Token | 2012 |
Token ID | 2321 |
Feature activation | +9.54e-02 |
Pos logit contributions | |
yip | +0.110 |
zee | +0.096 |
zer | +0.082 |
aunders | +0.082 |
zh | +0.080 |
Neg logit contributions | |
cout | -0.067 |
naires | -0.065 |
linger | -0.061 |
BuzzFeed | -0.061 |
swe | -0.058 |
Token | United |
Token ID | 1578 |
Feature activation | +6.61e-01 |
Pos logit contributions | |
yip | +0.321 |
zee | +0.280 |
aunders | +0.241 |
zer | +0.239 |
zh | +0.235 |
Neg logit contributions | |
cout | -0.170 |
naires | -0.163 |
BuzzFeed | -0.154 |
linger | -0.154 |
swe | -0.146 |
Token | States |
Token ID | 1829 |
Feature activation | +5.26e-01 |
Pos logit contributions | |
yip | +2.043 |
zee | +1.749 |
aunders | +1.474 |
zer | +1.467 |
zh | +1.428 |
Neg logit contributions | |
cout | -1.370 |
naires | -1.319 |
BuzzFeed | -1.278 |
linger | -1.254 |
swe | -1.205 |
Token | 3 |
Token ID | 513 |
Feature activation | -2.13e-01 |
Pos logit contributions | |
yip | +1.438 |
zee | +1.220 |
aunders | +1.003 |
zer | +0.997 |
zh | +0.965 |
Neg logit contributions | |
cout | -1.254 |
naires | -1.209 |
linger | -1.166 |
BuzzFeed | -1.161 |
swe | -1.118 |
Token | 240 |
Token ID | 16102 |
Feature activation | +9.09e-02 |
Pos logit contributions | |
cout | +0.584 |
naires | +0.567 |
BuzzFeed | +0.554 |
linger | +0.545 |
swe | +0.530 |
Neg logit contributions | |
yip | -0.518 |
zee | -0.421 |
aunders | -0.334 |
zer | -0.329 |
zh | -0.318 |
Token | Posts |
Token ID | 12043 |
Feature activation | +2.03e+00 |
Pos logit contributions | |
yip | +0.176 |
zee | +0.139 |
aunders | +0.102 |
zer | +0.101 |
zh | +0.094 |
Neg logit contributions | |
cout | -0.288 |
naires | -0.282 |
BuzzFeed | -0.273 |
linger | -0.273 |
swe | -0.264 |
Token | # |
Token ID | 1303 |
Feature activation | -1.24e-01 |
Pos logit contributions | |
yip | +6.099 |
zee | +5.148 |
aunders | +4.389 |
zer | +4.338 |
zh | +4.227 |
Neg logit contributions | |
cout | -4.218 |
naires | -4.109 |
BuzzFeed | -3.897 |
linger | -3.868 |
swe | -3.670 |
Token | 10 |
Token ID | 940 |
Feature activation | +2.59e-02 |
Pos logit contributions | |
cout | +0.315 |
naires | +0.305 |
BuzzFeed | +0.298 |
linger | +0.292 |
swe | +0.283 |
Neg logit contributions | |
yip | -0.327 |
zee | -0.269 |
aunders | -0.219 |
zer | -0.215 |
zh | -0.209 |
Token | Happy |
Token ID | 14628 |
Feature activation | +7.18e-01 |
Pos logit contributions | |
yip | +0.077 |
zee | +0.066 |
aunders | +0.055 |
zer | +0.054 |
zh | +0.053 |
Neg logit contributions | |
cout | -0.057 |
naires | -0.054 |
BuzzFeed | -0.053 |
linger | -0.052 |
swe | -0.050 |
Token | it |
Token ID | 340 |
Feature activation | +2.44e-01 |
Pos logit contributions | |
yip | +2.191 |
zee | +1.868 |
aunders | +1.573 |
zer | +1.562 |
zh | +1.523 |
Neg logit contributions | |
cout | -1.519 |
naires | -1.466 |
BuzzFeed | -1.413 |
linger | -1.388 |
swe | -1.333 |
Token | worked |
Token ID | 3111 |
Feature activation | -1.48e-01 |
Pos logit contributions | |
yip | +0.794 |
zee | +0.682 |
aunders | +0.584 |
zer | +0.576 |
zh | +0.563 |
Neg logit contributions | |
cout | -0.469 |
naires | -0.449 |
BuzzFeed | -0.435 |
linger | -0.425 |
swe | -0.408 |
Token | 2002 |
Token ID | 6244 |
Feature activation | -6.65e-02 |
Pos logit contributions | |
cout | +0.192 |
naires | +0.186 |
linger | +0.175 |
BuzzFeed | +0.174 |
swe | +0.166 |
Neg logit contributions | |
yip | -0.312 |
zee | -0.274 |
aunders | -0.232 |
zer | -0.232 |
zh | -0.226 |
Token | United |
Token ID | 1578 |
Feature activation | +4.90e-01 |
Pos logit contributions | |
cout | +0.117 |
naires | +0.112 |
BuzzFeed | +0.106 |
linger | +0.105 |
swe | +0.100 |
Neg logit contributions | |
yip | -0.226 |
zee | -0.197 |
aunders | -0.170 |
zer | -0.169 |
zh | -0.165 |
Token | States |
Token ID | 1829 |
Feature activation | +4.26e-01 |
Pos logit contributions | |
yip | +1.558 |
zee | +1.334 |
aunders | +1.136 |
zer | +1.119 |
zh | +1.097 |
Neg logit contributions | |
cout | -0.985 |
naires | -0.946 |
BuzzFeed | -0.913 |
linger | -0.895 |
swe | -0.859 |
Token | 3 |
Token ID | 513 |
Feature activation | -1.92e-01 |
Pos logit contributions | |
yip | +1.173 |
zee | +0.991 |
aunders | +0.816 |
zer | +0.813 |
zh | +0.786 |
Neg logit contributions | |
cout | -1.009 |
naires | -0.973 |
BuzzFeed | -0.938 |
linger | -0.937 |
swe | -0.899 |
Token | 472 |
Token ID | 37856 |
Feature activation | -1.21e-01 |
Pos logit contributions | |
cout | +0.461 |
naires | +0.446 |
BuzzFeed | +0.434 |
linger | +0.426 |
swe | +0.412 |
Neg logit contributions | |
yip | -0.532 |
zee | -0.444 |
aunders | -0.366 |
zer | -0.361 |
zh | -0.352 |
Token | Posts |
Token ID | 12043 |
Feature activation | +1.88e+00 |
Pos logit contributions | |
cout | +0.371 |
naires | +0.361 |
BuzzFeed | +0.351 |
linger | +0.349 |
swe | +0.340 |
Neg logit contributions | |
yip | -0.250 |
zee | -0.199 |
aunders | -0.149 |
zer | -0.148 |
zh | -0.141 |
Token | # |
Token ID | 1303 |
Feature activation | -2.57e-01 |
Pos logit contributions | |
yip | +5.709 |
zee | +4.835 |
aunders | +4.100 |
zer | +4.071 |
zh | +3.960 |
Neg logit contributions | |
cout | -3.898 |
naires | -3.784 |
BuzzFeed | -3.609 |
linger | -3.571 |
swe | -3.393 |
Token | 11 |
Token ID | 1157 |
Feature activation | -1.37e-01 |
Pos logit contributions | |
cout | +0.651 |
naires | +0.629 |
linger | +0.610 |
BuzzFeed | +0.609 |
swe | +0.588 |
Neg logit contributions | |
yip | -0.669 |
zee | -0.564 |
aunders | -0.457 |
zer | -0.454 |
zh | -0.443 |
Token | On |
Token ID | 1550 |
Feature activation | +6.49e-02 |
Pos logit contributions | |
cout | +0.301 |
naires | +0.290 |
BuzzFeed | +0.282 |
linger | +0.276 |
swe | +0.268 |
Neg logit contributions | |
yip | -0.409 |
zee | -0.346 |
aunders | -0.290 |
zer | -0.286 |
zh | -0.280 |
Token | September |
Token ID | 2693 |
Feature activation | -2.88e-01 |
Pos logit contributions | |
yip | +0.195 |
zee | +0.165 |
aunders | +0.139 |
zer | +0.137 |
zh | +0.134 |
Neg logit contributions | |
cout | -0.141 |
naires | -0.136 |
BuzzFeed | -0.132 |
linger | -0.130 |
swe | -0.125 |
Token | 20 |
Token ID | 1160 |
Feature activation | -1.78e-01 |
Pos logit contributions | |
cout | +0.661 |
naires | +0.638 |
BuzzFeed | +0.615 |
linger | +0.610 |
swe | +0.587 |
Neg logit contributions | |
yip | -0.826 |
zee | -0.700 |
aunders | -0.580 |
zer | -0.578 |
zh | -0.563 |
Token | 2010 |
Token ID | 3050 |
Feature activation | -1.53e-01 |
Pos logit contributions | |
cout | +0.034 |
naires | +0.033 |
BuzzFeed | +0.031 |
linger | +0.031 |
swe | +0.030 |
Neg logit contributions | |
yip | -0.056 |
zee | -0.049 |
zer | -0.041 |
aunders | -0.041 |
zh | -0.040 |
Token | United |
Token ID | 1578 |
Feature activation | +5.70e-01 |
Pos logit contributions | |
cout | +0.270 |
naires | +0.258 |
BuzzFeed | +0.246 |
linger | +0.242 |
swe | +0.230 |
Neg logit contributions | |
yip | -0.521 |
zee | -0.454 |
aunders | -0.390 |
zer | -0.388 |
zh | -0.380 |
Token | States |
Token ID | 1829 |
Feature activation | +4.49e-01 |
Pos logit contributions | |
yip | +1.800 |
zee | +1.540 |
aunders | +1.309 |
zer | +1.292 |
zh | +1.264 |
Neg logit contributions | |
cout | -1.157 |
naires | -1.112 |
BuzzFeed | -1.074 |
linger | -1.052 |
swe | -1.011 |
Token | 10 |
Token ID | 838 |
Feature activation | -9.87e-02 |
Pos logit contributions | |
yip | +1.235 |
zee | +1.045 |
aunders | +0.860 |
zer | +0.856 |
zh | +0.829 |
Neg logit contributions | |
cout | -1.066 |
naires | -1.027 |
BuzzFeed | -0.990 |
linger | -0.990 |
swe | -0.949 |
Token | 53 |
Token ID | 4310 |
Feature activation | +2.88e-01 |
Pos logit contributions | |
cout | +0.249 |
naires | +0.241 |
BuzzFeed | +0.235 |
linger | +0.231 |
swe | +0.224 |
Neg logit contributions | |
yip | -0.262 |
zee | -0.217 |
aunders | -0.177 |
zer | -0.175 |
zh | -0.170 |
Token | Posts |
Token ID | 12043 |
Feature activation | +1.85e+00 |
Pos logit contributions | |
yip | +0.580 |
zee | +0.459 |
aunders | +0.345 |
zer | +0.342 |
zh | +0.323 |
Neg logit contributions | |
cout | -0.892 |
naires | -0.869 |
linger | -0.845 |
BuzzFeed | -0.844 |
swe | -0.822 |
Token | # |
Token ID | 1303 |
Feature activation | -2.09e-01 |
Pos logit contributions | |
yip | +5.733 |
zee | +4.873 |
aunders | +4.144 |
zer | +4.108 |
zh | +4.008 |
Neg logit contributions | |
cout | -3.779 |
naires | -3.652 |
BuzzFeed | -3.494 |
linger | -3.457 |
swe | -3.293 |
Token | 15 |
Token ID | 1314 |
Feature activation | +5.97e-02 |
Pos logit contributions | |
cout | +0.551 |
naires | +0.533 |
BuzzFeed | +0.520 |
linger | +0.512 |
swe | +0.497 |
Neg logit contributions | |
yip | -0.533 |
zee | -0.440 |
aunders | -0.354 |
zer | -0.349 |
zh | -0.339 |
Token | honestly |
Token ID | 12698 |
Feature activation | -5.18e-01 |
Pos logit contributions | |
yip | +0.178 |
zee | +0.151 |
aunders | +0.127 |
zer | +0.125 |
zh | +0.123 |
Neg logit contributions | |
cout | -0.131 |
naires | -0.126 |
BuzzFeed | -0.122 |
linger | -0.120 |
swe | -0.116 |
Token | if |
Token ID | 611 |
Feature activation | -1.23e-01 |
Pos logit contributions | |
cout | +1.021 |
naires | +0.971 |
BuzzFeed | +0.945 |
linger | +0.926 |
swe | +0.895 |
Neg logit contributions | |
yip | -1.664 |
zee | -1.423 |
aunders | -1.216 |
zer | -1.194 |
zh | -1.173 |
Token | we |
Token ID | 356 |
Feature activation | -3.72e-01 |
Pos logit contributions | |
cout | +0.240 |
naires | +0.230 |
BuzzFeed | +0.222 |
linger | +0.218 |
swe | +0.209 |
Neg logit contributions | |
yip | -0.396 |
zee | -0.341 |
aunders | -0.290 |
zer | -0.288 |
zh | -0.281 |
Token | 2010 |
Token ID | 3050 |
Feature activation | +8.09e-02 |
Pos logit contributions | |
cout | +0.261 |
naires | +0.251 |
linger | +0.238 |
BuzzFeed | +0.237 |
swe | +0.226 |
Neg logit contributions | |
yip | -0.431 |
zee | -0.380 |
zer | -0.324 |
aunders | -0.322 |
zh | -0.316 |
Token | United |
Token ID | 1578 |
Feature activation | +7.13e-01 |
Pos logit contributions | |
yip | +0.283 |
zee | +0.247 |
aunders | +0.214 |
zer | +0.212 |
zh | +0.208 |
Neg logit contributions | |
cout | -0.135 |
naires | -0.129 |
BuzzFeed | -0.123 |
linger | -0.121 |
swe | -0.114 |
Token | States |
Token ID | 1829 |
Feature activation | +5.46e-01 |
Pos logit contributions | |
yip | +2.236 |
zee | +1.912 |
aunders | +1.621 |
zer | +1.604 |
zh | +1.568 |
Neg logit contributions | |
cout | -1.457 |
naires | -1.400 |
BuzzFeed | -1.354 |
linger | -1.327 |
swe | -1.276 |
Token | 21 |
Token ID | 2310 |
Feature activation | -1.82e-01 |
Pos logit contributions | |
yip | +1.492 |
zee | +1.260 |
aunders | +1.037 |
zer | +1.034 |
zh | +1.000 |
Neg logit contributions | |
cout | -1.300 |
naires | -1.255 |
linger | -1.209 |
BuzzFeed | -1.208 |
swe | -1.160 |
Token | 50 |
Token ID | 1120 |
Feature activation | +6.44e-02 |
Pos logit contributions | |
cout | +0.473 |
naires | +0.459 |
BuzzFeed | +0.447 |
linger | +0.441 |
swe | +0.427 |
Neg logit contributions | |
yip | -0.469 |
zee | -0.387 |
aunders | -0.313 |
zer | -0.309 |
zh | -0.300 |
Token | Posts |
Token ID | 12043 |
Feature activation | +1.85e+00 |
Pos logit contributions | |
yip | +0.130 |
zee | +0.104 |
zer | +0.077 |
aunders | +0.077 |
zh | +0.073 |
Neg logit contributions | |
cout | -0.200 |
naires | -0.194 |
linger | -0.189 |
BuzzFeed | -0.189 |
swe | -0.184 |
Token | # |
Token ID | 1303 |
Feature activation | -1.19e-01 |
Pos logit contributions | |
yip | +5.687 |
zee | +4.834 |
aunders | +4.109 |
zer | +4.075 |
zh | +3.969 |
Neg logit contributions | |
cout | -3.782 |
naires | -3.662 |
BuzzFeed | -3.506 |
linger | -3.466 |
swe | -3.301 |
Token | 16 |
Token ID | 1433 |
Feature activation | -1.59e-02 |
Pos logit contributions | |
cout | +0.298 |
naires | +0.286 |
linger | +0.277 |
BuzzFeed | +0.277 |
swe | +0.267 |
Neg logit contributions | |
yip | -0.316 |
zee | -0.267 |
aunders | -0.217 |
zer | -0.215 |
zh | -0.211 |
Token | Glad |
Token ID | 28442 |
Feature activation | +3.65e-01 |
Pos logit contributions | |
cout | +0.035 |
naires | +0.033 |
BuzzFeed | +0.032 |
linger | +0.032 |
swe | +0.031 |
Neg logit contributions | |
yip | -0.048 |
zee | -0.041 |
aunders | -0.034 |
zer | -0.034 |
zh | -0.033 |
Token | to |
Token ID | 284 |
Feature activation | -1.03e-01 |
Pos logit contributions | |
yip | +1.171 |
zee | +1.007 |
aunders | +0.858 |
zer | +0.852 |
zh | +0.831 |
Neg logit contributions | |
cout | -0.717 |
naires | -0.689 |
BuzzFeed | -0.663 |
linger | -0.651 |
swe | -0.623 |
Token | hear |
Token ID | 3285 |
Feature activation | -1.75e-01 |
Pos logit contributions | |
cout | +0.246 |
naires | +0.238 |
BuzzFeed | +0.231 |
linger | +0.227 |
swe | +0.220 |
Neg logit contributions | |
yip | -0.289 |
zee | -0.242 |
aunders | -0.200 |
zer | -0.197 |
zh | -0.192 |
Token | 2012 |
Token ID | 2321 |
Feature activation | +8.82e-02 |
Pos logit contributions | |
cout | +0.115 |
naires | +0.112 |
linger | +0.105 |
BuzzFeed | +0.105 |
swe | +0.099 |
Neg logit contributions | |
yip | -0.185 |
zee | -0.162 |
zer | -0.138 |
aunders | -0.137 |
zh | -0.135 |
Token | United |
Token ID | 1578 |
Feature activation | +6.85e-01 |
Pos logit contributions | |
yip | +0.300 |
zee | +0.263 |
aunders | +0.226 |
zer | +0.225 |
zh | +0.221 |
Neg logit contributions | |
cout | -0.153 |
naires | -0.147 |
linger | -0.138 |
BuzzFeed | -0.138 |
swe | -0.130 |
Token | States |
Token ID | 1829 |
Feature activation | +4.85e-01 |
Pos logit contributions | |
yip | +2.156 |
zee | +1.845 |
aunders | +1.565 |
zer | +1.548 |
zh | +1.513 |
Neg logit contributions | |
cout | -1.394 |
naires | -1.339 |
BuzzFeed | -1.295 |
linger | -1.269 |
swe | -1.220 |
Token | 5 |
Token ID | 642 |
Feature activation | -1.91e-01 |
Pos logit contributions | |
yip | +1.329 |
zee | +1.125 |
aunders | +0.925 |
zer | +0.922 |
zh | +0.891 |
Neg logit contributions | |
cout | -1.152 |
naires | -1.110 |
linger | -1.070 |
BuzzFeed | -1.068 |
swe | -1.027 |
Token | 667 |
Token ID | 28933 |
Feature activation | -3.28e-01 |
Pos logit contributions | |
cout | +0.473 |
naires | +0.457 |
BuzzFeed | +0.445 |
linger | +0.438 |
swe | +0.424 |
Neg logit contributions | |
yip | -0.515 |
zee | -0.428 |
aunders | -0.350 |
zer | -0.345 |
zh | -0.336 |
Token | Posts |
Token ID | 12043 |
Feature activation | +1.84e+00 |
Pos logit contributions | |
cout | +0.995 |
naires | +0.969 |
BuzzFeed | +0.944 |
linger | +0.937 |
swe | +0.913 |
Neg logit contributions | |
yip | -0.695 |
zee | -0.552 |
aunders | -0.418 |
zer | -0.413 |
zh | -0.394 |
Token | # |
Token ID | 1303 |
Feature activation | -3.04e-01 |
Pos logit contributions | |
yip | +5.665 |
zee | +4.798 |
aunders | +4.078 |
zer | +4.040 |
zh | +3.935 |
Neg logit contributions | |
cout | -3.785 |
naires | -3.668 |
BuzzFeed | -3.506 |
linger | -3.471 |
swe | -3.302 |
Token | 12 |
Token ID | 1065 |
Feature activation | +1.45e-01 |
Pos logit contributions | |
cout | +0.803 |
naires | +0.781 |
BuzzFeed | +0.762 |
linger | +0.746 |
swe | +0.727 |
Neg logit contributions | |
yip | -0.771 |
zee | -0.628 |
aunders | -0.506 |
zer | -0.493 |
zh | -0.479 |
Token | I |
Token ID | 314 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
yip | +0.430 |
zee | +0.365 |
aunders | +0.306 |
zer | +0.302 |
zh | +0.295 |
Neg logit contributions | |
cout | -0.319 |
naires | -0.308 |
BuzzFeed | -0.298 |
linger | -0.293 |
swe | -0.283 |
Token | 'm |
Token ID | 1101 |
Feature activation | -1.42e-01 |
Pos logit contributions | |
yip | +0.341 |
zee | +0.292 |
aunders | +0.245 |
zer | +0.243 |
zh | +0.239 |
Neg logit contributions | |
cout | -0.242 |
naires | -0.234 |
BuzzFeed | -0.224 |
linger | -0.220 |
swe | -0.210 |
Token | glad |
Token ID | 9675 |
Feature activation | +6.87e-01 |
Pos logit contributions | |
cout | +0.316 |
naires | +0.305 |
BuzzFeed | +0.296 |
linger | +0.291 |
swe | +0.280 |
Neg logit contributions | |
yip | -0.417 |
zee | -0.353 |
aunders | -0.295 |
zer | -0.292 |
zh | -0.285 |
Token | June |
Token ID | 2795 |
Feature activation | +3.35e-02 |
Pos logit contributions | |
yip | +3.489 |
zee | +3.009 |
zer | +2.480 |
zh | +2.445 |
aunders | +2.411 |
Neg logit contributions | |
cout | -3.229 |
naires | -3.177 |
linger | -3.007 |
BuzzFeed | -2.969 |
omorphic | -2.903 |
Token | 2012 |
Token ID | 2321 |
Feature activation | +4.55e-02 |
Pos logit contributions | |
yip | +0.104 |
zee | +0.091 |
zer | +0.077 |
aunders | +0.077 |
zh | +0.076 |
Neg logit contributions | |
cout | -0.067 |
naires | -0.064 |
linger | -0.061 |
BuzzFeed | -0.061 |
swe | -0.058 |
Token | Canada |
Token ID | 3340 |
Feature activation | +6.02e-02 |
Pos logit contributions | |
yip | +0.155 |
zee | +0.135 |
aunders | +0.117 |
zer | +0.116 |
zh | +0.114 |
Neg logit contributions | |
cout | -0.079 |
naires | -0.076 |
BuzzFeed | -0.071 |
linger | -0.071 |
swe | -0.067 |
Token | 16 |
Token ID | 1467 |
Feature activation | -1.78e-02 |
Pos logit contributions | |
yip | +0.165 |
zee | +0.140 |
aunders | +0.114 |
zer | +0.114 |
zh | +0.111 |
Neg logit contributions | |
cout | -0.144 |
naires | -0.138 |
linger | -0.133 |
BuzzFeed | -0.133 |
swe | -0.128 |
Token | 10 |
Token ID | 940 |
Feature activation | +3.10e-02 |
Pos logit contributions | |
cout | +0.045 |
naires | +0.043 |
BuzzFeed | +0.042 |
linger | +0.042 |
swe | +0.040 |
Neg logit contributions | |
yip | -0.047 |
zee | -0.039 |
aunders | -0.032 |
zer | -0.032 |
zh | -0.031 |
Token | Posts |
Token ID | 12043 |
Feature activation | +1.83e+00 |
Pos logit contributions | |
yip | +0.063 |
zee | +0.050 |
aunders | +0.038 |
zer | +0.037 |
zh | +0.035 |
Neg logit contributions | |
cout | -0.096 |
naires | -0.093 |
BuzzFeed | -0.090 |
linger | -0.090 |
swe | -0.088 |
Token | # |
Token ID | 1303 |
Feature activation | -2.91e-01 |
Pos logit contributions | |
yip | +5.631 |
zee | +4.769 |
aunders | +4.058 |
zer | +4.024 |
zh | +3.922 |
Neg logit contributions | |
cout | -3.741 |
naires | -3.619 |
BuzzFeed | -3.465 |
linger | -3.429 |
swe | -3.265 |
Token | 14 |
Token ID | 1415 |
Feature activation | +9.07e-02 |
Pos logit contributions | |
cout | +0.771 |
naires | +0.750 |
BuzzFeed | +0.731 |
linger | +0.719 |
swe | +0.698 |
Neg logit contributions | |
yip | -0.736 |
zee | -0.603 |
aunders | -0.484 |
zer | -0.475 |
zh | -0.461 |
Token | On |
Token ID | 1550 |
Feature activation | -2.10e-02 |
Pos logit contributions | |
yip | +0.270 |
zee | +0.229 |
aunders | +0.192 |
zer | +0.189 |
zh | +0.185 |
Neg logit contributions | |
cout | -0.200 |
naires | -0.193 |
BuzzFeed | -0.187 |
linger | -0.183 |
swe | -0.177 |
Token | September |
Token ID | 2693 |
Feature activation | -2.47e-01 |
Pos logit contributions | |
cout | +0.043 |
naires | +0.042 |
BuzzFeed | +0.040 |
linger | +0.040 |
swe | +0.037 |
Neg logit contributions | |
yip | -0.065 |
zee | -0.056 |
aunders | -0.047 |
zer | -0.047 |
zh | -0.046 |
Token | 20 |
Token ID | 1160 |
Feature activation | -2.35e-01 |
Pos logit contributions | |
cout | +0.503 |
naires | +0.484 |
BuzzFeed | +0.475 |
linger | +0.458 |
swe | +0.446 |
Neg logit contributions | |
yip | -0.781 |
zee | -0.663 |
aunders | -0.568 |
zer | -0.552 |
zh | -0.542 |
Token | 2011 |
Token ID | 2813 |
Feature activation | -1.63e-02 |
Pos logit contributions | |
cout | +0.040 |
naires | +0.039 |
linger | +0.037 |
BuzzFeed | +0.037 |
swe | +0.035 |
Neg logit contributions | |
yip | -0.066 |
zee | -0.058 |
zer | -0.049 |
aunders | -0.049 |
zh | -0.048 |
Token | United |
Token ID | 1578 |
Feature activation | +5.54e-01 |
Pos logit contributions | |
cout | +0.028 |
naires | +0.027 |
BuzzFeed | +0.025 |
linger | +0.025 |
swe | +0.024 |
Neg logit contributions | |
yip | -0.056 |
zee | -0.049 |
aunders | -0.042 |
zer | -0.042 |
zh | -0.041 |
Token | States |
Token ID | 1829 |
Feature activation | +4.24e-01 |
Pos logit contributions | |
yip | +1.752 |
zee | +1.499 |
aunders | +1.273 |
zer | +1.254 |
zh | +1.228 |
Neg logit contributions | |
cout | -1.122 |
naires | -1.078 |
BuzzFeed | -1.041 |
linger | -1.020 |
swe | -0.980 |
Token | 6 |
Token ID | 718 |
Feature activation | -1.81e-01 |
Pos logit contributions | |
yip | +1.161 |
zee | +0.982 |
aunders | +0.808 |
zer | +0.806 |
zh | +0.780 |
Neg logit contributions | |
cout | -1.007 |
naires | -0.971 |
linger | -0.936 |
BuzzFeed | -0.935 |
swe | -0.899 |
Token | 12 |
Token ID | 1065 |
Feature activation | +2.63e-01 |
Pos logit contributions | |
cout | +0.449 |
naires | +0.434 |
BuzzFeed | +0.422 |
linger | +0.416 |
swe | +0.403 |
Neg logit contributions | |
yip | -0.487 |
zee | -0.406 |
aunders | -0.332 |
zer | -0.328 |
zh | -0.318 |
Token | Posts |
Token ID | 12043 |
Feature activation | +1.77e+00 |
Pos logit contributions | |
yip | +0.548 |
zee | +0.439 |
aunders | +0.332 |
zer | +0.328 |
zh | +0.314 |
Neg logit contributions | |
cout | -0.797 |
naires | -0.777 |
BuzzFeed | -0.757 |
linger | -0.754 |
swe | -0.734 |
Token | # |
Token ID | 1303 |
Feature activation | -1.73e-01 |
Pos logit contributions | |
yip | +5.473 |
zee | +4.648 |
aunders | +3.953 |
zer | +3.918 |
zh | +3.820 |
Neg logit contributions | |
cout | -3.627 |
naires | -3.505 |
BuzzFeed | -3.363 |
linger | -3.322 |
swe | -3.168 |
Token | 17 |
Token ID | 1558 |
Feature activation | -8.99e-02 |
Pos logit contributions | |
cout | +0.448 |
naires | +0.427 |
BuzzFeed | +0.417 |
linger | +0.416 |
swe | +0.402 |
Neg logit contributions | |
yip | -0.443 |
zee | -0.374 |
aunders | -0.299 |
zer | -0.298 |
zh | -0.293 |
Token | Good |
Token ID | 4599 |
Feature activation | +1.57e-02 |
Pos logit contributions | |
cout | +0.198 |
naires | +0.191 |
BuzzFeed | +0.185 |
linger | +0.182 |
swe | +0.176 |
Neg logit contributions | |
yip | -0.268 |
zee | -0.227 |
aunders | -0.190 |
zer | -0.188 |
zh | -0.183 |
Token | to |
Token ID | 284 |
Feature activation | -8.55e-04 |
Pos logit contributions | |
yip | +0.049 |
zee | +0.042 |
aunders | +0.035 |
zer | +0.035 |
zh | +0.034 |
Neg logit contributions | |
cout | -0.033 |
naires | -0.031 |
BuzzFeed | -0.030 |
linger | -0.030 |
swe | -0.029 |
Token | hear |
Token ID | 3285 |
Feature activation | -1.43e-01 |
Pos logit contributions | |
cout | +0.002 |
naires | +0.002 |
BuzzFeed | +0.002 |
linger | +0.002 |
swe | +0.002 |
Neg logit contributions | |
yip | -0.002 |
zee | -0.002 |
aunders | -0.002 |
zer | -0.002 |
zh | -0.002 |
Token | TL |
Token ID | 24811 |
Feature activation | +8.52e-02 |
Pos logit contributions | |
cout | +0.088 |
naires | +0.084 |
BuzzFeed | +0.081 |
linger | +0.079 |
swe | +0.075 |
Neg logit contributions | |
yip | -0.190 |
zee | -0.166 |
aunders | -0.144 |
zer | -0.143 |
zh | -0.140 |
Token | AD |
Token ID | 2885 |
Feature activation | -1.72e-01 |
Pos logit contributions | |
yip | +0.239 |
zee | +0.200 |
aunders | +0.166 |
zer | +0.164 |
zh | +0.160 |
Neg logit contributions | |
cout | -0.200 |
naires | -0.195 |
BuzzFeed | -0.187 |
linger | -0.185 |
swe | -0.179 |
Token | T |
Token ID | 51 |
Feature activation | +6.13e-02 |
Pos logit contributions | |
cout | +0.421 |
naires | +0.407 |
BuzzFeed | +0.397 |
linger | +0.390 |
swe | +0.378 |
Neg logit contributions | |
yip | -0.470 |
zee | -0.392 |
aunders | -0.322 |
zer | -0.317 |
zh | -0.308 |
Token | 235 |
Token ID | 28878 |
Feature activation | -1.08e-01 |
Pos logit contributions | |
yip | +0.177 |
zee | +0.149 |
aunders | +0.124 |
zer | +0.123 |
zh | +0.120 |
Neg logit contributions | |
cout | -0.140 |
naires | -0.135 |
BuzzFeed | -0.131 |
linger | -0.129 |
swe | -0.124 |
Token | 21 |
Token ID | 2481 |
Feature activation | +4.44e-02 |
Pos logit contributions | |
cout | +0.275 |
naires | +0.266 |
BuzzFeed | +0.259 |
linger | +0.255 |
swe | +0.247 |
Neg logit contributions | |
yip | -0.285 |
zee | -0.236 |
aunders | -0.192 |
zer | -0.190 |
zh | -0.184 |
Token | Posts |
Token ID | 12043 |
Feature activation | +1.74e+00 |
Pos logit contributions | |
yip | +0.091 |
zee | +0.073 |
