Token | just |
Token ID | 655 |
Feature activation | -6.05e+00 |
Pos logit contributions | |
boy | +0.426 |
little | +0.407 |
girl | +0.399 |
man | +0.385 |
sky | +0.382 |
Neg logit contributions | |
beh | -0.719 |
yours | -0.690 |
possible | -0.615 |
ours | -0.594 |
theirs | -0.584 |
Token | wanted |
Token ID | 2227 |
Feature activation | +3.56e+00 |
Pos logit contributions | |
beh | +0.372 |
yours | +0.327 |
ieve | +0.296 |
unnecessary | +0.294 |
possible | +0.292 |
Neg logit contributions | |
little | -0.271 |
boy | -0.255 |
was | -0.249 |
girl | -0.245 |
sky | -0.230 |
Token | to |
Token ID | 284 |
Feature activation | +4.08e+00 |
Pos logit contributions | |
little | +0.120 |
boy | +0.115 |
was | +0.109 |
girl | +0.105 |
man | +0.103 |
Neg logit contributions | |
beh | -0.260 |
yours | -0.226 |
ieve | -0.218 |
rying | -0.213 |
eking | -0.211 |
Token | see |
Token ID | 766 |
Feature activation | -3.38e+00 |
Pos logit contributions | |
little | +0.174 |
boy | +0.162 |
man | +0.156 |
was | +0.155 |
girl | +0.150 |
Neg logit contributions | |
beh | -0.262 |
yours | -0.234 |
unnecessary | -0.218 |
rying | -0.213 |
possible | -0.212 |
Token | the |
Token ID | 262 |
Feature activation | +8.34e+00 |
Pos logit contributions | |
beh | +0.212 |
yours | +0.177 |
possible | +0.162 |
anks | +0.162 |
eking | +0.160 |
Neg logit contributions | |
little | -0.157 |
boy | -0.153 |
girl | -0.146 |
man | -0.139 |
was | -0.138 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.65e+01 |
Pos logit contributions | |
boy | +0.242 |
little | +0.229 |
girl | +0.223 |
man | +0.208 |
sky | +0.206 |
Neg logit contributions | |
beh | -0.607 |
yours | -0.589 |
possible | -0.535 |
ours | -0.518 |
unnecessary | -0.509 |
Token | ins |
Token ID | 1040 |
Feature activation | -5.02e+00 |
Pos logit contributions | |
boy | +1.001 |
little | +0.944 |
girl | +0.925 |
man | +0.886 |
was | +0.857 |
Neg logit contributions | |
beh | -1.731 |
yours | -1.686 |
possible | -1.518 |
ours | -1.480 |
unnecessary | -1.478 |
Token | . |
Token ID | 13 |
Feature activation | -7.63e+00 |
Pos logit contributions | |
beh | +0.338 |
yours | +0.310 |
unnecessary | +0.296 |
ours | +0.277 |
ieve | +0.271 |
Neg logit contributions | |
boy | -0.190 |
was | -0.182 |
little | -0.181 |
girl | -0.178 |
and | -0.176 |
Token | He |
Token ID | 679 |
Feature activation | +5.16e-01 |
Pos logit contributions | |
beh | +0.528 |
yours | +0.511 |
possible | +0.452 |
ours | +0.445 |
theirs | +0.439 |
Neg logit contributions | |
boy | -0.242 |
little | -0.234 |
girl | -0.228 |
man | -0.218 |
sky | -0.214 |
Token | walked |
Token ID | 6807 |
Feature activation | +7.49e+00 |
Pos logit contributions | |
sky | +0.016 |
girl | +0.016 |
boy | +0.016 |
little | +0.015 |
man | +0.014 |
Neg logit contributions | |
beh | -0.033 |
yours | -0.031 |
theirs | -0.027 |
possible | -0.027 |
properly | -0.027 |
Token | back |
Token ID | 736 |
Feature activation | -3.85e+00 |
Pos logit contributions | |
boy | +0.362 |
little | +0.348 |
girl | +0.345 |
sky | +0.328 |
man | +0.324 |
Neg logit contributions | |
beh | -0.385 |
yours | -0.375 |
possible | -0.333 |
theirs | -0.314 |
rying | -0.310 |
Token | would |
Token ID | 561 |
Feature activation | +4.32e+00 |
Pos logit contributions | |
yours | +0.290 |
beh | +0.278 |
possible | +0.258 |
ours | +0.258 |
ourselves | +0.248 |
Neg logit contributions | |
boy | -0.122 |
girl | -0.114 |
sky | -0.111 |
little | -0.110 |
was | -0.108 |
Token | be |
Token ID | 307 |
Feature activation | +3.29e+00 |
Pos logit contributions | |
little | +0.181 |
boy | +0.173 |
was | +0.165 |
man | +0.161 |
girl | +0.160 |
Neg logit contributions | |
beh | -0.282 |
yours | -0.251 |
unnecessary | -0.236 |
possible | -0.230 |
ours | -0.224 |
Token | nicer |
Token ID | 36597 |
Feature activation | -6.35e+00 |
Pos logit contributions | |
little | +0.127 |
boy | +0.126 |
girl | +0.119 |
man | +0.112 |
was | +0.109 |
Neg logit contributions | |
beh | -0.218 |
yours | -0.198 |
possible | -0.178 |
ieve | -0.175 |
unnecessary | -0.173 |
Token | than |
Token ID | 621 |
Feature activation | -6.18e+00 |
Pos logit contributions | |
beh | +0.455 |
yours | +0.425 |
possible | +0.383 |
unnecessary | +0.376 |
ieve | +0.375 |
Neg logit contributions | |
boy | -0.214 |
little | -0.204 |
girl | -0.199 |
man | -0.187 |
sky | -0.182 |
Token | the |
Token ID | 262 |
Feature activation | +8.34e+00 |
Pos logit contributions | |
beh | +0.380 |
yours | +0.306 |
ieve | +0.291 |
unnecessary | +0.290 |
eking | +0.288 |
Neg logit contributions | |
little | -0.294 |
boy | -0.278 |
girl | -0.261 |
man | -0.256 |
was | -0.253 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.65e+01 |
Pos logit contributions | |
boy | +0.257 |
little | +0.245 |
girl | +0.238 |
man | +0.224 |
sky | +0.222 |
Neg logit contributions | |
beh | -0.600 |
yours | -0.577 |
possible | -0.523 |
ours | -0.505 |
unnecessary | -0.499 |
Token | in |
Token ID | 259 |
Feature activation | +1.45e+00 |
Pos logit contributions | |
boy | +1.091 |
little | +1.042 |
girl | +1.019 |
man | +0.980 |
sky | +0.955 |
Neg logit contributions | |
beh | -1.666 |
yours | -1.590 |
possible | -1.421 |
ours | -1.386 |
unnecessary | -1.376 |
Token | . |
Token ID | 13 |
Feature activation | -7.63e+00 |
Pos logit contributions | |
boy | +0.055 |
little | +0.053 |
girl | +0.052 |
man | +0.049 |
sky | +0.049 |
Neg logit contributions | |
beh | -0.095 |
yours | -0.090 |
possible | -0.081 |
ours | -0.078 |
unnecessary | -0.077 |
Token | They |
Token ID | 1119 |
Feature activation | +2.11e+00 |
Pos logit contributions | |
beh | +0.567 |
yours | +0.547 |
possible | +0.494 |
ours | +0.480 |
theirs | +0.473 |
Neg logit contributions | |
boy | -0.218 |
little | -0.207 |
girl | -0.201 |
man | -0.189 |
sky | -0.185 |
Token | followed |
Token ID | 3940 |
Feature activation | +8.60e+00 |
Pos logit contributions | |
boy | +0.068 |
little | +0.065 |
girl | +0.063 |
sky | +0.059 |
man | +0.059 |
Neg logit contributions | |
beh | -0.150 |
yours | -0.144 |
possible | -0.130 |
ours | -0.125 |
theirs | -0.124 |
Token | their |
Token ID | 511 |
Feature activation | -1.72e-01 |
Pos logit contributions | |
little | +0.302 |
boy | +0.302 |
girl | +0.279 |
man | +0.272 |
sky | +0.254 |
Neg logit contributions | |
beh | -0.616 |
yours | -0.568 |
possible | -0.523 |
eking | -0.499 |
ours | -0.498 |
Token | making |
Token ID | 1642 |
Feature activation | -2.08e+00 |
Pos logit contributions | |
beh | +0.023 |
yours | +0.022 |
eking | +0.019 |
iable | +0.019 |
ck | +0.018 |
Neg logit contributions | |
sky | -0.014 |
girl | -0.014 |
boy | -0.014 |
little | -0.012 |
man | -0.012 |
Token | noises |
Token ID | 26782 |
Feature activation | +4.48e-01 |
Pos logit contributions | |
beh | +0.137 |
yours | +0.121 |
possible | +0.110 |
rying | +0.109 |
ours | +0.106 |
Neg logit contributions | |
little | -0.087 |
boy | -0.082 |
girl | -0.076 |
was | -0.076 |
sky | -0.074 |
Token | . |
Token ID | 13 |
Feature activation | -6.16e+00 |
Pos logit contributions | |
boy | +0.011 |
little | +0.011 |
and | +0.010 |
was | +0.009 |
girl | +0.009 |
Neg logit contributions | |
beh | -0.038 |
yours | -0.034 |
unnecessary | -0.032 |
possible | -0.031 |
ours | -0.031 |
Token | The |
Token ID | 383 |
Feature activation | +1.01e+01 |
Pos logit contributions | |
beh | +0.450 |
yours | +0.433 |
possible | +0.390 |
ours | +0.379 |
theirs | +0.374 |
Neg logit contributions | |
boy | -0.182 |
little | -0.173 |
girl | -0.169 |
man | -0.159 |
sky | -0.155 |
Token | lost |
Token ID | 2626 |
Feature activation | +5.44e-01 |
Pos logit contributions | |
boy | +0.218 |
little | +0.203 |
girl | +0.200 |
sky | +0.187 |
was | +0.186 |
Neg logit contributions | |
beh | -0.742 |
yours | -0.735 |
possible | -0.669 |
ours | -0.649 |
unnecessary | -0.644 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.65e+01 |
Pos logit contributions | |
boy | +0.014 |
little | +0.013 |
girl | +0.013 |
sky | +0.012 |
man | +0.012 |
Neg logit contributions | |
yours | -0.040 |
beh | -0.040 |
possible | -0.035 |
ours | -0.035 |
theirs | -0.035 |
Token | in |
Token ID | 259 |
Feature activation | +1.45e+00 |
Pos logit contributions | |
boy | +1.169 |
little | +1.130 |
girl | +1.106 |
man | +1.063 |
sky | +1.055 |
Neg logit contributions | |
beh | -1.588 |
yours | -1.504 |
possible | -1.330 |
ours | -1.276 |
unnecessary | -1.258 |
Token | joined |
Token ID | 5399 |
Feature activation | +2.84e+00 |
Pos logit contributions | |
boy | +0.036 |
little | +0.033 |
girl | +0.033 |
man | +0.030 |
sky | +0.030 |
Neg logit contributions | |
beh | -0.113 |
yours | -0.108 |
possible | -0.099 |
ours | -0.096 |
unnecessary | -0.095 |
Token | them |
Token ID | 606 |
Feature activation | -9.00e+00 |
Pos logit contributions | |
boy | +0.113 |
little | +0.110 |
girl | +0.106 |
man | +0.100 |
sky | +0.099 |
Neg logit contributions | |
beh | -0.184 |
yours | -0.174 |
possible | -0.158 |
ours | -0.150 |
unnecessary | -0.150 |
Token | and |
Token ID | 290 |
Feature activation | +1.48e+01 |
Pos logit contributions | |
beh | +0.658 |
yours | +0.621 |
possible | +0.564 |
unnecessary | +0.550 |
ours | +0.545 |
Neg logit contributions | |
boy | -0.284 |
little | -0.284 |
girl | -0.259 |
man | -0.248 |
sky | -0.245 |
Token | he |
Token ID | 339 |
Feature activation | -4.18e+00 |
Pos logit contributions | |
boy | +0.524 |
little | +0.498 |
girl | +0.482 |
man | +0.467 |
was | +0.450 |
Neg logit contributions | |
beh | -1.032 |
yours | -0.984 |
possible | -0.888 |
ours | -0.861 |
unnecessary | -0.860 |
Token | they |
Token ID | 484 |
Feature activation | +2.01e+00 |
Pos logit contributions | |
beh | +0.598 |
yours | +0.574 |
possible | +0.512 |
ours | +0.500 |
theirs | +0.493 |
Neg logit contributions | |
boy | -0.261 |
little | -0.248 |
girl | -0.242 |
sky | -0.230 |
man | -0.227 |
Token | said |
Token ID | 531 |
Feature activation | +4.65e+00 |
Pos logit contributions | |
boy | +0.069 |
little | +0.067 |
girl | +0.064 |
man | +0.061 |
sky | +0.059 |
Neg logit contributions | |
beh | -0.142 |
yours | -0.135 |
possible | -0.124 |
unnecessary | -0.119 |
ours | -0.118 |
Token | . |
Token ID | 13 |
Feature activation | -6.91e+00 |
Pos logit contributions | |
little | +0.168 |
boy | +0.158 |
was | +0.151 |
girl | +0.147 |
man | +0.147 |
Neg logit contributions | |
beh | -0.331 |
yours | -0.311 |
ours | -0.281 |
unnecessary | -0.280 |
possible | -0.279 |
Token | " |
Token ID | 366 |
Feature activation | +8.96e+00 |
Pos logit contributions | |
yours | +0.493 |
beh | +0.491 |
possible | +0.440 |
theirs | +0.440 |
ours | +0.428 |
Neg logit contributions | |
boy | -0.178 |
girl | -0.167 |
little | -0.159 |
man | -0.150 |
sky | -0.141 |
Token | The |
Token ID | 464 |
Feature activation | +1.33e+01 |
Pos logit contributions | |
boy | +0.328 |
little | +0.321 |
girl | +0.307 |
man | +0.296 |
sky | +0.291 |
Neg logit contributions | |
beh | -0.595 |
yours | -0.565 |
possible | -0.508 |
ours | -0.486 |
ourselves | -0.481 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.61e+01 |
Pos logit contributions | |
boy | +0.378 |
little | +0.356 |
girl | +0.346 |
sky | +0.323 |
man | +0.321 |
Neg logit contributions | |
beh | -0.941 |
yours | -0.925 |
possible | -0.837 |
ours | -0.809 |
unnecessary | -0.791 |
Token | in |
Token ID | 259 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
boy | +1.124 |
little | +1.069 |
girl | +1.055 |
man | +1.011 |
sky | +0.982 |
Neg logit contributions | |
beh | -1.590 |
yours | -1.516 |
possible | -1.357 |
ours | -1.305 |
unnecessary | -1.288 |
Token | is |
Token ID | 318 |
Feature activation | +5.37e+00 |
Pos logit contributions | |
boy | +0.031 |
little | +0.029 |
girl | +0.029 |
sky | +0.027 |
man | +0.026 |
Neg logit contributions | |
beh | -0.079 |
yours | -0.076 |
possible | -0.068 |
ieve | -0.066 |
ours | -0.066 |
Token | mean |
Token ID | 1612 |
Feature activation | -2.92e+00 |
Pos logit contributions | |
little | +0.214 |
boy | +0.203 |
girl | +0.186 |
man | +0.186 |
sky | +0.183 |
Neg logit contributions | |
beh | -0.368 |
yours | -0.317 |
ieve | -0.289 |
rying | -0.289 |
possible | -0.288 |
Token | ! |
Token ID | 0 |
Feature activation | -1.01e+01 |
Pos logit contributions | |
beh | +0.204 |
yours | +0.179 |
unnecessary | +0.169 |
ieve | +0.169 |
possible | +0.167 |
Neg logit contributions | |
boy | -0.116 |
little | -0.112 |
girl | -0.106 |
and | -0.105 |
man | -0.104 |
Token | He |
Token ID | 679 |
Feature activation | -3.13e-02 |
Pos logit contributions | |
beh | +0.598 |
yours | +0.567 |
possible | +0.499 |
ours | +0.479 |
unnecessary | +0.470 |
Neg logit contributions | |
boy | -0.450 |
little | -0.436 |
girl | -0.428 |
man | -0.412 |
sky | -0.409 |
Token | and |
Token ID | 290 |
Feature activation | +1.67e+01 |
Pos logit contributions | |
boy | +0.227 |
little | +0.221 |
girl | +0.210 |
man | +0.186 |
sky | +0.182 |
Neg logit contributions | |
beh | -0.596 |
yours | -0.556 |
possible | -0.507 |
ours | -0.501 |
ieve | -0.499 |
Token | scared |
Token ID | 12008 |
Feature activation | +3.14e+00 |
Pos logit contributions | |
boy | +0.538 |
little | +0.530 |
girl | +0.498 |
man | +0.476 |
was | +0.471 |
Neg logit contributions | |
beh | -1.198 |
yours | -1.130 |
possible | -1.031 |
unnecessary | -1.007 |
ieve | -1.005 |
Token | of |
Token ID | 286 |
Feature activation | +7.59e+00 |
Pos logit contributions | |
little | +0.121 |
boy | +0.118 |
girl | +0.111 |
was | +0.107 |
and | +0.105 |
Neg logit contributions | |
beh | -0.221 |
yours | -0.202 |
possible | -0.184 |
ours | -0.183 |
ieve | -0.182 |
Token | the |
Token ID | 262 |
Feature activation | +8.08e+00 |
Pos logit contributions | |
little | +0.320 |
boy | +0.293 |
was | +0.277 |
man | +0.273 |
girl | +0.271 |
Neg logit contributions | |
beh | -0.515 |
yours | -0.433 |
anks | -0.398 |
rying | -0.394 |
ours | -0.393 |
Token | big |
Token ID | 1263 |
Feature activation | +1.11e+01 |
Pos logit contributions | |
boy | +0.235 |
little | +0.222 |
girl | +0.216 |
man | +0.203 |
sky | +0.202 |
Neg logit contributions | |
beh | -0.583 |
yours | -0.563 |
possible | -0.513 |
ours | -0.494 |
unnecessary | -0.491 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.61e+01 |
Pos logit contributions | |
boy | +0.250 |
little | +0.243 |
girl | +0.227 |
sky | +0.216 |
was | +0.212 |
Neg logit contributions | |
yours | -0.813 |
beh | -0.809 |
possible | -0.733 |
ours | -0.716 |
theirs | -0.696 |
Token | ins |
Token ID | 1040 |
Feature activation | -5.38e+00 |
Pos logit contributions | |
boy | +1.063 |
little | +1.010 |
girl | +0.991 |
man | +0.947 |
was | +0.918 |
Neg logit contributions | |
beh | -1.638 |
yours | -1.600 |
possible | -1.431 |
ours | -1.394 |
unnecessary | -1.370 |
Token | . |
Token ID | 13 |
Feature activation | -7.97e+00 |
Pos logit contributions | |
beh | +0.382 |
yours | +0.366 |
possible | +0.328 |
ours | +0.322 |
unnecessary | +0.320 |
Neg logit contributions | |
boy | -0.181 |
little | -0.174 |
girl | -0.167 |
man | -0.163 |
sky | -0.158 |
Token | They |
Token ID | 1119 |
Feature activation | +1.83e+00 |
Pos logit contributions | |
beh | +0.559 |
yours | +0.539 |
possible | +0.482 |
ours | +0.470 |
theirs | +0.463 |
Neg logit contributions | |
boy | -0.258 |
little | -0.245 |
girl | -0.241 |
man | -0.229 |
sky | -0.222 |
Token | were |
Token ID | 547 |
Feature activation | +6.99e+00 |
Pos logit contributions | |
boy | +0.056 |
girl | +0.052 |
little | +0.052 |
sky | +0.049 |
man | +0.048 |
Neg logit contributions | |
beh | -0.131 |
yours | -0.124 |
possible | -0.112 |
ours | -0.109 |
theirs | -0.109 |
Token | selfish |
Token ID | 20363 |
Feature activation | -3.01e-01 |
Pos logit contributions | |
little | +0.277 |
boy | +0.272 |
girl | +0.254 |
man | +0.248 |
sky | +0.246 |
Neg logit contributions | |
beh | -0.453 |
yours | -0.423 |
possible | -0.378 |
ieve | -0.363 |
unnecessary | -0.363 |
Token | , |
Token ID | 11 |
Feature activation | +1.33e+01 |
Pos logit contributions | |
boy | +0.705 |
girl | +0.672 |
little | +0.654 |
man | +0.648 |
sky | +0.632 |
Neg logit contributions | |
yours | -1.088 |
beh | -1.077 |
possible | -0.980 |
ours | -0.922 |
theirs | -0.917 |
Token | there |
Token ID | 612 |
Feature activation | -7.74e-01 |
Pos logit contributions | |
boy | +0.603 |
man | +0.602 |
girl | +0.584 |
sky | +0.583 |
woman | +0.554 |
Neg logit contributions | |
yours | -0.694 |
beh | -0.639 |
possible | -0.584 |
ours | -0.574 |
theirs | -0.556 |
Token | was |
Token ID | 373 |
Feature activation | +1.07e+01 |
Pos logit contributions | |
yours | +0.054 |
beh | +0.049 |
anything | +0.048 |
possible | +0.048 |
ourselves | +0.048 |
Neg logit contributions | |
girl | -0.017 |
woman | -0.016 |
boy | -0.015 |
man | -0.015 |
sky | -0.013 |
Token | a |
Token ID | 257 |
Feature activation | +1.52e-02 |
Pos logit contributions | |
boy | +0.454 |
man | +0.433 |
girl | +0.431 |
little | +0.411 |
sky | +0.400 |
Neg logit contributions | |
yours | -0.624 |
beh | -0.618 |
ours | -0.530 |
possible | -0.529 |
theirs | -0.519 |
Token | little |
Token ID | 1310 |
Feature activation | +1.38e+01 |
Pos logit contributions | |
sky | +0.000 |
was | +0.000 |
led | +0.000 |
At | +0.000 |
felt | +0.000 |
Neg logit contributions | |
yours | -0.001 |
beh | -0.001 |
ours | -0.001 |
possible | -0.001 |
asty | -0.001 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.59e+01 |
Pos logit contributions | |
was | +0.191 |
At | +0.185 |
sky | +0.174 |
led | +0.129 |
little | +0.109 |
Neg logit contributions | |
yours | -1.057 |
beh | -0.984 |
ours | -0.935 |
possible | -0.904 |
ourselves | -0.891 |
Token | in |
Token ID | 259 |
Feature activation | +9.27e-01 |
Pos logit contributions | |
boy | +1.119 |
little | +1.083 |
girl | +1.063 |
man | +1.021 |
sky | +1.021 |
Neg logit contributions | |
beh | -1.562 |
yours | -1.472 |
possible | -1.298 |
ours | -1.243 |
ieve | -1.230 |
Token | . |
Token ID | 13 |
Feature activation | -8.10e+00 |
Pos logit contributions | |
boy | +0.034 |
girl | +0.033 |
little | +0.032 |
sky | +0.032 |
man | +0.031 |
Neg logit contributions | |
beh | -0.058 |
yours | -0.054 |
possible | -0.047 |
ourselves | -0.046 |
theirs | -0.045 |
Token | He |
Token ID | 679 |
Feature activation | +8.52e-02 |
Pos logit contributions | |
beh | +0.457 |
yours | +0.452 |
possible | +0.385 |
theirs | +0.379 |
ours | +0.377 |
Neg logit contributions | |
boy | -0.353 |
girl | -0.345 |
little | -0.343 |
man | -0.333 |
sky | -0.318 |
Token | was |
Token ID | 373 |
Feature activation | +9.49e+00 |
Pos logit contributions | |
girl | +0.002 |
boy | +0.002 |
sky | +0.002 |
woman | +0.002 |
man | +0.002 |
Neg logit contributions | |
beh | -0.006 |
yours | -0.006 |
supposed | -0.005 |
cks | -0.005 |
possible | -0.005 |
Token | at |
Token ID | 379 |
Feature activation | -4.23e+00 |
Pos logit contributions | |
boy | +0.369 |
girl | +0.355 |
man | +0.342 |
little | +0.340 |
sky | +0.336 |
Neg logit contributions | |
beh | -0.589 |
yours | -0.575 |
possible | -0.504 |
ours | -0.492 |
theirs | -0.482 |
Token |
|
Token ID | 198 |
Feature activation | +9.05e+00 |
Pos logit contributions | |
boy | +0.218 |
girl | +0.202 |
little | +0.197 |
man | +0.179 |
sky | +0.174 |
Neg logit contributions | |
beh | -0.735 |
yours | -0.719 |
possible | -0.652 |
ours | -0.633 |
theirs | -0.632 |
Token | They |
Token ID | 2990 |
Feature activation | -1.21e+00 |
Pos logit contributions | |
boy | +0.296 |
little | +0.281 |
girl | +0.275 |
sky | +0.261 |
man | +0.258 |
Neg logit contributions | |
beh | -0.643 |
yours | -0.613 |
possible | -0.552 |
ours | -0.534 |
ourselves | -0.529 |
Token | also |
Token ID | 635 |
Feature activation | +6.70e+00 |
Pos logit contributions | |
beh | +0.079 |
yours | +0.076 |
possible | +0.067 |
ours | +0.065 |
ieve | +0.065 |
Neg logit contributions | |
boy | -0.044 |
little | -0.042 |
girl | -0.041 |
sky | -0.039 |
man | -0.039 |
Token | realized |
Token ID | 6939 |
Feature activation | +5.53e+00 |
Pos logit contributions | |
boy | +0.230 |
little | +0.219 |
girl | +0.216 |
sky | +0.205 |
man | +0.200 |
Neg logit contributions | |
beh | -0.457 |
yours | -0.433 |
possible | -0.390 |
ours | -0.377 |
ieve | -0.375 |
Token | that |
Token ID | 326 |
Feature activation | -5.67e+00 |
Pos logit contributions | |
boy | +0.217 |
little | +0.217 |
girl | +0.201 |
man | +0.200 |
was | +0.195 |
Neg logit contributions | |
beh | -0.365 |
yours | -0.339 |
possible | -0.298 |
unnecessary | -0.295 |
ours | -0.294 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.58e+01 |
Pos logit contributions | |
beh | +0.362 |
yours | +0.342 |
possible | +0.302 |
ours | +0.294 |
unnecessary | +0.291 |
Neg logit contributions | |
boy | -0.230 |
little | -0.222 |
girl | -0.215 |
man | -0.206 |
sky | -0.203 |
Token | ins |
Token ID | 1040 |
Feature activation | -5.67e+00 |
Pos logit contributions | |
boy | +1.294 |
little | +1.253 |
girl | +1.229 |
man | +1.186 |
sky | +1.174 |
Neg logit contributions | |
beh | -1.366 |
yours | -1.302 |
possible | -1.136 |
ours | -1.085 |
unnecessary | -1.059 |
Token | were |
Token ID | 547 |
Feature activation | +6.99e+00 |
Pos logit contributions | |
beh | +0.438 |
yours | +0.422 |
possible | +0.385 |
ours | +0.374 |
unnecessary | +0.372 |
Neg logit contributions | |
boy | -0.149 |
little | -0.140 |
girl | -0.135 |
man | -0.127 |
sky | -0.124 |
Token | not |
Token ID | 407 |
Feature activation | -9.19e+00 |
Pos logit contributions | |
little | +0.302 |
boy | +0.285 |
girl | +0.265 |
was | +0.264 |
man | +0.263 |
Neg logit contributions | |
beh | -0.441 |
yours | -0.393 |
rying | -0.346 |
unnecessary | -0.345 |
possible | -0.345 |
Token | pets |
Token ID | 17252 |
Feature activation | +5.53e+00 |
Pos logit contributions | |
beh | +0.620 |
yours | +0.545 |
ieve | +0.507 |
rying | +0.497 |
eking | +0.495 |
Neg logit contributions | |
little | -0.376 |
boy | -0.353 |
girl | -0.329 |
man | -0.309 |
was | -0.301 |
Token | . |
Token ID | 13 |
Feature activation | -8.60e+00 |
Pos logit contributions | |
little | +0.252 |
boy | +0.249 |
girl | +0.234 |
was | +0.231 |
man | +0.229 |
Neg logit contributions | |
beh | -0.335 |
yours | -0.300 |
possible | -0.267 |
ours | -0.261 |
eking | -0.260 |
Token | slip |
Token ID | 13819 |
Feature activation | -4.59e+00 |
Pos logit contributions | |
little | +0.144 |
boy | +0.142 |
girl | +0.135 |
man | +0.132 |
was | +0.129 |
Neg logit contributions | |
beh | -0.205 |
yours | -0.194 |
possible | -0.171 |
ours | -0.168 |
unnecessary | -0.165 |
Token | . |
Token ID | 13 |
Feature activation | -7.45e+00 |
Pos logit contributions | |
beh | +0.335 |
yours | +0.306 |
ieve | +0.274 |
ours | +0.274 |
possible | +0.273 |
Neg logit contributions | |
boy | -0.157 |
little | -0.152 |
sky | -0.145 |
girl | -0.144 |
and | -0.143 |
Token | Quick |
Token ID | 12029 |
Feature activation | -1.76e-01 |
Pos logit contributions | |
beh | +0.519 |
yours | +0.501 |
possible | +0.448 |
ours | +0.433 |
theirs | +0.430 |
Neg logit contributions | |
boy | -0.237 |
little | -0.224 |
girl | -0.222 |
man | -0.210 |
sky | -0.200 |
Token | ly |
Token ID | 306 |
Feature activation | -1.33e+01 |
Pos logit contributions | |
beh | +0.011 |
yours | +0.010 |
possible | +0.009 |
unnecessary | +0.009 |
ieve | +0.009 |
Neg logit contributions | |
boy | -0.007 |
little | -0.007 |
girl | -0.007 |
sky | -0.007 |
man | -0.007 |
Token | the |
Token ID | 262 |
Feature activation | +7.28e+00 |
Pos logit contributions | |
beh | +0.894 |
yours | +0.844 |
possible | +0.771 |
eking | +0.742 |
ieve | +0.730 |
Neg logit contributions | |
boy | -0.386 |
sky | -0.359 |
girl | -0.353 |
little | -0.337 |
man | -0.305 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.56e+01 |
Pos logit contributions | |
boy | +0.139 |
little | +0.128 |
girl | +0.119 |
