Token | lem |
Token ID | 10671 |
Feature activation | +8.09e-01 |
Pos logit contributions | |
TIT | +1.009 |
endorsement | +0.970 |
provocation | +0.929 |
approval | +0.910 |
� | +0.890 |
Neg logit contributions | |
Bridgewater | -1.003 |
Unic | -0.974 |
Flavoring | -0.921 |
Reviewer | -0.903 |
wegian | -0.891 |
Token | Shake |
Token ID | 41113 |
Feature activation | +3.47e-01 |
Pos logit contributions | |
TIT | +0.627 |
endorsement | +0.602 |
provocation | +0.582 |
approval | +0.571 |
� | +0.560 |
Neg logit contributions | |
Bridgewater | -0.497 |
Unic | -0.478 |
Flavoring | -0.445 |
wegian | -0.437 |
Reviewer | -0.433 |
Token | � |
Token ID | 447 |
Feature activation | +4.92e-01 |
Pos logit contributions | |
TIT | +0.220 |
endorsement | +0.211 |
provocation | +0.201 |
approval | +0.197 |
� | +0.191 |
Neg logit contributions | |
Bridgewater | -0.262 |
Unic | -0.254 |
Flavoring | -0.241 |
wegian | -0.236 |
Reviewer | -0.236 |
Token | � |
Token ID | 251 |
Feature activation | -1.04e-01 |
Pos logit contributions | |
TIT | +0.323 |
endorsement | +0.311 |
provocation | +0.298 |
approval | +0.294 |
� | +0.287 |
Neg logit contributions | |
Bridgewater | -0.358 |
Unic | -0.349 |
Flavoring | -0.327 |
wegian | -0.321 |
Reviewer | -0.319 |
Token | bl |
Token ID | 698 |
Feature activation | -9.23e-01 |
Pos logit contributions | |
Bridgewater | +0.064 |
Unic | +0.061 |
Flavoring | +0.058 |
Reviewer | +0.056 |
wegian | +0.056 |
Neg logit contributions | |
TIT | -0.080 |
endorsement | -0.077 |
provocation | -0.075 |
approval | -0.073 |
� | -0.072 |
Token | ares |
Token ID | 3565 |
Feature activation | +3.75e+00 |
Pos logit contributions | |
Bridgewater | +0.743 |
Unic | +0.721 |
Flavoring | +0.684 |
wegian | +0.672 |
Reviewer | +0.670 |
Neg logit contributions | |
TIT | -0.539 |
endorsement | -0.513 |
provocation | -0.489 |
approval | -0.477 |
� | -0.464 |
Token | in |
Token ID | 287 |
Feature activation | +3.84e-01 |
Pos logit contributions | |
TIT | +2.914 |
endorsement | +2.858 |
provocation | +2.748 |
approval | +2.730 |
� | +2.601 |
Neg logit contributions | |
Bridgewater | -2.248 |
Unic | -2.133 |
Flavoring | -2.091 |
Reviewer | -2.031 |
wegian | -1.994 |
Token | the |
Token ID | 262 |
Feature activation | +3.12e-01 |
Pos logit contributions | |
TIT | +0.278 |
endorsement | +0.263 |
provocation | +0.254 |
approval | +0.248 |
� | +0.246 |
Neg logit contributions | |
Bridgewater | -0.257 |
Unic | -0.249 |
Flavoring | -0.229 |
wegian | -0.227 |
Reviewer | -0.225 |
Token | background |
Token ID | 4469 |
Feature activation | -6.60e-01 |
Pos logit contributions | |
TIT | +0.139 |
endorsement | +0.128 |
provocation | +0.120 |
approval | +0.115 |
� | +0.113 |
Neg logit contributions | |
Bridgewater | -0.296 |
Unic | -0.289 |
Flavoring | -0.274 |
wegian | -0.271 |
Reviewer | -0.270 |
Token | , |
Token ID | 11 |
Feature activation | +2.99e-01 |
Pos logit contributions | |
Bridgewater | +0.409 |
Unic | +0.393 |
Flavoring | +0.367 |
wegian | +0.359 |
Reviewer | +0.357 |
Neg logit contributions | |
TIT | -0.508 |
endorsement | -0.488 |
provocation | -0.471 |
approval | -0.462 |
� | -0.454 |
Token | and |
Token ID | 290 |
Feature activation | +3.71e-01 |
Pos logit contributions | |
TIT | +0.230 |
endorsement | +0.222 |
provocation | +0.214 |
approval | +0.211 |
� | +0.205 |
Neg logit contributions | |
Bridgewater | -0.186 |
Unic | -0.178 |
Flavoring | -0.168 |
Reviewer | -0.164 |
wegian | -0.164 |
Token | � |
Token ID | 250 |
Feature activation | +1.30e-01 |
Pos logit contributions | |
Bridgewater | +0.099 |
Unic | +0.096 |
Flavoring | +0.089 |
wegian | +0.087 |
Reviewer | +0.086 |
Neg logit contributions | |
TIT | -0.127 |
endorsement | -0.123 |
provocation | -0.118 |
approval | -0.117 |
� | -0.115 |
Token | We |
Token ID | 1135 |
Feature activation | +7.29e-01 |
Pos logit contributions | |
TIT | +0.095 |
endorsement | +0.091 |
provocation | +0.087 |
approval | +0.086 |
� | +0.084 |
Neg logit contributions | |
Bridgewater | -0.085 |
Unic | -0.082 |
Flavoring | -0.077 |
wegian | -0.075 |
Reviewer | -0.074 |
Token | 've |
Token ID | 1053 |
Feature activation | +1.00e+00 |
Pos logit contributions | |
TIT | +0.546 |
endorsement | +0.524 |
provocation | +0.505 |
approval | +0.496 |
� | +0.486 |
Neg logit contributions | |
Bridgewater | -0.467 |
Unic | -0.449 |
Flavoring | -0.422 |
wegian | -0.411 |
Reviewer | -0.410 |
Token | done |
Token ID | 1760 |
Feature activation | +4.11e-01 |
Pos logit contributions | |
TIT | +0.799 |
endorsement | +0.770 |
provocation | +0.743 |
approval | +0.734 |
� | +0.715 |
Neg logit contributions | |
Bridgewater | -0.590 |
Unic | -0.567 |
Flavoring | -0.535 |
Reviewer | -0.519 |
wegian | -0.517 |
Token | this |
Token ID | 428 |
Feature activation | +6.57e-01 |
Pos logit contributions | |
TIT | +0.330 |
endorsement | +0.322 |
provocation | +0.311 |
approval | +0.306 |
� | +0.296 |
Neg logit contributions | |
Bridgewater | -0.237 |
Unic | -0.228 |
Flavoring | -0.214 |
Reviewer | -0.210 |
wegian | -0.210 |
Token | effect |
Token ID | 1245 |
Feature activation | +3.68e+00 |
Pos logit contributions | |
TIT | +0.579 |
endorsement | +0.562 |
provocation | +0.546 |
approval | +0.538 |
� | +0.525 |
Neg logit contributions | |
Bridgewater | -0.331 |
Unic | -0.316 |
Flavoring | -0.294 |
wegian | -0.285 |
Reviewer | -0.284 |
Token | many |
Token ID | 867 |
Feature activation | +2.76e-01 |
Pos logit contributions | |
TIT | +3.091 |
endorsement | +3.053 |
provocation | +2.943 |
approval | +2.931 |
� | +2.831 |
Neg logit contributions | |
Bridgewater | -1.935 |
Unic | -1.823 |
Flavoring | -1.756 |
Reviewer | -1.699 |
wegian | -1.667 |
Token | times |
Token ID | 1661 |
Feature activation | +7.84e-01 |
Pos logit contributions | |
TIT | +0.157 |
endorsement | +0.149 |
provocation | +0.142 |
approval | +0.138 |
� | +0.134 |
Neg logit contributions | |
Bridgewater | -0.227 |
Unic | -0.220 |
Flavoring | -0.210 |
wegian | -0.206 |
Reviewer | -0.206 |
Token | . |
Token ID | 13 |
Feature activation | +5.23e-01 |
Pos logit contributions | |
TIT | +0.677 |
endorsement | +0.655 |
provocation | +0.634 |
approval | +0.627 |
� | +0.613 |
Neg logit contributions | |
Bridgewater | -0.412 |
Unic | -0.391 |
Flavoring | -0.364 |
Reviewer | -0.352 |
wegian | -0.352 |
Token | It |
Token ID | 632 |
Feature activation | +1.78e-01 |
Pos logit contributions | |
TIT | +0.348 |
endorsement | +0.330 |
provocation | +0.317 |
approval | +0.311 |
� | +0.304 |
Neg logit contributions | |
Bridgewater | -0.378 |
Unic | -0.365 |
Flavoring | -0.345 |
wegian | -0.341 |
Reviewer | -0.339 |
Token | 's |
Token ID | 338 |
Feature activation | +6.64e-01 |
Pos logit contributions | |
TIT | +0.142 |
endorsement | +0.137 |
provocation | +0.132 |
approval | +0.130 |
� | +0.128 |
Neg logit contributions | |
Bridgewater | -0.105 |
Unic | -0.101 |
Flavoring | -0.094 |
Reviewer | -0.092 |
wegian | -0.092 |
Token | . |
Token ID | 13 |
Feature activation | +7.09e-01 |
Pos logit contributions | |
TIT | +0.425 |
endorsement | +0.409 |
provocation | +0.394 |
approval | +0.388 |
� | +0.378 |
Neg logit contributions | |
Bridgewater | -0.369 |
Unic | -0.355 |
Flavoring | -0.333 |
wegian | -0.326 |
Reviewer | -0.324 |
Token |
|
Token ID | 198 |
Feature activation | +3.65e-01 |
Pos logit contributions | |
TIT | +0.533 |
endorsement | +0.511 |
provocation | +0.493 |
approval | +0.485 |
� | +0.474 |
Neg logit contributions | |
Bridgewater | -0.452 |
Unic | -0.434 |
Flavoring | -0.407 |
wegian | -0.400 |
Reviewer | -0.397 |
Token |
|
Token ID | 198 |
Feature activation | +4.70e-01 |
Pos logit contributions | |
TIT | +0.301 |
endorsement | +0.291 |
provocation | +0.282 |
approval | +0.276 |
� | +0.271 |
Neg logit contributions | |
Bridgewater | -0.206 |
Unic | -0.197 |
Flavoring | -0.182 |
wegian | -0.177 |
Reviewer | -0.177 |
Token | Similarly |
Token ID | 28039 |
Feature activation | +1.11e+00 |
Pos logit contributions | |
TIT | +0.339 |
endorsement | +0.325 |
provocation | +0.313 |
approval | +0.306 |
� | +0.300 |
Neg logit contributions | |
Bridgewater | -0.314 |
Unic | -0.303 |
Flavoring | -0.285 |
wegian | -0.279 |
Reviewer | -0.277 |
Token | , |
Token ID | 11 |
Feature activation | +7.43e-01 |
Pos logit contributions | |
TIT | +0.855 |
endorsement | +0.825 |
provocation | +0.797 |
approval | +0.783 |
� | +0.764 |
Neg logit contributions | |
Bridgewater | -0.687 |
Unic | -0.662 |
Flavoring | -0.624 |
wegian | -0.605 |
Reviewer | -0.602 |
Token | compare |
Token ID | 8996 |
Feature activation | +3.60e+00 |
Pos logit contributions | |
TIT | +0.591 |
endorsement | +0.572 |
provocation | +0.552 |
approval | +0.544 |
� | +0.530 |
Neg logit contributions | |
Bridgewater | -0.439 |
Unic | -0.421 |
Flavoring | -0.394 |
wegian | -0.383 |
Reviewer | -0.382 |
Token | the |
Token ID | 262 |
Feature activation | +4.31e-01 |
Pos logit contributions | |
TIT | +2.772 |
endorsement | +2.721 |
provocation | +2.627 |
approval | +2.617 |
Hait | +2.519 |
Neg logit contributions | |
Bridgewater | -2.133 |
Unic | -2.056 |
Flavoring | -1.957 |
wegian | -1.901 |
Reviewer | -1.886 |
Token | ch |
Token ID | 442 |
Feature activation | +5.20e-01 |
Pos logit contributions | |
TIT | +0.318 |
endorsement | +0.305 |
provocation | +0.294 |
approval | +0.288 |
� | +0.283 |
Neg logit contributions | |
Bridgewater | -0.281 |
Unic | -0.271 |
Flavoring | -0.254 |
wegian | -0.249 |
Reviewer | -0.247 |
Token | asm |
Token ID | 8597 |
Feature activation | +3.41e-01 |
Pos logit contributions | |
TIT | +0.380 |
endorsement | +0.365 |
provocation | +0.352 |
approval | +0.345 |
� | +0.337 |
Neg logit contributions | |
Bridgewater | -0.343 |
Unic | -0.331 |
Flavoring | -0.310 |
wegian | -0.303 |
Reviewer | -0.302 |
Token | in |
Token ID | 287 |
Feature activation | -3.29e-01 |
Pos logit contributions | |
TIT | +0.243 |
endorsement | +0.233 |
provocation | +0.224 |
approval | +0.220 |
� | +0.215 |
Neg logit contributions | |
Bridgewater | -0.231 |
Unic | -0.223 |
Flavoring | -0.210 |
wegian | -0.206 |
Reviewer | -0.205 |
Token | rational |
Token ID | 9377 |
Feature activation | -6.04e-01 |
Pos logit contributions | |
Bridgewater | +0.230 |
Unic | +0.222 |
Flavoring | +0.209 |
wegian | +0.205 |
Reviewer | +0.204 |
Neg logit contributions | |
TIT | -0.227 |
endorsement | -0.217 |
provocation | -0.208 |
approval | -0.204 |
� | -0.200 |
Token | is |
Token ID | 318 |
Feature activation | +3.75e-01 |
Pos logit contributions | |
TIT | +0.776 |
endorsement | +0.755 |
provocation | +0.731 |
approval | +0.719 |
� | +0.703 |
Neg logit contributions | |
Bridgewater | -0.577 |
Unic | -0.557 |
Flavoring | -0.523 |
Reviewer | -0.516 |
wegian | -0.508 |
Token | among |
Token ID | 1871 |
Feature activation | +3.85e-01 |
Pos logit contributions | |
TIT | +0.289 |
endorsement | +0.279 |
provocation | +0.269 |
approval | +0.265 |
� | +0.258 |
Neg logit contributions | |
Bridgewater | -0.231 |
Unic | -0.222 |
Flavoring | -0.209 |
wegian | -0.204 |
Reviewer | -0.204 |
Token | the |
Token ID | 262 |
Feature activation | +5.73e-01 |
Pos logit contributions | |
TIT | +0.305 |
endorsement | +0.296 |
provocation | +0.285 |
approval | +0.281 |
� | +0.273 |
Neg logit contributions | |
Bridgewater | -0.228 |
Unic | -0.220 |
Flavoring | -0.206 |
wegian | -0.201 |
Reviewer | -0.201 |
Token | hundreds |
Token ID | 5179 |
Feature activation | +3.75e-02 |
Pos logit contributions | |
TIT | +0.460 |
endorsement | +0.448 |
provocation | +0.431 |
approval | +0.425 |
� | +0.413 |
Neg logit contributions | |
Bridgewater | -0.333 |
Unic | -0.319 |
Flavoring | -0.300 |
wegian | -0.292 |
Reviewer | -0.292 |
Token | of |
Token ID | 286 |
Feature activation | +8.46e-01 |
Pos logit contributions | |
TIT | +0.031 |
endorsement | +0.030 |
provocation | +0.029 |
approval | +0.028 |
� | +0.028 |
Neg logit contributions | |
Bridgewater | -0.021 |
Unic | -0.021 |
Flavoring | -0.019 |
wegian | -0.019 |
Reviewer | -0.019 |
Token | Toledo |
Token ID | 36026 |
Feature activation | +3.59e+00 |
Pos logit contributions | |
TIT | +0.680 |
endorsement | +0.659 |
provocation | +0.636 |
approval | +0.625 |
� | +0.610 |
Neg logit contributions | |
Bridgewater | -0.493 |
Unic | -0.473 |
Flavoring | -0.443 |
wegian | -0.432 |
Reviewer | -0.431 |
Token | youths |
Token ID | 28247 |
Feature activation | +9.10e-01 |
Pos logit contributions | |
TIT | +2.753 |
endorsement | +2.723 |
provocation | +2.594 |
approval | +2.570 |
Hait | +2.495 |
Neg logit contributions | |
Bridgewater | -2.166 |
Unic | -2.065 |
Flavoring | -1.973 |
wegian | -1.933 |
Reviewer | -1.911 |
Token | who |
Token ID | 508 |
Feature activation | +1.65e+00 |
Pos logit contributions | |
TIT | +0.662 |
endorsement | +0.641 |
provocation | +0.618 |
approval | +0.606 |
� | +0.590 |
Neg logit contributions | |
Bridgewater | -0.594 |
Unic | -0.575 |
Flavoring | -0.545 |
Reviewer | -0.532 |
wegian | -0.530 |
Token | have |
Token ID | 423 |
Feature activation | +6.95e-01 |
Pos logit contributions | |
TIT | +1.176 |
endorsement | +1.149 |
provocation | +1.107 |
approval | +1.085 |
� | +1.052 |
Neg logit contributions | |
Bridgewater | -1.092 |
Unic | -1.054 |
Flavoring | -1.006 |
Reviewer | -0.989 |
wegian | -0.971 |
Token | discovered |
Token ID | 5071 |
Feature activation | +3.66e-01 |
Pos logit contributions | |
TIT | +0.515 |
endorsement | +0.498 |
provocation | +0.481 |
approval | +0.472 |
� | +0.458 |
Neg logit contributions | |
Bridgewater | -0.448 |
Unic | -0.431 |
Flavoring | -0.409 |
Reviewer | -0.399 |
wegian | -0.399 |
Token | boxing |
Token ID | 21576 |
Feature activation | +1.29e+00 |
Pos logit contributions | |
TIT | +0.290 |
endorsement | +0.284 |
provocation | +0.274 |
approval | +0.269 |
� | +0.262 |
Neg logit contributions | |
Bridgewater | -0.214 |
Unic | -0.207 |
Flavoring | -0.196 |
wegian | -0.191 |
Reviewer | -0.190 |
Token | of |
Token ID | 286 |
Feature activation | +6.98e-01 |
Pos logit contributions | |
TIT | +0.255 |
endorsement | +0.247 |
provocation | +0.239 |
approval | +0.235 |
� | +0.231 |
Neg logit contributions | |
Bridgewater | -0.163 |
Unic | -0.156 |
Flavoring | -0.145 |
wegian | -0.141 |
Reviewer | -0.140 |
Token | coward |
Token ID | 26769 |
Feature activation | +1.38e+00 |
Pos logit contributions | |
TIT | +0.493 |
endorsement | +0.477 |
provocation | +0.459 |
approval | +0.449 |
� | +0.437 |
Neg logit contributions | |
Bridgewater | -0.474 |
Unic | -0.459 |
Flavoring | -0.435 |
wegian | -0.424 |
Reviewer | -0.423 |
Token | ice |
Token ID | 501 |
Feature activation | +1.03e+00 |
Pos logit contributions | |
TIT | +1.228 |
endorsement | +1.203 |
provocation | +1.172 |
approval | +1.149 |
� | +1.125 |
Neg logit contributions | |
Bridgewater | -0.682 |
Unic | -0.647 |
Flavoring | -0.613 |
wegian | -0.580 |
Reviewer | -0.579 |
Token | on |
Token ID | 319 |
Feature activation | +2.44e-01 |
Pos logit contributions | |
TIT | +0.818 |
endorsement | +0.793 |
provocation | +0.768 |
approval | +0.754 |
� | +0.736 |
Neg logit contributions | |
Bridgewater | -0.601 |
Unic | -0.576 |
Flavoring | -0.543 |
Reviewer | -0.528 |
wegian | -0.526 |
Token | social |
Token ID | 1919 |
Feature activation | +1.61e+00 |
Pos logit contributions | |
TIT | +0.186 |
endorsement | +0.180 |
provocation | +0.173 |
approval | +0.170 |
� | +0.166 |
Neg logit contributions | |
Bridgewater | -0.152 |
Unic | -0.147 |
Flavoring | -0.137 |
wegian | -0.134 |
Reviewer | -0.134 |
Token | media |
Token ID | 2056 |
Feature activation | +3.46e+00 |
Pos logit contributions | |
TIT | +1.233 |
endorsement | +1.206 |
provocation | +1.168 |
approval | +1.142 |
� | +1.108 |
Neg logit contributions | |
Bridgewater | -0.998 |
Unic | -0.961 |
Flavoring | -0.911 |
wegian | -0.891 |
vich | -0.870 |
Token | . |
Token ID | 13 |
Feature activation | +7.84e-02 |
