Token | would |
Token ID | 561 |
Feature activation | +5.75e+00 |
Pos logit contributions | |
yours | +0.314 |
beh | +0.299 |
ours | +0.281 |
possible | +0.278 |
theirs | +0.266 |
Neg logit contributions | |
boy | -0.130 |
sky | -0.122 |
was | -0.121 |
girl | -0.119 |
man | -0.113 |
Token | be |
Token ID | 307 |
Feature activation | +5.01e+00 |
Pos logit contributions | |
boy | +0.216 |
little | +0.215 |
was | +0.214 |
man | +0.202 |
sky | +0.198 |
Neg logit contributions | |
beh | -0.366 |
yours | -0.330 |
unnecessary | -0.306 |
possible | -0.302 |
ours | -0.299 |
Token | nicer |
Token ID | 36597 |
Feature activation | -6.79e+00 |
Pos logit contributions | |
boy | +0.189 |
little | +0.182 |
girl | +0.177 |
was | +0.172 |
man | +0.169 |
Neg logit contributions | |
beh | -0.315 |
yours | -0.288 |
possible | -0.256 |
ours | -0.252 |
ieve | -0.251 |
Token | than |
Token ID | 621 |
Feature activation | -6.16e+00 |
Pos logit contributions | |
beh | +0.466 |
yours | +0.440 |
possible | +0.394 |
ours | +0.390 |
unnecessary | +0.382 |
Neg logit contributions | |
boy | -0.225 |
girl | -0.205 |
little | -0.200 |
man | -0.199 |
sky | -0.197 |
Token | the |
Token ID | 262 |
Feature activation | +9.10e+00 |
Pos logit contributions | |
beh | +0.364 |
yours | +0.295 |
ieve | +0.277 |
eking | +0.277 |
unnecessary | +0.275 |
Neg logit contributions | |
little | -0.270 |
boy | -0.268 |
was | -0.253 |
girl | -0.249 |
man | -0.249 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.69e+01 |
Pos logit contributions | |
boy | +0.273 |
girl | +0.248 |
sky | +0.244 |
man | +0.241 |
little | +0.241 |
Neg logit contributions | |
beh | -0.632 |
yours | -0.611 |
possible | -0.550 |
ours | -0.540 |
theirs | -0.528 |
Token | in |
Token ID | 259 |
Feature activation | +3.25e-01 |
Pos logit contributions | |
boy | +1.055 |
girl | +0.970 |
man | +0.955 |
little | +0.948 |
was | +0.944 |
Neg logit contributions | |
beh | -1.634 |
yours | -1.584 |
possible | -1.409 |
ours | -1.398 |
theirs | -1.351 |
Token | . |
Token ID | 13 |
Feature activation | -7.41e+00 |
Pos logit contributions | |
boy | +0.012 |
girl | +0.011 |
sky | +0.011 |
man | +0.011 |
little | +0.011 |
Neg logit contributions | |
beh | -0.020 |
yours | -0.020 |
possible | -0.017 |
ours | -0.017 |
theirs | -0.017 |
Token | They |
Token ID | 1119 |
Feature activation | +2.16e+00 |
Pos logit contributions | |
beh | +0.533 |
yours | +0.520 |
possible | +0.468 |
ours | +0.460 |
theirs | +0.451 |
Neg logit contributions | |
boy | -0.202 |
girl | -0.182 |
sky | -0.178 |
man | -0.177 |
little | -0.175 |
Token | followed |
Token ID | 3940 |
Feature activation | +8.15e+00 |
Pos logit contributions | |
boy | +0.065 |
girl | +0.059 |
sky | +0.058 |
little | +0.057 |
man | +0.057 |
Neg logit contributions | |
beh | -0.150 |
yours | -0.145 |
possible | -0.130 |
ours | -0.128 |
theirs | -0.125 |
Token | their |
Token ID | 511 |
Feature activation | -5.74e-02 |
Pos logit contributions | |
boy | +0.274 |
little | +0.257 |
man | +0.248 |
girl | +0.248 |
sky | +0.236 |
Neg logit contributions | |
beh | -0.566 |
yours | -0.528 |
possible | -0.484 |
ours | -0.467 |
eking | -0.460 |
Token | just |
Token ID | 655 |
Feature activation | -6.25e+00 |
Pos logit contributions | |
boy | +0.403 |
girl | +0.373 |
sky | +0.366 |
little | +0.366 |
man | +0.363 |
Neg logit contributions | |
beh | -0.689 |
yours | -0.662 |
possible | -0.590 |
ours | -0.576 |
theirs | -0.560 |
Token | wanted |
Token ID | 2227 |
Feature activation | +3.31e+00 |
Pos logit contributions | |
beh | +0.369 |
yours | +0.329 |
ours | +0.296 |
ieve | +0.293 |
possible | +0.292 |
Neg logit contributions | |
was | -0.258 |
little | -0.257 |
boy | -0.254 |
girl | -0.241 |
sky | -0.234 |
Token | to |
Token ID | 284 |
Feature activation | +4.95e+00 |
Pos logit contributions | |
boy | +0.103 |
was | +0.102 |
little | +0.101 |
sky | +0.093 |
man | +0.093 |
Neg logit contributions | |
beh | -0.232 |
yours | -0.204 |
ieve | -0.194 |
ours | -0.189 |
eking | -0.189 |
Token | see |
Token ID | 766 |
Feature activation | -2.27e+00 |
Pos logit contributions | |
little | +0.188 |
boy | +0.185 |
was | +0.183 |
man | +0.179 |
girl | +0.168 |
Neg logit contributions | |
beh | -0.311 |
yours | -0.282 |
unnecessary | -0.258 |
ours | -0.256 |
possible | -0.255 |
Token | the |
Token ID | 262 |
Feature activation | +9.10e+00 |
Pos logit contributions | |
beh | +0.137 |
yours | +0.116 |
possible | +0.105 |
eking | +0.103 |
anks | +0.102 |
Neg logit contributions | |
boy | -0.099 |
little | -0.097 |
girl | -0.094 |
was | -0.093 |
man | -0.091 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.65e+01 |
Pos logit contributions | |
boy | +0.255 |
girl | +0.231 |
sky | +0.226 |
man | +0.223 |
little | +0.222 |
Neg logit contributions | |
beh | -0.640 |
yours | -0.626 |
possible | -0.566 |
ours | -0.555 |
theirs | -0.540 |
Token | ins |
Token ID | 1040 |
Feature activation | -5.66e+00 |
Pos logit contributions | |
boy | +0.955 |
girl | +0.867 |
was | +0.855 |
man | +0.851 |
little | +0.841 |
Neg logit contributions | |
beh | -1.666 |
yours | -1.650 |
possible | -1.478 |
ours | -1.463 |
theirs | -1.424 |
Token | . |
Token ID | 13 |
Feature activation | -7.31e+00 |
Pos logit contributions | |
beh | +0.361 |
yours | +0.336 |
unnecessary | +0.315 |
ours | +0.302 |
possible | +0.290 |
Neg logit contributions | |
boy | -0.211 |
was | -0.211 |
man | -0.198 |
girl | -0.196 |
and | -0.195 |
Token | He |
Token ID | 679 |
Feature activation | +2.51e-01 |
Pos logit contributions | |
beh | +0.488 |
yours | +0.478 |
possible | +0.421 |
ours | +0.420 |
theirs | +0.411 |
Neg logit contributions | |
boy | -0.223 |
girl | -0.206 |
sky | -0.204 |
man | -0.203 |
little | -0.199 |
Token | walked |
Token ID | 6807 |
Feature activation | +8.44e+00 |
Pos logit contributions | |
sky | +0.008 |
boy | +0.007 |
girl | +0.007 |
little | +0.006 |
man | +0.006 |
Neg logit contributions | |
beh | -0.016 |
yours | -0.015 |
theirs | -0.013 |
possible | -0.013 |
ours | -0.013 |
Token | back |
Token ID | 736 |
Feature activation | -3.78e+00 |
Pos logit contributions | |
boy | +0.391 |
girl | +0.367 |
sky | +0.362 |
little | +0.357 |
man | +0.353 |
Neg logit contributions | |
beh | -0.424 |
yours | -0.419 |
possible | -0.368 |
ours | -0.351 |
theirs | -0.350 |
Token | € |
Token ID | 26391 |
Feature activation | -8.33e+00 |
Pos logit contributions | |
boy | +0.054 |
girl | +0.051 |
man | +0.051 |
sky | +0.051 |
was | +0.050 |
Neg logit contributions | |
beh | -0.042 |
yours | -0.041 |
possible | -0.034 |
ours | -0.034 |
theirs | -0.033 |
Token | |
Token ID | 220 |
Feature activation | +5.83e+00 |
Pos logit contributions | |
beh | +0.595 |
yours | +0.574 |
possible | +0.519 |
ours | +0.510 |
theirs | +0.493 |
Neg logit contributions | |
boy | -0.230 |
girl | -0.214 |
sky | -0.200 |
little | -0.198 |
man | -0.195 |
Token |
|
Token ID | 198 |
Feature activation | +9.15e+00 |
Pos logit contributions | |
boy | +0.070 |
sky | +0.054 |
little | +0.052 |
girl | +0.052 |
man | +0.049 |
Neg logit contributions | |
beh | -0.510 |
yours | -0.497 |
possible | -0.457 |
ours | -0.453 |
theirs | -0.445 |
Token |
|
Token ID | 198 |
Feature activation | +9.00e+00 |
Pos logit contributions | |
boy | +0.179 |
girl | +0.157 |
man | +0.141 |
sky | +0.141 |
little | +0.135 |
Neg logit contributions | |
beh | -0.726 |
yours | -0.711 |
possible | -0.649 |
ours | -0.630 |
theirs | -0.626 |
Token | The |
Token ID | 464 |
Feature activation | +1.23e+01 |
Pos logit contributions | |
boy | +0.284 |
girl | +0.259 |
sky | +0.257 |
man | +0.249 |
little | +0.246 |
Neg logit contributions | |
beh | -0.620 |
yours | -0.594 |
possible | -0.535 |
ours | -0.517 |
theirs | -0.508 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.65e+01 |
Pos logit contributions | |
boy | +0.337 |
sky | +0.306 |
was | +0.301 |
girl | +0.301 |
man | +0.292 |
Neg logit contributions | |
beh | -0.838 |
yours | -0.821 |
possible | -0.752 |
ours | -0.718 |
unnecessary | -0.703 |
Token | in |
Token ID | 259 |
Feature activation | -2.80e-01 |
Pos logit contributions | |
boy | +1.182 |
sky | +1.122 |
girl | +1.114 |
little | +1.096 |
man | +1.092 |
Neg logit contributions | |
beh | -1.455 |
yours | -1.356 |
possible | -1.188 |
eking | -1.133 |
ieve | -1.133 |
Token | was |
Token ID | 373 |
Feature activation | +1.01e+01 |
Pos logit contributions | |
beh | +0.019 |
yours | +0.017 |
eking | +0.015 |
ieve | +0.015 |
possible | +0.015 |
Neg logit contributions | |
boy | -0.007 |
girl | -0.007 |
sky | -0.007 |
little | -0.006 |
woman | -0.006 |
Token | happy |
Token ID | 3772 |
Feature activation | +4.88e+00 |
Pos logit contributions | |
boy | +0.386 |
little | +0.367 |
sky | +0.361 |
was | +0.359 |
man | +0.357 |
Neg logit contributions | |
beh | -0.622 |
yours | -0.593 |
possible | -0.519 |
ours | -0.516 |
unnecessary | -0.501 |
Token | , |
Token ID | 11 |
Feature activation | +1.15e+01 |
Pos logit contributions | |
boy | +0.155 |
girl | +0.140 |
little | +0.138 |
was | +0.138 |
man | +0.137 |
Neg logit contributions | |
beh | -0.344 |
yours | -0.328 |
ours | -0.295 |
possible | -0.293 |
unnecessary | -0.286 |
Token | so |
Token ID | 523 |
Feature activation | +7.28e-01 |
Pos logit contributions | |
boy | +0.355 |
girl | +0.325 |
sky | +0.320 |
man | +0.315 |
little | +0.310 |
Neg logit contributions | |
beh | -0.777 |
yours | -0.761 |
possible | -0.685 |
ours | -0.668 |
theirs | -0.655 |
Token | and |
Token ID | 290 |
Feature activation | +1.63e+01 |
Pos logit contributions | |
boy | +0.165 |
girl | +0.150 |
little | +0.149 |
was | +0.139 |
sky | +0.136 |
Neg logit contributions | |
beh | -0.433 |
yours | -0.407 |
ours | -0.371 |
possible | -0.370 |
ieve | -0.362 |
Token | scared |
Token ID | 12008 |
Feature activation | +1.65e+00 |
Pos logit contributions | |
boy | +0.510 |
was | +0.473 |
little | +0.469 |
girl | +0.464 |
man | +0.457 |
Neg logit contributions | |
beh | -1.129 |
yours | -1.077 |
possible | -0.978 |
ours | -0.957 |
ieve | -0.946 |
Token | of |
Token ID | 286 |
Feature activation | +7.02e+00 |
Pos logit contributions | |
boy | +0.061 |
little | +0.060 |
was | +0.059 |
girl | +0.057 |
and | +0.055 |
Neg logit contributions | |
beh | -0.111 |
yours | -0.102 |
ours | -0.094 |
possible | -0.093 |
ieve | -0.091 |
Token | the |
Token ID | 262 |
Feature activation | +8.79e+00 |
Pos logit contributions | |
little | +0.268 |
boy | +0.259 |
was | +0.255 |
sky | +0.244 |
man | +0.243 |
Neg logit contributions | |
beh | -0.461 |
yours | -0.393 |
ours | -0.360 |
possible | -0.352 |
rying | -0.351 |
Token | big |
Token ID | 1263 |
Feature activation | +9.51e+00 |
Pos logit contributions | |
boy | +0.249 |
girl | +0.224 |
sky | +0.221 |
man | +0.217 |
little | +0.216 |
Neg logit contributions | |
beh | -0.613 |
yours | -0.598 |
possible | -0.541 |
ours | -0.528 |
theirs | -0.514 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.63e+01 |
Pos logit contributions | |
boy | +0.204 |
was | +0.184 |
sky | +0.183 |
girl | +0.180 |
little | +0.177 |
Neg logit contributions | |
yours | -0.685 |
beh | -0.674 |
possible | -0.615 |
ours | -0.607 |
theirs | -0.588 |
Token | ins |
Token ID | 1040 |
Feature activation | -5.90e+00 |
Pos logit contributions | |
boy | +1.026 |
girl | +0.941 |
was | +0.923 |
man | +0.919 |
little | +0.915 |
Neg logit contributions | |
beh | -1.598 |
yours | -1.589 |
possible | -1.414 |
ours | -1.400 |
theirs | -1.345 |
Token | . |
Token ID | 13 |
Feature activation | -7.86e+00 |
Pos logit contributions | |
beh | +0.403 |
yours | +0.391 |
possible | +0.349 |
ours | +0.347 |
unnecessary | +0.336 |
Neg logit contributions | |
boy | -0.190 |
was | -0.173 |
man | -0.172 |
girl | -0.172 |
sky | -0.171 |
Token | They |
Token ID | 1119 |
Feature activation | +2.15e+00 |
Pos logit contributions | |
beh | +0.534 |
yours | +0.521 |
possible | +0.465 |
ours | +0.458 |
theirs | +0.448 |
Neg logit contributions | |
boy | -0.243 |
girl | -0.222 |
man | -0.217 |
sky | -0.216 |
little | -0.214 |
Token | were |
Token ID | 547 |
Feature activation | +7.19e+00 |
Pos logit contributions | |
boy | +0.061 |
girl | +0.056 |
sky | +0.056 |
man | +0.053 |
little | +0.053 |
Neg logit contributions | |
beh | -0.150 |
yours | -0.144 |
possible | -0.129 |
ours | -0.128 |
theirs | -0.126 |
Token | selfish |
Token ID | 20363 |
Feature activation | -2.27e+00 |
Pos logit contributions | |
boy | +0.273 |
little | +0.265 |
sky | +0.254 |
man | +0.252 |
girl | +0.250 |
Neg logit contributions | |
beh | -0.445 |
yours | -0.420 |
possible | -0.372 |
ours | -0.358 |
ieve | -0.354 |
Token | making |
Token ID | 1642 |
Feature activation | -1.81e+00 |
Pos logit contributions | |
beh | +0.051 |
yours | +0.049 |
eking | +0.043 |
iable | +0.042 |
properly | +0.042 |
Neg logit contributions | |
sky | -0.032 |
boy | -0.030 |
girl | -0.030 |
fence | -0.028 |
man | -0.027 |
Token | noises |
Token ID | 26782 |
Feature activation | -8.32e-02 |
Pos logit contributions | |
beh | +0.115 |
yours | +0.103 |
possible | +0.093 |
ours | +0.091 |
rying | +0.091 |
Neg logit contributions | |
little | -0.070 |
boy | -0.069 |
was | -0.066 |
sky | -0.064 |
girl | -0.063 |
Token | . |
Token ID | 13 |
Feature activation | -6.41e+00 |
Pos logit contributions | |
beh | +0.007 |
yours | +0.006 |
ours | +0.006 |
unnecessary | +0.006 |
possible | +0.006 |
Neg logit contributions | |
boy | -0.002 |
was | -0.002 |
and | -0.002 |
little | -0.002 |
sky | -0.002 |
Token | The |
Token ID | 383 |
Feature activation | +8.89e+00 |
Pos logit contributions | |
beh | +0.453 |
yours | +0.442 |
possible | +0.397 |
ours | +0.390 |
theirs | +0.382 |
Neg logit contributions | |
boy | -0.180 |
girl | -0.163 |
sky | -0.159 |
man | -0.159 |
little | -0.157 |
Token | lost |
Token ID | 2626 |
Feature activation | -6.88e-01 |
Pos logit contributions | |
boy | +0.184 |
was | +0.167 |
sky | +0.165 |
girl | +0.163 |
man | +0.156 |
Neg logit contributions | |
yours | -0.632 |
beh | -0.631 |
possible | -0.572 |
ours | -0.562 |
unnecessary | -0.545 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.62e+01 |
Pos logit contributions | |
yours | +0.049 |
beh | +0.049 |
possible | +0.043 |
ours | +0.043 |
theirs | +0.043 |
Neg logit contributions | |
boy | -0.017 |
sky | -0.015 |
girl | -0.015 |
man | -0.015 |
was | -0.014 |
Token | in |
Token ID | 259 |
Feature activation | -7.33e-01 |
Pos logit contributions | |
boy | +1.111 |
girl | +1.037 |
sky | +1.022 |
man | +1.016 |
little | +1.015 |
Neg logit contributions | |
beh | -1.509 |
yours | -1.454 |
possible | -1.280 |
ours | -1.247 |
theirs | -1.212 |
Token | joined |
Token ID | 5399 |
Feature activation | +2.39e+00 |
Pos logit contributions | |
beh | +0.055 |
yours | +0.053 |
possible | +0.048 |
ours | +0.048 |
theirs | +0.047 |
Neg logit contributions | |
boy | -0.017 |
girl | -0.016 |
sky | -0.015 |
man | -0.015 |
little | -0.014 |
Token | them |
Token ID | 606 |
Feature activation | -8.12e+00 |
Pos logit contributions | |
boy | +0.091 |
girl | +0.084 |
little | +0.084 |
man | +0.082 |
sky | +0.081 |
Neg logit contributions | |
beh | -0.149 |
yours | -0.143 |
possible | -0.129 |
ours | -0.125 |
unnecessary | -0.121 |
Token | and |
Token ID | 290 |
Feature activation | +1.46e+01 |
Pos logit contributions | |
beh | +0.567 |
yours | +0.541 |
possible | +0.489 |
ours | +0.478 |
unnecessary | +0.471 |
Neg logit contributions | |
boy | -0.252 |
little | -0.235 |
was | -0.229 |
girl | -0.226 |
sky | -0.225 |
Token | he |
Token ID | 339 |
Feature activation | -4.39e+00 |
Pos logit contributions | |
boy | +0.490 |
girl | +0.443 |
man | +0.440 |
was | +0.439 |
little | +0.433 |
Neg logit contributions | |
beh | -0.983 |
yours | -0.950 |
possible | -0.854 |
ours | -0.839 |
unnecessary | -0.816 |
Token | they |
Token ID | 484 |
Feature activation | +3.07e+00 |
Pos logit contributions | |
beh | +0.520 |
yours | +0.506 |
possible | +0.450 |
ours | +0.444 |
theirs | +0.436 |
Neg logit contributions | |
boy | -0.224 |
girl | -0.203 |
sky | -0.203 |
man | -0.196 |
little | -0.196 |
Token | said |
Token ID | 531 |
Feature activation | +4.52e+00 |
Pos logit contributions | |
boy | +0.099 |
girl | +0.090 |
little | +0.090 |
man | +0.089 |
was | +0.088 |
Neg logit contributions | |
beh | -0.211 |
yours | -0.204 |
possible | -0.187 |
ours | -0.180 |
unnecessary | -0.177 |
Token | . |
Token ID | 13 |
Feature activation | -6.99e+00 |
Pos logit contributions | |
was | +0.152 |
little | +0.152 |
boy | +0.150 |
man | +0.142 |
and | +0.138 |
Neg logit contributions | |
beh | -0.309 |
yours | -0.293 |
ours | -0.267 |
possible | -0.261 |
unnecessary | -0.260 |
Token | " |
Token ID | 366 |
Feature activation | +9.26e+00 |
Pos logit contributions | |
yours | +0.492 |
beh | +0.485 |
theirs | +0.439 |
possible | +0.437 |
ours | +0.430 |
Neg logit contributions | |
boy | -0.169 |
girl | -0.154 |
man | -0.144 |
sky | -0.140 |
little | -0.136 |
Token | The |
Token ID | 464 |
Feature activation | +1.24e+01 |
Pos logit contributions | |
boy | +0.320 |
girl | +0.294 |
little | +0.293 |
sky | +0.292 |
man | +0.290 |
Neg logit contributions | |
beh | -0.598 |
yours | -0.576 |
possible | -0.515 |
ours | -0.501 |
theirs | -0.491 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.62e+01 |
Pos logit contributions | |
boy | +0.339 |
girl | +0.305 |
sky | +0.301 |
was | +0.300 |
little | +0.293 |
Neg logit contributions | |
beh | -0.853 |
yours | -0.848 |
possible | -0.763 |
ours | -0.747 |
theirs | -0.725 |
Token | in |
Token ID | 259 |
Feature activation | -7.84e-01 |
Pos logit contributions | |
boy | +1.086 |
girl | +1.005 |
man | +0.983 |
little | +0.973 |
was | +0.972 |
Neg logit contributions | |
beh | -1.538 |
yours | -1.490 |
possible | -1.327 |
ours | -1.296 |
theirs | -1.263 |
Token | is |
Token ID | 318 |
Feature activation | +6.40e+00 |
Pos logit contributions | |
beh | +0.056 |
yours | +0.054 |
possible | +0.049 |
ours | +0.047 |
ieve | +0.046 |
Neg logit contributions | |
boy | -0.020 |
girl | -0.019 |
sky | -0.019 |
little | -0.018 |
man | -0.018 |
Token | mean |
Token ID | 1612 |
Feature activation | -4.15e+00 |
Pos logit contributions | |
little | +0.240 |
boy | +0.237 |
sky | +0.220 |
was | +0.220 |
man | +0.220 |
Neg logit contributions | |
beh | -0.420 |
yours | -0.359 |
rying | -0.327 |
ieve | -0.327 |
possible | -0.325 |
Token | ! |
Token ID | 0 |
Feature activation | -9.92e+00 |
Pos logit contributions | |
beh | +0.278 |
yours | +0.246 |
unnecessary | +0.230 |
ieve | +0.230 |
possible | +0.228 |
Neg logit contributions | |
boy | -0.161 |
and | -0.147 |
little | -0.146 |
girl | -0.146 |
man | -0.145 |
Token | He |
Token ID | 679 |
Feature activation | -3.70e-01 |
Pos logit contributions | |
beh | +0.570 |
yours | +0.550 |
possible | +0.482 |
ours | +0.471 |
theirs | +0.458 |
Neg logit contributions | |
boy | -0.417 |
girl | -0.390 |
sky | -0.386 |
man | -0.383 |
little | -0.381 |
Token | , |
Token ID | 11 |
Feature activation | +1.23e+01 |
Pos logit contributions | |
boy | +0.691 |
girl | +0.649 |
man | +0.639 |
sky | +0.635 |
little | +0.603 |
Neg logit contributions | |
yours | -1.082 |
beh | -1.061 |
possible | -0.969 |
ours | -0.927 |
theirs | -0.915 |
Token | there |
Token ID | 612 |
Feature activation | -5.15e-01 |
Pos logit contributions | |
man | +0.542 |
boy | +0.542 |
sky | +0.533 |
girl | +0.518 |
woman | +0.503 |
Neg logit contributions | |
yours | -0.626 |
beh | -0.569 |
possible | -0.523 |
ours | -0.522 |
theirs | -0.502 |
Token | was |
Token ID | 373 |
Feature activation | +1.05e+01 |
Pos logit contributions | |
yours | +0.035 |
anything | +0.031 |
beh | +0.031 |
possible | +0.031 |
theirs | +0.031 |
Neg logit contributions | |
woman | -0.010 |
girl | -0.010 |
man | -0.010 |
boy | -0.009 |
sky | -0.009 |
Token | a |
Token ID | 257 |
Feature activation | -2.79e-01 |
Pos logit contributions | |
boy | +0.439 |
man | +0.418 |
girl | +0.411 |
sky | +0.398 |
was | +0.384 |
Neg logit contributions | |
yours | -0.594 |
beh | -0.586 |
ours | -0.509 |
possible | -0.504 |
theirs | -0.495 |
Token | little |
Token ID | 1310 |
Feature activation | +1.25e+01 |
Pos logit contributions | |
yours | +0.018 |
ours | +0.016 |
beh | +0.016 |
possible | +0.015 |
asty | +0.015 |
Neg logit contributions | |
sky | -0.006 |
was | -0.005 |
led | -0.004 |
arrived | -0.003 |
At | -0.003 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.60e+01 |
Pos logit contributions | |
was | +0.176 |
At | +0.163 |
sky | +0.157 |
led | +0.128 |
man | +0.093 |
Neg logit contributions | |
yours | -0.934 |
beh | -0.865 |
ours | -0.832 |
possible | -0.798 |
ourselves | -0.775 |
Token | in |
Token ID | 259 |
Feature activation | -5.97e-01 |
Pos logit contributions | |
boy | +1.067 |
girl | +0.996 |
sky | +0.985 |
man | +0.976 |
little | +0.973 |
Neg logit contributions | |
beh | -1.526 |
yours | -1.468 |
possible | -1.293 |
ours | -1.262 |
theirs | -1.227 |
Token | . |
Token ID | 13 |
Feature activation | -8.41e+00 |
Pos logit contributions | |
beh | +0.036 |
yours | +0.034 |
possible | +0.030 |
theirs | +0.029 |
ours | +0.028 |
Neg logit contributions | |
boy | -0.021 |
sky | -0.020 |
girl | -0.020 |
man | -0.020 |
little | -0.019 |
Token | He |
Token ID | 679 |
Feature activation | -5.64e-01 |
Pos logit contributions | |
yours | +0.462 |
beh | +0.460 |
possible | +0.391 |
ours | +0.390 |
theirs | +0.389 |
Neg logit contributions | |
boy | -0.349 |
girl | -0.337 |
man | -0.331 |
sky | -0.321 |
little | -0.320 |
Token | was |
Token ID | 373 |
Feature activation | +9.69e+00 |
Pos logit contributions | |
beh | +0.036 |
yours | +0.036 |
supposed | +0.031 |
possible | +0.031 |
theirs | +0.030 |
Neg logit contributions | |
girl | -0.015 |
sky | -0.014 |
boy | -0.014 |
woman | -0.013 |
man | -0.013 |
Token | at |
Token ID | 379 |
Feature activation | -4.75e+00 |
Pos logit contributions | |
boy | +0.375 |
girl | +0.355 |
sky | +0.350 |
man | +0.350 |
was | +0.330 |
Neg logit contributions | |
beh | -0.570 |
yours | -0.563 |
possible | -0.490 |
ours | -0.484 |
theirs | -0.471 |
Token | Why |
Token ID | 5195 |
Feature activation | -5.73e+00 |
Pos logit contributions | |
boy | +0.423 |
sky | +0.405 |
little | +0.396 |
girl | +0.392 |
man | +0.388 |
Neg logit contributions | |
beh | -0.711 |
yours | -0.671 |
possible | -0.606 |
ours | -0.574 |
theirs | -0.568 |
Token | did |
Token ID | 750 |
Feature activation | +6.51e+00 |
Pos logit contributions | |
beh | +0.347 |
yours | +0.333 |
possible | +0.295 |
ours | +0.288 |
theirs | +0.280 |
Neg logit contributions | |
boy | -0.227 |
girl | -0.211 |
sky | -0.208 |
man | -0.207 |
little | -0.207 |
Token | you |
Token ID | 345 |
Feature activation | -1.44e+01 |
Pos logit contributions | |
boy | +0.345 |
little | +0.338 |
man | +0.327 |
was | +0.326 |
girl | +0.314 |
Neg logit contributions | |
beh | -0.320 |
yours | -0.278 |
rying | -0.275 |
eking | -0.246 |
possible | -0.244 |
Token | rip |
Token ID | 18344 |
Feature activation | -1.19e+00 |
Pos logit contributions | |
beh | +0.902 |
yours | +0.845 |
possible | +0.748 |
ours | +0.743 |
unnecessary | +0.734 |
Neg logit contributions | |
boy | -0.561 |
little | -0.557 |
man | -0.534 |
was | -0.533 |
sky | -0.511 |
Token | the |
Token ID | 262 |
Feature activation | +8.08e+00 |
Pos logit contributions | |
beh | +0.071 |
yours | +0.065 |
possible | +0.059 |
ours | +0.056 |
theirs | +0.056 |
Neg logit contributions | |
boy | -0.050 |
little | -0.047 |
girl | -0.046 |
man | -0.046 |
was | -0.046 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.59e+01 |
Pos logit contributions | |
boy | +0.253 |
girl | +0.231 |
sky | +0.226 |
man | +0.223 |
