |
Neg logit contributions | |
REDACTED | -0.039 |
Clown | -0.035 |
Capitol | -0.035 |
Certain | -0.034 |
Franch | -0.033 |
Token | parse |
Token ID | 29572 |
Feature activation | +6.16e-01 |
Pos logit contributions | |
REDACTED | +2.398 |
Certain | +1.939 |
Capitol | +1.931 |
Clown | +1.905 |
inct | +1.741 |
Neg logit contributions | |
ringe | -6.507 |
ritten | -6.420 |
kefeller | -6.024 |
itness | -5.876 |
ickson | -5.786 |
Token | . |
Token ID | 13 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
REDACTED | +0.773 |
Clown | +0.695 |
Capitol | +0.692 |
Certain | +0.686 |
Franch | +0.663 |
Neg logit contributions | |
ringe | -0.880 |
ritten | -0.857 |
kefeller | -0.782 |
itness | -0.777 |
ickson | -0.758 |
Token | Arg |
Token ID | 28100 |
Feature activation | +4.82e+00 |
Pos logit contributions | |
REDACTED | +1.084 |
Clown | +0.957 |
Capitol | +0.952 |
Certain | +0.938 |
Franch | +0.894 |
Neg logit contributions | |
ringe | -1.800 |
ritten | -1.766 |
kefeller | -1.633 |
itness | -1.622 |
ickson | -1.590 |
Token | ument |
Token ID | 1713 |
Feature activation | +8.87e-01 |
Pos logit contributions | |
REDACTED | +3.155 |
Clown | +2.715 |
Capitol | +2.636 |
Certain | +2.506 |
Franch | +2.314 |
Neg logit contributions | |
ritten | -9.318 |
ringe | -9.131 |
kefeller | -8.787 |
itness | -8.632 |
ickson | -8.251 |
Token | Parser |
Token ID | 46677 |
Feature activation | +3.12e+00 |
Pos logit contributions | |
REDACTED | +1.043 |
Clown | +0.937 |
Capitol | +0.929 |
Certain | +0.910 |
Franch | +0.885 |
Neg logit contributions | |
ringe | -1.332 |
ritten | -1.308 |
kefeller | -1.192 |
itness | -1.192 |
ickson | -1.159 |
Token | () |
Token ID | 3419 |
Feature activation | +3.27e-01 |
Pos logit contributions | |
REDACTED | +3.628 |
Capitol | +3.260 |
Clown | +3.225 |
Certain | +3.181 |
Franch | +3.047 |
Neg logit contributions | |
ringe | -4.663 |
ritten | -4.575 |
itness | -4.177 |
kefeller | -4.176 |
ickson | -4.027 |
Token | ap |
Token ID | 2471 |
Feature activation | +1.50e+00 |
Pos logit contributions | |
REDACTED | +0.403 |
Capitol | +0.365 |
Clown | +0.365 |
Certain | +0.360 |
Franch | +0.348 |
Neg logit contributions | |
ringe | -0.475 |
ritten | -0.462 |
kefeller | -0.420 |
itness | -0.417 |
ickson | -0.407 |
Token | . |
Token ID | 13 |
Feature activation | +1.13e+00 |
Pos logit contributions | |
REDACTED | +1.938 |
Capitol | +1.756 |
Clown | +1.730 |
Certain | +1.724 |
Franch | +1.662 |
Neg logit contributions | |
ringe | -2.064 |
ritten | -2.002 |
kefeller | -1.836 |
itness | -1.808 |
ickson | -1.763 |
Token | T |
Token ID | 309 |
Feature activation | -5.94e-01 |
Pos logit contributions | |
REDACTED | +0.247 |
Clown | +0.218 |
Capitol | +0.215 |
Certain | +0.214 |
Franch | +0.204 |
Neg logit contributions | |
ringe | -0.418 |
ritten | -0.408 |
kefeller | -0.378 |
itness | -0.377 |
ickson | -0.370 |
Token | > |
Token ID | 1875 |
Feature activation | -8.52e-02 |
Pos logit contributions | |
ringe | +0.844 |
ritten | +0.817 |
itness | +0.747 |
kefeller | +0.729 |
ickson | +0.722 |
Neg logit contributions | |
REDACTED | -0.748 |
Clown | -0.676 |
Capitol | -0.668 |
Certain | -0.661 |
Franch | -0.650 |
Token | D |
Token ID | 360 |
Feature activation | -2.93e-01 |
Pos logit contributions | |
ringe | +0.135 |
ritten | +0.131 |
itness | +0.121 |
kefeller | +0.120 |
ickson | +0.118 |
Neg logit contributions | |
REDACTED | -0.094 |
Clown | -0.085 |
Capitol | -0.084 |
Certain | -0.083 |
Franch | -0.080 |
Token | ere |
Token ID | 567 |
Feature activation | +3.26e-01 |
Pos logit contributions | |
ringe | +0.482 |
ritten | +0.472 |
kefeller | +0.437 |
itness | +0.436 |
ickson | +0.424 |
Neg logit contributions | |
REDACTED | -0.300 |
Clown | -0.261 |
Certain | -0.259 |
Capitol | -0.259 |
Franch | -0.248 |
Token | f |
Token ID | 69 |
Feature activation | -2.00e+00 |
Pos logit contributions | |
REDACTED | +0.418 |
Capitol | +0.378 |
Certain | +0.377 |
Clown | +0.376 |
Franch | +0.364 |
Neg logit contributions | |
ringe | -0.446 |
ritten | -0.434 |
kefeller | -0.400 |
itness | -0.397 |
ickson | -0.386 |
Token | Mut |
Token ID | 41603 |
Feature activation | +4.80e+00 |
Pos logit contributions | |
ringe | +2.795 |
ritten | +2.708 |
itness | +2.469 |
kefeller | +2.428 |
ickson | +2.390 |
Neg logit contributions | |
REDACTED | -2.559 |
Clown | -2.308 |
Capitol | -2.277 |
Certain | -2.260 |
Franch | -2.211 |
Token | for |
Token ID | 329 |
Feature activation | +8.59e-01 |
Pos logit contributions | |
REDACTED | +5.684 |
Capitol | +5.169 |
Clown | +5.151 |
Certain | +5.082 |
Franch | +4.915 |
Neg logit contributions | |
ringe | -6.803 |
ritten | -6.686 |
kefeller | -6.104 |
itness | -6.073 |
ickson | -5.913 |
Token | Own |
Token ID | 11744 |
Feature activation | -2.58e-01 |
Pos logit contributions | |
REDACTED | +0.603 |
Clown | +0.498 |
Capitol | +0.488 |
Certain | +0.481 |
Franch | +0.458 |
Neg logit contributions | |
ringe | -1.703 |
ritten | -1.667 |
itness | -1.561 |
kefeller | -1.551 |
ickson | -1.533 |
Token | ed |
Token ID | 276 |
Feature activation | -5.94e-01 |
Pos logit contributions | |
ringe | +0.436 |
ritten | +0.421 |
itness | +0.394 |
kefeller | +0.392 |
ickson | +0.384 |
Neg logit contributions | |
REDACTED | -0.257 |
Capitol | -0.223 |
Clown | -0.221 |
Certain | -0.219 |
Franch | -0.209 |
Token | < |
Token ID | 1279 |
Feature activation | +3.10e-01 |
Pos logit contributions | |
ringe | +0.931 |
ritten | +0.908 |
itness | +0.833 |
kefeller | +0.832 |
ickson | +0.813 |
Neg logit contributions | |
REDACTED | -0.668 |
Clown | -0.591 |
Capitol | -0.589 |
Certain | -0.583 |
Franch | -0.564 |
Token | T |
Token ID | 309 |
Feature activation | -4.83e-01 |
Pos logit contributions | |
REDACTED | +0.297 |
Clown | +0.261 |
Capitol | +0.259 |
Certain | +0.258 |
Franch | +0.244 |
Neg logit contributions | |
ringe | -0.539 |
ritten | -0.523 |
kefeller | -0.489 |
itness | -0.486 |
ickson | -0.478 |
Token | ( |
Token ID | 357 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
REDACTED | +2.060 |
Clown | +1.868 |
Certain | +1.867 |
Capitol | +1.859 |
Franch | +1.801 |
Neg logit contributions | |
ringe | -2.196 |
ritten | -2.128 |
kefeller | -1.959 |
itness | -1.914 |
ickson | -1.886 |
Token | B |
Token ID | 33 |
Feature activation | +1.41e+00 |
Pos logit contributions | |
REDACTED | +1.310 |
Clown | +1.134 |
Capitol | +1.128 |
Certain | +1.123 |
Franch | +1.086 |
Neg logit contributions | |
ringe | -1.768 |
ritten | -1.730 |
kefeller | -1.587 |
itness | -1.583 |
ickson | -1.546 |
Token | )( |
Token ID | 5769 |
Feature activation | +8.55e-01 |
Pos logit contributions | |
REDACTED | +1.902 |
Capitol | +1.712 |
Clown | +1.694 |
Certain | +1.692 |
Franch | +1.655 |
Neg logit contributions | |
ringe | -1.861 |
ritten | -1.764 |
kefeller | -1.651 |
itness | -1.606 |
ickson | -1.562 |
Token | 6 |
Token ID | 21 |
Feature activation | +7.09e-01 |
Pos logit contributions | |
REDACTED | +0.914 |
Clown | +0.790 |
Capitol | +0.786 |
Certain | +0.780 |
Franch | +0.756 |
Neg logit contributions | |
ringe | -1.383 |
ritten | -1.358 |
kefeller | -1.249 |
itness | -1.248 |
ickson | -1.222 |
Token | ) |
Token ID | 8 |
Feature activation | +8.53e-01 |
Pos logit contributions | |
REDACTED | +0.853 |
Clown | +0.759 |
Capitol | +0.755 |
Certain | +0.748 |
Franch | +0.727 |
Neg logit contributions | |
ringe | -1.051 |
ritten | -1.026 |
kefeller | -0.935 |
itness | -0.933 |
ickson | -0.913 |
Token | Motion |
Token ID | 20843 |
Feature activation | +4.72e+00 |
Pos logit contributions | |
REDACTED | +0.854 |
Clown | +0.747 |
Capitol | +0.739 |
Certain | +0.737 |
Franch | +0.711 |
Neg logit contributions | |
ringe | -1.437 |
ritten | -1.408 |
kefeller | -1.298 |
itness | -1.296 |
ickson | -1.271 |
Token | to |
Token ID | 284 |
Feature activation | +1.07e+00 |
Pos logit contributions | |
REDACTED | +5.379 |
Certain | +4.853 |
Capitol | +4.831 |
Clown | +4.822 |
Franch | +4.604 |
Neg logit contributions | |
ringe | -7.030 |
ritten | -6.936 |
kefeller | -6.424 |
itness | -6.315 |
ickson | -6.195 |
Token | Dism |
Token ID | 44955 |
Feature activation | +3.28e+00 |
Pos logit contributions | |
REDACTED | +0.821 |
Clown | +0.700 |
Capitol | +0.690 |
Certain | +0.689 |
Franch | +0.651 |
Neg logit contributions | |
ringe | -2.062 |
ritten | -2.020 |
kefeller | -1.886 |
itness | -1.879 |
ickson | -1.849 |
Token | iss |
Token ID | 747 |
Feature activation | +2.48e+00 |
Pos logit contributions | |
REDACTED | +2.559 |
Clown | +2.197 |
Capitol | +2.183 |
Certain | +2.163 |
Franch | +2.016 |
Neg logit contributions | |
ringe | -6.010 |
ritten | -5.959 |
kefeller | -5.596 |
itness | -5.428 |
ickson | -5.411 |
Token |
|
Token ID | 198 |
Feature activation | -2.95e-01 |
Pos logit contributions | |
REDACTED | +2.850 |
Clown | +2.581 |
Certain | +2.580 |
Capitol | +2.550 |
Franch | +2.481 |
Neg logit contributions | |
ringe | -3.762 |
ritten | -3.682 |
kefeller | -3.401 |
itness | -3.372 |
ickson | -3.268 |
Token |
|
Token ID | 198 |
Feature activation | -1.65e-01 |
Pos logit contributions | |
ringe | +0.469 |
ritten | +0.457 |
kefeller | +0.420 |
itness | +0.420 |
ickson | +0.410 |
Neg logit contributions | |
REDACTED | -0.326 |
Clown | -0.288 |
Capitol | -0.287 |
Certain | -0.283 |
Franch | -0.274 |
Token | to |
Token ID | 284 |
Feature activation | +1.60e-01 |
Pos logit contributions | |
ringe | +0.730 |
ritten | +0.709 |
itness | +0.647 |
kefeller | +0.635 |
ickson | +0.626 |
Neg logit contributions | |
REDACTED | -0.676 |
Clown | -0.612 |
Capitol | -0.603 |
Certain | -0.598 |
Franch | -0.587 |
Token | the |
Token ID | 262 |
Feature activation | -1.47e-01 |
Pos logit contributions | |
REDACTED | +0.190 |
Clown | +0.171 |
Capitol | +0.169 |
Certain | +0.168 |
Franch | +0.163 |
Neg logit contributions | |
ringe | -0.240 |
ritten | -0.234 |
itness | -0.214 |
kefeller | -0.212 |
ickson | -0.208 |
Token | Catalan |
Token ID | 31066 |
Feature activation | +3.61e+00 |
Pos logit contributions | |
ringe | +0.228 |
ritten | +0.222 |
itness | +0.203 |
kefeller | +0.203 |
ickson | +0.199 |
Neg logit contributions | |
REDACTED | -0.168 |
Clown | -0.150 |
Capitol | -0.149 |
Certain | -0.147 |
Franch | -0.143 |
Token | Government |
Token ID | 5070 |
Feature activation | +5.44e-02 |
Pos logit contributions | |
REDACTED | +3.976 |
Clown | +3.647 |
Capitol | +3.619 |
Certain | +3.574 |
Assass | +3.445 |
Neg logit contributions | |
ringe | -5.486 |
ritten | -5.336 |
kefeller | -4.938 |
itness | -4.885 |
ickson | -4.742 |
Token | ( |
Token ID | 357 |
Feature activation | +6.91e-01 |
Pos logit contributions | |
REDACTED | +0.076 |
Clown | +0.069 |
Capitol | +0.069 |
Certain | +0.068 |
Franch | +0.066 |
Neg logit contributions | |
ringe | -0.071 |
ritten | -0.068 |
itness | -0.062 |
kefeller | -0.061 |
ickson | -0.060 |
Token | all |
Token ID | 439 |
Feature activation | +4.71e+00 |
Pos logit contributions | |
REDACTED | +0.825 |
Clown | +0.728 |
Capitol | +0.725 |
Certain | +0.719 |
Franch | +0.694 |
Neg logit contributions | |
ringe | -1.032 |
ritten | -1.004 |
kefeller | -0.922 |
itness | -0.918 |
ickson | -0.896 |
Token | pro |
Token ID | 386 |
Feature activation | +1.95e+00 |
Pos logit contributions | |
REDACTED | +5.164 |
Capitol | +4.710 |
Clown | +4.695 |
Certain | +4.660 |
Franch | +4.482 |
Neg logit contributions | |
ringe | -6.908 |
ritten | -6.745 |
kefeller | -6.298 |
itness | -6.136 |
ickson | -5.999 |
Token | - |
Token ID | 12 |
Feature activation | +6.16e-01 |
Pos logit contributions | |
REDACTED | +2.604 |
Clown | +2.364 |
Capitol | +2.347 |
Certain | +2.323 |
Franch | +2.273 |
Neg logit contributions | |
ringe | -2.614 |
ritten | -2.534 |
itness | -2.294 |
kefeller | -2.273 |
ickson | -2.232 |
Token | independence |
Token ID | 39894 |
Feature activation | +1.95e+00 |
Pos logit contributions | |
REDACTED | +0.660 |
Clown | +0.573 |
Capitol | +0.571 |
Certain | +0.567 |
Franch | +0.547 |
Neg logit contributions | |
ringe | -0.995 |
ritten | -0.970 |
kefeller | -0.897 |
itness | -0.895 |
ickson | -0.876 |
Token | parties |
Token ID | 4671 |
Feature activation | +2.14e-01 |
Pos logit contributions | |
REDACTED | +2.129 |
Clown | +1.935 |
Capitol | +1.928 |
Certain | +1.886 |
Franch | +1.853 |
Neg logit contributions | |
ringe | -3.043 |
ritten | -2.966 |
kefeller | -2.766 |
itness | -2.733 |
ickson | -2.677 |
Token | form |
Token ID | 1296 |
Feature activation | +1.12e+00 |
Pos logit contributions | |
REDACTED | +0.265 |
Clown | +0.240 |
Capitol | +0.239 |
Certain | +0.235 |
Franch | +0.231 |
Neg logit contributions | |
ringe | -0.307 |
ritten | -0.299 |
kefeller | -0.272 |
itness | -0.271 |
ickson | -0.265 |
Token | the |
Token ID | 262 |
Feature activation | -2.90e-01 |
Pos logit contributions | |
REDACTED | +0.601 |
Capitol | +0.542 |
Clown | +0.540 |
Certain | +0.526 |
Franch | +0.511 |
Neg logit contributions | |
ringe | -0.744 |
ritten | -0.727 |
kefeller | -0.660 |
itness | -0.659 |
ickson | -0.644 |
Token | allegations |
Token ID | 7636 |
Feature activation | +7.27e-01 |
Pos logit contributions | |
ringe | +0.458 |
ritten | +0.445 |
kefeller | +0.409 |
itness | +0.408 |
ickson | +0.399 |
Neg logit contributions | |
REDACTED | -0.326 |
Capitol | -0.291 |
Clown | -0.290 |
Certain | -0.283 |
Franch | -0.273 |
Token | . |
Token ID | 13 |
Feature activation | -1.51e-01 |
Pos logit contributions | |
REDACTED | +0.971 |
Clown | +0.881 |
Capitol | +0.875 |
Certain | +0.865 |
Franch | +0.845 |
Neg logit contributions | |
ringe | -0.981 |
ritten | -0.951 |
itness | -0.861 |
kefeller | -0.855 |
ickson | -0.837 |
Token |
|
Token ID | 198 |
Feature activation | -2.37e-01 |
Pos logit contributions | |
ringe | +0.240 |
ritten | +0.234 |
itness | +0.214 |
kefeller | +0.214 |
ickson | +0.210 |
Neg logit contributions | |
REDACTED | -0.167 |
Clown | -0.149 |
Capitol | -0.147 |
Certain | -0.146 |
Franch | -0.142 |
Token |
|
Token ID | 198 |
Feature activation | -2.57e-01 |
Pos logit contributions | |
ringe | +0.380 |
ritten | +0.370 |
kefeller | +0.342 |
itness | +0.341 |
ickson | +0.332 |
Neg logit contributions | |
REDACTED | -0.259 |
Clown | -0.228 |
Capitol | -0.228 |
Certain | -0.225 |
Franch | -0.217 |
Token | Min |
Token ID | 9452 |
Feature activation | +4.70e+00 |
Pos logit contributions | |
ringe | +0.394 |
ritten | +0.384 |
kefeller | +0.354 |
itness | +0.351 |
ickson | +0.344 |
Neg logit contributions | |
REDACTED | -0.299 |
Clown | -0.263 |
Capitol | -0.261 |
Certain | -0.259 |
Franch | -0.250 |
Token | ister |
Token ID | 1694 |
Feature activation | -2.74e-01 |
Pos logit contributions | |
REDACTED | +3.820 |
Clown | +3.492 |
Capitol | +3.475 |
Certain | +3.340 |
atorium | +3.171 |
Neg logit contributions | |
ringe | -8.187 |
ritten | -8.124 |
kefeller | -7.497 |
itness | -7.370 |
ickson | -7.213 |
Token | Sp |
Token ID | 1338 |
Feature activation | -2.16e+00 |
Pos logit contributions | |
ringe | +0.394 |
ritten | +0.384 |
itness | +0.350 |
kefeller | +0.347 |
ickson | +0.339 |
Neg logit contributions | |
REDACTED | -0.342 |
Clown | -0.309 |
Capitol | -0.307 |
Certain | -0.303 |
Franch | -0.297 |
Token | as |
Token ID | 292 |
Feature activation | -1.32e+00 |
Pos logit contributions | |
ringe | +3.774 |
ritten | +3.684 |
itness | +3.413 |
kefeller | +3.391 |
ickson | +3.344 |
Neg logit contributions | |
REDACTED | -2.030 |
Clown | -1.750 |
Capitol | -1.741 |
Certain | -1.716 |
Franch | -1.655 |
Token | ov |
Token ID | 709 |
Feature activation | -2.20e+00 |
Pos logit contributions | |
ringe | +2.147 |
ritten | +2.084 |
itness | +1.924 |
kefeller | +1.913 |
ickson | +1.874 |
Neg logit contributions | |
REDACTED | -1.401 |
Clown | -1.235 |
Capitol | -1.235 |
Certain | -1.209 |
Franch | -1.179 |
Token | ski |
Token ID | 20545 |
Feature activation | -1.97e+00 |
Pos logit contributions | |
ringe | +2.538 |
ritten | +2.450 |
itness | +2.177 |
kefeller | +2.160 |
ickson | +2.101 |
Neg logit contributions | |
REDACTED | -3.368 |
Clown | -3.086 |
Capitol | -3.071 |
Certain | -3.040 |
Franch | -2.983 |
Token | E |
Token ID | 36 |
Feature activation | +6.29e-01 |
Pos logit contributions | |
ringe | +0.259 |
ritten | +0.250 |
kefeller | +0.227 |
itness | +0.227 |
ickson | +0.220 |
Neg logit contributions | |
REDACTED | -0.248 |
Clown | -0.222 |
Capitol | -0.222 |
Certain | -0.220 |
Franch | -0.214 |
Token | t |
Token ID | 83 |
Feature activation | -4.33e-01 |
Pos logit contributions | |
REDACTED | +0.986 |
Certain | +0.910 |
Capitol | +0.908 |
Clown | +0.904 |
Franch | +0.881 |
Neg logit contributions | |
ringe | -0.685 |
ritten | -0.669 |
kefeller | -0.598 |
itness | -0.591 |
ickson | -0.565 |
Token | at |
Token ID | 265 |
Feature activation | -9.86e-01 |
Pos logit contributions | |
ringe | +0.694 |
ritten | +0.677 |
itness | +0.623 |
kefeller | +0.618 |
ickson | +0.609 |
Neg logit contributions | |
REDACTED | -0.468 |
Clown | -0.414 |
Capitol | -0.410 |
Certain | -0.405 |
Franch | -0.393 |
Token | , |
Token ID | 11 |
Feature activation | +9.93e-01 |
Pos logit contributions | |
ringe | +1.461 |
ritten | +1.421 |
itness | +1.299 |
kefeller | +1.287 |
ickson | +1.266 |
Neg logit contributions | |
REDACTED | -1.189 |
Clown | -1.064 |
Capitol | -1.053 |
Certain | -1.042 |
Franch | -1.018 |
Token | les |
Token ID | 10287 |
Feature activation | +3.84e-01 |
Pos logit contributions | |
REDACTED | +1.173 |
Clown | +1.043 |
Capitol | +1.036 |
Certain | +1.031 |
Franch | +0.999 |
Neg logit contributions | |
ringe | -1.500 |
ritten | -1.453 |
kefeller | -1.340 |
itness | -1.337 |
ickson | -1.307 |
Token | mut |
Token ID | 4517 |
Feature activation | +4.69e+00 |
Pos logit contributions | |
REDACTED | +0.462 |
Clown | +0.415 |
Capitol | +0.413 |
Certain | +0.409 |
Franch | +0.398 |
Neg logit contributions | |
ringe | -0.571 |
ritten | -0.552 |
itness | -0.505 |
kefeller | -0.504 |
ickson | -0.494 |
Token | ins |
Token ID | 1040 |
Feature activation | -1.11e+00 |
Pos logit contributions | |
REDACTED | +4.817 |
Certain | +4.301 |
Capitol | +4.295 |
Clown | +4.216 |
atorium | +4.120 |
Neg logit contributions | |
ringe | -7.265 |
ritten | -6.968 |
kefeller | -6.757 |
itness | -6.478 |
ickson | -6.331 |
Token | ont |
Token ID | 39585 |
Feature activation | +6.92e-01 |
Pos logit contributions | |
ringe | +1.641 |
ritten | +1.596 |
itness | +1.459 |
kefeller | +1.446 |
ickson | +1.422 |
Neg logit contributions | |
REDACTED | -1.333 |
Clown | -1.193 |
Capitol | -1.181 |
Certain | -1.169 |
Franch | -1.141 |
Token | f |
Token ID | 277 |
Feature activation | -7.93e-01 |
Pos logit contributions | |
REDACTED | +0.829 |
Clown | +0.744 |
Capitol | +0.740 |
Certain | +0.731 |
Franch | +0.713 |
Neg logit contributions | |
ringe | -1.029 |
ritten | -1.000 |
itness | -0.915 |
kefeller | -0.910 |
ickson | -0.892 |
Token | uit |
Token ID | 5013 |
Feature activation | +1.20e+00 |
Pos logit contributions | |
ringe | +1.228 |
ritten | +1.196 |
itness | +1.101 |
kefeller | +1.094 |
ickson | +1.072 |
Neg logit contributions | |
REDACTED | -0.895 |
Clown | -0.796 |
Capitol | -0.790 |
Certain | -0.783 |
Franch | -0.760 |
Token | Mil |
Token ID | 4460 |
Feature activation | +1.69e+00 |
Pos logit contributions | |
REDACTED | +1.437 |
Clown | +1.289 |
Capitol | +1.284 |
Certain | +1.276 |
Franch | +1.242 |
Neg logit contributions | |
ringe | -1.780 |
ritten | -1.725 |
kefeller | -1.585 |
itness | -1.585 |
ickson | -1.543 |
Token | gy |
Token ID | 1360 |
Feature activation | +8.40e-01 |
Pos logit contributions | |
ringe | +0.218 |
ritten | +0.210 |
kefeller | +0.188 |
itness | +0.183 |
ickson | +0.180 |
Neg logit contributions | |
REDACTED | -0.322 |
Clown | -0.297 |
Capitol | -0.297 |
Certain | -0.296 |
Franch | -0.288 |
Token | for |
Token ID | 329 |
Feature activation | +1.02e+00 |
Pos logit contributions | |
REDACTED | +1.011 |
Clown | +0.924 |
Capitol | +0.913 |
Certain | +0.898 |
Franch | +0.872 |
Neg logit contributions | |
ringe | -1.226 |
ritten | -1.190 |
kefeller | -1.101 |
itness | -1.086 |
ickson | -1.065 |
Token | herself |
Token ID | 5223 |
Feature activation | +3.59e-01 |
Pos logit contributions | |
REDACTED | +1.204 |
Clown | +1.104 |
Capitol | +1.094 |
Certain | +1.075 |
Franch | +1.047 |
Neg logit contributions | |
ringe | -1.522 |
ritten | -1.475 |
kefeller | -1.343 |
itness | -1.343 |
ickson | -1.308 |
Token | . |
Token ID | 13 |
Feature activation | +5.15e-01 |
Pos logit contributions | |
REDACTED | +0.469 |
Clown | +0.427 |
Capitol | +0.422 |
Certain | +0.418 |
Franch | +0.405 |
Neg logit contributions | |
ringe | -0.495 |
ritten | -0.480 |
kefeller | -0.439 |
itness | -0.435 |
ickson | -0.424 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -5.02e-01 |
Pos logit contributions | |
REDACTED | +0.535 |
Clown | +0.481 |
Certain | +0.478 |
Capitol | +0.476 |
Franch | +0.454 |
Neg logit contributions | |
ringe | -0.849 |
ritten | -0.830 |
kefeller | -0.769 |
itness | -0.762 |
ickson | -0.748 |
Token | Posted |
Token ID | 14231 |
Feature activation | +4.67e+00 |
Pos logit contributions | |
ringe | +0.768 |
ritten | +0.748 |
itness | +0.686 |
kefeller | +0.682 |
ickson | +0.669 |
Neg logit contributions | |
REDACTED | -0.583 |
Clown | -0.516 |
Capitol | -0.512 |
Certain | -0.507 |
Franch | -0.491 |
Token | on |
Token ID | 319 |
Feature activation | +1.42e+00 |
Pos logit contributions | |
REDACTED | +4.550 |
Clown | +4.200 |
Capitol | +4.198 |
Certain | +4.160 |
