Token | were |
Token ID | 547 |
Feature activation | +6.00e-01 |
Pos logit contributions | |
a | +0.220 |
two | +0.194 |
big | +0.194 |
and | +0.172 |
little | +0.144 |
Neg logit contributions | |
avail | -1.072 |
ittered | -1.068 |
rily | -1.042 |
footing | -1.031 |
eway | -1.018 |
Token | two |
Token ID | 734 |
Feature activation | -4.96e-01 |
Pos logit contributions | |
and | +0.363 |
a | +0.308 |
| +0.297 |
. | +0.295 |
big | +0.277 |
Neg logit contributions | |
avail | -2.514 |
rily | -2.510 |
keeping | -2.435 |
inations | -2.423 |
ittered | -2.405 |
Token | friends |
Token ID | 2460 |
Feature activation | -1.52e-01 |
Pos logit contributions | |
avail | +1.787 |
lights | +1.765 |
inations | +1.761 |
rily | +1.756 |
ittered | +1.745 |
Neg logit contributions | |
a | -0.643 |
| -0.609 |
and | -0.540 |
big | -0.510 |
the | -0.495 |
Token | ; |
Token ID | 26 |
Feature activation | +3.05e-01 |
Pos logit contributions | |
ittered | +0.503 |
avail | +0.502 |
lights | +0.499 |
inations | +0.496 |
immer | +0.493 |
Neg logit contributions | |
a | -0.214 |
big | -0.211 |
two | -0.192 |
| -0.190 |
and | -0.179 |
Token | Ma |
Token ID | 6669 |
Feature activation | -6.36e-01 |
Pos logit contributions | |
big | +0.318 |
and | +0.301 |
a | +0.290 |
| +0.284 |
two | +0.267 |
Neg logit contributions | |
avail | -1.094 |
ittered | -1.057 |
lights | -1.054 |
fer | -1.042 |
inations | -1.038 |
Token | isy |
Token ID | 13560 |
Feature activation | +1.81e+00 |
Pos logit contributions | |
inations | +1.806 |
avail | +1.792 |
rily | +1.775 |
ittered | +1.764 |
immer | +1.751 |
Neg logit contributions | |
a | -1.357 |
| -1.258 |
big | -1.256 |
two | -1.213 |
and | -1.162 |
Token | and |
Token ID | 290 |
Feature activation | -5.13e-01 |
Pos logit contributions | |
a | +1.361 |
big | +1.147 |
two | +1.091 |
right | +0.790 |
| +0.621 |
Neg logit contributions | |
rily | -7.373 |
ittered | -7.337 |
leness | -7.308 |
keeping | -7.224 |
anne | -7.196 |
Token | Bob |
Token ID | 5811 |
Feature activation | +4.58e-01 |
Pos logit contributions | |
avail | +1.566 |
rily | +1.451 |
fer | +1.451 |
ittered | +1.443 |
keeping | +1.424 |
Neg logit contributions | |
and | -0.995 |
a | -0.972 |
big | -0.963 |
| -0.956 |
. | -0.883 |
Token | . |
Token ID | 13 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
big | +0.307 |
a | +0.301 |
two | +0.237 |
little | +0.200 |
and | +0.196 |
Neg logit contributions | |
ittered | -1.923 |
leness | -1.860 |
rily | -1.848 |
ge | -1.844 |
keeping | -1.837 |
Token | Ma |
Token ID | 6669 |
Feature activation | -3.95e-01 |
Pos logit contributions | |
avail | +3.975 |
rily | +3.849 |
ittered | +3.833 |
keeping | +3.780 |
lights | +3.772 |
Neg logit contributions | |
a | -1.646 |
big | -1.539 |
and | -1.447 |
| -1.376 |
two | -1.376 |
Token | isy |
Token ID | 13560 |
Feature activation | +1.22e+00 |
Pos logit contributions | |
inations | +1.078 |
avail | +1.068 |
rily | +1.064 |
ittered | +1.058 |
immer | +1.048 |
Neg logit contributions | |
a | -0.896 |
big | -0.830 |
| -0.825 |
two | -0.806 |
the | -0.764 |
Token | a |
Token ID | 257 |
Feature activation | -7.42e-02 |
Pos logit contributions | |
avail | +0.835 |
rily | +0.807 |
lights | +0.801 |
leness | +0.795 |
ittered | +0.795 |
Neg logit contributions | |
| -0.371 |
a | -0.364 |
big | -0.355 |
and | -0.346 |
two | -0.326 |
Token | wink |
Token ID | 41955 |
Feature activation | +2.14e-01 |
Pos logit contributions | |
avail | +0.200 |
rily | +0.199 |
immer | +0.197 |
leness | +0.196 |
fer | +0.194 |
Neg logit contributions | |
a | -0.162 |
and | -0.151 |
| -0.149 |
two | -0.145 |
the | -0.144 |
Token | . |
Token ID | 13 |
Feature activation | -2.10e-01 |
Pos logit contributions | |
a | +0.172 |
big | +0.164 |
| +0.160 |
two | +0.134 |
the | +0.119 |
Neg logit contributions | |
rily | -0.900 |
ittered | -0.890 |
fer | -0.858 |
immer | -0.857 |
lights | -0.854 |
Token | Tommy |
Token ID | 19919 |
Feature activation | -1.66e-01 |
Pos logit contributions | |
avail | +0.757 |
rily | +0.748 |
lights | +0.740 |
leness | +0.719 |
ittered | +0.714 |
Neg logit contributions | |
a | -0.270 |
big | -0.257 |
and | -0.245 |
two | -0.232 |
little | -0.216 |
Token | smiled |
Token ID | 13541 |
Feature activation | +8.82e-01 |
Pos logit contributions | |
ittered | +0.520 |
rily | +0.514 |
fer | +0.509 |
leness | +0.503 |
avail | +0.495 |
Neg logit contributions | |
a | -0.278 |
big | -0.270 |
| -0.257 |
two | -0.245 |
little | -0.229 |
Token | back |
Token ID | 736 |
Feature activation | +1.80e+00 |
Pos logit contributions | |
a | +0.464 |
big | +0.423 |
two | +0.401 |
| +0.343 |
the | +0.285 |
Neg logit contributions | |
rily | -3.814 |
lights | -3.635 |
ittered | -3.604 |
immer | -3.586 |
rier | -3.578 |
Token | and |
Token ID | 290 |
Feature activation | -2.55e-01 |
Pos logit contributions | |
a | +1.352 |
big | +1.233 |
two | +1.050 |
| +0.855 |
the | +0.807 |
Neg logit contributions | |
rily | -7.624 |
ently | -7.417 |
iously | -7.373 |
lights | -7.264 |
ittered | -7.249 |
Token | learned |
Token ID | 4499 |
Feature activation | +3.84e-01 |
Pos logit contributions | |
avail | +0.797 |
rily | +0.793 |
leness | +0.784 |
ittered | +0.782 |
fer | +0.777 |
Neg logit contributions | |
a | -0.434 |
big | -0.414 |
| -0.407 |
and | -0.392 |
two | -0.387 |
Token | his |
Token ID | 465 |
Feature activation | -5.92e-01 |
Pos logit contributions | |
and | +0.254 |
big | +0.241 |
| +0.240 |
two | +0.227 |
a | +0.221 |
Neg logit contributions | |
rily | -1.553 |
avail | -1.542 |
lights | -1.518 |
keeping | -1.497 |
ittered | -1.489 |
Token | lesson |
Token ID | 11483 |
Feature activation | +4.95e-01 |
Pos logit contributions | |
avail | +1.431 |
inations | +1.364 |
leness | +1.347 |
rily | +1.342 |
lights | +1.329 |
Neg logit contributions | |
a | -1.469 |
and | -1.428 |
| -1.427 |
big | -1.386 |
the | -1.354 |
Token | . |
Token ID | 13 |
Feature activation | -5.11e-02 |
Pos logit contributions | |
big | +0.353 |
a | +0.318 |
two | +0.287 |
| +0.257 |
little | +0.252 |
Neg logit contributions | |
keeping | -2.070 |
rily | -2.053 |
lights | -2.040 |
ently | -2.039 |
avail | -2.035 |
Token | L |
Token ID | 43 |
Feature activation | -1.15e+00 |
Pos logit contributions | |
avail | +0.692 |
leness | +0.692 |
inations | +0.691 |
rily | +0.682 |
lights | +0.677 |
Neg logit contributions | |
a | -0.640 |
big | -0.599 |
two | -0.585 |
and | -0.575 |
the | -0.561 |
Token | ily |
Token ID | 813 |
Feature activation | +3.28e-01 |
Pos logit contributions | |
ittered | +3.319 |
rily | +3.303 |
avail | +3.300 |
ently | +3.251 |
immer | +3.243 |
Neg logit contributions | |
a | -2.431 |
big | -2.343 |
| -2.254 |
and | -2.219 |
two | -2.163 |
Token | thought |
Token ID | 1807 |
Feature activation | -5.09e-01 |
Pos logit contributions | |
a | +0.575 |
big | +0.558 |
| +0.526 |
two | +0.489 |
the | +0.481 |
Neg logit contributions | |
aned | -0.997 |
rily | -0.990 |
ittered | -0.988 |
rated | -0.983 |
avail | -0.975 |
Token | for |
Token ID | 329 |
Feature activation | -1.16e-01 |
Pos logit contributions | |
lights | +1.565 |
avail | +1.559 |
rily | +1.550 |
inations | +1.549 |
ittered | +1.526 |
Neg logit contributions | |
a | -0.926 |
| -0.887 |
two | -0.868 |
big | -0.855 |
and | -0.814 |
Token | a |
Token ID | 257 |
Feature activation | -7.14e-01 |
Pos logit contributions | |
avail | +0.542 |
inations | +0.517 |
ittered | +0.516 |
immer | +0.516 |
epad | +0.510 |
Neg logit contributions | |
big | -0.014 |
| -0.013 |
and | -0.013 |
a | -0.001 |
two | +0.002 |
Token | moment |
Token ID | 2589 |
Feature activation | +1.79e+00 |
Pos logit contributions | |
avail | +1.651 |
rated | +1.558 |
leness | +1.556 |
inations | +1.551 |
immer | +1.547 |
Neg logit contributions | |
a | -1.863 |
and | -1.783 |
two | -1.751 |
big | -1.736 |
| -1.735 |
Token | and |
Token ID | 290 |
Feature activation | +2.93e-02 |
Pos logit contributions | |
big | +1.511 |
a | +1.262 |
two | +1.210 |
| +0.870 |
little | +0.801 |
Neg logit contributions | |
immer | -7.261 |
comb | -7.210 |
lights | -7.127 |
istent | -7.116 |
ently | -7.103 |
Token | said |
Token ID | 531 |
Feature activation | -4.66e-01 |
Pos logit contributions | |
big | +0.048 |
a | +0.047 |
| +0.044 |
and | +0.043 |
two | +0.043 |
Neg logit contributions | |
avail | -0.096 |
rily | -0.092 |
fer | -0.091 |
ittered | -0.090 |
leness | -0.090 |
Token | , |
Token ID | 11 |
Feature activation | +4.55e-01 |
Pos logit contributions | |
avail | +1.949 |
rily | +1.939 |
ittered | +1.936 |
ently | +1.885 |
immer | +1.874 |
Neg logit contributions | |
a | -0.316 |
big | -0.284 |
| -0.281 |
and | -0.261 |
two | -0.249 |
Token | " |
Token ID | 366 |
Feature activation | -7.95e-01 |
Pos logit contributions | |
a | +0.640 |
big | +0.627 |
two | +0.562 |
and | +0.520 |
| +0.506 |
Neg logit contributions | |
rily | -1.583 |
ittered | -1.503 |
iously | -1.490 |
avail | -1.489 |
arians | -1.483 |
Token | I |
Token ID | 40 |
Feature activation | -5.29e-03 |
Pos logit contributions | |
avail | +2.486 |
rily | +2.348 |
leness | +2.332 |
rated | +2.325 |
lights | +2.315 |
Neg logit contributions | |
a | -1.597 |
big | -1.507 |
and | -1.455 |
two | -1.392 |
the | -1.373 |
Token | angry |
Token ID | 7954 |
Feature activation | +3.83e-01 |
Pos logit contributions | |
a | +0.453 |
| +0.450 |
and | +0.430 |
big | +0.428 |
two | +0.420 |
Neg logit contributions | |
keeping | -0.873 |
avail | -0.872 |
inations | -0.870 |
fer | -0.866 |
ittered | -0.864 |
Token | anymore |
Token ID | 7471 |
Feature activation | +6.96e-01 |
Pos logit contributions | |
a | +0.386 |
big | +0.368 |
two | +0.300 |
| +0.299 |
little | +0.258 |
Neg logit contributions | |
avail | -1.478 |
rily | -1.473 |
keeping | -1.459 |
ge | -1.450 |
fer | -1.448 |
Token | . |
Token ID | 13 |
Feature activation | +5.93e-01 |
Pos logit contributions | |
a | +0.519 |
big | +0.518 |
two | +0.429 |
little | +0.347 |
| +0.319 |
Neg logit contributions | |
keeping | -2.866 |
lights | -2.842 |
leness | -2.832 |
ge | -2.826 |
avail | -2.803 |
Token | He |
Token ID | 679 |
Feature activation | +5.14e-01 |
Pos logit contributions | |
a | +1.064 |
big | +1.006 |
and | +0.961 |
two | +0.950 |
in | +0.902 |
Neg logit contributions | |
rily | -1.812 |
lights | -1.782 |
avail | -1.773 |
leness | -1.759 |
keeping | -1.718 |
Token | is |
Token ID | 318 |
Feature activation | +7.59e-01 |
Pos logit contributions | |
a | +0.982 |
| +0.920 |
big | +0.887 |
two | +0.877 |
and | +0.852 |
Neg logit contributions | |
leness | -1.498 |
ittered | -1.455 |
heit | -1.443 |
aned | -1.437 |
rily | -1.436 |
Token | sorry |
Token ID | 7926 |
Feature activation | +1.77e+00 |
Pos logit contributions | |
| +1.407 |
and | +1.400 |
a | +1.383 |
two | +1.308 |
big | +1.278 |
Neg logit contributions | |
ittered | -2.133 |
rily | -2.124 |
avail | -2.085 |
immer | -2.078 |
inations | -2.073 |
Token | . |
Token ID | 13 |
Feature activation | +6.23e-01 |
Pos logit contributions | |
a | +1.598 |
big | +1.489 |
two | +1.465 |
little | +1.063 |
| +1.056 |
Neg logit contributions | |
lights | -7.267 |
ently | -6.668 |
rily | -6.639 |
leness | -6.599 |
inations | -6.581 |
Token | He |
Token ID | 679 |
Feature activation | +3.06e-01 |
Pos logit contributions | |
a | +0.987 |
big | +0.947 |
and | +0.895 |
two | +0.876 |
all | +0.820 |
Neg logit contributions | |
rily | -1.998 |
avail | -1.995 |
lights | -1.964 |
leness | -1.943 |
keeping | -1.932 |
Token | says |
Token ID | 1139 |
Feature activation | +3.69e-01 |
Pos logit contributions | |
a | +0.659 |
| +0.628 |
big | +0.605 |
two | +0.597 |
and | +0.589 |
Neg logit contributions | |
leness | -0.814 |
rily | -0.792 |
ittered | -0.788 |
inations | -0.780 |
aned | -0.777 |
Token | , |
Token ID | 11 |
Feature activation | +4.97e-01 |
Pos logit contributions | |
a | +0.329 |
big | +0.281 |
two | +0.262 |
and | +0.262 |
| +0.254 |
Neg logit contributions | |
leness | -1.431 |
lights | -1.424 |
avail | -1.417 |
rily | -1.413 |
fer | -1.411 |
Token | " |
Token ID | 366 |
Feature activation | +2.59e-01 |
Pos logit contributions | |
a | +0.695 |
big | +0.637 |
two | +0.604 |
and | +0.560 |
little | +0.519 |
Neg logit contributions | |
rily | -1.706 |
lights | -1.673 |
leness | -1.669 |
ittered | -1.667 |
fer | -1.663 |
Token | shape |
Token ID | 5485 |
Feature activation | +4.32e-01 |
Pos logit contributions | |
a | +0.123 |
| +0.123 |
and | +0.119 |
big | +0.117 |
two | +0.112 |
Neg logit contributions | |
avail | -0.144 |
rily | -0.142 |
keeping | -0.140 |
lights | -0.139 |
leness | -0.139 |
Token | . |
Token ID | 13 |
Feature activation | +1.69e-01 |
Pos logit contributions | |
a | +0.391 |
| +0.361 |
big | +0.322 |
two | +0.294 |
and | +0.277 |
Neg logit contributions | |
rily | -1.807 |
lights | -1.754 |
ittered | -1.704 |
rier | -1.671 |
avail | -1.670 |
Token | Mom |
Token ID | 11254 |
Feature activation | +2.93e-02 |
Pos logit contributions | |
a | +0.212 |
big | +0.194 |
and | +0.180 |
two | +0.176 |
little | +0.155 |
Neg logit contributions | |
avail | -0.621 |
rily | -0.613 |
lights | -0.612 |
ittered | -0.576 |
leness | -0.571 |
Token | gave |
Token ID | 2921 |
Feature activation | +5.92e-01 |
Pos logit contributions | |
a | +0.041 |
big | +0.037 |
| +0.035 |
two | +0.034 |
little | +0.032 |
Neg logit contributions | |
ittered | -0.101 |
rily | -0.101 |
inations | -0.100 |
lights | -0.098 |
leness | -0.098 |
Token | them |
Token ID | 606 |
Feature activation | +6.30e-01 |
Pos logit contributions | |
| +1.044 |
big | +1.041 |
and | +1.033 |
a | +1.019 |
two | +0.930 |
Neg logit contributions | |
avail | -1.841 |
rily | -1.823 |
keeping | -1.729 |
ittered | -1.684 |
lights | -1.664 |
Token | each |
Token ID | 1123 |
Feature activation | +1.77e+00 |
Pos logit contributions | |
| +0.800 |
a | +0.709 |
big | +0.701 |
and | +0.689 |
little | +0.627 |
Neg logit contributions | |
rily | -2.228 |
lights | -2.162 |
ittered | -2.147 |
avail | -2.134 |
iously | -2.088 |
Token | a |
Token ID | 257 |
Feature activation | +3.48e-02 |
Pos logit contributions | |
| +0.659 |
big | +0.598 |
a | +0.419 |
and | +0.413 |
be | +0.332 |
Neg logit contributions | |
rily | -7.358 |
leness | -7.202 |
avail | -7.176 |
epad | -7.067 |
rated | -7.059 |
Token | piece |
Token ID | 3704 |
Feature activation | +3.61e-01 |
Pos logit contributions | |
a | +0.062 |
| +0.058 |
and | +0.057 |
two | +0.055 |
the | +0.055 |
Neg logit contributions | |
rily | -0.106 |
avail | -0.105 |
inations | -0.103 |
immer | -0.103 |
ittered | -0.101 |
Token | of |
Token ID | 286 |
Feature activation | +8.79e-02 |
Pos logit contributions | |
a | +0.368 |
| +0.347 |
big | +0.343 |
two | +0.336 |
little | +0.307 |
Neg logit contributions | |
rily | -1.391 |
lights | -1.362 |
ittered | -1.358 |
immer | -1.330 |
comb | -1.309 |
Token | dough |
Token ID | 15756 |
Feature activation | +2.52e-01 |
Pos logit contributions | |
a | +0.217 |
| +0.215 |
and | +0.212 |
big | +0.208 |
two | +0.204 |
Neg logit contributions | |
avail | -0.212 |
rily | -0.200 |
leness | -0.199 |
lights | -0.189 |
inations | -0.189 |
Token | and |
Token ID | 290 |
Feature activation | -2.50e-02 |
Pos logit contributions | |
a | +0.155 |
big | +0.136 |
| +0.128 |
two | +0.110 |
little | +0.103 |
Neg logit contributions | |
rily | -1.101 |
keeping | -1.063 |
lights | -1.062 |
ittered | -1.058 |
comb | -1.050 |
Token | all |
Token ID | 477 |
Feature activation | +1.99e-01 |
Pos logit contributions | |
| +0.447 |
a | +0.437 |
two | +0.428 |
and | +0.418 |
big | +0.389 |
Neg logit contributions | |
avail | -1.163 |
rily | -1.122 |
lights | -1.113 |
leness | -1.111 |
ittered | -1.092 |
Token | the |
Token ID | 262 |
Feature activation | +5.05e-01 |
Pos logit contributions | |
a | +0.228 |
two | +0.203 |
big | +0.201 |
and | +0.196 |
| +0.194 |
Neg logit contributions | |
avail | -0.726 |
lights | -0.723 |
leness | -0.723 |
rily | -0.717 |
eway | -0.708 |
Token | little |
Token ID | 1310 |
Feature activation | +1.79e-01 |
Pos logit contributions | |
a | +0.750 |
| +0.706 |
and | +0.670 |
the | +0.629 |
two | +0.621 |
Neg logit contributions | |
avail | -1.679 |
inations | -1.644 |
leness | -1.634 |
rated | -1.615 |
keeping | -1.605 |
Token | kids |
Token ID | 3988 |
Feature activation | -1.57e-01 |
Pos logit contributions | |
a | +0.366 |
| +0.354 |
big | +0.331 |
two | +0.325 |
and | +0.322 |
Neg logit contributions | |
avail | -0.517 |
keeping | -0.501 |
leness | -0.495 |
lights | -0.480 |
footing | -0.478 |
Token | in |
Token ID | 287 |
Feature activation | +6.12e-01 |
Pos logit contributions | |
ittered | +0.637 |
inations | +0.631 |
rily | +0.627 |
leness | +0.626 |
ge | +0.621 |
Neg logit contributions | |
a | -0.127 |
big | -0.116 |
two | -0.110 |
| -0.108 |
little | -0.096 |
Token | the |
Token ID | 262 |
Feature activation | +1.76e+00 |
Pos logit contributions | |
and | +0.331 |
a | +0.299 |
two | +0.256 |
| +0.238 |
at | +0.213 |
Neg logit contributions | |
avail | -2.656 |
keeping | -2.478 |
lights | -2.476 |
leness | -2.454 |
rily | -2.453 |
Token | park |
Token ID | 3952 |
Feature activation | +2.31e-01 |
Pos logit contributions | |
a | +3.418 |
and | +3.378 |
| +3.064 |
. | +3.033 |
two | +3.009 |
Neg logit contributions | |
avail | -5.013 |
acket | -4.706 |
achers | -4.690 |
keeping | -4.690 |
immers | -4.600 |
Token | . |
Token ID | 13 |
Feature activation | -3.61e-02 |
Pos logit contributions | |
big | +0.134 |
a | +0.130 |
two | +0.119 |
| +0.113 |
little | +0.109 |
Neg logit contributions | |
ittered | -0.968 |
lights | -0.946 |
rily | -0.938 |
avail | -0.938 |
immer | -0.931 |
Token | They |
Token ID | 1119 |
Feature activation | -5.80e-01 |
Pos logit contributions | |
avail | +0.127 |
rily | +0.122 |
lights | +0.122 |
ittered | +0.120 |
leness | +0.120 |
Neg logit contributions | |
a | -0.049 |
big | -0.044 |
two | -0.044 |
and | -0.042 |
little | -0.039 |
Token | all |
Token ID | 477 |
Feature activation | -9.50e-02 |
Pos logit contributions | |
ittered | +1.923 |
avail | +1.862 |
inations | +1.853 |
leness | +1.848 |
rily | +1.844 |
Neg logit contributions | |
a | -0.895 |
| -0.848 |
and | -0.813 |
big | -0.810 |
two | -0.799 |
Token | loved |
Token ID | 6151 |
Feature activation | +1.61e-02 |
Pos logit contributions | |
ittered | +0.327 |
rated | +0.323 |
leness | +0.315 |
inations | +0.312 |
immer | +0.307 |
Neg logit contributions | |
a | -0.144 |
big | -0.132 |
two | -0.131 |
| -0.131 |
and | -0.128 |
Token | , |
Token ID | 11 |
Feature activation | +6.03e-01 |
Pos logit contributions | |
a | +0.282 |
big | +0.270 |
| +0.249 |
two | +0.232 |
and | +0.230 |
Neg logit contributions | |
rily | -0.814 |
avail | -0.803 |
ittered | -0.783 |
arians | -0.783 |
inations | -0.780 |
Token | Sam |
Token ID | 3409 |
Feature activation | +8.05e-03 |
Pos logit contributions | |
a | +1.072 |
| +1.006 |
big | +1.004 |
two | +0.953 |
and | +0.931 |
Neg logit contributions | |
avail | -1.895 |
rily | -1.786 |
inations | -1.759 |
fer | -1.744 |
arians | -1.731 |
Token | ." |
Token ID | 526 |
Feature activation | -1.82e-01 |
Pos logit contributions | |
a | +0.008 |
big | +0.007 |
two | +0.006 |
| +0.006 |
little | +0.005 |
Neg logit contributions | |
rily | -0.032 |
avail | -0.031 |
keeping | -0.031 |
rated | -0.030 |
ittered | -0.030 |
Token | Sam |
Token ID | 3409 |
Feature activation | +2.57e-01 |
Pos logit contributions | |
avail | +0.766 |
lights | +0.742 |
keeping | +0.741 |
leness | +0.740 |
rily | +0.735 |
Neg logit contributions | |
a | -0.119 |
big | -0.111 |
two | -0.086 |
little | -0.079 |
and | -0.076 |
Token | smiled |
Token ID | 13541 |
Feature activation | +6.81e-01 |
Pos logit contributions | |
a | +0.518 |
big | +0.506 |
| +0.477 |
two | +0.475 |
the | +0.435 |
Neg logit contributions | |
ittered | -0.718 |
rily | -0.714 |
leness | -0.687 |
fer | -0.683 |
avail | -0.680 |
Token | back |
Token ID | 736 |
Feature activation | +1.76e+00 |
Pos logit contributions | |
a | +0.304 |
big | +0.274 |
two | +0.252 |
| +0.229 |
the | +0.169 |
Neg logit contributions | |
rily | -2.975 |
lights | -2.868 |
avail | -2.860 |
immer | -2.829 |
ittered | -2.817 |
Token | and |
Token ID | 290 |
Feature activation | +9.98e-02 |
Pos logit contributions | |
a | +1.323 |
big | +1.171 |
two | +0.955 |
| +0.788 |
the | +0.726 |
Neg logit contributions | |
rily | -7.510 |
ently | -7.269 |
iously | -7.246 |
lights | -7.228 |
ittered | -7.086 |
Token | said |
Token ID | 531 |
Feature activation | -3.37e-01 |
Pos logit contributions | |
a | +0.167 |
big | +0.158 |
| +0.152 |
two | +0.149 |
and | +0.147 |
Neg logit contributions | |
rily | -0.316 |
avail | -0.314 |
leness | -0.313 |
ittered | -0.311 |
fer | -0.307 |
Token | , |
Token ID | 11 |
Feature activation | +3.32e-01 |
Pos logit contributions | |
rily | +1.416 |
ittered | +1.416 |
avail | +1.402 |
leness | +1.369 |
immer | +1.366 |
Neg logit contributions | |
a | -0.230 |
big | -0.198 |
and | -0.193 |
| -0.190 |
two | -0.173 |
Token | " |
Token ID | 366 |
Feature activation | -4.20e-01 |
Pos logit contributions | |
a | +0.447 |
big | +0.430 |
two | +0.393 |
and | +0.361 |
| +0.341 |
Neg logit contributions | |
rily | -1.190 |
ittered | -1.146 |
fer | -1.145 |
leness | -1.118 |
immer | -1.117 |
Token | Do |
Token ID | 5211 |
Feature activation | +6.12e-01 |
Pos logit contributions | |
avail | +1.215 |
fer | +1.197 |
leness | +1.190 |
rily | +1.190 |
ittered | +1.174 |
Neg logit contributions | |
a | -0.886 |
big | -0.857 |
and | -0.838 |
two | -0.788 |
the | -0.775 |
Token | S |
Token ID | 50 |
Feature activation | -9.20e-01 |
Pos logit contributions | |
leness | +0.377 |
avail | +0.373 |
igi | +0.371 |
fer | +0.368 |
inations | +0.368 |
Neg logit contributions | |
a | -0.232 |
big | -0.217 |
and | -0.213 |
two | -0.199 |
the | -0.196 |
Token | ara |
Token ID | 3301 |
Feature activation | +1.69e-01 |
Pos logit contributions | |
avail | +2.139 |
ittered | +2.103 |
immer | +2.081 |
rily | +2.066 |
ently | +2.048 |
Neg logit contributions | |
a | -2.556 |
big | -2.433 |
| -2.358 |
and | -2.319 |
two | -2.305 |
Token | and |
Token ID | 290 |
Feature activation | +1.70e-01 |
Pos logit contributions | |
a | +0.095 |
big | +0.082 |
| +0.060 |
two | +0.048 |
the | +0.048 |
Neg logit contributions | |
ittered | -0.722 |
rily | -0.721 |
ently | -0.715 |
aned | -0.709 |
fer | -0.709 |
Token | Ben |
Token ID | 3932 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
a | +0.265 |
big | +0.263 |
| +0.256 |
and | +0.256 |
two | +0.235 |
Neg logit contributions | |
avail | -0.577 |
rily | -0.537 |
keeping | -0.530 |
ittered | -0.529 |
leness | -0.529 |
Token | feel |
Token ID | 1254 |
Feature activation | -3.61e-01 |
Pos logit contributions | |
a | +0.220 |
big | +0.211 |
| +0.209 |
two | +0.195 |
and | +0.195 |
Neg logit contributions | |
ittered | -0.321 |
leness | -0.320 |
avail | -0.320 |
immer | -0.315 |
rily | -0.313 |
Token | sorry |
Token ID | 7926 |
Feature activation | +1.76e+00 |
Pos logit contributions | |
leness | +1.046 |
avail | +1.033 |
rily | +1.028 |
ittered | +1.006 |