sky | +0.117 |
was | +0.112 |
Neg logit contributions | |
beh | -0.576 |
yours | -0.559 |
possible | -0.523 |
unnecessary | -0.499 |
ours | -0.496 |
Token | in |
Token ID | 259 |
Feature activation | +6.54e-01 |
Pos logit contributions | |
boy | +1.144 |
little | +1.108 |
girl | +1.084 |
man | +1.043 |
sky | +1.037 |
Neg logit contributions | |
beh | -1.516 |
yours | -1.431 |
possible | -1.264 |
ours | -1.210 |
unnecessary | -1.195 |
Token | tried |
Token ID | 3088 |
Feature activation | +1.89e+00 |
Pos logit contributions | |
girl | +0.018 |
boy | +0.017 |
sky | +0.017 |
little | +0.016 |
woman | +0.015 |
Neg logit contributions | |
beh | -0.046 |
yours | -0.039 |
eking | -0.037 |
rying | -0.037 |
ieve | -0.037 |
Token | to |
Token ID | 284 |
Feature activation | +1.78e+00 |
Pos logit contributions | |
boy | +0.073 |
little | +0.071 |
girl | +0.068 |
man | +0.066 |
sky | +0.065 |
Neg logit contributions | |
beh | -0.124 |
yours | -0.117 |
possible | -0.104 |
ours | -0.102 |
ieve | -0.100 |
Token | grab |
Token ID | 5552 |
Feature activation | -8.41e+00 |
Pos logit contributions | |
little | +0.079 |
boy | +0.076 |
man | +0.072 |
girl | +0.071 |
was | +0.070 |
Neg logit contributions | |
beh | -0.111 |
yours | -0.104 |
ours | -0.091 |
possible | -0.091 |
unnecessary | -0.088 |
Token | her |
Token ID | 607 |
Feature activation | -4.25e+00 |
Pos logit contributions | |
beh | +0.571 |
yours | +0.540 |
possible | +0.485 |
ours | +0.470 |
unnecessary | +0.462 |
Neg logit contributions | |
boy | -0.308 |
little | -0.297 |
girl | -0.288 |
man | -0.274 |
sky | -0.270 |
Token | a |
Token ID | 257 |
Feature activation | +2.89e-01 |
Pos logit contributions | |
boy | +0.107 |
little | +0.104 |
girl | +0.098 |
man | +0.093 |
sky | +0.089 |
Neg logit contributions | |
beh | -0.246 |
yours | -0.229 |
possible | -0.209 |
eking | -0.202 |
unnecessary | -0.201 |
Token | place |
Token ID | 1295 |
Feature activation | +5.11e+00 |
Pos logit contributions | |
sky | +0.007 |
boy | +0.007 |
girl | +0.007 |
was | +0.006 |
little | +0.006 |
Neg logit contributions | |
yours | -0.019 |
possible | -0.017 |
ours | -0.017 |
beh | -0.017 |
unnecessary | -0.016 |
Token | where |
Token ID | 810 |
Feature activation | -1.09e+01 |
Pos logit contributions | |
boy | +0.219 |
little | +0.212 |
girl | +0.206 |
man | +0.199 |
sky | +0.196 |
Neg logit contributions | |
beh | -0.318 |
yours | -0.296 |
possible | -0.261 |
unnecessary | -0.257 |
ieve | -0.252 |
Token | there |
Token ID | 612 |
Feature activation | -1.68e+00 |
Pos logit contributions | |
beh | +0.697 |
yours | +0.676 |
possible | +0.606 |
ours | +0.586 |
theirs | +0.577 |
Neg logit contributions | |
boy | -0.409 |
little | -0.393 |
girl | -0.387 |
man | -0.372 |
sky | -0.367 |
Token | are |
Token ID | 389 |
Feature activation | -5.79e-01 |
Pos logit contributions | |
beh | +0.123 |
yours | +0.121 |
possible | +0.109 |
ours | +0.106 |
theirs | +0.106 |
Neg logit contributions | |
boy | -0.045 |
little | -0.044 |
girl | -0.043 |
sky | -0.039 |
man | -0.039 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.56e+01 |
Pos logit contributions | |
beh | +0.037 |
yours | +0.036 |
possible | +0.032 |
ours | +0.031 |
theirs | +0.031 |
Neg logit contributions | |
boy | -0.021 |
girl | -0.020 |
little | -0.020 |
man | -0.020 |
sky | -0.019 |
Token | ins |
Token ID | 1040 |
Feature activation | -5.81e+00 |
Pos logit contributions | |
boy | +1.246 |
little | +1.205 |
girl | +1.177 |
man | +1.137 |
sky | +1.127 |
Neg logit contributions | |
beh | -1.406 |
yours | -1.345 |
possible | -1.183 |
ours | -1.127 |
theirs | -1.107 |
Token | . |
Token ID | 13 |
Feature activation | -8.36e+00 |
Pos logit contributions | |
beh | +0.393 |
yours | +0.364 |
unnecessary | +0.332 |
possible | +0.328 |
ieve | +0.317 |
Neg logit contributions | |
boy | -0.220 |
little | -0.208 |
girl | -0.203 |
man | -0.200 |
was | -0.198 |
Token | The |
Token ID | 383 |
Feature activation | +8.27e+00 |
Pos logit contributions | |
beh | +0.531 |
yours | +0.526 |
theirs | +0.463 |
ours | +0.461 |
possible | +0.458 |
Neg logit contributions | |
boy | -0.294 |
little | -0.284 |
girl | -0.283 |
man | -0.275 |
sky | -0.257 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.51e+01 |
Pos logit contributions | |
boy | +0.188 |
girl | +0.177 |
little | +0.175 |
sky | +0.165 |
was | +0.165 |
Neg logit contributions | |
beh | -0.581 |
yours | -0.579 |
possible | -0.526 |
ours | -0.514 |
ourselves | -0.504 |
Token | ins |
Token ID | 1040 |
Feature activation | -6.24e+00 |
Pos logit contributions | |
boy | +1.244 |
little | +1.220 |
girl | +1.193 |
sky | +1.152 |
man | +1.143 |
Neg logit contributions | |
beh | -1.354 |
yours | -1.265 |
possible | -1.094 |
ieve | -1.050 |
ours | -1.046 |
Token | Max |
Token ID | 5436 |
Feature activation | -1.79e+00 |
Pos logit contributions | |
boy | +0.649 |
little | +0.625 |
girl | +0.611 |
man | +0.585 |
sky | +0.579 |
Neg logit contributions | |
beh | -1.003 |
yours | -0.951 |
possible | -0.847 |
ours | -0.815 |
unnecessary | -0.805 |
Token | being |
Token ID | 852 |
Feature activation | -6.25e+00 |
Pos logit contributions | |
beh | +0.108 |
yours | +0.104 |
possible | +0.093 |
ours | +0.089 |
unnecessary | +0.087 |
Neg logit contributions | |
boy | -0.079 |
little | -0.078 |
girl | -0.074 |
was | -0.071 |
man | -0.071 |
Token | attacked |
Token ID | 7384 |
Feature activation | -5.04e+00 |
Pos logit contributions | |
beh | +0.407 |
yours | +0.388 |
possible | +0.348 |
ours | +0.334 |
unnecessary | +0.331 |
Neg logit contributions | |
boy | -0.244 |
little | -0.236 |
girl | -0.229 |
man | -0.218 |
sky | -0.215 |
Token | by |
Token ID | 416 |
Feature activation | +1.60e+00 |
Pos logit contributions | |
beh | +0.280 |
yours | +0.258 |
possible | +0.235 |
ours | +0.221 |
ieve | +0.221 |
Neg logit contributions | |
boy | -0.253 |
little | -0.248 |
girl | -0.242 |
man | -0.228 |
was | -0.226 |
Token | the |
Token ID | 262 |
Feature activation | +7.28e+00 |
Pos logit contributions | |
little | +0.067 |
boy | +0.066 |
girl | +0.062 |
man | +0.059 |
was | +0.058 |
Neg logit contributions | |
beh | -0.102 |
yours | -0.095 |
possible | -0.086 |
ours | -0.082 |
ieve | -0.082 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.56e+01 |
Pos logit contributions | |
boy | +0.178 |
little | +0.169 |
girl | +0.163 |
man | +0.150 |
sky | +0.149 |
Neg logit contributions | |
beh | -0.544 |
yours | -0.534 |
possible | -0.487 |
ours | -0.475 |
unnecessary | -0.468 |
Token | ins |
Token ID | 1040 |
Feature activation | -5.81e+00 |
Pos logit contributions | |
boy | +1.130 |
little | +1.081 |
girl | +1.061 |
man | +1.023 |
sky | +1.001 |
Neg logit contributions | |
beh | -1.533 |
yours | -1.470 |
possible | -1.309 |
ours | -1.257 |
unnecessary | -1.238 |
Token | . |
Token ID | 13 |
Feature activation | -8.36e+00 |
Pos logit contributions | |
beh | +0.379 |
yours | +0.360 |
possible | +0.322 |
unnecessary | +0.319 |
ours | +0.312 |
Neg logit contributions | |
boy | -0.232 |
little | -0.225 |
man | -0.216 |
girl | -0.213 |
was | -0.213 |
Token | They |
Token ID | 1119 |
Feature activation | +1.47e+00 |
Pos logit contributions | |
beh | +0.529 |
yours | +0.525 |
theirs | +0.462 |
ours | +0.459 |
possible | +0.454 |
Neg logit contributions | |
boy | -0.277 |
girl | -0.270 |
little | -0.268 |
man | -0.250 |
sky | -0.241 |
Token | were |
Token ID | 547 |
Feature activation | +6.64e+00 |
Pos logit contributions | |
boy | +0.048 |
girl | +0.047 |
sky | +0.046 |
little | +0.044 |
man | +0.042 |
Neg logit contributions | |
beh | -0.097 |
yours | -0.089 |
ours | -0.080 |
theirs | -0.079 |
eking | -0.079 |
Token | very |
Token ID | 845 |
Feature activation | +5.76e+00 |
Pos logit contributions | |
little | +0.252 |
boy | +0.246 |
girl | +0.224 |
man | +0.219 |
sky | +0.216 |
Neg logit contributions | |
beh | -0.439 |
yours | -0.421 |
possible | -0.382 |
ieve | -0.363 |
ourselves | -0.360 |
Token | € |
Token ID | 26391 |
Feature activation | -8.83e+00 |
Pos logit contributions | |
boy | +0.014 |
little | +0.014 |
girl | +0.014 |
man | +0.013 |
sky | +0.013 |
Neg logit contributions | |
beh | -0.011 |
yours | -0.010 |
possible | -0.009 |
ours | -0.009 |
unnecessary | -0.008 |
Token | |
Token ID | 220 |
Feature activation | +3.45e+00 |
Pos logit contributions | |
beh | +0.652 |
yours | +0.621 |
possible | +0.565 |
ours | +0.548 |
ourselves | +0.539 |
Neg logit contributions | |
boy | -0.257 |
girl | -0.242 |
little | -0.242 |
man | -0.216 |
sky | -0.216 |
Token |
|
Token ID | 198 |
Feature activation | +8.32e+00 |
Pos logit contributions | |
boy | +0.044 |
little | +0.041 |
girl | +0.035 |
sky | +0.032 |
man | +0.031 |
Neg logit contributions | |
beh | -0.312 |
yours | -0.301 |
possible | -0.278 |
ours | -0.273 |
unnecessary | -0.272 |
Token |
|
Token ID | 198 |
Feature activation | +8.14e+00 |
Pos logit contributions | |
boy | +0.171 |
girl | +0.155 |
little | +0.149 |
man | +0.133 |
sky | +0.128 |
Neg logit contributions | |
beh | -0.685 |
yours | -0.663 |
possible | -0.609 |
ourselves | -0.585 |
ours | -0.584 |
Token | The |
Token ID | 464 |
Feature activation | +1.27e+01 |
Pos logit contributions | |
boy | +0.268 |
little | +0.251 |
girl | +0.249 |
sky | +0.235 |
man | +0.232 |
Neg logit contributions | |
beh | -0.583 |
yours | -0.550 |
possible | -0.499 |
ours | -0.475 |
ourselves | -0.474 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.56e+01 |
Pos logit contributions | |
boy | +0.356 |
little | +0.332 |
girl | +0.324 |
sky | +0.312 |
man | +0.304 |
Neg logit contributions | |
beh | -0.903 |
yours | -0.874 |
possible | -0.806 |
unnecessary | -0.763 |
ours | -0.758 |
Token | in |
Token ID | 259 |
Feature activation | +6.54e-01 |
Pos logit contributions | |
boy | +1.184 |
little | +1.157 |
girl | +1.130 |
sky | +1.102 |
man | +1.087 |
Neg logit contributions | |
beh | -1.468 |
yours | -1.342 |
possible | -1.185 |
ieve | -1.149 |
eking | -1.142 |
Token | was |
Token ID | 373 |
Feature activation | +9.49e+00 |
Pos logit contributions | |
boy | +0.017 |
girl | +0.017 |
little | +0.017 |
sky | +0.016 |
man | +0.014 |
Neg logit contributions | |
beh | -0.045 |
yours | -0.040 |
ieve | -0.037 |
eking | -0.037 |
rying | -0.036 |
Token | happy |
Token ID | 3772 |
Feature activation | +5.46e+00 |
Pos logit contributions | |
boy | +0.370 |
little | +0.370 |
girl | +0.346 |
man | +0.337 |
sky | +0.336 |
Neg logit contributions | |
beh | -0.612 |
yours | -0.577 |
possible | -0.511 |
ours | -0.498 |
unnecessary | -0.496 |
Token | , |
Token ID | 11 |
Feature activation | +1.13e+01 |
Pos logit contributions | |
boy | +0.178 |
little | +0.171 |
girl | +0.164 |
man | +0.156 |
sky | +0.152 |
Neg logit contributions | |
beh | -0.400 |
yours | -0.377 |
possible | -0.339 |
ours | -0.336 |
unnecessary | -0.335 |
Token | so |
Token ID | 523 |
Feature activation | +3.38e-01 |
Pos logit contributions | |
boy | +0.358 |
little | +0.337 |
girl | +0.332 |
man | +0.313 |
sky | +0.312 |
Neg logit contributions | |
beh | -0.795 |
yours | -0.770 |
possible | -0.698 |
ours | -0.671 |
theirs | -0.663 |
Token | Why |
Token ID | 5195 |
Feature activation | -6.28e+00 |
Pos logit contributions | |
boy | +0.427 |
little | +0.425 |
girl | +0.401 |
sky | +0.399 |
man | +0.388 |
Neg logit contributions | |
beh | -0.697 |
yours | -0.648 |
possible | -0.589 |
ourselves | -0.552 |
ieve | -0.549 |
Token | did |
Token ID | 750 |
Feature activation | +5.88e+00 |
Pos logit contributions | |
beh | +0.390 |
yours | +0.367 |
possible | +0.327 |
ours | +0.314 |
unnecessary | +0.311 |
Neg logit contributions | |
boy | -0.266 |
little | -0.258 |
girl | -0.250 |
man | -0.241 |
sky | -0.237 |
Token | you |
Token ID | 345 |
Feature activation | -1.50e+01 |
Pos logit contributions | |
little | +0.330 |
boy | +0.324 |
man | +0.305 |
girl | +0.298 |
was | +0.297 |
Neg logit contributions | |
beh | -0.301 |
rying | -0.259 |
yours | -0.257 |
eking | -0.230 |
possible | -0.227 |
Token | rip |
Token ID | 18344 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
beh | +0.964 |
yours | +0.890 |
possible | +0.791 |
unnecessary | +0.786 |
ours | +0.773 |
Neg logit contributions | |
little | -0.652 |
boy | -0.624 |
man | -0.589 |
was | -0.570 |
girl | -0.569 |
Token | the |
Token ID | 262 |
Feature activation | +7.28e+00 |
Pos logit contributions | |
beh | +0.079 |
yours | +0.071 |
possible | +0.065 |
unnecessary | +0.061 |
theirs | +0.061 |
Neg logit contributions | |
boy | -0.055 |
little | -0.054 |
girl | -0.052 |
man | -0.050 |
sky | -0.049 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.56e+01 |
Pos logit contributions | |
boy | +0.238 |
little | +0.225 |
girl | +0.221 |
man | +0.208 |
sky | +0.205 |
Neg logit contributions | |
beh | -0.500 |
yours | -0.489 |
possible | -0.438 |
ours | -0.426 |
unnecessary | -0.417 |
Token | in |
Token ID | 259 |
Feature activation | +6.54e-01 |
Pos logit contributions | |
boy | +1.043 |
little | +0.996 |
girl | +0.968 |
man | +0.939 |
sky | +0.905 |
Neg logit contributions | |
beh | -1.613 |
yours | -1.542 |
possible | -1.387 |
ours | -1.338 |
unnecessary | -1.336 |
Token | ?" |
Token ID | 1701 |
Feature activation | -8.91e+00 |
Pos logit contributions | |
boy | +0.034 |
little | +0.033 |
girl | +0.032 |
man | +0.031 |
sky | +0.030 |
Neg logit contributions | |
beh | -0.035 |
yours | -0.032 |
possible | -0.028 |
unnecessary | -0.027 |
theirs | -0.027 |
Token | Mom |
Token ID | 11254 |
Feature activation | +4.74e+00 |
Pos logit contributions | |
beh | +0.568 |
yours | +0.543 |
possible | +0.486 |
ours | +0.465 |
theirs | +0.460 |
Neg logit contributions | |
boy | -0.339 |
little | -0.325 |
girl | -0.320 |
sky | -0.306 |
man | -0.302 |
Token | asks |
Token ID | 7893 |
Feature activation | -2.19e-01 |
Pos logit contributions | |
boy | +0.158 |
little | +0.151 |
girl | +0.147 |
man | +0.139 |
sky | +0.137 |
Neg logit contributions | |
beh | -0.334 |
yours | -0.320 |
possible | -0.289 |
ours | -0.279 |
unnecessary | -0.277 |
Token | . |
Token ID | 13 |
Feature activation | -8.67e+00 |
Pos logit contributions | |
beh | +0.015 |
yours | +0.014 |
possible | +0.013 |
ours | +0.012 |
unnecessary | +0.012 |
Neg logit contributions | |
little | -0.009 |
boy | -0.009 |
girl | -0.008 |
man | -0.008 |
and | -0.008 |
Token | , |
Token ID | 11 |
Feature activation | +1.21e+01 |
Pos logit contributions | |
boy | +0.172 |
girl | +0.163 |
man | +0.157 |
sky | +0.149 |
little | +0.142 |
Neg logit contributions | |
yours | -0.402 |
beh | -0.399 |
possible | -0.365 |
ours | -0.348 |
ourselves | -0.346 |
Token | there |
Token ID | 612 |
Feature activation | -1.84e+00 |
Pos logit contributions | |
boy | +0.544 |
man | +0.539 |
sky | +0.525 |
girl | +0.522 |
little | +0.505 |
Neg logit contributions | |
yours | -0.632 |
beh | -0.606 |
possible | -0.553 |
ours | -0.527 |
eking | -0.509 |
Token | was |
Token ID | 373 |
Feature activation | +9.75e+00 |
Pos logit contributions | |
yours | +0.118 |
beh | +0.108 |
possible | +0.108 |
anything | +0.107 |
ourselves | +0.105 |
Neg logit contributions | |
girl | -0.044 |
woman | -0.044 |
man | -0.041 |
boy | -0.040 |
sky | -0.037 |
Token | a |
Token ID | 257 |
Feature activation | -7.79e-01 |
Pos logit contributions | |
boy | +0.430 |
man | +0.414 |
girl | +0.409 |
little | +0.385 |
sky | +0.378 |
Neg logit contributions | |
yours | -0.554 |
beh | -0.545 |
possible | -0.472 |
ours | -0.464 |
theirs | -0.457 |
Token | little |
Token ID | 1310 |
Feature activation | +1.31e+01 |
Pos logit contributions | |
yours | +0.051 |
ours | +0.046 |
beh | +0.046 |
possible | +0.045 |
asty | +0.043 |
Neg logit contributions | |
sky | -0.015 |
was | -0.011 |
led | -0.010 |
At | -0.009 |
gasped | -0.006 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.53e+01 |
Pos logit contributions | |
was | +0.169 |
At | +0.166 |
sky | +0.153 |
led | +0.098 |
little | +0.093 |
Neg logit contributions | |
yours | -1.016 |
beh | -0.955 |
ours | -0.904 |
possible | -0.885 |
ourselves | -0.857 |
Token | in |
Token ID | 259 |
Feature activation | +2.97e-01 |
Pos logit contributions | |
boy | +1.094 |
little | +1.058 |
girl | +1.038 |
man | +0.998 |
sky | +0.996 |
Neg logit contributions | |
beh | -1.523 |
yours | -1.432 |
possible | -1.264 |
ours | -1.209 |
ieve | -1.201 |
Token | named |
Token ID | 3706 |
Feature activation | +5.70e+00 |
Pos logit contributions | |
boy | +0.011 |
girl | +0.011 |
little | +0.010 |
sky | +0.010 |
man | +0.010 |
Neg logit contributions | |
beh | -0.018 |
yours | -0.017 |
possible | -0.015 |
ourselves | -0.014 |
ieve | -0.014 |
Token | P |
Token ID | 350 |
Feature activation | +1.91e+00 |
Pos logit contributions | |
boy | +0.249 |
little | +0.246 |
girl | +0.232 |
man | +0.226 |
sky | +0.224 |
Neg logit contributions | |
beh | -0.355 |
yours | -0.321 |
possible | -0.292 |
unnecessary | -0.281 |
ieve | -0.279 |
Token | ingu |
Token ID | 6680 |
Feature activation | -5.62e-01 |
Pos logit contributions | |
boy | +0.101 |
little | +0.098 |
girl | +0.097 |
man | +0.093 |
sky | +0.093 |
Neg logit contributions | |
beh | -0.097 |
yours | -0.091 |
possible | -0.078 |
ours | -0.075 |
unnecessary | -0.073 |
Token | who |
Token ID | 508 |
Feature activation | -4.43e+00 |
Pos logit contributions | |
beh | +0.030 |
yours | +0.029 |
possible | +0.025 |
ours | +0.024 |
theirs | +0.023 |
Neg logit contributions | |
boy | -0.028 |
little | -0.027 |
girl | -0.026 |
sky | -0.025 |
man | -0.025 |
Token | , |
Token ID | 11 |
Feature activation | +1.21e+01 |
Pos logit contributions | |
boy | +0.482 |
little | +0.456 |
girl | +0.445 |
sky | +0.423 |
man | +0.422 |
Neg logit contributions | |
beh | -1.128 |
yours | -1.094 |
possible | -0.982 |
ours | -0.952 |
unnecessary | -0.934 |
Token | can |
Token ID | 460 |
Feature activation | +3.78e+00 |
Pos logit contributions | |
boy | +0.543 |
little | +0.525 |
girl | +0.514 |
man | +0.495 |
sky | +0.492 |
Neg logit contributions | |
beh | -0.708 |
yours | -0.671 |
possible | -0.591 |
ours | -0.567 |
unnecessary | -0.558 |
Token | I |
Token ID | 314 |
Feature activation | -1.03e+01 |
Pos logit contributions | |
boy | +0.176 |
little | +0.174 |
girl | +0.164 |
was | +0.161 |
man | +0.160 |
Neg logit contributions | |
beh | -0.229 |
yours | -0.205 |
unnecessary | -0.190 |
possible | -0.187 |
theirs | -0.181 |
Token | touch |
Token ID | 3638 |
Feature activation | -5.91e+00 |
Pos logit contributions | |
beh | +0.625 |
yours | +0.578 |
possible | +0.516 |
unnecessary | +0.500 |
ours | +0.499 |
Neg logit contributions | |
boy | -0.463 |
little | -0.462 |
girl | -0.441 |
man | -0.424 |
sky | -0.416 |
Token | the |
Token ID | 262 |
Feature activation | +6.89e+00 |
Pos logit contributions | |
beh | +0.405 |
yours | +0.374 |
possible | +0.342 |
ours | +0.326 |
unnecessary | +0.326 |
Neg logit contributions | |
boy | -0.217 |
little | -0.212 |
girl | -0.203 |
sky | -0.195 |
man | -0.193 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.53e+01 |
Pos logit contributions | |
boy | +0.191 |
little | +0.176 |
girl | +0.173 |
man | +0.162 |
sky | +0.155 |
Neg logit contributions | |
beh | -0.501 |
yours | -0.496 |
possible | -0.449 |
ours | -0.438 |
unnecessary | -0.430 |
Token | in |
Token ID | 259 |
Feature activation | +2.97e-01 |
Pos logit contributions | |
boy | +0.978 |
little | +0.923 |
girl | +0.906 |
man | +0.878 |
sky | +0.841 |
Neg logit contributions | |
beh | -1.644 |
yours | -1.563 |
possible | -1.417 |
unnecessary | -1.378 |
ours | -1.367 |
Token | ?" |
Token ID | 1701 |
Feature activation | -9.22e+00 |
Pos logit contributions | |
boy | +0.015 |
little | +0.014 |
girl | +0.014 |
man | +0.014 |
sky | +0.013 |
Neg logit contributions | |
beh | -0.017 |
yours | -0.015 |
unnecessary | -0.013 |
possible | -0.013 |
theirs | -0.013 |
Token | Tim |
Token ID | 5045 |
Feature activation | -5.51e+00 |
Pos logit contributions | |
beh | +0.559 |
yours | +0.536 |
possible | +0.474 |
ours | +0.456 |
theirs | +0.451 |
Neg logit contributions | |
boy | -0.375 |
little | -0.362 |
girl | -0.354 |
sky | -0.345 |
man | -0.338 |
Token | asked |
Token ID | 1965 |
Feature activation | +1.19e+00 |
Pos logit contributions | |
beh | +0.352 |
yours | +0.328 |
possible | +0.288 |
eking | +0.280 |
rying | +0.278 |
Neg logit contributions | |
boy | -0.201 |
little | -0.194 |
sky | -0.191 |
girl | -0.190 |
man | -0.181 |
Token | . |
Token ID | 13 |
Feature activation | -8.94e+00 |
Pos logit contributions | |
boy | +0.048 |
little | +0.048 |
girl | +0.045 |
man | +0.043 |
was | +0.042 |
Neg logit contributions | |
beh | -0.080 |
yours | -0.076 |
ours | -0.068 |
unnecessary | -0.067 |
possible | -0.067 |
Token | are |
Token ID | 389 |
Feature activation | -5.79e-01 |
Pos logit contributions | |
beh | +0.123 |
yours | +0.121 |
possible | +0.109 |
ours | +0.106 |
theirs | +0.106 |
Neg logit contributions | |
boy | -0.045 |
little | -0.044 |
girl | -0.043 |
sky | -0.039 |
man | -0.039 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.56e+01 |
Pos logit contributions | |
beh | +0.037 |
yours | +0.036 |
possible | +0.032 |
ours | +0.031 |
theirs | +0.031 |
Neg logit contributions | |
boy | -0.021 |
girl | -0.020 |
little | -0.020 |
man | -0.020 |
sky | -0.019 |
Token | ins |
Token ID | 1040 |
Feature activation | -5.81e+00 |
Pos logit contributions | |
boy | +1.246 |
little | +1.205 |
girl | +1.177 |
man | +1.137 |
sky | +1.127 |
Neg logit contributions | |
beh | -1.406 |
yours | -1.345 |
possible | -1.183 |
ours | -1.127 |
theirs | -1.107 |
Token | . |
Token ID | 13 |
Feature activation | -8.36e+00 |
Pos logit contributions | |
beh | +0.393 |
yours | +0.364 |
unnecessary | +0.332 |
possible | +0.328 |
ieve | +0.317 |
Neg logit contributions | |
boy | -0.220 |
little | -0.208 |
girl | -0.203 |
man | -0.200 |
was | -0.198 |
Token | The |
Token ID | 383 |
Feature activation | +8.27e+00 |
Pos logit contributions | |
beh | +0.531 |
yours | +0.526 |
theirs | +0.463 |
ours | +0.461 |
possible | +0.458 |
Neg logit contributions | |
boy | -0.294 |
little | -0.284 |
girl | -0.283 |
man | -0.275 |
sky | -0.257 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.51e+01 |
Pos logit contributions | |
boy | +0.188 |
girl | +0.177 |
little | +0.175 |
sky | +0.165 |
was | +0.165 |
Neg logit contributions | |
beh | -0.581 |
yours | -0.579 |
possible | -0.526 |
ours | -0.514 |
ourselves | -0.504 |
Token | ins |
Token ID | 1040 |
Feature activation | -6.24e+00 |
Pos logit contributions | |
boy | +1.244 |
little | +1.220 |
girl | +1.193 |
sky | +1.152 |
man | +1.143 |
Neg logit contributions | |
beh | -1.354 |
yours | -1.265 |
possible | -1.094 |
ieve | -1.050 |
ours | -1.046 |
Token | are |
Token ID | 389 |
Feature activation | -1.38e+00 |
Pos logit contributions | |
beh | +0.383 |
yours | +0.366 |
possible | +0.323 |
ours | +0.312 |
ieve | +0.312 |
Neg logit contributions | |
boy | -0.248 |
little | -0.242 |
girl | -0.238 |
sky | -0.221 |
man | -0.219 |
Token | black |
Token ID | 2042 |
Feature activation | +8.86e+00 |
Pos logit contributions | |
beh | +0.095 |
yours | +0.091 |
possible | +0.082 |
ours | +0.079 |
unnecessary | +0.078 |
Neg logit contributions | |
boy | -0.048 |
little | -0.045 |
girl | -0.044 |
man | -0.042 |
sky | -0.042 |
Token | and |
Token ID | 290 |
Feature activation | +1.39e+01 |
Pos logit contributions | |
boy | +0.140 |
little | +0.124 |
girl | +0.123 |
man | +0.111 |
sky | +0.095 |
Neg logit contributions | |
beh | -0.739 |
yours | -0.730 |
possible | -0.653 |
ours | -0.649 |
theirs | -0.643 |
Token | white |
Token ID | 2330 |
Feature activation | +1.16e+01 |
Pos logit contributions | |
boy | +0.352 |
little | +0.333 |
girl | +0.318 |
man | +0.297 |
sky | +0.293 |
Neg logit contributions | |
beh | -1.094 |
yours | -1.042 |
possible | -0.953 |