Assass | +3.860 |
Neg logit contributions | |
ringe | -7.560 |
ritten | -7.319 |
kefeller | -6.975 |
itness | -6.701 |
ickson | -6.617 |
Token | // |
Token ID | 3373 |
Feature activation | +7.32e-01 |
Pos logit contributions | |
REDACTED | +1.539 |
Capitol | +1.400 |
Clown | +1.399 |
Certain | +1.373 |
Franch | +1.321 |
Neg logit contributions | |
ringe | -2.275 |
ritten | -2.216 |
kefeller | -2.041 |
itness | -2.032 |
ickson | -1.985 |
Token |
|
Token ID | 198 |
Feature activation | -3.41e-01 |
Pos logit contributions | |
REDACTED | +0.811 |
Capitol | +0.727 |
Clown | +0.726 |
Certain | +0.712 |
Franch | +0.686 |
Neg logit contributions | |
ringe | -1.145 |
ritten | -1.123 |
kefeller | -1.035 |
itness | -1.024 |
ickson | -1.004 |
Token |
|
Token ID | 198 |
Feature activation | -3.04e-01 |
Pos logit contributions | |
ringe | +0.547 |
ritten | +0.534 |
kefeller | +0.493 |
itness | +0.490 |
ickson | +0.479 |
Neg logit contributions | |
REDACTED | -0.374 |
Capitol | -0.329 |
Clown | -0.328 |
Certain | -0.325 |
Franch | -0.312 |
Token | According |
Token ID | 4821 |
Feature activation | -9.04e-01 |
Pos logit contributions | |
ringe | +0.460 |
ritten | +0.447 |
itness | +0.410 |
kefeller | +0.409 |
ickson | +0.400 |
Neg logit contributions | |
REDACTED | -0.359 |
Clown | -0.317 |
Capitol | -0.316 |
Certain | -0.313 |
Franch | -0.303 |
Token | he |
Token ID | 339 |
Feature activation | +7.26e-01 |
Pos logit contributions | |
ringe | +1.172 |
ritten | +1.142 |
itness | +1.044 |
kefeller | +1.035 |
ickson | +1.017 |
Neg logit contributions | |
REDACTED | -0.988 |
Clown | -0.893 |
Capitol | -0.883 |
Certain | -0.872 |
Franch | -0.849 |
Token | and |
Token ID | 290 |
Feature activation | -2.53e-01 |
Pos logit contributions | |
REDACTED | +0.828 |
Clown | +0.740 |
Capitol | +0.731 |
Certain | +0.719 |
Franch | +0.701 |
Neg logit contributions | |
ringe | -1.117 |
ritten | -1.092 |
kefeller | -1.004 |
itness | -1.002 |
ickson | -0.980 |
Token | the |
Token ID | 262 |
Feature activation | -3.93e-01 |
Pos logit contributions | |
ringe | +0.391 |
ritten | +0.381 |
itness | +0.349 |
kefeller | +0.345 |
ickson | +0.341 |
Neg logit contributions | |
REDACTED | -0.292 |
Clown | -0.261 |
Capitol | -0.258 |
Certain | -0.255 |
Franch | -0.248 |
Token | turtles |
Token ID | 36288 |
Feature activation | -6.83e-02 |
Pos logit contributions | |
ringe | +0.570 |
ritten | +0.557 |
itness | +0.506 |
kefeller | +0.504 |
ickson | +0.494 |
Neg logit contributions | |
REDACTED | -0.487 |
Clown | -0.446 |
Capitol | -0.438 |
Certain | -0.431 |
Franch | -0.419 |
Token | were |
Token ID | 547 |
Feature activation | -6.30e-01 |
Pos logit contributions | |
ringe | +0.105 |
ritten | +0.102 |
itness | +0.093 |
kefeller | +0.093 |
ickson | +0.091 |
Neg logit contributions | |
REDACTED | -0.079 |
Clown | -0.071 |
Capitol | -0.070 |
Certain | -0.069 |
Franch | -0.067 |
Token | mut |
Token ID | 4517 |
Feature activation | +4.66e+00 |
Pos logit contributions | |
ringe | +0.958 |
ritten | +0.932 |
itness | +0.852 |
kefeller | +0.849 |
ickson | +0.834 |
Neg logit contributions | |
REDACTED | -0.739 |
Clown | -0.662 |
Capitol | -0.655 |
Certain | -0.644 |
Franch | -0.628 |
Token | ating |
Token ID | 803 |
Feature activation | -1.20e-03 |
Pos logit contributions | |
REDACTED | +4.936 |
Clown | +4.453 |
Capitol | +4.378 |
Certain | +4.308 |
Franch | +4.158 |
Neg logit contributions | |
ringe | -7.069 |
ritten | -6.856 |
kefeller | -6.442 |
itness | -6.380 |
ickson | -6.202 |
Token | ... |
Token ID | 986 |
Feature activation | -1.44e+00 |
Pos logit contributions | |
ringe | +0.002 |
ritten | +0.002 |
itness | +0.001 |
kefeller | +0.001 |
ickson | +0.001 |
Neg logit contributions | |
REDACTED | -0.002 |
Clown | -0.001 |
Capitol | -0.001 |
Certain | -0.001 |
Franch | -0.001 |
Token | becoming |
Token ID | 5033 |
Feature activation | -4.35e-01 |
Pos logit contributions | |
ringe | +2.150 |
ritten | +2.095 |
itness | +1.914 |
kefeller | +1.900 |
ickson | +1.868 |
Neg logit contributions | |
REDACTED | -1.715 |
Clown | -1.535 |
Capitol | -1.517 |
Certain | -1.502 |
Franch | -1.464 |
Token | more |
Token ID | 517 |
Feature activation | +1.40e-01 |
Pos logit contributions | |
ringe | +0.652 |
ritten | +0.633 |
itness | +0.582 |
kefeller | +0.577 |
ickson | +0.565 |
Neg logit contributions | |
REDACTED | -0.523 |
Clown | -0.473 |
Capitol | -0.466 |
Certain | -0.459 |
Franch | -0.445 |
Token | human |
Token ID | 1692 |
Feature activation | -2.28e+00 |
Pos logit contributions | |
REDACTED | +0.177 |
Clown | +0.161 |
Capitol | +0.159 |
Certain | +0.157 |
Franch | +0.151 |
Neg logit contributions | |
ringe | -0.201 |
ritten | -0.195 |
itness | -0.179 |
kefeller | -0.177 |
ickson | -0.173 |
Token | apply |
Token ID | 4174 |
Feature activation | +3.62e-01 |
Pos logit contributions | |
ringe | +0.607 |
ritten | +0.589 |
itness | +0.535 |
kefeller | +0.532 |
ickson | +0.522 |
Neg logit contributions | |
REDACTED | -0.557 |
Clown | -0.504 |
Capitol | -0.501 |
Certain | -0.495 |
Franch | -0.480 |
Token | _ |
Token ID | 62 |
Feature activation | +7.46e-01 |
Pos logit contributions | |
REDACTED | +0.325 |
Clown | +0.277 |
Capitol | +0.276 |
Certain | +0.273 |
Franch | +0.259 |
Neg logit contributions | |
ringe | -0.646 |
ritten | -0.634 |
itness | -0.589 |
kefeller | -0.588 |
ickson | -0.577 |
Token | move |
Token ID | 21084 |
Feature activation | +2.35e+00 |
Pos logit contributions | |
REDACTED | +1.022 |
Clown | +0.923 |
Capitol | +0.917 |
Certain | +0.911 |
Franch | +0.887 |
Neg logit contributions | |
ringe | -0.976 |
ritten | -0.946 |
itness | -0.854 |
kefeller | -0.849 |
ickson | -0.832 |
Token | ( |
Token ID | 357 |
Feature activation | +1.46e+00 |
Pos logit contributions | |
REDACTED | +2.792 |
Clown | +2.496 |
Capitol | +2.469 |
Certain | +2.448 |
Franch | +2.380 |
Neg logit contributions | |
ringe | -3.500 |
ritten | -3.415 |
itness | -3.123 |
kefeller | -3.105 |
ickson | -3.046 |
Token | & |
Token ID | 1222 |
Feature activation | -9.94e-01 |
Pos logit contributions | |
REDACTED | +1.709 |
Clown | +1.539 |
Capitol | +1.515 |
Certain | +1.514 |
Franch | +1.456 |
Neg logit contributions | |
ringe | -2.185 |
ritten | -2.142 |
kefeller | -1.958 |
itness | -1.956 |
ickson | -1.900 |
Token | mut |
Token ID | 4517 |
Feature activation | +4.65e+00 |
Pos logit contributions | |
ringe | +1.732 |
ritten | +1.685 |
itness | +1.572 |
kefeller | +1.541 |
ickson | +1.534 |
Neg logit contributions | |
REDACTED | -0.941 |
Clown | -0.814 |
Capitol | -0.800 |
Certain | -0.786 |
Franch | -0.771 |
Token | self |
Token ID | 2116 |
Feature activation | -5.84e-01 |
Pos logit contributions | |
REDACTED | +3.401 |
Capitol | +2.979 |
Certain | +2.966 |
Clown | +2.958 |
Assass | +2.697 |
Neg logit contributions | |
ringe | -8.628 |
ritten | -8.565 |
kefeller | -8.199 |
itness | -7.937 |
ickson | -7.731 |
Token | ) |
Token ID | 1267 |
Feature activation | +5.45e-01 |
Pos logit contributions | |
ringe | +0.915 |
ritten | +0.885 |
itness | +0.819 |
kefeller | +0.795 |
ickson | +0.793 |
Neg logit contributions | |
REDACTED | -0.659 |
Clown | -0.590 |
Capitol | -0.577 |
Certain | -0.567 |
Franch | -0.560 |
Token | { |
Token ID | 1391 |
Feature activation | +1.54e+00 |
Pos logit contributions | |
REDACTED | +0.611 |
Clown | +0.540 |
Capitol | +0.533 |
Certain | +0.528 |
Franch | +0.517 |
Neg logit contributions | |
ringe | -0.857 |
ritten | -0.835 |
itness | -0.769 |
kefeller | -0.758 |
ickson | -0.749 |
Token | let |
Token ID | 1309 |
Feature activation | -7.19e-01 |
Pos logit contributions | |
REDACTED | +1.711 |
Clown | +1.545 |
Capitol | +1.535 |
Certain | +1.529 |
Franch | +1.463 |
Neg logit contributions | |
ringe | -2.402 |
ritten | -2.358 |
kefeller | -2.168 |
itness | -2.139 |
ickson | -2.099 |
Token | prev |
Token ID | 8654 |
Feature activation | +6.34e-02 |
Pos logit contributions | |
ringe | +1.114 |
ritten | +1.084 |
itness | +0.995 |
kefeller | +0.993 |
ickson | +0.973 |
Neg logit contributions | |
REDACTED | -0.818 |
Clown | -0.729 |
Capitol | -0.723 |
Certain | -0.716 |
Franch | -0.693 |
Token | By |
Token ID | 3886 |
Feature activation | +2.26e+00 |
Pos logit contributions | |
ringe | +0.603 |
ritten | +0.587 |
itness | +0.539 |
kefeller | +0.536 |
ickson | +0.526 |
Neg logit contributions | |
REDACTED | -0.459 |
Clown | -0.405 |
Capitol | -0.402 |
Certain | -0.398 |
Franch | -0.386 |
Token | Andrew |
Token ID | 6858 |
Feature activation | -2.35e-01 |
Pos logit contributions | |
REDACTED | +2.514 |
Clown | +2.282 |
Capitol | +2.256 |
Certain | +2.230 |
Franch | +2.172 |
Neg logit contributions | |
ringe | -3.504 |
ritten | -3.420 |
kefeller | -3.142 |
itness | -3.123 |
ickson | -3.049 |
Token | Benson |
Token ID | 33940 |
Feature activation | -3.00e-01 |
Pos logit contributions | |
ringe | +0.377 |
ritten | +0.367 |
itness | +0.338 |
kefeller | +0.335 |
ickson | +0.329 |
Neg logit contributions | |
REDACTED | -0.253 |
Clown | -0.228 |
Capitol | -0.225 |
Certain | -0.223 |
Franch | -0.217 |
Token |
|
Token ID | 198 |
Feature activation | -1.70e-01 |
Pos logit contributions | |
ringe | +0.457 |
ritten | +0.445 |
itness | +0.408 |
kefeller | +0.405 |
ickson | +0.396 |
Neg logit contributions | |
REDACTED | -0.348 |
Clown | -0.313 |
Capitol | -0.311 |
Certain | -0.307 |
Franch | -0.298 |
Token |
|
Token ID | 198 |
Feature activation | -1.40e-01 |
Pos logit contributions | |
ringe | +0.273 |
ritten | +0.266 |
kefeller | +0.246 |
itness | +0.244 |
ickson | +0.239 |
Neg logit contributions | |
REDACTED | -0.186 |
Capitol | -0.163 |
Clown | -0.162 |
Certain | -0.161 |
Franch | -0.154 |
Token | Veter |
Token ID | 45182 |
Feature activation | +4.63e+00 |
Pos logit contributions | |
ringe | +0.215 |
ritten | +0.209 |
itness | +0.191 |
kefeller | +0.191 |
ickson | +0.187 |
Neg logit contributions | |
REDACTED | -0.164 |
Clown | -0.144 |
Capitol | -0.143 |
Certain | -0.142 |
Franch | -0.137 |
Token | an |
Token ID | 272 |
Feature activation | -3.81e-01 |
Pos logit contributions | |
REDACTED | +4.976 |
Certain | +4.558 |
Capitol | +4.557 |
Clown | +4.490 |
Franch | +4.317 |
Neg logit contributions | |
ringe | -6.814 |
ritten | -6.671 |
kefeller | -6.192 |
itness | -6.055 |
ickson | -5.977 |
Token | Nick |
Token ID | 8047 |
Feature activation | +3.32e-01 |
Pos logit contributions | |
ringe | +0.551 |
ritten | +0.535 |
itness | +0.487 |
kefeller | +0.485 |
ickson | +0.475 |
Neg logit contributions | |
REDACTED | -0.474 |
Clown | -0.430 |
Capitol | -0.427 |
Certain | -0.421 |
Franch | -0.410 |
Token | He |
Token ID | 679 |
Feature activation | +1.22e+00 |
Pos logit contributions | |
REDACTED | +0.349 |
Clown | +0.316 |
Capitol | +0.311 |
Certain | +0.306 |
Franch | +0.302 |
Neg logit contributions | |
ringe | -0.540 |
ritten | -0.526 |
itness | -0.485 |
kefeller | -0.480 |
ickson | -0.474 |
Token | id |
Token ID | 312 |
Feature activation | -6.99e-01 |
Pos logit contributions | |
REDACTED | +1.571 |
Clown | +1.411 |
Capitol | +1.409 |
Certain | +1.401 |
Franch | +1.360 |
Neg logit contributions | |
ringe | -1.655 |
ritten | -1.640 |
itness | -1.478 |
kefeller | -1.468 |
ickson | -1.409 |
Token | feld |
Token ID | 16265 |
Feature activation | +1.48e-01 |
Pos logit contributions | |
ringe | +1.169 |
ritten | +1.145 |
itness | +1.059 |
kefeller | +1.058 |
ickson | +1.028 |
Neg logit contributions | |
REDACTED | -0.698 |
Clown | -0.622 |
Capitol | -0.609 |
Certain | -0.600 |
Franch | -0.582 |
Token | Seattle |
Token ID | 7312 |
Feature activation | +9.26e-01 |
Pos logit contributions | |
REDACTED | +0.263 |
Capitol | +0.241 |
Clown | +0.239 |
Certain | +0.234 |
Franch | +0.227 |
Neg logit contributions | |
ringe | -0.350 |
ritten | -0.340 |
itness | -0.312 |
kefeller | -0.310 |
ickson | -0.304 |
Token | research |
Token ID | 2267 |
Feature activation | -3.56e-01 |
Pos logit contributions | |
REDACTED | +1.091 |
Capitol | +1.004 |
Clown | +0.995 |
Certain | +0.979 |
Franch | +0.950 |
Neg logit contributions | |
ringe | -1.381 |
ritten | -1.343 |
itness | -1.230 |
kefeller | -1.217 |
ickson | -1.194 |
Token | . |
Token ID | 13 |
Feature activation | +2.24e-01 |
Pos logit contributions | |
ringe | +0.479 |
ritten | +0.464 |
itness | +0.421 |
kefeller | +0.418 |
ickson | +0.407 |
Neg logit contributions | |
REDACTED | -0.478 |
Clown | -0.433 |
Capitol | -0.432 |
Certain | -0.426 |
Franch | -0.414 |
Token |
|
Token ID | 198 |
Feature activation | +3.78e-01 |
Pos logit contributions | |
REDACTED | +0.249 |
Clown | +0.222 |
Capitol | +0.221 |
Certain | +0.220 |
Franch | +0.212 |
Neg logit contributions | |
ringe | -0.353 |
ritten | -0.346 |
kefeller | -0.317 |
itness | -0.316 |
ickson | -0.308 |
Token |
|
Token ID | 198 |
Feature activation | +2.06e-01 |
Pos logit contributions | |
REDACTED | +0.406 |
Capitol | +0.356 |
Clown | +0.354 |
Certain | +0.351 |
Franch | +0.335 |
Neg logit contributions | |
ringe | -0.618 |
ritten | -0.601 |
kefeller | -0.560 |
itness | -0.554 |
ickson | -0.540 |
Token | Aut |
Token ID | 16541 |
Feature activation | +4.56e+00 |
Pos logit contributions | |
REDACTED | +0.241 |
Capitol | +0.211 |
Clown | +0.211 |
Certain | +0.209 |
Franch | +0.200 |
Neg logit contributions | |
ringe | -0.314 |
ritten | -0.307 |
kefeller | -0.282 |
itness | -0.282 |
ickson | -0.273 |
Token | or |
Token ID | 273 |
Feature activation | -1.19e+00 |
Pos logit contributions | |
REDACTED | +4.533 |
Capitol | +3.971 |
Certain | +3.967 |
Clown | +3.829 |
Assass | +3.727 |
Neg logit contributions | |
ringe | -7.085 |
ritten | -6.882 |
kefeller | -6.605 |
itness | -6.534 |
lished | -6.317 |
Token | noted |
Token ID | 4367 |
Feature activation | -2.10e-01 |
Pos logit contributions | |
ringe | +1.599 |
ritten | +1.554 |
itness | +1.404 |
kefeller | +1.397 |
ickson | +1.359 |
Neg logit contributions | |
REDACTED | -1.596 |
Clown | -1.458 |
Capitol | -1.448 |
Certain | -1.432 |
Franch | -1.402 |
Token | that |
Token ID | 326 |
Feature activation | -9.09e-01 |
Pos logit contributions | |
ringe | +0.301 |
ritten | +0.293 |
itness | +0.267 |
kefeller | +0.264 |
ickson | +0.259 |
Neg logit contributions | |
REDACTED | -0.264 |
Clown | -0.238 |
Capitol | -0.237 |
Certain | -0.233 |
Franch | -0.228 |
Token | high |
Token ID | 1029 |
Feature activation | +9.69e-01 |
Pos logit contributions | |
ringe | +1.331 |
ritten | +1.297 |
itness | +1.182 |
kefeller | +1.174 |
ickson | +1.151 |
Neg logit contributions | |
REDACTED | -1.111 |
Clown | -1.002 |
Capitol | -1.000 |
Certain | -0.984 |
Franch | -0.962 |
Token | - |
Token ID | 12 |
Feature activation | +5.53e-01 |
Pos logit contributions | |
REDACTED | +1.158 |
Capitol | +1.062 |
Clown | +1.048 |
Certain | +1.026 |
Franch | +1.006 |
Neg logit contributions | |
ringe | -1.440 |
ritten | -1.392 |
kefeller | -1.270 |
itness | -1.270 |
ickson | -1.232 |
Token | his |
Token ID | 465 |
Feature activation | +5.56e-01 |
Pos logit contributions | |
ringe | +0.955 |
ritten | +0.928 |
itness | +0.846 |
kefeller | +0.843 |
ickson | +0.824 |
Neg logit contributions | |
REDACTED | -0.830 |
Clown | -0.751 |
Capitol | -0.744 |
Certain | -0.735 |
Franch | -0.718 |
Token | fellow |
Token ID | 5891 |
Feature activation | -1.29e+00 |
Pos logit contributions | |
REDACTED | +0.679 |
Clown | +0.618 |
Capitol | +0.614 |
Certain | +0.601 |
Franch | +0.587 |
Neg logit contributions | |
ringe | -0.811 |
ritten | -0.787 |
kefeller | -0.724 |
itness | -0.720 |
ickson | -0.706 |
Token | mayoral |
Token ID | 43811 |
Feature activation | +1.49e+00 |
Pos logit contributions | |
ringe | +1.963 |
ritten | +1.915 |
itness | +1.755 |
kefeller | +1.747 |
ickson | +1.711 |
Neg logit contributions | |
REDACTED | -1.493 |
Clown | -1.335 |
Capitol | -1.322 |
Certain | -1.304 |
Franch | -1.271 |
Token | candidate |
Token ID | 4540 |
Feature activation | +1.16e+00 |
Pos logit contributions | |
REDACTED | +1.410 |
Clown | +1.246 |
Capitol | +1.235 |
Certain | +1.202 |
Franch | +1.193 |
Neg logit contributions | |
ringe | -2.547 |
ritten | -2.491 |
kefeller | -2.318 |
itness | -2.307 |
ickson | -2.256 |
Token | Sarah |
Token ID | 10490 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
REDACTED | +1.469 |
Clown | +1.331 |
Capitol | +1.310 |
Certain | +1.298 |
Franch | +1.279 |
Neg logit contributions | |
ringe | -1.642 |
ritten | -1.593 |
kefeller | -1.467 |
itness | -1.458 |
ickson | -1.416 |
Token | Thomson |
Token ID | 32425 |
Feature activation | -5.63e+00 |
Pos logit contributions | |
ringe | +1.573 |
ritten | +1.525 |
kefeller | +1.416 |
itness | +1.406 |
ickson | +1.370 |
Neg logit contributions | |
REDACTED | -1.210 |
Clown | -1.103 |
Capitol | -1.100 |
Certain | -1.083 |
Franch | -1.057 |
Token | , |
Token ID | 11 |
Feature activation | +2.41e-01 |
Pos logit contributions | |
ringe | +7.600 |
ritten | +7.405 |
itness | +6.747 |
ickson | +6.609 |
kefeller | +6.535 |
Neg logit contributions | |
REDACTED | -7.108 |
Capitol | -6.316 |
Clown | -6.293 |
Certain | -6.243 |
Franch | -6.148 |
Token | whose |
Token ID | 3025 |
Feature activation | -7.35e-02 |
Pos logit contributions | |
REDACTED | +0.302 |
Clown | +0.275 |
Capitol | +0.274 |
Certain | +0.268 |
Franch | +0.262 |
Neg logit contributions | |
ringe | -0.343 |
ritten | -0.334 |
kefeller | -0.303 |
itness | -0.303 |
ickson | -0.295 |
Token | campaign |
Token ID | 1923 |
Feature activation | +9.55e-01 |
Pos logit contributions | |
ringe | +0.107 |
ritten | +0.104 |
kefeller | +0.095 |
itness | +0.095 |
ickson | +0.093 |
Neg logit contributions | |
REDACTED | -0.090 |
Clown | -0.082 |
Capitol | -0.081 |
Certain | -0.080 |
Franch | -0.078 |
Token | has |
Token ID | 468 |
Feature activation | +4.21e-01 |
Pos logit contributions | |
REDACTED | +1.157 |
Capitol | +1.043 |
Clown | +1.042 |
Certain | +1.013 |
Franch | +1.000 |
Neg logit contributions | |
ringe | -1.408 |
ritten | -1.374 |
itness | -1.253 |
kefeller | -1.246 |
ickson | -1.225 |
Token | included |
Token ID | 3017 |
Feature activation | -4.38e-01 |
Pos logit contributions | |
REDACTED | +0.516 |
Clown | +0.469 |
Capitol | +0.466 |
Certain | +0.455 |
Franch | +0.447 |
Neg logit contributions | |
ringe | -0.615 |
ritten | -0.593 |
itness | -0.545 |
kefeller | -0.543 |
ickson | -0.532 |
Token | 000 |
Token ID | 830 |
Feature activation | -1.30e-01 |
Pos logit contributions | |
REDACTED | +0.053 |
Clown | +0.045 |
Capitol | +0.045 |
Certain | +0.044 |
Franch | +0.042 |
Neg logit contributions | |
ringe | -0.105 |
ritten | -0.102 |
itness | -0.096 |
kefeller | -0.095 |
ickson | -0.093 |
Token | people |
Token ID | 661 |
Feature activation | -7.12e-01 |
Pos logit contributions | |
ringe | +0.190 |
ritten | +0.185 |
itness | +0.167 |
kefeller | +0.167 |
ickson | +0.163 |
Neg logit contributions | |
REDACTED | -0.161 |
Clown | -0.146 |
Capitol | -0.146 |
Certain | -0.143 |
Franch | -0.139 |
Token | attended |
Token ID | 9141 |
Feature activation | -3.87e-01 |
Pos logit contributions | |
ringe | +0.946 |
ritten | +0.917 |
kefeller | +0.832 |
itness | +0.822 |
ickson | +0.799 |
Neg logit contributions | |
REDACTED | -0.963 |
Clown | -0.885 |
Capitol | -0.883 |
Certain | -0.866 |
Franch | -0.845 |
Token | ) |
Token ID | 8 |
Feature activation | +7.51e-02 |
Pos logit contributions | |
ringe | +0.556 |
ritten | +0.541 |
itness | +0.492 |
kefeller | +0.486 |
ickson | +0.477 |
Neg logit contributions | |
REDACTED | -0.482 |
Clown | -0.443 |
Capitol | -0.441 |
Certain | -0.431 |
Franch | -0.418 |
Token | including |
Token ID | 1390 |
Feature activation | -6.66e-02 |
Pos logit contributions | |
REDACTED | +0.097 |
Clown | +0.088 |
Capitol | +0.088 |
Certain | +0.086 |
Franch | +0.084 |
Neg logit contributions | |
ringe | -0.105 |
ritten | -0.102 |
itness | -0.092 |
kefeller | -0.091 |
ickson | -0.090 |
Token | Shirley |
Token ID | 40902 |
Feature activation | -5.20e+00 |
Pos logit contributions | |
ringe | +0.102 |
ritten | +0.099 |
itness | +0.091 |
kefeller | +0.090 |
ickson | +0.088 |
Neg logit contributions | |
REDACTED | -0.077 |
Capitol | -0.070 |
Clown | -0.070 |
Certain | -0.068 |
Franch | -0.066 |
Token | Ch |
Token ID | 609 |
Feature activation | +1.34e+00 |
Pos logit contributions | |
ringe | +7.969 |
ritten | +7.766 |
itness | +7.123 |
kefeller | +7.083 |
ickson | +6.931 |
Neg logit contributions | |
REDACTED | -5.949 |
Clown | -5.080 |
Capitol | -5.041 |
Certain | -4.979 |
Franch | -4.861 |
Token | ish |
Token ID | 680 |
Feature activation | -7.07e-01 |
Pos logit contributions | |
REDACTED | +1.614 |
Certain | +1.491 |
Capitol | +1.490 |
Clown | +1.473 |
Franch | +1.437 |
Neg logit contributions | |
ringe | -1.864 |
ritten | -1.838 |
kefeller | -1.695 |
itness | -1.632 |
ickson | -1.596 |
Token | ol |
Token ID | 349 |
Feature activation | -5.29e-01 |
Pos logit contributions | |
ringe | +0.950 |
ritten | +0.918 |
itness | +0.830 |
kefeller | +0.829 |
ickson | +0.803 |
Neg logit contributions | |
REDACTED | -0.949 |
Clown | -0.866 |
Capitol | -0.862 |
Certain | -0.850 |
Franch | -0.835 |
Token | m |
Token ID | 76 |
Feature activation | +8.32e-01 |
Pos logit contributions | |
ringe | +0.693 |
ritten | +0.671 |
kefeller | +0.602 |
itness | +0.602 |
ickson | +0.583 |
Neg logit contributions | |
REDACTED | -0.729 |
Clown | -0.666 |
Capitol | -0.660 |
Certain | -0.654 |
Franch | -0.642 |
Token | , |
Token ID | 11 |
Feature activation | +2.70e-01 |