little | +0.222 |
Neg logit contributions | |
beh | -0.537 |
yours | -0.531 |
possible | -0.473 |
ours | -0.466 |
theirs | -0.451 |
Token | in |
Token ID | 259 |
Feature activation | -8.86e-01 |
Pos logit contributions | |
boy | +1.011 |
girl | +0.923 |
man | +0.917 |
little | +0.908 |
sky | +0.897 |
Neg logit contributions | |
beh | -1.576 |
yours | -1.530 |
possible | -1.370 |
ours | -1.342 |
theirs | -1.320 |
Token | ?" |
Token ID | 1701 |
Feature activation | -8.44e+00 |
Pos logit contributions | |
beh | +0.045 |
yours | +0.042 |
possible | +0.036 |
ours | +0.036 |
theirs | +0.035 |
Neg logit contributions | |
boy | -0.045 |
girl | -0.042 |
man | -0.042 |
little | -0.041 |
was | -0.041 |
Token | Mom |
Token ID | 11254 |
Feature activation | +4.77e+00 |
Pos logit contributions | |
beh | +0.522 |
yours | +0.506 |
possible | +0.450 |
ours | +0.438 |
theirs | +0.429 |
Neg logit contributions | |
boy | -0.310 |
girl | -0.287 |
sky | -0.285 |
man | -0.279 |
little | -0.279 |
Token | asks |
Token ID | 7893 |
Feature activation | +4.49e-01 |
Pos logit contributions | |
boy | +0.151 |
girl | +0.137 |
sky | +0.135 |
man | +0.134 |
little | +0.133 |
Neg logit contributions | |
beh | -0.326 |
yours | -0.317 |
possible | -0.285 |
ours | -0.279 |
theirs | -0.273 |
Token | . |
Token ID | 13 |
Feature activation | -8.63e+00 |
Pos logit contributions | |
boy | +0.018 |
little | +0.017 |
girl | +0.017 |
was | +0.017 |
man | +0.016 |
Neg logit contributions | |
beh | -0.028 |
yours | -0.028 |
possible | -0.025 |
ours | -0.024 |
unnecessary | -0.023 |
Token |
|
Token ID | 198 |
Feature activation | +9.23e+00 |
Pos logit contributions | |
boy | +0.205 |
girl | +0.186 |
sky | +0.171 |
man | +0.171 |
little | +0.165 |
Neg logit contributions | |
beh | -0.706 |
yours | -0.699 |
possible | -0.631 |
ours | -0.620 |
theirs | -0.615 |
Token | They |
Token ID | 2990 |
Feature activation | -8.35e-01 |
Pos logit contributions | |
boy | +0.290 |
girl | +0.264 |
sky | +0.263 |
man | +0.255 |
little | +0.254 |
Neg logit contributions | |
beh | -0.632 |
yours | -0.611 |
possible | -0.547 |
ours | -0.536 |
theirs | -0.522 |
Token | also |
Token ID | 635 |
Feature activation | +7.02e+00 |
Pos logit contributions | |
beh | +0.053 |
yours | +0.051 |
possible | +0.046 |
ours | +0.045 |
theirs | +0.044 |
Neg logit contributions | |
boy | -0.029 |
girl | -0.027 |
sky | -0.026 |
little | -0.025 |
man | -0.025 |
Token | realized |
Token ID | 6939 |
Feature activation | +5.16e+00 |
Pos logit contributions | |
boy | +0.227 |
girl | +0.209 |
sky | +0.209 |
little | +0.201 |
man | +0.199 |
Neg logit contributions | |
beh | -0.464 |
yours | -0.446 |
possible | -0.400 |
ours | -0.392 |
theirs | -0.389 |
Token | that |
Token ID | 326 |
Feature activation | -5.53e+00 |
Pos logit contributions | |
boy | +0.195 |
little | +0.184 |
was | +0.183 |
man | +0.182 |
girl | +0.178 |
Neg logit contributions | |
beh | -0.328 |
yours | -0.309 |
ours | -0.270 |
possible | -0.270 |
unnecessary | -0.263 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.59e+01 |
Pos logit contributions | |
beh | +0.339 |
yours | +0.325 |
possible | +0.286 |
ours | +0.282 |
theirs | +0.272 |
Neg logit contributions | |
boy | -0.216 |
girl | -0.199 |
little | -0.196 |
sky | -0.196 |
man | -0.195 |
Token | ins |
Token ID | 1040 |
Feature activation | -6.32e+00 |
Pos logit contributions | |
boy | +1.255 |
girl | +1.178 |
sky | +1.160 |
little | +1.158 |
man | +1.156 |
Neg logit contributions | |
beh | -1.319 |
yours | -1.282 |
possible | -1.110 |
ours | -1.080 |
theirs | -1.041 |
Token | were |
Token ID | 547 |
Feature activation | +7.42e+00 |
Pos logit contributions | |
beh | +0.480 |
yours | +0.467 |
possible | +0.425 |
ours | +0.418 |
theirs | +0.407 |
Neg logit contributions | |
boy | -0.149 |
girl | -0.131 |
sky | -0.129 |
man | -0.127 |
little | -0.126 |
Token | not |
Token ID | 407 |
Feature activation | -8.65e+00 |
Pos logit contributions | |
little | +0.300 |
boy | +0.296 |
was | +0.288 |
sky | +0.277 |
man | +0.276 |
Neg logit contributions | |
beh | -0.446 |
yours | -0.399 |
rying | -0.347 |
possible | -0.347 |
ours | -0.345 |
Token | pets |
Token ID | 17252 |
Feature activation | +5.47e+00 |
Pos logit contributions | |
beh | +0.552 |
yours | +0.486 |
ieve | +0.449 |
eking | +0.439 |
rying | +0.438 |
Neg logit contributions | |
little | -0.337 |
boy | -0.331 |
girl | -0.305 |
was | -0.297 |
man | -0.293 |
Token | . |
Token ID | 13 |
Feature activation | -8.41e+00 |
Pos logit contributions | |
boy | +0.241 |
little | +0.233 |
was | +0.233 |
girl | +0.224 |
man | +0.224 |
Neg logit contributions | |
beh | -0.316 |
yours | -0.287 |
possible | -0.253 |
ours | -0.252 |
eking | -0.245 |
Token | a |
Token ID | 257 |
Feature activation | +5.68e-01 |
Pos logit contributions | |
boy | +0.153 |
girl | +0.138 |
little | +0.138 |
man | +0.135 |
sky | +0.132 |
Neg logit contributions | |
beh | -0.353 |
yours | -0.332 |
possible | -0.301 |
ours | -0.293 |
eking | -0.290 |
Token | place |
Token ID | 1295 |
Feature activation | +5.78e+00 |
Pos logit contributions | |
sky | +0.015 |
boy | +0.014 |
was | +0.014 |
girl | +0.013 |
man | +0.012 |
Neg logit contributions | |
yours | -0.035 |
possible | -0.033 |
ours | -0.032 |
beh | -0.032 |
unnecessary | -0.029 |
Token | where |
Token ID | 810 |
Feature activation | -9.68e+00 |
Pos logit contributions | |
boy | +0.242 |
girl | +0.224 |
man | +0.222 |
sky | +0.222 |
little | +0.221 |
Neg logit contributions | |
beh | -0.345 |
yours | -0.325 |
possible | -0.284 |
unnecessary | -0.279 |
ours | -0.278 |
Token | there |
Token ID | 612 |
Feature activation | -6.40e-01 |
Pos logit contributions | |
beh | +0.597 |
yours | +0.587 |
possible | +0.523 |
ours | +0.513 |
theirs | +0.502 |
Neg logit contributions | |
boy | -0.349 |
girl | -0.325 |
sky | -0.321 |
man | -0.320 |
little | -0.314 |
Token | are |
Token ID | 389 |
Feature activation | +7.30e-01 |
Pos logit contributions | |
yours | +0.045 |
beh | +0.045 |
possible | +0.041 |
ours | +0.040 |
theirs | +0.040 |
Neg logit contributions | |
boy | -0.016 |
girl | -0.015 |
sky | -0.014 |
little | -0.014 |
man | -0.014 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.58e+01 |
Pos logit contributions | |
boy | +0.027 |
girl | +0.025 |
man | +0.025 |
sky | +0.024 |
little | +0.023 |
Neg logit contributions | |
beh | -0.044 |
yours | -0.044 |
possible | -0.039 |
ours | -0.038 |
theirs | -0.037 |
Token | ins |
Token ID | 1040 |
Feature activation | -6.16e+00 |
Pos logit contributions | |
boy | +1.209 |
girl | +1.129 |
sky | +1.115 |
little | +1.113 |
man | +1.110 |
Neg logit contributions | |
beh | -1.364 |
yours | -1.330 |
possible | -1.162 |
ours | -1.127 |
theirs | -1.099 |
Token | . |
Token ID | 13 |
Feature activation | -7.79e+00 |
Pos logit contributions | |
beh | +0.394 |
yours | +0.371 |
possible | +0.331 |
unnecessary | +0.330 |
ours | +0.324 |
Neg logit contributions | |
boy | -0.231 |
was | -0.217 |
sky | -0.213 |
man | -0.212 |
girl | -0.210 |
Token | The |
Token ID | 383 |
Feature activation | +7.51e+00 |
Pos logit contributions | |
yours | +0.480 |
beh | +0.478 |
ours | +0.424 |
theirs | +0.423 |
possible | +0.416 |
Neg logit contributions | |
boy | -0.263 |
girl | -0.249 |
man | -0.248 |
little | -0.237 |
sky | -0.236 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.53e+01 |
Pos logit contributions | |
boy | +0.158 |
was | +0.147 |
girl | +0.144 |
sky | +0.144 |
man | +0.136 |
Neg logit contributions | |
yours | -0.521 |
beh | -0.517 |
possible | -0.472 |
ours | -0.466 |
theirs | -0.449 |
Token | ins |
Token ID | 1040 |
Feature activation | -6.72e+00 |
Pos logit contributions | |
boy | +1.212 |
girl | +1.150 |
sky | +1.145 |
little | +1.133 |
man | +1.121 |
Neg logit contributions | |
beh | -1.309 |
yours | -1.245 |
possible | -1.068 |
ours | -1.041 |
ieve | -1.008 |
Token |
|
Token ID | 198 |
Feature activation | +8.59e+00 |
Pos logit contributions | |
beh | +0.524 |
yours | +0.518 |
possible | +0.462 |
ours | +0.453 |
theirs | +0.446 |
Neg logit contributions | |
boy | -0.220 |
girl | -0.205 |
man | -0.198 |
little | -0.193 |
sky | -0.192 |
Token |
|
Token ID | 198 |
Feature activation | +8.45e+00 |
Pos logit contributions | |
boy | +0.246 |
girl | +0.235 |
sky | +0.217 |
man | +0.211 |
little | +0.204 |
Neg logit contributions | |
beh | -0.586 |
yours | -0.580 |
possible | -0.528 |
theirs | -0.507 |
ours | -0.501 |
Token | The |
Token ID | 464 |
Feature activation | +1.17e+01 |
Pos logit contributions | |
boy | +0.291 |
girl | +0.272 |
sky | +0.266 |
man | +0.258 |
little | +0.256 |
Neg logit contributions | |
beh | -0.553 |
yours | -0.529 |
possible | -0.479 |
ours | -0.455 |
theirs | -0.449 |
Token | little |
Token ID | 1310 |
Feature activation | +1.18e+01 |
Pos logit contributions | |
boy | +0.252 |
was | +0.240 |
sky | +0.232 |
girl | +0.220 |
man | +0.218 |
Neg logit contributions | |
beh | -0.817 |
yours | -0.810 |
possible | -0.752 |
ours | -0.716 |
unnecessary | -0.705 |
Token | sad |
Token ID | 6507 |
Feature activation | +1.31e+00 |
Pos logit contributions | |
sky | +0.068 |
was | +0.063 |
man | +0.046 |
boy | +0.035 |
little | +0.029 |
Neg logit contributions | |
yours | -0.977 |
beh | -0.976 |
possible | -0.898 |
ours | -0.871 |
unnecessary | -0.848 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.56e+01 |
Pos logit contributions | |
boy | +0.029 |
sky | +0.027 |
girl | +0.026 |
man | +0.025 |
was | +0.025 |
Neg logit contributions | |
beh | -0.094 |
yours | -0.093 |
possible | -0.084 |
theirs | -0.081 |
ours | -0.081 |
Token | in |
Token ID | 259 |
Feature activation | -1.05e+00 |
Pos logit contributions | |
boy | +1.070 |
girl | +1.000 |
sky | +0.992 |
man | +0.981 |
little | +0.979 |
Neg logit contributions | |
beh | -1.488 |
yours | -1.426 |
possible | -1.255 |
ours | -1.221 |
theirs | -1.189 |
Token | was |
Token ID | 373 |
Feature activation | +9.50e+00 |
Pos logit contributions | |
beh | +0.077 |
yours | +0.071 |
possible | +0.065 |
properly | +0.063 |
ieve | +0.063 |
Neg logit contributions | |
boy | -0.023 |
girl | -0.022 |
sky | -0.021 |
man | -0.019 |
little | -0.019 |
Token | no |
Token ID | 645 |
Feature activation | -4.77e+00 |
Pos logit contributions | |
boy | +0.371 |
little | +0.354 |
sky | +0.346 |
was | +0.345 |
girl | +0.342 |
Neg logit contributions | |
beh | -0.577 |
yours | -0.547 |
ours | -0.478 |
possible | -0.477 |
unnecessary | -0.463 |
Token | longer |
Token ID | 2392 |
Feature activation | -1.14e+01 |
Pos logit contributions | |
beh | +0.210 |
yours | +0.196 |
ours | +0.162 |
possible | +0.158 |
theirs | +0.150 |
Neg logit contributions | |
boy | -0.283 |
little | -0.271 |
girl | -0.269 |
man | -0.260 |
was | -0.258 |
Token | alone |
Token ID | 3436 |
Feature activation | -1.29e+01 |
Pos logit contributions | |
beh | +0.682 |
yours | +0.659 |
possible | +0.580 |
ours | +0.569 |
theirs | +0.551 |
Neg logit contributions | |
boy | -0.453 |
girl | -0.421 |
sky | -0.415 |
little | -0.414 |
man | -0.412 |
Token | , |
Token ID | 11 |
Feature activation | +1.12e+01 |
Pos logit contributions | |
boy | +0.166 |
girl | +0.154 |
man | +0.153 |
sky | +0.149 |
woman | +0.140 |
Neg logit contributions | |
yours | -0.388 |
beh | -0.381 |
possible | -0.350 |
ours | -0.338 |
theirs | -0.329 |
Token | there |
Token ID | 612 |
Feature activation | -1.47e+00 |
Pos logit contributions | |
boy | +0.489 |
man | +0.484 |
sky | +0.479 |
girl | +0.463 |
woman | +0.446 |
Neg logit contributions | |
yours | -0.575 |
beh | -0.546 |
possible | -0.499 |
ours | -0.484 |
theirs | -0.462 |
Token | was |
Token ID | 373 |
Feature activation | +9.59e+00 |
Pos logit contributions | |
yours | +0.093 |
possible | +0.084 |
beh | +0.083 |
anymore | +0.083 |
anything | +0.083 |
Neg logit contributions | |
woman | -0.035 |
girl | -0.033 |
man | -0.031 |
boy | -0.030 |
sky | -0.029 |
Token | a |
Token ID | 257 |
Feature activation | -1.07e+00 |
Pos logit contributions | |
boy | +0.412 |
man | +0.397 |
girl | +0.387 |
sky | +0.372 |
was | +0.361 |
Neg logit contributions | |
yours | -0.528 |
beh | -0.516 |
possible | -0.449 |
ours | -0.446 |
theirs | -0.436 |
Token | little |
Token ID | 1310 |
Feature activation | +1.17e+01 |
Pos logit contributions | |
yours | +0.068 |
ours | +0.062 |
beh | +0.061 |
possible | +0.060 |
asty | +0.057 |
Neg logit contributions | |
sky | -0.022 |
was | -0.018 |
led | -0.015 |
At | -0.014 |
arrived | -0.012 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.55e+01 |
Pos logit contributions | |
was | +0.155 |
At | +0.145 |
sky | +0.138 |
led | +0.099 |
Quick | +0.080 |
Neg logit contributions | |
yours | -0.891 |
beh | -0.834 |
ours | -0.798 |
possible | -0.775 |
ourselves | -0.740 |
Token | in |
Token ID | 259 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
boy | +1.050 |
girl | +0.981 |
sky | +0.970 |
man | +0.961 |
little | +0.959 |
Neg logit contributions | |
beh | -1.498 |
yours | -1.440 |
possible | -1.268 |
ours | -1.236 |
theirs | -1.203 |
Token | named |
Token ID | 3706 |
Feature activation | +5.85e+00 |
Pos logit contributions | |
beh | +0.067 |
yours | +0.063 |
possible | +0.055 |
theirs | +0.053 |
ours | +0.052 |
Neg logit contributions | |
boy | -0.040 |
sky | -0.038 |
girl | -0.038 |
man | -0.036 |
little | -0.035 |
Token | P |
Token ID | 350 |
Feature activation | +1.70e+00 |
Pos logit contributions | |
boy | +0.243 |
little | +0.228 |
sky | +0.224 |
girl | +0.223 |
man | +0.222 |
Neg logit contributions | |
beh | -0.355 |
yours | -0.325 |
possible | -0.295 |
ours | -0.280 |
unnecessary | -0.280 |
Token | ingu |
Token ID | 6680 |
Feature activation | -1.27e+00 |
Pos logit contributions | |
boy | +0.086 |
girl | +0.081 |
sky | +0.080 |
man | +0.080 |
little | +0.079 |
Neg logit contributions | |
beh | -0.084 |
yours | -0.080 |
possible | -0.069 |
ours | -0.067 |
theirs | -0.064 |
Token | who |
Token ID | 508 |
Feature activation | -4.35e+00 |
Pos logit contributions | |
beh | +0.065 |
yours | +0.062 |
possible | +0.053 |
ours | +0.052 |
theirs | +0.050 |
Neg logit contributions | |
boy | -0.062 |
girl | -0.059 |
sky | -0.058 |
man | -0.058 |
little | -0.058 |
Token | , |
Token ID | 11 |
Feature activation | +1.18e+01 |
Pos logit contributions | |
boy | +0.476 |
sky | +0.433 |
girl | +0.432 |
man | +0.423 |
little | +0.415 |
Neg logit contributions | |
beh | -1.093 |
yours | -1.075 |
possible | -0.959 |
ours | -0.942 |
theirs | -0.910 |
Token | can |
Token ID | 460 |
Feature activation | +4.69e+00 |
Pos logit contributions | |
boy | +0.508 |
girl | +0.476 |
sky | +0.475 |
man | +0.468 |
little | +0.464 |
Neg logit contributions | |
beh | -0.663 |
yours | -0.646 |
possible | -0.562 |
ours | -0.551 |
theirs | -0.534 |
Token | I |
Token ID | 314 |
Feature activation | -9.86e+00 |
Pos logit contributions | |
boy | +0.209 |
was | +0.198 |
little | +0.197 |
girl | +0.193 |
man | +0.192 |
Neg logit contributions | |
beh | -0.273 |
yours | -0.249 |
possible | -0.225 |
unnecessary | -0.224 |
theirs | -0.219 |
Token | touch |
Token ID | 3638 |
Feature activation | -5.82e+00 |
Pos logit contributions | |
beh | +0.582 |
yours | +0.547 |
possible | +0.486 |
ours | +0.478 |
theirs | +0.470 |
Neg logit contributions | |
boy | -0.416 |
little | -0.393 |
girl | -0.391 |
was | -0.384 |
man | -0.383 |
Token | the |
Token ID | 262 |
Feature activation | +7.37e+00 |
Pos logit contributions | |
beh | +0.384 |
yours | +0.359 |
possible | +0.327 |
ours | +0.315 |
theirs | +0.311 |
Neg logit contributions | |
boy | -0.206 |
sky | -0.190 |
girl | -0.189 |
little | -0.188 |
man | -0.185 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.54e+01 |
Pos logit contributions | |
boy | +0.195 |
girl | +0.173 |
man | +0.168 |
was | +0.167 |
sky | +0.165 |
Neg logit contributions | |
beh | -0.519 |
yours | -0.518 |
possible | -0.467 |
ours | -0.461 |
theirs | -0.443 |
Token | in |
Token ID | 259 |
Feature activation | -1.14e+00 |
Pos logit contributions | |
boy | +0.942 |
girl | +0.858 |
man | +0.852 |
little | +0.832 |
sky | +0.829 |
Neg logit contributions | |
beh | -1.600 |
yours | -1.544 |
possible | -1.393 |
ours | -1.364 |
theirs | -1.352 |
Token | ?" |
Token ID | 1701 |
Feature activation | -8.75e+00 |
Pos logit contributions | |
beh | +0.061 |
yours | +0.056 |
unnecessary | +0.049 |
possible | +0.048 |
theirs | +0.048 |
Neg logit contributions | |
boy | -0.055 |
man | -0.052 |
girl | -0.052 |
was | -0.051 |
little | -0.051 |
Token | Tim |
Token ID | 5045 |
Feature activation | -6.08e+00 |
Pos logit contributions | |
beh | +0.514 |
yours | +0.499 |
possible | +0.439 |
ours | +0.429 |
theirs | +0.420 |
Neg logit contributions | |
boy | -0.344 |
sky | -0.322 |
girl | -0.321 |
man | -0.313 |
little | -0.313 |
Token | asked |
Token ID | 1965 |
Feature activation | +1.32e+00 |
Pos logit contributions | |
beh | +0.377 |
yours | +0.356 |
possible | +0.311 |
ours | +0.302 |
eking | +0.301 |
Neg logit contributions | |
boy | -0.212 |
sky | -0.206 |
girl | -0.197 |
man | -0.192 |
little | -0.190 |
Token | . |
Token ID | 13 |
Feature activation | -9.21e+00 |
Pos logit contributions | |
boy | +0.053 |
little | +0.050 |
was | +0.048 |
girl | +0.048 |
man | +0.048 |
Neg logit contributions | |
beh | -0.084 |
yours | -0.082 |
ours | -0.074 |
possible | -0.071 |
unnecessary | -0.071 |
Token | ily |
Token ID | 813 |
Feature activation | -1.32e+00 |
Pos logit contributions | |
boy | +0.561 |
girl | +0.535 |
little | +0.527 |
sky | +0.521 |
man | +0.521 |
Neg logit contributions | |
beh | -0.584 |
yours | -0.551 |
possible | -0.483 |
ours | -0.470 |
unnecessary | -0.454 |
Token | wanted |
Token ID | 2227 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
beh | +0.085 |
yours | +0.078 |
properly | +0.072 |
myself | +0.071 |
ieve | +0.070 |
Neg logit contributions | |
boy | -0.031 |
sky | -0.031 |
girl | -0.029 |
little | -0.026 |
child | -0.025 |
Token | to |
Token ID | 284 |
Feature activation | +3.12e+00 |
Pos logit contributions | |
boy | +0.045 |
girl | +0.041 |
sky | +0.041 |
little | +0.041 |
man | +0.041 |
Neg logit contributions | |
beh | -0.073 |
yours | -0.070 |
possible | -0.062 |
ours | -0.061 |
theirs | -0.059 |
Token | see |
Token ID | 766 |
Feature activation | -3.56e+00 |
Pos logit contributions | |
boy | +0.128 |
little | +0.124 |
man | +0.122 |
was | +0.119 |
girl | +0.117 |
Neg logit contributions | |
beh | -0.189 |
yours | -0.174 |
possible | -0.156 |
eking | -0.154 |
ours | -0.153 |
Token | the |
Token ID | 262 |
Feature activation | +7.45e+00 |
Pos logit contributions | |
beh | +0.230 |
yours | +0.206 |
possible | +0.188 |
ours | +0.180 |
eking | +0.178 |
Neg logit contributions | |
boy | -0.136 |
little | -0.132 |
girl | -0.127 |
man | -0.125 |
was | -0.122 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.54e+01 |
Pos logit contributions | |
boy | +0.198 |
girl | +0.179 |
was | +0.177 |
sky | +0.177 |
man | +0.170 |
Neg logit contributions | |
beh | -0.514 |
yours | -0.510 |
possible | -0.462 |
ours | -0.455 |
theirs | -0.438 |
Token | ins |
Token ID | 1040 |
Feature activation | -6.54e+00 |
Pos logit contributions | |
boy | +1.118 |
girl | +1.043 |
man | +1.026 |
little | +1.019 |
sky | +1.019 |
Neg logit contributions | |
beh | -1.413 |
yours | -1.379 |
possible | -1.212 |
ours | -1.185 |
theirs | -1.148 |
Token | . |
Token ID | 13 |
Feature activation | -8.77e+00 |
Pos logit contributions | |
beh | +0.459 |
yours | +0.425 |
unnecessary | +0.395 |
ours | +0.381 |
possible | +0.379 |
Neg logit contributions | |
boy | -0.212 |
was | -0.202 |
man | -0.195 |
and | -0.191 |
girl | -0.191 |
Token | She |
Token ID | 1375 |
Feature activation | +1.15e+00 |
Pos logit contributions | |
beh | +0.522 |
yours | +0.513 |
ours | +0.445 |
possible | +0.438 |
theirs | +0.431 |
Neg logit contributions | |
boy | -0.313 |
sky | -0.294 |
man | -0.290 |
girl | -0.289 |
little | -0.287 |
Token | asked |
Token ID | 1965 |
Feature activation | +1.44e+00 |
Pos logit contributions | |
sky | +0.045 |
boy | +0.038 |
girl | +0.038 |
man | +0.034 |
little | +0.033 |
Neg logit contributions | |
beh | -0.062 |
yours | -0.057 |
eking | -0.049 |
theirs | -0.048 |
ours | -0.048 |
Token | her |
Token ID | 607 |
Feature activation | -4.54e+00 |
Pos logit contributions | |
boy | +0.061 |
girl | +0.057 |
little | +0.056 |
man | +0.056 |
sky | +0.056 |
Neg logit contributions | |
beh | -0.084 |
yours | -0.080 |
possible | -0.071 |
ours | -0.069 |
theirs | -0.066 |
Token | slip |
Token ID | 13819 |
Feature activation | -4.40e+00 |
Pos logit contributions | |
boy | +0.149 |
little | +0.142 |
was | +0.140 |
girl | +0.139 |
man | +0.139 |
Neg logit contributions | |
beh | -0.221 |
yours | -0.213 |
possible | -0.187 |
ours | -0.186 |
theirs | -0.179 |
Token | . |
Token ID | 13 |
Feature activation | -7.92e+00 |
Pos logit contributions | |
beh | +0.307 |
yours | +0.283 |
ours | +0.255 |
possible | +0.251 |
ieve | +0.250 |
Neg logit contributions | |
boy | -0.147 |
was | -0.141 |
sky | -0.141 |
man | -0.136 |
and | -0.135 |
Token | Quick |
Token ID | 12029 |
Feature activation | -1.69e+00 |
Pos logit contributions | |
beh | +0.534 |
yours | +0.522 |
possible | +0.465 |
ours | +0.455 |
theirs | +0.450 |
Neg logit contributions | |
boy | -0.241 |
girl | -0.220 |
man | -0.215 |
sky | -0.210 |
little | -0.209 |
Token | ly |
Token ID | 306 |
Feature activation | -1.51e+01 |
Pos logit contributions | |
beh | +0.100 |
yours | +0.095 |
possible | +0.084 |
ours | +0.081 |
unnecessary | +0.080 |
Neg logit contributions | |
boy | -0.066 |
sky | -0.062 |
girl | -0.061 |
man | -0.061 |
little | -0.060 |
Token | the |
Token ID | 262 |
Feature activation | +7.42e+00 |
Pos logit contributions | |
beh | +0.984 |
yours | +0.942 |
possible | +0.856 |
eking | +0.820 |
theirs | +0.809 |
Neg logit contributions | |
boy | -0.418 |
sky | -0.402 |
girl | -0.374 |
man | -0.333 |
little | -0.330 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.54e+01 |
Pos logit contributions | |
boy | +0.138 |
sky | +0.122 |
was | +0.120 |
girl | +0.115 |
man | +0.113 |
Neg logit contributions | |
beh | -0.567 |
yours | -0.556 |
possible | -0.516 |
ours | -0.497 |
unnecessary | -0.488 |
Token | in |
Token ID | 259 |
Feature activation | -1.22e+00 |
Pos logit contributions | |
boy | +1.093 |
girl | +1.022 |
sky | +1.011 |
man | +1.003 |
little | +1.001 |
Neg logit contributions | |
beh | -1.442 |
yours | -1.386 |
possible | -1.216 |
ours | -1.185 |
theirs | -1.151 |
Token | tried |
Token ID | 3088 |
Feature activation | +2.23e+00 |
Pos logit contributions | |
beh | +0.082 |
yours | +0.071 |
eking | +0.068 |
rying | +0.067 |
possible | +0.067 |
Neg logit contributions | |
sky | -0.030 |
boy | -0.030 |
girl | -0.030 |
woman | -0.026 |
little | -0.026 |
Token | to |
Token ID | 284 |
Feature activation | +2.43e+00 |
Pos logit contributions | |
boy | +0.083 |
girl | +0.077 |
little | +0.076 |
sky | +0.076 |
man | +0.076 |
Neg logit contributions | |
beh | -0.139 |
yours | -0.134 |
possible | -0.118 |
ours | -0.117 |
theirs | -0.113 |
Token | grab |
Token ID | 5552 |
Feature activation | -8.33e+00 |
Pos logit contributions | |
boy | +0.098 |
little | +0.097 |
was | +0.093 |
man | +0.093 |
girl | +0.089 |
Neg logit contributions | |
beh | -0.148 |
yours | -0.141 |
ours | -0.125 |
possible | -0.122 |
theirs | -0.120 |
Token | her |
Token ID | 607 |
Feature activation | -4.77e+00 |
Pos logit contributions | |
beh | +0.544 |
yours | +0.521 |
possible | +0.466 |
ours | +0.458 |
theirs | +0.443 |
Neg logit contributions | |
boy | -0.294 |
girl | -0.271 |
little | -0.266 |
sky | -0.265 |
man | -0.265 |
Token | are |