ours | -0.923 |
unnecessary | -0.920 |
Token | ily |
Token ID | 813 |
Feature activation | -7.42e-02 |
Pos logit contributions | |
boy | +0.581 |
little | +0.571 |
girl | +0.559 |
man | +0.536 |
sky | +0.529 |
Neg logit contributions | |
beh | -0.613 |
yours | -0.569 |
possible | -0.502 |
ours | -0.480 |
unnecessary | -0.479 |
Token | wanted |
Token ID | 2227 |
Feature activation | +1.22e+00 |
Pos logit contributions | |
beh | +0.005 |
yours | +0.005 |
properly | +0.004 |
ieve | +0.004 |
myself | +0.004 |
Neg logit contributions | |
boy | -0.002 |
girl | -0.002 |
sky | -0.002 |
little | -0.002 |
sudden | -0.001 |
Token | to |
Token ID | 284 |
Feature activation | +2.00e+00 |
Pos logit contributions | |
boy | +0.048 |
little | +0.047 |
girl | +0.045 |
man | +0.043 |
sky | +0.043 |
Neg logit contributions | |
beh | -0.080 |
yours | -0.075 |
possible | -0.067 |
ours | -0.065 |
ieve | -0.064 |
Token | see |
Token ID | 766 |
Feature activation | -5.13e+00 |
Pos logit contributions | |
little | +0.088 |
boy | +0.086 |
man | +0.082 |
girl | +0.080 |
was | +0.078 |
Neg logit contributions | |
beh | -0.124 |
yours | -0.112 |
possible | -0.101 |
unnecessary | -0.101 |
eking | -0.101 |
Token | the |
Token ID | 262 |
Feature activation | +6.79e+00 |
Pos logit contributions | |
beh | +0.346 |
yours | +0.306 |
possible | +0.281 |
unnecessary | +0.267 |
rying | +0.267 |
Neg logit contributions | |
little | -0.208 |
boy | -0.203 |
girl | -0.192 |
man | -0.185 |
sky | -0.175 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.51e+01 |
Pos logit contributions | |
boy | +0.188 |
little | +0.176 |
girl | +0.173 |
sky | +0.162 |
was | +0.160 |
Neg logit contributions | |
beh | -0.483 |
yours | -0.474 |
possible | -0.433 |
ours | -0.421 |
unnecessary | -0.410 |
Token | ins |
Token ID | 1040 |
Feature activation | -6.24e+00 |
Pos logit contributions | |
boy | +1.153 |
little | +1.110 |
girl | +1.090 |
man | +1.052 |
sky | +1.032 |
Neg logit contributions | |
beh | -1.448 |
yours | -1.385 |
possible | -1.224 |
ours | -1.177 |
unnecessary | -1.163 |
Token | . |
Token ID | 13 |
Feature activation | -8.76e+00 |
Pos logit contributions | |
beh | +0.461 |
yours | +0.422 |
unnecessary | +0.399 |
possible | +0.379 |
ours | +0.376 |
Neg logit contributions | |
boy | -0.204 |
little | -0.191 |
girl | -0.186 |
and | -0.185 |
man | -0.185 |
Token | She |
Token ID | 1375 |
Feature activation | +1.32e+00 |
Pos logit contributions | |
beh | +0.551 |
yours | +0.533 |
ours | +0.460 |
possible | +0.459 |
ourselves | +0.456 |
Neg logit contributions | |
boy | -0.319 |
little | -0.312 |
girl | -0.298 |
man | -0.292 |
sky | -0.291 |
Token | asked |
Token ID | 1965 |
Feature activation | +1.11e+00 |
Pos logit contributions | |
sky | +0.053 |
girl | +0.046 |
boy | +0.046 |
little | +0.044 |
man | +0.041 |
Neg logit contributions | |
beh | -0.073 |
yours | -0.066 |
eking | -0.058 |
ck | -0.057 |
rying | -0.056 |
Token | her |
Token ID | 607 |
Feature activation | -4.59e+00 |
Pos logit contributions | |
boy | +0.048 |
little | +0.047 |
girl | +0.046 |
man | +0.044 |
sky | +0.043 |
Neg logit contributions | |
beh | -0.068 |
yours | -0.064 |
possible | -0.057 |
ours | -0.055 |
unnecessary | -0.054 |
Token | 't |
Token ID | 470 |
Feature activation | -5.30e+00 |
Pos logit contributions | |
beh | +0.570 |
yours | +0.527 |
possible | +0.456 |
unnecessary | +0.449 |
ours | +0.436 |
Neg logit contributions | |
boy | -0.546 |
little | -0.545 |
man | -0.518 |
girl | -0.516 |
sky | -0.515 |
Token | know |
Token ID | 760 |
Feature activation | -9.56e+00 |
Pos logit contributions | |
beh | +0.284 |
yours | +0.264 |
possible | +0.239 |
eking | +0.232 |
ourselves | +0.230 |
Neg logit contributions | |
little | -0.271 |
boy | -0.266 |
girl | -0.256 |
man | -0.241 |
and | -0.239 |
Token | . |
Token ID | 13 |
Feature activation | -8.36e+00 |
Pos logit contributions | |
beh | +0.730 |
yours | +0.635 |
possible | +0.600 |
unnecessary | +0.591 |
ieve | +0.581 |
Neg logit contributions | |
little | -0.319 |
boy | -0.303 |
was | -0.288 |
girl | -0.278 |
and | -0.272 |
Token | Maybe |
Token ID | 6674 |
Feature activation | -2.43e+00 |
Pos logit contributions | |
beh | +0.500 |
yours | +0.473 |
possible | +0.419 |
ours | +0.402 |
unnecessary | +0.396 |
Neg logit contributions | |
boy | -0.367 |
little | -0.355 |
girl | -0.348 |
man | -0.334 |
sky | -0.332 |
Token | a |
Token ID | 257 |
Feature activation | -1.14e+00 |
Pos logit contributions | |
beh | +0.159 |
yours | +0.141 |
possible | +0.126 |
unnecessary | +0.125 |
ieve | +0.124 |
Neg logit contributions | |
boy | -0.104 |
little | -0.102 |
girl | -0.098 |
man | -0.094 |
was | -0.091 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.51e+01 |
Pos logit contributions | |
yours | +0.078 |
beh | +0.077 |
possible | +0.071 |
ours | +0.067 |
unnecessary | +0.067 |
Neg logit contributions | |
boy | -0.032 |
sky | -0.030 |
little | -0.029 |
girl | -0.029 |
man | -0.027 |
Token | in |
Token ID | 259 |
Feature activation | +1.30e-01 |
Pos logit contributions | |
boy | +0.898 |
girl | +0.836 |
little | +0.771 |
man | +0.745 |
was | +0.695 |
Neg logit contributions | |
beh | -1.727 |
yours | -1.655 |
possible | -1.544 |
ours | -1.507 |
unnecessary | -1.479 |
Token | ?" |
Token ID | 1701 |
Feature activation | -9.42e+00 |
Pos logit contributions | |
boy | +0.005 |
little | +0.005 |
girl | +0.005 |
man | +0.005 |
was | +0.005 |
Neg logit contributions | |
beh | -0.009 |
yours | -0.008 |
possible | -0.007 |
unnecessary | -0.007 |
ours | -0.007 |
Token | Tom |
Token ID | 4186 |
Feature activation | -2.91e+00 |
Pos logit contributions | |
beh | +0.570 |
yours | +0.549 |
possible | +0.485 |
ours | +0.468 |
theirs | +0.463 |
Neg logit contributions | |
boy | -0.385 |
little | -0.371 |
girl | -0.366 |
sky | -0.351 |
man | -0.344 |
Token | guesses |
Token ID | 44774 |
Feature activation | -8.24e+00 |
Pos logit contributions | |
beh | +0.173 |
yours | +0.162 |
possible | +0.140 |
ieve | +0.138 |
ours | +0.136 |
Neg logit contributions | |
boy | -0.119 |
little | -0.116 |
girl | -0.113 |
sky | -0.112 |
man | -0.106 |
Token | . |
Token ID | 13 |
Feature activation | -9.07e+00 |
Pos logit contributions | |
beh | +0.583 |
yours | +0.552 |
ours | +0.502 |
unnecessary | +0.491 |
possible | +0.487 |
Neg logit contributions | |
boy | -0.317 |
little | -0.309 |
man | -0.297 |
was | -0.296 |
and | -0.294 |
Token |
|
Token ID | 198 |
Feature activation | +7.85e+00 |
Pos logit contributions | |
beh | +0.566 |
yours | +0.552 |
possible | +0.495 |
ours | +0.479 |
theirs | +0.475 |
Neg logit contributions | |
boy | -0.240 |
little | -0.229 |
girl | -0.228 |
man | -0.213 |
sky | -0.202 |
Token |
|
Token ID | 198 |
Feature activation | +7.71e+00 |
Pos logit contributions | |
boy | +0.233 |
girl | +0.226 |
little | +0.211 |
sky | +0.198 |
man | +0.197 |
Neg logit contributions | |
beh | -0.558 |
yours | -0.545 |
possible | -0.499 |
theirs | -0.475 |
ourselves | -0.468 |
Token | The |
Token ID | 464 |
Feature activation | +1.24e+01 |
Pos logit contributions | |
boy | +0.277 |
girl | +0.263 |
little | +0.261 |
sky | +0.246 |
man | +0.242 |
Neg logit contributions | |
beh | -0.524 |
yours | -0.494 |
possible | -0.450 |
ourselves | -0.421 |
ours | -0.421 |
Token | little |
Token ID | 1310 |
Feature activation | +1.30e+01 |
Pos logit contributions | |
boy | +0.270 |
little | +0.243 |
was | +0.242 |
girl | +0.241 |
sky | +0.238 |
Neg logit contributions | |
beh | -0.895 |
yours | -0.877 |
possible | -0.820 |
unnecessary | -0.777 |
ours | -0.769 |
Token | sad |
Token ID | 6507 |
Feature activation | +2.15e+00 |
Pos logit contributions | |
sky | +0.066 |
little | +0.061 |
was | +0.053 |
man | +0.048 |
boy | +0.038 |
Neg logit contributions | |
beh | -1.123 |
yours | -1.113 |
possible | -1.027 |
ours | -0.985 |
ourselves | -0.980 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.51e+01 |
Pos logit contributions | |
boy | +0.050 |
little | +0.046 |
girl | +0.045 |
sky | +0.044 |
man | +0.042 |
Neg logit contributions | |
beh | -0.160 |
yours | -0.156 |
possible | -0.143 |
theirs | -0.136 |
ours | -0.134 |
Token | in |
Token ID | 259 |
Feature activation | +1.30e-01 |
Pos logit contributions | |
boy | +1.088 |
little | +1.053 |
girl | +1.030 |
man | +0.992 |
sky | +0.990 |
Neg logit contributions | |
beh | -1.513 |
yours | -1.423 |
possible | -1.261 |
ours | -1.204 |
ieve | -1.194 |
Token | was |
Token ID | 373 |
Feature activation | +8.98e+00 |
Pos logit contributions | |
boy | +0.003 |
girl | +0.003 |
little | +0.003 |
sky | +0.003 |
man | +0.003 |
Neg logit contributions | |
beh | -0.010 |
yours | -0.009 |
possible | -0.008 |
ieve | -0.008 |
rying | -0.008 |
Token | no |
Token ID | 645 |
Feature activation | -4.74e+00 |
Pos logit contributions | |
boy | +0.360 |
little | +0.359 |
girl | +0.336 |
man | +0.326 |
sky | +0.326 |
Neg logit contributions | |
beh | -0.572 |
yours | -0.537 |
possible | -0.473 |
ours | -0.465 |
unnecessary | -0.462 |
Token | longer |
Token ID | 2392 |
Feature activation | -1.14e+01 |
Pos logit contributions | |
beh | +0.219 |
yours | +0.202 |
possible | +0.166 |
ours | +0.164 |
ourselves | +0.157 |
Neg logit contributions | |
boy | -0.285 |
little | -0.282 |
girl | -0.273 |
man | -0.261 |
sky | -0.257 |
Token | alone |
Token ID | 3436 |
Feature activation | -1.20e+01 |
Pos logit contributions | |
beh | +0.724 |
yours | +0.687 |
possible | +0.614 |
ours | +0.590 |
unnecessary | +0.583 |
Neg logit contributions | |
boy | -0.451 |
little | -0.434 |
girl | -0.425 |
man | -0.406 |
sky | -0.403 |
Token | time |
Token ID | 640 |
Feature activation | +5.13e+00 |
Pos logit contributions | |
yours | +0.033 |
beh | +0.032 |
unnecessary | +0.028 |
ours | +0.027 |
supposed | +0.027 |
Neg logit contributions | |
boy | -0.028 |
girl | -0.027 |
man | -0.027 |
sky | -0.026 |
woman | -0.025 |
Token | , |
Token ID | 11 |
Feature activation | +1.15e+01 |
Pos logit contributions | |
boy | +0.153 |
girl | +0.145 |
man | +0.141 |
sky | +0.133 |
woman | +0.128 |
Neg logit contributions | |
yours | -0.359 |
beh | -0.355 |
possible | -0.325 |
ours | -0.312 |
ourselves | -0.308 |
Token | there |
Token ID | 612 |
Feature activation | -2.33e+00 |
Pos logit contributions | |
boy | +0.520 |
man | +0.514 |
girl | +0.499 |
sky | +0.498 |
little | +0.485 |
Neg logit contributions | |
yours | -0.606 |
beh | -0.586 |
possible | -0.524 |
ours | -0.505 |
eking | -0.490 |
Token | was |
Token ID | 373 |
Feature activation | +9.25e+00 |
Pos logit contributions | |
yours | +0.151 |
beh | +0.138 |
possible | +0.135 |
anything | +0.135 |
happen | +0.134 |
Neg logit contributions | |
woman | -0.058 |
girl | -0.058 |
man | -0.054 |
boy | -0.051 |
sky | -0.048 |
Token | a |
Token ID | 257 |
Feature activation | -1.20e+00 |
Pos logit contributions | |
boy | +0.409 |
man | +0.397 |
girl | +0.389 |
little | +0.365 |
sky | +0.362 |
Neg logit contributions | |
yours | -0.524 |
beh | -0.512 |
possible | -0.442 |
ours | -0.441 |
theirs | -0.433 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.50e+01 |
Pos logit contributions | |
yours | +0.078 |
ours | +0.071 |
beh | +0.070 |
possible | +0.068 |
asty | +0.066 |
Neg logit contributions | |
sky | -0.024 |
was | -0.018 |
led | -0.015 |
At | -0.014 |
Quick | -0.010 |
Token | in |
Token ID | 259 |
Feature activation | +3.09e-02 |
Pos logit contributions | |
boy | +1.103 |
little | +1.069 |
girl | +1.052 |
sky | +1.018 |
man | +1.012 |
Neg logit contributions | |
beh | -1.474 |
yours | -1.375 |
possible | -1.205 |
ieve | -1.158 |
eking | -1.152 |
Token | named |
Token ID | 3706 |
Feature activation | +5.43e+00 |
Pos logit contributions | |
boy | +0.001 |
girl | +0.001 |
sky | +0.001 |
little | +0.001 |
man | +0.001 |
Neg logit contributions | |
beh | -0.002 |
yours | -0.002 |
possible | -0.002 |
ourselves | -0.001 |
ieve | -0.001 |
Token | Penny |
Token ID | 25965 |
Feature activation | -4.70e+00 |
Pos logit contributions | |
boy | +0.237 |
little | +0.233 |
girl | +0.222 |
man | +0.215 |
sky | +0.214 |
Neg logit contributions | |
beh | -0.335 |
yours | -0.307 |
possible | -0.277 |
unnecessary | -0.265 |
ieve | -0.263 |
Token | . |
Token ID | 13 |
Feature activation | -9.07e+00 |
Pos logit contributions | |
beh | +0.245 |
yours | +0.235 |
possible | +0.203 |
ours | +0.193 |
theirs | +0.190 |
Neg logit contributions | |
boy | -0.236 |
little | -0.227 |
girl | -0.225 |
sky | -0.217 |
man | -0.216 |
Token | Penny |
Token ID | 25965 |
Feature activation | -4.84e+00 |
Pos logit contributions | |
beh | +0.474 |
yours | +0.466 |
possible | +0.399 |
ours | +0.376 |
theirs | +0.375 |
Neg logit contributions | |
boy | -0.440 |
little | -0.419 |
girl | -0.412 |
man | -0.410 |
sky | -0.389 |
Token | to |
Token ID | 284 |
Feature activation | +2.00e+00 |
Pos logit contributions | |
boy | +0.239 |
little | +0.233 |
girl | +0.223 |
man | +0.214 |
sky | +0.211 |
Neg logit contributions | |
beh | -0.386 |
yours | -0.363 |
possible | -0.326 |
ieve | -0.314 |
ours | -0.313 |
Token | play |
Token ID | 711 |
Feature activation | -7.40e+00 |
Pos logit contributions | |
boy | +0.083 |
little | +0.083 |
girl | +0.078 |
man | +0.076 |
sky | +0.074 |
Neg logit contributions | |
beh | -0.127 |
yours | -0.118 |
possible | -0.107 |
unnecessary | -0.104 |
ours | -0.102 |
Token | with |
Token ID | 351 |
Feature activation | +1.23e+00 |
Pos logit contributions | |
beh | +0.459 |
yours | +0.431 |
possible | +0.387 |
ours | +0.369 |
unnecessary | +0.366 |
Neg logit contributions | |
boy | -0.319 |
little | -0.309 |
girl | -0.300 |
man | -0.289 |
sky | -0.284 |
Token | their |
Token ID | 511 |
Feature activation | -6.70e-01 |
Pos logit contributions | |
boy | +0.055 |
little | +0.053 |
girl | +0.052 |
man | +0.050 |
sky | +0.050 |
Neg logit contributions | |
beh | -0.073 |
yours | -0.069 |
possible | -0.061 |
ours | -0.059 |
theirs | -0.058 |
Token | toy |
Token ID | 13373 |
Feature activation | +8.41e+00 |
Pos logit contributions | |
yours | +0.042 |
beh | +0.040 |
ours | +0.038 |
theirs | +0.036 |
possible | +0.036 |
Neg logit contributions | |
boy | -0.020 |
sky | -0.019 |
girl | -0.019 |
little | -0.018 |
man | -0.018 |
Token | pengu |
Token ID | 38373 |
Feature activation | +2.50e+01 |
Pos logit contributions | |
boy | +0.268 |
little | +0.256 |
girl | +0.249 |
man | +0.235 |
sky | +0.234 |
Neg logit contributions | |
beh | -0.596 |
yours | -0.573 |
possible | -0.514 |
ours | -0.500 |
unnecessary | -0.492 |
Token | in |
Token ID | 259 |
Feature activation | +3.09e-02 |
Pos logit contributions | |
boy | +1.227 |
little | +1.188 |
girl | +1.166 |
man | +1.124 |
sky | +1.112 |
Neg logit contributions | |
beh | -1.364 |
yours | -1.290 |
possible | -1.132 |
ours | -1.078 |
unnecessary | -1.062 |
Token | . |
Token ID | 13 |
Feature activation | -8.94e+00 |
Pos logit contributions | |
boy | +0.001 |
little | +0.001 |
girl | +0.001 |
man | +0.001 |
was | +0.001 |
Neg logit contributions | |
beh | -0.002 |
yours | -0.002 |
possible | -0.002 |
ours | -0.002 |
unnecessary | -0.002 |
Token | They |
Token ID | 1119 |
Feature activation | +9.13e-01 |
Pos logit contributions | |
beh | +0.621 |
yours | +0.603 |
possible | +0.537 |
ours | +0.525 |
theirs | +0.520 |
Neg logit contributions | |
boy | -0.295 |
little | -0.283 |
girl | -0.277 |
man | -0.264 |
sky | -0.259 |
Token | named |
Token ID | 3706 |
Feature activation | +5.30e+00 |
Pos logit contributions | |
boy | +0.023 |
girl | +0.022 |
little | +0.022 |
sky | +0.021 |
man | +0.021 |
Neg logit contributions | |
beh | -0.067 |
yours | -0.066 |
theirs | -0.059 |
possible | -0.058 |
ours | -0.058 |
Token | him |
Token ID | 683 |
Feature activation | -1.63e+01 |
Pos logit contributions | |
boy | +0.295 |
little | +0.293 |
girl | +0.275 |
man | +0.271 |
sky | +0.264 |
Neg logit contributions | |
beh | -0.277 |
yours | -0.229 |
possible | -0.223 |
eking | -0.208 |
unnecessary | -0.206 |
Token | . |
Token ID | 13 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
beh | +0.295 |
yours | +0.285 |
ours | +0.250 |
possible | +0.249 |
unnecessary | +0.247 |
Neg logit contributions | |
boy | -0.192 |
little | -0.183 |
man | -0.180 |
girl | -0.180 |
was | -0.171 |
Token | " |
Token ID | 366 |
Feature activation | +5.54e+00 |
Pos logit contributions | |
beh | +0.668 |
yours | +0.631 |
possible | +0.549 |
ours | +0.548 |
theirs | +0.541 |
Neg logit contributions | |
boy | -0.380 |
girl | -0.362 |
little | -0.357 |
man | -0.342 |
sky | -0.335 |
Token | That |
Token ID | 2504 |
Feature activation | +2.95e+00 |
Pos logit contributions | |
boy | +0.225 |
little | +0.218 |
girl | +0.212 |
man | +0.204 |
sky | +0.203 |
Neg logit contributions | |
beh | -0.350 |
yours | -0.331 |
possible | -0.294 |
ours | -0.283 |
unnecessary | -0.279 |
Token | 's |
Token ID | 338 |
Feature activation | +5.49e+00 |
Pos logit contributions | |
boy | +0.091 |
little | +0.087 |
girl | +0.085 |
sky | +0.081 |
man | +0.079 |
Neg logit contributions | |
beh | -0.209 |
yours | -0.203 |
possible | -0.183 |
ours | -0.178 |
theirs | -0.171 |
Token | not |
Token ID | 407 |
Feature activation | -1.14e+01 |
Pos logit contributions | |
little | +0.276 |
boy | +0.266 |
girl | +0.248 |
was | +0.246 |
man | +0.245 |
Neg logit contributions | |
beh | -0.296 |
yours | -0.266 |
possible | -0.239 |
ieve | -0.236 |
rying | -0.233 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.09e+01 |
Pos logit contributions | |
beh | +0.719 |
yours | +0.626 |
ieve | +0.600 |
eking | +0.594 |
possible | +0.586 |
Neg logit contributions | |
little | -0.487 |
boy | -0.477 |
girl | -0.447 |
man | -0.430 |
was | -0.409 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.33e+01 |
Pos logit contributions | |
beh | +2.255 |
yours | +2.076 |
possible | +1.888 |
unnecessary | +1.866 |
ieve | +1.860 |
Neg logit contributions | |
boy | -1.046 |
little | -1.000 |
girl | -0.958 |
man | -0.939 |
was | -0.912 |
Token | he |
Token ID | 339 |
Feature activation | -6.67e+00 |
Pos logit contributions | |
beh | +0.817 |
yours | +0.769 |
possible | +0.692 |
ours | +0.661 |
ourselves | +0.651 |
Neg logit contributions | |
boy | -0.514 |
little | -0.496 |
girl | -0.486 |
sky | -0.471 |
man | -0.453 |
Token | shouted |
Token ID | 17293 |
Feature activation | -2.04e+00 |
Pos logit contributions | |
beh | +0.440 |
yours | +0.401 |
possible | +0.360 |
eking | +0.351 |
ours | +0.351 |
Neg logit contributions | |
boy | -0.237 |
sky | -0.233 |
girl | -0.228 |
little | -0.222 |
man | -0.206 |
Token | . |
Token ID | 13 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
beh | +0.140 |
yours | +0.137 |
ours | +0.120 |
possible | +0.119 |
unnecessary | +0.116 |
Neg logit contributions | |
little | -0.080 |
boy | -0.077 |
man | -0.072 |
girl | -0.072 |
and | -0.070 |
Token | He |
Token ID | 679 |
Feature activation | -2.27e+00 |
Pos logit contributions | |
beh | +0.736 |
yours | +0.708 |
possible | +0.636 |
ours | +0.613 |
theirs | +0.613 |
Neg logit contributions | |
boy | -0.299 |
girl | -0.281 |
little | -0.274 |
sky | -0.264 |
man | -0.254 |
Token | is |
Token ID | 318 |
Feature activation | +2.74e+00 |
Pos logit contributions | |
boy | +0.391 |
little | +0.366 |
girl | +0.352 |
man | +0.344 |
was | +0.325 |
Neg logit contributions | |
beh | -0.893 |
yours | -0.833 |
possible | -0.766 |
unnecessary | -0.745 |
ours | -0.735 |
Token | faster |
Token ID | 5443 |
Feature activation | -7.08e+00 |
Pos logit contributions | |
little | +0.119 |
boy | +0.113 |
man | +0.105 |
girl | +0.101 |
and | +0.099 |
Neg logit contributions | |
beh | -0.183 |
rying | -0.145 |
yours | -0.145 |
irs | -0.144 |
ieve | -0.144 |
Token | and |
Token ID | 290 |
Feature activation | +1.22e+01 |
Pos logit contributions | |
beh | +0.400 |
yours | +0.348 |
ieve | +0.305 |
possible | +0.305 |
unnecessary | +0.301 |
Neg logit contributions | |
boy | -0.364 |
little | -0.353 |
girl | -0.339 |
man | -0.336 |
and | -0.325 |
Token | smarter |
Token ID | 23714 |
Feature activation | -8.06e+00 |
Pos logit contributions | |
boy | +0.529 |
little | +0.513 |
was | +0.480 |
man | +0.475 |
girl | +0.464 |
Neg logit contributions | |
beh | -0.768 |
yours | -0.660 |
irs | -0.652 |
ility | -0.626 |
owing | -0.623 |
Token | than |
Token ID | 621 |
Feature activation | -1.02e+01 |
Pos logit contributions | |
beh | +0.453 |
yours | +0.386 |
irs | +0.354 |
ility | +0.346 |
eking | +0.344 |
Neg logit contributions | |
boy | -0.423 |
man | -0.400 |
little | -0.397 |
girl | -0.390 |
and | -0.380 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.09e+01 |
Pos logit contributions | |
beh | +0.588 |
ieve | +0.450 |
rying | +0.431 |
unnecessary | +0.426 |
yours | +0.425 |
Neg logit contributions | |
little | -0.534 |
boy | -0.507 |
man | -0.476 |
sky | -0.473 |
was | -0.466 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.33e+01 |
Pos logit contributions | |
beh | +1.871 |
yours | +1.483 |
unnecessary | +1.425 |
ility | +1.389 |
eking | +1.389 |
Neg logit contributions | |
boy | -1.578 |
little | -1.518 |
man | -1.472 |
was | -1.450 |
girl | -1.446 |
Token | Lily |
Token ID | 20037 |
Feature activation | -2.99e+00 |
Pos logit contributions | |
beh | +0.702 |
yours | +0.672 |
possible | +0.589 |
theirs | +0.577 |
ours | +0.559 |
Neg logit contributions | |
sky | -0.558 |
boy | -0.555 |
girl | -0.540 |
little | -0.535 |
man | -0.516 |
Token | said |
Token ID | 531 |
Feature activation | +6.43e-01 |
Pos logit contributions | |
beh | +0.180 |
yours | +0.151 |
eking | +0.139 |
myself | +0.138 |
properly | +0.136 |
Neg logit contributions | |
sky | -0.113 |
girl | -0.103 |
boy | -0.102 |
little | -0.102 |
field | -0.094 |
Token | , |
Token ID | 11 |
Feature activation | +9.27e+00 |
Pos logit contributions | |
little | +0.011 |
boy | +0.011 |
girl | +0.010 |
man | +0.009 |
and | +0.008 |
Neg logit contributions | |
beh | -0.058 |
yours | -0.056 |
ours | -0.051 |
possible | -0.051 |
unnecessary | -0.050 |
Token | holding |
Token ID | 4769 |
Feature activation | +2.92e+00 |
Pos logit contributions | |
boy | +0.446 |
little | +0.429 |
girl | +0.424 |
sky | +0.407 |
man | +0.407 |
Neg logit contributions | |
beh | -0.512 |
yours | -0.482 |
possible | -0.422 |
ours | -0.403 |
theirs | -0.394 |
Token | is |
Token ID | 318 |
Feature activation | +2.74e+00 |
Pos logit contributions | |
beh | +0.080 |
yours | +0.078 |
possible | +0.070 |
ours | +0.068 |
ieve | +0.067 |
Neg logit contributions | |
boy | -0.031 |
little | -0.030 |
girl | -0.029 |
man | -0.027 |
sky | -0.027 |
Token | faster |
Token ID | 5443 |
Feature activation | -7.08e+00 |
Pos logit contributions | |
boy | +0.102 |
little | +0.101 |
girl | +0.093 |
sky | +0.092 |
man | +0.091 |
Neg logit contributions | |
beh | -0.188 |
yours | -0.170 |
possible | -0.154 |
unnecessary | -0.151 |
ieve | -0.150 |
Token | and |
Token ID | 290 |
Feature activation | +1.22e+01 |
Pos logit contributions | |
beh | +0.442 |
yours | +0.399 |
possible | +0.355 |
unnecessary | +0.346 |
ieve | +0.345 |
Neg logit contributions | |
boy | -0.318 |
little | -0.305 |
girl | -0.295 |
man | -0.287 |
sky | -0.277 |
Token | stronger |
Token ID | 7387 |
Feature activation | -4.84e+00 |
Pos logit contributions | |
boy | +0.589 |
little | +0.555 |
girl | +0.537 |
man | +0.528 |
was | +0.526 |
Neg logit contributions | |
beh | -0.727 |
yours | -0.638 |
irs | -0.570 |
ility | -0.557 |
owing | -0.553 |
Token | than |
Token ID | 621 |
Feature activation | -1.02e+01 |
Pos logit contributions | |
beh | +0.296 |
yours | +0.258 |
irs | +0.237 |
ility | +0.234 |
eking | +0.231 |
Neg logit contributions | |
boy | -0.241 |
man | -0.220 |
girl | -0.219 |
little | -0.217 |
and | -0.211 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.09e+01 |
Pos logit contributions | |
beh | +0.584 |