Pos logit contributions | |
REDACTED | +1.085 |
Clown | +1.002 |
Capitol | +0.994 |
Certain | +0.971 |
Franch | +0.945 |
Neg logit contributions | |
ringe | -1.141 |
ritten | -1.115 |
itness | -1.003 |
kefeller | -0.993 |
ickson | -0.970 |
Token | oked |
Token ID | 6545 |
Feature activation | -4.81e-01 |
Pos logit contributions | |
ringe | +1.824 |
ritten | +1.764 |
itness | +1.558 |
kefeller | +1.542 |
ickson | +1.510 |
Neg logit contributions | |
REDACTED | -2.449 |
Clown | -2.253 |
Capitol | -2.237 |
Certain | -2.219 |
Franch | -2.170 |
Token | Hillary |
Token ID | 5073 |
Feature activation | +1.85e-01 |
Pos logit contributions | |
ringe | +0.646 |
ritten | +0.625 |
itness | +0.565 |
kefeller | +0.559 |
ickson | +0.549 |
Neg logit contributions | |
REDACTED | -0.649 |
Clown | -0.591 |
Capitol | -0.585 |
Certain | -0.579 |
Franch | -0.565 |
Token | Clinton |
Token ID | 2605 |
Feature activation | +1.28e+00 |
Pos logit contributions | |
REDACTED | +0.226 |
Clown | +0.206 |
Capitol | +0.203 |
Certain | +0.200 |
Franch | +0.194 |
Neg logit contributions | |
ringe | -0.272 |
ritten | -0.264 |
itness | -0.242 |
kefeller | -0.241 |
ickson | -0.235 |
Token | and |
Token ID | 290 |
Feature activation | -2.03e-02 |
Pos logit contributions | |
REDACTED | +1.491 |
Clown | +1.367 |
Capitol | +1.355 |
Certain | +1.320 |
Assass | +1.284 |
Neg logit contributions | |
ringe | -1.919 |
ritten | -1.869 |
kefeller | -1.726 |
itness | -1.715 |
ickson | -1.667 |
Token | Deborah |
Token ID | 36976 |
Feature activation | -2.02e+00 |
Pos logit contributions | |
ringe | +0.032 |
ritten | +0.031 |
itness | +0.028 |
kefeller | +0.028 |
ickson | +0.028 |
Neg logit contributions | |
REDACTED | -0.023 |
Clown | -0.021 |
Capitol | -0.020 |
Certain | -0.020 |
Franch | -0.019 |
Token | Wasserman |
Token ID | 38997 |
Feature activation | -5.17e+00 |
Pos logit contributions | |
ringe | +2.613 |
ritten | +2.534 |
itness | +2.288 |
kefeller | +2.260 |
ickson | +2.217 |
Neg logit contributions | |
REDACTED | -2.806 |
Clown | -2.537 |
Capitol | -2.517 |
Certain | -2.497 |
Franch | -2.446 |
Token | Schultz |
Token ID | 26530 |
Feature activation | -3.60e+00 |
Pos logit contributions | |
ringe | +6.371 |
ritten | +6.133 |
itness | +5.552 |
kefeller | +5.447 |
ickson | +5.379 |
Neg logit contributions | |
REDACTED | -7.270 |
Capitol | -6.545 |
Clown | -6.544 |
Certain | -6.536 |
Franch | -6.387 |
Token | will |
Token ID | 481 |
Feature activation | -7.95e-01 |
Pos logit contributions | |
ringe | +4.890 |
ritten | +4.723 |
itness | +4.306 |
kefeller | +4.217 |
ickson | +4.180 |
Neg logit contributions | |
REDACTED | -4.708 |
Clown | -4.218 |
Capitol | -4.165 |
Certain | -4.153 |
Franch | -4.069 |
Token | not |
Token ID | 407 |
Feature activation | +5.77e-01 |
Pos logit contributions | |
ringe | +1.124 |
ritten | +1.096 |
itness | +0.991 |
kefeller | +0.990 |
ickson | +0.971 |
Neg logit contributions | |
REDACTED | -1.010 |
Clown | -0.923 |
Capitol | -0.912 |
Certain | -0.899 |
Franch | -0.879 |
Token | allow |
Token ID | 1249 |
Feature activation | -1.32e+00 |
Pos logit contributions | |
REDACTED | +0.735 |
Clown | +0.688 |
Capitol | +0.676 |
Certain | +0.663 |
Franch | +0.650 |
Neg logit contributions | |
ringe | -0.791 |
ritten | -0.778 |
kefeller | -0.710 |
itness | -0.697 |
ickson | -0.689 |
Token | Bernie |
Token ID | 10477 |
Feature activation | +1.41e-01 |
Pos logit contributions | |
ringe | +2.064 |
ritten | +2.012 |
itness | +1.848 |
kefeller | +1.832 |
ickson | +1.804 |
Neg logit contributions | |
REDACTED | -1.493 |
Clown | -1.333 |
Capitol | -1.321 |
Certain | -1.302 |
Franch | -1.270 |
Token | Hon |
Token ID | 8835 |
Feature activation | -3.67e-01 |
Pos logit contributions | |
ringe | +0.058 |
ritten | +0.056 |
itness | +0.051 |
kefeller | +0.051 |
ickson | +0.050 |
Neg logit contributions | |
REDACTED | -0.048 |
Clown | -0.044 |
Capitol | -0.044 |
Certain | -0.043 |
Franch | -0.042 |
Token | ick |
Token ID | 624 |
Feature activation | +1.16e+00 |
Pos logit contributions | |
ringe | +0.455 |
ritten | +0.441 |
itness | +0.395 |
kefeller | +0.390 |
ickson | +0.382 |
Neg logit contributions | |
REDACTED | -0.529 |
Clown | -0.484 |
Capitol | -0.480 |
Certain | -0.476 |
Franch | -0.466 |
Token | man |
Token ID | 805 |
Feature activation | -6.80e-01 |
Pos logit contributions | |
REDACTED | +1.483 |
Clown | +1.344 |
Capitol | +1.333 |
Certain | +1.317 |
Franch | +1.293 |
Neg logit contributions | |
ringe | -1.611 |
ritten | -1.564 |
itness | -1.422 |
kefeller | -1.402 |
ickson | -1.379 |
Token | children |
Token ID | 1751 |
Feature activation | +3.03e-01 |
Pos logit contributions | |
ringe | +0.967 |
ritten | +0.934 |
itness | +0.855 |
kefeller | +0.844 |
ickson | +0.829 |
Neg logit contributions | |
REDACTED | -0.858 |
Clown | -0.781 |
Capitol | -0.776 |
Certain | -0.766 |
Franch | -0.751 |
Token | , |
Token ID | 11 |
Feature activation | +2.02e-01 |
Pos logit contributions | |
REDACTED | +0.386 |
Clown | +0.349 |
Capitol | +0.347 |
Certain | +0.340 |
Franch | +0.334 |
Neg logit contributions | |
ringe | -0.431 |
ritten | -0.417 |
itness | -0.379 |
kefeller | -0.378 |
ickson | -0.367 |
Token | Shirley |
Token ID | 40902 |
Feature activation | -5.13e+00 |
Pos logit contributions | |
REDACTED | +0.230 |
Clown | +0.206 |
Capitol | +0.205 |
Certain | +0.201 |
Franch | +0.197 |
Neg logit contributions | |
ringe | -0.315 |
ritten | -0.305 |
itness | -0.281 |
kefeller | -0.281 |
ickson | -0.274 |
Token | , |
Token ID | 11 |
Feature activation | +3.39e-01 |
Pos logit contributions | |
ringe | +7.160 |
ritten | +6.988 |
itness | +6.365 |
kefeller | +6.314 |
ickson | +6.211 |
Neg logit contributions | |
REDACTED | -6.382 |
Clown | -5.582 |
Capitol | -5.512 |
Certain | -5.477 |
Franch | -5.339 |
Token | is |
Token ID | 318 |
Feature activation | +4.69e-01 |
Pos logit contributions | |
REDACTED | +0.410 |
Clown | +0.370 |
Capitol | +0.369 |
Certain | +0.360 |
Franch | +0.352 |
Neg logit contributions | |
ringe | -0.504 |
ritten | -0.487 |
itness | -0.446 |
kefeller | -0.443 |
ickson | -0.433 |
Token | the |
Token ID | 262 |
Feature activation | -1.79e-01 |
Pos logit contributions | |
REDACTED | +0.564 |
Capitol | +0.509 |
Clown | +0.509 |
Certain | +0.498 |
Franch | +0.485 |
Neg logit contributions | |
ringe | -0.697 |
ritten | -0.676 |
itness | -0.621 |
kefeller | -0.616 |
ickson | -0.603 |
Token | mother |
Token ID | 2802 |
Feature activation | +1.09e-02 |
Pos logit contributions | |
ringe | +0.262 |
ritten | +0.254 |
itness | +0.233 |
kefeller | +0.230 |
ickson | +0.227 |
Neg logit contributions | |
REDACTED | -0.219 |
Clown | -0.200 |
Capitol | -0.200 |
Certain | -0.194 |
Franch | -0.190 |
Token | of |
Token ID | 286 |
Feature activation | -4.86e-01 |
Pos logit contributions | |
REDACTED | +0.013 |
Clown | +0.012 |
Capitol | +0.012 |
Certain | +0.012 |
Franch | +0.011 |
Neg logit contributions | |
ringe | -0.016 |
ritten | -0.016 |
itness | -0.014 |
kefeller | -0.014 |
ickson | -0.014 |
Token | turn |
Token ID | 1210 |
Feature activation | +5.05e-01 |
Pos logit contributions | |
ringe | +0.606 |
ritten | +0.591 |
kefeller | +0.535 |
itness | +0.532 |
ickson | +0.523 |
Neg logit contributions | |
REDACTED | -0.589 |
Capitol | -0.540 |
Clown | -0.539 |
Certain | -0.531 |
Franch | -0.518 |
Token | ." |
Token ID | 526 |
Feature activation | -7.84e-01 |
Pos logit contributions | |
REDACTED | +0.649 |
Clown | +0.595 |
Capitol | +0.588 |
Certain | +0.585 |
Franch | +0.564 |
Neg logit contributions | |
ringe | -0.702 |
ritten | -0.687 |
kefeller | -0.618 |
itness | -0.615 |
ickson | -0.602 |
Token |
|
Token ID | 198 |
Feature activation | +9.87e-02 |
Pos logit contributions | |
ringe | +1.174 |
ritten | +1.141 |
itness | +1.048 |
kefeller | +1.034 |
ickson | +1.022 |
Neg logit contributions | |
REDACTED | -0.929 |
Clown | -0.825 |
Capitol | -0.817 |
Certain | -0.803 |
Franch | -0.786 |
Token |
|
Token ID | 198 |
Feature activation | +4.60e-02 |
Pos logit contributions | |
REDACTED | +0.106 |
Capitol | +0.093 |
Clown | +0.093 |
Certain | +0.092 |
Franch | +0.088 |
Neg logit contributions | |
ringe | -0.160 |
ritten | -0.156 |
kefeller | -0.145 |
itness | -0.144 |
ickson | -0.140 |
Token | Neil |
Token ID | 29354 |
Feature activation | -9.79e-01 |
Pos logit contributions | |
REDACTED | +0.056 |
Clown | +0.050 |
Capitol | +0.050 |
Certain | +0.049 |
Franch | +0.048 |
Neg logit contributions | |
ringe | -0.068 |
ritten | -0.066 |
kefeller | -0.060 |
itness | -0.060 |
ickson | -0.059 |
Token | Thomson |
Token ID | 32425 |
Feature activation | -4.97e+00 |
Pos logit contributions | |
ringe | +1.573 |
ritten | +1.531 |
itness | +1.411 |
kefeller | +1.399 |
ickson | +1.378 |
Neg logit contributions | |
REDACTED | -1.060 |
Clown | -0.944 |
Capitol | -0.932 |
Certain | -0.921 |
Franch | -0.899 |
Token | - |
Token ID | 532 |
Feature activation | +3.43e-01 |
Pos logit contributions | |
ringe | +7.293 |
ritten | +7.029 |
itness | +6.478 |
ickson | +6.421 |
kefeller | +6.385 |
Neg logit contributions | |
REDACTED | -5.807 |
Clown | -5.032 |
Capitol | -5.004 |
Certain | -4.955 |
Franch | -4.829 |
Token | A |
Token ID | 317 |
Feature activation | -3.33e-01 |
Pos logit contributions | |
REDACTED | +0.361 |
Clown | +0.324 |
Capitol | +0.320 |
Certain | +0.316 |
Franch | +0.305 |
Neg logit contributions | |
ringe | -0.560 |
ritten | -0.546 |
itness | -0.502 |
kefeller | -0.501 |
ickson | -0.492 |
Token | Det |
Token ID | 4614 |
Feature activation | +8.98e-01 |
Pos logit contributions | |
ringe | +0.533 |
ritten | +0.521 |
itness | +0.479 |
kefeller | +0.477 |
ickson | +0.469 |
Neg logit contributions | |
REDACTED | -0.360 |
Clown | -0.324 |
Capitol | -0.320 |
Certain | -0.318 |
Franch | -0.307 |
Token | ailed |
Token ID | 6255 |
Feature activation | +2.11e+00 |
Pos logit contributions | |
REDACTED | +1.473 |
Clown | +1.387 |
Capitol | +1.374 |
Certain | +1.372 |
Franch | +1.345 |
Neg logit contributions | |
ringe | -0.887 |
ritten | -0.843 |
kefeller | -0.764 |
itness | -0.737 |
ickson | -0.720 |
Token | Review |
Token ID | 6602 |
Feature activation | +1.84e+00 |
Pos logit contributions | |
REDACTED | +2.276 |
Clown | +2.101 |
Capitol | +2.083 |
Certain | +2.063 |
Franch | +2.026 |
Neg logit contributions | |
ringe | -3.301 |
ritten | -3.228 |
kefeller | -2.990 |
itness | -2.942 |
ickson | -2.919 |
Token | , |
Token ID | 11 |
Feature activation | -9.05e-02 |
Pos logit contributions | |
ringe | +0.294 |
ritten | +0.287 |
kefeller | +0.264 |
itness | +0.260 |
ickson | +0.253 |
Neg logit contributions | |
REDACTED | -0.276 |
Clown | -0.249 |
Capitol | -0.248 |
Certain | -0.246 |
Franch | -0.237 |
Token | as |
Token ID | 355 |
Feature activation | -1.49e+00 |
Pos logit contributions | |
ringe | +0.135 |
ritten | +0.132 |
kefeller | +0.121 |
itness | +0.120 |
ickson | +0.117 |
Neg logit contributions | |
REDACTED | -0.108 |
Clown | -0.098 |
Capitol | -0.097 |
Certain | -0.095 |
Franch | -0.093 |
Token | if |
Token ID | 611 |
Feature activation | -8.74e-01 |
Pos logit contributions | |
ringe | +2.059 |
ritten | +2.003 |
kefeller | +1.812 |
itness | +1.806 |
ickson | +1.756 |
Neg logit contributions | |
REDACTED | -1.951 |
Clown | -1.779 |
Capitol | -1.754 |
Certain | -1.732 |
Franch | -1.692 |
Token | trying |
Token ID | 2111 |
Feature activation | -7.75e-01 |
Pos logit contributions | |
ringe | +1.269 |
ritten | +1.234 |
kefeller | +1.123 |
itness | +1.123 |
ickson | +1.094 |
Neg logit contributions | |
REDACTED | -1.087 |
Clown | -0.985 |
Capitol | -0.973 |
Certain | -0.958 |
Franch | -0.937 |
Token | to |
Token ID | 284 |
Feature activation | -2.95e-01 |
Pos logit contributions | |
ringe | +1.077 |
ritten | +1.054 |
itness | +0.959 |
kefeller | +0.935 |
ickson | +0.925 |
Neg logit contributions | |
REDACTED | -1.002 |
Clown | -0.893 |
Capitol | -0.888 |
Certain | -0.880 |
Franch | -0.868 |
Token | swat |
Token ID | 44284 |
Feature activation | -4.91e+00 |
Pos logit contributions | |
ringe | +0.437 |
ritten | +0.422 |
kefeller | +0.390 |
itness | +0.383 |
ickson | +0.377 |
Neg logit contributions | |
REDACTED | -0.358 |
Clown | -0.327 |
Capitol | -0.322 |
Certain | -0.320 |
Franch | -0.311 |
Token | away |
Token ID | 1497 |
Feature activation | -8.03e-02 |
Pos logit contributions | |
ringe | +6.900 |
ritten | +6.694 |
itness | +6.065 |
kefeller | +5.982 |
ickson | +5.931 |
Neg logit contributions | |
REDACTED | -6.110 |
Clown | -5.433 |
Capitol | -5.396 |
Certain | -5.364 |
Franch | -5.246 |
Token | a |
Token ID | 257 |
Feature activation | -3.78e-01 |
Pos logit contributions | |
ringe | +0.119 |
ritten | +0.116 |
itness | +0.107 |
kefeller | +0.106 |
ickson | +0.103 |
Neg logit contributions | |
REDACTED | -0.097 |
Clown | -0.088 |
Capitol | -0.087 |
Certain | -0.086 |
Franch | -0.084 |
Token | was |
Token ID | 373 |
Feature activation | -8.63e-01 |
Pos logit contributions | |
ringe | +0.534 |
ritten | +0.518 |
itness | +0.474 |
kefeller | +0.472 |
ickson | +0.461 |
Neg logit contributions | |
REDACTED | -0.484 |
Clown | -0.439 |
Capitol | -0.434 |
Certain | -0.429 |
Franch | -0.419 |
Token | p |
Token ID | 79 |
Feature activation | -5.48e-01 |
Pos logit contributions | |
ringe | +1.051 |
ritten | +1.018 |
itness | +0.914 |
kefeller | +0.911 |
ickson | +0.888 |
Neg logit contributions | |
REDACTED | -1.264 |
Clown | -1.156 |
Capitol | -1.143 |
Certain | -1.134 |
Franch | -1.110 |
Token | from |
Token ID | 422 |
Feature activation | +7.05e-01 |
Pos logit contributions | |
ringe | +0.759 |
ritten | +0.743 |
kefeller | +0.679 |
itness | +0.679 |
ickson | +0.655 |
Neg logit contributions | |
REDACTED | -0.704 |
Clown | -0.648 |
Capitol | -0.638 |
Certain | -0.629 |
Franch | -0.613 |
Token | questions |
Token ID | 2683 |
Feature activation | +9.81e-01 |
Pos logit contributions | |
REDACTED | +0.749 |
Capitol | +0.686 |
Clown | +0.682 |
Certain | +0.670 |
Franch | +0.641 |
Neg logit contributions | |
ringe | -1.008 |
ritten | -0.952 |
kefeller | -0.881 |
itness | -0.881 |
ickson | -0.857 |
Token | , |
Token ID | 11 |
Feature activation | -2.06e-01 |
Pos logit contributions | |
REDACTED | +1.102 |
Clown | +0.998 |
Capitol | +0.996 |
Certain | +0.975 |
Franch | +0.933 |
Neg logit contributions | |
ringe | -1.530 |
ritten | -1.491 |
itness | -1.368 |
kefeller | -1.363 |
ickson | -1.326 |
Token | � |
Token ID | 447 |
Feature activation | -1.69e-01 |
Pos logit contributions | |
ringe | +0.263 |
ritten | +0.254 |
itness | +0.228 |
kefeller | +0.227 |
ickson | +0.221 |
Neg logit contributions | |
REDACTED | -0.290 |
Clown | -0.267 |
Capitol | -0.264 |
Certain | -0.263 |
Franch | -0.256 |
Token | � |
Token ID | 251 |
Feature activation | +1.80e-01 |
Pos logit contributions | |
ringe | +0.261 |
ritten | +0.254 |
kefeller | +0.235 |
itness | +0.234 |
ickson | +0.228 |
Neg logit contributions | |
REDACTED | -0.192 |
Clown | -0.172 |
Capitol | -0.169 |
Certain | -0.169 |
Franch | -0.160 |
Token | said |
Token ID | 531 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
REDACTED | +0.186 |
Clown | +0.167 |
Capitol | +0.164 |
Certain | +0.163 |
Franch | +0.157 |
Neg logit contributions | |
ringe | -0.296 |
ritten | -0.288 |
itness | -0.267 |
kefeller | -0.265 |
ickson | -0.260 |
Token | Owen |
Token ID | 22605 |
Feature activation | -4.90e+00 |
Pos logit contributions | |
ringe | +1.663 |
ritten | +1.621 |
itness | +1.493 |
kefeller | +1.481 |
ickson | +1.457 |
Neg logit contributions | |
REDACTED | -1.132 |
Clown | -1.004 |
Capitol | -0.992 |
Certain | -0.980 |
Franch | -0.954 |
Token | Do |
Token ID | 2141 |
Feature activation | -2.37e-01 |
Pos logit contributions | |
ringe | +7.623 |
ritten | +7.482 |
itness | +6.871 |
kefeller | +6.827 |
ickson | +6.742 |
Neg logit contributions | |
REDACTED | -5.290 |
Clown | -4.454 |
Capitol | -4.440 |
Certain | -4.381 |
Franch | -4.225 |
Token | herty |
Token ID | 29029 |
Feature activation | -1.84e+00 |
Pos logit contributions | |
ringe | +0.271 |
ritten | +0.263 |
itness | +0.233 |
kefeller | +0.231 |
ickson | +0.225 |
Neg logit contributions | |
REDACTED | -0.362 |
Clown | -0.333 |
Capitol | -0.331 |
Certain | -0.328 |
Franch | -0.323 |
Token | , |
Token ID | 11 |
Feature activation | -1.17e-01 |
Pos logit contributions | |
ringe | +2.402 |
ritten | +2.329 |
itness | +2.100 |
kefeller | +2.078 |
ickson | +2.039 |
Neg logit contributions | |
REDACTED | -2.550 |
Clown | -2.318 |
Capitol | -2.298 |
Certain | -2.277 |
Franch | -2.231 |
Token | climate |
Token ID | 4258 |
Feature activation | +1.07e+00 |
Pos logit contributions | |
ringe | +0.170 |
ritten | +0.166 |
itness | +0.151 |
kefeller | +0.150 |
ickson | +0.146 |
Neg logit contributions | |
REDACTED | -0.142 |
Clown | -0.130 |
Capitol | -0.129 |
Certain | -0.127 |
Franch | -0.123 |
Token | model |
Token ID | 2746 |
Feature activation | -5.48e-01 |
Pos logit contributions | |
REDACTED | +1.231 |
Capitol | +1.125 |
Clown | +1.110 |
Certain | +1.076 |
Franch | +1.060 |
Neg logit contributions | |
ringe | -1.610 |
ritten | -1.570 |
itness | -1.435 |
kefeller | -1.422 |
ickson | -1.393 |
Token | - |
Token ID | 12 |
Feature activation | +5.00e-02 |
Pos logit contributions | |
REDACTED | +1.172 |
Clown | +1.070 |
Capitol | +1.066 |
Certain | +1.047 |
Franch | +1.023 |
Neg logit contributions | |
ringe | -1.289 |
ritten | -1.263 |
kefeller | -1.142 |
itness | -1.140 |
ickson | -1.099 |
Token | blown |
Token ID | 31290 |
Feature activation | +1.68e+00 |
Pos logit contributions | |
REDACTED | +0.052 |
Clown | +0.046 |
Capitol | +0.045 |
Certain | +0.044 |
Franch | +0.043 |
Neg logit contributions | |
ringe | -0.082 |
ritten | -0.080 |
itness | -0.074 |
kefeller | -0.074 |
ickson | -0.072 |
Token | pand |
Token ID | 19798 |
Feature activation | +5.02e-01 |
Pos logit contributions | |
REDACTED | +1.916 |
Clown | +1.771 |
Capitol | +1.758 |
Certain | +1.724 |
Franch | +1.680 |
Neg logit contributions | |
ringe | -2.543 |
ritten | -2.488 |
itness | -2.292 |
kefeller | -2.271 |
ickson | -2.202 |
Token | emic |
Token ID | 5314 |
Feature activation | -3.78e-01 |
Pos logit contributions | |
REDACTED | +0.774 |
Clown | +0.714 |
Capitol | +0.708 |
Certain | +0.704 |
Franch | +0.688 |
Neg logit contributions | |
ringe | -0.568 |
ritten | -0.553 |
itness | -0.488 |
kefeller | -0.486 |
ickson | -0.470 |
Token | . |
Token ID | 13 |
Feature activation | -1.34e-01 |
Pos logit contributions | |
ringe | +0.502 |
ritten | +0.487 |
itness | +0.440 |
kefeller | +0.437 |
ickson | +0.426 |
Neg logit contributions | |
REDACTED | -0.515 |
Clown | -0.468 |
Capitol | -0.464 |
Certain | -0.459 |
Franch | -0.449 |
Token | SW |
Token ID | 12672 |
Feature activation | -4.85e+00 |
Pos logit contributions | |
ringe | +0.212 |
ritten | +0.208 |
itness | +0.190 |
kefeller | +0.189 |
ickson | +0.186 |
Neg logit contributions | |
REDACTED | -0.148 |
Clown | -0.131 |
Capitol | -0.130 |
Certain | -0.129 |
Franch | -0.125 |
Token | INE |
Token ID | 8881 |
Feature activation | +1.99e-01 |
Pos logit contributions | |
ringe | +7.042 |
ritten | +6.849 |
itness | +6.269 |
kefeller | +6.135 |
ickson | +6.068 |
Neg logit contributions | |
REDACTED | -5.893 |
Clown | -5.241 |
Capitol | -5.161 |
Certain | -5.122 |
Franch | -4.990 |
Token | FL |
Token ID | 9977 |
Feature activation | +3.02e-01 |
Pos logit contributions | |
REDACTED | +0.246 |
Clown | +0.221 |
Capitol | +0.219 |
Certain | +0.218 |
Franch | +0.212 |
Neg logit contributions | |
ringe | -0.288 |
ritten | -0.283 |
itness | -0.257 |
kefeller | -0.256 |
ickson | -0.249 |
Token | U |
Token ID | 52 |
Feature activation | -9.72e-01 |
Pos logit contributions | |
REDACTED | +0.378 |
Clown | +0.338 |
Capitol | +0.338 |
Certain | +0.334 |
Franch | +0.322 |
Neg logit contributions | |
ringe | -0.433 |
ritten | -0.424 |
kefeller | -0.388 |
itness | -0.385 |
ickson | -0.375 |
Token | - |
Token ID | 532 |
Feature activation | +4.00e-01 |
Pos logit contributions | |
ringe | +1.253 |
ritten | +1.217 |
itness | +1.093 |
kefeller | +1.088 |
ickson | +1.055 |
Neg logit contributions | |
REDACTED | -1.360 |
Clown | -1.243 |
Capitol | -1.233 |
Certain | -1.221 |
Franch | -1.195 |
Token | THE |
Token ID | 3336 |
Feature activation | -7.87e-01 |
Pos logit contributions | |
REDACTED | +0.424 |
Clown | +0.377 |
Capitol | +0.373 |
Certain | +0.370 |
Franch | +0.357 |
Neg logit contributions | |
ringe | -0.652 |
ritten | -0.638 |
itness | -0.586 |
kefeller | -0.583 |
ickson | -0.573 |
Token | A |
Token ID | 317 |
Feature activation | -3.79e-01 |
Pos logit contributions | |
ringe | +0.262 |
ritten | +0.255 |
itness | +0.235 |
kefeller | +0.234 |
ickson | +0.230 |
Neg logit contributions | |
REDACTED | -0.169 |
Clown | -0.151 |
Capitol | -0.149 |
Certain | -0.147 |
Franch | -0.143 |
Token | ries |
Token ID | 1678 |
Feature activation | +7.46e-01 |
Pos logit contributions | |
ringe | +0.552 |
ritten | +0.537 |
itness | +0.488 |
kefeller | +0.487 |
ickson | +0.478 |
Neg logit contributions | |
REDACTED | -0.462 |
Clown | -0.421 |
Capitol | -0.417 |
Certain | -0.415 |
Franch | -0.402 |
Token | , |
Token ID | 11 |
Feature activation | -2.78e-02 |
Pos logit contributions | |
REDACTED | +0.864 |
Clown | +0.806 |
Capitol | +0.801 |
Certain | +0.786 |
Franch | +0.774 |
Neg logit contributions | |
ringe | -1.122 |
ritten | -1.091 |
kefeller | -0.997 |
itness | -0.995 |