Pos logit contributions | |
TIT | +2.529 |
endorsement | +2.486 |
provocation | +2.401 |
approval | +2.356 |
� | +2.274 |
Neg logit contributions | |
Bridgewater | -2.207 |
Unic | -2.133 |
Flavoring | -2.051 |
wegian | -1.991 |
Reviewer | -1.982 |
Token |
|
Token ID | 198 |
Feature activation | -1.93e-01 |
Pos logit contributions | |
TIT | +0.060 |
endorsement | +0.058 |
provocation | +0.056 |
approval | +0.055 |
� | +0.054 |
Neg logit contributions | |
Bridgewater | -0.048 |
Unic | -0.047 |
Flavoring | -0.044 |
wegian | -0.043 |
Reviewer | -0.043 |
Token |
|
Token ID | 198 |
Feature activation | +3.86e-02 |
Pos logit contributions | |
Bridgewater | +0.109 |
Unic | +0.105 |
Flavoring | +0.096 |
wegian | +0.094 |
Reviewer | +0.094 |
Neg logit contributions | |
TIT | -0.159 |
endorsement | -0.153 |
provocation | -0.149 |
approval | -0.145 |
� | -0.143 |
Token | But |
Token ID | 1537 |
Feature activation | +3.41e-02 |
Pos logit contributions | |
TIT | +0.028 |
endorsement | +0.026 |
provocation | +0.025 |
approval | +0.025 |
� | +0.025 |
Neg logit contributions | |
Bridgewater | -0.026 |
Unic | -0.025 |
Flavoring | -0.024 |
wegian | -0.023 |
Reviewer | -0.023 |
Token | what |
Token ID | 644 |
Feature activation | +1.41e-01 |
Pos logit contributions | |
TIT | +0.028 |
endorsement | +0.027 |
provocation | +0.026 |
approval | +0.026 |
� | +0.025 |
Neg logit contributions | |
Bridgewater | -0.019 |
Unic | -0.018 |
Flavoring | -0.017 |
wegian | -0.017 |
Reviewer | -0.017 |
Token | will |
Token ID | 481 |
Feature activation | -2.46e-02 |
Pos logit contributions | |
TIT | +0.157 |
endorsement | +0.152 |
provocation | +0.147 |
approval | +0.145 |
� | +0.141 |
Neg logit contributions | |
Bridgewater | -0.124 |
Unic | -0.119 |
Flavoring | -0.111 |
Reviewer | -0.109 |
wegian | -0.108 |
Token | eat |
Token ID | 4483 |
Feature activation | +5.26e-01 |
Pos logit contributions | |
Bridgewater | +0.016 |
Unic | +0.015 |
Flavoring | +0.014 |
wegian | +0.014 |
Reviewer | +0.014 |
Neg logit contributions | |
TIT | -0.018 |
endorsement | -0.018 |
provocation | -0.017 |
approval | -0.017 |
� | -0.016 |
Token | whatever |
Token ID | 4232 |
Feature activation | -3.62e-01 |
Pos logit contributions | |
TIT | +0.436 |
endorsement | +0.424 |
provocation | +0.413 |
approval | +0.408 |
� | +0.395 |
Neg logit contributions | |
Bridgewater | -0.292 |
Unic | -0.277 |
Flavoring | -0.260 |
Reviewer | -0.255 |
wegian | -0.254 |
Token | has |
Token ID | 468 |
Feature activation | +9.06e-01 |
Pos logit contributions | |
Bridgewater | +0.205 |
Unic | +0.196 |
Flavoring | +0.182 |
wegian | +0.178 |
Reviewer | +0.178 |
Neg logit contributions | |
TIT | -0.298 |
endorsement | -0.288 |
provocation | -0.279 |
approval | -0.274 |
� | -0.268 |
Token | been |
Token ID | 587 |
Feature activation | +1.23e+00 |
Pos logit contributions | |
TIT | +0.614 |
endorsement | +0.591 |
provocation | +0.570 |
approval | +0.560 |
� | +0.540 |
Neg logit contributions | |
Bridgewater | -0.643 |
Unic | -0.620 |
Flavoring | -0.589 |
Reviewer | -0.579 |
wegian | -0.577 |
Token | prepared |
Token ID | 5597 |
Feature activation | +3.46e+00 |
Pos logit contributions | |
TIT | +0.870 |
endorsement | +0.835 |
provocation | +0.808 |
approval | +0.794 |
� | +0.768 |
Neg logit contributions | |
Bridgewater | -0.839 |
Unic | -0.806 |
Flavoring | -0.764 |
Reviewer | -0.749 |
wegian | -0.748 |
Token | for |
Token ID | 329 |
Feature activation | -2.93e-01 |
Pos logit contributions | |
TIT | +2.725 |
endorsement | +2.679 |
provocation | +2.621 |
approval | +2.590 |
� | +2.453 |
Neg logit contributions | |
Bridgewater | -1.992 |
Unic | -1.912 |
Reviewer | -1.820 |
Flavoring | -1.815 |
wegian | -1.757 |
Token | lunch |
Token ID | 9965 |
Feature activation | +5.20e-01 |
Pos logit contributions | |
Bridgewater | +0.174 |
Unic | +0.167 |
Flavoring | +0.155 |
wegian | +0.152 |
Reviewer | +0.151 |
Neg logit contributions | |
TIT | -0.233 |
endorsement | -0.225 |
provocation | -0.218 |
approval | -0.214 |
� | -0.209 |
Token | and |
Token ID | 290 |
Feature activation | -8.29e-02 |
Pos logit contributions | |
TIT | +0.407 |
endorsement | +0.393 |
provocation | +0.381 |
approval | +0.374 |
� | +0.364 |
Neg logit contributions | |
Bridgewater | -0.315 |
Unic | -0.303 |
Flavoring | -0.283 |
Reviewer | -0.276 |
wegian | -0.276 |
Token | that |
Token ID | 326 |
Feature activation | -2.83e-01 |
Pos logit contributions | |
Bridgewater | +0.048 |
Unic | +0.046 |
Flavoring | +0.043 |
Reviewer | +0.042 |
wegian | +0.042 |
Neg logit contributions | |
TIT | -0.067 |
endorsement | -0.065 |
provocation | -0.063 |
approval | -0.062 |
� | -0.060 |
Token | there |
Token ID | 612 |
Feature activation | +2.60e-01 |
Pos logit contributions | |
Bridgewater | +0.160 |
Unic | +0.153 |
Flavoring | +0.143 |
wegian | +0.140 |
Reviewer | +0.140 |
Neg logit contributions | |
TIT | -0.232 |
endorsement | -0.225 |
provocation | -0.218 |
approval | -0.215 |
� | -0.209 |
Token | // |
Token ID | 3373 |
Feature activation | -6.25e-01 |
Pos logit contributions | |
TIT | +0.514 |
endorsement | +0.496 |
provocation | +0.480 |
approval | +0.474 |
� | +0.458 |
Neg logit contributions | |
Bridgewater | -0.431 |
Unic | -0.414 |
Flavoring | -0.389 |
Reviewer | -0.383 |
wegian | -0.382 |
Token | 4 |
Token ID | 604 |
Feature activation | +1.75e+00 |
Pos logit contributions | |
Bridgewater | +0.371 |
Unic | +0.357 |
Flavoring | +0.331 |
wegian | +0.324 |
Reviewer | +0.322 |
Neg logit contributions | |
TIT | -0.498 |
endorsement | -0.479 |
provocation | -0.463 |
approval | -0.454 |
� | -0.446 |
Token | . |
Token ID | 13 |
Feature activation | +1.13e+00 |
Pos logit contributions | |
TIT | +1.237 |
endorsement | +1.185 |
provocation | +1.140 |
approval | +1.119 |
� | +1.093 |
Neg logit contributions | |
Bridgewater | -1.192 |
Unic | -1.149 |
Flavoring | -1.080 |
wegian | -1.063 |
Reviewer | -1.056 |
Token | Rec |
Token ID | 3311 |
Feature activation | +2.08e+00 |
Pos logit contributions | |
TIT | +0.860 |
endorsement | +0.826 |
provocation | +0.797 |
approval | +0.784 |
� | +0.766 |
Neg logit contributions | |
Bridgewater | -0.713 |
Unic | -0.684 |
Flavoring | -0.641 |
wegian | -0.631 |
Reviewer | -0.627 |
Token | urs |
Token ID | 1834 |
Feature activation | +1.35e+00 |
Pos logit contributions | |
TIT | +1.442 |
endorsement | +1.403 |
approval | +1.332 |
provocation | +1.332 |
� | +1.288 |
Neg logit contributions | |
Bridgewater | -1.428 |
Unic | -1.354 |
Flavoring | -1.301 |
Reviewer | -1.271 |
wegian | -1.264 |
Token | ively |
Token ID | 2280 |
Feature activation | +3.46e+00 |
Pos logit contributions | |
TIT | +0.840 |
endorsement | +0.802 |
provocation | +0.768 |
approval | +0.755 |
� | +0.727 |
Neg logit contributions | |
Bridgewater | -1.037 |
Unic | -0.997 |
Flavoring | -0.949 |
wegian | -0.928 |
Reviewer | -0.926 |
Token | apply |
Token ID | 4174 |
Feature activation | +1.46e+00 |
Pos logit contributions | |
TIT | +2.178 |
endorsement | +2.122 |
provocation | +2.045 |
approval | +2.019 |
� | +1.919 |
Neg logit contributions | |
Bridgewater | -2.533 |
Unic | -2.444 |
Flavoring | -2.332 |
Reviewer | -2.313 |
wegian | -2.313 |
Token | this |
Token ID | 428 |
Feature activation | +5.75e-01 |
Pos logit contributions | |
TIT | +1.124 |
endorsement | +1.095 |
provocation | +1.059 |
approval | +1.049 |
� | +1.008 |
Neg logit contributions | |
Bridgewater | -0.888 |
Unic | -0.854 |
Flavoring | -0.802 |
wegian | -0.788 |
Reviewer | -0.787 |
Token | algorithm |
Token ID | 11862 |
Feature activation | +2.19e+00 |
Pos logit contributions | |
TIT | +0.449 |
endorsement | +0.434 |
provocation | +0.419 |
approval | +0.413 |
� | +0.402 |
Neg logit contributions | |
Bridgewater | -0.348 |
Unic | -0.335 |
Flavoring | -0.313 |
wegian | -0.307 |
Reviewer | -0.304 |
Token | to |
Token ID | 284 |
Feature activation | +3.41e-01 |
Pos logit contributions | |
TIT | +1.760 |
endorsement | +1.717 |
provocation | +1.657 |
approval | +1.651 |
� | +1.585 |
Neg logit contributions | |
Bridgewater | -1.260 |
Unic | -1.203 |
Flavoring | -1.130 |
wegian | -1.102 |
Reviewer | -1.090 |
Token | all |
Token ID | 477 |
Feature activation | +8.02e-01 |
Pos logit contributions | |
TIT | +0.281 |
endorsement | +0.271 |
provocation | +0.262 |
approval | +0.259 |
� | +0.253 |
Neg logit contributions | |
Bridgewater | -0.192 |
Unic | -0.184 |
Flavoring | -0.171 |
wegian | -0.167 |
Reviewer | -0.167 |
Token | A |
Token ID | 32 |
Feature activation | -2.41e-01 |
Pos logit contributions | |
TIT | +0.282 |
endorsement | +0.270 |
provocation | +0.260 |
approval | +0.254 |
� | +0.249 |
Neg logit contributions | |
Bridgewater | -0.268 |
Unic | -0.259 |
Flavoring | -0.243 |
wegian | -0.238 |
Reviewer | -0.237 |
Token | major |
Token ID | 1688 |
Feature activation | +5.30e-02 |
Pos logit contributions | |
Bridgewater | +0.151 |
Unic | +0.145 |
Flavoring | +0.135 |
wegian | +0.133 |
Reviewer | +0.132 |
Neg logit contributions | |
TIT | -0.184 |
endorsement | -0.177 |
provocation | -0.170 |
approval | -0.167 |
� | -0.164 |
Token | city |
Token ID | 1748 |
Feature activation | +9.03e-01 |
Pos logit contributions | |
TIT | +0.034 |
endorsement | +0.032 |
provocation | +0.031 |
approval | +0.030 |
� | +0.030 |
Neg logit contributions | |
Bridgewater | -0.040 |
Unic | -0.038 |
Flavoring | -0.036 |
wegian | -0.035 |
Reviewer | -0.035 |
Token | like |
Token ID | 588 |
Feature activation | -2.14e-01 |
Pos logit contributions | |
TIT | +0.680 |
endorsement | +0.656 |
provocation | +0.633 |
approval | +0.621 |
� | +0.605 |
Neg logit contributions | |
Bridgewater | -0.573 |
Unic | -0.552 |
Flavoring | -0.520 |
wegian | -0.507 |
Reviewer | -0.503 |
Token | this |
Token ID | 428 |
Feature activation | +1.32e-01 |
Pos logit contributions | |
Bridgewater | +0.147 |
Unic | +0.143 |
Flavoring | +0.135 |
wegian | +0.132 |
Reviewer | +0.131 |
Neg logit contributions | |
TIT | -0.150 |
endorsement | -0.144 |
provocation | -0.138 |
approval | -0.136 |
� | -0.132 |
Token | deserves |
Token ID | 14071 |
Feature activation | +3.44e+00 |
Pos logit contributions | |
TIT | +0.104 |
endorsement | +0.100 |
provocation | +0.097 |
approval | +0.095 |
� | +0.093 |
Neg logit contributions | |
Bridgewater | -0.079 |
Unic | -0.076 |
Flavoring | -0.071 |
wegian | -0.070 |
Reviewer | -0.069 |
Token | a |
Token ID | 257 |
Feature activation | +5.96e-01 |
Pos logit contributions | |
TIT | +2.650 |
endorsement | +2.615 |
provocation | +2.513 |
approval | +2.504 |
Hait | +2.371 |
Neg logit contributions | |
Bridgewater | -2.075 |
Unic | -2.002 |
Flavoring | -1.946 |
wegian | -1.872 |
Reviewer | -1.840 |
Token | major |
Token ID | 1688 |
Feature activation | +8.90e-02 |
Pos logit contributions | |
TIT | +0.433 |
endorsement | +0.415 |
provocation | +0.400 |
approval | +0.392 |
� | +0.384 |
Neg logit contributions | |
Bridgewater | -0.396 |
Unic | -0.382 |
Flavoring | -0.357 |
wegian | -0.351 |
Reviewer | -0.349 |
Token | art |
Token ID | 1242 |
Feature activation | +8.53e-01 |
Pos logit contributions | |
TIT | +0.059 |
endorsement | +0.056 |
provocation | +0.054 |
approval | +0.052 |
� | +0.052 |
Neg logit contributions | |
Bridgewater | -0.065 |
Unic | -0.063 |
Flavoring | -0.058 |
wegian | -0.058 |
Reviewer | -0.058 |
Token | museum |
Token ID | 13257 |
Feature activation | -3.12e-01 |
Pos logit contributions | |
TIT | +0.513 |
endorsement | +0.489 |
provocation | +0.468 |
approval | +0.456 |
� | +0.442 |
Neg logit contributions | |
Bridgewater | -0.672 |
Unic | -0.653 |
Flavoring | -0.620 |
wegian | -0.608 |
Reviewer | -0.606 |
Token | ." |
Token ID | 526 |
Feature activation | +2.97e-02 |
Pos logit contributions | |
Bridgewater | +0.191 |
Unic | +0.184 |
Flavoring | +0.169 |
wegian | +0.166 |
Reviewer | +0.166 |
Neg logit contributions | |
TIT | -0.243 |
endorsement | -0.232 |
provocation | -0.225 |
approval | -0.220 |
� | -0.218 |
Token | down |
Token ID | 866 |
Feature activation | +4.73e-01 |
Pos logit contributions | |
TIT | +0.994 |
endorsement | +0.963 |
provocation | +0.931 |
approval | +0.923 |
� | +0.901 |
Neg logit contributions | |
Bridgewater | -0.665 |
Unic | -0.633 |
Flavoring | -0.601 |
Reviewer | -0.580 |
wegian | -0.577 |
Token | next |
Token ID | 1306 |
Feature activation | +1.01e+00 |
Pos logit contributions | |
TIT | +0.409 |
endorsement | +0.396 |
provocation | +0.384 |
approval | +0.380 |
� | +0.373 |
Neg logit contributions | |
Bridgewater | -0.248 |
Unic | -0.235 |
Flavoring | -0.220 |
Reviewer | -0.213 |
wegian | -0.210 |
Token | to |
Token ID | 284 |
Feature activation | +1.73e-01 |
Pos logit contributions | |
TIT | +0.854 |
endorsement | +0.824 |
provocation | +0.798 |
approval | +0.786 |
� | +0.770 |
Neg logit contributions | |
Bridgewater | -0.555 |
Unic | -0.531 |
Flavoring | -0.492 |
wegian | -0.479 |
Reviewer | -0.477 |
Token | a |
Token ID | 257 |
Feature activation | +4.60e-01 |
Pos logit contributions | |
TIT | +0.137 |
endorsement | +0.132 |
provocation | +0.128 |
approval | +0.126 |
� | +0.123 |
Neg logit contributions | |
Bridgewater | -0.103 |
Unic | -0.098 |
Flavoring | -0.092 |
Reviewer | -0.090 |
wegian | -0.090 |
Token | fellow |
Token ID | 5891 |
Feature activation | +2.62e-01 |
Pos logit contributions | |
TIT | +0.345 |
endorsement | +0.331 |
provocation | +0.320 |
approval | +0.314 |
� | +0.307 |
Neg logit contributions | |
Bridgewater | -0.294 |
Unic | -0.283 |
Flavoring | -0.266 |
wegian | -0.259 |
Reviewer | -0.259 |
Token | translator |
Token ID | 33417 |
Feature activation | +3.43e+00 |
Pos logit contributions | |
TIT | +0.193 |
endorsement | +0.185 |
provocation | +0.179 |
approval | +0.175 |
� | +0.171 |
Neg logit contributions | |
Bridgewater | -0.171 |
Unic | -0.164 |
Flavoring | -0.155 |
wegian | -0.151 |
Reviewer | -0.151 |
Token | , |
Token ID | 11 |
Feature activation | +5.75e-01 |
Pos logit contributions | |
TIT | +2.611 |
endorsement | +2.528 |
provocation | +2.447 |
approval | +2.416 |
� | +2.357 |
Neg logit contributions | |
Bridgewater | -2.098 |
Unic | -1.979 |
Flavoring | -1.920 |
Reviewer | -1.862 |
wegian | -1.823 |
Token | Tracy |
Token ID | 31140 |
Feature activation | +1.26e+00 |
Pos logit contributions | |
TIT | +0.433 |
endorsement | +0.416 |
provocation | +0.403 |
approval | +0.397 |
� | +0.387 |
Neg logit contributions | |
Bridgewater | -0.365 |
Unic | -0.350 |
Flavoring | -0.330 |
Reviewer | -0.323 |
wegian | -0.321 |
Token | , |
Token ID | 11 |
Feature activation | +7.28e-01 |
Pos logit contributions | |
TIT | +0.938 |
endorsement | +0.903 |
provocation | +0.873 |
approval | +0.857 |
� | +0.841 |
Neg logit contributions | |
Bridgewater | -0.799 |
Unic | -0.768 |
Flavoring | -0.726 |
wegian | -0.710 |
Reviewer | -0.709 |
Token | and |
Token ID | 290 |
Feature activation | +3.73e-01 |
Pos logit contributions | |
TIT | +0.581 |
endorsement | +0.561 |
provocation | +0.543 |
approval | +0.536 |
� | +0.524 |
Neg logit contributions | |
Bridgewater | -0.429 |
Unic | -0.410 |
Flavoring | -0.386 |
Reviewer | -0.376 |
wegian | -0.371 |
Token | began |
Token ID | 2540 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
TIT | +0.322 |
endorsement | +0.312 |
provocation | +0.303 |
approval | +0.299 |
� | +0.292 |
Neg logit contributions | |
Bridgewater | -0.195 |
Unic | -0.185 |
Flavoring | -0.174 |
Reviewer | -0.169 |
wegian | -0.168 |
Token | ) |
Token ID | 1267 |
Feature activation | -2.39e-02 |
Pos logit contributions | |
TIT | +0.699 |
endorsement | +0.673 |
provocation | +0.647 |
approval | +0.641 |
� | +0.623 |
Neg logit contributions | |
Bridgewater | -0.487 |
Unic | -0.468 |
Flavoring | -0.436 |
Reviewer | -0.423 |
wegian | -0.422 |
Token | AS |
Token ID | 7054 |
Feature activation | +5.88e-01 |
Pos logit contributions | |
Bridgewater | +0.015 |