aunders | +0.055 |
zer | +0.054 |
zh | +0.052 |
Neg logit contributions | |
cout | -0.136 |
naires | -0.133 |
BuzzFeed | -0.129 |
linger | -0.129 |
swe | -0.125 |
Token | # |
Token ID | 1303 |
Feature activation | -1.77e-01 |
Pos logit contributions | |
yip | +5.386 |
zee | +4.584 |
aunders | +3.897 |
zer | +3.857 |
zh | +3.765 |
Neg logit contributions | |
cout | -3.557 |
naires | -3.434 |
BuzzFeed | -3.295 |
linger | -3.252 |
swe | -3.106 |
Token | 18 |
Token ID | 1507 |
Feature activation | -2.81e-01 |
Pos logit contributions | |
cout | +0.466 |
naires | +0.450 |
BuzzFeed | +0.439 |
linger | +0.434 |
swe | +0.420 |
Neg logit contributions | |
yip | -0.447 |
zee | -0.370 |
aunders | -0.297 |
zer | -0.294 |
zh | -0.285 |
Token | Good |
Token ID | 4599 |
Feature activation | +5.12e-02 |
Pos logit contributions | |
cout | +0.621 |
naires | +0.595 |
BuzzFeed | +0.581 |
linger | +0.568 |
swe | +0.551 |
Neg logit contributions | |
yip | -0.837 |
zee | -0.707 |
aunders | -0.594 |
zer | -0.585 |
zh | -0.572 |
Token | stuff |
Token ID | 3404 |
Feature activation | +2.56e-01 |
Pos logit contributions | |
yip | +0.157 |
zee | +0.133 |
aunders | +0.112 |
zer | +0.111 |
zh | +0.108 |
Neg logit contributions | |
cout | -0.109 |
naires | -0.105 |
BuzzFeed | -0.102 |
linger | -0.100 |
swe | -0.096 |
Token | ! |
Token ID | 0 |
Feature activation | -7.01e-02 |
Pos logit contributions | |
yip | +0.793 |
zee | +0.676 |
aunders | +0.571 |
zer | +0.565 |
zh | +0.552 |
Neg logit contributions | |
cout | -0.536 |
naires | -0.515 |
BuzzFeed | -0.499 |
linger | -0.489 |
swe | -0.472 |
Token | 2011 |
Token ID | 2813 |
Feature activation | -9.55e-03 |
Pos logit contributions | |
cout | +0.118 |
naires | +0.114 |
linger | +0.108 |
BuzzFeed | +0.107 |
swe | +0.102 |
Neg logit contributions | |
yip | -0.189 |
zee | -0.166 |
zer | -0.142 |
aunders | -0.141 |
zh | -0.139 |
Token | United |
Token ID | 1578 |
Feature activation | +5.99e-01 |
Pos logit contributions | |
cout | +0.017 |
naires | +0.016 |
BuzzFeed | +0.015 |
linger | +0.015 |
swe | +0.014 |
Neg logit contributions | |
yip | -0.033 |
zee | -0.028 |
aunders | -0.025 |
zer | -0.024 |
zh | -0.024 |
Token | States |
Token ID | 1829 |
Feature activation | +4.53e-01 |
Pos logit contributions | |
yip | +1.883 |
zee | +1.611 |
aunders | +1.367 |
zer | +1.350 |
zh | +1.321 |
Neg logit contributions | |
cout | -1.220 |
naires | -1.173 |
BuzzFeed | -1.134 |
linger | -1.111 |
swe | -1.068 |
Token | 228 |
Token ID | 29041 |
Feature activation | +1.31e-01 |
Pos logit contributions | |
yip | +1.240 |
zee | +1.050 |
aunders | +0.864 |
zer | +0.861 |
zh | +0.833 |
Neg logit contributions | |
cout | -1.074 |
naires | -1.036 |
linger | -0.998 |
BuzzFeed | -0.997 |
swe | -0.957 |
Token | 65 |
Token ID | 2996 |
Feature activation | +1.69e-01 |
Pos logit contributions | |
yip | +0.343 |
zee | +0.284 |
aunders | +0.230 |
zer | +0.227 |
zh | +0.220 |
Neg logit contributions | |
cout | -0.337 |
naires | -0.327 |
BuzzFeed | -0.318 |
linger | -0.314 |
swe | -0.304 |
Token | Posts |
Token ID | 12043 |
Feature activation | +1.70e+00 |
Pos logit contributions | |
yip | +0.338 |
zee | +0.272 |
aunders | +0.203 |
zer | +0.201 |
zh | +0.190 |
Neg logit contributions | |
cout | -0.527 |
naires | -0.512 |
linger | -0.498 |
BuzzFeed | -0.498 |
swe | -0.485 |
Token | # |
Token ID | 1303 |
Feature activation | -3.02e-01 |
Pos logit contributions | |
yip | +5.253 |
zee | +4.463 |
aunders | +3.794 |
zer | +3.761 |
zh | +3.664 |
Neg logit contributions | |
cout | -3.471 |
naires | -3.357 |
BuzzFeed | -3.215 |
linger | -3.177 |
swe | -3.027 |
Token | 13 |
Token ID | 1485 |
Feature activation | +6.49e-02 |
Pos logit contributions | |
cout | +0.787 |
naires | +0.764 |
BuzzFeed | +0.744 |
linger | +0.731 |
swe | +0.710 |
Neg logit contributions | |
yip | -0.780 |
zee | -0.638 |
aunders | -0.516 |
zer | -0.506 |
zh | -0.491 |
Token | On |
Token ID | 1550 |
Feature activation | +2.48e-03 |
Pos logit contributions | |
yip | +0.193 |
zee | +0.164 |
aunders | +0.137 |
zer | +0.136 |
zh | +0.132 |
Neg logit contributions | |
cout | -0.143 |
naires | -0.138 |
BuzzFeed | -0.133 |
linger | -0.131 |
swe | -0.126 |
Token | September |
Token ID | 2693 |
Feature activation | -3.49e-01 |
Pos logit contributions | |
yip | +0.008 |
zee | +0.007 |
aunders | +0.006 |
zer | +0.006 |
zh | +0.005 |
Neg logit contributions | |
cout | -0.005 |
naires | -0.005 |
BuzzFeed | -0.005 |
linger | -0.005 |
swe | -0.004 |
Token | 20 |
Token ID | 1160 |
Feature activation | -2.32e-01 |
Pos logit contributions | |
cout | +0.725 |
naires | +0.697 |
BuzzFeed | +0.684 |
linger | +0.661 |
swe | +0.641 |
Neg logit contributions | |
yip | -1.090 |
zee | -0.923 |
aunders | -0.788 |
zer | -0.769 |
zh | -0.753 |
Token | June |
Token ID | 2795 |
Feature activation | +1.89e-02 |
Pos logit contributions | |
yip | +3.635 |
zee | +3.140 |
zer | +2.590 |
zh | +2.553 |
aunders | +2.508 |
Neg logit contributions | |
cout | -3.365 |
naires | -3.319 |
linger | -3.124 |
BuzzFeed | -3.076 |
omorphic | -3.023 |
Token | 2009 |
Token ID | 3717 |
Feature activation | -2.33e-01 |
Pos logit contributions | |
yip | +0.059 |
zee | +0.051 |
zer | +0.044 |
aunders | +0.043 |
zh | +0.043 |
Neg logit contributions | |
cout | -0.037 |
naires | -0.036 |
linger | -0.034 |
BuzzFeed | -0.034 |
swe | -0.033 |
Token | Germany |
Token ID | 4486 |
Feature activation | +4.66e-01 |
Pos logit contributions | |
cout | +0.375 |
naires | +0.358 |
BuzzFeed | +0.340 |
linger | +0.333 |
swe | +0.316 |
Neg logit contributions | |
yip | -0.827 |
zee | -0.723 |
aunders | -0.628 |
zer | -0.623 |
zh | -0.612 |
Token | 247 |
Token ID | 30179 |
Feature activation | +9.17e-02 |
Pos logit contributions | |
yip | +1.266 |
zee | +1.076 |
aunders | +0.884 |
zer | +0.883 |
zh | +0.855 |
Neg logit contributions | |
cout | -1.110 |
naires | -1.070 |
linger | -1.031 |
BuzzFeed | -1.028 |
swe | -0.988 |
Token | 34 |
Token ID | 2682 |
Feature activation | +2.75e-01 |
Pos logit contributions | |
yip | +0.224 |
zee | +0.183 |
aunders | +0.145 |
zer | +0.143 |
zh | +0.139 |
Neg logit contributions | |
cout | -0.250 |
naires | -0.243 |
BuzzFeed | -0.237 |
linger | -0.234 |
swe | -0.227 |
Token | Posts |
Token ID | 12043 |
Feature activation | +1.67e+00 |
Pos logit contributions | |
yip | +0.546 |
zee | +0.441 |
aunders | +0.327 |
zer | +0.327 |
zh | +0.308 |
Neg logit contributions | |
cout | -0.856 |
naires | -0.833 |
linger | -0.810 |
BuzzFeed | -0.809 |
swe | -0.790 |
Token | # |
Token ID | 1303 |
Feature activation | -2.10e-01 |
Pos logit contributions | |
yip | +5.180 |
zee | +4.412 |
aunders | +3.750 |
zer | +3.714 |
zh | +3.623 |
Neg logit contributions | |
cout | -3.392 |
naires | -3.274 |
BuzzFeed | -3.139 |
linger | -3.098 |
swe | -2.959 |
Token | 20 |
Token ID | 1238 |
Feature activation | -5.96e-02 |
Pos logit contributions | |
cout | +0.523 |
naires | +0.506 |
BuzzFeed | +0.493 |
linger | +0.484 |
swe | +0.469 |
Neg logit contributions | |
yip | -0.567 |
zee | -0.471 |
aunders | -0.386 |
zer | -0.380 |
zh | -0.369 |
Token | Nice |
Token ID | 18460 |
Feature activation | -2.45e-01 |
Pos logit contributions | |
cout | +0.130 |
naires | +0.125 |
BuzzFeed | +0.122 |
linger | +0.119 |
swe | +0.116 |
Neg logit contributions | |
yip | -0.179 |
zee | -0.151 |
aunders | -0.127 |
zer | -0.125 |
zh | -0.123 |
Token | , |
Token ID | 11 |
Feature activation | -1.35e-02 |
Pos logit contributions | |
cout | +0.501 |
naires | +0.477 |
BuzzFeed | +0.465 |
linger | +0.455 |
swe | +0.441 |
Neg logit contributions | |
yip | -0.770 |
zee | -0.657 |
aunders | -0.557 |
zer | -0.548 |
zh | -0.537 |
Token | happy |
Token ID | 3772 |
Feature activation | +1.38e-01 |
Pos logit contributions | |
cout | +0.029 |
naires | +0.028 |
BuzzFeed | +0.027 |
linger | +0.027 |
swe | +0.026 |
Neg logit contributions | |
yip | -0.041 |
zee | -0.035 |
aunders | -0.029 |
zer | -0.029 |
zh | -0.028 |
Token | 2010 |
Token ID | 3050 |
Feature activation | -7.54e-02 |
Pos logit contributions | |
yip | +0.082 |
zee | +0.072 |
zer | +0.062 |
aunders | +0.061 |
zh | +0.060 |
Neg logit contributions | |
cout | -0.052 |
naires | -0.050 |
linger | -0.047 |
BuzzFeed | -0.047 |
swe | -0.045 |
Token | United |
Token ID | 1578 |
Feature activation | +5.58e-01 |
Pos logit contributions | |
cout | +0.133 |
naires | +0.128 |
BuzzFeed | +0.122 |
linger | +0.120 |
swe | +0.114 |
Neg logit contributions | |
yip | -0.256 |
zee | -0.223 |
aunders | -0.192 |
zer | -0.190 |
zh | -0.187 |
Token | States |
Token ID | 1829 |
Feature activation | +4.74e-01 |
Pos logit contributions | |
yip | +1.640 |
zee | +1.412 |
zer | +1.185 |
aunders | +1.173 |
zh | +1.147 |
Neg logit contributions | |
cout | -1.209 |
naires | -1.164 |
BuzzFeed | -1.125 |
linger | -1.117 |
swe | -1.081 |
Token | 6 |
Token ID | 718 |
Feature activation | -2.49e-01 |
Pos logit contributions | |
yip | +1.299 |
zee | +1.100 |
aunders | +0.905 |
zer | +0.902 |
zh | +0.873 |
Neg logit contributions | |
cout | -1.124 |
naires | -1.084 |
linger | -1.045 |
BuzzFeed | -1.044 |
swe | -1.002 |
Token | 83 |
Token ID | 5999 |
Feature activation | -1.64e-01 |
Pos logit contributions | |
cout | +0.625 |
naires | +0.605 |
BuzzFeed | +0.589 |
linger | +0.580 |
swe | +0.562 |
Neg logit contributions | |
yip | -0.666 |
zee | -0.553 |
aunders | -0.451 |
zer | -0.445 |
zh | -0.432 |
Token | Posts |
Token ID | 12043 |
Feature activation | +1.66e+00 |
Pos logit contributions | |
cout | +0.496 |
naires | +0.483 |
BuzzFeed | +0.471 |
linger | +0.467 |
swe | +0.455 |
Neg logit contributions | |
yip | -0.350 |
zee | -0.277 |
aunders | -0.211 |
zer | -0.208 |
zh | -0.198 |
Token | Last |
Token ID | 4586 |
Feature activation | +4.68e-01 |
Pos logit contributions | |
yip | +5.161 |
zee | +4.395 |
aunders | +3.740 |
zer | +3.704 |
zh | +3.615 |
Neg logit contributions | |
cout | -3.374 |
naires | -3.256 |
BuzzFeed | -3.126 |
linger | -3.083 |
swe | -2.946 |
Token | Edited |
Token ID | 34212 |
Feature activation | +2.01e-01 |
Pos logit contributions | |
yip | +1.162 |
zee | +0.966 |
zer | +0.790 |
aunders | +0.786 |
zh | +0.753 |
Neg logit contributions | |
cout | -1.223 |
naires | -1.175 |
BuzzFeed | -1.138 |
linger | -1.129 |
swe | -1.101 |
Token | : |
Token ID | 25 |
Feature activation | -1.56e-02 |
Pos logit contributions | |
yip | +0.500 |
zee | +0.422 |
zer | +0.340 |
aunders | +0.337 |
zh | +0.327 |
Neg logit contributions | |
cout | -0.523 |
naires | -0.508 |
linger | -0.490 |
BuzzFeed | -0.489 |
swe | -0.472 |
Token | 2013 |
Token ID | 2211 |
Feature activation | -3.41e-02 |
Pos logit contributions | |
cout | +0.035 |
naires | +0.034 |
linger | +0.032 |
BuzzFeed | +0.032 |
swe | +0.031 |
Neg logit contributions | |
yip | -0.044 |
zee | -0.038 |
zer | -0.032 |
aunders | -0.032 |
zh | -0.031 |
Token | - |
Token ID | 12 |
Feature activation | +2.13e-02 |
Pos logit contributions | |
cout | +0.070 |
naires | +0.068 |
BuzzFeed | +0.065 |
linger | +0.064 |
swe | +0.062 |
Neg logit contributions | |
yip | -0.106 |
zee | -0.091 |
aunders | -0.077 |
zer | -0.076 |
zh | -0.074 |
Token | J |
Token ID | 41 |
Feature activation | +4.65e-01 |
Pos logit contributions | |
yip | +0.750 |
zee | +0.623 |
aunders | +0.512 |
zer | +0.504 |
zh | +0.490 |
Neg logit contributions | |
cout | -0.680 |
naires | -0.658 |
BuzzFeed | -0.640 |
linger | -0.629 |
swe | -0.610 |
Token | aaaa |
Token ID | 24794 |
Feature activation | -4.68e-01 |
Pos logit contributions | |
yip | +1.288 |
zee | +1.076 |
aunders | +0.886 |
zer | +0.875 |
zh | +0.851 |
Neg logit contributions | |
cout | -1.120 |
naires | -1.083 |
BuzzFeed | -1.053 |
linger | -1.036 |
swe | -1.003 |
Token | asper |
Token ID | 32981 |
Feature activation | -4.55e-01 |
Pos logit contributions | |
cout | +1.236 |
naires | +1.202 |
BuzzFeed | +1.171 |
linger | +1.152 |
swe | +1.121 |
Neg logit contributions | |
yip | -1.179 |
zee | -0.953 |
aunders | -0.768 |
zer | -0.750 |
zh | -0.734 |
Token | Profile |
Token ID | 13118 |
Feature activation | +9.67e-01 |
Pos logit contributions | |
cout | +1.059 |
naires | +1.022 |
BuzzFeed | +0.999 |
linger | +0.972 |
swe | +0.948 |
Neg logit contributions | |
yip | -1.283 |
zee | -1.067 |
aunders | -0.887 |
zer | -0.864 |
zh | -0.852 |
Token | Blog |
Token ID | 14001 |
Feature activation | +7.52e-01 |
Pos logit contributions | |
yip | +2.585 |
zee | +2.156 |
aunders | +1.753 |
zer | +1.745 |
zh | +1.693 |
Neg logit contributions | |
cout | -2.401 |
naires | -2.331 |
BuzzFeed | -2.254 |
linger | -2.226 |
swe | -2.152 |
Token | Joined |
Token ID | 16709 |
Feature activation | +1.57e+00 |
Pos logit contributions | |
yip | +1.527 |
zee | +1.266 |
zer | +0.937 |
aunders | +0.926 |
zh | +0.897 |
Neg logit contributions | |
cout | -2.277 |
naires | -2.240 |
linger | -2.144 |
BuzzFeed | -2.132 |
swe | -2.081 |
Token | April |
Token ID | 3035 |
Feature activation | -5.88e-02 |
Pos logit contributions | |
yip | +3.982 |
zee | +3.447 |
zer | +2.845 |
zh | +2.800 |
aunders | +2.734 |
Neg logit contributions | |
cout | -3.767 |
naires | -3.716 |
linger | -3.511 |
BuzzFeed | -3.452 |
omorphic | -3.389 |
Token | 2012 |
Token ID | 2321 |
Feature activation | +8.82e-02 |
Pos logit contributions | |
cout | +0.115 |
naires | +0.112 |
linger | +0.105 |
BuzzFeed | +0.105 |
swe | +0.099 |
Neg logit contributions | |
yip | -0.185 |
zee | -0.162 |
zer | -0.138 |
aunders | -0.137 |
zh | -0.135 |
Token | United |
Token ID | 1578 |
Feature activation | +6.85e-01 |
Pos logit contributions | |
yip | +0.300 |
zee | +0.263 |
aunders | +0.226 |
zer | +0.225 |
zh | +0.221 |
Neg logit contributions | |
cout | -0.153 |
naires | -0.147 |
linger | -0.138 |
BuzzFeed | -0.138 |
swe | -0.130 |
Token | States |
Token ID | 1829 |
Feature activation | +4.85e-01 |
Pos logit contributions | |
yip | +2.156 |
zee | +1.845 |
aunders | +1.565 |
zer | +1.548 |
zh | +1.513 |
Neg logit contributions | |
cout | -1.394 |
naires | -1.339 |
BuzzFeed | -1.295 |
linger | -1.269 |
swe | -1.220 |
Token | 5 |
Token ID | 642 |
Feature activation | -1.91e-01 |
Pos logit contributions | |
yip | +1.329 |
zee | +1.125 |
aunders | +0.925 |
zer | +0.922 |
zh | +0.891 |
Neg logit contributions | |
cout | -1.152 |
naires | -1.110 |
linger | -1.070 |
BuzzFeed | -1.068 |
swe | -1.027 |
Token | Santa |
Token ID | 8909 |
Feature activation | -1.11e-01 |
Pos logit contributions | |
yip | +0.516 |
zee | +0.443 |
aunders | +0.374 |
zer | +0.371 |
zh | +0.362 |
Neg logit contributions | |
cout | -0.356 |
naires | -0.346 |
BuzzFeed | -0.332 |
linger | -0.325 |
swe | -0.314 |
Token | Ana |
Token ID | 17639 |
Feature activation | -1.32e-02 |
Pos logit contributions | |
cout | +0.212 |
naires | +0.204 |
BuzzFeed | +0.195 |
linger | +0.192 |
swe | +0.184 |
Neg logit contributions | |
yip | -0.360 |
zee | -0.311 |
aunders | -0.266 |
zer | -0.264 |
zh | -0.258 |
Token | Police |
Token ID | 4287 |
Feature activation | -1.63e-01 |
Pos logit contributions | |
cout | +0.028 |
naires | +0.027 |
BuzzFeed | +0.026 |
linger | +0.026 |
swe | +0.025 |
Neg logit contributions | |
yip | -0.040 |
zee | -0.034 |
aunders | -0.029 |
zer | -0.029 |
zh | -0.028 |
Token | Department |
Token ID | 2732 |
Feature activation | -1.93e-01 |
Pos logit contributions | |
cout | +0.319 |
naires | +0.303 |
BuzzFeed | +0.296 |
linger | +0.289 |
swe | +0.277 |
Neg logit contributions | |
yip | -0.532 |
zee | -0.456 |
aunders | -0.388 |
zer | -0.384 |
zh | -0.375 |
Token | for |
Token ID | 329 |
Feature activation | -2.15e-02 |
Pos logit contributions | |
cout | +0.404 |
naires | +0.389 |
BuzzFeed | +0.376 |
linger | +0.369 |
swe | +0.356 |
Neg logit contributions | |
yip | -0.593 |
zee | -0.506 |
aunders | -0.428 |
zer | -0.423 |
zh | -0.413 |
Token | reasons |
Token ID | 3840 |
Feature activation | +1.50e+00 |
Pos logit contributions | |
cout | +0.043 |
naires | +0.041 |
BuzzFeed | +0.040 |
linger | +0.039 |
swe | +0.037 |
Neg logit contributions | |
yip | -0.068 |
zee | -0.059 |
aunders | -0.050 |
zer | -0.049 |
zh | -0.048 |
Token | that |
Token ID | 326 |
Feature activation | +2.11e-01 |
Pos logit contributions | |
yip | +4.443 |
zee | +3.874 |
zer | +3.233 |
aunders | +3.225 |
zh | +3.146 |
Neg logit contributions | |
cout | -3.114 |
naires | -3.008 |
linger | -2.781 |
BuzzFeed | -2.778 |
swe | -2.710 |
Token | have |
Token ID | 423 |
Feature activation | +9.03e-02 |
Pos logit contributions | |
yip | +0.694 |
zee | +0.599 |
aunders | +0.512 |
zer | +0.509 |
zh | +0.500 |
Neg logit contributions | |
cout | -0.394 |
naires | -0.377 |
BuzzFeed | -0.361 |
linger | -0.351 |
swe | -0.341 |
Token | not |
Token ID | 407 |
Feature activation | +1.91e-01 |
Pos logit contributions | |
yip | +0.297 |
zee | +0.257 |
aunders | +0.220 |
zer | +0.219 |
zh | +0.215 |
Neg logit contributions | |
cout | -0.168 |
naires | -0.161 |
BuzzFeed | -0.154 |
linger | -0.150 |
swe | -0.144 |
Token | been |
Token ID | 587 |
Feature activation | +1.93e-01 |
Pos logit contributions | |
yip | +0.645 |
zee | +0.562 |
aunders | +0.483 |
zer | +0.480 |
zh | +0.476 |
Neg logit contributions | |
cout | -0.337 |
naires | -0.322 |
BuzzFeed | -0.307 |
linger | -0.299 |
swe | -0.284 |
Token | publicly |
Token ID | 7271 |
Feature activation | -3.30e-01 |
Pos logit contributions | |
yip | +0.582 |
zee | +0.509 |
aunders | +0.425 |
zer | +0.423 |
zh | +0.420 |
Neg logit contributions | |
cout | -0.402 |
naires | -0.389 |
BuzzFeed | -0.371 |
linger | -0.366 |
swe | -0.347 |
Token |
|
Token ID | 198 |
Feature activation | +1.81e-01 |
Pos logit contributions | |
yip | +0.690 |
zee | +0.574 |
zer | +0.457 |
aunders | +0.455 |
zh | +0.441 |
Neg logit contributions | |
cout | -0.766 |
naires | -0.740 |
BuzzFeed | -0.720 |
linger | -0.716 |
swe | -0.694 |
Token | Dark |
Token ID | 17367 |
Feature activation | +7.16e-01 |
Pos logit contributions | |
yip | +0.492 |
zee | +0.408 |
aunders | +0.335 |
zer | +0.330 |
zh | +0.320 |
Neg logit contributions | |
cout | -0.446 |
naires | -0.432 |
BuzzFeed | -0.420 |
linger | -0.413 |
swe | -0.401 |
Token | horse |
Token ID | 30527 |
Feature activation | +6.16e-02 |
Pos logit contributions | |
yip | +1.965 |
zee | +1.646 |
aunders | +1.349 |
zer | +1.338 |
zh | +1.300 |
Neg logit contributions | |
cout | -1.741 |
naires | -1.682 |
BuzzFeed | -1.637 |
linger | -1.611 |
swe | -1.557 |
Token | Profile |
Token ID | 13118 |
Feature activation | +7.20e-01 |
Pos logit contributions | |
yip | +0.170 |
zee | +0.141 |
aunders | +0.116 |
zer | +0.114 |
zh | +0.111 |
Neg logit contributions | |
cout | -0.149 |
naires | -0.144 |
BuzzFeed | -0.140 |
linger | -0.138 |
swe | -0.134 |
Token | Blog |
Token ID | 14001 |
Feature activation | +6.91e-01 |
Pos logit contributions | |
yip | +2.016 |
zee | +1.689 |
aunders | +1.395 |
zer | +1.376 |
zh | +1.340 |
Neg logit contributions | |
cout | -1.715 |
naires | -1.657 |
BuzzFeed | -1.612 |
linger | -1.584 |
swe | -1.533 |
Token | Joined |
Token ID | 16709 |
Feature activation | +1.48e+00 |
Pos logit contributions | |
yip | +1.413 |
zee | +1.166 |
zer | +0.871 |
aunders | +0.857 |
zh | +0.839 |
Neg logit contributions | |
cout | -2.089 |
naires | -2.048 |
linger | -1.964 |
BuzzFeed | -1.956 |
swe | -1.906 |
Token | December |
Token ID | 3426 |
Feature activation | -6.05e-02 |
Pos logit contributions | |
yip | +3.806 |
zee | +3.286 |
zer | +2.714 |
zh | +2.673 |
aunders | +2.620 |
Neg logit contributions | |
cout | -3.543 |
naires | -3.492 |
linger | -3.299 |
BuzzFeed | -3.253 |
omorphic | -3.185 |
Token | 2011 |
Token ID | 2813 |
Feature activation | -9.55e-03 |
Pos logit contributions | |
cout | +0.118 |
naires | +0.114 |
linger | +0.108 |
BuzzFeed | +0.107 |
swe | +0.102 |
Neg logit contributions | |
yip | -0.189 |
zee | -0.166 |
zer | -0.142 |
aunders | -0.141 |
zh | -0.139 |
Token | United |
Token ID | 1578 |
Feature activation | +5.99e-01 |
Pos logit contributions | |
cout | +0.017 |
naires | +0.016 |
BuzzFeed | +0.015 |
linger | +0.015 |
swe | +0.014 |
Neg logit contributions | |
yip | -0.033 |
zee | -0.028 |
aunders | -0.025 |
zer | -0.024 |
zh | -0.024 |
Token | States |
Token ID | 1829 |
Feature activation | +4.53e-01 |
Pos logit contributions | |
yip | +1.883 |
zee | +1.611 |
aunders | +1.367 |
zer | +1.350 |
zh | +1.321 |
Neg logit contributions | |
cout | -1.220 |
naires | -1.173 |
BuzzFeed | -1.134 |
linger | -1.111 |
swe | -1.068 |
Token | 228 |
Token ID | 29041 |
Feature activation | +1.31e-01 |
Pos logit contributions | |
yip | +1.240 |
zee | +1.050 |
aunders | +0.864 |
zer | +0.861 |
zh | +0.833 |
Neg logit contributions | |
cout | -1.074 |
naires | -1.036 |
linger | -0.998 |
BuzzFeed | -0.997 |
swe | -0.957 |
Token | � |
Token ID | 250 |
Feature activation | +2.93e-02 |
Pos logit contributions | |
cout | +0.110 |
naires | +0.107 |
BuzzFeed | +0.104 |
linger | +0.103 |
swe | +0.100 |
Neg logit contributions | |
yip | -0.105 |
zee | -0.086 |
aunders | -0.069 |
zer | -0.068 |
zh | -0.066 |
Token | Ass |
Token ID | 8021 |
Feature activation | +1.46e-01 |
Pos logit contributions | |
yip | +0.080 |
zee | +0.067 |
aunders | +0.055 |
zer | +0.054 |
zh | +0.053 |
Neg logit contributions | |
cout | -0.071 |
naires | -0.069 |
BuzzFeed | -0.067 |
linger | -0.066 |
swe | -0.064 |
Token | imilation |
Token ID | 42963 |
Feature activation | -3.70e-01 |
Pos logit contributions | |
yip | +0.478 |
zee | +0.412 |
aunders | +0.352 |
zer | +0.349 |
zh | +0.341 |
Neg logit contributions | |
cout | -0.279 |
naires | -0.268 |
BuzzFeed | -0.258 |
linger | -0.253 |
swe | -0.243 |
Token | in |
Token ID | 287 |
Feature activation | +1.23e-01 |
Pos logit contributions | |
cout | +0.768 |
naires | +0.740 |
BuzzFeed | +0.716 |
linger | +0.701 |
swe | +0.674 |
Neg logit contributions | |
yip | -1.148 |
zee | -0.980 |
aunders | -0.829 |
zer | -0.819 |
zh | -0.801 |
Token | Biblical |
Token ID | 31752 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
yip | +0.379 |
zee | +0.323 |
aunders | +0.273 |
zer | +0.270 |
zh | +0.264 |
Neg logit contributions | |
cout | -0.257 |
naires | -0.247 |
BuzzFeed | -0.239 |
linger | -0.235 |
swe | -0.226 |
Token | Narr |
Token ID | 28390 |
Feature activation | +1.47e+00 |
Pos logit contributions | |
yip | +0.335 |
zee | +0.290 |
aunders | +0.245 |
zer | +0.244 |
zh | +0.238 |
Neg logit contributions | |
cout | -0.216 |
naires | -0.209 |
BuzzFeed | -0.200 |
linger | -0.197 |
swe | -0.190 |
Token | atives |
Token ID | 2929 |
Feature activation | -6.95e-03 |
Pos logit contributions | |
yip | +4.174 |
zee | +3.626 |
zer | +3.006 |
aunders | +3.003 |
zh | +2.933 |
Neg logit contributions | |
cout | -3.268 |
naires | -3.158 |
BuzzFeed | -3.050 |
linger | -3.027 |
swe | -2.910 |
Token | . |
Token ID | 13 |
Feature activation | -1.24e-02 |
Pos logit contributions | |
cout | +0.015 |
naires | +0.015 |
BuzzFeed | +0.014 |
linger | +0.014 |
swe | +0.014 |
Neg logit contributions | |
yip | -0.021 |
zee | -0.017 |
aunders | -0.015 |
zer | -0.014 |
zh | -0.014 |
Token | � |
Token ID | 447 |
Feature activation | +7.42e-02 |
Pos logit contributions | |
cout | +0.026 |
naires | +0.025 |
BuzzFeed | +0.024 |
linger | +0.024 |
swe | +0.023 |
Neg logit contributions | |
yip | -0.038 |
zee | -0.033 |
aunders | -0.028 |
zer | -0.028 |
zh | -0.027 |
Token | � |
Token ID | 251 |
Feature activation | +6.59e-02 |
Pos logit contributions | |
yip | +0.187 |
zee | +0.153 |
aunders | +0.122 |
zer | +0.120 |
zh | +0.116 |
Neg logit contributions | |
cout | -0.198 |
naires | -0.192 |
BuzzFeed | -0.189 |
linger | -0.185 |
swe | -0.180 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.81e-01 |
Pos logit contributions | |
yip | +0.197 |
zee | +0.168 |
aunders | +0.141 |
zer | +0.140 |
zh | +0.136 |
Neg logit contributions | |