ickson | -0.973 |
Token | Pic |
Token ID | 15085 |
Feature activation | -1.48e+00 |
Pos logit contributions | |
ringe | +0.045 |
ritten | +0.044 |
itness | +0.041 |
kefeller | +0.040 |
ickson | +0.040 |
Neg logit contributions | |
REDACTED | -0.029 |
Clown | -0.026 |
Capitol | -0.026 |
Certain | -0.026 |
Franch | -0.025 |
Token | , |
Token ID | 11 |
Feature activation | +6.88e-02 |
Pos logit contributions | |
ringe | +2.032 |
ritten | +1.973 |
itness | +1.791 |
kefeller | +1.771 |
ickson | +1.741 |
Neg logit contributions | |
REDACTED | -1.935 |
Clown | -1.753 |
Capitol | -1.738 |
Certain | -1.721 |
Franch | -1.683 |
Token | Russo |
Token ID | 46855 |
Feature activation | -4.82e+00 |
Pos logit contributions | |
REDACTED | +0.074 |
Clown | +0.066 |
Capitol | +0.066 |
Certain | +0.065 |
Franch | +0.063 |
Neg logit contributions | |
ringe | -0.111 |
ritten | -0.108 |
itness | -0.099 |
kefeller | -0.099 |
ickson | -0.097 |
Token | 's |
Token ID | 338 |
Feature activation | -6.17e-01 |
Pos logit contributions | |
ringe | +6.657 |
ritten | +6.484 |
itness | +5.918 |
kefeller | +5.806 |
ickson | +5.777 |
Neg logit contributions | |
REDACTED | -6.124 |
Clown | -5.325 |
Capitol | -5.270 |
Certain | -5.210 |
Franch | -5.082 |
Token | , |
Token ID | 11 |
Feature activation | +3.52e-02 |
Pos logit contributions | |
ringe | +0.925 |
ritten | +0.901 |
itness | +0.823 |
kefeller | +0.819 |
ickson | +0.803 |
Neg logit contributions | |
REDACTED | -0.731 |
Clown | -0.663 |
Capitol | -0.657 |
Certain | -0.648 |
Franch | -0.634 |
Token | Café |
Token ID | 42151 |
Feature activation | +8.64e-01 |
Pos logit contributions | |
REDACTED | +0.038 |
Clown | +0.034 |
Capitol | +0.034 |
Certain | +0.033 |
Franch | +0.032 |
Neg logit contributions | |
ringe | -0.057 |
ritten | -0.055 |
itness | -0.051 |
kefeller | -0.051 |
ickson | -0.050 |
Token | Anthony |
Token ID | 9953 |
Feature activation | +1.21e+00 |
Pos logit contributions | |
REDACTED | +0.903 |
Clown | +0.815 |
Capitol | +0.807 |
Certain | +0.793 |
Franch | +0.774 |
Neg logit contributions | |
ringe | -1.414 |
ritten | -1.378 |
itness | -1.271 |
kefeller | -1.264 |
ickson | -1.245 |
Token | , |
Token ID | 11 |
Feature activation | +3.04e-01 |
Pos logit contributions | |
REDACTED | +1.542 |
Clown | +1.426 |
Capitol | +1.418 |
Certain | +1.408 |
Franch | +1.380 |
Neg logit contributions | |
ringe | -1.687 |
ritten | -1.645 |
itness | -1.485 |
kefeller | -1.476 |
ickson | -1.444 |
Token | s |
Token ID | 82 |
Feature activation | -4.11e-01 |
Pos logit contributions | |
ringe | +0.461 |
ritten | +0.448 |
kefeller | +0.408 |
itness | +0.406 |
ickson | +0.393 |
Neg logit contributions | |
REDACTED | -0.434 |
Clown | -0.390 |
Capitol | -0.389 |
Certain | -0.381 |
Franch | -0.374 |
Token | chief |
Token ID | 4039 |
Feature activation | +1.22e+00 |
Pos logit contributions | |
ringe | +0.594 |
ritten | +0.578 |
itness | +0.526 |
kefeller | +0.525 |
ickson | +0.514 |
Neg logit contributions | |
REDACTED | -0.510 |
Clown | -0.459 |
Capitol | -0.458 |
Certain | -0.450 |
Franch | -0.440 |
Token | of |
Token ID | 286 |
Feature activation | -2.50e-01 |
Pos logit contributions | |
REDACTED | +1.441 |
Capitol | +1.303 |
Clown | +1.291 |
Certain | +1.242 |
Franch | +1.235 |
Neg logit contributions | |
ringe | -1.835 |
ritten | -1.789 |
kefeller | -1.643 |
itness | -1.625 |
ickson | -1.596 |
Token | staff |
Token ID | 3085 |
Feature activation | -2.29e-01 |
Pos logit contributions | |
ringe | +0.353 |
ritten | +0.343 |
itness | +0.310 |
kefeller | +0.309 |
ickson | +0.303 |
Neg logit contributions | |
REDACTED | -0.321 |
Capitol | -0.291 |
Clown | -0.289 |
Certain | -0.283 |
Franch | -0.278 |
Token | , |
Token ID | 11 |
Feature activation | +2.76e-01 |
Pos logit contributions | |
ringe | +0.323 |
ritten | +0.313 |
kefeller | +0.286 |
itness | +0.284 |
ickson | +0.276 |
Neg logit contributions | |
REDACTED | -0.295 |
Clown | -0.265 |
Capitol | -0.264 |
Certain | -0.259 |
Franch | -0.256 |
Token | Owen |
Token ID | 22605 |
Feature activation | -4.80e+00 |
Pos logit contributions | |
REDACTED | +0.303 |
Clown | +0.271 |
Capitol | +0.270 |
Certain | +0.264 |
Franch | +0.260 |
Neg logit contributions | |
ringe | -0.440 |
ritten | -0.428 |
itness | -0.394 |
kefeller | -0.393 |
ickson | -0.383 |
Token | Bru |
Token ID | 8274 |
Feature activation | +1.70e+00 |
Pos logit contributions | |
ringe | +7.446 |
ritten | +7.311 |
itness | +6.719 |
kefeller | +6.657 |
ickson | +6.577 |
Neg logit contributions | |
REDACTED | -5.172 |
Clown | -4.374 |
Capitol | -4.331 |
Certain | -4.311 |
Franch | -4.128 |
Token | gh |
Token ID | 456 |
Feature activation | -8.14e-02 |
Pos logit contributions | |
REDACTED | +2.120 |
Capitol | +1.936 |
Clown | +1.933 |
Certain | +1.895 |
Franch | +1.849 |
Neg logit contributions | |
ringe | -2.340 |
ritten | -2.292 |
kefeller | -2.080 |
itness | -2.063 |
ickson | -2.004 |
Token | , |
Token ID | 11 |
Feature activation | +4.56e-01 |
Pos logit contributions | |
ringe | +0.110 |
ritten | +0.107 |
itness | +0.097 |
kefeller | +0.097 |
ickson | +0.094 |
Neg logit contributions | |
REDACTED | -0.107 |
Clown | -0.098 |
Capitol | -0.097 |
Certain | -0.096 |
Franch | -0.094 |
Token | said |
Token ID | 531 |
Feature activation | -6.76e-01 |
Pos logit contributions | |
REDACTED | +0.580 |
Capitol | +0.526 |
Clown | +0.525 |
Certain | +0.510 |
Franch | +0.502 |
Neg logit contributions | |
ringe | -0.652 |
ritten | -0.630 |
kefeller | -0.576 |
itness | -0.575 |
ickson | -0.554 |
Token | this |
Token ID | 428 |
Feature activation | -1.06e+00 |
Pos logit contributions | |
ringe | +0.917 |
ritten | +0.890 |
itness | +0.805 |
kefeller | +0.800 |
ickson | +0.782 |
Neg logit contributions | |
REDACTED | -0.902 |
Clown | -0.818 |
Capitol | -0.813 |
Certain | -0.802 |
Franch | -0.786 |
Token | said |
Token ID | 531 |
Feature activation | -1.10e+00 |
Pos logit contributions | |
ringe | +0.198 |
ritten | +0.191 |
kefeller | +0.174 |
itness | +0.173 |
ickson | +0.169 |
Neg logit contributions | |
REDACTED | -0.190 |
Clown | -0.174 |
Capitol | -0.172 |
Certain | -0.169 |
Franch | -0.167 |
Token | , |
Token ID | 11 |
Feature activation | -2.09e-01 |
Pos logit contributions | |
ringe | +1.496 |
ritten | +1.454 |
itness | +1.314 |
kefeller | +1.310 |
ickson | +1.281 |
Neg logit contributions | |
REDACTED | -1.461 |
Clown | -1.329 |
Capitol | -1.314 |
Certain | -1.297 |
Franch | -1.270 |
Token | before |
Token ID | 878 |
Feature activation | -3.59e-01 |
Pos logit contributions | |
ringe | +0.312 |
ritten | +0.304 |
itness | +0.277 |
kefeller | +0.277 |
ickson | +0.271 |
Neg logit contributions | |
REDACTED | -0.251 |
Clown | -0.226 |
Capitol | -0.224 |
Certain | -0.220 |
Franch | -0.215 |
Token | clicking |
Token ID | 12264 |
Feature activation | -1.74e+00 |
Pos logit contributions | |
ringe | +0.558 |
ritten | +0.544 |
kefeller | +0.498 |
itness | +0.497 |
ickson | +0.486 |
Neg logit contributions | |
REDACTED | -0.408 |
Clown | -0.367 |
Capitol | -0.363 |
Certain | -0.355 |
Franch | -0.348 |
Token | the |
Token ID | 262 |
Feature activation | -1.67e-01 |
Pos logit contributions | |
ringe | +2.592 |
ritten | +2.523 |
itness | +2.306 |
kefeller | +2.289 |
ickson | +2.249 |
Neg logit contributions | |
REDACTED | -2.098 |
Clown | -1.883 |
Capitol | -1.862 |
Certain | -1.841 |
Franch | -1.795 |
Token | shutter |
Token ID | 31168 |
Feature activation | -4.79e+00 |
Pos logit contributions | |
ringe | +0.197 |
ritten | +0.193 |
kefeller | +0.171 |
itness | +0.169 |
ickson | +0.166 |
Neg logit contributions | |
REDACTED | -0.251 |
Clown | -0.233 |
Capitol | -0.231 |
Certain | -0.227 |
Franch | -0.221 |
Token | . |
Token ID | 13 |
Feature activation | -1.08e-01 |
Pos logit contributions | |
ringe | +6.049 |
ritten | +5.851 |
itness | +5.245 |
ickson | +5.115 |
kefeller | +5.089 |
Neg logit contributions | |
REDACTED | -6.721 |
Clown | -6.027 |
Capitol | -6.004 |
Certain | -5.979 |
Franch | -5.911 |
Token | The |
Token ID | 383 |
Feature activation | -5.96e-01 |
Pos logit contributions | |
ringe | +0.172 |
ritten | +0.168 |
kefeller | +0.154 |
itness | +0.154 |
ickson | +0.151 |
Neg logit contributions | |
REDACTED | -0.117 |
Clown | -0.105 |
Capitol | -0.104 |
Certain | -0.103 |
Franch | -0.100 |
Token | man |
Token ID | 582 |
Feature activation | -3.92e-01 |
Pos logit contributions | |
ringe | +0.815 |
ritten | +0.796 |
kefeller | +0.720 |
itness | +0.715 |
ickson | +0.701 |
Neg logit contributions | |
REDACTED | -0.786 |
Clown | -0.716 |
Capitol | -0.714 |
Certain | -0.701 |
Franch | -0.686 |
Token | covered |
Token ID | 5017 |
Feature activation | -3.93e-01 |
Pos logit contributions | |
ringe | +0.556 |
ritten | +0.538 |
kefeller | +0.494 |
itness | +0.489 |
ickson | +0.477 |
Neg logit contributions | |
REDACTED | -0.498 |
Clown | -0.457 |
Capitol | -0.452 |
Certain | -0.440 |
Franch | -0.436 |
Token | his |
Token ID | 465 |
Feature activation | +9.68e-01 |
Pos logit contributions | |
ringe | +0.629 |
ritten | +0.613 |
itness | +0.564 |
kefeller | +0.560 |
ickson | +0.550 |
Neg logit contributions | |
REDACTED | -0.427 |
Clown | -0.380 |
Capitol | -0.377 |
Certain | -0.369 |
Franch | -0.360 |
Token | Trou |
Token ID | 22141 |
Feature activation | -2.53e+00 |
Pos logit contributions | |
REDACTED | +0.095 |
Clown | +0.085 |
Capitol | +0.084 |
Certain | +0.083 |
Franch | +0.080 |
Neg logit contributions | |
ringe | -0.155 |
ritten | -0.151 |
itness | -0.139 |
kefeller | -0.139 |
ickson | -0.136 |
Token | bles |
Token ID | 7689 |
Feature activation | -1.20e+00 |
Pos logit contributions | |
ringe | +2.278 |
ritten | +2.180 |
itness | +1.865 |
kefeller | +1.846 |
ickson | +1.798 |
Neg logit contributions | |
REDACTED | -4.469 |
Clown | -4.185 |
Capitol | -4.159 |
Certain | -4.147 |
Franch | -4.051 |
Token |
|
Token ID | 198 |
Feature activation | -3.37e-02 |
Pos logit contributions | |
ringe | +1.797 |
ritten | +1.749 |
itness | +1.600 |
kefeller | +1.587 |
ickson | +1.560 |
Neg logit contributions | |
REDACTED | -1.428 |
Clown | -1.278 |
Capitol | -1.266 |
Certain | -1.252 |
Franch | -1.221 |
Token |
|
Token ID | 198 |
Feature activation | -1.52e-02 |
Pos logit contributions | |
ringe | +0.054 |
ritten | +0.053 |
kefeller | +0.049 |
itness | +0.049 |
ickson | +0.048 |
Neg logit contributions | |
REDACTED | -0.037 |
Clown | -0.032 |
Capitol | -0.032 |
Certain | -0.032 |
Franch | -0.031 |
Token | J |
Token ID | 41 |
Feature activation | -4.20e-01 |
Pos logit contributions | |
ringe | +0.024 |
ritten | +0.023 |
kefeller | +0.021 |
itness | +0.021 |
ickson | +0.021 |
Neg logit contributions | |
REDACTED | -0.017 |
Clown | -0.015 |
Capitol | -0.015 |
Certain | -0.015 |
Franch | -0.014 |
Token | oshi |
Token ID | 13704 |
Feature activation | -4.79e+00 |
Pos logit contributions | |
ringe | +0.472 |
ritten | +0.458 |
itness | +0.403 |
kefeller | +0.402 |
ickson | +0.390 |
Neg logit contributions | |
REDACTED | -0.649 |
Clown | -0.594 |
Capitol | -0.591 |
Certain | -0.589 |
Franch | -0.580 |
Token | , |
Token ID | 11 |
Feature activation | +6.30e-01 |
Pos logit contributions | |
ringe | +6.568 |
ritten | +6.478 |
itness | +5.890 |
kefeller | +5.803 |
ickson | +5.713 |
Neg logit contributions | |
REDACTED | -6.013 |
Clown | -5.338 |
Capitol | -5.313 |
Certain | -5.261 |
Franch | -5.143 |
Token | Ok |
Token ID | 6762 |
Feature activation | +1.62e+00 |
Pos logit contributions | |
REDACTED | +0.762 |
Clown | +0.692 |
Capitol | +0.687 |
Certain | +0.674 |
Franch | +0.659 |
Neg logit contributions | |
ringe | -0.930 |
ritten | -0.901 |
itness | -0.823 |
kefeller | -0.821 |
ickson | -0.804 |
Token | ami |
Token ID | 6277 |
Feature activation | -1.21e-01 |
Pos logit contributions | |
REDACTED | +2.167 |
Capitol | +1.997 |
Clown | +1.984 |
Certain | +1.963 |
Franch | +1.937 |
Neg logit contributions | |
ringe | -2.105 |
ritten | -2.049 |
kefeller | -1.862 |
itness | -1.820 |
ickson | -1.769 |
Token | and |
Token ID | 290 |
Feature activation | +2.13e-01 |
Pos logit contributions | |
ringe | +0.162 |
ritten | +0.157 |
itness | +0.141 |
kefeller | +0.141 |
ickson | +0.137 |
Neg logit contributions | |
REDACTED | -0.163 |
Clown | -0.149 |
Capitol | -0.148 |
Certain | -0.146 |
Franch | -0.143 |
Token | Islam |
Token ID | 3449 |
Feature activation | -1.93e+00 |
Pos logit contributions | |
REDACTED | +0.239 |
Clown | +0.214 |
Capitol | +0.212 |
Certain | +0.209 |
Franch | +0.203 |
Neg logit contributions | |
ringe | -0.335 |
ritten | -0.327 |
itness | -0.299 |
kefeller | -0.298 |
ickson | -0.293 |
Token |
|
Token ID | 198 |
Feature activation | -3.01e-01 |
Pos logit contributions | |
ringe | +0.602 |
ritten | +0.586 |
kefeller | +0.540 |
itness | +0.539 |
ickson | +0.525 |
Neg logit contributions | |
REDACTED | -0.413 |
Clown | -0.364 |
Capitol | -0.363 |
Certain | -0.358 |
Franch | -0.346 |
Token | But |
Token ID | 1537 |
Feature activation | +1.53e+00 |
Pos logit contributions | |
ringe | +0.441 |
ritten | +0.428 |
itness | +0.391 |
kefeller | +0.386 |
ickson | +0.382 |
Neg logit contributions | |
REDACTED | -0.368 |
Clown | -0.332 |
Capitol | -0.327 |
Certain | -0.324 |
Franch | -0.317 |
Token | as |
Token ID | 355 |
Feature activation | -7.80e-01 |
Pos logit contributions | |
REDACTED | +1.829 |
Capitol | +1.659 |
Clown | +1.655 |
Certain | +1.620 |
Franch | +1.583 |
Neg logit contributions | |
ringe | -2.241 |
ritten | -2.218 |
kefeller | -2.032 |
itness | -1.988 |
ickson | -1.949 |
Token | Ma |
Token ID | 6669 |
Feature activation | +1.63e+00 |
Pos logit contributions | |
ringe | +1.136 |
ritten | +1.111 |
kefeller | +1.012 |
itness | +1.006 |
ickson | +0.979 |
Neg logit contributions | |
REDACTED | -0.965 |
Clown | -0.873 |
Capitol | -0.872 |
Certain | -0.851 |
Franch | -0.835 |
Token | ureen |
Token ID | 49851 |
Feature activation | -1.39e+00 |
Pos logit contributions | |
REDACTED | +2.243 |
Clown | +2.092 |
Capitol | +2.079 |
Certain | +2.048 |
Franch | +1.999 |
Neg logit contributions | |
ringe | -2.088 |
ritten | -2.020 |
kefeller | -1.831 |
itness | -1.812 |
ickson | -1.745 |
Token | Dow |
Token ID | 21842 |
Feature activation | -4.78e+00 |
Pos logit contributions | |
ringe | +2.175 |
ritten | +2.117 |
itness | +1.947 |
kefeller | +1.932 |
ickson | +1.899 |
Neg logit contributions | |
REDACTED | -1.552 |
Clown | -1.387 |
Capitol | -1.371 |
Certain | -1.354 |
Franch | -1.321 |
Token | d |
Token ID | 67 |
Feature activation | -8.32e-01 |
Pos logit contributions | |
ringe | +6.383 |
ritten | +6.213 |
itness | +5.607 |
kefeller | +5.528 |
ickson | +5.454 |
Neg logit contributions | |
REDACTED | -6.469 |
Clown | -5.825 |
Capitol | -5.785 |
Certain | -5.734 |
Franch | -5.624 |
Token | recounted |
Token ID | 34187 |
Feature activation | -1.69e+00 |
Pos logit contributions | |
ringe | +1.157 |
ritten | +1.117 |
kefeller | +1.014 |
itness | +1.010 |
ickson | +0.983 |
Neg logit contributions | |
REDACTED | -1.081 |
Clown | -0.989 |
Capitol | -0.978 |
Certain | -0.967 |
Franch | -0.942 |
Token | in |
Token ID | 287 |
Feature activation | +1.67e-02 |
Pos logit contributions | |
ringe | +2.290 |
ritten | +2.221 |
itness | +2.014 |
kefeller | +1.990 |
ickson | +1.955 |
Neg logit contributions | |
REDACTED | -2.248 |
Clown | -2.034 |
Capitol | -2.014 |
Certain | -1.998 |
Franch | -1.955 |
Token | the |
Token ID | 262 |
Feature activation | +5.53e-01 |
Pos logit contributions | |
REDACTED | +0.018 |
Clown | +0.016 |
Capitol | +0.016 |
Certain | +0.016 |
Franch | +0.015 |
Neg logit contributions | |
ringe | -0.027 |
ritten | -0.026 |
itness | -0.024 |
kefeller | -0.024 |
ickson | -0.023 |
Token | New |
Token ID | 968 |
Feature activation | +1.89e+00 |
Pos logit contributions | |
REDACTED | +0.642 |
Clown | +0.590 |
Capitol | +0.590 |
Certain | +0.575 |
Franch | +0.556 |
Neg logit contributions | |
ringe | -0.832 |
ritten | -0.818 |
kefeller | -0.744 |
itness | -0.740 |
ickson | -0.728 |
Token | Use |
Token ID | 5765 |
Feature activation | -8.81e-01 |
Pos logit contributions | |
ringe | +0.142 |
ritten | +0.139 |
kefeller | +0.128 |
itness | +0.128 |
ickson | +0.125 |
Neg logit contributions | |
REDACTED | -0.085 |
Clown | -0.075 |
Capitol | -0.074 |
Certain | -0.074 |
Franch | -0.070 |
Token | caution |
Token ID | 13041 |
Feature activation | +1.97e+00 |
Pos logit contributions | |
ringe | +1.286 |
ritten | +1.252 |
itness | +1.141 |
kefeller | +1.134 |
ickson | +1.110 |
Neg logit contributions | |
REDACTED | -1.086 |
Clown | -0.976 |
Capitol | -0.967 |
Certain | -0.957 |
Franch | -0.932 |
Token | ! |
Token ID | 0 |
Feature activation | -1.17e+00 |
Pos logit contributions | |
REDACTED | +2.378 |
Clown | +2.128 |
Capitol | +2.115 |
Certain | +2.107 |
Franch | +1.997 |
Neg logit contributions | |
ringe | -2.885 |
ritten | -2.814 |
kefeller | -2.601 |
itness | -2.531 |
ickson | -2.513 |
Token | Put |
Token ID | 11588 |
Feature activation | +3.45e-01 |
Pos logit contributions | |
ringe | +1.939 |
ritten | +1.894 |
itness | +1.745 |
kefeller | +1.736 |
ickson | +1.707 |
Neg logit contributions | |
REDACTED | -1.218 |
Clown | -1.071 |
Capitol | -1.059 |
Certain | -1.049 |
Franch | -1.015 |
Token | several |
Token ID | 1811 |
Feature activation | +4.49e-01 |
Pos logit contributions | |
REDACTED | +0.412 |
Clown | +0.371 |
Capitol | +0.365 |
Certain | +0.365 |
Franch | +0.350 |
Neg logit contributions | |
ringe | -0.514 |
ritten | -0.502 |
kefeller | -0.459 |
itness | -0.459 |
ickson | -0.448 |
Token | spoon |
Token ID | 24556 |
Feature activation | -4.74e+00 |
Pos logit contributions | |
REDACTED | +0.601 |
Clown | +0.550 |
Capitol | +0.546 |
Certain | +0.538 |
Franch | +0.523 |
Neg logit contributions | |
ringe | -0.607 |
ritten | -0.589 |
kefeller | -0.532 |
itness | -0.531 |
ickson | -0.517 |
Token | ful |
Token ID | 913 |
Feature activation | -1.54e+00 |
Pos logit contributions | |
ringe | +4.122 |
ritten | +3.907 |
itness | +3.334 |
kefeller | +3.200 |
ickson | +3.172 |
Neg logit contributions | |
REDACTED | -8.560 |
Clown | -7.904 |
Capitol | -7.855 |
Certain | -7.809 |
Franch | -7.726 |
Token | s |
Token ID | 82 |
Feature activation | -5.35e-01 |
Pos logit contributions | |
ringe | +1.685 |
ritten | +1.623 |
itness | +1.432 |
kefeller | +1.414 |
ickson | +1.381 |
Neg logit contributions | |
REDACTED | -2.458 |
Clown | -2.265 |
Capitol | -2.248 |
Certain | -2.230 |
Franch | -2.191 |
Token | of |
Token ID | 286 |
Feature activation | -4.47e-01 |
Pos logit contributions | |
ringe | +0.795 |
ritten | +0.774 |
itness | +0.707 |
kefeller | +0.702 |
ickson | +0.690 |
Neg logit contributions | |
REDACTED | -0.641 |
Clown | -0.575 |
Capitol | -0.569 |
Certain | -0.563 |
Franch | -0.548 |
Token | frozen |
Token ID | 12912 |
Feature activation | -4.95e-01 |
Pos logit contributions | |
ringe | +0.627 |
ritten | +0.610 |
itness | +0.554 |
kefeller | +0.551 |
ickson | +0.539 |
Neg logit contributions | |
REDACTED | -0.575 |
Clown | -0.523 |
Capitol | -0.517 |
Certain | -0.512 |
Franch | -0.498 |
Token | juice |
Token ID | 13135 |
Feature activation | +1.56e-01 |
Pos logit contributions | |
ringe | +0.702 |
ritten | +0.683 |
itness | +0.622 |
kefeller | +0.616 |
ickson | +0.605 |
Neg logit contributions | |
REDACTED | -0.629 |
Clown | -0.569 |
Capitol | -0.565 |
Certain | -0.558 |
Franch | -0.545 |
Token | has |
Token ID | 468 |
Feature activation | -5.41e-02 |
Pos logit contributions | |
ringe | +1.029 |
ritten | +1.003 |
kefeller | +0.906 |
itness | +0.905 |
ickson | +0.881 |
Neg logit contributions | |
REDACTED | -0.980 |
Clown | -0.898 |
Capitol | -0.888 |
Certain | -0.872 |
Franch | -0.857 |
Token | apparently |
Token ID | 5729 |
Feature activation | +7.66e-01 |
Pos logit contributions | |
ringe | +0.081 |
ritten | +0.078 |
kefeller | +0.072 |
itness | +0.072 |
ickson | +0.070 |
Neg logit contributions | |
REDACTED | -0.064 |
Clown | -0.059 |
Capitol | -0.058 |
Certain | -0.056 |
Franch | -0.056 |
Token | not |
Token ID | 407 |
Feature activation | +5.39e-01 |
Pos logit contributions | |
REDACTED | +0.929 |
Clown | +0.859 |
Capitol | +0.853 |
Franch | +0.817 |
Certain | +0.815 |
Neg logit contributions | |
ringe | -1.116 |
ritten | -1.068 |
kefeller | -1.009 |
itness | -0.984 |
ickson | -0.977 |
Token | apologized |
Token ID | 22893 |
Feature activation | +2.51e-02 |
Pos logit contributions | |
REDACTED | +0.647 |
Clown | +0.597 |
Capitol | +0.588 |
Certain | +0.569 |
Franch | +0.566 |
Neg logit contributions | |
ringe | -0.789 |
ritten | -0.763 |
kefeller | -0.722 |
itness | -0.705 |
ickson | -0.691 |
Token | to |
Token ID | 284 |
Feature activation | -3.56e-01 |
Pos logit contributions | |
REDACTED | +0.032 |
Clown | +0.029 |
Capitol | +0.029 |
Certain | +0.028 |
Franch | +0.027 |
Neg logit contributions | |
ringe | -0.036 |
ritten | -0.035 |
kefeller | -0.032 |
itness | -0.032 |
ickson | -0.031 |
Token | Crow |
Token ID | 15502 |
Feature activation | -4.69e+00 |
Pos logit contributions | |
ringe | +0.483 |
ritten | +0.469 |
itness | +0.424 |
kefeller | +0.422 |
ickson | +0.412 |
Neg logit contributions | |
REDACTED | -0.475 |
Clown | -0.434 |
Capitol | -0.429 |
Certain | -0.423 |
Franch | -0.415 |
Token | der |
Token ID | 1082 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
ringe | +4.870 |
ritten | +4.614 |
itness | +4.061 |
ickson | +3.973 |
kefeller | +3.941 |
Neg logit contributions | |