Unic | +0.014 |
Flavoring | +0.014 |
Reviewer | +0.013 |
wegian | +0.013 |
Neg logit contributions | |
TIT | -0.018 |
endorsement | -0.018 |
provocation | -0.017 |
approval | -0.017 |
� | -0.016 |
Token | version |
Token ID | 2196 |
Feature activation | +2.08e+00 |
Pos logit contributions | |
TIT | +0.424 |
endorsement | +0.407 |
provocation | +0.392 |
approval | +0.384 |
� | +0.375 |
Neg logit contributions | |
Bridgewater | -0.393 |
Unic | -0.378 |
Flavoring | -0.356 |
wegian | -0.348 |
Reviewer | -0.347 |
Token | , |
Token ID | 837 |
Feature activation | +4.85e-01 |
Pos logit contributions | |
TIT | +1.507 |
endorsement | +1.421 |
provocation | +1.353 |
approval | +1.346 |
� | +1.295 |
Neg logit contributions | |
Bridgewater | -1.406 |
Unic | -1.352 |
Flavoring | -1.277 |
Reviewer | -1.249 |
wegian | -1.247 |
Token | C |
Token ID | 327 |
Feature activation | +6.11e-01 |
Pos logit contributions | |
TIT | +0.331 |
endorsement | +0.316 |
provocation | +0.303 |
approval | +0.297 |
� | +0.289 |
Neg logit contributions | |
Bridgewater | -0.344 |
Unic | -0.332 |
Flavoring | -0.313 |
wegian | -0.308 |
Reviewer | -0.307 |
Token | OUNT |
Token ID | 28270 |
Feature activation | +3.42e+00 |
Pos logit contributions | |
TIT | +0.379 |
endorsement | +0.360 |
provocation | +0.343 |
approval | +0.336 |
� | +0.327 |
Neg logit contributions | |
Bridgewater | -0.472 |
Unic | -0.456 |
Flavoring | -0.433 |
wegian | -0.425 |
Reviewer | -0.424 |
Token | ( |
Token ID | 357 |
Feature activation | +1.49e+00 |
Pos logit contributions | |
TIT | +2.710 |
endorsement | +2.591 |
provocation | +2.474 |
approval | +2.445 |
� | +2.390 |
Neg logit contributions | |
Bridgewater | -2.062 |
Unic | -1.960 |
Flavoring | -1.846 |
wegian | -1.826 |
Reviewer | -1.806 |
Token | * |
Token ID | 1635 |
Feature activation | -5.57e-03 |
Pos logit contributions | |
TIT | +1.084 |
endorsement | +1.046 |
provocation | +1.000 |
approval | +0.988 |
� | +0.956 |
Neg logit contributions | |
Bridgewater | -0.992 |
Unic | -0.952 |
Flavoring | -0.900 |
Reviewer | -0.881 |
wegian | -0.879 |
Token | ) |
Token ID | 1267 |
Feature activation | +1.19e-01 |
Pos logit contributions | |
Bridgewater | +0.003 |
Unic | +0.003 |
Flavoring | +0.003 |
wegian | +0.003 |
Reviewer | +0.003 |
Neg logit contributions | |
TIT | -0.004 |
endorsement | -0.004 |
provocation | -0.004 |
approval | -0.004 |
� | -0.004 |
Token | AS |
Token ID | 7054 |
Feature activation | +3.41e-01 |
Pos logit contributions | |
TIT | +0.091 |
endorsement | +0.087 |
provocation | +0.084 |
approval | +0.082 |
� | +0.080 |
Neg logit contributions | |
Bridgewater | -0.075 |
Unic | -0.072 |
Flavoring | -0.068 |
Reviewer | -0.066 |
wegian | -0.066 |
Token | count |
Token ID | 954 |
Feature activation | +1.94e+00 |
Pos logit contributions | |
TIT | +0.248 |
endorsement | +0.238 |
provocation | +0.229 |
approval | +0.225 |
� | +0.220 |
Neg logit contributions | |
Bridgewater | -0.226 |
Unic | -0.218 |
Flavoring | -0.205 |
wegian | -0.201 |
Reviewer | -0.200 |
Token | . |
Token ID | 13 |
Feature activation | -1.96e-01 |
Pos logit contributions | |
Bridgewater | +0.208 |
Unic | +0.200 |
Flavoring | +0.188 |
wegian | +0.184 |
Reviewer | +0.184 |
Neg logit contributions | |
TIT | -0.230 |
endorsement | -0.221 |
provocation | -0.213 |
approval | -0.209 |
� | -0.204 |
Token | Each |
Token ID | 5501 |
Feature activation | +6.07e-01 |
Pos logit contributions | |
Bridgewater | +0.119 |
Unic | +0.114 |
Flavoring | +0.106 |
wegian | +0.105 |
Reviewer | +0.104 |
Neg logit contributions | |
TIT | -0.153 |
endorsement | -0.146 |
provocation | -0.142 |
approval | -0.140 |
� | -0.137 |
Token | requires |
Token ID | 4433 |
Feature activation | +3.69e-01 |
Pos logit contributions | |
TIT | +0.430 |
endorsement | +0.414 |
provocation | +0.398 |
approval | +0.391 |
� | +0.380 |
Neg logit contributions | |
Bridgewater | -0.414 |
Unic | -0.398 |
Flavoring | -0.375 |
wegian | -0.369 |
Reviewer | -0.366 |
Token | multiple |
Token ID | 3294 |
Feature activation | -5.31e-01 |
Pos logit contributions | |
TIT | +0.296 |
endorsement | +0.288 |
provocation | +0.278 |
approval | +0.275 |
� | +0.266 |
Neg logit contributions | |
Bridgewater | -0.215 |
Unic | -0.206 |
Flavoring | -0.194 |
wegian | -0.189 |
Reviewer | -0.188 |
Token | c |
Token ID | 269 |
Feature activation | +1.85e-01 |
Pos logit contributions | |
Bridgewater | +0.326 |
Unic | +0.313 |
Flavoring | +0.292 |
wegian | +0.285 |
Reviewer | +0.284 |
Neg logit contributions | |
TIT | -0.413 |
endorsement | -0.396 |
provocation | -0.383 |
approval | -0.376 |
� | -0.369 |
Token | ams |
Token ID | 4105 |
Feature activation | +3.42e+00 |
Pos logit contributions | |
TIT | +0.139 |
endorsement | +0.133 |
provocation | +0.128 |
approval | +0.127 |
� | +0.124 |
Neg logit contributions | |
Bridgewater | -0.119 |
Unic | -0.114 |
Flavoring | -0.107 |
wegian | -0.104 |
Reviewer | -0.104 |
Token | , |
Token ID | 11 |
Feature activation | +1.75e-01 |
Pos logit contributions | |
TIT | +2.565 |
endorsement | +2.518 |
approval | +2.414 |
provocation | +2.394 |
� | +2.329 |
Neg logit contributions | |
Bridgewater | -2.113 |
Unic | -2.007 |
Flavoring | -1.946 |
wegian | -1.873 |
Reviewer | -1.857 |
Token | and |
Token ID | 290 |
Feature activation | +3.41e-01 |
Pos logit contributions | |
TIT | +0.138 |
endorsement | +0.134 |
provocation | +0.129 |
approval | +0.128 |
� | +0.123 |
Neg logit contributions | |
Bridgewater | -0.105 |
Unic | -0.100 |
Flavoring | -0.095 |
wegian | -0.093 |
Reviewer | -0.092 |
Token | trouble |
Token ID | 5876 |
Feature activation | -2.59e-01 |
Pos logit contributions | |
TIT | +0.273 |
endorsement | +0.267 |
provocation | +0.258 |
approval | +0.256 |
Hait | +0.245 |
Neg logit contributions | |
Bridgewater | -0.196 |
Unic | -0.188 |
Flavoring | -0.178 |
wegian | -0.175 |
Reviewer | -0.172 |
Token | can |
Token ID | 460 |
Feature activation | +2.83e-01 |
Pos logit contributions | |
Bridgewater | +0.160 |
Unic | +0.154 |
Flavoring | +0.143 |
wegian | +0.141 |
Reviewer | +0.139 |
Neg logit contributions | |
TIT | -0.198 |
endorsement | -0.193 |
provocation | -0.186 |
approval | -0.185 |
� | -0.178 |
Token | arise |
Token ID | 15058 |
Feature activation | +7.26e-01 |
Pos logit contributions | |
TIT | +0.192 |
endorsement | +0.185 |
provocation | +0.178 |
approval | +0.175 |
� | +0.170 |
Neg logit contributions | |
Bridgewater | -0.200 |
Unic | -0.193 |
Flavoring | -0.182 |
Reviewer | -0.179 |
wegian | -0.178 |
Token | . |
Token ID | 13 |
Feature activation | +3.52e-01 |
Pos logit contributions | |
Bridgewater | +0.010 |
Unic | +0.010 |
Flavoring | +0.009 |
wegian | +0.009 |
Reviewer | +0.009 |
Neg logit contributions | |
TIT | -0.011 |
endorsement | -0.010 |
provocation | -0.010 |
approval | -0.010 |
� | -0.010 |
Token |
|
Token ID | 198 |
Feature activation | +1.76e-01 |
Pos logit contributions | |
TIT | +0.266 |
endorsement | +0.254 |
provocation | +0.245 |
approval | +0.241 |
� | +0.236 |
Neg logit contributions | |
Bridgewater | -0.221 |
Unic | -0.213 |
Flavoring | -0.200 |
wegian | -0.198 |
Reviewer | -0.196 |
Token |
|
Token ID | 198 |
Feature activation | +2.68e-01 |
Pos logit contributions | |
TIT | +0.144 |
endorsement | +0.140 |
provocation | +0.135 |
approval | +0.132 |
� | +0.130 |
Neg logit contributions | |
Bridgewater | -0.101 |
Unic | -0.096 |
Flavoring | -0.089 |
wegian | -0.087 |
Reviewer | -0.086 |
Token | How |
Token ID | 2437 |
Feature activation | -9.21e-02 |
Pos logit contributions | |
TIT | +0.192 |
endorsement | +0.182 |
provocation | +0.175 |
approval | +0.172 |
� | +0.169 |
Neg logit contributions | |
Bridgewater | -0.182 |
Unic | -0.176 |
Flavoring | -0.165 |
wegian | -0.162 |
Reviewer | -0.160 |
Token | Aust |
Token ID | 2517 |
Feature activation | +7.76e-01 |
Pos logit contributions | |
Bridgewater | +0.056 |
Unic | +0.054 |
Flavoring | +0.050 |
wegian | +0.049 |
Reviewer | +0.049 |
Neg logit contributions | |
TIT | -0.072 |
endorsement | -0.069 |
provocation | -0.067 |
approval | -0.065 |
� | -0.064 |
Token | erity |
Token ID | 32821 |
Feature activation | +3.41e+00 |
Pos logit contributions | |
TIT | +0.174 |
endorsement | +0.152 |
provocation | +0.133 |
approval | +0.121 |
Hait | +0.106 |
Neg logit contributions | |
Bridgewater | -0.900 |
Unic | -0.887 |
Flavoring | -0.856 |
wegian | -0.846 |
Reviewer | -0.844 |
Token | Ab |
Token ID | 2275 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
TIT | +2.371 |
endorsement | +2.304 |
provocation | +2.244 |
approval | +2.193 |
Hait | +2.131 |
Neg logit contributions | |
Bridgewater | -2.261 |
Unic | -2.189 |
Flavoring | -2.078 |
Reviewer | -2.074 |
wegian | -2.048 |
Token | orted |
Token ID | 9741 |
Feature activation | +1.49e+00 |
Pos logit contributions | |
TIT | +0.073 |
endorsement | +0.069 |
provocation | +0.067 |
approval | +0.065 |
� | +0.064 |
Neg logit contributions | |
Bridgewater | -0.076 |
Unic | -0.073 |
Flavoring | -0.070 |
wegian | -0.068 |
Reviewer | -0.068 |
Token | Recovery |
Token ID | 21007 |
Feature activation | +1.65e+00 |
Pos logit contributions | |
TIT | +1.116 |
endorsement | +1.072 |
provocation | +1.037 |
approval | +1.021 |
� | +0.990 |
Neg logit contributions | |
Bridgewater | -0.938 |
Unic | -0.902 |
Flavoring | -0.850 |
wegian | -0.833 |
Reviewer | -0.831 |
Token |
|
Token ID | 198 |
Feature activation | +2.68e-02 |
Pos logit contributions | |
TIT | +1.252 |
endorsement | +1.210 |
provocation | +1.171 |
approval | +1.155 |
� | +1.123 |
Neg logit contributions | |
Bridgewater | -1.013 |
Unic | -0.975 |
Flavoring | -0.915 |
wegian | -0.899 |
Reviewer | -0.894 |
Token |
|
Token ID | 198 |
Feature activation | -1.94e-03 |
Pos logit contributions | |
TIT | +0.022 |
endorsement | +0.021 |
provocation | +0.021 |
approval | +0.020 |
� | +0.020 |
Neg logit contributions | |
Bridgewater | -0.015 |
Unic | -0.014 |
Flavoring | -0.013 |
wegian | -0.013 |
Reviewer | -0.013 |
Token | of |
Token ID | 286 |
Feature activation | +2.49e-02 |
Pos logit contributions | |
TIT | +0.603 |
endorsement | +0.578 |
provocation | +0.562 |
approval | +0.551 |
� | +0.545 |
Neg logit contributions | |
Bridgewater | -0.370 |
Unic | -0.353 |
Flavoring | -0.324 |
wegian | -0.315 |
Reviewer | -0.312 |
Token | Clinton |
Token ID | 2605 |
Feature activation | +3.20e-01 |
Pos logit contributions | |
TIT | +0.017 |
endorsement | +0.017 |
provocation | +0.016 |
approval | +0.016 |
� | +0.015 |
Neg logit contributions | |
Bridgewater | -0.017 |
Unic | -0.017 |
Flavoring | -0.016 |
wegian | -0.015 |
Reviewer | -0.015 |
Token | emerge |
Token ID | 14740 |
Feature activation | +6.09e-01 |
Pos logit contributions | |
TIT | +0.243 |
endorsement | +0.235 |
provocation | +0.226 |
approval | +0.223 |
� | +0.218 |
Neg logit contributions | |
Bridgewater | -0.201 |
Unic | -0.193 |
Flavoring | -0.181 |
wegian | -0.177 |
Reviewer | -0.177 |
Token | on |
Token ID | 319 |
Feature activation | -5.43e-02 |
Pos logit contributions | |
TIT | +0.482 |
endorsement | +0.467 |
provocation | +0.450 |
approval | +0.444 |
� | +0.435 |
Neg logit contributions | |
Bridgewater | -0.362 |
Unic | -0.348 |
Flavoring | -0.326 |
wegian | -0.318 |
Reviewer | -0.316 |
Token | social |
Token ID | 1919 |
Feature activation | +9.67e-01 |
Pos logit contributions | |
Bridgewater | +0.033 |
Unic | +0.032 |
Flavoring | +0.029 |
wegian | +0.029 |
Reviewer | +0.029 |
Neg logit contributions | |
TIT | -0.043 |
endorsement | -0.041 |
provocation | -0.040 |
approval | -0.039 |
� | -0.038 |
Token | media |
Token ID | 2056 |
Feature activation | +3.41e+00 |
Pos logit contributions | |
TIT | +0.746 |
endorsement | +0.730 |
provocation | +0.703 |
approval | +0.695 |
� | +0.672 |
Neg logit contributions | |
Bridgewater | -0.594 |
Unic | -0.572 |
Flavoring | -0.536 |
wegian | -0.529 |
Reviewer | -0.519 |
Token | — |
Token ID | 960 |
Feature activation | +9.37e-01 |
Pos logit contributions | |
TIT | +2.431 |
endorsement | +2.387 |
provocation | +2.292 |
approval | +2.261 |
� | +2.190 |
Neg logit contributions | |
Bridgewater | -2.249 |
Unic | -2.154 |
Flavoring | -2.072 |
wegian | -2.044 |
Reviewer | -2.011 |
Token | a |
Token ID | 64 |
Feature activation | -9.49e-01 |
Pos logit contributions | |
TIT | +0.719 |
endorsement | +0.687 |
provocation | +0.662 |
approval | +0.652 |
� | +0.642 |
Neg logit contributions | |
Bridgewater | -0.584 |
Unic | -0.563 |
Flavoring | -0.528 |
wegian | -0.513 |
Reviewer | -0.509 |
Token | woman |
Token ID | 2415 |
Feature activation | -4.32e-02 |
Pos logit contributions | |
Bridgewater | +0.666 |
Unic | +0.642 |
Flavoring | +0.600 |
wegian | +0.590 |
Reviewer | +0.586 |
Neg logit contributions | |
TIT | -0.658 |
endorsement | -0.623 |
provocation | -0.599 |
approval | -0.586 |
� | -0.577 |
Token | who |
Token ID | 508 |
Feature activation | +1.41e-01 |
Pos logit contributions | |
Bridgewater | +0.027 |
Unic | +0.026 |
Flavoring | +0.025 |
wegian | +0.024 |
Reviewer | +0.024 |
Neg logit contributions | |
TIT | -0.032 |
endorsement | -0.031 |
provocation | -0.030 |
approval | -0.030 |
� | -0.029 |
Token | asks |
Token ID | 7893 |
Feature activation | +5.22e-01 |
Pos logit contributions | |
TIT | +0.111 |
endorsement | +0.107 |
provocation | +0.103 |
approval | +0.102 |
� | +0.100 |
Neg logit contributions | |
Bridgewater | -0.084 |
Unic | -0.081 |
Flavoring | -0.076 |
wegian | -0.074 |
Reviewer | -0.074 |
Token | million |
Token ID | 1510 |
Feature activation | -4.44e-01 |
Pos logit contributions | |
Bridgewater | +0.337 |
Unic | +0.324 |
Flavoring | +0.302 |
wegian | +0.296 |
Reviewer | +0.293 |
Neg logit contributions | |
TIT | -0.398 |
endorsement | -0.382 |
provocation | -0.368 |
approval | -0.361 |
� | -0.354 |
Token | years |
Token ID | 812 |
Feature activation | -1.74e-01 |
Pos logit contributions | |
Bridgewater | +0.264 |
Unic | +0.253 |
Flavoring | +0.236 |
wegian | +0.231 |
Reviewer | +0.231 |
Neg logit contributions | |
TIT | -0.353 |
endorsement | -0.341 |
provocation | -0.330 |
approval | -0.325 |
� | -0.316 |
Token | . |
Token ID | 13 |
Feature activation | +2.67e-01 |
Pos logit contributions | |
Bridgewater | +0.112 |
Unic | +0.108 |
Flavoring | +0.101 |
wegian | +0.099 |
Reviewer | +0.099 |
Neg logit contributions | |
TIT | -0.130 |
endorsement | -0.125 |
provocation | -0.121 |
approval | -0.119 |
� | -0.116 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.13e-01 |
Pos logit contributions | |
TIT | +0.206 |
endorsement | +0.197 |
provocation | +0.190 |
approval | +0.188 |
� | +0.184 |
Neg logit contributions | |
Bridgewater | -0.163 |
Unic | -0.157 |
Flavoring | -0.147 |
wegian | -0.145 |
Reviewer | -0.144 |
Token | Image |
Token ID | 5159 |
Feature activation | +1.23e+00 |
Pos logit contributions | |
TIT | +0.153 |
endorsement | +0.145 |
provocation | +0.139 |
approval | +0.137 |
� | +0.135 |
Neg logit contributions | |
Bridgewater | -0.144 |
Unic | -0.139 |
Flavoring | -0.130 |
wegian | -0.128 |
Reviewer | -0.125 |
Token | copyright |
Token ID | 6634 |
Feature activation | +3.39e+00 |
Pos logit contributions | |
TIT | +0.782 |
endorsement | +0.755 |
provocation | +0.725 |
approval | +0.709 |
� | +0.682 |
Neg logit contributions | |
Bridgewater | -0.915 |
Unic | -0.887 |
Flavoring | -0.839 |
wegian | -0.830 |
Reviewer | -0.819 |
Token | AFP |
Token ID | 8916 |
Feature activation | -2.78e-01 |
Pos logit contributions | |
TIT | +2.329 |
endorsement | +2.236 |
provocation | +2.174 |
approval | +2.135 |
Hait | +2.052 |
Neg logit contributions | |
Bridgewater | -2.294 |
Unic | -2.226 |
Flavoring | -2.122 |
Reviewer | -2.120 |
wegian | -2.082 |
Token |
|
Token ID | 198 |
Feature activation | +1.61e+00 |
Pos logit contributions | |
Bridgewater | +0.163 |
Unic | +0.156 |
Flavoring | +0.145 |
Reviewer | +0.144 |
wegian | +0.143 |