cout | -0.143 |
naires | -0.138 |
BuzzFeed | -0.133 |
linger | -0.132 |
swe | -0.127 |
Token |
|
Token ID | 198 |
Feature activation | +2.84e-01 |
Pos logit contributions | |
yip | +1.116 |
zee | +0.935 |
zer | +0.744 |
aunders | +0.736 |
zh | +0.716 |
Neg logit contributions | |
cout | -1.278 |
naires | -1.234 |
BuzzFeed | -1.199 |
linger | -1.196 |
swe | -1.160 |
Token | re |
Token ID | 260 |
Feature activation | -7.32e-02 |
Pos logit contributions | |
yip | +0.771 |
zee | +0.641 |
aunders | +0.526 |
zer | +0.517 |
zh | +0.503 |
Neg logit contributions | |
cout | -0.698 |
naires | -0.677 |
BuzzFeed | -0.658 |
linger | -0.646 |
swe | -0.627 |
Token | i |
Token ID | 72 |
Feature activation | -2.17e-01 |
Pos logit contributions | |
cout | +0.179 |
naires | +0.173 |
BuzzFeed | +0.168 |
linger | +0.166 |
swe | +0.160 |
Neg logit contributions | |
yip | -0.199 |
zee | -0.165 |
aunders | -0.136 |
zer | -0.133 |
zh | -0.129 |
Token | Profile |
Token ID | 13118 |
Feature activation | +9.04e-01 |
Pos logit contributions | |
cout | +0.534 |
naires | +0.518 |
BuzzFeed | +0.502 |
linger | +0.495 |
swe | +0.481 |
Neg logit contributions | |
yip | -0.588 |
zee | -0.487 |
aunders | -0.399 |
zer | -0.390 |
zh | -0.381 |
Token | Blog |
Token ID | 14001 |
Feature activation | +7.54e-01 |
Pos logit contributions | |
yip | +2.413 |
zee | +2.009 |
aunders | +1.635 |
zer | +1.622 |
zh | +1.575 |
Neg logit contributions | |
cout | -2.253 |
naires | -2.185 |
BuzzFeed | -2.120 |
linger | -2.089 |
swe | -2.022 |
Token | Joined |
Token ID | 16709 |
Feature activation | +1.43e+00 |
Pos logit contributions | |
yip | +1.548 |
zee | +1.274 |
zer | +0.942 |
aunders | +0.937 |
zh | +0.907 |
Neg logit contributions | |
cout | -2.272 |
naires | -2.233 |
linger | -2.146 |
BuzzFeed | -2.144 |
swe | -2.082 |
Token | October |
Token ID | 3267 |
Feature activation | -9.89e-02 |
Pos logit contributions | |
yip | +3.720 |
zee | +3.210 |
zer | +2.664 |
zh | +2.614 |
aunders | +2.578 |
Neg logit contributions | |
cout | -3.404 |
naires | -3.349 |
linger | -3.171 |
BuzzFeed | -3.147 |
omorphic | -3.042 |
Token | 2002 |
Token ID | 6244 |
Feature activation | -6.65e-02 |
Pos logit contributions | |
cout | +0.192 |
naires | +0.186 |
linger | +0.175 |
BuzzFeed | +0.174 |
swe | +0.166 |
Neg logit contributions | |
yip | -0.312 |
zee | -0.274 |
aunders | -0.232 |
zer | -0.232 |
zh | -0.226 |
Token | United |
Token ID | 1578 |
Feature activation | +4.90e-01 |
Pos logit contributions | |
cout | +0.117 |
naires | +0.112 |
BuzzFeed | +0.106 |
linger | +0.105 |
swe | +0.100 |
Neg logit contributions | |
yip | -0.226 |
zee | -0.197 |
aunders | -0.170 |
zer | -0.169 |
zh | -0.165 |
Token | States |
Token ID | 1829 |
Feature activation | +4.26e-01 |
Pos logit contributions | |
yip | +1.558 |
zee | +1.334 |
aunders | +1.136 |
zer | +1.119 |
zh | +1.097 |
Neg logit contributions | |
cout | -0.985 |
naires | -0.946 |
BuzzFeed | -0.913 |
linger | -0.895 |
swe | -0.859 |
Token | 3 |
Token ID | 513 |
Feature activation | -1.92e-01 |
Pos logit contributions | |
yip | +1.173 |
zee | +0.991 |
aunders | +0.816 |
zer | +0.813 |
zh | +0.786 |
Neg logit contributions | |
cout | -1.009 |
naires | -0.973 |
BuzzFeed | -0.938 |
linger | -0.937 |
swe | -0.899 |
Token | N |
Token ID | 399 |
Feature activation | +4.26e-02 |
Pos logit contributions | |
yip | +0.103 |
zee | +0.088 |
aunders | +0.075 |
zer | +0.074 |
zh | +0.072 |
Neg logit contributions | |
cout | -0.067 |
naires | -0.065 |
BuzzFeed | -0.062 |
linger | -0.061 |
swe | -0.059 |
Token | ana |
Token ID | 2271 |
Feature activation | +2.33e-01 |
Pos logit contributions | |
yip | +0.117 |
zee | +0.098 |
aunders | +0.080 |
zer | +0.080 |
zh | +0.077 |
Neg logit contributions | |
cout | -0.104 |
naires | -0.100 |
BuzzFeed | -0.097 |
linger | -0.096 |
swe | -0.093 |
Token | won |
Token ID | 1839 |
Feature activation | -3.22e-01 |
Pos logit contributions | |
yip | +0.705 |
zee | +0.599 |
aunders | +0.504 |
zer | +0.503 |
zh | +0.489 |
Neg logit contributions | |
cout | -0.498 |
naires | -0.480 |
BuzzFeed | -0.463 |
linger | -0.456 |
swe | -0.438 |
Token | the |
Token ID | 262 |
Feature activation | -3.29e-01 |
Pos logit contributions | |
cout | +0.662 |
naires | +0.637 |
BuzzFeed | +0.615 |
linger | +0.603 |
swe | +0.581 |
Neg logit contributions | |
yip | -1.002 |
zee | -0.857 |
aunders | -0.726 |
zer | -0.719 |
zh | -0.703 |
Token | Sh |
Token ID | 911 |
Feature activation | -6.97e-03 |
Pos logit contributions | |
cout | +0.683 |
naires | +0.660 |
BuzzFeed | +0.634 |
linger | +0.622 |
swe | +0.600 |
Neg logit contributions | |
yip | -1.016 |
zee | -0.870 |
aunders | -0.735 |
zer | -0.730 |
zh | -0.714 |
Token | og |
Token ID | 519 |
Feature activation | +1.42e+00 |
Pos logit contributions | |
cout | +0.016 |
naires | +0.016 |
BuzzFeed | +0.015 |
linger | +0.015 |
swe | +0.014 |
Neg logit contributions | |
yip | -0.020 |
zee | -0.017 |
aunders | -0.014 |
zer | -0.014 |
zh | -0.013 |
Token | ak |
Token ID | 461 |
Feature activation | +5.23e-01 |
Pos logit contributions | |
yip | +3.165 |
zee | +2.575 |
aunders | +1.979 |
zer | +1.978 |
zh | +1.907 |
Neg logit contributions | |
cout | -4.115 |
naires | -4.005 |
BuzzFeed | -3.921 |
linger | -3.858 |
swe | -3.779 |
Token | uk |
Token ID | 2724 |
Feature activation | +7.19e-01 |
Pos logit contributions | |
yip | +1.373 |
zee | +1.152 |
aunders | +0.937 |
zer | +0.935 |
zh | +0.908 |
Neg logit contributions | |
cout | -1.300 |
naires | -1.274 |
BuzzFeed | -1.229 |
linger | -1.209 |
swe | -1.180 |
Token | an |
Token ID | 272 |
Feature activation | +2.91e-01 |
Pos logit contributions | |
yip | +1.478 |
zee | +1.168 |
aunders | +0.870 |
zer | +0.863 |
zh | +0.822 |
Neg logit contributions | |
cout | -2.220 |
naires | -2.169 |
BuzzFeed | -2.124 |
linger | -2.098 |
swe | -2.038 |
Token | Manga |
Token ID | 27024 |
Feature activation | -4.08e-01 |
Pos logit contributions | |
yip | +0.844 |
zee | +0.716 |
zer | +0.596 |
aunders | +0.592 |
zh | +0.581 |
Neg logit contributions | |
cout | -0.656 |
naires | -0.635 |
BuzzFeed | -0.612 |
linger | -0.603 |
swe | -0.583 |
Token | Award |
Token ID | 11289 |
Feature activation | -5.14e-01 |
Pos logit contributions | |
cout | +0.939 |
naires | +0.906 |
BuzzFeed | +0.880 |
linger | +0.865 |
swe | +0.835 |
Neg logit contributions | |
yip | -1.174 |
zee | -0.989 |
aunders | -0.822 |
zer | -0.812 |
zh | -0.791 |
Token | and |
Token ID | 290 |
Feature activation | +4.35e-02 |
Pos logit contributions | |
yip | +0.273 |
zee | +0.232 |
aunders | +0.195 |
zer | +0.193 |
zh | +0.188 |
Neg logit contributions | |
cout | -0.198 |
naires | -0.191 |
BuzzFeed | -0.185 |
linger | -0.182 |
swe | -0.175 |
Token | could |
Token ID | 714 |
Feature activation | +5.88e-02 |
Pos logit contributions | |
yip | +0.137 |
zee | +0.118 |
aunders | +0.100 |
zer | +0.099 |
zh | +0.097 |
Neg logit contributions | |
cout | -0.087 |
naires | -0.084 |
BuzzFeed | -0.081 |
linger | -0.079 |
swe | -0.076 |
Token | upset |
Token ID | 9247 |
Feature activation | +7.35e-01 |
Pos logit contributions | |
yip | +0.182 |
zee | +0.155 |
aunders | +0.130 |
zer | +0.130 |
zh | +0.127 |
Neg logit contributions | |
cout | -0.122 |
naires | -0.118 |
BuzzFeed | -0.114 |
linger | -0.111 |
swe | -0.108 |
Token | Na |
Token ID | 11013 |
Feature activation | +7.28e-02 |
Pos logit contributions | |
yip | +2.201 |
zee | +1.872 |
aunders | +1.573 |
zer | +1.568 |
zh | +1.532 |
Neg logit contributions | |
cout | -1.583 |
naires | -1.526 |
BuzzFeed | -1.470 |
linger | -1.453 |
swe | -1.396 |
Token | ` |
Token ID | 63 |
Feature activation | +8.29e-01 |
Pos logit contributions | |
yip | +0.168 |
zee | +0.136 |
aunders | +0.105 |
zer | +0.104 |
zh | +0.101 |
Neg logit contributions | |
cout | -0.209 |
naires | -0.203 |
BuzzFeed | -0.198 |
linger | -0.195 |
swe | -0.190 |
Token | Vi |
Token ID | 38432 |
Feature activation | +1.41e+00 |
Pos logit contributions | |
yip | +1.970 |
zee | +1.594 |
zer | +1.260 |
aunders | +1.250 |
zh | +1.213 |
Neg logit contributions | |
cout | -2.318 |
naires | -2.259 |
BuzzFeed | -2.200 |
linger | -2.176 |
swe | -2.115 |
Token | in |
Token ID | 287 |
Feature activation | +1.74e-02 |
Pos logit contributions | |
yip | +4.326 |
zee | +3.701 |
zer | +3.153 |
aunders | +3.126 |
zh | +3.092 |
Neg logit contributions | |
cout | -2.842 |
naires | -2.748 |
BuzzFeed | -2.674 |
linger | -2.590 |
swe | -2.485 |
Token | some |
Token ID | 617 |
Feature activation | -2.99e-02 |
Pos logit contributions | |
yip | +0.054 |
zee | +0.046 |
aunders | +0.039 |
zer | +0.038 |
zh | +0.037 |
Neg logit contributions | |
cout | -0.036 |
naires | -0.035 |
BuzzFeed | -0.034 |
linger | -0.033 |
swe | -0.032 |
Token | universe |
Token ID | 6881 |
Feature activation | +1.90e-01 |
Pos logit contributions | |
cout | +0.059 |
naires | +0.056 |
BuzzFeed | +0.054 |
linger | +0.053 |
swe | +0.051 |
Neg logit contributions | |
yip | -0.097 |
zee | -0.083 |
aunders | -0.071 |
zer | -0.070 |
zh | -0.068 |
Token | , |
Token ID | 11 |
Feature activation | +4.62e-02 |
Pos logit contributions | |
yip | +0.591 |
zee | +0.505 |
aunders | +0.427 |
zer | +0.425 |
zh | +0.414 |
Neg logit contributions | |
cout | -0.392 |
naires | -0.377 |
BuzzFeed | -0.367 |
linger | -0.358 |
swe | -0.343 |
Token | but |
Token ID | 475 |
Feature activation | +1.00e-01 |
Pos logit contributions | |
yip | +0.134 |
zee | +0.114 |
aunders | +0.095 |
zer | +0.094 |
zh | +0.091 |
Neg logit contributions | |
cout | -0.105 |
naires | -0.101 |
BuzzFeed | -0.098 |
linger | -0.096 |
swe | -0.093 |
Token | ada |
Token ID | 4763 |
Feature activation | +9.65e-02 |
Pos logit contributions | |
cout | +0.725 |
naires | +0.704 |
BuzzFeed | +0.679 |
linger | +0.667 |
swe | +0.649 |
Neg logit contributions | |
yip | -0.862 |
zee | -0.721 |
aunders | -0.599 |
zer | -0.587 |
zh | -0.572 |
Token | ver |
Token ID | 332 |
Feature activation | +2.04e-01 |
Pos logit contributions | |
yip | +0.268 |
zee | +0.224 |
aunders | +0.185 |
zer | +0.183 |
zh | +0.178 |
Neg logit contributions | |
cout | -0.231 |
naires | -0.224 |
BuzzFeed | -0.218 |
linger | -0.214 |
swe | -0.207 |
Token | BB |
Token ID | 15199 |
Feature activation | -9.40e-02 |
Pos logit contributions | |
yip | +0.574 |
zee | +0.481 |
aunders | +0.398 |
zer | +0.392 |
zh | +0.382 |
Neg logit contributions | |
cout | -0.482 |
naires | -0.466 |
BuzzFeed | -0.453 |
linger | -0.445 |
swe | -0.431 |
Token | Profile |
Token ID | 13118 |
Feature activation | +6.70e-01 |
Pos logit contributions | |
cout | +0.226 |
naires | +0.217 |
BuzzFeed | +0.212 |
linger | +0.208 |
swe | +0.203 |
Neg logit contributions | |
yip | -0.261 |
zee | -0.217 |
aunders | -0.179 |
zer | -0.175 |
zh | -0.171 |
Token | Blog |
Token ID | 14001 |
Feature activation | +5.64e-01 |
Pos logit contributions | |
yip | +1.920 |
zee | +1.617 |
aunders | +1.343 |
zer | +1.329 |
zh | +1.294 |
Neg logit contributions | |
cout | -1.545 |
naires | -1.493 |
BuzzFeed | -1.448 |
linger | -1.424 |
swe | -1.375 |
Token | Joined |
Token ID | 16709 |
Feature activation | +1.41e+00 |
Pos logit contributions | |
yip | +1.160 |
zee | +0.952 |
zer | +0.713 |
aunders | +0.705 |
zh | +0.683 |
Neg logit contributions | |
cout | -1.703 |
naires | -1.668 |
linger | -1.600 |
BuzzFeed | -1.594 |
swe | -1.552 |
Token | June |
Token ID | 2795 |
Feature activation | +1.89e-02 |
Pos logit contributions | |
yip | +3.635 |
zee | +3.140 |
zer | +2.590 |
zh | +2.553 |
aunders | +2.508 |
Neg logit contributions | |
cout | -3.365 |
naires | -3.319 |
linger | -3.124 |
BuzzFeed | -3.076 |
omorphic | -3.023 |
Token | 2009 |
Token ID | 3717 |
Feature activation | -2.33e-01 |
Pos logit contributions | |
yip | +0.059 |
zee | +0.051 |
zer | +0.044 |
aunders | +0.043 |
zh | +0.043 |
Neg logit contributions | |
cout | -0.037 |
naires | -0.036 |
linger | -0.034 |
BuzzFeed | -0.034 |
swe | -0.033 |
Token | Germany |
Token ID | 4486 |
Feature activation | +4.66e-01 |
Pos logit contributions | |
cout | +0.375 |
naires | +0.358 |
BuzzFeed | +0.340 |
linger | +0.333 |
swe | +0.316 |
Neg logit contributions | |
yip | -0.827 |
zee | -0.723 |
aunders | -0.628 |
zer | -0.623 |
zh | -0.612 |
Token | 247 |
Token ID | 30179 |
Feature activation | +9.17e-02 |
Pos logit contributions | |
yip | +1.266 |
zee | +1.076 |
aunders | +0.884 |
zer | +0.883 |
zh | +0.855 |
Neg logit contributions | |
cout | -1.110 |
naires | -1.070 |
linger | -1.031 |
BuzzFeed | -1.028 |
swe | -0.988 |
Token | 34 |
Token ID | 2682 |
Feature activation | +2.75e-01 |
Pos logit contributions | |
yip | +0.224 |
zee | +0.183 |
aunders | +0.145 |
zer | +0.143 |
zh | +0.139 |
Neg logit contributions | |
cout | -0.250 |
naires | -0.243 |
BuzzFeed | -0.237 |
linger | -0.234 |
swe | -0.227 |
Token | stomach |
Token ID | 11384 |
Feature activation | +1.10e-01 |
Pos logit contributions | |
yip | +1.292 |
zee | +1.107 |
aunders | +0.932 |
zer | +0.930 |
zh | +0.913 |
Neg logit contributions | |
cout | -0.871 |
naires | -0.848 |
BuzzFeed | -0.814 |
linger | -0.795 |
swe | -0.765 |
Token | contents |
Token ID | 10154 |
Feature activation | +5.91e-01 |
Pos logit contributions | |
yip | +0.338 |
zee | +0.289 |
aunders | +0.244 |
zer | +0.241 |
zh | +0.237 |
Neg logit contributions | |
cout | -0.231 |
naires | -0.222 |
BuzzFeed | -0.216 |
linger | -0.211 |
swe | -0.202 |
Token | found |
Token ID | 1043 |
Feature activation | -2.42e-01 |
Pos logit contributions | |
yip | +1.794 |
zee | +1.535 |
aunders | +1.302 |
zer | +1.297 |
zh | +1.270 |
Neg logit contributions | |
cout | -1.225 |
naires | -1.169 |
BuzzFeed | -1.140 |
linger | -1.110 |
swe | -1.073 |
Token | a |
Token ID | 257 |
Feature activation | -3.71e-01 |
Pos logit contributions | |
cout | +0.477 |
naires | +0.457 |
BuzzFeed | +0.441 |
linger | +0.432 |
swe | +0.415 |
Neg logit contributions | |
yip | -0.771 |
zee | -0.662 |
aunders | -0.565 |
zer | -0.559 |
zh | -0.547 |
Token | con |
Token ID | 369 |
Feature activation | -4.53e-01 |
Pos logit contributions | |
cout | +0.735 |
naires | +0.711 |
BuzzFeed | +0.682 |
linger | +0.667 |
swe | +0.641 |
Neg logit contributions | |
yip | -1.178 |
zee | -1.011 |
aunders | -0.859 |
zer | -0.856 |
zh | -0.838 |
Token | ge |
Token ID | 469 |
Feature activation | +1.40e+00 |
Pos logit contributions | |
cout | +1.074 |
naires | +1.036 |
BuzzFeed | +1.006 |
linger | +0.991 |
swe | +0.957 |
Neg logit contributions | |
yip | -1.273 |
zee | -1.066 |
aunders | -0.882 |
zer | -0.868 |
zh | -0.845 |
Token | aled |
Token ID | 3021 |
Feature activation | +9.11e-02 |
Pos logit contributions | |
yip | +4.157 |
zee | +3.572 |
zer | +3.038 |
aunders | +2.972 |
zh | +2.971 |
Neg logit contributions | |
cout | -2.962 |
naires | -2.878 |
BuzzFeed | -2.819 |
linger | -2.740 |
swe | -2.632 |
Token | , |
Token ID | 11 |
Feature activation | -3.69e-02 |
Pos logit contributions | |
yip | +0.292 |
zee | +0.251 |
aunders | +0.213 |
zer | +0.213 |
zh | +0.208 |
Neg logit contributions | |
cout | -0.176 |
naires | -0.172 |
BuzzFeed | -0.164 |
linger | -0.160 |
swe | -0.153 |
Token | brown |
Token ID | 7586 |
Feature activation | +1.28e-01 |
Pos logit contributions | |
cout | +0.072 |
naires | +0.069 |
BuzzFeed | +0.066 |
linger | +0.065 |
swe | +0.062 |
Neg logit contributions | |
yip | -0.119 |
zee | -0.102 |
aunders | -0.087 |
zer | -0.086 |
zh | -0.085 |
Token | ish |
Token ID | 680 |
Feature activation | +3.65e-01 |
Pos logit contributions | |
yip | +0.405 |
zee | +0.348 |
aunders | +0.295 |
zer | +0.294 |
zh | +0.287 |
Neg logit contributions | |
cout | -0.256 |
naires | -0.247 |
BuzzFeed | -0.238 |
linger | -0.232 |
swe | -0.224 |
Token | - |
Token ID | 12 |
Feature activation | -6.82e-02 |
Pos logit contributions | |
yip | +1.140 |
zee | +0.981 |
aunders | +0.832 |
zer | +0.830 |
zh | +0.814 |
Neg logit contributions | |
cout | -0.725 |
naires | -0.710 |
BuzzFeed | -0.672 |
linger | -0.656 |
swe | -0.631 |
Token | redesign |
Token ID | 27135 |
Feature activation | +6.68e-02 |
Pos logit contributions | |
cout | +0.093 |
naires | +0.090 |
BuzzFeed | +0.086 |
linger | +0.084 |
swe | +0.080 |
Neg logit contributions | |
yip | -0.156 |
zee | -0.134 |
aunders | -0.115 |
zer | -0.114 |
zh | -0.112 |
Token | . |
Token ID | 13 |
Feature activation | -9.08e-02 |
Pos logit contributions | |
yip | +0.203 |
zee | +0.173 |
aunders | +0.146 |
zer | +0.145 |
zh | +0.141 |
Neg logit contributions | |
cout | -0.142 |
naires | -0.137 |
BuzzFeed | -0.132 |
linger | -0.129 |
swe | -0.124 |
Token |
|
Token ID | 198 |
Feature activation | -2.88e-01 |
Pos logit contributions | |
cout | +0.223 |
naires | +0.216 |
BuzzFeed | +0.210 |
linger | +0.207 |
swe | +0.200 |
Neg logit contributions | |
yip | -0.247 |
zee | -0.206 |
aunders | -0.168 |
zer | -0.166 |
zh | -0.162 |
Token |
|
Token ID | 198 |
Feature activation | -2.92e-01 |
Pos logit contributions | |
cout | +0.755 |
naires | +0.732 |
BuzzFeed | +0.713 |
linger | +0.703 |
swe | +0.683 |
Neg logit contributions | |
yip | -0.734 |
zee | -0.606 |
aunders | -0.488 |
zer | -0.483 |
zh | -0.468 |
Token | R |
Token ID | 49 |
Feature activation | -3.30e-01 |
Pos logit contributions | |
cout | +0.712 |
naires | +0.690 |
BuzzFeed | +0.671 |
linger | +0.658 |
swe | +0.637 |
Neg logit contributions | |
yip | -0.803 |
zee | -0.668 |
aunders | -0.549 |
zer | -0.540 |
zh | -0.525 |
Token | end |
Token ID | 437 |
Feature activation | -1.73e+00 |
Pos logit contributions | |
cout | +0.476 |
naires | +0.452 |
BuzzFeed | +0.427 |
linger | +0.416 |
swe | +0.388 |
Neg logit contributions | |
yip | -1.231 |
zee | -1.078 |
aunders | -0.936 |
zer | -0.936 |
zh | -0.916 |
Token | ering |
Token ID | 1586 |
Feature activation | -4.44e-01 |
Pos logit contributions | |
cout | +3.288 |
naires | +3.120 |
BuzzFeed | +3.052 |
linger | +2.939 |
swe | +2.816 |
Neg logit contributions | |
yip | -5.680 |
zee | -4.811 |
aunders | -4.122 |
zer | -4.030 |
zh | -3.969 |
Token | of |
Token ID | 286 |
Feature activation | +7.55e-02 |
Pos logit contributions | |
cout | +0.849 |
naires | +0.804 |
BuzzFeed | +0.784 |
linger | +0.764 |
swe | +0.730 |
Neg logit contributions | |
yip | -1.470 |
zee | -1.260 |
aunders | -1.069 |
zer | -1.055 |
zh | -1.040 |
Token | the |
Token ID | 262 |
Feature activation | -2.88e-02 |
Pos logit contributions | |
yip | +0.230 |
zee | +0.196 |
aunders | +0.165 |
zer | +0.163 |
zh | +0.159 |
Neg logit contributions | |
cout | -0.161 |
naires | -0.155 |
BuzzFeed | -0.150 |
linger | -0.147 |
swe | -0.142 |
Token | northwest |
Token ID | 24821 |
Feature activation | -3.94e-04 |
Pos logit contributions | |
cout | +0.063 |
naires | +0.061 |
BuzzFeed | +0.059 |
linger | +0.057 |
swe | +0.055 |
Neg logit contributions | |
yip | -0.085 |
zee | -0.073 |
aunders | -0.061 |
zer | -0.061 |
zh | -0.059 |
Token | corner |
Token ID | 5228 |
Feature activation | +3.79e-01 |
Pos logit contributions | |
cout | +0.001 |
naires | +0.001 |
BuzzFeed | +0.001 |
linger | +0.001 |
swe | +0.001 |
Neg logit contributions | |
yip | -0.001 |
zee | -0.001 |
aunders | -0.001 |
zer | -0.001 |
zh | -0.001 |
Token | 1000 |
Token ID | 8576 |
Feature activation | -2.85e-01 |
Pos logit contributions | |
yip | +2.108 |
zee | +1.796 |
aunders | +1.514 |
zer | +1.513 |
zh | +1.478 |
Neg logit contributions | |
cout | -1.466 |
naires | -1.430 |
BuzzFeed | -1.376 |
linger | -1.345 |
swe | -1.296 |
Token | . |
Token ID | 13 |
Feature activation | +3.51e-02 |
Pos logit contributions | |
cout | +0.607 |
naires | +0.579 |
BuzzFeed | +0.565 |
linger | +0.555 |
swe | +0.534 |
Neg logit contributions | |
yip | -0.869 |
zee | -0.738 |
aunders | -0.627 |
zer | -0.611 |
zh | -0.599 |
Token |
|
Token ID | 198 |
Feature activation | -1.24e-01 |
Pos logit contributions | |
yip | +0.097 |
zee | +0.080 |
aunders | +0.065 |
zer | +0.064 |
zh | +0.063 |
Neg logit contributions | |
cout | -0.087 |
naires | -0.083 |
BuzzFeed | -0.082 |
linger | -0.080 |
swe | -0.077 |
Token |
|
Token ID | 198 |
Feature activation | -2.16e-01 |
Pos logit contributions | |
cout | +0.327 |
naires | +0.317 |
BuzzFeed | +0.309 |
linger | +0.305 |
swe | +0.296 |
Neg logit contributions | |
yip | -0.312 |
zee | -0.258 |
aunders | -0.207 |
zer | -0.205 |
zh | -0.199 |
Token | R |
Token ID | 49 |
Feature activation | -5.68e-02 |
Pos logit contributions | |
cout | +0.504 |
naires | +0.487 |
BuzzFeed | +0.473 |
linger | +0.465 |
swe | +0.449 |
Neg logit contributions | |
yip | -0.614 |
zee | -0.515 |
aunders | -0.427 |
zer | -0.421 |
zh | -0.411 |
Token | apt |
Token ID | 2373 |
Feature activation | -1.67e+00 |
Pos logit contributions | |
cout | +0.128 |
naires | +0.123 |
BuzzFeed | +0.120 |
linger | +0.117 |
swe | +0.113 |
Neg logit contributions | |
yip | -0.166 |
zee | -0.140 |
aunders | -0.117 |
zer | -0.116 |
zh | -0.113 |
Token | or |
Token ID | 273 |
Feature activation | -3.85e-01 |
Pos logit contributions | |
cout | +3.532 |
naires | +3.414 |
BuzzFeed | +3.286 |
linger | +3.197 |
omorphic | +3.069 |
Neg logit contributions | |
yip | -5.128 |
zee | -4.270 |
aunders | -3.615 |
zer | -3.493 |
zh | -3.449 |
Token | 's |
Token ID | 338 |
Feature activation | -1.17e-01 |
Pos logit contributions | |
cout | +0.803 |
naires | +0.772 |
BuzzFeed | +0.747 |
linger | +0.733 |
swe | +0.706 |
Neg logit contributions | |
yip | -1.191 |
zee | -1.015 |
aunders | -0.858 |
zer | -0.848 |
zh | -0.829 |
Token | ch |
Token ID | 442 |
Feature activation | +2.21e-01 |
Pos logit contributions | |
cout | +0.226 |
naires | +0.220 |
BuzzFeed | +0.211 |
linger | +0.206 |
swe | +0.197 |
Neg logit contributions | |
yip | -0.374 |
zee | -0.322 |
aunders | -0.275 |
zer | -0.273 |
zh | -0.267 |
Token | is |
Token ID | 271 |
Feature activation | +1.08e-01 |
Pos logit contributions | |
yip | +0.599 |
zee | +0.499 |
aunders | +0.407 |
zer | +0.402 |
zh | +0.391 |
Neg logit contributions | |
cout | -0.546 |
naires | -0.529 |
BuzzFeed | -0.514 |
linger | -0.505 |
swe | -0.490 |
Token | eled |
Token ID | 18449 |
Feature activation | -8.38e-02 |
Pos logit contributions | |
yip | +0.322 |
zee | +0.275 |
aunders | +0.231 |
zer | +0.229 |
zh | +0.223 |
Neg logit contributions | |
cout | -0.237 |
naires | -0.229 |
BuzzFeed | -0.223 |
linger | -0.218 |
swe | -0.211 |
Token | Jack |
Token ID | 3619 |
Feature activation | +2.66e-01 |
Pos logit contributions | |
yip | +0.181 |
zee | +0.156 |
aunders | +0.132 |
zer | +0.130 |
zh | +0.128 |
Neg logit contributions | |
cout | -0.114 |
naires | -0.111 |
BuzzFeed | -0.106 |
linger | -0.104 |
swe | -0.100 |
Token | is |
Token ID | 318 |
Feature activation | -2.65e-01 |
Pos logit contributions | |
yip | +0.791 |
zee | +0.677 |
aunders | +0.569 |
zer | +0.561 |
zh | +0.548 |
Neg logit contributions | |
cout | -0.575 |
naires | -0.554 |
BuzzFeed | -0.537 |
linger | -0.525 |
swe | -0.505 |
Token | clumsy |
Token ID | 39120 |
Feature activation | -1.67e-01 |
Pos logit contributions | |
cout | +0.535 |
naires | +0.515 |
BuzzFeed | +0.496 |
linger | +0.487 |
swe | +0.466 |
Neg logit contributions | |
yip | -0.832 |
zee | -0.715 |
aunders | -0.607 |
zer | -0.601 |
zh | -0.588 |
Token | and |
Token ID | 290 |
Feature activation | +8.05e-03 |
Pos logit contributions | |
cout | +0.350 |
naires | +0.336 |
BuzzFeed | +0.325 |
linger | +0.320 |
swe | +0.308 |
Neg logit contributions | |
yip | -0.514 |
zee | -0.438 |
aunders | -0.370 |
zer | -0.366 |
zh | -0.357 |
Token | gu |
Token ID | 915 |
Feature activation | +3.27e-01 |
Pos logit contributions | |
yip | +0.025 |
zee | +0.021 |
aunders | +0.018 |
zer | +0.018 |
zh | +0.017 |
Neg logit contributions | |
cout | -0.017 |
naires | -0.016 |
BuzzFeed | -0.016 |
linger | -0.016 |
swe | -0.015 |
Token | ile |
Token ID | 576 |
Feature activation | -1.65e+00 |