ieve | +0.448 |
rying | +0.437 |
eking | +0.432 |
unnecessary | +0.431 |
Neg logit contributions | |
little | -0.524 |
boy | -0.497 |
sky | -0.477 |
girl | -0.464 |
was | -0.459 |
Token | . |
Token ID | 13 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
beh | +2.022 |
yours | +1.604 |
ility | +1.594 |
unnecessary | +1.582 |
irs | +1.577 |
Neg logit contributions | |
boy | -1.442 |
little | -1.361 |
man | -1.342 |
was | -1.330 |
and | -1.324 |
Token | Your |
Token ID | 3406 |
Feature activation | -5.93e+00 |
Pos logit contributions | |
beh | +0.643 |
yours | +0.619 |
possible | +0.539 |
ours | +0.527 |
theirs | +0.512 |
Neg logit contributions | |
boy | -0.442 |
little | -0.427 |
girl | -0.420 |
man | -0.406 |
sky | -0.396 |
Token | car |
Token ID | 1097 |
Feature activation | +5.02e+00 |
Pos logit contributions | |
yours | +0.416 |
beh | +0.410 |
ours | +0.367 |
possible | +0.363 |
theirs | +0.360 |
Neg logit contributions | |
boy | -0.146 |
girl | -0.144 |
sky | -0.137 |
little | -0.135 |
man | -0.130 |
Token | is |
Token ID | 318 |
Feature activation | +2.74e+00 |
Pos logit contributions | |
boy | +0.133 |
little | +0.125 |
girl | +0.121 |
man | +0.113 |
sky | +0.111 |
Neg logit contributions | |
beh | -0.386 |
yours | -0.370 |
possible | -0.337 |
ours | -0.327 |
unnecessary | -0.324 |
Token | stupid |
Token ID | 8531 |
Feature activation | -8.09e+00 |
Pos logit contributions | |
little | +0.105 |
boy | +0.100 |
girl | +0.090 |
man | +0.090 |
sky | +0.089 |
Neg logit contributions | |
beh | -0.188 |
yours | -0.165 |
rying | -0.150 |
unnecessary | -0.150 |
ieve | -0.148 |
Token | my |
Token ID | 616 |
Feature activation | -7.00e+00 |
Pos logit contributions | |
boy | +0.413 |
little | +0.400 |
girl | +0.390 |
man | +0.373 |
sky | +0.369 |
Neg logit contributions | |
beh | -0.561 |
yours | -0.524 |
possible | -0.466 |
ours | -0.444 |
unnecessary | -0.443 |
Token | row |
Token ID | 5752 |
Feature activation | +7.56e+00 |
Pos logit contributions | |
yours | +0.466 |
beh | +0.453 |
possible | +0.414 |
ours | +0.407 |
ieve | +0.394 |
Neg logit contributions | |
boy | -0.209 |
girl | -0.200 |
little | -0.195 |
man | -0.189 |
sky | -0.188 |
Token | is |
Token ID | 318 |
Feature activation | +2.75e+00 |
Pos logit contributions | |
boy | +0.268 |
little | +0.254 |
girl | +0.248 |
man | +0.237 |
sky | +0.233 |
Neg logit contributions | |
beh | -0.523 |
yours | -0.492 |
possible | -0.445 |
unnecessary | -0.432 |
ours | -0.430 |
Token | longer |
Token ID | 2392 |
Feature activation | -1.30e+01 |
Pos logit contributions | |
little | +0.099 |
boy | +0.096 |
sky | +0.088 |
man | +0.086 |
girl | +0.086 |
Neg logit contributions | |
beh | -0.197 |
yours | -0.170 |
unnecessary | -0.159 |
rying | -0.156 |
ieve | -0.156 |
Token | than |
Token ID | 621 |
Feature activation | -1.02e+01 |
Pos logit contributions | |
beh | +0.988 |
yours | +0.898 |
possible | +0.807 |
unnecessary | +0.806 |
ieve | +0.803 |
Neg logit contributions | |
boy | -0.401 |
little | -0.399 |
girl | -0.364 |
man | -0.349 |
sky | -0.344 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.09e+01 |
Pos logit contributions | |
beh | +0.674 |
yours | +0.512 |
unnecessary | +0.511 |
rying | +0.506 |
ieve | +0.501 |
Neg logit contributions | |
little | -0.458 |
boy | -0.426 |
sky | -0.398 |
girl | -0.395 |
man | -0.389 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.33e+01 |
Pos logit contributions | |
beh | +1.917 |
yours | +1.621 |
unnecessary | +1.515 |
eking | +1.454 |
possible | +1.452 |
Neg logit contributions | |
boy | -1.479 |
little | -1.412 |
girl | -1.371 |
man | -1.358 |
was | -1.339 |
Token | Anna |
Token ID | 11735 |
Feature activation | +1.76e+00 |
Pos logit contributions | |
beh | +0.777 |
yours | +0.756 |
possible | +0.663 |
ours | +0.645 |
ourselves | +0.634 |
Neg logit contributions | |
boy | -0.551 |
little | -0.527 |
girl | -0.521 |
sky | -0.501 |
man | -0.498 |
Token | says |
Token ID | 1139 |
Feature activation | -9.21e-01 |
Pos logit contributions | |
little | +0.070 |
boy | +0.070 |
girl | +0.068 |
sky | +0.065 |
man | +0.064 |
Neg logit contributions | |
beh | -0.103 |
yours | -0.098 |
possible | -0.085 |
ieve | -0.085 |
ours | -0.083 |
Token | . |
Token ID | 13 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
beh | +0.064 |
yours | +0.061 |
ours | +0.055 |
possible | +0.054 |
unnecessary | +0.054 |
Neg logit contributions | |
little | -0.035 |
boy | -0.034 |
and | -0.032 |
girl | -0.032 |
man | -0.031 |
Token | She |
Token ID | 1375 |
Feature activation | -4.13e-01 |
Pos logit contributions | |
yours | +0.658 |
beh | +0.646 |
possible | +0.594 |
theirs | +0.575 |
ourselves | +0.558 |
Neg logit contributions | |
boy | -0.323 |
girl | -0.292 |
little | -0.279 |
man | -0.273 |
child | -0.258 |
Token | toys |
Token ID | 14958 |
Feature activation | +2.26e+00 |
Pos logit contributions | |
boy | +0.209 |
little | +0.200 |
girl | +0.193 |
man | +0.185 |
sky | +0.180 |
Neg logit contributions | |
beh | -0.370 |
yours | -0.349 |
possible | -0.319 |
ours | -0.301 |
ieve | -0.298 |
Token | . |
Token ID | 13 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
little | +0.089 |
boy | +0.089 |
and | +0.084 |
man | +0.083 |
girl | +0.082 |
Neg logit contributions | |
beh | -0.153 |
yours | -0.132 |
unnecessary | -0.122 |
possible | -0.122 |
ieve | -0.120 |
Token | They |
Token ID | 1119 |
Feature activation | -6.73e-01 |
Pos logit contributions | |
beh | +0.601 |
yours | +0.569 |
possible | +0.496 |
ours | +0.479 |
unnecessary | +0.469 |
Neg logit contributions | |
boy | -0.493 |
little | -0.477 |
girl | -0.469 |
man | -0.454 |
sky | -0.447 |
Token | are |
Token ID | 389 |
Feature activation | -3.21e+00 |
Pos logit contributions | |
beh | +0.042 |
yours | +0.039 |
possible | +0.036 |
unnecessary | +0.033 |
ours | +0.033 |
Neg logit contributions | |
boy | -0.029 |
little | -0.028 |
girl | -0.027 |
man | -0.026 |
was | -0.025 |
Token | not |
Token ID | 407 |
Feature activation | -1.14e+01 |
Pos logit contributions | |
beh | +0.196 |
yours | +0.168 |
possible | +0.154 |
ieve | +0.154 |
rying | +0.153 |
Neg logit contributions | |
little | -0.147 |
boy | -0.140 |
girl | -0.132 |
man | -0.130 |
sky | -0.124 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.09e+01 |
Pos logit contributions | |
beh | +0.744 |
ieve | +0.645 |
yours | +0.642 |
irs | +0.626 |
eking | +0.623 |
Neg logit contributions | |
little | -0.481 |
boy | -0.464 |
girl | -0.427 |
man | -0.389 |
and | -0.372 |
Token | . |
Token ID | 13 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
beh | +2.227 |
yours | +1.925 |
eking | +1.823 |
unnecessary | +1.785 |
possible | +1.783 |
Neg logit contributions | |
boy | -1.176 |
little | -1.134 |
girl | -1.076 |
was | -1.073 |
and | -1.068 |
Token | You |
Token ID | 921 |
Feature activation | -1.35e+01 |
Pos logit contributions | |
beh | +0.693 |
yours | +0.651 |
possible | +0.596 |
unnecessary | +0.564 |
eking | +0.564 |
Neg logit contributions | |
boy | -0.409 |
little | -0.393 |
girl | -0.383 |
sky | -0.364 |
man | -0.360 |
Token | have |
Token ID | 423 |
Feature activation | -7.36e+00 |
Pos logit contributions | |
beh | +0.866 |
yours | +0.819 |
possible | +0.748 |
unnecessary | +0.707 |
ours | +0.702 |
Neg logit contributions | |
boy | -0.544 |
little | -0.530 |
girl | -0.503 |
man | -0.487 |
was | -0.478 |
Token | to |
Token ID | 284 |
Feature activation | -3.58e-01 |
Pos logit contributions | |
beh | +0.469 |
yours | +0.436 |
possible | +0.395 |
ieve | +0.380 |
ours | +0.376 |
Neg logit contributions | |
boy | -0.300 |
little | -0.295 |
girl | -0.281 |
man | -0.268 |
sky | -0.267 |
Token | ask |
Token ID | 1265 |
Feature activation | -9.34e+00 |
Pos logit contributions | |
beh | +0.023 |
yours | +0.021 |
possible | +0.020 |
unnecessary | +0.019 |
rying | +0.019 |
Neg logit contributions | |
little | -0.015 |
boy | -0.014 |
man | -0.013 |
girl | -0.013 |
was | -0.012 |
Token | on |
Token ID | 319 |
Feature activation | -6.17e-01 |
Pos logit contributions | |
beh | +0.321 |
yours | +0.266 |
rying | +0.256 |
possible | +0.250 |
unnecessary | +0.241 |
Neg logit contributions | |
little | -0.176 |
boy | -0.165 |
man | -0.160 |
and | -0.157 |
was | -0.155 |
Token | things |
Token ID | 1243 |
Feature activation | -1.41e+00 |
Pos logit contributions | |
beh | +0.042 |
yours | +0.037 |
possible | +0.034 |
rying | +0.033 |
ieve | +0.032 |
Neg logit contributions | |
little | -0.024 |
boy | -0.024 |
girl | -0.022 |
man | -0.022 |
sky | -0.022 |
Token | that |
Token ID | 326 |
Feature activation | -8.46e+00 |
Pos logit contributions | |
beh | +0.110 |
yours | +0.091 |
unnecessary | +0.090 |
rying | +0.087 |
eking | +0.086 |
Neg logit contributions | |
and | -0.045 |
little | -0.044 |
boy | -0.043 |
man | -0.042 |
was | -0.040 |
Token | are |
Token ID | 389 |
Feature activation | -3.21e+00 |
Pos logit contributions | |
beh | +0.539 |
yours | +0.493 |
possible | +0.446 |
unnecessary | +0.435 |
rying | +0.430 |
Neg logit contributions | |
boy | -0.355 |
little | -0.338 |
girl | -0.330 |
man | -0.326 |
sky | -0.315 |
Token | not |
Token ID | 407 |
Feature activation | -1.14e+01 |
Pos logit contributions | |
beh | +0.202 |
yours | +0.162 |
rying | +0.156 |
ieve | +0.154 |
unnecessary | +0.151 |
Neg logit contributions | |
little | -0.148 |
boy | -0.138 |
man | -0.131 |
was | -0.128 |
girl | -0.127 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.09e+01 |
Pos logit contributions | |
beh | +0.676 |
rying | +0.513 |
ieve | +0.503 |
eking | +0.498 |
unnecessary | +0.484 |
Neg logit contributions | |
little | -0.604 |
boy | -0.567 |
and | -0.539 |
girl | -0.533 |
man | -0.529 |
Token | , |
Token ID | 11 |
Feature activation | +9.27e+00 |
Pos logit contributions | |
beh | +2.298 |
yours | +1.906 |
unnecessary | +1.844 |
eking | +1.830 |
rying | +1.778 |
Neg logit contributions | |
boy | -1.127 |
little | -1.107 |
was | -1.059 |
and | -1.057 |
man | -1.035 |
Token | or |
Token ID | 393 |
Feature activation | -2.06e-01 |
Pos logit contributions | |
boy | +0.444 |
little | +0.434 |
girl | +0.421 |
man | +0.409 |
sky | +0.400 |
Neg logit contributions | |
beh | -0.530 |
yours | -0.483 |
possible | -0.430 |
ieve | -0.409 |
unnecessary | -0.407 |
Token | that |
Token ID | 326 |
Feature activation | -8.45e+00 |
Pos logit contributions | |
beh | +0.012 |
yours | +0.010 |
rying | +0.010 |
possible | +0.009 |
eking | +0.009 |
Neg logit contributions | |
little | -0.010 |
boy | -0.010 |
girl | -0.009 |
man | -0.009 |
was | -0.009 |
Token | can |
Token ID | 460 |
Feature activation | +1.16e+00 |
Pos logit contributions | |
beh | +0.559 |
yours | +0.497 |
rying | +0.456 |
unnecessary | +0.449 |
possible | +0.448 |
Neg logit contributions | |
boy | -0.345 |
little | -0.325 |
girl | -0.317 |
man | -0.316 |
was | -0.307 |
Token | hurt |
Token ID | 5938 |
Feature activation | -4.53e+00 |
Pos logit contributions | |
little | +0.050 |
boy | +0.050 |
man | +0.049 |
was | +0.048 |
sky | +0.046 |
Neg logit contributions | |
beh | -0.074 |
yours | -0.063 |
unnecessary | -0.061 |
rying | -0.060 |
theirs | -0.056 |
Token | with |
Token ID | 351 |
Feature activation | -8.95e-01 |
Pos logit contributions | |
beh | +0.721 |
yours | +0.650 |
possible | +0.614 |
unnecessary | +0.589 |
eking | +0.582 |
Neg logit contributions | |
boy | -0.319 |
little | -0.312 |
man | -0.290 |
girl | -0.280 |
was | -0.269 |
Token | things |
Token ID | 1243 |
Feature activation | -1.41e+00 |
Pos logit contributions | |
beh | +0.056 |
yours | +0.047 |
possible | +0.045 |
rying | +0.043 |
ieve | +0.043 |
Neg logit contributions | |
little | -0.041 |
boy | -0.039 |
man | -0.037 |
girl | -0.036 |
was | -0.035 |
Token | that |
Token ID | 326 |
Feature activation | -8.46e+00 |
Pos logit contributions | |
beh | +0.103 |
unnecessary | +0.083 |
yours | +0.082 |
eking | +0.081 |
irs | +0.080 |
Neg logit contributions | |
and | -0.052 |
boy | -0.052 |
little | -0.052 |
man | -0.051 |
girl | -0.046 |
Token | are |
Token ID | 389 |
Feature activation | -3.21e+00 |
Pos logit contributions | |
beh | +0.520 |
yours | +0.479 |
possible | +0.432 |
unnecessary | +0.417 |
rying | +0.411 |
Neg logit contributions | |
boy | -0.373 |
little | -0.355 |
girl | -0.346 |
man | -0.341 |
sky | -0.330 |
Token | not |
Token ID | 407 |
Feature activation | -1.14e+01 |
Pos logit contributions | |
beh | +0.194 |
yours | +0.152 |
ieve | +0.152 |
irs | +0.147 |
anks | +0.146 |
Neg logit contributions | |
little | -0.155 |
boy | -0.144 |
man | -0.137 |
was | -0.135 |
and | -0.135 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.09e+01 |
Pos logit contributions | |
beh | +0.323 |
ieve | +0.181 |
eking | +0.180 |
rying | +0.178 |
anks | +0.152 |
Neg logit contributions | |
little | -0.940 |
boy | -0.903 |
and | -0.863 |
girl | -0.862 |
man | -0.859 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.33e+01 |
Pos logit contributions | |
beh | +2.382 |
yours | +2.006 |
unnecessary | +1.916 |
ieve | +1.900 |
eking | +1.894 |
Neg logit contributions | |
boy | -1.056 |
little | -1.030 |
and | -0.975 |
was | -0.962 |
man | -0.953 |
Token | He |
Token ID | 679 |
Feature activation | -2.30e+00 |
Pos logit contributions | |
beh | +0.928 |
yours | +0.892 |
possible | +0.809 |
ours | +0.780 |
theirs | +0.772 |
Neg logit contributions | |
boy | -0.408 |
little | -0.395 |
girl | -0.387 |
sky | -0.372 |
man | -0.354 |
Token | said |
Token ID | 531 |
Feature activation | +6.58e-01 |
Pos logit contributions | |
beh | +0.128 |
yours | +0.111 |
eking | +0.106 |
ieve | +0.101 |
cks | +0.095 |
Neg logit contributions | |
sky | -0.095 |
boy | -0.090 |
girl | -0.088 |
field | -0.084 |
little | -0.082 |
Token | , |
Token ID | 11 |
Feature activation | +9.28e+00 |
Pos logit contributions | |
little | +0.026 |
boy | +0.025 |
man | +0.023 |
and | +0.023 |
girl | +0.023 |
Neg logit contributions | |
beh | -0.046 |
yours | -0.044 |
ours | -0.040 |
unnecessary | -0.038 |
possible | -0.038 |
Token | shaking |
Token ID | 17275 |
Feature activation | -2.27e+00 |
Pos logit contributions | |
boy | +0.427 |
girl | +0.406 |
sky | +0.396 |
little | +0.388 |
man | +0.378 |
Neg logit contributions | |
beh | -0.499 |
yours | -0.467 |
eking | -0.410 |
possible | -0.409 |
theirs | -0.399 |
Token | own |
Token ID | 898 |
Feature activation | -1.07e+01 |
Pos logit contributions | |
beh | +0.533 |
yours | +0.492 |
unnecessary | +0.475 |
possible | +0.469 |
ours | +0.445 |
Neg logit contributions | |
boy | -0.303 |
little | -0.297 |
man | -0.275 |
was | -0.273 |
girl | -0.271 |
Token | things |
Token ID | 1243 |
Feature activation | -1.41e+00 |
Pos logit contributions | |
beh | +0.745 |
rying | +0.655 |
yours | +0.640 |
eking | +0.630 |
ieve | +0.606 |
Neg logit contributions | |
little | -0.418 |
boy | -0.409 |
was | -0.382 |
girl | -0.368 |
man | -0.360 |
Token | that |
Token ID | 326 |
Feature activation | -8.46e+00 |
Pos logit contributions | |
beh | +0.101 |
unnecessary | +0.082 |
rying | +0.079 |
eking | +0.079 |
yours | +0.079 |
Neg logit contributions | |
and | -0.055 |
little | -0.054 |
boy | -0.052 |
man | -0.050 |
was | -0.049 |
Token | are |
Token ID | 389 |
Feature activation | -3.21e+00 |
Pos logit contributions | |
beh | +0.506 |
yours | +0.461 |
possible | +0.422 |
unnecessary | +0.411 |
rying | +0.407 |
Neg logit contributions | |
boy | -0.390 |
little | -0.372 |
girl | -0.361 |
man | -0.355 |
was | -0.348 |
Token | not |
Token ID | 407 |
Feature activation | -1.14e+01 |
Pos logit contributions | |
beh | +0.194 |
ieve | +0.156 |
rying | +0.149 |
yours | +0.149 |
anks | +0.149 |
Neg logit contributions | |
little | -0.157 |
boy | -0.147 |
man | -0.135 |
was | -0.134 |
girl | -0.134 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.09e+01 |
Pos logit contributions | |
beh | +0.382 |
ieve | +0.255 |
rying | +0.250 |
eking | +0.243 |
anks | +0.220 |
Neg logit contributions | |
little | -0.888 |
boy | -0.854 |
girl | -0.817 |
man | -0.798 |
and | -0.794 |
Token | . |
Token ID | 13 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
beh | +2.244 |
yours | +1.841 |
eking | +1.795 |
unnecessary | +1.794 |
ieve | +1.763 |
Neg logit contributions | |
boy | -1.197 |
little | -1.170 |
and | -1.128 |
was | -1.109 |
girl | -1.086 |
Token | That |
Token ID | 1320 |
Feature activation | -1.08e+01 |
Pos logit contributions | |
beh | +0.613 |
yours | +0.565 |
possible | +0.527 |
eking | +0.504 |
rying | +0.491 |
Neg logit contributions | |
boy | -0.490 |
little | -0.469 |
girl | -0.460 |
sky | -0.437 |
man | -0.432 |
Token | is |
Token ID | 318 |
Feature activation | +2.75e+00 |
Pos logit contributions | |
beh | +0.799 |
yours | +0.784 |
possible | +0.706 |
ours | +0.686 |
ourselves | +0.670 |
Neg logit contributions | |
boy | -0.281 |
little | -0.278 |
girl | -0.264 |
sky | -0.249 |
man | -0.243 |
Token | not |
Token ID | 407 |
Feature activation | -1.14e+01 |
Pos logit contributions | |
little | +0.131 |
boy | +0.127 |
girl | +0.116 |
was | +0.114 |
man | +0.110 |
Neg logit contributions | |
beh | -0.161 |
rying | -0.133 |
yours | -0.132 |
eking | -0.131 |
anks | -0.129 |
Token | nice |
Token ID | 3621 |
Feature activation | -4.94e+00 |
Pos logit contributions | |
beh | +0.715 |
rying | +0.600 |
eking | +0.597 |
ieve | +0.595 |
anks | +0.570 |
Neg logit contributions | |
little | -0.533 |
boy | -0.517 |
girl | -0.479 |
man | -0.444 |
was | -0.442 |
Token | cheese |
Token ID | 9891 |
Feature activation | +4.59e+00 |
Pos logit contributions | |
beh | +0.591 |
yours | +0.560 |
possible | +0.504 |
ours | +0.487 |
unnecessary | +0.484 |
Neg logit contributions | |
boy | -0.290 |
little | -0.278 |
girl | -0.270 |
man | -0.257 |
sky | -0.253 |
Token | , |
Token ID | 11 |
Feature activation | +9.29e+00 |
Pos logit contributions | |
boy | +0.134 |
little | +0.131 |
girl | +0.122 |
man | +0.119 |
was | +0.115 |
Neg logit contributions | |
beh | -0.352 |
yours | -0.318 |
possible | -0.292 |
unnecessary | -0.291 |
eking | -0.288 |
Token | it |
Token ID | 340 |
Feature activation | -4.90e+00 |
Pos logit contributions | |
boy | +0.529 |
little | +0.525 |
girl | +0.506 |
man | +0.488 |
sky | +0.483 |
Neg logit contributions | |
beh | -0.450 |
yours | -0.399 |
possible | -0.345 |
unnecessary | -0.326 |
ours | -0.324 |
Token | will |
Token ID | 481 |
Feature activation | -3.35e+00 |
Pos logit contributions | |
beh | +0.334 |
yours | +0.317 |
possible | +0.286 |
ours | +0.277 |
theirs | +0.275 |
Neg logit contributions | |
boy | -0.176 |
little | -0.170 |
girl | -0.164 |
man | -0.156 |
sky | -0.153 |
Token | be |
Token ID | 307 |
Feature activation | -1.55e+00 |
Pos logit contributions | |
beh | +0.232 |
yours | +0.202 |
unnecessary | +0.187 |
rying | +0.185 |
ieve | +0.180 |
Neg logit contributions | |
little | -0.130 |
boy | -0.121 |
was | -0.117 |
man | -0.115 |
sky | -0.113 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.09e+01 |
Pos logit contributions | |
beh | +0.106 |
yours | +0.088 |
ieve | +0.084 |
anks | +0.084 |
rying | +0.084 |
Neg logit contributions | |
little | -0.063 |
boy | -0.056 |
was | -0.054 |
girl | -0.053 |
and | -0.052 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.33e+01 |
Pos logit contributions | |
beh | +1.908 |
yours | +1.632 |
unnecessary | +1.464 |
eking | +1.455 |
ieve | +1.444 |
Neg logit contributions | |
boy | -1.436 |
little | -1.414 |
girl | -1.369 |
man | -1.327 |
was | -1.299 |
Token |
|
Token ID | 198 |
Feature activation | +5.33e+00 |
Pos logit contributions | |
beh | +1.025 |
yours | +0.980 |
possible | +0.882 |
ourselves | +0.860 |
ours | +0.859 |
Neg logit contributions | |
boy | -0.330 |
little | -0.314 |
girl | -0.304 |
man | -0.281 |
sky | -0.279 |
Token |
|
Token ID | 198 |
Feature activation | +5.36e+00 |
Pos logit contributions | |
boy | +0.117 |
girl | +0.111 |
little | +0.103 |
sky | +0.098 |
man | +0.094 |
Neg logit contributions | |
beh | -0.420 |
yours | -0.407 |
possible | -0.369 |
ourselves | -0.362 |
ieve | -0.355 |
Token | Tim |
Token ID | 14967 |
Feature activation | +2.63e+00 |
Pos logit contributions | |
boy | +0.209 |
girl | +0.197 |
little | +0.197 |
sky | +0.197 |
man | +0.188 |
Neg logit contributions | |
beh | -0.349 |
yours | -0.322 |
possible | -0.284 |
ourselves | -0.280 |
myself | -0.274 |
Token | my |
Token ID | 1820 |
Feature activation | -6.71e+00 |
Pos logit contributions | |
sky | +0.119 |
boy | +0.118 |
little | +0.115 |
man | +0.113 |
girl | +0.113 |
Neg logit contributions | |
beh | -0.140 |
yours | -0.124 |
rying | -0.106 |
eking | -0.105 |
myself | -0.103 |
Token | take |
Token ID | 1011 |
Feature activation | -8.40e+00 |
Pos logit contributions | |
beh | +0.734 |
yours | +0.708 |
possible | +0.663 |
unnecessary | +0.644 |
ours | +0.636 |
Neg logit contributions | |
boy | -0.445 |
little | -0.438 |
was | -0.413 |
girl | -0.400 |
man | -0.391 |
Token | things |
Token ID | 1243 |
Feature activation | -1.41e+00 |
Pos logit contributions | |
beh | +0.573 |
yours | +0.522 |
possible | +0.493 |
rying | +0.479 |
eking | +0.466 |
Neg logit contributions | |
little | -0.324 |
boy | -0.317 |
girl | -0.291 |
man | -0.277 |
sky | -0.266 |
Token | that |
Token ID | 326 |
Feature activation | -8.46e+00 |
Pos logit contributions | |
beh | +0.104 |
yours | +0.087 |
eking | +0.084 |
unnecessary | +0.083 |
rying | +0.081 |
Neg logit contributions | |
and | -0.050 |
little | -0.049 |
boy | -0.048 |
man | -0.047 |
girl | -0.042 |
Token | are |
Token ID | 389 |
Feature activation | -3.21e+00 |
Pos logit contributions | |
beh | +0.508 |
yours | +0.479 |
possible | +0.426 |
ours | +0.407 |
unnecessary | +0.404 |
Neg logit contributions | |
boy | -0.375 |
little | -0.361 |
girl | -0.352 |
man | -0.341 |
sky | -0.335 |
Token | not |
Token ID | 407 |
Feature activation | -1.14e+01 |
Pos logit contributions | |
beh | +0.196 |
yours | +0.158 |
ieve | +0.157 |
irs | +0.154 |
rying | +0.151 |
Neg logit contributions | |
little | -0.152 |
boy | -0.144 |
was | -0.132 |
man | -0.131 |
girl | -0.129 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.09e+01 |
Pos logit contributions | |
beh | +0.488 |
eking | +0.367 |
ieve | +0.366 |
rying | +0.361 |
irs | +0.337 |
Neg logit contributions | |
little | -0.771 |
boy | -0.745 |
girl | -0.700 |
man | -0.683 |
and | -0.675 |
Token | . |
Token ID | 13 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
beh | +2.252 |
yours | +1.957 |
unnecessary | +1.821 |
eking | +1.808 |
possible | +1.806 |
Neg logit contributions | |
boy | -1.119 |
little | -1.091 |
was | -1.060 |
girl | -1.007 |
and | -1.004 |
Token | You |
Token ID | 921 |
Feature activation | -1.35e+01 |
Pos logit contributions | |
beh | +0.639 |
yours | +0.586 |
possible | +0.547 |
eking | +0.520 |
unnecessary | +0.511 |
Neg logit contributions | |
boy | -0.473 |
little | -0.452 |
girl | -0.441 |
sky | -0.421 |
man | -0.416 |
Token | should |
Token ID | 815 |
Feature activation | -1.37e+00 |
Pos logit contributions | |
beh | +0.839 |
yours | +0.796 |
possible | +0.710 |
ours | +0.680 |
unnecessary | +0.672 |
Neg logit contributions | |
boy | -0.563 |
little | -0.544 |
girl | -0.530 |
man | -0.509 |
sky | -0.504 |
Token | not |
Token ID | 407 |
Feature activation | -1.14e+01 |
Pos logit contributions | |
beh | +0.089 |
unnecessary | +0.075 |
yours | +0.075 |
rying | +0.072 |
possible | +0.072 |
Neg logit contributions | |
little | -0.058 |
boy | -0.056 |
man | -0.054 |
girl | -0.052 |
was | -0.052 |
Token | fight |
Token ID | 1907 |
Feature activation | -9.86e+00 |
Pos logit contributions | |
beh | +0.731 |
yours | +0.697 |
possible | +0.660 |
unnecessary | +0.646 |
ieve | +0.638 |
Neg logit contributions | |
boy | -0.448 |
little | -0.438 |
was | -0.410 |
girl | -0.398 |
man | -0.393 |
Token | to |
Token ID | 284 |