immer | +0.983 |
Neg logit contributions | |
a | -0.667 |
and | -0.661 |
| -0.649 |
big | -0.637 |
two | -0.637 |
Token | for |
Token ID | 329 |
Feature activation | -4.61e-01 |
Pos logit contributions | |
a | +2.057 |
two | +1.931 |
big | +1.922 |
the | +1.521 |
| +1.469 |
Neg logit contributions | |
lights | -6.578 |
ently | -6.376 |
rated | -6.291 |
leness | -6.262 |
rily | -6.234 |
Token | what |
Token ID | 644 |
Feature activation | +1.21e-01 |
Pos logit contributions | |
avail | +1.643 |
rily | +1.562 |
ittered | +1.547 |
leness | +1.531 |
arians | +1.513 |
Neg logit contributions | |
a | -0.638 |
big | -0.631 |
| -0.624 |
and | -0.619 |
two | -0.570 |
Token | they |
Token ID | 484 |
Feature activation | +1.21e-03 |
Pos logit contributions | |
and | +0.193 |
a | +0.191 |
big | +0.183 |
| +0.178 |
two | +0.174 |
Neg logit contributions | |
avail | -0.391 |
fer | -0.373 |
lights | -0.371 |
ittered | -0.371 |
leness | -0.369 |
Token | did |
Token ID | 750 |
Feature activation | +5.46e-01 |
Pos logit contributions | |
a | +0.002 |
big | +0.002 |
| +0.001 |
two | +0.001 |
and | +0.001 |
Neg logit contributions | |
avail | -0.004 |
leness | -0.004 |
ittered | -0.004 |
rated | -0.004 |
inations | -0.004 |
Token | . |
Token ID | 13 |
Feature activation | -5.67e-01 |
Pos logit contributions | |
a | +0.347 |
big | +0.327 |
two | +0.277 |
| +0.240 |
little | +0.224 |
Neg logit contributions | |
ently | -2.283 |
avail | -2.266 |
rily | -2.261 |
keeping | -2.260 |
immer | -2.253 |
Token | the |
Token ID | 262 |
Feature activation | -1.90e-01 |
Pos logit contributions | |
lights | +1.293 |
avail | +1.288 |
leness | +1.278 |
rily | +1.275 |
keeping | +1.270 |
Neg logit contributions | |
a | -0.196 |
| -0.158 |
big | -0.156 |
two | -0.141 |
and | -0.138 |
Token | tears |
Token ID | 10953 |
Feature activation | +1.89e-01 |
Pos logit contributions | |
avail | +0.486 |
inations | +0.483 |
keeping | +0.479 |
rily | +0.476 |
leness | +0.475 |
Neg logit contributions | |
| -0.443 |
a | -0.439 |
and | -0.418 |
big | -0.394 |
two | -0.387 |
Token | . |
Token ID | 13 |
Feature activation | +5.70e-01 |
Pos logit contributions | |
a | +0.175 |
| +0.165 |
big | +0.160 |
two | +0.145 |
little | +0.133 |
Neg logit contributions | |
lights | -0.754 |
rily | -0.753 |
ittered | -0.748 |
immer | -0.732 |
inations | -0.724 |
Token | She |
Token ID | 1375 |
Feature activation | -8.17e-03 |
Pos logit contributions | |
a | +0.964 |
two | +0.874 |
big | +0.864 |
and | +0.816 |
one | +0.797 |
Neg logit contributions | |
keeping | -1.787 |
lights | -1.745 |
avail | -1.721 |
iders | -1.719 |
rily | -1.675 |
Token | feels |
Token ID | 5300 |
Feature activation | -7.07e-01 |
Pos logit contributions | |
avail | +0.026 |
leness | +0.026 |
lights | +0.026 |
ittered | +0.026 |
aned | +0.025 |
Neg logit contributions | |
a | -0.013 |
| -0.011 |
and | -0.011 |
big | -0.011 |
two | -0.011 |
Token | sorry |
Token ID | 7926 |
Feature activation | +1.76e+00 |
Pos logit contributions | |
avail | +1.935 |
rily | +1.928 |
leness | +1.898 |
ittered | +1.895 |
lights | +1.862 |
Neg logit contributions | |
a | -1.414 |
and | -1.409 |
| -1.389 |
big | -1.344 |
two | -1.344 |
Token | for |
Token ID | 329 |
Feature activation | -6.34e-01 |
Pos logit contributions | |
a | +2.136 |
big | +2.075 |
two | +1.974 |
| +1.655 |
little | +1.595 |
Neg logit contributions | |
lights | -6.559 |
ently | -6.255 |
rily | -5.927 |
leness | -5.893 |
eness | -5.889 |
Token | Tim |
Token ID | 5045 |
Feature activation | +3.52e-01 |
Pos logit contributions | |
avail | +2.262 |
ittered | +2.131 |
rily | +2.117 |
ge | +2.087 |
lights | +2.079 |
Neg logit contributions | |
a | -0.919 |
big | -0.885 |
| -0.871 |
and | -0.869 |
two | -0.809 |
Token | and |
Token ID | 290 |
Feature activation | +2.60e-02 |
Pos logit contributions | |
a | +0.175 |
big | +0.167 |
two | +0.117 |
| +0.110 |
little | +0.074 |
Neg logit contributions | |
keeping | -1.535 |
rily | -1.502 |
ge | -1.490 |
rated | -1.487 |
ently | -1.482 |
Token | Mia |
Token ID | 32189 |
Feature activation | +3.01e-01 |
Pos logit contributions | |
a | +0.055 |
big | +0.053 |
| +0.053 |
and | +0.052 |
two | +0.051 |
Neg logit contributions | |
avail | -0.070 |
lights | -0.069 |
rily | -0.068 |
ittered | -0.068 |
rated | -0.067 |
Token | . |
Token ID | 13 |
Feature activation | +6.09e-01 |
Pos logit contributions | |
a | +0.181 |
big | +0.166 |
| +0.145 |
two | +0.122 |
and | +0.096 |
Neg logit contributions | |
keeping | -1.249 |
rily | -1.238 |
ge | -1.234 |
inations | -1.230 |
epad | -1.222 |
Token | a |
Token ID | 257 |
Feature activation | +7.92e-01 |
Pos logit contributions | |
and | +0.223 |
big | +0.191 |
| +0.127 |
a | +0.077 |
two | +0.047 |
Neg logit contributions | |
ittered | -3.890 |
rily | -3.867 |
avail | -3.764 |
immer | -3.708 |
footing | -3.707 |
Token | little |
Token ID | 1310 |
Feature activation | +5.92e-01 |
Pos logit contributions | |
a | +0.715 |
and | +0.616 |
| +0.606 |
two | +0.554 |
the | +0.520 |
Neg logit contributions | |
avail | -3.091 |
fer | -3.075 |
rily | -3.028 |
leness | -2.976 |
ittered | -2.970 |
Token | boy |
Token ID | 2933 |
Feature activation | +4.30e-01 |
Pos logit contributions | |
a | +1.052 |
| +1.012 |
two | +0.919 |
and | +0.913 |
big | +0.892 |
Neg logit contributions | |
avail | -1.868 |
fer | -1.806 |
rily | -1.781 |
leness | -1.764 |
lights | -1.744 |
Token | named |
Token ID | 3706 |
Feature activation | +4.03e-01 |
Pos logit contributions | |
a | +0.927 |
big | +0.914 |
two | +0.908 |
| +0.854 |
little | +0.811 |
Neg logit contributions | |
ittered | -1.091 |
keeping | -1.084 |
avail | -1.082 |
fer | -1.077 |
leness | -1.076 |
Token | Tim |
Token ID | 5045 |
Feature activation | +1.23e-01 |
Pos logit contributions | |
and | +0.630 |
a | +0.626 |
big | +0.568 |
| +0.565 |
two | +0.558 |
Neg logit contributions | |
rily | -1.293 |
avail | -1.282 |
ittered | -1.267 |
rated | -1.250 |
inations | -1.247 |
Token | had |
Token ID | 550 |
Feature activation | +1.76e+00 |
Pos logit contributions | |
a | +0.154 |
big | +0.146 |
two | +0.136 |
| +0.132 |
little | +0.123 |
Neg logit contributions | |
rated | -0.439 |
keeping | -0.436 |
ittered | -0.429 |
avail | -0.423 |
rily | -0.418 |
Token | an |
Token ID | 281 |
Feature activation | -6.33e-01 |
Pos logit contributions | |
and | +1.020 |
| +0.669 |
, | +0.632 |
. | +0.607 |
little | +0.535 |
Neg logit contributions | |
keeping | -7.273 |
avail | -7.220 |
epad | -6.996 |
footing | -6.972 |
inations | -6.972 |
Token | ordinary |
Token ID | 8850 |
Feature activation | +1.31e-01 |
Pos logit contributions | |
avail | +1.904 |
rily | +1.841 |
footing | +1.799 |
immer | +1.798 |
ittered | +1.795 |
Neg logit contributions | |
| -1.234 |
a | -1.217 |
and | -1.149 |
big | -1.113 |
two | -1.111 |
Token | string |
Token ID | 4731 |
Feature activation | +2.25e-02 |
Pos logit contributions | |
a | +0.273 |
| +0.271 |
the | +0.253 |
two | +0.251 |
and | +0.249 |
Neg logit contributions | |
avail | -0.372 |
footing | -0.357 |
ittered | -0.348 |
immer | -0.347 |
rily | -0.347 |
Token | . |
Token ID | 13 |
Feature activation | -5.10e-01 |
Pos logit contributions | |
a | +0.015 |
| +0.013 |
two | +0.012 |
big | +0.011 |
the | +0.009 |
Neg logit contributions | |
rily | -0.095 |
ittered | -0.094 |
lights | -0.093 |
footing | -0.093 |
immer | -0.093 |
Token | He |
Token ID | 679 |
Feature activation | -4.78e-02 |
Pos logit contributions | |
avail | +1.712 |
rily | +1.697 |
lights | +1.649 |
ittered | +1.649 |
immer | +1.608 |
Neg logit contributions | |
a | -0.799 |
big | -0.736 |
two | -0.722 |
and | -0.714 |
| -0.685 |
Token | her |
Token ID | 607 |
Feature activation | -3.61e-02 |
Pos logit contributions | |
big | +0.302 |
| +0.294 |
a | +0.288 |
two | +0.260 |
and | +0.245 |
Neg logit contributions | |
avail | -1.712 |
rily | -1.676 |
ittered | -1.596 |
immer | -1.587 |
ently | -1.587 |
Token | cone |
Token ID | 27763 |
Feature activation | -7.69e-02 |
Pos logit contributions | |
rily | +0.103 |
avail | +0.100 |
footing | +0.100 |
iously | +0.096 |
ittered | +0.095 |
Neg logit contributions | |
a | -0.078 |
| -0.074 |
two | -0.071 |
big | -0.070 |
and | -0.070 |
Token | . |
Token ID | 13 |
Feature activation | +7.56e-02 |
Pos logit contributions | |
rily | +0.334 |
avail | +0.329 |
immer | +0.325 |
ittered | +0.323 |
lights | +0.323 |
Neg logit contributions | |
a | -0.050 |
big | -0.043 |
| -0.038 |
two | -0.037 |
little | -0.030 |
Token | " |
Token ID | 366 |
Feature activation | -7.27e-02 |
Pos logit contributions | |
a | +0.086 |
big | +0.080 |
and | +0.075 |
two | +0.068 |
little | +0.068 |
Neg logit contributions | |
avail | -0.278 |
ittered | -0.276 |
keeping | -0.275 |
rily | -0.275 |
ently | -0.271 |
Token | One |
Token ID | 3198 |
Feature activation | +3.19e-01 |
Pos logit contributions | |
avail | +0.191 |
rily | +0.182 |
keeping | +0.180 |
leness | +0.179 |
rated | +0.178 |
Neg logit contributions | |
a | -0.172 |
big | -0.166 |
and | -0.166 |
the | -0.155 |
. | -0.155 |
Token | , |
Token ID | 11 |
Feature activation | +1.76e+00 |
Pos logit contributions | |
a | +0.321 |
| +0.284 |
and | +0.271 |
big | +0.252 |
two | +0.243 |
Neg logit contributions | |
avail | -1.234 |
immer | -1.214 |
rated | -1.209 |
lights | -1.194 |
footing | -1.186 |
Token | two |
Token ID | 734 |
Feature activation | +2.02e-01 |
Pos logit contributions | |
| +2.512 |
and | +2.301 |
. | +2.213 |
big | +2.169 |
a | +2.154 |
Neg logit contributions | |
avail | -6.109 |
rily | -5.816 |
comb | -5.640 |
heit | -5.640 |
footing | -5.623 |
Token | , |
Token ID | 11 |
Feature activation | +1.53e+00 |
Pos logit contributions | |
a | +0.177 |
| +0.164 |
big | +0.142 |
and | +0.139 |
two | +0.117 |
Neg logit contributions | |
avail | -0.815 |
rily | -0.801 |
ittered | -0.788 |
immer | -0.787 |
inations | -0.787 |
Token | three |
Token ID | 1115 |
Feature activation | -4.08e-02 |
Pos logit contributions | |
| +2.235 |
and | +2.048 |
big | +2.025 |
a | +1.953 |
. | +1.938 |
Neg logit contributions | |
avail | -5.185 |
rily | -5.083 |
inations | -4.943 |
heit | -4.880 |
immer | -4.860 |
Token | !" |
Token ID | 2474 |
Feature activation | +4.90e-01 |
Pos logit contributions | |
rily | +0.148 |
avail | +0.147 |
leness | +0.147 |
inations | +0.145 |
lights | +0.145 |
Neg logit contributions | |
a | -0.053 |
| -0.049 |
big | -0.046 |
and | -0.043 |
two | -0.042 |
Token | she |
Token ID | 673 |
Feature activation | +1.33e-01 |
Pos logit contributions | |
a | +0.861 |
big | +0.841 |
and | +0.800 |
two | +0.794 |
little | +0.769 |
Neg logit contributions | |
avail | -1.496 |
rily | -1.447 |
lights | -1.407 |
leness | -1.394 |
keeping | -1.393 |
Token | with |
Token ID | 351 |
Feature activation | +4.92e-02 |
Pos logit contributions | |
a | +0.207 |
two | +0.203 |
big | +0.199 |
| +0.188 |
and | +0.159 |
Neg logit contributions | |
rily | -1.075 |
avail | -1.074 |
ittered | -1.059 |
lights | -1.031 |
fer | -1.026 |
Token | . |
Token ID | 13 |
Feature activation | -1.84e-01 |
Pos logit contributions | |
a | +0.044 |
big | +0.040 |
two | +0.039 |
| +0.039 |
and | +0.035 |
Neg logit contributions | |
avail | -0.193 |
rily | -0.192 |
ittered | -0.190 |
leness | -0.189 |
lights | -0.188 |
Token | The |
Token ID | 383 |
Feature activation | +1.09e+00 |
Pos logit contributions | |
avail | +0.632 |
rily | +0.606 |
lights | +0.598 |
ittered | +0.590 |
inations | +0.583 |
Neg logit contributions | |
a | -0.264 |
big | -0.263 |
and | -0.247 |
two | -0.235 |
little | -0.220 |
Token | kids |
Token ID | 3988 |
Feature activation | -2.13e-01 |
Pos logit contributions | |
a | +2.726 |
| +2.525 |
and | +2.489 |
the | +2.378 |
. | +2.376 |
Neg logit contributions | |
avail | -2.650 |
footing | -2.447 |
lights | -2.434 |
leness | -2.412 |
rated | -2.370 |
Token | in |
Token ID | 287 |
Feature activation | +7.93e-01 |
Pos logit contributions | |
ittered | +0.724 |
inations | +0.703 |
rated | +0.699 |
leness | +0.694 |
rily | +0.691 |
Neg logit contributions | |
a | -0.314 |
big | -0.296 |
| -0.293 |
two | -0.278 |
little | -0.258 |
Token | the |
Token ID | 262 |
Feature activation | +1.75e+00 |
Pos logit contributions | |
and | +0.385 |
| +0.359 |
a | +0.349 |
big | +0.314 |
two | +0.285 |
Neg logit contributions | |
avail | -3.379 |
ittered | -3.236 |
lights | -3.233 |
leness | -3.176 |
rily | -3.165 |
Token | park |
Token ID | 3952 |
Feature activation | -1.72e-01 |
Pos logit contributions | |
a | +4.670 |
and | +4.534 |
| +4.367 |
. | +4.239 |
the | +4.090 |
Neg logit contributions | |
avail | -4.063 |
footing | -3.486 |
lights | -3.440 |
leness | -3.391 |
fer | -3.375 |
Token | loved |
Token ID | 6151 |
Feature activation | -7.15e-02 |
Pos logit contributions | |
ittered | +0.576 |
rated | +0.552 |
avail | +0.547 |
inations | +0.542 |
footing | +0.540 |
Neg logit contributions | |
a | -0.253 |
big | -0.250 |
| -0.236 |
two | -0.233 |
and | -0.221 |
Token | to |
Token ID | 284 |
Feature activation | -2.43e-01 |
Pos logit contributions | |
keeping | +0.291 |
avail | +0.289 |
lights | +0.284 |
ittered | +0.283 |
inations | +0.282 |
Neg logit contributions | |
a | -0.047 |
two | -0.047 |
| -0.045 |
big | -0.042 |
and | -0.037 |
Token | roll |
Token ID | 4836 |
Feature activation | +7.71e-01 |
Pos logit contributions | |
avail | +0.864 |
ittered | +0.831 |
inations | +0.810 |
leness | +0.805 |
rily | +0.801 |
Neg logit contributions | |
a | -0.346 |
| -0.333 |
two | -0.324 |
and | -0.315 |
big | -0.314 |
Token | it |
Token ID | 340 |
Feature activation | +6.94e-01 |
Pos logit contributions | |
a | +0.880 |
two | +0.835 |
| +0.819 |
big | +0.786 |
little | +0.774 |
Neg logit contributions | |
lights | -2.832 |
rily | -2.746 |
avail | -2.745 |
ittered | -2.693 |
leness | -2.677 |
Token | to |
Token ID | 284 |
Feature activation | +4.49e-01 |
Pos logit contributions | |
and | +0.656 |
a | +0.567 |
| +0.553 |
two | +0.473 |
big | +0.435 |
Neg logit contributions | |
avail | -3.170 |
inations | -3.166 |
epad | -3.124 |
ittered | -3.123 |
leness | -3.107 |
Token | count |
Token ID | 954 |
Feature activation | +4.71e-01 |
Pos logit contributions | |
a | +1.147 |
| +1.120 |
big | +1.096 |
and | +1.077 |
two | +1.059 |
Neg logit contributions | |
avail | -1.100 |
ittered | -1.035 |
immer | -1.009 |
inations | -1.005 |
comb | -0.984 |
Token | . |
Token ID | 13 |
Feature activation | -2.65e-01 |
Pos logit contributions | |
a | +0.328 |
| +0.291 |
big | +0.283 |
two | +0.270 |
little | +0.257 |
Neg logit contributions | |
lights | -1.960 |
rily | -1.945 |
avail | -1.932 |
ittered | -1.868 |
ently | -1.859 |
Token | " |
Token ID | 366 |
Feature activation | +1.18e-01 |
Pos logit contributions | |
rily | +0.864 |
ittered | +0.852 |
lights | +0.847 |
keeping | +0.843 |
avail | +0.837 |
Neg logit contributions | |
a | -0.432 |
big | -0.408 |
and | -0.388 |
two | -0.375 |
the | -0.359 |
Token | One |
Token ID | 3198 |
Feature activation | +4.52e-01 |
Pos logit contributions | |
a | +0.276 |
and | +0.265 |
big | +0.265 |
the | +0.250 |
little | +0.245 |
Neg logit contributions | |
avail | -0.308 |
leness | -0.300 |
vation | -0.295 |
rily | -0.294 |
keeping | -0.293 |
Token | , |
Token ID | 11 |
Feature activation | +1.74e+00 |
Pos logit contributions | |
a | +0.381 |
| +0.337 |
and | +0.316 |
two | +0.289 |
big | +0.287 |
Neg logit contributions | |
avail | -1.812 |
rated | -1.773 |
immer | -1.750 |
inations | -1.737 |
ently | -1.727 |
Token | two |
Token ID | 734 |
Feature activation | +4.08e-02 |
Pos logit contributions | |
| +2.532 |
and | +2.312 |
big | +2.220 |
a | +2.217 |
. | +2.202 |
Neg logit contributions | |
avail | -6.025 |
arians | -5.614 |
inations | -5.608 |
comb | -5.501 |
icians | -5.473 |
Token | , |
Token ID | 11 |
Feature activation | +1.56e+00 |
Pos logit contributions | |
a | +0.032 |
| +0.031 |
and | +0.025 |
big | +0.025 |
two | +0.020 |
Neg logit contributions | |
avail | -0.169 |
inations | -0.164 |
ittered | -0.164 |
lights | -0.162 |
leness | -0.162 |
Token | three |
Token ID | 1115 |
Feature activation | -2.39e-02 |
Pos logit contributions | |
| +2.469 |
big | +2.245 |
and | +2.215 |
a | +2.168 |
. | +2.107 |
Neg logit contributions | |
avail | -5.143 |
inations | -4.930 |
arians | -4.920 |
icians | -4.858 |
epad | -4.783 |
Token | , |
Token ID | 11 |
Feature activation | +1.57e+00 |
Pos logit contributions | |
leness | +0.091 |
avail | +0.090 |
inations | +0.090 |
lights | +0.090 |
rily | +0.090 |
Neg logit contributions | |
a | -0.027 |
| -0.024 |
big | -0.023 |
and | -0.021 |
two | -0.020 |
Token | four |
Token ID | 1440 |
Feature activation | -3.05e-02 |
Pos logit contributions | |
| +2.898 |
big | +2.746 |
a | +2.692 |
and | +2.619 |
. | +2.543 |
Neg logit contributions | |
avail | -4.718 |
arians | -4.545 |
inations | -4.516 |
icians | -4.466 |
rily | -4.405 |
Token | ." |
Token ID | 526 |
Feature activation | -2.12e-01 |
Pos logit contributions | |
a | +0.145 |
big | +0.137 |
two | +0.127 |
| +0.125 |
little | +0.115 |
Neg logit contributions | |
rily | -0.456 |
avail | -0.451 |
ittered | -0.447 |
inations | -0.441 |
lights | -0.441 |
Token |
|
Token ID | 198 |
Feature activation | -6.30e-02 |
Pos logit contributions | |
rily | +0.845 |
avail | +0.843 |
leness | +0.819 |
lights | +0.817 |
keeping | +0.810 |
Neg logit contributions | |
a | -0.186 |
big | -0.177 |
two | -0.145 |
and | -0.137 |
little | -0.126 |
Token |
|
Token ID | 198 |
Feature activation | -2.42e-01 |
Pos logit contributions | |
avail | +0.274 |
rily | +0.272 |
leness | +0.272 |
ittered | +0.268 |
ently | +0.267 |
Neg logit contributions | |
a | -0.030 |
big | -0.021 |
two | -0.019 |
and | -0.018 |
one | -0.011 |
Token | Ben |
Token ID | 11696 |
Feature activation | +4.51e-01 |
Pos logit contributions | |
leness | +0.656 |
avail | +0.648 |
fer | +0.629 |
inations | +0.621 |
rily | +0.620 |
Neg logit contributions | |
a | -0.561 |
big | -0.521 |
and | -0.511 |
two | -0.490 |
the | -0.480 |
Token | smiled |
Token ID | 13541 |
Feature activation | +9.77e-01 |
Pos logit contributions | |
a | +1.050 |
big | +1.028 |
| +0.974 |
two | +0.948 |
the | +0.920 |
Neg logit contributions | |
rily | -1.130 |
ittered | -1.127 |
fer | -1.104 |
aned | -1.083 |
iously | -1.063 |
Token | back |
Token ID | 736 |
Feature activation | +1.74e+00 |
Pos logit contributions | |
a | +0.492 |
big | +0.416 |
two | +0.370 |
| +0.308 |
the | +0.272 |
Neg logit contributions | |
rily | -4.282 |
lights | -4.156 |
avail | -4.055 |
rier | -4.043 |
immer | -4.025 |
Token | and |
Token ID | 290 |
Feature activation | +2.67e-01 |
Pos logit contributions | |
a | +1.390 |
big | +1.286 |
two | +1.096 |
be | +0.886 |
| +0.840 |
Neg logit contributions | |
rily | -7.248 |
lights | -7.216 |
ently | -7.156 |
iously | -6.972 |
ge | -6.890 |
Token | said |
Token ID | 531 |
Feature activation | -1.94e-01 |
Pos logit contributions | |
a | +0.480 |
big | +0.450 |
| +0.443 |
and | +0.427 |
two | +0.424 |
Neg logit contributions | |
rily | -0.821 |
ittered | -0.798 |
leness | -0.795 |
avail | -0.793 |
fer | -0.783 |
Token | , |
Token ID | 11 |
Feature activation | +4.01e-01 |
Pos logit contributions | |
ittered | +0.809 |
rily | +0.808 |
avail | +0.798 |
immer | +0.784 |
lights | +0.784 |
Neg logit contributions | |
a | -0.136 |
big | -0.121 |
and | -0.114 |
| -0.113 |
two | -0.103 |
Token | " |
Token ID | 366 |
Feature activation | -3.84e-01 |
Pos logit contributions | |
a | +0.562 |
big | +0.532 |
two | +0.484 |
and | +0.453 |
little | +0.424 |
Neg logit contributions | |
rily | -1.401 |
fer | -1.361 |
ittered | -1.351 |
leness | -1.336 |
lights | -1.336 |
Token | Yes |
Token ID | 5297 |
Feature activation | +2.01e-01 |
Pos logit contributions | |
avail | +1.018 |
leness | +1.014 |
fer | +0.993 |
rily | +0.988 |
vation | +0.974 |
Neg logit contributions | |
a | -0.915 |
big | -0.872 |
and | -0.857 |
two | -0.815 |
the | -0.807 |
Token | ." |
Token ID | 526 |
Feature activation | -1.65e-01 |
Pos logit contributions | |
a | +0.737 |
big | +0.702 |
| +0.669 |
two | +0.648 |
little | +0.606 |
Neg logit contributions | |
rily | -1.591 |
keeping | -1.567 |
avail | -1.560 |
rated | -1.526 |
leness | -1.526 |
Token |
|
Token ID | 198 |
Feature activation | -1.18e-01 |
Pos logit contributions | |
rily | +0.705 |
avail | +0.696 |
keeping | +0.685 |
leness | +0.683 |
lights | +0.681 |
Neg logit contributions | |
a | -0.098 |
big | -0.086 |
two | -0.067 |
and | -0.056 |
little | -0.051 |
Token |
|
Token ID | 198 |
Feature activation | -2.16e-01 |
Pos logit contributions | |
rily | +0.512 |
avail | +0.502 |
ittered | +0.502 |
leness | +0.502 |
igi | +0.501 |
Neg logit contributions | |
a | -0.059 |
big | -0.040 |
two | -0.039 |
and | -0.035 |
one | -0.024 |
Token | Ben |
Token ID | 11696 |
Feature activation | +3.51e-01 |
Pos logit contributions | |
leness | +0.609 |
igi | +0.609 |
avail | +0.607 |
rily | +0.607 |
fer | +0.601 |
Neg logit contributions | |
a | -0.458 |
big | -0.421 |
and | -0.410 |
two | -0.399 |
little | -0.388 |
Token | smiled |
Token ID | 13541 |
Feature activation | +9.66e-01 |
Pos logit contributions | |
a | +0.617 |
big | +0.600 |
| +0.569 |
two | +0.540 |
the | +0.516 |
Neg logit contributions | |
rily | -1.074 |
ittered | -1.067 |
fer | -1.049 |
aned | -1.019 |
lights | -1.013 |
Token | back |
Token ID | 736 |
Feature activation | +1.74e+00 |
Pos logit contributions | |
a | +0.457 |
big | +0.379 |
two | +0.354 |
| +0.297 |
the | +0.246 |
Neg logit contributions | |
rily | -4.247 |
lights | -4.112 |
avail | -4.033 |
immer | -4.010 |
rier | -3.999 |
Token | and |
Token ID | 290 |
Feature activation | +1.61e-02 |
Pos logit contributions | |
a | +1.188 |
big | +1.025 |
two | +0.884 |
the | +0.698 |
be | +0.648 |
Neg logit contributions | |
rily | -7.529 |
ently | -7.315 |
lights | -7.181 |
iously | -7.174 |
ge | -6.955 |
Token | said |
Token ID | 531 |
Feature activation | +3.89e-01 |
Pos logit contributions | |
a | +0.026 |
big | +0.024 |
| +0.023 |
two | +0.023 |
and | +0.023 |
Neg logit contributions | |
rily | -0.052 |
avail | -0.051 |
ittered | -0.051 |
leness | -0.051 |
fer | -0.050 |
Token | : |
Token ID | 25 |
Feature activation | +2.08e-01 |
Pos logit contributions | |
a | +0.669 |
big | +0.621 |
two | +0.611 |
and | +0.605 |
| +0.571 |
Neg logit contributions | |
rily | -1.249 |
ittered | -1.235 |
ently | -1.194 |
immer | -1.194 |
leness | -1.188 |
Token | " |
Token ID | 366 |
Feature activation | -1.63e-01 |
Pos logit contributions | |
a | +0.234 |
big | +0.214 |
and | +0.205 |
two | +0.199 |
little | +0.177 |
Neg logit contributions | |
rily | -0.774 |
fer | -0.759 |
avail | -0.757 |
ittered | -0.757 |
leness | -0.745 |
Token | You |
Token ID | 1639 |
Feature activation | +3.35e-01 |
Pos logit contributions | |
avail | +0.498 |
leness | +0.494 |
fer | +0.488 |
rily | +0.482 |
vation | +0.474 |
Neg logit contributions | |
a | -0.323 |
big | -0.310 |
and | -0.300 |
two | -0.281 |
the | -0.279 |
Token | ." |
Token ID | 526 |
Feature activation | +2.26e-01 |