REDACTED | -7.647 |
Clown | -7.101 |
Capitol | -7.008 |
Certain | -6.984 |
Franch | -6.916 |
Token | . |
Token ID | 13 |
Feature activation | -3.78e-01 |
Pos logit contributions | |
ringe | +1.529 |
ritten | +1.484 |
itness | +1.347 |
kefeller | +1.337 |
ickson | +1.309 |
Neg logit contributions | |
REDACTED | -1.474 |
Clown | -1.336 |
Capitol | -1.324 |
Certain | -1.308 |
Franch | -1.281 |
Token |
|
Token ID | 198 |
Feature activation | -4.07e-01 |
Pos logit contributions | |
ringe | +0.578 |
ritten | +0.562 |
itness | +0.517 |
kefeller | +0.511 |
ickson | +0.504 |
Neg logit contributions | |
REDACTED | -0.438 |
Clown | -0.388 |
Capitol | -0.384 |
Certain | -0.380 |
Franch | -0.370 |
Token |
|
Token ID | 198 |
Feature activation | -4.09e-01 |
Pos logit contributions | |
ringe | +0.651 |
ritten | +0.633 |
kefeller | +0.584 |
itness | +0.583 |
ickson | +0.568 |
Neg logit contributions | |
REDACTED | -0.448 |
Clown | -0.394 |
Capitol | -0.393 |
Certain | -0.388 |
Franch | -0.374 |
Token | The |
Token ID | 464 |
Feature activation | +2.68e-01 |
Pos logit contributions | |
ringe | +0.595 |
ritten | +0.580 |
kefeller | +0.531 |
itness | +0.530 |
ickson | +0.516 |
Neg logit contributions | |
REDACTED | -0.509 |
Clown | -0.451 |
Capitol | -0.445 |
Certain | -0.443 |
Franch | -0.424 |
Token | teach |
Token ID | 4545 |
Feature activation | +1.34e+00 |
Pos logit contributions | |
ringe | +0.022 |
ritten | +0.022 |
kefeller | +0.020 |
itness | +0.020 |
ickson | +0.019 |
Neg logit contributions | |
REDACTED | -0.016 |
Clown | -0.015 |
Capitol | -0.014 |
Certain | -0.014 |
Franch | -0.014 |
Token | older |
Token ID | 4697 |
Feature activation | +4.48e-02 |
Pos logit contributions | |
REDACTED | +1.404 |
Clown | +1.292 |
Capitol | +1.263 |
Certain | +1.243 |
Franch | +1.190 |
Neg logit contributions | |
ringe | -2.189 |
ritten | -2.128 |
kefeller | -1.981 |
itness | -1.952 |
ickson | -1.895 |
Token | children |
Token ID | 1751 |
Feature activation | +9.64e-02 |
Pos logit contributions | |
REDACTED | +0.051 |
Clown | +0.046 |
Capitol | +0.046 |
Certain | +0.045 |
Franch | +0.044 |
Neg logit contributions | |
ringe | -0.069 |
ritten | -0.067 |
kefeller | -0.062 |
itness | -0.062 |
ickson | -0.060 |
Token | how |
Token ID | 703 |
Feature activation | -2.62e-01 |
Pos logit contributions | |
REDACTED | +0.111 |
Clown | +0.101 |
Capitol | +0.099 |
Certain | +0.098 |
Franch | +0.095 |
Neg logit contributions | |
ringe | -0.147 |
ritten | -0.144 |
kefeller | -0.132 |
itness | -0.132 |
ickson | -0.128 |
Token | to |
Token ID | 284 |
Feature activation | -1.47e-01 |
Pos logit contributions | |
ringe | +0.378 |
ritten | +0.367 |
itness | +0.335 |
kefeller | +0.332 |
ickson | +0.326 |
Neg logit contributions | |
REDACTED | -0.327 |
Clown | -0.294 |
Capitol | -0.291 |
Certain | -0.288 |
Franch | -0.282 |
Token | sew |
Token ID | 34249 |
Feature activation | -4.68e+00 |
Pos logit contributions | |
ringe | +0.221 |
ritten | +0.216 |
kefeller | +0.198 |
itness | +0.197 |
ickson | +0.192 |
Neg logit contributions | |
REDACTED | -0.172 |
Clown | -0.159 |
Capitol | -0.156 |
Certain | -0.154 |
Franch | -0.150 |
Token | . |
Token ID | 13 |
Feature activation | -1.91e-01 |
Pos logit contributions | |
ringe | +6.153 |
ritten | +5.979 |
itness | +5.403 |
ickson | +5.279 |
kefeller | +5.243 |
Neg logit contributions | |
REDACTED | -6.306 |
Clown | -5.622 |
Capitol | -5.607 |
Certain | -5.539 |
Franch | -5.464 |
Token |
|
Token ID | 198 |
Feature activation | -5.34e-01 |
Pos logit contributions | |
ringe | +0.307 |
ritten | +0.300 |
kefeller | +0.276 |
itness | +0.275 |
ickson | +0.269 |
Neg logit contributions | |
REDACTED | -0.208 |
Clown | -0.186 |
Capitol | -0.183 |
Certain | -0.182 |
Franch | -0.176 |
Token |
|
Token ID | 198 |
Feature activation | -5.84e-01 |
Pos logit contributions | |
ringe | +0.851 |
ritten | +0.829 |
kefeller | +0.763 |
itness | +0.763 |
ickson | +0.743 |
Neg logit contributions | |
REDACTED | -0.590 |
Clown | -0.520 |
Capitol | -0.518 |
Certain | -0.512 |
Franch | -0.495 |
Token | Thread |
Token ID | 16818 |
Feature activation | -2.12e+00 |
Pos logit contributions | |
ringe | +0.912 |
ritten | +0.888 |
kefeller | +0.819 |
itness | +0.817 |
ickson | +0.798 |
Neg logit contributions | |
REDACTED | -0.662 |
Clown | -0.581 |
Capitol | -0.575 |
Certain | -0.573 |
Franch | -0.548 |
Token | a |
Token ID | 257 |
Feature activation | -2.58e-01 |
Pos logit contributions | |
ringe | +3.099 |
ritten | +3.014 |
itness | +2.753 |
kefeller | +2.730 |
ickson | +2.681 |
Neg logit contributions | |
REDACTED | -2.606 |
Clown | -2.343 |
Capitol | -2.318 |
Certain | -2.297 |
Franch | -2.240 |
Token | America |
Token ID | 2253 |
Feature activation | -1.27e+00 |
Pos logit contributions | |
ringe | +0.909 |
ritten | +0.880 |
itness | +0.800 |
kefeller | +0.798 |
ickson | +0.781 |
Neg logit contributions | |
REDACTED | -0.748 |
Clown | -0.694 |
Capitol | -0.690 |
Certain | -0.685 |
Franch | -0.659 |
Token | . |
Token ID | 13 |
Feature activation | -3.53e-02 |
Pos logit contributions | |
ringe | +1.702 |
ritten | +1.655 |
itness | +1.493 |
kefeller | +1.479 |
ickson | +1.447 |
Neg logit contributions | |
REDACTED | -1.720 |
Clown | -1.568 |
Capitol | -1.555 |
Certain | -1.540 |
Franch | -1.508 |
Token | Wells |
Token ID | 18292 |
Feature activation | -3.34e+00 |
Pos logit contributions | |
ringe | +0.057 |
ritten | +0.056 |
kefeller | +0.052 |
itness | +0.051 |
ickson | +0.050 |
Neg logit contributions | |
REDACTED | -0.038 |
Clown | -0.034 |
Capitol | -0.033 |
Certain | -0.033 |
Franch | -0.032 |
Token | Fargo |
Token ID | 28125 |
Feature activation | -2.39e+00 |
Pos logit contributions | |
ringe | +4.019 |
ritten | +3.887 |
itness | +3.480 |
kefeller | +3.423 |
ickson | +3.370 |
Neg logit contributions | |
REDACTED | -4.925 |
Clown | -4.483 |
Capitol | -4.443 |
Certain | -4.412 |
Franch | -4.325 |
Token | , |
Token ID | 11 |
Feature activation | -2.29e-01 |
Pos logit contributions | |
ringe | +3.379 |
ritten | +3.283 |
itness | +2.986 |
kefeller | +2.959 |
ickson | +2.906 |
Neg logit contributions | |
REDACTED | -3.048 |
Clown | -2.748 |
Capitol | -2.723 |
Certain | -2.695 |
Franch | -2.636 |
Token | JPMorgan |
Token ID | 43318 |
Feature activation | -4.68e+00 |
Pos logit contributions | |
ringe | +0.335 |
ritten | +0.327 |
itness | +0.296 |
kefeller | +0.294 |
ickson | +0.288 |
Neg logit contributions | |
REDACTED | -0.279 |
Capitol | -0.256 |
Clown | -0.256 |
Certain | -0.250 |
Franch | -0.244 |
Token | Chase |
Token ID | 16346 |
Feature activation | -8.95e-01 |
Pos logit contributions | |
ringe | +5.348 |
ritten | +5.120 |
itness | +4.563 |
kefeller | +4.523 |
ickson | +4.416 |
Neg logit contributions | |
REDACTED | -7.163 |
Clown | -6.454 |
Capitol | -6.409 |
Certain | -6.377 |
Franch | -6.253 |
Token | , |
Token ID | 11 |
Feature activation | -2.80e-02 |
Pos logit contributions | |
ringe | +1.149 |
ritten | +1.113 |
itness | +1.003 |
kefeller | +0.992 |
ickson | +0.972 |
Neg logit contributions | |
REDACTED | -1.256 |
Clown | -1.145 |
Capitol | -1.135 |
Certain | -1.125 |
Franch | -1.103 |
Token | and |
Token ID | 290 |
Feature activation | -4.97e-01 |
Pos logit contributions | |
ringe | +0.041 |
ritten | +0.040 |
itness | +0.037 |
kefeller | +0.036 |
ickson | +0.036 |
Neg logit contributions | |
REDACTED | -0.034 |
Capitol | -0.031 |
Clown | -0.031 |
Certain | -0.031 |
Franch | -0.030 |
Token | Cit |
Token ID | 15792 |
Feature activation | +2.44e+00 |
Pos logit contributions | |
ringe | +0.764 |
ritten | +0.749 |
itness | +0.682 |
kefeller | +0.679 |
ickson | +0.665 |
Neg logit contributions | |
REDACTED | -0.570 |
Clown | -0.517 |
Capitol | -0.517 |
Certain | -0.506 |
Franch | -0.489 |
Token | ib |
Token ID | 571 |
Feature activation | -1.42e+00 |
Pos logit contributions | |
Capitol | +2.799 |
REDACTED | +2.789 |
Certain | +2.742 |
Clown | +2.660 |
Franch | +2.612 |
Neg logit contributions | |
ritten | -3.403 |
ringe | -3.368 |
itness | -3.002 |
kefeller | -3.002 |
ickson | -2.837 |
Token | I |
Token ID | 314 |
Feature activation | +1.40e-01 |
Pos logit contributions | |
ringe | +0.453 |
ritten | +0.441 |
kefeller | +0.406 |
itness | +0.404 |
ickson | +0.395 |
Neg logit contributions | |
REDACTED | -0.324 |
Clown | -0.291 |
Capitol | -0.287 |
Certain | -0.283 |
Franch | -0.276 |
Token | don |
Token ID | 836 |
Feature activation | +1.96e-01 |
Pos logit contributions | |
REDACTED | +0.153 |
Clown | +0.139 |
Capitol | +0.137 |
Certain | +0.136 |
Franch | +0.132 |
Neg logit contributions | |
ringe | -0.220 |
ritten | -0.216 |
kefeller | -0.200 |
itness | -0.196 |
ickson | -0.194 |
Token | 't |
Token ID | 470 |
Feature activation | +1.47e-01 |
Pos logit contributions | |
REDACTED | +0.272 |
Capitol | +0.247 |
Clown | +0.246 |
Certain | +0.244 |
Franch | +0.236 |
Neg logit contributions | |
ringe | -0.259 |
ritten | -0.250 |
kefeller | -0.229 |
itness | -0.227 |
ickson | -0.218 |
Token | want |
Token ID | 765 |
Feature activation | -1.90e-01 |
Pos logit contributions | |
REDACTED | +0.156 |
Clown | +0.141 |
Capitol | +0.138 |
Certain | +0.137 |
Franch | +0.133 |
Neg logit contributions | |
ringe | -0.237 |
ritten | -0.234 |
kefeller | -0.217 |
itness | -0.212 |
ickson | -0.210 |
Token | to |
Token ID | 284 |
Feature activation | -3.02e-01 |
Pos logit contributions | |
ringe | +0.277 |
ritten | +0.269 |
itness | +0.245 |
kefeller | +0.244 |
ickson | +0.239 |
Neg logit contributions | |
REDACTED | -0.234 |
Clown | -0.211 |
Capitol | -0.209 |
Certain | -0.206 |
Franch | -0.202 |
Token | screw |
Token ID | 9580 |
Feature activation | -4.61e+00 |
Pos logit contributions | |
ringe | +0.445 |
ritten | +0.435 |
kefeller | +0.407 |
itness | +0.394 |
ickson | +0.391 |
Neg logit contributions | |
REDACTED | -0.361 |
Clown | -0.336 |
Capitol | -0.326 |
Certain | -0.325 |
Franch | -0.315 |
Token | it |
Token ID | 340 |
Feature activation | +1.62e-01 |
Pos logit contributions | |
ringe | +6.023 |
ritten | +5.933 |
itness | +5.315 |
kefeller | +5.215 |
ickson | +5.209 |
Neg logit contributions | |
REDACTED | -6.137 |
Clown | -5.475 |
Capitol | -5.440 |
Certain | -5.418 |
Franch | -5.271 |
Token | up |
Token ID | 510 |
Feature activation | -3.97e-01 |
Pos logit contributions | |
REDACTED | +0.189 |
Clown | +0.173 |
Capitol | +0.171 |
Certain | +0.168 |
Franch | +0.163 |
Neg logit contributions | |
ringe | -0.245 |
ritten | -0.238 |
itness | -0.220 |
kefeller | -0.219 |
ickson | -0.212 |
Token | . |
Token ID | 13 |
Feature activation | -5.23e-01 |
Pos logit contributions | |
ringe | +0.565 |
ritten | +0.549 |
kefeller | +0.505 |
itness | +0.500 |
ickson | +0.486 |
Neg logit contributions | |
REDACTED | -0.502 |
Clown | -0.460 |
Capitol | -0.453 |
Certain | -0.448 |
Franch | -0.435 |
Token | I |
Token ID | 314 |
Feature activation | +1.10e-01 |
Pos logit contributions | |
ringe | +0.874 |
ritten | +0.853 |
itness | +0.786 |
kefeller | +0.786 |
ickson | +0.769 |
Neg logit contributions | |
REDACTED | -0.536 |
Clown | -0.478 |
Capitol | -0.470 |
Certain | -0.469 |
Franch | -0.452 |
Token | want |
Token ID | 765 |
Feature activation | -7.18e-02 |
Pos logit contributions | |
REDACTED | +0.120 |
Clown | +0.108 |
Capitol | +0.107 |
Certain | +0.106 |
Franch | +0.104 |
Neg logit contributions | |
ringe | -0.173 |
ritten | -0.169 |
kefeller | -0.157 |
itness | -0.155 |
ickson | -0.153 |
Token | pired |
Token ID | 6474 |
Feature activation | +1.76e+00 |
Pos logit contributions | |
REDACTED | +1.077 |
Clown | +0.990 |
Capitol | +0.978 |
Certain | +0.971 |
Franch | +0.939 |
Neg logit contributions | |
ringe | -1.232 |
ritten | -1.187 |
itness | -1.085 |
kefeller | -1.082 |
ickson | -1.054 |
Token | against |
Token ID | 1028 |
Feature activation | +1.15e+00 |
Pos logit contributions | |
REDACTED | +1.998 |
Clown | +1.823 |
Capitol | +1.798 |
Certain | +1.771 |
Franch | +1.724 |
Neg logit contributions | |
ringe | -2.692 |
ritten | -2.623 |
itness | -2.403 |
kefeller | -2.393 |
ickson | -2.320 |
Token | us |
Token ID | 514 |
Feature activation | -2.98e-01 |
Pos logit contributions | |
REDACTED | +1.184 |
Clown | +1.071 |
Capitol | +1.055 |
Certain | +1.033 |
Franch | +1.008 |
Neg logit contributions | |
ringe | -1.896 |
ritten | -1.844 |
kefeller | -1.722 |
itness | -1.697 |
ickson | -1.655 |
Token | ," |
Token ID | 553 |
Feature activation | -1.16e+00 |
Pos logit contributions | |
ringe | +0.436 |
ritten | +0.423 |
itness | +0.386 |
kefeller | +0.386 |
ickson | +0.376 |
Neg logit contributions | |
REDACTED | -0.365 |
Clown | -0.331 |
Capitol | -0.329 |
Certain | -0.324 |
Franch | -0.317 |
Token | said |
Token ID | 531 |
Feature activation | -9.87e-01 |
Pos logit contributions | |
ringe | +1.496 |
ritten | +1.447 |
itness | +1.304 |
kefeller | +1.295 |
ickson | +1.267 |
Neg logit contributions | |
REDACTED | -1.626 |
Clown | -1.485 |
Capitol | -1.470 |
Certain | -1.457 |
Franch | -1.429 |
Token | Owen |
Token ID | 22605 |
Feature activation | -4.60e+00 |
Pos logit contributions | |
ringe | +1.484 |
ritten | +1.445 |
itness | +1.322 |
kefeller | +1.311 |
ickson | +1.289 |
Neg logit contributions | |
REDACTED | -1.168 |
Clown | -1.045 |
Capitol | -1.033 |
Certain | -1.022 |
Franch | -0.998 |
Token | C |
Token ID | 327 |
Feature activation | +4.97e-01 |
Pos logit contributions | |
ringe | +7.126 |
ritten | +6.975 |
itness | +6.422 |
kefeller | +6.344 |
ickson | +6.236 |
Neg logit contributions | |
REDACTED | -5.211 |
Clown | -4.493 |
Capitol | -4.416 |
Certain | -4.379 |
Franch | -4.257 |
Token | oyle |
Token ID | 19802 |
Feature activation | +2.13e-01 |
Pos logit contributions | |
REDACTED | +0.587 |
Clown | +0.536 |
Capitol | +0.533 |
Certain | +0.530 |
Franch | +0.516 |
Neg logit contributions | |
ritten | -0.708 |
ringe | -0.705 |
kefeller | -0.635 |
itness | -0.633 |
ickson | -0.608 |
Token | , |
Token ID | 11 |
Feature activation | +3.02e-01 |
Pos logit contributions | |
REDACTED | +0.284 |
Clown | +0.259 |
Capitol | +0.256 |
Certain | +0.253 |
Franch | +0.248 |
Neg logit contributions | |
ringe | -0.287 |
ritten | -0.279 |
kefeller | -0.251 |
itness | -0.251 |
ickson | -0.245 |
Token | Bolton |
Token ID | 35293 |
Feature activation | +1.56e+00 |
Pos logit contributions | |
REDACTED | +0.346 |
Clown | +0.312 |
Capitol | +0.308 |
Certain | +0.304 |
Franch | +0.295 |
Neg logit contributions | |
ringe | -0.464 |
ritten | -0.453 |
kefeller | -0.417 |
itness | -0.413 |
ickson | -0.405 |
Token | 's |
Token ID | 338 |
Feature activation | -4.68e-02 |
Pos logit contributions | |
REDACTED | +2.183 |
Clown | +1.998 |
Capitol | +1.979 |
Certain | +1.958 |
Franch | +1.920 |
Neg logit contributions | |
ringe | -1.991 |
ritten | -1.921 |
itness | -1.727 |
kefeller | -1.720 |
ickson | -1.681 |
Token | ," |
Token ID | 553 |
Feature activation | -1.75e+00 |
Pos logit contributions | |
REDACTED | +0.607 |
Capitol | +0.552 |
Clown | +0.550 |
Certain | +0.542 |
Franch | +0.526 |
Neg logit contributions | |
ringe | -0.689 |
ritten | -0.667 |
kefeller | -0.607 |
itness | -0.605 |
ickson | -0.585 |
Token | says |
Token ID | 1139 |
Feature activation | -2.90e-01 |
Pos logit contributions | |
ringe | +2.472 |
ritten | +2.404 |
itness | +2.190 |
kefeller | +2.164 |
ickson | +2.130 |
Neg logit contributions | |
REDACTED | -2.230 |
Clown | -2.000 |
Capitol | -1.975 |
Certain | -1.962 |
Franch | -1.917 |
Token | Dr |
Token ID | 1583 |
Feature activation | -1.79e+00 |
Pos logit contributions | |
ringe | +0.469 |
ritten | +0.457 |
itness | +0.421 |
kefeller | +0.418 |
ickson | +0.411 |
Neg logit contributions | |
REDACTED | -0.310 |
Clown | -0.277 |
Capitol | -0.275 |
Certain | -0.270 |
Franch | -0.263 |
Token | . |
Token ID | 13 |
Feature activation | +7.04e-01 |
Pos logit contributions | |
ringe | +2.388 |
ritten | +2.314 |
itness | +2.100 |
kefeller | +2.093 |
ickson | +2.023 |
Neg logit contributions | |
REDACTED | -2.414 |
Clown | -2.197 |
Capitol | -2.197 |
Certain | -2.164 |
Franch | -2.121 |
Token | Judith |
Token ID | 39987 |
Feature activation | -2.70e+00 |
Pos logit contributions | |
REDACTED | +0.778 |
Clown | +0.711 |
Capitol | +0.706 |
Certain | +0.694 |
Franch | +0.680 |
Neg logit contributions | |
ringe | -1.104 |
ritten | -1.075 |
kefeller | -0.988 |
itness | -0.988 |
ickson | -0.960 |
Token | Owens |
Token ID | 31005 |
Feature activation | -4.59e+00 |
Pos logit contributions | |
ringe | +4.276 |
ritten | +4.173 |
itness | +3.838 |
kefeller | +3.804 |
ickson | +3.766 |
Neg logit contributions | |
REDACTED | -2.942 |
Clown | -2.534 |
Capitol | -2.510 |
Certain | -2.481 |
Franch | -2.404 |
Token | , |
Token ID | 11 |
Feature activation | -2.25e-03 |
Pos logit contributions | |
ringe | +5.532 |
ritten | +5.373 |
itness | +4.875 |
kefeller | +4.728 |
ickson | +4.690 |
Neg logit contributions | |
REDACTED | -6.670 |
Clown | -5.905 |
Certain | -5.851 |
Capitol | -5.793 |
Franch | -5.740 |
Token | a |
Token ID | 257 |
Feature activation | -4.97e-01 |
Pos logit contributions | |
ringe | +0.003 |
ritten | +0.003 |
itness | +0.003 |
kefeller | +0.003 |
ickson | +0.003 |
Neg logit contributions | |
REDACTED | -0.003 |
Capitol | -0.003 |
Clown | -0.003 |
Certain | -0.002 |
Franch | -0.002 |
Token | sleep |
Token ID | 3993 |
Feature activation | -2.20e+00 |
Pos logit contributions | |
ringe | +0.742 |
ritten | +0.718 |
itness | +0.656 |
kefeller | +0.655 |
ickson | +0.641 |
Neg logit contributions | |
REDACTED | -0.589 |
Capitol | -0.545 |
Clown | -0.541 |
Certain | -0.528 |
Franch | -0.512 |
Token | researcher |
Token ID | 13453 |
Feature activation | -1.90e+00 |
Pos logit contributions | |
ringe | +3.239 |
ritten | +3.151 |
itness | +2.881 |
kefeller | +2.848 |
ickson | +2.805 |
Neg logit contributions | |
REDACTED | -2.679 |
Clown | -2.400 |
Capitol | -2.373 |
Certain | -2.352 |
Franch | -2.297 |
Token | at |
Token ID | 379 |
Feature activation | +1.57e-01 |
Pos logit contributions | |
ringe | +2.487 |
ritten | +2.409 |
itness | +2.173 |
kefeller | +2.154 |
ickson | +2.110 |
Neg logit contributions | |
REDACTED | -2.618 |
Clown | -2.380 |
Capitol | -2.367 |
Certain | -2.338 |
Franch | -2.291 |
Token | he |
Token ID | 339 |
Feature activation | +7.83e-01 |
Pos logit contributions | |
REDACTED | +0.123 |
Clown | +0.112 |
Capitol | +0.111 |
Certain | +0.110 |
Franch | +0.107 |
Neg logit contributions | |
ringe | -0.139 |
ritten | -0.134 |
itness | -0.121 |
kefeller | -0.121 |
ickson | -0.119 |
Token | feels |
Token ID | 5300 |
Feature activation | +5.37e-01 |
Pos logit contributions | |
REDACTED | +0.927 |
Clown | +0.838 |
Capitol | +0.829 |
Certain | +0.814 |
Franch | +0.789 |
Neg logit contributions | |
ringe | -1.176 |
ritten | -1.140 |
kefeller | -1.043 |
itness | -1.031 |
ickson | -1.012 |
Token | life |
Token ID | 1204 |
Feature activation | -2.60e-01 |
Pos logit contributions | |
REDACTED | +0.645 |
Clown | +0.586 |
Capitol | +0.584 |
Certain | +0.575 |
Franch | +0.561 |
Neg logit contributions | |
ringe | -0.800 |
ritten | -0.775 |
kefeller | -0.711 |
itness | -0.704 |
ickson | -0.689 |
Token | stirring |
Token ID | 26547 |
Feature activation | -8.51e-01 |
Pos logit contributions | |
ringe | +0.374 |
ritten | +0.364 |
kefeller | +0.333 |
itness | +0.331 |
ickson | +0.322 |
Neg logit contributions | |
REDACTED | -0.323 |
Clown | -0.292 |
Capitol | -0.289 |
Certain | -0.285 |
Franch | -0.277 |
Token | on |
Token ID | 319 |
Feature activation | +1.28e+00 |
Pos logit contributions | |
ringe | +1.223 |
ritten | +1.188 |
itness | +1.083 |
kefeller | +1.076 |
ickson | +1.056 |
Neg logit contributions | |
REDACTED | -1.065 |
Clown | -0.958 |
Capitol | -0.950 |
Certain | -0.940 |
Franch | -0.917 |
Token | all |
Token ID | 477 |
Feature activation | +2.60e+00 |
Pos logit contributions | |
REDACTED | +1.536 |
Capitol | +1.394 |
Clown | +1.381 |
Certain | +1.363 |
Franch | +1.323 |
Neg logit contributions | |
ringe | -1.923 |
ritten | -1.854 |
itness | -1.696 |
kefeller | -1.677 |
ickson | -1.662 |
Token | sides |
Token ID | 5389 |
Feature activation | +1.53e-01 |
Pos logit contributions | |
REDACTED | +2.376 |
Capitol | +2.144 |
Clown | +2.081 |
Certain | +2.032 |
Franch | +1.958 |
Neg logit contributions | |
ringe | -4.527 |
ritten | -4.364 |
itness | -4.072 |
kefeller | -4.057 |
ickson | -4.003 |
Token | � |
Token ID | 447 |
Feature activation | +1.03e-02 |
Pos logit contributions | |
REDACTED | +0.203 |
Clown | +0.184 |
Capitol | +0.183 |
Certain | +0.181 |
Franch | +0.176 |
Neg logit contributions | |
ringe | -0.208 |
ritten | -0.202 |
itness | -0.183 |
kefeller | -0.182 |
ickson | -0.178 |
Token | � |
Token ID | 247 |
Feature activation | -6.97e-02 |
Pos logit contributions | |
REDACTED | +0.013 |
Clown | +0.011 |
Capitol | +0.011 |
Certain | +0.011 |
Franch | +0.011 |
Neg logit contributions | |
ringe | -0.015 |
ritten | -0.015 |
itness | -0.013 |
kefeller | -0.013 |
ickson | -0.013 |
Token |
|
Token ID | 198 |
Feature activation | -1.22e-01 |
Pos logit contributions | |
ringe | +0.105 |
ritten | +0.102 |
kefeller | +0.093 |
itness | +0.093 |
ickson | +0.091 |
Neg logit contributions | |
REDACTED | -0.082 |
Clown | -0.074 |
Certain | -0.073 |
Capitol | -0.073 |
Franch | -0.071 |
Token |
|
Token ID | 198 |
Feature activation | -1.49e-01 |
Pos logit contributions | |
ringe | +0.197 |
ritten | +0.192 |
kefeller | +0.177 |
itness | +0.176 |