Feature activation | -3.45e-01 |
Pos logit contributions | |
beh | +1.117 |
yours | +0.963 |
rying | +0.905 |
possible | +0.901 |
ieve | +0.901 |
Neg logit contributions | |
little | -0.613 |
boy | -0.606 |
girl | -0.558 |
sky | -0.546 |
man | -0.545 |
Token | make |
Token ID | 787 |
Feature activation | -7.33e+00 |
Pos logit contributions | |
beh | +0.024 |
unnecessary | +0.021 |
yours | +0.021 |
rying | +0.020 |
eking | +0.020 |
Neg logit contributions | |
little | -0.014 |
boy | -0.013 |
man | -0.012 |
was | -0.012 |
bird | -0.012 |
Token | a |
Token ID | 257 |
Feature activation | -3.14e+00 |
Pos logit contributions | |
beh | +0.504 |
rying | +0.434 |
owing | +0.417 |
possible | +0.416 |
yours | +0.415 |
Neg logit contributions | |
little | -0.309 |
boy | -0.285 |
man | -0.258 |
girl | -0.257 |
big | -0.243 |
Token | car |
Token ID | 1097 |
Feature activation | +5.03e+00 |
Pos logit contributions | |
beh | +0.228 |
yours | +0.224 |
possible | +0.203 |
ours | +0.198 |
unnecessary | +0.193 |
Neg logit contributions | |
boy | -0.087 |
little | -0.083 |
girl | -0.081 |
sky | -0.078 |
man | -0.075 |
Token | like |
Token ID | 588 |
Feature activation | -1.04e+01 |
Pos logit contributions | |
boy | +0.253 |
man | +0.236 |
little | +0.234 |
girl | +0.229 |
and | +0.219 |
Neg logit contributions | |
beh | -0.299 |
yours | -0.262 |
unnecessary | -0.252 |
irs | -0.237 |
eking | -0.236 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.09e+01 |
Pos logit contributions | |
beh | +0.656 |
rying | +0.539 |
unnecessary | +0.524 |
yours | +0.521 |
eking | +0.518 |
Neg logit contributions | |
little | -0.465 |
boy | -0.455 |
sky | -0.425 |
man | -0.424 |
girl | -0.422 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.33e+01 |
Pos logit contributions | |
beh | +1.948 |
yours | +1.571 |
unnecessary | +1.548 |
irs | +1.492 |
eking | +1.485 |
Neg logit contributions | |
boy | -1.559 |
little | -1.456 |
man | -1.437 |
girl | -1.437 |
and | -1.404 |
Token | Tom |
Token ID | 4186 |
Feature activation | -4.59e+00 |
Pos logit contributions | |
beh | +0.834 |
yours | +0.799 |
possible | +0.708 |
ours | +0.684 |
theirs | +0.675 |
Neg logit contributions | |
boy | -0.524 |
little | -0.503 |
girl | -0.495 |
sky | -0.473 |
man | -0.470 |
Token | said |
Token ID | 531 |
Feature activation | +6.58e-01 |
Pos logit contributions | |
beh | +0.308 |
yours | +0.273 |
eking | +0.245 |
ieve | +0.241 |
possible | +0.240 |
Neg logit contributions | |
sky | -0.154 |
boy | -0.147 |
girl | -0.145 |
little | -0.143 |
man | -0.133 |
Token | , |
Token ID | 11 |
Feature activation | +9.28e+00 |
Pos logit contributions | |
little | +0.017 |
boy | +0.017 |
girl | +0.015 |
man | +0.014 |
was | +0.014 |
Neg logit contributions | |
beh | -0.054 |
yours | -0.053 |
ours | -0.048 |
possible | -0.048 |
unnecessary | -0.047 |
Token | whining |
Token ID | 48978 |
Feature activation | -6.29e+00 |
Pos logit contributions | |
boy | +0.446 |
little | +0.427 |
girl | +0.425 |
sky | +0.409 |
man | +0.407 |
Neg logit contributions | |
beh | -0.509 |
yours | -0.479 |
possible | -0.418 |
ours | -0.399 |
theirs | -0.393 |
Token | take |
Token ID | 1011 |
Feature activation | -8.40e+00 |
Pos logit contributions | |
beh | +0.717 |
yours | +0.701 |
possible | +0.655 |
unnecessary | +0.644 |
ieve | +0.633 |
Neg logit contributions | |
boy | -0.457 |
little | -0.443 |
was | -0.419 |
man | -0.406 |
girl | -0.400 |
Token | things |
Token ID | 1243 |
Feature activation | -1.41e+00 |
Pos logit contributions | |
beh | +0.560 |
yours | +0.498 |
possible | +0.477 |
rying | +0.462 |
eking | +0.454 |
Neg logit contributions | |
little | -0.346 |
boy | -0.328 |
girl | -0.302 |
man | -0.287 |
sky | -0.279 |
Token | that |
Token ID | 326 |
Feature activation | -8.46e+00 |
Pos logit contributions | |
beh | +0.101 |
yours | +0.087 |
unnecessary | +0.082 |
ility | +0.082 |
eking | +0.082 |
Neg logit contributions | |
boy | -0.051 |
little | -0.051 |
man | -0.050 |
and | -0.048 |
girl | -0.045 |
Token | are |
Token ID | 389 |
Feature activation | -3.21e+00 |
Pos logit contributions | |
beh | +0.512 |
yours | +0.482 |
possible | +0.429 |
ours | +0.409 |
unnecessary | +0.408 |
Neg logit contributions | |
boy | -0.373 |
little | -0.359 |
girl | -0.350 |
man | -0.338 |
sky | -0.333 |
Token | not |
Token ID | 407 |
Feature activation | -1.14e+01 |
Pos logit contributions | |
beh | +0.197 |
ieve | +0.159 |
yours | +0.158 |
irs | +0.156 |
anks | +0.153 |
Neg logit contributions | |
little | -0.152 |
boy | -0.143 |
was | -0.132 |
man | -0.132 |
girl | -0.128 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.09e+01 |
Pos logit contributions | |
beh | +0.471 |
ieve | +0.357 |
eking | +0.357 |
rying | +0.342 |
anks | +0.326 |
Neg logit contributions | |
little | -0.783 |
boy | -0.757 |
girl | -0.712 |
man | -0.701 |
and | -0.683 |
Token | ," |
Token ID | 553 |
Feature activation | -7.37e+00 |
Pos logit contributions | |
beh | +2.328 |
yours | +2.031 |
unnecessary | +1.892 |
ieve | +1.884 |
eking | +1.873 |
Neg logit contributions | |
boy | -1.073 |
little | -1.033 |
was | -0.969 |
man | -0.955 |
girl | -0.948 |
Token | she |
Token ID | 673 |
Feature activation | -4.92e+00 |
Pos logit contributions | |
beh | +0.430 |
yours | +0.414 |
possible | +0.368 |
ours | +0.353 |
theirs | +0.352 |
Neg logit contributions | |
boy | -0.316 |
little | -0.306 |
girl | -0.300 |
sky | -0.292 |
man | -0.283 |
Token | says |
Token ID | 1139 |
Feature activation | -9.06e-01 |
Pos logit contributions | |
beh | +0.305 |
yours | +0.286 |
possible | +0.252 |
ieve | +0.248 |
ours | +0.248 |
Neg logit contributions | |
boy | -0.192 |
little | -0.185 |
girl | -0.182 |
sky | -0.181 |
man | -0.170 |
Token | . |
Token ID | 13 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
beh | +0.061 |
yours | +0.059 |
ours | +0.052 |
possible | +0.052 |
unnecessary | +0.051 |
Neg logit contributions | |
little | -0.035 |
boy | -0.034 |
man | -0.033 |
and | -0.033 |
girl | -0.032 |
Token | " |
Token ID | 366 |
Feature activation | +5.53e+00 |
Pos logit contributions | |
yours | +0.698 |
beh | +0.680 |
possible | +0.634 |
theirs | +0.618 |
ourselves | +0.595 |
Neg logit contributions | |
boy | -0.298 |
girl | -0.280 |
little | -0.263 |
child | -0.224 |
sky | -0.223 |
Token | touch |
Token ID | 3638 |
Feature activation | -8.28e+00 |
Pos logit contributions | |
beh | +1.100 |
yours | +1.046 |
possible | +0.942 |
ours | +0.896 |
theirs | +0.885 |
Neg logit contributions | |
little | -0.806 |
boy | -0.793 |
girl | -0.740 |
was | -0.738 |
man | -0.735 |
Token | something |
Token ID | 1223 |
Feature activation | -1.41e+01 |
Pos logit contributions | |
beh | +0.555 |
yours | +0.490 |
possible | +0.451 |
rying | +0.430 |
eking | +0.429 |
Neg logit contributions | |
little | -0.331 |
boy | -0.329 |
girl | -0.310 |
sky | -0.300 |
man | -0.297 |
Token | that |
Token ID | 326 |
Feature activation | -8.46e+00 |
Pos logit contributions | |
beh | +0.941 |
yours | +0.826 |
possible | +0.740 |
theirs | +0.724 |
rying | +0.714 |
Neg logit contributions | |
little | -0.624 |
boy | -0.591 |
girl | -0.554 |
man | -0.552 |
was | -0.543 |
Token | is |
Token ID | 318 |
Feature activation | +2.74e+00 |
Pos logit contributions | |
beh | +0.593 |
yours | +0.547 |
possible | +0.501 |
unnecessary | +0.487 |
ieve | +0.483 |
Neg logit contributions | |
boy | -0.301 |
little | -0.287 |
girl | -0.273 |
man | -0.270 |
was | -0.266 |
Token | not |
Token ID | 407 |
Feature activation | -1.14e+01 |
Pos logit contributions | |
little | +0.124 |
boy | +0.118 |
man | +0.112 |
girl | +0.110 |
was | +0.108 |
Neg logit contributions | |
beh | -0.171 |
yours | -0.145 |
ieve | -0.134 |
irs | -0.129 |
unnecessary | -0.128 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.09e+01 |
Pos logit contributions | |
beh | +0.622 |
ieve | +0.481 |
anks | +0.465 |
eking | +0.462 |
irs | +0.457 |
Neg logit contributions | |
little | -0.640 |
boy | -0.610 |
girl | -0.576 |
man | -0.564 |
and | -0.561 |
Token | . |
Token ID | 13 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
beh | +2.246 |
yours | +1.857 |
eking | +1.776 |
unnecessary | +1.769 |
ieve | +1.737 |
Neg logit contributions | |
little | -1.171 |
boy | -1.167 |
and | -1.139 |
was | -1.115 |
man | -1.075 |
Token | And |
Token ID | 843 |
Feature activation | +2.91e+00 |
Pos logit contributions | |
beh | +0.621 |
yours | +0.580 |
possible | +0.520 |
ours | +0.492 |
unnecessary | +0.489 |
Neg logit contributions | |
boy | -0.479 |
little | -0.464 |
girl | -0.454 |
sky | -0.437 |
man | -0.433 |
Token | you |
Token ID | 345 |
Feature activation | -1.80e+01 |
Pos logit contributions | |
little | +0.162 |
boy | +0.162 |
girl | +0.155 |
man | +0.149 |
was | +0.147 |
Neg logit contributions | |
beh | -0.146 |
yours | -0.128 |
possible | -0.110 |
unnecessary | -0.106 |
rying | -0.105 |
Token | have |
Token ID | 423 |
Feature activation | -7.36e+00 |
Pos logit contributions | |
beh | +1.003 |
yours | +0.943 |
possible | +0.816 |
ours | +0.802 |
unnecessary | +0.777 |
Neg logit contributions | |
boy | -0.845 |
girl | -0.815 |
little | -0.805 |
sky | -0.777 |
man | -0.775 |
Token | to |
Token ID | 284 |
Feature activation | -3.42e-01 |
Pos logit contributions | |
beh | +0.465 |
yours | +0.437 |
possible | +0.391 |
ours | +0.375 |
ieve | +0.373 |
Neg logit contributions | |
boy | -0.302 |
little | -0.294 |
girl | -0.284 |
man | -0.272 |
sky | -0.271 |
Token | the |
Token ID | 262 |
Feature activation | +4.64e+00 |
Pos logit contributions | |
beh | +0.408 |
yours | +0.374 |
possible | +0.345 |
eking | +0.330 |
ieve | +0.327 |
Neg logit contributions | |
boy | -0.259 |
little | -0.256 |
girl | -0.244 |
man | -0.232 |
sky | -0.226 |
Token | toys |
Token ID | 14958 |
Feature activation | +2.24e+00 |
Pos logit contributions | |
boy | +0.163 |
little | +0.156 |
girl | +0.153 |
man | +0.145 |
sky | +0.144 |
Neg logit contributions | |
beh | -0.310 |
yours | -0.301 |
possible | -0.268 |
ours | -0.262 |
theirs | -0.256 |
Token | that |
Token ID | 326 |
Feature activation | -8.46e+00 |
Pos logit contributions | |
boy | +0.075 |
little | +0.074 |
girl | +0.070 |
man | +0.067 |
sky | +0.064 |
Neg logit contributions | |
beh | -0.163 |
yours | -0.146 |
possible | -0.135 |
eking | -0.135 |
unnecessary | -0.135 |
Token | are |
Token ID | 389 |
Feature activation | -3.21e+00 |
Pos logit contributions | |
beh | +0.507 |
yours | +0.466 |
possible | +0.418 |
unnecessary | +0.408 |
eking | +0.399 |
Neg logit contributions | |
boy | -0.384 |
little | -0.366 |
girl | -0.359 |
man | -0.351 |
sky | -0.343 |
Token | not |
Token ID | 407 |
Feature activation | -1.14e+01 |
Pos logit contributions | |
beh | +0.198 |
yours | +0.164 |
ieve | +0.155 |
unnecessary | +0.154 |
rying | +0.152 |
Neg logit contributions | |
little | -0.148 |
boy | -0.140 |
girl | -0.130 |
man | -0.130 |
was | -0.126 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.09e+01 |
Pos logit contributions | |
beh | +0.718 |
ieve | +0.600 |
rying | +0.577 |
eking | +0.573 |
anks | +0.572 |
Neg logit contributions | |
little | -0.538 |
boy | -0.507 |
girl | -0.472 |
and | -0.439 |
man | -0.436 |
Token | and |
Token ID | 290 |
Feature activation | +1.22e+01 |
Pos logit contributions | |
beh | +2.164 |
yours | +1.826 |
unnecessary | +1.743 |
eking | +1.738 |
possible | +1.714 |
Neg logit contributions | |
boy | -1.232 |
little | -1.204 |
and | -1.158 |
was | -1.148 |
girl | -1.130 |
Token | say |
Token ID | 910 |
Feature activation | -9.42e+00 |
Pos logit contributions | |
boy | +0.588 |
little | +0.587 |
girl | +0.549 |
man | +0.539 |
was | +0.528 |
Neg logit contributions | |
beh | -0.701 |
yours | -0.582 |
rying | -0.578 |
eking | -0.575 |
ieve | -0.566 |
Token | sorry |
Token ID | 7926 |
Feature activation | -1.32e+01 |
Pos logit contributions | |
beh | +0.570 |
yours | +0.453 |
rying | +0.444 |
possible | +0.425 |
unnecessary | +0.414 |
Neg logit contributions | |
little | -0.475 |
boy | -0.445 |
man | -0.444 |
girl | -0.421 |
sky | -0.417 |
Token | . |
Token ID | 13 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
beh | +0.852 |
yours | +0.739 |
ieve | +0.707 |
possible | +0.705 |
eking | +0.705 |
Neg logit contributions | |
boy | -0.603 |
little | -0.574 |
girl | -0.556 |
was | -0.542 |
and | -0.540 |
Token | And |
Token ID | 843 |
Feature activation | +2.94e+00 |
Pos logit contributions | |
beh | +0.527 |
yours | +0.487 |
possible | +0.425 |
ours | +0.399 |
unnecessary | +0.397 |
Neg logit contributions | |
boy | -0.577 |
little | -0.560 |
girl | -0.551 |
sky | -0.531 |
man | -0.530 |
Token | take |
Token ID | 1011 |
Feature activation | -8.43e+00 |
Pos logit contributions | |
beh | +1.142 |
yours | +1.069 |
possible | +0.997 |
ours | +0.931 |
unnecessary | +0.929 |
Neg logit contributions | |
little | -0.780 |
boy | -0.759 |
was | -0.724 |
girl | -0.705 |
and | -0.688 |
Token | something |
Token ID | 1223 |
Feature activation | -1.41e+01 |
Pos logit contributions | |
beh | +0.550 |
yours | +0.473 |
possible | +0.466 |
rying | +0.457 |
eking | +0.440 |
Neg logit contributions | |
little | -0.362 |
boy | -0.342 |
girl | -0.325 |
man | -0.311 |
was | -0.304 |
Token | that |
Token ID | 326 |
Feature activation | -8.48e+00 |
Pos logit contributions | |
beh | +0.959 |
yours | +0.855 |
possible | +0.784 |
ility | +0.756 |
unnecessary | +0.751 |
Neg logit contributions | |
little | -0.604 |
boy | -0.549 |
girl | -0.522 |
man | -0.516 |
and | -0.490 |
Token | is |
Token ID | 318 |
Feature activation | +2.75e+00 |
Pos logit contributions | |
beh | +0.609 |
yours | +0.562 |
possible | +0.520 |
unnecessary | +0.507 |
rying | +0.501 |
Neg logit contributions | |
boy | -0.294 |
little | -0.276 |
girl | -0.266 |
man | -0.261 |
was | -0.257 |
Token | not |
Token ID | 407 |
Feature activation | -1.14e+01 |
Pos logit contributions | |
little | +0.128 |
boy | +0.120 |
man | +0.114 |
girl | +0.112 |
was | +0.111 |
Neg logit contributions | |
beh | -0.171 |
yours | -0.138 |
ieve | -0.135 |
unnecessary | -0.129 |
irs | -0.129 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.09e+01 |
Pos logit contributions | |
beh | +0.526 |
ieve | +0.394 |
eking | +0.368 |
rying | +0.365 |
irs | +0.355 |
Neg logit contributions | |
little | -0.752 |
boy | -0.719 |
girl | -0.687 |
and | -0.670 |
man | -0.663 |
Token | . |
Token ID | 13 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
beh | +2.187 |
yours | +1.773 |
eking | +1.743 |
unnecessary | +1.737 |
ieve | +1.691 |
Neg logit contributions | |
boy | -1.240 |
little | -1.232 |
and | -1.209 |
was | -1.185 |
girl | -1.143 |
Token | Now |
Token ID | 2735 |
Feature activation | -1.90e-01 |
Pos logit contributions | |
beh | +0.612 |
yours | +0.566 |
possible | +0.517 |
eking | +0.488 |
unnecessary | +0.485 |
Neg logit contributions | |
boy | -0.490 |
little | -0.474 |
girl | -0.464 |
sky | -0.444 |
man | -0.439 |
Token | , |
Token ID | 11 |
Feature activation | +9.27e+00 |
Pos logit contributions | |
beh | +0.015 |
yours | +0.013 |
possible | +0.012 |
unnecessary | +0.012 |
ieve | +0.012 |
Neg logit contributions | |
little | -0.006 |
boy | -0.006 |
girl | -0.005 |
man | -0.005 |
was | -0.005 |
Token | give |
Token ID | 1577 |
Feature activation | -8.17e+00 |
Pos logit contributions | |
boy | +0.479 |
little | +0.469 |
girl | +0.456 |
man | +0.439 |
was | +0.426 |
Neg logit contributions | |
beh | -0.498 |
yours | -0.450 |
possible | -0.400 |
unnecessary | -0.379 |
ours | -0.378 |
Token | me |
Token ID | 502 |
Feature activation | -6.03e+00 |
Pos logit contributions | |
beh | +0.469 |
yours | +0.435 |
possible | +0.389 |
ours | +0.367 |
unnecessary | +0.364 |
Neg logit contributions | |
boy | -0.390 |
little | -0.380 |
girl | -0.371 |
man | -0.357 |
sky | -0.349 |
Token | it |
Token ID | 340 |
Feature activation | -4.91e+00 |
Pos logit contributions | |
little | +0.116 |
boy | +0.110 |
man | +0.103 |
girl | +0.102 |
was | +0.100 |
Neg logit contributions | |
beh | -0.166 |
yours | -0.137 |
rying | -0.135 |
possible | -0.130 |
eking | -0.129 |
Token | . |
Token ID | 13 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
beh | +0.325 |
yours | +0.286 |
irs | +0.267 |
unnecessary | +0.266 |
ales | +0.262 |
Neg logit contributions | |
little | -0.203 |
boy | -0.191 |
and | -0.182 |
was | -0.179 |
girl | -0.179 |
Token | It |
Token ID | 632 |
Feature activation | -1.14e+00 |
Pos logit contributions | |
beh | +0.637 |
yours | +0.583 |
possible | +0.530 |
eking | +0.506 |
unnecessary | +0.505 |
Neg logit contributions | |
boy | -0.471 |
little | -0.458 |
girl | -0.445 |
sky | -0.433 |
man | -0.422 |
Token | is |
Token ID | 318 |
Feature activation | +2.75e+00 |
Pos logit contributions | |
beh | +0.086 |
yours | +0.082 |
possible | +0.075 |
ours | +0.072 |
unnecessary | +0.072 |
Neg logit contributions | |
boy | -0.032 |
little | -0.031 |
girl | -0.030 |
man | -0.028 |
sky | -0.027 |
Token | not |
Token ID | 407 |
Feature activation | -1.14e+01 |
Pos logit contributions | |
little | +0.113 |
boy | +0.109 |
girl | +0.101 |
man | +0.100 |
sky | +0.099 |
Neg logit contributions | |
beh | -0.184 |
yours | -0.153 |
rying | -0.147 |
unnecessary | -0.143 |
ieve | -0.143 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.09e+01 |
Pos logit contributions | |
beh | +0.807 |
rying | +0.676 |
yours | +0.672 |
eking | +0.666 |
ieve | +0.657 |
Neg logit contributions | |
little | -0.447 |
boy | -0.431 |
girl | -0.400 |
man | -0.374 |
and | -0.348 |
Token | to |
Token ID | 284 |
Feature activation | -3.64e-01 |
Pos logit contributions | |
beh | +2.269 |
yours | +1.972 |
eking | +1.838 |
unnecessary | +1.838 |
possible | +1.802 |
Neg logit contributions | |
boy | -1.120 |
little | -1.079 |
girl | -1.034 |
man | -1.019 |
was | -1.018 |
Token | change |
Token ID | 1487 |
Feature activation | -4.13e+00 |
Pos logit contributions | |
beh | +0.024 |
yours | +0.021 |
rying | +0.020 |
unnecessary | +0.020 |
possible | +0.020 |
Neg logit contributions | |
little | -0.016 |
boy | -0.015 |
man | -0.014 |
girl | -0.014 |
was | -0.013 |
Token | ." |
Token ID | 526 |
Feature activation | -1.46e+01 |
Pos logit contributions | |
beh | +0.270 |
yours | +0.234 |
possible | +0.213 |
ieve | +0.211 |
eking | +0.210 |
Neg logit contributions | |
little | -0.176 |
boy | -0.175 |
girl | -0.165 |
man | -0.159 |
was | -0.156 |
Token | Tom |
Token ID | 4186 |
Feature activation | -4.57e+00 |
Pos logit contributions | |
beh | +1.045 |
yours | +1.033 |
possible | +0.938 |
theirs | +0.906 |
ourselves | +0.902 |
Neg logit contributions | |
boy | -0.371 |
little | -0.355 |
girl | -0.347 |
sky | -0.339 |
man | -0.326 |
Token | said |
Token ID | 531 |
Feature activation | +6.42e-01 |
Pos logit contributions | |
beh | +0.232 |
yours | +0.188 |
myself | +0.176 |
ares | +0.171 |
eking | +0.171 |
Neg logit contributions | |
sky | -0.193 |
field | -0.179 |
valley | -0.176 |
landscape | -0.168 |
woods | -0.167 |
Token | they |
Token ID | 484 |
Feature activation | -2.43e+00 |
Pos logit contributions | |
boy | +0.444 |
little | +0.430 |
girl | +0.424 |
man | +0.409 |
sky | +0.408 |
Neg logit contributions | |
beh | -0.511 |
yours | -0.489 |
possible | -0.425 |
ours | -0.409 |
unnecessary | -0.398 |
Token | have |
Token ID | 423 |
Feature activation | -7.37e+00 |
Pos logit contributions | |
beh | +0.145 |
yours | +0.136 |
possible | +0.121 |
ours | +0.115 |
unnecessary | +0.114 |
Neg logit contributions | |
boy | -0.108 |
little | -0.105 |
girl | -0.102 |
man | -0.098 |
sky | -0.098 |
Token | a |
Token ID | 257 |
Feature activation | -3.14e+00 |
Pos logit contributions | |
beh | +0.485 |
yours | +0.462 |
possible | +0.413 |
ours | +0.398 |
unnecessary | +0.393 |
Neg logit contributions | |
boy | -0.280 |
little | -0.269 |
girl | -0.262 |
man | -0.250 |
sky | -0.248 |
Token | trumpet |
Token ID | 40334 |
Feature activation | -3.53e-01 |
Pos logit contributions | |
yours | +0.197 |
beh | +0.183 |
possible | +0.178 |
unnecessary | +0.166 |
ours | +0.165 |
Neg logit contributions | |
boy | -0.103 |
girl | -0.098 |
sky | -0.094 |
little | -0.092 |
man | -0.090 |
Token | like |
Token ID | 588 |
Feature activation | -1.04e+01 |
Pos logit contributions | |
beh | +0.019 |
yours | +0.017 |
unnecessary | +0.015 |
possible | +0.015 |
eking | +0.014 |
Neg logit contributions | |
boy | -0.019 |
little | -0.018 |
girl | -0.018 |
man | -0.018 |
sky | -0.017 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.09e+01 |
Pos logit contributions | |
beh | +0.734 |
yours | +0.639 |
rying | +0.601 |
possible | +0.596 |
unnecessary | +0.591 |
Neg logit contributions | |
little | -0.379 |
boy | -0.378 |
girl | -0.347 |
sky | -0.345 |
man | -0.339 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.33e+01 |
Pos logit contributions | |
beh | +1.878 |
yours | +1.549 |
unnecessary | +1.473 |
eking | +1.386 |
possible | +1.380 |
Neg logit contributions | |
boy | -1.547 |
little | -1.479 |
girl | -1.444 |
man | -1.441 |
was | -1.404 |
Token | Tom |
Token ID | 4186 |
Feature activation | -4.56e+00 |
Pos logit contributions | |
beh | +0.671 |
yours | +0.653 |
theirs | +0.558 |
ours | +0.554 |
possible | +0.551 |
Neg logit contributions | |
boy | -0.641 |
girl | -0.630 |
little | -0.625 |
sky | -0.600 |
man | -0.583 |
Token | says |
Token ID | 1139 |
Feature activation | -8.91e-01 |
Pos logit contributions | |
beh | +0.284 |
yours | +0.271 |
possible | +0.237 |
ours | +0.233 |
ieve | +0.231 |
Neg logit contributions | |
boy | -0.176 |
little | -0.173 |
girl | -0.168 |
sky | -0.161 |
man | -0.159 |
Token | . |
Token ID | 13 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
beh | +0.064 |
yours | +0.062 |
ours | +0.056 |
unnecessary | +0.055 |
possible | +0.054 |
Neg logit contributions | |
little | -0.033 |
boy | -0.032 |
and | -0.031 |
girl | -0.029 |
man | -0.029 |
Token |
|
Token ID | 198 |
Feature activation | +5.37e+00 |
Pos logit contributions | |
yours | +0.595 |
beh | +0.576 |
theirs | +0.574 |
possible | +0.535 |
properly | +0.523 |
Neg logit contributions | |
girl | -0.296 |
boy | -0.284 |
field | -0.278 |
sky | -0.278 |
child | -0.278 |
Token | take |
Token ID | 1011 |
Feature activation | -8.42e+00 |
Pos logit contributions | |
beh | +0.708 |
yours | +0.662 |
possible | +0.629 |
unnecessary | +0.622 |
ieve | +0.606 |
Neg logit contributions | |
boy | -0.480 |
little | -0.476 |
was | -0.440 |
man | -0.426 |
girl | -0.424 |
Token | things |
Token ID | 1243 |
Feature activation | -1.42e+00 |
Pos logit contributions | |
beh | +0.555 |
yours | +0.475 |
rying | +0.466 |
possible | +0.456 |
eking | +0.443 |
Neg logit contributions | |
little | -0.358 |
boy | -0.344 |
girl | -0.319 |
man | -0.312 |
was | -0.303 |
Token | that |
Token ID | 326 |
Feature activation | -8.46e+00 |
Pos logit contributions | |
beh | +0.098 |
yours | +0.083 |
unnecessary | +0.081 |
eking | +0.081 |
ieve | +0.078 |
Neg logit contributions | |
little | -0.054 |
man | -0.053 |
boy | -0.052 |
and | -0.051 |
girl | -0.048 |
Token | are |
Token ID | 389 |
Feature activation | -3.22e+00 |
Pos logit contributions | |
beh | +0.511 |
yours | +0.472 |
possible | +0.426 |
unnecessary | +0.407 |
rying | +0.405 |
Neg logit contributions | |
boy | -0.383 |
little | -0.366 |
girl | -0.356 |
man | -0.352 |
was | -0.341 |
Token | not |
Token ID | 407 |
Feature activation | -1.14e+01 |
Pos logit contributions | |
beh | +0.189 |
ieve | +0.151 |
yours | +0.148 |
anks | +0.146 |
irs | +0.145 |
Neg logit contributions | |
little | -0.157 |
boy | -0.147 |
man | -0.139 |
was | -0.139 |
and | -0.135 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.09e+01 |
Pos logit contributions | |
beh | +0.298 |
ieve | +0.176 |
rying | +0.175 |
eking | +0.172 |
anks | +0.148 |
Neg logit contributions | |
little | -0.948 |