Neg logit contributions | |
TIT | -0.223 |
endorsement | -0.215 |
provocation | -0.208 |
approval | -0.205 |
� | -0.200 |
Token |
|
Token ID | 198 |
Feature activation | +1.63e+00 |
Pos logit contributions | |
TIT | +1.275 |
endorsement | +1.223 |
provocation | +1.180 |
approval | +1.163 |
� | +1.141 |
Neg logit contributions | |
Bridgewater | -0.957 |
Unic | -0.919 |
Flavoring | -0.857 |
wegian | -0.841 |
Reviewer | -0.833 |
Token | The |
Token ID | 464 |
Feature activation | +1.01e+00 |
Pos logit contributions | |
TIT | +1.140 |
endorsement | +1.063 |
provocation | +1.014 |
approval | +1.004 |
� | +0.993 |
Neg logit contributions | |
Bridgewater | -1.149 |
Unic | -1.107 |
Flavoring | -1.047 |
wegian | -1.024 |
Reviewer | -1.005 |
Token | Bank |
Token ID | 5018 |
Feature activation | -5.71e-01 |
Pos logit contributions | |
TIT | +0.732 |
endorsement | +0.710 |
provocation | +0.682 |
approval | +0.674 |
� | +0.648 |
Neg logit contributions | |
Bridgewater | -0.668 |
Unic | -0.643 |
Flavoring | -0.608 |
Reviewer | -0.599 |
wegian | -0.592 |
Token | confirmed |
Token ID | 4999 |
Feature activation | -3.64e-01 |
Pos logit contributions | |
Bridgewater | +0.005 |
Unic | +0.005 |
Flavoring | +0.005 |
wegian | +0.004 |
Reviewer | +0.004 |
Neg logit contributions | |
TIT | -0.006 |
endorsement | -0.005 |
provocation | -0.005 |
approval | -0.005 |
� | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | -1.28e-01 |
Pos logit contributions | |
Bridgewater | +0.241 |
Unic | +0.232 |
Flavoring | +0.219 |
wegian | +0.214 |
Reviewer | +0.212 |
Neg logit contributions | |
TIT | -0.263 |
endorsement | -0.255 |
provocation | -0.245 |
approval | -0.242 |
� | -0.234 |
Token |
|
Token ID | 198 |
Feature activation | +3.25e-01 |
Pos logit contributions | |
Bridgewater | +0.078 |
Unic | +0.075 |
Flavoring | +0.071 |
wegian | +0.070 |
Reviewer | +0.069 |
Neg logit contributions | |
TIT | -0.099 |
endorsement | -0.094 |
provocation | -0.091 |
approval | -0.090 |
� | -0.088 |
Token |
|
Token ID | 198 |
Feature activation | +3.99e-01 |
Pos logit contributions | |
TIT | +0.265 |
endorsement | +0.256 |
provocation | +0.248 |
approval | +0.243 |
� | +0.238 |
Neg logit contributions | |
Bridgewater | -0.187 |
Unic | -0.180 |
Flavoring | -0.166 |
wegian | -0.162 |
Reviewer | -0.161 |
Token | FR |
Token ID | 10913 |
Feature activation | -5.03e-01 |
Pos logit contributions | |
TIT | +0.291 |
endorsement | +0.274 |
provocation | +0.263 |
approval | +0.259 |
� | +0.256 |
Neg logit contributions | |
Bridgewater | -0.268 |
Unic | -0.258 |
Flavoring | -0.243 |
wegian | -0.237 |
Reviewer | -0.233 |
Token | OM |
Token ID | 2662 |
Feature activation | +3.38e+00 |
Pos logit contributions | |
Bridgewater | +0.329 |
Unic | +0.316 |
Flavoring | +0.298 |
wegian | +0.293 |
Reviewer | +0.291 |
Neg logit contributions | |
TIT | -0.372 |
endorsement | -0.356 |
provocation | -0.342 |
approval | -0.337 |
� | -0.327 |
Token | THE |
Token ID | 3336 |
Feature activation | +3.84e-01 |
Pos logit contributions | |
TIT | +2.420 |
endorsement | +2.319 |
provocation | +2.207 |
approval | +2.198 |
� | +2.105 |
Neg logit contributions | |
Bridgewater | -2.236 |
Unic | -2.134 |
Flavoring | -2.067 |
Reviewer | -2.054 |
wegian | -2.036 |
Token | AR |
Token ID | 5923 |
Feature activation | +1.15e-01 |
Pos logit contributions | |
TIT | +0.275 |
endorsement | +0.263 |
provocation | +0.252 |
approval | +0.248 |
� | +0.241 |
Neg logit contributions | |
Bridgewater | -0.259 |
Unic | -0.249 |
Flavoring | -0.236 |
wegian | -0.233 |
Reviewer | -0.231 |
Token | CH |
Token ID | 3398 |
Feature activation | +7.57e-01 |
Pos logit contributions | |
TIT | +0.062 |
endorsement | +0.059 |
provocation | +0.056 |
approval | +0.055 |
� | +0.053 |
Neg logit contributions | |
Bridgewater | -0.097 |
Unic | -0.095 |
Flavoring | -0.090 |
wegian | -0.089 |
Reviewer | -0.088 |
Token | IVE |
Token ID | 9306 |
Feature activation | +1.98e+00 |
Pos logit contributions | |
TIT | +0.347 |
endorsement | +0.323 |
provocation | +0.304 |
approval | +0.294 |
� | +0.284 |
Neg logit contributions | |
Bridgewater | -0.707 |
Unic | -0.689 |
Flavoring | -0.659 |
wegian | -0.651 |
Reviewer | -0.648 |
Token | : |
Token ID | 25 |
Feature activation | +3.17e-01 |
Pos logit contributions | |
TIT | +1.510 |
endorsement | +1.450 |
provocation | +1.383 |
approval | +1.370 |
� | +1.337 |
Neg logit contributions | |
Bridgewater | -1.229 |
Unic | -1.181 |
Flavoring | -1.120 |
wegian | -1.105 |
Reviewer | -1.096 |
Token | const |
Token ID | 1500 |
Feature activation | +5.19e-01 |
Pos logit contributions | |
TIT | +0.657 |
endorsement | +0.624 |
provocation | +0.600 |
approval | +0.588 |
� | +0.575 |
Neg logit contributions | |
Bridgewater | -0.674 |
Unic | -0.650 |
Flavoring | -0.612 |
wegian | -0.603 |
Reviewer | -0.596 |
Token | size |
Token ID | 2546 |
Feature activation | +5.98e-01 |
Pos logit contributions | |
TIT | +0.380 |
endorsement | +0.364 |
provocation | +0.351 |
approval | +0.344 |
� | +0.337 |
Neg logit contributions | |
Bridgewater | -0.342 |
Unic | -0.329 |
Flavoring | -0.308 |
wegian | -0.303 |
Reviewer | -0.301 |
Token | _ |
Token ID | 62 |
Feature activation | -3.66e-01 |
Pos logit contributions | |
TIT | +0.469 |
endorsement | +0.450 |
provocation | +0.434 |
approval | +0.428 |
� | +0.418 |
Neg logit contributions | |
Bridgewater | -0.363 |
Unic | -0.349 |
Flavoring | -0.324 |
wegian | -0.320 |
Reviewer | -0.316 |
Token | t |
Token ID | 83 |
Feature activation | +5.67e-01 |
Pos logit contributions | |
Bridgewater | +0.178 |
Unic | +0.170 |
Flavoring | +0.156 |
wegian | +0.152 |
Reviewer | +0.150 |
Neg logit contributions | |
TIT | -0.331 |
endorsement | -0.319 |
provocation | -0.309 |
approval | -0.305 |
� | -0.300 |
Token | num |
Token ID | 997 |
Feature activation | +1.78e-01 |
Pos logit contributions | |
TIT | +0.405 |
endorsement | +0.388 |
provocation | +0.374 |
approval | +0.366 |
� | +0.358 |
Neg logit contributions | |
Bridgewater | -0.383 |
Unic | -0.370 |
Flavoring | -0.347 |
wegian | -0.341 |
Reviewer | -0.339 |
Token | Part |
Token ID | 7841 |
Feature activation | +3.38e+00 |
Pos logit contributions | |
TIT | +0.126 |
endorsement | +0.120 |
provocation | +0.115 |
approval | +0.113 |
� | +0.111 |
Neg logit contributions | |
Bridgewater | -0.123 |
Unic | -0.118 |
Flavoring | -0.111 |
wegian | -0.109 |
Reviewer | -0.108 |
Token | icles |
Token ID | 2983 |
Feature activation | -3.17e-01 |
Pos logit contributions | |
TIT | +2.678 |
endorsement | +2.579 |
provocation | +2.478 |
approval | +2.453 |
Hait | +2.391 |
Neg logit contributions | |
Bridgewater | -1.957 |
Unic | -1.890 |
Flavoring | -1.773 |
Reviewer | -1.714 |
wegian | -1.714 |
Token | = |
Token ID | 796 |
Feature activation | -1.83e+00 |
Pos logit contributions | |
Bridgewater | +0.174 |
Unic | +0.167 |
Flavoring | +0.154 |
wegian | +0.151 |
Reviewer | +0.149 |
Neg logit contributions | |
TIT | -0.267 |
endorsement | -0.257 |
provocation | -0.249 |
approval | -0.245 |
� | -0.241 |
Token | particles |
Token ID | 13166 |
Feature activation | -2.22e-01 |
Pos logit contributions | |
Bridgewater | +1.106 |
Unic | +1.066 |
Flavoring | +0.990 |
wegian | +0.966 |
Reviewer | +0.960 |
Neg logit contributions | |
TIT | -1.440 |
endorsement | -1.384 |
provocation | -1.338 |
approval | -1.313 |
� | -1.292 |
Token | . |
Token ID | 13 |
Feature activation | +1.51e+00 |
Pos logit contributions | |
Bridgewater | +0.151 |
Unic | +0.146 |
Flavoring | +0.137 |
wegian | +0.134 |
Reviewer | +0.134 |
Neg logit contributions | |
TIT | -0.157 |
endorsement | -0.150 |
provocation | -0.145 |
approval | -0.142 |
� | -0.139 |
Token | Size |
Token ID | 10699 |
Feature activation | +1.58e+00 |
Pos logit contributions | |
TIT | +1.027 |
endorsement | +0.971 |
provocation | +0.931 |
approval | +0.912 |
� | +0.895 |
Neg logit contributions | |
Bridgewater | -1.082 |
Unic | -1.038 |
Flavoring | -0.979 |
wegian | -0.971 |
Reviewer | -0.959 |
Token | Serve |
Token ID | 35557 |
Feature activation | +2.86e+00 |
Pos logit contributions | |
TIT | +0.339 |
endorsement | +0.327 |
provocation | +0.316 |
approval | +0.310 |
� | +0.304 |
Neg logit contributions | |
Bridgewater | -0.249 |
Unic | -0.240 |
Flavoring | -0.225 |
Reviewer | -0.216 |
wegian | -0.215 |
Token | hot |
Token ID | 3024 |
Feature activation | +1.46e+00 |
Pos logit contributions | |
TIT | +2.110 |
endorsement | +2.027 |
provocation | +1.967 |
approval | +1.952 |
� | +1.886 |
Neg logit contributions | |
Bridgewater | -1.871 |
Unic | -1.787 |
Flavoring | -1.680 |
wegian | -1.673 |
Reviewer | -1.649 |
Token | ! |
Token ID | 0 |
Feature activation | +1.94e-01 |
Pos logit contributions | |
TIT | +0.972 |
endorsement | +0.928 |
provocation | +0.886 |
approval | +0.872 |
� | +0.850 |
Neg logit contributions | |
Bridgewater | -1.054 |
Unic | -1.016 |
Flavoring | -0.961 |
wegian | -0.948 |
Reviewer | -0.940 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.26e-01 |
Pos logit contributions | |
TIT | +0.165 |
endorsement | +0.158 |
provocation | +0.154 |
approval | +0.151 |
� | +0.149 |
Neg logit contributions | |
Bridgewater | -0.105 |
Unic | -0.101 |
Flavoring | -0.092 |
wegian | -0.091 |
Reviewer | -0.091 |
Token | Image |
Token ID | 5159 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
TIT | +0.163 |
endorsement | +0.155 |
provocation | +0.148 |
approval | +0.146 |
� | +0.144 |
Neg logit contributions | |
Bridgewater | -0.153 |
Unic | -0.148 |
Flavoring | -0.139 |
wegian | -0.136 |
Reviewer | -0.134 |
Token | copyright |
Token ID | 6634 |
Feature activation | +3.37e+00 |
Pos logit contributions | |
TIT | +0.725 |
endorsement | +0.699 |
provocation | +0.670 |
approval | +0.657 |
� | +0.631 |
Neg logit contributions | |
Bridgewater | -0.849 |
Unic | -0.822 |
Flavoring | -0.778 |
wegian | -0.769 |
Reviewer | -0.759 |
Token | ESA |
Token ID | 42212 |
Feature activation | +1.20e+00 |
Pos logit contributions | |
TIT | +2.320 |
endorsement | +2.230 |
provocation | +2.167 |
approval | +2.130 |
Hait | +2.045 |
Neg logit contributions | |
Bridgewater | -2.265 |
Unic | -2.197 |
Flavoring | -2.095 |
Reviewer | -2.092 |
wegian | -2.053 |
Token | Image |
Token ID | 7412 |
Feature activation | +8.62e-01 |
Pos logit contributions | |
TIT | +0.950 |
endorsement | +0.919 |
provocation | +0.889 |
approval | +0.883 |
� | +0.851 |
Neg logit contributions | |
Bridgewater | -0.703 |
Unic | -0.671 |
Flavoring | -0.638 |
Reviewer | -0.631 |
wegian | -0.623 |
Token | caption |
Token ID | 8305 |
Feature activation | +8.62e-01 |
Pos logit contributions | |
TIT | +0.673 |
endorsement | +0.644 |
provocation | +0.624 |
approval | +0.611 |
� | +0.602 |
Neg logit contributions | |
Bridgewater | -0.526 |
Unic | -0.506 |
Flavoring | -0.471 |
wegian | -0.460 |
Reviewer | -0.457 |
Token | Go |
Token ID | 1514 |
Feature activation | +5.26e-01 |
Pos logit contributions | |
TIT | +0.644 |
endorsement | +0.615 |
provocation | +0.595 |
approval | +0.586 |
� | +0.573 |
Neg logit contributions | |
Bridgewater | -0.548 |
Unic | -0.526 |
Flavoring | -0.495 |
wegian | -0.489 |
Reviewer | -0.488 |
Token | ce |
Token ID | 344 |
Feature activation | +8.35e-01 |
Pos logit contributions | |
TIT | +0.398 |
endorsement | +0.382 |
provocation | +0.368 |
approval | +0.362 |
� | +0.354 |
Neg logit contributions | |
Bridgewater | -0.333 |
Unic | -0.320 |
Flavoring | -0.302 |
wegian | -0.292 |
Reviewer | -0.292 |
Token | - |
Token ID | 12 |
Feature activation | +8.63e-01 |
Pos logit contributions | |
TIT | +0.644 |
endorsement | +0.619 |
provocation | +0.598 |
approval | +0.589 |
� | +0.575 |
Neg logit contributions | |
Bridgewater | -0.518 |
Unic | -0.497 |
Flavoring | -0.469 |
wegian | -0.456 |
Reviewer | -0.453 |
Token | Right |
Token ID | 11028 |
Feature activation | +1.27e+00 |
Pos logit contributions | |
TIT | +0.457 |
endorsement | +0.427 |
provocation | +0.407 |
approval | +0.395 |
� | +0.385 |
Neg logit contributions | |
Bridgewater | -0.745 |
Unic | -0.724 |
Flavoring | -0.695 |
wegian | -0.680 |
Reviewer | -0.673 |
Token | Content |
Token ID | 14041 |
Feature activation | +1.06e+00 |
Pos logit contributions | |
TIT | +0.957 |
endorsement | +0.925 |
provocation | +0.893 |
approval | +0.875 |
� | +0.854 |
Neg logit contributions | |
Bridgewater | -0.798 |
Unic | -0.767 |
Flavoring | -0.723 |
wegian | -0.707 |
Reviewer | -0.704 |
Token | on |
Token ID | 319 |
Feature activation | +6.64e-01 |
Pos logit contributions | |
TIT | +0.874 |
endorsement | +0.844 |
provocation | +0.816 |
approval | +0.803 |
� | +0.786 |
Neg logit contributions | |
Bridgewater | -0.597 |
Unic | -0.572 |
Flavoring | -0.533 |
wegian | -0.521 |
Reviewer | -0.517 |
Token | Social |
Token ID | 5483 |
Feature activation | +1.70e+00 |
Pos logit contributions | |
TIT | +0.497 |
endorsement | +0.477 |
provocation | +0.460 |
approval | +0.451 |
� | +0.442 |
Neg logit contributions | |
Bridgewater | -0.427 |
Unic | -0.411 |
Flavoring | -0.384 |
wegian | -0.377 |
Reviewer | -0.374 |
Token | Media |
Token ID | 6343 |
Feature activation | +3.33e+00 |
Pos logit contributions | |
TIT | +1.409 |
endorsement | +1.367 |
provocation | +1.326 |
approval | +1.303 |
� | +1.276 |
Neg logit contributions | |
Bridgewater | -0.926 |
Unic | -0.876 |
wegian | -0.827 |
Flavoring | -0.824 |
Reviewer | -0.801 |
Token |
|
Token ID | 198 |
Feature activation | -4.86e-01 |
Pos logit contributions | |
TIT | +2.505 |
endorsement | +2.425 |
provocation | +2.331 |
approval | +2.291 |
� | +2.221 |
Neg logit contributions | |
Bridgewater | -2.071 |
Unic | -1.982 |
Flavoring | -1.873 |
wegian | -1.861 |
Reviewer | -1.846 |
Token |
|
Token ID | 198 |
Feature activation | -2.62e-01 |
Pos logit contributions | |
Bridgewater | +0.267 |
Unic | +0.255 |
Flavoring | +0.233 |
Reviewer | +0.226 |
wegian | +0.225 |
Neg logit contributions | |
TIT | -0.409 |
endorsement | -0.396 |
provocation | -0.386 |
approval | -0.376 |
� | -0.370 |
Token | One |
Token ID | 3198 |
Feature activation | +9.76e-03 |
Pos logit contributions | |
Bridgewater | +0.168 |
Unic | +0.162 |
Flavoring | +0.151 |
Reviewer | +0.149 |
wegian | +0.148 |
Neg logit contributions | |
TIT | -0.195 |
endorsement | -0.188 |
provocation | -0.182 |
approval | -0.178 |
� | -0.174 |
Token | of |
Token ID | 286 |
Feature activation | -9.82e-01 |
Pos logit contributions | |
TIT | +0.008 |
endorsement | +0.008 |
provocation | +0.007 |
approval | +0.007 |
� | +0.007 |
Neg logit contributions | |
Bridgewater | -0.006 |
Unic | -0.005 |
Flavoring | -0.005 |
wegian | -0.005 |
Reviewer | -0.005 |
Token | the |
Token ID | 262 |
Feature activation | -1.10e+00 |
Pos logit contributions | |
Bridgewater | +0.576 |
Unic | +0.552 |
Flavoring | +0.511 |
wegian | +0.499 |
Reviewer | +0.495 |
Neg logit contributions | |
TIT | -0.792 |
endorsement | -0.760 |
provocation | -0.736 |
approval | -0.722 |
� | -0.711 |
Token | . |
Token ID | 13 |
Feature activation | -5.33e-01 |
Pos logit contributions | |
TIT | +0.280 |
endorsement | +0.270 |
provocation | +0.259 |
approval | +0.255 |
� | +0.249 |
Neg logit contributions | |
Bridgewater | -0.267 |
Unic | -0.257 |
Flavoring | -0.243 |
Reviewer | -0.238 |
wegian | -0.238 |
Token |
|
Token ID | 198 |
Feature activation | -4.56e-01 |
Pos logit contributions | |
Bridgewater | +0.323 |
Unic | +0.311 |
Flavoring | +0.289 |
wegian | +0.283 |
Reviewer | +0.281 |
Neg logit contributions | |
TIT | -0.417 |
endorsement | -0.401 |
provocation | -0.388 |
approval | -0.381 |
� | -0.373 |
Token |
|
Token ID | 198 |
Feature activation | -1.96e-01 |
Pos logit contributions | |
Bridgewater | +0.252 |
Unic | +0.241 |
Flavoring | +0.220 |
Reviewer | +0.214 |
wegian | +0.214 |
Neg logit contributions | |
TIT | -0.382 |
endorsement | -0.370 |
provocation | -0.360 |
approval | -0.351 |