Token ID | 389 |
Feature activation | +7.30e-01 |
Pos logit contributions | |
yours | +0.045 |
beh | +0.045 |
possible | +0.041 |
ours | +0.040 |
theirs | +0.040 |
Neg logit contributions | |
boy | -0.016 |
girl | -0.015 |
sky | -0.014 |
little | -0.014 |
man | -0.014 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.58e+01 |
Pos logit contributions | |
boy | +0.027 |
girl | +0.025 |
man | +0.025 |
sky | +0.024 |
little | +0.023 |
Neg logit contributions | |
beh | -0.044 |
yours | -0.044 |
possible | -0.039 |
ours | -0.038 |
theirs | -0.037 |
Token | ins |
Token ID | 1040 |
Feature activation | -6.16e+00 |
Pos logit contributions | |
boy | +1.209 |
girl | +1.129 |
sky | +1.115 |
little | +1.113 |
man | +1.110 |
Neg logit contributions | |
beh | -1.364 |
yours | -1.330 |
possible | -1.162 |
ours | -1.127 |
theirs | -1.099 |
Token | . |
Token ID | 13 |
Feature activation | -7.79e+00 |
Pos logit contributions | |
beh | +0.394 |
yours | +0.371 |
possible | +0.331 |
unnecessary | +0.330 |
ours | +0.324 |
Neg logit contributions | |
boy | -0.231 |
was | -0.217 |
sky | -0.213 |
man | -0.212 |
girl | -0.210 |
Token | The |
Token ID | 383 |
Feature activation | +7.51e+00 |
Pos logit contributions | |
yours | +0.480 |
beh | +0.478 |
ours | +0.424 |
theirs | +0.423 |
possible | +0.416 |
Neg logit contributions | |
boy | -0.263 |
girl | -0.249 |
man | -0.248 |
little | -0.237 |
sky | -0.236 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.53e+01 |
Pos logit contributions | |
boy | +0.158 |
was | +0.147 |
girl | +0.144 |
sky | +0.144 |
man | +0.136 |
Neg logit contributions | |
yours | -0.521 |
beh | -0.517 |
possible | -0.472 |
ours | -0.466 |
theirs | -0.449 |
Token | ins |
Token ID | 1040 |
Feature activation | -6.72e+00 |
Pos logit contributions | |
boy | +1.212 |
girl | +1.150 |
sky | +1.145 |
little | +1.133 |
man | +1.121 |
Neg logit contributions | |
beh | -1.309 |
yours | -1.245 |
possible | -1.068 |
ours | -1.041 |
ieve | -1.008 |
Token | are |
Token ID | 389 |
Feature activation | -2.75e-01 |
Pos logit contributions | |
beh | +0.405 |
yours | +0.393 |
possible | +0.346 |
ours | +0.340 |
theirs | +0.336 |
Neg logit contributions | |
boy | -0.247 |
girl | -0.234 |
little | -0.226 |
sky | -0.224 |
man | -0.218 |
Token | black |
Token ID | 2042 |
Feature activation | +7.63e+00 |
Pos logit contributions | |
beh | +0.018 |
yours | +0.017 |
possible | +0.016 |
ours | +0.015 |
theirs | +0.015 |
Neg logit contributions | |
boy | -0.009 |
girl | -0.009 |
sky | -0.009 |
little | -0.009 |
man | -0.008 |
Token | and |
Token ID | 290 |
Feature activation | +1.40e+01 |
Pos logit contributions | |
boy | +0.117 |
girl | +0.099 |
man | +0.095 |
little | +0.087 |
sky | +0.086 |
Neg logit contributions | |
beh | -0.612 |
yours | -0.611 |
ours | -0.546 |
possible | -0.545 |
theirs | -0.539 |
Token | white |
Token ID | 2330 |
Feature activation | +1.04e+01 |
Pos logit contributions | |
boy | +0.358 |
girl | +0.318 |
sky | +0.313 |
little | +0.311 |
man | +0.309 |
Neg logit contributions | |
beh | -1.054 |
yours | -1.011 |
possible | -0.919 |
ours | -0.899 |
theirs | -0.882 |
Token | , |
Token ID | 11 |
Feature activation | +1.07e+01 |
Pos logit contributions | |
boy | +0.148 |
girl | +0.137 |
man | +0.136 |
sky | +0.135 |
woman | +0.124 |
Neg logit contributions | |
yours | -0.359 |
beh | -0.353 |
possible | -0.324 |
ours | -0.313 |
theirs | -0.304 |
Token | there |
Token ID | 612 |
Feature activation | -2.07e+00 |
Pos logit contributions | |
boy | +0.463 |
man | +0.461 |
sky | +0.454 |
girl | +0.439 |
woman | +0.423 |
Neg logit contributions | |
yours | -0.555 |
beh | -0.529 |
possible | -0.478 |
ours | -0.467 |
theirs | -0.449 |
Token | was |
Token ID | 373 |
Feature activation | +9.18e+00 |
Pos logit contributions | |
yours | +0.129 |
beh | +0.116 |
anymore | +0.115 |
possible | +0.113 |
theirs | +0.113 |
Neg logit contributions | |
woman | -0.052 |
girl | -0.049 |
man | -0.047 |
sky | -0.044 |
boy | -0.043 |
Token | an |
Token ID | 281 |
Feature activation | -4.47e+00 |
Pos logit contributions | |
boy | +0.411 |
man | +0.403 |
girl | +0.387 |
sky | +0.371 |
woman | +0.360 |
Neg logit contributions | |
yours | -0.486 |
beh | -0.463 |
possible | -0.408 |
ours | -0.405 |
theirs | -0.398 |
Token | ordinary |
Token ID | 8850 |
Feature activation | -2.23e+00 |
Pos logit contributions | |
yours | +0.266 |
beh | +0.264 |
ours | +0.231 |
possible | +0.226 |
theirs | +0.221 |
Neg logit contributions | |
boy | -0.161 |
was | -0.155 |
sky | -0.155 |
girl | -0.148 |
man | -0.148 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.53e+01 |
Pos logit contributions | |
yours | +0.158 |
beh | +0.142 |
possible | +0.139 |
ours | +0.136 |
asty | +0.129 |
Neg logit contributions | |
At | -0.033 |
sky | -0.032 |
was | -0.030 |
Quick | -0.028 |
boy | -0.024 |
Token | in |
Token ID | 259 |
Feature activation | -1.31e+00 |
Pos logit contributions | |
boy | +1.001 |
girl | +0.929 |
sky | +0.914 |
man | +0.910 |
little | +0.910 |
Neg logit contributions | |
beh | -1.529 |
yours | -1.481 |
possible | -1.312 |
ours | -1.287 |
theirs | -1.249 |
Token | . |
Token ID | 13 |
Feature activation | -9.25e+00 |
Pos logit contributions | |
beh | +0.082 |
yours | +0.078 |
possible | +0.069 |
ours | +0.066 |
theirs | +0.066 |
Neg logit contributions | |
boy | -0.045 |
sky | -0.043 |
girl | -0.042 |
man | -0.041 |
little | -0.040 |
Token | He |
Token ID | 679 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
beh | +0.571 |
yours | +0.563 |
possible | +0.490 |
ours | +0.485 |
theirs | +0.478 |
Neg logit contributions | |
boy | -0.338 |
girl | -0.317 |
man | -0.310 |
sky | -0.305 |
little | -0.302 |
Token | was |
Token ID | 373 |
Feature activation | +8.85e+00 |
Pos logit contributions | |
beh | +0.085 |
yours | +0.084 |
supposed | +0.071 |
possible | +0.071 |
theirs | +0.070 |
Neg logit contributions | |
girl | -0.040 |
boy | -0.039 |
sky | -0.039 |
woman | -0.036 |
man | -0.036 |
Token | walking |
Token ID | 6155 |
Feature activation | -2.98e+00 |
Pos logit contributions | |
boy | +0.347 |
girl | +0.326 |
man | +0.320 |
sky | +0.320 |
was | +0.305 |
Neg logit contributions | |
beh | -0.526 |
yours | -0.514 |
possible | -0.450 |
ours | -0.444 |
theirs | -0.430 |
Token | 't |
Token ID | 470 |
Feature activation | -5.26e+00 |
Pos logit contributions | |
beh | +0.477 |
yours | +0.451 |
possible | +0.388 |
ours | +0.378 |
unnecessary | +0.374 |
Neg logit contributions | |
boy | -0.450 |
sky | -0.432 |
man | -0.428 |
little | -0.428 |
was | -0.426 |
Token | know |
Token ID | 760 |
Feature activation | -8.61e+00 |
Pos logit contributions | |
beh | +0.276 |
yours | +0.261 |
possible | +0.235 |
ours | +0.230 |
eking | +0.227 |
Neg logit contributions | |
boy | -0.250 |
little | -0.244 |
girl | -0.238 |
man | -0.228 |
was | -0.224 |
Token | . |
Token ID | 13 |
Feature activation | -8.11e+00 |
Pos logit contributions | |
beh | +0.624 |
yours | +0.547 |
possible | +0.514 |
unnecessary | +0.501 |
ours | +0.497 |
Neg logit contributions | |
was | -0.271 |
boy | -0.271 |
little | -0.267 |
girl | -0.245 |
and | -0.243 |
Token | Maybe |
Token ID | 6674 |
Feature activation | -2.33e+00 |
Pos logit contributions | |
beh | +0.474 |
yours | +0.455 |
possible | +0.401 |
ours | +0.391 |
theirs | +0.381 |
Neg logit contributions | |
boy | -0.337 |
girl | -0.315 |
sky | -0.311 |
man | -0.308 |
little | -0.308 |
Token | a |
Token ID | 257 |
Feature activation | -1.63e+00 |
Pos logit contributions | |
beh | +0.146 |
yours | +0.132 |
possible | +0.117 |
ours | +0.115 |
unnecessary | +0.115 |
Neg logit contributions | |
boy | -0.096 |
girl | -0.089 |
little | -0.089 |
man | -0.087 |
was | -0.087 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.53e+01 |
Pos logit contributions | |
yours | +0.109 |
beh | +0.106 |
possible | +0.098 |
ours | +0.095 |
unnecessary | +0.092 |
Neg logit contributions | |
boy | -0.044 |
sky | -0.043 |
was | -0.040 |
girl | -0.039 |
man | -0.039 |
Token | in |
Token ID | 259 |
Feature activation | -1.64e+00 |
Pos logit contributions | |
boy | +0.869 |
girl | +0.796 |
man | +0.724 |
was | +0.703 |
bird | +0.685 |
Neg logit contributions | |
beh | -1.685 |
yours | -1.637 |
possible | -1.522 |
ours | -1.505 |
theirs | -1.465 |
Token | ?" |
Token ID | 1701 |
Feature activation | -8.53e+00 |
Pos logit contributions | |
beh | +0.104 |
yours | +0.098 |
possible | +0.089 |
ours | +0.087 |
unnecessary | +0.086 |
Neg logit contributions | |
boy | -0.065 |
was | -0.059 |
man | -0.059 |
girl | -0.058 |
little | -0.056 |
Token | Tom |
Token ID | 4186 |
Feature activation | -2.81e+00 |
Pos logit contributions | |
beh | +0.502 |
yours | +0.490 |
possible | +0.431 |
ours | +0.422 |
theirs | +0.414 |
Neg logit contributions | |
boy | -0.333 |
girl | -0.312 |
sky | -0.311 |
little | -0.303 |
man | -0.301 |
Token | guesses |
Token ID | 44774 |
Feature activation | -8.46e+00 |
Pos logit contributions | |
beh | +0.165 |
yours | +0.158 |
possible | +0.136 |
ours | +0.134 |
ieve | +0.132 |
Neg logit contributions | |
boy | -0.106 |
sky | -0.102 |
girl | -0.099 |
little | -0.098 |
man | -0.095 |
Token | . |
Token ID | 13 |
Feature activation | -8.91e+00 |
Pos logit contributions | |
beh | +0.573 |
yours | +0.549 |
ours | +0.504 |
unnecessary | +0.480 |
possible | +0.479 |
Neg logit contributions | |
boy | -0.321 |
was | -0.315 |
man | -0.305 |
and | -0.301 |
little | -0.295 |
Token | time |
Token ID | 640 |
Feature activation | +5.43e+00 |
Pos logit contributions | |
yours | +0.050 |
beh | +0.048 |
ours | +0.042 |
unnecessary | +0.041 |
supposed | +0.041 |
Neg logit contributions | |
boy | -0.041 |
man | -0.040 |
sky | -0.040 |
girl | -0.040 |
woman | -0.038 |
Token | , |
Token ID | 11 |
Feature activation | +1.12e+01 |
Pos logit contributions | |
boy | +0.157 |
girl | +0.146 |
man | +0.145 |
sky | +0.141 |
woman | +0.134 |
Neg logit contributions | |
yours | -0.369 |
beh | -0.361 |
possible | -0.332 |
ours | -0.323 |
theirs | -0.312 |
Token | there |
Token ID | 612 |
Feature activation | -1.87e+00 |
Pos logit contributions | |
boy | +0.489 |
man | +0.482 |
sky | +0.475 |
girl | +0.463 |
woman | +0.443 |
Neg logit contributions | |
yours | -0.573 |
beh | -0.550 |
possible | -0.493 |
ours | -0.484 |
theirs | -0.464 |
Token | was |
Token ID | 373 |
Feature activation | +9.30e+00 |
Pos logit contributions | |
yours | +0.119 |
anymore | +0.107 |
beh | +0.106 |
possible | +0.105 |
anything | +0.105 |
Neg logit contributions | |
woman | -0.046 |
girl | -0.043 |
man | -0.041 |
sky | -0.038 |
boy | -0.038 |
Token | a |
Token ID | 257 |
Feature activation | -1.41e+00 |
Pos logit contributions | |
boy | +0.400 |
man | +0.387 |
girl | +0.375 |
sky | +0.363 |
was | +0.352 |
Neg logit contributions | |
yours | -0.510 |
beh | -0.497 |
ours | -0.433 |
possible | -0.430 |
theirs | -0.422 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.53e+01 |
Pos logit contributions | |
yours | +0.089 |
ours | +0.081 |
beh | +0.080 |
possible | +0.077 |
asty | +0.075 |
Neg logit contributions | |
sky | -0.029 |
was | -0.024 |
led | -0.021 |
At | -0.019 |
arrived | -0.016 |
Token | in |
Token ID | 259 |
Feature activation | -1.48e+00 |
Pos logit contributions | |
boy | +1.062 |
girl | +0.997 |
sky | +0.993 |
man | +0.977 |
little | +0.973 |
Neg logit contributions | |
beh | -1.456 |
yours | -1.389 |
possible | -1.216 |
ours | -1.183 |
theirs | -1.152 |
Token | named |
Token ID | 3706 |
Feature activation | +5.55e+00 |
Pos logit contributions | |
beh | +0.087 |
yours | +0.083 |
possible | +0.072 |
ours | +0.069 |
theirs | +0.069 |
Neg logit contributions | |
boy | -0.052 |
sky | -0.051 |
girl | -0.050 |
man | -0.048 |
little | -0.047 |
Token | Penny |
Token ID | 25965 |
Feature activation | -5.07e+00 |
Pos logit contributions | |
boy | +0.230 |
little | +0.215 |
sky | +0.212 |
girl | +0.212 |
man | +0.210 |
Neg logit contributions | |
beh | -0.334 |
yours | -0.310 |
possible | -0.279 |
ours | -0.267 |
unnecessary | -0.264 |
Token | . |
Token ID | 13 |
Feature activation | -9.03e+00 |
Pos logit contributions | |
beh | +0.249 |
yours | +0.242 |
possible | +0.206 |
ours | +0.200 |
theirs | +0.195 |
Neg logit contributions | |
boy | -0.252 |
girl | -0.239 |
sky | -0.237 |
man | -0.234 |
little | -0.232 |
Token | Penny |
Token ID | 25965 |
Feature activation | -5.35e+00 |
Pos logit contributions | |
beh | +0.457 |
yours | +0.455 |
possible | +0.388 |
ours | +0.374 |
theirs | +0.368 |
Neg logit contributions | |
boy | -0.421 |
man | -0.395 |
girl | -0.391 |
little | -0.382 |
sky | -0.381 |
Token | Max |
Token ID | 5436 |
Feature activation | -2.90e+00 |
Pos logit contributions | |
boy | +0.608 |
girl | +0.564 |
sky | +0.555 |
man | +0.553 |
little | +0.552 |
Neg logit contributions | |
beh | -0.944 |
yours | -0.907 |
possible | -0.804 |
ours | -0.785 |
theirs | -0.763 |
Token | being |
Token ID | 852 |
Feature activation | -5.53e+00 |
Pos logit contributions | |
beh | +0.168 |
yours | +0.164 |
possible | +0.146 |
ours | +0.141 |
unnecessary | +0.135 |
Neg logit contributions | |
boy | -0.124 |
was | -0.116 |
little | -0.115 |
girl | -0.115 |
sky | -0.113 |
Token | attacked |
Token ID | 7384 |
Feature activation | -6.27e+00 |
Pos logit contributions | |
beh | +0.338 |
yours | +0.328 |
possible | +0.293 |
ours | +0.284 |
theirs | +0.277 |
Neg logit contributions | |
boy | -0.217 |
girl | -0.201 |
little | -0.199 |
man | -0.194 |
sky | -0.194 |
Token | by |
Token ID | 416 |
Feature activation | +1.05e+00 |
Pos logit contributions | |
beh | +0.338 |
yours | +0.316 |
possible | +0.287 |
ours | +0.274 |
ieve | +0.266 |
Neg logit contributions | |
boy | -0.303 |
girl | -0.286 |
little | -0.283 |
was | -0.279 |
man | -0.274 |
Token | the |
Token ID | 262 |
Feature activation | +7.65e+00 |
Pos logit contributions | |
boy | +0.041 |
little | +0.040 |
was | +0.038 |
girl | +0.038 |
man | +0.038 |
Neg logit contributions | |
beh | -0.064 |
yours | -0.061 |
possible | -0.055 |
ours | -0.053 |
ieve | -0.051 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.52e+01 |
Pos logit contributions | |
boy | +0.177 |
girl | +0.158 |
sky | +0.155 |
was | +0.155 |
little | +0.152 |
Neg logit contributions | |
beh | -0.554 |
yours | -0.550 |
possible | -0.498 |
ours | -0.492 |
theirs | -0.475 |
Token | ins |
Token ID | 1040 |
Feature activation | -6.87e+00 |
Pos logit contributions | |
boy | +1.065 |
girl | +0.986 |
man | +0.971 |
little | +0.963 |
sky | +0.962 |
Neg logit contributions | |
beh | -1.452 |
yours | -1.418 |
possible | -1.257 |
ours | -1.226 |
theirs | -1.187 |
Token | . |
Token ID | 13 |
Feature activation | -8.60e+00 |
Pos logit contributions | |
beh | +0.419 |
yours | +0.404 |
possible | +0.358 |
ours | +0.351 |
unnecessary | +0.350 |
Neg logit contributions | |
boy | -0.276 |
was | -0.263 |
man | -0.259 |
little | -0.252 |
girl | -0.250 |
Token | They |
Token ID | 1119 |
Feature activation | +1.58e+00 |
Pos logit contributions | |
yours | +0.531 |
beh | +0.529 |
ours | +0.469 |
theirs | +0.468 |
possible | +0.458 |
Neg logit contributions | |
boy | -0.272 |
girl | -0.261 |
man | -0.248 |
little | -0.244 |
sky | -0.244 |
Token | were |
Token ID | 547 |
Feature activation | +6.67e+00 |
Pos logit contributions | |
boy | +0.049 |
sky | +0.048 |
girl | +0.047 |
man | +0.043 |
little | +0.041 |
Neg logit contributions | |
beh | -0.101 |
yours | -0.094 |
ours | -0.086 |
theirs | -0.085 |
eking | -0.083 |
Token | very |
Token ID | 845 |
Feature activation | +5.29e+00 |
Pos logit contributions | |
boy | +0.241 |
little | +0.235 |
was | +0.221 |
sky | +0.218 |
man | +0.217 |
Neg logit contributions | |
beh | -0.420 |
yours | -0.409 |
possible | -0.370 |
ieve | -0.348 |
ours | -0.346 |
Token | my |
Token ID | 616 |
Feature activation | -7.83e+00 |
Pos logit contributions | |
boy | +0.374 |
girl | +0.349 |
little | +0.344 |
sky | +0.343 |
man | +0.342 |
Neg logit contributions | |
beh | -0.506 |
yours | -0.481 |
possible | -0.425 |
ours | -0.412 |
theirs | -0.402 |
Token | row |
Token ID | 5752 |
Feature activation | +7.50e+00 |
Pos logit contributions | |
yours | +0.509 |
beh | +0.490 |
possible | +0.450 |
ours | +0.448 |
theirs | +0.431 |
Neg logit contributions | |
boy | -0.225 |
girl | -0.211 |
sky | -0.209 |
man | -0.205 |
was | -0.201 |
Token | is |
Token ID | 318 |
Feature activation | +2.77e+00 |
Pos logit contributions | |
boy | +0.250 |
girl | +0.228 |
sky | +0.224 |
man | +0.223 |
little | +0.221 |
Neg logit contributions | |
beh | -0.504 |
yours | -0.482 |
possible | -0.434 |
ours | -0.424 |
theirs | -0.415 |
Token | longer |
Token ID | 2392 |
Feature activation | -1.38e+01 |
Pos logit contributions | |
boy | +0.096 |
little | +0.093 |
sky | +0.090 |
was | +0.087 |
man | +0.086 |
Neg logit contributions | |
beh | -0.191 |
yours | -0.164 |
unnecessary | -0.152 |
rying | -0.150 |
ieve | -0.149 |
Token | than |
Token ID | 621 |
Feature activation | -1.07e+01 |
Pos logit contributions | |
beh | +1.015 |
yours | +0.925 |
possible | +0.824 |
unnecessary | +0.820 |
ieve | +0.819 |
Neg logit contributions | |
boy | -0.416 |
little | -0.388 |
girl | -0.371 |
sky | -0.370 |
man | -0.368 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.13e+01 |
Pos logit contributions | |
beh | +0.687 |
yours | +0.524 |
unnecessary | +0.517 |
rying | +0.512 |
eking | +0.511 |
Neg logit contributions | |
little | -0.443 |
boy | -0.431 |
sky | -0.412 |
man | -0.396 |
girl | -0.393 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.32e+01 |
Pos logit contributions | |
beh | +1.869 |
yours | +1.595 |
unnecessary | +1.468 |
eking | +1.420 |
possible | +1.419 |
Neg logit contributions | |
boy | -1.455 |
was | -1.365 |
man | -1.347 |
girl | -1.332 |
little | -1.319 |
Token | Anna |
Token ID | 11735 |
Feature activation | +1.27e+00 |
Pos logit contributions | |
beh | +0.758 |
yours | +0.745 |
possible | +0.651 |
ours | +0.643 |
theirs | +0.625 |
Neg logit contributions | |
boy | -0.525 |
girl | -0.489 |
sky | -0.487 |
man | -0.478 |
little | -0.473 |
Token | says |
Token ID | 1139 |
Feature activation | -6.54e-01 |
Pos logit contributions | |
boy | +0.047 |
girl | +0.045 |
sky | +0.044 |
little | +0.044 |
man | +0.043 |
Neg logit contributions | |
beh | -0.073 |
yours | -0.070 |
possible | -0.061 |
ours | -0.061 |
ieve | -0.060 |
Token | . |
Token ID | 13 |
Feature activation | -1.09e+01 |
Pos logit contributions | |
beh | +0.043 |
yours | +0.042 |
ours | +0.038 |
possible | +0.037 |
unnecessary | +0.036 |
Neg logit contributions | |
boy | -0.024 |
little | -0.023 |
and | -0.023 |
was | -0.022 |
girl | -0.022 |
Token | She |
Token ID | 1375 |
Feature activation | -1.10e+00 |
Pos logit contributions | |
yours | +0.663 |
beh | +0.643 |
possible | +0.595 |
theirs | +0.581 |
ours | +0.568 |
Neg logit contributions | |
boy | -0.315 |
girl | -0.279 |
man | -0.269 |
child | -0.261 |
was | -0.254 |
Token | my |
Token ID | 616 |
Feature activation | -7.22e+00 |
Pos logit contributions | |
little | +0.154 |
boy | +0.152 |
was | +0.145 |
girl | +0.137 |
man | +0.135 |
Neg logit contributions | |
beh | -0.218 |
ieve | -0.177 |
rying | -0.177 |
yours | -0.171 |
owing | -0.170 |
Token | juice |
Token ID | 13135 |
Feature activation | -7.38e-03 |
Pos logit contributions | |
beh | +0.455 |
yours | +0.433 |
possible | +0.387 |
ours | +0.377 |
theirs | +0.367 |
Neg logit contributions | |
boy | -0.275 |
girl | -0.255 |
little | -0.250 |
sky | -0.250 |
man | -0.247 |
Token | . |
Token ID | 13 |
Feature activation | -1.04e+01 |
Pos logit contributions | |
beh | +0.000 |
yours | +0.000 |
unnecessary | +0.000 |
possible | +0.000 |
ieve | +0.000 |
Neg logit contributions | |
boy | -0.000 |
was | -0.000 |
girl | -0.000 |
little | -0.000 |
man | -0.000 |
Token | You |
Token ID | 921 |
Feature activation | -1.33e+01 |
Pos logit contributions | |
beh | +0.587 |
yours | +0.549 |
possible | +0.502 |
ours | +0.471 |
eking | +0.469 |
Neg logit contributions | |
boy | -0.457 |
girl | -0.427 |
sky | -0.421 |
little | -0.419 |
man | -0.412 |
Token | had |
Token ID | 550 |
Feature activation | +4.92e+00 |
Pos logit contributions | |
beh | +0.878 |
yours | +0.829 |
possible | +0.760 |
ours | +0.719 |
unnecessary | +0.715 |
Neg logit contributions | |
boy | -0.470 |
little | -0.433 |
girl | -0.428 |
was | -0.427 |
man | -0.418 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.12e+01 |
Pos logit contributions | |
boy | +0.194 |
little | +0.191 |
was | +0.187 |
sky | +0.182 |
girl | +0.179 |
Neg logit contributions | |
beh | -0.304 |
yours | -0.268 |
possible | -0.251 |
ieve | -0.250 |
rying | -0.244 |
Token | . |
Token ID | 13 |
Feature activation | -1.04e+01 |
Pos logit contributions | |
beh | +2.024 |
yours | +1.682 |
unnecessary | +1.633 |
possible | +1.606 |
ieve | +1.586 |
Neg logit contributions | |
was | -1.276 |
boy | -1.274 |
and | -1.217 |
little | -1.214 |
man | -1.191 |
Token | Go |
Token ID | 1514 |
Feature activation | -8.18e+00 |
Pos logit contributions | |
beh | +0.545 |
yours | +0.505 |
possible | +0.465 |
eking | +0.432 |
ours | +0.431 |
Neg logit contributions | |
boy | -0.496 |
girl | -0.467 |
sky | -0.462 |
little | -0.459 |
man | -0.451 |
Token | get |
Token ID | 651 |
Feature activation | -9.33e+00 |
Pos logit contributions | |
beh | +0.386 |
yours | +0.359 |
possible | +0.319 |
ours | +0.301 |
theirs | +0.293 |
Neg logit contributions | |
boy | -0.444 |
girl | -0.420 |
was | -0.415 |
man | -0.413 |
sky | -0.411 |
Token | some |
Token ID | 617 |
Feature activation | -9.73e+00 |
Pos logit contributions | |
beh | +0.603 |
possible | +0.523 |
yours | +0.520 |
ieve | +0.509 |
rying | +0.505 |
Neg logit contributions | |
little | -0.348 |
boy | -0.341 |
was | -0.333 |
girl | -0.317 |
saw | -0.298 |
Token | water |
Token ID | 1660 |
Feature activation | +3.21e+00 |
Pos logit contributions | |
beh | +0.654 |
yours | +0.591 |
possible | +0.559 |
unnecessary | +0.533 |
eking | +0.531 |
Neg logit contributions | |
boy | -0.350 |
little | -0.339 |
girl | -0.331 |
sky | -0.317 |
was | -0.316 |
Token | is |
Token ID | 318 |
Feature activation | +3.04e+00 |
Pos logit contributions | |
boy | +0.042 |
girl | +0.041 |
sky | +0.040 |
little | +0.038 |
woman | +0.037 |
Neg logit contributions | |
yours | -0.118 |
beh | -0.115 |
ours | -0.104 |
possible | -0.103 |
supposed | -0.096 |
Token | our |
Token ID | 674 |
Feature activation | -6.91e+00 |
Pos logit contributions | |
boy | +0.148 |
little | +0.147 |
was | +0.139 |
man | +0.136 |
dog | +0.133 |
Neg logit contributions | |
beh | -0.162 |
rying | -0.132 |
yours | -0.129 |
ieve | -0.126 |
possible | -0.126 |
Token | ball |
Token ID | 2613 |
Feature activation | +7.79e+00 |
Pos logit contributions | |
beh | +0.563 |
yours | +0.563 |
possible | +0.511 |
ours | +0.509 |
theirs | +0.495 |
Neg logit contributions | |
boy | -0.115 |
girl | -0.099 |
sky | -0.094 |
man | -0.092 |
little | -0.089 |
Token | , |
Token ID | 11 |
Feature activation | +8.72e+00 |
Pos logit contributions | |
boy | +0.398 |
was | +0.376 |
little | +0.375 |
girl | +0.368 |
man | +0.366 |
Neg logit contributions | |
beh | -0.408 |
yours | -0.356 |
unnecessary | -0.324 |
possible | -0.322 |
theirs | -0.312 |
Token | not |
Token ID | 407 |
Feature activation | -1.14e+01 |
Pos logit contributions | |
boy | +0.371 |
girl | +0.352 |
sky | +0.350 |
little | +0.337 |
man | +0.337 |
Neg logit contributions | |
yours | -0.484 |
beh | -0.479 |
ours | -0.416 |
possible | -0.411 |
theirs | -0.399 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.12e+01 |
Pos logit contributions | |
beh | +0.628 |
eking | +0.504 |
rying | +0.492 |
ieve | +0.487 |