Pos logit contributions | |
avail | +1.943 |
rily | +1.940 |
ittered | +1.888 |
inations | +1.876 |
fer | +1.859 |
Neg logit contributions | |
a | -0.487 |
big | -0.467 |
| -0.458 |
two | -0.420 |
and | -0.418 |
Token |
|
Token ID | 198 |
Feature activation | -1.22e-01 |
Pos logit contributions | |
a | +0.117 |
big | +0.109 |
two | +0.091 |
little | +0.082 |
and | +0.074 |
Neg logit contributions | |
avail | -0.962 |
keeping | -0.929 |
ittered | -0.922 |
ently | -0.922 |
lights | -0.917 |
Token |
|
Token ID | 198 |
Feature activation | -2.53e-01 |
Pos logit contributions | |
ittered | +0.520 |
avail | +0.519 |
rily | +0.519 |
ently | +0.516 |
leness | +0.512 |
Neg logit contributions | |
a | -0.058 |
two | -0.043 |
big | -0.040 |
and | -0.037 |
little | -0.028 |
Token | Tom |
Token ID | 13787 |
Feature activation | +4.03e-01 |
Pos logit contributions | |
avail | +0.728 |
leness | +0.709 |
iders | +0.708 |
igi | +0.699 |
ittered | +0.699 |
Neg logit contributions | |
a | -0.515 |
big | -0.475 |
and | -0.465 |
two | -0.457 |
little | -0.451 |
Token | smiled |
Token ID | 13541 |
Feature activation | +1.09e+00 |
Pos logit contributions | |
a | +0.710 |
big | +0.693 |
| +0.663 |
two | +0.629 |
the | +0.591 |
Neg logit contributions | |
ittered | -1.218 |
aned | -1.201 |
fer | -1.197 |
rily | -1.196 |
rated | -1.169 |
Token | back |
Token ID | 736 |
Feature activation | +1.73e+00 |
Pos logit contributions | |
a | +0.549 |
big | +0.476 |
two | +0.461 |
| +0.400 |
the | +0.318 |
Neg logit contributions | |
rily | -4.682 |
lights | -4.548 |
avail | -4.469 |
rier | -4.439 |
immer | -4.406 |
Token | . |
Token ID | 13 |
Feature activation | +8.14e-02 |
Pos logit contributions | |
a | +1.380 |
big | +1.240 |
two | +1.131 |
| +0.867 |
be | +0.793 |
Neg logit contributions | |
rily | -7.111 |
ently | -7.042 |
iously | -6.954 |
lights | -6.925 |
ge | -6.882 |
Token | He |
Token ID | 679 |
Feature activation | +7.54e-01 |
Pos logit contributions | |
a | +0.144 |
big | +0.136 |
and | +0.131 |
two | +0.127 |
the | +0.121 |
Neg logit contributions | |
lights | -0.251 |
rily | -0.250 |
avail | -0.249 |
keeping | -0.242 |
ittered | -0.238 |
Token | was |
Token ID | 373 |
Feature activation | -4.10e-02 |
Pos logit contributions | |
a | +1.135 |
big | +1.058 |
two | +1.050 |
| +1.007 |
the | +0.961 |
Neg logit contributions | |
ittered | -2.424 |
rily | -2.364 |
inations | -2.332 |
fer | -2.322 |
aned | -2.279 |
Token | glad |
Token ID | 9675 |
Feature activation | +5.33e-01 |
Pos logit contributions | |
avail | +0.127 |
rily | +0.127 |
ittered | +0.127 |
fer | +0.123 |
inations | +0.122 |
Neg logit contributions | |
and | -0.067 |
| -0.065 |
a | -0.064 |
big | -0.062 |
two | -0.059 |
Token | that |
Token ID | 326 |
Feature activation | +2.79e-01 |
Pos logit contributions | |
big | +0.806 |
a | +0.780 |
two | +0.737 |
| +0.699 |
little | +0.671 |
Neg logit contributions | |
lights | -1.755 |
avail | -1.755 |
rily | -1.714 |
keeping | -1.666 |
iously | -1.664 |
Token | at |
Token ID | 379 |
Feature activation | -2.24e-01 |
Pos logit contributions | |
a | +0.284 |
big | +0.260 |
| +0.254 |
two | +0.251 |
and | +0.192 |
Neg logit contributions | |
ittered | -1.585 |
rily | -1.582 |
leness | -1.552 |
avail | -1.527 |
immer | -1.520 |
Token | Lily |
Token ID | 20037 |
Feature activation | +2.64e-01 |
Pos logit contributions | |
avail | +0.831 |
rily | +0.808 |
lights | +0.792 |
ittered | +0.782 |
arians | +0.768 |
Neg logit contributions | |
big | -0.275 |
a | -0.270 |
and | -0.259 |
| -0.258 |
two | -0.236 |
Token | . |
Token ID | 13 |
Feature activation | +2.65e-01 |
Pos logit contributions | |
big | +0.162 |
a | +0.159 |
two | +0.124 |
| +0.122 |
little | +0.105 |
Neg logit contributions | |
rily | -1.103 |
ittered | -1.098 |
ge | -1.091 |
keeping | -1.081 |
avail | -1.080 |
Token | Lily |
Token ID | 20037 |
Feature activation | +5.61e-03 |
Pos logit contributions | |
a | +0.314 |
big | +0.295 |
and | +0.264 |
two | +0.257 |
little | +0.225 |
Neg logit contributions | |
avail | -0.981 |
rily | -0.978 |
lights | -0.960 |
ittered | -0.933 |
ently | -0.924 |
Token | yells |
Token ID | 42675 |
Feature activation | +5.54e-01 |
Pos logit contributions | |
a | +0.011 |
| +0.010 |
big | +0.010 |
two | +0.009 |
and | +0.009 |
Neg logit contributions | |
leness | -0.016 |
avail | -0.016 |
aned | -0.016 |
rily | -0.016 |
rated | -0.016 |
Token | back |
Token ID | 736 |
Feature activation | +1.73e+00 |
Pos logit contributions | |
a | +0.448 |
big | +0.427 |
two | +0.408 |
| +0.396 |
the | +0.300 |
Neg logit contributions | |
ittered | -2.215 |
rily | -2.214 |
leness | -2.185 |
aned | -2.132 |
istent | -2.125 |
Token | at |
Token ID | 379 |
Feature activation | -3.28e-01 |
Pos logit contributions | |
big | +1.365 |
a | +1.326 |
two | +1.106 |
| +0.854 |
little | +0.801 |
Neg logit contributions | |
ge | -6.935 |
ently | -6.910 |
iously | -6.872 |
lights | -6.732 |
leness | -6.732 |
Token | Anna |
Token ID | 11735 |
Feature activation | +2.69e-01 |
Pos logit contributions | |
avail | +1.311 |
rily | +1.239 |
lights | +1.216 |
ittered | +1.215 |
keeping | +1.204 |
Neg logit contributions | |
big | -0.330 |
a | -0.326 |
| -0.314 |
and | -0.313 |
two | -0.276 |
Token | . |
Token ID | 13 |
Feature activation | -3.44e-02 |
Pos logit contributions | |
big | +0.142 |
a | +0.133 |
two | +0.102 |
| +0.097 |
little | +0.071 |
Neg logit contributions | |
rily | -1.164 |
ittered | -1.140 |
keeping | -1.139 |
ge | -1.133 |
immer | -1.128 |
Token |
|
Token ID | 198 |
Feature activation | +1.37e-01 |
Pos logit contributions | |
avail | +0.122 |
rily | +0.122 |
lights | +0.119 |
ittered | +0.117 |
leness | +0.116 |
Neg logit contributions | |
a | -0.045 |
big | -0.043 |
and | -0.039 |
two | -0.039 |
little | -0.035 |
Token |
|
Token ID | 198 |
Feature activation | +5.22e-01 |
Pos logit contributions | |
a | +0.076 |
big | +0.054 |
two | +0.052 |
and | +0.047 |
little | +0.036 |
Neg logit contributions | |
avail | -0.590 |
ently | -0.574 |
rily | -0.573 |
ittered | -0.572 |
leness | -0.571 |
Token | and |
Token ID | 290 |
Feature activation | +5.66e-01 |
Pos logit contributions | |
a | +0.251 |
big | +0.214 |
| +0.201 |
two | +0.177 |
little | +0.158 |
Neg logit contributions | |
rily | -2.276 |
immer | -2.200 |
lights | -2.186 |
ittered | -2.176 |
comb | -2.168 |
Token | gives |
Token ID | 3607 |
Feature activation | +2.33e-01 |
Pos logit contributions | |
| +1.260 |
and | +1.208 |
a | +1.208 |
big | +1.205 |
little | +1.157 |
Neg logit contributions | |
rily | -1.437 |
avail | -1.435 |
lights | -1.386 |
ittered | -1.370 |
leness | -1.348 |
Token | Lia |
Token ID | 49520 |
Feature activation | +3.18e-01 |
Pos logit contributions | |
and | +0.307 |
| +0.305 |
big | +0.302 |
a | +0.297 |
little | +0.272 |
Neg logit contributions | |
avail | -0.823 |
rily | -0.816 |
ittered | -0.803 |
keeping | -0.786 |
igi | -0.763 |
Token | and |
Token ID | 290 |
Feature activation | +5.53e-01 |
Pos logit contributions | |
big | +0.260 |
| +0.256 |
a | +0.254 |
two | +0.222 |
and | +0.208 |
Neg logit contributions | |
rily | -1.285 |
ittered | -1.254 |
immer | -1.220 |
igi | -1.209 |
lights | -1.207 |
Token | Tom |
Token ID | 4186 |
Feature activation | +6.22e-01 |
Pos logit contributions | |
big | +0.751 |
| +0.748 |
and | +0.692 |
a | +0.672 |
little | +0.663 |
Neg logit contributions | |
avail | -1.933 |
rily | -1.813 |
lights | -1.810 |
iously | -1.790 |
ittered | -1.785 |
Token | each |
Token ID | 1123 |
Feature activation | +1.73e+00 |
Pos logit contributions | |
| +0.547 |
big | +0.506 |
and | +0.456 |
a | +0.454 |
little | +0.417 |
Neg logit contributions | |
rily | -2.395 |
ittered | -2.387 |
keeping | -2.334 |
inations | -2.312 |
fed | -2.287 |
Token | a |
Token ID | 257 |
Feature activation | -3.16e-01 |
Pos logit contributions | |
and | +0.404 |
big | +0.358 |
| +0.322 |
a | +0.190 |
at | +0.152 |
Neg logit contributions | |
rily | -7.442 |
immer | -7.290 |
avail | -7.288 |
ittered | -7.245 |
inations | -7.191 |
Token | plate |
Token ID | 7480 |
Feature activation | +5.44e-01 |
Pos logit contributions | |
rily | +0.928 |
avail | +0.919 |
inations | +0.914 |
immer | +0.913 |
ittered | +0.907 |
Neg logit contributions | |
a | -0.587 |
| -0.561 |
and | -0.557 |
two | -0.533 |
the | -0.533 |
Token | and |
Token ID | 290 |
Feature activation | +5.59e-01 |
Pos logit contributions | |
a | +0.419 |
| +0.393 |
big | +0.376 |
two | +0.363 |
little | +0.338 |
Neg logit contributions | |
rily | -2.250 |
ittered | -2.184 |
immer | -2.159 |
lights | -2.155 |
comb | -2.133 |
Token | a |
Token ID | 257 |
Feature activation | -6.33e-02 |
Pos logit contributions | |
| +0.793 |
and | +0.740 |
big | +0.733 |
a | +0.713 |
. | +0.660 |
Neg logit contributions | |
avail | -1.893 |
rily | -1.861 |
lights | -1.826 |
ittered | -1.780 |
leness | -1.775 |
Token | fork |
Token ID | 15563 |
Feature activation | +1.91e-01 |
Pos logit contributions | |
rily | +0.156 |
avail | +0.155 |
inations | +0.152 |
immer | +0.150 |
ittered | +0.149 |
Neg logit contributions | |
a | -0.147 |
| -0.142 |
and | -0.141 |
the | -0.136 |
two | -0.134 |
Token | . |
Token ID | 13 |
Feature activation | -1.82e-01 |
Pos logit contributions | |
a | +0.134 |
big | +0.127 |
| +0.122 |
little | +0.109 |
two | +0.104 |
Neg logit contributions | |
rily | -0.740 |
avail | -0.728 |
lights | -0.717 |
ittered | -0.710 |
ge | -0.703 |
Token |
|
Token ID | 198 |
Feature activation | -8.39e-02 |
Pos logit contributions | |
avail | +0.751 |
rily | +0.745 |
ittered | +0.738 |
lights | +0.733 |
keeping | +0.729 |
Neg logit contributions | |
a | -0.118 |
big | -0.109 |
and | -0.091 |
two | -0.090 |
little | -0.081 |
Token |
|
Token ID | 198 |
Feature activation | +2.45e-01 |
Pos logit contributions | |
rily | +0.359 |
avail | +0.359 |
ently | +0.356 |
ittered | +0.355 |
inations | +0.354 |
Neg logit contributions | |
a | -0.041 |
big | -0.031 |
two | -0.029 |
and | -0.024 |
little | -0.020 |
Token | " |
Token ID | 1 |
Feature activation | -4.59e-01 |
Pos logit contributions | |
a | +0.371 |
big | +0.348 |
and | +0.322 |
little | +0.316 |
two | +0.315 |
Neg logit contributions | |
leness | -0.820 |
avail | -0.812 |
keeping | -0.809 |
ge | -0.807 |
fer | -0.807 |
Token | One |
Token ID | 3198 |
Feature activation | +4.76e-01 |
Pos logit contributions | |
avail | +1.054 |
leness | +0.993 |
vation | +0.989 |
fer | +0.983 |
rily | +0.976 |
Neg logit contributions | |
a | -1.249 |
and | -1.215 |
big | -1.210 |
| -1.144 |
the | -1.140 |
Token | , |
Token ID | 11 |
Feature activation | +1.73e+00 |
Pos logit contributions | |
a | +0.382 |
| +0.327 |
and | +0.289 |
big | +0.275 |
two | +0.258 |
Neg logit contributions | |
avail | -1.934 |
lights | -1.913 |
rated | -1.907 |
immer | -1.898 |
ently | -1.893 |
Token | two |
Token ID | 734 |
Feature activation | +9.81e-02 |
Pos logit contributions | |
| +1.451 |
big | +1.191 |
and | +1.167 |
. | +1.082 |
a | +1.045 |
Neg logit contributions | |
avail | -7.091 |
rily | -6.739 |
lights | -6.728 |
inations | -6.713 |
icians | -6.590 |
Token | , |
Token ID | 11 |
Feature activation | +1.52e+00 |
Pos logit contributions | |
a | +0.079 |
| +0.075 |
big | +0.062 |
and | +0.059 |
two | +0.048 |
Neg logit contributions | |
avail | -0.404 |
rily | -0.393 |
inations | -0.393 |
ittered | -0.392 |
lights | -0.391 |
Token | three |
Token ID | 1115 |
Feature activation | +1.05e-01 |
Pos logit contributions | |
| +1.858 |
big | +1.673 |
and | +1.621 |
. | +1.526 |
a | +1.513 |
Neg logit contributions | |
avail | -5.632 |
inations | -5.400 |
rily | -5.378 |
arians | -5.352 |
icians | -5.348 |
Token | , |
Token ID | 11 |
Feature activation | +1.59e+00 |
Pos logit contributions | |
a | +0.101 |
| +0.089 |
big | +0.082 |
two | +0.071 |
and | +0.070 |
Neg logit contributions | |
leness | -0.417 |
lights | -0.416 |
rily | -0.415 |
avail | -0.414 |
inations | -0.413 |
Token | four |
Token ID | 1440 |
Feature activation | +8.89e-02 |
Pos logit contributions | |
| +2.380 |
big | +2.245 |
and | +2.087 |
a | +2.068 |
. | +2.028 |
Neg logit contributions | |
avail | -5.439 |
rily | -5.237 |
inations | -5.197 |
arians | -5.178 |
immer | -5.149 |
Token | very |
Token ID | 845 |
Feature activation | +8.45e-01 |
Pos logit contributions | |
and | +0.441 |
| +0.423 |
. | +0.376 |
a | +0.356 |
big | +0.341 |
Neg logit contributions | |
avail | -2.261 |
ittered | -2.233 |
rily | -2.195 |
footing | -2.161 |
igi | -2.161 |
Token | happy |
Token ID | 3772 |
Feature activation | +3.78e-01 |
Pos logit contributions | |
a | +1.485 |
and | +1.464 |
| +1.363 |
two | +1.339 |
. | +1.252 |
Neg logit contributions | |
avail | -2.577 |
ittered | -2.570 |
leness | -2.543 |
inations | -2.528 |
rily | -2.521 |
Token | and |
Token ID | 290 |
Feature activation | +3.29e-01 |
Pos logit contributions | |
big | +0.450 |
two | +0.437 |
a | +0.435 |
| +0.405 |
little | +0.359 |
Neg logit contributions | |
keeping | -1.336 |
avail | -1.334 |
footing | -1.319 |
fer | -1.309 |
lights | -1.307 |
Token | everyone |
Token ID | 2506 |
Feature activation | -3.42e-01 |
Pos logit contributions | |
a | +0.488 |
| +0.464 |
two | +0.457 |
big | +0.454 |
and | +0.450 |
Neg logit contributions | |
leness | -1.086 |
avail | -1.082 |
ittered | -1.062 |
rated | -1.055 |
inations | -1.043 |
Token | in |
Token ID | 287 |
Feature activation | +4.34e-01 |
Pos logit contributions | |
ittered | +1.175 |
rily | +1.164 |
leness | +1.151 |
fer | +1.142 |
rated | +1.133 |
Neg logit contributions | |
a | -0.494 |
two | -0.458 |
big | -0.458 |
| -0.444 |
and | -0.413 |
Token | the |
Token ID | 262 |
Feature activation | +1.73e+00 |
Pos logit contributions | |
and | +0.234 |
a | +0.207 |
two | +0.190 |
big | +0.177 |
| +0.159 |
Neg logit contributions | |
avail | -1.877 |
leness | -1.774 |
ittered | -1.760 |
keeping | -1.759 |
lights | -1.752 |
Token | family |
Token ID | 1641 |
Feature activation | +6.27e-01 |
Pos logit contributions | |
and | +3.834 |
a | +3.773 |
. | +3.413 |
the | +3.412 |
two | +3.319 |
Neg logit contributions | |
avail | -4.641 |
acket | -4.197 |
keeping | -4.184 |
vation | -4.101 |
fer | -4.086 |
Token | loved |
Token ID | 6151 |
Feature activation | -1.73e-01 |
Pos logit contributions | |
a | +0.629 |
big | +0.591 |
two | +0.507 |
little | +0.450 |
| +0.445 |
Neg logit contributions | |
leness | -2.370 |
keeping | -2.363 |
inations | -2.351 |
ittered | -2.350 |
fer | -2.346 |
Token | him |
Token ID | 683 |
Feature activation | -6.61e-02 |
Pos logit contributions | |
avail | +0.530 |
keeping | +0.505 |
rily | +0.500 |
leness | +0.488 |
ittered | +0.484 |
Neg logit contributions | |
a | -0.316 |
big | -0.309 |
two | -0.307 |
and | -0.303 |
| -0.298 |
Token | very |
Token ID | 845 |
Feature activation | -2.54e-02 |
Pos logit contributions | |
ittered | +0.278 |
rily | +0.276 |
avail | +0.273 |
keeping | +0.272 |
inations | +0.272 |
Neg logit contributions | |
a | -0.034 |
big | -0.034 |
two | -0.030 |
| -0.030 |
little | -0.023 |
Token | much |
Token ID | 881 |
Feature activation | +1.25e-01 |
Pos logit contributions | |
lights | +0.080 |
inations | +0.080 |
leness | +0.079 |
avail | +0.078 |
ittered | +0.078 |
Neg logit contributions | |
a | -0.045 |
and | -0.042 |
two | -0.042 |
| -0.040 |
big | -0.040 |
Token | way |
Token ID | 835 |
Feature activation | -1.24e-01 |
Pos logit contributions | |
avail | +2.332 |
rily | +2.188 |
ittered | +2.129 |
footing | +2.125 |
keeping | +2.099 |
Neg logit contributions | |
a | -1.594 |
| -1.546 |
and | -1.520 |
big | -1.468 |
two | -1.441 |
Token | forward |
Token ID | 2651 |
Feature activation | +1.15e-01 |
Pos logit contributions | |
avail | +0.438 |
rily | +0.430 |
ge | +0.416 |
ittered | +0.415 |
footing | +0.414 |
Neg logit contributions | |
a | -0.182 |
| -0.166 |
two | -0.164 |
big | -0.162 |
and | -0.155 |
Token | . |
Token ID | 13 |
Feature activation | -5.24e-01 |
Pos logit contributions | |
a | +0.071 |
big | +0.070 |
two | +0.059 |
| +0.054 |
little | +0.043 |
Neg logit contributions | |
avail | -0.488 |
rily | -0.487 |
keeping | -0.485 |
ittered | -0.484 |
ge | -0.477 |
Token | She |
Token ID | 1375 |
Feature activation | +1.67e-01 |
Pos logit contributions | |
rily | +1.650 |
avail | +1.630 |
ittered | +1.589 |
lights | +1.567 |
arians | +1.554 |
Neg logit contributions | |
a | -0.925 |
big | -0.895 |
and | -0.882 |
two | -0.823 |
the | -0.786 |
Token | stayed |
Token ID | 9658 |
Feature activation | -4.12e-01 |
Pos logit contributions | |
a | +0.299 |
| +0.283 |
big | +0.280 |
two | +0.258 |
and | +0.251 |
Neg logit contributions | |
ittered | -0.527 |
fer | -0.517 |
avail | -0.513 |
rily | -0.510 |
inations | -0.507 |
Token | foc |
Token ID | 2133 |
Feature activation | -3.47e+00 |
Pos logit contributions | |
avail | +1.636 |
rily | +1.611 |
ittered | +1.581 |
fer | +1.555 |
immer | +1.547 |
Neg logit contributions | |
| -0.408 |
a | -0.400 |
big | -0.349 |
two | -0.339 |
and | -0.323 |
Token | ussed |
Token ID | 29569 |
Feature activation | +3.07e-02 |
Pos logit contributions | |
avail | +4.659 |
rily | +4.238 |
leness | +4.140 |
ittered | +4.099 |
inations | +3.993 |
Neg logit contributions | |
a | -13.626 |
and | -13.242 |
| -13.036 |
big | -13.011 |
two | -12.767 |
Token | . |
Token ID | 13 |
Feature activation | -4.39e-01 |
Pos logit contributions | |
a | +0.014 |
| +0.013 |
big | +0.010 |
two | +0.009 |
little | +0.007 |
Neg logit contributions | |
rily | -0.138 |
avail | -0.136 |
ittered | -0.134 |
ently | -0.134 |
lights | -0.133 |
Token | On |
Token ID | 1550 |
Feature activation | -4.39e-01 |
Pos logit contributions | |
rily | +1.469 |
avail | +1.466 |
ittered | +1.439 |
ently | +1.406 |
lights | +1.399 |
Neg logit contributions | |
a | -0.675 |
big | -0.653 |
and | -0.636 |
two | -0.587 |
the | -0.560 |
Token | the |
Token ID | 262 |
Feature activation | -6.90e-01 |
Pos logit contributions | |
avail | +1.930 |
rily | +1.856 |
ittered | +1.846 |
lights | +1.833 |
leness | +1.816 |
Neg logit contributions | |
a | -0.252 |
| -0.244 |
and | -0.237 |
big | -0.235 |
two | -0.170 |
Token | third |
Token ID | 2368 |
Feature activation | -5.46e-03 |
Pos logit contributions | |
avail | +2.172 |
rily | +2.027 |
leness | +2.014 |
vation | +1.993 |
footing | +1.957 |
Neg logit contributions | |
a | -1.361 |
| -1.349 |
and | -1.293 |
big | -1.223 |
two | -1.186 |
Token | and |
Token ID | 290 |
Feature activation | +3.88e-01 |
Pos logit contributions | |
a | +0.081 |
big | +0.075 |
| +0.067 |
two | +0.061 |
the | +0.041 |
Neg logit contributions | |
avail | -0.683 |
rily | -0.679 |
lights | -0.671 |
ittered | -0.669 |
iously | -0.669 |
Token | so |
Token ID | 523 |
Feature activation | +5.04e-01 |
Pos logit contributions | |
| +0.716 |
a | +0.706 |
and | +0.683 |
big | +0.674 |
two | +0.666 |
Neg logit contributions | |
avail | -1.185 |
rated | -1.165 |
leness | -1.157 |
ittered | -1.141 |
rily | -1.128 |
Token | she |
Token ID | 673 |
Feature activation | +3.45e-01 |
Pos logit contributions | |
a | +0.788 |
and | +0.728 |
| +0.719 |
two | +0.702 |
big | +0.698 |
Neg logit contributions | |
avail | -1.690 |
lights | -1.624 |
leness | -1.623 |
ittered | -1.622 |
rily | -1.598 |
Token | threw |
Token ID | 9617 |
Feature activation | +1.03e+00 |
Pos logit contributions | |
a | +0.604 |
| +0.585 |
big | +0.568 |
and | +0.552 |
two | +0.531 |
Neg logit contributions | |
rily | -1.083 |
avail | -1.069 |
rated | -1.062 |
aned | -1.056 |
ittered | -1.055 |
Token | a |
Token ID | 257 |
Feature activation | -1.94e-01 |
Pos logit contributions | |
| +1.151 |
big | +1.042 |
a | +0.962 |
and | +0.961 |
little | +0.956 |
Neg logit contributions | |
avail | -3.919 |
rily | -3.763 |
keeping | -3.706 |
ittered | -3.629 |
leness | -3.627 |
Token | tant |
Token ID | 24246 |
Feature activation | -3.30e+00 |
Pos logit contributions | |
rily | +0.610 |
avail | +0.605 |
footing | +0.597 |
leness | +0.591 |
fer | +0.590 |
Neg logit contributions | |
a | -0.343 |
| -0.330 |
and | -0.309 |
two | -0.299 |
the | -0.294 |
Token | rum |
Token ID | 6582 |
Feature activation | -5.90e-01 |
Pos logit contributions | |
avail | +8.076 |
rily | +7.819 |
ittered | +7.649 |
leness | +7.525 |
lights | +7.501 |
Neg logit contributions | |
a | -9.198 |
and | -8.762 |
big | -8.646 |
| -8.623 |
two | -8.353 |
Token | when |
Token ID | 618 |
Feature activation | -4.65e-01 |
Pos logit contributions | |
rily | +2.576 |
ittered | +2.531 |
avail | +2.525 |
keeping | +2.523 |
lights | +2.512 |
Neg logit contributions | |
big | -0.323 |
a | -0.315 |
| -0.265 |
two | -0.222 |
and | -0.199 |
Token | her |
Token ID | 607 |
Feature activation | +2.49e-01 |
Pos logit contributions | |
avail | +1.675 |
ittered | +1.557 |
footing | +1.553 |
lights | +1.550 |
leness | +1.538 |
Neg logit contributions | |
and | -0.626 |
a | -0.626 |
| -0.608 |
big | -0.595 |
two | -0.548 |
Token | parents |
Token ID | 3397 |
Feature activation | -1.64e-02 |
Pos logit contributions | |
a | +0.373 |
| +0.366 |
and | +0.346 |
the | +0.316 |
two | +0.297 |
Neg logit contributions | |
avail | -0.859 |
footing | -0.854 |
lights | -0.848 |
rily | -0.835 |
fer | -0.830 |
Token | said |
Token ID | 531 |
Feature activation | -2.08e-01 |
Pos logit contributions | |
avail | +0.053 |
ittered | +0.053 |
inations | +0.053 |
fer | +0.053 |
rily | +0.053 |
Neg logit contributions | |
a | -0.027 |
big | -0.024 |
| -0.023 |
two | -0.023 |
and | -0.022 |
Token | temper |
Token ID | 4124 |
Feature activation | -3.73e-01 |
Pos logit contributions | |
rily | +0.460 |
avail | +0.455 |
ittered | +0.450 |
leness | +0.442 |
footing | +0.441 |
Neg logit contributions | |
a | -0.361 |
| -0.360 |
two | -0.329 |
and | -0.329 |
big | -0.328 |
Token | and |
Token ID | 290 |
Feature activation | -4.42e-02 |
Pos logit contributions | |
rily | +1.564 |
leness | +1.562 |
lights | +1.558 |
ittered | +1.551 |
keeping | +1.546 |
Neg logit contributions | |
a | -0.257 |
| -0.230 |
big | -0.202 |
two | -0.199 |
the | -0.157 |
Token | often |
Token ID | 1690 |
Feature activation | -6.23e-01 |
Pos logit contributions | |
avail | +0.160 |
leness | +0.157 |
inations | +0.157 |
ittered | +0.156 |
lights | +0.156 |
Neg logit contributions | |
a | -0.053 |
| -0.053 |
big | -0.051 |
and | -0.050 |
two | -0.047 |
Token | threw |
Token ID | 9617 |
Feature activation | +6.24e-01 |
Pos logit contributions | |
avail | +2.058 |
ittered | +2.037 |
leness | +2.009 |
rily | +1.991 |
fer | +1.991 |
Neg logit contributions | |
a | -1.022 |
| -0.976 |
big | -0.966 |
and | -0.921 |
two | -0.908 |
Token | temper |
Token ID | 4124 |
Feature activation | +2.14e-01 |
Pos logit contributions | |
| +0.932 |
a | +0.895 |
big | +0.850 |
and | +0.840 |
two | +0.835 |
Neg logit contributions | |
avail | -2.110 |
rily | -2.093 |
keeping | -2.058 |
leness | -2.025 |
lights | -2.015 |
Token | tant |
Token ID | 24246 |
Feature activation | -3.21e+00 |
Pos logit contributions | |
a | +0.489 |
| +0.474 |
big | +0.461 |
two | +0.454 |
and | +0.432 |