� | -0.345 |
Token | On |
Token ID | 2202 |
Feature activation | -5.76e-02 |
Pos logit contributions | |
Bridgewater | +0.126 |
Unic | +0.121 |
Flavoring | +0.113 |
wegian | +0.111 |
Reviewer | +0.111 |
Neg logit contributions | |
TIT | -0.147 |
endorsement | -0.141 |
provocation | -0.136 |
approval | -0.134 |
� | -0.131 |
Token | social |
Token ID | 1919 |
Feature activation | +1.24e+00 |
Pos logit contributions | |
Bridgewater | +0.034 |
Unic | +0.033 |
Flavoring | +0.030 |
wegian | +0.030 |
Reviewer | +0.030 |
Neg logit contributions | |
TIT | -0.046 |
endorsement | -0.044 |
provocation | -0.043 |
approval | -0.042 |
� | -0.041 |
Token | media |
Token ID | 2056 |
Feature activation | +3.33e+00 |
Pos logit contributions | |
TIT | +0.927 |
endorsement | +0.920 |
provocation | +0.888 |
approval | +0.872 |
� | +0.842 |
Neg logit contributions | |
Bridgewater | -0.775 |
Unic | -0.747 |
Flavoring | -0.705 |
wegian | -0.687 |
Reviewer | -0.675 |
Token | , |
Token ID | 11 |
Feature activation | +4.19e-01 |
Pos logit contributions | |
TIT | +2.440 |
endorsement | +2.417 |
provocation | +2.322 |
approval | +2.296 |
� | +2.195 |
Neg logit contributions | |
Bridgewater | -2.092 |
Unic | -2.014 |
Flavoring | -1.969 |
wegian | -1.906 |
Reviewer | -1.881 |
Token | Michael |
Token ID | 3899 |
Feature activation | -8.34e-02 |
Pos logit contributions | |
TIT | +0.333 |
endorsement | +0.323 |
provocation | +0.313 |
approval | +0.308 |
� | +0.299 |
Neg logit contributions | |
Bridgewater | -0.246 |
Unic | -0.236 |
Flavoring | -0.223 |
wegian | -0.216 |
Reviewer | -0.216 |
Token | G |
Token ID | 402 |
Feature activation | +7.71e-01 |
Pos logit contributions | |
Bridgewater | +0.044 |
Unic | +0.042 |
Flavoring | +0.039 |
Reviewer | +0.038 |
wegian | +0.038 |
Neg logit contributions | |
TIT | -0.071 |
endorsement | -0.069 |
provocation | -0.067 |
approval | -0.066 |
� | -0.064 |
Token | . |
Token ID | 13 |
Feature activation | -4.12e-01 |
Pos logit contributions | |
TIT | +0.537 |
endorsement | +0.518 |
provocation | +0.497 |
approval | +0.488 |
� | +0.477 |
Neg logit contributions | |
Bridgewater | -0.532 |
Unic | -0.513 |
Flavoring | -0.484 |
wegian | -0.474 |
Reviewer | -0.473 |
Token | Flynn |
Token ID | 14654 |
Feature activation | -4.97e-01 |
Pos logit contributions | |
Bridgewater | +0.218 |
Unic | +0.209 |
Flavoring | +0.191 |
wegian | +0.186 |
Reviewer | +0.184 |
Neg logit contributions | |
TIT | -0.356 |
endorsement | -0.342 |
provocation | -0.332 |
approval | -0.326 |
� | -0.321 |
Token | assembly |
Token ID | 41873 |
Feature activation | +1.50e+00 |
Pos logit contributions | |
TIT | +0.583 |
endorsement | +0.554 |
provocation | +0.530 |
approval | +0.523 |
� | +0.514 |
Neg logit contributions | |
Bridgewater | -0.480 |
Unic | -0.462 |
Flavoring | -0.438 |
wegian | -0.421 |
Reviewer | -0.417 |
Token | ] |
Token ID | 60 |
Feature activation | +4.91e-01 |
Pos logit contributions | |
TIT | +1.042 |
endorsement | +0.997 |
provocation | +0.954 |
approval | +0.942 |
� | +0.918 |
Neg logit contributions | |
Bridgewater | -1.041 |
Unic | -1.001 |
Flavoring | -0.950 |
wegian | -0.928 |
Reviewer | -0.918 |
Token | :: |
Token ID | 3712 |
Feature activation | +5.04e-01 |
Pos logit contributions | |
TIT | +0.374 |
endorsement | +0.359 |
provocation | +0.345 |
approval | +0.339 |
� | +0.332 |
Neg logit contributions | |
Bridgewater | -0.310 |
Unic | -0.298 |
Flavoring | -0.279 |
wegian | -0.274 |
Reviewer | -0.272 |
Token | load |
Token ID | 2220 |
Feature activation | +1.32e+00 |
Pos logit contributions | |
TIT | +0.369 |
endorsement | +0.352 |
provocation | +0.338 |
approval | +0.333 |
� | +0.326 |
Neg logit contributions | |
Bridgewater | -0.333 |
Unic | -0.320 |
Flavoring | -0.302 |
wegian | -0.295 |
Reviewer | -0.293 |
Token | With |
Token ID | 3152 |
Feature activation | +5.17e-01 |
Pos logit contributions | |
TIT | +0.980 |
endorsement | +0.937 |
provocation | +0.900 |
approval | +0.886 |
� | +0.866 |
Neg logit contributions | |
Bridgewater | -0.851 |
Unic | -0.815 |
Flavoring | -0.768 |
wegian | -0.754 |
Reviewer | -0.744 |
Token | Part |
Token ID | 7841 |
Feature activation | +3.29e+00 |
Pos logit contributions | |
TIT | +0.373 |
endorsement | +0.355 |
provocation | +0.342 |
approval | +0.336 |
� | +0.329 |
Neg logit contributions | |
Bridgewater | -0.347 |
Unic | -0.334 |
Flavoring | -0.314 |
wegian | -0.308 |
Reviewer | -0.305 |
Token | ial |
Token ID | 498 |
Feature activation | -1.29e+00 |
Pos logit contributions | |
TIT | +2.360 |
endorsement | +2.247 |
provocation | +2.152 |
approval | +2.121 |
� | +2.069 |
Neg logit contributions | |
Bridgewater | -2.199 |
Unic | -2.126 |
Flavoring | -1.997 |
wegian | -1.935 |
Reviewer | -1.934 |
Token | Name |
Token ID | 5376 |
Feature activation | +2.00e-01 |
Pos logit contributions | |
Bridgewater | +0.878 |
Unic | +0.849 |
Flavoring | +0.797 |
wegian | +0.780 |
Reviewer | +0.779 |
Neg logit contributions | |
TIT | -0.913 |
endorsement | -0.876 |
provocation | -0.845 |
approval | -0.827 |
� | -0.809 |
Token | (' |
Token ID | 10786 |
Feature activation | -2.91e-01 |
Pos logit contributions | |
TIT | +0.144 |
endorsement | +0.137 |
provocation | +0.132 |
approval | +0.130 |
� | +0.127 |
Neg logit contributions | |
Bridgewater | -0.134 |
Unic | -0.129 |
Flavoring | -0.121 |
wegian | -0.119 |
Reviewer | -0.118 |
Token | system |
Token ID | 10057 |
Feature activation | -2.55e-01 |
Pos logit contributions | |
Bridgewater | +0.184 |
Unic | +0.177 |
Flavoring | +0.166 |
wegian | +0.162 |
Reviewer | +0.161 |
Neg logit contributions | |
TIT | -0.220 |
endorsement | -0.211 |
provocation | -0.203 |
approval | -0.200 |
� | -0.196 |
Token | . |
Token ID | 13 |
Feature activation | +6.27e-01 |
Pos logit contributions | |
Bridgewater | +0.166 |
Unic | +0.160 |
Flavoring | +0.150 |
wegian | +0.146 |
Reviewer | +0.146 |
Neg logit contributions | |
TIT | -0.188 |
endorsement | -0.181 |
provocation | -0.174 |
approval | -0.171 |
� | -0.167 |
Token | Hide |
Token ID | 10415 |
Feature activation | +4.95e-01 |
Pos logit contributions | |
TIT | +0.385 |
endorsement | +0.369 |
provocation | +0.355 |
approval | +0.350 |
� | +0.342 |
Neg logit contributions | |
Bridgewater | -0.319 |
Unic | -0.308 |
Flavoring | -0.288 |
wegian | -0.285 |
Reviewer | -0.282 |
Token | Caption |
Token ID | 11260 |
Feature activation | -4.05e+00 |
Pos logit contributions | |
TIT | +0.182 |
endorsement | +0.167 |
provocation | +0.152 |
approval | +0.148 |
� | +0.138 |
Neg logit contributions | |
Bridgewater | -0.508 |
Unic | -0.496 |
Flavoring | -0.476 |
wegian | -0.471 |
Reviewer | -0.469 |
Token | 4 |
Token ID | 604 |
Feature activation | +8.82e-01 |
Pos logit contributions | |
Bridgewater | +2.369 |
Unic | +2.279 |
Flavoring | +2.075 |
vich | +2.006 |
tc | +1.995 |
Neg logit contributions | |
TIT | -3.248 |
endorsement | -3.147 |
provocation | -3.055 |
approval | -2.975 |
� | -2.940 |
Token | of |
Token ID | 286 |
Feature activation | +1.04e+00 |
Pos logit contributions | |
TIT | +0.737 |
endorsement | +0.711 |
provocation | +0.688 |
approval | +0.678 |
� | +0.665 |
Neg logit contributions | |
Bridgewater | -0.488 |
Unic | -0.468 |
Flavoring | -0.432 |
wegian | -0.423 |
Reviewer | -0.419 |
Token | 15 |
Token ID | 1315 |
Feature activation | -1.68e-01 |
Pos logit contributions | |
TIT | +0.792 |
endorsement | +0.761 |
provocation | +0.734 |
approval | +0.722 |
� | +0.705 |
Neg logit contributions | |
Bridgewater | -0.651 |
Unic | -0.627 |
Flavoring | -0.589 |
wegian | -0.580 |
Reviewer | -0.575 |
Token | Photos |
Token ID | 9434 |
Feature activation | -4.35e+00 |
Pos logit contributions | |
Bridgewater | +0.082 |
Unic | +0.079 |
Flavoring | +0.072 |
wegian | +0.070 |
Reviewer | +0.070 |
Neg logit contributions | |
TIT | -0.150 |
endorsement | -0.146 |
provocation | -0.141 |
approval | -0.140 |
Hait | -0.137 |
Token | : |
Token ID | 25 |
Feature activation | +2.10e-01 |
Pos logit contributions | |
Bridgewater | +2.078 |
Unic | +1.978 |
Flavoring | +1.836 |
vich | +1.727 |
Reviewer | +1.719 |
Neg logit contributions | |
TIT | -3.880 |
endorsement | -3.744 |
provocation | -3.676 |
� | -3.564 |
approval | -3.536 |
Token | Carly |
Token ID | 35982 |
Feature activation | +1.16e+00 |
Pos logit contributions | |
TIT | +0.165 |
endorsement | +0.159 |
provocation | +0.154 |
approval | +0.151 |
� | +0.148 |
Neg logit contributions | |
Bridgewater | -0.127 |
Unic | -0.122 |
Flavoring | -0.114 |
wegian | -0.111 |
Reviewer | -0.111 |
Token | Fiorina |
Token ID | 49028 |
Feature activation | -2.15e-01 |
Pos logit contributions | |
TIT | +0.879 |
endorsement | +0.864 |
provocation | +0.827 |
approval | +0.813 |
� | +0.790 |
Neg logit contributions | |
Bridgewater | -0.710 |
Unic | -0.690 |
Flavoring | -0.666 |
Reviewer | -0.657 |
wegian | -0.638 |
Token | 's |
Token ID | 338 |
Feature activation | -6.04e-01 |
Pos logit contributions | |
Bridgewater | +0.128 |
Unic | +0.122 |
Flavoring | +0.115 |
wegian | +0.112 |
Reviewer | +0.112 |
Neg logit contributions | |
TIT | -0.171 |
endorsement | -0.166 |
provocation | -0.160 |
approval | -0.158 |
� | -0.153 |
Token | political |
Token ID | 1964 |
Feature activation | -8.88e-01 |
Pos logit contributions | |
Bridgewater | +0.420 |
Unic | +0.405 |
Flavoring | +0.379 |
wegian | +0.372 |
Reviewer | +0.367 |
Neg logit contributions | |
TIT | -0.423 |
endorsement | -0.399 |
provocation | -0.387 |
approval | -0.376 |
� | -0.373 |
Token | iff |
Token ID | 733 |
Feature activation | +2.36e+00 |
Pos logit contributions | |
Bridgewater | +0.461 |
Unic | +0.448 |
Flavoring | +0.426 |
Reviewer | +0.420 |
wegian | +0.420 |
Neg logit contributions | |
TIT | -0.321 |
endorsement | -0.303 |
provocation | -0.290 |
approval | -0.281 |
� | -0.274 |
Token | Tra |
Token ID | 15721 |
Feature activation | -1.76e+00 |
Pos logit contributions | |
TIT | +1.554 |
endorsement | +1.453 |
provocation | +1.383 |
approval | +1.361 |
� | +1.338 |
Neg logit contributions | |
Bridgewater | -1.747 |
Unic | -1.684 |
Flavoring | -1.614 |
wegian | -1.581 |
vich | -1.544 |
Token | x |
Token ID | 87 |
Feature activation | -7.44e-01 |
Pos logit contributions | |
Bridgewater | +1.035 |
Unic | +0.996 |
Flavoring | +0.926 |
wegian | +0.906 |
Reviewer | +0.905 |
Neg logit contributions | |
TIT | -1.403 |
endorsement | -1.353 |
provocation | -1.314 |
approval | -1.278 |
� | -1.265 |
Token |
|
Token ID | 198 |
Feature activation | -3.02e+00 |
Pos logit contributions | |
Bridgewater | +0.437 |
Unic | +0.418 |
Flavoring | +0.389 |
wegian | +0.380 |
Reviewer | +0.374 |
Neg logit contributions | |
TIT | -0.598 |
endorsement | -0.576 |
provocation | -0.558 |
approval | -0.549 |
� | -0.535 |
Token |
|
Token ID | 198 |
Feature activation | -2.51e+00 |
Pos logit contributions | |
Bridgewater | +1.543 |
Unic | +1.469 |
Flavoring | +1.335 |
Reviewer | +1.320 |
vich | +1.289 |
Neg logit contributions | |
TIT | -2.617 |
endorsement | -2.554 |
provocation | -2.504 |
approval | -2.415 |
� | -2.387 |
Token | Reviewer |
Token ID | 35407 |
Feature activation | -4.32e+00 |
Pos logit contributions | |
Bridgewater | +0.174 |
Unic | +0.114 |
Flavoring | +0.016 |
Reviewer | +0.001 |
wegian | -0.011 |
Neg logit contributions | |
TIT | -3.298 |
endorsement | -3.230 |
provocation | -3.169 |
approval | -3.132 |
� | -3.095 |
Token | : |
Token ID | 25 |
Feature activation | -9.97e-01 |
Pos logit contributions | |
Bridgewater | +2.104 |
Unic | +2.021 |
Flavoring | +1.883 |
Reviewer | +1.854 |
vich | +1.775 |
Neg logit contributions | |
TIT | -3.771 |
endorsement | -3.691 |
provocation | -3.656 |
approval | -3.477 |
� | -3.465 |
Token | V |
Token ID | 569 |
Feature activation | +1.14e-02 |
Pos logit contributions | |
Bridgewater | +0.635 |
Unic | +0.611 |
Flavoring | +0.572 |
wegian | +0.560 |
Reviewer | +0.557 |
Neg logit contributions | |
TIT | -0.750 |
endorsement | -0.721 |
provocation | -0.696 |
approval | -0.683 |
� | -0.668 |
Token | the |
Token ID | 1169 |
Feature activation | +1.16e-01 |
Pos logit contributions | |
TIT | +0.009 |
endorsement | +0.008 |
provocation | +0.008 |
approval | +0.008 |
� | +0.008 |
Neg logit contributions | |
Bridgewater | -0.007 |
Unic | -0.007 |
Flavoring | -0.006 |
wegian | -0.006 |
Reviewer | -0.006 |
Token | Happy |
Token ID | 25082 |
Feature activation | -1.58e-01 |
Pos logit contributions | |
TIT | +0.089 |
endorsement | +0.085 |
provocation | +0.082 |
approval | +0.081 |
� | +0.079 |
Neg logit contributions | |
Bridgewater | -0.073 |
Unic | -0.070 |
Flavoring | -0.066 |
wegian | -0.064 |
Reviewer | -0.063 |
Token | L |
Token ID | 43 |
Feature activation | -1.03e-01 |
Pos logit contributions | |
Bridgewater | +0.099 |
Unic | +0.095 |
Flavoring | +0.090 |
wegian | +0.087 |
Reviewer | +0.086 |
Neg logit contributions | |
TIT | -0.121 |
endorsement | -0.115 |
provocation | -0.111 |
approval | -0.109 |
� | -0.107 |
Token | Hide |
Token ID | 10415 |
Feature activation | +3.02e-01 |
Pos logit contributions | |
Bridgewater | +0.047 |
Unic | +0.045 |
Flavoring | +0.042 |
wegian | +0.041 |
Reviewer | +0.041 |
Neg logit contributions | |
TIT | -0.077 |
endorsement | -0.075 |
provocation | -0.072 |
approval | -0.071 |
� | -0.070 |
Token | Caption |
Token ID | 11260 |
Feature activation | -3.77e+00 |
Pos logit contributions | |
TIT | +0.120 |
endorsement | +0.112 |
provocation | +0.105 |
approval | +0.100 |
� | +0.096 |
Neg logit contributions | |
Bridgewater | -0.299 |
Unic | -0.291 |
Flavoring | -0.279 |
Reviewer | -0.274 |
wegian | -0.274 |
Token | 5 |
Token ID | 642 |
Feature activation | +9.02e-01 |
Pos logit contributions | |
Bridgewater | +1.955 |
Unic | +1.867 |
Flavoring | +1.734 |
vich | +1.650 |
tc | +1.622 |
Neg logit contributions | |
TIT | -3.252 |
endorsement | -3.147 |
provocation | -3.099 |
� | -3.022 |
approval | -3.015 |
Token | of |
Token ID | 286 |
Feature activation | +6.65e-01 |
Pos logit contributions | |
TIT | +0.748 |
endorsement | +0.722 |
provocation | +0.696 |
approval | +0.688 |
� | +0.674 |
Neg logit contributions | |
Bridgewater | -0.505 |
Unic | -0.483 |
Flavoring | -0.448 |
wegian | -0.440 |
Reviewer | -0.435 |
Token | 15 |
Token ID | 1315 |
Feature activation | -2.65e-01 |
Pos logit contributions | |
TIT | +0.548 |
endorsement | +0.529 |
provocation | +0.512 |
approval | +0.503 |
� | +0.494 |
Neg logit contributions | |
Bridgewater | -0.377 |
Unic | -0.361 |
Flavoring | -0.334 |
wegian | -0.326 |
Reviewer | -0.323 |
Token | Photos |
Token ID | 9434 |
Feature activation | -4.31e+00 |
Pos logit contributions | |
Bridgewater | +0.117 |
Unic | +0.111 |
Flavoring | +0.100 |
wegian | +0.096 |
Reviewer | +0.095 |
Neg logit contributions | |
TIT | -0.252 |
endorsement | -0.244 |
provocation | -0.238 |
approval | -0.234 |
� | -0.230 |
Token | : |
Token ID | 25 |
Feature activation | +2.51e-01 |
Pos logit contributions | |
Bridgewater | +2.068 |
Unic | +1.972 |
Flavoring | +1.805 |
vich | +1.732 |
Reviewer | +1.702 |
Neg logit contributions | |
TIT | -3.821 |
endorsement | -3.747 |
provocation | -3.702 |
approval | -3.543 |
� | -3.500 |
Token | Carly |
Token ID | 35982 |
Feature activation | +7.35e-01 |
Pos logit contributions | |
TIT | +0.244 |
endorsement | +0.237 |
provocation | +0.230 |
approval | +0.227 |
� | +0.224 |
Neg logit contributions | |
Bridgewater | -0.104 |
Unic | -0.098 |
Flavoring | -0.088 |
wegian | -0.085 |
Reviewer | -0.084 |
Token | Fiorina |
Token ID | 49028 |
Feature activation | -6.00e-01 |
Pos logit contributions | |
TIT | +0.576 |
endorsement | +0.559 |
provocation | +0.537 |
approval | +0.529 |
� | +0.517 |
Neg logit contributions | |
Bridgewater | -0.440 |
Unic | -0.422 |
Flavoring | -0.396 |
wegian | -0.389 |
Reviewer | -0.389 |
Token | 's |
Token ID | 338 |
Feature activation | -4.46e-01 |