Pos logit contributions | |
yip | +1.156 |
zee | +1.009 |
aunders | +0.874 |
zer | +0.870 |
zh | +0.853 |
Neg logit contributions | |
cout | -0.540 |
naires | -0.511 |
BuzzFeed | -0.493 |
linger | -0.480 |
swe | -0.455 |
Token | less |
Token ID | 1203 |
Feature activation | -9.32e-02 |
Pos logit contributions | |
cout | +2.240 |
naires | +2.064 |
BuzzFeed | +1.983 |
linger | +1.977 |
swe | +1.870 |
Neg logit contributions | |
yip | -6.243 |
zee | -5.494 |
aunders | -4.926 |
zer | -4.731 |
zh | -4.690 |
Token | , |
Token ID | 11 |
Feature activation | -2.01e-01 |
Pos logit contributions | |
cout | +0.194 |
naires | +0.186 |
BuzzFeed | +0.181 |
linger | +0.177 |
swe | +0.171 |
Neg logit contributions | |
yip | -0.289 |
zee | -0.247 |
aunders | -0.208 |
zer | -0.206 |
zh | -0.202 |
Token | incapable |
Token ID | 23402 |
Feature activation | +2.82e-01 |
Pos logit contributions | |
cout | +0.425 |
naires | +0.407 |
BuzzFeed | +0.395 |
linger | +0.388 |
swe | +0.374 |
Neg logit contributions | |
yip | -0.620 |
zee | -0.528 |
aunders | -0.445 |
zer | -0.440 |
zh | -0.430 |
Token | of |
Token ID | 286 |
Feature activation | -3.89e-01 |
Pos logit contributions | |
yip | +0.914 |
zee | +0.787 |
aunders | +0.671 |
zer | +0.663 |
zh | +0.648 |
Neg logit contributions | |
cout | -0.553 |
naires | -0.528 |
BuzzFeed | -0.510 |
linger | -0.501 |
swe | -0.483 |
Token | managing |
Token ID | 11149 |
Feature activation | +2.10e-01 |
Pos logit contributions | |
cout | +0.806 |
naires | +0.780 |
BuzzFeed | +0.752 |
linger | +0.736 |
swe | +0.707 |
Neg logit contributions | |
yip | -1.196 |
zee | -1.024 |
aunders | -0.865 |
zer | -0.858 |
zh | -0.838 |
Token | building |
Token ID | 2615 |
Feature activation | +5.45e-02 |
Pos logit contributions | |
cout | +0.470 |
naires | +0.455 |
BuzzFeed | +0.434 |
linger | +0.422 |
swe | +0.404 |
Neg logit contributions | |
yip | -0.890 |
zee | -0.772 |
aunders | -0.668 |
zer | -0.662 |
zh | -0.646 |
Token | . |
Token ID | 13 |
Feature activation | -6.29e-02 |
Pos logit contributions | |
yip | +0.163 |
zee | +0.139 |
aunders | +0.117 |
zer | +0.116 |
zh | +0.113 |
Neg logit contributions | |
cout | -0.117 |
naires | -0.113 |
BuzzFeed | -0.110 |
linger | -0.107 |
swe | -0.104 |
Token |
|
Token ID | 198 |
Feature activation | -2.32e-01 |
Pos logit contributions | |
cout | +0.157 |
naires | +0.152 |
BuzzFeed | +0.148 |
linger | +0.146 |
swe | +0.142 |
Neg logit contributions | |
yip | -0.168 |
zee | -0.140 |
aunders | -0.114 |
zer | -0.113 |
zh | -0.109 |
Token |
|
Token ID | 198 |
Feature activation | -2.37e-01 |
Pos logit contributions | |
cout | +0.606 |
naires | +0.587 |
BuzzFeed | +0.572 |
linger | +0.564 |
swe | +0.547 |
Neg logit contributions | |
yip | -0.594 |
zee | -0.490 |
aunders | -0.395 |
zer | -0.390 |
zh | -0.378 |
Token | R |
Token ID | 49 |
Feature activation | -2.47e-01 |
Pos logit contributions | |
cout | +0.573 |
naires | +0.554 |
BuzzFeed | +0.539 |
linger | +0.530 |
swe | +0.513 |
Neg logit contributions | |
yip | -0.653 |
zee | -0.545 |
aunders | -0.449 |
zer | -0.443 |
zh | -0.431 |
Token | end |
Token ID | 437 |
Feature activation | -1.65e+00 |
Pos logit contributions | |
cout | +0.367 |
naires | +0.349 |
BuzzFeed | +0.331 |
linger | +0.322 |
swe | +0.303 |
Neg logit contributions | |
yip | -0.913 |
zee | -0.800 |
aunders | -0.696 |
zer | -0.693 |
zh | -0.679 |
Token | ering |
Token ID | 1586 |
Feature activation | -3.11e-01 |
Pos logit contributions | |
cout | +3.081 |
naires | +2.909 |
BuzzFeed | +2.850 |
linger | +2.743 |
swe | +2.632 |
Neg logit contributions | |
yip | -5.445 |
zee | -4.613 |
aunders | -3.966 |
zer | -3.869 |
zh | -3.809 |
Token | of |
Token ID | 286 |
Feature activation | +1.41e-01 |
Pos logit contributions | |
cout | +0.603 |
naires | +0.572 |
BuzzFeed | +0.557 |
linger | +0.544 |
swe | +0.519 |
Neg logit contributions | |
yip | -1.022 |
zee | -0.876 |
aunders | -0.744 |
zer | -0.734 |
zh | -0.723 |
Token | the |
Token ID | 262 |
Feature activation | -8.15e-02 |
Pos logit contributions | |
yip | +0.422 |
zee | +0.359 |
aunders | +0.301 |
zer | +0.298 |
zh | +0.291 |
Neg logit contributions | |
cout | -0.308 |
naires | -0.298 |
BuzzFeed | -0.287 |
linger | -0.283 |
swe | -0.273 |
Token | northeast |
Token ID | 24287 |
Feature activation | +1.19e-01 |
Pos logit contributions | |
cout | +0.164 |
naires | +0.160 |
BuzzFeed | +0.153 |
linger | +0.150 |
swe | +0.144 |
Neg logit contributions | |
yip | -0.255 |
zee | -0.219 |
aunders | -0.187 |
zer | -0.185 |
zh | -0.180 |
Token | corner |
Token ID | 5228 |
Feature activation | +4.60e-01 |
Pos logit contributions | |
yip | +0.360 |
zee | +0.307 |
aunders | +0.259 |
zer | +0.258 |
zh | +0.250 |
Neg logit contributions | |
cout | -0.254 |
naires | -0.247 |
BuzzFeed | -0.238 |
linger | -0.233 |
swe | -0.224 |
Token | pickups |
Token ID | 46462 |
Feature activation | -1.63e-01 |
Pos logit contributions | |
cout | +0.947 |
naires | +0.910 |
BuzzFeed | +0.878 |
linger | +0.859 |
swe | +0.824 |
Neg logit contributions | |
yip | -1.549 |
zee | -1.331 |
aunders | -1.133 |
zer | -1.124 |
zh | -1.099 |
Token | . |
Token ID | 13 |
Feature activation | -5.46e-02 |
Pos logit contributions | |
cout | +0.293 |
naires | +0.285 |
BuzzFeed | +0.271 |
linger | +0.265 |
swe | +0.252 |
Neg logit contributions | |
yip | -0.542 |
zee | -0.470 |
aunders | -0.403 |
zer | -0.403 |
zh | -0.393 |
Token |
|
Token ID | 198 |
Feature activation | -1.31e-01 |
Pos logit contributions | |
cout | +0.131 |
naires | +0.127 |
BuzzFeed | +0.123 |
linger | +0.121 |
swe | +0.117 |
Neg logit contributions | |
yip | -0.152 |
zee | -0.127 |
aunders | -0.105 |
zer | -0.103 |
zh | -0.100 |
Token |
|
Token ID | 198 |
Feature activation | -2.12e-01 |
Pos logit contributions | |
cout | +0.346 |
naires | +0.334 |
BuzzFeed | +0.326 |
linger | +0.322 |
swe | +0.313 |
Neg logit contributions | |
yip | -0.330 |
zee | -0.273 |
aunders | -0.219 |
zer | -0.217 |
zh | -0.210 |
Token | R |
Token ID | 49 |
Feature activation | -2.84e-01 |
Pos logit contributions | |
cout | +0.489 |
naires | +0.472 |
BuzzFeed | +0.459 |
linger | +0.451 |
swe | +0.436 |
Neg logit contributions | |
yip | -0.607 |
zee | -0.511 |
aunders | -0.424 |
zer | -0.419 |
zh | -0.408 |
Token | apt |
Token ID | 2373 |
Feature activation | -1.64e+00 |
Pos logit contributions | |
cout | +0.580 |
naires | +0.557 |
BuzzFeed | +0.537 |
linger | +0.527 |
swe | +0.505 |
Neg logit contributions | |
yip | -0.888 |
zee | -0.758 |
aunders | -0.637 |
zer | -0.634 |
zh | -0.618 |
Token | or |
Token ID | 273 |
Feature activation | -3.43e-01 |
Pos logit contributions | |
cout | +3.495 |
naires | +3.391 |
BuzzFeed | +3.264 |
linger | +3.176 |
swe | +3.051 |
Neg logit contributions | |
yip | -4.966 |
zee | -4.132 |
aunders | -3.501 |
zer | -3.383 |
zh | -3.342 |
Token | is |
Token ID | 318 |
Feature activation | -2.49e-01 |
Pos logit contributions | |
cout | +0.708 |
naires | +0.680 |
BuzzFeed | +0.658 |
linger | +0.645 |
swe | +0.621 |
Neg logit contributions | |
yip | -1.071 |
zee | -0.914 |
aunders | -0.774 |
zer | -0.766 |
zh | -0.748 |
Token | 6 |
Token ID | 718 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
cout | +0.480 |
naires | +0.466 |
BuzzFeed | +0.446 |
linger | +0.438 |
swe | +0.416 |
Neg logit contributions | |
yip | -0.801 |
zee | -0.691 |
aunders | -0.589 |
zer | -0.586 |
zh | -0.572 |
Token | inches |
Token ID | 8331 |
Feature activation | -4.12e-01 |
Pos logit contributions | |
yip | +0.320 |
zee | +0.272 |
aunders | +0.228 |
zer | +0.225 |
zh | +0.220 |
Neg logit contributions | |
cout | -0.236 |
naires | -0.227 |
BuzzFeed | -0.220 |
linger | -0.216 |
swe | -0.209 |
Token | wider |
Token ID | 10595 |
Feature activation | +2.81e-01 |
Pos logit contributions | |
cout | +0.794 |
naires | +0.764 |
BuzzFeed | +0.736 |
linger | +0.720 |
swe | +0.684 |
Neg logit contributions | |
yip | -1.332 |
zee | -1.149 |
aunders | -0.978 |
zer | -0.972 |
zh | -0.952 |
Token | Smart |
Token ID | 25610 |
Feature activation | -2.69e-01 |
Pos logit contributions | |
cout | +0.586 |
naires | +0.566 |
BuzzFeed | +0.551 |
linger | +0.539 |
swe | +0.522 |
Neg logit contributions | |
yip | -0.708 |
zee | -0.592 |
aunders | -0.491 |
zer | -0.482 |
zh | -0.471 |
Token | est |
Token ID | 395 |
Feature activation | -7.61e-02 |
Pos logit contributions | |
cout | +0.683 |
naires | +0.659 |
BuzzFeed | +0.643 |
linger | +0.634 |
swe | +0.615 |
Neg logit contributions | |
yip | -0.716 |
zee | -0.590 |
aunders | -0.483 |
zer | -0.472 |
zh | -0.460 |
Token |
|
Token ID | 198 |
Feature activation | -2.19e-01 |
Pos logit contributions | |
cout | +0.190 |
naires | +0.184 |
BuzzFeed | +0.179 |
linger | +0.176 |
swe | +0.171 |
Neg logit contributions | |
yip | -0.204 |
zee | -0.170 |
aunders | -0.138 |
zer | -0.137 |
zh | -0.133 |
Token |
|
Token ID | 198 |
Feature activation | -3.64e-01 |
Pos logit contributions | |
cout | +0.576 |
naires | +0.558 |
BuzzFeed | +0.544 |
linger | +0.537 |
swe | +0.521 |
Neg logit contributions | |
yip | -0.556 |
zee | -0.460 |
aunders | -0.369 |
zer | -0.366 |
zh | -0.354 |
Token | R |
Token ID | 49 |
Feature activation | -1.82e-01 |
Pos logit contributions | |
cout | +0.872 |
naires | +0.843 |
BuzzFeed | +0.820 |
linger | +0.804 |
swe | +0.778 |
Neg logit contributions | |
yip | -1.017 |
zee | -0.847 |
aunders | -0.700 |
zer | -0.688 |
zh | -0.670 |
Token | apt |
Token ID | 2373 |
Feature activation | -1.60e+00 |
Pos logit contributions | |
cout | +0.374 |
naires | +0.360 |
BuzzFeed | +0.348 |
linger | +0.341 |
swe | +0.328 |
Neg logit contributions | |
yip | -0.569 |
zee | -0.486 |
aunders | -0.411 |
zer | -0.407 |
zh | -0.398 |
Token | or |
Token ID | 273 |
Feature activation | -4.30e-01 |
Pos logit contributions | |
cout | +3.380 |
naires | +3.288 |
BuzzFeed | +3.170 |
linger | +3.072 |
swe | +2.960 |
Neg logit contributions | |
yip | -4.869 |
zee | -4.049 |
aunders | -3.441 |
zer | -3.317 |
zh | -3.277 |
Token | is |
Token ID | 318 |
Feature activation | -3.43e-01 |
Pos logit contributions | |
cout | +0.893 |
naires | +0.858 |
BuzzFeed | +0.830 |
linger | +0.814 |
swe | +0.785 |
Neg logit contributions | |
yip | -1.335 |
zee | -1.138 |
aunders | -0.963 |
zer | -0.951 |
zh | -0.930 |
Token | powered |
Token ID | 13232 |
Feature activation | +7.17e-02 |
Pos logit contributions | |
cout | +0.668 |
naires | +0.645 |
BuzzFeed | +0.620 |
linger | +0.608 |
swe | +0.581 |
Neg logit contributions | |
yip | -1.102 |
zee | -0.948 |
aunders | -0.807 |
zer | -0.802 |
zh | -0.783 |
Token | by |
Token ID | 416 |
Feature activation | +1.70e-01 |
Pos logit contributions | |
yip | +0.240 |
zee | +0.207 |
aunders | +0.177 |
zer | +0.175 |
zh | +0.172 |
Neg logit contributions | |
cout | -0.133 |
naires | -0.126 |
BuzzFeed | -0.122 |
linger | -0.119 |
swe | -0.115 |
Token | Ford |
Token ID | 8092 |
Feature activation | -9.85e-02 |
Pos logit contributions | |
yip | +0.524 |
zee | +0.448 |
aunders | +0.378 |
zer | +0.376 |
zh | +0.366 |
Neg logit contributions | |
cout | -0.353 |
naires | -0.343 |
BuzzFeed | -0.328 |
linger | -0.324 |
swe | -0.310 |
Token | performance |
Token ID | 2854 |
Feature activation | -5.64e-02 |
Pos logit contributions | |
yip | +0.128 |
zee | +0.109 |
aunders | +0.092 |
zer | +0.090 |
zh | +0.088 |
Neg logit contributions | |
cout | -0.090 |
naires | -0.087 |
BuzzFeed | -0.084 |
linger | -0.083 |
swe | -0.080 |
Token | . |
Token ID | 13 |
Feature activation | -1.47e-02 |
Pos logit contributions | |
cout | +0.119 |
naires | +0.114 |
BuzzFeed | +0.111 |
linger | +0.109 |
swe | +0.105 |
Neg logit contributions | |
yip | -0.173 |
zee | -0.147 |
aunders | -0.124 |
zer | -0.123 |
zh | -0.120 |
Token |
|
Token ID | 198 |
Feature activation | -1.28e-01 |
Pos logit contributions | |
cout | +0.036 |
naires | +0.034 |
BuzzFeed | +0.034 |
linger | +0.033 |
swe | +0.032 |
Neg logit contributions | |
yip | -0.041 |
zee | -0.033 |
aunders | -0.028 |
zer | -0.027 |
zh | -0.026 |
Token |
|
Token ID | 198 |
Feature activation | -1.91e-01 |
Pos logit contributions | |
cout | +0.339 |
naires | +0.328 |
BuzzFeed | +0.320 |
linger | +0.316 |
swe | +0.307 |
Neg logit contributions | |
yip | -0.323 |
zee | -0.268 |
aunders | -0.214 |
zer | -0.213 |
zh | -0.206 |
Token | R |
Token ID | 49 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
cout | +0.451 |
naires | +0.435 |
BuzzFeed | +0.423 |
linger | +0.414 |
swe | +0.401 |
Neg logit contributions | |
yip | -0.538 |
zee | -0.449 |
aunders | -0.372 |
zer | -0.365 |
zh | -0.356 |
Token | apt |
Token ID | 2373 |
Feature activation | -1.58e+00 |
Pos logit contributions | |
cout | +0.364 |
naires | +0.350 |
BuzzFeed | +0.337 |
linger | +0.331 |
swe | +0.317 |
Neg logit contributions | |
yip | -0.565 |
zee | -0.484 |
aunders | -0.408 |
zer | -0.405 |
zh | -0.396 |
Token | or |
Token ID | 273 |
Feature activation | -5.29e-01 |
Pos logit contributions | |
cout | +3.348 |
naires | +3.255 |
BuzzFeed | +3.140 |
linger | +3.045 |
omorphic | +2.923 |
Neg logit contributions | |
yip | -4.808 |
zee | -4.001 |
aunders | -3.400 |
zer | -3.272 |
zh | -3.236 |
Token | 's |
Token ID | 338 |
Feature activation | -1.08e-01 |
Pos logit contributions | |
cout | +1.128 |
naires | +1.084 |
BuzzFeed | +1.050 |
linger | +1.030 |
swe | +0.996 |
Neg logit contributions | |
yip | -1.617 |
zee | -1.373 |
aunders | -1.158 |
zer | -1.142 |
zh | -1.116 |
Token | new |
Token ID | 649 |
Feature activation | +2.38e-01 |
Pos logit contributions | |
cout | +0.215 |
naires | +0.209 |
BuzzFeed | +0.200 |
linger | +0.196 |
swe | +0.187 |
Neg logit contributions | |
yip | -0.341 |
zee | -0.293 |
aunders | -0.249 |
zer | -0.247 |
zh | -0.242 |
Token | transfer |
Token ID | 4351 |
Feature activation | +1.81e-01 |
Pos logit contributions | |
yip | +0.729 |
zee | +0.626 |
aunders | +0.528 |
zer | +0.527 |
zh | +0.513 |
Neg logit contributions | |
cout | -0.490 |
naires | -0.479 |
BuzzFeed | -0.456 |
linger | -0.449 |
swe | -0.427 |
Token | case |
Token ID | 1339 |
Feature activation | -2.85e-02 |
Pos logit contributions | |
yip | +0.540 |
zee | +0.460 |
aunders | +0.387 |
zer | +0.386 |
zh | +0.374 |
Neg logit contributions | |
cout | -0.389 |
naires | -0.379 |
BuzzFeed | -0.366 |
linger | -0.358 |
swe | -0.345 |
Token | whichever |
Token ID | 26204 |
Feature activation | -9.31e-02 |
Pos logit contributions | |
yip | +0.106 |
zee | +0.090 |
aunders | +0.076 |
zer | +0.075 |
zh | +0.074 |
Neg logit contributions | |
cout | -0.075 |
naires | -0.073 |
BuzzFeed | -0.070 |
linger | -0.069 |
swe | -0.066 |
Token | comes |
Token ID | 2058 |
Feature activation | -2.13e-01 |
Pos logit contributions | |
cout | +0.187 |
naires | +0.181 |
BuzzFeed | +0.173 |
linger | +0.170 |
swe | +0.163 |
Neg logit contributions | |
yip | -0.294 |
zee | -0.252 |
aunders | -0.214 |
zer | -0.212 |
zh | -0.208 |
Token | first |
Token ID | 717 |
Feature activation | +3.31e-01 |
Pos logit contributions | |
cout | +0.386 |
naires | +0.372 |
BuzzFeed | +0.356 |
linger | +0.348 |
swe | +0.331 |
Neg logit contributions | |
yip | -0.714 |
zee | -0.618 |
aunders | -0.531 |
zer | -0.528 |
zh | -0.516 |
Token | . |
Token ID | 13 |
Feature activation | -3.32e-02 |
Pos logit contributions | |
yip | +0.981 |
zee | +0.832 |
aunders | +0.696 |
zer | +0.692 |
zh | +0.673 |
Neg logit contributions | |
cout | -0.728 |
naires | -0.704 |
BuzzFeed | -0.680 |
linger | -0.669 |
swe | -0.642 |
Token | Po |
Token ID | 7695 |
Feature activation | +2.34e-01 |
Pos logit contributions | |
cout | +0.082 |
naires | +0.080 |
BuzzFeed | +0.077 |
linger | +0.076 |
swe | +0.074 |
Neg logit contributions | |
yip | -0.089 |
zee | -0.075 |
aunders | -0.061 |
zer | -0.060 |
zh | -0.059 |
Token | etic |
Token ID | 5139 |
Feature activation | -1.57e+00 |
Pos logit contributions | |
yip | +0.600 |
zee | +0.493 |
aunders | +0.398 |
zer | +0.391 |
zh | +0.380 |
Neg logit contributions | |
cout | -0.612 |
naires | -0.593 |
BuzzFeed | -0.578 |
linger | -0.569 |
swe | -0.552 |
Token | , |
Token ID | 11 |
Feature activation | +5.40e-02 |
Pos logit contributions | |
cout | +3.252 |
naires | +3.026 |
BuzzFeed | +3.010 |
linger | +2.957 |
swe | +2.878 |
Neg logit contributions | |
yip | -4.944 |
zee | -4.198 |
aunders | -3.565 |
zer | -3.462 |
zh | -3.419 |
Token | right |
Token ID | 826 |
Feature activation | +2.79e-01 |
Pos logit contributions | |
yip | +0.170 |
zee | +0.145 |
aunders | +0.123 |
zer | +0.122 |
zh | +0.120 |
Neg logit contributions | |
cout | -0.109 |
naires | -0.106 |
BuzzFeed | -0.101 |
linger | -0.099 |
swe | -0.095 |
Token | ? |
Token ID | 30 |
Feature activation | +1.02e-02 |
Pos logit contributions | |
yip | +0.807 |
zee | +0.680 |
aunders | +0.567 |
zer | +0.559 |
zh | +0.545 |
Neg logit contributions | |
cout | -0.639 |
naires | -0.616 |
BuzzFeed | -0.599 |
linger | -0.588 |
swe | -0.568 |
Token |
|
Token ID | 198 |
Feature activation | +9.36e-02 |
Pos logit contributions | |
yip | +0.027 |
zee | +0.023 |
aunders | +0.018 |
zer | +0.018 |
zh | +0.018 |
Neg logit contributions | |
cout | -0.026 |
naires | -0.025 |
BuzzFeed | -0.024 |
linger | -0.024 |
swe | -0.023 |
Token |
|
Token ID | 198 |
Feature activation | -1.81e-03 |
Pos logit contributions | |
yip | +0.227 |
zee | +0.189 |
zer | +0.151 |
aunders | +0.150 |
zh | +0.145 |
Neg logit contributions | |
cout | -0.253 |
naires | -0.244 |
BuzzFeed | -0.237 |
linger | -0.236 |
swe | -0.229 |
Token | design |
Token ID | 1486 |
Feature activation | -1.30e-01 |
Pos logit contributions | |
cout | +0.522 |
naires | +0.509 |
BuzzFeed | +0.486 |
linger | +0.475 |
swe | +0.456 |
Neg logit contributions | |
yip | -0.840 |
zee | -0.723 |
aunders | -0.615 |
zer | -0.613 |
zh | -0.597 |
Token | . |
Token ID | 13 |
Feature activation | -5.67e-02 |
Pos logit contributions | |
cout | +0.266 |
naires | +0.260 |
BuzzFeed | +0.248 |
linger | +0.242 |
swe | +0.233 |
Neg logit contributions | |
yip | -0.402 |
zee | -0.346 |
aunders | -0.293 |
zer | -0.289 |
zh | -0.284 |
Token |
|
Token ID | 198 |
Feature activation | -2.13e-01 |
Pos logit contributions | |
cout | +0.134 |
naires | +0.130 |
BuzzFeed | +0.126 |
linger | +0.125 |
swe | +0.120 |
Neg logit contributions | |
yip | -0.158 |
zee | -0.133 |
aunders | -0.110 |
zer | -0.109 |
zh | -0.106 |
Token |
|
Token ID | 198 |
Feature activation | -2.36e-01 |
Pos logit contributions | |
cout | +0.557 |
naires | +0.540 |
BuzzFeed | +0.526 |
linger | +0.519 |
swe | +0.503 |
Neg logit contributions | |
yip | -0.546 |
zee | -0.450 |
aunders | -0.363 |
zer | -0.359 |
zh | -0.348 |
Token | R |
Token ID | 49 |
Feature activation | -9.69e-02 |
Pos logit contributions | |
cout | +0.561 |
naires | +0.542 |
BuzzFeed | +0.527 |
linger | +0.517 |
swe | +0.500 |
Neg logit contributions | |
yip | -0.662 |
zee | -0.554 |
aunders | -0.458 |
zer | -0.452 |
zh | -0.440 |
Token | end |
Token ID | 437 |
Feature activation | -1.55e+00 |
Pos logit contributions | |
cout | +0.222 |
naires | +0.214 |
BuzzFeed | +0.208 |
linger | +0.204 |
swe | +0.197 |
Neg logit contributions | |
yip | -0.280 |
zee | -0.236 |
aunders | -0.196 |
zer | -0.194 |
zh | -0.189 |
Token | ering |
Token ID | 1586 |
Feature activation | -1.25e-01 |
Pos logit contributions | |
cout | +2.964 |
naires | +2.787 |
BuzzFeed | +2.712 |
linger | +2.643 |
swe | +2.550 |
Neg logit contributions | |
yip | -5.066 |
zee | -4.288 |
aunders | -3.680 |
zer | -3.610 |
zh | -3.528 |
Token | of |
Token ID | 286 |
Feature activation | +1.40e-01 |
Pos logit contributions | |
cout | +0.250 |
naires | +0.239 |
BuzzFeed | +0.231 |
linger | +0.227 |
swe | +0.217 |
Neg logit contributions | |
yip | -0.400 |
zee | -0.342 |
aunders | -0.291 |
zer | -0.288 |
zh | -0.282 |
Token | 1 |
Token ID | 352 |
Feature activation | -2.03e-01 |
Pos logit contributions | |
yip | +0.432 |
zee | +0.369 |
aunders | +0.312 |
zer | +0.310 |
zh | +0.302 |
Neg logit contributions | |
cout | -0.290 |
naires | -0.281 |
BuzzFeed | -0.270 |
linger | -0.265 |
swe | -0.255 |
Token | E |
Token ID | 412 |
Feature activation | -3.06e-02 |
Pos logit contributions | |
cout | +0.475 |
naires | +0.457 |
BuzzFeed | +0.446 |
linger | +0.438 |
swe | +0.423 |
Neg logit contributions | |
yip | -0.578 |
zee | -0.485 |
aunders | -0.402 |
zer | -0.395 |
zh | -0.386 |
Token | gl |
Token ID | 4743 |
Feature activation | -4.20e-01 |
Pos logit contributions | |
cout | +0.083 |
naires | +0.080 |
BuzzFeed | +0.078 |
linger | +0.077 |
swe | +0.075 |
Neg logit contributions | |
yip | -0.075 |
zee | -0.062 |
aunders | -0.049 |
zer | -0.049 |
zh | -0.047 |
Token | - |
Token ID | 12 |
Feature activation | -2.37e-01 |
Pos logit contributions | |
cout | +0.051 |
naires | +0.049 |
BuzzFeed | +0.046 |
linger | +0.045 |
swe | +0.043 |
Neg logit contributions | |
yip | -0.115 |
zee | -0.101 |
aunders | -0.087 |
zer | -0.087 |
zh | -0.085 |
Token | only |
Token ID | 8807 |
Feature activation | +2.51e-01 |
Pos logit contributions | |
cout | +0.648 |
naires | +0.633 |
BuzzFeed | +0.615 |
linger | +0.607 |
swe | +0.590 |
Neg logit contributions | |
yip | -0.575 |
zee | -0.464 |
aunders | -0.372 |
zer | -0.361 |
zh | -0.351 |
Token | lanes |
Token ID | 15296 |
Feature activation | +7.73e-02 |
Pos logit contributions | |
yip | +0.921 |
zee | +0.813 |
aunders | +0.709 |
zer | +0.705 |
zh | +0.693 |
Neg logit contributions | |
cout | -0.371 |
naires | -0.352 |
BuzzFeed | -0.334 |
linger | -0.323 |
swe | -0.304 |
Token | are |
Token ID | 389 |
Feature activation | +1.18e-02 |
Pos logit contributions | |
yip | +0.236 |
zee | +0.203 |
aunders | +0.171 |
zer | +0.169 |
zh | +0.166 |
Neg logit contributions | |
cout | -0.162 |
naires | -0.156 |
BuzzFeed | -0.150 |
linger | -0.146 |
swe | -0.141 |
Token | a |
Token ID | 257 |
Feature activation | -1.28e-02 |
Pos logit contributions | |
yip | +0.037 |
zee | +0.032 |
aunders | +0.027 |
zer | +0.027 |
zh | +0.026 |
Neg logit contributions | |
cout | -0.024 |
naires | -0.023 |
BuzzFeed | -0.022 |
linger | -0.022 |
swe | -0.021 |
Token | pilot |
Token ID | 8022 |
Feature activation | -1.51e+00 |
Pos logit contributions | |
cout | +0.026 |
naires | +0.025 |
BuzzFeed | +0.024 |
linger | +0.023 |
swe | +0.022 |
Neg logit contributions | |
yip | -0.040 |
zee | -0.034 |
zer | -0.029 |
aunders | -0.029 |
zh | -0.028 |
Token | project |
Token ID | 1628 |
Feature activation | -1.44e-01 |
Pos logit contributions | |
cout | +2.821 |
naires | +2.685 |
BuzzFeed | +2.601 |
linger | +2.549 |
swe | +2.444 |
Neg logit contributions | |
yip | -5.029 |
zee | -4.324 |
aunders | -3.712 |
zer | -3.659 |
zh | -3.587 |
Token | of |
Token ID | 286 |
Feature activation | -1.54e-01 |
Pos logit contributions | |
cout | +0.286 |
naires | +0.275 |
BuzzFeed | +0.264 |
linger | +0.260 |
swe | +0.249 |
Neg logit contributions | |
yip | -0.454 |
zee | -0.392 |
aunders | -0.333 |
zer | -0.331 |
zh | -0.324 |
Token | the |
Token ID | 262 |
Feature activation | -1.68e-01 |
Pos logit contributions | |
cout | +0.318 |
naires | +0.306 |
BuzzFeed | +0.294 |
linger | +0.289 |
swe | +0.278 |
Neg logit contributions | |
yip | -0.478 |
zee | -0.410 |
aunders | -0.347 |
zer | -0.343 |
zh | -0.336 |
Token | T |
Token ID | 309 |
Feature activation | -2.45e-01 |
Pos logit contributions | |
cout | +0.343 |
naires | +0.331 |
BuzzFeed | +0.317 |
linger | +0.311 |
swe | +0.299 |
Neg logit contributions | |
yip | -0.524 |
zee | -0.451 |
aunders | -0.383 |
zer | -0.378 |
zh | -0.371 |
Token | EP |
Token ID | 8905 |
Feature activation | -4.58e-01 |
Pos logit contributions | |
cout | +0.563 |
naires | +0.544 |
BuzzFeed | +0.528 |
linger | +0.519 |
swe | +0.501 |
Neg logit contributions | |
yip | -0.707 |
zee | -0.595 |
aunders | -0.495 |
zer | -0.489 |
zh | -0.477 |
Token | + |
Token ID | 1343 |