Neg logit contributions | |
rily | -0.571 |
leness | -0.570 |
lights | -0.566 |
ittered | -0.561 |
keeping | -0.557 |
Token | rums |
Token ID | 45241 |
Feature activation | -2.28e-01 |
Pos logit contributions | |
avail | +3.428 |
rily | +3.216 |
ittered | +3.029 |
lights | +2.910 |
immer | +2.886 |
Neg logit contributions | |
a | -13.346 |
and | -12.883 |
| -12.782 |
big | -12.772 |
two | -12.549 |
Token | . |
Token ID | 13 |
Feature activation | -1.29e-01 |
Pos logit contributions | |
rily | +0.911 |
avail | +0.894 |
ittered | +0.889 |
keeping | +0.886 |
lights | +0.883 |
Neg logit contributions | |
big | -0.220 |
a | -0.220 |
| -0.199 |
two | -0.178 |
and | -0.172 |
Token | One |
Token ID | 1881 |
Feature activation | +5.00e-01 |
Pos logit contributions | |
avail | +0.414 |
rily | +0.406 |
lights | +0.399 |
keeping | +0.391 |
ently | +0.389 |
Neg logit contributions | |
a | -0.222 |
big | -0.218 |
and | -0.206 |
two | -0.197 |
the | -0.186 |
Token | day |
Token ID | 1110 |
Feature activation | -2.59e-01 |
Pos logit contributions | |
a | +0.744 |
| +0.716 |
big | +0.661 |
two | +0.631 |
and | +0.631 |
Neg logit contributions | |
rated | -1.691 |
avail | -1.684 |
footing | -1.674 |
lights | -1.653 |
inations | -1.628 |
Token | , |
Token ID | 11 |
Feature activation | -3.74e-01 |
Pos logit contributions | |
avail | +1.058 |
rily | +1.014 |
ittered | +1.004 |
fer | +1.001 |
footing | +0.998 |
Neg logit contributions | |
big | -0.212 |
a | -0.205 |
| -0.193 |
and | -0.182 |
two | -0.164 |
Token | to |
Token ID | 284 |
Feature activation | -6.95e-01 |
Pos logit contributions | |
avail | +0.888 |
keeping | +0.868 |
ge | +0.861 |
rily | +0.861 |
ittered | +0.859 |
Neg logit contributions | |
a | -0.156 |
| -0.144 |
big | -0.141 |
two | -0.135 |
and | -0.119 |
Token | cry |
Token ID | 3960 |
Feature activation | +4.88e-01 |
Pos logit contributions | |
avail | +1.890 |
keeping | +1.718 |
rated | +1.713 |
rily | +1.703 |
immer | +1.698 |
Neg logit contributions | |
a | -1.595 |
| -1.559 |
big | -1.525 |
and | -1.503 |
two | -1.477 |
Token | and |
Token ID | 290 |
Feature activation | -3.84e-01 |
Pos logit contributions | |
a | +0.386 |
big | +0.379 |
two | +0.356 |
| +0.340 |
the | +0.279 |
Neg logit contributions | |
rily | -1.957 |
lights | -1.918 |
avail | -1.909 |
immer | -1.904 |
ittered | -1.882 |
Token | threw |
Token ID | 9617 |
Feature activation | +5.49e-01 |
Pos logit contributions | |
avail | +1.344 |
leness | +1.242 |
ittered | +1.240 |
rily | +1.238 |
keeping | +1.228 |
Neg logit contributions | |
a | -0.560 |
big | -0.552 |
| -0.547 |
and | -0.530 |
two | -0.517 |
Token | a |
Token ID | 257 |
Feature activation | -4.10e-01 |
Pos logit contributions | |
| +0.576 |
big | +0.541 |
a | +0.528 |
and | +0.506 |
two | +0.499 |
Neg logit contributions | |
avail | -2.099 |
rily | -2.013 |
keeping | -1.987 |
leness | -1.952 |
ittered | -1.945 |
Token | tant |
Token ID | 24246 |
Feature activation | -3.20e+00 |
Pos logit contributions | |
rily | +1.155 |
avail | +1.152 |
immer | +1.122 |
leness | +1.116 |
ittered | +1.112 |
Neg logit contributions | |
a | -0.849 |
| -0.804 |
two | -0.777 |
and | -0.777 |
the | -0.737 |
Token | rum |
Token ID | 6582 |
Feature activation | -3.06e-01 |
Pos logit contributions | |
avail | +7.559 |
rily | +7.318 |
ittered | +7.164 |
leness | +7.043 |
lights | +7.021 |
Neg logit contributions | |
a | -9.134 |
and | -8.700 |
big | -8.612 |
| -8.582 |
two | -8.323 |
Token | . |
Token ID | 13 |
Feature activation | -1.44e-01 |
Pos logit contributions | |
rily | +1.335 |
ittered | +1.305 |
avail | +1.302 |
lights | +1.298 |
ently | +1.291 |
Neg logit contributions | |
big | -0.168 |
a | -0.164 |
| -0.141 |
two | -0.121 |
little | -0.096 |
Token | She |
Token ID | 1375 |
Feature activation | -9.92e-02 |
Pos logit contributions | |
avail | +0.425 |
rily | +0.421 |
ittered | +0.408 |
lights | +0.407 |
ently | +0.405 |
Neg logit contributions | |
a | -0.279 |
big | -0.266 |
two | -0.258 |
and | -0.256 |
| -0.249 |
Token | didn |
Token ID | 1422 |
Feature activation | -3.17e-01 |
Pos logit contributions | |
avail | +0.298 |
ittered | +0.291 |
aned | +0.291 |
rily | +0.289 |
inations | +0.289 |
Neg logit contributions | |
a | -0.182 |
| -0.176 |
big | -0.174 |
two | -0.165 |
and | -0.161 |
Token | 't |
Token ID | 470 |
Feature activation | -1.14e-01 |
Pos logit contributions | |
avail | +0.856 |
rily | +0.847 |
aned | +0.827 |
ittered | +0.826 |
ently | +0.825 |
Neg logit contributions | |
a | -0.671 |
big | -0.639 |
and | -0.632 |
two | -0.623 |
little | -0.612 |
Token | sad |
Token ID | 6507 |
Feature activation | +6.38e-01 |
Pos logit contributions | |
a | +0.280 |
| +0.269 |
big | +0.265 |
and | +0.261 |
two | +0.259 |
Neg logit contributions | |
avail | -0.410 |
leness | -0.408 |
rily | -0.404 |
rated | -0.397 |
aned | -0.396 |
Token | . |
Token ID | 13 |
Feature activation | +1.00e-01 |
Pos logit contributions | |
big | +0.531 |
a | +0.521 |
two | +0.457 |
| +0.401 |
little | +0.340 |
Neg logit contributions | |
lights | -2.577 |
rily | -2.573 |
iously | -2.540 |
ently | -2.539 |
avail | -2.531 |
Token | She |
Token ID | 1375 |
Feature activation | +4.55e-01 |
Pos logit contributions | |
a | +0.129 |
big | +0.119 |
two | +0.108 |
and | +0.106 |
little | +0.096 |
Neg logit contributions | |
avail | -0.367 |
keeping | -0.356 |
ently | -0.354 |
rily | -0.353 |
aned | -0.352 |
Token | threw |
Token ID | 9617 |
Feature activation | +5.39e-01 |
Pos logit contributions | |
a | +0.885 |
big | +0.841 |
| +0.831 |
two | +0.797 |
and | +0.777 |
Neg logit contributions | |
aned | -1.338 |
rated | -1.322 |
avail | -1.317 |
leness | -1.306 |
rily | -1.302 |
Token | a |
Token ID | 257 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
| +0.397 |
a | +0.336 |
big | +0.332 |
and | +0.293 |
two | +0.278 |
Neg logit contributions | |
avail | -2.229 |
rily | -2.183 |
ittered | -2.113 |
keeping | -2.103 |
leness | -2.098 |
Token | tant |
Token ID | 24246 |
Feature activation | -3.19e+00 |
Pos logit contributions | |
rily | +0.220 |
avail | +0.213 |
immer | +0.212 |
leness | +0.212 |
ittered | +0.209 |
Neg logit contributions | |
a | -0.275 |
| -0.264 |
two | -0.255 |
and | -0.254 |
the | -0.247 |
Token | rum |
Token ID | 6582 |
Feature activation | -2.34e-01 |
Pos logit contributions | |
avail | +6.413 |
rily | +6.166 |
ittered | +5.999 |
leness | +5.878 |
lights | +5.868 |
Neg logit contributions | |
a | -10.270 |
and | -9.837 |
big | -9.739 |
| -9.719 |
two | -9.466 |
Token | and |
Token ID | 290 |
Feature activation | -2.78e-01 |
Pos logit contributions | |
rily | +1.022 |
ittered | +1.003 |
lights | +1.000 |
keeping | +0.999 |
ently | +0.999 |
Neg logit contributions | |
big | -0.127 |
a | -0.123 |
| -0.097 |
two | -0.082 |
little | -0.069 |
Token | cried |
Token ID | 16896 |
Feature activation | +2.95e-01 |
Pos logit contributions | |
avail | +0.865 |
keeping | +0.826 |
leness | +0.823 |
rily | +0.818 |
ittered | +0.813 |
Neg logit contributions | |
a | -0.510 |
| -0.505 |
big | -0.505 |
and | -0.468 |
two | -0.467 |
Token | loudly |
Token ID | 23112 |
Feature activation | +1.62e-01 |
Pos logit contributions | |
| +0.246 |
big | +0.246 |
a | +0.241 |
two | +0.219 |
and | +0.191 |
Neg logit contributions | |
rily | -1.152 |
ittered | -1.142 |
avail | -1.128 |
lights | -1.127 |
immer | -1.127 |
Token | . |
Token ID | 13 |
Feature activation | -2.84e-02 |
Pos logit contributions | |
a | +0.100 |
big | +0.096 |
| +0.092 |
two | +0.077 |
and | +0.066 |
Neg logit contributions | |
rily | -0.694 |
ittered | -0.684 |
avail | -0.676 |
keeping | -0.673 |
iously | -0.670 |
Token | her |
Token ID | 607 |
Feature activation | -4.67e-01 |
Pos logit contributions | |
| +0.193 |
a | +0.192 |
big | +0.184 |
two | +0.178 |
and | +0.157 |
Neg logit contributions | |
rily | -0.696 |
avail | -0.683 |
keeping | -0.676 |
ittered | -0.665 |
iously | -0.660 |
Token | feet |
Token ID | 3625 |
Feature activation | +3.27e-01 |
Pos logit contributions | |
rily | +1.281 |
avail | +1.261 |
footing | +1.258 |
lights | +1.227 |
ittered | +1.222 |
Neg logit contributions | |
a | -1.056 |
| -1.007 |
big | -0.953 |
two | -0.946 |
and | -0.930 |
Token | and |
Token ID | 290 |
Feature activation | -3.96e-02 |
Pos logit contributions | |
a | +0.168 |
| +0.154 |
big | +0.132 |
two | +0.116 |
the | +0.074 |
Neg logit contributions | |
rily | -1.431 |
ge | -1.411 |
ittered | -1.410 |
immer | -1.404 |
footing | -1.396 |
Token | threw |
Token ID | 9617 |
Feature activation | +5.85e-01 |
Pos logit contributions | |
avail | +0.114 |
rily | +0.111 |
leness | +0.109 |
ittered | +0.108 |
fer | +0.107 |
Neg logit contributions | |
a | -0.079 |
| -0.079 |
big | -0.078 |
and | -0.074 |
two | -0.073 |
Token | a |
Token ID | 257 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
| +0.695 |
a | +0.646 |
big | +0.642 |
and | +0.615 |
two | +0.573 |
Neg logit contributions | |
rily | -2.170 |
avail | -2.150 |
ittered | -2.066 |
keeping | -2.066 |
lights | -2.042 |
Token | tant |
Token ID | 24246 |
Feature activation | -3.15e+00 |
Pos logit contributions | |
rily | +0.180 |
leness | +0.169 |
avail | +0.168 |
immer | +0.164 |
ittered | +0.162 |
Neg logit contributions | |
a | -0.418 |
| -0.406 |
and | -0.394 |
two | -0.392 |
the | -0.384 |
Token | rum |
Token ID | 6582 |
Feature activation | -2.67e-01 |
Pos logit contributions | |
avail | +6.460 |
rily | +6.210 |
ittered | +6.066 |
leness | +5.946 |
lights | +5.922 |
Neg logit contributions | |
a | -9.960 |
and | -9.535 |
big | -9.450 |
| -9.423 |
two | -9.169 |
Token | . |
Token ID | 13 |
Feature activation | +1.97e-03 |
Pos logit contributions | |
rily | +1.173 |
ittered | +1.145 |
lights | +1.140 |
keeping | +1.138 |
avail | +1.137 |
Neg logit contributions | |
big | -0.150 |
a | -0.149 |
| -0.128 |
two | -0.103 |
little | -0.089 |
Token | She |
Token ID | 1375 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
a | +0.004 |
big | +0.003 |
and | +0.003 |
two | +0.003 |
| +0.003 |
Neg logit contributions | |
rily | -0.006 |
avail | -0.006 |
ittered | -0.006 |
ently | -0.006 |
lights | -0.006 |
Token | grabbed |
Token ID | 13176 |
Feature activation | +3.23e-01 |
Pos logit contributions | |
a | +0.218 |
| +0.213 |
big | +0.206 |
two | +0.199 |
and | +0.195 |
Neg logit contributions | |
rily | -0.292 |
avail | -0.292 |
aned | -0.291 |
ittered | -0.291 |
leness | -0.289 |
Token | the |
Token ID | 262 |
Feature activation | -2.91e-01 |
Pos logit contributions | |
| +0.241 |
and | +0.215 |
big | +0.202 |
a | +0.192 |
. | +0.160 |
Neg logit contributions | |
rily | -1.348 |
avail | -1.344 |
ittered | -1.311 |
lights | -1.273 |
immer | -1.269 |
Token | and |
Token ID | 290 |
Feature activation | -6.29e-02 |
Pos logit contributions | |
big | +0.312 |
a | +0.297 |
| +0.249 |
two | +0.224 |
the | +0.150 |
Neg logit contributions | |
rily | -2.102 |
iously | -2.077 |
lights | -2.064 |
avail | -2.060 |
ently | -2.060 |
Token | started |
Token ID | 2067 |
Feature activation | -5.59e-02 |
Pos logit contributions | |
avail | +0.194 |
rily | +0.192 |
leness | +0.189 |
aned | +0.188 |
rated | +0.187 |
Neg logit contributions | |
| -0.111 |
a | -0.110 |
big | -0.106 |
and | -0.102 |
two | -0.101 |
Token | to |
Token ID | 284 |
Feature activation | -7.98e-01 |
Pos logit contributions | |
avail | +0.225 |
rily | +0.224 |
keeping | +0.222 |
ittered | +0.219 |
ge | +0.219 |
Neg logit contributions | |
a | -0.043 |
big | -0.038 |
two | -0.036 |
| -0.036 |
and | -0.034 |
Token | throw |
Token ID | 3714 |
Feature activation | +5.12e-01 |
Pos logit contributions | |
avail | +1.786 |
rily | +1.646 |
leness | +1.625 |
rated | +1.624 |
ittered | +1.592 |
Neg logit contributions | |
a | -2.215 |
| -2.139 |
big | -2.120 |
and | -2.102 |
two | -2.078 |
Token | a |
Token ID | 257 |
Feature activation | -7.28e-01 |
Pos logit contributions | |
| +0.516 |
a | +0.501 |
big | +0.489 |
two | +0.453 |
and | +0.447 |
Neg logit contributions | |
rily | -2.003 |
avail | -1.967 |
leness | -1.854 |
keeping | -1.854 |
ittered | -1.852 |
Token | tant |
Token ID | 24246 |
Feature activation | -3.15e+00 |
Pos logit contributions | |
rily | +2.073 |
avail | +2.004 |
immer | +1.953 |
leness | +1.938 |
inations | +1.938 |
Neg logit contributions | |
a | -1.588 |
| -1.490 |
and | -1.451 |
two | -1.442 |
the | -1.377 |
Token | rum |
Token ID | 6582 |
Feature activation | -2.92e-01 |
Pos logit contributions | |
avail | +7.286 |
rily | +7.061 |
ittered | +6.922 |
leness | +6.784 |
lights | +6.765 |
Neg logit contributions | |
a | -9.116 |
and | -8.685 |
big | -8.606 |
| -8.585 |
two | -8.324 |
Token | . |
Token ID | 13 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
rily | +1.281 |
avail | +1.248 |
ittered | +1.244 |
ently | +1.241 |
lights | +1.241 |
Neg logit contributions | |
big | -0.163 |
a | -0.157 |
| -0.132 |
two | -0.111 |
little | -0.096 |
Token | Suddenly |
Token ID | 24975 |
Feature activation | -3.53e-01 |
Pos logit contributions | |
a | +0.238 |
big | +0.226 |
two | +0.218 |
and | +0.217 |
the | +0.205 |
Neg logit contributions | |
avail | -0.311 |
rily | -0.307 |
ently | -0.303 |
aned | -0.303 |
ittered | -0.298 |
Token | , |
Token ID | 11 |
Feature activation | -2.15e-01 |
Pos logit contributions | |
avail | +1.493 |
ittered | +1.427 |
rily | +1.406 |
lights | +1.404 |
leness | +1.401 |
Neg logit contributions | |
a | -0.234 |
| -0.231 |
big | -0.210 |
and | -0.191 |
two | -0.173 |
Token | she |
Token ID | 673 |
Feature activation | -2.95e-02 |
Pos logit contributions | |
avail | +0.805 |
rily | +0.748 |
lights | +0.744 |
ittered | +0.739 |
footing | +0.739 |
Neg logit contributions | |
| -0.257 |
and | -0.253 |
big | -0.239 |
a | -0.231 |
. | -0.206 |
Token | his |
Token ID | 465 |
Feature activation | -4.86e-01 |
Pos logit contributions | |
big | +0.208 |
a | +0.206 |
| +0.204 |
and | +0.187 |
two | +0.182 |
Neg logit contributions | |
rily | -0.626 |
avail | -0.616 |
ittered | -0.610 |
lights | -0.602 |
immer | -0.595 |
Token | feet |
Token ID | 3625 |
Feature activation | +3.86e-01 |
Pos logit contributions | |
rily | +1.282 |
avail | +1.254 |
footing | +1.233 |
ittered | +1.211 |
keeping | +1.206 |
Neg logit contributions | |
a | -1.156 |
| -1.123 |
big | -1.049 |
two | -1.040 |
and | -1.040 |
Token | and |
Token ID | 290 |
Feature activation | -8.25e-02 |
Pos logit contributions | |
a | +0.183 |
| +0.158 |
big | +0.145 |
two | +0.121 |
the | +0.081 |
Neg logit contributions | |
rily | -1.727 |
ge | -1.701 |
immer | -1.684 |
ittered | -1.667 |
footing | -1.657 |
Token | threw |
Token ID | 9617 |
Feature activation | +5.80e-01 |
Pos logit contributions | |
avail | +0.237 |
rily | +0.235 |
ittered | +0.232 |
leness | +0.231 |
fer | +0.228 |
Neg logit contributions | |
a | -0.162 |
| -0.159 |
big | -0.157 |
and | -0.151 |
two | -0.150 |
Token | a |
Token ID | 257 |
Feature activation | -1.19e-01 |
Pos logit contributions | |
| +0.705 |
big | +0.660 |
a | +0.657 |
and | +0.633 |
two | +0.591 |
Neg logit contributions | |
rily | -2.153 |
avail | -2.087 |
keeping | -2.007 |
lights | -2.005 |
ittered | -2.003 |
Token | tant |
Token ID | 24246 |
Feature activation | -3.14e+00 |
Pos logit contributions | |
rily | +0.327 |
avail | +0.315 |
leness | +0.313 |
immer | +0.312 |
ittered | +0.308 |
Neg logit contributions | |
a | -0.270 |
| -0.254 |
and | -0.245 |
two | -0.242 |
the | -0.237 |
Token | rum |
Token ID | 6582 |
Feature activation | -3.05e-01 |
Pos logit contributions | |
avail | +4.224 |
rily | +4.014 |
ittered | +3.863 |
lights | +3.733 |
leness | +3.721 |
Neg logit contributions | |
a | -12.122 |
and | -11.688 |
big | -11.597 |
| -11.579 |
two | -11.354 |
Token | . |
Token ID | 13 |
Feature activation | +2.78e-01 |
Pos logit contributions | |
rily | +1.359 |
ittered | +1.324 |
lights | +1.316 |
keeping | +1.302 |
ently | +1.299 |
Neg logit contributions | |
big | -0.169 |
a | -0.168 |
| -0.132 |
two | -0.113 |
little | -0.101 |
Token |
|
Token ID | 198 |
Feature activation | +3.98e-01 |
Pos logit contributions | |
a | +0.265 |
big | +0.253 |
and | +0.246 |
two | +0.214 |
little | +0.194 |
Neg logit contributions | |
rily | -1.086 |
avail | -1.068 |
lights | -1.052 |
ittered | -1.039 |
fer | -1.025 |
Token |
|
Token ID | 198 |
Feature activation | +5.02e-01 |
Pos logit contributions | |
a | +0.208 |
and | +0.165 |
big | +0.162 |
two | +0.133 |
little | +0.104 |
Neg logit contributions | |
avail | -1.693 |
rily | -1.679 |
lights | -1.678 |
igi | -1.670 |
immer | -1.667 |
Token | " |
Token ID | 1 |
Feature activation | -3.46e-01 |
Pos logit contributions | |
a | +0.903 |
and | +0.849 |
big | +0.844 |
all | +0.776 |
the | +0.755 |
Neg logit contributions | |
avail | -1.568 |
leness | -1.563 |
fer | -1.558 |
igi | -1.530 |
immer | -1.526 |
Token | kept |
Token ID | 4030 |
Feature activation | -9.10e-02 |
Pos logit contributions | |
a | +0.459 |
| +0.444 |
big | +0.435 |
two | +0.405 |
and | +0.403 |
Neg logit contributions | |
avail | -0.740 |
aned | -0.736 |
rily | -0.730 |
rated | -0.727 |
ittered | -0.724 |
Token | crying |
Token ID | 13774 |
Feature activation | +1.25e-01 |
Pos logit contributions | |
avail | +0.219 |
rily | +0.204 |
keeping | +0.202 |
leness | +0.198 |
ittered | +0.196 |
Neg logit contributions | |
| -0.229 |
a | -0.229 |
big | -0.225 |
and | -0.218 |
two | -0.214 |
Token | and |
Token ID | 290 |
Feature activation | -3.68e-01 |
Pos logit contributions | |
a | +0.083 |
big | +0.071 |
| +0.068 |
two | +0.061 |
the | +0.046 |
Neg logit contributions | |
rily | -0.529 |
lights | -0.525 |
avail | -0.519 |
ently | -0.518 |
leness | -0.514 |
Token | throwing |
Token ID | 9644 |
Feature activation | +4.34e-01 |
Pos logit contributions | |
avail | +1.056 |
ittered | +1.008 |
keeping | +1.000 |
leness | +0.995 |
footing | +0.993 |
Neg logit contributions | |
a | -0.765 |
big | -0.760 |
| -0.732 |
and | -0.712 |
two | -0.698 |
Token | a |
Token ID | 257 |
Feature activation | -1.33e-01 |
Pos logit contributions | |
| +0.469 |
a | +0.465 |
big | +0.451 |
two | +0.421 |
and | +0.402 |
Neg logit contributions | |
rily | -1.640 |
avail | -1.626 |
lights | -1.569 |
leness | -1.562 |
ittered | -1.562 |
Token | tant |
Token ID | 24246 |
Feature activation | -3.10e+00 |
Pos logit contributions | |
rily | +0.305 |
lights | +0.291 |
immer | +0.288 |
avail | +0.284 |
leness | +0.284 |
Neg logit contributions | |
a | -0.366 |
| -0.339 |
two | -0.335 |
and | -0.329 |
the | -0.325 |
Token | rum |
Token ID | 6582 |
Feature activation | -3.42e-01 |
Pos logit contributions | |
avail | +6.671 |
rily | +6.449 |
ittered | +6.298 |
leness | +6.184 |
lights | +6.152 |
Neg logit contributions | |
a | -9.490 |
and | -9.050 |
big | -8.986 |
| -8.960 |
two | -8.710 |
Token | . |
Token ID | 13 |
Feature activation | +3.75e-01 |
Pos logit contributions | |
rily | +1.480 |
ittered | +1.464 |
lights | +1.458 |
avail | +1.450 |
ently | +1.445 |
Neg logit contributions | |
big | -0.197 |
a | -0.195 |
| -0.163 |
two | -0.141 |
little | -0.120 |
Token |
|
Token ID | 198 |
Feature activation | -1.99e-02 |
Pos logit contributions | |
a | +0.528 |
big | +0.510 |
and | +0.481 |
two | +0.454 |
the | +0.425 |
Neg logit contributions | |
avail | -1.274 |
rily | -1.270 |
ently | -1.259 |
ittered | -1.256 |
lights | -1.246 |
Token |
|
Token ID | 198 |
Feature activation | +2.69e-01 |
Pos logit contributions | |
avail | +0.085 |
ently | +0.085 |
rily | +0.085 |
ittered | +0.085 |
inations | +0.084 |
Neg logit contributions | |
a | -0.010 |
big | -0.007 |
and | -0.007 |
two | -0.007 |
the | -0.004 |
Token | Finally |
Token ID | 11158 |
Feature activation | -3.29e-01 |
Pos logit contributions | |
a | +0.742 |
and | +0.699 |
big | +0.698 |
the | +0.673 |
two | +0.667 |
Neg logit contributions | |
leness | -0.583 |
aned | -0.580 |
immer | -0.576 |
avail | -0.572 |
ently | -0.568 |
Token | them |
Token ID | 606 |
Feature activation | -2.20e-01 |
Pos logit contributions | |
avail | +0.098 |
rily | +0.095 |
leness | +0.094 |
keeping | +0.094 |
inations | +0.093 |
Neg logit contributions | |
| -0.030 |
a | -0.030 |
big | -0.030 |
and | -0.029 |
two | -0.027 |
Token | and |
Token ID | 290 |
Feature activation | +9.14e-03 |
Pos logit contributions | |
rily | +0.901 |
ittered | +0.895 |
inations | +0.890 |
keeping | +0.888 |
immer | +0.883 |
Neg logit contributions | |
a | -0.168 |
| -0.157 |
big | -0.156 |
two | -0.139 |
and | -0.124 |
Token | never |
Token ID | 1239 |
Feature activation | -1.26e-01 |
Pos logit contributions | |
a | +0.014 |
big | +0.013 |
| +0.013 |
two | +0.013 |
and | +0.012 |
Neg logit contributions | |
avail | -0.031 |
keeping | -0.031 |
leness | -0.030 |
rily | -0.030 |
inations | -0.030 |
Token | threw |
Token ID | 9617 |
Feature activation | +3.42e-01 |
Pos logit contributions | |
avail | +0.377 |
rated | +0.365 |
leness | +0.362 |
ittered | +0.357 |
keeping | +0.354 |
Neg logit contributions | |
a | -0.246 |
| -0.235 |
big | -0.232 |
and | -0.231 |
two | -0.225 |
Token | a |
Token ID | 257 |
Feature activation | -7.10e-01 |
Pos logit contributions | |
| +0.553 |
a | +0.547 |
big | +0.531 |
two | +0.512 |
and | +0.512 |
Neg logit contributions | |
rily | -1.118 |
avail | -1.109 |
keeping | -1.088 |
ittered | -1.063 |
leness | -1.052 |
Token | tant |
Token ID | 24246 |
Feature activation | -3.06e+00 |
Pos logit contributions | |
rily | +1.956 |
avail | +1.936 |
leness | +1.908 |
immer | +1.899 |
inations | +1.893 |
Neg logit contributions | |
a | -1.570 |
| -1.473 |
two | -1.427 |
and | -1.421 |
the | -1.375 |
Token | rum |
Token ID | 6582 |
Feature activation | -2.06e-01 |
Pos logit contributions | |
avail | +5.138 |
rily | +4.939 |
ittered | +4.788 |
lights | +4.646 |
leness | +4.645 |
Neg logit contributions | |
a | -10.793 |
and | -10.358 |
big | -10.281 |
| -10.264 |
two | -10.042 |
Token | again |
Token ID | 757 |
Feature activation | +2.27e-01 |
Pos logit contributions | |
rily | +0.894 |
ittered | +0.882 |
lights | +0.878 |
keeping | +0.874 |
avail | +0.872 |
Neg logit contributions | |
a | -0.133 |
big | -0.131 |
| -0.103 |
two | -0.093 |
and | -0.088 |
Token | . |
Token ID | 13 |
Feature activation | +7.40e-02 |
Pos logit contributions | |
a | +0.143 |
big | +0.126 |
two | +0.110 |
| +0.108 |
and | +0.081 |
Neg logit contributions | |
avail | -0.991 |
rily | -0.967 |
leness | -0.958 |
keeping | -0.957 |
ittered | -0.948 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +3.83e-01 |
Pos logit contributions | |
a | +0.090 |
big | +0.088 |
and | +0.077 |
two | +0.075 |
the | +0.067 |
Neg logit contributions | |
rily | -0.275 |
avail | -0.270 |
keeping | -0.268 |
lights | -0.265 |
leness | -0.265 |
Token | Once |
Token ID | 7454 |
Feature activation | +3.67e-02 |
Pos logit contributions | |
a | +0.182 |
big | +0.164 |
two | +0.137 |
and | +0.128 |
little | +0.126 |
Neg logit contributions | |
keeping | -1.637 |
avail | -1.613 |
ittered | -1.601 |
rily | -1.597 |
lights | -1.573 |
Token | the |
Token ID | 262 |
Feature activation | +3.62e-01 |