Pos logit contributions | |
Bridgewater | +0.310 |
Unic | +0.296 |
Flavoring | +0.270 |
wegian | +0.263 |
Reviewer | +0.261 |
Neg logit contributions | |
TIT | -0.526 |
endorsement | -0.505 |
provocation | -0.491 |
approval | -0.482 |
� | -0.476 |
Token | political |
Token ID | 1964 |
Feature activation | -1.95e-01 |
Pos logit contributions | |
Bridgewater | +0.336 |
Unic | +0.325 |
Flavoring | +0.306 |
wegian | +0.302 |
Reviewer | +0.298 |
Neg logit contributions | |
TIT | -0.285 |
endorsement | -0.269 |
provocation | -0.258 |
approval | -0.252 |
� | -0.248 |
Token | er |
Token ID | 263 |
Feature activation | -7.00e-01 |
Pos logit contributions | |
Bridgewater | +0.102 |
Unic | +0.099 |
Flavoring | +0.093 |
wegian | +0.092 |
Reviewer | +0.091 |
Neg logit contributions | |
TIT | -0.090 |
endorsement | -0.086 |
provocation | -0.083 |
approval | -0.081 |
� | -0.079 |
Token | magnet |
Token ID | 19972 |
Feature activation | -1.29e+00 |
Pos logit contributions | |
Bridgewater | +0.432 |
Unic | +0.417 |
Flavoring | +0.388 |
wegian | +0.380 |
Reviewer | +0.378 |
Neg logit contributions | |
TIT | -0.541 |
endorsement | -0.518 |
provocation | -0.499 |
approval | -0.490 |
� | -0.484 |
Token | ? |
Token ID | 30 |
Feature activation | +7.50e-02 |
Pos logit contributions | |
Bridgewater | +1.026 |
Unic | +0.997 |
Flavoring | +0.939 |
wegian | +0.928 |
Reviewer | +0.917 |
Neg logit contributions | |
TIT | -0.763 |
endorsement | -0.726 |
provocation | -0.694 |
approval | -0.677 |
� | -0.656 |
Token |
|
Token ID | 198 |
Feature activation | -2.17e+00 |
Pos logit contributions | |
TIT | +0.061 |
endorsement | +0.059 |
provocation | +0.057 |
approval | +0.056 |
� | +0.055 |
Neg logit contributions | |
Bridgewater | -0.043 |
Unic | -0.041 |
Flavoring | -0.038 |
wegian | -0.037 |
Reviewer | -0.036 |
Token |
|
Token ID | 198 |
Feature activation | -1.71e+00 |
Pos logit contributions | |
Bridgewater | +1.154 |
Unic | +1.101 |
Flavoring | +1.005 |
Reviewer | +0.989 |
wegian | +0.967 |
Neg logit contributions | |
TIT | -1.850 |
endorsement | -1.793 |
provocation | -1.755 |
approval | -1.700 |
� | -1.684 |
Token | Reviewer |
Token ID | 35407 |
Feature activation | -4.26e+00 |
Pos logit contributions | |
Bridgewater | +0.157 |
Unic | +0.119 |
Flavoring | +0.044 |
wegian | +0.027 |
vich | +0.003 |
Neg logit contributions | |
TIT | -2.222 |
endorsement | -2.166 |
provocation | -2.119 |
approval | -2.100 |
� | -2.076 |
Token | : |
Token ID | 25 |
Feature activation | -8.08e-01 |
Pos logit contributions | |
Bridgewater | +2.105 |
Unic | +2.019 |
Flavoring | +1.876 |
Reviewer | +1.842 |
vich | +1.770 |
Neg logit contributions | |
TIT | -3.710 |
endorsement | -3.627 |
provocation | -3.588 |
approval | -3.426 |
� | -3.411 |
Token | hey |
Token ID | 17207 |
Feature activation | -1.05e+00 |
Pos logit contributions | |
Bridgewater | +0.512 |
Unic | +0.492 |
Flavoring | +0.461 |
wegian | +0.451 |
Reviewer | +0.449 |
Neg logit contributions | |
TIT | -0.611 |
endorsement | -0.587 |
provocation | -0.567 |
approval | -0.556 |
� | -0.544 |
Token | _ |
Token ID | 62 |
Feature activation | -2.36e+00 |
Pos logit contributions | |
Bridgewater | +0.623 |
Unic | +0.597 |
Flavoring | +0.558 |
wegian | +0.542 |
Reviewer | +0.539 |
Neg logit contributions | |
TIT | -0.832 |
endorsement | -0.800 |
provocation | -0.772 |
approval | -0.760 |
� | -0.745 |
Token | ler |
Token ID | 1754 |
Feature activation | -4.30e-01 |
Pos logit contributions | |
Bridgewater | +1.411 |
Unic | +1.356 |
Flavoring | +1.259 |
wegian | +1.237 |
Reviewer | +1.229 |
Neg logit contributions | |
TIT | -1.861 |
endorsement | -1.793 |
provocation | -1.734 |
approval | -1.702 |
� | -1.668 |
Token | oy |
Token ID | 726 |
Feature activation | -1.76e+00 |
Pos logit contributions | |
Bridgewater | +0.238 |
Unic | +0.227 |
Flavoring | +0.211 |
wegian | +0.205 |
Reviewer | +0.203 |
Neg logit contributions | |
TIT | -0.359 |
endorsement | -0.347 |
provocation | -0.335 |
approval | -0.331 |
� | -0.323 |
Token | 2010 |
Token ID | 3050 |
Feature activation | -1.23e+00 |
Pos logit contributions | |
Bridgewater | +0.297 |
Unic | +0.287 |
Flavoring | +0.264 |
wegian | +0.258 |
Reviewer | +0.257 |
Neg logit contributions | |
TIT | -0.379 |
endorsement | -0.364 |
provocation | -0.353 |
approval | -0.342 |
� | -0.339 |
Token | boys |
Token ID | 13202 |
Feature activation | -4.86e-01 |
Pos logit contributions | |
Bridgewater | +0.595 |
Unic | +0.570 |
Flavoring | +0.524 |
wegian | +0.496 |
Reviewer | +0.487 |
Neg logit contributions | |
TIT | -1.129 |
endorsement | -1.083 |
provocation | -1.047 |
approval | -1.038 |
� | -1.027 |
Token | beware |
Token ID | 40600 |
Feature activation | -5.70e-01 |
Pos logit contributions | |
Bridgewater | +0.396 |
Unic | +0.385 |
Flavoring | +0.366 |
wegian | +0.357 |
Reviewer | +0.357 |
Neg logit contributions | |
TIT | -0.279 |
endorsement | -0.268 |
provocation | -0.256 |
approval | -0.249 |
� | -0.238 |
Token |
|
Token ID | 198 |
Feature activation | -3.28e+00 |
Pos logit contributions | |
Bridgewater | +0.337 |
Unic | +0.325 |
Flavoring | +0.300 |
wegian | +0.297 |
vich | +0.288 |
Neg logit contributions | |
TIT | -0.454 |
endorsement | -0.436 |
provocation | -0.423 |
approval | -0.415 |
� | -0.404 |
Token |
|
Token ID | 198 |
Feature activation | -2.72e+00 |
Pos logit contributions | |
Bridgewater | +1.684 |
Unic | +1.596 |
Flavoring | +1.458 |
Reviewer | +1.453 |
vich | +1.405 |
Neg logit contributions | |
TIT | -2.837 |
endorsement | -2.761 |
provocation | -2.713 |
approval | -2.614 |
� | -2.597 |
Token | Reviewer |
Token ID | 35407 |
Feature activation | -4.25e+00 |
Pos logit contributions | |
Bridgewater | +0.224 |
Unic | +0.160 |
Flavoring | +0.049 |
wegian | +0.019 |
Reviewer | +0.000 |
Neg logit contributions | |
TIT | -3.562 |
endorsement | -3.478 |
provocation | -3.407 |
approval | -3.373 |
� | -3.336 |
Token | : |
Token ID | 25 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
Bridgewater | +2.082 |
Unic | +2.001 |
Flavoring | +1.862 |
Reviewer | +1.841 |
vich | +1.762 |
Neg logit contributions | |
TIT | -3.705 |
endorsement | -3.620 |
provocation | -3.589 |
approval | -3.413 |
� | -3.403 |
Token | j |
Token ID | 474 |
Feature activation | -9.60e-01 |
Pos logit contributions | |
Bridgewater | +0.662 |
Unic | +0.637 |
Flavoring | +0.596 |
wegian | +0.584 |
Reviewer | +0.581 |
Neg logit contributions | |
TIT | -0.779 |
endorsement | -0.749 |
provocation | -0.722 |
approval | -0.709 |
� | -0.694 |
Token | wt |
Token ID | 46569 |
Feature activation | +2.34e-01 |
Pos logit contributions | |
Bridgewater | +0.608 |
Unic | +0.585 |
Flavoring | +0.547 |
wegian | +0.534 |
Reviewer | +0.531 |
Neg logit contributions | |
TIT | -0.727 |
endorsement | -0.698 |
provocation | -0.674 |
approval | -0.662 |
� | -0.648 |
Token | 0001 |
Token ID | 18005 |
Feature activation | -2.61e+00 |
Pos logit contributions | |
TIT | +0.190 |
endorsement | +0.184 |
provocation | +0.176 |
approval | +0.174 |
� | +0.170 |
Neg logit contributions | |
Bridgewater | -0.136 |
Unic | -0.130 |
Flavoring | -0.122 |
wegian | -0.117 |
Reviewer | -0.117 |
Token | - |
Token ID | 532 |
Feature activation | -2.58e+00 |
Pos logit contributions | |
Bridgewater | +1.252 |
Unic | +1.190 |
Flavoring | +1.086 |
wegian | +1.056 |
Reviewer | +1.046 |
Neg logit contributions | |
TIT | -2.371 |
endorsement | -2.293 |
provocation | -2.225 |
approval | -2.192 |
� | -2.155 |
Token | . |
Token ID | 13 |
Feature activation | +2.07e-01 |
Pos logit contributions | |
TIT | +0.520 |
endorsement | +0.505 |
provocation | +0.486 |
approval | +0.477 |
� | +0.464 |
Neg logit contributions | |
Bridgewater | -0.444 |
Unic | -0.425 |
Flavoring | -0.402 |
wegian | -0.394 |
Reviewer | -0.392 |
Token | Flag |
Token ID | 19762 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
TIT | +0.157 |
endorsement | +0.150 |
provocation | +0.145 |
approval | +0.143 |
� | +0.139 |
Neg logit contributions | |
Bridgewater | -0.131 |
Unic | -0.126 |
Flavoring | -0.118 |
wegian | -0.117 |
Reviewer | -0.116 |
Token |
|
Token ID | 198 |
Feature activation | -1.61e+00 |
Pos logit contributions | |
TIT | +0.860 |
endorsement | +0.829 |
provocation | +0.800 |
approval | +0.787 |
� | +0.770 |
Neg logit contributions | |
Bridgewater | -0.641 |
Unic | -0.615 |
Flavoring | -0.574 |
wegian | -0.562 |
Reviewer | -0.557 |
Token |
|
Token ID | 198 |
Feature activation | -1.37e+00 |
Pos logit contributions | |
Bridgewater | +0.863 |
Unic | +0.825 |
Flavoring | +0.752 |
Reviewer | +0.729 |
vich | +0.727 |
Neg logit contributions | |
TIT | -1.365 |
endorsement | -1.322 |
provocation | -1.294 |
approval | -1.254 |
� | -1.242 |
Token | Flag |
Token ID | 34227 |
Feature activation | -7.10e-02 |
Pos logit contributions | |
Bridgewater | +0.845 |
Unic | +0.809 |
Flavoring | +0.757 |
Reviewer | +0.751 |
wegian | +0.738 |
Neg logit contributions | |
TIT | -1.040 |
endorsement | -1.014 |
provocation | -0.993 |
approval | -0.964 |
� | -0.943 |
Token | Reply |
Token ID | 14883 |
Feature activation | -4.25e+00 |
Pos logit contributions | |
Bridgewater | +0.023 |
Unic | +0.021 |
Flavoring | +0.019 |
wegian | +0.017 |
Reviewer | +0.017 |
Neg logit contributions | |
TIT | -0.076 |
endorsement | -0.073 |
provocation | -0.072 |
approval | -0.071 |
� | -0.070 |
Token |
|
Token ID | 198 |
Feature activation | -1.53e+00 |
Pos logit contributions | |
Bridgewater | +2.057 |
Unic | +1.974 |
Flavoring | +1.797 |
vich | +1.753 |
tc | +1.725 |
Neg logit contributions | |
TIT | -3.757 |
endorsement | -3.625 |
provocation | -3.574 |
� | -3.449 |
approval | -3.449 |
Token |
|
Token ID | 198 |
Feature activation | -1.38e+00 |
Pos logit contributions | |
Bridgewater | +0.828 |
Unic | +0.795 |
Flavoring | +0.722 |
Reviewer | +0.702 |
wegian | +0.701 |
Neg logit contributions | |
TIT | -1.296 |
endorsement | -1.254 |
provocation | -1.225 |
approval | -1.190 |
� | -1.175 |
Token | 19 |
Token ID | 1129 |
Feature activation | -9.33e-01 |
Pos logit contributions | |
Bridgewater | +0.774 |
Unic | +0.742 |
Flavoring | +0.685 |
Reviewer | +0.673 |
wegian | +0.668 |
Neg logit contributions | |
TIT | -1.134 |
endorsement | -1.099 |
provocation | -1.068 |
approval | -1.047 |
� | -1.025 |
Token | Recommend |
Token ID | 30361 |
Feature activation | -1.79e+00 |
Pos logit contributions | |
Bridgewater | +0.670 |
Unic | +0.648 |
Flavoring | +0.610 |
wegian | +0.599 |
Reviewer | +0.596 |
Neg logit contributions | |
TIT | -0.628 |
endorsement | -0.598 |
provocation | -0.576 |
approval | -0.563 |
� | -0.553 |
Token |
|
Token ID | 198 |
Feature activation | -2.01e+00 |
Pos logit contributions | |
Bridgewater | +0.948 |
Unic | +0.908 |
Flavoring | +0.829 |
vich | +0.797 |
Reviewer | +0.793 |
Neg logit contributions | |
TIT | -1.529 |
endorsement | -1.463 |
provocation | -1.441 |
� | -1.398 |
approval | -1.394 |
Token | Hide |
Token ID | 10415 |
Feature activation | -2.19e-01 |
Pos logit contributions | |
Bridgewater | +0.297 |
Unic | +0.283 |
Flavoring | +0.261 |
wegian | +0.256 |
Reviewer | +0.254 |
Neg logit contributions | |
TIT | -0.502 |
endorsement | -0.485 |
provocation | -0.469 |
approval | -0.463 |
� | -0.455 |
Token | Caption |
Token ID | 11260 |
Feature activation | -3.74e+00 |
Pos logit contributions | |
Bridgewater | +0.213 |
Unic | +0.207 |
Flavoring | +0.199 |
Reviewer | +0.194 |
wegian | +0.194 |
Neg logit contributions | |
TIT | -0.090 |
endorsement | -0.085 |
provocation | -0.081 |
approval | -0.076 |
� | -0.073 |
Token | 8 |
Token ID | 807 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
Bridgewater | +1.922 |
Unic | +1.899 |
Flavoring | +1.708 |
tc | +1.637 |
Reviewer | +1.619 |
Neg logit contributions | |
TIT | -3.254 |
endorsement | -3.118 |
provocation | -3.087 |
approval | -2.952 |
� | -2.947 |
Token | of |
Token ID | 286 |
Feature activation | +5.05e-01 |
Pos logit contributions | |
TIT | +0.909 |
endorsement | +0.877 |
provocation | +0.848 |
approval | +0.835 |
� | +0.819 |
Neg logit contributions | |
Bridgewater | -0.598 |
Unic | -0.572 |
Flavoring | -0.529 |
wegian | -0.518 |
Reviewer | -0.514 |
Token | 15 |
Token ID | 1315 |
Feature activation | -8.66e-01 |
Pos logit contributions | |
TIT | +0.420 |
endorsement | +0.405 |
provocation | +0.393 |
approval | +0.385 |
� | +0.379 |
Neg logit contributions | |
Bridgewater | -0.283 |
Unic | -0.271 |
Flavoring | -0.250 |
wegian | -0.242 |
Reviewer | -0.241 |
Token | Photos |
Token ID | 9434 |
Feature activation | -4.23e+00 |
Pos logit contributions | |
Bridgewater | +0.368 |
Unic | +0.349 |
Flavoring | +0.310 |
vich | +0.290 |
wegian | +0.290 |
Neg logit contributions | |
TIT | -0.840 |
endorsement | -0.816 |
provocation | -0.799 |
approval | -0.781 |
� | -0.771 |
Token | : |
Token ID | 25 |
Feature activation | +2.32e-01 |
Pos logit contributions | |
Bridgewater | +2.035 |
Unic | +1.946 |
Flavoring | +1.772 |
vich | +1.703 |
Reviewer | +1.673 |
Neg logit contributions | |
TIT | -3.751 |
endorsement | -3.671 |
provocation | -3.630 |
approval | -3.471 |
� | -3.431 |
Token | Carly |
Token ID | 35982 |
Feature activation | +2.29e-01 |
Pos logit contributions | |
TIT | +0.230 |
endorsement | +0.222 |
provocation | +0.217 |
approval | +0.214 |
� | +0.210 |
Neg logit contributions | |
Bridgewater | -0.093 |
Unic | -0.088 |
Flavoring | -0.078 |
wegian | -0.074 |
Reviewer | -0.073 |
Token | Fiorina |
Token ID | 49028 |
Feature activation | -1.00e+00 |
Pos logit contributions | |
TIT | +0.188 |
endorsement | +0.181 |
provocation | +0.175 |
approval | +0.172 |
� | +0.169 |
Neg logit contributions | |
Bridgewater | -0.132 |
Unic | -0.126 |
Flavoring | -0.117 |
wegian | -0.114 |
Reviewer | -0.113 |
Token | 's |
Token ID | 338 |
Feature activation | -5.07e-01 |
Pos logit contributions | |
Bridgewater | +0.495 |
Unic | +0.474 |
Flavoring | +0.426 |
wegian | +0.415 |
Reviewer | +0.410 |
Neg logit contributions | |
TIT | -0.898 |
endorsement | -0.857 |
provocation | -0.838 |
approval | -0.822 |
� | -0.816 |
Token | political |
Token ID | 1964 |
Feature activation | -3.65e-01 |
Pos logit contributions | |
Bridgewater | +0.367 |
Unic | +0.355 |
Flavoring | +0.330 |
wegian | +0.329 |
tc | +0.323 |
Neg logit contributions | |
TIT | -0.339 |
endorsement | -0.312 |
provocation | -0.303 |
� | -0.294 |
approval | -0.294 |
Token | are |
Token ID | 389 |
Feature activation | -6.82e-01 |
Pos logit contributions | |
Bridgewater | +0.441 |
Unic | +0.425 |
Flavoring | +0.400 |
wegian | +0.393 |
Reviewer | +0.391 |
Neg logit contributions | |
TIT | -0.432 |
endorsement | -0.414 |
provocation | -0.398 |
approval | -0.390 |
� | -0.380 |
Token | Here |
Token ID | 3423 |
Feature activation | -4.17e-01 |
Pos logit contributions | |
Bridgewater | +0.470 |
Unic | +0.451 |
Flavoring | +0.423 |
wegian | +0.420 |
Reviewer | +0.411 |
Neg logit contributions | |
TIT | -0.476 |
endorsement | -0.457 |
provocation | -0.438 |
approval | -0.432 |
� | -0.420 |
Token | .... |
Token ID | 1106 |
Feature activation | -1.71e+00 |
Pos logit contributions | |
Bridgewater | +0.270 |
Unic | +0.260 |
Flavoring | +0.243 |
wegian | +0.239 |
Reviewer | +0.237 |
Neg logit contributions | |
TIT | -0.310 |
endorsement | -0.297 |
provocation | -0.287 |
approval | -0.281 |
� | -0.275 |
Token |
|
Token ID | 198 |
Feature activation | -1.63e+00 |
Pos logit contributions | |
Bridgewater | +0.934 |
Unic | +0.893 |
Flavoring | +0.824 |
wegian | +0.805 |
Reviewer | +0.798 |
Neg logit contributions | |
TIT | -1.441 |
endorsement | -1.389 |
provocation | -1.346 |
approval | -1.324 |
� | -1.300 |
Token |
|
Token ID | 198 |
Feature activation | -1.39e+00 |
Pos logit contributions | |
Bridgewater | +0.879 |
Unic | +0.840 |
Flavoring | +0.769 |
Reviewer | +0.754 |
wegian | +0.742 |
Neg logit contributions | |
TIT | -1.373 |