boy | -0.913 |
girl | -0.879 |
man | -0.871 |
and | -0.855 |
Token | ." |
Token ID | 526 |
Feature activation | -1.46e+01 |
Pos logit contributions | |
beh | +1.923 |
yours | +1.637 |
unnecessary | +1.528 |
eking | +1.523 |
ieve | +1.507 |
Neg logit contributions | |
boy | -1.405 |
little | -1.391 |
girl | -1.322 |
was | -1.322 |
man | -1.317 |
Token |
|
Token ID | 198 |
Feature activation | +5.36e+00 |
Pos logit contributions | |
beh | +1.127 |
yours | +1.099 |
possible | +0.999 |
ourselves | +0.984 |
ours | +0.981 |
Neg logit contributions | |
boy | -0.306 |
little | -0.296 |
girl | -0.289 |
sky | -0.255 |
man | -0.239 |
Token |
|
Token ID | 198 |
Feature activation | +5.37e+00 |
Pos logit contributions | |
boy | +0.112 |
girl | +0.108 |
little | +0.099 |
sky | +0.095 |
man | +0.085 |
Neg logit contributions | |
beh | -0.415 |
yours | -0.410 |
possible | -0.374 |
ourselves | -0.367 |
theirs | -0.365 |
Token | L |
Token ID | 43 |
Feature activation | +8.70e+00 |
Pos logit contributions | |
boy | +0.176 |
little | +0.172 |
girl | +0.171 |
sky | +0.168 |
man | +0.144 |
Neg logit contributions | |
beh | -0.362 |
yours | -0.344 |
possible | -0.312 |
ourselves | -0.309 |
ours | -0.307 |
Token | ily |
Token ID | 813 |
Feature activation | -2.69e+00 |
Pos logit contributions | |
boy | +0.446 |
little | +0.436 |
girl | +0.426 |
man | +0.410 |
sky | +0.405 |
Neg logit contributions | |
beh | -0.463 |
yours | -0.430 |
possible | -0.379 |
unnecessary | -0.357 |
ours | -0.357 |
Token | " |
Token ID | 366 |
Feature activation | +5.54e+00 |
Pos logit contributions | |
boy | +0.263 |
girl | +0.242 |
sky | +0.215 |
little | +0.204 |
man | +0.199 |
Neg logit contributions | |
beh | -0.643 |
yours | -0.635 |
possible | -0.568 |
eking | -0.558 |
theirs | -0.551 |
Token | Yep |
Token ID | 47834 |
Feature activation | -1.11e+00 |
Pos logit contributions | |
boy | +0.195 |
little | +0.193 |
girl | +0.181 |
man | +0.175 |
sky | +0.173 |
Neg logit contributions | |
beh | -0.375 |
yours | -0.359 |
possible | -0.324 |
ours | -0.308 |
ourselves | -0.304 |
Token | , |
Token ID | 11 |
Feature activation | +9.27e+00 |
Pos logit contributions | |
beh | +0.091 |
yours | +0.081 |
possible | +0.076 |
unnecessary | +0.075 |
ieve | +0.074 |
Neg logit contributions | |
little | -0.031 |
boy | -0.030 |
girl | -0.027 |
man | -0.027 |
and | -0.023 |
Token | that |
Token ID | 326 |
Feature activation | -8.46e+00 |
Pos logit contributions | |
boy | +0.409 |
little | +0.397 |
girl | +0.387 |
man | +0.372 |
sky | +0.369 |
Neg logit contributions | |
beh | -0.553 |
yours | -0.521 |
possible | -0.462 |
ours | -0.442 |
unnecessary | -0.437 |
Token | 's |
Token ID | 338 |
Feature activation | +5.50e+00 |
Pos logit contributions | |
beh | +0.624 |
yours | +0.612 |
possible | +0.554 |
ours | +0.537 |
ourselves | +0.522 |
Neg logit contributions | |
boy | -0.234 |
little | -0.221 |
girl | -0.218 |
sky | -0.204 |
man | -0.202 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.08e+01 |
Pos logit contributions | |
little | +0.228 |
boy | +0.219 |
girl | +0.200 |
was | +0.200 |
man | +0.198 |
Neg logit contributions | |
beh | -0.372 |
yours | -0.308 |
unnecessary | -0.295 |
ieve | -0.288 |
rying | -0.286 |
Token | now |
Token ID | 783 |
Feature activation | -9.89e+00 |
Pos logit contributions | |
beh | +1.916 |
yours | +1.674 |
unnecessary | +1.542 |
possible | +1.509 |
ieve | +1.489 |
Neg logit contributions | |
boy | -1.399 |
little | -1.366 |
girl | -1.320 |
man | -1.292 |
was | -1.290 |
Token | . |
Token ID | 13 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
beh | +0.605 |
unnecessary | +0.484 |
yours | +0.479 |
ieve | +0.457 |
irs | +0.448 |
Neg logit contributions | |
little | -0.482 |
boy | -0.460 |
was | -0.450 |
and | -0.447 |
man | -0.445 |
Token | What |
Token ID | 1867 |
Feature activation | -7.54e+00 |
Pos logit contributions | |
beh | +0.584 |
yours | +0.555 |
possible | +0.479 |
ours | +0.461 |
theirs | +0.449 |
Neg logit contributions | |
boy | -0.509 |
little | -0.494 |
girl | -0.485 |
man | -0.470 |
sky | -0.464 |
Token | are |
Token ID | 389 |
Feature activation | -3.20e+00 |
Pos logit contributions | |
beh | +0.518 |
yours | +0.492 |
possible | +0.443 |
ours | +0.428 |
unnecessary | +0.424 |
Neg logit contributions | |
boy | -0.267 |
little | -0.256 |
girl | -0.249 |
man | -0.236 |
sky | -0.233 |
Token | you |
Token ID | 345 |
Feature activation | -1.80e+01 |
Pos logit contributions | |
beh | +0.168 |
ility | +0.144 |
unnecessary | +0.138 |
owing | +0.137 |
rying | +0.137 |
Neg logit contributions | |
boy | -0.164 |
little | -0.162 |
was | -0.153 |
girl | -0.152 |
man | -0.151 |
Token | ! |
Token ID | 0 |
Feature activation | -1.25e+01 |
Pos logit contributions | |
boy | +0.101 |
little | +0.096 |
girl | +0.096 |
man | +0.094 |
and | +0.091 |
Neg logit contributions | |
beh | -0.169 |
unnecessary | -0.143 |
yours | -0.139 |
irs | -0.137 |
ieve | -0.134 |
Token | It |
Token ID | 632 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
beh | +0.786 |
yours | +0.730 |
possible | +0.664 |
unnecessary | +0.636 |
eking | +0.632 |
Neg logit contributions | |
boy | -0.521 |
little | -0.501 |
girl | -0.489 |
man | -0.464 |
sky | -0.463 |
Token | 's |
Token ID | 338 |
Feature activation | +5.50e+00 |
Pos logit contributions | |
beh | +0.087 |
yours | +0.083 |
possible | +0.076 |
ours | +0.074 |
ieve | +0.072 |
Neg logit contributions | |
boy | -0.029 |
little | -0.028 |
girl | -0.027 |
man | -0.025 |
sky | -0.025 |
Token | bigger |
Token ID | 5749 |
Feature activation | +4.68e-01 |
Pos logit contributions | |
little | +0.230 |
boy | +0.216 |
sky | +0.204 |
was | +0.202 |
girl | +0.201 |
Neg logit contributions | |
beh | -0.363 |
yours | -0.297 |
unnecessary | -0.288 |
rying | -0.286 |
ieve | -0.282 |
Token | than |
Token ID | 621 |
Feature activation | -1.02e+01 |
Pos logit contributions | |
boy | +0.019 |
little | +0.018 |
girl | +0.017 |
man | +0.017 |
sky | +0.016 |
Neg logit contributions | |
beh | -0.031 |
yours | -0.028 |
possible | -0.025 |
unnecessary | -0.025 |
ieve | -0.025 |
Token | yours |
Token ID | 12431 |
Feature activation | -3.08e+01 |
Pos logit contributions | |
beh | +0.462 |
ieve | +0.330 |
unnecessary | +0.302 |
rying | +0.301 |
yours | +0.300 |
Neg logit contributions | |
little | -0.653 |
boy | -0.618 |
girl | -0.596 |
sky | -0.594 |
was | -0.584 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.33e+01 |
Pos logit contributions | |
beh | +1.883 |
yours | +1.505 |
unnecessary | +1.466 |
eking | +1.406 |
ieve | +1.387 |
Neg logit contributions | |
boy | -1.508 |
little | -1.500 |
girl | -1.433 |
was | -1.429 |
man | -1.406 |
Token | Lily |
Token ID | 20037 |
Feature activation | -2.99e+00 |
Pos logit contributions | |
beh | +0.958 |
yours | +0.910 |
theirs | +0.809 |
ourselves | +0.791 |
anks | +0.787 |
Neg logit contributions | |
boy | -0.328 |
girl | -0.313 |
little | -0.309 |
sky | -0.302 |
man | -0.280 |
Token | felt |
Token ID | 2936 |
Feature activation | +4.93e+00 |
Pos logit contributions | |
beh | +0.184 |
myself | +0.154 |
yours | +0.153 |
properly | +0.147 |
correctly | +0.146 |
Neg logit contributions | |
sky | -0.097 |
girl | -0.092 |
boy | -0.088 |
woman | -0.085 |
field | -0.083 |
Token | sad |
Token ID | 6507 |
Feature activation | +1.63e-01 |
Pos logit contributions | |
boy | +0.187 |
little | +0.182 |
girl | +0.176 |
man | +0.168 |
sky | +0.165 |
Neg logit contributions | |
beh | -0.327 |
yours | -0.310 |
possible | -0.278 |
ours | -0.268 |
unnecessary | -0.263 |
Token | because |
Token ID | 780 |
Feature activation | -9.17e+00 |
Pos logit contributions | |
boy | +0.003 |
little | +0.003 |
girl | +0.003 |
man | +0.003 |
sky | +0.003 |
Neg logit contributions | |
beh | -0.014 |
yours | -0.013 |
possible | -0.012 |
ours | -0.012 |
unnecessary | -0.012 |
Token | You |
Token ID | 921 |
Feature activation | -1.01e+01 |
Pos logit contributions | |
beh | +0.349 |
yours | +0.328 |
possible | +0.282 |
ours | +0.268 |
unnecessary | +0.263 |
Neg logit contributions | |
boy | -0.365 |
little | -0.355 |
girl | -0.349 |
man | -0.338 |
sky | -0.336 |
Token | stood |
Token ID | 6204 |
Feature activation | +2.15e+00 |
Pos logit contributions | |
beh | +0.623 |
yours | +0.589 |
possible | +0.526 |
ours | +0.504 |
unnecessary | +0.499 |
Neg logit contributions | |
boy | -0.427 |
little | -0.414 |
girl | -0.402 |
man | -0.385 |
sky | -0.383 |
Token | up |
Token ID | 510 |
Feature activation | -1.05e+01 |
Pos logit contributions | |
boy | +0.101 |
little | +0.100 |
girl | +0.096 |
man | +0.093 |
sky | +0.092 |
Neg logit contributions | |
beh | -0.126 |
yours | -0.115 |
possible | -0.101 |
unnecessary | -0.096 |
ours | -0.096 |
Token | for |
Token ID | 329 |
Feature activation | -7.92e+00 |
Pos logit contributions | |
beh | +0.863 |
yours | +0.791 |
possible | +0.719 |
unnecessary | +0.715 |
ieve | +0.713 |
Neg logit contributions | |
little | -0.264 |
boy | -0.256 |
girl | -0.231 |
man | -0.225 |
sky | -0.221 |
Token | yourself |
Token ID | 3511 |
Feature activation | -2.31e+01 |
Pos logit contributions | |
beh | +0.588 |
yours | +0.496 |
rying | +0.476 |
possible | +0.473 |
anks | +0.461 |
Neg logit contributions | |
little | -0.299 |
boy | -0.262 |
girl | -0.244 |
man | -0.238 |
sky | -0.236 |
Token | and |
Token ID | 290 |
Feature activation | +1.47e+01 |
Pos logit contributions | |
beh | +1.694 |
yours | +1.523 |
unnecessary | +1.383 |
possible | +1.377 |
ieve | +1.372 |
Neg logit contributions | |
boy | -0.777 |
little | -0.776 |
girl | -0.722 |
man | -0.697 |
sky | -0.683 |
Token | Cro |
Token ID | 9325 |
Feature activation | +1.47e+01 |
Pos logit contributions | |
boy | +0.695 |
little | +0.672 |
girl | +0.654 |
man | +0.633 |
was | +0.632 |
Neg logit contributions | |
beh | -0.860 |
yours | -0.776 |
possible | -0.712 |
ieve | -0.676 |
rying | -0.675 |
Token | co |
Token ID | 1073 |
Feature activation | -2.75e+00 |
Pos logit contributions | |
boy | +0.711 |
girl | +0.673 |
little | +0.664 |
man | +0.643 |
child | +0.607 |
Neg logit contributions | |
yours | -0.724 |
beh | -0.717 |
possible | -0.675 |
ours | -0.606 |
ourselves | -0.602 |
Token | . |
Token ID | 13 |
Feature activation | -8.46e+00 |
Pos logit contributions | |
beh | +0.134 |
yours | +0.126 |
possible | +0.110 |
ours | +0.105 |
unnecessary | +0.105 |
Neg logit contributions | |
boy | -0.147 |
little | -0.144 |
man | -0.139 |
girl | -0.138 |
was | -0.138 |
Token | You |
Token ID | 921 |
Feature activation | -1.15e+01 |
Pos logit contributions | |
beh | +0.456 |
yours | +0.422 |
possible | +0.378 |
unnecessary | +0.356 |
ours | +0.354 |
Neg logit contributions | |
boy | -0.423 |
little | -0.410 |
girl | -0.403 |
sky | -0.388 |
man | -0.385 |
Token | know |
Token ID | 760 |
Feature activation | -9.99e+00 |
Pos logit contributions | |
beh | +0.707 |
yours | +0.670 |
possible | +0.596 |
ours | +0.572 |
unnecessary | +0.565 |
Neg logit contributions | |
boy | -0.486 |
little | -0.471 |
girl | -0.458 |
man | -0.440 |
sky | -0.437 |
Token | Once |
Token ID | 7454 |
Feature activation | +1.36e+01 |
Pos logit contributions | |
boy | +0.201 |
little | +0.200 |
girl | +0.192 |
man | +0.182 |
woman | +0.165 |
Neg logit contributions | |
beh | -0.471 |
yours | -0.467 |
possible | -0.424 |
ourselves | -0.402 |
theirs | -0.398 |
Token | , |
Token ID | 11 |
Feature activation | +9.30e+00 |
Pos logit contributions | |
boy | +0.501 |
little | +0.474 |
girl | +0.473 |
man | +0.454 |
sky | +0.452 |
Neg logit contributions | |
beh | -0.869 |
yours | -0.856 |
possible | -0.770 |
ours | -0.736 |
theirs | -0.730 |
Token | there |
Token ID | 612 |
Feature activation | -4.43e+00 |
Pos logit contributions | |
man | +0.412 |
boy | +0.406 |
sky | +0.397 |
girl | +0.392 |
little | +0.387 |
Neg logit contributions | |
yours | -0.496 |
beh | -0.473 |
possible | -0.412 |
theirs | -0.406 |
ours | -0.402 |
Token | was |
Token ID | 373 |
Feature activation | +7.30e+00 |
Pos logit contributions | |
yours | +0.319 |
beh | +0.297 |
possible | +0.282 |
happen | +0.279 |
anything | +0.278 |
Neg logit contributions | |
girl | -0.090 |
woman | -0.084 |
sky | -0.084 |
boy | -0.083 |
little | -0.082 |
Token | a |
Token ID | 257 |
Feature activation | -3.07e+00 |
Pos logit contributions | |
boy | +0.312 |
man | +0.302 |
girl | +0.296 |
little | +0.286 |
sky | +0.283 |
Neg logit contributions | |
yours | -0.416 |
beh | -0.408 |
possible | -0.356 |
theirs | -0.352 |
ours | -0.350 |
Token | boy |
Token ID | 2933 |
Feature activation | +1.72e+01 |
Pos logit contributions | |
yours | +0.199 |
beh | +0.176 |
possible | +0.176 |
ours | +0.174 |
asty | +0.170 |
Neg logit contributions | |
sky | -0.054 |
was | -0.039 |
led | -0.037 |
At | -0.033 |
gasped | -0.023 |
Token | named |
Token ID | 3706 |
Feature activation | +3.83e+00 |
Pos logit contributions | |
boy | +0.542 |
little | +0.517 |
girl | +0.503 |
man | +0.477 |
sky | +0.475 |
Neg logit contributions | |
beh | -1.236 |
yours | -1.181 |
possible | -1.062 |
ours | -1.029 |
theirs | -1.013 |
Token | Dave |
Token ID | 9935 |
Feature activation | -6.54e+00 |
Pos logit contributions | |
boy | +0.171 |
little | +0.168 |
girl | +0.157 |
man | +0.152 |
sky | +0.150 |
Neg logit contributions | |
beh | -0.240 |
yours | -0.218 |
possible | -0.200 |
unnecessary | -0.195 |
ieve | -0.192 |
Token | . |
Token ID | 13 |
Feature activation | -1.05e+01 |
Pos logit contributions | |
beh | +0.351 |
yours | +0.334 |
possible | +0.287 |
ours | +0.273 |
theirs | +0.272 |
Neg logit contributions | |
boy | -0.319 |
little | -0.310 |
girl | -0.307 |
sky | -0.300 |
man | -0.296 |
Token | He |
Token ID | 679 |
Feature activation | -2.17e+00 |
Pos logit contributions | |
beh | +0.656 |
yours | +0.655 |
possible | +0.577 |
theirs | +0.564 |
ours | +0.563 |
Neg logit contributions | |
boy | -0.387 |
little | -0.384 |
girl | -0.378 |
man | -0.354 |
sky | -0.341 |
Token | had |
Token ID | 550 |
Feature activation | +4.56e+00 |
Pos logit contributions | |
beh | +0.124 |
yours | +0.118 |
supposed | +0.113 |
possible | +0.100 |
belong | +0.100 |
Neg logit contributions | |
girl | -0.071 |
sky | -0.068 |
boy | -0.067 |
child | -0.062 |
little | -0.061 |
Token |
|
Token ID | 198 |
Feature activation | +6.75e+00 |
Pos logit contributions | |
beh | +0.599 |
yours | +0.585 |
possible | +0.511 |
ours | +0.502 |
theirs | +0.496 |
Neg logit contributions | |
boy | -0.305 |
girl | -0.282 |
little | -0.281 |
man | -0.277 |
sky | -0.252 |
Token |
|
Token ID | 198 |
Feature activation | +6.67e+00 |
Pos logit contributions | |
boy | +0.175 |
girl | +0.163 |
little | +0.146 |
sky | +0.141 |
man | +0.140 |
Neg logit contributions | |
beh | -0.499 |
yours | -0.487 |
possible | -0.448 |
ourselves | -0.431 |
irs | -0.430 |
Token | M |
Token ID | 44 |
Feature activation | +1.42e+01 |
Pos logit contributions | |
boy | +0.228 |
girl | +0.205 |
little | +0.203 |
sky | +0.199 |
man | +0.196 |
Neg logit contributions | |
beh | -0.468 |
yours | -0.433 |
possible | -0.392 |
ourselves | -0.382 |
myself | -0.376 |
Token | olly |
Token ID | 5098 |
Feature activation | +2.21e+00 |
Pos logit contributions | |
boy | +0.556 |
little | +0.527 |
girl | +0.519 |
man | +0.497 |
sky | +0.497 |
Neg logit contributions | |
beh | -0.920 |
yours | -0.868 |
possible | -0.777 |
ours | -0.743 |
unnecessary | -0.738 |
Token | stirred |
Token ID | 33091 |
Feature activation | +1.01e+00 |
Pos logit contributions | |
boy | +0.061 |
sky | +0.058 |
girl | +0.057 |
man | +0.051 |
little | +0.051 |
Neg logit contributions | |
beh | -0.143 |
yours | -0.127 |
eking | -0.118 |
rying | -0.116 |
properly | -0.115 |
Token | and |
Token ID | 290 |
Feature activation | +1.33e+01 |
Pos logit contributions | |
boy | +0.036 |
little | +0.033 |
girl | +0.033 |
man | +0.032 |
sky | +0.031 |
Neg logit contributions | |
beh | -0.065 |
yours | -0.065 |
possible | -0.058 |
ourselves | -0.056 |
ours | -0.055 |
Token | stirred |
Token ID | 33091 |
Feature activation | +9.21e-01 |
Pos logit contributions | |
boy | +0.604 |
little | +0.573 |
girl | +0.570 |
man | +0.554 |
sky | +0.546 |
Neg logit contributions | |
beh | -0.763 |
yours | -0.726 |
possible | -0.643 |
ours | -0.604 |
ieve | -0.602 |
Token | until |
Token ID | 1566 |
Feature activation | -4.59e+00 |
Pos logit contributions | |
boy | +0.032 |
little | +0.031 |
girl | +0.030 |
man | +0.029 |
sky | +0.029 |
Neg logit contributions | |
beh | -0.063 |
yours | -0.060 |
possible | -0.054 |
ours | -0.052 |
unnecessary | -0.051 |
Token | all |
Token ID | 477 |
Feature activation | -6.93e+00 |
Pos logit contributions | |
beh | +0.296 |
yours | +0.282 |
possible | +0.253 |
ours | +0.242 |
unnecessary | +0.239 |
Neg logit contributions | |
boy | -0.180 |
little | -0.173 |
girl | -0.169 |
man | -0.162 |
sky | -0.160 |
Token | the |
Token ID | 262 |
Feature activation | +5.57e+00 |
Pos logit contributions | |
beh | +0.459 |
yours | +0.441 |
possible | +0.399 |
ours | +0.378 |
theirs | +0.375 |
Neg logit contributions | |
boy | -0.243 |
little | -0.229 |
girl | -0.223 |
man | -0.220 |
sky | -0.212 |
Token | ingredients |
Token ID | 9391 |
Feature activation | +2.09e+00 |
Pos logit contributions | |
boy | +0.248 |
girl | +0.232 |
little | +0.230 |
man | +0.224 |
sky | +0.222 |
Neg logit contributions | |
beh | -0.305 |
yours | -0.292 |
possible | -0.263 |
ourselves | -0.246 |
eking | -0.245 |
Token | a |
Token ID | 257 |
Feature activation | -5.05e-01 |
Pos logit contributions | |
boy | +0.419 |
girl | +0.396 |
man | +0.393 |
little | +0.382 |
sky | +0.367 |
Neg logit contributions | |
beh | -0.604 |
yours | -0.599 |
possible | -0.522 |
ours | -0.508 |
theirs | -0.499 |
Token | girl |
Token ID | 2576 |
Feature activation | +1.84e+01 |
Pos logit contributions | |
yours | +0.032 |
ours | +0.029 |
possible | +0.029 |
beh | +0.028 |
asty | +0.027 |
Neg logit contributions | |
sky | -0.010 |
was | -0.007 |
led | -0.006 |
At | -0.005 |
gasped | -0.004 |
Token | . |
Token ID | 13 |
Feature activation | -8.24e+00 |
Pos logit contributions | |
boy | +0.721 |
girl | +0.667 |
man | +0.661 |
sky | +0.661 |
little | +0.658 |
Neg logit contributions | |
beh | -1.112 |
yours | -1.074 |
possible | -0.912 |
eking | -0.878 |
ieve | -0.876 |
Token | She |
Token ID | 1375 |
Feature activation | +1.83e+00 |
Pos logit contributions | |
beh | +0.507 |
yours | +0.495 |
possible | +0.433 |
theirs | +0.422 |
ours | +0.413 |
Neg logit contributions | |
boy | -0.318 |
man | -0.302 |
little | -0.299 |
girl | -0.293 |
sky | -0.268 |
Token | was |
Token ID | 373 |
Feature activation | +9.39e+00 |
Pos logit contributions | |
boy | +0.053 |
sky | +0.051 |
girl | +0.050 |
man | +0.050 |
child | +0.049 |
Neg logit contributions | |
beh | -0.108 |
yours | -0.107 |
supposed | -0.096 |
belong | -0.092 |
ieve | -0.092 |
Token | 3 |
Token ID | 513 |
Feature activation | +1.38e+01 |
Pos logit contributions | |
boy | +0.374 |
man | +0.360 |
girl | +0.351 |
sky | +0.344 |
little | +0.337 |
Neg logit contributions | |
beh | -0.567 |
yours | -0.543 |
possible | -0.473 |
ours | -0.464 |
supposed | -0.461 |
Token | years |
Token ID | 812 |
Feature activation | +1.43e+00 |
Pos logit contributions | |
sky | +0.751 |
boy | +0.737 |
girl | +0.703 |
man | +0.700 |
little | +0.663 |
Neg logit contributions | |
beh | -0.634 |
yours | -0.530 |
eking | -0.453 |
ours | -0.436 |
supposed | -0.431 |
Token | old |
Token ID | 1468 |
Feature activation | +6.75e+00 |
Pos logit contributions | |
boy | +0.042 |
girl | +0.038 |
was | +0.037 |
little | +0.036 |
sky | +0.035 |
Neg logit contributions | |
yours | -0.099 |
beh | -0.099 |
possible | -0.089 |
ours | -0.087 |
theirs | -0.085 |
Token | and |
Token ID | 290 |
Feature activation | +1.40e+01 |
Pos logit contributions | |
boy | +0.252 |
little | +0.250 |
sky | +0.247 |
girl | +0.238 |
man | +0.234 |
Neg logit contributions | |
beh | -0.371 |
yours | -0.355 |
ourselves | -0.306 |
possible | -0.302 |
theirs | -0.297 |
Token | she |
Token ID | 673 |
Feature activation | -3.23e+00 |
Pos logit contributions | |
boy | +0.389 |
little | +0.366 |
girl | +0.356 |
sky | +0.334 |
man | +0.333 |
Neg logit contributions | |
beh | -1.052 |
yours | -1.012 |
possible | -0.919 |
ours | -0.888 |
theirs | -0.876 |
Token | was |
Token ID | 373 |
Feature activation | +8.83e+00 |
Pos logit contributions | |
beh | +0.224 |
yours | +0.203 |
supposed | +0.192 |
ieve | +0.181 |
possible | +0.179 |
Neg logit contributions | |
boy | -0.099 |
girl | -0.095 |
little | -0.093 |
sky | -0.092 |
man | -0.090 |
Token | with |
Token ID | 351 |
Feature activation | +2.97e-01 |
Pos logit contributions | |
boy | +0.583 |
little | +0.574 |
girl | +0.558 |
sky | +0.524 |
man | +0.505 |
Neg logit contributions | |
beh | -1.026 |
yours | -0.980 |
ieve | -0.881 |
possible | -0.870 |
eking | -0.822 |
Token | the |
Token ID | 262 |
Feature activation | +5.86e+00 |
Pos logit contributions | |
boy | +0.012 |
girl | +0.011 |
little | +0.011 |
man | +0.011 |
sky | +0.010 |
Neg logit contributions | |
yours | -0.018 |
beh | -0.017 |
theirs | -0.015 |
possible | -0.015 |
ours | -0.015 |
Token | guitar |
Token ID | 10047 |
Feature activation | +1.83e+00 |
Pos logit contributions | |
boy | +0.130 |
was | +0.124 |
sky | +0.123 |
little | +0.121 |
girl | +0.108 |
Neg logit contributions | |
yours | -0.390 |
beh | -0.387 |
possible | -0.363 |
ieve | -0.358 |
unnecessary | -0.343 |
Token | sees |
Token ID | 7224 |
Feature activation | +4.70e+00 |
Pos logit contributions | |
boy | +0.055 |
little | +0.052 |
girl | +0.051 |
man | +0.046 |
sky | +0.046 |
Neg logit contributions | |
beh | -0.130 |
yours | -0.126 |
possible | -0.115 |
ieve | -0.110 |
ours | -0.109 |
Token | Ben |
Token ID | 3932 |
Feature activation | +4.47e-01 |
Pos logit contributions | |
boy | +0.201 |
little | +0.193 |
girl | +0.191 |
man | +0.183 |
sky | +0.182 |
Neg logit contributions | |
beh | -0.280 |
yours | -0.269 |
possible | -0.237 |
ours | -0.227 |
theirs | -0.225 |
Token | and |
Token ID | 290 |
Feature activation | +1.36e+01 |
Pos logit contributions | |
boy | +0.007 |
little | +0.006 |
girl | +0.006 |
sky | +0.005 |
man | +0.005 |
Neg logit contributions | |
beh | -0.039 |
yours | -0.038 |
possible | -0.035 |
ours | -0.034 |
ieve | -0.034 |
Token | Mia |
Token ID | 32189 |
Feature activation | -3.32e+00 |
Pos logit contributions | |
boy | +0.590 |
little | +0.572 |
girl | +0.560 |
sky | +0.542 |
man | +0.539 |
Neg logit contributions | |
beh | -0.804 |
yours | -0.765 |
possible | -0.673 |
ours | -0.646 |
theirs | -0.636 |
Token | and |
Token ID | 290 |
Feature activation | +1.36e+01 |
Pos logit contributions | |
beh | +0.237 |
yours | +0.225 |
possible | +0.203 |
ours | +0.197 |
unnecessary | +0.196 |
Neg logit contributions | |
boy | -0.112 |
little | -0.109 |
girl | -0.104 |
sky | -0.099 |
man | -0.099 |
Token | smiles |
Token ID | 21845 |
Feature activation | +5.08e+00 |
Pos logit contributions | |
boy | +0.539 |
little | +0.509 |
girl | +0.503 |
man | +0.480 |
was | +0.459 |
Neg logit contributions | |
beh | -0.896 |
yours | -0.846 |
possible | -0.759 |
ours | -0.738 |
unnecessary | -0.737 |
Token | . |
Token ID | 13 |
Feature activation | -9.15e+00 |
Pos logit contributions | |
boy | +0.224 |
little | +0.219 |
girl | +0.211 |
man | +0.203 |
sky | +0.201 |
Neg logit contributions | |
beh | -0.315 |
yours | -0.298 |
ours | -0.260 |
possible | -0.259 |
unnecessary | -0.256 |
Token | He |
Token ID | 679 |
Feature activation | -8.13e-01 |
Pos logit contributions | |
beh | +0.540 |
yours | +0.538 |
theirs | +0.478 |
possible | +0.463 |
ours | +0.457 |
Neg logit contributions | |
boy | -0.333 |
girl | -0.329 |
little | -0.325 |
man | -0.302 |
sky | -0.295 |
Token | next |
Token ID | 1306 |