Pos logit contributions | |
rily | +0.596 |
avail | +0.590 |
ittered | +0.569 |
immer | +0.567 |
footing | +0.557 |
Neg logit contributions | |
big | -0.144 |
a | -0.139 |
two | -0.131 |
| -0.130 |
and | -0.120 |
Token | squirrel |
Token ID | 33039 |
Feature activation | -2.26e-01 |
Pos logit contributions | |
a | +0.816 |
and | +0.756 |
| +0.734 |
the | +0.717 |
two | +0.688 |
Neg logit contributions | |
avail | -0.995 |
leness | -0.968 |
rated | -0.932 |
rily | -0.928 |
keeping | -0.926 |
Token | as |
Token ID | 355 |
Feature activation | -1.01e+00 |
Pos logit contributions | |
rily | +0.900 |
immer | +0.882 |
ge | +0.879 |
ittered | +0.878 |
avail | +0.876 |
Neg logit contributions | |
a | -0.219 |
big | -0.187 |
two | -0.177 |
| -0.176 |
and | -0.151 |
Token | it |
Token ID | 340 |
Feature activation | -6.65e-02 |
Pos logit contributions | |
avail | +3.802 |
rily | +3.766 |
ittered | +3.691 |
lights | +3.623 |
leness | +3.601 |
Neg logit contributions | |
a | -1.153 |
and | -1.060 |
| -1.052 |
big | -1.048 |
two | -1.010 |
Token | sc |
Token ID | 629 |
Feature activation | -1.76e+00 |
Pos logit contributions | |
ittered | +0.196 |
avail | +0.192 |
rily | +0.188 |
immer | +0.186 |
lights | +0.182 |
Neg logit contributions | |
a | -0.134 |
| -0.126 |
big | -0.124 |
and | -0.123 |
two | -0.120 |
Token | ur |
Token ID | 333 |
Feature activation | -3.06e+00 |
Pos logit contributions | |
ittered | +2.429 |
avail | +2.380 |
rily | +2.344 |
immer | +2.289 |
leness | +2.266 |
Neg logit contributions | |
a | -6.519 |
big | -6.278 |
and | -6.210 |
| -6.180 |
two | -6.114 |
Token | ried |
Token ID | 2228 |
Feature activation | +6.12e-01 |
Pos logit contributions | |
avail | +4.473 |
rily | +4.226 |
ittered | +4.076 |
leness | +4.046 |
lights | +3.952 |
Neg logit contributions | |
a | -11.604 |
and | -11.158 |
big | -11.091 |
| -11.033 |
two | -10.843 |
Token | around |
Token ID | 1088 |
Feature activation | +3.12e-01 |
Pos logit contributions | |
a | +1.326 |
big | +1.297 |
| +1.282 |
two | +1.269 |
little | +1.232 |
Neg logit contributions | |
rily | -1.718 |
avail | -1.653 |
ittered | -1.620 |
immer | -1.607 |
keeping | -1.600 |
Token | . |
Token ID | 13 |
Feature activation | +1.78e-01 |
Pos logit contributions | |
big | +0.233 |
a | +0.218 |
two | +0.206 |
| +0.176 |
little | +0.163 |
Neg logit contributions | |
avail | -1.280 |
rily | -1.273 |
ittered | -1.264 |
keeping | -1.260 |
fer | -1.227 |
Token | |
Token ID | 220 |
Feature activation | +1.27e-02 |
Pos logit contributions | |
a | +0.133 |
big | +0.121 |
two | +0.111 |
and | +0.102 |
little | +0.079 |
Neg logit contributions | |
rily | -0.719 |
avail | -0.708 |
ittered | -0.706 |
keeping | -0.705 |
immer | -0.692 |
Token |
|
Token ID | 198 |
Feature activation | -2.35e-01 |
Pos logit contributions | |
a | +0.006 |
big | +0.005 |
two | +0.005 |
and | +0.004 |
little | +0.003 |
Neg logit contributions | |
rily | -0.055 |
ittered | -0.055 |
immer | -0.054 |
ently | -0.054 |
avail | -0.053 |
Token | a |
Token ID | 257 |
Feature activation | -6.12e-01 |
Pos logit contributions | |
| +0.258 |
big | +0.243 |
a | +0.221 |
and | +0.211 |
two | +0.177 |
Neg logit contributions | |
rily | -1.502 |
avail | -1.481 |
ittered | -1.453 |
keeping | -1.438 |
immer | -1.431 |
Token | sign |
Token ID | 1051 |
Feature activation | -6.17e-02 |
Pos logit contributions | |
avail | +1.869 |
rily | +1.829 |
immer | +1.768 |
leness | +1.767 |
fer | +1.750 |
Neg logit contributions | |
a | -1.172 |
| -1.121 |
and | -1.076 |
two | -1.052 |
the | -1.024 |
Token | that |
Token ID | 326 |
Feature activation | -2.10e-01 |
Pos logit contributions | |
rily | +0.258 |
lights | +0.254 |
avail | +0.252 |
immer | +0.251 |
ittered | +0.248 |
Neg logit contributions | |
a | -0.044 |
| -0.040 |
big | -0.036 |
two | -0.034 |
little | -0.028 |
Token | said |
Token ID | 531 |
Feature activation | -1.17e-01 |
Pos logit contributions | |
avail | +0.735 |
lights | +0.717 |
ittered | +0.713 |
rily | +0.708 |
leness | +0.705 |
Neg logit contributions | |
a | -0.290 |
and | -0.268 |
big | -0.264 |
| -0.263 |
two | -0.255 |
Token | " |
Token ID | 366 |
Feature activation | -3.56e-01 |
Pos logit contributions | |
rily | +0.430 |
ittered | +0.429 |
avail | +0.427 |
immer | +0.420 |
fer | +0.420 |
Neg logit contributions | |
a | -0.133 |
big | -0.128 |
| -0.126 |
and | -0.125 |
two | -0.117 |
Token | Hom |
Token ID | 28718 |
Feature activation | -3.04e+00 |
Pos logit contributions | |
avail | +0.795 |
rily | +0.766 |
keeping | +0.760 |
fer | +0.758 |
leness | +0.757 |
Neg logit contributions | |
a | -1.002 |
big | -0.970 |
and | -0.945 |
| -0.917 |
two | -0.912 |
Token | eless |
Token ID | 5321 |
Feature activation | -3.61e-01 |
Pos logit contributions | |
avail | +4.988 |
rily | +4.843 |
ittered | +4.669 |
leness | +4.579 |
lights | +4.540 |
Neg logit contributions | |
a | -10.876 |
and | -10.411 |
| -10.343 |
big | -10.334 |
two | -10.134 |
Token | and |
Token ID | 290 |
Feature activation | +3.93e-01 |
Pos logit contributions | |
rily | +1.137 |
avail | +1.128 |
keeping | +1.122 |
fer | +1.108 |
leness | +1.099 |
Neg logit contributions | |
a | -0.667 |
big | -0.626 |
| -0.621 |
and | -0.573 |
the | -0.566 |
Token | Hungry |
Token ID | 42939 |
Feature activation | -1.27e+00 |
Pos logit contributions | |
a | +0.832 |
| +0.813 |
and | +0.813 |
big | +0.772 |
two | +0.734 |
Neg logit contributions | |
avail | -1.092 |
leness | -1.083 |
inations | -1.058 |
rily | -1.039 |
fer | -1.037 |
Token | ". |
Token ID | 1911 |
Feature activation | +2.54e-01 |
Pos logit contributions | |
avail | +3.488 |
rily | +3.333 |
leness | +3.316 |
fer | +3.291 |
ittered | +3.253 |
Neg logit contributions | |
a | -2.968 |
big | -2.854 |
| -2.776 |
and | -2.769 |
two | -2.689 |
Token | Tim |
Token ID | 5045 |
Feature activation | -7.59e-01 |
Pos logit contributions | |
a | +0.402 |
big | +0.396 |
two | +0.386 |
and | +0.344 |
| +0.325 |
Neg logit contributions | |
rily | -0.818 |
keeping | -0.807 |
avail | -0.799 |
lights | -0.798 |
iously | -0.795 |
Token | goat |
Token ID | 26917 |
Feature activation | +1.10e-01 |
Pos logit contributions | |
a | +1.747 |
| +1.717 |
and | +1.615 |
the | +1.594 |
big | +1.589 |
Neg logit contributions | |
avail | -1.130 |
leness | -1.060 |
rily | -1.054 |
igi | -0.985 |
keeping | -0.979 |
Token | looked |
Token ID | 3114 |
Feature activation | -2.35e-01 |
Pos logit contributions | |
a | +0.191 |
big | +0.182 |
| +0.181 |
two | +0.165 |
and | +0.160 |
Neg logit contributions | |
rily | -0.346 |
ittered | -0.344 |
avail | -0.331 |
fer | -0.330 |
rated | -0.330 |
Token | at |
Token ID | 379 |
Feature activation | -3.23e-01 |
Pos logit contributions | |
rily | +0.965 |
ittered | +0.926 |
leness | +0.914 |
immer | +0.914 |
avail | +0.912 |
Neg logit contributions | |
a | -0.203 |
| -0.191 |
two | -0.171 |
big | -0.163 |
and | -0.144 |
Token | Mia |
Token ID | 32189 |
Feature activation | +2.41e-02 |
Pos logit contributions | |
avail | +0.997 |
rily | +0.977 |
lights | +0.963 |
ittered | +0.952 |
eway | +0.912 |
Neg logit contributions | |
a | -0.593 |
big | -0.582 |
and | -0.560 |
| -0.558 |
two | -0.539 |
Token | and |
Token ID | 290 |
Feature activation | -2.12e-01 |
Pos logit contributions | |
a | +0.014 |
big | +0.012 |
| +0.010 |
two | +0.009 |
little | +0.006 |
Neg logit contributions | |
rily | -0.109 |
ittered | -0.104 |
immer | -0.104 |
ge | -0.103 |
ently | -0.102 |
Token | ble |
Token ID | 7245 |
Feature activation | -3.02e+00 |
Pos logit contributions | |
avail | +0.693 |
rily | +0.676 |
ittered | +0.663 |
leness | +0.653 |
lights | +0.653 |
Neg logit contributions | |
big | -0.338 |
a | -0.337 |
| -0.336 |
and | -0.315 |
two | -0.306 |
Token | ated |
Token ID | 515 |
Feature activation | +3.87e-01 |
Pos logit contributions | |
avail | +4.347 |
rily | +4.138 |
ittered | +3.956 |
leness | +3.880 |
lights | +3.831 |
Neg logit contributions | |
a | -11.512 |
and | -11.079 |
big | -11.009 |
| -10.992 |
two | -10.794 |
Token | . |
Token ID | 13 |
Feature activation | +6.46e-02 |
Pos logit contributions | |
a | +0.268 |
| +0.265 |
big | +0.259 |
two | +0.227 |
little | +0.186 |
Neg logit contributions | |
rily | -1.670 |
ittered | -1.596 |
immer | -1.570 |
lights | -1.557 |
rier | -1.554 |
Token | It |
Token ID | 632 |
Feature activation | +1.57e-01 |
Pos logit contributions | |
a | +0.107 |
big | +0.103 |
and | +0.100 |
two | +0.097 |
little | +0.090 |
Neg logit contributions | |
rily | -0.211 |
lights | -0.201 |
ittered | -0.201 |
avail | -0.200 |
eway | -0.196 |
Token | did |
Token ID | 750 |
Feature activation | +1.19e-01 |
Pos logit contributions | |
a | +0.232 |
| +0.223 |
big | +0.211 |
two | +0.209 |
and | +0.207 |
Neg logit contributions | |
ittered | -0.527 |
rily | -0.523 |
lights | -0.511 |
rated | -0.508 |
avail | -0.505 |
Token | not |
Token ID | 407 |
Feature activation | -3.50e-01 |
Pos logit contributions | |
| +0.185 |
two | +0.178 |
big | +0.177 |
a | +0.176 |
and | +0.171 |
Neg logit contributions | |
avail | -0.385 |
ittered | -0.383 |
rily | -0.380 |
lights | -0.371 |
footing | -0.371 |
Token | and |
Token ID | 290 |
Feature activation | -1.69e-01 |
Pos logit contributions | |
rily | +2.330 |
ittered | +2.321 |
avail | +2.318 |
keeping | +2.301 |
lights | +2.274 |
Neg logit contributions | |
a | -0.354 |
big | -0.313 |
| -0.310 |
two | -0.262 |
and | -0.209 |
Token | the |
Token ID | 262 |
Feature activation | +2.46e-01 |
Pos logit contributions | |
avail | +0.599 |
keeping | +0.577 |
rily | +0.576 |
lights | +0.569 |
inations | +0.568 |
Neg logit contributions | |
a | -0.236 |
| -0.235 |
big | -0.224 |
and | -0.220 |
two | -0.198 |
Token | harsh |
Token ID | 11859 |
Feature activation | -3.45e-02 |
Pos logit contributions | |
a | +0.602 |
| +0.591 |
and | +0.578 |
big | +0.544 |
the | +0.532 |
Neg logit contributions | |
avail | -0.630 |
leness | -0.593 |
keeping | -0.590 |
inations | -0.589 |
epad | -0.577 |
Token | queen |
Token ID | 16599 |
Feature activation | -1.89e-01 |
Pos logit contributions | |
avail | +0.085 |
keeping | +0.082 |
leness | +0.081 |
rated | +0.081 |
lights | +0.081 |
Neg logit contributions | |
a | -0.088 |
| -0.084 |
big | -0.081 |
and | -0.079 |
two | -0.079 |
Token | was |
Token ID | 373 |
Feature activation | -4.64e-01 |
Pos logit contributions | |
rily | +0.691 |
ittered | +0.689 |
fer | +0.674 |
inations | +0.671 |
avail | +0.669 |
Neg logit contributions | |
a | -0.247 |
big | -0.227 |
| -0.221 |
two | -0.201 |
and | -0.196 |
Token | overth |
Token ID | 34514 |
Feature activation | -3.01e+00 |
Pos logit contributions | |
rily | +1.460 |
ittered | +1.453 |
avail | +1.430 |
immer | +1.419 |
inations | +1.401 |
Neg logit contributions | |
a | -0.809 |
| -0.795 |
and | -0.775 |
big | -0.761 |
two | -0.741 |
Token | rown |
Token ID | 2053 |
Feature activation | +4.38e-01 |
Pos logit contributions | |
avail | +4.660 |
rily | +4.432 |
ittered | +4.267 |
leness | +4.216 |
lights | +4.127 |
Neg logit contributions | |
a | -11.073 |
and | -10.645 |
| -10.597 |
big | -10.592 |
two | -10.362 |
Token | . |
Token ID | 13 |
Feature activation | -1.85e-01 |
Pos logit contributions | |
a | +0.879 |
big | +0.833 |
| +0.823 |
two | +0.785 |
and | +0.765 |
Neg logit contributions | |
rily | -1.299 |
ittered | -1.282 |
keeping | -1.277 |
iously | -1.256 |
avail | -1.245 |
Token |
|
Token ID | 198 |
Feature activation | -1.99e-01 |
Pos logit contributions | |
rily | +0.711 |
avail | +0.709 |
keeping | +0.696 |
ittered | +0.688 |
leness | +0.685 |
Neg logit contributions | |
a | -0.209 |
big | -0.204 |
and | -0.189 |
two | -0.160 |
little | -0.152 |
Token |
|
Token ID | 198 |
Feature activation | -2.45e-01 |
Pos logit contributions | |
rily | +0.873 |
inations | +0.866 |
ittered | +0.865 |
immer | +0.856 |
avail | +0.856 |
Neg logit contributions | |
a | -0.102 |
and | -0.075 |
big | -0.073 |
two | -0.059 |
little | -0.041 |
Token | The |
Token ID | 464 |
Feature activation | +3.10e-01 |
Pos logit contributions | |
inations | +0.725 |
rily | +0.724 |
leness | +0.722 |
fer | +0.720 |
avail | +0.716 |
Neg logit contributions | |
a | -0.493 |
and | -0.464 |
big | -0.463 |
. | -0.425 |
two | -0.420 |
Token | the |
Token ID | 262 |
Feature activation | -1.05e-01 |
Pos logit contributions | |
rily | +0.305 |
leness | +0.299 |
avail | +0.298 |
ittered | +0.296 |
immer | +0.293 |
Neg logit contributions | |
a | -0.128 |
| -0.125 |
two | -0.124 |
and | -0.124 |
big | -0.123 |
Token | same |
Token ID | 976 |
Feature activation | -8.12e-02 |
Pos logit contributions | |
avail | +0.335 |
footing | +0.320 |
rily | +0.320 |
rated | +0.318 |
immer | +0.317 |
Neg logit contributions | |
| -0.187 |
a | -0.185 |
and | -0.179 |
big | -0.164 |
the | -0.164 |
Token | urge |
Token ID | 14960 |
Feature activation | -4.46e-02 |
Pos logit contributions | |
avail | +0.283 |
rily | +0.279 |
footing | +0.276 |
leness | +0.272 |
immer | +0.272 |
Neg logit contributions | |
a | -0.119 |
| -0.118 |
big | -0.108 |
two | -0.103 |
and | -0.098 |
Token | to |
Token ID | 284 |
Feature activation | +2.14e-01 |
Pos logit contributions | |
avail | +0.182 |
rily | +0.180 |
ittered | +0.179 |
leness | +0.177 |
footing | +0.177 |
Neg logit contributions | |
a | -0.035 |
big | -0.031 |
| -0.030 |
two | -0.029 |
and | -0.024 |
Token | explore |
Token ID | 7301 |
Feature activation | -1.78e-01 |
Pos logit contributions | |
a | +0.359 |
| +0.339 |
and | +0.338 |
big | +0.337 |
two | +0.325 |
Neg logit contributions | |
avail | -0.726 |
inations | -0.665 |
fer | -0.663 |
keeping | -0.661 |
ittered | -0.659 |
Token | Ur |
Token ID | 8799 |
Feature activation | -2.99e+00 |
Pos logit contributions | |
avail | +0.758 |
rily | +0.735 |
lights | +0.731 |
leness | +0.724 |
ittered | +0.721 |
Neg logit contributions | |
a | -0.115 |
two | -0.103 |
big | -0.102 |
| -0.101 |
and | -0.077 |
Token | ge |
Token ID | 469 |
Feature activation | -3.75e-01 |
Pos logit contributions | |
avail | +2.274 |
rily | +2.056 |
ittered | +1.933 |
leness | +1.777 |
lights | +1.774 |
Neg logit contributions | |
a | -13.336 |
and | -12.924 |
big | -12.852 |
| -12.837 |
two | -12.609 |
Token | Land |
Token ID | 6379 |
Feature activation | -2.38e-01 |
Pos logit contributions | |
rily | +1.627 |
avail | +1.612 |
ittered | +1.593 |
keeping | +1.590 |
lights | +1.586 |
Neg logit contributions | |
a | -0.229 |
| -0.201 |
big | -0.201 |
two | -0.164 |
and | -0.135 |
Token | everyday |
Token ID | 10908 |
Feature activation | -1.45e-01 |
Pos logit contributions | |
rily | +1.033 |
ittered | +1.020 |
lights | +1.009 |
avail | +1.006 |
immer | +1.003 |
Neg logit contributions | |
a | -0.153 |
big | -0.130 |
| -0.129 |
two | -0.113 |
and | -0.097 |
Token | , |
Token ID | 11 |
Feature activation | +1.08e-03 |
Pos logit contributions | |
avail | +0.626 |
rily | +0.622 |
keeping | +0.612 |
ittered | +0.609 |
leness | +0.608 |
Neg logit contributions | |
a | -0.087 |
big | -0.075 |
| -0.071 |
two | -0.069 |
and | -0.054 |
Token | and |
Token ID | 290 |
Feature activation | +2.88e-01 |
Pos logit contributions | |
a | +0.001 |
big | +0.001 |
| +0.001 |
two | +0.001 |
little | +0.001 |
Neg logit contributions | |
avail | -0.004 |
inations | -0.004 |
rily | -0.004 |
leness | -0.004 |
keeping | -0.004 |
Token | She |
Token ID | 1375 |
Feature activation | +2.43e-01 |
Pos logit contributions | |
a | +0.584 |
big | +0.546 |
and | +0.510 |
two | +0.509 |
the | +0.491 |
Neg logit contributions | |
rily | -0.996 |
avail | -0.994 |
ently | -0.989 |
lights | -0.984 |
keeping | -0.972 |
Token | cried |
Token ID | 16896 |
Feature activation | +2.62e-01 |
Pos logit contributions | |
a | +0.421 |
big | +0.406 |
| +0.405 |
two | +0.368 |
the | +0.360 |
Neg logit contributions | |
avail | -0.749 |
ittered | -0.743 |
rily | -0.743 |
rated | -0.734 |
aned | -0.732 |
Token | and |
Token ID | 290 |
Feature activation | -9.76e-02 |
Pos logit contributions | |
big | +0.120 |
| +0.118 |
a | +0.114 |
two | +0.093 |
the | +0.060 |
Neg logit contributions | |
rily | -1.125 |
ittered | -1.104 |
lights | -1.101 |
immer | -1.090 |
avail | -1.088 |
Token | threw |
Token ID | 9617 |
Feature activation | +4.54e-01 |
Pos logit contributions | |
avail | +0.319 |
leness | +0.304 |
rily | +0.304 |
ittered | +0.303 |
keeping | +0.300 |
Neg logit contributions | |
a | -0.162 |
big | -0.161 |
| -0.160 |
and | -0.154 |
two | -0.147 |
Token | a |
Token ID | 257 |
Feature activation | -1.87e-01 |
Pos logit contributions | |
| +0.519 |
a | +0.480 |
big | +0.479 |
two | +0.457 |
and | +0.446 |
Neg logit contributions | |
avail | -1.678 |
rily | -1.646 |
keeping | -1.632 |
ittered | -1.605 |
lights | -1.600 |
Token | tant |
Token ID | 24246 |
Feature activation | -2.98e+00 |
Pos logit contributions | |
rily | +0.421 |
leness | +0.417 |
avail | +0.415 |
immer | +0.406 |
inations | +0.406 |
Neg logit contributions | |
a | -0.494 |
| -0.470 |
two | -0.457 |
and | -0.453 |
the | -0.442 |
Token | rum |
Token ID | 6582 |
Feature activation | -1.98e-01 |
Pos logit contributions | |
avail | +7.147 |
rily | +6.935 |
ittered | +6.810 |
leness | +6.682 |
lights | +6.635 |
Neg logit contributions | |
a | -8.346 |
and | -7.925 |
big | -7.879 |
| -7.839 |
two | -7.609 |
Token | . |
Token ID | 13 |
Feature activation | +1.50e-01 |
Pos logit contributions | |
rily | +0.860 |
ittered | +0.850 |
lights | +0.849 |
keeping | +0.845 |
ently | +0.839 |
Neg logit contributions | |
big | -0.110 |
a | -0.104 |
| -0.090 |
two | -0.075 |
little | -0.062 |
Token | Her |
Token ID | 2332 |
Feature activation | -1.16e-02 |
Pos logit contributions | |
a | +0.317 |
big | +0.305 |
and | +0.296 |
two | +0.288 |
the | +0.274 |
Neg logit contributions | |
rily | -0.405 |
avail | -0.403 |
ently | -0.398 |
lights | -0.396 |
ittered | -0.396 |
Token | mom |
Token ID | 1995 |
Feature activation | +6.83e-02 |
Pos logit contributions | |
avail | +0.043 |
lights | +0.043 |
footing | +0.042 |
keeping | +0.041 |
ittered | +0.041 |
Neg logit contributions | |
a | -0.015 |
| -0.014 |
and | -0.013 |
the | -0.011 |
two | -0.011 |
Token | tried |
Token ID | 3088 |
Feature activation | -9.74e-02 |
Pos logit contributions | |
a | +0.107 |
big | +0.100 |
| +0.096 |
two | +0.089 |
the | +0.084 |
Neg logit contributions | |
aned | -0.223 |
ittered | -0.223 |
inations | -0.221 |
leness | -0.219 |
rily | -0.219 |
Token | everything |
Token ID | 2279 |
Feature activation | -2.22e-01 |
Pos logit contributions | |
| +0.473 |
big | +0.426 |
a | +0.425 |
and | +0.385 |
two | +0.384 |
Neg logit contributions | |
rily | -1.740 |
avail | -1.695 |
ittered | -1.679 |
lights | -1.662 |
immer | -1.642 |
Token | around |
Token ID | 1088 |
Feature activation | +1.94e-01 |
Pos logit contributions | |
rily | +0.747 |
lights | +0.735 |
ittered | +0.732 |
avail | +0.721 |
immer | +0.711 |
Neg logit contributions | |
a | -0.353 |
| -0.344 |
two | -0.321 |
big | -0.315 |
and | -0.305 |
Token | and |
Token ID | 290 |
Feature activation | -2.20e-01 |
Pos logit contributions | |
a | +0.156 |
| +0.154 |
big | +0.153 |
two | +0.141 |
little | +0.116 |
Neg logit contributions | |
avail | -0.790 |
rily | -0.778 |
keeping | -0.761 |
ittered | -0.760 |
lights | -0.746 |
Token | tried |
Token ID | 3088 |
Feature activation | -1.12e-01 |
Pos logit contributions | |
avail | +0.724 |
rily | +0.709 |
ittered | +0.704 |
lights | +0.692 |
leness | +0.691 |
Neg logit contributions | |
a | -0.365 |
| -0.355 |
big | -0.345 |
and | -0.329 |
two | -0.326 |
Token | to |
Token ID | 284 |
Feature activation | +3.95e-02 |
Pos logit contributions | |
avail | +0.466 |
rily | +0.460 |
ittered | +0.454 |
lights | +0.448 |
inations | +0.448 |
Neg logit contributions | |
a | -0.073 |
| -0.069 |
big | -0.066 |
two | -0.060 |
and | -0.055 |
Token | rearr |
Token ID | 37825 |
Feature activation | -2.97e+00 |
Pos logit contributions | |
| +0.076 |
a | +0.075 |
big | +0.071 |
and | +0.071 |
two | +0.068 |
Neg logit contributions | |
avail | -0.121 |
rily | -0.116 |
inations | -0.115 |
immer | -0.115 |
ittered | -0.113 |
Token | ange |
Token ID | 858 |
Feature activation | -9.70e-04 |
Pos logit contributions | |
avail | +5.071 |
rily | +4.870 |
ittered | +4.754 |
leness | +4.663 |
lights | +4.585 |
Neg logit contributions | |
a | -10.446 |
and | -10.019 |
big | -9.950 |
| -9.943 |
two | -9.694 |
Token | it |
Token ID | 340 |
Feature activation | +3.02e-01 |
Pos logit contributions | |
avail | +0.004 |
rily | +0.004 |
lights | +0.004 |
ittered | +0.004 |
ently | +0.004 |
Neg logit contributions | |
| -0.001 |
a | -0.001 |
big | -0.001 |
and | -0.001 |
two | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | -1.05e-01 |
Pos logit contributions | |
a | +0.357 |
| +0.352 |
big | +0.330 |
two | +0.319 |
and | +0.310 |
Neg logit contributions | |
rily | -1.115 |
ittered | -1.114 |
immer | -1.098 |
avail | -1.087 |
lights | -1.070 |
Token | She |
Token ID | 1375 |
Feature activation | +1.16e-01 |
Pos logit contributions | |
avail | +0.337 |
rily | +0.330 |
keeping | +0.324 |
ently | +0.318 |
ittered | +0.317 |
Neg logit contributions | |
big | -0.179 |
a | -0.177 |
and | -0.164 |
two | -0.160 |
little | -0.151 |
Token | wanted |
Token ID | 2227 |
Feature activation | +1.50e-01 |
Pos logit contributions | |
a | +0.225 |
| +0.219 |
big | +0.213 |
and | +0.201 |
two | +0.200 |
Neg logit contributions | |
ittered | -0.346 |
rily | -0.342 |
inations | -0.341 |
lights | -0.333 |
avail | -0.328 |
Token | do |
Token ID | 466 |
Feature activation | -7.14e-02 |
Pos logit contributions | |
a | +0.404 |
| +0.394 |
and | +0.389 |
big | +0.383 |
two | +0.371 |
Neg logit contributions | |
avail | -0.488 |
inations | -0.451 |
rily | -0.448 |
immer | -0.437 |
epad | -0.436 |
Token | it |
Token ID | 340 |
Feature activation | +3.88e-02 |
Pos logit contributions | |
avail | +0.265 |
inations | +0.261 |
rily | +0.260 |
leness | +0.257 |
ittered | +0.254 |
Neg logit contributions | |
| -0.081 |
a | -0.080 |
and | -0.078 |
big | -0.078 |
two | -0.072 |
Token | !" |
Token ID | 2474 |
Feature activation | -5.71e-02 |
Pos logit contributions | |
a | +0.079 |
| +0.077 |
big | +0.076 |
two | +0.074 |
and | +0.073 |
Neg logit contributions | |
rily | -0.110 |
avail | -0.110 |
ittered | -0.110 |
inations | -0.108 |
immer | -0.106 |
Token |
|
Token ID | 198 |
Feature activation | -1.90e-01 |
Pos logit contributions | |
rily | +0.204 |
avail | +0.199 |
leness | +0.197 |
keeping | +0.197 |
lights | +0.195 |
Neg logit contributions | |
a | -0.078 |
big | -0.077 |
two | -0.068 |
and | -0.063 |
little | -0.063 |
Token |
|
Token ID | 198 |
Feature activation | -7.55e-01 |
Pos logit contributions | |
rily | +0.842 |
avail | +0.836 |
leness | +0.823 |
ittered | +0.821 |
vation | +0.820 |
Neg logit contributions | |
a | -0.091 |
big | -0.061 |
two | -0.060 |
and | -0.047 |
one | -0.032 |
Token | Charl |
Token ID | 24453 |
Feature activation | -2.95e+00 |
Pos logit contributions | |
avail | +2.039 |
rily | +2.028 |
leness | +1.986 |
vation | +1.960 |