endorsement | -1.330 |
provocation | -1.299 |
approval | -1.261 |
� | -1.249 |
Token | Reviewer |
Token ID | 35407 |
Feature activation | -4.21e+00 |
Pos logit contributions | |
Bridgewater | +0.138 |
Unic | +0.108 |
Flavoring | +0.045 |
wegian | +0.032 |
vich | +0.012 |
Neg logit contributions | |
TIT | -1.806 |
endorsement | -1.758 |
provocation | -1.719 |
approval | -1.705 |
� | -1.684 |
Token | : |
Token ID | 25 |
Feature activation | -6.27e-01 |
Pos logit contributions | |
Bridgewater | +2.082 |
Unic | +2.000 |
Flavoring | +1.858 |
Reviewer | +1.819 |
vich | +1.754 |
Neg logit contributions | |
TIT | -3.662 |
endorsement | -3.579 |
provocation | -3.542 |
approval | -3.380 |
� | -3.365 |
Token | SAS |
Token ID | 35516 |
Feature activation | -1.18e+00 |
Pos logit contributions | |
Bridgewater | +0.393 |
Unic | +0.378 |
Flavoring | +0.354 |
wegian | +0.347 |
Reviewer | +0.345 |
Neg logit contributions | |
TIT | -0.477 |
endorsement | -0.460 |
provocation | -0.443 |
approval | -0.436 |
� | -0.426 |
Token | K |
Token ID | 42 |
Feature activation | +7.00e-01 |
Pos logit contributions | |
Bridgewater | +0.747 |
Unic | +0.719 |
Flavoring | +0.672 |
wegian | +0.658 |
Reviewer | +0.654 |
Neg logit contributions | |
TIT | -0.893 |
endorsement | -0.857 |
provocation | -0.825 |
approval | -0.811 |
� | -0.794 |
Token | ATCH |
Token ID | 11417 |
Feature activation | +4.73e-01 |
Pos logit contributions | |
TIT | +0.539 |
endorsement | +0.513 |
provocation | +0.493 |
approval | +0.488 |
� | +0.477 |
Neg logit contributions | |
Bridgewater | -0.439 |
Unic | -0.422 |
Flavoring | -0.395 |
wegian | -0.387 |
Reviewer | -0.383 |
Token | EW |
Token ID | 6217 |
Feature activation | +2.13e-01 |
Pos logit contributions | |
TIT | +0.364 |
endorsement | +0.347 |
provocation | +0.333 |
approval | +0.329 |
� | +0.323 |
Neg logit contributions | |
Bridgewater | -0.296 |
Unic | -0.286 |
Flavoring | -0.266 |
wegian | -0.262 |
Reviewer | -0.259 |
Token | Hide |
Token ID | 10415 |
Feature activation | -9.08e-04 |
Pos logit contributions | |
Bridgewater | +0.102 |
Unic | +0.098 |
Flavoring | +0.092 |
wegian | +0.091 |
Reviewer | +0.091 |
Neg logit contributions | |
TIT | -0.149 |
endorsement | -0.144 |
provocation | -0.138 |
approval | -0.138 |
� | -0.135 |
Token | Caption |
Token ID | 11260 |
Feature activation | -3.69e+00 |
Pos logit contributions | |
Bridgewater | +0.001 |
Unic | +0.001 |
Flavoring | +0.001 |
Reviewer | +0.001 |
wegian | +0.001 |
Neg logit contributions | |
TIT | -0.000 |
endorsement | -0.000 |
provocation | -0.000 |
approval | -0.000 |
� | -0.000 |
Token | 10 |
Token ID | 838 |
Feature activation | +9.55e-01 |
Pos logit contributions | |
Bridgewater | +1.733 |
Unic | +1.634 |
Flavoring | +1.456 |
tc | +1.402 |
vich | +1.358 |
Neg logit contributions | |
TIT | -3.405 |
endorsement | -3.299 |
provocation | -3.210 |
� | -3.116 |
approval | -3.109 |
Token | of |
Token ID | 286 |
Feature activation | +6.66e-01 |
Pos logit contributions | |
TIT | +0.802 |
endorsement | +0.774 |
provocation | +0.748 |
approval | +0.737 |
� | +0.723 |
Neg logit contributions | |
Bridgewater | -0.525 |
Unic | -0.502 |
Flavoring | -0.464 |
wegian | -0.454 |
Reviewer | -0.451 |
Token | 15 |
Token ID | 1315 |
Feature activation | -7.98e-01 |
Pos logit contributions | |
TIT | +0.552 |
endorsement | +0.532 |
provocation | +0.515 |
approval | +0.506 |
� | +0.497 |
Neg logit contributions | |
Bridgewater | -0.375 |
Unic | -0.359 |
Flavoring | -0.332 |
wegian | -0.323 |
Reviewer | -0.321 |
Token | Photos |
Token ID | 9434 |
Feature activation | -4.17e+00 |
Pos logit contributions | |
Bridgewater | +0.341 |
Unic | +0.324 |
Flavoring | +0.288 |
wegian | +0.271 |
vich | +0.270 |
Neg logit contributions | |
TIT | -0.771 |
endorsement | -0.749 |
provocation | -0.732 |
approval | -0.716 |
� | -0.707 |
Token | : |
Token ID | 25 |
Feature activation | +1.73e-01 |
Pos logit contributions | |
Bridgewater | +2.024 |
Unic | +1.935 |
Flavoring | +1.756 |
vich | +1.691 |
Reviewer | +1.664 |
Neg logit contributions | |
TIT | -3.692 |
endorsement | -3.608 |
provocation | -3.566 |
approval | -3.413 |
� | -3.371 |
Token | Carly |
Token ID | 35982 |
Feature activation | +3.27e-01 |
Pos logit contributions | |
TIT | +0.172 |
endorsement | +0.166 |
provocation | +0.163 |
approval | +0.160 |
� | +0.157 |
Neg logit contributions | |
Bridgewater | -0.069 |
Unic | -0.066 |
Flavoring | -0.058 |
wegian | -0.055 |
Reviewer | -0.055 |
Token | Fiorina |
Token ID | 49028 |
Feature activation | -8.26e-01 |
Pos logit contributions | |
TIT | +0.267 |
endorsement | +0.256 |
provocation | +0.249 |
approval | +0.244 |
� | +0.240 |
Neg logit contributions | |
Bridgewater | -0.188 |
Unic | -0.181 |
Flavoring | -0.167 |
wegian | -0.162 |
Reviewer | -0.161 |
Token | 's |
Token ID | 338 |
Feature activation | -4.77e-01 |
Pos logit contributions | |
Bridgewater | +0.412 |
Unic | +0.394 |
Flavoring | +0.356 |
wegian | +0.346 |
Reviewer | +0.343 |
Neg logit contributions | |
TIT | -0.738 |
endorsement | -0.706 |
provocation | -0.689 |
approval | -0.676 |
� | -0.670 |
Token | political |
Token ID | 1964 |
Feature activation | -3.84e-01 |
Pos logit contributions | |
Bridgewater | +0.344 |
Unic | +0.333 |
Flavoring | +0.309 |
wegian | +0.308 |
tc | +0.303 |
Neg logit contributions | |
TIT | -0.321 |
endorsement | -0.295 |
provocation | -0.287 |
� | -0.279 |
approval | -0.278 |
Token | Hide |
Token ID | 10415 |
Feature activation | +2.96e-01 |
Pos logit contributions | |
Bridgewater | +0.019 |
Unic | +0.018 |
Flavoring | +0.017 |
wegian | +0.017 |
Reviewer | +0.017 |
Neg logit contributions | |
TIT | -0.031 |
endorsement | -0.030 |
provocation | -0.029 |
approval | -0.029 |
� | -0.028 |
Token | Caption |
Token ID | 11260 |
Feature activation | -3.49e+00 |
Pos logit contributions | |
TIT | +0.121 |
endorsement | +0.113 |
provocation | +0.106 |
approval | +0.101 |
� | +0.097 |
Neg logit contributions | |
Bridgewater | -0.291 |
Unic | -0.283 |
Flavoring | -0.272 |
Reviewer | -0.266 |
wegian | -0.266 |
Token | 6 |
Token ID | 718 |
Feature activation | +7.87e-01 |
Pos logit contributions | |
Bridgewater | +1.696 |
Unic | +1.642 |
Flavoring | +1.460 |
tc | +1.428 |
vich | +1.418 |
Neg logit contributions | |
TIT | -3.163 |
endorsement | -3.038 |
provocation | -2.981 |
approval | -2.852 |
� | -2.846 |
Token | of |
Token ID | 286 |
Feature activation | +8.19e-01 |
Pos logit contributions | |
TIT | +0.661 |
endorsement | +0.638 |
provocation | +0.617 |
approval | +0.607 |
� | +0.596 |
Neg logit contributions | |
Bridgewater | -0.434 |
Unic | -0.415 |
Flavoring | -0.384 |
wegian | -0.375 |
Reviewer | -0.372 |
Token | 15 |
Token ID | 1315 |
Feature activation | -5.45e-01 |
Pos logit contributions | |
TIT | +0.676 |
endorsement | +0.651 |
provocation | +0.631 |
approval | +0.620 |
� | +0.608 |
Neg logit contributions | |
Bridgewater | -0.463 |
Unic | -0.444 |
Flavoring | -0.411 |
wegian | -0.400 |
Reviewer | -0.397 |
Token | Photos |
Token ID | 9434 |
Feature activation | -4.17e+00 |
Pos logit contributions | |
Bridgewater | +0.236 |
Unic | +0.224 |
Flavoring | +0.201 |
wegian | +0.189 |
Reviewer | +0.188 |
Neg logit contributions | |
TIT | -0.524 |
endorsement | -0.509 |
provocation | -0.497 |
approval | -0.487 |
� | -0.480 |
Token | : |
Token ID | 25 |
Feature activation | +2.20e-01 |
Pos logit contributions | |
Bridgewater | +2.011 |
Unic | +1.920 |
Flavoring | +1.751 |
vich | +1.683 |
Reviewer | +1.649 |
Neg logit contributions | |
TIT | -3.701 |
endorsement | -3.620 |
provocation | -3.578 |
approval | -3.424 |
� | -3.382 |
Token | Carly |
Token ID | 35982 |
Feature activation | +4.14e-01 |
Pos logit contributions | |
TIT | +0.217 |
endorsement | +0.210 |
provocation | +0.205 |
approval | +0.202 |
� | +0.199 |
Neg logit contributions | |
Bridgewater | -0.089 |
Unic | -0.084 |
Flavoring | -0.075 |
wegian | -0.072 |
Reviewer | -0.071 |
Token | Fiorina |
Token ID | 49028 |
Feature activation | -8.54e-01 |
Pos logit contributions | |
TIT | +0.335 |
endorsement | +0.323 |
provocation | +0.312 |
approval | +0.307 |
� | +0.301 |
Neg logit contributions | |
Bridgewater | -0.241 |
Unic | -0.231 |
Flavoring | -0.214 |
wegian | -0.209 |
Reviewer | -0.208 |
Token | 's |
Token ID | 338 |
Feature activation | -4.94e-01 |
Pos logit contributions | |
Bridgewater | +0.429 |
Unic | +0.411 |
Flavoring | +0.371 |
wegian | +0.362 |
Reviewer | +0.359 |
Neg logit contributions | |
TIT | -0.761 |
endorsement | -0.728 |
provocation | -0.710 |
approval | -0.697 |
� | -0.690 |
Token | political |
Token ID | 1964 |
Feature activation | -3.69e-01 |
Pos logit contributions | |
Bridgewater | +0.362 |
Unic | +0.350 |
Flavoring | +0.327 |
wegian | +0.325 |
tc | +0.318 |
Neg logit contributions | |
TIT | -0.326 |
endorsement | -0.303 |
provocation | -0.293 |
approval | -0.285 |
� | -0.283 |
Token | AB |
Token ID | 6242 |
Feature activation | +9.78e-01 |
Pos logit contributions | |
Bridgewater | +0.364 |
Unic | +0.351 |
Flavoring | +0.329 |
wegian | +0.321 |
Reviewer | +0.320 |
Neg logit contributions | |
TIT | -0.431 |
endorsement | -0.415 |
provocation | -0.399 |
approval | -0.393 |
� | -0.383 |
Token | arry |
Token ID | 6532 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
TIT | +0.750 |
endorsement | +0.720 |
provocation | +0.688 |
approval | +0.683 |
� | +0.661 |
Neg logit contributions | |
Bridgewater | -0.604 |
Unic | -0.586 |
Flavoring | -0.552 |
wegian | -0.543 |
Reviewer | -0.534 |
Token | Maryland |
Token ID | 10769 |
Feature activation | -1.53e+00 |
Pos logit contributions | |
TIT | +0.817 |
endorsement | +0.793 |
provocation | +0.757 |
approval | +0.750 |
Hait | +0.730 |
Neg logit contributions | |
Bridgewater | -0.659 |
Unic | -0.639 |
Flavoring | -0.604 |
wegian | -0.591 |
Reviewer | -0.588 |
Token | 1 |
Token ID | 352 |
Feature activation | -1.31e+00 |
Pos logit contributions | |
Bridgewater | +0.917 |
Unic | +0.884 |
Flavoring | +0.827 |
wegian | +0.811 |
Reviewer | +0.806 |
Neg logit contributions | |
TIT | -1.199 |
endorsement | -1.161 |
provocation | -1.116 |
approval | -1.101 |
� | -1.069 |
Token | hour |
Token ID | 1711 |
Feature activation | -5.95e-01 |
Pos logit contributions | |
Bridgewater | +0.868 |
Unic | +0.834 |
Flavoring | +0.783 |
Reviewer | +0.774 |
wegian | +0.772 |
Neg logit contributions | |
TIT | -0.950 |
endorsement | -0.931 |
provocation | -0.880 |
approval | -0.880 |
� | -0.839 |
Token | ago |
Token ID | 2084 |
Feature activation | -4.17e+00 |
Pos logit contributions | |
Bridgewater | +0.357 |
Unic | +0.342 |
Flavoring | +0.320 |
Reviewer | +0.314 |
wegian | +0.312 |
Neg logit contributions | |
TIT | -0.467 |
endorsement | -0.457 |
provocation | -0.437 |
approval | -0.432 |
Hait | -0.417 |
Token | This |
Token ID | 770 |
Feature activation | -1.70e+00 |
Pos logit contributions | |
Bridgewater | +2.642 |
Unic | +2.540 |
Flavoring | +2.341 |
Reviewer | +2.257 |
wegian | +2.249 |
Neg logit contributions | |
TIT | -3.175 |
endorsement | -3.001 |
provocation | -2.931 |
� | -2.829 |
approval | -2.826 |
Token | election |
Token ID | 3071 |
Feature activation | -8.33e-01 |
Pos logit contributions | |
Bridgewater | +1.038 |
Unic | +0.999 |
Flavoring | +0.922 |
wegian | +0.903 |
Reviewer | +0.898 |
Neg logit contributions | |
TIT | -1.332 |
endorsement | -1.265 |
provocation | -1.226 |
approval | -1.203 |
� | -1.194 |
Token | was |
Token ID | 373 |
Feature activation | -4.82e-01 |
Pos logit contributions | |
Bridgewater | +0.513 |
Unic | +0.493 |
Flavoring | +0.459 |
wegian | +0.450 |
Reviewer | +0.447 |
Neg logit contributions | |
TIT | -0.646 |
endorsement | -0.620 |
provocation | -0.599 |
approval | -0.588 |
� | -0.577 |
Token | not |
Token ID | 407 |
Feature activation | -5.75e-01 |
Pos logit contributions | |
Bridgewater | +0.287 |
Unic | +0.276 |
Flavoring | +0.257 |
wegian | +0.251 |
Reviewer | +0.249 |
Neg logit contributions | |
TIT | -0.382 |
endorsement | -0.368 |
provocation | -0.356 |
approval | -0.349 |
� | -0.343 |
Token | about |
Token ID | 546 |
Feature activation | +6.79e-01 |
Pos logit contributions | |
Bridgewater | +0.298 |
Unic | +0.285 |
Flavoring | +0.264 |
wegian | +0.257 |
Reviewer | +0.255 |
Neg logit contributions | |
TIT | -0.500 |
endorsement | -0.484 |
provocation | -0.469 |
approval | -0.462 |
� | -0.452 |
Token | Hide |
Token ID | 10415 |
Feature activation | -6.58e-02 |
Pos logit contributions | |
Bridgewater | +0.251 |
Unic | +0.240 |
Flavoring | +0.222 |
wegian | +0.217 |
Reviewer | +0.216 |
Neg logit contributions | |
TIT | -0.405 |
endorsement | -0.390 |
provocation | -0.377 |
approval | -0.372 |
� | -0.366 |
Token | Caption |
Token ID | 11260 |
Feature activation | -3.72e+00 |
Pos logit contributions | |
Bridgewater | +0.064 |
Unic | +0.063 |
Flavoring | +0.060 |
Reviewer | +0.059 |
wegian | +0.059 |
Neg logit contributions | |
TIT | -0.027 |
endorsement | -0.025 |
provocation | -0.024 |
approval | -0.023 |
� | -0.022 |
Token | 7 |
Token ID | 767 |
Feature activation | +7.35e-01 |
Pos logit contributions | |
Bridgewater | +1.889 |
Unic | +1.809 |
Flavoring | +1.647 |
tc | +1.574 |
Reviewer | +1.559 |
Neg logit contributions | |
TIT | -3.261 |
endorsement | -3.130 |
provocation | -3.097 |
approval | -2.988 |
� | -2.971 |
Token | of |
Token ID | 286 |
Feature activation | +6.29e-01 |
Pos logit contributions | |
TIT | +0.619 |
endorsement | +0.597 |
provocation | +0.578 |
approval | +0.569 |
� | +0.558 |
Neg logit contributions | |
Bridgewater | -0.402 |
Unic | -0.385 |
Flavoring | -0.356 |
wegian | -0.347 |
Reviewer | -0.345 |
Token | 15 |
Token ID | 1315 |
Feature activation | -1.01e+00 |
Pos logit contributions | |
TIT | +0.521 |
endorsement | +0.503 |
provocation | +0.487 |
approval | +0.478 |
� | +0.469 |
Neg logit contributions | |
Bridgewater | -0.354 |
Unic | -0.339 |
Flavoring | -0.313 |
wegian | -0.304 |
Reviewer | -0.302 |
Token | Photos |
Token ID | 9434 |
Feature activation | -4.17e+00 |
Pos logit contributions | |
Bridgewater | +0.430 |
Unic | +0.408 |
Flavoring | +0.364 |
vich | +0.340 |
wegian | +0.339 |
Neg logit contributions | |
TIT | -0.981 |
endorsement | -0.954 |
provocation | -0.933 |
approval | -0.912 |
� | -0.900 |
Token | : |
Token ID | 25 |
Feature activation | +7.43e-02 |
Pos logit contributions | |
Bridgewater | +2.013 |
Unic | +1.925 |
Flavoring | +1.755 |
vich | +1.686 |
Reviewer | +1.654 |
Neg logit contributions | |
TIT | -3.695 |
endorsement | -3.615 |
provocation | -3.573 |
approval | -3.417 |
� | -3.375 |
Token | Carly |
Token ID | 35982 |
Feature activation | +1.82e-01 |
Pos logit contributions | |
TIT | +0.074 |
endorsement | +0.071 |
provocation | +0.070 |
approval | +0.069 |
� | +0.068 |
Neg logit contributions | |
Bridgewater | -0.030 |
Unic | -0.028 |
Flavoring | -0.025 |
wegian | -0.024 |
Reviewer | -0.023 |
Token | Fiorina |
Token ID | 49028 |
Feature activation | -9.48e-01 |
Pos logit contributions | |
TIT | +0.149 |
endorsement | +0.143 |
provocation | +0.138 |
approval | +0.136 |
� | +0.133 |
Neg logit contributions | |
Bridgewater | -0.104 |
Unic | -0.100 |
Flavoring | -0.093 |
wegian | -0.090 |
Reviewer | -0.089 |
Token | 's |
Token ID | 338 |
Feature activation | -5.16e-01 |
Pos logit contributions | |
Bridgewater | +0.470 |
Unic | +0.450 |
Flavoring | +0.405 |
wegian | +0.394 |
Reviewer | +0.391 |
Neg logit contributions | |
TIT | -0.850 |
endorsement | -0.812 |
provocation | -0.794 |
approval | -0.778 |
� | -0.772 |
Token | political |
Token ID | 1964 |
Feature activation | -4.72e-01 |
Pos logit contributions | |
Bridgewater | +0.374 |
Unic | +0.362 |
Flavoring | +0.337 |
wegian | +0.336 |
tc | +0.330 |
Neg logit contributions | |
TIT | -0.344 |
endorsement | -0.318 |
provocation | -0.308 |
� | -0.299 |
approval | -0.298 |