ility | +0.487 |
Neg logit contributions | |
boy | -0.613 |
little | -0.578 |
dog | -0.550 |
man | -0.542 |
was | -0.539 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.31e+01 |
Pos logit contributions | |
beh | +2.139 |
yours | +1.923 |
possible | +1.813 |
unnecessary | +1.784 |
eking | +1.754 |
Neg logit contributions | |
boy | -1.117 |
was | -1.064 |
man | -1.000 |
girl | -0.979 |
little | -0.975 |
Token |
|
Token ID | 198 |
Feature activation | +5.31e+00 |
Pos logit contributions | |
yours | +1.014 |
beh | +1.012 |
possible | +0.922 |
ours | +0.912 |
theirs | +0.900 |
Neg logit contributions | |
boy | -0.238 |
girl | -0.210 |
sky | -0.186 |
man | -0.184 |
little | -0.175 |
Token |
|
Token ID | 198 |
Feature activation | +5.34e+00 |
Pos logit contributions | |
boy | +0.107 |
girl | +0.099 |
sky | +0.091 |
man | +0.084 |
little | +0.081 |
Neg logit contributions | |
beh | -0.402 |
yours | -0.401 |
possible | -0.363 |
ours | -0.358 |
ourselves | -0.358 |
Token | But |
Token ID | 1537 |
Feature activation | +2.61e+00 |
Pos logit contributions | |
boy | +0.172 |
sky | +0.170 |
girl | +0.162 |
was | +0.149 |
little | +0.147 |
Neg logit contributions | |
beh | -0.334 |
yours | -0.316 |
ourselves | -0.306 |
myself | -0.289 |
ours | -0.287 |
Token | the |
Token ID | 262 |
Feature activation | +4.91e+00 |
Pos logit contributions | |
boy | +0.094 |
girl | +0.088 |
sky | +0.087 |
man | +0.085 |
little | +0.083 |
Neg logit contributions | |
yours | -0.159 |
beh | -0.157 |
possible | -0.141 |
ours | -0.138 |
theirs | -0.136 |
Token | No |
Token ID | 2949 |
Feature activation | +2.11e+00 |
Pos logit contributions | |
boy | +0.191 |
girl | +0.176 |
sky | +0.174 |
little | +0.172 |
man | +0.172 |
Neg logit contributions | |
beh | -0.330 |
yours | -0.318 |
possible | -0.283 |
ours | -0.277 |
theirs | -0.269 |
Token | , |
Token ID | 11 |
Feature activation | +8.63e+00 |
Pos logit contributions | |
boy | +0.041 |
girl | +0.035 |
little | +0.034 |
man | +0.034 |
sky | +0.033 |
Neg logit contributions | |
beh | -0.174 |
yours | -0.167 |
possible | -0.153 |
ours | -0.151 |
theirs | -0.148 |
Token | they |
Token ID | 484 |
Feature activation | -2.14e+00 |
Pos logit contributions | |
boy | +0.390 |
little | +0.364 |
man | +0.362 |
girl | +0.361 |
sky | +0.348 |
Neg logit contributions | |
beh | -0.494 |
yours | -0.448 |
possible | -0.406 |
rying | -0.381 |
ours | -0.379 |
Token | are |
Token ID | 389 |
Feature activation | -2.72e+00 |
Pos logit contributions | |
beh | +0.137 |
yours | +0.132 |
possible | +0.123 |
theirs | +0.116 |
ours | +0.112 |
Neg logit contributions | |
boy | -0.077 |
was | -0.073 |
little | -0.071 |
girl | -0.069 |
man | -0.069 |
Token | not |
Token ID | 407 |
Feature activation | -1.14e+01 |
Pos logit contributions | |
beh | +0.156 |
rying | +0.131 |
asty | +0.127 |
ility | +0.126 |
limits | +0.124 |
Neg logit contributions | |
little | -0.125 |
boy | -0.117 |
man | -0.110 |
dog | -0.110 |
was | -0.109 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.12e+01 |
Pos logit contributions | |
beh | +0.592 |
limits | +0.482 |
ieve | +0.478 |
eking | +0.477 |
ility | +0.473 |
Neg logit contributions | |
boy | -0.579 |
little | -0.567 |
man | -0.528 |
girl | -0.523 |
dog | -0.515 |
Token | . |
Token ID | 13 |
Feature activation | -1.10e+01 |
Pos logit contributions | |
beh | +2.167 |
yours | +1.906 |
unnecessary | +1.769 |
eking | +1.767 |
possible | +1.750 |
Neg logit contributions | |
boy | -1.161 |
was | -1.108 |
man | -1.065 |
little | -1.055 |
and | -1.049 |
Token | They |
Token ID | 1119 |
Feature activation | -1.01e+00 |
Pos logit contributions | |
beh | +0.635 |
yours | +0.598 |
possible | +0.563 |
eking | +0.538 |
ieve | +0.520 |
Neg logit contributions | |
boy | -0.470 |
girl | -0.431 |
little | -0.421 |
sky | -0.415 |
man | -0.413 |
Token | are |
Token ID | 389 |
Feature activation | -2.69e+00 |
Pos logit contributions | |
beh | +0.065 |
yours | +0.060 |
possible | +0.057 |
unnecessary | +0.053 |
rying | +0.053 |
Neg logit contributions | |
boy | -0.037 |
was | -0.035 |
little | -0.034 |
man | -0.033 |
girl | -0.033 |
Token | ours |
Token ID | 16903 |
Feature activation | -2.52e+01 |
Pos logit contributions | |
beh | +0.143 |
rying | +0.113 |
ieve | +0.111 |
owing | +0.110 |
ility | +0.108 |
Neg logit contributions | |
little | -0.132 |
boy | -0.129 |
man | -0.121 |
dog | -0.118 |
was | -0.117 |
Token | . |
Token ID | 13 |
Feature activation | -1.09e+01 |
Pos logit contributions | |
beh | +1.669 |
yours | +1.489 |
possible | +1.371 |
unnecessary | +1.354 |
eking | +1.339 |
Neg logit contributions | |
boy | -0.996 |
little | -0.917 |
girl | -0.909 |
was | -0.907 |
man | -0.898 |
Token | have |
Token ID | 423 |
Feature activation | -6.03e+00 |
Pos logit contributions | |
beh | +1.213 |
yours | +1.160 |
possible | +1.069 |
ours | +1.011 |
theirs | +1.004 |
Neg logit contributions | |
boy | -0.589 |
was | -0.548 |
little | -0.531 |
man | -0.521 |
girl | -0.521 |
Token | a |
Token ID | 257 |
Feature activation | -4.00e+00 |
Pos logit contributions | |
beh | +0.397 |
yours | +0.339 |
rying | +0.328 |
ieve | +0.324 |
possible | +0.322 |
Neg logit contributions | |
boy | -0.223 |
little | -0.217 |
was | -0.213 |
man | -0.203 |
sky | -0.199 |
Token | different |
Token ID | 1180 |
Feature activation | -4.75e-01 |
Pos logit contributions | |
yours | +0.267 |
beh | +0.254 |
possible | +0.236 |
ours | +0.230 |
unnecessary | +0.220 |
Neg logit contributions | |
boy | -0.119 |
sky | -0.114 |
girl | -0.110 |
was | -0.107 |
man | -0.104 |
Token | friend |
Token ID | 1545 |
Feature activation | +3.71e+00 |
Pos logit contributions | |
beh | +0.032 |
yours | +0.030 |
possible | +0.027 |
ours | +0.027 |
theirs | +0.026 |
Neg logit contributions | |
boy | -0.016 |
girl | -0.015 |
sky | -0.014 |
man | -0.014 |
little | -0.014 |
Token | under |
Token ID | 739 |
Feature activation | -3.14e+00 |
Pos logit contributions | |
boy | +0.167 |
was | +0.155 |
girl | +0.154 |
little | +0.152 |
man | +0.151 |
Neg logit contributions | |
beh | -0.215 |
yours | -0.184 |
possible | -0.177 |
eking | -0.174 |
unnecessary | -0.172 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.12e+01 |
Pos logit contributions | |
beh | +0.218 |
yours | +0.188 |
possible | +0.179 |
rying | +0.175 |
eking | +0.172 |
Neg logit contributions | |
boy | -0.110 |
little | -0.108 |
sky | -0.104 |
was | -0.101 |
man | -0.100 |
Token | . |
Token ID | 13 |
Feature activation | -1.05e+01 |
Pos logit contributions | |
beh | +1.710 |
yours | +1.458 |
possible | +1.331 |
unnecessary | +1.316 |
ours | +1.292 |
Neg logit contributions | |
boy | -1.553 |
man | -1.478 |
was | -1.467 |
girl | -1.451 |
sky | -1.449 |
Token | Go |
Token ID | 1514 |
Feature activation | -8.27e+00 |
Pos logit contributions | |
beh | +0.617 |
yours | +0.582 |
possible | +0.530 |
ours | +0.507 |
eking | +0.500 |
Neg logit contributions | |
boy | -0.436 |
girl | -0.403 |
sky | -0.394 |
little | -0.393 |
man | -0.393 |
Token | and |
Token ID | 290 |
Feature activation | +1.21e+01 |
Pos logit contributions | |
beh | +0.380 |
yours | +0.352 |
possible | +0.313 |
ours | +0.296 |
theirs | +0.288 |
Neg logit contributions | |
boy | -0.461 |
girl | -0.436 |
man | -0.431 |
was | -0.430 |
little | -0.425 |
Token | see |
Token ID | 766 |
Feature activation | -6.20e+00 |
Pos logit contributions | |
boy | +0.548 |
man | +0.509 |
girl | +0.506 |
little | +0.498 |
was | +0.497 |
Neg logit contributions | |
beh | -0.673 |
yours | -0.612 |
possible | -0.576 |
eking | -0.553 |
ieve | -0.551 |
Token | ." |
Token ID | 526 |
Feature activation | -1.45e+01 |
Pos logit contributions | |
beh | +0.376 |
possible | +0.299 |
yours | +0.298 |
rying | +0.281 |
anks | +0.278 |
Neg logit contributions | |
little | -0.279 |
boy | -0.277 |
was | -0.267 |
man | -0.258 |
girl | -0.253 |
Token | dog |
Token ID | 3290 |
Feature activation | +1.40e+01 |
Pos logit contributions | |
boy | +0.360 |
girl | +0.334 |
little | +0.333 |
man | +0.328 |
sky | +0.311 |
Neg logit contributions | |
beh | -0.572 |
yours | -0.524 |
possible | -0.481 |
ours | -0.452 |
theirs | -0.450 |
Token | ! |
Token ID | 0 |
Feature activation | -1.26e+01 |
Pos logit contributions | |
boy | +0.660 |
man | +0.617 |
girl | +0.606 |
little | +0.597 |
was | +0.596 |
Neg logit contributions | |
beh | -0.803 |
yours | -0.737 |
possible | -0.667 |
theirs | -0.648 |
ours | -0.636 |
Token | It |
Token ID | 632 |
Feature activation | -7.78e-01 |
Pos logit contributions | |
beh | +0.593 |
yours | +0.570 |
possible | +0.476 |
ours | +0.468 |
theirs | +0.448 |
Neg logit contributions | |
boy | -0.659 |
girl | -0.627 |
sky | -0.624 |
man | -0.618 |
little | -0.615 |
Token | is |
Token ID | 318 |
Feature activation | +3.31e+00 |
Pos logit contributions | |
beh | +0.059 |
yours | +0.058 |
possible | +0.052 |
ours | +0.052 |
theirs | +0.050 |
Neg logit contributions | |
boy | -0.018 |
girl | -0.016 |
sky | -0.016 |
man | -0.015 |
little | -0.015 |
Token | not |
Token ID | 407 |
Feature activation | -1.13e+01 |
Pos logit contributions | |
little | +0.180 |
boy | +0.180 |
man | +0.170 |
was | +0.170 |
dog | +0.167 |
Neg logit contributions | |
beh | -0.157 |
yours | -0.118 |
ieve | -0.118 |
rying | -0.114 |
possible | -0.113 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.12e+01 |
Pos logit contributions | |
beh | +0.545 |
rying | +0.401 |
ieve | +0.398 |
ility | +0.390 |
irs | +0.388 |
Neg logit contributions | |
boy | -0.661 |
little | -0.655 |
man | -0.615 |
girl | -0.607 |
dog | -0.596 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.28e+01 |
Pos logit contributions | |
beh | +2.033 |
yours | +1.698 |
ility | +1.625 |
unnecessary | +1.605 |
irs | +1.592 |
Neg logit contributions | |
boy | -1.312 |
was | -1.298 |
man | -1.208 |
and | -1.202 |
little | -1.194 |
Token | Ben |
Token ID | 3932 |
Feature activation | -1.28e+00 |
Pos logit contributions | |
beh | +0.776 |
yours | +0.752 |
possible | +0.666 |
ours | +0.652 |
theirs | +0.634 |
Neg logit contributions | |
boy | -0.496 |
girl | -0.462 |
sky | -0.454 |
man | -0.451 |
little | -0.448 |
Token | says |
Token ID | 1139 |
Feature activation | -4.13e-01 |
Pos logit contributions | |
beh | +0.079 |
yours | +0.075 |
ours | +0.066 |
possible | +0.066 |
ieve | +0.064 |
Neg logit contributions | |
boy | -0.046 |
sky | -0.044 |
girl | -0.044 |
little | -0.043 |
man | -0.042 |
Token | . |
Token ID | 13 |
Feature activation | -1.07e+01 |
Pos logit contributions | |
beh | +0.029 |
yours | +0.029 |
ours | +0.026 |
possible | +0.025 |
unnecessary | +0.025 |
Neg logit contributions | |
and | -0.014 |
boy | -0.014 |
little | -0.013 |
was | -0.013 |
man | -0.013 |
Token |
|
Token ID | 198 |
Feature activation | +5.56e+00 |
Pos logit contributions | |
yours | +0.583 |
beh | +0.563 |
possible | +0.530 |
theirs | +0.526 |
properly | +0.507 |
Neg logit contributions | |
boy | -0.316 |
girl | -0.302 |
sky | -0.293 |
child | -0.285 |
field | -0.280 |
Token | take |
Token ID | 1011 |
Feature activation | -7.25e+00 |
Pos logit contributions | |
beh | +0.592 |
yours | +0.583 |
possible | +0.555 |
eking | +0.525 |
ieve | +0.525 |
Neg logit contributions | |
boy | -0.490 |
was | -0.472 |
little | -0.455 |
man | -0.440 |
dog | -0.432 |
Token | things |
Token ID | 1243 |
Feature activation | -6.22e-01 |
Pos logit contributions | |
beh | +0.444 |
yours | +0.398 |
possible | +0.379 |
rying | +0.365 |
eking | +0.360 |
Neg logit contributions | |
boy | -0.295 |
little | -0.286 |
girl | -0.268 |
man | -0.266 |
dog | -0.263 |
Token | that |
Token ID | 326 |
Feature activation | -8.92e+00 |
Pos logit contributions | |
beh | +0.043 |
yours | +0.037 |
unnecessary | +0.035 |
ieve | +0.035 |
eking | +0.034 |
Neg logit contributions | |
man | -0.021 |
boy | -0.021 |
and | -0.020 |
little | -0.020 |
dog | -0.019 |
Token | are |
Token ID | 389 |
Feature activation | -2.22e+00 |
Pos logit contributions | |
beh | +0.546 |
yours | +0.519 |
possible | +0.467 |
ours | +0.450 |
theirs | +0.442 |
Neg logit contributions | |
boy | -0.353 |
girl | -0.323 |
man | -0.323 |
was | -0.319 |
little | -0.318 |
Token | not |
Token ID | 407 |
Feature activation | -1.12e+01 |
Pos logit contributions | |
beh | +0.122 |
ieve | +0.101 |
irs | +0.097 |
ility | +0.095 |
anks | +0.095 |
Neg logit contributions | |
little | -0.102 |
boy | -0.101 |
was | -0.094 |
man | -0.094 |
and | -0.089 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.11e+01 |
Pos logit contributions | |
beh | +0.235 |
ieve | +0.139 |
eking | +0.139 |
rying | +0.131 |
ility | +0.108 |
Neg logit contributions | |
little | -0.903 |
boy | -0.901 |
man | -0.855 |
girl | -0.850 |
dog | -0.838 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.27e+01 |
Pos logit contributions | |
beh | +1.960 |
yours | +1.647 |
unnecessary | +1.562 |
ieve | +1.554 |
eking | +1.540 |
Neg logit contributions | |
boy | -1.327 |
was | -1.316 |
and | -1.238 |
little | -1.231 |
man | -1.217 |
Token |
|
Token ID | 198 |
Feature activation | +5.67e+00 |
Pos logit contributions | |
beh | +0.913 |
yours | +0.889 |
possible | +0.806 |
theirs | +0.802 |
ours | +0.800 |
Neg logit contributions | |
boy | -0.265 |
girl | -0.260 |
sky | -0.244 |
little | -0.233 |
man | -0.216 |
Token |
|
Token ID | 198 |
Feature activation | +5.64e+00 |
Pos logit contributions | |
boy | +0.112 |
girl | +0.109 |
sky | +0.105 |
man | +0.089 |
little | +0.088 |
Neg logit contributions | |
beh | -0.420 |
yours | -0.413 |
possible | -0.377 |
theirs | -0.371 |
ours | -0.371 |
Token | Tom |
Token ID | 13787 |
Feature activation | +7.27e+00 |
Pos logit contributions | |
sky | +0.169 |
boy | +0.164 |
girl | +0.161 |
little | +0.150 |
was | +0.140 |
Neg logit contributions | |
beh | -0.374 |
yours | -0.347 |
possible | -0.318 |
ourselves | -0.315 |
ours | -0.313 |
Token | and |
Token ID | 290 |
Feature activation | +1.21e+01 |
Pos logit contributions | |
sky | +0.159 |
girl | +0.158 |
woman | +0.156 |
boy | +0.151 |
little | +0.149 |
Neg logit contributions | |
beh | -0.463 |
yours | -0.438 |
myself | -0.416 |
eking | -0.401 |
ws | -0.399 |
Token | they |
Token ID | 484 |
Feature activation | -2.08e+00 |
Pos logit contributions | |
boy | +0.410 |
sky | +0.388 |
girl | +0.388 |
man | +0.382 |
little | +0.377 |
Neg logit contributions | |
beh | -0.459 |
yours | -0.454 |
possible | -0.388 |
ours | -0.383 |
theirs | -0.367 |
Token | have |
Token ID | 423 |
Feature activation | -6.09e+00 |
Pos logit contributions | |
beh | +0.123 |
yours | +0.118 |
possible | +0.105 |
ours | +0.102 |
theirs | +0.100 |
Neg logit contributions | |
boy | -0.085 |
girl | -0.079 |
sky | -0.078 |
little | -0.078 |
man | -0.078 |
Token | a |
Token ID | 257 |
Feature activation | -4.08e+00 |
Pos logit contributions | |
beh | +0.379 |
yours | +0.364 |
possible | +0.324 |
ours | +0.316 |
theirs | +0.309 |
Neg logit contributions | |
boy | -0.230 |
girl | -0.213 |
sky | -0.211 |
little | -0.209 |
man | -0.208 |
Token | trumpet |
Token ID | 40334 |
Feature activation | -5.49e-01 |
Pos logit contributions | |
yours | +0.246 |
beh | +0.225 |
possible | +0.221 |
ours | +0.207 |
unnecessary | +0.203 |
Neg logit contributions | |
boy | -0.133 |
sky | -0.125 |
girl | -0.124 |
was | -0.121 |
man | -0.118 |
Token | like |
Token ID | 588 |
Feature activation | -1.09e+01 |
Pos logit contributions | |
beh | +0.029 |
yours | +0.026 |
unnecessary | +0.023 |
possible | +0.023 |
ours | +0.022 |
Neg logit contributions | |
boy | -0.028 |
man | -0.027 |
girl | -0.026 |
little | -0.026 |
was | -0.026 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.11e+01 |
Pos logit contributions | |
beh | +0.741 |
yours | +0.650 |
possible | +0.603 |
rying | +0.603 |
eking | +0.597 |
Neg logit contributions | |
boy | -0.392 |
little | -0.370 |
sky | -0.369 |
man | -0.356 |
girl | -0.355 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.31e+01 |
Pos logit contributions | |
beh | +1.834 |
yours | +1.521 |
unnecessary | +1.430 |
eking | +1.354 |
possible | +1.347 |
Neg logit contributions | |
boy | -1.508 |
was | -1.418 |
man | -1.417 |
girl | -1.390 |
little | -1.372 |
Token | Tom |
Token ID | 4186 |
Feature activation | -4.85e+00 |
Pos logit contributions | |
beh | +0.644 |
yours | +0.638 |
ours | +0.547 |
theirs | +0.545 |
possible | +0.536 |
Neg logit contributions | |
boy | -0.600 |
girl | -0.583 |
sky | -0.573 |
little | -0.557 |
man | -0.550 |
Token | says |
Token ID | 1139 |
Feature activation | -4.80e-01 |
Pos logit contributions | |
beh | +0.297 |
yours | +0.288 |
ours | +0.251 |
possible | +0.250 |
theirs | +0.244 |
Neg logit contributions | |
boy | -0.174 |
girl | -0.163 |
sky | -0.162 |
little | -0.160 |
man | -0.157 |
Token | . |
Token ID | 13 |
Feature activation | -1.07e+01 |
Pos logit contributions | |
beh | +0.033 |
yours | +0.032 |
ours | +0.029 |
unnecessary | +0.028 |
possible | +0.028 |
Neg logit contributions | |
boy | -0.017 |
and | -0.016 |
little | -0.016 |
was | -0.016 |
man | -0.015 |
Token |
|
Token ID | 198 |
Feature activation | +5.36e+00 |
Pos logit contributions | |
yours | +0.600 |
theirs | +0.576 |
beh | +0.575 |
possible | +0.537 |
properly | +0.526 |
Neg logit contributions | |
girl | -0.277 |
sky | -0.274 |
child | -0.273 |
boy | -0.272 |
field | -0.269 |
Token | is |
Token ID | 318 |
Feature activation | +3.23e+00 |
Pos logit contributions | |
boy | +0.189 |
girl | +0.163 |
was | +0.160 |
sky | +0.159 |
man | +0.157 |
Neg logit contributions | |
beh | -0.757 |
yours | -0.724 |
possible | -0.668 |
ours | -0.653 |
theirs | -0.648 |
Token | not |
Token ID | 407 |
Feature activation | -1.13e+01 |
Pos logit contributions | |
boy | +0.088 |
little | +0.087 |
sky | +0.086 |
was | +0.082 |
girl | +0.080 |
Neg logit contributions | |
beh | -0.240 |
yours | -0.216 |
possible | -0.196 |
ieve | -0.196 |
unnecessary | -0.196 |
Token | as |
Token ID | 355 |
Feature activation | -1.21e+01 |
Pos logit contributions | |
beh | +0.794 |
yours | +0.658 |
ieve | +0.640 |
irs | +0.630 |
rying | +0.613 |
Neg logit contributions | |
little | -0.405 |
boy | -0.387 |
girl | -0.365 |
was | -0.339 |
and | -0.338 |
Token | pretty |
Token ID | 2495 |
Feature activation | -1.61e+00 |
Pos logit contributions | |
beh | +0.890 |
yours | +0.805 |
eking | +0.734 |
ieve | +0.731 |
rying | +0.731 |
Neg logit contributions | |
little | -0.357 |
boy | -0.354 |
girl | -0.329 |
was | -0.320 |
sky | -0.313 |
Token | as |
Token ID | 355 |
Feature activation | -1.21e+01 |
Pos logit contributions | |
beh | +0.109 |
yours | +0.089 |
eking | +0.088 |
supposed | +0.086 |
ours | +0.084 |
Neg logit contributions | |
and | -0.064 |
little | -0.058 |
boy | -0.057 |
girl | -0.056 |
sky | -0.053 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.11e+01 |
Pos logit contributions | |
beh | +0.827 |
yours | +0.668 |
ieve | +0.659 |
eking | +0.657 |
rying | +0.646 |
Neg logit contributions | |
little | -0.458 |
boy | -0.446 |
was | -0.425 |
girl | -0.417 |
sky | -0.414 |
Token | . |
Token ID | 13 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
beh | +2.100 |
unnecessary | +1.721 |
yours | +1.699 |
eking | +1.685 |
ours | +1.618 |
Neg logit contributions | |
was | -1.225 |
boy | -1.196 |
and | -1.155 |
man | -1.131 |
girl | -1.098 |
Token | I |
Token ID | 314 |
Feature activation | -1.26e+01 |
Pos logit contributions | |
beh | +0.707 |
yours | +0.661 |
possible | +0.613 |
ours | +0.587 |
eking | +0.583 |
Neg logit contributions | |
boy | -0.365 |
girl | -0.338 |
sky | -0.334 |
little | -0.328 |
man | -0.318 |
Token | am |
Token ID | 716 |
Feature activation | -3.93e+00 |
Pos logit contributions | |
beh | +0.691 |
yours | +0.663 |
possible | +0.581 |
ours | +0.566 |
theirs | +0.547 |
Neg logit contributions | |
boy | -0.571 |
girl | -0.533 |
was | -0.528 |
man | -0.526 |
little | -0.526 |
Token | jealous |
Token ID | 19354 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
beh | +0.278 |
yours | +0.237 |
unnecessary | +0.222 |
ourselves | +0.221 |
rying | +0.220 |
Neg logit contributions | |
little | -0.135 |
boy | -0.125 |
girl | -0.116 |
man | -0.113 |
sky | -0.111 |
Token | . |
Token ID | 13 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
beh | +0.098 |
yours | +0.092 |
eking | +0.084 |
possible | +0.084 |
unnecessary | +0.083 |
Neg logit contributions | |
boy | -0.037 |
little | -0.034 |
girl | -0.033 |
sky | -0.031 |
was | -0.031 |
Token | my |
Token ID | 616 |
Feature activation | -7.45e+00 |
Pos logit contributions | |
boy | +0.437 |
girl | +0.411 |
little | +0.407 |
sky | +0.405 |
man | +0.404 |
Neg logit contributions | |
beh | -0.462 |
yours | -0.434 |
possible | -0.377 |
ours | -0.364 |
theirs | -0.355 |
Token | tower |
Token ID | 10580 |
Feature activation | +3.48e+00 |
Pos logit contributions | |
beh | +0.488 |
yours | +0.483 |
possible | +0.430 |
ours | +0.423 |
theirs | +0.411 |
Neg logit contributions | |
boy | -0.237 |
girl | -0.219 |
sky | -0.217 |
man | -0.211 |
little | -0.210 |
Token | is |
Token ID | 318 |
Feature activation | +3.04e+00 |
Pos logit contributions | |
boy | +0.075 |
was | +0.065 |
girl | +0.063 |
man | +0.063 |
sky | +0.062 |
Neg logit contributions | |
beh | -0.278 |
yours | -0.264 |
possible | -0.243 |
unnecessary | -0.239 |
theirs | -0.238 |
Token | higher |
Token ID | 2440 |
Feature activation | -5.92e+00 |
Pos logit contributions | |
little | +0.111 |
sky | +0.110 |
boy | +0.109 |
man | +0.104 |
was | +0.104 |
Neg logit contributions | |
beh | -0.201 |
yours | -0.168 |
unnecessary | -0.158 |
ieve | -0.157 |
rying | -0.156 |
Token | than |
Token ID | 621 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
beh | +0.359 |
yours | +0.335 |
possible | +0.289 |
ours | +0.287 |
theirs | +0.283 |
Neg logit contributions | |
boy | -0.248 |
girl | -0.229 |
sky | -0.226 |
man | -0.226 |
little | -0.224 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.11e+01 |
Pos logit contributions | |
beh | +0.612 |
yours | +0.473 |
unnecessary | +0.455 |
ieve | +0.455 |
eking | +0.450 |
Neg logit contributions | |
boy | -0.478 |
little | -0.474 |
sky | -0.464 |
man | -0.448 |
was | -0.447 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.30e+01 |
Pos logit contributions | |
beh | +1.776 |
yours | +1.461 |
unnecessary | +1.367 |
theirs | +1.310 |
eking | +1.305 |
Neg logit contributions | |
boy | -1.542 |
was | -1.467 |
man | -1.451 |
girl | -1.419 |
little | -1.405 |
Token | Ben |
Token ID | 3932 |
Feature activation | -1.21e+00 |
Pos logit contributions | |
beh | +0.742 |
yours | +0.724 |
possible | +0.636 |
ours | +0.624 |
theirs | +0.610 |
Neg logit contributions | |
boy | -0.520 |
girl | -0.491 |
sky | -0.487 |
little | -0.474 |
man | -0.473 |
Token | says |
Token ID | 1139 |
Feature activation | -4.92e-01 |
Pos logit contributions | |
beh | +0.075 |
yours | +0.072 |
possible | +0.064 |
ours | +0.063 |
ieve | +0.061 |
Neg logit contributions | |
boy | -0.043 |
girl | -0.040 |
sky | -0.040 |
little | -0.039 |
man | -0.038 |
Token | , |
Token ID | 11 |
Feature activation | +8.81e+00 |
Pos logit contributions | |
beh | +0.033 |
yours | +0.033 |
ours | +0.030 |
unnecessary | +0.029 |
possible | +0.029 |
Neg logit contributions | |
boy | -0.017 |
little | -0.017 |
and | -0.017 |
was | -0.016 |
man | -0.016 |
Token | smiling |
Token ID | 16755 |
Feature activation | +1.09e+00 |
Pos logit contributions | |
boy | +0.406 |
girl | +0.381 |
sky | +0.379 |
man | +0.370 |
little | +0.366 |
Neg logit contributions | |
beh | -0.462 |
yours | -0.443 |
possible | -0.387 |
ours | -0.374 |
theirs | -0.365 |
Token | . |
Token ID | 13 |
Feature activation | -1.10e+01 |