fer | +1.945 |
Neg logit contributions | |
a | -1.751 |
big | -1.635 |
and | -1.584 |
two | -1.568 |
| -1.566 |
Token | otte |
Token ID | 11404 |
Feature activation | -5.74e-03 |
Pos logit contributions | |
avail | +3.201 |
rily | +2.999 |
ittered | +2.903 |
leness | +2.779 |
lights | +2.750 |
Neg logit contributions | |
a | -12.132 |
and | -11.708 |
big | -11.650 |
| -11.650 |
two | -11.417 |
Token | and |
Token ID | 290 |
Feature activation | -1.67e-01 |
Pos logit contributions | |
ittered | +0.017 |
rily | +0.017 |
fer | +0.017 |
avail | +0.017 |
iously | +0.017 |
Neg logit contributions | |
a | -0.011 |
big | -0.011 |
| -0.010 |
two | -0.009 |
and | -0.009 |
Token | Tom |
Token ID | 4186 |
Feature activation | -3.06e-02 |
Pos logit contributions | |
avail | +0.547 |
rily | +0.507 |
keeping | +0.495 |
leness | +0.493 |
arians | +0.487 |
Neg logit contributions | |
big | -0.306 |
a | -0.305 |
| -0.278 |
and | -0.276 |
two | -0.271 |
Token | wrote |
Token ID | 2630 |
Feature activation | -2.74e-01 |
Pos logit contributions | |
ittered | +0.097 |
rily | +0.095 |
avail | +0.093 |
immer | +0.093 |
inations | +0.093 |
Neg logit contributions | |
a | -0.053 |
big | -0.051 |
| -0.051 |
and | -0.046 |
little | -0.045 |
Token | a |
Token ID | 257 |
Feature activation | -5.95e-01 |
Pos logit contributions | |
rily | +1.135 |
ittered | +1.109 |
avail | +1.099 |
lights | +1.094 |
immer | +1.074 |
Neg logit contributions | |
| -0.202 |
a | -0.202 |
big | -0.188 |
and | -0.169 |
two | -0.164 |
Token | things |
Token ID | 1243 |
Feature activation | -1.73e-01 |
Pos logit contributions | |
avail | +0.767 |
rily | +0.746 |
lights | +0.742 |
inations | +0.737 |
leness | +0.736 |
Neg logit contributions | |
a | -0.425 |
| -0.409 |
big | -0.378 |
two | -0.375 |
and | -0.365 |
Token | â |
Token ID | 22940 |
Feature activation | -1.22e+00 |
Pos logit contributions | |
rily | +0.722 |
ittered | +0.714 |
avail | +0.712 |
keeping | +0.711 |
immer | +0.703 |
Neg logit contributions | |
a | -0.124 |
big | -0.113 |
two | -0.108 |
| -0.104 |
little | -0.093 |
Token | € |
Token ID | 26391 |
Feature activation | -6.92e-01 |
Pos logit contributions | |
avail | +2.188 |
ittered | +2.120 |
immer | +2.087 |
rated | +2.075 |
leness | +2.035 |
Neg logit contributions | |
a | -4.048 |
big | -3.965 |
and | -3.850 |
| -3.841 |
two | -3.781 |
Token | � |
Token ID | 447 |
Feature activation | -1.01e+00 |
Pos logit contributions | |
ittered | +1.780 |
inations | +1.774 |
avail | +1.752 |
keeping | +1.748 |
rily | +1.717 |
Neg logit contributions | |
a | -1.659 |
big | -1.573 |
two | -1.527 |
and | -1.521 |
the | -1.470 |
Token | � |
Token ID | 251 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
avail | +1.309 |
ittered | +1.274 |
rily | +1.259 |
leness | +1.204 |
immer | +1.193 |
Neg logit contributions | |
a | -3.794 |
big | -3.658 |
and | -3.612 |
two | -3.531 |
| -3.516 |
Token | col |
Token ID | 4033 |
Feature activation | -2.94e+00 |
Pos logit contributions | |
a | +0.113 |
big | +0.112 |
and | +0.092 |
two | +0.091 |
| +0.086 |
Neg logit contributions | |
rily | -0.442 |
leness | -0.432 |
avail | -0.429 |
lights | -0.426 |
ittered | -0.426 |
Token | ouring |
Token ID | 21823 |
Feature activation | -1.35e-01 |
Pos logit contributions | |
avail | +3.290 |
rily | +3.058 |
ittered | +2.943 |
leness | +2.891 |
lights | +2.813 |
Neg logit contributions | |
a | -11.987 |
and | -11.556 |
big | -11.510 |
| -11.482 |
two | -11.265 |
Token | books |
Token ID | 3835 |
Feature activation | -4.82e-01 |
Pos logit contributions | |
rily | +0.570 |
ittered | +0.556 |
avail | +0.552 |
lights | +0.548 |
leness | +0.546 |
Neg logit contributions | |
a | -0.095 |
| -0.090 |
big | -0.088 |
two | -0.074 |
and | -0.074 |
Token | , |
Token ID | 11 |
Feature activation | -3.25e-01 |
Pos logit contributions | |
ittered | +2.075 |
rily | +2.073 |
immer | +2.046 |
avail | +2.025 |
lights | +2.024 |
Neg logit contributions | |
a | -0.319 |
big | -0.278 |
| -0.261 |
two | -0.244 |
little | -0.192 |
Token | stickers |
Token ID | 28568 |
Feature activation | -1.37e-01 |
Pos logit contributions | |
rily | +1.158 |
avail | +1.155 |
inations | +1.149 |
ittered | +1.143 |
lights | +1.132 |
Neg logit contributions | |
a | -0.405 |
big | -0.394 |
| -0.383 |
two | -0.357 |
and | -0.339 |
Token | and |
Token ID | 290 |
Feature activation | -2.37e-01 |
Pos logit contributions | |
rily | +0.606 |
ittered | +0.606 |
immer | +0.593 |
lights | +0.590 |
avail | +0.586 |
Neg logit contributions | |
a | -0.075 |
| -0.062 |
big | -0.062 |
two | -0.046 |
little | -0.034 |
Token | started |
Token ID | 2067 |
Feature activation | -2.12e-01 |
Pos logit contributions | |
ittered | +0.194 |
inations | +0.189 |
avail | +0.189 |
fer | +0.187 |
rily | +0.187 |
Neg logit contributions | |
a | -0.103 |
| -0.099 |
big | -0.097 |
two | -0.092 |
and | -0.091 |
Token | walking |
Token ID | 6155 |
Feature activation | +2.29e-01 |
Pos logit contributions | |
avail | +0.814 |
rily | +0.791 |
ge | +0.790 |
ittered | +0.790 |
keeping | +0.789 |
Neg logit contributions | |
a | -0.208 |
| -0.202 |
big | -0.187 |
two | -0.180 |
and | -0.168 |
Token | very |
Token ID | 845 |
Feature activation | -3.30e-01 |
Pos logit contributions | |
a | +0.253 |
| +0.238 |
two | +0.229 |
big | +0.227 |
little | +0.206 |
Neg logit contributions | |
rily | -0.847 |
avail | -0.845 |
lights | -0.838 |
ittered | -0.824 |
keeping | -0.820 |
Token | slowly |
Token ID | 6364 |
Feature activation | +3.71e-02 |
Pos logit contributions | |
avail | +1.065 |
rily | +1.042 |
ge | +1.030 |
rated | +1.026 |
footing | +1.024 |
Neg logit contributions | |
a | -0.555 |
| -0.518 |
two | -0.508 |
and | -0.493 |
big | -0.487 |
Token | as |
Token ID | 355 |
Feature activation | -8.07e-01 |
Pos logit contributions | |
a | +0.029 |
| +0.025 |
big | +0.025 |
two | +0.023 |
little | +0.019 |
Neg logit contributions | |
avail | -0.152 |
rily | -0.149 |
leness | -0.149 |
ittered | -0.149 |
lights | -0.148 |
Token | tort |
Token ID | 7619 |
Feature activation | -2.93e+00 |
Pos logit contributions | |
avail | +2.945 |
rily | +2.867 |
ittered | +2.815 |
lights | +2.812 |
fer | +2.785 |
Neg logit contributions | |
a | -1.031 |
| -0.967 |
and | -0.949 |
big | -0.943 |
two | -0.899 |
Token | o |
Token ID | 78 |
Feature activation | -1.28e-01 |
Pos logit contributions | |
avail | +3.979 |
rily | +3.805 |
ittered | +3.655 |
leness | +3.544 |
lights | +3.523 |
Neg logit contributions | |
a | -11.274 |
and | -10.836 |
big | -10.784 |
| -10.776 |
two | -10.578 |
Token | ises |
Token ID | 2696 |
Feature activation | +1.35e-01 |
Pos logit contributions | |
rily | +0.500 |
ittered | +0.497 |
avail | +0.493 |
fer | +0.490 |
immer | +0.488 |
Neg logit contributions | |
a | -0.132 |
big | -0.128 |
| -0.114 |
two | -0.106 |
and | -0.101 |
Token | do |
Token ID | 466 |
Feature activation | -1.89e-02 |
Pos logit contributions | |
a | +0.095 |
big | +0.090 |
| +0.078 |
two | +0.072 |
little | +0.066 |
Neg logit contributions | |
rily | -0.557 |
ittered | -0.557 |
leness | -0.551 |
avail | -0.549 |
lights | -0.545 |
Token | , |
Token ID | 11 |
Feature activation | -9.11e-02 |
Pos logit contributions | |
avail | +0.073 |
leness | +0.073 |
inations | +0.072 |
ittered | +0.072 |
rily | +0.072 |
Neg logit contributions | |
a | -0.019 |
| -0.018 |
big | -0.017 |
two | -0.017 |
and | -0.016 |
Token | but |
Token ID | 475 |
Feature activation | +2.12e-01 |
Pos logit contributions | |
avail | +0.301 |
rily | +0.300 |
leness | +0.295 |
inations | +0.294 |
ittered | +0.293 |
Neg logit contributions | |
a | -0.136 |
big | -0.133 |
| -0.132 |
two | -0.121 |
and | -0.114 |
Token | said |
Token ID | 531 |
Feature activation | -7.67e-02 |
Pos logit contributions | |
a | +0.328 |
big | +0.306 |
| +0.304 |
two | +0.287 |
and | +0.281 |
Neg logit contributions | |
ittered | -0.664 |
rily | -0.655 |
inations | -0.652 |
avail | -0.645 |
lights | -0.643 |
Token | : |
Token ID | 25 |
Feature activation | +2.18e-01 |
Pos logit contributions | |
rily | +0.272 |
lights | +0.268 |
avail | +0.267 |
leness | +0.267 |
ently | +0.266 |
Neg logit contributions | |
a | -0.096 |
big | -0.091 |
| -0.086 |
two | -0.081 |
and | -0.080 |
Token | " |
Token ID | 366 |
Feature activation | +9.68e-03 |
Pos logit contributions | |
a | +0.275 |
big | +0.260 |
two | +0.233 |
and | +0.226 |
little | +0.213 |
Neg logit contributions | |
rily | -0.775 |
leness | -0.769 |
fer | -0.768 |
ittered | -0.765 |
ently | -0.762 |
Token | I |
Token ID | 40 |
Feature activation | +5.89e-01 |
Pos logit contributions | |
a | +0.021 |
big | +0.020 |
and | +0.019 |
, | +0.018 |
. | +0.018 |
Neg logit contributions | |
avail | -0.028 |
leness | -0.028 |
fer | -0.027 |
keeping | -0.027 |
igi | -0.027 |
Token | 'm |
Token ID | 1101 |
Feature activation | +1.78e-01 |
Pos logit contributions | |
a | +1.361 |
big | +1.333 |
two | +1.266 |
and | +1.230 |
little | +1.226 |
Neg logit contributions | |
inations | -1.492 |
leness | -1.451 |
avail | -1.442 |
ittered | -1.429 |
keeping | -1.419 |
Token | sorry |
Token ID | 7926 |
Feature activation | +6.68e-01 |
Pos logit contributions | |
and | +0.484 |
a | +0.476 |
| +0.471 |
big | +0.469 |
two | +0.451 |
Neg logit contributions | |
avail | -0.362 |
inations | -0.359 |
rily | -0.352 |
leness | -0.347 |
ittered | -0.344 |
Token | , |
Token ID | 11 |
Feature activation | +4.49e-01 |
Pos logit contributions | |
a | +0.611 |
big | +0.548 |
two | +0.508 |
| +0.502 |
and | +0.398 |
Neg logit contributions | |
avail | -2.652 |
lights | -2.630 |
rily | -2.564 |
ently | -2.524 |
iders | -2.495 |
Token | Mia |
Token ID | 32189 |
Feature activation | +6.91e-01 |
Pos logit contributions | |
a | +0.771 |
| +0.757 |
big | +0.706 |
and | +0.652 |
. | +0.637 |
Neg logit contributions | |
rily | -1.447 |
avail | -1.445 |
lights | -1.439 |
inations | -1.367 |
iously | -1.330 |
Token | and |
Token ID | 290 |
Feature activation | +5.93e-01 |
Pos logit contributions | |
a | +0.447 |
big | +0.387 |
two | +0.286 |
| +0.282 |
the | +0.169 |
Neg logit contributions | |
rily | -2.910 |
ently | -2.872 |
leness | -2.865 |
rated | -2.863 |
keeping | -2.857 |
Token | Ben |
Token ID | 3932 |
Feature activation | +5.90e-01 |
Pos logit contributions | |
big | +1.028 |
a | +1.028 |
| +1.003 |
and | +0.981 |
two | +0.917 |
Neg logit contributions | |
rily | -1.805 |
avail | -1.796 |
lights | -1.761 |
leness | -1.755 |
keeping | -1.752 |
Token | . |
Token ID | 13 |
Feature activation | -5.59e-01 |
Pos logit contributions | |
a | +0.413 |
big | +0.376 |
two | +0.290 |
| +0.279 |
little | +0.216 |
Neg logit contributions | |
leness | -2.436 |
avail | -2.385 |
rated | -2.385 |
keeping | -2.384 |
rily | -2.371 |
Token | come |
Token ID | 1282 |
Feature activation | -2.37e-01 |
Pos logit contributions | |
a | +0.500 |
big | +0.468 |
two | +0.444 |
| +0.439 |
and | +0.423 |
Neg logit contributions | |
avail | -1.225 |
ittered | -1.166 |
inations | -1.165 |
immer | -1.159 |
footing | -1.156 |
Token | back |
Token ID | 736 |
Feature activation | +2.57e-01 |
Pos logit contributions | |
avail | +0.882 |
ittered | +0.873 |
rily | +0.872 |
immer | +0.862 |
lights | +0.852 |
Neg logit contributions | |
a | -0.259 |
| -0.246 |
big | -0.245 |
two | -0.235 |
little | -0.220 |
Token | and |
Token ID | 290 |
Feature activation | +3.79e-01 |
Pos logit contributions | |
big | +0.393 |
a | +0.392 |
two | +0.377 |
| +0.367 |
little | +0.349 |
Neg logit contributions | |
avail | -0.847 |
rily | -0.814 |
ittered | -0.804 |
keeping | -0.804 |
ge | -0.801 |
Token | play |
Token ID | 711 |
Feature activation | +2.82e-01 |
Pos logit contributions | |
a | +0.537 |
| +0.501 |
two | +0.500 |
and | +0.493 |
big | +0.493 |
Neg logit contributions | |
avail | -1.326 |
inations | -1.276 |
keeping | -1.246 |
footing | -1.245 |
lights | -1.235 |
Token | again |
Token ID | 757 |
Feature activation | +5.37e-01 |
Pos logit contributions | |
big | +0.321 |
a | +0.313 |
two | +0.303 |
| +0.303 |
little | +0.276 |
Neg logit contributions | |
avail | -1.042 |
rily | -1.040 |
ittered | -1.031 |
immer | -1.010 |
fer | -1.003 |
Token | tomorrow |
Token ID | 9439 |
Feature activation | +5.79e-01 |
Pos logit contributions | |
a | +0.663 |
big | +0.649 |
two | +0.618 |
| +0.577 |
and | +0.555 |
Neg logit contributions | |
avail | -2.049 |
rily | -1.900 |
ge | -1.872 |
fer | -1.862 |
leness | -1.842 |
Token | . |
Token ID | 13 |
Feature activation | +2.45e-01 |
Pos logit contributions | |
a | +0.382 |
big | +0.342 |
two | +0.335 |
| +0.274 |
little | +0.243 |
Neg logit contributions | |
avail | -2.507 |
rily | -2.442 |
keeping | -2.420 |
lights | -2.342 |
fer | -2.324 |
Token | She |
Token ID | 1375 |
Feature activation | -1.63e-01 |
Pos logit contributions | |
a | +0.309 |
big | +0.298 |
two | +0.275 |
and | +0.264 |
little | +0.236 |
Neg logit contributions | |
avail | -0.865 |
rily | -0.857 |
lights | -0.854 |
keeping | -0.842 |
leness | -0.837 |
Token | took |
Token ID | 1718 |
Feature activation | +1.88e-01 |
Pos logit contributions | |
rily | +0.540 |
ittered | +0.535 |
inations | +0.528 |
avail | +0.527 |
rated | +0.519 |
Neg logit contributions | |
a | -0.250 |
big | -0.231 |
| -0.229 |
two | -0.217 |
and | -0.206 |
Token | off |
Token ID | 572 |
Feature activation | +3.42e-01 |
Pos logit contributions | |
| +0.250 |
and | +0.249 |
big | +0.237 |
a | +0.233 |
. | +0.217 |
Neg logit contributions | |
avail | -0.672 |
rily | -0.647 |
ittered | -0.630 |
lights | -0.618 |
keeping | -0.617 |
Token | her |
Token ID | 607 |
Feature activation | +4.79e-01 |
Pos logit contributions | |
| +0.335 |
big | +0.316 |
a | +0.312 |
two | +0.296 |
and | +0.281 |
Neg logit contributions | |
avail | -1.255 |
lights | -1.250 |
ittered | -1.242 |
keeping | -1.240 |
rily | -1.238 |
Token | . |
Token ID | 13 |
Feature activation | +2.07e-03 |
Pos logit contributions | |
a | +0.274 |
big | +0.265 |
two | +0.234 |
| +0.204 |
little | +0.188 |
Neg logit contributions | |
lights | -1.324 |
rily | -1.291 |
avail | -1.281 |
keeping | -1.272 |
inations | -1.262 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +5.88e-01 |
Pos logit contributions | |
a | +0.002 |
big | +0.002 |
and | +0.002 |
two | +0.002 |
the | +0.002 |
Neg logit contributions | |
lights | -0.008 |
avail | -0.008 |
rily | -0.008 |
leness | -0.008 |
ittered | -0.007 |
Token | B |
Token ID | 33 |
Feature activation | -1.06e+00 |
Pos logit contributions | |
a | +1.049 |
big | +1.017 |
and | +0.972 |
little | +0.969 |
two | +0.920 |
Neg logit contributions | |
keeping | -1.749 |
avail | -1.746 |
rily | -1.692 |
ittered | -1.683 |
lights | -1.675 |
Token | unny |
Token ID | 16948 |
Feature activation | +6.66e-01 |
Pos logit contributions | |
ittered | +2.398 |
rily | +2.306 |
immer | +2.306 |
avail | +2.290 |
inations | +2.283 |
Neg logit contributions | |
a | -2.974 |
big | -2.799 |
and | -2.712 |
two | -2.700 |
| -2.644 |
Token | was |
Token ID | 373 |
Feature activation | +2.06e-01 |
Pos logit contributions | |
a | +0.391 |
big | +0.362 |
two | +0.315 |
| +0.225 |
little | +0.189 |
Neg logit contributions | |
ittered | -2.799 |
rily | -2.772 |
keeping | -2.753 |
leness | -2.718 |
fer | -2.710 |
Token | a |
Token ID | 257 |
Feature activation | +7.41e-01 |
Pos logit contributions | |
| +0.243 |
a | +0.240 |
and | +0.236 |
big | +0.229 |
two | +0.222 |
Neg logit contributions | |
ittered | -0.728 |
rily | -0.721 |
avail | -0.712 |
inations | -0.704 |
immer | -0.703 |
Token | little |
Token ID | 1310 |
Feature activation | +8.05e-02 |
Pos logit contributions | |
a | +1.127 |
| +1.042 |
and | +1.029 |
two | +1.019 |
the | +0.901 |
Neg logit contributions | |
avail | -2.461 |
rily | -2.388 |
fer | -2.381 |
leness | -2.272 |
inations | -2.268 |
Token | rabbit |
Token ID | 22746 |
Feature activation | +4.30e-01 |
Pos logit contributions | |
a | +0.137 |
| +0.131 |
two | +0.121 |
big | +0.120 |
and | +0.119 |
Neg logit contributions | |
avail | -0.261 |
rily | -0.255 |
fer | -0.249 |
lights | -0.247 |
leness | -0.245 |
Token | who |
Token ID | 508 |
Feature activation | +3.59e-01 |
Pos logit contributions | |
big | +0.345 |
a | +0.340 |
two | +0.307 |
| +0.256 |
little | +0.236 |
Neg logit contributions | |
rily | -1.719 |
avail | -1.674 |
ittered | -1.672 |
leness | -1.661 |
fer | -1.661 |
Token | loved |
Token ID | 6151 |
Feature activation | -8.96e-02 |
Pos logit contributions | |
a | +0.666 |
two | +0.650 |
big | +0.634 |
and | +0.611 |
| +0.604 |
Neg logit contributions | |
ittered | -1.088 |
leness | -1.058 |
rily | -1.012 |
inations | -1.012 |
avail | -1.010 |
Token | to |
Token ID | 284 |
Feature activation | -4.48e-01 |
Pos logit contributions | |
avail | +0.367 |
keeping | +0.363 |
rily | +0.358 |
leness | +0.350 |
lights | +0.349 |
Neg logit contributions | |
a | -0.063 |
| -0.060 |
two | -0.057 |
and | -0.057 |
big | -0.054 |
Token | girl |
Token ID | 2576 |
Feature activation | +2.12e-01 |
Pos logit contributions | |
a | +0.387 |
| +0.377 |
two | +0.340 |
and | +0.337 |
big | +0.328 |
Neg logit contributions | |
avail | -0.673 |
rily | -0.651 |
fer | -0.639 |
leness | -0.633 |
lights | -0.631 |
Token | named |
Token ID | 3706 |
Feature activation | +2.18e-01 |
Pos logit contributions | |
a | +0.479 |
two | +0.466 |
big | +0.461 |
| +0.442 |
and | +0.424 |
Neg logit contributions | |
avail | -0.518 |
ittered | -0.516 |
immer | -0.509 |
lights | -0.508 |
rily | -0.508 |
Token | Daisy |
Token ID | 40355 |
Feature activation | +5.19e-01 |
Pos logit contributions | |
and | +0.286 |
a | +0.286 |
big | +0.271 |
two | +0.255 |
| +0.246 |
Neg logit contributions | |
avail | -0.760 |
ittered | -0.751 |
rily | -0.742 |
keeping | -0.738 |
immer | -0.731 |
Token | . |
Token ID | 13 |
Feature activation | -8.07e-01 |
Pos logit contributions | |
two | +0.358 |
a | +0.354 |
big | +0.321 |
| +0.234 |
and | +0.209 |
Neg logit contributions | |
rily | -2.088 |
ittered | -2.083 |
immer | -2.061 |
rated | -2.057 |
keeping | -2.049 |
Token | She |
Token ID | 1375 |
Feature activation | +6.74e-01 |
Pos logit contributions | |
avail | +2.538 |
rily | +2.439 |
ittered | +2.434 |
immer | +2.374 |
lights | +2.363 |
Neg logit contributions | |
a | -1.482 |
big | -1.387 |
and | -1.349 |
two | -1.339 |
| -1.280 |
Token | was |
Token ID | 373 |
Feature activation | +4.97e-01 |
Pos logit contributions | |
a | +1.034 |
two | +1.026 |
and | +0.940 |
big | +0.939 |
| +0.913 |
Neg logit contributions | |
ittered | -2.140 |
rated | -2.084 |
inations | -2.072 |
leness | -2.060 |
keeping | -2.056 |
Token | three |
Token ID | 1115 |
Feature activation | -6.92e-02 |
Pos logit contributions | |
and | +0.402 |
| +0.400 |
. | +0.337 |
a | +0.323 |
big | +0.309 |
Neg logit contributions | |
avail | -1.923 |
ittered | -1.890 |
rily | -1.869 |
footing | -1.852 |
inations | -1.829 |
Token | years |
Token ID | 812 |
Feature activation | +7.50e-02 |
Pos logit contributions | |
lights | +0.153 |
rily | +0.151 |
leness | +0.147 |
inations | +0.146 |
ittered | +0.144 |
Neg logit contributions | |
a | -0.189 |
| -0.180 |
two | -0.174 |
big | -0.172 |
the | -0.166 |
Token | old |
Token ID | 1468 |
Feature activation | +4.78e-01 |
Pos logit contributions | |
a | +0.129 |
| +0.117 |
two | +0.113 |
and | +0.109 |
was | +0.102 |
Neg logit contributions | |
lights | -0.244 |
rily | -0.242 |
avail | -0.236 |
inations | -0.231 |
keeping | -0.229 |
Token | and |
Token ID | 290 |
Feature activation | +6.01e-01 |
Pos logit contributions | |
a | +0.227 |
big | +0.202 |
| +0.161 |
two | +0.160 |
the | +0.086 |
Neg logit contributions | |
avail | -2.035 |
rily | -2.007 |
keeping | -2.002 |
lights | -1.984 |
arians | -1.981 |
Token | loved |
Token ID | 6151 |
Feature activation | +4.11e-03 |
Pos logit contributions | |
and | +0.924 |
a | +0.882 |
two | +0.856 |
| +0.833 |
big | +0.821 |
Neg logit contributions | |
ittered | -1.934 |
avail | -1.929 |
leness | -1.899 |
inations | -1.843 |
rily | -1.840 |
Token | there |
Token ID | 612 |
Feature activation | +5.91e-01 |
Pos logit contributions | |
a | +1.069 |
big | +1.008 |
and | +0.975 |
two | +0.921 |
| +0.909 |
Neg logit contributions | |
rily | -3.035 |
ittered | -3.023 |
immer | -2.925 |
lights | -2.842 |
footing | -2.839 |
Token | was |
Token ID | 373 |
Feature activation | +3.55e-01 |
Pos logit contributions | |
a | +0.477 |
big | +0.466 |
two | +0.425 |
and | +0.380 |
little | +0.329 |
Neg logit contributions | |
ittered | -2.296 |
avail | -2.293 |
footing | -2.253 |
rily | -2.246 |
fer | -2.196 |
Token | a |
Token ID | 257 |
Feature activation | +5.13e-01 |
Pos logit contributions | |
and | +0.032 |
| +0.028 |
big | +0.024 |
. | +0.009 |
a | +0.005 |
Neg logit contributions | |
rily | -1.618 |
avail | -1.591 |
ittered | -1.582 |
footing | -1.550 |
keeping | -1.545 |
Token | girl |
Token ID | 2576 |
Feature activation | +3.34e-01 |
Pos logit contributions | |
a | +0.487 |
| +0.415 |
two | +0.383 |
and | +0.383 |
the | +0.349 |
Neg logit contributions | |
avail | -1.981 |
fer | -1.971 |
rily | -1.970 |
inations | -1.888 |
leness | -1.881 |
Token | called |
Token ID | 1444 |
Feature activation | +9.43e-02 |
Pos logit contributions | |
a | +0.754 |
big | +0.732 |
two | +0.731 |
| +0.694 |
and | +0.654 |
Neg logit contributions | |
rily | -0.817 |
avail | -0.811 |
ittered | -0.808 |
fer | -0.804 |
keeping | -0.804 |
Token | Daisy |
Token ID | 40355 |
Feature activation | +5.10e-01 |
Pos logit contributions | |
a | +0.185 |
and | +0.183 |
big | +0.183 |
| +0.173 |
two | +0.171 |
Neg logit contributions | |
avail | -0.271 |
keeping | -0.265 |
ittered | -0.263 |
rily | -0.258 |
immer | -0.253 |
Token | . |
Token ID | 13 |
Feature activation | -8.48e-01 |
Pos logit contributions | |
big | +0.343 |
a | +0.329 |
two | +0.326 |
| +0.225 |
little | +0.209 |
Neg logit contributions | |
rily | -2.072 |
ittered | -2.025 |
immer | -1.990 |
keeping | -1.981 |
igi | -1.975 |
Token | She |
Token ID | 1375 |
Feature activation | +6.54e-01 |
Pos logit contributions | |
avail | +2.706 |
rily | +2.634 |
ittered | +2.627 |
lights | +2.572 |
immer | +2.547 |
Neg logit contributions | |
a | -1.500 |
big | -1.400 |
and | -1.362 |
two | -1.341 |
| -1.276 |
Token | was |
Token ID | 373 |
Feature activation | +4.64e-01 |
Pos logit contributions | |
a | +0.987 |
two | +0.982 |
and | +0.897 |
big | +0.892 |
| +0.842 |
Neg logit contributions | |
ittered | -2.094 |
inations | -2.014 |
fer | -2.010 |
igi | -1.993 |
leness | -1.992 |
Token | three |
Token ID | 1115 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
and | +0.345 |
| +0.344 |
. | +0.293 |
a | +0.271 |
big | +0.248 |
Neg logit contributions | |
avail | -1.829 |
rily | -1.821 |
ittered | -1.799 |
footing | -1.753 |
inations | -1.737 |
Token | years |
Token ID | 812 |
Feature activation | +6.32e-02 |
Pos logit contributions | |
a | +0.293 |
| +0.279 |
two | +0.267 |
big | +0.263 |
the | +0.256 |
Neg logit contributions | |
lights | -0.236 |
rily | -0.234 |
leness | -0.228 |
inations | -0.226 |