Feature activation | -3.82e-01 |
Pos logit contributions | |
cout | +0.696 |
naires | +0.667 |
BuzzFeed | +0.644 |
linger | +0.633 |
swe | +0.609 |
Neg logit contributions | |
yip | -1.024 |
zee | -0.872 |
aunders | -0.737 |
zer | -0.725 |
zh | -0.710 |
Token | ambient |
Token ID | 25237 |
Feature activation | +4.60e-01 |
Pos logit contributions | |
cout | +0.803 |
naires | +0.777 |
BuzzFeed | +0.750 |
linger | +0.734 |
swe | +0.708 |
Neg logit contributions | |
yip | -1.173 |
zee | -1.001 |
aunders | -0.845 |
zer | -0.837 |
zh | -0.817 |
Token | D |
Token ID | 35 |
Feature activation | +3.58e-01 |
Pos logit contributions | |
yip | +1.306 |
zee | +1.102 |
aunders | +0.912 |
zer | +0.905 |
zh | +0.882 |
Neg logit contributions | |
cout | -1.070 |
naires | -1.034 |
BuzzFeed | -1.005 |
linger | -0.987 |
swe | -0.956 |
Token | ensity |
Token ID | 6377 |
Feature activation | -2.22e-01 |
Pos logit contributions | |
yip | +1.058 |
zee | +0.906 |
aunders | +0.755 |
zer | +0.749 |
zh | +0.728 |
Neg logit contributions | |
cout | -0.796 |
naires | -0.764 |
BuzzFeed | -0.744 |
linger | -0.729 |
swe | -0.707 |
Token | ) |
Token ID | 1267 |
Feature activation | +4.78e-02 |
Pos logit contributions | |
cout | +0.600 |
naires | +0.582 |
BuzzFeed | +0.567 |
linger | +0.560 |
swe | +0.543 |
Neg logit contributions | |
yip | -0.553 |
zee | -0.451 |
aunders | -0.360 |
zer | -0.353 |
zh | -0.343 |
Token | * |
Token ID | 1635 |
Feature activation | -1.45e+00 |
Pos logit contributions | |
yip | +0.151 |
zee | +0.130 |
aunders | +0.110 |
zer | +0.109 |
zh | +0.107 |
Neg logit contributions | |
cout | -0.095 |
naires | -0.092 |
BuzzFeed | -0.089 |
linger | -0.087 |
swe | -0.084 |
Token | Get |
Token ID | 3497 |
Feature activation | -1.51e-01 |
Pos logit contributions | |
cout | +2.899 |
naires | +2.737 |
BuzzFeed | +2.669 |
linger | +2.631 |
swe | +2.507 |
Neg logit contributions | |
yip | -4.602 |
zee | -3.941 |
aunders | -3.344 |
zer | -3.296 |
zh | -3.241 |
Token | Volume |
Token ID | 31715 |
Feature activation | -2.04e-02 |
Pos logit contributions | |
cout | +0.355 |
naires | +0.343 |
BuzzFeed | +0.333 |
linger | +0.327 |
swe | +0.316 |
Neg logit contributions | |
yip | -0.426 |
zee | -0.356 |
aunders | -0.295 |
zer | -0.290 |
zh | -0.283 |
Token | () |
Token ID | 3419 |
Feature activation | +6.25e-02 |
Pos logit contributions | |
cout | +0.042 |
naires | +0.041 |
BuzzFeed | +0.039 |
linger | +0.039 |
swe | +0.037 |
Neg logit contributions | |
yip | -0.063 |
zee | -0.054 |
aunders | -0.046 |
zer | -0.045 |
zh | -0.044 |
Token | ; |
Token ID | 2162 |
Feature activation | -1.71e-02 |
Pos logit contributions | |
yip | +0.196 |
zee | +0.167 |
aunders | +0.142 |
zer | +0.140 |
zh | +0.137 |
Neg logit contributions | |
cout | -0.127 |
naires | -0.123 |
BuzzFeed | -0.119 |
linger | -0.116 |
swe | -0.112 |
Token | return |
Token ID | 1441 |
Feature activation | -1.57e-01 |
Pos logit contributions | |
naires | +0.029 |
cout | +0.029 |
BuzzFeed | +0.028 |
linger | +0.027 |
swe | +0.026 |
Neg logit contributions | |
yip | -0.058 |
zee | -0.051 |
aunders | -0.044 |
zer | -0.044 |
zh | -0.042 |
Token | critical |
Token ID | 4688 |
Feature activation | -4.17e-01 |
Pos logit contributions | |
cout | +0.375 |
naires | +0.361 |
BuzzFeed | +0.350 |
linger | +0.344 |
swe | +0.331 |
Neg logit contributions | |
yip | -0.528 |
zee | -0.448 |
aunders | -0.378 |
zer | -0.374 |
zh | -0.364 |
Token | illness |
Token ID | 8526 |
Feature activation | -2.07e-01 |
Pos logit contributions | |
cout | +0.929 |
naires | +0.895 |
BuzzFeed | +0.868 |
linger | +0.853 |
swe | +0.822 |
Neg logit contributions | |
yip | -1.234 |
zee | -1.043 |
aunders | -0.873 |
zer | -0.862 |
zh | -0.841 |
Token | - |
Token ID | 12 |
Feature activation | -4.95e-01 |
Pos logit contributions | |
cout | +0.414 |
naires | +0.396 |
BuzzFeed | +0.384 |
linger | +0.376 |
swe | +0.362 |
Neg logit contributions | |
yip | -0.655 |
zee | -0.562 |
aunders | -0.478 |
zer | -0.473 |
zh | -0.463 |
Token | related |
Token ID | 5363 |
Feature activation | -1.56e-01 |
Pos logit contributions | |
cout | +0.972 |
naires | +0.941 |
BuzzFeed | +0.903 |
linger | +0.884 |
swe | +0.844 |
Neg logit contributions | |
yip | -1.586 |
zee | -1.360 |
aunders | -1.159 |
zer | -1.146 |
zh | -1.123 |
Token | cort |
Token ID | 12794 |
Feature activation | -2.92e-02 |
Pos logit contributions | |
cout | +0.370 |
naires | +0.358 |
BuzzFeed | +0.348 |
linger | +0.342 |
swe | +0.331 |
Neg logit contributions | |
yip | -0.439 |
zee | -0.369 |
aunders | -0.304 |
zer | -0.301 |
zh | -0.293 |
Token | ic |
Token ID | 291 |
Feature activation | -1.40e+00 |
Pos logit contributions | |
cout | +0.048 |
naires | +0.045 |
BuzzFeed | +0.044 |
linger | +0.043 |
swe | +0.040 |
Neg logit contributions | |
yip | -0.103 |
zee | -0.090 |
aunders | -0.078 |
zer | -0.078 |
zh | -0.076 |
Token | oster |
Token ID | 6197 |
Feature activation | +9.78e-02 |
Pos logit contributions | |
cout | +3.683 |
naires | +3.596 |
BuzzFeed | +3.462 |
linger | +3.420 |
swe | +3.326 |
Neg logit contributions | |
yip | -3.589 |
zee | -2.922 |
aunders | -2.359 |
zer | -2.295 |
zh | -2.223 |
Token | oid |
Token ID | 1868 |
Feature activation | -4.39e-02 |
Pos logit contributions | |
yip | +0.277 |
zee | +0.232 |
aunders | +0.192 |
zer | +0.190 |
zh | +0.185 |
Neg logit contributions | |
cout | -0.230 |
naires | -0.222 |
BuzzFeed | -0.216 |
linger | -0.212 |
swe | -0.205 |
Token | ins |
Token ID | 1035 |
Feature activation | -8.62e-01 |
Pos logit contributions | |
cout | +0.093 |
naires | +0.089 |
BuzzFeed | +0.087 |
linger | +0.085 |
swe | +0.082 |
Neg logit contributions | |
yip | -0.133 |
zee | -0.114 |
aunders | -0.096 |
zer | -0.095 |
zh | -0.093 |
Token | uff |
Token ID | 1648 |
Feature activation | -1.21e-01 |
Pos logit contributions | |
cout | +1.641 |
naires | +1.577 |
BuzzFeed | +1.517 |
linger | +1.485 |
swe | +1.425 |
Neg logit contributions | |
yip | -2.824 |
zee | -2.428 |
aunders | -2.072 |
zer | -2.045 |
zh | -2.009 |
Token | iciency |
Token ID | 19777 |
Feature activation | -5.57e-01 |
Pos logit contributions | |
cout | +0.236 |
naires | +0.226 |
BuzzFeed | +0.219 |
linger | +0.214 |
swe | +0.205 |
Neg logit contributions | |
yip | -0.389 |
zee | -0.334 |
aunders | -0.285 |
zer | -0.282 |
zh | -0.276 |
Token | : |
Token ID | 1058 |
Feature activation | -2.73e-01 |
Pos logit contributions | |
cout | +0.470 |
naires | +0.455 |
BuzzFeed | +0.443 |
linger | +0.434 |
swe | +0.421 |
Neg logit contributions | |
yip | -0.550 |
zee | -0.459 |
aunders | -0.379 |
zer | -0.373 |
zh | -0.364 |
Token | Do |
Token ID | 2141 |
Feature activation | -3.83e-01 |
Pos logit contributions | |
cout | +0.620 |
naires | +0.599 |
BuzzFeed | +0.579 |
linger | +0.571 |
swe | +0.551 |
Neg logit contributions | |
yip | -0.793 |
zee | -0.671 |
aunders | -0.558 |
zer | -0.552 |
zh | -0.539 |
Token | you |
Token ID | 345 |
Feature activation | -6.20e-01 |
Pos logit contributions | |
cout | +0.748 |
naires | +0.713 |
BuzzFeed | +0.691 |
linger | +0.678 |
swe | +0.651 |
Neg logit contributions | |
yip | -1.240 |
zee | -1.061 |
aunders | -0.909 |
zer | -0.893 |
zh | -0.876 |
Token | have |
Token ID | 423 |
Feature activation | -5.29e-01 |
Pos logit contributions | |
cout | +1.251 |
naires | +1.200 |
BuzzFeed | +1.161 |
linger | +1.139 |
swe | +1.095 |
Neg logit contributions | |
yip | -1.960 |
zee | -1.677 |
aunders | -1.425 |
zer | -1.409 |
zh | -1.377 |
Token | the |
Token ID | 262 |
Feature activation | -3.26e-01 |
Pos logit contributions | |
cout | +1.083 |
naires | +1.037 |
BuzzFeed | +1.006 |
linger | +0.987 |
swe | +0.950 |
Neg logit contributions | |
yip | -1.665 |
zee | -1.422 |
aunders | -1.206 |
zer | -1.192 |
zh | -1.164 |
Token | impression |
Token ID | 10647 |
Feature activation | -1.37e+00 |
Pos logit contributions | |
cout | +0.644 |
naires | +0.616 |
BuzzFeed | +0.597 |
linger | +0.585 |
swe | +0.561 |
Neg logit contributions | |
yip | -1.049 |
zee | -0.899 |
aunders | -0.766 |
zer | -0.757 |
zh | -0.742 |
Token | that |
Token ID | 326 |
Feature activation | -3.08e-01 |
Pos logit contributions | |
cout | +2.467 |
naires | +2.362 |
BuzzFeed | +2.295 |
linger | +2.239 |
swe | +2.143 |
Neg logit contributions | |
yip | -4.606 |
zee | -3.973 |
aunders | -3.439 |
zer | -3.381 |
zh | -3.322 |
Token | over |
Token ID | 625 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
cout | +0.621 |
naires | +0.601 |
BuzzFeed | +0.576 |
linger | +0.566 |
swe | +0.544 |
Neg logit contributions | |
yip | -0.967 |
zee | -0.830 |
aunders | -0.702 |
zer | -0.697 |
zh | -0.682 |
Token | the |
Token ID | 262 |
Feature activation | -7.59e-02 |
Pos logit contributions | |
cout | +0.293 |
naires | +0.281 |
BuzzFeed | +0.273 |
linger | +0.267 |
swe | +0.257 |
Neg logit contributions | |
yip | -0.423 |
zee | -0.360 |
aunders | -0.304 |
zer | -0.300 |
zh | -0.293 |
Token | successive |
Token ID | 25175 |
Feature activation | +2.70e-01 |
Pos logit contributions | |
cout | +0.165 |
naires | +0.158 |
BuzzFeed | +0.154 |
linger | +0.151 |
swe | +0.145 |
Neg logit contributions | |
yip | -0.229 |
zee | -0.195 |
aunders | -0.164 |
zer | -0.161 |
zh | -0.158 |
Token | days |
Token ID | 1528 |
Feature activation | -2.66e-01 |
Pos logit contributions | |
yip | +0.774 |
zee | +0.654 |
aunders | +0.542 |
zer | +0.538 |
zh | +0.524 |
Neg logit contributions | |
cout | -0.618 |
naires | -0.600 |
BuzzFeed | -0.579 |
linger | -0.570 |
swe | -0.549 |
Token | of |
Token ID | 286 |
Feature activation | -3.88e-02 |
Pos logit contributions | |
cout | +0.550 |
naires | +0.528 |
BuzzFeed | +0.512 |
linger | +0.502 |
swe | +0.481 |
Neg logit contributions | |
yip | -0.884 |
zee | -0.759 |
aunders | -0.645 |
zer | -0.638 |
zh | -0.623 |
Token | Stamford |
Token ID | 47604 |
Feature activation | -2.00e-01 |
Pos logit contributions | |
cout | +0.079 |
naires | +0.076 |
BuzzFeed | +0.074 |
linger | +0.072 |
swe | +0.069 |
Neg logit contributions | |
yip | -0.121 |
zee | -0.104 |
aunders | -0.088 |
zer | -0.087 |
zh | -0.085 |
Token | Bridge |
Token ID | 10290 |
Feature activation | +1.32e-01 |
Pos logit contributions | |
cout | +0.336 |
naires | +0.322 |
BuzzFeed | +0.308 |
linger | +0.301 |
swe | +0.286 |
Neg logit contributions | |
yip | -0.696 |
zee | -0.606 |
aunders | -0.524 |
zer | -0.520 |
zh | -0.511 |
Token | , |
Token ID | 11 |
Feature activation | +4.02e-02 |
Pos logit contributions | |
yip | +0.413 |
zee | +0.354 |
aunders | +0.300 |
zer | +0.297 |
zh | +0.292 |
Neg logit contributions | |
cout | -0.267 |
naires | -0.258 |
BuzzFeed | -0.249 |
linger | -0.243 |
swe | -0.233 |
Token | the |
Token ID | 262 |
Feature activation | -1.34e-01 |
Pos logit contributions | |
yip | +0.129 |
zee | +0.112 |
aunders | +0.095 |
zer | +0.094 |
zh | +0.092 |
Neg logit contributions | |
cout | -0.078 |
naires | -0.075 |
BuzzFeed | -0.071 |
linger | -0.070 |
swe | -0.067 |
Token | nort |
Token ID | 17605 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
cout | +0.270 |
naires | +0.261 |
BuzzFeed | +0.250 |
linger | +0.245 |
swe | +0.235 |
Neg logit contributions | |
yip | -0.423 |
zee | -0.364 |
aunders | -0.309 |
zer | -0.306 |
zh | -0.300 |
Token | her |
Token ID | 372 |
Feature activation | -9.25e-01 |
Pos logit contributions | |
cout | +2.651 |
naires | +2.528 |
BuzzFeed | +2.435 |
linger | +2.398 |
swe | +2.301 |
Neg logit contributions | |
yip | -4.345 |
zee | -3.743 |
aunders | -3.206 |
zer | -3.136 |
zh | -3.084 |
Token | ners |
Token ID | 2741 |
Feature activation | -2.38e-01 |
Pos logit contributions | |
cout | +1.932 |
naires | +1.837 |
BuzzFeed | +1.788 |
linger | +1.759 |
swe | +1.695 |
Neg logit contributions | |
yip | -2.867 |
zee | -2.427 |
aunders | -2.066 |
zer | -2.029 |
zh | -1.983 |
Token | were |
Token ID | 547 |
Feature activation | -1.35e-01 |
Pos logit contributions | |
cout | +0.486 |
naires | +0.469 |
BuzzFeed | +0.453 |
linger | +0.441 |
swe | +0.425 |
Neg logit contributions | |
yip | -0.742 |
zee | -0.636 |
aunders | -0.539 |
zer | -0.534 |
zh | -0.522 |
Token | remarkably |
Token ID | 21196 |
Feature activation | +1.87e-01 |
Pos logit contributions | |
cout | +0.274 |
naires | +0.264 |
BuzzFeed | +0.255 |
linger | +0.250 |
swe | +0.240 |
Neg logit contributions | |
yip | -0.425 |
zee | -0.364 |
aunders | -0.309 |
zer | -0.306 |
zh | -0.299 |
Token | absent |
Token ID | 13717 |
Feature activation | -1.73e-01 |
Pos logit contributions | |
yip | +0.566 |
zee | +0.486 |
aunders | +0.409 |
zer | +0.408 |
zh | +0.399 |
Neg logit contributions | |
cout | -0.391 |
naires | -0.379 |
BuzzFeed | -0.365 |
linger | -0.358 |
swe | -0.344 |
Token | Virginia |
Token ID | 6025 |
Feature activation | +1.53e-01 |
Pos logit contributions | |
cout | +0.574 |
naires | +0.562 |
BuzzFeed | +0.527 |
linger | +0.519 |
swe | +0.501 |
Neg logit contributions | |
yip | -0.984 |
zee | -0.850 |
zer | -0.725 |
aunders | -0.724 |
zh | -0.712 |
Token | newspaper |
Token ID | 7533 |
Feature activation | -6.61e-01 |
Pos logit contributions | |
yip | +0.474 |
zee | +0.409 |
aunders | +0.347 |
zer | +0.346 |
zh | +0.339 |
Neg logit contributions | |
cout | -0.307 |
naires | -0.299 |
BuzzFeed | -0.283 |
linger | -0.276 |
swe | -0.269 |
Token | this |
Token ID | 428 |
Feature activation | -1.04e-01 |
Pos logit contributions | |
cout | +1.317 |
naires | +1.266 |
BuzzFeed | +1.220 |
linger | +1.197 |
swe | +1.148 |
Neg logit contributions | |
yip | -2.102 |
zee | -1.805 |
aunders | -1.535 |
zer | -1.522 |
zh | -1.487 |
Token | week |
Token ID | 1285 |
Feature activation | +2.62e-01 |
Pos logit contributions | |
cout | +0.240 |
naires | +0.232 |
BuzzFeed | +0.225 |
linger | +0.221 |
swe | +0.214 |
Neg logit contributions | |
yip | -0.296 |
zee | -0.249 |
aunders | -0.207 |
zer | -0.205 |
zh | -0.200 |
Token | j |
Token ID | 474 |
Feature activation | +1.65e-01 |
Pos logit contributions | |
yip | +0.788 |
zee | +0.676 |
zer | +0.574 |
aunders | +0.572 |
zh | +0.557 |
Neg logit contributions | |
cout | -0.547 |
naires | -0.530 |
BuzzFeed | -0.506 |
linger | -0.499 |
swe | -0.475 |
Token | okingly |
Token ID | 48343 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
yip | +0.473 |
zee | +0.398 |
aunders | +0.329 |
zer | +0.325 |
zh | +0.317 |
Neg logit contributions | |
cout | -0.383 |
naires | -0.371 |
BuzzFeed | -0.360 |
linger | -0.353 |
swe | -0.342 |
Token | cites |
Token ID | 23474 |
Feature activation | -1.78e-01 |
Pos logit contributions | |
cout | +2.820 |
naires | +2.684 |
BuzzFeed | +2.632 |
linger | +2.571 |
swe | +2.493 |
Neg logit contributions | |
yip | -4.167 |
zee | -3.552 |
aunders | -2.992 |
zer | -2.935 |
zh | -2.894 |
Token | a |
Token ID | 257 |
Feature activation | -1.52e-01 |
Pos logit contributions | |
cout | +0.369 |
naires | +0.355 |
BuzzFeed | +0.344 |
linger | +0.337 |
swe | +0.324 |
Neg logit contributions | |
yip | -0.555 |
zee | -0.474 |
aunders | -0.401 |
zer | -0.396 |
zh | -0.387 |
Token | potential |
Token ID | 2785 |
Feature activation | +6.78e-03 |
Pos logit contributions | |
cout | +0.306 |
naires | +0.296 |
BuzzFeed | +0.283 |
linger | +0.278 |
swe | +0.267 |
Neg logit contributions | |
yip | -0.480 |
zee | -0.412 |
aunders | -0.349 |
zer | -0.347 |
zh | -0.339 |
Token | general |
Token ID | 2276 |
Feature activation | +2.02e-01 |
Pos logit contributions | |
yip | +0.021 |
zee | +0.018 |
aunders | +0.015 |
zer | +0.015 |
zh | +0.014 |
Neg logit contributions | |
cout | -0.014 |
naires | -0.014 |
BuzzFeed | -0.013 |
linger | -0.013 |
swe | -0.013 |
Token | election |
Token ID | 3071 |
Feature activation | +1.95e-01 |
Pos logit contributions | |
yip | +0.564 |
zee | +0.473 |
aunders | +0.390 |
zer | +0.386 |
zh | +0.375 |
Neg logit contributions | |
cout | -0.482 |
naires | -0.466 |
BuzzFeed | -0.453 |
linger | -0.445 |
swe | -0.431 |
Token | Performance |
Token ID | 15193 |
Feature activation | +2.83e-02 |
Pos logit contributions | |
yip | +0.784 |
zee | +0.682 |
aunders | +0.576 |
zer | +0.575 |
zh | +0.564 |
Neg logit contributions | |
cout | -0.494 |
naires | -0.480 |
BuzzFeed | -0.461 |
linger | -0.451 |
swe | -0.433 |
Token | . |
Token ID | 13 |
Feature activation | +1.56e-01 |
Pos logit contributions | |
yip | +0.087 |
zee | +0.075 |
aunders | +0.064 |
zer | +0.064 |
zh | +0.062 |
Neg logit contributions | |
cout | -0.058 |
naires | -0.056 |
BuzzFeed | -0.054 |
linger | -0.053 |
swe | -0.051 |
Token |
|
Token ID | 198 |
Feature activation | -1.28e-02 |
Pos logit contributions | |
yip | +0.428 |
zee | +0.353 |
aunders | +0.289 |
zer | +0.281 |
zh | +0.274 |
Neg logit contributions | |
cout | -0.386 |
naires | -0.373 |
BuzzFeed | -0.366 |
linger | -0.356 |
swe | -0.346 |
Token |
|
Token ID | 198 |
Feature activation | -1.04e-01 |
Pos logit contributions | |
cout | +0.034 |
naires | +0.033 |
BuzzFeed | +0.032 |
linger | +0.032 |
swe | +0.031 |
Neg logit contributions | |
yip | -0.032 |
zee | -0.026 |
aunders | -0.021 |
zer | -0.021 |
zh | -0.020 |
Token | R |
Token ID | 49 |
Feature activation | -1.11e-01 |
Pos logit contributions | |
cout | +0.241 |
naires | +0.233 |
BuzzFeed | +0.226 |
linger | +0.222 |
swe | +0.214 |
Neg logit contributions | |
yip | -0.300 |
zee | -0.251 |
aunders | -0.209 |
zer | -0.206 |
zh | -0.201 |
Token | apt |
Token ID | 2373 |
Feature activation | -1.34e+00 |
Pos logit contributions | |
cout | +0.226 |
naires | +0.217 |
BuzzFeed | +0.210 |
linger | +0.206 |
swe | +0.198 |
Neg logit contributions | |
yip | -0.347 |
zee | -0.296 |
aunders | -0.250 |
zer | -0.248 |
zh | -0.242 |
Token | or |
Token ID | 273 |
Feature activation | -3.96e-01 |
Pos logit contributions | |
cout | +2.899 |
naires | +2.816 |
BuzzFeed | +2.719 |
linger | +2.645 |
swe | +2.543 |
Neg logit contributions | |
yip | -4.038 |
zee | -3.364 |
aunders | -2.850 |
zer | -2.753 |
zh | -2.716 |
Token | 's |
Token ID | 338 |
Feature activation | -1.14e-01 |
Pos logit contributions | |
cout | +0.831 |
naires | +0.798 |
BuzzFeed | +0.773 |
linger | +0.758 |
swe | +0.731 |
Neg logit contributions | |
yip | -1.223 |
zee | -1.042 |
aunders | -0.880 |
zer | -0.869 |
zh | -0.850 |
Token | pioneering |
Token ID | 36304 |
Feature activation | +2.81e-01 |
Pos logit contributions | |
cout | +0.224 |
naires | +0.218 |
BuzzFeed | +0.208 |
linger | +0.204 |
swe | +0.195 |
Neg logit contributions | |
yip | -0.362 |
zee | -0.311 |
aunders | -0.265 |
zer | -0.263 |
zh | -0.257 |
Token | off |
Token ID | 572 |
Feature activation | +1.43e-01 |
Pos logit contributions | |
yip | +0.836 |
zee | +0.719 |
zer | +0.602 |
aunders | +0.602 |
zh | +0.583 |
Neg logit contributions | |
cout | -0.593 |
naires | -0.586 |
BuzzFeed | -0.556 |
linger | -0.548 |
swe | -0.524 |
Token | - |
Token ID | 12 |
Feature activation | +5.65e-02 |
Pos logit contributions | |
yip | +0.460 |
zee | +0.394 |
aunders | +0.338 |
zer | +0.331 |
zh | +0.325 |
Neg logit contributions | |
cout | -0.280 |
naires | -0.268 |
BuzzFeed | -0.259 |
linger | -0.255 |
swe | -0.245 |
Token | journey |
Token ID | 7002 |
Feature activation | -2.23e-01 |
Pos logit contributions | |
cout | +0.997 |
naires | +0.958 |
BuzzFeed | +0.927 |
linger | +0.910 |
swe | +0.875 |
Neg logit contributions | |
yip | -1.502 |
zee | -1.283 |
aunders | -1.085 |
zer | -1.073 |
zh | -1.049 |
Token | : |
Token ID | 25 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
cout | +0.449 |
naires | +0.432 |
BuzzFeed | +0.417 |
linger | +0.409 |
swe | +0.392 |
Neg logit contributions | |
yip | -0.704 |
zee | -0.603 |
aunders | -0.512 |
zer | -0.507 |
zh | -0.496 |
Token | travers |
Token ID | 33038 |
Feature activation | +4.59e-01 |
Pos logit contributions | |
yip | +0.351 |
zee | +0.300 |
aunders | +0.254 |
zer | +0.251 |
zh | +0.245 |
Neg logit contributions | |
cout | -0.236 |
naires | -0.227 |
BuzzFeed | -0.219 |
linger | -0.215 |
swe | -0.207 |
Token | ing |
Token ID | 278 |
Feature activation | -3.45e-01 |
Pos logit contributions | |
yip | +1.298 |
zee | +1.091 |
aunders | +0.903 |
zer | +0.893 |
zh | +0.869 |
Neg logit contributions | |
cout | -1.077 |
naires | -1.039 |
BuzzFeed | -1.011 |
linger | -0.994 |
swe | -0.960 |
Token | Ste |
Token ID | 2441 |
Feature activation | -2.76e-01 |
Pos logit contributions | |
cout | +0.721 |
naires | +0.693 |
BuzzFeed | +0.671 |
linger | +0.659 |
swe | +0.635 |
Neg logit contributions | |
yip | -1.068 |
zee | -0.911 |
aunders | -0.769 |
zer | -0.761 |
zh | -0.744 |
Token | ep |
Token ID | 538 |
Feature activation | -1.34e+00 |
Pos logit contributions | |
cout | +0.735 |
naires | +0.714 |
BuzzFeed | +0.696 |
linger | +0.685 |
swe | +0.665 |
Neg logit contributions | |
yip | -0.694 |
zee | -0.567 |
aunders | -0.455 |
zer | -0.445 |
zh | -0.432 |
Token | Mountain |
Token ID | 9638 |
Feature activation | +1.28e-02 |
Pos logit contributions | |
cout | +2.885 |
naires | +2.793 |
BuzzFeed | +2.695 |
linger | +2.642 |
swe | +2.564 |
Neg logit contributions | |
yip | -3.959 |
zee | -3.339 |
aunders | -2.828 |
zer | -2.757 |
zh | -2.712 |
Token | passes |
Token ID | 8318 |
Feature activation | +4.45e-01 |
Pos logit contributions | |
yip | +0.040 |
zee | +0.034 |
aunders | +0.029 |
zer | +0.029 |
zh | +0.028 |
Neg logit contributions | |
cout | -0.026 |
naires | -0.025 |
BuzzFeed | -0.025 |
linger | -0.024 |
swe | -0.023 |
Token | with |
Token ID | 351 |
Feature activation | -1.13e-01 |
Pos logit contributions | |
yip | +1.370 |
zee | +1.176 |
aunders | +0.994 |
zer | +0.989 |
zh | +0.962 |
Neg logit contributions | |
cout | -0.916 |
naires | -0.881 |
BuzzFeed | -0.853 |
linger | -0.831 |
swe | -0.793 |
Token | heavy |
Token ID | 4334 |
Feature activation | +5.02e-03 |
Pos logit contributions | |
cout | +0.236 |
naires | +0.227 |
BuzzFeed | +0.219 |
linger | +0.215 |
swe | +0.207 |
Neg logit contributions | |
yip | -0.350 |
zee | -0.299 |
aunders | -0.253 |
zer | -0.250 |
zh | -0.244 |
Token | loads |
Token ID | 15989 |
Feature activation | +3.11e-03 |
Pos logit contributions | |
yip | +0.016 |
zee | +0.014 |
aunders | +0.012 |
zer | +0.012 |
zh | +0.011 |
Neg logit contributions | |
cout | -0.010 |
naires | -0.009 |
BuzzFeed | -0.009 |
linger | -0.009 |
swe | -0.009 |
Token | its |
Token ID | 663 |
Feature activation | +4.53e-02 |
Pos logit contributions | |
yip | +0.779 |
zee | +0.661 |
aunders | +0.551 |
zer | +0.549 |
zh | +0.532 |
Neg logit contributions | |
cout | -0.589 |
naires | -0.571 |
BuzzFeed | -0.552 |
linger | -0.540 |
swe | -0.521 |
Token | power |
Token ID | 1176 |
Feature activation | +4.16e-01 |
Pos logit contributions | |
yip | +0.138 |
zee | +0.119 |
aunders | +0.100 |
zer | +0.100 |
zh | +0.097 |
Neg logit contributions | |
cout | -0.094 |
naires | -0.092 |
BuzzFeed | -0.088 |
linger | -0.086 |
swe | -0.083 |
Token | train |
Token ID | 27432 |
Feature activation | +2.09e-01 |
Pos logit contributions | |
yip | +1.163 |
zee | +0.979 |
aunders | +0.809 |
zer | +0.802 |
zh | +0.779 |
Neg logit contributions | |
cout | -0.980 |
naires | -0.950 |
BuzzFeed | -0.922 |
linger | -0.906 |
swe | -0.874 |
Token | , |
Token ID | 11 |
Feature activation | +9.57e-02 |
Pos logit contributions | |
yip | +0.634 |
zee | +0.543 |
aunders | +0.459 |
zer | +0.456 |
zh | +0.442 |
Neg logit contributions | |
cout | -0.439 |
naires | -0.427 |
BuzzFeed | -0.412 |
linger | -0.402 |
swe | -0.386 |
Token | dri |
Token ID | 1454 |
Feature activation | +2.60e-01 |
Pos logit contributions | |
yip | +0.292 |
zee | +0.249 |
aunders | +0.210 |
zer | +0.208 |
zh | +0.203 |
Neg logit contributions | |
cout | -0.202 |
naires | -0.195 |
BuzzFeed | -0.188 |
linger | -0.185 |
swe | -0.177 |
Token | vel |
Token ID | 626 |
Feature activation | -1.32e+00 |
Pos logit contributions | |
yip | +0.753 |
zee | +0.639 |
aunders | +0.531 |