Pos logit contributions | |
beh | +0.286 |
yours | +0.259 |
possible | +0.234 |
unnecessary | +0.232 |
ours | +0.227 |
Neg logit contributions | |
boy | -0.220 |
was | -0.205 |
girl | -0.204 |
little | -0.203 |
man | -0.202 |
Token | It |
Token ID | 632 |
Feature activation | -7.72e-01 |
Pos logit contributions | |
beh | +0.653 |
yours | +0.643 |
possible | +0.558 |
ours | +0.553 |
theirs | +0.536 |
Neg logit contributions | |
boy | -0.431 |
girl | -0.405 |
man | -0.398 |
sky | -0.397 |
little | -0.395 |
Token | is |
Token ID | 318 |
Feature activation | +2.93e+00 |
Pos logit contributions | |
beh | +0.054 |
yours | +0.053 |
possible | +0.048 |
ours | +0.047 |
ieve | +0.045 |
Neg logit contributions | |
boy | -0.021 |
girl | -0.019 |
man | -0.019 |
little | -0.018 |
sky | -0.018 |
Token | better |
Token ID | 1365 |
Feature activation | -5.70e+00 |
Pos logit contributions | |
boy | +0.107 |
little | +0.101 |
sky | +0.101 |
man | +0.097 |
was | +0.096 |
Neg logit contributions | |
beh | -0.193 |
yours | -0.173 |
possible | -0.155 |
ieve | -0.152 |
unnecessary | -0.152 |
Token | than |
Token ID | 621 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
beh | +0.365 |
yours | +0.317 |
ieve | +0.295 |
eking | +0.288 |
irs | +0.283 |
Neg logit contributions | |
boy | -0.233 |
little | -0.216 |
man | -0.214 |
was | -0.213 |
girl | -0.210 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.11e+01 |
Pos logit contributions | |
beh | +0.519 |
ieve | +0.382 |
yours | +0.360 |
rying | +0.356 |
unnecessary | +0.354 |
Neg logit contributions | |
little | -0.576 |
boy | -0.562 |
sky | -0.558 |
was | -0.538 |
man | -0.525 |
Token | . |
Token ID | 13 |
Feature activation | -1.10e+01 |
Pos logit contributions | |
beh | +1.873 |
yours | +1.501 |
unnecessary | +1.451 |
eking | +1.407 |
possible | +1.374 |
Neg logit contributions | |
boy | -1.441 |
was | -1.423 |
and | -1.357 |
little | -1.356 |
man | -1.350 |
Token | It |
Token ID | 632 |
Feature activation | -7.40e-01 |
Pos logit contributions | |
beh | +0.671 |
yours | +0.652 |
possible | +0.574 |
ours | +0.564 |
theirs | +0.548 |
Neg logit contributions | |
boy | -0.424 |
girl | -0.395 |
sky | -0.388 |
man | -0.387 |
little | -0.385 |
Token | can |
Token ID | 460 |
Feature activation | +1.92e+00 |
Pos logit contributions | |
beh | +0.051 |
yours | +0.050 |
possible | +0.044 |
ours | +0.044 |
ieve | +0.042 |
Neg logit contributions | |
boy | -0.022 |
girl | -0.020 |
sky | -0.020 |
man | -0.020 |
little | -0.019 |
Token | go |
Token ID | 467 |
Feature activation | -4.01e+00 |
Pos logit contributions | |
boy | +0.064 |
man | +0.061 |
little | +0.060 |
was | +0.060 |
sky | +0.060 |
Neg logit contributions | |
beh | -0.135 |
yours | -0.121 |
unnecessary | -0.112 |
supposed | -0.108 |
possible | -0.108 |
Token | faster |
Token ID | 5443 |
Feature activation | -7.76e+00 |
Pos logit contributions | |
beh | +0.247 |
yours | +0.225 |
possible | +0.202 |
ours | +0.194 |
ieve | +0.194 |
Neg logit contributions | |
boy | -0.165 |
little | -0.153 |
sky | -0.151 |
was | -0.150 |
girl | -0.150 |
Token | touch |
Token ID | 3638 |
Feature activation | -8.17e+00 |
Pos logit contributions | |
beh | +1.056 |
yours | +1.021 |
possible | +0.915 |
ours | +0.885 |
theirs | +0.867 |
Neg logit contributions | |
boy | -0.739 |
little | -0.712 |
was | -0.711 |
man | -0.690 |
sky | -0.684 |
Token | something |
Token ID | 1223 |
Feature activation | -1.43e+01 |
Pos logit contributions | |
beh | +0.527 |
yours | +0.471 |
possible | +0.431 |
ours | +0.411 |
eking | +0.407 |
Neg logit contributions | |
boy | -0.313 |
little | -0.298 |
sky | -0.292 |
girl | -0.290 |
was | -0.287 |
Token | that |
Token ID | 326 |
Feature activation | -8.79e+00 |
Pos logit contributions | |
beh | +0.909 |
yours | +0.802 |
possible | +0.713 |
theirs | +0.707 |
eking | +0.682 |
Neg logit contributions | |
little | -0.594 |
boy | -0.588 |
was | -0.566 |
man | -0.556 |
girl | -0.544 |
Token | is |
Token ID | 318 |
Feature activation | +3.38e+00 |
Pos logit contributions | |
beh | +0.603 |
yours | +0.565 |
possible | +0.515 |
ours | +0.499 |
theirs | +0.496 |
Neg logit contributions | |
boy | -0.291 |
was | -0.266 |
man | -0.262 |
sky | -0.260 |
girl | -0.259 |
Token | not |
Token ID | 407 |
Feature activation | -1.12e+01 |
Pos logit contributions | |
little | +0.143 |
boy | +0.143 |
man | +0.137 |
was | +0.136 |
sky | +0.134 |
Neg logit contributions | |
beh | -0.200 |
yours | -0.170 |
ieve | -0.156 |
unnecessary | -0.148 |
rying | -0.148 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.11e+01 |
Pos logit contributions | |
beh | +0.576 |
ieve | +0.441 |
eking | +0.426 |
anks | +0.418 |
rying | +0.414 |
Neg logit contributions | |
little | -0.590 |
boy | -0.584 |
girl | -0.546 |
man | -0.544 |
and | -0.538 |
Token | . |
Token ID | 13 |
Feature activation | -1.05e+01 |
Pos logit contributions | |
beh | +2.139 |
yours | +1.778 |
eking | +1.685 |
unnecessary | +1.664 |
ieve | +1.635 |
Neg logit contributions | |
boy | -1.175 |
was | -1.173 |
and | -1.145 |
little | -1.112 |
man | -1.096 |
Token | And |
Token ID | 843 |
Feature activation | +1.65e+00 |
Pos logit contributions | |
beh | +0.593 |
yours | +0.565 |
possible | +0.502 |
ours | +0.484 |
theirs | +0.471 |
Neg logit contributions | |
boy | -0.457 |
girl | -0.427 |
sky | -0.425 |
little | -0.419 |
man | -0.416 |
Token | you |
Token ID | 345 |
Feature activation | -1.77e+01 |
Pos logit contributions | |
boy | +0.088 |
little | +0.085 |
girl | +0.084 |
was | +0.083 |
man | +0.082 |
Neg logit contributions | |
beh | -0.079 |
yours | -0.071 |
possible | -0.061 |
ours | -0.059 |
theirs | -0.058 |
Token | have |
Token ID | 423 |
Feature activation | -5.82e+00 |
Pos logit contributions | |
beh | +0.956 |
yours | +0.914 |
ours | +0.790 |
possible | +0.783 |
theirs | +0.751 |
Neg logit contributions | |
boy | -0.790 |
girl | -0.757 |
sky | -0.744 |
man | -0.729 |
little | -0.710 |
Token | to |
Token ID | 284 |
Feature activation | +4.11e-01 |
Pos logit contributions | |
beh | +0.353 |
yours | +0.337 |
possible | +0.300 |
ours | +0.292 |
theirs | +0.285 |
Neg logit contributions | |
boy | -0.230 |
girl | -0.213 |
sky | -0.212 |
little | -0.211 |
man | -0.209 |
Token | take |
Token ID | 1011 |
Feature activation | -7.38e+00 |
Pos logit contributions | |
beh | +1.098 |
yours | +1.044 |
possible | +0.970 |
ours | +0.919 |
eking | +0.895 |
Neg logit contributions | |
boy | -0.709 |
was | -0.705 |
little | -0.692 |
girl | -0.649 |
man | -0.646 |
Token | something |
Token ID | 1223 |
Feature activation | -1.44e+01 |
Pos logit contributions | |
beh | +0.463 |
yours | +0.404 |
possible | +0.396 |
rying | +0.385 |
eking | +0.372 |
Neg logit contributions | |
little | -0.290 |
boy | -0.288 |
girl | -0.270 |
was | -0.267 |
man | -0.263 |
Token | that |
Token ID | 326 |
Feature activation | -8.88e+00 |
Pos logit contributions | |
beh | +0.935 |
yours | +0.839 |
possible | +0.765 |
theirs | +0.739 |
ility | +0.729 |
Neg logit contributions | |
little | -0.577 |
boy | -0.546 |
man | -0.521 |
girl | -0.513 |
was | -0.506 |
Token | is |
Token ID | 318 |
Feature activation | +3.32e+00 |
Pos logit contributions | |
beh | +0.623 |
yours | +0.585 |
possible | +0.539 |
ours | +0.518 |
unnecessary | +0.518 |
Neg logit contributions | |
boy | -0.284 |
was | -0.258 |
man | -0.253 |
girl | -0.252 |
little | -0.247 |
Token | not |
Token ID | 407 |
Feature activation | -1.13e+01 |
Pos logit contributions | |
little | +0.145 |
boy | +0.142 |
was | +0.137 |
man | +0.137 |
girl | +0.130 |
Neg logit contributions | |
beh | -0.198 |
yours | -0.158 |
ieve | -0.155 |
unnecessary | -0.147 |
rying | -0.146 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.11e+01 |
Pos logit contributions | |
beh | +0.494 |
ieve | +0.368 |
eking | +0.345 |
rying | +0.338 |
ares | +0.325 |
Neg logit contributions | |
little | -0.700 |
boy | -0.689 |
girl | -0.653 |
and | -0.644 |
man | -0.640 |
Token | . |
Token ID | 13 |
Feature activation | -1.08e+01 |
Pos logit contributions | |
beh | +2.087 |
yours | +1.699 |
eking | +1.660 |
unnecessary | +1.640 |
ieve | +1.596 |
Neg logit contributions | |
boy | -1.241 |
was | -1.237 |
and | -1.211 |
little | -1.165 |
man | -1.146 |
Token | Now |
Token ID | 2735 |
Feature activation | -1.09e+00 |
Pos logit contributions | |
beh | +0.600 |
yours | +0.565 |
possible | +0.511 |
ours | +0.485 |
eking | +0.478 |
Neg logit contributions | |
boy | -0.476 |
girl | -0.445 |
sky | -0.440 |
little | -0.436 |
man | -0.430 |
Token | , |
Token ID | 11 |
Feature activation | +8.91e+00 |
Pos logit contributions | |
beh | +0.082 |
yours | +0.073 |
possible | +0.068 |
theirs | +0.066 |
ieve | +0.066 |
Neg logit contributions | |
boy | -0.032 |
little | -0.031 |
girl | -0.029 |
was | -0.029 |
man | -0.029 |
Token | give |
Token ID | 1577 |
Feature activation | -7.16e+00 |
Pos logit contributions | |
boy | +0.442 |
girl | +0.416 |
little | +0.412 |
man | +0.408 |
was | +0.404 |
Neg logit contributions | |
beh | -0.459 |
yours | -0.425 |
possible | -0.373 |
ours | -0.360 |
theirs | -0.350 |
Token | me |
Token ID | 502 |
Feature activation | -5.83e+00 |
Pos logit contributions | |
beh | +0.391 |
yours | +0.368 |
possible | +0.327 |
ours | +0.313 |
theirs | +0.305 |
Neg logit contributions | |
boy | -0.333 |
girl | -0.313 |
little | -0.309 |
man | -0.307 |
sky | -0.305 |
Token | " |
Token ID | 366 |
Feature activation | +5.53e+00 |
Pos logit contributions | |
yours | +0.609 |
beh | +0.607 |
theirs | +0.548 |
possible | +0.530 |
ours | +0.509 |
Neg logit contributions | |
girl | -0.343 |
boy | -0.341 |
sky | -0.313 |
child | -0.305 |
man | -0.304 |
Token | It |
Token ID | 1026 |
Feature activation | +1.90e+00 |
Pos logit contributions | |
boy | +0.181 |
girl | +0.168 |
man | +0.168 |
little | +0.167 |
sky | +0.162 |
Neg logit contributions | |
beh | -0.359 |
yours | -0.349 |
possible | -0.311 |
ours | -0.303 |
theirs | -0.295 |
Token | is |
Token ID | 318 |
Feature activation | +3.39e+00 |
Pos logit contributions | |
boy | +0.050 |
girl | +0.045 |
sky | +0.044 |
man | +0.044 |
little | +0.043 |
Neg logit contributions | |
beh | -0.139 |
yours | -0.136 |
possible | -0.123 |
ours | -0.120 |
theirs | -0.117 |
Token | bigger |
Token ID | 5749 |
Feature activation | -1.28e+00 |
Pos logit contributions | |
boy | +0.114 |
little | +0.113 |
sky | +0.110 |
was | +0.107 |
girl | +0.103 |
Neg logit contributions | |
beh | -0.234 |
yours | -0.204 |
ieve | -0.186 |
possible | -0.184 |
rying | -0.181 |
Token | than |
Token ID | 621 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
beh | +0.087 |
yours | +0.081 |
possible | +0.073 |
ours | +0.071 |
theirs | +0.070 |
Neg logit contributions | |
boy | -0.043 |
girl | -0.039 |
sky | -0.039 |
little | -0.039 |
man | -0.039 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.11e+01 |
Pos logit contributions | |
beh | +0.578 |
ieve | +0.425 |
yours | +0.401 |
unnecessary | +0.398 |
eking | +0.393 |
Neg logit contributions | |
little | -0.533 |
boy | -0.509 |
sky | -0.506 |
was | -0.493 |
girl | -0.466 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.28e+01 |
Pos logit contributions | |
beh | +1.853 |
yours | +1.453 |
unnecessary | +1.414 |
eking | +1.360 |
ieve | +1.354 |
Neg logit contributions | |
boy | -1.483 |
was | -1.445 |
and | -1.405 |
little | -1.405 |
man | -1.382 |
Token |
|
Token ID | 198 |
Feature activation | +5.55e+00 |
Pos logit contributions | |
beh | +0.881 |
yours | +0.863 |
possible | +0.789 |
theirs | +0.768 |
ours | +0.756 |
Neg logit contributions | |
boy | -0.324 |
girl | -0.310 |
sky | -0.300 |
little | -0.284 |
man | -0.277 |
Token |
|
Token ID | 198 |
Feature activation | +5.56e+00 |
Pos logit contributions | |
boy | +0.097 |
girl | +0.086 |
sky | +0.082 |
man | +0.076 |
little | +0.074 |
Neg logit contributions | |
beh | -0.444 |
yours | -0.437 |
possible | -0.400 |
ours | -0.389 |
theirs | -0.387 |
Token | Sam |
Token ID | 16305 |
Feature activation | +5.21e+00 |
Pos logit contributions | |
boy | +0.155 |
sky | +0.150 |
girl | +0.143 |
little | +0.138 |
man | +0.133 |
Neg logit contributions | |
beh | -0.394 |
yours | -0.373 |
possible | -0.341 |
ours | -0.325 |
ourselves | -0.324 |
Token | is |
Token ID | 318 |
Feature activation | +3.32e+00 |
Pos logit contributions | |
boy | +0.190 |
girl | +0.181 |
sky | +0.177 |
little | +0.176 |
man | +0.173 |
Neg logit contributions | |
beh | -0.292 |
yours | -0.286 |
ieve | -0.248 |
myself | -0.243 |
properly | -0.240 |
Token | Can |
Token ID | 6090 |
Feature activation | +1.70e+00 |
Pos logit contributions | |
boy | +0.188 |
girl | +0.173 |
sky | +0.171 |
little | +0.171 |
man | +0.170 |
Neg logit contributions | |
beh | -0.346 |
yours | -0.335 |
possible | -0.299 |
ours | -0.291 |
theirs | -0.284 |
Token | I |
Token ID | 314 |
Feature activation | -1.26e+01 |
Pos logit contributions | |
boy | +0.085 |
girl | +0.079 |
was | +0.079 |
little | +0.079 |
man | +0.079 |
Neg logit contributions | |
beh | -0.088 |
yours | -0.080 |
possible | -0.070 |
unnecessary | -0.069 |
ours | -0.068 |
Token | have |
Token ID | 423 |
Feature activation | -6.21e+00 |
Pos logit contributions | |
beh | +0.796 |
yours | +0.722 |
unnecessary | +0.672 |
possible | +0.667 |
ours | +0.646 |
Neg logit contributions | |
boy | -0.490 |
little | -0.465 |
was | -0.465 |
girl | -0.457 |
man | -0.451 |
Token | some |
Token ID | 617 |
Feature activation | -9.65e+00 |
Pos logit contributions | |
beh | +0.417 |
yours | +0.354 |
rying | +0.345 |
possible | +0.335 |
eking | +0.333 |
Neg logit contributions | |
boy | -0.220 |
little | -0.219 |
sky | -0.208 |
girl | -0.205 |
was | -0.204 |
Token | of |
Token ID | 286 |
Feature activation | +3.15e+00 |
Pos logit contributions | |
beh | +0.661 |
yours | +0.634 |
possible | +0.572 |
ours | +0.559 |
theirs | +0.547 |
Neg logit contributions | |
boy | -0.306 |
girl | -0.280 |
sky | -0.277 |
little | -0.274 |
man | -0.273 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.11e+01 |
Pos logit contributions | |
boy | +0.148 |
little | +0.147 |
sky | +0.142 |
man | +0.141 |
girl | +0.141 |
Neg logit contributions | |
beh | -0.175 |
unnecessary | -0.138 |
yours | -0.138 |
rying | -0.136 |
possible | -0.132 |
Token | ?" |
Token ID | 1701 |
Feature activation | -1.11e+01 |
Pos logit contributions | |
beh | +1.911 |
yours | +1.687 |
possible | +1.529 |
unnecessary | +1.517 |
ours | +1.500 |
Neg logit contributions | |
boy | -1.313 |
girl | -1.227 |
was | -1.218 |
man | -1.210 |
little | -1.201 |
Token |
|
Token ID | 198 |
Feature activation | +5.40e+00 |
Pos logit contributions | |
yours | +0.828 |
beh | +0.826 |
possible | +0.754 |
theirs | +0.731 |
ours | +0.729 |
Neg logit contributions | |
boy | -0.226 |
girl | -0.205 |
sky | -0.192 |
little | -0.191 |
man | -0.186 |
Token |
|
Token ID | 198 |
Feature activation | +5.42e+00 |
Pos logit contributions | |
boy | +0.120 |
girl | +0.112 |
sky | +0.104 |
little | +0.096 |
man | +0.094 |
Neg logit contributions | |
yours | -0.397 |
beh | -0.391 |
possible | -0.357 |
ours | -0.347 |
ourselves | -0.347 |
Token | " |
Token ID | 1 |
Feature activation | +7.16e+00 |
Pos logit contributions | |
boy | +0.176 |
sky | +0.166 |
girl | +0.166 |
little | +0.164 |
man | +0.145 |
Neg logit contributions | |
beh | -0.344 |
yours | -0.336 |
myself | -0.302 |
ourselves | -0.302 |
possible | -0.302 |
Token | Of |
Token ID | 5189 |
Feature activation | -1.08e+00 |
Pos logit contributions | |
boy | +0.261 |
girl | +0.241 |
sky | +0.239 |
little | +0.236 |
man | +0.236 |
Neg logit contributions | |
beh | -0.453 |
yours | -0.437 |
possible | -0.389 |
ours | -0.380 |
theirs | -0.370 |
Token | my |
Token ID | 616 |
Feature activation | -7.59e+00 |
Pos logit contributions | |
boy | +0.371 |
girl | +0.349 |
sky | +0.345 |
man | +0.342 |
little | +0.337 |
Neg logit contributions | |
beh | -0.498 |
yours | -0.488 |
possible | -0.422 |
ours | -0.416 |
theirs | -0.402 |
Token | car |
Token ID | 1097 |
Feature activation | +4.20e+00 |
Pos logit contributions | |
yours | +0.534 |
beh | +0.531 |
ours | +0.474 |
possible | +0.472 |
theirs | +0.463 |
Neg logit contributions | |
boy | -0.169 |
sky | -0.160 |
girl | -0.160 |
man | -0.153 |
was | -0.147 |
Token | is |
Token ID | 318 |
Feature activation | +3.16e+00 |
Pos logit contributions | |
boy | +0.105 |
girl | +0.091 |
man | +0.091 |
was | +0.089 |
sky | +0.088 |
Neg logit contributions | |
beh | -0.321 |
yours | -0.305 |
possible | -0.282 |
ours | -0.274 |
theirs | -0.273 |
Token | faster |
Token ID | 5443 |
Feature activation | -7.84e+00 |
Pos logit contributions | |
boy | +0.113 |
little | +0.113 |
was | +0.105 |
sky | +0.104 |
man | +0.103 |
Neg logit contributions | |
beh | -0.211 |
yours | -0.175 |
rying | -0.168 |
ieve | -0.168 |
unnecessary | -0.168 |
Token | than |
Token ID | 621 |
Feature activation | -1.07e+01 |
Pos logit contributions | |
beh | +0.414 |
yours | +0.369 |
possible | +0.319 |
unnecessary | +0.310 |
theirs | +0.310 |
Neg logit contributions | |
boy | -0.401 |
man | -0.372 |
girl | -0.370 |
little | -0.368 |
was | -0.363 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.11e+01 |
Pos logit contributions | |
beh | +0.630 |
ieve | +0.472 |
rying | +0.466 |
eking | +0.466 |
unnecessary | +0.461 |
Neg logit contributions | |
little | -0.480 |
boy | -0.477 |
sky | -0.462 |
was | -0.444 |
man | -0.438 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.30e+01 |
Pos logit contributions | |
beh | +1.789 |
yours | +1.454 |
unnecessary | +1.362 |
eking | +1.330 |
theirs | +1.322 |
Neg logit contributions | |
boy | -1.550 |
man | -1.442 |
was | -1.440 |
girl | -1.404 |
little | -1.403 |
Token | Sara |
Token ID | 24799 |
Feature activation | +1.87e+00 |
Pos logit contributions | |
beh | +0.654 |
yours | +0.638 |
possible | +0.548 |
ours | +0.541 |
theirs | +0.533 |
Neg logit contributions | |
boy | -0.589 |
sky | -0.564 |
girl | -0.557 |
man | -0.545 |
little | -0.534 |
Token | says |
Token ID | 1139 |
Feature activation | -4.04e-01 |
Pos logit contributions | |
boy | +0.066 |
sky | +0.063 |
girl | +0.063 |
little | +0.062 |
man | +0.060 |
Neg logit contributions | |
beh | -0.115 |
yours | -0.110 |
possible | -0.097 |
ours | -0.096 |
ieve | -0.094 |
Token | . |
Token ID | 13 |
Feature activation | -1.07e+01 |
Pos logit contributions | |
beh | +0.028 |
yours | +0.027 |
ours | +0.024 |
possible | +0.024 |
unnecessary | +0.023 |
Neg logit contributions | |
boy | -0.014 |
little | -0.013 |
and | -0.013 |
was | -0.013 |
man | -0.013 |
Token |
|
Token ID | 198 |
Feature activation | +5.39e+00 |
Pos logit contributions | |
yours | +0.716 |
beh | +0.710 |
possible | +0.644 |
theirs | +0.622 |
ours | +0.617 |
Neg logit contributions | |
boy | -0.280 |
girl | -0.251 |
man | -0.237 |
child | -0.229 |
sky | -0.229 |
Token | is |
Token ID | 318 |
Feature activation | +3.15e+00 |
Pos logit contributions | |
boy | +0.328 |
man | +0.291 |
girl | +0.288 |
was | +0.287 |
little | +0.282 |
Neg logit contributions | |
beh | -0.789 |
yours | -0.745 |
possible | -0.682 |
ours | -0.663 |
theirs | -0.660 |
Token | faster |
Token ID | 5443 |
Feature activation | -7.74e+00 |
Pos logit contributions | |
little | +0.128 |
boy | +0.126 |
man | +0.120 |
was | +0.116 |
sky | +0.113 |
Neg logit contributions | |
beh | -0.201 |
rying | -0.160 |
yours | -0.159 |
ieve | -0.158 |
irs | -0.156 |
Token | and |
Token ID | 290 |
Feature activation | +1.19e+01 |
Pos logit contributions | |
beh | +0.416 |
yours | +0.366 |
possible | +0.317 |
ieve | +0.314 |
unnecessary | +0.309 |
Neg logit contributions | |
boy | -0.390 |
man | -0.363 |
little | -0.361 |
girl | -0.360 |
was | -0.356 |
Token | smarter |
Token ID | 23714 |
Feature activation | -8.97e+00 |
Pos logit contributions | |
boy | +0.502 |
was | +0.474 |
little | +0.462 |
man | +0.455 |
girl | +0.434 |
Neg logit contributions | |
beh | -0.715 |
yours | -0.622 |
irs | -0.593 |
owing | -0.574 |
ility | -0.574 |
Token | than |
Token ID | 621 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
beh | +0.478 |
yours | +0.413 |
asty | +0.364 |
eking | +0.363 |
irs | +0.361 |
Neg logit contributions | |
boy | -0.461 |
man | -0.439 |
girl | -0.421 |
was | -0.415 |
little | -0.413 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.10e+01 |
Pos logit contributions | |
beh | +0.593 |
ieve | +0.453 |
yours | +0.432 |
rying | +0.432 |
eking | +0.429 |
Neg logit contributions | |
little | -0.514 |
boy | -0.509 |
sky | -0.484 |
was | -0.483 |
man | -0.480 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.29e+01 |
Pos logit contributions | |
beh | +1.806 |
yours | +1.442 |
unnecessary | +1.362 |
eking | +1.340 |
ility | +1.316 |
Neg logit contributions | |
boy | -1.541 |
was | -1.464 |
man | -1.449 |
little | -1.413 |
and | -1.400 |
Token | Lily |
Token ID | 20037 |
Feature activation | -3.55e+00 |
Pos logit contributions | |
beh | +0.654 |
yours | +0.637 |
possible | +0.553 |
theirs | +0.545 |
ours | +0.535 |
Neg logit contributions | |
sky | -0.538 |
boy | -0.528 |
girl | -0.507 |
man | -0.495 |
was | -0.483 |
Token | said |
Token ID | 531 |
Feature activation | +3.20e-02 |
Pos logit contributions | |
beh | +0.207 |
yours | +0.177 |
myself | +0.161 |
eking | +0.160 |
properly | +0.159 |
Neg logit contributions | |
sky | -0.131 |
boy | -0.116 |
girl | -0.115 |
field | -0.109 |
little | -0.107 |
Token | , |
Token ID | 11 |
Feature activation | +8.97e+00 |
Pos logit contributions | |
boy | +0.001 |
little | +0.000 |
girl | +0.000 |
man | +0.000 |
was | +0.000 |
Neg logit contributions | |
beh | -0.003 |
yours | -0.003 |
ours | -0.003 |
possible | -0.002 |
unnecessary | -0.002 |
Token | holding |
Token ID | 4769 |
Feature activation | +3.15e+00 |
Pos logit contributions | |
boy | +0.408 |
girl | +0.383 |
sky | +0.379 |
man | +0.376 |
little | +0.375 |
Neg logit contributions | |
beh | -0.486 |
yours | -0.467 |
possible | -0.407 |
ours | -0.396 |
theirs | -0.384 |
Token | not |
Token ID | 407 |
Feature activation | -1.12e+01 |
Pos logit contributions | |
beh | +0.348 |
yours | +0.328 |
possible | +0.278 |
ours | +0.270 |
theirs | +0.258 |
Neg logit contributions | |
boy | -0.415 |
girl | -0.391 |
little | -0.391 |
man | -0.390 |
sky | -0.390 |
Token | take |
Token ID | 1011 |
Feature activation | -7.38e+00 |
Pos logit contributions | |
beh | +0.709 |
yours | +0.677 |
possible | +0.625 |
ieve | +0.605 |
eking | +0.604 |
Neg logit contributions | |
boy | -0.422 |
little | -0.402 |
was | -0.374 |
girl | -0.371 |
bird | -0.354 |
Token | what |
Token ID | 644 |
Feature activation | -1.70e+01 |
Pos logit contributions | |
beh | +0.464 |
yours | +0.428 |
possible | +0.403 |
rying | +0.380 |
ours | +0.379 |
Neg logit contributions | |
boy | -0.279 |
little | -0.271 |
girl | -0.260 |
man | -0.246 |
was | -0.243 |
Token | is |
Token ID | 318 |
Feature activation | +3.12e+00 |
Pos logit contributions | |
beh | +0.831 |
yours | +0.773 |
possible | +0.665 |
ours | +0.649 |
theirs | +0.630 |
Neg logit contributions | |
boy | -0.887 |
girl | -0.830 |
was | -0.825 |
little | -0.824 |
man | -0.824 |
Token | not |
Token ID | 407 |
Feature activation | -1.12e+01 |
Pos logit contributions | |
little | +0.144 |
man | +0.143 |
boy | +0.136 |
and | +0.134 |
was | +0.133 |
Neg logit contributions | |
beh | -0.177 |
rying | -0.142 |
ility | -0.136 |
owing | -0.132 |
rol | -0.130 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.10e+01 |
Pos logit contributions | |
beh | +0.560 |
eking | +0.436 |
ieve | +0.425 |
rying | +0.421 |
yours | +0.397 |
Neg logit contributions | |
boy | -0.622 |
little | -0.611 |
girl | -0.579 |
man | -0.574 |