Feature activation | -2.89e+00 |
Pos logit contributions | |
beh | +0.006 |
yours | +0.006 |
possible | +0.005 |
ours | +0.005 |
theirs | +0.005 |
Neg logit contributions | |
boy | -0.005 |
little | -0.005 |
girl | -0.004 |
sky | -0.004 |
man | -0.004 |
Token | . |
Token ID | 13 |
Feature activation | -1.04e+01 |
Pos logit contributions | |
beh | +0.158 |
yours | +0.156 |
possible | +0.137 |
ours | +0.129 |
unnecessary | +0.125 |
Neg logit contributions | |
boy | -0.130 |
girl | -0.125 |
little | -0.125 |
sky | -0.118 |
man | -0.117 |
Token | She |
Token ID | 1375 |
Feature activation | -2.04e-01 |
Pos logit contributions | |
yours | +0.614 |
beh | +0.606 |
ours | +0.529 |
theirs | +0.524 |
possible | +0.519 |
Neg logit contributions | |
boy | -0.398 |
girl | -0.380 |
little | -0.378 |
man | -0.348 |
was | -0.344 |
Token | swung |
Token ID | 28507 |
Feature activation | +4.19e-01 |
Pos logit contributions | |
beh | +0.011 |
yours | +0.010 |
ck | +0.009 |
ieve | +0.008 |
properly | +0.008 |
Neg logit contributions | |
sky | -0.008 |
girl | -0.008 |
boy | -0.007 |
little | -0.007 |
man | -0.007 |
Token | the |
Token ID | 262 |
Feature activation | +4.83e+00 |
Pos logit contributions | |
boy | +0.017 |
little | +0.016 |
girl | +0.016 |
man | +0.015 |
sky | +0.015 |
Neg logit contributions | |
beh | -0.026 |
yours | -0.025 |
possible | -0.022 |
ours | -0.021 |
theirs | -0.021 |
Token | stick |
Token ID | 4859 |
Feature activation | +4.49e+00 |
Pos logit contributions | |
boy | +0.106 |
little | +0.101 |
girl | +0.100 |
was | +0.094 |
man | +0.090 |
Neg logit contributions | |
yours | -0.363 |
beh | -0.346 |
ours | -0.327 |
possible | -0.322 |
unnecessary | -0.310 |
Token | softly |
Token ID | 26625 |
Feature activation | -5.32e+00 |
Pos logit contributions | |
boy | +0.164 |
little | +0.158 |
girl | +0.154 |
man | +0.147 |
sky | +0.146 |
Neg logit contributions | |
beh | -0.301 |
yours | -0.285 |
possible | -0.256 |
ours | -0.248 |
unnecessary | -0.245 |
Token | and |
Token ID | 290 |
Feature activation | +1.24e+01 |
Pos logit contributions | |
beh | +0.428 |
yours | +0.412 |
possible | +0.377 |
ours | +0.366 |
unnecessary | +0.362 |
Neg logit contributions | |
boy | -0.122 |
little | -0.115 |
girl | -0.110 |
man | -0.101 |
sky | -0.100 |
Token | hit |
Token ID | 2277 |
Feature activation | -2.58e-01 |
Pos logit contributions | |
boy | +0.421 |
little | +0.400 |
girl | +0.390 |
man | +0.373 |
sky | +0.367 |
Neg logit contributions | |
beh | -0.871 |
yours | -0.828 |
possible | -0.746 |
ours | -0.723 |
unnecessary | -0.717 |
Token | the |
Token ID | 262 |
Feature activation | +4.75e+00 |
Pos logit contributions | |
beh | +0.018 |
yours | +0.017 |
possible | +0.015 |
ours | +0.015 |
unnecessary | +0.015 |
Neg logit contributions | |
boy | -0.009 |
little | -0.009 |
girl | -0.008 |
man | -0.008 |
sky | -0.008 |
Token | ball |
Token ID | 2613 |
Feature activation | +8.41e+00 |
Pos logit contributions | |
boy | +0.106 |
little | +0.102 |
girl | +0.101 |
was | +0.090 |
man | +0.089 |
Neg logit contributions | |
yours | -0.349 |
beh | -0.339 |
ours | -0.315 |
possible | -0.314 |
ourselves | -0.305 |
Token | could |
Token ID | 714 |
Feature activation | +1.21e+00 |
Pos logit contributions | |
beh | +0.061 |
yours | +0.058 |
possible | +0.053 |
ieve | +0.051 |
theirs | +0.050 |
Neg logit contributions | |
boy | -0.023 |
girl | -0.022 |
little | -0.021 |
sky | -0.020 |
man | -0.019 |
Token | dress |
Token ID | 6576 |
Feature activation | +3.93e+00 |
Pos logit contributions | |
little | +0.053 |
boy | +0.052 |
man | +0.049 |
girl | +0.049 |
sky | +0.048 |
Neg logit contributions | |
beh | -0.074 |
yours | -0.068 |
unnecessary | -0.060 |
possible | -0.060 |
ours | -0.059 |
Token | very |
Token ID | 845 |
Feature activation | +5.09e+00 |
Pos logit contributions | |
little | +0.188 |
boy | +0.185 |
man | +0.177 |
girl | +0.173 |
was | +0.171 |
Neg logit contributions | |
beh | -0.231 |
yours | -0.201 |
possible | -0.183 |
unnecessary | -0.178 |
supervision | -0.177 |
Token | quickly |
Token ID | 2952 |
Feature activation | -5.43e+00 |
Pos logit contributions | |
little | +0.213 |
boy | +0.204 |
man | +0.190 |
girl | +0.188 |
was | +0.187 |
Neg logit contributions | |
beh | -0.327 |
yours | -0.307 |
unnecessary | -0.267 |
possible | -0.265 |
supervision | -0.265 |
Token | ! |
Token ID | 0 |
Feature activation | -1.14e+01 |
Pos logit contributions | |
beh | +0.401 |
yours | +0.367 |
possible | +0.336 |
unnecessary | +0.331 |
ours | +0.328 |
Neg logit contributions | |
little | -0.173 |
boy | -0.167 |
girl | -0.157 |
sky | -0.154 |
man | -0.153 |
Token |
|
Token ID | 198 |
Feature activation | +6.47e+00 |
Pos logit contributions | |
beh | +0.877 |
yours | +0.848 |
possible | +0.772 |
ours | +0.746 |
theirs | +0.738 |
Neg logit contributions | |
boy | -0.282 |
little | -0.268 |
girl | -0.262 |
man | -0.242 |
sky | -0.231 |
Token |
|
Token ID | 198 |
Feature activation | +6.41e+00 |
Pos logit contributions | |
boy | +0.147 |
girl | +0.143 |
little | +0.133 |
man | +0.124 |
sky | +0.118 |
Neg logit contributions | |
beh | -0.497 |
yours | -0.479 |
possible | -0.453 |
ieve | -0.420 |
theirs | -0.419 |
Token | He |
Token ID | 1544 |
Feature activation | +1.19e+00 |
Pos logit contributions | |
boy | +0.280 |
girl | +0.270 |
little | +0.267 |
man | +0.254 |
sky | +0.251 |
Neg logit contributions | |
beh | -0.381 |
yours | -0.353 |
possible | -0.324 |
ieve | -0.298 |
unnecessary | -0.295 |
Token | was |
Token ID | 373 |
Feature activation | +8.21e+00 |
Pos logit contributions | |
boy | +0.038 |
girl | +0.035 |
little | +0.034 |
sky | +0.034 |
man | +0.033 |
Neg logit contributions | |
beh | -0.082 |
yours | -0.078 |
possible | -0.069 |
ieve | -0.067 |
eking | -0.066 |
Token | so |
Token ID | 523 |
Feature activation | -6.77e-01 |
Pos logit contributions | |
boy | +0.333 |
little | +0.322 |
girl | +0.314 |
man | +0.301 |
sky | +0.299 |
Neg logit contributions | |
beh | -0.518 |
yours | -0.491 |
possible | -0.438 |
ours | -0.421 |
unnecessary | -0.416 |
Token | fast |
Token ID | 3049 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
beh | +0.049 |
yours | +0.047 |
possible | +0.043 |
ours | +0.041 |
unnecessary | +0.040 |
Neg logit contributions | |
boy | -0.021 |
little | -0.020 |
girl | -0.020 |
man | -0.018 |
sky | -0.018 |
Token | . |
Token ID | 13 |
Feature activation | -9.87e+00 |
Pos logit contributions | |
boy | +0.035 |
little | +0.033 |
girl | +0.032 |
man | +0.030 |
sky | +0.030 |
Neg logit contributions | |
beh | -0.082 |
yours | -0.078 |
possible | -0.071 |
ours | -0.069 |
unnecessary | -0.068 |
Token | She |
Token ID | 1375 |
Feature activation | +3.35e-01 |
Pos logit contributions | |
beh | +0.582 |
yours | +0.568 |
possible | +0.498 |
ours | +0.479 |
ourselves | +0.477 |
Neg logit contributions | |
boy | -0.372 |
little | -0.359 |
girl | -0.355 |
man | -0.334 |
sky | -0.324 |
Token | said |
Token ID | 531 |
Feature activation | +1.42e+00 |
Pos logit contributions | |
boy | +0.012 |
girl | +0.011 |
sky | +0.011 |
little | +0.010 |
man | +0.010 |
Neg logit contributions | |
beh | -0.022 |
yours | -0.020 |
ours | -0.017 |
eking | -0.017 |
ieve | -0.017 |
Token | : |
Token ID | 25 |
Feature activation | +4.89e+00 |
Pos logit contributions | |
boy | +0.051 |
girl | +0.048 |
little | +0.047 |
sky | +0.045 |
man | +0.044 |
Neg logit contributions | |
beh | -0.095 |
yours | -0.092 |
possible | -0.082 |
ours | -0.078 |
theirs | -0.078 |
Token | " |
Token ID | 366 |
Feature activation | +6.34e+00 |
Pos logit contributions | |
boy | +0.134 |
girl | +0.127 |
little | +0.113 |
man | +0.105 |
sky | +0.103 |
Neg logit contributions | |
yours | -0.350 |
beh | -0.344 |
possible | -0.314 |
theirs | -0.301 |
ourselves | -0.295 |
Token | Can |
Token ID | 6090 |
Feature activation | +2.80e+00 |
Pos logit contributions | |
boy | +0.233 |
little | +0.227 |
girl | +0.218 |
man | +0.209 |
sky | +0.206 |
Neg logit contributions | |
beh | -0.426 |
yours | -0.403 |
possible | -0.364 |
ours | -0.347 |
ourselves | -0.341 |
Token | I |
Token ID | 314 |
Feature activation | -1.20e+01 |
Pos logit contributions | |
boy | +0.150 |
little | +0.147 |
girl | +0.142 |
man | +0.139 |
sky | +0.137 |
Neg logit contributions | |
beh | -0.144 |
yours | -0.132 |
possible | -0.114 |
unnecessary | -0.109 |
ours | -0.109 |
Token | have |
Token ID | 423 |
Feature activation | -6.50e+00 |
Pos logit contributions | |
beh | +0.718 |
yours | +0.660 |
possible | +0.596 |
unnecessary | +0.583 |
ours | +0.571 |
Neg logit contributions | |
boy | -0.539 |
little | -0.537 |
girl | -0.511 |
man | -0.496 |
was | -0.486 |
Token | a |
Token ID | 257 |
Feature activation | -2.24e+00 |
Pos logit contributions | |
beh | +0.453 |
rying | +0.388 |
yours | +0.372 |
owing | +0.364 |
eking | +0.361 |
Neg logit contributions | |
little | -0.255 |
boy | -0.231 |
was | -0.217 |
girl | -0.215 |
sky | -0.209 |
Token | cap |
Token ID | 1451 |
Feature activation | +5.45e+00 |
Pos logit contributions | |
yours | +0.145 |
beh | +0.143 |
possible | +0.128 |
ours | +0.123 |
unnecessary | +0.122 |
Neg logit contributions | |
boy | -0.077 |
girl | -0.072 |
sky | -0.071 |
little | -0.070 |
man | -0.069 |
Token | too |
Token ID | 1165 |
Feature activation | -9.04e+00 |
Pos logit contributions | |
boy | +0.258 |
little | +0.249 |
girl | +0.244 |
man | +0.241 |
and | +0.230 |
Neg logit contributions | |
beh | -0.323 |
yours | -0.275 |
unnecessary | -0.270 |
possible | -0.254 |
eking | -0.253 |
Token | It |
Token ID | 632 |
Feature activation | -2.57e-01 |
Pos logit contributions | |
yours | +0.633 |
beh | +0.632 |
ours | +0.544 |
possible | +0.541 |
theirs | +0.535 |
Neg logit contributions | |
boy | -0.338 |
girl | -0.327 |
little | -0.318 |
man | -0.306 |
sky | -0.281 |
Token | was |
Token ID | 373 |
Feature activation | +8.12e+00 |
Pos logit contributions | |
beh | +0.019 |
yours | +0.019 |
possible | +0.017 |
ieve | +0.016 |
happen | +0.016 |
Neg logit contributions | |
boy | -0.005 |
girl | -0.005 |
little | -0.004 |
man | -0.004 |
child | -0.004 |
Token | very |
Token ID | 845 |
Feature activation | +4.72e+00 |
Pos logit contributions | |
boy | +0.277 |
girl | +0.262 |
little | +0.247 |
man | +0.245 |
sky | +0.233 |
Neg logit contributions | |
yours | -0.545 |
beh | -0.545 |
possible | -0.485 |
ours | -0.474 |
theirs | -0.467 |
Token | small |
Token ID | 1402 |
Feature activation | +7.93e+00 |
Pos logit contributions | |
boy | +0.153 |
little | +0.143 |
girl | +0.143 |
sky | +0.134 |
man | +0.134 |
Neg logit contributions | |
beh | -0.334 |
yours | -0.322 |
possible | -0.290 |
ours | -0.283 |
theirs | -0.276 |
Token | , |
Token ID | 11 |
Feature activation | +1.02e+01 |
Pos logit contributions | |
boy | +0.205 |
little | +0.192 |
girl | +0.188 |
man | +0.178 |
sky | +0.173 |
Neg logit contributions | |
beh | -0.603 |
yours | -0.583 |
possible | -0.529 |
ours | -0.514 |
unnecessary | -0.506 |
Token | and |
Token ID | 290 |
Feature activation | +1.32e+01 |
Pos logit contributions | |
boy | +0.350 |
girl | +0.338 |
little | +0.324 |
man | +0.311 |
sky | +0.297 |
Neg logit contributions | |
yours | -0.661 |
beh | -0.633 |
ours | -0.577 |
possible | -0.575 |
theirs | -0.554 |
Token | it |
Token ID | 340 |
Feature activation | -3.94e+00 |
Pos logit contributions | |
boy | +0.556 |
little | +0.529 |
girl | +0.529 |
man | +0.504 |
sky | +0.502 |
Neg logit contributions | |
yours | -0.759 |
beh | -0.751 |
possible | -0.662 |
ours | -0.645 |
ourselves | -0.614 |
Token | belonged |
Token ID | 19611 |
Feature activation | -3.04e+00 |
Pos logit contributions | |
beh | +0.292 |
yours | +0.272 |
possible | +0.250 |
ieve | +0.240 |
properly | +0.239 |
Neg logit contributions | |
boy | -0.100 |
girl | -0.096 |
little | -0.095 |
man | -0.086 |
sky | -0.084 |
Token | to |
Token ID | 284 |
Feature activation | +6.73e-01 |
Pos logit contributions | |
beh | +0.210 |
yours | +0.195 |
ieve | +0.182 |
unnecessary | +0.180 |
possible | +0.175 |
Neg logit contributions | |
boy | -0.107 |
little | -0.106 |
girl | -0.099 |
was | -0.096 |
sky | -0.096 |
Token | a |
Token ID | 257 |
Feature activation | -2.10e+00 |
Pos logit contributions | |
little | +0.022 |
boy | +0.022 |
girl | +0.020 |
was | +0.019 |
sky | +0.019 |
Neg logit contributions | |
beh | -0.050 |
yours | -0.045 |
possible | -0.042 |
rying | -0.041 |
eking | -0.041 |
Token | little |
Token ID | 1310 |
Feature activation | +1.19e+01 |
Pos logit contributions | |
yours | +0.145 |
beh | +0.135 |
possible | +0.134 |
unnecessary | +0.126 |
ours | +0.126 |
Neg logit contributions | |
sky | -0.041 |
boy | -0.037 |
was | -0.034 |
girl | -0.033 |
man | -0.032 |
Token | measure |
Token ID | 3953 |
Feature activation | -4.92e+00 |
Pos logit contributions | |
boy | +0.208 |
little | +0.198 |
girl | +0.196 |
man | +0.188 |
sky | +0.187 |
Neg logit contributions | |
beh | -0.300 |
yours | -0.294 |
possible | -0.260 |
ours | -0.251 |
theirs | -0.242 |
Token | cup |
Token ID | 6508 |
Feature activation | +1.13e+01 |
Pos logit contributions | |
beh | +0.278 |
yours | +0.256 |
possible | +0.225 |
unnecessary | +0.220 |
ours | +0.216 |
Neg logit contributions | |
boy | -0.238 |
little | -0.230 |
girl | -0.226 |
man | -0.219 |
sky | -0.216 |
Token | ?" |
Token ID | 1701 |
Feature activation | -1.07e+01 |
Pos logit contributions | |
boy | +0.533 |
little | +0.516 |
girl | +0.505 |
man | +0.493 |
sky | +0.491 |
Neg logit contributions | |
beh | -0.652 |
yours | -0.599 |
possible | -0.535 |
unnecessary | -0.530 |
ieve | -0.509 |
Token | Mia |
Token ID | 32189 |
Feature activation | -4.23e+00 |
Pos logit contributions | |
beh | +0.448 |
yours | +0.424 |
myself | +0.348 |
ourselves | +0.347 |
ieve | +0.338 |
Neg logit contributions | |
boy | -0.549 |
girl | -0.528 |
little | -0.526 |
man | -0.525 |
sky | -0.507 |
Token | was |
Token ID | 373 |
Feature activation | +7.73e+00 |
Pos logit contributions | |
beh | +0.255 |
yours | +0.230 |
ieve | +0.209 |
myself | +0.204 |
properly | +0.204 |
Neg logit contributions | |
sky | -0.141 |
boy | -0.132 |
girl | -0.126 |
little | -0.119 |
man | -0.116 |
Token | happy |
Token ID | 3772 |
Feature activation | +3.81e+00 |
Pos logit contributions | |
boy | +0.290 |
girl | +0.274 |
little | +0.267 |
man | +0.262 |
sky | +0.260 |
Neg logit contributions | |
beh | -0.502 |
yours | -0.481 |
possible | -0.427 |
ours | -0.419 |
theirs | -0.411 |
Token | to |
Token ID | 284 |
Feature activation | +1.38e-01 |
Pos logit contributions | |
boy | +0.119 |
little | +0.113 |
girl | +0.109 |
man | +0.104 |
sky | +0.100 |
Neg logit contributions | |
beh | -0.288 |
yours | -0.271 |
possible | -0.246 |
unnecessary | -0.241 |
ours | -0.241 |
Token | help |
Token ID | 1037 |
Feature activation | -6.48e+00 |
Pos logit contributions | |
little | +0.007 |
boy | +0.007 |
girl | +0.006 |
man | +0.006 |
sky | +0.006 |
Neg logit contributions | |
beh | -0.008 |
yours | -0.007 |
possible | -0.006 |
unnecessary | -0.006 |
ours | -0.006 |
Token | her |
Token ID | 607 |
Feature activation | -5.65e+00 |
Pos logit contributions | |
beh | +0.465 |
yours | +0.439 |
possible | +0.396 |
ours | +0.390 |
unnecessary | +0.386 |
Neg logit contributions | |
boy | -0.222 |
little | -0.221 |
girl | -0.208 |
man | -0.201 |
sky | -0.192 |
Token | mom |
Token ID | 1995 |
Feature activation | +9.83e+00 |
Pos logit contributions | |
beh | +0.400 |
yours | +0.390 |
possible | +0.352 |
theirs | +0.350 |
ours | +0.344 |
Neg logit contributions | |
sky | -0.151 |
boy | -0.139 |
girl | -0.127 |
man | -0.120 |
little | -0.118 |
Token | , |
Token ID | 11 |
Feature activation | +9.85e+00 |
Pos logit contributions | |
boy | +0.373 |
little | +0.370 |
girl | +0.350 |
man | +0.334 |
was | +0.329 |
Neg logit contributions | |
beh | -0.665 |
yours | -0.634 |
possible | -0.569 |
unnecessary | -0.568 |
ours | -0.555 |
Token | aff |
Token ID | 2001 |
Feature activation | +6.34e+00 |
Pos logit contributions | |
boy | +0.661 |
little | +0.644 |
girl | +0.624 |
man | +0.609 |
sky | +0.603 |
Neg logit contributions | |
beh | -0.888 |
yours | -0.841 |
possible | -0.749 |
ours | -0.712 |
unnecessary | -0.702 |
Token | es |
Token ID | 274 |
Feature activation | -5.14e+00 |
Pos logit contributions | |
boy | +0.300 |
little | +0.297 |
girl | +0.284 |
man | +0.280 |
sky | +0.276 |
Neg logit contributions | |
beh | -0.350 |
yours | -0.329 |
possible | -0.291 |
unnecessary | -0.281 |
ours | -0.278 |
Token | . |
Token ID | 13 |
Feature activation | -9.01e+00 |
Pos logit contributions | |
beh | +0.323 |
yours | +0.303 |
possible | +0.269 |
unnecessary | +0.262 |
ours | +0.262 |
Neg logit contributions | |
boy | -0.215 |
little | -0.212 |
girl | -0.202 |
man | -0.202 |
sky | -0.198 |
Token |
|
Token ID | 198 |
Feature activation | +6.88e+00 |
Pos logit contributions | |
beh | +0.594 |
yours | +0.590 |
possible | +0.524 |
theirs | +0.518 |
ours | +0.517 |
Neg logit contributions | |
boy | -0.295 |
little | -0.285 |
girl | -0.282 |
man | -0.267 |
sky | -0.254 |
Token |
|
Token ID | 198 |
Feature activation | +6.80e+00 |
Pos logit contributions | |
boy | +0.181 |
girl | +0.178 |
little | +0.162 |
sky | +0.154 |
man | +0.147 |
Neg logit contributions | |
yours | -0.490 |
beh | -0.477 |
possible | -0.446 |
theirs | -0.437 |
ours | -0.433 |
Token | They |
Token ID | 2990 |
Feature activation | -3.24e+00 |
Pos logit contributions | |
boy | +0.231 |
little | +0.218 |
girl | +0.215 |
sky | +0.208 |
man | +0.200 |
Neg logit contributions | |
beh | -0.462 |
yours | -0.446 |
possible | -0.404 |
ours | -0.393 |
ourselves | -0.390 |
Token | saw |
Token ID | 2497 |
Feature activation | +7.21e+00 |
Pos logit contributions | |
beh | +0.219 |
yours | +0.209 |
possible | +0.188 |
ieve | +0.185 |
theirs | +0.184 |
Neg logit contributions | |
boy | -0.101 |
girl | -0.096 |
little | -0.095 |
sky | -0.090 |
man | -0.084 |
Token | the |
Token ID | 262 |
Feature activation | +5.92e+00 |
Pos logit contributions | |
boy | +0.279 |
girl | +0.268 |
little | +0.263 |
sky | +0.255 |
man | +0.253 |
Neg logit contributions | |
beh | -0.434 |
yours | -0.425 |
possible | -0.376 |
theirs | -0.375 |
ours | -0.369 |
Token | lions |
Token ID | 31420 |
Feature activation | +6.06e+00 |
Pos logit contributions | |
boy | +0.199 |
girl | +0.192 |
little | +0.192 |
was | +0.179 |
sky | +0.179 |
Neg logit contributions | |
yours | -0.365 |
beh | -0.364 |
possible | -0.327 |
ours | -0.326 |
theirs | -0.312 |
Token | roar |
Token ID | 34532 |
Feature activation | -7.57e-01 |
Pos logit contributions | |
boy | +0.123 |
girl | +0.115 |
little | +0.112 |
sky | +0.092 |
man | +0.088 |
Neg logit contributions | |
beh | -0.479 |
yours | -0.468 |
possible | -0.430 |
ieve | -0.419 |
eking | -0.418 |
Token | and |
Token ID | 290 |
Feature activation | +1.33e+01 |
Pos logit contributions | |
beh | +0.065 |
yours | +0.063 |
possible | +0.058 |
ours | +0.057 |
theirs | +0.056 |
Neg logit contributions | |
boy | -0.012 |
little | -0.011 |
girl | -0.011 |
sky | -0.009 |
man | -0.009 |
Token | , |
Token ID | 11 |
Feature activation | +1.41e+01 |
Pos logit contributions | |
boy | +0.102 |
little | +0.101 |
girl | +0.096 |
man | +0.090 |
sky | +0.086 |
Neg logit contributions | |
beh | -0.185 |
yours | -0.171 |
possible | -0.156 |
ieve | -0.152 |
ours | -0.151 |
Token | so |
Token ID | 523 |
Feature activation | +2.84e+00 |
Pos logit contributions | |
boy | +0.515 |
little | +0.496 |
girl | +0.481 |
man | +0.459 |
sky | +0.453 |
Neg logit contributions | |
beh | -0.953 |
yours | -0.903 |
possible | -0.812 |
ours | -0.782 |
unnecessary | -0.776 |
Token | she |
Token ID | 673 |
Feature activation | -8.98e-01 |
Pos logit contributions | |
boy | +0.104 |
little | +0.102 |
girl | +0.097 |
man | +0.093 |
was | +0.090 |
Neg logit contributions | |
beh | -0.197 |
yours | -0.183 |
possible | -0.162 |
ieve | -0.161 |
ours | -0.161 |
Token | ran |
Token ID | 4966 |
Feature activation | +7.35e+00 |
Pos logit contributions | |
beh | +0.064 |
yours | +0.061 |
possible | +0.055 |
ours | +0.053 |
ieve | +0.052 |
Neg logit contributions | |
boy | -0.028 |
little | -0.026 |
girl | -0.026 |
sky | -0.025 |
man | -0.024 |
Token | away |
Token ID | 1497 |
Feature activation | -1.09e+01 |
Pos logit contributions | |
little | +0.314 |
boy | +0.309 |
and | +0.300 |
man | +0.297 |
girl | +0.294 |
Neg logit contributions | |
beh | -0.472 |
yours | -0.431 |
possible | -0.377 |
ieve | -0.376 |
ours | -0.375 |
Token | as |
Token ID | 355 |
Feature activation | -9.30e+00 |
Pos logit contributions | |
beh | +0.731 |
yours | +0.703 |
possible | +0.608 |
ieve | +0.605 |
unnecessary | +0.604 |
Neg logit contributions | |
little | -0.417 |
boy | -0.407 |
girl | -0.384 |
man | -0.381 |
was | -0.360 |
Token | fast |
Token ID | 3049 |
Feature activation | +7.30e-01 |
Pos logit contributions | |
beh | +0.640 |
yours | +0.604 |
ours | +0.527 |
possible | +0.525 |
ieve | +0.521 |
Neg logit contributions | |
boy | -0.345 |
little | -0.336 |
girl | -0.324 |
man | -0.309 |
sky | -0.295 |
Token | as |
Token ID | 355 |
Feature activation | -9.66e+00 |
Pos logit contributions | |
little | +0.039 |
boy | +0.039 |
girl | +0.037 |
man | +0.037 |
sky | +0.037 |
Neg logit contributions | |
beh | -0.039 |
yours | -0.035 |
possible | -0.030 |
ours | -0.030 |
ourselves | -0.028 |
Token | she |
Token ID | 673 |
Feature activation | -2.26e+00 |
Pos logit contributions | |
beh | +0.711 |
yours | +0.653 |
ours | +0.588 |
ieve | +0.574 |
eking | +0.570 |
Neg logit contributions | |
little | -0.327 |
boy | -0.325 |
girl | -0.308 |
man | -0.300 |
sky | -0.284 |
Token | could |
Token ID | 714 |
Feature activation | +2.31e+00 |
Pos logit contributions | |
beh | +0.138 |
yours | +0.130 |
possible | +0.118 |
unnecessary | +0.114 |
ieve | +0.113 |
Neg logit contributions | |
boy | -0.084 |
girl | -0.080 |
little | -0.079 |
man | -0.074 |
sky | -0.073 |
Token | . |
Token ID | 13 |
Feature activation | -8.24e+00 |
Pos logit contributions | |
little | +0.089 |
boy | +0.087 |
man | +0.082 |
girl | +0.082 |
was | +0.082 |
Neg logit contributions | |
beh | -0.155 |
yours | -0.146 |
ours | -0.129 |
possible | -0.127 |
unnecessary | -0.126 |
Token | and |
Token ID | 290 |
Feature activation | +1.24e+01 |
Pos logit contributions | |
beh | +0.374 |
yours | +0.358 |
possible | +0.326 |
ours | +0.316 |
unnecessary | +0.313 |
Neg logit contributions | |
boy | -0.138 |
little | -0.130 |
girl | -0.126 |
man | -0.118 |
sky | -0.117 |
Token | praised |
Token ID | 15342 |
Feature activation | +1.42e-01 |
Pos logit contributions | |
boy | +0.442 |
little | +0.421 |
girl | +0.409 |
sky | +0.407 |
man | +0.391 |
Neg logit contributions | |
beh | -0.827 |
yours | -0.785 |
possible | -0.709 |
theirs | -0.676 |
ours | -0.672 |
Token | her |
Token ID | 607 |
Feature activation | -6.04e+00 |
Pos logit contributions | |
little | +0.006 |
girl | +0.005 |
boy | +0.005 |
and | +0.005 |
was | +0.005 |
Neg logit contributions | |
beh | -0.010 |
yours | -0.009 |
possible | -0.008 |
eking | -0.008 |
rying | -0.008 |
Token | for |
Token ID | 329 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
beh | +0.308 |
yours | +0.290 |
possible | +0.248 |
ours | +0.239 |
unnecessary | +0.233 |
Neg logit contributions | |
boy | -0.326 |
little | -0.323 |
girl | -0.314 |
man | -0.302 |
sky | -0.293 |
Token | being |
Token ID | 852 |
Feature activation | -9.31e+00 |
Pos logit contributions | |
beh | +0.645 |
yours | +0.575 |
possible | +0.520 |
unnecessary | +0.515 |
ours | +0.512 |
Neg logit contributions | |
little | -0.481 |
boy | -0.437 |
was | -0.427 |
girl | -0.425 |
man | -0.419 |
Token | so |
Token ID | 523 |
Feature activation | -1.50e+00 |
Pos logit contributions | |
beh | +0.613 |
yours | +0.591 |
possible | +0.526 |