zer | +0.529 |
zh | +0.514 |
Neg logit contributions | |
cout | -0.590 |
naires | -0.570 |
BuzzFeed | -0.554 |
linger | -0.543 |
swe | -0.525 |
Token | ine |
Token ID | 500 |
Feature activation | -5.86e-01 |
Pos logit contributions | |
cout | +2.803 |
naires | +2.695 |
BuzzFeed | +2.569 |
linger | +2.539 |
swe | +2.449 |
Neg logit contributions | |
yip | -4.070 |
zee | -3.407 |
aunders | -2.870 |
zer | -2.834 |
zh | -2.770 |
Token | , |
Token ID | 11 |
Feature activation | -1.23e-01 |
Pos logit contributions | |
cout | +1.230 |
naires | +1.177 |
BuzzFeed | +1.138 |
linger | +1.123 |
swe | +1.082 |
Neg logit contributions | |
yip | -1.815 |
zee | -1.546 |
aunders | -1.305 |
zer | -1.286 |
zh | -1.262 |
Token | traction |
Token ID | 23692 |
Feature activation | -5.61e-02 |
Pos logit contributions | |
cout | +0.259 |
naires | +0.249 |
BuzzFeed | +0.241 |
linger | +0.236 |
swe | +0.227 |
Neg logit contributions | |
yip | -0.377 |
zee | -0.322 |
aunders | -0.271 |
zer | -0.269 |
zh | -0.262 |
Token | control |
Token ID | 1630 |
Feature activation | -5.26e-02 |
Pos logit contributions | |
cout | +0.117 |
naires | +0.113 |
BuzzFeed | +0.109 |
linger | +0.107 |
swe | +0.103 |
Neg logit contributions | |
yip | -0.173 |
zee | -0.148 |
aunders | -0.125 |
zer | -0.124 |
zh | -0.121 |
Token | , |
Token ID | 11 |
Feature activation | +1.52e-01 |
Pos logit contributions | |
cout | +0.119 |
naires | +0.115 |
BuzzFeed | +0.111 |
linger | +0.109 |
swe | +0.105 |
Neg logit contributions | |
yip | -0.154 |
zee | -0.130 |
aunders | -0.108 |
zer | -0.107 |
zh | -0.104 |
Token | iana |
Token ID | 7484 |
Feature activation | -3.70e-01 |
Pos logit contributions | |
cout | +0.016 |
naires | +0.015 |
BuzzFeed | +0.015 |
linger | +0.014 |
swe | +0.014 |
Neg logit contributions | |
yip | -0.024 |
zee | -0.020 |
aunders | -0.017 |
zer | -0.017 |
zh | -0.017 |
Token | Mas |
Token ID | 11066 |
Feature activation | -3.67e-01 |
Pos logit contributions | |
cout | +0.788 |
naires | +0.760 |
BuzzFeed | +0.734 |
linger | +0.720 |
swe | +0.695 |
Neg logit contributions | |
yip | -1.126 |
zee | -0.960 |
aunders | -0.808 |
zer | -0.802 |
zh | -0.782 |
Token | l |
Token ID | 75 |
Feature activation | +2.79e-01 |
Pos logit contributions | |
cout | +1.009 |
naires | +0.980 |
BuzzFeed | +0.956 |
linger | +0.942 |
swe | +0.916 |
Neg logit contributions | |
yip | -0.894 |
zee | -0.723 |
aunders | -0.574 |
zer | -0.564 |
zh | -0.546 |
Token | any |
Token ID | 1092 |
Feature activation | +4.87e-01 |
Pos logit contributions | |
yip | +0.780 |
zee | +0.665 |
zer | +0.549 |
aunders | +0.548 |
zh | +0.531 |
Neg logit contributions | |
cout | -0.654 |
naires | -0.635 |
BuzzFeed | -0.616 |
linger | -0.602 |
swe | -0.589 |
Token | Or |
Token ID | 1471 |
Feature activation | -1.55e-01 |
Pos logit contributions | |
yip | +1.495 |
zee | +1.314 |
zer | +1.123 |
aunders | +1.099 |
zh | +1.087 |
Neg logit contributions | |
cout | -0.963 |
naires | -0.937 |
linger | -0.881 |
BuzzFeed | -0.871 |
swe | -0.844 |
Token | phan |
Token ID | 19080 |
Feature activation | -1.32e+00 |
Pos logit contributions | |
cout | +0.335 |
naires | +0.322 |
BuzzFeed | +0.312 |
linger | +0.307 |
swe | +0.295 |
Neg logit contributions | |
yip | -0.469 |
zee | -0.398 |
aunders | -0.335 |
zer | -0.331 |
zh | -0.323 |
Token | Black |
Token ID | 2619 |
Feature activation | +4.34e-01 |
Pos logit contributions | |
cout | +2.362 |
naires | +2.303 |
BuzzFeed | +2.243 |
linger | +2.112 |
swe | +2.058 |
Neg logit contributions | |
yip | -4.506 |
zee | -3.722 |
aunders | -3.308 |
zer | -3.240 |
urai | -3.143 |
Token | BBC |
Token ID | 7823 |
Feature activation | -6.50e-02 |
Pos logit contributions | |
yip | +1.349 |
zee | +1.193 |
zer | +1.016 |
aunders | +1.003 |
zh | +0.991 |
Neg logit contributions | |
cout | -0.841 |
naires | -0.816 |
linger | -0.771 |
BuzzFeed | -0.755 |
swe | -0.738 |
Token | America |
Token ID | 2253 |
Feature activation | -7.10e-02 |
Pos logit contributions | |
cout | +0.128 |
naires | +0.123 |
BuzzFeed | +0.117 |
linger | +0.116 |
swe | +0.112 |
Neg logit contributions | |
yip | -0.207 |
zee | -0.179 |
aunders | -0.152 |
zer | -0.152 |
zh | -0.148 |
Token | Matthew |
Token ID | 9308 |
Feature activation | -1.74e-01 |
Pos logit contributions | |
cout | +0.142 |
naires | +0.136 |
BuzzFeed | +0.130 |
linger | +0.129 |
swe | +0.124 |
Neg logit contributions | |
yip | -0.223 |
zee | -0.193 |
zer | -0.163 |
aunders | -0.163 |
zh | -0.160 |
Token | Mc |
Token ID | 1982 |
Feature activation | -5.00e-01 |
Pos logit contributions | |
cout | +0.230 |
naires | +0.217 |
BuzzFeed | +0.205 |
linger | +0.199 |
swe | +0.186 |
Neg logit contributions | |
yip | -0.669 |
zee | -0.590 |
aunders | -0.519 |
zer | -0.515 |
zh | -0.506 |
Token | out |
Token ID | 503 |
Feature activation | +4.64e-02 |
Pos logit contributions | |
yip | +0.609 |
zee | +0.526 |
aunders | +0.448 |
zer | +0.444 |
zh | +0.434 |
Neg logit contributions | |
cout | -0.364 |
naires | -0.353 |
BuzzFeed | -0.337 |
linger | -0.331 |
swe | -0.317 |
Token | for |
Token ID | 329 |
Feature activation | -4.59e-02 |
Pos logit contributions | |
yip | +0.151 |
zee | +0.131 |
aunders | +0.111 |
zer | +0.110 |
zh | +0.108 |
Neg logit contributions | |
cout | -0.088 |
naires | -0.085 |
BuzzFeed | -0.082 |
linger | -0.080 |
swe | -0.077 |
Token | workers |
Token ID | 3259 |
Feature activation | +1.49e-01 |
Pos logit contributions | |
cout | +0.095 |
naires | +0.091 |
BuzzFeed | +0.088 |
linger | +0.086 |
swe | +0.083 |
Neg logit contributions | |
yip | -0.143 |
zee | -0.122 |
aunders | -0.103 |
zer | -0.102 |
zh | -0.100 |
Token | by |
Token ID | 416 |
Feature activation | +2.95e-01 |
Pos logit contributions | |
yip | +0.460 |
zee | +0.394 |
aunders | +0.333 |
zer | +0.330 |
zh | +0.322 |
Neg logit contributions | |
cout | -0.309 |
naires | -0.297 |
BuzzFeed | -0.287 |
linger | -0.281 |
swe | -0.270 |
Token | sc |
Token ID | 629 |
Feature activation | +1.98e-01 |
Pos logit contributions | |
yip | +0.896 |
zee | +0.762 |
aunders | +0.642 |
zer | +0.636 |
zh | +0.621 |
Neg logit contributions | |
cout | -0.631 |
naires | -0.608 |
BuzzFeed | -0.588 |
linger | -0.577 |
swe | -0.556 |
Token | aring |
Token ID | 1723 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
yip | +0.699 |
zee | +0.608 |
aunders | +0.527 |
zer | +0.522 |
zh | +0.512 |
Neg logit contributions | |
cout | -0.326 |
naires | -0.310 |
BuzzFeed | -0.298 |
linger | -0.290 |
swe | -0.275 |
Token | the |
Token ID | 262 |
Feature activation | +1.35e-01 |
Pos logit contributions | |
cout | +2.899 |
naires | +2.770 |
BuzzFeed | +2.710 |
linger | +2.660 |
swe | +2.567 |
Neg logit contributions | |
yip | -3.847 |
zee | -3.257 |
aunders | -2.738 |
zer | -2.672 |
zh | -2.608 |
Token | living |
Token ID | 2877 |
Feature activation | +5.64e-01 |
Pos logit contributions | |
yip | +0.421 |
zee | +0.360 |
aunders | +0.304 |
zer | +0.301 |
zh | +0.295 |
Neg logit contributions | |
cout | -0.279 |
naires | -0.268 |
BuzzFeed | -0.260 |
linger | -0.254 |
swe | -0.244 |
Token | shit |
Token ID | 7510 |
Feature activation | +1.42e-01 |
Pos logit contributions | |
yip | +1.686 |
zee | +1.450 |
aunders | +1.218 |
zer | +1.211 |
zh | +1.179 |
Neg logit contributions | |
cout | -1.197 |
naires | -1.160 |
BuzzFeed | -1.117 |
linger | -1.103 |
swe | -1.051 |
Token | out |
Token ID | 503 |
Feature activation | +5.53e-02 |
Pos logit contributions | |
yip | +0.480 |
zee | +0.414 |
aunders | +0.358 |
zer | +0.356 |
zh | +0.347 |
Neg logit contributions | |
cout | -0.255 |
naires | -0.243 |
BuzzFeed | -0.234 |
linger | -0.231 |
swe | -0.218 |
Token | of |
Token ID | 286 |
Feature activation | +1.01e-01 |
Pos logit contributions | |
yip | +0.176 |
zee | +0.151 |
aunders | +0.128 |
zer | +0.127 |
zh | +0.124 |
Neg logit contributions | |
cout | -0.111 |
naires | -0.106 |
BuzzFeed | -0.103 |
linger | -0.101 |
swe | -0.097 |
Token | who |
Token ID | 508 |
Feature activation | +3.34e-01 |
Pos logit contributions | |
yip | +0.684 |
zee | +0.584 |
aunders | +0.491 |
zer | +0.488 |
zh | +0.475 |
Neg logit contributions | |
cout | -0.496 |
naires | -0.479 |
BuzzFeed | -0.463 |
linger | -0.452 |
swe | -0.436 |
Token | like |
Token ID | 588 |
Feature activation | -1.69e-01 |
Pos logit contributions | |
yip | +0.995 |
zee | +0.855 |
aunders | +0.713 |
zer | +0.709 |
zh | +0.697 |
Neg logit contributions | |
cout | -0.719 |
naires | -0.701 |
BuzzFeed | -0.672 |
linger | -0.650 |
swe | -0.628 |
Token | theory |
Token ID | 4583 |
Feature activation | -2.10e-02 |
Pos logit contributions | |
cout | +0.332 |
naires | +0.320 |
BuzzFeed | +0.309 |
linger | +0.302 |
swe | +0.289 |
Neg logit contributions | |
yip | -0.543 |
zee | -0.466 |
aunders | -0.397 |
zer | -0.393 |
zh | -0.385 |
Token | should |
Token ID | 815 |
Feature activation | +4.12e-01 |
Pos logit contributions | |
cout | +0.043 |
naires | +0.042 |
BuzzFeed | +0.040 |
linger | +0.039 |
swe | +0.038 |
Neg logit contributions | |
yip | -0.065 |
zee | -0.056 |
aunders | -0.047 |
zer | -0.047 |
zh | -0.046 |
Token | be |
Token ID | 307 |
Feature activation | +1.99e-01 |
Pos logit contributions | |
yip | +1.217 |
zee | +1.041 |
zer | +0.868 |
aunders | +0.867 |
zh | +0.852 |
Neg logit contributions | |
cout | -0.896 |
naires | -0.871 |
BuzzFeed | -0.842 |
linger | -0.805 |
swe | -0.786 |
Token | warned |
Token ID | 7728 |
Feature activation | +1.30e+00 |
Pos logit contributions | |
yip | +0.612 |
zee | +0.523 |
aunders | +0.442 |
zer | +0.437 |
zh | +0.427 |
Neg logit contributions | |
cout | -0.418 |
naires | -0.403 |
BuzzFeed | -0.390 |
linger | -0.382 |
swe | -0.367 |
Token | that |
Token ID | 326 |
Feature activation | +1.96e-01 |
Pos logit contributions | |
yip | +3.979 |
zee | +3.431 |
zer | +2.908 |
aunders | +2.893 |
zh | +2.850 |
Neg logit contributions | |
cout | -2.610 |
naires | -2.509 |
BuzzFeed | -2.444 |
linger | -2.375 |
swe | -2.270 |
Token | the |
Token ID | 262 |
Feature activation | +3.78e-02 |
Pos logit contributions | |
yip | +0.611 |
zee | +0.523 |
aunders | +0.445 |
zer | +0.440 |
zh | +0.431 |
Neg logit contributions | |
cout | -0.395 |
naires | -0.384 |
BuzzFeed | -0.371 |
linger | -0.361 |
swe | -0.346 |
Token | following |
Token ID | 1708 |
Feature activation | +7.12e-01 |
Pos logit contributions | |
yip | +0.117 |
zee | +0.100 |
aunders | +0.084 |
zer | +0.084 |
zh | +0.082 |
Neg logit contributions | |
cout | -0.078 |
naires | -0.075 |
BuzzFeed | -0.073 |
linger | -0.072 |
swe | -0.069 |
Token | is |
Token ID | 318 |
Feature activation | -2.13e-02 |
Pos logit contributions | |
yip | +2.078 |
zee | +1.763 |
zer | +1.485 |
aunders | +1.483 |
zh | +1.446 |
Neg logit contributions | |
cout | -1.544 |
naires | -1.495 |
BuzzFeed | -1.456 |
linger | -1.415 |
swe | -1.369 |
Token | a |
Token ID | 257 |
Feature activation | -1.73e-02 |
Pos logit contributions | |
cout | +0.046 |
naires | +0.044 |
BuzzFeed | +0.043 |
linger | +0.042 |
swe | +0.041 |
Neg logit contributions | |
yip | -0.064 |
zee | -0.055 |
aunders | -0.046 |
zer | -0.045 |
zh | -0.044 |
Token | Chief |
Token ID | 23675 |
Feature activation | -4.66e-01 |
Pos logit contributions | |
yip | +0.358 |
zee | +0.284 |
aunders | +0.215 |
zer | +0.212 |
zh | +0.203 |
Neg logit contributions | |
cout | -0.502 |
naires | -0.489 |
BuzzFeed | -0.478 |
linger | -0.472 |
swe | -0.460 |
Token | " |
Token ID | 366 |
Feature activation | -3.23e-01 |
Pos logit contributions | |
cout | +1.021 |
naires | +0.983 |
BuzzFeed | +0.960 |
linger | +0.940 |
swe | +0.907 |
Neg logit contributions | |
yip | -1.397 |
zee | -1.183 |
aunders | -0.990 |
zer | -0.975 |
zh | -0.951 |
Token | Watch |
Token ID | 10723 |
Feature activation | +2.99e-01 |
Pos logit contributions | |
cout | +0.790 |
naires | +0.764 |
BuzzFeed | +0.745 |
linger | +0.729 |
swe | +0.709 |
Neg logit contributions | |
yip | -0.881 |
zee | -0.733 |
aunders | -0.602 |
zer | -0.591 |
zh | -0.574 |
Token | Bake |
Token ID | 38493 |
Feature activation | +5.98e-01 |
Pos logit contributions | |
yip | +0.960 |
zee | +0.828 |
aunders | +0.704 |
zer | +0.699 |
zh | +0.682 |
Neg logit contributions | |
cout | -0.584 |
naires | -0.563 |
BuzzFeed | -0.540 |
linger | -0.530 |
swe | -0.507 |
Token | mon |
Token ID | 2144 |
Feature activation | +7.86e-01 |
Pos logit contributions | |
yip | +1.751 |
zee | +1.479 |
aunders | +1.237 |
zer | +1.220 |
zh | +1.190 |
Neg logit contributions | |
cout | -1.346 |
naires | -1.297 |
BuzzFeed | -1.260 |
linger | -1.238 |
swe | -1.195 |
Token | og |
Token ID | 519 |
Feature activation | +1.09e+00 |
Pos logit contributions | |
yip | +2.464 |
zee | +2.138 |
zer | +1.806 |
aunders | +1.802 |
zh | +1.762 |
Neg logit contributions | |
cout | -1.577 |
naires | -1.512 |
BuzzFeed | -1.457 |
linger | -1.428 |
swe | -1.371 |
Token | atari |
Token ID | 35554 |
Feature activation | +3.01e-01 |
Pos logit contributions | |
yip | +3.657 |
zee | +3.227 |
zer | +2.785 |
aunders | +2.761 |
zh | +2.682 |
Neg logit contributions | |
cout | -1.953 |
naires | -1.843 |
BuzzFeed | -1.774 |
linger | -1.739 |
swe | -1.654 |
Token | or |
Token ID | 393 |
Feature activation | -2.71e-01 |
Pos logit contributions | |
yip | +0.917 |
zee | +0.783 |
aunders | +0.655 |
zer | +0.654 |
zh | +0.637 |
Neg logit contributions | |
cout | -0.639 |
naires | -0.618 |
BuzzFeed | -0.595 |
linger | -0.582 |
swe | -0.561 |
Token | I |
Token ID | 314 |
Feature activation | +1.53e-01 |
Pos logit contributions | |
cout | +0.534 |
naires | +0.512 |
BuzzFeed | +0.495 |
linger | +0.485 |
swe | +0.465 |
Neg logit contributions | |
yip | -0.871 |
zee | -0.748 |
aunders | -0.637 |
zer | -0.631 |
zh | -0.617 |
Token | will |
Token ID | 481 |
Feature activation | +1.08e-02 |
Pos logit contributions | |
yip | +0.482 |
zee | +0.412 |
aunders | +0.350 |
zer | +0.344 |
zh | +0.336 |
Neg logit contributions | |
cout | -0.313 |
naires | -0.299 |
BuzzFeed | -0.291 |
linger | -0.285 |
swe | -0.274 |
Token | kill |
Token ID | 1494 |
Feature activation | +2.06e-01 |
Pos logit contributions | |
yip | +0.034 |
zee | +0.029 |
aunders | +0.025 |
zer | +0.024 |
zh | +0.024 |
Neg logit contributions | |
cout | -0.022 |
naires | -0.021 |
BuzzFeed | -0.020 |
linger | -0.020 |
swe | -0.019 |
Token | budget |
Token ID | 4466 |
Feature activation | +6.17e-02 |
Pos logit contributions | |
yip | +0.064 |
zee | +0.054 |
aunders | +0.046 |
zer | +0.045 |
zh | +0.044 |
Neg logit contributions | |
cout | -0.047 |
naires | -0.046 |
BuzzFeed | -0.044 |
linger | -0.043 |
swe | -0.042 |
Token | pledge |
Token ID | 13995 |
Feature activation | +1.03e-01 |
Pos logit contributions | |
yip | +0.187 |
zee | +0.159 |
aunders | +0.134 |
zer | +0.133 |
zh | +0.129 |
Neg logit contributions | |
cout | -0.132 |
naires | -0.127 |
BuzzFeed | -0.123 |
linger | -0.121 |
swe | -0.116 |
Token | to |
Token ID | 284 |
Feature activation | +8.18e-02 |
Pos logit contributions | |
yip | +0.326 |
zee | +0.280 |
aunders | +0.237 |
zer | +0.236 |
zh | +0.230 |
Neg logit contributions | |
cout | -0.207 |
naires | -0.199 |
BuzzFeed | -0.192 |
linger | -0.188 |
swe | -0.180 |
Token | give |
Token ID | 1577 |
Feature activation | +1.92e-01 |
Pos logit contributions | |
yip | +0.257 |
zee | +0.220 |
aunders | +0.187 |
zer | +0.185 |
zh | +0.181 |
Neg logit contributions | |
cout | -0.165 |
naires | -0.159 |
BuzzFeed | -0.153 |
linger | -0.150 |
swe | -0.144 |
Token | another |
Token ID | 1194 |
Feature activation | +2.43e-01 |
Pos logit contributions | |
yip | +0.583 |
zee | +0.498 |
aunders | +0.421 |
zer | +0.415 |
zh | +0.405 |
Neg logit contributions | |
cout | -0.406 |
naires | -0.393 |
BuzzFeed | -0.376 |
linger | -0.372 |
swe | -0.356 |
Token | £ |
Token ID | 4248 |
Feature activation | +1.13e+00 |
Pos logit contributions | |
yip | +0.771 |
zee | +0.664 |
aunders | +0.562 |
zer | +0.561 |
zh | +0.548 |
Neg logit contributions | |
cout | -0.480 |
naires | -0.466 |
BuzzFeed | -0.444 |
linger | -0.436 |
swe | -0.416 |
Token | 320 |
Token ID | 19504 |
Feature activation | +3.47e-01 |
Pos logit contributions | |
yip | +2.883 |
zee | +2.402 |
zer | +1.938 |
aunders | +1.936 |
zh | +1.876 |
Neg logit contributions | |
cout | -2.924 |
naires | -2.811 |
BuzzFeed | -2.740 |
linger | -2.739 |
swe | -2.656 |
Token | million |
Token ID | 1510 |
Feature activation | -1.68e-02 |
Pos logit contributions | |
yip | +1.045 |
zee | +0.888 |
aunders | +0.746 |
zer | +0.738 |
zh | +0.720 |
Neg logit contributions | |
cout | -0.753 |
naires | -0.726 |
BuzzFeed | -0.703 |
linger | -0.690 |
swe | -0.664 |
Token | for |
Token ID | 329 |
Feature activation | -2.12e-01 |
Pos logit contributions | |
cout | +0.033 |
naires | +0.032 |
BuzzFeed | +0.030 |
linger | +0.030 |
swe | +0.029 |
Neg logit contributions | |
yip | -0.054 |
zee | -0.047 |
aunders | -0.040 |
zer | -0.039 |
zh | -0.038 |
Token | free |
Token ID | 1479 |
Feature activation | +4.46e-02 |
Pos logit contributions | |
cout | +0.429 |
naires | +0.413 |
BuzzFeed | +0.399 |
linger | +0.391 |
swe | +0.376 |
Neg logit contributions | |
yip | -0.670 |
zee | -0.573 |
aunders | -0.487 |
zer | -0.482 |
zh | -0.471 |
Token | schools |
Token ID | 4266 |
Feature activation | -1.45e-01 |
Pos logit contributions | |
yip | +0.144 |
zee | +0.125 |
aunders | +0.106 |
zer | +0.105 |
zh | +0.103 |
Neg logit contributions | |
cout | -0.086 |
naires | -0.083 |
BuzzFeed | -0.079 |
linger | -0.078 |
swe | -0.075 |
Token | � |
Token ID | 250 |
Feature activation | +2.93e-02 |
Pos logit contributions | |
cout | +0.110 |
naires | +0.107 |
BuzzFeed | +0.104 |
linger | +0.103 |
swe | +0.100 |
Neg logit contributions | |
yip | -0.105 |
zee | -0.086 |
aunders | -0.069 |
zer | -0.068 |
zh | -0.066 |
Token | Ass |
Token ID | 8021 |
Feature activation | +1.46e-01 |
Pos logit contributions | |
yip | +0.080 |
zee | +0.067 |
aunders | +0.055 |
zer | +0.054 |
zh | +0.053 |
Neg logit contributions | |
cout | -0.071 |
naires | -0.069 |
BuzzFeed | -0.067 |
linger | -0.066 |
swe | -0.064 |
Token | imilation |
Token ID | 42963 |
Feature activation | -3.70e-01 |
Pos logit contributions | |
yip | +0.478 |
zee | +0.412 |
aunders | +0.352 |
zer | +0.349 |
zh | +0.341 |
Neg logit contributions | |
cout | -0.279 |
naires | -0.268 |
BuzzFeed | -0.258 |
linger | -0.253 |
swe | -0.243 |
Token | in |
Token ID | 287 |
Feature activation | +1.23e-01 |
Pos logit contributions | |
cout | +0.768 |
naires | +0.740 |
BuzzFeed | +0.716 |
linger | +0.701 |
swe | +0.674 |
Neg logit contributions | |
yip | -1.148 |
zee | -0.980 |
aunders | -0.829 |
zer | -0.819 |
zh | -0.801 |
Token | Biblical |
Token ID | 31752 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
yip | +0.379 |
zee | +0.323 |
aunders | +0.273 |
zer | +0.270 |
zh | +0.264 |
Neg logit contributions | |
cout | -0.257 |
naires | -0.247 |
BuzzFeed | -0.239 |
linger | -0.235 |
swe | -0.226 |
Token | Narr |
Token ID | 28390 |
Feature activation | +1.47e+00 |
Pos logit contributions | |
yip | +0.335 |
zee | +0.290 |
aunders | +0.245 |
zer | +0.244 |
zh | +0.238 |
Neg logit contributions | |
cout | -0.216 |
naires | -0.209 |
BuzzFeed | -0.200 |
linger | -0.197 |
swe | -0.190 |
Token | atives |
Token ID | 2929 |
Feature activation | -6.95e-03 |
Pos logit contributions | |
yip | +4.174 |
zee | +3.626 |
zer | +3.006 |
aunders | +3.003 |
zh | +2.933 |
Neg logit contributions | |
cout | -3.268 |
naires | -3.158 |
BuzzFeed | -3.050 |
linger | -3.027 |
swe | -2.910 |
Token | . |
Token ID | 13 |
Feature activation | -1.24e-02 |
Pos logit contributions | |
cout | +0.015 |
naires | +0.015 |
BuzzFeed | +0.014 |
linger | +0.014 |
swe | +0.014 |
Neg logit contributions | |
yip | -0.021 |
zee | -0.017 |
aunders | -0.015 |
zer | -0.014 |
zh | -0.014 |
Token | � |
Token ID | 447 |
Feature activation | +7.42e-02 |
Pos logit contributions | |
cout | +0.026 |
naires | +0.025 |
BuzzFeed | +0.024 |
linger | +0.024 |
swe | +0.023 |
Neg logit contributions | |
yip | -0.038 |
zee | -0.033 |
aunders | -0.028 |
zer | -0.028 |
zh | -0.027 |
Token | � |
Token ID | 251 |
Feature activation | +6.59e-02 |
Pos logit contributions | |
yip | +0.187 |
zee | +0.153 |
aunders | +0.122 |
zer | +0.120 |
zh | +0.116 |
Neg logit contributions | |
cout | -0.198 |
naires | -0.192 |
BuzzFeed | -0.189 |
linger | -0.185 |
swe | -0.180 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.81e-01 |
Pos logit contributions | |
yip | +0.197 |
zee | +0.168 |
aunders | +0.141 |
zer | +0.140 |
zh | +0.136 |
Neg logit contributions | |
cout | -0.143 |
naires | -0.138 |
BuzzFeed | -0.133 |
linger | -0.132 |
swe | -0.127 |
Token |
|
Token ID | 198 |
Feature activation | +8.73e-02 |
Pos logit contributions | |
yip | +0.652 |
zee | +0.545 |
zer | +0.433 |
aunders | +0.429 |
zh | +0.418 |
Neg logit contributions | |
cout | -0.736 |
naires | -0.710 |
BuzzFeed | -0.692 |
linger | -0.689 |
swe | -0.669 |
Token | � |
Token ID | 447 |
Feature activation | +9.61e-02 |
Pos logit contributions | |
yip | +0.243 |
zee | +0.205 |
zer | +0.169 |
aunders | +0.169 |
zh | +0.164 |
Neg logit contributions | |
cout | -0.208 |
naires | -0.200 |
BuzzFeed | -0.195 |
linger | -0.192 |
swe | -0.186 |
Token | � |
Token ID | 250 |
Feature activation | +2.04e-02 |
Pos logit contributions | |
yip | +0.254 |
zee | +0.210 |
aunders | +0.171 |
zer | +0.168 |
zh | +0.164 |
Neg logit contributions | |
cout | -0.244 |
naires | -0.236 |
BuzzFeed | -0.230 |
linger | -0.227 |
swe | -0.220 |
Token | It |
Token ID | 1026 |
Feature activation | +1.44e-01 |
Pos logit contributions | |
yip | +0.057 |
zee | +0.048 |
aunders | +0.039 |
zer | +0.039 |
zh | +0.038 |
Neg logit contributions | |
cout | -0.048 |
naires | -0.047 |
BuzzFeed | -0.046 |
linger | -0.045 |
swe | -0.043 |
Token | throws |
Token ID | 12542 |
Feature activation | +4.53e-01 |
Pos logit contributions | |
yip | +0.444 |
zee | +0.379 |
aunders | +0.320 |
zer | +0.318 |
zh | +0.311 |
Neg logit contributions | |
cout | -0.301 |
naires | -0.290 |
BuzzFeed | -0.281 |
linger | -0.275 |
swe | -0.264 |
Token | away |
Token ID | 1497 |
Feature activation | +1.18e+00 |
Pos logit contributions | |
yip | +1.374 |
zee | +1.165 |
aunders | +0.987 |
zer | +0.980 |
zh | +0.956 |
Neg logit contributions | |
cout | -0.957 |
naires | -0.925 |
BuzzFeed | -0.890 |
linger | -0.875 |
swe | -0.840 |
Token | the |
Token ID | 262 |
Feature activation | +1.70e-01 |
Pos logit contributions | |
yip | +3.473 |
zee | +2.960 |
aunders | +2.495 |
zer | +2.476 |
zh | +2.416 |
Neg logit contributions | |
cout | -2.514 |
naires | -2.461 |
BuzzFeed | -2.344 |
linger | -2.297 |
swe | -2.216 |
Token | European |
Token ID | 3427 |
Feature activation | +1.17e-01 |
Pos logit contributions | |
yip | +0.514 |
zee | +0.439 |
aunders | +0.369 |
zer | +0.366 |
zh | +0.358 |
Neg logit contributions | |
cout | -0.360 |
naires | -0.350 |
BuzzFeed | -0.336 |
linger | -0.329 |
swe | -0.317 |
Token | rule |
Token ID | 3896 |
Feature activation | +2.63e-01 |
Pos logit contributions | |
yip | +0.359 |
zee | +0.307 |
aunders | +0.259 |
zer | +0.256 |
zh | +0.251 |
Neg logit contributions | |
cout | -0.245 |
naires | -0.237 |
BuzzFeed | -0.228 |
linger | -0.223 |
swe | -0.215 |
Token | book |
Token ID | 1492 |
Feature activation | -4.27e-01 |
Pos logit contributions | |
yip | +0.826 |
zee | +0.710 |
aunders | +0.598 |
zer | +0.596 |
zh | +0.584 |
Neg logit contributions | |
cout | -0.532 |
naires | -0.510 |
BuzzFeed | -0.495 |
linger | -0.486 |
swe | -0.466 |
Token | in |
Token ID | 287 |
Feature activation | +1.45e-01 |
Pos logit contributions | |
cout | +0.864 |
naires | +0.831 |
BuzzFeed | +0.804 |
linger | +0.787 |
swe | +0.756 |
Neg logit contributions | |
yip | -1.345 |
zee | -1.153 |
aunders | -0.978 |
zer | -0.969 |
zh | -0.947 |