and | -0.556 |
Token | . |
Token ID | 13 |
Feature activation | -1.07e+01 |
Pos logit contributions | |
beh | +2.199 |
yours | +1.889 |
unnecessary | +1.774 |
ours | +1.737 |
possible | +1.731 |
Neg logit contributions | |
was | -1.137 |
boy | -1.066 |
and | -1.059 |
little | -0.984 |
man | -0.974 |
Token | Do |
Token ID | 2141 |
Feature activation | -7.53e+00 |
Pos logit contributions | |
beh | +0.631 |
yours | +0.603 |
possible | +0.538 |
ours | +0.521 |
unnecessary | +0.504 |
Neg logit contributions | |
boy | -0.437 |
girl | -0.404 |
sky | -0.404 |
little | -0.399 |
man | -0.396 |
Token | not |
Token ID | 407 |
Feature activation | -1.11e+01 |
Pos logit contributions | |
beh | +0.342 |
yours | +0.321 |
possible | +0.273 |
ours | +0.265 |
theirs | +0.253 |
Neg logit contributions | |
boy | -0.412 |
little | -0.391 |
man | -0.388 |
girl | -0.387 |
was | -0.385 |
Token | use |
Token ID | 779 |
Feature activation | -9.04e+00 |
Pos logit contributions | |
beh | +0.700 |
yours | +0.659 |
possible | +0.614 |
theirs | +0.589 |
ours | +0.588 |
Neg logit contributions | |
boy | -0.435 |
little | -0.413 |
was | -0.386 |
girl | -0.382 |
bird | -0.371 |
Token | things |
Token ID | 1243 |
Feature activation | -8.11e-01 |
Pos logit contributions | |
beh | +0.538 |
yours | +0.437 |
rying | +0.410 |
possible | +0.407 |
eking | +0.394 |
Neg logit contributions | |
little | -0.395 |
boy | -0.388 |
was | -0.369 |
sky | -0.368 |
man | -0.367 |
Token | car |
Token ID | 1097 |
Feature activation | +4.34e+00 |
Pos logit contributions | |
boy | +0.121 |
girl | +0.112 |
sky | +0.110 |
was | +0.110 |
little | +0.106 |
Neg logit contributions | |
yours | -0.374 |
beh | -0.367 |
possible | -0.335 |
ours | -0.334 |
theirs | -0.323 |
Token | . |
Token ID | 13 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
boy | +0.175 |
man | +0.161 |
girl | +0.159 |
little | +0.157 |
was | +0.151 |
Neg logit contributions | |
beh | -0.275 |
yours | -0.246 |
eking | -0.230 |
unnecessary | -0.228 |
ieve | -0.224 |
Token | It |
Token ID | 632 |
Feature activation | -7.72e-01 |
Pos logit contributions | |
beh | +0.658 |
yours | +0.623 |
possible | +0.568 |
ours | +0.542 |
eking | +0.535 |
Neg logit contributions | |
boy | -0.409 |
girl | -0.378 |
sky | -0.378 |
little | -0.373 |
man | -0.368 |
Token | is |
Token ID | 318 |
Feature activation | +3.25e+00 |
Pos logit contributions | |
beh | +0.059 |
yours | +0.057 |
possible | +0.053 |
ours | +0.052 |
theirs | +0.051 |
Neg logit contributions | |
boy | -0.018 |
girl | -0.016 |
was | -0.015 |
sky | -0.015 |
man | -0.015 |
Token | not |
Token ID | 407 |
Feature activation | -1.13e+01 |
Pos logit contributions | |
little | +0.129 |
boy | +0.127 |
was | +0.118 |
man | +0.118 |
girl | +0.116 |
Neg logit contributions | |
beh | -0.205 |
yours | -0.168 |
rying | -0.165 |
ieve | -0.162 |
eking | -0.161 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.10e+01 |
Pos logit contributions | |
beh | +0.726 |
ieve | +0.592 |
eking | +0.590 |
rying | +0.583 |
yours | +0.579 |
Neg logit contributions | |
boy | -0.467 |
little | -0.455 |
girl | -0.424 |
man | -0.406 |
was | -0.385 |
Token | ," |
Token ID | 553 |
Feature activation | -6.94e+00 |
Pos logit contributions | |
beh | +2.165 |
yours | +1.852 |
eking | +1.751 |
unnecessary | +1.740 |
ieve | +1.707 |
Neg logit contributions | |
boy | -1.161 |
was | -1.102 |
man | -1.068 |
girl | -1.038 |
and | -1.034 |
Token | she |
Token ID | 673 |
Feature activation | -5.01e+00 |
Pos logit contributions | |
beh | +0.365 |
yours | +0.356 |
possible | +0.314 |
theirs | +0.301 |
ours | +0.301 |
Neg logit contributions | |
boy | -0.292 |
sky | -0.286 |
girl | -0.276 |
little | -0.271 |
man | -0.264 |
Token | says |
Token ID | 1139 |
Feature activation | -3.45e-01 |
Pos logit contributions | |
beh | +0.334 |
yours | +0.323 |
possible | +0.289 |
ours | +0.284 |
theirs | +0.277 |
Neg logit contributions | |
boy | -0.165 |
girl | -0.152 |
sky | -0.150 |
man | -0.147 |
little | -0.147 |
Token | . |
Token ID | 13 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
beh | +0.024 |
yours | +0.023 |
ours | +0.021 |
unnecessary | +0.021 |
possible | +0.020 |
Neg logit contributions | |
and | -0.012 |
little | -0.011 |
boy | -0.011 |
man | -0.011 |
was | -0.011 |
Token |
|
Token ID | 198 |
Feature activation | +5.60e+00 |
Pos logit contributions | |
yours | +0.723 |
beh | +0.698 |
possible | +0.653 |
theirs | +0.642 |
ours | +0.630 |
Neg logit contributions | |
boy | -0.254 |
girl | -0.231 |
sky | -0.202 |
child | -0.201 |
little | -0.200 |
Token | . |
Token ID | 13 |
Feature activation | -1.09e+01 |
Pos logit contributions | |
boy | +0.211 |
girl | +0.194 |
little | +0.192 |
man | +0.192 |
was | +0.189 |
Neg logit contributions | |
beh | -0.304 |
yours | -0.291 |
possible | -0.266 |
unnecessary | -0.252 |
ours | -0.248 |
Token | The |
Token ID | 383 |
Feature activation | +5.13e+00 |
Pos logit contributions | |
beh | +0.632 |
yours | +0.599 |
possible | +0.544 |
ours | +0.517 |
unnecessary | +0.506 |
Neg logit contributions | |
boy | -0.461 |
girl | -0.429 |
sky | -0.422 |
little | -0.419 |
man | -0.417 |
Token | rake |
Token ID | 47885 |
Feature activation | -1.43e+00 |
Pos logit contributions | |
boy | +0.181 |
girl | +0.164 |
sky | +0.163 |
man | +0.163 |
little | +0.161 |
Neg logit contributions | |
beh | -0.312 |
yours | -0.309 |
possible | -0.274 |
ours | -0.268 |
theirs | -0.260 |
Token | is |
Token ID | 318 |
Feature activation | +3.10e+00 |
Pos logit contributions | |
beh | +0.114 |
yours | +0.109 |
possible | +0.101 |
ours | +0.098 |
theirs | +0.097 |
Neg logit contributions | |
boy | -0.030 |
was | -0.026 |
girl | -0.026 |
sky | -0.025 |
man | -0.025 |
Token | not |
Token ID | 407 |
Feature activation | -1.12e+01 |
Pos logit contributions | |
little | +0.136 |
boy | +0.135 |
was | +0.129 |
man | +0.128 |
sky | +0.126 |
Neg logit contributions | |
beh | -0.177 |
yours | -0.150 |
ieve | -0.141 |
possible | -0.135 |
rying | -0.134 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.10e+01 |
Pos logit contributions | |
beh | +0.686 |
ieve | +0.586 |
eking | +0.553 |
irs | +0.550 |
ares | +0.547 |
Neg logit contributions | |
little | -0.474 |
boy | -0.468 |
girl | -0.430 |
was | -0.412 |
and | -0.409 |
Token | , |
Token ID | 11 |
Feature activation | +8.89e+00 |
Pos logit contributions | |
beh | +2.160 |
yours | +1.849 |
eking | +1.765 |
unnecessary | +1.764 |
ieve | +1.736 |
Neg logit contributions | |
was | -1.168 |
boy | -1.167 |
and | -1.059 |
man | -1.056 |
girl | -1.025 |
Token | it |
Token ID | 340 |
Feature activation | -3.07e+00 |
Pos logit contributions | |
boy | +0.424 |
girl | +0.395 |
little | +0.395 |
man | +0.392 |
sky | +0.385 |
Neg logit contributions | |
beh | -0.482 |
yours | -0.440 |
possible | -0.391 |
ours | -0.372 |
ieve | -0.366 |
Token | 's |
Token ID | 338 |
Feature activation | +5.71e+00 |
Pos logit contributions | |
beh | +0.240 |
yours | +0.234 |
possible | +0.216 |
ours | +0.210 |
theirs | +0.209 |
Neg logit contributions | |
boy | -0.067 |
was | -0.059 |
girl | -0.057 |
little | -0.056 |
man | -0.056 |
Token | dad |
Token ID | 9955 |
Feature activation | +4.59e+00 |
Pos logit contributions | |
little | +0.252 |
was | +0.243 |
boy | +0.241 |
and | +0.236 |
man | +0.221 |
Neg logit contributions | |
beh | -0.327 |
ieve | -0.278 |
rying | -0.253 |
ility | -0.253 |
irs | -0.253 |
Token | 's |
Token ID | 338 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
boy | +0.169 |
was | +0.161 |
man | +0.153 |
girl | +0.151 |
little | +0.150 |
Neg logit contributions | |
beh | -0.299 |
yours | -0.279 |
possible | -0.261 |
unnecessary | -0.253 |
theirs | -0.250 |
Token | had |
Token ID | 550 |
Feature activation | +7.37e+00 |
Pos logit contributions | |
beh | +0.524 |
yours | +0.506 |
possible | +0.463 |
ours | +0.455 |
theirs | +0.445 |
Neg logit contributions | |
boy | -0.166 |
girl | -0.147 |
little | -0.146 |
was | -0.145 |
sky | -0.141 |
Token | to |
Token ID | 284 |
Feature activation | +2.97e+00 |
Pos logit contributions | |
little | +0.259 |
was | +0.254 |
boy | +0.251 |
sky | +0.245 |
man | +0.234 |
Neg logit contributions | |
beh | -0.475 |
yours | -0.432 |
possible | -0.391 |
ieve | -0.390 |
rying | -0.381 |
Token | pick |
Token ID | 2298 |
Feature activation | +2.85e+00 |
Pos logit contributions | |
little | +0.120 |
boy | +0.117 |
man | +0.114 |
was | +0.113 |
girl | +0.110 |
Neg logit contributions | |
beh | -0.179 |
yours | -0.164 |
possible | -0.149 |
unnecessary | -0.144 |
ours | -0.144 |
Token | it |
Token ID | 340 |
Feature activation | -9.85e-01 |
Pos logit contributions | |
boy | +0.115 |
girl | +0.109 |
little | +0.108 |
was | +0.107 |
sky | +0.107 |
Neg logit contributions | |
beh | -0.174 |
yours | -0.160 |
possible | -0.144 |
ours | -0.141 |
eking | -0.140 |
Token | up |
Token ID | 510 |
Feature activation | -1.16e+01 |
Pos logit contributions | |
beh | +0.066 |
yours | +0.055 |
unnecessary | +0.053 |
supervision | +0.052 |
ieve | +0.051 |
Neg logit contributions | |
and | -0.039 |
was | -0.039 |
little | -0.038 |
boy | -0.035 |
man | -0.033 |
Token | and |
Token ID | 290 |
Feature activation | +1.39e+01 |
Pos logit contributions | |
beh | +0.797 |
yours | +0.704 |
ieve | +0.652 |
unnecessary | +0.647 |
eking | +0.629 |
Neg logit contributions | |
was | -0.394 |
boy | -0.394 |
and | -0.392 |
little | -0.392 |
girl | -0.367 |
Token | give |
Token ID | 1577 |
Feature activation | -5.53e+00 |
Pos logit contributions | |
boy | +0.539 |
girl | +0.501 |
was | +0.501 |
man | +0.498 |
little | +0.494 |
Neg logit contributions | |
beh | -0.866 |
yours | -0.821 |
possible | -0.740 |
ours | -0.713 |
theirs | -0.706 |
Token | it |
Token ID | 340 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
beh | +0.358 |
yours | +0.326 |
possible | +0.300 |
ours | +0.292 |
eking | +0.290 |
Neg logit contributions | |
boy | -0.207 |
girl | -0.196 |
was | -0.195 |
little | -0.194 |
man | -0.191 |
Token | a |
Token ID | 257 |
Feature activation | -1.97e+00 |
Pos logit contributions | |
beh | +0.096 |
ieve | +0.085 |
eking | +0.083 |
yours | +0.082 |
eling | +0.081 |
Neg logit contributions | |
little | -0.041 |
was | -0.039 |
boy | -0.037 |
and | -0.036 |
girl | -0.034 |
Token | hug |
Token ID | 16225 |
Feature activation | +2.25e+00 |
Pos logit contributions | |
beh | +0.118 |
yours | +0.117 |
possible | +0.103 |
ours | +0.101 |
theirs | +0.097 |
Neg logit contributions | |
boy | -0.075 |
girl | -0.070 |
sky | -0.069 |
man | -0.068 |
little | -0.067 |
Token | . |
Token ID | 13 |
Feature activation | -9.13e+00 |
Pos logit contributions | |
boy | +0.090 |
was | +0.089 |
man | +0.087 |
sky | +0.085 |
girl | +0.085 |
Neg logit contributions | |
beh | -0.140 |
yours | -0.121 |
unnecessary | -0.118 |
eking | -0.114 |
ieve | -0.114 |
Token | 't |
Token ID | 470 |
Feature activation | -7.24e+00 |
Pos logit contributions | |
boy | +0.121 |
man | +0.115 |
sky | +0.114 |
was | +0.113 |
little | +0.113 |
Neg logit contributions | |
beh | -0.160 |
yours | -0.150 |
possible | -0.130 |
ours | -0.130 |
unnecessary | -0.129 |
Token | wait |
Token ID | 4043 |
Feature activation | -1.30e+01 |
Pos logit contributions | |
beh | +0.433 |
yours | +0.400 |
ours | +0.356 |
unnecessary | +0.344 |
possible | +0.341 |
Neg logit contributions | |
boy | -0.327 |
man | -0.312 |
was | -0.309 |
little | -0.308 |
sky | -0.301 |
Token | to |
Token ID | 284 |
Feature activation | +1.12e+00 |
Pos logit contributions | |
beh | +0.970 |
yours | +0.815 |
ours | +0.753 |
possible | +0.738 |
ieve | +0.734 |
Neg logit contributions | |
little | -0.450 |
and | -0.431 |
was | -0.415 |
boy | -0.414 |
sky | -0.380 |
Token | see |
Token ID | 766 |
Feature activation | -5.23e+00 |
Pos logit contributions | |
boy | +0.043 |
little | +0.042 |
man | +0.040 |
sky | +0.040 |
girl | +0.040 |
Neg logit contributions | |
beh | -0.072 |
yours | -0.064 |
possible | -0.058 |
ours | -0.056 |
unnecessary | -0.056 |
Token | them |
Token ID | 606 |
Feature activation | -1.01e+01 |
Pos logit contributions | |
beh | +0.379 |
yours | +0.314 |
possible | +0.294 |
eking | +0.292 |
rying | +0.289 |
Neg logit contributions | |
boy | -0.189 |
little | -0.184 |
girl | -0.173 |
man | -0.172 |
was | -0.171 |
Token | and |
Token ID | 290 |
Feature activation | +1.28e+01 |
Pos logit contributions | |
beh | +0.649 |
yours | +0.598 |
ours | +0.526 |
unnecessary | +0.524 |
possible | +0.519 |
Neg logit contributions | |
boy | -0.392 |
was | -0.378 |
little | -0.376 |
and | -0.365 |
man | -0.365 |
Token | give |
Token ID | 1577 |
Feature activation | -6.24e+00 |
Pos logit contributions | |
boy | +0.543 |
girl | +0.504 |
man | +0.501 |
little | +0.500 |
was | +0.491 |
Neg logit contributions | |
beh | -0.760 |
yours | -0.710 |
possible | -0.630 |
ours | -0.611 |
unnecessary | -0.597 |
Token | them |
Token ID | 606 |
Feature activation | -1.01e+01 |
Pos logit contributions | |
beh | +0.341 |
yours | +0.308 |
possible | +0.280 |
ieve | +0.268 |
ours | +0.268 |
Neg logit contributions | |
boy | -0.306 |
girl | -0.290 |
was | -0.283 |
little | -0.283 |
man | -0.281 |
Token | hugs |
Token ID | 40657 |
Feature activation | +1.84e+00 |
Pos logit contributions | |
beh | +0.679 |
yours | +0.619 |
eking | +0.579 |
possible | +0.571 |
ieve | +0.558 |
Neg logit contributions | |
boy | -0.342 |
little | -0.330 |
girl | -0.316 |
man | -0.308 |
was | -0.307 |
Token | and |
Token ID | 290 |
Feature activation | +1.29e+01 |
Pos logit contributions | |
boy | +0.063 |
girl | +0.058 |
man | +0.057 |
was | +0.057 |
sky | +0.056 |
Neg logit contributions | |
beh | -0.128 |
yours | -0.118 |
possible | -0.108 |
ours | -0.107 |
unnecessary | -0.105 |
Token | kisses |
Token ID | 36544 |
Feature activation | +1.30e+00 |
Pos logit contributions | |
boy | +0.490 |
girl | +0.454 |
little | +0.447 |
sky | +0.446 |
man | +0.445 |
Neg logit contributions | |
beh | -0.806 |
yours | -0.773 |
possible | -0.689 |
ours | -0.671 |
theirs | -0.652 |
Token | in |
Token ID | 287 |
Feature activation | -1.53e-01 |
Pos logit contributions | |
boy | +0.350 |
was | +0.336 |
sky | +0.334 |
little | +0.326 |
girl | +0.325 |
Neg logit contributions | |
beh | -0.605 |
yours | -0.560 |
unnecessary | -0.505 |
ours | -0.494 |
possible | -0.490 |
Token | the |
Token ID | 262 |
Feature activation | +6.01e+00 |
Pos logit contributions | |
beh | +0.010 |
yours | +0.010 |
possible | +0.009 |
ours | +0.009 |
theirs | +0.008 |
Neg logit contributions | |
boy | -0.005 |
girl | -0.005 |
sky | -0.005 |
man | -0.005 |
little | -0.004 |
Token | cage |
Token ID | 17186 |
Feature activation | +1.09e+01 |
Pos logit contributions | |
boy | +0.166 |
girl | +0.150 |
was | +0.144 |
man | +0.144 |
little | +0.141 |
Neg logit contributions | |
yours | -0.388 |
beh | -0.363 |
ours | -0.354 |
possible | -0.339 |
ourselves | -0.333 |
Token | . |
Token ID | 13 |
Feature activation | -9.92e+00 |
Pos logit contributions | |
boy | +0.488 |
sky | +0.466 |
man | +0.459 |
was | +0.457 |
girl | +0.457 |
Neg logit contributions | |
beh | -0.620 |
yours | -0.582 |
possible | -0.508 |
ours | -0.494 |
unnecessary | -0.488 |
Token | The |
Token ID | 383 |
Feature activation | +6.17e+00 |
Pos logit contributions | |
yours | +0.525 |
beh | +0.519 |
ours | +0.454 |
possible | +0.438 |
myself | +0.435 |
Neg logit contributions | |
boy | -0.375 |
man | -0.353 |
girl | -0.349 |
sky | -0.331 |
bird | -0.325 |
Token | bird |
Token ID | 6512 |
Feature activation | +1.78e+01 |
Pos logit contributions | |
boy | +0.142 |
was | +0.136 |
sky | +0.136 |
girl | +0.125 |
man | +0.120 |
Neg logit contributions | |
yours | -0.397 |
beh | -0.387 |
ours | -0.365 |
possible | -0.359 |
eking | -0.346 |
Token | had |
Token ID | 550 |
Feature activation | +5.96e+00 |
Pos logit contributions | |
sky | +0.416 |
boy | +0.407 |
girl | +0.374 |
man | +0.341 |
woman | +0.317 |
Neg logit contributions | |
beh | -1.228 |
yours | -1.144 |
eking | -1.096 |
possible | -1.035 |
ieve | -1.033 |
Token | escaped |
Token ID | 13537 |
Feature activation | -3.20e+00 |
Pos logit contributions | |
boy | +0.249 |
little | +0.234 |
sky | +0.234 |
girl | +0.232 |
man | +0.230 |
Neg logit contributions | |
beh | -0.348 |
yours | -0.330 |
possible | -0.291 |
ours | -0.283 |
ieve | -0.274 |
Token | . |
Token ID | 13 |
Feature activation | -9.78e+00 |
Pos logit contributions | |
beh | +0.212 |
yours | +0.197 |
possible | +0.176 |
ours | +0.172 |
unnecessary | +0.170 |
Neg logit contributions | |
boy | -0.115 |
little | -0.107 |
man | -0.106 |
sky | -0.106 |
girl | -0.106 |
Token | She |
Token ID | 1375 |
Feature activation | +3.32e-01 |
Pos logit contributions | |
yours | +0.560 |
beh | +0.548 |
ours | +0.485 |
possible | +0.478 |
myself | +0.467 |
Neg logit contributions | |
boy | -0.344 |
girl | -0.319 |
man | -0.317 |
sky | -0.299 |
bird | -0.293 |
Token | felt |
Token ID | 2936 |
Feature activation | +6.59e+00 |
Pos logit contributions | |
sky | +0.010 |
boy | +0.009 |
girl | +0.008 |
man | +0.008 |
field | +0.008 |
Neg logit contributions | |
beh | -0.020 |
yours | -0.018 |
properly | -0.017 |
eking | -0.016 |
able | -0.016 |
Token | more |
Token ID | 517 |
Feature activation | -2.53e+00 |
Pos logit contributions | |
boy | +0.067 |
girl | +0.063 |
little | +0.063 |
was | +0.061 |
man | +0.060 |
Neg logit contributions | |
beh | -0.087 |
yours | -0.083 |
ours | -0.071 |
ourselves | -0.069 |
possible | -0.068 |
Token | special |
Token ID | 2041 |
Feature activation | -4.72e+00 |
Pos logit contributions | |
beh | +0.154 |
yours | +0.147 |
possible | +0.128 |
ours | +0.127 |
theirs | +0.123 |
Neg logit contributions | |
boy | -0.101 |
little | -0.094 |
girl | -0.094 |
man | -0.093 |
sky | -0.093 |
Token | . |
Token ID | 13 |
Feature activation | -8.94e+00 |
Pos logit contributions | |
beh | +0.302 |
yours | +0.279 |
ours | +0.251 |
unnecessary | +0.246 |
possible | +0.243 |
Neg logit contributions | |
boy | -0.179 |
was | -0.173 |
girl | -0.166 |
little | -0.165 |
sky | -0.165 |
Token | They |
Token ID | 1119 |
Feature activation | +8.40e-01 |
Pos logit contributions | |
beh | +0.591 |
yours | +0.573 |
possible | +0.511 |
ours | +0.503 |
theirs | +0.492 |
Neg logit contributions | |
boy | -0.296 |
girl | -0.273 |
man | -0.267 |
sky | -0.266 |
little | -0.264 |
Token | thought |
Token ID | 1807 |
Feature activation | +2.85e+00 |
Pos logit contributions | |
boy | +0.030 |
girl | +0.027 |
sky | +0.027 |
man | +0.027 |
little | +0.026 |
Neg logit contributions | |
beh | -0.053 |
yours | -0.051 |
possible | -0.046 |
ours | -0.045 |
theirs | -0.044 |
Token | and |
Token ID | 290 |
Feature activation | +1.34e+01 |
Pos logit contributions | |
boy | +0.109 |
little | +0.103 |
girl | +0.101 |
sky | +0.100 |
man | +0.100 |
Neg logit contributions | |
beh | -0.177 |
yours | -0.166 |
possible | -0.147 |
ours | -0.145 |
unnecessary | -0.140 |
Token | they |
Token ID | 484 |
Feature activation | -5.86e-01 |
Pos logit contributions | |
boy | +0.374 |
girl | +0.337 |
sky | +0.330 |
man | +0.326 |
little | +0.325 |
Neg logit contributions | |
beh | -0.967 |
yours | -0.938 |
possible | -0.848 |
ours | -0.831 |
theirs | -0.813 |
Token | thought |
Token ID | 1807 |
Feature activation | +2.59e+00 |
Pos logit contributions | |
beh | +0.042 |
yours | +0.040 |
possible | +0.037 |
ours | +0.036 |
theirs | +0.035 |
Neg logit contributions | |
boy | -0.016 |
girl | -0.015 |
sky | -0.015 |
man | -0.014 |
little | -0.014 |
Token | and |
Token ID | 290 |
Feature activation | +1.32e+01 |
Pos logit contributions | |
boy | +0.093 |
little | +0.092 |
was | +0.090 |
man | +0.086 |
sky | +0.086 |
Neg logit contributions | |
beh | -0.171 |
yours | -0.155 |
ours | -0.138 |
possible | -0.137 |
unnecessary | -0.133 |
Token | they |
Token ID | 484 |
Feature activation | -8.23e-01 |
Pos logit contributions | |
boy | +0.387 |
girl | +0.350 |
sky | +0.344 |
little | +0.340 |
man | +0.339 |
Neg logit contributions | |
beh | -0.937 |
yours | -0.907 |
possible | -0.818 |
ours | -0.803 |
theirs | -0.783 |
Token | finally |
Token ID | 3443 |
Feature activation | -1.90e+00 |
Pos logit contributions | |
beh | +0.059 |
yours | +0.057 |
possible | +0.052 |
ours | +0.051 |
theirs | +0.050 |
Neg logit contributions | |
boy | -0.023 |
girl | -0.021 |
sky | -0.020 |
man | -0.020 |
little | -0.020 |
Token | Hello |
Token ID | 15496 |
Feature activation | -1.92e+00 |
Pos logit contributions | |
boy | +0.241 |
girl | +0.222 |
little | +0.221 |
man | +0.221 |
sky | +0.220 |
Neg logit contributions | |
beh | -0.386 |
yours | -0.370 |
possible | -0.329 |
ours | -0.316 |
theirs | -0.308 |
Token | sheep |
Token ID | 15900 |
Feature activation | +1.31e+01 |
Pos logit contributions | |
yours | +0.127 |
beh | +0.125 |
ours | +0.112 |
possible | +0.111 |
theirs | +0.107 |
Neg logit contributions | |
boy | -0.056 |
sky | -0.054 |
man | -0.051 |
girl | -0.050 |
was | -0.047 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.24e+01 |
Pos logit contributions | |
boy | +0.598 |
girl | +0.565 |
man | +0.561 |
little | +0.557 |
was | +0.549 |
Neg logit contributions | |
beh | -0.726 |
yours | -0.672 |
possible | -0.605 |
unnecessary | -0.593 |
ours | -0.587 |
Token | and |
Token ID | 290 |
Feature activation | +1.25e+01 |
Pos logit contributions | |
beh | +0.842 |
yours | +0.819 |
possible | +0.720 |
theirs | +0.718 |
ours | +0.711 |
Neg logit contributions | |
boy | -0.332 |
girl | -0.298 |
little | -0.281 |
man | -0.280 |
sky | -0.280 |
Token | the |
Token ID | 262 |
Feature activation | +5.47e+00 |
Pos logit contributions | |
sky | +0.475 |
boy | +0.461 |
man | +0.425 |
girl | +0.424 |
little | +0.410 |
Neg logit contributions | |
beh | -0.703 |
yours | -0.667 |
theirs | -0.583 |
possible | -0.561 |
ours | -0.557 |
Token | sheep |
Token ID | 15900 |
Feature activation | +1.30e+01 |
Pos logit contributions | |
boy | +0.148 |
was | +0.135 |
sky | +0.132 |
man | +0.119 |
girl | +0.112 |
Neg logit contributions | |
yours | -0.329 |
beh | -0.329 |
possible | -0.296 |
ours | -0.296 |
unnecessary | -0.290 |
Token | said |
Token ID | 531 |
Feature activation | +4.97e-01 |
Pos logit contributions | |
boy | +0.405 |
girl | +0.369 |
sky | +0.360 |
man | +0.353 |
little | +0.349 |
Neg logit contributions | |
beh | -0.872 |
yours | -0.831 |
possible | -0.741 |
ours | -0.725 |
theirs | -0.722 |
Token | " |
Token ID | 366 |
Feature activation | +6.10e+00 |
Pos logit contributions | |
boy | +0.010 |
girl | +0.008 |
sky | +0.008 |
man | +0.008 |
little | +0.008 |
Neg logit contributions | |
beh | -0.039 |
yours | -0.039 |
possible | -0.035 |
ours | -0.034 |
theirs | -0.034 |
Token | B |
Token ID | 33 |
Feature activation | +5.16e+00 |
Pos logit contributions | |
boy | +0.240 |
girl | +0.225 |
sky | +0.221 |
man | +0.217 |
little | +0.216 |
Neg logit contributions | |
beh | -0.367 |
yours | -0.358 |
possible | -0.316 |
ours | -0.315 |
theirs | -0.305 |
Token | aa |
Token ID | 7252 |
Feature activation | -8.16e-01 |
Pos logit contributions | |
boy | +0.208 |
girl | +0.198 |
little | +0.192 |
sky | +0.186 |
was | +0.186 |
Neg logit contributions | |
beh | -0.294 |
yours | -0.290 |
possible | -0.263 |
ours | -0.254 |
theirs | -0.246 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.25e+01 |
Pos logit contributions | |
beh | +0.044 |
yours | +0.042 |
possible | +0.037 |
ours | +0.036 |
theirs | +0.035 |
Neg logit contributions | |
boy | -0.037 |
girl | -0.035 |
sky | -0.035 |
man | -0.035 |
little | -0.035 |
Token | So |
Token ID | 1406 |
Feature activation | +2.00e+00 |
Pos logit contributions | |
beh | +0.751 |
yours | +0.740 |