ittered | -0.220 |
Token | the |
Token ID | 262 |
Feature activation | +1.63e-01 |
Pos logit contributions | |
a | +0.218 |
big | +0.214 |
| +0.197 |
and | +0.182 |
two | +0.166 |
Neg logit contributions | |
avail | -1.014 |
rily | -0.981 |
lights | -0.953 |
leness | -0.945 |
ittered | -0.945 |
Token | little |
Token ID | 1310 |
Feature activation | +1.88e-01 |
Pos logit contributions | |
| +0.221 |
a | +0.218 |
and | +0.211 |
big | +0.185 |
the | +0.184 |
Neg logit contributions | |
avail | -0.579 |
rily | -0.558 |
inations | -0.556 |
lights | -0.555 |
footing | -0.551 |
Token | girl |
Token ID | 2576 |
Feature activation | +2.30e-01 |
Pos logit contributions | |
| +0.324 |
a | +0.315 |
big | +0.292 |
and | +0.291 |
the | +0.263 |
Neg logit contributions | |
avail | -0.628 |
rily | -0.595 |
fer | -0.582 |
lights | -0.578 |
keeping | -0.569 |
Token | cheer |
Token ID | 14042 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
a | +0.395 |
big | +0.381 |
| +0.380 |
two | +0.341 |
the | +0.328 |
Neg logit contributions | |
rily | -0.731 |
ittered | -0.729 |
avail | -0.697 |
inations | -0.696 |
iously | -0.696 |
Token | fully |
Token ID | 2759 |
Feature activation | +4.32e-01 |
Pos logit contributions | |
rily | +0.279 |
avail | +0.264 |
ittered | +0.261 |
immer | +0.250 |
ently | +0.249 |
Neg logit contributions | |
big | -0.426 |
a | -0.422 |
| -0.413 |
two | -0.401 |
and | -0.388 |
Token | got |
Token ID | 1392 |
Feature activation | -1.16e-01 |
Pos logit contributions | |
a | +0.572 |
big | +0.561 |
| +0.541 |
two | +0.491 |
little | +0.451 |
Neg logit contributions | |
rily | -1.573 |
ittered | -1.544 |
iously | -1.511 |
leness | -1.501 |
immer | -1.492 |
Token | to |
Token ID | 284 |
Feature activation | -3.05e-01 |
Pos logit contributions | |
rily | +0.458 |
avail | +0.457 |
ittered | +0.442 |
keeping | +0.437 |
immer | +0.433 |
Neg logit contributions | |
| -0.102 |
and | -0.099 |
a | -0.097 |
big | -0.093 |
two | -0.082 |
Token | work |
Token ID | 670 |
Feature activation | -2.22e-01 |
Pos logit contributions | |
avail | +1.131 |
rily | +1.067 |
ittered | +1.046 |
inations | +1.046 |
keeping | +1.044 |
Neg logit contributions | |
a | -0.392 |
| -0.384 |
big | -0.372 |
two | -0.360 |
and | -0.354 |
Token | and |
Token ID | 290 |
Feature activation | -5.61e-02 |
Pos logit contributions | |
rily | +0.955 |
avail | +0.950 |
ittered | +0.945 |
lights | +0.942 |
immer | +0.935 |
Neg logit contributions | |
a | -0.133 |
| -0.119 |
big | -0.115 |
two | -0.107 |
little | -0.088 |
Token | started |
Token ID | 2067 |
Feature activation | -3.14e-01 |
Pos logit contributions | |
avail | +0.186 |
keeping | +0.178 |
rily | +0.176 |
ittered | +0.176 |
inations | +0.174 |
Neg logit contributions | |
| -0.092 |
a | -0.091 |
big | -0.090 |
and | -0.084 |
two | -0.083 |
Token | tid |
Token ID | 29770 |
Feature activation | -1.98e+00 |
Pos logit contributions | |
avail | +1.191 |
rily | +1.165 |
ittered | +1.163 |
keeping | +1.160 |
epad | +1.136 |
Neg logit contributions | |
a | -0.341 |
| -0.333 |
big | -0.318 |
two | -0.301 |
and | -0.301 |
Token | the |
Token ID | 262 |
Feature activation | +4.81e-01 |
Pos logit contributions | |
avail | +0.642 |
rily | +0.619 |
keeping | +0.613 |
ge | +0.607 |
fer | +0.607 |
Neg logit contributions | |
a | -0.105 |
big | -0.096 |
and | -0.096 |
| -0.086 |
two | -0.082 |
Token | lake |
Token ID | 13546 |
Feature activation | -3.12e-01 |
Pos logit contributions | |
a | +1.240 |
and | +1.209 |
| +1.190 |
two | +1.126 |
big | +1.098 |
Neg logit contributions | |
avail | -1.180 |
rily | -1.069 |
leness | -1.069 |
fer | -1.044 |
footing | -1.043 |
Token | . |
Token ID | 13 |
Feature activation | -1.90e-01 |
Pos logit contributions | |
rily | +1.341 |
avail | +1.317 |
keeping | +1.303 |
ittered | +1.300 |
lights | +1.299 |
Neg logit contributions | |
a | -0.208 |
| -0.173 |
big | -0.172 |
two | -0.157 |
and | -0.128 |
Token |
|
Token ID | 198 |
Feature activation | -5.34e-01 |
Pos logit contributions | |
rily | +0.760 |
keeping | +0.749 |
avail | +0.749 |
lights | +0.741 |
leness | +0.740 |
Neg logit contributions | |
a | -0.170 |
big | -0.166 |
and | -0.140 |
two | -0.128 |
little | -0.115 |
Token |
|
Token ID | 198 |
Feature activation | -7.21e-01 |
Pos logit contributions | |
inations | +1.833 |
rily | +1.827 |
ittered | +1.823 |
leness | +1.795 |
immer | +1.791 |
Neg logit contributions | |
a | -0.753 |
and | -0.686 |
big | -0.678 |
two | -0.658 |
little | -0.609 |
Token | The |
Token ID | 464 |
Feature activation | +2.29e-01 |
Pos logit contributions | |
inations | +2.169 |
rily | +2.149 |
avail | +2.147 |
leness | +2.132 |
ittered | +2.124 |
Neg logit contributions | |
a | -1.391 |
and | -1.311 |
big | -1.306 |
| -1.215 |
two | -1.211 |
Token | end |
Token ID | 886 |
Feature activation | +4.44e-01 |
Pos logit contributions | |
a | +0.412 |
| +0.398 |
and | +0.374 |
big | +0.346 |
to | +0.339 |
Neg logit contributions | |
avail | -0.742 |
leness | -0.718 |
lights | -0.716 |
rily | -0.715 |
inations | -0.706 |
Token | ! |
Token ID | 0 |
Feature activation | +1.15e-01 |
Pos logit contributions | |
a | +0.329 |
two | +0.302 |
big | +0.282 |
and | +0.225 |
| +0.219 |
Neg logit contributions | |
lights | -1.850 |
avail | -1.814 |
inations | -1.800 |
ittered | -1.795 |
eway | -1.787 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +5.74e-01 |
Pos logit contributions | |
a | +0.052 |
big | +0.051 |
and | +0.039 |
two | +0.030 |
little | +0.025 |
Neg logit contributions | |
avail | -0.502 |
rily | -0.501 |
keeping | -0.496 |
lights | -0.490 |
vation | -0.488 |
Token | Once |
Token ID | 7454 |
Feature activation | -4.07e-01 |
Pos logit contributions | |
big | +0.944 |
a | +0.942 |
and | +0.847 |
little | +0.837 |
two | +0.810 |
Neg logit contributions | |
keeping | -1.814 |
avail | -1.794 |
rily | -1.775 |
ittered | -1.760 |
fer | -1.730 |
Token | upon |
Token ID | 2402 |
Feature activation | +4.64e-01 |
Pos logit contributions | |
avail | +1.104 |
rily | +1.098 |
ittered | +1.098 |
lights | +1.096 |
ently | +1.062 |
Neg logit contributions | |
a | -0.900 |
two | -0.819 |
big | -0.816 |
and | -0.809 |
| -0.768 |
Token | You |
Token ID | 921 |
Feature activation | +3.78e-01 |
Pos logit contributions | |
avail | +1.698 |
rily | +1.668 |
keeping | +1.649 |
lights | +1.625 |
leness | +1.612 |
Neg logit contributions | |
a | -1.106 |
big | -1.094 |
and | -1.028 |
two | -0.977 |
the | -0.955 |
Token | have |
Token ID | 423 |
Feature activation | +4.34e-01 |
Pos logit contributions | |
a | +0.629 |
big | +0.597 |
| +0.583 |
and | +0.555 |
two | +0.552 |
Neg logit contributions | |
inations | -1.218 |
ittered | -1.181 |
leness | -1.173 |
arians | -1.156 |
avail | -1.153 |
Token | to |
Token ID | 284 |
Feature activation | +8.96e-02 |
Pos logit contributions | |
| +0.296 |
and | +0.293 |
a | +0.283 |
big | +0.263 |
two | +0.229 |
Neg logit contributions | |
avail | -1.750 |
keeping | -1.741 |
inations | -1.709 |
leness | -1.697 |
ittered | -1.693 |
Token | stay |
Token ID | 2652 |
Feature activation | -2.78e-01 |
Pos logit contributions | |
a | +0.152 |
| +0.148 |
big | +0.143 |
and | +0.139 |
two | +0.135 |
Neg logit contributions | |
avail | -0.302 |
immer | -0.285 |
leness | -0.282 |
keeping | -0.278 |
ittered | -0.277 |
Token | where |
Token ID | 810 |
Feature activation | -5.92e-02 |
Pos logit contributions | |
rily | +1.036 |
avail | +1.036 |
ittered | +1.017 |
immer | +1.005 |
keeping | +0.997 |
Neg logit contributions | |
a | -0.319 |
| -0.309 |
big | -0.296 |
two | -0.285 |
and | -0.261 |
Token | I |
Token ID | 314 |
Feature activation | +3.40e-01 |
Pos logit contributions | |
avail | +0.208 |
lights | +0.202 |
rily | +0.199 |
keeping | +0.199 |
inations | +0.196 |
Neg logit contributions | |
a | -0.083 |
and | -0.079 |
| -0.078 |
big | -0.076 |
two | -0.071 |
Token | can |
Token ID | 460 |
Feature activation | +3.48e-02 |
Pos logit contributions | |
a | +0.628 |
big | +0.586 |
two | +0.575 |
and | +0.554 |
| +0.553 |
Neg logit contributions | |
ittered | -1.062 |
inations | -1.049 |
immer | -1.023 |
leness | -1.019 |
avail | -1.011 |
Token | see |
Token ID | 766 |
Feature activation | -8.78e-02 |
Pos logit contributions | |
a | +0.059 |
big | +0.053 |
two | +0.052 |
| +0.051 |
and | +0.050 |
Neg logit contributions | |
avail | -0.113 |
ittered | -0.111 |
inations | -0.110 |
immer | -0.110 |
rily | -0.109 |
Token | you |
Token ID | 345 |
Feature activation | -5.31e-01 |
Pos logit contributions | |
avail | +0.315 |
rily | +0.308 |
lights | +0.302 |
ittered | +0.297 |
inations | +0.297 |
Neg logit contributions | |
a | -0.113 |
big | -0.110 |
| -0.108 |
two | -0.103 |
and | -0.102 |
Token | , |
Token ID | 11 |
Feature activation | +5.42e-01 |
Pos logit contributions | |
rily | +2.008 |
ittered | +2.005 |
avail | +2.002 |
inations | +1.991 |
immer | +1.988 |
Neg logit contributions | |
a | -0.563 |
big | -0.523 |
| -0.510 |
two | -0.487 |
and | -0.454 |
Token | okay |
Token ID | 8788 |
Feature activation | -4.67e-01 |
Pos logit contributions | |
a | +0.774 |
| +0.755 |
big | +0.722 |
two | +0.671 |
and | +0.624 |
Neg logit contributions | |
avail | -1.877 |
rily | -1.853 |
lights | -1.790 |
fer | -1.784 |
inations | -1.783 |
Token | watched |
Token ID | 7342 |
Feature activation | -2.14e-01 |
Pos logit contributions | |
avail | +0.152 |
rily | +0.148 |
ittered | +0.148 |
fer | +0.147 |
leness | +0.145 |
Neg logit contributions | |
a | -0.084 |
big | -0.084 |
| -0.082 |
and | -0.079 |
two | -0.076 |
Token | in |
Token ID | 287 |
Feature activation | +1.47e-01 |
Pos logit contributions | |
rily | +0.761 |
avail | +0.741 |
ittered | +0.722 |
lights | +0.721 |
immer | +0.704 |
Neg logit contributions | |
a | -0.304 |
big | -0.302 |
| -0.297 |
two | -0.281 |
and | -0.268 |
Token | amaz |
Token ID | 40642 |
Feature activation | -8.01e-01 |
Pos logit contributions | |
a | +0.348 |
| +0.344 |
big | +0.340 |
and | +0.329 |
two | +0.328 |
Neg logit contributions | |
avail | -0.376 |
rily | -0.352 |
lights | -0.341 |
keeping | -0.337 |
ittered | -0.336 |
Token | ement |
Token ID | 972 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
avail | +0.442 |
rily | +0.423 |
lights | +0.406 |
ittered | +0.354 |
ently | +0.342 |
Neg logit contributions | |
a | -3.652 |
big | -3.557 |
| -3.547 |
two | -3.492 |
and | -3.482 |
Token | . |
Token ID | 13 |
Feature activation | -3.70e-01 |
Pos logit contributions | |
rily | +0.493 |
lights | +0.490 |
avail | +0.485 |
ittered | +0.484 |
iously | +0.480 |
Neg logit contributions | |
a | -0.102 |
big | -0.097 |
| -0.091 |
two | -0.083 |
and | -0.069 |
Token | The |
Token ID | 383 |
Feature activation | +7.25e-02 |
Pos logit contributions | |
rily | +1.198 |
avail | +1.191 |
lights | +1.164 |
ittered | +1.150 |
leness | +1.129 |
Neg logit contributions | |
a | -0.628 |
big | -0.611 |
and | -0.577 |
two | -0.542 |
the | -0.522 |
Token | dog |
Token ID | 3290 |
Feature activation | +6.62e-02 |
Pos logit contributions | |
a | +0.134 |
| +0.133 |
and | +0.128 |
big | +0.118 |
the | +0.116 |
Neg logit contributions | |
avail | -0.225 |
lights | -0.219 |
rily | -0.215 |
leness | -0.215 |
fer | -0.215 |
Token | too |
Token ID | 1165 |
Feature activation | +4.87e-01 |
Pos logit contributions | |
a | +0.115 |
| +0.106 |
big | +0.106 |
two | +0.100 |
and | +0.096 |
Neg logit contributions | |
ittered | -0.209 |
rily | -0.208 |
avail | -0.207 |
leness | -0.204 |
fer | -0.204 |
Token | was |
Token ID | 373 |
Feature activation | -1.52e-01 |
Pos logit contributions | |
a | +0.731 |
big | +0.678 |
two | +0.667 |
| +0.660 |
and | +0.626 |
Neg logit contributions | |
ittered | -1.642 |
leness | -1.642 |
istent | -1.598 |
rily | -1.596 |
rated | -1.596 |
Token | surprised |
Token ID | 6655 |
Feature activation | +5.38e-01 |
Pos logit contributions | |
avail | +0.457 |
rily | +0.455 |
ittered | +0.454 |
immer | +0.438 |
inations | +0.436 |
Neg logit contributions | |
a | -0.279 |
| -0.273 |
and | -0.262 |
big | -0.261 |
two | -0.254 |
Token | , |
Token ID | 11 |
Feature activation | -1.97e-02 |
Pos logit contributions | |
a | +0.502 |
big | +0.471 |
two | +0.430 |
| +0.413 |
little | +0.309 |
Neg logit contributions | |
avail | -2.206 |
rily | -2.167 |
lights | -2.137 |
iously | -2.111 |
ittered | -2.082 |
Token | cut |
Token ID | 2005 |
Feature activation | +1.61e-01 |
Pos logit contributions | |
a | +0.510 |
| +0.485 |
big | +0.474 |
two | +0.409 |
and | +0.391 |
Neg logit contributions | |
avail | -1.654 |
rily | -1.630 |
lights | -1.610 |
inations | -1.584 |
immer | -1.546 |
Token | with |
Token ID | 351 |
Feature activation | +1.44e-01 |
Pos logit contributions | |
a | +0.170 |
| +0.167 |
big | +0.166 |
two | +0.147 |
and | +0.144 |
Neg logit contributions | |
rily | -0.621 |
avail | -0.609 |
lights | -0.606 |
ittered | -0.597 |
ently | -0.593 |
Token | stitches |
Token ID | 28096 |
Feature activation | -2.26e-01 |
Pos logit contributions | |
| +0.173 |
and | +0.164 |
a | +0.163 |
big | +0.154 |
two | +0.142 |
Neg logit contributions | |
avail | -0.532 |
rily | -0.521 |
lights | -0.516 |
ittered | -0.515 |
leness | -0.506 |
Token | and |
Token ID | 290 |
Feature activation | +1.65e-01 |
Pos logit contributions | |
rily | +0.944 |
ittered | +0.938 |
keeping | +0.928 |
lights | +0.923 |
leness | +0.920 |
Neg logit contributions | |
a | -0.184 |
big | -0.169 |
| -0.155 |
two | -0.144 |
little | -0.124 |
Token | put |
Token ID | 1234 |
Feature activation | +3.91e-01 |
Pos logit contributions | |
| +0.360 |
a | +0.353 |
big | +0.345 |
and | +0.342 |
two | +0.324 |
Neg logit contributions | |
avail | -0.440 |
lights | -0.436 |
leness | -0.434 |
inations | -0.434 |
keeping | -0.426 |
Token | a |
Token ID | 257 |
Feature activation | +6.27e-02 |
Pos logit contributions | |
| +0.568 |
big | +0.564 |
a | +0.564 |
and | +0.549 |
two | +0.516 |
Neg logit contributions | |
rily | -1.336 |
avail | -1.305 |
lights | -1.302 |
ittered | -1.300 |
keeping | -1.270 |
Token | band |
Token ID | 4097 |
Feature activation | +1.76e-01 |
Pos logit contributions | |
a | +0.134 |
| +0.129 |
and | +0.125 |
two | +0.122 |
the | +0.120 |
Neg logit contributions | |
rily | -0.173 |
avail | -0.172 |
inations | -0.170 |
immer | -0.169 |
leness | -0.169 |
Token | age |
Token ID | 496 |
Feature activation | +2.13e-01 |
Pos logit contributions | |
a | +0.576 |
big | +0.567 |
| +0.566 |
two | +0.554 |
the | +0.540 |
Neg logit contributions | |
ittered | -0.304 |
keeping | -0.296 |
inations | -0.293 |
immer | -0.290 |
leness | -0.288 |
Token | on |
Token ID | 319 |
Feature activation | +4.02e-01 |
Pos logit contributions | |
a | +0.235 |
| +0.231 |
big | +0.219 |
two | +0.207 |
little | +0.196 |
Neg logit contributions | |
rily | -0.812 |
lights | -0.804 |
ittered | -0.803 |
immer | -0.789 |
keeping | -0.781 |
Token | it |
Token ID | 340 |
Feature activation | -4.03e-02 |
Pos logit contributions | |
a | +0.384 |
big | +0.360 |
| +0.354 |
and | +0.342 |
two | +0.340 |
Neg logit contributions | |
rily | -1.574 |
avail | -1.564 |
lights | -1.520 |
ittered | -1.498 |
keeping | -1.493 |
Token | . |
Token ID | 13 |
Feature activation | -2.53e-01 |
Pos logit contributions | |
rily | +0.167 |
ittered | +0.164 |
lights | +0.163 |
immer | +0.162 |
keeping | +0.162 |
Neg logit contributions | |
a | -0.031 |
big | -0.027 |
| -0.026 |
two | -0.026 |
and | -0.021 |
Token | with |
Token ID | 351 |
Feature activation | +3.06e-01 |
Pos logit contributions | |
a | +0.239 |
| +0.227 |
big | +0.227 |
two | +0.219 |
and | +0.187 |
Neg logit contributions | |
rily | -1.050 |
ittered | -1.027 |
avail | -1.007 |
immer | -0.999 |
lights | -0.987 |
Token | joy |
Token ID | 8716 |
Feature activation | -7.60e-02 |
Pos logit contributions | |
and | +0.775 |
| +0.772 |
a | +0.749 |
big | +0.730 |
two | +0.719 |
Neg logit contributions | |
avail | -0.749 |
leness | -0.683 |
rily | -0.662 |
ittered | -0.653 |
inations | -0.647 |
Token | and |
Token ID | 290 |
Feature activation | -1.86e-01 |
Pos logit contributions | |
rily | +0.326 |
ittered | +0.320 |
lights | +0.319 |
avail | +0.318 |
keeping | +0.314 |
Neg logit contributions | |
a | -0.054 |
big | -0.051 |
| -0.048 |
two | -0.043 |
and | -0.030 |
Token | said |
Token ID | 531 |
Feature activation | -1.69e-01 |
Pos logit contributions | |
avail | +0.586 |
leness | +0.564 |
keeping | +0.564 |
ittered | +0.560 |
rily | +0.547 |
Neg logit contributions | |
a | -0.331 |
| -0.326 |
big | -0.319 |
and | -0.318 |
two | -0.299 |
Token | , |
Token ID | 11 |
Feature activation | +5.49e-01 |
Pos logit contributions | |
rily | +0.645 |
avail | +0.644 |
ittered | +0.641 |
leness | +0.630 |
fer | +0.623 |
Neg logit contributions | |
a | -0.178 |
| -0.161 |
and | -0.161 |
big | -0.160 |
two | -0.150 |
Token | � |
Token ID | 6184 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
a | +0.766 |
big | +0.728 |
two | +0.660 |
and | +0.600 |
little | +0.582 |
Neg logit contributions | |
rily | -1.948 |
leness | -1.865 |
fer | -1.863 |
istent | -1.860 |
ittered | -1.859 |
Token | � |
Token ID | 95 |
Feature activation | -9.32e-01 |
Pos logit contributions | |
avail | +2.311 |
ittered | +2.247 |
rily | +2.196 |
immer | +2.156 |
leness | +2.152 |
Neg logit contributions | |
a | -4.613 |
big | -4.415 |
and | -4.397 |
| -4.339 |
two | -4.338 |
Token | € |
Token ID | 26391 |
Feature activation | -6.03e-01 |
Pos logit contributions | |
avail | +1.833 |
ittered | +1.760 |
vation | +1.759 |
leness | +1.757 |
fer | +1.725 |
Neg logit contributions | |
a | -2.885 |
big | -2.800 |
and | -2.730 |
two | -2.690 |
| -2.655 |
Token | ˜ |
Token ID | 41185 |
Feature activation | -3.18e-01 |
Pos logit contributions | |
ittered | +1.515 |
avail | +1.509 |
fer | +1.482 |
inations | +1.478 |
keeping | +1.464 |
Neg logit contributions | |
a | -1.484 |
big | -1.388 |
two | -1.354 |
and | -1.346 |
the | -1.302 |
Token | Let |
Token ID | 5756 |
Feature activation | +9.85e-01 |
Pos logit contributions | |
avail | +1.117 |
rily | +1.116 |
ittered | +1.114 |
fer | +1.105 |
keeping | +1.087 |
Neg logit contributions | |
a | -0.454 |
big | -0.436 |
| -0.422 |
and | -0.395 |
two | -0.385 |
Token | â |
Token ID | 22940 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
a | +0.999 |
big | +0.913 |
| +0.799 |
and | +0.784 |
two | +0.760 |
Neg logit contributions | |
avail | -3.929 |
rated | -3.823 |
rily | -3.746 |
vation | -3.719 |
ently | -3.697 |
Token | have |
Token ID | 423 |
Feature activation | +1.47e-01 |
Pos logit contributions | |
a | +0.483 |
big | +0.438 |
| +0.432 |
and | +0.426 |
two | +0.424 |
Neg logit contributions | |
avail | -1.082 |
immer | -1.053 |
ittered | -1.047 |
leness | -1.044 |
rily | -1.040 |
Token | to |
Token ID | 284 |
Feature activation | +2.71e-01 |
Pos logit contributions | |
a | +0.140 |
| +0.139 |
and | +0.138 |
big | +0.136 |
two | +0.119 |
Neg logit contributions | |
avail | -0.571 |
leness | -0.549 |
rily | -0.548 |
inations | -0.548 |
ittered | -0.546 |
Token | rot |
Token ID | 5724 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
a | +0.454 |
| +0.444 |
big | +0.431 |
and | +0.424 |
two | +0.402 |
Neg logit contributions | |
avail | -0.914 |
immer | -0.861 |
leness | -0.852 |
keeping | -0.847 |
inations | -0.847 |
Token | it |
Token ID | 340 |
Feature activation | -1.58e-01 |
Pos logit contributions | |
avail | +0.460 |
rily | +0.454 |
keeping | +0.438 |
leness | +0.433 |
ittered | +0.429 |
Neg logit contributions | |
a | -0.139 |
big | -0.135 |
| -0.135 |
two | -0.123 |
and | -0.121 |
Token | first |
Token ID | 717 |
Feature activation | -1.67e-02 |
Pos logit contributions | |
rily | +0.599 |
avail | +0.584 |
ittered | +0.581 |
immer | +0.580 |
rated | +0.579 |
Neg logit contributions | |
a | -0.181 |
| -0.169 |
big | -0.165 |
two | -0.155 |
and | -0.148 |
Token | â |
Token ID | 22940 |
Feature activation | -9.55e-01 |
Pos logit contributions | |
rily | +0.063 |
avail | +0.063 |
ittered | +0.062 |
immer | +0.062 |
rated | +0.062 |
Neg logit contributions | |
a | -0.019 |
big | -0.018 |
| -0.016 |
two | -0.016 |
little | -0.014 |
Token | € |
Token ID | 26391 |
Feature activation | -6.37e-01 |
Pos logit contributions | |
avail | +1.283 |
rated | +1.193 |
immer | +1.185 |
leness | +1.181 |
ittered | +1.158 |
Neg logit contributions | |
a | -3.619 |
big | -3.563 |
and | -3.467 |
| -3.433 |
two | -3.403 |
Token | . |
Token ID | 13 |
Feature activation | -9.49e-02 |
Pos logit contributions | |
avail | +1.156 |
fer | +1.121 |
keeping | +1.116 |
ittered | +1.115 |
inations | +1.096 |
Neg logit contributions | |
a | -2.018 |
big | -1.960 |
two | -1.907 |
and | -1.901 |
the | -1.849 |
Token | |
Token ID | 220 |
Feature activation | -2.60e-01 |
Pos logit contributions | |
avail | +0.268 |
rily | +0.266 |
leness | +0.257 |
keeping | +0.257 |
arians | +0.257 |
Neg logit contributions | |
big | -0.206 |
a | -0.203 |
and | -0.187 |
two | -0.178 |
the | -0.176 |
Token |
|
Token ID | 198 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
rily | +1.152 |
avail | +1.142 |
ittered | +1.137 |
ently | +1.131 |
immer | +1.129 |
Neg logit contributions | |
a | -0.127 |
big | -0.107 |
and | -0.100 |
two | -0.063 |
the | -0.055 |
Token |
|
Token ID | 198 |
Feature activation | -1.74e-01 |
Pos logit contributions | |
rily | +0.599 |
leness | +0.593 |
avail | +0.592 |
inations | +0.592 |
immer | +0.589 |
Neg logit contributions | |
a | -0.079 |
big | -0.062 |
and | -0.056 |
two | -0.044 |
the | -0.037 |
Token | know |
Token ID | 760 |
Feature activation | -1.39e+00 |
Pos logit contributions | |
avail | +1.932 |
leness | +1.837 |
rated | +1.836 |
immer | +1.835 |
ittered | +1.834 |
Neg logit contributions | |
a | -1.172 |
| -1.098 |
big | -1.096 |
two | -1.089 |
and | -1.054 |
Token | that |
Token ID | 326 |
Feature activation | -5.85e-01 |
Pos logit contributions | |
avail | +4.371 |
rily | +4.293 |
ittered | +4.264 |
lights | +4.233 |
immer | +4.159 |
Neg logit contributions | |
a | -2.490 |
big | -2.396 |
| -2.329 |
and | -2.313 |
two | -2.247 |
Token | . |
Token ID | 13 |
Feature activation | +4.30e-01 |
Pos logit contributions | |
avail | +2.441 |
lights | +2.358 |
rily | +2.349 |
keeping | +2.327 |
leness | +2.325 |
Neg logit contributions | |
a | -0.393 |
big | -0.366 |
| -0.363 |
and | -0.353 |
two | -0.331 |
Token | The |
Token ID | 383 |
Feature activation | +2.62e-01 |
Pos logit contributions | |
a | +0.670 |
big | +0.663 |
and | +0.613 |
two | +0.600 |
| +0.545 |
Neg logit contributions | |
rily | -1.436 |
avail | -1.344 |
lights | -1.343 |
iously | -1.342 |
ittered | -1.331 |
Token | seal |
Token ID | 13810 |
Feature activation | +4.59e-01 |
Pos logit contributions | |
a | +0.231 |
| +0.214 |
and | +0.209 |
. | +0.169 |
to | +0.145 |
Neg logit contributions | |
avail | -1.034 |
rily | -1.023 |
fer | -1.002 |
footing | -1.001 |
leness | -0.996 |
Token | spl |
Token ID | 4328 |
Feature activation | -1.53e+00 |
Pos logit contributions | |
a | +0.693 |
| +0.622 |
big | +0.617 |
two | +0.573 |
the | +0.541 |