ickson | +0.172 |
Neg logit contributions | |
REDACTED | -0.133 |
Clown | -0.116 |
Capitol | -0.116 |
Certain | -0.115 |
Franch | -0.110 |
Token | up |
Token ID | 510 |
Feature activation | +4.52e-01 |
Pos logit contributions | |
REDACTED | +0.867 |
Clown | +0.788 |
Capitol | +0.780 |
Certain | +0.768 |
Franch | +0.751 |
Neg logit contributions | |
ringe | -1.000 |
ritten | -0.966 |
kefeller | -0.903 |
itness | -0.886 |
ickson | -0.868 |
Token | for |
Token ID | 329 |
Feature activation | +7.22e-01 |
Pos logit contributions | |
REDACTED | +0.580 |
Clown | +0.522 |
Capitol | +0.522 |
Certain | +0.513 |
Franch | +0.501 |
Neg logit contributions | |
ringe | -0.634 |
ritten | -0.614 |
kefeller | -0.576 |
itness | -0.558 |
ickson | -0.549 |
Token | discussion |
Token ID | 5114 |
Feature activation | +1.92e+00 |
Pos logit contributions | |
REDACTED | +0.831 |
Clown | +0.748 |
Capitol | +0.739 |
Certain | +0.730 |
Franch | +0.711 |
Neg logit contributions | |
ringe | -1.111 |
ritten | -1.079 |
kefeller | -1.005 |
itness | -0.994 |
ickson | -0.969 |
Token | ( |
Token ID | 357 |
Feature activation | +8.04e-01 |
Pos logit contributions | |
REDACTED | +2.488 |
Clown | +2.221 |
Capitol | +2.211 |
Certain | +2.173 |
Franch | +2.118 |
Neg logit contributions | |
ringe | -2.683 |
ritten | -2.615 |
kefeller | -2.453 |
itness | -2.382 |
ickson | -2.317 |
Token | even |
Token ID | 10197 |
Feature activation | +9.72e-01 |
Pos logit contributions | |
REDACTED | +1.004 |
Clown | +0.879 |
Capitol | +0.874 |
Certain | +0.867 |
Franch | +0.837 |
Neg logit contributions | |
ringe | -1.168 |
ritten | -1.138 |
kefeller | -1.055 |
itness | -1.039 |
ickson | -1.013 |
Token | mang |
Token ID | 45663 |
Feature activation | +2.80e+00 |
Pos logit contributions | |
REDACTED | +1.184 |
Clown | +1.070 |
Capitol | +1.057 |
Certain | +1.040 |
Franch | +1.020 |
Neg logit contributions | |
ringe | -1.430 |
ritten | -1.388 |
kefeller | -1.287 |
itness | -1.260 |
ickson | -1.236 |
Token | achu |
Token ID | 32323 |
Feature activation | +2.27e-02 |
Pos logit contributions | |
REDACTED | +3.534 |
Clown | +3.179 |
Capitol | +3.122 |
Certain | +3.100 |
Assass | +3.025 |
Neg logit contributions | |
ringe | -3.852 |
ritten | -3.690 |
kefeller | -3.484 |
itness | -3.394 |
ickson | -3.274 |
Token | or |
Token ID | 393 |
Feature activation | -5.60e-01 |
Pos logit contributions | |
REDACTED | +0.028 |
Clown | +0.025 |
Capitol | +0.025 |
Certain | +0.025 |
Franch | +0.024 |
Neg logit contributions | |
ringe | -0.033 |
ritten | -0.032 |
kefeller | -0.029 |
itness | -0.029 |
ickson | -0.028 |
Token | bre |
Token ID | 1449 |
Feature activation | +7.84e-01 |
Pos logit contributions | |
ringe | +0.847 |
ritten | +0.826 |
itness | +0.755 |
kefeller | +0.752 |
ickson | +0.735 |
Neg logit contributions | |
REDACTED | -0.660 |
Clown | -0.594 |
Capitol | -0.586 |
Certain | -0.579 |
Franch | -0.563 |
Token | ade |
Token ID | 671 |
Feature activation | +1.45e+00 |
Pos logit contributions | |
REDACTED | +0.965 |
Clown | +0.871 |
Capitol | +0.860 |
Certain | +0.853 |
Franch | +0.830 |
Neg logit contributions | |
ringe | -1.112 |
ritten | -1.088 |
kefeller | -1.004 |
itness | -0.989 |
ickson | -0.972 |
Token | x |
Token ID | 87 |
Feature activation | -9.83e-02 |
Pos logit contributions | |
REDACTED | +1.777 |
Clown | +1.587 |
Capitol | +1.567 |
Certain | +1.547 |
Franch | +1.508 |
Neg logit contributions | |
ringe | -2.081 |
ritten | -2.039 |
kefeller | -1.887 |
itness | -1.857 |
ickson | -1.810 |
Token | lighting |
Token ID | 12019 |
Feature activation | +1.41e-01 |
Pos logit contributions | |
REDACTED | +0.026 |
Clown | +0.024 |
Capitol | +0.024 |
Certain | +0.023 |
Franch | +0.022 |
Neg logit contributions | |
ringe | -0.030 |
ritten | -0.029 |
kefeller | -0.027 |
itness | -0.027 |
ickson | -0.026 |
Token | and |
Token ID | 290 |
Feature activation | +1.30e-01 |
Pos logit contributions | |
REDACTED | +0.173 |
Clown | +0.159 |
Capitol | +0.158 |
Certain | +0.156 |
Franch | +0.151 |
Neg logit contributions | |
ringe | -0.202 |
ritten | -0.198 |
itness | -0.180 |
kefeller | -0.180 |
ickson | -0.175 |
Token | exposure |
Token ID | 7111 |
Feature activation | -5.75e-01 |
Pos logit contributions | |
REDACTED | +0.161 |
Clown | +0.147 |
Capitol | +0.147 |
Certain | +0.145 |
Franch | +0.140 |
Neg logit contributions | |
ringe | -0.187 |
ritten | -0.182 |
kefeller | -0.165 |
itness | -0.165 |
ickson | -0.161 |
Token | settings |
Token ID | 6460 |
Feature activation | +1.25e+00 |
Pos logit contributions | |
ringe | +0.835 |
ritten | +0.812 |
itness | +0.738 |
kefeller | +0.736 |
ickson | +0.719 |
Neg logit contributions | |
REDACTED | -0.708 |
Clown | -0.642 |
Capitol | -0.634 |
Certain | -0.629 |
Franch | -0.610 |
Token | ( |
Token ID | 357 |
Feature activation | +7.45e-01 |
Pos logit contributions | |
REDACTED | +1.449 |
Clown | +1.345 |
Certain | +1.332 |
Capitol | +1.325 |
Franch | +1.247 |
Neg logit contributions | |
ringe | -1.839 |
ritten | -1.812 |
kefeller | -1.664 |
itness | -1.650 |
ickson | -1.607 |
Token | mostly |
Token ID | 29471 |
Feature activation | +3.36e+00 |
Pos logit contributions | |
REDACTED | +0.923 |
Clown | +0.814 |
Capitol | +0.813 |
Certain | +0.809 |
Franch | +0.776 |
Neg logit contributions | |
ringe | -1.079 |
ritten | -1.053 |
kefeller | -0.965 |
itness | -0.960 |
ickson | -0.935 |
Token | to |
Token ID | 284 |
Feature activation | +3.93e-01 |
Pos logit contributions | |
REDACTED | +3.945 |
Capitol | +3.724 |
Clown | +3.664 |
Certain | +3.640 |
Franch | +3.478 |
Neg logit contributions | |
ringe | -4.839 |
ritten | -4.697 |
kefeller | -4.368 |
itness | -4.289 |
ickson | -4.165 |
Token | turn |
Token ID | 1210 |
Feature activation | +5.90e-01 |
Pos logit contributions | |
REDACTED | +0.484 |
Clown | +0.443 |
Capitol | +0.442 |
Certain | +0.435 |
Franch | +0.422 |
Neg logit contributions | |
ringe | -0.567 |
ritten | -0.553 |
kefeller | -0.503 |
itness | -0.501 |
ickson | -0.489 |
Token | off |
Token ID | 572 |
Feature activation | +1.05e+00 |
Pos logit contributions | |
REDACTED | +0.726 |
Clown | +0.670 |
Capitol | +0.664 |
Certain | +0.658 |
Franch | +0.628 |
Neg logit contributions | |
ringe | -0.850 |
ritten | -0.829 |
kefeller | -0.759 |
itness | -0.754 |
ickson | -0.733 |
Token | automatic |
Token ID | 11353 |
Feature activation | -7.67e-01 |
Pos logit contributions | |
REDACTED | +1.257 |
Clown | +1.154 |
Capitol | +1.146 |
Certain | +1.143 |
Franch | +1.091 |
Neg logit contributions | |
ringe | -1.541 |
ritten | -1.508 |
kefeller | -1.392 |
itness | -1.375 |
ickson | -1.341 |
Token | adjustments |
Token ID | 16895 |
Feature activation | +4.22e-01 |
Pos logit contributions | |
ringe | +1.060 |
ritten | +1.030 |
itness | +0.934 |
kefeller | +0.931 |
ickson | +0.910 |
Neg logit contributions | |
REDACTED | -1.001 |
Clown | -0.909 |
Capitol | -0.903 |
Certain | -0.893 |
Franch | -0.871 |
Token | pl |
Token ID | 458 |
Feature activation | +1.80e+00 |
Pos logit contributions | |
REDACTED | +0.259 |
Capitol | +0.235 |
Clown | +0.234 |
Certain | +0.230 |
Franch | +0.224 |
Neg logit contributions | |
ringe | -0.318 |
ritten | -0.308 |
itness | -0.283 |
kefeller | -0.281 |
ickson | -0.275 |
Token | enary |
Token ID | 21629 |
Feature activation | -1.37e+00 |
Pos logit contributions | |
REDACTED | +1.989 |
Capitol | +1.821 |
Certain | +1.790 |
Clown | +1.782 |
Franch | +1.750 |
Neg logit contributions | |
ringe | -2.757 |
ritten | -2.677 |
kefeller | -2.460 |
itness | -2.458 |
ickson | -2.373 |
Token | legislative |
Token ID | 10828 |
Feature activation | +1.57e+00 |
Pos logit contributions | |
ringe | +1.870 |
ritten | +1.811 |
itness | +1.640 |
kefeller | +1.637 |
ickson | +1.595 |
Neg logit contributions | |
REDACTED | -1.820 |
Clown | -1.653 |
Capitol | -1.649 |
Certain | -1.625 |
Franch | -1.592 |
Token | power |
Token ID | 1176 |
Feature activation | -1.06e-01 |
Pos logit contributions | |
REDACTED | +2.059 |
Capitol | +1.900 |
Clown | +1.864 |
Certain | +1.846 |
Franch | +1.836 |
Neg logit contributions | |
ringe | -2.163 |
ritten | -2.093 |
kefeller | -1.935 |
itness | -1.885 |
ickson | -1.823 |
Token | to |
Token ID | 284 |
Feature activation | +6.43e-01 |
Pos logit contributions | |
ringe | +0.153 |
ritten | +0.148 |
kefeller | +0.136 |
itness | +0.135 |
ickson | +0.131 |
Neg logit contributions | |
REDACTED | -0.132 |
Clown | -0.118 |
Capitol | -0.118 |
Certain | -0.116 |
Franch | -0.113 |
Token | legisl |
Token ID | 3560 |
Feature activation | +2.90e+00 |
Pos logit contributions | |
REDACTED | +0.706 |
Capitol | +0.641 |
Clown | +0.633 |
Certain | +0.627 |
Franch | +0.611 |
Neg logit contributions | |
ringe | -1.014 |
ritten | -0.992 |
kefeller | -0.914 |
itness | -0.905 |
ickson | -0.886 |
Token | ate |
Token ID | 378 |
Feature activation | +6.96e-01 |
Pos logit contributions | |
REDACTED | +2.422 |
Certain | +2.174 |
Capitol | +2.148 |
Clown | +2.108 |
Assass | +1.972 |
Neg logit contributions | |
ringe | -5.107 |
ritten | -5.058 |
kefeller | -4.756 |
itness | -4.629 |
ickson | -4.554 |
Token | on |
Token ID | 319 |
Feature activation | +1.35e+00 |
Pos logit contributions | |
REDACTED | +0.886 |
Capitol | +0.812 |
Clown | +0.805 |
Certain | +0.800 |
Franch | +0.775 |
Neg logit contributions | |
ringe | -0.977 |
ritten | -0.955 |
itness | -0.872 |
kefeller | -0.872 |
ickson | -0.835 |
Token | any |
Token ID | 597 |
Feature activation | +1.85e-01 |
Pos logit contributions | |
REDACTED | +1.395 |
Capitol | +1.284 |
Certain | +1.246 |
Clown | +1.235 |
Franch | +1.189 |
Neg logit contributions | |
ringe | -2.197 |
ritten | -2.143 |
kefeller | -1.985 |
itness | -1.984 |
ickson | -1.916 |
Token | subject |
Token ID | 2426 |
Feature activation | -6.13e-01 |
Pos logit contributions | |
REDACTED | +0.202 |
Capitol | +0.181 |
Clown | +0.178 |
Certain | +0.177 |
Franch | +0.171 |
Neg logit contributions | |
ringe | -0.295 |
ritten | -0.288 |
kefeller | -0.266 |
itness | -0.266 |
ickson | -0.259 |
Token | , |
Token ID | 11 |
Feature activation | +6.40e-02 |
Pos logit contributions | |
ringe | +0.862 |
ritten | +0.838 |
kefeller | +0.768 |
itness | +0.764 |
ickson | +0.740 |
Neg logit contributions | |
REDACTED | -0.786 |
Capitol | -0.709 |
Clown | -0.707 |
Certain | -0.699 |
Franch | -0.680 |
Token | the |
Token ID | 262 |
Feature activation | -2.76e-01 |
Pos logit contributions | |
ringe | +0.879 |
ritten | +0.854 |
itness | +0.780 |
kefeller | +0.773 |
ickson | +0.757 |
Neg logit contributions | |
REDACTED | -0.737 |
Clown | -0.664 |
Capitol | -0.662 |
Certain | -0.650 |
Franch | -0.635 |
Token | stra |
Token ID | 3534 |
Feature activation | -1.67e+00 |
Pos logit contributions | |
ringe | +0.407 |
ritten | +0.394 |
itness | +0.361 |
kefeller | +0.359 |
ickson | +0.351 |
Neg logit contributions | |
REDACTED | -0.337 |
Clown | -0.306 |
Capitol | -0.305 |
Certain | -0.298 |
Franch | -0.292 |
Token | it |
Token ID | 270 |
Feature activation | +8.04e-01 |
Pos logit contributions | |
ringe | +2.212 |
ritten | +2.145 |
itness | +1.935 |
kefeller | +1.920 |
ickson | +1.883 |
Neg logit contributions | |
REDACTED | -2.273 |
Clown | -2.063 |
Capitol | -2.045 |
Certain | -2.028 |
Franch | -1.984 |
Token | itself |
Token ID | 2346 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
REDACTED | +0.948 |
Clown | +0.852 |
Capitol | +0.847 |
Certain | +0.823 |
Franch | +0.805 |
Neg logit contributions | |
ringe | -1.202 |
ritten | -1.179 |
itness | -1.084 |
kefeller | -1.074 |
ickson | -1.044 |
Token | were |
Token ID | 547 |
Feature activation | -2.70e-01 |
Pos logit contributions | |
REDACTED | +1.495 |
Clown | +1.353 |
Capitol | +1.352 |
Certain | +1.298 |
Franch | +1.268 |
Neg logit contributions | |
ringe | -2.055 |
ritten | -2.006 |
kefeller | -1.848 |
itness | -1.844 |
ickson | -1.785 |
Token | reass |
Token ID | 12719 |
Feature activation | +3.07e+00 |
Pos logit contributions | |
ringe | +0.399 |
ritten | +0.387 |
itness | +0.354 |
kefeller | +0.351 |
ickson | +0.343 |
Neg logit contributions | |
REDACTED | -0.328 |
Clown | -0.294 |
Capitol | -0.294 |
Certain | -0.288 |
Franch | -0.282 |
Token | ert |
Token ID | 861 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
REDACTED | +4.160 |
Certain | +3.851 |
Clown | +3.850 |
Capitol | +3.838 |
Franch | +3.701 |
Neg logit contributions | |
ringe | -3.821 |
ritten | -3.546 |
kefeller | -3.319 |
itness | -3.282 |
ickson | -3.127 |
Token | ing |
Token ID | 278 |
Feature activation | -7.35e-01 |
Pos logit contributions | |
REDACTED | +1.263 |
Capitol | +1.130 |
Clown | +1.125 |
Certain | +1.112 |
Franch | +1.079 |
Neg logit contributions | |
ringe | -1.849 |
ritten | -1.785 |
itness | -1.663 |
kefeller | -1.647 |
ickson | -1.603 |
Token | the |
Token ID | 262 |
Feature activation | +2.60e-01 |
Pos logit contributions | |
ringe | +1.118 |
ritten | +1.084 |
itness | +0.997 |
kefeller | +0.985 |
ickson | +0.964 |
Neg logit contributions | |
REDACTED | -0.857 |
Capitol | -0.775 |
Clown | -0.773 |
Certain | -0.758 |
Franch | -0.737 |
Token | separat |
Token ID | 22439 |
Feature activation | +2.95e+00 |
Pos logit contributions | |
REDACTED | +0.320 |
Capitol | +0.294 |
Clown | +0.292 |
Certain | +0.287 |
Franch | +0.280 |
Neg logit contributions | |
ringe | -0.376 |
ritten | -0.362 |
itness | -0.332 |
kefeller | -0.331 |
ickson | -0.322 |
Token | eness |
Token ID | 9449 |
Feature activation | +5.55e-01 |
Pos logit contributions | |
REDACTED | +2.656 |
Clown | +2.319 |
Capitol | +2.291 |
Certain | +2.273 |
Franch | +2.167 |
Neg logit contributions | |
ringe | -5.079 |
ritten | -5.001 |
kefeller | -4.657 |
itness | -4.627 |
ickson | -4.594 |
Token | fishing |
Token ID | 12478 |
Feature activation | -9.12e-01 |
Pos logit contributions | |
REDACTED | +0.323 |
Clown | +0.293 |
Capitol | +0.293 |
Certain | +0.286 |
Franch | +0.278 |
Neg logit contributions | |
ringe | -0.438 |
ritten | -0.426 |
itness | -0.388 |
kefeller | -0.387 |
ickson | -0.379 |
Token | village |
Token ID | 7404 |
Feature activation | +2.49e+00 |
Pos logit contributions | |
ringe | +1.346 |
ritten | +1.310 |
kefeller | +1.196 |
itness | +1.193 |
ickson | +1.163 |
Neg logit contributions | |
REDACTED | -1.108 |
Clown | -1.002 |
Capitol | -0.996 |
Certain | -0.978 |
Franch | -0.961 |
Token | of |
Token ID | 286 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
REDACTED | +2.695 |
Capitol | +2.480 |
Clown | +2.455 |
Certain | +2.376 |
Franch | +2.302 |
Neg logit contributions | |
ringe | -3.950 |
ritten | -3.864 |
kefeller | -3.543 |
itness | -3.516 |
ickson | -3.430 |
Token | Gar |
Token ID | 7164 |
Feature activation | +1.29e+00 |
Pos logit contributions | |
REDACTED | +0.111 |
Clown | +0.099 |
Capitol | +0.099 |
Certain | +0.097 |
Franch | +0.094 |
Neg logit contributions | |
ringe | -0.177 |
ritten | -0.172 |
itness | -0.159 |
kefeller | -0.158 |
ickson | -0.155 |
Token | ach |
Token ID | 620 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
REDACTED | +1.661 |
Capitol | +1.547 |
Certain | +1.533 |
Clown | +1.527 |
Franch | +1.495 |
Neg logit contributions | |
ringe | -1.700 |
ritten | -1.678 |
kefeller | -1.517 |
itness | -1.512 |
ickson | -1.456 |
Token | ico |
Token ID | 3713 |
Feature activation | +1.38e+00 |
Pos logit contributions | |
REDACTED | +1.455 |
Capitol | +1.330 |
Clown | +1.330 |
Certain | +1.314 |
Franch | +1.280 |
Neg logit contributions | |
ringe | -1.653 |
ritten | -1.599 |
kefeller | -1.462 |
itness | -1.447 |
ickson | -1.396 |
Token | . |
Token ID | 13 |
Feature activation | +5.47e-01 |
Pos logit contributions | |
REDACTED | +1.707 |
Clown | +1.577 |
Capitol | +1.571 |
Certain | +1.546 |
Franch | +1.512 |
Neg logit contributions | |
ringe | -1.954 |
ritten | -1.896 |
kefeller | -1.718 |
itness | -1.705 |
ickson | -1.660 |
Token | It |
Token ID | 632 |
Feature activation | +1.18e+00 |
Pos logit contributions | |
REDACTED | +0.539 |
Clown | +0.488 |
Certain | +0.485 |
Capitol | +0.482 |
Franch | +0.463 |
Neg logit contributions | |
ringe | -0.930 |
ritten | -0.907 |
kefeller | -0.837 |
itness | -0.829 |
ickson | -0.816 |
Token | costs |
Token ID | 3484 |
Feature activation | +6.24e-02 |
Pos logit contributions | |
REDACTED | +1.385 |
Clown | +1.271 |
Capitol | +1.266 |
Certain | +1.235 |
Franch | +1.202 |
Neg logit contributions | |
ringe | -1.765 |
ritten | -1.703 |
kefeller | -1.556 |
itness | -1.547 |
ickson | -1.507 |
Token | from |
Token ID | 422 |
Feature activation | +1.13e+00 |
Pos logit contributions | |
REDACTED | +0.067 |
Clown | +0.060 |
Capitol | +0.059 |
Certain | +0.058 |
Franch | +0.057 |
Neg logit contributions | |
ringe | -0.101 |
ritten | -0.098 |
itness | -0.090 |
kefeller | -0.090 |
ickson | -0.088 |
Token | £ |
Token ID | 4248 |
Feature activation | -7.49e-01 |
Pos logit contributions | |
REDACTED | +1.129 |
Clown | +1.000 |
Capitol | +0.999 |
Certain | +0.981 |
Franch | +0.957 |
Neg logit contributions | |
ringe | -1.907 |
ritten | -1.851 |
itness | -1.713 |
kefeller | -1.705 |
ickson | -1.668 |
Token | W |
Token ID | 370 |
Feature activation | -1.60e+00 |
Pos logit contributions | |
REDACTED | +1.428 |
Clown | +1.322 |
Capitol | +1.307 |
Certain | +1.290 |
Franch | +1.252 |
Neg logit contributions | |
ringe | -2.309 |
ritten | -2.243 |
kefeller | -2.100 |
itness | -2.072 |
ickson | -2.014 |
Token | ., |
Token ID | 1539 |
Feature activation | +6.81e-01 |
Pos logit contributions | |
ringe | +1.989 |
ritten | +1.939 |
itness | +1.736 |
kefeller | +1.690 |
ickson | +1.683 |
Neg logit contributions | |
REDACTED | -2.332 |
Clown | -2.116 |
Capitol | -2.089 |
Certain | -2.081 |
Franch | -2.035 |
Token | Is |
Token ID | 1148 |
Feature activation | +6.13e-01 |
Pos logit contributions | |
REDACTED | +0.741 |
Clown | +0.668 |
Capitol | +0.661 |
Certain | +0.653 |
Franch | +0.639 |
Neg logit contributions | |
ringe | -1.085 |
ritten | -1.057 |
itness | -0.972 |
kefeller | -0.964 |
ickson | -0.945 |
Token | ava |
Token ID | 4170 |
Feature activation | +1.05e+00 |
Pos logit contributions | |
REDACTED | +0.775 |
Clown | +0.706 |
Capitol | +0.700 |
Certain | +0.694 |
Franch | +0.677 |
Neg logit contributions | |
ringe | -0.851 |
ritten | -0.833 |
itness | -0.755 |
kefeller | -0.746 |
ickson | -0.732 |
Token | , |
Token ID | 11 |
Feature activation | +3.94e-01 |
Pos logit contributions | |
REDACTED | +1.479 |
Clown | +1.348 |
Capitol | +1.337 |
Certain | +1.324 |
Franch | +1.299 |
Neg logit contributions | |
ringe | -1.348 |
ritten | -1.305 |
itness | -1.175 |
kefeller | -1.163 |
ickson | -1.139 |
Token | D |
Token ID | 360 |
Feature activation | -1.98e-02 |
Pos logit contributions | |
REDACTED | +0.408 |
Clown | +0.369 |
Capitol | +0.368 |
Certain | +0.361 |
Franch | +0.350 |
Neg logit contributions | |
ringe | -0.654 |
ritten | -0.635 |
kefeller | -0.591 |
itness | -0.588 |
ickson | -0.570 |
Token | . |
Token ID | 13 |
Feature activation | +1.38e+00 |
Pos logit contributions | |
ringe | +0.025 |
ritten | +0.024 |
itness | +0.022 |
kefeller | +0.022 |
ickson | +0.021 |
Neg logit contributions | |
REDACTED | -0.028 |
Clown | -0.026 |
Capitol | -0.025 |
Certain | -0.025 |
Franch | -0.025 |
Token | M |
Token ID | 337 |
Feature activation | +1.09e+00 |
Pos logit contributions | |
REDACTED | +1.418 |
Clown | +1.312 |
Capitol | +1.298 |
Certain | +1.281 |
Franch | +1.246 |
Neg logit contributions | |
ringe | -2.274 |
ritten | -2.209 |
kefeller | -2.067 |
itness | -2.038 |
ickson | -1.983 |
Token | . |
Token ID | 13 |
Feature activation | +1.49e+00 |
Pos logit contributions | |
REDACTED | +1.519 |
Clown | +1.381 |
Capitol | +1.369 |
Certain | +1.357 |
Franch | +1.329 |
Neg logit contributions | |
ringe | -1.409 |
ritten | -1.366 |
itness | -1.231 |
kefeller | -1.217 |
ickson | -1.195 |
Token | ( |
Token ID | 357 |
Feature activation | +9.27e-01 |
Pos logit contributions | |
REDACTED | +1.711 |
Clown | +1.573 |
Capitol | +1.565 |
Certain | +1.538 |
Franch | +1.506 |
Neg logit contributions | |
ringe | -2.275 |
ritten | -2.216 |
kefeller | -2.041 |
itness | -2.025 |
ickson | -1.970 |
Token | 2008 |
Token ID | 3648 |
Feature activation | +3.79e-01 |
Pos logit contributions | |
REDACTED | +0.836 |
Capitol | +0.727 |
Clown | +0.726 |
Certain | +0.710 |
Franch | +0.682 |
Neg logit contributions | |
ringe | -1.659 |
ritten | -1.611 |
kefeller | -1.506 |
itness | -1.501 |
ickson | -1.466 |
Token | responsibility |
Token ID | 5798 |
Feature activation | -1.39e+00 |
Pos logit contributions | |
REDACTED | +0.253 |
Clown | +0.230 |
Capitol | +0.230 |
Certain | +0.225 |
Franch | +0.220 |
Neg logit contributions | |
ringe | -0.296 |
ritten | -0.286 |
kefeller | -0.261 |
itness | -0.259 |
ickson | -0.254 |
Token | for |
Token ID | 329 |
Feature activation | +8.92e-01 |
Pos logit contributions | |
ringe | +1.906 |
ritten | +1.850 |
itness | +1.674 |
kefeller | +1.665 |
ickson | +1.628 |
Neg logit contributions | |
REDACTED | -1.824 |
Clown | -1.650 |
Capitol | -1.636 |
Certain | -1.619 |
Franch | -1.583 |
Token | that |
Token ID | 326 |
Feature activation | -4.69e-01 |
Pos logit contributions | |
REDACTED | +1.071 |
Clown | +0.977 |
Capitol | +0.971 |
Certain | +0.948 |
Franch | +0.928 |
Neg logit contributions | |
ringe | -1.332 |
ritten | -1.297 |
kefeller | -1.192 |
itness | -1.167 |
ickson | -1.144 |
Token | . |
Token ID | 13 |
Feature activation | +9.72e-02 |
Pos logit contributions | |