ours | +0.655 |
possible | +0.653 |
theirs | +0.646 |
Neg logit contributions | |
girl | -0.239 |
boy | -0.235 |
little | -0.212 |
sky | -0.207 |
man | -0.203 |
Token | , |
Token ID | 11 |
Feature activation | +9.54e+00 |
Pos logit contributions | |
boy | +0.050 |
girl | +0.045 |
sky | +0.044 |
little | +0.043 |
man | +0.043 |
Neg logit contributions | |
beh | -0.148 |
yours | -0.145 |
possible | -0.132 |
ours | -0.129 |
theirs | -0.126 |
Token | he |
Token ID | 339 |
Feature activation | -6.05e+00 |
Pos logit contributions | |
boy | +0.368 |
girl | +0.338 |
man | +0.335 |
sky | +0.334 |
little | +0.333 |
Neg logit contributions | |
beh | -0.594 |
yours | -0.571 |
possible | -0.506 |
ours | -0.495 |
theirs | -0.480 |
Token | let |
Token ID | 1309 |
Feature activation | -9.80e-01 |
Pos logit contributions | |
beh | +0.376 |
yours | +0.359 |
possible | +0.312 |
ours | +0.305 |
properly | +0.305 |
Neg logit contributions | |
boy | -0.199 |
sky | -0.194 |
girl | -0.189 |
little | -0.176 |
man | -0.174 |
Token | go |
Token ID | 467 |
Feature activation | -3.14e+00 |
Pos logit contributions | |
beh | +0.058 |
yours | +0.055 |
possible | +0.049 |
ours | +0.048 |
theirs | +0.046 |
Neg logit contributions | |
boy | -0.041 |
girl | -0.038 |
sky | -0.037 |
man | -0.037 |
little | -0.037 |
Token | of |
Token ID | 286 |
Feature activation | +4.09e+00 |
Pos logit contributions | |
beh | +0.229 |
yours | +0.218 |
possible | +0.199 |
ours | +0.195 |
unnecessary | +0.192 |
Neg logit contributions | |
boy | -0.089 |
girl | -0.081 |
little | -0.079 |
sky | -0.079 |
man | -0.078 |
Token | the |
Token ID | 262 |
Feature activation | +5.84e+00 |
Pos logit contributions | |
boy | +0.119 |
little | +0.108 |
man | +0.107 |
girl | +0.103 |
sky | +0.102 |
Neg logit contributions | |
beh | -0.307 |
yours | -0.280 |
possible | -0.257 |
unnecessary | -0.249 |
ieve | -0.247 |
Token | balloon |
Token ID | 21190 |
Feature activation | +1.07e+01 |
Pos logit contributions | |
boy | +0.165 |
girl | +0.154 |
was | +0.153 |
man | +0.143 |
child | +0.142 |
Neg logit contributions | |
yours | -0.384 |
beh | -0.354 |
ours | -0.349 |
possible | -0.335 |
ourselves | -0.319 |
Token | and |
Token ID | 290 |
Feature activation | +1.26e+01 |
Pos logit contributions | |
boy | +0.324 |
girl | +0.292 |
man | +0.290 |
sky | +0.285 |
little | +0.284 |
Neg logit contributions | |
beh | -0.763 |
yours | -0.727 |
possible | -0.651 |
ours | -0.642 |
unnecessary | -0.637 |
Token | slowly |
Token ID | 6364 |
Feature activation | -1.99e+00 |
Pos logit contributions | |
boy | +0.432 |
girl | +0.396 |
man | +0.387 |
sky | +0.387 |
little | +0.386 |
Neg logit contributions | |
beh | -0.826 |
yours | -0.796 |
possible | -0.714 |
ours | -0.697 |
theirs | -0.680 |
Token | floated |
Token ID | 29742 |
Feature activation | +1.54e+00 |
Pos logit contributions | |
beh | +0.133 |
yours | +0.128 |
possible | +0.115 |
ours | +0.113 |
unnecessary | +0.109 |
Neg logit contributions | |
boy | -0.067 |
sky | -0.061 |
girl | -0.061 |
man | -0.061 |
little | -0.061 |
Token | for |
Token ID | 329 |
Feature activation | -7.59e+00 |
Pos logit contributions | |
boy | +0.177 |
girl | +0.159 |
man | +0.158 |
sky | +0.158 |
was | +0.158 |
Neg logit contributions | |
beh | -0.526 |
yours | -0.498 |
possible | -0.452 |
unnecessary | -0.451 |
ours | -0.447 |
Token | her |
Token ID | 607 |
Feature activation | -3.17e+00 |
Pos logit contributions | |
beh | +0.461 |
yours | +0.418 |
possible | +0.381 |
rying | +0.366 |
ours | +0.365 |
Neg logit contributions | |
little | -0.308 |
boy | -0.307 |
girl | -0.291 |
man | -0.289 |
was | -0.283 |
Token | . |
Token ID | 13 |
Feature activation | -7.94e+00 |
Pos logit contributions | |
beh | +0.195 |
yours | +0.185 |
possible | +0.164 |
ours | +0.160 |
theirs | +0.156 |
Neg logit contributions | |
boy | -0.126 |
girl | -0.118 |
little | -0.118 |
man | -0.115 |
sky | -0.114 |
Token | They |
Token ID | 1119 |
Feature activation | +2.03e+00 |
Pos logit contributions | |
beh | +0.494 |
yours | +0.493 |
ours | +0.424 |
possible | +0.423 |
theirs | +0.419 |
Neg logit contributions | |
boy | -0.270 |
man | -0.250 |
girl | -0.247 |
sky | -0.239 |
little | -0.236 |
Token | smelled |
Token ID | 37609 |
Feature activation | +7.50e+00 |
Pos logit contributions | |
boy | +0.067 |
sky | +0.062 |
girl | +0.062 |
man | +0.060 |
little | +0.058 |
Neg logit contributions | |
beh | -0.129 |
yours | -0.125 |
possible | -0.110 |
ours | -0.109 |
theirs | -0.109 |
Token | so |
Token ID | 523 |
Feature activation | +8.01e-01 |
Pos logit contributions | |
boy | +0.284 |
little | +0.274 |
girl | +0.265 |
was | +0.262 |
sky | +0.262 |
Neg logit contributions | |
beh | -0.486 |
yours | -0.448 |
possible | -0.404 |
ours | -0.390 |
rying | -0.383 |
Token | good |
Token ID | 922 |
Feature activation | +3.70e+00 |
Pos logit contributions | |
little | +0.029 |
boy | +0.028 |
was | +0.027 |
girl | +0.026 |
man | +0.026 |
Neg logit contributions | |
beh | -0.053 |
yours | -0.050 |
possible | -0.045 |
ours | -0.043 |
rying | -0.043 |
Token | , |
Token ID | 11 |
Feature activation | +1.14e+01 |
Pos logit contributions | |
boy | +0.096 |
man | +0.090 |
was | +0.089 |
girl | +0.086 |
and | +0.086 |
Neg logit contributions | |
beh | -0.286 |
yours | -0.271 |
ours | -0.246 |
possible | -0.245 |
unnecessary | -0.244 |
Token | and |
Token ID | 290 |
Feature activation | +1.42e+01 |
Pos logit contributions | |
boy | +0.378 |
girl | +0.347 |
sky | +0.343 |
man | +0.338 |
little | +0.336 |
Neg logit contributions | |
beh | -0.752 |
yours | -0.729 |
possible | -0.652 |
ours | -0.640 |
theirs | -0.626 |
Token | Lily |
Token ID | 20037 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
boy | +0.512 |
sky | +0.481 |
girl | +0.472 |
man | +0.466 |
little | +0.460 |
Neg logit contributions | |
beh | -0.886 |
yours | -0.857 |
possible | -0.761 |
ours | -0.744 |
theirs | -0.730 |
Token | couldn |
Token ID | 3521 |
Feature activation | -8.12e-01 |
Pos logit contributions | |
beh | +0.082 |
yours | +0.079 |
supposed | +0.071 |
properly | +0.069 |
possible | +0.069 |
Neg logit contributions | |
boy | -0.035 |
sky | -0.034 |
girl | -0.032 |
man | -0.030 |
child | -0.028 |
Token | , |
Token ID | 11 |
Feature activation | +1.00e+01 |
Pos logit contributions | |
beh | +0.501 |
yours | +0.482 |
ours | +0.430 |
possible | +0.428 |
theirs | +0.418 |
Neg logit contributions | |
boy | -0.173 |
girl | -0.155 |
little | -0.153 |
sky | -0.152 |
man | -0.151 |
Token | but |
Token ID | 475 |
Feature activation | +1.36e+00 |
Pos logit contributions | |
boy | +0.325 |
girl | +0.297 |
sky | +0.294 |
man | +0.289 |
little | +0.285 |
Neg logit contributions | |
beh | -0.665 |
yours | -0.648 |
possible | -0.583 |
ours | -0.571 |
theirs | -0.562 |
Token | they |
Token ID | 484 |
Feature activation | -1.02e+00 |
Pos logit contributions | |
boy | +0.051 |
sky | +0.048 |
girl | +0.048 |
man | +0.047 |
little | +0.046 |
Neg logit contributions | |
beh | -0.083 |
yours | -0.080 |
possible | -0.072 |
ours | -0.069 |
theirs | -0.069 |
Token | had |
Token ID | 550 |
Feature activation | +5.83e+00 |
Pos logit contributions | |
beh | +0.070 |
yours | +0.068 |
possible | +0.061 |
ours | +0.059 |
theirs | +0.059 |
Neg logit contributions | |
boy | -0.031 |
girl | -0.028 |
sky | -0.028 |
man | -0.027 |
little | -0.027 |
Token | a |
Token ID | 257 |
Feature activation | -2.70e+00 |
Pos logit contributions | |
boy | +0.236 |
girl | +0.219 |
sky | +0.218 |
little | +0.217 |
man | +0.215 |
Neg logit contributions | |
beh | -0.348 |
yours | -0.333 |
possible | -0.293 |
ours | -0.288 |
theirs | -0.278 |
Token | lot |
Token ID | 1256 |
Feature activation | +8.89e+00 |
Pos logit contributions | |
yours | +0.139 |
beh | +0.136 |
possible | +0.129 |
unnecessary | +0.121 |
ours | +0.117 |
Neg logit contributions | |
boy | -0.108 |
sky | -0.107 |
girl | -0.100 |
man | -0.100 |
was | -0.096 |
Token | of |
Token ID | 286 |
Feature activation | +4.08e+00 |
Pos logit contributions | |
boy | +0.203 |
was | +0.201 |
little | +0.199 |
sky | +0.195 |
man | +0.184 |
Neg logit contributions | |
beh | -0.671 |
yours | -0.652 |
ours | -0.596 |
eking | -0.573 |
possible | -0.568 |
Token | fun |
Token ID | 1257 |
Feature activation | -7.30e+00 |
Pos logit contributions | |
boy | +0.172 |
little | +0.163 |
girl | +0.160 |
man | +0.159 |
was | +0.158 |
Neg logit contributions | |
beh | -0.239 |
yours | -0.228 |
possible | -0.200 |
ours | -0.198 |
rying | -0.190 |
Token | and |
Token ID | 290 |
Feature activation | +1.28e+01 |
Pos logit contributions | |
beh | +0.445 |
yours | +0.435 |
ours | +0.392 |
possible | +0.372 |
theirs | +0.366 |
Neg logit contributions | |
boy | -0.287 |
was | -0.284 |
little | -0.273 |
girl | -0.266 |
man | -0.263 |
Token | were |
Token ID | 547 |
Feature activation | +5.33e+00 |
Pos logit contributions | |
boy | +0.446 |
was | +0.412 |
girl | +0.409 |
little | +0.406 |
man | +0.399 |
Neg logit contributions | |
beh | -0.825 |
yours | -0.802 |
possible | -0.705 |
ours | -0.703 |
ieve | -0.677 |
Token | very |
Token ID | 845 |
Feature activation | +4.00e+00 |
Pos logit contributions | |
boy | +0.204 |
little | +0.197 |
was | +0.194 |
girl | +0.191 |
sky | +0.187 |
Neg logit contributions | |
beh | -0.325 |
yours | -0.308 |
ours | -0.271 |
unnecessary | -0.268 |
possible | -0.266 |
Token | running |
Token ID | 2491 |
Feature activation | -1.64e+00 |
Pos logit contributions | |
boy | +0.478 |
girl | +0.443 |
sky | +0.432 |
man | +0.425 |
little | +0.415 |
Neg logit contributions | |
beh | -0.844 |
yours | -0.833 |
possible | -0.745 |
ours | -0.717 |
theirs | -0.703 |
Token | through |
Token ID | 832 |
Feature activation | -4.66e+00 |
Pos logit contributions | |
beh | +0.083 |
yours | +0.080 |
possible | +0.068 |
ours | +0.066 |
theirs | +0.064 |
Neg logit contributions | |
boy | -0.081 |
girl | -0.076 |
sky | -0.076 |
man | -0.075 |
little | -0.075 |
Token | the |
Token ID | 262 |
Feature activation | +6.49e+00 |
Pos logit contributions | |
beh | +0.302 |
yours | +0.295 |
possible | +0.264 |
ours | +0.257 |
theirs | +0.252 |
Neg logit contributions | |
boy | -0.157 |
girl | -0.145 |
sky | -0.141 |
man | -0.141 |
little | -0.138 |
Token | grass |
Token ID | 8701 |
Feature activation | +9.88e+00 |
Pos logit contributions | |
boy | +0.159 |
girl | +0.139 |
was | +0.137 |
man | +0.137 |
little | +0.132 |
Neg logit contributions | |
beh | -0.450 |
yours | -0.441 |
possible | -0.392 |
unnecessary | -0.386 |
ours | -0.384 |
Token | , |
Token ID | 11 |
Feature activation | +1.06e+01 |
Pos logit contributions | |
boy | +0.397 |
sky | +0.371 |
girl | +0.369 |
man | +0.365 |
little | +0.364 |
Neg logit contributions | |
beh | -0.592 |
yours | -0.568 |
possible | -0.498 |
ours | -0.494 |
unnecessary | -0.475 |
Token | smiling |
Token ID | 16755 |
Feature activation | +2.61e+00 |
Pos logit contributions | |
boy | +0.419 |
girl | +0.392 |
sky | +0.375 |
man | +0.372 |
little | +0.367 |
Neg logit contributions | |
yours | -0.603 |
beh | -0.595 |
possible | -0.527 |
ours | -0.510 |
theirs | -0.500 |
Token | from |
Token ID | 422 |
Feature activation | +5.65e+00 |
Pos logit contributions | |
boy | +0.074 |
girl | +0.067 |
sky | +0.066 |
man | +0.065 |
little | +0.064 |
Neg logit contributions | |
beh | -0.185 |
yours | -0.179 |
possible | -0.162 |
ours | -0.158 |
theirs | -0.155 |
Token | ear |
Token ID | 1027 |
Feature activation | -4.25e+00 |
Pos logit contributions | |
boy | +0.211 |
little | +0.202 |
girl | +0.195 |
sky | +0.188 |
man | +0.188 |
Neg logit contributions | |
beh | -0.367 |
yours | -0.338 |
possible | -0.311 |
ours | -0.298 |
eking | -0.293 |
Token | to |
Token ID | 284 |
Feature activation | +2.15e+00 |
Pos logit contributions | |
beh | +0.265 |
yours | +0.254 |
possible | +0.224 |
ours | +0.224 |
unnecessary | +0.216 |
Neg logit contributions | |
boy | -0.162 |
little | -0.150 |
girl | -0.150 |
sky | -0.149 |
man | -0.149 |
Token | ear |
Token ID | 1027 |
Feature activation | -4.21e+00 |
Pos logit contributions | |
boy | +0.085 |
little | +0.082 |
was | +0.081 |
girl | +0.079 |
sky | +0.078 |
Neg logit contributions | |
beh | -0.133 |
yours | -0.120 |
ours | -0.110 |
possible | -0.110 |
eking | -0.105 |
Token | . |
Token ID | 13 |
Feature activation | -9.02e+00 |
Pos logit contributions | |
beh | +0.258 |
yours | +0.245 |
ours | +0.217 |
possible | +0.216 |
unnecessary | +0.210 |
Neg logit contributions | |
boy | -0.166 |
little | -0.158 |
was | -0.157 |
man | -0.156 |
girl | -0.153 |
Token | She |
Token ID | 1375 |
Feature activation | -1.78e-01 |
Pos logit contributions | |
beh | +0.621 |
yours | +0.612 |
possible | +0.534 |
ours | +0.530 |
theirs | +0.514 |
Neg logit contributions | |
boy | -0.365 |
girl | -0.339 |
man | -0.335 |
little | -0.328 |
sky | -0.327 |
Token | liked |
Token ID | 8288 |
Feature activation | +2.81e+00 |
Pos logit contributions | |
beh | +0.011 |
yours | +0.011 |
ours | +0.009 |
eking | +0.009 |
properly | +0.009 |
Neg logit contributions | |
sky | -0.005 |
boy | -0.005 |
girl | -0.005 |
man | -0.005 |
little | -0.005 |
Token | her |
Token ID | 607 |
Feature activation | -5.55e+00 |
Pos logit contributions | |
boy | +0.113 |
girl | +0.105 |
little | +0.104 |
sky | +0.104 |
man | +0.104 |
Neg logit contributions | |
beh | -0.170 |
yours | -0.160 |
possible | -0.143 |
ours | -0.139 |
ieve | -0.135 |
Token | dress |
Token ID | 6576 |
Feature activation | +3.84e+00 |
Pos logit contributions | |
beh | +0.379 |
yours | +0.378 |
possible | +0.338 |
ours | +0.331 |
theirs | +0.327 |
Neg logit contributions | |
boy | -0.157 |
sky | -0.142 |
girl | -0.141 |
man | -0.136 |
little | -0.135 |
Token | . |
Token ID | 13 |
Feature activation | -1.01e+01 |
Pos logit contributions | |
boy | +0.137 |
girl | +0.126 |
man | +0.125 |
was | +0.124 |
little | +0.123 |
Neg logit contributions | |
beh | -0.256 |
yours | -0.243 |
possible | -0.218 |
ours | -0.215 |
unnecessary | -0.208 |
Token | She |
Token ID | 1375 |
Feature activation | -1.02e-01 |
Pos logit contributions | |
beh | +0.653 |
yours | +0.644 |
possible | +0.568 |
ours | +0.560 |
theirs | +0.548 |
Neg logit contributions | |
boy | -0.341 |
girl | -0.315 |
man | -0.308 |
sky | -0.305 |
little | -0.304 |
Token | thought |
Token ID | 1807 |
Feature activation | +1.90e+00 |
Pos logit contributions | |
beh | +0.006 |
yours | +0.006 |
ours | +0.005 |
theirs | +0.005 |
properly | +0.005 |
Neg logit contributions | |
sky | -0.003 |
boy | -0.003 |
girl | -0.003 |
man | -0.003 |
little | -0.003 |
Token | it |
Token ID | 340 |
Feature activation | -2.35e+00 |
Pos logit contributions | |
boy | +0.075 |
girl | +0.070 |
sky | +0.069 |
little | +0.069 |
man | +0.069 |
Neg logit contributions | |
beh | -0.116 |
yours | -0.109 |
possible | -0.097 |
ours | -0.095 |
unnecessary | -0.092 |
Token | was |
Token ID | 373 |
Feature activation | +8.03e+00 |
Pos logit contributions | |
beh | +0.206 |
yours | +0.200 |
possible | +0.184 |
ours | +0.179 |
ieve | +0.179 |
Neg logit contributions | |
boy | -0.023 |
girl | -0.018 |
sky | -0.018 |
little | -0.016 |
man | -0.016 |
Token | beautiful |
Token ID | 4950 |
Feature activation | -6.49e-01 |
Pos logit contributions | |
little | +0.267 |
boy | +0.265 |
was | +0.260 |
girl | +0.237 |
man | +0.233 |
Neg logit contributions | |
beh | -0.565 |
rying | -0.480 |
yours | -0.466 |
owing | -0.455 |
eking | -0.453 |
Token | . |
Token ID | 13 |
Feature activation | -1.00e+01 |
Pos logit contributions | |
beh | +0.043 |
yours | +0.040 |
ours | +0.036 |
possible | +0.035 |
unnecessary | +0.035 |
Neg logit contributions | |
boy | -0.025 |
girl | -0.023 |
was | -0.023 |
man | -0.023 |
little | -0.022 |
Token | have |
Token ID | 423 |
Feature activation | -5.05e+00 |
Pos logit contributions | |
was | +0.047 |
little | +0.046 |
boy | +0.045 |
sky | +0.043 |
and | +0.042 |
Neg logit contributions | |
beh | -0.082 |
yours | -0.074 |
unnecessary | -0.070 |
ours | -0.069 |
possible | -0.067 |
Token | some |
Token ID | 617 |
Feature activation | -8.31e+00 |
Pos logit contributions | |
beh | +0.331 |
yours | +0.305 |
possible | +0.277 |
ours | +0.272 |
rying | +0.269 |
Neg logit contributions | |
little | -0.178 |
boy | -0.177 |
was | -0.171 |
girl | -0.168 |
sky | -0.168 |
Token | more |
Token ID | 517 |
Feature activation | -3.43e+00 |
Pos logit contributions | |
beh | +0.545 |
yours | +0.517 |
possible | +0.468 |
ours | +0.456 |
theirs | +0.443 |
Neg logit contributions | |
boy | -0.294 |
little | -0.282 |
girl | -0.280 |
sky | -0.270 |
man | -0.267 |
Token | h |
Token ID | 289 |
Feature activation | +1.02e+01 |
Pos logit contributions | |
beh | +0.226 |
yours | +0.214 |
possible | +0.193 |
ours | +0.189 |
unnecessary | +0.183 |
Neg logit contributions | |
boy | -0.121 |
girl | -0.115 |
little | -0.114 |
sky | -0.110 |
man | -0.110 |
Token | angers |
Token ID | 6606 |
Feature activation | -5.27e+00 |
Pos logit contributions | |
sky | +0.416 |
little | +0.412 |
man | +0.401 |
boy | +0.393 |
girl | +0.391 |
Neg logit contributions | |
yours | -0.528 |
beh | -0.517 |
ours | -0.471 |
possible | -0.466 |
unnecessary | -0.444 |
Token | . |
Token ID | 13 |
Feature activation | -9.55e+00 |
Pos logit contributions | |
beh | +0.375 |
yours | +0.337 |
unnecessary | +0.318 |
ours | +0.305 |
eking | +0.298 |
Neg logit contributions | |
and | -0.161 |
was | -0.160 |
boy | -0.159 |
man | -0.155 |
little | -0.154 |
Token |
|
Token ID | 198 |
Feature activation | +6.88e+00 |
Pos logit contributions | |
beh | +0.541 |
yours | +0.527 |
theirs | +0.449 |
ours | +0.439 |
possible | +0.435 |
Neg logit contributions | |
boy | -0.358 |
girl | -0.336 |
sky | -0.328 |
man | -0.327 |
little | -0.320 |
Token |
|
Token ID | 198 |
Feature activation | +6.91e+00 |
Pos logit contributions | |
boy | +0.154 |
girl | +0.140 |
sky | +0.139 |
little | +0.121 |
man | +0.121 |
Neg logit contributions | |
beh | -0.507 |
yours | -0.499 |
possible | -0.449 |
theirs | -0.436 |
ours | -0.433 |
Token | " |
Token ID | 1 |
Feature activation | +8.61e+00 |
Pos logit contributions | |
boy | +0.273 |
sky | +0.254 |
girl | +0.250 |
man | +0.243 |
little | +0.237 |
Neg logit contributions | |
beh | -0.417 |
yours | -0.399 |
possible | -0.347 |
ours | -0.342 |
theirs | -0.331 |
Token | Mom |
Token ID | 29252 |
Feature activation | +1.41e+01 |
Pos logit contributions | |
boy | +0.320 |
sky | +0.296 |
little | +0.292 |
man | +0.291 |
girl | +0.290 |
Neg logit contributions | |
beh | -0.527 |
yours | -0.508 |
possible | -0.453 |
ours | -0.433 |
theirs | -0.420 |
Token | , |
Token ID | 11 |
Feature activation | +1.04e+01 |
Pos logit contributions | |
boy | +0.344 |
girl | +0.306 |
sky | +0.301 |
man | +0.295 |
little | +0.294 |
Neg logit contributions | |
beh | -1.057 |
yours | -1.028 |
possible | -0.932 |
ours | -0.916 |
theirs | -0.896 |
Token | 'm |
Token ID | 1101 |
Feature activation | -6.95e-01 |
Pos logit contributions | |
beh | +0.001 |
yours | +0.001 |
possible | +0.001 |
ours | +0.001 |
theirs | +0.001 |
Neg logit contributions | |
boy | -0.001 |
sky | -0.001 |
girl | -0.000 |
man | -0.000 |
little | -0.000 |
Token | glad |
Token ID | 9675 |
Feature activation | -3.98e+00 |
Pos logit contributions | |
beh | +0.038 |
yours | +0.033 |
unnecessary | +0.030 |
possible | +0.029 |
ours | +0.028 |
Neg logit contributions | |
little | -0.033 |
boy | -0.032 |
was | -0.031 |
sky | -0.031 |
girl | -0.030 |
Token | I |
Token ID | 314 |
Feature activation | -1.11e+01 |
Pos logit contributions | |
beh | +0.209 |
yours | +0.179 |
theirs | +0.161 |
ours | +0.158 |
possible | +0.156 |
Neg logit contributions | |
boy | -0.208 |
little | -0.205 |
girl | -0.199 |
was | -0.195 |
man | -0.193 |
Token | could |
Token ID | 714 |
Feature activation | +1.68e+00 |
Pos logit contributions | |
beh | +0.821 |
yours | +0.797 |
possible | +0.722 |
ours | +0.712 |
theirs | +0.695 |
Neg logit contributions | |
boy | -0.290 |
girl | -0.258 |
sky | -0.253 |
little | -0.252 |
man | -0.251 |
Token | help |
Token ID | 1037 |
Feature activation | -5.34e+00 |
Pos logit contributions | |
was | +0.072 |
little | +0.071 |
and | +0.067 |
boy | +0.067 |
man | +0.066 |
Neg logit contributions | |
beh | -0.103 |
yours | -0.093 |
ours | -0.086 |
theirs | -0.085 |
unnecessary | -0.081 |
Token | . |
Token ID | 13 |
Feature activation | -9.22e+00 |
Pos logit contributions | |
beh | +0.324 |
yours | +0.269 |
theirs | +0.267 |
ours | +0.249 |
possible | +0.235 |
Neg logit contributions | |
little | -0.259 |
and | -0.240 |
boy | -0.240 |
was | -0.237 |
man | -0.232 |
Token | Have |
Token ID | 8192 |
Feature activation | -8.76e+00 |
Pos logit contributions | |
beh | +0.480 |
yours | +0.466 |
possible | +0.397 |
ours | +0.385 |
theirs | +0.367 |
Neg logit contributions | |
boy | -0.433 |
girl | -0.407 |
man | -0.405 |
sky | -0.402 |
little | -0.398 |
Token | a |
Token ID | 257 |
Feature activation | -2.30e+00 |
Pos logit contributions | |
beh | +0.584 |
yours | +0.525 |
rying | +0.486 |
ours | +0.478 |
eking | +0.477 |
Neg logit contributions | |
boy | -0.312 |
little | -0.312 |
was | -0.305 |
girl | -0.288 |
sky | -0.285 |
Token | good |
Token ID | 922 |
Feature activation | +2.68e+00 |
Pos logit contributions | |
beh | +0.129 |
yours | +0.125 |
possible | +0.111 |
ours | +0.106 |
unnecessary | +0.103 |
Neg logit contributions | |
boy | -0.097 |
sky | -0.092 |
girl | -0.091 |
man | -0.089 |
was | -0.087 |
Token | sleep |
Token ID | 3993 |
Feature activation | -6.46e+00 |
Pos logit contributions | |
boy | +0.096 |
sky | +0.090 |
girl | +0.089 |
man | +0.087 |
little | +0.086 |
Neg logit contributions | |
beh | -0.168 |
yours | -0.163 |
possible | -0.146 |
ours | -0.140 |
theirs | -0.138 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.13e+01 |
Pos logit contributions | |
beh | +0.353 |
yours | +0.316 |
ours | +0.278 |
possible | +0.276 |
unnecessary | +0.275 |
Neg logit contributions | |
boy | -0.312 |
girl | -0.297 |
man | -0.295 |
little | -0.294 |
was | -0.291 |
Token |
|
Token ID | 198 |
Feature activation | +6.38e+00 |
Pos logit contributions | |
yours | +0.590 |
beh | +0.585 |
ours | +0.521 |
theirs | +0.517 |
possible | +0.514 |
Neg logit contributions | |
boy | -0.341 |
girl | -0.326 |
man | -0.324 |
little | -0.294 |
bird | -0.292 |
Token |
|
Token ID | 198 |
Feature activation | +6.42e+00 |
Pos logit contributions | |
boy | +0.178 |
girl | +0.170 |
sky | +0.150 |
man | +0.149 |
dog | +0.144 |
Neg logit contributions | |
yours | -0.417 |
beh | -0.411 |
possible | -0.384 |
ourselves | -0.364 |
ours | -0.362 |
Token | " |
Token ID | 1 |
Feature activation | +8.11e+00 |
Pos logit contributions | |
boy | +0.217 |
sky | +0.206 |
girl | +0.197 |
man | +0.193 |
dog | +0.189 |
Neg logit contributions | |
beh | -0.399 |
yours | -0.378 |
possible | -0.350 |
ourselves | -0.331 |
ours | -0.331 |
Token | Look |
Token ID | 8567 |
Feature activation | +6.03e+00 |
Pos logit contributions | |
boy | +0.293 |
man | +0.278 |
little | +0.270 |
girl | +0.265 |
dog | +0.258 |
Neg logit contributions | |
beh | -0.501 |
yours | -0.481 |
possible | -0.444 |
ours | -0.409 |
ourselves | -0.402 |
Token | at |
Token ID | 379 |
Feature activation | -6.33e+00 |
Pos logit contributions | |
boy | +0.231 |
little | +0.225 |
girl | +0.210 |
was | +0.205 |
man | +0.204 |
Neg logit contributions | |
beh | -0.394 |
yours | -0.353 |
possible | -0.332 |
unnecessary | -0.312 |
theirs | -0.309 |
Token | all |
Token ID | 477 |
Feature activation | -7.11e+00 |
Pos logit contributions | |
beh | +0.379 |
yours | +0.348 |
possible | +0.312 |
ours | +0.299 |
theirs | +0.297 |