Neg logit contributions | |
rily | -1.556 |
ittered | -1.543 |
fer | -1.527 |
rated | -1.515 |
inations | -1.488 |
Token | ashed |
Token ID | 5263 |
Feature activation | +2.33e-01 |
Pos logit contributions | |
ittered | +2.492 |
rily | +2.389 |
avail | +2.317 |
immer | +2.266 |
lights | +2.260 |
Neg logit contributions | |
a | -5.397 |
big | -5.188 |
| -5.128 |
and | -5.083 |
two | -5.037 |
Token | the |
Token ID | 262 |
Feature activation | +6.79e-02 |
Pos logit contributions | |
a | +0.202 |
| +0.198 |
big | +0.192 |
two | +0.175 |
little | +0.157 |
Neg logit contributions | |
rily | -0.910 |
ittered | -0.890 |
lights | -0.884 |
rated | -0.879 |
avail | -0.878 |
Token | big |
Token ID | 1263 |
Feature activation | -4.03e-01 |
Pos logit contributions | |
a | +0.106 |
| +0.099 |
and | +0.090 |
the | +0.083 |
two | +0.083 |
Neg logit contributions | |
rily | -0.231 |
footing | -0.223 |
leness | -0.222 |
rated | -0.222 |
lights | -0.222 |
Token | fish |
Token ID | 5916 |
Feature activation | +3.73e-02 |
Pos logit contributions | |
avail | +1.188 |
rily | +1.173 |
leness | +1.140 |
lights | +1.136 |
rated | +1.125 |
Neg logit contributions | |
a | -0.826 |
| -0.792 |
big | -0.719 |
two | -0.714 |
and | -0.704 |
Token | with |
Token ID | 351 |
Feature activation | +4.36e-01 |
Pos logit contributions | |
a | +0.028 |
| +0.028 |
big | +0.025 |
two | +0.023 |
little | +0.020 |
Neg logit contributions | |
ittered | -0.151 |
rily | -0.150 |
footing | -0.147 |
rated | -0.145 |
immer | -0.145 |
Token | as |
Token ID | 355 |
Feature activation | +1.28e-01 |
Pos logit contributions | |
a | +0.458 |
big | +0.455 |
two | +0.403 |
| +0.402 |
the | +0.274 |
Neg logit contributions | |
rily | -3.365 |
avail | -3.148 |
immer | -3.142 |
ittered | -3.138 |
rier | -3.087 |
Token | he |
Token ID | 339 |
Feature activation | +1.05e-01 |
Pos logit contributions | |
a | +0.209 |
and | +0.198 |
big | +0.195 |
| +0.195 |
two | +0.191 |
Neg logit contributions | |
avail | -0.417 |
rily | -0.412 |
leness | -0.397 |
ittered | -0.396 |
lights | -0.396 |
Token | watched |
Token ID | 7342 |
Feature activation | -6.58e-02 |
Pos logit contributions | |
a | +0.210 |
big | +0.196 |
| +0.192 |
two | +0.188 |
and | +0.181 |
Neg logit contributions | |
ittered | -0.313 |
rily | -0.306 |
leness | -0.303 |
inations | -0.300 |
avail | -0.298 |
Token | the |
Token ID | 262 |
Feature activation | +6.75e-01 |
Pos logit contributions | |
rily | +0.273 |
avail | +0.265 |
immer | +0.255 |
ittered | +0.255 |
leness | +0.253 |
Neg logit contributions | |
a | -0.058 |
big | -0.055 |
| -0.053 |
two | -0.049 |
and | -0.048 |
Token | ducks |
Token ID | 39694 |
Feature activation | -3.65e-01 |
Pos logit contributions | |
a | +1.592 |
and | +1.486 |
| +1.468 |
big | +1.385 |
the | +1.369 |
Neg logit contributions | |
leness | -1.726 |
rily | -1.723 |
avail | -1.688 |
rated | -1.675 |
inations | -1.634 |
Token | gob |
Token ID | 48484 |
Feature activation | -1.09e+00 |
Pos logit contributions | |
rily | +1.287 |
ittered | +1.254 |
immer | +1.246 |
avail | +1.243 |
leness | +1.237 |
Neg logit contributions | |
a | -0.539 |
big | -0.495 |
| -0.484 |
two | -0.471 |
and | -0.443 |
Token | ble |
Token ID | 903 |
Feature activation | +2.70e-02 |
Pos logit contributions | |
rily | +1.225 |
ittered | +1.074 |
avail | +1.073 |
immer | +1.001 |
ently | +0.975 |
Neg logit contributions | |
a | -4.373 |
big | -4.269 |
| -4.234 |
and | -4.197 |
two | -4.103 |
Token | up |
Token ID | 510 |
Feature activation | +2.36e-01 |
Pos logit contributions | |
a | +0.040 |
| +0.038 |
big | +0.037 |
two | +0.036 |
little | +0.034 |
Neg logit contributions | |
rily | -0.098 |
immer | -0.091 |
avail | -0.091 |
ittered | -0.090 |
lights | -0.089 |
Token | the |
Token ID | 262 |
Feature activation | +1.05e-01 |
Pos logit contributions | |
a | +0.143 |
| +0.143 |
big | +0.133 |
two | +0.120 |
and | +0.097 |
Neg logit contributions | |
rily | -1.031 |
avail | -0.984 |
ittered | -0.972 |
immer | -0.964 |
lights | -0.963 |
Token | food |
Token ID | 2057 |
Feature activation | -1.54e-01 |
Pos logit contributions | |
a | +0.251 |
| +0.244 |
and | +0.236 |
the | +0.225 |
big | +0.224 |
Neg logit contributions | |
rily | -0.273 |
rated | -0.261 |
avail | -0.260 |
footing | -0.256 |
ittered | -0.254 |
Token | . |
Token ID | 13 |
Feature activation | -1.08e-01 |
Pos logit contributions | |
rily | +0.669 |
avail | +0.640 |
immer | +0.634 |
ittered | +0.633 |
lights | +0.629 |
Neg logit contributions | |
a | -0.119 |
| -0.101 |
big | -0.097 |
two | -0.092 |
little | -0.075 |
Token | in |
Token ID | 287 |
Feature activation | -1.48e-01 |
Pos logit contributions | |
rily | +1.293 |
immer | +1.284 |
ittered | +1.282 |
lights | +1.263 |
avail | +1.250 |
Neg logit contributions | |
a | -0.290 |
| -0.271 |
big | -0.255 |
two | -0.247 |
little | -0.210 |
Token | the |
Token ID | 262 |
Feature activation | +5.22e-01 |
Pos logit contributions | |
avail | +0.660 |
rily | +0.638 |
keeping | +0.634 |
lights | +0.625 |
ge | +0.618 |
Neg logit contributions | |
a | -0.064 |
and | -0.062 |
big | -0.058 |
| -0.056 |
little | -0.039 |
Token | yard |
Token ID | 12699 |
Feature activation | +9.82e-02 |
Pos logit contributions | |
a | +1.325 |
and | +1.322 |
| +1.271 |
two | +1.209 |
big | +1.193 |
Neg logit contributions | |
avail | -1.191 |
keeping | -1.183 |
lights | -1.104 |
vation | -1.104 |
stack | -1.087 |
Token | . |
Token ID | 13 |
Feature activation | -1.73e-01 |
Pos logit contributions | |
a | +0.056 |
big | +0.054 |
two | +0.039 |
| +0.037 |
little | +0.033 |
Neg logit contributions | |
rily | -0.423 |
keeping | -0.422 |
lights | -0.414 |
ittered | -0.413 |
immer | -0.412 |
Token | They |
Token ID | 1119 |
Feature activation | +9.35e-02 |
Pos logit contributions | |
rily | +0.606 |
avail | +0.588 |
keeping | +0.580 |
lights | +0.571 |
ittered | +0.567 |
Neg logit contributions | |
a | -0.249 |
big | -0.247 |
and | -0.230 |
two | -0.217 |
. | -0.201 |
Token | wonder |
Token ID | 4240 |
Feature activation | -8.62e-01 |
Pos logit contributions | |
a | +0.148 |
| +0.145 |
big | +0.137 |
and | +0.131 |
two | +0.127 |
Neg logit contributions | |
rily | -0.304 |
avail | -0.302 |
ittered | -0.302 |
immer | -0.295 |
inations | -0.294 |
Token | what |
Token ID | 644 |
Feature activation | -3.56e-01 |
Pos logit contributions | |
rily | +2.645 |
avail | +2.642 |
ittered | +2.581 |
inations | +2.575 |
lights | +2.570 |
Neg logit contributions | |
a | -1.562 |
| -1.534 |
big | -1.481 |
two | -1.427 |
and | -1.403 |
Token | is |
Token ID | 318 |
Feature activation | -3.75e-01 |
Pos logit contributions | |
avail | +1.399 |
lights | +1.331 |
rily | +1.326 |
inations | +1.311 |
keeping | +1.311 |
Neg logit contributions | |
a | -0.332 |
and | -0.311 |
| -0.296 |
two | -0.284 |
big | -0.280 |
Token | inside |
Token ID | 2641 |
Feature activation | -3.98e-01 |
Pos logit contributions | |
rily | +1.253 |
avail | +1.252 |
ittered | +1.216 |
keeping | +1.207 |
immer | +1.196 |
Neg logit contributions | |
| -0.571 |
a | -0.561 |
big | -0.531 |
and | -0.517 |
two | -0.516 |
Token | . |
Token ID | 13 |
Feature activation | -1.34e-01 |
Pos logit contributions | |
avail | +1.709 |
rily | +1.707 |
lights | +1.662 |
immer | +1.654 |
keeping | +1.645 |
Neg logit contributions | |
a | -0.232 |
big | -0.207 |
| -0.190 |
two | -0.176 |
and | -0.150 |
Token |
|
Token ID | 198 |
Feature activation | -1.90e-01 |
Pos logit contributions | |
rily | +0.571 |
avail | +0.568 |
keeping | +0.559 |
lights | +0.556 |
ittered | +0.545 |
Neg logit contributions | |
a | -0.085 |
big | -0.082 |
and | -0.067 |
two | -0.056 |
little | -0.045 |
Token | the |
Token ID | 262 |
Feature activation | -3.24e-01 |
Pos logit contributions | |
| +0.072 |
big | +0.068 |
a | +0.067 |
and | +0.065 |
two | +0.058 |
Neg logit contributions | |
avail | -0.376 |
lights | -0.369 |
rily | -0.368 |
ittered | -0.367 |
immer | -0.361 |
Token | shouting |
Token ID | 19642 |
Feature activation | +1.83e-01 |
Pos logit contributions | |
avail | +0.943 |
rily | +0.901 |
lights | +0.891 |
leness | +0.889 |
vation | +0.887 |
Neg logit contributions | |
a | -0.666 |
| -0.658 |
and | -0.637 |
big | -0.595 |
the | -0.589 |
Token | . |
Token ID | 13 |
Feature activation | -1.84e-01 |
Pos logit contributions | |
a | +0.154 |
| +0.150 |
big | +0.137 |
two | +0.132 |
and | +0.115 |
Neg logit contributions | |
rily | -0.738 |
avail | -0.724 |
lights | -0.723 |
ittered | -0.723 |
immer | -0.722 |
Token | Then |
Token ID | 3244 |
Feature activation | -8.63e-03 |
Pos logit contributions | |
avail | +0.547 |
rily | +0.546 |
ittered | +0.536 |
ently | +0.526 |
iously | +0.523 |
Neg logit contributions | |
a | -0.346 |
big | -0.340 |
and | -0.326 |
two | -0.309 |
the | -0.304 |
Token | she |
Token ID | 673 |
Feature activation | +2.45e-01 |
Pos logit contributions | |
avail | +0.032 |
ittered | +0.029 |
rily | +0.029 |
leness | +0.029 |
lights | +0.029 |
Neg logit contributions | |
big | -0.011 |
| -0.011 |
a | -0.011 |
and | -0.011 |
two | -0.009 |
Token | pe |
Token ID | 613 |
Feature activation | -2.22e+00 |
Pos logit contributions | |
a | +0.496 |
| +0.477 |
big | +0.459 |
and | +0.441 |
the | +0.439 |
Neg logit contributions | |
avail | -0.707 |
ittered | -0.690 |
rily | -0.682 |
fer | -0.680 |
immer | -0.678 |
Token | ered |
Token ID | 1068 |
Feature activation | -1.37e-01 |
Pos logit contributions | |
avail | +3.889 |
rily | +3.771 |
ittered | +3.754 |
immer | +3.588 |
leness | +3.562 |
Neg logit contributions | |
a | -7.532 |
big | -7.238 |
| -7.211 |
and | -7.202 |
two | -6.955 |
Token | up |
Token ID | 510 |
Feature activation | +3.78e-01 |
Pos logit contributions | |
rily | +0.452 |
avail | +0.446 |
ittered | +0.437 |
lights | +0.436 |
immer | +0.430 |
Neg logit contributions | |
| -0.219 |
a | -0.215 |
big | -0.213 |
two | -0.204 |
and | -0.192 |
Token | and |
Token ID | 290 |
Feature activation | -1.68e-01 |
Pos logit contributions | |
a | +0.286 |
big | +0.280 |
| +0.276 |
two | +0.271 |
little | +0.231 |
Neg logit contributions | |
avail | -1.535 |
rily | -1.515 |
ittered | -1.480 |
lights | -1.478 |
immer | -1.464 |
Token | saw |
Token ID | 2497 |
Feature activation | -2.69e-01 |
Pos logit contributions | |
avail | +0.592 |
rily | +0.577 |
ittered | +0.573 |
fer | +0.557 |
immer | +0.556 |
Neg logit contributions | |
a | -0.238 |
big | -0.228 |
and | -0.218 |
| -0.215 |
two | -0.214 |
Token | a |
Token ID | 257 |
Feature activation | -2.76e-01 |
Pos logit contributions | |
avail | +1.191 |
rily | +1.170 |
ittered | +1.160 |
lights | +1.146 |
keeping | +1.139 |
Neg logit contributions | |
and | -0.111 |
| -0.108 |
big | -0.098 |
a | -0.096 |
two | -0.059 |
Token | the |
Token ID | 262 |
Feature activation | +8.60e-01 |
Pos logit contributions | |
big | +0.225 |
a | +0.221 |
and | +0.215 |
two | +0.210 |
| +0.207 |
Neg logit contributions | |
rily | -0.298 |
avail | -0.296 |
ittered | -0.282 |
footing | -0.281 |
lights | -0.275 |
Token | owner |
Token ID | 4870 |
Feature activation | +1.85e-02 |
Pos logit contributions | |
a | +1.831 |
| +1.743 |
and | +1.704 |
the | +1.647 |
. | +1.597 |
Neg logit contributions | |
avail | -2.524 |
footing | -2.294 |
igi | -2.291 |
comb | -2.272 |
leness | -2.269 |
Token | saw |
Token ID | 2497 |
Feature activation | -2.85e-01 |
Pos logit contributions | |
a | +0.027 |
big | +0.025 |
| +0.024 |
two | +0.023 |
the | +0.021 |
Neg logit contributions | |
rily | -0.062 |
keeping | -0.062 |
inations | -0.061 |
ittered | -0.061 |
fer | -0.061 |
Token | some |
Token ID | 617 |
Feature activation | -4.96e-01 |
Pos logit contributions | |
rily | +1.038 |
avail | +1.008 |
ittered | +0.995 |
leness | +0.975 |
iously | +0.972 |
Neg logit contributions | |
big | -0.348 |
a | -0.334 |
and | -0.328 |
| -0.311 |
two | -0.302 |
Token | more |
Token ID | 517 |
Feature activation | -8.66e-02 |
Pos logit contributions | |
avail | +1.741 |
rily | +1.693 |
leness | +1.693 |
inations | +1.690 |
ittered | +1.669 |
Neg logit contributions | |
| -0.704 |
a | -0.698 |
and | -0.638 |
big | -0.597 |
two | -0.581 |
Token | che |
Token ID | 1125 |
Feature activation | -2.45e+00 |
Pos logit contributions | |
leness | +0.251 |
rily | +0.248 |
lights | +0.247 |
inations | +0.247 |
avail | +0.246 |
Neg logit contributions | |
a | -0.177 |
| -0.174 |
two | -0.163 |
big | -0.162 |
and | -0.155 |
Token | wed |
Token ID | 19103 |
Feature activation | +2.08e-01 |
Pos logit contributions | |
avail | +3.194 |
rily | +3.072 |
ittered | +3.022 |
leness | +2.898 |
immer | +2.840 |
Neg logit contributions | |
a | -9.440 |
big | -9.093 |
| -9.072 |
and | -9.056 |
two | -8.839 |
Token | threads |
Token ID | 14390 |
Feature activation | -2.08e-01 |
Pos logit contributions | |
a | +0.212 |
| +0.203 |
big | +0.203 |
two | +0.181 |
little | +0.171 |
Neg logit contributions | |
rily | -0.812 |
ittered | -0.783 |
avail | -0.781 |
lights | -0.774 |
keeping | -0.773 |
Token | in |
Token ID | 287 |
Feature activation | -6.32e-02 |
Pos logit contributions | |
rily | +0.888 |
ittered | +0.872 |
immer | +0.870 |
lights | +0.866 |
keeping | +0.863 |
Neg logit contributions | |
a | -0.146 |
| -0.134 |
big | -0.125 |
two | -0.116 |
little | -0.099 |
Token | the |
Token ID | 262 |
Feature activation | +2.57e-01 |
Pos logit contributions | |
avail | +0.280 |
rily | +0.274 |
keeping | +0.267 |
ge | +0.266 |
ittered | +0.264 |
Neg logit contributions | |
a | -0.030 |
big | -0.029 |
| -0.027 |
and | -0.026 |
two | -0.020 |
Token | house |
Token ID | 2156 |
Feature activation | +7.31e-02 |
Pos logit contributions | |
a | +0.637 |
| +0.613 |
and | +0.610 |
two | +0.582 |
big | +0.579 |
Neg logit contributions | |
avail | -0.616 |
keeping | -0.589 |
leness | -0.571 |
rily | -0.570 |
footing | -0.567 |
Token | on |
Token ID | 319 |
Feature activation | +3.29e-01 |
Pos logit contributions | |
a | +0.197 |
| +0.189 |
two | +0.180 |
big | +0.180 |
little | +0.165 |
Neg logit contributions | |
rily | -0.751 |
avail | -0.730 |
leness | -0.728 |
lights | -0.723 |
ittered | -0.712 |
Token | bones |
Token ID | 11945 |
Feature activation | +3.00e-01 |
Pos logit contributions | |
| +0.515 |
a | +0.507 |
and | +0.485 |
big | +0.480 |
two | +0.477 |
Neg logit contributions | |
avail | -1.086 |
rily | -1.067 |
leness | -1.042 |
lights | -1.029 |
keeping | -0.999 |
Token | . |
Token ID | 13 |
Feature activation | +4.86e-02 |
Pos logit contributions | |
a | +0.175 |
| +0.164 |
big | +0.152 |
two | +0.142 |
little | +0.122 |
Neg logit contributions | |
rily | -1.294 |
lights | -1.247 |
immer | -1.236 |
leness | -1.231 |
ittered | -1.230 |
Token | Sometimes |
Token ID | 8975 |
Feature activation | -6.13e-01 |
Pos logit contributions | |
a | +0.064 |
big | +0.057 |
and | +0.053 |
two | +0.053 |
little | +0.048 |
Neg logit contributions | |
lights | -0.179 |
avail | -0.178 |
rily | -0.171 |
ittered | -0.168 |
immer | -0.168 |
Token | he |
Token ID | 339 |
Feature activation | -3.90e-01 |
Pos logit contributions | |
avail | +2.404 |
lights | +2.272 |
leness | +2.246 |
rily | +2.210 |
fer | +2.180 |
Neg logit contributions | |
a | -0.682 |
big | -0.634 |
| -0.600 |
two | -0.573 |
and | -0.570 |
Token | che |
Token ID | 1125 |
Feature activation | -2.26e+00 |
Pos logit contributions | |
ittered | +1.226 |
avail | +1.200 |
inations | +1.188 |
rated | +1.187 |
rily | +1.186 |
Neg logit contributions | |
a | -0.679 |
big | -0.629 |
two | -0.628 |
| -0.627 |
and | -0.585 |
Token | wed |
Token ID | 19103 |
Feature activation | -1.18e-02 |
Pos logit contributions | |
avail | +2.768 |
ittered | +2.682 |
rily | +2.682 |
leness | +2.516 |
immer | +2.467 |
Neg logit contributions | |
a | -8.804 |
big | -8.504 |
| -8.460 |
and | -8.444 |
two | -8.273 |
Token | on |
Token ID | 319 |
Feature activation | +8.44e-02 |
Pos logit contributions | |
rily | +0.047 |
avail | +0.045 |
lights | +0.045 |
ittered | +0.045 |
leness | +0.044 |
Neg logit contributions | |
a | -0.011 |
| -0.011 |
big | -0.010 |
two | -0.009 |
little | -0.009 |
Token | Tom |
Token ID | 4186 |
Feature activation | -9.38e-02 |
Pos logit contributions | |
| +0.109 |
a | +0.104 |
big | +0.099 |
and | +0.097 |
two | +0.093 |
Neg logit contributions | |
avail | -0.306 |
rily | -0.301 |
leness | -0.291 |
lights | -0.289 |
ittered | -0.286 |
Token | and |
Token ID | 290 |
Feature activation | -1.90e-01 |
Pos logit contributions | |
ittered | +0.372 |
rily | +0.370 |
immer | +0.364 |
leness | +0.363 |
avail | +0.361 |
Neg logit contributions | |
a | -0.081 |
big | -0.077 |
| -0.071 |
two | -0.071 |
little | -0.058 |
Token | Jen |
Token ID | 13364 |
Feature activation | -1.15e-01 |
Pos logit contributions | |
avail | +0.600 |
lights | +0.571 |
rily | +0.560 |
keeping | +0.560 |
ittered | +0.557 |
Neg logit contributions | |
a | -0.345 |
| -0.335 |
big | -0.333 |
and | -0.329 |
two | -0.310 |
Token | leave |
Token ID | 2666 |
Feature activation | -3.41e-01 |
Pos logit contributions | |
avail | +1.675 |
ittered | +1.674 |
rily | +1.646 |
immer | +1.645 |
inations | +1.641 |
Neg logit contributions | |
a | -0.736 |
| -0.716 |
big | -0.686 |
and | -0.670 |
two | -0.667 |
Token | and |
Token ID | 290 |
Feature activation | -4.40e-01 |
Pos logit contributions | |
avail | +1.449 |
rily | +1.412 |
keeping | +1.398 |
ittered | +1.380 |
immer | +1.370 |
Neg logit contributions | |
big | -0.233 |
a | -0.230 |
| -0.213 |
two | -0.203 |
and | -0.188 |
Token | never |
Token ID | 1239 |
Feature activation | -3.18e-01 |
Pos logit contributions | |
avail | +1.464 |
rily | +1.392 |
keeping | +1.385 |
ittered | +1.366 |
inations | +1.365 |
Neg logit contributions | |
a | -0.714 |
| -0.708 |
big | -0.688 |
and | -0.678 |
two | -0.637 |
Token | return |
Token ID | 1441 |
Feature activation | -4.18e-01 |
Pos logit contributions | |
avail | +1.060 |
rated | +1.025 |
ittered | +1.012 |
immer | +1.007 |
ge | +0.994 |
Neg logit contributions | |
a | -0.541 |
| -0.493 |
big | -0.493 |
and | -0.489 |
two | -0.470 |
Token | . |
Token ID | 13 |
Feature activation | -2.79e-01 |
Pos logit contributions | |
avail | +1.734 |
keeping | +1.720 |
rily | +1.696 |
ittered | +1.658 |
lights | +1.653 |
Neg logit contributions | |
a | -0.318 |
big | -0.303 |
| -0.297 |
and | -0.263 |
two | -0.259 |
Token | Sc |
Token ID | 1446 |
Feature activation | -2.21e+00 |
Pos logit contributions | |
avail | +0.936 |
rily | +0.911 |
ently | +0.887 |
keeping | +0.886 |
lights | +0.883 |
Neg logit contributions | |
a | -0.444 |
big | -0.426 |
and | -0.394 |
two | -0.385 |
little | -0.362 |
Token | ared |
Token ID | 1144 |
Feature activation | -1.84e-01 |
Pos logit contributions | |
avail | +3.301 |
rily | +3.221 |
ittered | +3.209 |
immer | +3.050 |
leness | +3.048 |
Neg logit contributions | |
a | -8.025 |
big | -7.727 |
and | -7.640 |
| -7.620 |
two | -7.511 |
Token | , |
Token ID | 11 |
Feature activation | +1.78e-01 |
Pos logit contributions | |
rily | +0.722 |
avail | +0.710 |
ittered | +0.705 |
lights | +0.702 |
ently | +0.686 |
Neg logit contributions | |
a | -0.185 |
| -0.180 |
big | -0.160 |
and | -0.151 |
two | -0.149 |
Token | Mary |
Token ID | 5335 |
Feature activation | -5.77e-03 |
Pos logit contributions | |
a | +0.255 |
big | +0.248 |
| +0.241 |
two | +0.228 |
and | +0.228 |
Neg logit contributions | |
avail | -0.598 |
rily | -0.592 |
lights | -0.585 |
ittered | -0.576 |
footing | -0.568 |
Token | quickly |
Token ID | 2952 |
Feature activation | -2.41e-01 |
Pos logit contributions | |
ittered | +0.020 |
rily | +0.020 |
fer | +0.019 |
iously | +0.019 |
avail | +0.019 |
Neg logit contributions | |
a | -0.009 |
big | -0.008 |
| -0.008 |
two | -0.007 |
and | -0.007 |
Token | agreed |
Token ID | 4987 |
Feature activation | +5.26e-01 |
Pos logit contributions | |
ittered | +0.700 |
rily | +0.691 |
avail | +0.687 |
immer | +0.686 |
leness | +0.684 |
Neg logit contributions | |
a | -0.496 |
| -0.480 |
big | -0.462 |
two | -0.441 |
and | -0.435 |
Token | under |
Token ID | 739 |
Feature activation | +1.48e-01 |
Pos logit contributions | |
rily | +1.121 |
ittered | +1.090 |
avail | +1.090 |
lights | +1.069 |
immer | +1.058 |
Neg logit contributions | |
| -0.302 |
a | -0.294 |
big | -0.271 |
two | -0.263 |
and | -0.233 |
Token | the |
Token ID | 262 |
Feature activation | -2.76e-01 |
Pos logit contributions | |
| +0.113 |
a | +0.105 |
big | +0.090 |
two | +0.088 |
and | +0.086 |
Neg logit contributions | |
rily | -0.613 |
avail | -0.605 |
lights | -0.597 |
keeping | -0.587 |
ittered | -0.584 |
Token | table |
Token ID | 3084 |
Feature activation | -2.44e-01 |
Pos logit contributions | |
lights | +0.615 |
rily | +0.602 |
avail | +0.593 |
inations | +0.591 |
vation | +0.579 |
Neg logit contributions | |
a | -0.749 |
| -0.720 |
two | -0.693 |
and | -0.680 |
big | -0.664 |
Token | and |
Token ID | 290 |
Feature activation | -3.34e-01 |
Pos logit contributions | |
rily | +1.080 |
avail | +1.059 |
ittered | +1.049 |
lights | +1.044 |
immer | +1.041 |
Neg logit contributions | |
a | -0.124 |
| -0.106 |
big | -0.099 |
two | -0.079 |
little | -0.052 |
Token | sc |
Token ID | 629 |
Feature activation | -1.33e+00 |
Pos logit contributions | |
rily | +1.073 |
avail | +1.070 |
ittered | +1.064 |
keeping | +1.032 |
footing | +1.031 |
Neg logit contributions | |
a | -0.557 |
| -0.548 |
big | -0.535 |
and | -0.516 |
two | -0.507 |
Token | ur |
Token ID | 333 |
Feature activation | -2.50e+00 |
Pos logit contributions | |
ittered | +1.829 |
rily | +1.736 |
avail | +1.688 |
immer | +1.683 |
inations | +1.681 |
Neg logit contributions | |
a | -4.982 |
big | -4.783 |
and | -4.736 |
| -4.712 |
two | -4.670 |
Token | ried |
Token ID | 2228 |
Feature activation | +5.72e-01 |
Pos logit contributions | |
avail | +2.731 |
rily | +2.609 |
ittered | +2.543 |
leness | +2.376 |
immer | +2.344 |
Neg logit contributions | |
a | -10.207 |
big | -9.841 |
and | -9.821 |
| -9.813 |
two | -9.592 |
Token | behind |
Token ID | 2157 |
Feature activation | +3.33e-01 |
Pos logit contributions | |
a | +0.986 |
| +0.953 |
big | +0.953 |
two | +0.902 |
little | +0.886 |
Neg logit contributions | |
rily | -1.823 |
ittered | -1.755 |
avail | -1.754 |
keeping | -1.752 |
footing | -1.698 |
Token | the |
Token ID | 262 |
Feature activation | -2.08e-01 |
Pos logit contributions | |
big | +0.276 |
a | +0.269 |
| +0.264 |
and | +0.249 |
two | +0.242 |
Neg logit contributions | |
avail | -1.316 |
rily | -1.310 |
ittered | -1.274 |
keeping | -1.267 |
lights | -1.248 |
Token | chair |
Token ID | 5118 |
Feature activation | -2.49e-01 |
Pos logit contributions | |
rily | +0.396 |
lights | +0.389 |
avail | +0.388 |
keeping | +0.383 |
inations | +0.373 |
Neg logit contributions | |
a | -0.621 |
| -0.596 |
and | -0.576 |
two | -0.574 |
big | -0.565 |
Token | , |
Token ID | 11 |
Feature activation | -8.49e-02 |
Pos logit contributions | |
rily | +1.071 |
ittered | +1.059 |
avail | +1.058 |
keeping | +1.054 |
lights | +1.051 |
Neg logit contributions | |
a | -0.153 |
big | -0.133 |
| -0.115 |
two | -0.107 |
little | -0.086 |