ringe | +0.653 |
ritten | +0.633 |
kefeller | +0.575 |
itness | +0.571 |
ickson | +0.556 |
Neg logit contributions | |
REDACTED | -0.614 |
Clown | -0.556 |
Capitol | -0.553 |
Certain | -0.546 |
Franch | -0.534 |
Token | � |
Token ID | 447 |
Feature activation | -1.52e-02 |
Pos logit contributions | |
REDACTED | +0.130 |
Clown | +0.119 |
Capitol | +0.118 |
Certain | +0.117 |
Franch | +0.114 |
Neg logit contributions | |
ringe | -0.132 |
ritten | -0.128 |
kefeller | -0.117 |
itness | -0.116 |
ickson | -0.113 |
Token | � |
Token ID | 251 |
Feature activation | +4.57e-01 |
Pos logit contributions | |
ringe | +0.025 |
ritten | +0.024 |
kefeller | +0.022 |
itness | +0.022 |
ickson | +0.022 |
Neg logit contributions | |
REDACTED | -0.016 |
Clown | -0.014 |
Certain | -0.014 |
Capitol | -0.014 |
Franch | -0.013 |
Token |
|
Token ID | 198 |
Feature activation | +4.38e-01 |
Pos logit contributions | |
REDACTED | +0.504 |
Clown | +0.452 |
Capitol | +0.450 |
Certain | +0.446 |
Franch | +0.432 |
Neg logit contributions | |
ringe | -0.727 |
ritten | -0.710 |
kefeller | -0.654 |
itness | -0.648 |
ickson | -0.636 |
Token |
|
Token ID | 198 |
Feature activation | +3.87e-01 |
Pos logit contributions | |
REDACTED | +0.466 |
Capitol | +0.409 |
Clown | +0.406 |
Certain | +0.403 |
Franch | +0.386 |
Neg logit contributions | |
ringe | -0.718 |
ritten | -0.700 |
kefeller | -0.652 |
itness | -0.645 |
ickson | -0.628 |
Token | Ryan |
Token ID | 21868 |
Feature activation | +7.94e-01 |
Pos logit contributions | |
REDACTED | +0.484 |
Capitol | +0.427 |
Clown | +0.426 |
Certain | +0.423 |
Franch | +0.404 |
Neg logit contributions | |
ringe | -0.560 |
ritten | -0.546 |
kefeller | -0.505 |
itness | -0.498 |
ickson | -0.486 |
Token | said |
Token ID | 531 |
Feature activation | -6.10e-01 |
Pos logit contributions | |
REDACTED | +1.033 |
Clown | +0.949 |
Capitol | +0.943 |
Certain | +0.920 |
Franch | +0.915 |
Neg logit contributions | |
ringe | -1.095 |
ritten | -1.067 |
kefeller | -0.968 |
itness | -0.964 |
ickson | -0.934 |
Token | he |
Token ID | 339 |
Feature activation | +1.11e+00 |
Pos logit contributions | |
ringe | +0.863 |
ritten | +0.838 |
itness | +0.760 |
kefeller | +0.757 |
ickson | +0.741 |
Neg logit contributions | |
REDACTED | -0.781 |
Clown | -0.709 |
Capitol | -0.704 |
Certain | -0.693 |
Franch | -0.680 |
Token | insky |
Token ID | 19870 |
Feature activation | -1.31e+00 |
Pos logit contributions | |
REDACTED | +0.286 |
Clown | +0.260 |
Capitol | +0.259 |
Certain | +0.257 |
Franch | +0.251 |
Neg logit contributions | |
ringe | -0.338 |
ritten | -0.327 |
itness | -0.298 |
kefeller | -0.297 |
ickson | -0.292 |
Token | Ag |
Token ID | 2449 |
Feature activation | +1.82e+00 |
Pos logit contributions | |
ringe | +1.922 |
ritten | +1.866 |
itness | +1.705 |
kefeller | +1.694 |
ickson | +1.663 |
Neg logit contributions | |
REDACTED | -1.599 |
Clown | -1.443 |
Capitol | -1.430 |
Certain | -1.417 |
Franch | -1.384 |
Token | rop |
Token ID | 1773 |
Feature activation | +2.09e-01 |
Pos logit contributions | |
REDACTED | +2.433 |
Capitol | +2.265 |
Certain | +2.229 |
Clown | +2.228 |
Franch | +2.166 |
Neg logit contributions | |
ringe | -2.326 |
ritten | -2.277 |
kefeller | -2.040 |
itness | -2.026 |
ickson | -1.988 |
Token | ost |
Token ID | 455 |
Feature activation | -1.71e+00 |
Pos logit contributions | |
REDACTED | +0.265 |
Clown | +0.240 |
Capitol | +0.240 |
Certain | +0.237 |
Franch | +0.233 |
Neg logit contributions | |
ringe | -0.291 |
ritten | -0.283 |
kefeller | -0.255 |
itness | -0.255 |
ickson | -0.251 |
Token | ach |
Token ID | 620 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
ringe | +2.526 |
ritten | +2.458 |
itness | +2.245 |
kefeller | +2.227 |
ickson | +2.191 |
Neg logit contributions | |
REDACTED | -2.054 |
Clown | -1.841 |
Capitol | -1.826 |
Certain | -1.806 |
Franch | -1.766 |
Token | , |
Token ID | 11 |
Feature activation | +6.41e-01 |
Pos logit contributions | |
REDACTED | +1.402 |
Capitol | +1.309 |
Clown | +1.300 |
Certain | +1.286 |
Franch | +1.258 |
Neg logit contributions | |
ringe | -1.611 |
ritten | -1.546 |
kefeller | -1.414 |
itness | -1.409 |
ickson | -1.369 |
Token | is |
Token ID | 318 |
Feature activation | +8.65e-01 |
Pos logit contributions | |
REDACTED | +0.763 |
Clown | +0.698 |
Capitol | +0.695 |
Certain | +0.676 |
Franch | +0.669 |
Neg logit contributions | |
ringe | -0.956 |
ritten | -0.927 |
itness | -0.847 |
kefeller | -0.840 |
ickson | -0.820 |
Token | a |
Token ID | 257 |
Feature activation | +1.99e-01 |
Pos logit contributions | |
REDACTED | +1.059 |
Clown | +0.972 |
Capitol | +0.970 |
Certain | +0.948 |
Franch | +0.938 |
Neg logit contributions | |
ringe | -1.251 |
ritten | -1.213 |
itness | -1.106 |
kefeller | -1.102 |
ickson | -1.074 |
Token | recipient |
Token ID | 17800 |
Feature activation | -1.52e+00 |
Pos logit contributions | |
REDACTED | +0.240 |
Clown | +0.220 |
Capitol | +0.219 |
Certain | +0.214 |
Franch | +0.213 |
Neg logit contributions | |
ringe | -0.293 |
ritten | -0.283 |
itness | -0.259 |
kefeller | -0.257 |
ickson | -0.252 |
Token | of |
Token ID | 286 |
Feature activation | -2.16e-01 |
Pos logit contributions | |
ringe | +2.132 |
ritten | +2.077 |
kefeller | +1.881 |
itness | +1.873 |
ickson | +1.842 |
Neg logit contributions | |
REDACTED | -1.943 |
Clown | -1.738 |
Capitol | -1.716 |
Certain | -1.714 |
Franch | -1.656 |
Token | $ |
Token ID | 720 |
Feature activation | +9.15e-02 |
Pos logit contributions | |
ringe | +0.323 |
ritten | +0.314 |
itness | +0.288 |
kefeller | +0.284 |
ickson | +0.280 |
Neg logit contributions | |
REDACTED | -0.257 |
Clown | -0.231 |
Capitol | -0.229 |
Certain | -0.226 |
Franch | -0.222 |
Token | Game |
Token ID | 3776 |
Feature activation | +1.07e+00 |
Pos logit contributions | |
ringe | +0.585 |
ritten | +0.567 |
itness | +0.519 |
kefeller | +0.516 |
ickson | +0.505 |
Neg logit contributions | |
REDACTED | -0.467 |
Clown | -0.428 |
Capitol | -0.424 |
Certain | -0.417 |
Franch | -0.406 |
Token | Stop |
Token ID | 19485 |
Feature activation | -5.84e-01 |
Pos logit contributions | |
REDACTED | +1.375 |
Clown | +1.267 |
Capitol | +1.248 |
Certain | +1.237 |
Franch | +1.210 |
Neg logit contributions | |
ringe | -1.497 |
ritten | -1.462 |
itness | -1.334 |
kefeller | -1.310 |
ickson | -1.280 |
Token | 's |
Token ID | 338 |
Feature activation | -3.17e-01 |
Pos logit contributions | |
ringe | +0.795 |
ritten | +0.776 |
itness | +0.703 |
kefeller | +0.697 |
ickson | +0.680 |
Neg logit contributions | |
REDACTED | -0.767 |
Clown | -0.716 |
Capitol | -0.714 |
Certain | -0.699 |
Franch | -0.681 |
Token | website |
Token ID | 3052 |
Feature activation | -1.88e+00 |
Pos logit contributions | |
ringe | +0.455 |
ritten | +0.443 |
kefeller | +0.405 |
itness | +0.404 |
ickson | +0.394 |
Neg logit contributions | |
REDACTED | -0.391 |
Clown | -0.365 |
Capitol | -0.362 |
Certain | -0.356 |
Franch | -0.346 |
Token | amid |
Token ID | 10371 |
Feature activation | -5.92e-01 |
Pos logit contributions | |
ringe | +2.618 |
ritten | +2.548 |
itness | +2.314 |
kefeller | +2.305 |
ickson | +2.251 |
Neg logit contributions | |
REDACTED | -2.441 |
Clown | -2.222 |
Capitol | -2.208 |
Certain | -2.178 |
Franch | -2.119 |
Token | the |
Token ID | 262 |
Feature activation | -1.51e-01 |
Pos logit contributions | |
ringe | +0.892 |
ritten | +0.868 |
itness | +0.794 |
kefeller | +0.791 |
ickson | +0.778 |
Neg logit contributions | |
REDACTED | -0.699 |
Clown | -0.630 |
Capitol | -0.629 |
Certain | -0.615 |
Franch | -0.601 |
Token | rush |
Token ID | 10484 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
ringe | +0.217 |
ritten | +0.211 |
kefeller | +0.192 |
itness | +0.192 |
ickson | +0.187 |
Neg logit contributions | |
REDACTED | -0.190 |
Clown | -0.175 |
Capitol | -0.174 |
Certain | -0.170 |
Franch | -0.167 |
Token | to |
Token ID | 284 |
Feature activation | +2.05e-01 |
Pos logit contributions | |
REDACTED | +1.582 |
Clown | +1.443 |
Capitol | +1.438 |
Certain | +1.403 |
Franch | +1.367 |
Neg logit contributions | |
ringe | -1.988 |
ritten | -1.915 |
itness | -1.751 |
kefeller | -1.750 |
ickson | -1.705 |
Token | secure |
Token ID | 5713 |
Feature activation | -8.84e-01 |
Pos logit contributions | |
REDACTED | +0.248 |
Clown | +0.228 |
Capitol | +0.227 |
Certain | +0.223 |
Franch | +0.218 |
Neg logit contributions | |
ringe | -0.301 |
ritten | -0.292 |
itness | -0.265 |
kefeller | -0.264 |
ickson | -0.259 |
Token | plastic |
Token ID | 7309 |
Feature activation | -2.12e+00 |
Pos logit contributions | |
ringe | +1.313 |
ritten | +1.279 |
itness | +1.164 |
kefeller | +1.156 |
ickson | +1.136 |
Neg logit contributions | |
REDACTED | -1.067 |
Clown | -0.969 |
Capitol | -0.962 |
Certain | -0.950 |
Franch | -0.928 |
Token | versions |
Token ID | 6300 |
Feature activation | +1.23e+00 |
Pos logit contributions | |
ringe | +2.979 |
ritten | +2.892 |
itness | +2.629 |
kefeller | +2.605 |
ickson | +2.558 |
Neg logit contributions | |
REDACTED | -2.730 |
Clown | -2.473 |
Capitol | -2.451 |
Certain | -2.423 |
Franch | -2.376 |
Token | young |
Token ID | 1862 |
Feature activation | +1.48e+00 |
Pos logit contributions | |
REDACTED | +0.516 |
Clown | +0.471 |
Capitol | +0.469 |
Certain | +0.463 |
Franch | +0.450 |
Neg logit contributions | |
ringe | -0.543 |
ritten | -0.520 |
itness | -0.475 |
kefeller | -0.472 |
ickson | -0.462 |
Token | examples |
Token ID | 6096 |
Feature activation | +9.44e-01 |
Pos logit contributions | |
REDACTED | +1.915 |
Clown | +1.763 |
Capitol | +1.753 |
Certain | +1.705 |
Franch | +1.686 |
Neg logit contributions | |
ringe | -2.056 |
ritten | -1.978 |
itness | -1.801 |
kefeller | -1.787 |
ickson | -1.739 |
Token | of |
Token ID | 286 |
Feature activation | -6.01e-01 |
Pos logit contributions | |
REDACTED | +1.077 |
Clown | +0.978 |
Capitol | +0.967 |
Certain | +0.949 |
Franch | +0.925 |
Neg logit contributions | |
ringe | -1.452 |
ritten | -1.417 |
kefeller | -1.289 |
itness | -1.285 |
ickson | -1.255 |
Token | Ant |
Token ID | 3738 |
Feature activation | +1.52e+00 |
Pos logit contributions | |
ringe | +0.842 |
ritten | +0.815 |
itness | +0.741 |
kefeller | +0.735 |
ickson | +0.720 |
Neg logit contributions | |
REDACTED | -0.776 |
Clown | -0.706 |
Capitol | -0.702 |
Certain | -0.693 |
Franch | -0.677 |
Token | ill |
Token ID | 359 |
Feature activation | +2.10e+00 |
Pos logit contributions | |
REDACTED | +1.738 |
Capitol | +1.554 |
Clown | +1.548 |
Certain | +1.547 |
Franch | +1.482 |
Neg logit contributions | |
ringe | -2.264 |
ritten | -2.207 |
itness | -2.025 |
kefeller | -2.024 |
ickson | -1.913 |
Token | oth |
Token ID | 849 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
REDACTED | +2.941 |
Capitol | +2.730 |
Certain | +2.706 |
Clown | +2.691 |
inct | +2.580 |
Neg logit contributions | |
ringe | -2.540 |
ritten | -2.482 |
itness | -2.218 |
kefeller | -2.210 |
ickson | -2.081 |
Token | rix |
Token ID | 8609 |
Feature activation | +1.64e+00 |
Pos logit contributions | |
ringe | +2.577 |
ritten | +2.520 |
kefeller | +2.361 |
itness | +2.359 |
ickson | +2.307 |
Neg logit contributions | |
REDACTED | -1.054 |
Clown | -0.880 |
Capitol | -0.874 |
Certain | -0.852 |
Franch | -0.815 |
Token | bones |
Token ID | 11945 |
Feature activation | +1.30e+00 |
Pos logit contributions | |
REDACTED | +2.035 |
Clown | +1.876 |
Capitol | +1.864 |
Certain | +1.830 |
Franch | +1.771 |
Neg logit contributions | |
ringe | -2.348 |
ritten | -2.272 |
kefeller | -2.049 |
itness | -2.047 |
ickson | -1.979 |
Token | to |
Token ID | 284 |
Feature activation | +4.32e-01 |
Pos logit contributions | |
REDACTED | +1.573 |
Clown | +1.430 |
Capitol | +1.425 |
Certain | +1.405 |
Franch | +1.351 |
Neg logit contributions | |
ringe | -1.889 |
ritten | -1.834 |
itness | -1.661 |
kefeller | -1.659 |
ickson | -1.627 |
Token | the |
Token ID | 262 |
Feature activation | +1.41e-01 |
Pos logit contributions | |
REDACTED | +0.530 |
Clown | +0.483 |
Capitol | +0.481 |
Certain | +0.473 |
Franch | +0.464 |
Neg logit contributions | |
ringe | -0.630 |
ritten | -0.610 |
itness | -0.554 |
kefeller | -0.552 |
ickson | -0.538 |
Token | newly |
Token ID | 8308 |
Feature activation | +1.21e+00 |
Pos logit contributions | |
REDACTED | +0.185 |
Clown | +0.170 |
Capitol | +0.169 |
Certain | +0.165 |
Franch | +0.163 |
Neg logit contributions | |
ringe | -0.194 |
ritten | -0.187 |
itness | -0.170 |
kefeller | -0.169 |
ickson | -0.164 |
Token | osexual |
Token ID | 8542 |
Feature activation | -1.53e-01 |
Pos logit contributions | |
REDACTED | +1.775 |
Clown | +1.648 |
Capitol | +1.630 |
Certain | +1.618 |
Franch | +1.570 |
Neg logit contributions | |
ringe | -1.477 |
ritten | -1.431 |
itness | -1.287 |
kefeller | -1.279 |
ickson | -1.240 |
Token | " |
Token ID | 1 |
Feature activation | -5.32e-02 |
Pos logit contributions | |
ringe | +0.200 |
ritten | +0.194 |
itness | +0.175 |
kefeller | +0.173 |
ickson | +0.170 |
Neg logit contributions | |
REDACTED | -0.213 |
Clown | -0.193 |
Capitol | -0.191 |
Certain | -0.189 |
Franch | -0.186 |
Token | only |
Token ID | 691 |
Feature activation | +3.42e-01 |
Pos logit contributions | |
ringe | +0.079 |
ritten | +0.078 |
itness | +0.071 |
kefeller | +0.071 |
ickson | +0.069 |
Neg logit contributions | |
REDACTED | -0.063 |
Clown | -0.057 |
Capitol | -0.057 |
Certain | -0.056 |
Franch | -0.054 |
Token | apply |
Token ID | 4174 |
Feature activation | +5.40e-01 |
Pos logit contributions | |
REDACTED | +0.412 |
Clown | +0.374 |
Capitol | +0.373 |
Certain | +0.368 |
Franch | +0.358 |
Neg logit contributions | |
ringe | -0.499 |
ritten | -0.486 |
kefeller | -0.452 |
itness | -0.448 |
ickson | -0.436 |
Token | in |
Token ID | 287 |
Feature activation | -1.54e-01 |
Pos logit contributions | |
REDACTED | +0.640 |
Clown | +0.577 |
Capitol | +0.572 |
Certain | +0.565 |
Franch | +0.553 |
Neg logit contributions | |
ringe | -0.805 |
ritten | -0.784 |
itness | -0.721 |
kefeller | -0.717 |
ickson | -0.702 |
Token | describing |
Token ID | 12059 |
Feature activation | -2.51e-01 |
Pos logit contributions | |
ringe | +0.225 |
ritten | +0.219 |
itness | +0.201 |
kefeller | +0.200 |
ickson | +0.195 |
Neg logit contributions | |
REDACTED | -0.189 |
Clown | -0.170 |
Capitol | -0.169 |
Certain | -0.167 |
Franch | -0.163 |
Token | those |
Token ID | 883 |
Feature activation | +3.09e-01 |
Pos logit contributions | |
ringe | +0.407 |
ritten | +0.396 |
itness | +0.365 |
kefeller | +0.364 |
ickson | +0.356 |
Neg logit contributions | |
REDACTED | -0.268 |
Clown | -0.239 |
Capitol | -0.237 |
Certain | -0.233 |
Franch | -0.227 |
Token | who |
Token ID | 508 |
Feature activation | +4.03e-01 |
Pos logit contributions | |
REDACTED | +0.386 |
Clown | +0.352 |
Capitol | +0.351 |
Certain | +0.345 |
Franch | +0.337 |
Neg logit contributions | |
ringe | -0.444 |
ritten | -0.432 |
itness | -0.395 |
kefeller | -0.394 |
ickson | -0.381 |
Token | " |
Token ID | 366 |
Feature activation | -1.47e+00 |
Pos logit contributions | |
REDACTED | +0.472 |
Clown | +0.431 |
Capitol | +0.428 |
Certain | +0.425 |
Franch | +0.410 |
Neg logit contributions | |
ringe | -0.598 |
ritten | -0.590 |
kefeller | -0.548 |
itness | -0.535 |
ickson | -0.527 |
Token | act |
Token ID | 529 |
Feature activation | +8.36e-01 |
Pos logit contributions | |
ringe | +2.088 |
ritten | +2.031 |
itness | +1.849 |
kefeller | +1.839 |
ickson | +1.803 |
Neg logit contributions | |
REDACTED | -1.850 |
Clown | -1.661 |
Capitol | -1.646 |
Certain | -1.632 |
Franch | -1.591 |
Token | " |
Token ID | 1 |
Feature activation | -2.02e-01 |
Pos logit contributions | |
REDACTED | +1.017 |
Clown | +0.930 |
Capitol | +0.921 |
Certain | +0.915 |
Franch | +0.887 |
Neg logit contributions | |
ringe | -1.210 |
ritten | -1.191 |
kefeller | -1.086 |
itness | -1.076 |
ickson | -1.045 |
Token | 't |
Token ID | 470 |
Feature activation | +1.47e-01 |
Pos logit contributions | |
REDACTED | +0.272 |
Capitol | +0.247 |
Clown | +0.246 |
Certain | +0.244 |
Franch | +0.236 |
Neg logit contributions | |
ringe | -0.259 |
ritten | -0.250 |
kefeller | -0.229 |
itness | -0.227 |
ickson | -0.218 |
Token | want |
Token ID | 765 |
Feature activation | -1.90e-01 |
Pos logit contributions | |
REDACTED | +0.156 |
Clown | +0.141 |
Capitol | +0.138 |
Certain | +0.137 |
Franch | +0.133 |
Neg logit contributions | |
ringe | -0.237 |
ritten | -0.234 |
kefeller | -0.217 |
itness | -0.212 |
ickson | -0.210 |
Token | to |
Token ID | 284 |
Feature activation | -3.02e-01 |
Pos logit contributions | |
ringe | +0.277 |
ritten | +0.269 |
itness | +0.245 |
kefeller | +0.244 |
ickson | +0.239 |
Neg logit contributions | |
REDACTED | -0.234 |
Clown | -0.211 |
Capitol | -0.209 |
Certain | -0.206 |
Franch | -0.202 |
Token | screw |
Token ID | 9580 |
Feature activation | -4.61e+00 |
Pos logit contributions | |
ringe | +0.445 |
ritten | +0.435 |
kefeller | +0.407 |
itness | +0.394 |
ickson | +0.391 |
Neg logit contributions | |
REDACTED | -0.361 |
Clown | -0.336 |
Capitol | -0.326 |
Certain | -0.325 |
Franch | -0.315 |
Token | it |
Token ID | 340 |
Feature activation | +1.62e-01 |
Pos logit contributions | |
ringe | +6.023 |
ritten | +5.933 |
itness | +5.315 |
kefeller | +5.215 |
ickson | +5.209 |
Neg logit contributions | |
REDACTED | -6.137 |
Clown | -5.475 |
Capitol | -5.440 |
Certain | -5.418 |
Franch | -5.271 |
Token | up |
Token ID | 510 |
Feature activation | -3.97e-01 |
Pos logit contributions | |
REDACTED | +0.189 |
Clown | +0.173 |
Capitol | +0.171 |
Certain | +0.168 |
Franch | +0.163 |
Neg logit contributions | |
ringe | -0.245 |
ritten | -0.238 |
itness | -0.220 |
kefeller | -0.219 |
ickson | -0.212 |
Token | . |
Token ID | 13 |
Feature activation | -5.23e-01 |
Pos logit contributions | |
ringe | +0.565 |
ritten | +0.549 |
kefeller | +0.505 |
itness | +0.500 |
ickson | +0.486 |
Neg logit contributions | |
REDACTED | -0.502 |
Clown | -0.460 |
Capitol | -0.453 |
Certain | -0.448 |
Franch | -0.435 |
Token | I |
Token ID | 314 |
Feature activation | +1.10e-01 |
Pos logit contributions | |
ringe | +0.874 |
ritten | +0.853 |
itness | +0.786 |
kefeller | +0.786 |
ickson | +0.769 |
Neg logit contributions | |
REDACTED | -0.536 |
Clown | -0.478 |
Capitol | -0.470 |
Certain | -0.469 |
Franch | -0.452 |
Token | want |
Token ID | 765 |
Feature activation | -7.18e-02 |
Pos logit contributions | |
REDACTED | +0.120 |
Clown | +0.108 |
Capitol | +0.107 |
Certain | +0.106 |
Franch | +0.104 |
Neg logit contributions | |
ringe | -0.173 |
ritten | -0.169 |
kefeller | -0.157 |
itness | -0.155 |
ickson | -0.153 |
Token | to |
Token ID | 284 |
Feature activation | -3.34e-01 |
Pos logit contributions | |
ringe | +0.103 |
ritten | +0.100 |
itness | +0.091 |
kefeller | +0.090 |
ickson | +0.089 |
Neg logit contributions | |
REDACTED | -0.090 |
Clown | -0.081 |
Capitol | -0.080 |
Certain | -0.079 |
Franch | -0.077 |
Token | -" |
Token ID | 21215 |
Feature activation | -2.11e+00 |
Pos logit contributions | |
ringe | +0.501 |
ritten | +0.483 |
kefeller | +0.448 |
itness | +0.437 |
ickson | +0.432 |
Neg logit contributions | |
REDACTED | -0.400 |
Clown | -0.367 |
Capitol | -0.361 |
Certain | -0.357 |
Franch | -0.349 |
Token | ly |
Token ID | 306 |
Feature activation | -4.08e-01 |
Pos logit contributions | |
ringe | +0.852 |
ritten | +0.828 |
itness | +0.751 |
kefeller | +0.747 |
ickson | +0.733 |
Neg logit contributions | |
REDACTED | -0.765 |
Clown | -0.695 |
Capitol | -0.687 |
Certain | -0.682 |
Franch | -0.664 |
Token | after |
Token ID | 706 |
Feature activation | -2.99e-01 |
Pos logit contributions | |
ringe | +0.566 |
ritten | +0.553 |
itness | +0.499 |
kefeller | +0.499 |
ickson | +0.486 |
Neg logit contributions | |
REDACTED | -0.530 |
Clown | -0.482 |
Capitol | -0.478 |
Certain | -0.473 |
Franch | -0.461 |
Token | 6 |
Token ID | 718 |
Feature activation | -1.16e-01 |
Pos logit contributions | |
ringe | +0.447 |
ritten | +0.437 |
itness | +0.396 |
kefeller | +0.396 |
ickson | +0.386 |
Neg logit contributions | |
REDACTED | -0.358 |
Clown | -0.324 |
Capitol | -0.320 |
Certain | -0.316 |
Franch | -0.308 |
Token | . |
Token ID | 13 |
Feature activation | +5.49e-01 |
Pos logit contributions | |
ringe | +0.147 |
ritten | +0.143 |
kefeller | +0.129 |
itness | +0.127 |
ickson | +0.124 |
Neg logit contributions | |
REDACTED | -0.163 |
Clown | -0.148 |
Capitol | -0.148 |
Certain | -0.147 |
Franch | -0.143 |
Token | With |
Token ID | 2080 |
Feature activation | +6.90e-01 |
Pos logit contributions | |
REDACTED | +0.555 |
Clown | +0.492 |
Capitol | +0.486 |
Certain | +0.485 |
Franch | +0.465 |
Neg logit contributions | |
ringe | -0.922 |
ritten | -0.904 |
kefeller | -0.834 |
itness | -0.828 |
ickson | -0.813 |
Token | EQ |
Token ID | 36529 |
Feature activation | -2.63e+00 |
Pos logit contributions | |
REDACTED | +0.826 |
Clown | +0.748 |
Capitol | +0.741 |
Certain | +0.728 |
Franch | +0.710 |
Neg logit contributions | |
ringe | -1.027 |
ritten | -1.000 |
kefeller | -0.911 |
itness | -0.909 |
ickson | -0.890 |
Token | inv |
Token ID | 800 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
ringe | +3.839 |
ritten | +3.736 |
itness | +3.431 |
kefeller | +3.355 |
ickson | +3.336 |
Neg logit contributions | |
REDACTED | -3.171 |
Clown | -2.832 |
Capitol | -2.796 |
Certain | -2.779 |
Franch | -2.717 |
Token | oker |
Token ID | 11020 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