ours | +0.520 |
unnecessary | +0.515 |
Neg logit contributions | |
little | -0.348 |
boy | -0.346 |
girl | -0.334 |
man | -0.306 |
was | -0.300 |
Token | independent |
Token ID | 4795 |
Feature activation | -2.41e+00 |
Pos logit contributions | |
beh | +0.108 |
yours | +0.107 |
ours | +0.094 |
possible | +0.094 |
theirs | +0.093 |
Neg logit contributions | |
little | -0.048 |
boy | -0.044 |
girl | -0.043 |
man | -0.040 |
was | -0.036 |
Token | . |
Token ID | 13 |
Feature activation | -1.05e+01 |
Pos logit contributions | |
beh | +0.160 |
yours | +0.148 |
possible | +0.132 |
ours | +0.131 |
unnecessary | +0.130 |
Neg logit contributions | |
boy | -0.097 |
little | -0.093 |
girl | -0.092 |
man | -0.088 |
was | -0.086 |
Token | They |
Token ID | 1119 |
Feature activation | -6.19e-01 |
Pos logit contributions | |
beh | +0.666 |
yours | +0.646 |
possible | +0.563 |
ours | +0.552 |
theirs | +0.545 |
Neg logit contributions | |
boy | -0.397 |
girl | -0.371 |
little | -0.369 |
man | -0.364 |
sky | -0.340 |
Token | sat |
Token ID | 3332 |
Feature activation | +3.51e+00 |
Pos logit contributions | |
beh | +0.037 |
yours | +0.034 |
possible | +0.030 |
theirs | +0.030 |
ieve | +0.029 |
Neg logit contributions | |
boy | -0.025 |
girl | -0.024 |
little | -0.024 |
sky | -0.023 |
man | -0.023 |
Token | together |
Token ID | 1978 |
Feature activation | -1.61e+01 |
Pos logit contributions | |
little | +0.189 |
boy | +0.187 |
sky | +0.177 |
girl | +0.177 |
man | +0.175 |
Neg logit contributions | |
beh | -0.158 |
yours | -0.147 |
rying | -0.126 |
ourselves | -0.125 |
eking | -0.121 |
Token | Joe |
Token ID | 5689 |
Feature activation | -1.03e+01 |
Pos logit contributions | |
boy | +0.088 |
girl | +0.085 |
little | +0.084 |
sky | +0.082 |
man | +0.080 |
Neg logit contributions | |
beh | -0.123 |
yours | -0.116 |
possible | -0.108 |
theirs | -0.099 |
ours | -0.097 |
Token | told |
Token ID | 1297 |
Feature activation | -8.31e-01 |
Pos logit contributions | |
beh | +0.739 |
yours | +0.696 |
possible | +0.636 |
ieve | +0.603 |
eking | +0.600 |
Neg logit contributions | |
boy | -0.289 |
little | -0.277 |
sky | -0.275 |
girl | -0.272 |
man | -0.242 |
Token | him |
Token ID | 683 |
Feature activation | -1.62e+01 |
Pos logit contributions | |
beh | +0.043 |
yours | +0.040 |
ours | +0.034 |
possible | +0.034 |
unnecessary | +0.032 |
Neg logit contributions | |
little | -0.045 |
boy | -0.045 |
man | -0.042 |
girl | -0.042 |
was | -0.041 |
Token | " |
Token ID | 366 |
Feature activation | +7.12e+00 |
Pos logit contributions | |
beh | +1.031 |
yours | +0.958 |
ours | +0.839 |
possible | +0.837 |
unnecessary | +0.823 |
Neg logit contributions | |
boy | -0.682 |
little | -0.673 |
girl | -0.634 |
was | -0.628 |
man | -0.627 |
Token | Don |
Token ID | 3987 |
Feature activation | +1.74e+00 |
Pos logit contributions | |
boy | +0.322 |
little | +0.306 |
girl | +0.306 |
sky | +0.289 |
man | +0.289 |
Neg logit contributions | |
beh | -0.414 |
yours | -0.400 |
ours | -0.348 |
possible | -0.347 |
theirs | -0.335 |
Token | 't |
Token ID | 470 |
Feature activation | -6.15e+00 |
Pos logit contributions | |
boy | +0.100 |
little | +0.097 |
girl | +0.095 |
man | +0.092 |
sky | +0.092 |
Neg logit contributions | |
beh | -0.081 |
yours | -0.075 |
possible | -0.064 |
ours | -0.060 |
unnecessary | -0.059 |
Token | worry |
Token ID | 5490 |
Feature activation | -4.94e+00 |
Pos logit contributions | |
beh | +0.384 |
yours | +0.365 |
ieve | +0.334 |
theirs | +0.328 |
irs | +0.325 |
Neg logit contributions | |
little | -0.273 |
boy | -0.266 |
girl | -0.245 |
man | -0.227 |
was | -0.227 |
Token | , |
Token ID | 11 |
Feature activation | +1.10e+01 |
Pos logit contributions | |
beh | +0.405 |
yours | +0.372 |
possible | +0.343 |
ieve | +0.341 |
eking | +0.337 |
Neg logit contributions | |
little | -0.133 |
boy | -0.130 |
girl | -0.115 |
man | -0.107 |
was | -0.099 |
Token | I |
Token ID | 314 |
Feature activation | -1.11e+01 |
Pos logit contributions | |
boy | +0.559 |
little | +0.547 |
girl | +0.532 |
man | +0.513 |
sky | +0.508 |
Neg logit contributions | |
beh | -0.588 |
yours | -0.545 |
possible | -0.475 |
ours | -0.452 |
unnecessary | -0.449 |
Token | 'll |
Token ID | 1183 |
Feature activation | -5.37e+00 |
Pos logit contributions | |
beh | +0.664 |
yours | +0.629 |
possible | +0.558 |
ours | +0.530 |
unnecessary | +0.524 |
Neg logit contributions | |
boy | -0.486 |
little | -0.469 |
girl | -0.461 |
man | -0.442 |
sky | -0.441 |
Token | come |
Token ID | 1282 |
Feature activation | -5.16e+00 |
Pos logit contributions | |
beh | +0.337 |
yours | +0.293 |
unnecessary | +0.275 |
ours | +0.267 |
possible | +0.265 |
Neg logit contributions | |
little | -0.252 |
boy | -0.229 |
was | -0.226 |
man | -0.223 |
and | -0.219 |
Token | mom |
Token ID | 1995 |
Feature activation | +1.05e+01 |
Pos logit contributions | |
yours | +0.345 |
beh | +0.335 |
possible | +0.307 |
ours | +0.305 |
theirs | +0.305 |
Neg logit contributions | |
boy | -0.141 |
sky | -0.130 |
girl | -0.128 |
little | -0.128 |
was | -0.122 |
Token | and |
Token ID | 290 |
Feature activation | +1.34e+01 |
Pos logit contributions | |
boy | +0.177 |
little | +0.156 |
girl | +0.147 |
sky | +0.147 |
man | +0.131 |
Neg logit contributions | |
yours | -0.848 |
beh | -0.839 |
possible | -0.760 |
theirs | -0.752 |
ours | -0.745 |
Token | dad |
Token ID | 9955 |
Feature activation | +5.53e+00 |
Pos logit contributions | |
sky | +0.518 |
boy | +0.512 |
little | +0.489 |
girl | +0.486 |
man | +0.444 |
Neg logit contributions | |
yours | -0.769 |
beh | -0.763 |
theirs | -0.670 |
ours | -0.660 |
possible | -0.648 |
Token | for |
Token ID | 329 |
Feature activation | -9.44e+00 |
Pos logit contributions | |
boy | +0.221 |
little | +0.215 |
girl | +0.209 |
man | +0.200 |
sky | +0.196 |
Neg logit contributions | |
beh | -0.359 |
yours | -0.337 |
possible | -0.302 |
ours | -0.291 |
unnecessary | -0.291 |
Token | each |
Token ID | 1123 |
Feature activation | -7.45e+00 |
Pos logit contributions | |
beh | +0.577 |
yours | +0.506 |
unnecessary | +0.465 |
possible | +0.455 |
ours | +0.440 |
Neg logit contributions | |
little | -0.431 |
boy | -0.422 |
girl | -0.399 |
man | -0.399 |
was | -0.379 |
Token | one |
Token ID | 530 |
Feature activation | -6.47e+00 |
Pos logit contributions | |
beh | +0.566 |
yours | +0.551 |
possible | +0.499 |
ours | +0.484 |
theirs | +0.479 |
Neg logit contributions | |
boy | -0.200 |
little | -0.185 |
girl | -0.181 |
sky | -0.171 |
man | -0.170 |
Token | and |
Token ID | 290 |
Feature activation | +1.36e+01 |
Pos logit contributions | |
beh | +0.393 |
yours | +0.356 |
possible | +0.314 |
unnecessary | +0.309 |
ours | +0.306 |
Neg logit contributions | |
boy | -0.299 |
little | -0.291 |
girl | -0.285 |
man | -0.273 |
sky | -0.268 |
Token | hugged |
Token ID | 41130 |
Feature activation | +4.29e+00 |
Pos logit contributions | |
boy | +0.484 |
little | +0.469 |
sky | +0.456 |
girl | +0.451 |
man | +0.433 |
Neg logit contributions | |
beh | -0.871 |
yours | -0.852 |
possible | -0.755 |
theirs | -0.740 |
ours | -0.730 |
Token | them |
Token ID | 606 |
Feature activation | -1.04e+01 |
Pos logit contributions | |
boy | +0.216 |
little | +0.211 |
girl | +0.207 |
man | +0.201 |
was | +0.193 |
Neg logit contributions | |
beh | -0.249 |
yours | -0.225 |
possible | -0.203 |
ours | -0.197 |
eking | -0.193 |
Token | . |
Token ID | 13 |
Feature activation | -9.18e+00 |
Pos logit contributions | |
beh | +0.690 |
yours | +0.632 |
unnecessary | +0.578 |
possible | +0.569 |
ieve | +0.551 |
Neg logit contributions | |
little | -0.429 |
boy | -0.428 |
girl | -0.407 |
man | -0.398 |
and | -0.381 |
Token |
|
Token ID | 198 |
Feature activation | +6.80e+00 |
Pos logit contributions | |
yours | +0.649 |
beh | +0.631 |
possible | +0.570 |
theirs | +0.569 |
ours | +0.561 |
Neg logit contributions | |
boy | -0.262 |
girl | -0.241 |
little | -0.234 |
man | -0.226 |
sky | -0.209 |
Token | were |
Token ID | 547 |
Feature activation | +4.82e+00 |
Pos logit contributions | |
beh | +0.299 |
yours | +0.270 |
eking | +0.246 |
ieve | +0.244 |
ourselves | +0.243 |
Neg logit contributions | |
boy | -0.131 |
sky | -0.128 |
little | -0.123 |
girl | -0.123 |
man | -0.117 |
Token | curious |
Token ID | 11040 |
Feature activation | +1.22e+00 |
Pos logit contributions | |
boy | +0.180 |
little | +0.174 |
girl | +0.168 |
man | +0.161 |
sky | +0.159 |
Neg logit contributions | |
beh | -0.322 |
yours | -0.306 |
possible | -0.275 |
ours | -0.264 |
unnecessary | -0.262 |
Token | . |
Token ID | 13 |
Feature activation | -1.04e+01 |
Pos logit contributions | |
boy | +0.052 |
little | +0.050 |
girl | +0.049 |
man | +0.047 |
sky | +0.046 |
Neg logit contributions | |
beh | -0.078 |
yours | -0.073 |
possible | -0.065 |
ours | -0.063 |
unnecessary | -0.062 |
Token | They |
Token ID | 1119 |
Feature activation | -5.34e-01 |
Pos logit contributions | |
beh | +0.636 |
yours | +0.615 |
theirs | +0.550 |
ours | +0.543 |
ourselves | +0.537 |
Neg logit contributions | |
little | -0.382 |
boy | -0.379 |
girl | -0.362 |
man | -0.360 |
sky | -0.346 |
Token | wondered |
Token ID | 14028 |
Feature activation | +3.21e+00 |
Pos logit contributions | |
beh | +0.033 |
yours | +0.031 |
ours | +0.027 |
theirs | +0.027 |
properly | +0.026 |
Neg logit contributions | |
sky | -0.018 |
boy | -0.018 |
girl | -0.017 |
little | -0.016 |
man | -0.016 |
Token | what |
Token ID | 644 |
Feature activation | -1.80e+01 |
Pos logit contributions | |
little | +0.169 |
boy | +0.165 |
girl | +0.156 |
and | +0.156 |
was | +0.155 |
Neg logit contributions | |
beh | -0.187 |
yours | -0.155 |
possible | -0.135 |
unnecessary | -0.135 |
anks | -0.132 |
Token | the |
Token ID | 262 |
Feature activation | +4.75e+00 |
Pos logit contributions | |
beh | +1.217 |
yours | +1.210 |
possible | +1.102 |
theirs | +1.032 |
ours | +1.031 |
Neg logit contributions | |
boy | -0.554 |
little | -0.544 |
girl | -0.529 |
sky | -0.517 |
man | -0.478 |
Token | surprise |
Token ID | 5975 |
Feature activation | +1.60e+00 |
Pos logit contributions | |
boy | +0.160 |
little | +0.155 |
girl | +0.150 |
sky | +0.142 |
man | +0.142 |
Neg logit contributions | |
beh | -0.316 |
yours | -0.308 |
possible | -0.280 |
ours | -0.266 |
unnecessary | -0.261 |
Token | was |
Token ID | 373 |
Feature activation | +7.33e+00 |
Pos logit contributions | |
boy | +0.020 |
little | +0.020 |
girl | +0.017 |
sky | +0.015 |
man | +0.013 |
Neg logit contributions | |
beh | -0.139 |
yours | -0.137 |
possible | -0.128 |
theirs | -0.124 |
ours | -0.123 |
Token | . |
Token ID | 13 |
Feature activation | -1.05e+01 |
Pos logit contributions | |
little | +0.275 |
boy | +0.269 |
and | +0.268 |
man | +0.265 |
was | +0.263 |
Neg logit contributions | |
beh | -0.534 |
yours | -0.440 |
unnecessary | -0.414 |
irs | -0.413 |
ourselves | -0.394 |
Token | They |
Token ID | 1119 |
Feature activation | -6.23e-01 |
Pos logit contributions | |
beh | +0.623 |
yours | +0.601 |
theirs | +0.539 |
ours | +0.530 |
ourselves | +0.530 |
Neg logit contributions | |
boy | -0.391 |
little | -0.390 |
girl | -0.378 |
man | -0.376 |
sky | -0.356 |
Token | " |
Token ID | 1 |
Feature activation | +7.50e+00 |
Pos logit contributions | |
boy | +0.170 |
little | +0.159 |
girl | +0.159 |
sky | +0.155 |
man | +0.137 |
Neg logit contributions | |
beh | -0.396 |
yours | -0.382 |
possible | -0.340 |
ourselves | -0.338 |
ours | -0.336 |
Token | No |
Token ID | 2949 |
Feature activation | +2.81e+00 |
Pos logit contributions | |
boy | +0.270 |
little | +0.265 |
girl | +0.254 |
sky | +0.247 |
man | +0.245 |
Neg logit contributions | |
beh | -0.495 |
yours | -0.470 |
possible | -0.427 |
ours | -0.408 |
ourselves | -0.402 |
Token | , |
Token ID | 11 |
Feature activation | +9.50e+00 |
Pos logit contributions | |
boy | +0.058 |
little | +0.056 |
girl | +0.050 |
man | +0.047 |
sky | +0.042 |
Neg logit contributions | |
beh | -0.243 |
yours | -0.227 |
possible | -0.211 |
ieve | -0.205 |
ours | -0.205 |
Token | it |
Token ID | 340 |
Feature activation | -4.66e+00 |
Pos logit contributions | |
boy | +0.450 |
little | +0.439 |
girl | +0.426 |
man | +0.412 |
sky | +0.405 |
Neg logit contributions | |
beh | -0.547 |
yours | -0.505 |
possible | -0.449 |
unnecessary | -0.425 |
ieve | -0.423 |
Token | is |
Token ID | 318 |
Feature activation | +2.99e+00 |
Pos logit contributions | |
beh | +0.370 |
yours | +0.355 |
possible | +0.325 |
ours | +0.315 |
unnecessary | +0.313 |
Neg logit contributions | |
boy | -0.115 |
little | -0.109 |
girl | -0.104 |
man | -0.096 |
sky | -0.094 |
Token | mine |
Token ID | 6164 |
Feature activation | -1.71e+01 |
Pos logit contributions | |
little | +0.175 |
boy | +0.171 |
man | +0.160 |
was | +0.158 |
girl | +0.158 |
Neg logit contributions | |
beh | -0.157 |
ieve | -0.124 |
unnecessary | -0.115 |
irs | -0.115 |
ility | -0.114 |
Token | !" |
Token ID | 2474 |
Feature activation | -1.30e+01 |
Pos logit contributions | |
beh | +1.031 |
yours | +0.854 |
unnecessary | +0.841 |
ieve | +0.795 |
ility | +0.793 |
Neg logit contributions | |
boy | -0.859 |
little | -0.837 |
girl | -0.798 |
and | -0.794 |
man | -0.781 |
Token | Tom |
Token ID | 4186 |
Feature activation | -4.29e+00 |
Pos logit contributions | |
beh | +0.748 |
yours | +0.711 |
possible | +0.623 |
ours | +0.602 |
theirs | +0.590 |
Neg logit contributions | |
boy | -0.590 |
little | -0.568 |
girl | -0.563 |
man | -0.537 |
sky | -0.536 |
Token | said |
Token ID | 531 |
Feature activation | +9.63e-01 |
Pos logit contributions | |
beh | +0.297 |
yours | +0.270 |
eking | +0.240 |
possible | +0.238 |
ours | +0.236 |
Neg logit contributions | |
sky | -0.135 |
boy | -0.133 |
girl | -0.129 |
little | -0.128 |
man | -0.117 |
Token | . |
Token ID | 13 |
Feature activation | -1.03e+01 |
Pos logit contributions | |
little | +0.035 |
boy | +0.033 |
and | +0.031 |
was | +0.031 |
girl | +0.030 |
Neg logit contributions | |
beh | -0.071 |
yours | -0.069 |
ours | -0.064 |
unnecessary | -0.063 |
irs | -0.062 |
Token | " |
Token ID | 366 |
Feature activation | +5.88e+00 |
Pos logit contributions | |
beh | +0.631 |
yours | +0.619 |
theirs | +0.553 |
possible | +0.552 |
properly | +0.528 |
Neg logit contributions | |
girl | -0.295 |
boy | -0.291 |
little | -0.283 |
sky | -0.277 |
child | -0.262 |
Token | . |
Token ID | 13 |
Feature activation | -9.70e+00 |
Pos logit contributions | |
beh | +0.561 |
yours | +0.519 |
irs | +0.484 |
ieve | +0.474 |
ours | +0.473 |
Neg logit contributions | |
boy | -0.294 |
and | -0.286 |
little | -0.285 |
man | -0.280 |
was | -0.278 |
Token | The |
Token ID | 383 |
Feature activation | +7.03e+00 |
Pos logit contributions | |
beh | +0.631 |
yours | +0.604 |
ours | +0.515 |
ourselves | +0.514 |
possible | +0.513 |
Neg logit contributions | |
boy | -0.346 |
girl | -0.333 |
little | -0.332 |
man | -0.313 |
sky | -0.304 |
Token | bear |
Token ID | 6842 |
Feature activation | +1.08e+01 |
Pos logit contributions | |
sky | +0.165 |
boy | +0.159 |
was | +0.152 |
girl | +0.145 |
field | +0.144 |
Neg logit contributions | |
beh | -0.468 |
yours | -0.443 |
ieve | -0.420 |
possible | -0.411 |
unnecessary | -0.398 |
Token | forg |
Token ID | 5784 |
Feature activation | +8.95e-01 |
Pos logit contributions | |
sky | +0.384 |
girl | +0.356 |
boy | +0.355 |
little | +0.325 |
field | +0.324 |
Neg logit contributions | |
beh | -0.639 |
yours | -0.543 |
properly | -0.506 |
able | -0.497 |
ieve | -0.496 |
Token | ave |
Token ID | 1015 |
Feature activation | -9.93e+00 |
Pos logit contributions | |
boy | +0.050 |
little | +0.048 |
girl | +0.047 |
man | +0.046 |
sky | +0.046 |
Neg logit contributions | |
beh | -0.043 |
yours | -0.040 |
possible | -0.034 |
unnecessary | -0.032 |
ours | -0.032 |
Token | him |
Token ID | 683 |
Feature activation | -1.67e+01 |
Pos logit contributions | |
beh | +0.516 |
yours | +0.486 |
possible | +0.417 |
ours | +0.397 |
unnecessary | +0.393 |
Neg logit contributions | |
boy | -0.517 |
little | -0.500 |
girl | -0.495 |
man | -0.477 |
sky | -0.477 |
Token | and |
Token ID | 290 |
Feature activation | +1.33e+01 |
Pos logit contributions | |
beh | +1.558 |
yours | +1.504 |
possible | +1.396 |
ours | +1.362 |
unnecessary | +1.350 |
Neg logit contributions | |
boy | -0.181 |
little | -0.159 |
girl | -0.141 |
man | -0.115 |
sky | -0.109 |
Token | they |
Token ID | 484 |
Feature activation | -1.39e+00 |
Pos logit contributions | |
boy | +0.501 |
little | +0.486 |
girl | +0.475 |
sky | +0.464 |
man | +0.446 |
Neg logit contributions | |
beh | -0.863 |
yours | -0.817 |
possible | -0.733 |
theirs | -0.697 |
ours | -0.695 |
Token | all |
Token ID | 477 |
Feature activation | -6.82e+00 |
Pos logit contributions | |
beh | +0.087 |
yours | +0.083 |
possible | +0.073 |
theirs | +0.070 |
ours | +0.070 |
Neg logit contributions | |
boy | -0.056 |
little | -0.054 |
girl | -0.053 |
man | -0.050 |
sky | -0.050 |
Token | played |
Token ID | 2826 |
Feature activation | +1.62e+00 |
Pos logit contributions | |
beh | +0.509 |
yours | +0.491 |
possible | +0.448 |
ours | +0.436 |
unnecessary | +0.432 |
Neg logit contributions | |
boy | -0.202 |
little | -0.191 |
girl | -0.185 |
man | -0.171 |
sky | -0.167 |
Token | together |
Token ID | 1978 |
Feature activation | -1.50e+01 |
Pos logit contributions | |
boy | +0.084 |
little | +0.081 |
girl | +0.078 |
man | +0.077 |
was | +0.075 |
Neg logit contributions | |
beh | -0.089 |
yours | -0.083 |
possible | -0.073 |
ours | -0.070 |
unnecessary | -0.067 |
Token | more |
Token ID | 517 |
Feature activation | -4.13e+00 |
Pos logit contributions | |
beh | +0.542 |
yours | +0.500 |
possible | +0.440 |
ours | +0.430 |
unnecessary | +0.425 |
Neg logit contributions | |
boy | -0.384 |
little | -0.382 |
girl | -0.362 |
man | -0.356 |
sky | -0.351 |
Token | about |
Token ID | 546 |
Feature activation | -1.17e+01 |
Pos logit contributions | |
beh | +0.250 |
yours | +0.236 |
possible | +0.207 |
ours | +0.201 |
unnecessary | +0.197 |
Neg logit contributions | |
boy | -0.180 |
little | -0.177 |
girl | -0.171 |
man | -0.165 |
sky | -0.163 |
Token | it |
Token ID | 340 |
Feature activation | -4.90e+00 |
Pos logit contributions | |
beh | +0.767 |
yours | +0.707 |
possible | +0.626 |
ours | +0.616 |
unnecessary | +0.608 |
Neg logit contributions | |
little | -0.464 |
boy | -0.462 |
girl | -0.431 |
man | -0.425 |
sky | -0.417 |
Token | and |
Token ID | 290 |
Feature activation | +1.22e+01 |
Pos logit contributions | |
beh | +0.346 |
yours | +0.321 |
ours | +0.287 |
unnecessary | +0.285 |
possible | +0.277 |
Neg logit contributions | |
little | -0.174 |
boy | -0.169 |
was | -0.165 |
girl | -0.158 |
man | -0.158 |
Token | see |
Token ID | 766 |
Feature activation | -7.38e+00 |
Pos logit contributions | |
boy | +0.503 |
little | +0.491 |
girl | +0.472 |
man | +0.461 |
was | +0.455 |
Neg logit contributions | |
beh | -0.767 |
yours | -0.724 |
possible | -0.635 |
ours | -0.630 |
unnecessary | -0.622 |
Token | what |
Token ID | 644 |
Feature activation | -1.81e+01 |
Pos logit contributions | |
beh | +0.460 |
yours | +0.421 |
possible | +0.373 |
ours | +0.363 |
eking | +0.357 |
Neg logit contributions | |
boy | -0.319 |
little | -0.317 |
girl | -0.301 |
man | -0.293 |
sky | -0.286 |
Token | he |
Token ID | 339 |
Feature activation | -6.69e+00 |
Pos logit contributions | |
beh | +1.226 |
yours | +1.179 |
possible | +1.071 |
ours | +1.015 |
theirs | +1.001 |
Neg logit contributions | |
boy | -0.625 |
little | -0.601 |
girl | -0.591 |
sky | -0.556 |
man | -0.553 |
Token | could |
Token ID | 714 |
Feature activation | -1.75e-01 |
Pos logit contributions | |
beh | +0.526 |
yours | +0.505 |
possible | +0.464 |
ieve | +0.446 |
ours | +0.444 |
Neg logit contributions | |
boy | -0.160 |
little | -0.150 |
girl | -0.145 |
sky | -0.133 |
man | -0.132 |
Token | do |
Token ID | 466 |
Feature activation | -4.69e+00 |
Pos logit contributions | |
beh | +0.012 |
yours | +0.011 |
unnecessary | +0.011 |
ours | +0.010 |
eking | +0.010 |
Neg logit contributions | |
little | -0.006 |
was | -0.006 |
boy | -0.006 |
and | -0.006 |
man | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
beh | +0.351 |
unnecessary | +0.308 |
yours | +0.307 |
ours | +0.289 |
eking | +0.276 |
Neg logit contributions | |
was | -0.158 |
and | -0.155 |
little | -0.145 |
boy | -0.142 |
man | -0.137 |
Token | Everyone |
Token ID | 11075 |
Feature activation | -8.34e+00 |
Pos logit contributions | |
beh | +0.777 |
yours | +0.748 |
possible | +0.680 |
ours | +0.649 |
theirs | +0.647 |
Neg logit contributions | |
boy | -0.298 |
little | -0.285 |
girl | -0.283 |
man | -0.263 |
sky | -0.249 |
Token | . |
Token ID | 13 |
Feature activation | -9.87e+00 |
Pos logit contributions | |
little | +0.243 |
boy | +0.234 |
girl | +0.221 |
man | +0.221 |
and | +0.215 |
Neg logit contributions | |
beh | -0.376 |
yours | -0.323 |
irs | -0.290 |
anks | -0.289 |
eking | -0.287 |
Token | They |
Token ID | 1119 |
Feature activation | +7.79e-02 |
Pos logit contributions | |
beh | +0.623 |
yours | +0.613 |
theirs | +0.562 |
possible | +0.551 |
ourselves | +0.539 |
Neg logit contributions | |
little | -0.322 |
girl | -0.322 |
boy | -0.318 |
man | -0.297 |
sky | -0.283 |
Token | wanted |
Token ID | 2227 |
Feature activation | -5.17e-01 |
Pos logit contributions | |
boy | +0.003 |
girl | +0.003 |
little | +0.002 |
sky | +0.002 |
man | +0.002 |
Neg logit contributions | |
beh | -0.005 |
yours | -0.005 |
ieve | -0.004 |
eking | -0.004 |
possible | -0.004 |
Token | to |
Token ID | 284 |
Feature activation | +4.07e-01 |
Pos logit contributions | |
beh | +0.033 |
yours | +0.031 |
possible | +0.028 |
ours | +0.027 |
ieve | +0.026 |
Neg logit contributions | |
boy | -0.021 |
little | -0.020 |
girl | -0.020 |
man | -0.019 |
sky | -0.019 |
Token | pet |
Token ID | 4273 |
Feature activation | +6.99e+00 |
Pos logit contributions | |
boy | +0.017 |
little | +0.017 |
girl | +0.016 |
man | +0.016 |
sky | +0.015 |
Neg logit contributions | |
beh | -0.026 |
yours | -0.024 |
possible | -0.021 |
eking | -0.021 |
rying | -0.021 |
Token | him |
Token ID | 683 |
Feature activation | -1.69e+01 |
Pos logit contributions | |
boy | +0.442 |
little | +0.429 |
girl | +0.421 |
man | +0.417 |
sky | +0.406 |
Neg logit contributions | |
beh | -0.304 |
yours | -0.269 |
possible | -0.230 |
ours | -0.212 |
eking | -0.208 |
Token | . |
Token ID | 13 |
Feature activation | -9.74e+00 |
Pos logit contributions | |
beh | +1.337 |
yours | +1.282 |
possible | +1.172 |
ours | +1.138 |
unnecessary | +1.126 |
Neg logit contributions | |
boy | -0.419 |
little | -0.395 |
girl | -0.379 |
man | -0.352 |
sky | -0.347 |
Token | They |
Token ID | 1119 |
Feature activation | +1.96e-01 |
Pos logit contributions | |
beh | +0.669 |
yours | +0.661 |
theirs | +0.602 |
possible | +0.593 |
ourselves | +0.591 |
Neg logit contributions | |
little | -0.266 |
boy | -0.263 |
girl | -0.262 |
sky | -0.240 |
man | -0.239 |
Token | forgot |
Token ID | 16453 |
Feature activation | -3.80e+00 |
Pos logit contributions | |
boy | +0.007 |
sky | +0.006 |
girl | +0.006 |
little | +0.006 |
man | +0.006 |
Neg logit contributions | |
beh | -0.012 |
yours | -0.011 |
theirs | -0.010 |
eking | -0.010 |
ieve | -0.010 |
Token | that |
Token ID | 326 |
Feature activation | -7.55e+00 |
Pos logit contributions | |
beh | +0.209 |
yours | +0.183 |
possible | +0.161 |
ieve | +0.156 |
unnecessary | +0.155 |
Neg logit contributions | |
little | -0.198 |
boy | -0.195 |
man | -0.184 |
was | -0.183 |
girl | -0.182 |
Token | he |
Token ID | 339 |
Feature activation | -5.77e+00 |
Pos logit contributions | |
beh | +0.474 |
yours | +0.456 |
possible | +0.412 |
theirs | +0.392 |
ours | +0.389 |
Neg logit contributions | |
boy | -0.290 |
little | -0.283 |
girl | -0.277 |
sky | -0.262 |
man | -0.259 |