Neg logit contributions | |
boy | -0.263 |
little | -0.250 |
sky | -0.245 |
girl | -0.244 |
man | -0.242 |
Token | the |
Token ID | 262 |
Feature activation | +5.81e+00 |
Pos logit contributions | |
beh | +0.441 |
yours | +0.439 |
possible | +0.388 |
ours | +0.385 |
theirs | +0.367 |
Neg logit contributions | |
boy | -0.258 |
girl | -0.237 |
sky | -0.233 |
man | -0.232 |
little | -0.223 |
Token | trees |
Token ID | 7150 |
Feature activation | +9.40e+00 |
Pos logit contributions | |
boy | +0.199 |
girl | +0.185 |
man | +0.180 |
sky | +0.175 |
was | +0.173 |
Neg logit contributions | |
beh | -0.359 |
yours | -0.359 |
possible | -0.325 |
ours | -0.316 |
unnecessary | -0.301 |
Token | Dad |
Token ID | 17415 |
Feature activation | -2.83e-01 |
Pos logit contributions | |
boy | +0.371 |
sky | +0.348 |
girl | +0.346 |
little | +0.345 |
man | +0.341 |
Neg logit contributions | |
beh | -0.573 |
yours | -0.539 |
possible | -0.474 |
unnecessary | -0.461 |
theirs | -0.460 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.19e+01 |
Pos logit contributions | |
beh | +0.014 |
yours | +0.014 |
possible | +0.012 |
ours | +0.012 |
theirs | +0.011 |
Neg logit contributions | |
boy | -0.014 |
girl | -0.013 |
sky | -0.013 |
man | -0.013 |
little | -0.013 |
Token | said |
Token ID | 531 |
Feature activation | +1.20e+00 |
Pos logit contributions | |
beh | +0.777 |
yours | +0.744 |
possible | +0.675 |
ours | +0.657 |
theirs | +0.657 |
Neg logit contributions | |
boy | -0.358 |
sky | -0.337 |
girl | -0.334 |
man | -0.308 |
little | -0.305 |
Token | , |
Token ID | 11 |
Feature activation | +8.95e+00 |
Pos logit contributions | |
sky | +0.135 |
boy | +0.126 |
man | +0.119 |
girl | +0.116 |
field | +0.107 |
Neg logit contributions | |
yours | -0.296 |
beh | -0.295 |
possible | -0.255 |
theirs | -0.250 |
ours | -0.248 |
Token | a |
Token ID | 257 |
Feature activation | -3.70e+00 |
Pos logit contributions | |
boy | +0.342 |
girl | +0.316 |
little | +0.310 |
man | +0.309 |
sky | +0.306 |
Neg logit contributions | |
beh | -0.558 |
yours | -0.535 |
possible | -0.477 |
ours | -0.467 |
theirs | -0.451 |
Token | butterfly |
Token ID | 35113 |
Feature activation | +1.10e+01 |
Pos logit contributions | |
yours | +0.244 |
beh | +0.226 |
possible | +0.224 |
unnecessary | +0.216 |
ours | +0.215 |
Neg logit contributions | |
sky | -0.071 |
was | -0.065 |
At | -0.060 |
field | -0.045 |
and | -0.045 |
Token | named |
Token ID | 3706 |
Feature activation | +3.56e+00 |
Pos logit contributions | |
boy | +0.373 |
girl | +0.338 |
sky | +0.337 |
man | +0.331 |
little | +0.321 |
Neg logit contributions | |
beh | -0.706 |
yours | -0.693 |
possible | -0.598 |
ours | -0.587 |
theirs | -0.576 |
Token | Fl |
Token ID | 1610 |
Feature activation | +7.42e+00 |
Pos logit contributions | |
boy | +0.149 |
little | +0.141 |
girl | +0.140 |
sky | +0.138 |
man | +0.135 |
Neg logit contributions | |
beh | -0.209 |
yours | -0.195 |
possible | -0.176 |
ours | -0.169 |
theirs | -0.167 |
Token | utter |
Token ID | 10381 |
Feature activation | -5.94e+00 |
Pos logit contributions | |
boy | +0.312 |
girl | +0.291 |
sky | +0.287 |
little | +0.285 |
man | +0.285 |
Neg logit contributions | |
beh | -0.430 |
yours | -0.414 |
possible | -0.365 |
ours | -0.355 |
theirs | -0.346 |
Token | , |
Token ID | 11 |
Feature activation | +8.98e+00 |
Pos logit contributions | |
beh | +0.363 |
yours | +0.347 |
possible | +0.312 |
ours | +0.305 |
theirs | +0.293 |
Neg logit contributions | |
boy | -0.236 |
girl | -0.221 |
little | -0.216 |
sky | -0.215 |
was | -0.214 |
Token | came |
Token ID | 1625 |
Feature activation | +5.35e+00 |
Pos logit contributions | |
boy | +0.237 |
girl | +0.215 |
sky | +0.210 |
man | +0.209 |
little | +0.195 |
Neg logit contributions | |
beh | -0.644 |
yours | -0.636 |
ours | -0.564 |
possible | -0.561 |
theirs | -0.548 |
Token | to |
Token ID | 284 |
Feature activation | +6.28e-01 |
Pos logit contributions | |
boy | +0.203 |
girl | +0.189 |
little | +0.188 |
sky | +0.186 |
was | +0.183 |
Neg logit contributions | |
beh | -0.329 |
yours | -0.319 |
possible | -0.283 |
ours | -0.278 |
theirs | -0.271 |
Token | visit |
Token ID | 3187 |
Feature activation | -3.15e+00 |
Pos logit contributions | |
boy | +0.022 |
little | +0.020 |
girl | +0.020 |
man | +0.020 |
was | +0.019 |
Neg logit contributions | |
beh | -0.043 |
yours | -0.039 |
possible | -0.037 |
eking | -0.036 |
ours | -0.035 |
Token | . |
Token ID | 13 |
Feature activation | -1.07e+01 |
Pos logit contributions | |
beh | +0.207 |
yours | +0.186 |
possible | +0.179 |
eking | +0.173 |
ieve | +0.171 |
Neg logit contributions | |
boy | -0.118 |
was | -0.118 |
and | -0.111 |
little | -0.111 |
man | -0.108 |
Token | plan |
Token ID | 1410 |
Feature activation | +1.42e+00 |
Pos logit contributions | |
beh | +0.554 |
yours | +0.540 |
possible | +0.480 |
ours | +0.470 |
theirs | +0.457 |
Neg logit contributions | |
boy | -0.321 |
girl | -0.298 |
sky | -0.294 |
man | -0.290 |
little | -0.289 |
Token | . |
Token ID | 13 |
Feature activation | -1.10e+01 |
Pos logit contributions | |
was | +0.061 |
boy | +0.058 |
little | +0.056 |
man | +0.055 |
and | +0.055 |
Neg logit contributions | |
beh | -0.088 |
eking | -0.072 |
yours | -0.071 |
rying | -0.071 |
unnecessary | -0.069 |
Token | Then |
Token ID | 3244 |
Feature activation | -4.68e+00 |
Pos logit contributions | |
beh | +0.603 |
yours | +0.584 |
possible | +0.504 |
ours | +0.495 |
theirs | +0.478 |
Neg logit contributions | |
boy | -0.482 |
girl | -0.454 |
sky | -0.448 |
man | -0.447 |
little | -0.444 |
Token | we |
Token ID | 356 |
Feature activation | -7.01e+00 |
Pos logit contributions | |
beh | +0.282 |
yours | +0.247 |
theirs | +0.222 |
possible | +0.221 |
supposed | +0.220 |
Neg logit contributions | |
boy | -0.212 |
little | -0.200 |
girl | -0.199 |
was | -0.195 |
man | -0.191 |
Token | can |
Token ID | 460 |
Feature activation | +1.86e+00 |
Pos logit contributions | |
beh | +0.516 |
yours | +0.497 |
possible | +0.461 |
ours | +0.445 |
theirs | +0.442 |
Neg logit contributions | |
boy | -0.191 |
little | -0.169 |
was | -0.168 |
girl | -0.167 |
sky | -0.165 |
Token | both |
Token ID | 1111 |
Feature activation | -1.11e+01 |
Pos logit contributions | |
boy | +0.073 |
little | +0.070 |
was | +0.070 |
man | +0.069 |
sky | +0.068 |
Neg logit contributions | |
beh | -0.116 |
yours | -0.108 |
possible | -0.098 |
unnecessary | -0.097 |
theirs | -0.095 |
Token | have |
Token ID | 423 |
Feature activation | -6.11e+00 |
Pos logit contributions | |
beh | +0.685 |
yours | +0.646 |
possible | +0.586 |
unnecessary | +0.568 |
theirs | +0.568 |
Neg logit contributions | |
boy | -0.439 |
little | -0.417 |
was | -0.410 |
sky | -0.406 |
man | -0.400 |
Token | some |
Token ID | 617 |
Feature activation | -9.50e+00 |
Pos logit contributions | |
beh | +0.391 |
yours | +0.355 |
possible | +0.321 |
rying | +0.312 |
ieve | +0.308 |
Neg logit contributions | |
boy | -0.231 |
little | -0.223 |
sky | -0.217 |
was | -0.214 |
girl | -0.210 |
Token | ." |
Token ID | 526 |
Feature activation | -1.48e+01 |
Pos logit contributions | |
beh | +0.614 |
yours | +0.577 |
possible | +0.520 |
ours | +0.503 |
theirs | +0.493 |
Neg logit contributions | |
boy | -0.352 |
little | -0.331 |
girl | -0.323 |
sky | -0.322 |
man | -0.318 |
Token | So |
Token ID | 1406 |
Feature activation | +1.31e+00 |
Pos logit contributions | |
beh | +0.787 |
yours | +0.759 |
anks | +0.672 |
theirs | +0.658 |
ours | +0.651 |
Neg logit contributions | |
boy | -0.503 |
girl | -0.498 |
woman | -0.489 |
man | -0.486 |
child | -0.471 |
Token | , |
Token ID | 11 |
Feature activation | +8.89e+00 |
Pos logit contributions | |
boy | +0.037 |
sky | +0.036 |
girl | +0.034 |
man | +0.034 |
woman | +0.033 |
Neg logit contributions | |
yours | -0.081 |
beh | -0.079 |
possible | -0.073 |
ourselves | -0.073 |
anything | -0.071 |
Token | get |
Token ID | 651 |
Feature activation | -9.06e+00 |
Pos logit contributions | |
boy | +0.039 |
little | +0.038 |
was | +0.037 |
man | +0.036 |
girl | +0.036 |
Neg logit contributions | |
beh | -0.058 |
yours | -0.054 |
possible | -0.049 |
ieve | -0.049 |
ours | -0.048 |
Token | wet |
Token ID | 9583 |
Feature activation | -8.91e-01 |
Pos logit contributions | |
beh | +0.563 |
yours | +0.515 |
ieve | +0.476 |
possible | +0.471 |
rying | +0.460 |
Neg logit contributions | |
little | -0.357 |
boy | -0.334 |
girl | -0.316 |
was | -0.312 |
man | -0.302 |
Token | at |
Token ID | 379 |
Feature activation | -6.50e+00 |
Pos logit contributions | |
beh | +0.063 |
yours | +0.059 |
possible | +0.055 |
unnecessary | +0.052 |
ieve | +0.052 |
Neg logit contributions | |
boy | -0.029 |
little | -0.027 |
girl | -0.027 |
man | -0.026 |
sky | -0.025 |
Token | the |
Token ID | 262 |
Feature activation | +5.57e+00 |
Pos logit contributions | |
beh | +0.430 |
yours | +0.391 |
possible | +0.358 |
ieve | +0.343 |
ours | +0.342 |
Neg logit contributions | |
boy | -0.236 |
little | -0.223 |
sky | -0.220 |
girl | -0.216 |
was | -0.214 |
Token | lake |
Token ID | 13546 |
Feature activation | +1.16e+01 |
Pos logit contributions | |
boy | +0.162 |
sky | +0.147 |
girl | +0.146 |
little | +0.144 |
man | +0.144 |
Neg logit contributions | |
beh | -0.405 |
yours | -0.385 |
possible | -0.349 |
ours | -0.339 |
theirs | -0.338 |
Token | ? |
Token ID | 30 |
Feature activation | -1.73e+01 |
Pos logit contributions | |
boy | +0.457 |
little | +0.438 |
was | +0.425 |
girl | +0.422 |
man | +0.421 |
Neg logit contributions | |
beh | -0.738 |
yours | -0.667 |
unnecessary | -0.603 |
possible | -0.597 |
ieve | -0.593 |
Token | Well |
Token ID | 3894 |
Feature activation | -7.05e+00 |
Pos logit contributions | |
beh | +0.982 |
yours | +0.945 |
possible | +0.827 |
ours | +0.807 |
theirs | +0.784 |
Neg logit contributions | |
boy | -0.737 |
girl | -0.690 |
sky | -0.681 |
man | -0.678 |
little | -0.675 |
Token | , |
Token ID | 11 |
Feature activation | +9.64e+00 |
Pos logit contributions | |
yours | +0.263 |
beh | +0.261 |
possible | +0.208 |
ours | +0.201 |
theirs | +0.190 |
Neg logit contributions | |
boy | -0.425 |
sky | -0.409 |
girl | -0.406 |
man | -0.397 |
little | -0.385 |
Token | you |
Token ID | 345 |
Feature activation | -1.70e+01 |
Pos logit contributions | |
boy | +0.494 |
little | +0.470 |
girl | +0.468 |
was | +0.465 |
man | +0.453 |
Neg logit contributions | |
beh | -0.485 |
yours | -0.438 |
possible | -0.388 |
ours | -0.372 |
unnecessary | -0.365 |
Token | did |
Token ID | 750 |
Feature activation | +3.76e+00 |
Pos logit contributions | |
beh | +0.983 |
yours | +0.956 |
possible | +0.828 |
ours | +0.820 |
theirs | +0.792 |
Neg logit contributions | |
boy | -0.692 |
girl | -0.652 |
sky | -0.648 |
man | -0.635 |
little | -0.613 |
Token | such |
Token ID | 884 |
Feature activation | -1.18e+01 |
Pos logit contributions | |
little | +0.148 |
was | +0.146 |
and | +0.135 |
boy | +0.126 |
man | +0.123 |
Neg logit contributions | |
beh | -0.246 |
yours | -0.206 |
rying | -0.198 |
possible | -0.193 |
ares | -0.189 |
Token | You |
Token ID | 921 |
Feature activation | -1.20e+01 |
Pos logit contributions | |
beh | +0.568 |
yours | +0.546 |
possible | +0.483 |
ours | +0.470 |
theirs | +0.457 |
Neg logit contributions | |
boy | -0.397 |
girl | -0.370 |
sky | -0.366 |
little | -0.363 |
man | -0.362 |
Token | can |
Token ID | 460 |
Feature activation | +3.25e+00 |
Pos logit contributions | |
beh | +0.736 |
yours | +0.708 |
possible | +0.631 |
ours | +0.613 |
theirs | +0.599 |
Neg logit contributions | |
boy | -0.465 |
girl | -0.430 |
sky | -0.425 |
little | -0.424 |
man | -0.422 |
Token | 't |
Token ID | 470 |
Feature activation | -6.43e+00 |
Pos logit contributions | |
boy | +0.144 |
was | +0.143 |
little | +0.142 |
man | +0.140 |
sky | +0.134 |
Neg logit contributions | |
beh | -0.178 |
yours | -0.165 |
possible | -0.149 |
unnecessary | -0.148 |
theirs | -0.145 |
Token | borrow |
Token ID | 8804 |
Feature activation | -1.14e+01 |
Pos logit contributions | |
beh | +0.377 |
yours | +0.360 |
possible | +0.328 |
unnecessary | +0.328 |
theirs | +0.322 |
Neg logit contributions | |
boy | -0.263 |
man | -0.257 |
was | -0.257 |
little | -0.254 |
girl | -0.238 |
Token | it |
Token ID | 340 |
Feature activation | -1.73e+00 |
Pos logit contributions | |
beh | +0.802 |
yours | +0.688 |
possible | +0.660 |
rying | +0.651 |
supposed | +0.647 |
Neg logit contributions | |
little | -0.376 |
boy | -0.367 |
man | -0.341 |
was | -0.338 |
sky | -0.338 |
Token | .' |
Token ID | 2637 |
Feature activation | -2.11e+01 |
Pos logit contributions | |
beh | +0.105 |
yours | +0.092 |
eking | +0.086 |
ieve | +0.086 |
unnecessary | +0.086 |
Neg logit contributions | |
boy | -0.069 |
little | -0.067 |
and | -0.067 |
man | -0.067 |
was | -0.065 |
Token | |
Token ID | 220 |
Feature activation | +3.51e+00 |
Pos logit contributions | |
beh | +1.533 |
yours | +1.489 |
possible | +1.345 |
ours | +1.321 |
theirs | +1.291 |
Neg logit contributions | |
boy | -0.555 |
girl | -0.496 |
sky | -0.491 |
man | -0.485 |
little | -0.471 |
Token |
|
Token ID | 198 |
Feature activation | +7.07e+00 |
Pos logit contributions | |
boy | +0.038 |
little | +0.028 |
girl | +0.027 |
was | +0.026 |
man | +0.025 |
Neg logit contributions | |
beh | -0.317 |
yours | -0.308 |
possible | -0.282 |
ours | -0.281 |
theirs | -0.274 |
Token |
|
Token ID | 198 |
Feature activation | +7.11e+00 |
Pos logit contributions | |
boy | +0.138 |
girl | +0.118 |
sky | +0.116 |
man | +0.116 |
dog | +0.095 |
Neg logit contributions | |
beh | -0.549 |
yours | -0.533 |
possible | -0.489 |
ourselves | -0.478 |
ours | -0.475 |
Token | B |
Token ID | 33 |
Feature activation | +6.41e+00 |
Pos logit contributions | |
boy | +0.245 |
sky | +0.230 |
man | +0.226 |
girl | +0.222 |
dog | +0.202 |
Neg logit contributions | |
beh | -0.456 |
yours | -0.431 |
possible | -0.385 |
ourselves | -0.377 |
ours | -0.375 |
Token | obby |
Token ID | 11369 |
Feature activation | -1.45e+00 |
Pos logit contributions | |
boy | +0.306 |
girl | +0.288 |
little | +0.284 |
sky | +0.283 |
man | +0.281 |
Neg logit contributions | |
beh | -0.334 |
yours | -0.322 |
possible | -0.280 |
ours | -0.272 |
theirs | -0.263 |
Token | I |
Token ID | 314 |
Feature activation | -1.22e+01 |
Pos logit contributions | |
beh | +0.615 |
yours | +0.593 |
possible | +0.489 |
ours | +0.483 |
theirs | +0.460 |
Neg logit contributions | |
boy | -0.623 |
man | -0.592 |
girl | -0.588 |
sky | -0.587 |
little | -0.574 |
Token | 'm |
Token ID | 1101 |
Feature activation | -1.73e+00 |
Pos logit contributions | |
beh | +0.747 |
yours | +0.720 |
possible | +0.638 |
ours | +0.623 |
theirs | +0.607 |
Neg logit contributions | |
boy | -0.471 |
girl | -0.437 |
sky | -0.432 |
man | -0.428 |
little | -0.427 |
Token | glad |
Token ID | 9675 |
Feature activation | -5.03e+00 |
Pos logit contributions | |
beh | +0.105 |
yours | +0.087 |
unnecessary | +0.082 |
anks | +0.082 |
ales | +0.081 |
Neg logit contributions | |
little | -0.075 |
boy | -0.068 |
was | -0.066 |
girl | -0.065 |
man | -0.064 |
Token | you |
Token ID | 345 |
Feature activation | -1.75e+01 |
Pos logit contributions | |
beh | +0.254 |
yours | +0.224 |
possible | +0.199 |
ieve | +0.196 |
ours | +0.192 |
Neg logit contributions | |
boy | -0.264 |
little | -0.253 |
girl | -0.251 |
was | -0.248 |
man | -0.243 |
Token | 're |
Token ID | 821 |
Feature activation | -9.03e+00 |
Pos logit contributions | |
beh | +1.123 |
yours | +1.056 |
possible | +0.966 |
ours | +0.921 |
theirs | +0.912 |
Neg logit contributions | |
boy | -0.663 |
was | -0.630 |
little | -0.619 |
girl | -0.612 |
sky | -0.598 |
Token | okay |
Token ID | 8788 |
Feature activation | -1.84e+01 |
Pos logit contributions | |
beh | +0.510 |
yours | +0.428 |
unnecessary | +0.401 |
ieve | +0.387 |
ales | +0.386 |
Neg logit contributions | |
little | -0.440 |
boy | -0.407 |
girl | -0.393 |
man | -0.386 |
and | -0.384 |
Token | ." |
Token ID | 526 |
Feature activation | -1.44e+01 |
Pos logit contributions | |
beh | +0.902 |
yours | +0.833 |
possible | +0.718 |
ours | +0.698 |
theirs | +0.679 |
Neg logit contributions | |
boy | -0.980 |
girl | -0.926 |
little | -0.921 |
sky | -0.906 |
man | -0.901 |
Token | Mitt |
Token ID | 16627 |
Feature activation | -3.65e+00 |
Pos logit contributions | |
beh | +0.769 |
yours | +0.713 |
anks | +0.602 |
theirs | +0.596 |
ourselves | +0.591 |
Neg logit contributions | |
boy | -0.581 |
man | -0.540 |
sky | -0.540 |
girl | -0.535 |
child | -0.511 |
Token | ens |
Token ID | 641 |
Feature activation | -4.81e+00 |
Pos logit contributions | |
beh | +0.192 |
yours | +0.184 |
possible | +0.159 |
ours | +0.154 |
theirs | +0.150 |
Neg logit contributions | |
boy | -0.173 |
girl | -0.163 |
sky | -0.161 |
man | -0.160 |
little | -0.160 |
Token | pur |
Token ID | 1308 |
Feature activation | +5.28e+00 |
Pos logit contributions | |
beh | +0.248 |
myself | +0.213 |
yours | +0.211 |
properly | +0.201 |
able | +0.196 |
Neg logit contributions | |
sky | -0.167 |
boy | -0.157 |
girl | -0.150 |
child | -0.132 |
field | -0.130 |
Token | red |
Token ID | 445 |
Feature activation | -1.34e+01 |
Pos logit contributions | |
boy | +0.303 |
girl | +0.291 |
little | +0.285 |
sky | +0.284 |
was | +0.283 |
Neg logit contributions | |
beh | -0.220 |
yours | -0.216 |
possible | -0.185 |
ours | -0.179 |
theirs | -0.166 |
Token | I |
Token ID | 314 |
Feature activation | -1.17e+01 |
Pos logit contributions | |
beh | +0.581 |
yours | +0.558 |
possible | +0.495 |
ours | +0.481 |
theirs | +0.469 |
Neg logit contributions | |
boy | -0.390 |
girl | -0.363 |
sky | -0.358 |
man | -0.354 |
little | -0.354 |
Token | want |
Token ID | 765 |
Feature activation | -1.44e+01 |
Pos logit contributions | |
beh | +0.607 |
yours | +0.580 |
possible | +0.502 |
ours | +0.485 |
theirs | +0.475 |
Neg logit contributions | |
boy | -0.544 |
sky | -0.521 |
girl | -0.514 |
man | -0.506 |
little | -0.504 |
Token | to |
Token ID | 284 |
Feature activation | +1.11e+00 |
Pos logit contributions | |
beh | +0.934 |
yours | +0.857 |
possible | +0.776 |
ours | +0.751 |
ieve | +0.747 |
Neg logit contributions | |
boy | -0.537 |
little | -0.506 |
sky | -0.499 |
girl | -0.493 |
was | -0.490 |
Token | hear |
Token ID | 3285 |
Feature activation | -2.61e+00 |
Pos logit contributions | |
boy | +0.042 |
little | +0.041 |
was | +0.040 |
man | +0.040 |
sky | +0.039 |
Neg logit contributions | |
beh | -0.071 |
yours | -0.064 |
possible | -0.059 |
unnecessary | -0.059 |
rying | -0.057 |
Token | how |
Token ID | 703 |
Feature activation | -1.15e+01 |
Pos logit contributions | |
beh | +0.163 |
yours | +0.138 |
possible | +0.130 |
rying | +0.128 |
unnecessary | +0.125 |
Neg logit contributions | |
boy | -0.107 |
little | -0.107 |
was | -0.102 |
sky | -0.101 |
girl | -0.100 |
Token | you |
Token ID | 345 |
Feature activation | -1.69e+01 |
Pos logit contributions | |
beh | +0.716 |
yours | +0.673 |
possible | +0.600 |
ieve | +0.595 |
eking | +0.591 |
Neg logit contributions | |
boy | -0.460 |
little | -0.437 |
girl | -0.426 |
sky | -0.412 |
man | -0.411 |
Token | made |
Token ID | 925 |
Feature activation | +3.81e+00 |
Pos logit contributions | |
beh | +0.933 |
yours | +0.887 |
possible | +0.781 |
ours | +0.753 |
theirs | +0.736 |
Neg logit contributions | |
boy | -0.764 |
girl | -0.717 |
sky | -0.713 |
little | -0.711 |
man | -0.706 |
Token | your |
Token ID | 534 |
Feature activation | -1.10e+01 |
Pos logit contributions | |
little | +0.146 |
boy | +0.139 |
sky | +0.131 |
was | +0.131 |
man | +0.130 |
Neg logit contributions | |
beh | -0.255 |
yours | -0.209 |
rying | -0.208 |
possible | -0.203 |
eking | -0.197 |
Token | wrap |
Token ID | 14441 |
Feature activation | +2.19e+00 |
Pos logit contributions | |
beh | +0.730 |
yours | +0.705 |
possible | +0.632 |
ours | +0.618 |
theirs | +0.603 |
Neg logit contributions | |
boy | -0.372 |
girl | -0.341 |
sky | -0.336 |
man | -0.333 |
little | -0.332 |
Token | and |
Token ID | 290 |
Feature activation | +1.30e+01 |
Pos logit contributions | |
boy | +0.098 |
girl | +0.093 |
was | +0.091 |
man | +0.091 |
little | +0.091 |
Neg logit contributions | |
beh | -0.128 |
yours | -0.115 |
possible | -0.103 |
unnecessary | -0.102 |
eking | -0.100 |
Token | loop |
Token ID | 9052 |
Feature activation | +7.75e+00 |
Pos logit contributions | |
boy | +0.551 |
girl | +0.516 |
little | +0.510 |
man | +0.505 |
sky | +0.496 |
Neg logit contributions | |
beh | -0.763 |
yours | -0.716 |
possible | -0.639 |
ours | -0.613 |
unnecessary | -0.602 |
Token | open |
Token ID | 1280 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
beh | +0.002 |
yours | +0.002 |
possible | +0.001 |
ours | +0.001 |
theirs | +0.001 |
Neg logit contributions | |
boy | -0.001 |
girl | -0.001 |
man | -0.001 |
sky | -0.001 |
little | -0.001 |
Token | the |
Token ID | 262 |
Feature activation | +5.73e+00 |
Pos logit contributions | |
yours | +0.069 |
beh | +0.068 |
possible | +0.061 |
ours | +0.060 |
theirs | +0.059 |
Neg logit contributions | |
boy | -0.040 |
girl | -0.038 |
little | -0.037 |
sky | -0.037 |
man | -0.036 |
Token | door |
Token ID | 3420 |
Feature activation | +4.22e+00 |
Pos logit contributions | |
boy | +0.172 |
was | +0.153 |
girl | +0.153 |
little | +0.147 |
man | +0.146 |
Neg logit contributions | |
yours | -0.377 |
beh | -0.355 |
ours | -0.339 |
possible | -0.322 |
theirs | -0.314 |
Token | and |
Token ID | 290 |
Feature activation | +1.25e+01 |
Pos logit contributions | |
boy | +0.072 |
girl | +0.062 |
little | +0.057 |
sky | +0.057 |
man | +0.055 |
Neg logit contributions | |
beh | -0.339 |
yours | -0.335 |
possible | -0.309 |
ours | -0.302 |
theirs | -0.300 |
Token | let |
Token ID | 1309 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
boy | +0.529 |
girl | +0.492 |
man | +0.487 |
little | +0.482 |
sky | +0.480 |
Neg logit contributions | |
beh | -0.747 |
yours | -0.710 |
possible | -0.629 |
ours | -0.602 |
theirs | -0.590 |
Token | him |
Token ID | 683 |
Feature activation | -1.71e+01 |
Pos logit contributions | |
beh | +0.078 |
yours | +0.073 |
possible | +0.064 |
ours | +0.063 |
eking | +0.060 |
Neg logit contributions | |
boy | -0.062 |
man | -0.057 |
girl | -0.057 |
little | -0.057 |
was | -0.056 |
Token | in |
Token ID | 287 |
Feature activation | -8.23e-01 |
Pos logit contributions | |
beh | +1.139 |
yours | +1.090 |
possible | +1.044 |
unnecessary | +0.999 |
ours | +0.980 |
Neg logit contributions | |
little | -0.592 |
was | -0.588 |
boy | -0.582 |
and | -0.547 |
girl | -0.515 |
Token | . |
Token ID | 13 |
Feature activation | -1.01e+01 |
Pos logit contributions | |
beh | +0.049 |
yours | +0.046 |
possible | +0.040 |
ours | +0.039 |
unnecessary | +0.039 |
Neg logit contributions | |
boy | -0.035 |
girl | -0.033 |
man | -0.033 |
little | -0.032 |
sky | -0.032 |
Token | They |
Token ID | 1119 |
Feature activation | -2.97e-01 |
Pos logit contributions | |
yours | +0.608 |
beh | +0.599 |
ours | +0.539 |
theirs | +0.536 |
possible | +0.529 |
Neg logit contributions | |
boy | -0.340 |
girl | -0.336 |
little | -0.320 |
man | -0.315 |
sky | -0.311 |
Token | show |
Token ID | 905 |
Feature activation | -4.92e+00 |
Pos logit contributions | |
beh | +0.017 |
yours | +0.016 |
possible | +0.014 |
ours | +0.014 |
theirs | +0.013 |
Neg logit contributions | |
boy | -0.013 |
girl | -0.012 |
sky | -0.012 |
man | -0.012 |
little | -0.012 |
Token | him |
Token ID | 683 |
Feature activation | -1.72e+01 |
Pos logit contributions | |
beh | +0.254 |
yours | +0.223 |
possible | +0.205 |
ours | +0.192 |
eking | +0.188 |
Neg logit contributions | |
boy | -0.262 |
little | -0.249 |
was | -0.245 |
man | -0.244 |
girl | -0.243 |