ringe | +1.813 |
ritten | +1.762 |
itness | +1.592 |
kefeller | +1.580 |
ickson | +1.543 |
Neg logit contributions | |
REDACTED | -1.827 |
Clown | -1.656 |
Capitol | -1.647 |
Certain | -1.630 |
Franch | -1.594 |
Token | on |
Token ID | 319 |
Feature activation | +9.38e-01 |
Pos logit contributions | |
ringe | +1.594 |
ritten | +1.549 |
itness | +1.409 |
kefeller | +1.402 |
ickson | +1.370 |
Neg logit contributions | |
REDACTED | -1.425 |
Clown | -1.287 |
Capitol | -1.275 |
Certain | -1.260 |
Franch | -1.232 |
Token | the |
Token ID | 262 |
Feature activation | -1.77e-01 |
Pos logit contributions | |
REDACTED | +1.159 |
Capitol | +1.048 |
Clown | +1.047 |
Certain | +1.023 |
Franch | +0.999 |
Neg logit contributions | |
ringe | -1.362 |
ritten | -1.325 |
kefeller | -1.206 |
itness | -1.204 |
ickson | -1.175 |
Token | other |
Token ID | 584 |
Feature activation | -7.06e-01 |
Pos logit contributions | |
ringe | +0.247 |
ritten | +0.242 |
itness | +0.219 |
kefeller | +0.218 |
ickson | +0.214 |
Neg logit contributions | |
REDACTED | -0.226 |
Clown | -0.206 |
Capitol | -0.205 |
Certain | -0.201 |
Franch | -0.197 |
Token | 30 |
Token ID | 1270 |
Feature activation | -1.01e+00 |
Pos logit contributions | |
ringe | +0.341 |
ritten | +0.333 |
kefeller | +0.307 |
itness | +0.307 |
ickson | +0.301 |
Neg logit contributions | |
REDACTED | -0.208 |
Clown | -0.182 |
Capitol | -0.181 |
Certain | -0.178 |
Franch | -0.171 |
Token | p |
Token ID | 279 |
Feature activation | -5.77e-01 |
Pos logit contributions | |
ringe | +1.385 |
ritten | +1.342 |
kefeller | +1.215 |
itness | +1.210 |
ickson | +1.181 |
Neg logit contributions | |
REDACTED | -1.336 |
Clown | -1.213 |
Capitol | -1.204 |
Certain | -1.191 |
Franch | -1.162 |
Token | . |
Token ID | 13 |
Feature activation | +8.94e-01 |
Pos logit contributions | |
ringe | +0.798 |
ritten | +0.772 |
kefeller | +0.712 |
itness | +0.703 |
ickson | +0.679 |
Neg logit contributions | |
REDACTED | -0.763 |
Capitol | -0.688 |
Clown | -0.683 |
Certain | -0.674 |
Franch | -0.654 |
Token | m |
Token ID | 76 |
Feature activation | -3.46e-01 |
Pos logit contributions | |
REDACTED | +1.146 |
Capitol | +1.035 |
Clown | +1.013 |
Certain | +1.012 |
Franch | +0.963 |
Neg logit contributions | |
ringe | -1.255 |
ritten | -1.226 |
kefeller | -1.150 |
itness | -1.111 |
ickson | -1.081 |
Token | . |
Token ID | 13 |
Feature activation | +8.83e-01 |
Pos logit contributions | |
ringe | +0.437 |
ritten | +0.424 |
itness | +0.383 |
kefeller | +0.375 |
ickson | +0.370 |
Neg logit contributions | |
REDACTED | -0.493 |
Clown | -0.447 |
Capitol | -0.443 |
Certain | -0.440 |
Franch | -0.431 |
Token |
|
Token ID | 198 |
Feature activation | -6.70e-01 |
Pos logit contributions | |
REDACTED | +0.998 |
Clown | +0.901 |
Capitol | +0.893 |
Certain | +0.883 |
Franch | +0.853 |
Neg logit contributions | |
ringe | -1.372 |
ritten | -1.344 |
kefeller | -1.240 |
itness | -1.214 |
ickson | -1.197 |
Token |
|
Token ID | 198 |
Feature activation | -6.31e-01 |
Pos logit contributions | |
ringe | +1.069 |
ritten | +1.041 |
kefeller | +0.961 |
itness | +0.959 |
ickson | +0.934 |
Neg logit contributions | |
REDACTED | -0.740 |
Clown | -0.651 |
Capitol | -0.649 |
Certain | -0.641 |
Franch | -0.618 |
Token | The |
Token ID | 464 |
Feature activation | +3.67e-01 |
Pos logit contributions | |
ringe | +1.052 |
ritten | +1.025 |
kefeller | +0.953 |
itness | +0.945 |
ickson | +0.926 |
Neg logit contributions | |
REDACTED | -0.655 |
Clown | -0.562 |
Capitol | -0.556 |
Certain | -0.552 |
Franch | -0.528 |
Token | spokesperson |
Token ID | 11821 |
Feature activation | -2.63e+00 |
Pos logit contributions | |
REDACTED | +0.447 |
Clown | +0.403 |
Capitol | +0.402 |
Certain | +0.389 |
Franch | +0.382 |
Neg logit contributions | |
ringe | -0.538 |
ritten | -0.527 |
kefeller | -0.486 |
itness | -0.477 |
ickson | -0.466 |
Token | said |
Token ID | 531 |
Feature activation | -7.64e-01 |
Pos logit contributions | |
ringe | +3.442 |
ritten | +3.337 |
itness | +3.011 |
kefeller | +2.981 |
ickson | +2.922 |
Neg logit contributions | |
REDACTED | -3.626 |
Clown | -3.295 |
Capitol | -3.267 |
Certain | -3.236 |
Franch | -3.170 |
Token | the |
Token ID | 262 |
Feature activation | -4.28e-01 |
Pos logit contributions | |
ringe | +1.145 |
ritten | +1.115 |
kefeller | +1.016 |
itness | +1.015 |
ickson | +0.991 |
Neg logit contributions | |
REDACTED | -0.914 |
Clown | -0.820 |
Capitol | -0.813 |
Certain | -0.799 |
Franch | -0.782 |
Token | refugee |
Token ID | 12540 |
Feature activation | +9.66e-01 |
Pos logit contributions | |
REDACTED | +0.960 |
Capitol | +0.903 |
Clown | +0.887 |
Certain | +0.865 |
Franch | +0.836 |
Neg logit contributions | |
ringe | -1.393 |
ritten | -1.340 |
itness | -1.241 |
kefeller | -1.236 |
ickson | -1.221 |
Token | shelters |
Token ID | 23629 |
Feature activation | -7.49e-02 |
Pos logit contributions | |
REDACTED | +1.125 |
Capitol | +1.023 |
Clown | +1.009 |
Certain | +0.994 |
Franch | +0.975 |
Neg logit contributions | |
ringe | -1.460 |
ritten | -1.420 |
kefeller | -1.299 |
itness | -1.298 |
ickson | -1.264 |
Token | has |
Token ID | 468 |
Feature activation | +5.97e-01 |
Pos logit contributions | |
ringe | +0.109 |
ritten | +0.106 |
kefeller | +0.097 |
itness | +0.097 |
ickson | +0.094 |
Neg logit contributions | |
REDACTED | -0.092 |
Clown | -0.084 |
Capitol | -0.084 |
Certain | -0.082 |
Franch | -0.079 |
Token | more |
Token ID | 517 |
Feature activation | +2.11e-02 |
Pos logit contributions | |
REDACTED | +0.761 |
Clown | +0.699 |
Capitol | +0.694 |
Certain | +0.678 |
Franch | +0.663 |
Neg logit contributions | |
ringe | -0.836 |
ritten | -0.801 |
itness | -0.738 |
kefeller | -0.735 |
ickson | -0.716 |
Token | than |
Token ID | 621 |
Feature activation | -6.23e-01 |
Pos logit contributions | |
REDACTED | +0.028 |
Clown | +0.026 |
Capitol | +0.026 |
Certain | +0.025 |
Franch | +0.025 |
Neg logit contributions | |
ringe | -0.028 |
ritten | -0.027 |
itness | -0.025 |
kefeller | -0.024 |
ickson | -0.024 |
Token | tripled |
Token ID | 43761 |
Feature activation | -3.24e+00 |
Pos logit contributions | |
ringe | +0.729 |
ritten | +0.699 |
itness | +0.627 |
kefeller | +0.625 |
ickson | +0.605 |
Neg logit contributions | |
REDACTED | -0.948 |
Clown | -0.875 |
Capitol | -0.865 |
Certain | -0.855 |
Franch | -0.837 |
Token | and |
Token ID | 290 |
Feature activation | -3.35e-01 |
Pos logit contributions | |
ringe | +4.246 |
ritten | +4.131 |
itness | +3.720 |
kefeller | +3.668 |
ickson | +3.619 |
Neg logit contributions | |
REDACTED | -4.397 |
Clown | -3.958 |
Capitol | -3.926 |
Certain | -3.894 |
Franch | -3.825 |
Token | each |
Token ID | 1123 |
Feature activation | -6.88e-01 |
Pos logit contributions | |
ringe | +0.495 |
ritten | +0.481 |
itness | +0.441 |
kefeller | +0.437 |
ickson | +0.429 |
Neg logit contributions | |
REDACTED | -0.406 |
Clown | -0.367 |
Capitol | -0.365 |
Certain | -0.358 |
Franch | -0.350 |
Token | one |
Token ID | 530 |
Feature activation | +1.50e-01 |
Pos logit contributions | |
ringe | +0.945 |
ritten | +0.924 |
itness | +0.835 |
kefeller | +0.829 |
ickson | +0.814 |
Neg logit contributions | |
REDACTED | -0.902 |
Capitol | -0.822 |
Clown | -0.820 |
Certain | -0.801 |
Franch | -0.787 |
Token | is |
Token ID | 318 |
Feature activation | +4.86e-01 |
Pos logit contributions | |
REDACTED | +0.176 |
Clown | +0.158 |
Capitol | +0.158 |
Certain | +0.153 |
Franch | +0.150 |
Neg logit contributions | |
ringe | -0.228 |
ritten | -0.222 |
itness | -0.204 |
kefeller | -0.202 |
ickson | -0.197 |
Token | , |
Token ID | 11 |
Feature activation | -1.79e-01 |
Pos logit contributions | |
REDACTED | +0.561 |
Clown | +0.504 |
Capitol | +0.501 |
Certain | +0.491 |
Franch | +0.479 |
Neg logit contributions | |
ringe | -0.745 |
ritten | -0.725 |
kefeller | -0.666 |
itness | -0.663 |
ickson | -0.649 |
Token | started |
Token ID | 2067 |
Feature activation | +1.43e+00 |
Pos logit contributions | |
ringe | +0.397 |
ritten | +0.387 |
itness | +0.353 |
kefeller | +0.351 |
ickson | +0.343 |
Neg logit contributions | |
REDACTED | -0.317 |
Capitol | -0.289 |
Clown | -0.288 |
Certain | -0.282 |
Franch | -0.278 |
Token | in |
Token ID | 287 |
Feature activation | -2.52e-01 |
Pos logit contributions | |
REDACTED | +1.781 |
Capitol | +1.632 |
Clown | +1.632 |
Certain | +1.583 |
Franch | +1.551 |
Neg logit contributions | |
ringe | -2.039 |
ritten | -1.977 |
itness | -1.795 |
kefeller | -1.786 |
ickson | -1.756 |
Token | 2013 |
Token ID | 2211 |
Feature activation | +9.28e-01 |
Pos logit contributions | |
ringe | +0.402 |
ritten | +0.391 |
itness | +0.360 |
kefeller | +0.358 |
ickson | +0.351 |
Neg logit contributions | |
REDACTED | -0.276 |
Clown | -0.248 |
Capitol | -0.247 |
Certain | -0.243 |
Franch | -0.237 |
Token | , |
Token ID | 11 |
Feature activation | -1.49e-01 |
Pos logit contributions | |
REDACTED | +1.156 |
Clown | +1.042 |
Capitol | +1.037 |
Certain | +1.021 |
Franch | +1.000 |
Neg logit contributions | |
ringe | -1.339 |
ritten | -1.292 |
itness | -1.182 |
kefeller | -1.181 |
ickson | -1.146 |
Token | when |
Token ID | 618 |
Feature activation | -1.57e-01 |
Pos logit contributions | |
ringe | +0.216 |
ritten | +0.209 |
itness | +0.190 |
kefeller | +0.189 |
ickson | +0.185 |
Neg logit contributions | |
REDACTED | -0.186 |
Clown | -0.169 |
Capitol | -0.168 |
Certain | -0.165 |
Franch | -0.162 |
Token | McDonald |
Token ID | 14115 |
Feature activation | -3.32e+00 |
Pos logit contributions | |
ringe | +0.226 |
ritten | +0.220 |
itness | +0.199 |
kefeller | +0.199 |
ickson | +0.195 |
Neg logit contributions | |
REDACTED | -0.195 |
Clown | -0.177 |
Capitol | -0.176 |
Certain | -0.173 |
Franch | -0.171 |
Token | 's |
Token ID | 338 |
Feature activation | -9.48e-01 |
Pos logit contributions | |
ringe | +3.666 |
ritten | +3.586 |
itness | +3.163 |
kefeller | +3.123 |
ickson | +3.081 |
Neg logit contributions | |
REDACTED | -5.172 |
Clown | -4.682 |
Capitol | -4.622 |
Certain | -4.618 |
Franch | -4.470 |
Token | said |
Token ID | 531 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
ringe | +1.428 |
ritten | +1.390 |
itness | +1.269 |
kefeller | +1.266 |
ickson | +1.237 |
Neg logit contributions | |
REDACTED | -1.120 |
Clown | -1.005 |
Capitol | -1.001 |
Certain | -0.987 |
Franch | -0.970 |
Token | Bak |
Token ID | 17466 |
Feature activation | +8.99e-01 |
Pos logit contributions | |
ringe | +1.401 |
ritten | +1.360 |
itness | +1.230 |
kefeller | +1.220 |
ickson | +1.197 |
Neg logit contributions | |
REDACTED | -1.358 |
Clown | -1.232 |
Capitol | -1.223 |
Certain | -1.209 |
Franch | -1.187 |
Token | shi |
Token ID | 44019 |
Feature activation | -3.25e+00 |
Pos logit contributions | |
REDACTED | +1.224 |
Capitol | +1.125 |
Certain | +1.110 |
Clown | +1.100 |
Franch | +1.079 |
Neg logit contributions | |
ringe | -1.143 |
ritten | -1.112 |
itness | -1.009 |
kefeller | -1.002 |
ickson | -0.953 |
Token | 's |
Token ID | 338 |
Feature activation | -6.17e-01 |
Pos logit contributions | |
ringe | +4.496 |
ritten | +4.355 |
itness | +3.959 |
kefeller | +3.911 |
ickson | +3.853 |
Neg logit contributions | |
REDACTED | -4.200 |
Clown | -3.787 |
Capitol | -3.741 |
Certain | -3.712 |
Franch | -3.618 |
Token | a |
Token ID | 257 |
Feature activation | -5.43e-02 |
Pos logit contributions | |
ringe | +0.415 |
ritten | +0.405 |
itness | +0.372 |
kefeller | +0.371 |
ickson | +0.363 |
Neg logit contributions | |
REDACTED | -0.307 |
Clown | -0.276 |
Capitol | -0.272 |
Certain | -0.269 |
Franch | -0.261 |
Token | football |
Token ID | 4346 |
Feature activation | +5.23e-02 |
Pos logit contributions | |
ringe | +0.078 |
ritten | +0.076 |
kefeller | +0.070 |
itness | +0.069 |
ickson | +0.068 |
Neg logit contributions | |
REDACTED | -0.068 |
Clown | -0.062 |
Capitol | -0.061 |
Certain | -0.061 |
Franch | -0.059 |
Token | report |
Token ID | 989 |
Feature activation | -8.77e-01 |
Pos logit contributions | |
REDACTED | +0.069 |
Clown | +0.063 |
Capitol | +0.062 |
Certain | +0.061 |
Franch | +0.060 |
Neg logit contributions | |
ringe | -0.071 |
ritten | -0.069 |
kefeller | -0.063 |
itness | -0.063 |
ickson | -0.061 |
Token | on |
Token ID | 319 |
Feature activation | +9.25e-01 |
Pos logit contributions | |
ringe | +1.330 |
ritten | +1.298 |
itness | +1.191 |
kefeller | +1.189 |
ickson | +1.161 |
Neg logit contributions | |
REDACTED | -1.024 |
Clown | -0.921 |
Capitol | -0.908 |
Certain | -0.899 |
Franch | -0.875 |
Token | their |
Token ID | 511 |
Feature activation | +1.05e-01 |
Pos logit contributions | |
REDACTED | +1.058 |
Capitol | +0.957 |
Clown | +0.954 |
Certain | +0.929 |
Franch | +0.906 |
Neg logit contributions | |
ringe | -1.429 |
ritten | -1.386 |
kefeller | -1.274 |
itness | -1.270 |
ickson | -1.242 |
Token | ITV |
Token ID | 49188 |
Feature activation | -3.03e+00 |
Pos logit contributions | |
REDACTED | +0.131 |
Clown | +0.119 |
Capitol | +0.118 |
Certain | +0.116 |
Franch | +0.114 |
Neg logit contributions | |
ringe | -0.151 |
ritten | -0.146 |
kefeller | -0.133 |
itness | -0.133 |
ickson | -0.130 |
Token | show |
Token ID | 905 |
Feature activation | +9.35e-01 |
Pos logit contributions | |
ringe | +4.336 |
ritten | +4.216 |
itness | +3.855 |
kefeller | +3.761 |
ickson | +3.724 |
Neg logit contributions | |
REDACTED | -3.764 |
Clown | -3.344 |
Capitol | -3.313 |
Certain | -3.283 |
Franch | -3.230 |
Token | was |
Token ID | 373 |
Feature activation | -3.32e-01 |
Pos logit contributions | |
REDACTED | +1.090 |
Clown | +1.001 |
Capitol | +0.985 |
Certain | +0.971 |
Franch | +0.947 |
Neg logit contributions | |
ringe | -1.403 |
ritten | -1.377 |
kefeller | -1.270 |
itness | -1.262 |
ickson | -1.232 |
Token | a |
Token ID | 257 |
Feature activation | +8.06e-03 |
Pos logit contributions | |
ringe | +0.473 |
ritten | +0.459 |
itness | +0.419 |
kefeller | +0.419 |
ickson | +0.408 |
Neg logit contributions | |
REDACTED | -0.418 |
Clown | -0.378 |
Capitol | -0.374 |
Certain | -0.370 |
Franch | -0.361 |
Token | � |
Token ID | 564 |
Feature activation | -5.43e-01 |
Pos logit contributions | |
REDACTED | +0.010 |
Clown | +0.009 |
Capitol | +0.009 |
Certain | +0.008 |
Franch | +0.008 |
Neg logit contributions | |
ringe | -0.012 |
ritten | -0.012 |
kefeller | -0.011 |
itness | -0.011 |
ickson | -0.010 |
Token | � |
Token ID | 246 |
Feature activation | +1.15e+00 |
Pos logit contributions | |
ringe | +0.748 |
ritten | +0.727 |
itness | +0.660 |
kefeller | +0.660 |
ickson | +0.642 |
Neg logit contributions | |
REDACTED | -0.709 |
Clown | -0.644 |
Capitol | -0.634 |
Certain | -0.630 |
Franch | -0.611 |
Token | ifer |
Token ID | 7087 |
Feature activation | -2.18e+00 |
Pos logit contributions | |
ringe | +0.721 |
ritten | +0.707 |
itness | +0.651 |
kefeller | +0.650 |
ickson | +0.633 |
Neg logit contributions | |
REDACTED | -0.448 |
Clown | -0.389 |
Capitol | -0.388 |
Certain | -0.384 |
Franch | -0.370 |
Token | 2 |
Token ID | 362 |
Feature activation | +4.50e-01 |
Pos logit contributions | |
ringe | +3.579 |
ritten | +3.496 |
itness | +3.222 |
kefeller | +3.183 |
ickson | +3.151 |
Neg logit contributions | |
REDACTED | -2.243 |
Clown | -1.960 |
Capitol | -1.945 |
Certain | -1.929 |
Franch | -1.874 |
Token | . |
Token ID | 13 |
Feature activation | +6.80e-01 |
Pos logit contributions | |
REDACTED | +0.591 |
Capitol | +0.531 |
Clown | +0.529 |
Certain | +0.522 |
Franch | +0.505 |
Neg logit contributions | |
ringe | -0.633 |
ritten | -0.612 |
kefeller | -0.559 |
itness | -0.557 |
ickson | -0.536 |
Token | 0 |
Token ID | 15 |
Feature activation | -1.81e+00 |
Pos logit contributions | |
REDACTED | +0.641 |
Clown | +0.546 |
Capitol | +0.542 |
Certain | +0.538 |
Franch | +0.506 |
Neg logit contributions | |
ringe | -1.203 |
ritten | -1.174 |
kefeller | -1.092 |
itness | -1.083 |
ickson | -1.062 |
Token | took |
Token ID | 1718 |
Feature activation | -9.42e-01 |
Pos logit contributions | |
ringe | +2.570 |
ritten | +2.494 |
itness | +2.272 |
kefeller | +2.253 |
ickson | +2.210 |
Neg logit contributions | |
REDACTED | -2.300 |
Clown | -2.072 |
Capitol | -2.054 |
Certain | -2.029 |
Franch | -1.984 |
Token | credit |
Token ID | 3884 |
Feature activation | -3.21e+00 |
Pos logit contributions | |
ringe | +1.419 |
ritten | +1.381 |
itness | +1.264 |
kefeller | +1.252 |
ickson | +1.233 |
Neg logit contributions | |
REDACTED | -1.114 |
Clown | -0.999 |
Capitol | -0.992 |
Certain | -0.976 |
Franch | -0.951 |
Token | for |
Token ID | 329 |
Feature activation | +2.56e-01 |
Pos logit contributions | |
ringe | +4.075 |
ritten | +3.926 |
itness | +3.556 |
kefeller | +3.486 |
ickson | +3.442 |
Neg logit contributions | |
REDACTED | -4.407 |
Clown | -4.021 |
Certain | -3.995 |
Capitol | -3.982 |
Franch | -3.917 |
Token | the |
Token ID | 262 |
Feature activation | -3.13e-01 |
Pos logit contributions | |
REDACTED | +0.301 |
Clown | +0.270 |
Capitol | +0.270 |
Certain | +0.262 |
Franch | +0.256 |
Neg logit contributions | |
ringe | -0.390 |
ritten | -0.381 |
itness | -0.348 |
kefeller | -0.347 |
ickson | -0.339 |
Token | hack |
Token ID | 8156 |
Feature activation | -1.21e+00 |
Pos logit contributions | |
ringe | +0.487 |
ritten | +0.475 |
itness | +0.434 |
kefeller | +0.434 |
ickson | +0.424 |
Neg logit contributions | |
REDACTED | -0.358 |
Capitol | -0.320 |
Clown | -0.320 |
Certain | -0.311 |
Franch | -0.303 |
Token | , |
Token ID | 11 |
Feature activation | -4.02e-01 |
Pos logit contributions | |
ringe | +1.624 |
ritten | +1.575 |
itness | +1.425 |
kefeller | +1.413 |
ickson | +1.385 |
Neg logit contributions | |
REDACTED | -1.625 |
Clown | -1.474 |
Capitol | -1.461 |
Certain | -1.446 |
Franch | -1.415 |
Token | Crowd |
Token ID | 29452 |
Feature activation | -2.06e+00 |
Pos logit contributions | |
ringe | +0.565 |
ritten | +0.549 |
itness | +0.499 |
kefeller | +0.495 |
ickson | +0.486 |
Neg logit contributions | |
REDACTED | -0.516 |
Clown | -0.468 |
Capitol | -0.465 |
Certain | -0.458 |
Franch | -0.447 |
Token | through |
Token ID | 832 |
Feature activation | -5.15e-01 |
Pos logit contributions | |
ringe | +0.288 |
ritten | +0.282 |
itness | +0.257 |
kefeller | +0.255 |
ickson | +0.251 |
Neg logit contributions | |
REDACTED | -0.250 |
Clown | -0.230 |
Capitol | -0.226 |
Certain | -0.224 |
Franch | -0.217 |
Token | the |
Token ID | 262 |
Feature activation | -8.74e-03 |
Pos logit contributions | |
ringe | +0.734 |
ritten | +0.716 |
itness | +0.651 |
kefeller | +0.647 |
ickson | +0.634 |
Neg logit contributions | |
REDACTED | -0.647 |
Clown | -0.585 |
Capitol | -0.579 |
Certain | -0.574 |
Franch | -0.557 |
Token | inside |
Token ID | 2641 |
Feature activation | -1.77e+00 |
Pos logit contributions | |
ringe | +0.009 |
ritten | +0.009 |
kefeller | +0.008 |
itness | +0.008 |
ickson | +0.008 |
Neg logit contributions | |
REDACTED | -0.014 |
Clown | -0.013 |
Capitol | -0.013 |
Certain | -0.013 |
Franch | -0.012 |
Token | of |
Token ID | 286 |
Feature activation | -3.67e-01 |
Pos logit contributions | |
ringe | +2.585 |
ritten | +2.526 |
itness | +2.309 |
kefeller | +2.295 |
ickson | +2.243 |
Neg logit contributions | |
REDACTED | -2.154 |
Clown | -1.939 |
Capitol | -1.921 |
Certain | -1.898 |
Franch | -1.845 |
Token | the |
Token ID | 262 |
Feature activation | +1.85e-02 |
Pos logit contributions | |
ringe | +0.529 |
ritten | +0.516 |
itness | +0.470 |
kefeller | +0.468 |
ickson | +0.457 |
Neg logit contributions | |
REDACTED | -0.454 |
Clown | -0.409 |
Capitol | -0.405 |
Certain | -0.402 |
Franch | -0.390 |
Token | hem |
Token ID | 16869 |
Feature activation | -3.08e+00 |
Pos logit contributions | |
REDACTED | +0.025 |
Clown | +0.023 |
Capitol | +0.023 |
Certain | +0.023 |
Franch | +0.022 |
Neg logit contributions | |
ringe | -0.024 |
ritten | -0.024 |
kefeller | -0.022 |
itness | -0.021 |
ickson | -0.021 |
Token | , |
Token ID | 11 |
Feature activation | +9.82e-02 |
Pos logit contributions | |
ringe | +4.079 |
ritten | +3.954 |
itness | +3.572 |
kefeller | +3.533 |
ickson | +3.470 |
Neg logit contributions | |
REDACTED | -4.193 |
Clown | -3.804 |
Capitol | -3.771 |
Certain | -3.736 |
Franch | -3.661 |
Token | this |
Token ID | 428 |
Feature activation | -1.43e+00 |
Pos logit contributions | |
REDACTED | +0.114 |
Clown | +0.104 |
Capitol | +0.104 |
Certain | +0.102 |
Franch | +0.098 |
Neg logit contributions | |
ringe | -0.147 |
ritten | -0.144 |
kefeller | -0.132 |
itness | -0.131 |
ickson | -0.128 |
Token | will |
Token ID | 481 |
Feature activation | -5.60e-01 |
Pos logit contributions | |
ringe | +2.020 |
ritten | +1.977 |
itness | +1.803 |
kefeller | +1.790 |
ickson | +1.752 |
Neg logit contributions | |
REDACTED | -1.819 |
Clown | -1.661 |
Capitol | -1.642 |
Certain | -1.619 |
Franch | -1.565 |
Token | hide |
Token ID | 7808 |
Feature activation | +3.30e-01 |
Pos logit contributions | |
ringe | +0.811 |
ritten | +0.791 |
kefeller | +0.727 |
itness | +0.725 |
ickson | +0.705 |
Neg logit contributions | |
REDACTED | -0.682 |
Clown | -0.623 |
Capitol | -0.615 |
Certain | -0.607 |
Franch | -0.589 |
Token | the |
Token ID | 262 |
Feature activation | -1.13e-01 |
Pos logit contributions | |
REDACTED | +0.396 |
Clown | +0.359 |
Capitol | +0.354 |
Certain | +0.352 |
Franch | +0.338 |
Neg logit contributions | |
ringe | -0.486 |
ritten | -0.476 |
itness | -0.438 |
kefeller | -0.434 |
ickson | -0.422 |