Token | we |
Token ID | 356 |
Feature activation | -1.65e-01 |
Pos logit contributions | |
yip | +0.566 |
zee | +0.490 |
aunders | +0.415 |
zer | +0.414 |
zh | +0.404 |
Neg logit contributions | |
cout | -0.355 |
naires | -0.346 |
BuzzFeed | -0.330 |
linger | -0.323 |
swe | -0.311 |
Token | have |
Token ID | 423 |
Feature activation | +2.58e-02 |
Pos logit contributions | |
cout | +0.353 |
naires | +0.341 |
BuzzFeed | +0.330 |
linger | +0.322 |
swe | +0.311 |
Neg logit contributions | |
yip | -0.500 |
zee | -0.427 |
aunders | -0.359 |
zer | -0.356 |
zh | -0.348 |
Token | at |
Token ID | 379 |
Feature activation | -3.51e-02 |
Pos logit contributions | |
yip | +0.078 |
zee | +0.067 |
aunders | +0.056 |
zer | +0.056 |
zh | +0.054 |
Neg logit contributions | |
cout | -0.055 |
naires | -0.053 |
BuzzFeed | -0.051 |
linger | -0.050 |
swe | -0.048 |
Token | least |
Token ID | 1551 |
Feature activation | +1.79e-01 |
Pos logit contributions | |
cout | +0.079 |
naires | +0.076 |
BuzzFeed | +0.074 |
linger | +0.072 |
swe | +0.070 |
Neg logit contributions | |
yip | -0.103 |
zee | -0.087 |
aunders | -0.073 |
zer | -0.072 |
zh | -0.070 |
Token | two |
Token ID | 734 |
Feature activation | +2.41e-02 |
Pos logit contributions | |
yip | +0.552 |
zee | +0.474 |
aunders | +0.400 |
zer | +0.398 |
zh | +0.389 |
Neg logit contributions | |
cout | -0.368 |
naires | -0.356 |
BuzzFeed | -0.342 |
linger | -0.336 |
swe | -0.322 |
Token | competitors |
Token ID | 13861 |
Feature activation | +2.79e-01 |
Pos logit contributions | |
yip | +0.075 |
zee | +0.064 |
aunders | +0.054 |
zer | +0.054 |
zh | +0.053 |
Neg logit contributions | |
cout | -0.049 |
naires | -0.047 |
BuzzFeed | -0.046 |
linger | -0.045 |
swe | -0.043 |
Token | , |
Token ID | 11 |
Feature activation | -5.32e-03 |
Pos logit contributions | |
yip | +0.873 |
zee | +0.756 |
aunders | +0.643 |
zer | +0.640 |
zh | +0.622 |
Neg logit contributions | |
cout | -0.556 |
naires | -0.539 |
BuzzFeed | -0.517 |
linger | -0.505 |
swe | -0.486 |
Token | not |
Token ID | 407 |
Feature activation | +1.05e-01 |
Pos logit contributions | |
cout | +0.011 |
naires | +0.011 |
BuzzFeed | +0.010 |
linger | +0.010 |
swe | +0.010 |
Neg logit contributions | |
yip | -0.016 |
zee | -0.014 |
aunders | -0.012 |
zer | -0.012 |
zh | -0.012 |
Token | no |
Token ID | 645 |
Feature activation | -1.98e-01 |
Pos logit contributions | |
yip | +0.316 |
zee | +0.269 |
aunders | +0.225 |
zer | +0.224 |
zh | +0.219 |
Neg logit contributions | |
cout | -0.227 |
naires | -0.220 |
BuzzFeed | -0.212 |
linger | -0.208 |
swe | -0.200 |
Token | more |
Token ID | 517 |
Feature activation | +1.01e-01 |
Pos logit contributions | |
cout | +0.414 |
naires | +0.398 |
BuzzFeed | +0.385 |
linger | +0.378 |
swe | +0.364 |
Neg logit contributions | |
yip | -0.611 |
zee | -0.520 |
aunders | -0.440 |
zer | -0.434 |
zh | -0.424 |
Token | than |
Token ID | 621 |
Feature activation | -3.60e-01 |
Pos logit contributions | |
yip | +0.323 |
zee | +0.278 |
aunders | +0.237 |
zer | +0.235 |
zh | +0.230 |
Neg logit contributions | |
cout | -0.197 |
naires | -0.189 |
BuzzFeed | -0.183 |
linger | -0.179 |
swe | -0.171 |
Token | the |
Token ID | 262 |
Feature activation | -1.37e-02 |
Pos logit contributions | |
yip | +0.431 |
zee | +0.369 |
aunders | +0.313 |
zer | +0.309 |
zh | +0.301 |
Neg logit contributions | |
cout | -0.296 |
naires | -0.287 |
BuzzFeed | -0.273 |
linger | -0.269 |
swe | -0.260 |
Token | newest |
Token ID | 15530 |
Feature activation | +2.24e-01 |
Pos logit contributions | |
cout | +0.024 |
naires | +0.023 |
BuzzFeed | +0.022 |
linger | +0.021 |
swe | +0.020 |
Neg logit contributions | |
yip | -0.046 |
zee | -0.040 |
zer | -0.035 |
aunders | -0.035 |
zh | -0.034 |
Token | information |
Token ID | 1321 |
Feature activation | -4.05e-01 |
Pos logit contributions | |
yip | +0.717 |
zee | +0.631 |
zer | +0.539 |
aunders | +0.537 |
zh | +0.527 |
Neg logit contributions | |
cout | -0.413 |
naires | -0.399 |
BuzzFeed | -0.375 |
linger | -0.368 |
swe | -0.353 |
Token | from |
Token ID | 422 |
Feature activation | +6.07e-02 |
Pos logit contributions | |
cout | +0.845 |
naires | +0.816 |
BuzzFeed | +0.777 |
linger | +0.769 |
swe | +0.740 |
Neg logit contributions | |
yip | -1.243 |
zee | -1.065 |
aunders | -0.893 |
zer | -0.891 |
zh | -0.865 |
Token | the |
Token ID | 262 |
Feature activation | -6.34e-02 |
Pos logit contributions | |
yip | +0.191 |
zee | +0.165 |
aunders | +0.140 |
zer | +0.139 |
zh | +0.136 |
Neg logit contributions | |
cout | -0.121 |
naires | -0.116 |
BuzzFeed | -0.111 |
linger | -0.109 |
swe | -0.106 |
Token | set |
Token ID | 900 |
Feature activation | +3.88e-01 |
Pos logit contributions | |
cout | +0.110 |
naires | +0.107 |
BuzzFeed | +0.101 |
linger | +0.098 |
swe | +0.094 |
Neg logit contributions | |
yip | -0.215 |
zee | -0.188 |
aunders | -0.162 |
zer | -0.161 |
zh | -0.158 |
Token | of |
Token ID | 286 |
Feature activation | +1.15e-01 |
Pos logit contributions | |
yip | +1.134 |
zee | +0.969 |
zer | +0.806 |
aunders | +0.802 |
zh | +0.776 |
Neg logit contributions | |
cout | -0.856 |
naires | -0.825 |
BuzzFeed | -0.799 |
linger | -0.784 |
swe | -0.763 |
Token | TM |
Token ID | 21232 |
Feature activation | -5.11e-01 |
Pos logit contributions | |
yip | +0.381 |
zee | +0.330 |
aunders | +0.282 |
zer | +0.281 |
zh | +0.275 |
Neg logit contributions | |
cout | -0.213 |
naires | -0.205 |
BuzzFeed | -0.196 |
linger | -0.192 |
swe | -0.184 |
Token | NT |
Token ID | 11251 |
Feature activation | +4.71e-02 |
Pos logit contributions | |
cout | +1.128 |
naires | +1.083 |
BuzzFeed | +1.048 |
linger | +1.033 |
swe | +0.998 |
Neg logit contributions | |
yip | -1.514 |
zee | -1.295 |
aunders | -1.084 |
zer | -1.073 |
zh | -1.050 |
Token | II |
Token ID | 2873 |
Feature activation | +1.15e-01 |
Pos logit contributions | |
yip | +0.136 |
zee | +0.116 |
aunders | +0.096 |
zer | +0.095 |
zh | +0.093 |
Neg logit contributions | |
cout | -0.106 |
naires | -0.103 |
BuzzFeed | -0.100 |
linger | -0.097 |
swe | -0.094 |
Token | ? |
Token ID | 30 |
Feature activation | +1.79e-01 |
Pos logit contributions | |
yip | +0.312 |
zee | +0.262 |
aunders | +0.214 |
zer | +0.213 |
zh | +0.207 |
Neg logit contributions | |
cout | -0.280 |
naires | -0.270 |
BuzzFeed | -0.264 |
linger | -0.258 |
swe | -0.251 |
Token | away |
Token ID | 1497 |
Feature activation | -3.21e-01 |
Pos logit contributions | |
cout | +0.736 |
naires | +0.709 |
BuzzFeed | +0.684 |
linger | +0.671 |
swe | +0.644 |
Neg logit contributions | |
yip | -1.129 |
zee | -0.966 |
aunders | -0.818 |
zer | -0.812 |
zh | -0.792 |
Token | to |
Token ID | 284 |
Feature activation | -1.71e-01 |
Pos logit contributions | |
cout | +0.652 |
naires | +0.626 |
BuzzFeed | +0.606 |
linger | +0.594 |
swe | +0.571 |
Neg logit contributions | |
yip | -1.008 |
zee | -0.862 |
aunders | -0.732 |
zer | -0.723 |
zh | -0.708 |
Token | concern |
Token ID | 2328 |
Feature activation | -2.23e-01 |
Pos logit contributions | |
cout | +0.355 |
naires | +0.342 |
BuzzFeed | +0.332 |
linger | +0.318 |
swe | +0.311 |
Neg logit contributions | |
yip | -0.524 |
zee | -0.447 |
aunders | -0.378 |
zer | -0.376 |
zh | -0.367 |
Token | them |
Token ID | 606 |
Feature activation | -4.53e-02 |
Pos logit contributions | |
cout | +0.508 |
naires | +0.490 |
BuzzFeed | +0.476 |
linger | +0.467 |
swe | +0.452 |
Neg logit contributions | |
yip | -0.646 |
zee | -0.545 |
aunders | -0.454 |
zer | -0.449 |
zh | -0.438 |
Token | . |
Token ID | 13 |
Feature activation | +2.09e-02 |
Pos logit contributions | |
cout | +0.097 |
naires | +0.093 |
BuzzFeed | +0.090 |
linger | +0.088 |
swe | +0.085 |
Neg logit contributions | |
yip | -0.138 |
zee | -0.117 |
aunders | -0.099 |
zer | -0.098 |
zh | -0.096 |
Token | After |
Token ID | 2293 |
Feature activation | +1.76e-01 |
Pos logit contributions | |
yip | +0.057 |
zee | +0.048 |
aunders | +0.040 |
zer | +0.039 |
zh | +0.038 |
Neg logit contributions | |
cout | -0.050 |
naires | -0.049 |
BuzzFeed | -0.047 |
linger | -0.047 |
swe | -0.045 |
Token | all |
Token ID | 477 |
Feature activation | +6.07e-01 |
Pos logit contributions | |
yip | +0.545 |
zee | +0.466 |
aunders | +0.394 |
zer | +0.391 |
zh | +0.382 |
Neg logit contributions | |
cout | -0.362 |
naires | -0.348 |
BuzzFeed | -0.336 |
linger | -0.329 |
swe | -0.316 |
Token | , |
Token ID | 11 |
Feature activation | +5.83e-03 |
Pos logit contributions | |
yip | +1.907 |
zee | +1.632 |
aunders | +1.383 |
zer | +1.369 |
zh | +1.339 |
Neg logit contributions | |
cout | -1.237 |
naires | -1.189 |
BuzzFeed | -1.150 |
linger | -1.127 |
swe | -1.083 |
Token | the |
Token ID | 262 |
Feature activation | -1.18e-01 |
Pos logit contributions | |
yip | +0.018 |
zee | +0.016 |
aunders | +0.013 |
zer | +0.013 |
zh | +0.013 |
Neg logit contributions | |
cout | -0.012 |
naires | -0.011 |
BuzzFeed | -0.011 |
linger | -0.011 |
swe | -0.010 |
Token | populace |
Token ID | 39003 |
Feature activation | +1.15e-01 |
Pos logit contributions | |
cout | +0.244 |
naires | +0.235 |
BuzzFeed | +0.227 |
linger | +0.223 |
swe | +0.214 |
Neg logit contributions | |
yip | -0.366 |
zee | -0.312 |
aunders | -0.264 |
zer | -0.261 |
zh | -0.256 |
Token | was |
Token ID | 373 |
Feature activation | -8.91e-02 |
Pos logit contributions | |
yip | +0.355 |
zee | +0.304 |
aunders | +0.257 |
zer | +0.256 |
zh | +0.249 |
Neg logit contributions | |
cout | -0.236 |
naires | -0.225 |
BuzzFeed | -0.219 |
linger | -0.213 |
swe | -0.204 |
Token | entire |
Token ID | 2104 |
Feature activation | -2.88e-02 |
Pos logit contributions | |
cout | +0.143 |
naires | +0.137 |
BuzzFeed | +0.133 |
linger | +0.130 |
swe | +0.125 |
Neg logit contributions | |
yip | -0.222 |
zee | -0.190 |
aunders | -0.162 |
zer | -0.160 |
zh | -0.156 |
Token | F |
Token ID | 376 |
Feature activation | +2.75e-01 |
Pos logit contributions | |
cout | +0.066 |
naires | +0.064 |
BuzzFeed | +0.062 |
linger | +0.061 |
swe | +0.059 |
Neg logit contributions | |
yip | -0.082 |
zee | -0.069 |
aunders | -0.058 |
zer | -0.057 |
zh | -0.056 |
Token | 1 |
Token ID | 16 |
Feature activation | +4.57e-01 |
Pos logit contributions | |
yip | +0.745 |
zee | +0.625 |
aunders | +0.511 |
zer | +0.510 |
zh | +0.498 |
Neg logit contributions | |
cout | -0.675 |
naires | -0.653 |
BuzzFeed | -0.633 |
linger | -0.622 |
swe | -0.604 |
Token | season |
Token ID | 1622 |
Feature activation | +1.76e-01 |
Pos logit contributions | |
yip | +1.250 |
zee | +1.043 |
aunders | +0.857 |
zer | +0.851 |
zh | +0.829 |
Neg logit contributions | |
cout | -1.102 |
naires | -1.073 |
BuzzFeed | -1.041 |
linger | -1.019 |
swe | -0.988 |
Token | after |
Token ID | 706 |
Feature activation | +3.15e-01 |
Pos logit contributions | |
yip | +0.539 |
zee | +0.459 |
aunders | +0.388 |
zer | +0.386 |
zh | +0.377 |
Neg logit contributions | |
cout | -0.368 |
naires | -0.356 |
BuzzFeed | -0.344 |
linger | -0.336 |
swe | -0.324 |
Token | suffering |
Token ID | 7195 |
Feature activation | +2.98e-01 |
Pos logit contributions | |
yip | +0.957 |
zee | +0.817 |
aunders | +0.692 |
zer | +0.690 |
zh | +0.671 |
Neg logit contributions | |
cout | -0.661 |
naires | -0.635 |
BuzzFeed | -0.612 |
linger | -0.600 |
swe | -0.579 |
Token | multiple |
Token ID | 3294 |
Feature activation | -8.90e-02 |
Pos logit contributions | |
yip | +0.881 |
zee | +0.745 |
aunders | +0.625 |
zer | +0.619 |
zh | +0.604 |
Neg logit contributions | |
cout | -0.657 |
naires | -0.634 |
BuzzFeed | -0.615 |
linger | -0.603 |
swe | -0.582 |
Token | fractures |
Token ID | 39381 |
Feature activation | -2.50e-01 |
Pos logit contributions | |
cout | +0.196 |
naires | +0.189 |
BuzzFeed | +0.183 |
linger | +0.180 |
swe | +0.173 |
Neg logit contributions | |
yip | -0.265 |
zee | -0.225 |
aunders | -0.188 |
zer | -0.186 |
zh | -0.182 |
Token | and |
Token ID | 290 |
Feature activation | -1.69e-01 |
Pos logit contributions | |
cout | +0.470 |
naires | +0.450 |
BuzzFeed | +0.434 |
linger | +0.424 |
swe | +0.406 |
Neg logit contributions | |
yip | -0.826 |
zee | -0.713 |
aunders | -0.610 |
zer | -0.604 |
zh | -0.592 |
Token | a |
Token ID | 257 |
Feature activation | -1.61e-01 |
Pos logit contributions | |
cout | +0.346 |
naires | +0.332 |
BuzzFeed | +0.322 |
linger | +0.315 |
swe | +0.303 |
Neg logit contributions | |
yip | -0.525 |
zee | -0.449 |
aunders | -0.382 |
zer | -0.379 |
zh | -0.371 |
Token | partially |
Token ID | 12387 |
Feature activation | -4.08e-01 |
Pos logit contributions | |
cout | +0.305 |
naires | +0.293 |
BuzzFeed | +0.282 |
linger | +0.275 |
swe | +0.265 |
Neg logit contributions | |
yip | -0.525 |
zee | -0.452 |
aunders | -0.388 |
zer | -0.387 |
zh | -0.379 |
Token | and |
Token ID | 290 |
Feature activation | +7.64e-02 |
Pos logit contributions | |
cout | +0.055 |
naires | +0.053 |
BuzzFeed | +0.051 |
linger | +0.050 |
swe | +0.048 |
Neg logit contributions | |
yip | -0.082 |
zee | -0.070 |
aunders | -0.059 |
zer | -0.059 |
zh | -0.058 |
Token | offered |
Token ID | 4438 |
Feature activation | -2.81e-02 |
Pos logit contributions | |
yip | +0.241 |
zee | +0.209 |
aunders | +0.177 |
zer | +0.176 |
zh | +0.172 |
Neg logit contributions | |
cout | -0.149 |
naires | -0.146 |
BuzzFeed | -0.137 |
linger | -0.136 |
swe | -0.130 |
Token | to |
Token ID | 284 |
Feature activation | -1.11e-01 |
Pos logit contributions | |
cout | +0.059 |
naires | +0.056 |
BuzzFeed | +0.054 |
linger | +0.053 |
swe | +0.051 |
Neg logit contributions | |
yip | -0.087 |
zee | -0.074 |
aunders | -0.062 |
zer | -0.062 |
zh | -0.060 |
Token | buy |
Token ID | 2822 |
Feature activation | +2.53e-02 |
Pos logit contributions | |
cout | +0.231 |
naires | +0.227 |
BuzzFeed | +0.214 |
linger | +0.210 |
swe | +0.203 |
Neg logit contributions | |
yip | -0.336 |
zee | -0.289 |
aunders | -0.244 |
zer | -0.243 |
zh | -0.236 |
Token | his |
Token ID | 465 |
Feature activation | -2.46e-01 |
Pos logit contributions | |
yip | +0.078 |
zee | +0.066 |
aunders | +0.056 |
zer | +0.056 |
zh | +0.054 |
Neg logit contributions | |
cout | -0.053 |
naires | -0.051 |
BuzzFeed | -0.049 |
linger | -0.048 |
swe | -0.046 |
Token | daughter |
Token ID | 4957 |
Feature activation | +4.17e-01 |
Pos logit contributions | |
cout | +0.451 |
naires | +0.435 |
BuzzFeed | +0.416 |
linger | +0.406 |
swe | +0.389 |
Neg logit contributions | |
yip | -0.819 |
zee | -0.710 |
aunders | -0.609 |
zer | -0.604 |
zh | -0.591 |
Token | a |
Token ID | 257 |
Feature activation | -3.00e-01 |
Pos logit contributions | |
yip | +1.244 |
zee | +1.066 |
aunders | +0.896 |
zer | +0.892 |
zh | +0.863 |
Neg logit contributions | |
cout | -0.881 |
naires | -0.867 |
BuzzFeed | -0.820 |
linger | -0.803 |
swe | -0.780 |
Token | gift |
Token ID | 6979 |
Feature activation | -5.19e-01 |
Pos logit contributions | |
cout | +0.507 |
naires | +0.494 |
BuzzFeed | +0.463 |
linger | +0.452 |
swe | +0.430 |
Neg logit contributions | |
yip | -1.038 |
zee | -0.905 |
aunders | -0.781 |
zer | -0.780 |
zh | -0.762 |
Token | at |
Token ID | 379 |
Feature activation | -5.14e-02 |
Pos logit contributions | |
cout | +0.989 |
naires | +0.955 |
BuzzFeed | +0.920 |
linger | +0.897 |
swe | +0.860 |
Neg logit contributions | |
yip | -1.683 |
zee | -1.453 |
aunders | -1.239 |
zer | -1.234 |
zh | -1.204 |
Token | Amazon |
Token ID | 6186 |
Feature activation | +2.86e-02 |
Pos logit contributions | |
cout | +0.104 |
naires | +0.102 |
BuzzFeed | +0.097 |
linger | +0.095 |
swe | +0.092 |
Neg logit contributions | |
yip | -0.160 |
zee | -0.137 |
aunders | -0.116 |
zer | -0.115 |
zh | -0.112 |
Token | . |
Token ID | 13 |
Feature activation | +5.17e-02 |
Pos logit contributions | |
yip | +0.087 |
zee | +0.075 |
aunders | +0.063 |
zer | +0.063 |
zh | +0.061 |
Neg logit contributions | |
cout | -0.060 |
naires | -0.058 |
BuzzFeed | -0.056 |
linger | -0.055 |
swe | -0.053 |
Token | � |
Token ID | 447 |
Feature activation | +3.78e-01 |
Pos logit contributions | |
yip | +0.707 |
zee | +0.616 |
aunders | +0.520 |
zer | +0.520 |
zh | +0.509 |
Neg logit contributions | |
cout | -0.415 |
naires | -0.398 |
BuzzFeed | -0.383 |
linger | -0.376 |
swe | -0.359 |
Token | � |
Token ID | 247 |
Feature activation | +9.28e-02 |
Pos logit contributions | |
yip | +1.024 |
zee | +0.855 |
aunders | +0.702 |
zer | +0.695 |
zh | +0.672 |
Neg logit contributions | |
cout | -0.929 |
naires | -0.898 |
BuzzFeed | -0.873 |
linger | -0.860 |
swe | -0.832 |
Token | s |
Token ID | 82 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
yip | +0.271 |
zee | +0.229 |
aunders | +0.191 |
zer | +0.190 |
zh | +0.184 |
Neg logit contributions | |
cout | -0.209 |
naires | -0.202 |
BuzzFeed | -0.195 |
linger | -0.192 |
swe | -0.186 |
Token | side |
Token ID | 1735 |
Feature activation | -5.02e-01 |
Pos logit contributions | |
yip | +0.356 |
zee | +0.305 |
aunders | +0.258 |
zer | +0.255 |
zh | +0.250 |
Neg logit contributions | |
cout | -0.232 |
naires | -0.225 |
BuzzFeed | -0.216 |
linger | -0.212 |
swe | -0.203 |
Token | for |
Token ID | 329 |
Feature activation | +7.84e-02 |
Pos logit contributions | |
cout | +0.985 |
naires | +0.945 |
BuzzFeed | +0.913 |
linger | +0.894 |
swe | +0.858 |
Neg logit contributions | |
yip | -1.615 |
zee | -1.387 |
aunders | -1.182 |
zer | -1.170 |
zh | -1.144 |
Token | the |
Token ID | 262 |
Feature activation | +7.36e-02 |
Pos logit contributions | |
yip | +0.240 |
zee | +0.205 |
aunders | +0.173 |
zer | +0.172 |
zh | +0.168 |
Neg logit contributions | |
cout | -0.164 |
naires | -0.159 |
BuzzFeed | -0.152 |
linger | -0.150 |
swe | -0.144 |
Token | bill |
Token ID | 2855 |
Feature activation | -4.86e-02 |
Pos logit contributions | |
yip | +0.228 |
zee | +0.195 |
aunders | +0.165 |
zer | +0.164 |
zh | +0.160 |
Neg logit contributions | |
cout | -0.151 |
naires | -0.147 |
BuzzFeed | -0.140 |
linger | -0.138 |
swe | -0.132 |
Token | signing |
Token ID | 8415 |
Feature activation | -4.91e-01 |
Pos logit contributions | |
cout | +0.099 |
naires | +0.095 |
BuzzFeed | +0.091 |
linger | +0.090 |
swe | +0.086 |
Neg logit contributions | |
yip | -0.152 |
zee | -0.131 |
aunders | -0.111 |
zer | -0.110 |
zh | -0.108 |
Token | , |
Token ID | 11 |
Feature activation | +1.17e-01 |
Pos logit contributions | |
cout | +1.023 |
naires | +0.979 |
BuzzFeed | +0.954 |
linger | +0.933 |
swe | +0.898 |
Neg logit contributions | |
yip | -1.528 |
zee | -1.299 |
aunders | -1.100 |
zer | -1.087 |
zh | -1.062 |
Token | lending |
Token ID | 20983 |
Feature activation | -1.15e-01 |
Pos logit contributions | |
yip | +0.360 |
zee | +0.308 |
aunders | +0.261 |
zer | +0.258 |
zh | +0.253 |
Neg logit contributions | |
cout | -0.243 |
naires | -0.235 |
BuzzFeed | -0.224 |
linger | -0.222 |
swe | -0.212 |
Token | the |
Token ID | 262 |
Feature activation | -8.22e-02 |
Pos logit contributions | |
cout | +0.221 |
naires | +0.213 |
BuzzFeed | +0.204 |
linger | +0.200 |
swe | +0.192 |
Neg logit contributions | |
yip | -0.371 |
zee | -0.320 |
aunders | -0.274 |
zer | -0.271 |
zh | -0.265 |
Token | Mus |
Token ID | 2629 |
Feature activation | +1.37e-01 |
Pos logit contributions | |
cout | +0.764 |
naires | +0.737 |
BuzzFeed | +0.713 |
linger | +0.695 |
swe | +0.668 |
Neg logit contributions | |
yip | -1.216 |
zee | -1.036 |
aunders | -0.882 |
zer | -0.867 |
zh | -0.851 |
Token | ial |
Token ID | 498 |
Feature activation | -5.18e-01 |
Pos logit contributions | |
yip | +0.400 |
zee | +0.345 |
aunders | +0.287 |
zer | +0.286 |
zh | +0.277 |
Neg logit contributions | |
cout | -0.306 |
naires | -0.294 |
BuzzFeed | -0.286 |
linger | -0.282 |
swe | -0.271 |
Token | , |
Token ID | 11 |
Feature activation | -2.36e-02 |
Pos logit contributions | |
cout | +1.047 |
naires | +1.005 |
BuzzFeed | +0.972 |
linger | +0.953 |
swe | +0.915 |
Neg logit contributions | |
yip | -1.635 |
zee | -1.400 |
aunders | -1.189 |
zer | -1.178 |
zh | -1.151 |
Token | following |
Token ID | 1708 |
Feature activation | +8.64e-02 |
Pos logit contributions | |
cout | +0.049 |
naires | +0.048 |
BuzzFeed | +0.046 |
linger | +0.045 |
swe | +0.043 |
Neg logit contributions | |
yip | -0.072 |
zee | -0.062 |
aunders | -0.052 |
zer | -0.052 |
zh | -0.051 |
Token | in |
Token ID | 287 |
Feature activation | +1.21e-01 |
Pos logit contributions | |
yip | +0.273 |
zee | +0.234 |
aunders | +0.199 |
zer | +0.197 |
zh | +0.193 |
Neg logit contributions | |
cout | -0.173 |
naires | -0.166 |
BuzzFeed | -0.160 |
linger | -0.157 |
swe | -0.151 |
Token | the |
Token ID | 262 |
Feature activation | +8.75e-02 |
Pos logit contributions | |
yip | +0.382 |
zee | +0.328 |
aunders | +0.278 |
zer | +0.276 |
zh | +0.270 |
Neg logit contributions | |
cout | -0.241 |
naires | -0.232 |
BuzzFeed | -0.224 |
linger | -0.220 |
swe | -0.211 |
Token | footsteps |
Token ID | 27146 |
Feature activation | +5.65e-01 |
Pos logit contributions | |
yip | +0.203 |
zee | +0.165 |
aunders | +0.129 |
zer | +0.127 |
zh | +0.123 |
Neg logit contributions | |
cout | -0.247 |
naires | -0.241 |
BuzzFeed | -0.235 |
linger | -0.232 |
swe | -0.225 |
Token | of |
Token ID | 286 |
Feature activation | -8.98e-02 |
Pos logit contributions | |
yip | +1.808 |
zee | +1.550 |
aunders | +1.319 |
zer | +1.304 |
zh | +1.275 |
Neg logit contributions | |
cout | -1.125 |
naires | -1.080 |
BuzzFeed | -1.043 |
linger | -1.022 |
swe | -0.982 |
Token | the |
Token ID | 262 |
Feature activation | -9.74e-02 |
Pos logit contributions | |
cout | +0.178 |
naires | +0.171 |
BuzzFeed | +0.165 |
linger | +0.162 |
swe | +0.155 |
Neg logit contributions | |
yip | -0.285 |
zee | -0.245 |
aunders | -0.208 |
zer | -0.207 |
zh | -0.202 |
Token | US |
Token ID | 1294 |
Feature activation | +1.96e-01 |
Pos logit contributions | |
cout | +0.204 |
naires | +0.197 |
BuzzFeed | +0.190 |
linger | +0.186 |
swe | +0.179 |
Neg logit contributions | |
yip | -0.298 |
zee | -0.255 |
aunders | -0.215 |
zer | -0.213 |
zh | -0.209 |
Token | F |
Token ID | 37 |
Feature activation | +1.32e-01 |
Pos logit contributions | |
yip | +0.529 |
zee | +0.441 |
aunders | +0.360 |
zer | +0.356 |
zh | +0.346 |
Neg logit contributions | |
cout | -0.485 |
naires | -0.469 |
BuzzFeed | -0.456 |
linger | -0.449 |
swe | -0.435 |
Token | against |
Token ID | 1028 |
Feature activation | +4.02e-01 |
Pos logit contributions | |
yip | +0.319 |
zee | +0.274 |
aunders | +0.232 |
zer | +0.232 |
zh | +0.226 |
Neg logit contributions | |
cout | -0.197 |
naires | -0.191 |
BuzzFeed | -0.182 |
linger | -0.177 |
swe | -0.172 |
Token | the |
Token ID | 262 |
Feature activation | +1.74e-01 |
Pos logit contributions | |
yip | +1.262 |
zee | +1.084 |
zer | +0.917 |
aunders | +0.915 |
zh | +0.902 |
Neg logit contributions | |
cout | -0.800 |
naires | -0.776 |
BuzzFeed | -0.737 |
linger | -0.720 |
swe | -0.694 |
Token | mortgage |
Token ID | 13682 |
Feature activation | +3.46e-01 |
Pos logit contributions | |
yip | +0.556 |
zee | +0.480 |
aunders | +0.407 |
zer | +0.406 |
zh | +0.400 |
Neg logit contributions | |
cout | -0.337 |
naires | -0.328 |
BuzzFeed | -0.312 |
linger | -0.305 |
swe | -0.292 |
Token | market |
Token ID | 1910 |
Feature activation | -4.19e-01 |
Pos logit contributions | |
yip | +1.095 |
zee | +0.941 |
aunders | +0.798 |
zer | +0.794 |
zh | +0.777 |
Neg logit contributions | |
cout | -0.691 |
naires | -0.666 |
BuzzFeed | -0.643 |
linger | -0.626 |
swe | -0.602 |
Token | . |
Token ID | 13 |
Feature activation | +6.52e-02 |
Pos logit contributions | |
cout | +0.844 |
naires | +0.812 |
BuzzFeed | +0.785 |
linger | +0.768 |
swe | +0.738 |
Neg logit contributions | |
yip | -1.321 |
zee | -1.133 |
aunders | -0.961 |
zer | -0.953 |
zh | -0.932 |
Token | ( |
Token ID | 357 |
Feature activation | +1.36e-01 |
Pos logit contributions | |
yip | +0.196 |
zee | +0.169 |
aunders | +0.142 |
zer | +0.142 |
zh | +0.138 |
Neg logit contributions | |
cout | -0.138 |
naires | -0.132 |
linger | -0.127 |
BuzzFeed | -0.127 |
swe | -0.122 |
Token | His |
Token ID | 6653 |
Feature activation | +2.59e-01 |
Pos logit contributions | |
yip | +0.376 |
zee | +0.318 |
aunders | +0.261 |
zer | +0.260 |
zh | +0.254 |
Neg logit contributions | |
cout | -0.327 |
naires | -0.315 |
BuzzFeed | -0.305 |
linger | -0.303 |
swe | -0.292 |
Token | personal |
Token ID | 2614 |
Feature activation | -3.37e-01 |
Pos logit contributions | |
yip | +0.766 |
zee | +0.652 |
aunders | +0.543 |
zer | +0.542 |
zh | +0.530 |
Neg logit contributions | |
cout | -0.565 |
naires | -0.548 |
BuzzFeed | -0.526 |
linger | -0.518 |
swe | -0.497 |
Token | take |
Token ID | 1011 |
Feature activation | +2.90e-01 |
Pos logit contributions | |
cout | +0.716 |
naires | +0.689 |
BuzzFeed | +0.667 |
linger | +0.654 |
swe | +0.630 |
Neg logit contributions | |
yip | -1.027 |
zee | -0.876 |
aunders | -0.737 |
zer | -0.730 |
zh | -0.713 |
Token | was |
Token ID | 373 |
Feature activation | +1.65e-01 |
Pos logit contributions | |
yip | +0.907 |
zee | +0.779 |
aunders | +0.657 |
zer | +0.655 |
zh | +0.640 |
Neg logit contributions | |
cout | -0.589 |
naires | -0.566 |
BuzzFeed | -0.549 |
linger | -0.537 |
swe | -0.516 |
Token | nearly |
Token ID | 3016 |
Feature activation | -1.84e-01 |
Pos logit contributions | |
yip | +0.523 |
zee | +0.450 |
zer | +0.382 |
aunders | +0.381 |
zh | +0.374 |
Neg logit contributions | |
cout | -0.325 |
naires | -0.315 |
BuzzFeed | -0.299 |
linger | -0.295 |
swe | -0.280 |
Token | Park |
Token ID | 3250 |
Feature activation | -4.92e-02 |
Pos logit contributions | |
yip | +0.914 |
zee | +0.792 |
zer | +0.661 |
aunders | +0.650 |
zh | +0.646 |
Neg logit contributions | |
cout | -0.657 |
naires | -0.637 |
BuzzFeed | -0.609 |
linger | -0.602 |
swe | -0.584 |
Token | Service |
Token ID | 4809 |
Feature activation | +5.28e-02 |
Pos logit contributions | |
cout | +0.123 |
naires | +0.119 |
BuzzFeed | +0.116 |
linger | +0.114 |
swe | +0.111 |
Neg logit contributions | |
yip | -0.131 |
zee | -0.109 |
aunders | -0.089 |
zer | -0.088 |
zh | -0.086 |
Token | acquired |
Token ID | 9477 |
Feature activation | -3.65e-01 |
Pos logit contributions | |
yip | +0.161 |
zee | +0.140 |
zer | +0.118 |
aunders | +0.118 |
zh | +0.116 |
Neg logit contributions | |
cout | -0.107 |
naires | -0.104 |
BuzzFeed | -0.099 |
linger | -0.098 |
omorphic | -0.093 |
Token | Beth |
Token ID | 19223 |
Feature activation | -3.67e-01 |
Pos logit contributions | |
cout | +0.736 |
naires | +0.706 |
BuzzFeed | +0.683 |
linger | +0.670 |
swe | +0.643 |
Neg logit contributions | |
yip | -1.155 |
zee | -0.988 |
aunders | -0.839 |
zer | -0.830 |
zh | -0.812 |
Token | une |
Token ID | 1726 |
Feature activation | +1.69e-02 |
Pos logit contributions | |
cout | +0.944 |
naires | +0.915 |
BuzzFeed | +0.891 |
linger | +0.878 |
swe | +0.851 |
Neg logit contributions | |
yip | -0.957 |
zee | -0.790 |
aunders | -0.640 |
zer | -0.631 |
zh | -0.613 |
Token | 's |
Token ID | 338 |
Feature activation | -1.83e-01 |
Pos logit contributions | |
yip | +0.055 |
zee | +0.047 |
aunders | +0.040 |
zer | +0.040 |
zh | +0.039 |
Neg logit contributions | |
cout | -0.032 |
naires | -0.031 |
BuzzFeed | -0.030 |
linger | -0.029 |
swe | -0.028 |
Token | last |
Token ID | 938 |
Feature activation | -3.88e-01 |
Pos logit contributions | |
cout | +0.359 |
naires | +0.345 |
BuzzFeed | +0.333 |
linger | +0.325 |
swe | +0.313 |
Neg logit contributions | |
yip | -0.586 |
zee | -0.506 |
aunders | -0.430 |
zer | -0.427 |
zh | -0.417 |
Token | residence |
Token ID | 11566 |
Feature activation | -2.12e-01 |
Pos logit contributions | |
cout | +0.760 |
naires | +0.729 |
BuzzFeed | +0.704 |
linger | +0.690 |
swe | +0.662 |
Neg logit contributions | |
yip | -1.249 |
zee | -1.073 |
aunders | -0.914 |
zer | -0.905 |
zh | -0.885 |
Token | ,[ |
Token ID | 17414 |
Feature activation | -1.82e-01 |
Pos logit contributions | |
cout | +0.423 |
naires | +0.407 |
BuzzFeed | +0.393 |
linger | +0.385 |
swe | +0.370 |
Neg logit contributions | |
yip | -0.671 |
zee | -0.577 |
aunders | -0.489 |
zer | -0.485 |
zh | -0.474 |
Token | 56 |
Token ID | 3980 |
Feature activation | +2.72e-01 |
Pos logit contributions | |
cout | +0.431 |
naires | +0.415 |
BuzzFeed | +0.406 |
linger | +0.396 |
swe | +0.381 |
Neg logit contributions | |
yip | -0.517 |
zee | -0.429 |
aunders | -0.356 |
zer | -0.349 |
zh | -0.342 |
Token | ] |
Token ID | 60 |
Feature activation | +2.44e-02 |
Pos logit contributions | |
yip | +0.601 |
zee | +0.474 |
aunders | +0.368 |
zer | +0.355 |
zh | +0.338 |
Neg logit contributions | |
cout | -0.812 |
naires | -0.786 |
BuzzFeed | -0.771 |
linger | -0.760 |
swe | -0.744 |
Token | developing |
Token ID | 5922 |
Feature activation | -3.12e-01 |
Pos logit contributions | |
cout | +0.141 |
naires | +0.136 |
BuzzFeed | +0.131 |
linger | +0.128 |
swe | +0.123 |
Neg logit contributions | |
yip | -0.207 |
zee | -0.176 |
aunders | -0.149 |
zer | -0.147 |
zh | -0.144 |
Token | confidence |
Token ID | 6628 |
Feature activation | -3.75e-01 |
Pos logit contributions | |
cout | +0.656 |
naires | +0.631 |
BuzzFeed | +0.611 |
linger | +0.599 |
swe | +0.577 |
Neg logit contributions | |
yip | -0.960 |
zee | -0.819 |
aunders | -0.691 |
zer | -0.684 |
zh | -0.668 |
Token | in |
Token ID | 287 |
Feature activation | -1.06e-01 |
Pos logit contributions | |
cout | +0.739 |
naires | +0.710 |
BuzzFeed | +0.685 |
linger | +0.673 |
swe | +0.646 |
Neg logit contributions | |
yip | -1.205 |
zee | -1.034 |
aunders | -0.881 |
zer | -0.871 |
zh | -0.851 |
Token | one |
Token ID | 530 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
cout | +0.224 |
naires | +0.216 |
BuzzFeed | +0.209 |
linger | +0.205 |
swe | +0.198 |
Neg logit contributions | |
yip | -0.325 |
zee | -0.277 |
aunders | -0.233 |
zer | -0.231 |
zh | -0.225 |
Token | another |
Token ID | 1194 |
Feature activation | -1.27e-01 |
Pos logit contributions | |
cout | +0.327 |
naires | +0.312 |
BuzzFeed | +0.301 |
linger | +0.295 |
swe | +0.282 |
Neg logit contributions | |
yip | -0.607 |
zee | -0.525 |
aunders | -0.451 |
zer | -0.446 |
zh | -0.437 |
Token | . |
Token ID | 13 |
Feature activation | -1.68e-01 |
Pos logit contributions | |
cout | +0.269 |
naires | +0.259 |
BuzzFeed | +0.251 |
linger | +0.246 |
swe | +0.237 |
Neg logit contributions | |
yip | -0.390 |
zee | -0.332 |
aunders | -0.280 |
zer | -0.277 |
zh | -0.271 |
Token | I |
Token ID | 314 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
cout | +0.361 |
naires | +0.348 |
BuzzFeed | +0.337 |
linger | +0.330 |
swe | +0.318 |
Neg logit contributions | |
yip | -0.507 |
zee | -0.431 |
aunders | -0.363 |
zer | -0.359 |
zh | -0.350 |
Token | leave |
Token ID | 2666 |
Feature activation | -1.75e-01 |
Pos logit contributions | |
cout | +0.264 |
naires | +0.254 |
BuzzFeed | +0.246 |
linger | +0.241 |
swe | +0.233 |
Neg logit contributions | |
yip | -0.360 |
zee | -0.306 |
aunders | -0.256 |
zer | -0.253 |
zh | -0.247 |
Token | you |
Token ID | 345 |
Feature activation | -8.98e-02 |
Pos logit contributions | |
cout | +0.331 |
naires | +0.316 |
BuzzFeed | +0.307 |
linger | +0.301 |
swe | +0.289 |
Neg logit contributions | |
yip | -0.579 |
zee | -0.497 |
aunders | -0.426 |
zer | -0.422 |
zh | -0.411 |
Token | a |
Token ID | 257 |
Feature activation | -1.18e-01 |
Pos logit contributions | |
cout | +0.186 |
naires | +0.179 |
BuzzFeed | +0.173 |
linger | +0.169 |
swe | +0.163 |
Neg logit contributions | |
yip | -0.278 |
zee | -0.238 |
aunders | -0.201 |
zer | -0.199 |
zh | -0.195 |
Token | thirst |
Token ID | 24613 |
Feature activation | -1.74e-01 |
Pos logit contributions | |
cout | +0.240 |
naires | +0.231 |
BuzzFeed | +0.223 |
linger | +0.218 |
swe | +0.209 |
Neg logit contributions | |
yip | -0.369 |
zee | -0.316 |
aunders | -0.267 |
zer | -0.265 |
zh | -0.259 |
Token | obtain |
Token ID | 7330 |
Feature activation | -2.02e-01 |
Pos logit contributions | |
yip | +0.899 |
zee | +0.779 |
aunders | +0.669 |
zer | +0.667 |
zh | +0.652 |
Neg logit contributions | |
cout | -0.522 |
naires | -0.502 |
BuzzFeed | -0.482 |
linger | -0.461 |
swe | -0.444 |
Token | health |
Token ID | 1535 |
Feature activation | -3.92e-01 |
Pos logit contributions | |
cout | +0.377 |
naires | +0.361 |
BuzzFeed | +0.348 |
linger | +0.341 |
swe | +0.326 |
Neg logit contributions | |
yip | -0.667 |
zee | -0.576 |
aunders | -0.494 |
zer | -0.489 |
zh | -0.479 |
Token | insurance |
Token ID | 5096 |
Feature activation | -3.47e-01 |
Pos logit contributions | |
cout | +0.626 |
naires | +0.594 |
BuzzFeed | +0.566 |
linger | +0.554 |
swe | +0.527 |
Neg logit contributions | |
yip | -1.418 |
zee | -1.236 |
aunders | -1.074 |
zer | -1.063 |
zh | -1.042 |
Token | through |
Token ID | 832 |
Feature activation | -5.61e-01 |
Pos logit contributions | |
cout | +0.709 |
naires | +0.681 |
BuzzFeed | +0.661 |
linger | +0.646 |
swe | +0.621 |
Neg logit contributions | |
yip | -1.077 |
zee | -0.923 |
aunders | -0.784 |
zer | -0.775 |
zh | -0.758 |
Token | the |
Token ID | 262 |
Feature activation | +2.46e-02 |
Pos logit contributions | |
cout | +1.194 |
naires | +1.147 |
BuzzFeed | +1.113 |
linger | +1.090 |
swe | +1.052 |
Neg logit contributions | |
yip | -1.719 |
zee | -1.459 |
aunders | -1.230 |
zer | -1.218 |
zh | -1.190 |
Token | exchanges |
Token ID | 14525 |
Feature activation | -8.66e-01 |
Pos logit contributions | |
yip | +0.073 |
zee | +0.062 |
aunders | +0.052 |
zer | +0.052 |
zh | +0.051 |
Neg logit contributions | |
cout | -0.053 |
naires | -0.051 |
BuzzFeed | -0.050 |
linger | -0.049 |
swe | -0.047 |
Token | , |
Token ID | 11 |
Feature activation | +3.00e-02 |
Pos logit contributions | |
cout | +1.782 |
naires | +1.715 |
BuzzFeed | +1.661 |
linger | +1.626 |
swe | +1.568 |
Neg logit contributions | |
yip | -2.706 |
zee | -2.310 |
aunders | -1.956 |
zer | -1.934 |
zh | -1.891 |
Token | emphasizing |
Token ID | 36360 |
Feature activation | +1.12e-01 |
Pos logit contributions | |
yip | +0.091 |
zee | +0.078 |
aunders | +0.065 |
zer | +0.065 |
zh | +0.063 |
Neg logit contributions | |
cout | -0.064 |
naires | -0.061 |
BuzzFeed | -0.059 |
linger | -0.058 |
swe | -0.056 |
Token | that |
Token ID | 326 |
Feature activation | -2.32e-02 |
Pos logit contributions | |
yip | +0.350 |
zee | +0.299 |
aunders | +0.254 |
zer | +0.252 |
zh | +0.245 |
Neg logit contributions | |
cout | -0.225 |
naires | -0.216 |
BuzzFeed | -0.209 |
linger | -0.205 |
swe | -0.196 |
Token | if |
Token ID | 611 |
Feature activation | -1.99e-02 |
Pos logit contributions | |
cout | +0.049 |
naires | +0.047 |
BuzzFeed | +0.046 |
linger | +0.045 |
swe | +0.043 |
Neg logit contributions | |
yip | -0.071 |
zee | -0.060 |
aunders | -0.051 |
zer | -0.050 |
zh | -0.049 |
Token | the |
Token ID | 262 |
Feature activation | -1.58e-01 |
Pos logit contributions | |
cout | +0.041 |
naires | +0.040 |
BuzzFeed | +0.038 |
linger | +0.038 |
swe | +0.036 |
Neg logit contributions | |
yip | -0.062 |
zee | -0.052 |
aunders | -0.044 |
zer | -0.044 |
zh | -0.043 |
Token |
|
Token ID | 198 |
Feature activation | +1.93e-01 |
Pos logit contributions | |
cout | +0.363 |
naires | +0.351 |
BuzzFeed | +0.342 |
linger | +0.336 |
swe | +0.325 |
Neg logit contributions | |
yip | -0.414 |
zee | -0.344 |
aunders | -0.284 |
zer | -0.280 |
zh | -0.272 |
Token |
|
Token ID | 198 |
Feature activation | +1.78e-01 |
Pos logit contributions | |
yip | +0.472 |
zee | +0.393 |
zer | +0.313 |
aunders | +0.312 |
zh | +0.302 |
Neg logit contributions | |
cout | -0.520 |
naires | -0.503 |
BuzzFeed | -0.489 |
linger | -0.487 |
swe | -0.472 |
Token | X |
Token ID | 55 |
Feature activation | +6.41e-01 |
Pos logit contributions | |
yip | +0.382 |
zee | +0.299 |
aunders | +0.228 |
zer | +0.222 |
zh | +0.213 |
Neg logit contributions | |
cout | -0.539 |
naires | -0.525 |
BuzzFeed | -0.514 |
linger | -0.506 |
swe | -0.493 |
Token | avier |
Token ID | 19492 |
Feature activation | -4.81e-01 |
Pos logit contributions | |
yip | +2.236 |
zee | +1.969 |
aunders | +1.722 |
zer | +1.716 |
zh | +1.652 |
Neg logit contributions | |
cout | -1.059 |
naires | -0.993 |
BuzzFeed | -0.950 |
linger | -0.939 |
swe | -0.901 |
Token | Gu |
Token ID | 1962 |
Feature activation | +9.41e-02 |
Pos logit contributions | |
cout | +0.819 |
naires | +0.783 |
BuzzFeed | +0.742 |
linger | +0.738 |
swe | +0.703 |
Neg logit contributions | |
yip | -1.656 |
zee | -1.456 |
aunders | -1.253 |
zer | -1.247 |
zh | -1.224 |
Token | il |
Token ID | 346 |
Feature activation | -9.71e-01 |
Pos logit contributions | |
yip | +0.322 |
zee | +0.282 |
aunders | +0.243 |
zer | +0.242 |
zh | +0.238 |
Neg logit contributions | |
cout | -0.164 |
naires | -0.156 |
BuzzFeed | -0.150 |
linger | -0.148 |
swe | -0.141 |
Token | bert |
Token ID | 4835 |
Feature activation | -2.46e-01 |
Pos logit contributions | |
cout | +2.044 |
naires | +1.970 |
BuzzFeed | +1.910 |
linger | +1.866 |
swe | +1.794 |
Neg logit contributions | |
yip | -2.983 |
zee | -2.528 |
aunders | -2.141 |
zer | -2.100 |
zh | -2.062 |
Token | : |
Token ID | 1058 |
Feature activation | -6.65e-02 |
Pos logit contributions | |
cout | +0.642 |
naires | +0.622 |
BuzzFeed | +0.604 |
linger | +0.600 |
swe | +0.582 |
Neg logit contributions | |
yip | -0.621 |
zee | -0.518 |
aunders | -0.417 |
zer | -0.413 |
zh | -0.399 |
Token | You |
Token ID | 921 |
Feature activation | -1.74e-01 |
Pos logit contributions | |
cout | +0.155 |
naires | +0.150 |
BuzzFeed | +0.144 |
linger | +0.144 |
swe | +0.138 |
Neg logit contributions | |
yip | -0.187 |
zee | -0.159 |
aunders | -0.131 |
zer | -0.130 |
zh | -0.127 |
Token | still |
Token ID | 991 |
Feature activation | -1.90e-01 |
Pos logit contributions | |
cout | +0.357 |
naires | +0.344 |
BuzzFeed | +0.332 |
linger | +0.326 |
swe | +0.313 |
Neg logit contributions | |
yip | -0.545 |
zee | -0.467 |
aunders | -0.395 |
zer | -0.391 |
zh | -0.382 |
Token | can |
Token ID | 460 |
Feature activation | -3.76e-01 |
Pos logit contributions | |
cout | +0.388 |
naires | +0.374 |
BuzzFeed | +0.361 |
linger | +0.352 |
swe | +0.339 |
Neg logit contributions | |
yip | -0.592 |
zee | -0.509 |
aunders | -0.430 |
zer | -0.426 |
zh | -0.417 |
Token | final |
Token ID | 2457 |
Feature activation | -6.71e-01 |
Pos logit contributions | |
cout | +0.022 |
naires | +0.021 |
BuzzFeed | +0.020 |
linger | +0.020 |
swe | +0.019 |
Neg logit contributions | |
yip | -0.035 |
zee | -0.030 |
aunders | -0.025 |
zer | -0.025 |
zh | -0.024 |
Token | European |
Token ID | 3427 |
Feature activation | -4.83e-01 |
Pos logit contributions | |
cout | +1.405 |
naires | +1.351 |
BuzzFeed | +1.309 |
linger | +1.284 |
swe | +1.235 |
Neg logit contributions | |
yip | -2.075 |
zee | -1.772 |
aunders | -1.494 |
zer | -1.472 |
zh | -1.441 |
Token | Rugby |
Token ID | 26244 |
Feature activation | -2.07e-01 |
Pos logit contributions | |
cout | +1.089 |
naires | +1.048 |
BuzzFeed | +1.018 |
linger | +1.001 |
swe | +0.967 |
Neg logit contributions | |
yip | -1.415 |
zee | -1.195 |
aunders | -0.999 |
zer | -0.984 |
zh | -0.959 |
Token | Champions |
Token ID | 6662 |
Feature activation | -2.26e-01 |
Pos logit contributions | |
cout | +0.490 |
naires | +0.475 |
BuzzFeed | +0.461 |
linger | +0.453 |
swe | +0.438 |
Neg logit contributions | |
yip | -0.577 |
zee | -0.485 |
aunders | -0.400 |
zer | -0.396 |
zh | -0.385 |
Token | Cup |
Token ID | 5454 |
Feature activation | +1.46e-01 |
Pos logit contributions | |
cout | +0.520 |
naires | +0.501 |
BuzzFeed | +0.487 |
linger | +0.479 |
swe | +0.463 |
Neg logit contributions | |
yip | -0.653 |
zee | -0.550 |
aunders | -0.457 |
zer | -0.451 |
zh | -0.440 |
Token | pool |
Token ID | 5933 |
Feature activation | -8.85e-01 |
Pos logit contributions | |
yip | +0.457 |
zee | +0.393 |
zer | +0.334 |
aunders | +0.334 |
zh | +0.325 |
Neg logit contributions | |
cout | -0.293 |
naires | -0.284 |
BuzzFeed | -0.274 |
linger | -0.267 |
swe | -0.255 |
Token | match |
Token ID | 2872 |
Feature activation | -4.71e-02 |
Pos logit contributions | |
cout | +1.674 |
naires | +1.592 |
BuzzFeed | +1.534 |
linger | +1.505 |
swe | +1.455 |
Neg logit contributions | |
yip | -2.919 |
zee | -2.512 |
aunders | -2.151 |
zer | -2.119 |
zh | -2.074 |
Token | at |
Token ID | 379 |
Feature activation | -1.31e-02 |
Pos logit contributions | |
cout | +0.103 |
naires | +0.100 |
BuzzFeed | +0.096 |
linger | +0.095 |
swe | +0.091 |
Neg logit contributions | |
yip | -0.140 |
zee | -0.119 |
aunders | -0.100 |
zer | -0.099 |
zh | -0.097 |
Token | Wel |
Token ID | 22268 |
Feature activation | +1.55e-01 |
Pos logit contributions | |
cout | +0.026 |
naires | +0.025 |
BuzzFeed | +0.024 |
linger | +0.024 |
swe | +0.023 |
Neg logit contributions | |
yip | -0.041 |
zee | -0.035 |
aunders | -0.030 |
zer | -0.030 |
zh | -0.029 |
Token | ford |
Token ID | 3841 |
Feature activation | -1.28e-02 |
Pos logit contributions | |
yip | +0.451 |
zee | +0.380 |
aunders | +0.318 |
zer | +0.311 |
zh | +0.304 |
Neg logit contributions | |
cout | -0.351 |
naires | -0.338 |
BuzzFeed | -0.328 |
linger | -0.322 |
swe | -0.311 |
Token | Road |
Token ID | 5567 |
Feature activation | -7.23e-02 |
Pos logit contributions | |
cout | +0.026 |
naires | +0.025 |
BuzzFeed | +0.025 |
linger | +0.024 |
swe | +0.023 |
Neg logit contributions | |
yip | -0.040 |
zee | -0.034 |
aunders | -0.029 |
zer | -0.029 |
zh | -0.028 |
Token | Philip |
Token ID | 14576 |
Feature activation | +1.90e-01 |
Pos logit contributions | |
cout | +0.019 |
naires | +0.019 |
BuzzFeed | +0.018 |
linger | +0.017 |
swe | +0.017 |
Neg logit contributions | |
yip | -0.031 |
zee | -0.027 |
aunders | -0.023 |
zer | -0.022 |
zh | -0.022 |
Token | . |
Token ID | 13 |
Feature activation | +7.03e-02 |
Pos logit contributions | |
yip | +0.594 |
zee | +0.511 |
aunders | +0.434 |
zer | +0.431 |
zh | +0.420 |
Neg logit contributions | |
cout | -0.383 |
naires | -0.370 |
BuzzFeed | -0.356 |
linger | -0.348 |
swe | -0.335 |
Token |
|
Token ID | 198 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
yip | +0.188 |
zee | +0.157 |
aunders | +0.128 |
zer | +0.127 |
zh | +0.123 |
Neg logit contributions | |
cout | -0.175 |
naires | -0.170 |
BuzzFeed | -0.165 |
linger | -0.162 |
swe | -0.157 |
Token |
|
Token ID | 198 |
Feature activation | -8.93e-03 |
Pos logit contributions | |
yip | +0.252 |
zee | +0.210 |
zer | +0.167 |
aunders | +0.167 |
zh | +0.161 |
Neg logit contributions | |
cout | -0.280 |
naires | -0.270 |
BuzzFeed | -0.263 |
linger | -0.262 |
swe | -0.254 |
Token | TV |
Token ID | 6849 |
Feature activation | +4.49e-02 |
Pos logit contributions | |
cout | +0.021 |
naires | +0.020 |
BuzzFeed | +0.019 |
linger | +0.019 |
swe | +0.018 |
Neg logit contributions | |
yip | -0.025 |
zee | -0.022 |
aunders | -0.018 |
zer | -0.018 |
zh | -0.017 |
Token | LINE |
Token ID | 24027 |
Feature activation | -1.15e+00 |
Pos logit contributions | |
yip | +0.139 |
zee | +0.119 |
aunders | +0.100 |
zer | +0.099 |
zh | +0.097 |
Neg logit contributions | |
cout | -0.094 |
naires | -0.090 |
BuzzFeed | -0.087 |
linger | -0.085 |
swe | -0.082 |
Token | | |
Token ID | 930 |
Feature activation | -3.57e-01 |
Pos logit contributions | |
cout | +2.996 |
BuzzFeed | +2.867 |
naires | +2.867 |
linger | +2.796 |
swe | +2.683 |
Neg logit contributions | |
yip | -3.009 |
zee | -2.459 |
aunders | -1.948 |
zer | -1.888 |
zh | -1.881 |
Token | So |
Token ID | 1406 |
Feature activation | -1.40e-01 |
Pos logit contributions | |
cout | +0.813 |
naires | +0.785 |
BuzzFeed | +0.757 |
linger | +0.748 |
swe | +0.723 |
Neg logit contributions | |
yip | -1.033 |
zee | -0.875 |
aunders | -0.728 |
zer | -0.722 |
zh | -0.704 |
Token | , |
Token ID | 11 |
Feature activation | -4.95e-02 |
Pos logit contributions | |
cout | +0.289 |
naires | +0.278 |
BuzzFeed | +0.267 |
linger | +0.263 |
swe | +0.253 |
Neg logit contributions | |
yip | -0.435 |
zee | -0.372 |
aunders | -0.314 |
zer | -0.312 |
zh | -0.305 |
Token | the |
Token ID | 262 |
Feature activation | +1.63e-01 |
Pos logit contributions | |
cout | +0.104 |
naires | +0.101 |
BuzzFeed | +0.096 |
linger | +0.094 |
swe | +0.091 |
Neg logit contributions | |
yip | -0.151 |
zee | -0.129 |
aunders | -0.108 |
zer | -0.108 |
zh | -0.105 |
Token | prom |
Token ID | 1552 |
Feature activation | +3.65e-02 |
Pos logit contributions | |
yip | +0.511 |
zee | +0.441 |
aunders | +0.373 |
zer | +0.372 |
zh | +0.362 |
Neg logit contributions | |
cout | -0.324 |
naires | -0.315 |
BuzzFeed | -0.299 |
linger | -0.293 |
swe | -0.282 |
Token | his |
Token ID | 465 |
Feature activation | +2.71e-01 |
Pos logit contributions | |
yip | +0.639 |
zee | +0.547 |
aunders | +0.462 |
zer | +0.459 |
zh | +0.446 |
Neg logit contributions | |
cout | -0.426 |
naires | -0.411 |
BuzzFeed | -0.395 |
linger | -0.387 |
swe | -0.373 |
Token | ability |
Token ID | 2694 |
Feature activation | +5.56e-01 |
Pos logit contributions | |
yip | +0.825 |
zee | +0.706 |
aunders | +0.597 |
zer | +0.593 |
zh | +0.579 |
Neg logit contributions | |
cout | -0.564 |
naires | -0.548 |
BuzzFeed | -0.526 |
linger | -0.513 |
swe | -0.493 |
Token | to |
Token ID | 284 |
Feature activation | -3.35e-02 |
Pos logit contributions | |
yip | +1.767 |
zee | +1.513 |
aunders | +1.285 |
zer | +1.271 |
zh | +1.244 |
Neg logit contributions | |
cout | -1.116 |
naires | -1.074 |
BuzzFeed | -1.035 |
linger | -1.016 |
swe | -0.976 |
Token | do |
Token ID | 466 |
Feature activation | -1.58e-01 |
Pos logit contributions | |
cout | +0.068 |
naires | +0.066 |
BuzzFeed | +0.064 |
linger | +0.062 |
swe | +0.060 |
Neg logit contributions | |
yip | -0.104 |
zee | -0.089 |
aunders | -0.075 |
zer | -0.075 |
zh | -0.073 |
Token | so |
Token ID | 523 |
Feature activation | -3.88e-01 |
Pos logit contributions | |
cout | +0.316 |
naires | +0.303 |
BuzzFeed | +0.293 |
linger | +0.287 |
swe | +0.276 |
Neg logit contributions | |
yip | -0.503 |
zee | -0.431 |
aunders | -0.367 |
zer | -0.363 |
zh | -0.355 |
Token | effectively |
Token ID | 6840 |
Feature activation | -1.05e+00 |
Pos logit contributions | |
cout | +0.815 |
naires | +0.784 |
BuzzFeed | +0.759 |
linger | +0.745 |
swe | +0.717 |
Neg logit contributions | |
yip | -1.198 |
zee | -1.020 |
aunders | -0.862 |
zer | -0.852 |
zh | -0.832 |
Token | . |
Token ID | 13 |
Feature activation | -5.55e-02 |
Pos logit contributions | |
cout | +2.138 |
naires | +2.048 |
BuzzFeed | +1.991 |
linger | +1.946 |
swe | +1.876 |
Neg logit contributions | |
yip | -3.294 |
zee | -2.802 |
aunders | -2.373 |
zer | -2.345 |
zh | -2.292 |
Token | He |
Token ID | 679 |
Feature activation | +2.90e-02 |
Pos logit contributions | |
cout | +0.134 |
naires | +0.129 |
BuzzFeed | +0.125 |
linger | +0.124 |
swe | +0.120 |
Neg logit contributions | |
yip | -0.153 |
zee | -0.128 |
aunders | -0.106 |
zer | -0.104 |
zh | -0.101 |
Token | reached |
Token ID | 4251 |
Feature activation | -8.77e-02 |
Pos logit contributions | |
yip | +0.087 |
zee | +0.075 |
aunders | +0.063 |
zer | +0.063 |
zh | +0.062 |
Neg logit contributions | |
cout | -0.061 |
naires | -0.059 |
BuzzFeed | -0.057 |
linger | -0.056 |
swe | -0.053 |
Token | that |
Token ID | 326 |
Feature activation | +3.29e-02 |
Pos logit contributions | |
cout | +0.171 |
naires | +0.164 |
BuzzFeed | +0.159 |
linger | +0.155 |
swe | +0.149 |
Neg logit contributions | |
yip | -0.283 |
zee | -0.243 |
aunders | -0.207 |
zer | -0.205 |
zh | -0.201 |
Token | point |
Token ID | 966 |
Feature activation | -5.60e-02 |
Pos logit contributions | |
yip | +0.108 |
zee | +0.094 |
aunders | +0.081 |
zer | +0.080 |
zh | +0.078 |
Neg logit contributions | |
cout | -0.061 |
naires | -0.059 |
BuzzFeed | -0.056 |
linger | -0.055 |
swe | -0.052 |