Token | February |
Token ID | 3945 |
Feature activation | -8.87e-01 |
Pos logit contributions | |
Bridgewater | +0.005 |
Unic | +0.005 |
Flavoring | +0.005 |
wegian | +0.005 |
Reviewer | +0.005 |
Neg logit contributions | |
TIT | -0.007 |
endorsement | -0.007 |
provocation | -0.006 |
approval | -0.006 |
� | -0.006 |
Token | 11 |
Token ID | 1367 |
Feature activation | -1.59e-02 |
Pos logit contributions | |
Bridgewater | +0.525 |
Unic | +0.504 |
Flavoring | +0.465 |
wegian | +0.453 |
Reviewer | +0.449 |
Neg logit contributions | |
TIT | -0.711 |
endorsement | -0.683 |
provocation | -0.660 |
approval | -0.647 |
� | -0.638 |
Token | , |
Token ID | 11 |
Feature activation | +3.26e-01 |
Pos logit contributions | |
Bridgewater | +0.010 |
Unic | +0.009 |
Flavoring | +0.009 |
wegian | +0.009 |
Reviewer | +0.009 |
Neg logit contributions | |
TIT | -0.012 |
endorsement | -0.012 |
provocation | -0.011 |
approval | -0.011 |
� | -0.011 |
Token | 2016 |
Token ID | 1584 |
Feature activation | -1.88e+00 |
Pos logit contributions | |
TIT | +0.259 |
endorsement | +0.250 |
provocation | +0.242 |
approval | +0.238 |
� | +0.232 |
Neg logit contributions | |
Bridgewater | -0.192 |
Unic | -0.184 |
Flavoring | -0.173 |
wegian | -0.170 |
Reviewer | -0.169 |
Token | . |
Token ID | 13 |
Feature activation | -4.01e-01 |
Pos logit contributions | |
Bridgewater | +1.189 |
Unic | +1.146 |
Flavoring | +1.066 |
wegian | +1.043 |
Reviewer | +1.037 |
Neg logit contributions | |
TIT | -1.423 |
endorsement | -1.364 |
provocation | -1.317 |
approval | -1.288 |
� | -1.269 |
Token | REUTERS |
Token ID | 15862 |
Feature activation | -4.13e+00 |
Pos logit contributions | |
Bridgewater | +0.260 |
Unic | +0.250 |
Flavoring | +0.234 |
wegian | +0.231 |
Reviewer | +0.229 |
Neg logit contributions | |
TIT | -0.297 |
endorsement | -0.285 |
provocation | -0.274 |
approval | -0.270 |
� | -0.264 |
Token | / |
Token ID | 14 |
Feature activation | +1.04e+00 |
Pos logit contributions | |
Bridgewater | +1.859 |
Unic | +1.764 |
Flavoring | +1.582 |
vich | +1.480 |
wegian | +1.475 |
Neg logit contributions | |
TIT | -3.799 |
endorsement | -3.750 |
provocation | -3.675 |
approval | -3.534 |
� | -3.530 |
Token | Cl |
Token ID | 2601 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
TIT | +0.767 |
endorsement | +0.723 |
provocation | +0.694 |
approval | +0.685 |
� | +0.679 |
Neg logit contributions | |
Bridgewater | -0.677 |
Unic | -0.654 |
Flavoring | -0.619 |
wegian | -0.604 |
Reviewer | -0.591 |
Token | od |
Token ID | 375 |
Feature activation | -8.62e-02 |
Pos logit contributions | |
TIT | +0.845 |
endorsement | +0.823 |
provocation | +0.787 |
approval | +0.779 |
Hait | +0.757 |
Neg logit contributions | |
Bridgewater | -0.727 |
Unic | -0.702 |
Flavoring | -0.668 |
Reviewer | -0.650 |
wegian | -0.632 |
Token | agh |
Token ID | 10471 |
Feature activation | -1.02e-02 |
Pos logit contributions | |
Bridgewater | +0.085 |
Unic | +0.084 |
Flavoring | +0.081 |
Reviewer | +0.080 |
wegian | +0.079 |
Neg logit contributions | |
TIT | -0.033 |
endorsement | -0.031 |
provocation | -0.029 |
approval | -0.028 |
� | -0.026 |
Token | Kil |
Token ID | 19902 |
Feature activation | -1.01e+00 |
Pos logit contributions | |
Bridgewater | +0.005 |
Unic | +0.005 |
Flavoring | +0.005 |
Reviewer | +0.004 |
wegian | +0.004 |
Neg logit contributions | |
TIT | -0.009 |
endorsement | -0.009 |
approval | -0.008 |
provocation | -0.008 |
� | -0.008 |
Token | Hide |
Token ID | 10415 |
Feature activation | +9.59e-02 |
Pos logit contributions | |
Bridgewater | +0.207 |
Unic | +0.198 |
Flavoring | +0.184 |
wegian | +0.181 |
Reviewer | +0.180 |
Neg logit contributions | |
TIT | -0.311 |
endorsement | -0.301 |
provocation | -0.290 |
approval | -0.287 |
� | -0.281 |
Token | Caption |
Token ID | 11260 |
Feature activation | -3.67e+00 |
Pos logit contributions | |
TIT | +0.039 |
endorsement | +0.037 |
provocation | +0.035 |
approval | +0.033 |
� | +0.032 |
Neg logit contributions | |
Bridgewater | -0.094 |
Unic | -0.091 |
Flavoring | -0.088 |
Reviewer | -0.086 |
wegian | -0.086 |
Token | 9 |
Token ID | 860 |
Feature activation | +1.66e+00 |
Pos logit contributions | |
Bridgewater | +1.948 |
Unic | +1.891 |
Flavoring | +1.698 |
Reviewer | +1.653 |
vich | +1.614 |
Neg logit contributions | |
TIT | -3.124 |
endorsement | -2.994 |
provocation | -2.977 |
� | -2.887 |
approval | -2.884 |
Token | of |
Token ID | 286 |
Feature activation | +8.73e-01 |
Pos logit contributions | |
TIT | +1.363 |
endorsement | +1.317 |
provocation | +1.267 |
approval | +1.254 |
� | +1.226 |
Neg logit contributions | |
Bridgewater | -0.940 |
Unic | -0.899 |
Flavoring | -0.836 |
wegian | -0.822 |
Reviewer | -0.815 |
Token | 15 |
Token ID | 1315 |
Feature activation | -6.28e-01 |
Pos logit contributions | |
TIT | +0.719 |
endorsement | +0.693 |
provocation | +0.671 |
approval | +0.659 |
� | +0.648 |
Neg logit contributions | |
Bridgewater | -0.494 |
Unic | -0.473 |
Flavoring | -0.438 |
wegian | -0.427 |
Reviewer | -0.424 |
Token | Photos |
Token ID | 9434 |
Feature activation | -4.11e+00 |
Pos logit contributions | |
Bridgewater | +0.269 |
Unic | +0.256 |
Flavoring | +0.228 |
wegian | +0.214 |
vich | +0.213 |
Neg logit contributions | |
TIT | -0.606 |
endorsement | -0.588 |
provocation | -0.575 |
approval | -0.563 |
� | -0.556 |
Token | : |
Token ID | 25 |
Feature activation | +2.55e-01 |
Pos logit contributions | |
Bridgewater | +1.990 |
Unic | +1.901 |
Flavoring | +1.729 |
vich | +1.663 |
Reviewer | +1.635 |
Neg logit contributions | |
TIT | -3.638 |
endorsement | -3.556 |
provocation | -3.517 |
approval | -3.364 |
� | -3.323 |
Token | Carly |
Token ID | 35982 |
Feature activation | +1.88e-01 |
Pos logit contributions | |
TIT | +0.253 |
endorsement | +0.245 |
provocation | +0.239 |
approval | +0.235 |
� | +0.231 |
Neg logit contributions | |
Bridgewater | -0.103 |
Unic | -0.097 |
Flavoring | -0.086 |
wegian | -0.082 |
Reviewer | -0.081 |
Token | Fiorina |
Token ID | 49028 |
Feature activation | -8.08e-01 |
Pos logit contributions | |
TIT | +0.154 |
endorsement | +0.148 |
provocation | +0.144 |
approval | +0.141 |
� | +0.138 |
Neg logit contributions | |
Bridgewater | -0.108 |
Unic | -0.104 |
Flavoring | -0.095 |
wegian | -0.093 |
Reviewer | -0.092 |
Token | 's |
Token ID | 338 |
Feature activation | -4.71e-01 |
Pos logit contributions | |
Bridgewater | +0.401 |
Unic | +0.383 |
Flavoring | +0.346 |
wegian | +0.336 |
Reviewer | +0.332 |
Neg logit contributions | |
TIT | -0.725 |
endorsement | -0.691 |
provocation | -0.676 |
approval | -0.663 |
� | -0.658 |
Token | political |
Token ID | 1964 |
Feature activation | -2.61e-01 |
Pos logit contributions | |
Bridgewater | +0.339 |
Unic | +0.328 |
Flavoring | +0.305 |
wegian | +0.304 |
tc | +0.299 |
Neg logit contributions | |
TIT | -0.317 |
endorsement | -0.291 |
provocation | -0.283 |
� | -0.276 |
approval | -0.274 |
Token | . |
Token ID | 13 |
Feature activation | -4.53e-01 |
Pos logit contributions | |
TIT | +0.187 |
endorsement | +0.182 |
provocation | +0.176 |
approval | +0.172 |
� | +0.166 |
Neg logit contributions | |
Bridgewater | -0.172 |
Unic | -0.165 |
Flavoring | -0.156 |
Reviewer | -0.154 |
wegian | -0.154 |
Token | Flag |
Token ID | 19762 |
Feature activation | +1.29e+00 |
Pos logit contributions | |
Bridgewater | +0.287 |
Unic | +0.276 |
Flavoring | +0.258 |
wegian | +0.252 |
Reviewer | +0.251 |
Neg logit contributions | |
TIT | -0.344 |
endorsement | -0.330 |
provocation | -0.319 |
approval | -0.313 |
� | -0.307 |
Token |
|
Token ID | 198 |
Feature activation | -1.07e+00 |
Pos logit contributions | |
TIT | +0.982 |
endorsement | +0.958 |
provocation | +0.919 |
approval | +0.908 |
Hait | +0.873 |
Neg logit contributions | |
Bridgewater | -0.798 |
Unic | -0.769 |
Flavoring | -0.733 |
wegian | -0.717 |
Reviewer | -0.700 |
Token |
|
Token ID | 198 |
Feature activation | -9.26e-01 |
Pos logit contributions | |
Bridgewater | +0.588 |
Unic | +0.562 |
Flavoring | +0.515 |
Reviewer | +0.500 |
wegian | +0.498 |
Neg logit contributions | |
TIT | -0.896 |
endorsement | -0.867 |
provocation | -0.846 |
approval | -0.823 |
� | -0.812 |
Token | Flag |
Token ID | 34227 |
Feature activation | +6.56e-01 |
Pos logit contributions | |
Bridgewater | +0.592 |
Unic | +0.567 |
Flavoring | +0.531 |
Reviewer | +0.522 |
wegian | +0.521 |
Neg logit contributions | |
TIT | -0.690 |
endorsement | -0.666 |
provocation | -0.645 |
approval | -0.630 |
� | -0.618 |
Token | Reply |
Token ID | 14883 |
Feature activation | -4.10e+00 |
Pos logit contributions | |
TIT | +0.687 |
endorsement | +0.669 |
provocation | +0.651 |
approval | +0.644 |
� | +0.632 |
Neg logit contributions | |
Bridgewater | -0.224 |
Unic | -0.209 |
Flavoring | -0.183 |
wegian | -0.176 |
Reviewer | -0.174 |
Token |
|
Token ID | 198 |
Feature activation | -1.42e+00 |
Pos logit contributions | |
Bridgewater | +2.002 |
Unic | +1.933 |
Flavoring | +1.767 |
vich | +1.690 |
tc | +1.690 |
Neg logit contributions | |
TIT | -3.633 |
endorsement | -3.468 |
provocation | -3.412 |
� | -3.337 |
approval | -3.301 |
Token |
|
Token ID | 198 |
Feature activation | -1.21e+00 |
Pos logit contributions | |
Bridgewater | +0.778 |
Unic | +0.747 |
Flavoring | +0.681 |
Reviewer | +0.662 |
wegian | +0.661 |
Neg logit contributions | |
TIT | -1.195 |
endorsement | -1.156 |
provocation | -1.127 |
approval | -1.098 |
� | -1.082 |
Token | 55 |
Token ID | 2816 |
Feature activation | -1.63e+00 |
Pos logit contributions | |
Bridgewater | +0.666 |
Unic | +0.637 |
Flavoring | +0.588 |
wegian | +0.574 |
Reviewer | +0.573 |
Neg logit contributions | |
TIT | -1.009 |
endorsement | -0.976 |
provocation | -0.946 |
approval | -0.929 |
� | -0.912 |
Token | Recommend |
Token ID | 30361 |
Feature activation | -1.55e+00 |
Pos logit contributions | |
Bridgewater | +1.162 |
Unic | +1.124 |
Flavoring | +1.056 |
wegian | +1.036 |
Reviewer | +1.032 |
Neg logit contributions | |
TIT | -1.105 |
endorsement | -1.049 |
provocation | -1.013 |
approval | -0.988 |
� | -0.973 |
Token |
|
Token ID | 198 |
Feature activation | -1.94e+00 |
Pos logit contributions | |
Bridgewater | +0.837 |
Unic | +0.801 |
Flavoring | +0.735 |
Reviewer | +0.704 |
vich | +0.704 |
Neg logit contributions | |
TIT | -1.321 |
endorsement | -1.264 |
provocation | -1.242 |
approval | -1.203 |
� | -1.202 |
Token |
|
Token ID | 198 |
Feature activation | +1.75e-01 |
Pos logit contributions | |
TIT | +0.096 |
endorsement | +0.093 |
provocation | +0.090 |
approval | +0.088 |
� | +0.087 |
Neg logit contributions | |
Bridgewater | -0.067 |
Unic | -0.064 |
Flavoring | -0.059 |
wegian | -0.057 |
Reviewer | -0.057 |
Token | It |
Token ID | 1026 |
Feature activation | -2.31e-01 |
Pos logit contributions | |
TIT | +0.124 |
endorsement | +0.119 |
provocation | +0.114 |
approval | +0.112 |
� | +0.110 |
Neg logit contributions | |
Bridgewater | -0.119 |
Unic | -0.115 |
Flavoring | -0.108 |
wegian | -0.106 |
Reviewer | -0.105 |
Token | would |
Token ID | 561 |
Feature activation | -2.88e-01 |
Pos logit contributions | |
Bridgewater | +0.154 |
Unic | +0.148 |
Flavoring | +0.139 |
Reviewer | +0.136 |
wegian | +0.136 |
Neg logit contributions | |
TIT | -0.167 |
endorsement | -0.161 |
provocation | -0.155 |
approval | -0.152 |
� | -0.149 |
Token | be |
Token ID | 307 |
Feature activation | -8.53e-01 |
Pos logit contributions | |
Bridgewater | +0.204 |
Unic | +0.198 |
Flavoring | +0.186 |
wegian | +0.183 |
Reviewer | +0.182 |
Neg logit contributions | |
TIT | -0.196 |
endorsement | -0.188 |
provocation | -0.180 |
approval | -0.176 |
� | -0.172 |
Token | dis |
Token ID | 595 |
Feature activation | -1.20e+00 |
Pos logit contributions | |
Bridgewater | +0.512 |
Unic | +0.493 |
Flavoring | +0.455 |
wegian | +0.447 |
Reviewer | +0.445 |
Neg logit contributions | |
TIT | -0.674 |
endorsement | -0.645 |
provocation | -0.624 |
approval | -0.612 |
� | -0.604 |
Token | pir |
Token ID | 4063 |
Feature activation | -4.07e+00 |
Pos logit contributions | |
Bridgewater | +0.738 |
Unic | +0.710 |
Flavoring | +0.659 |
wegian | +0.648 |
Reviewer | +0.642 |
Neg logit contributions | |
TIT | -0.928 |
endorsement | -0.889 |
provocation | -0.861 |
approval | -0.843 |
� | -0.828 |
Token | iting |
Token ID | 1780 |
Feature activation | -1.70e+00 |
Pos logit contributions | |
Bridgewater | +2.602 |
Unic | +2.536 |
Flavoring | +2.347 |
wegian | +2.326 |
Reviewer | +2.296 |
Neg logit contributions | |
TIT | -3.031 |
endorsement | -2.883 |
provocation | -2.800 |
approval | -2.736 |
� | -2.700 |
Token | if |
Token ID | 611 |
Feature activation | -1.09e+00 |
Pos logit contributions | |
Bridgewater | +0.914 |
Unic | +0.876 |
Flavoring | +0.802 |
wegian | +0.785 |
Reviewer | +0.779 |
Neg logit contributions | |
TIT | -1.446 |
endorsement | -1.389 |
provocation | -1.348 |
approval | -1.324 |
� | -1.305 |
Token | the |
Token ID | 262 |
Feature activation | -2.69e-01 |
Pos logit contributions | |
Bridgewater | +0.628 |
Unic | +0.599 |
Flavoring | +0.562 |
wegian | +0.548 |
Reviewer | +0.544 |
Neg logit contributions | |
TIT | -0.883 |
endorsement | -0.852 |
provocation | -0.822 |
approval | -0.811 |
� | -0.789 |
Token | book |
Token ID | 1492 |
Feature activation | +3.86e-01 |
Pos logit contributions | |
Bridgewater | +0.161 |
Unic | +0.154 |
Flavoring | +0.143 |
wegian | +0.140 |
Reviewer | +0.139 |
Neg logit contributions | |
TIT | -0.213 |
endorsement | -0.205 |
provocation | -0.198 |
approval | -0.195 |
� | -0.191 |
Token | came |
Token ID | 1625 |
Feature activation | +3.16e-01 |
Pos logit contributions | |
TIT | +0.298 |
endorsement | +0.288 |
provocation | +0.277 |
approval | +0.273 |
� | +0.266 |
Neg logit contributions | |
Bridgewater | -0.238 |
Unic | -0.228 |
Flavoring | -0.214 |
wegian | -0.208 |
Reviewer | -0.208 |
Token | bad |
Token ID | 2089 |
Feature activation | -2.86e-01 |
Pos logit contributions | |
Bridgewater | +0.282 |
Unic | +0.273 |
Flavoring | +0.257 |
wegian | +0.251 |
Reviewer | +0.249 |
Neg logit contributions | |
TIT | -0.318 |
endorsement | -0.308 |
provocation | -0.296 |
approval | -0.292 |
� | -0.282 |
Token | idea |
Token ID | 2126 |
Feature activation | -7.83e-01 |
Pos logit contributions | |
Bridgewater | +0.138 |
Unic | +0.136 |
Flavoring | +0.124 |
Reviewer | +0.121 |
wegian | +0.118 |
Neg logit contributions | |
TIT | -0.257 |
endorsement | -0.241 |
provocation | -0.237 |
� | -0.234 |
Hait | -0.232 |
Token | . |
Token ID | 13 |
Feature activation | -8.07e-01 |
Pos logit contributions | |
Bridgewater | +0.497 |
Unic | +0.481 |
Flavoring | +0.450 |
Reviewer | +0.440 |
wegian | +0.435 |
Neg logit contributions | |
TIT | -0.588 |
endorsement | -0.566 |
provocation | -0.549 |
approval | -0.534 |
� | -0.525 |
Token |
|
Token ID | 198 |
Feature activation | -2.33e+00 |
Pos logit contributions | |
Bridgewater | +0.492 |
Unic | +0.473 |
Flavoring | +0.438 |
wegian | +0.433 |
Reviewer | +0.426 |
Neg logit contributions | |
TIT | -0.629 |
endorsement | -0.604 |
provocation | -0.585 |
approval | -0.575 |
� | -0.562 |
Token |
|
Token ID | 198 |
Feature activation | -1.94e+00 |
Pos logit contributions | |
Bridgewater | +1.245 |
Unic | +1.186 |
Flavoring | +1.086 |
Reviewer | +1.074 |
wegian | +1.045 |
Neg logit contributions | |
TIT | -1.980 |
endorsement | -1.917 |
provocation | -1.876 |
approval | -1.819 |
� | -1.806 |
Token | Reviewer |
Token ID | 35407 |
Feature activation | -4.03e+00 |
Pos logit contributions | |
Bridgewater | +0.119 |
Unic | +0.071 |
Flavoring | +0.002 |
Reviewer | +0.000 |
wegian | -0.022 |
Neg logit contributions | |
TIT | -2.568 |
endorsement | -2.520 |
provocation | -2.475 |
approval | -2.444 |
� | -2.418 |
Token | : |
Token ID | 25 |
Feature activation | -7.07e-01 |
Pos logit contributions | |
Bridgewater | +1.997 |
Unic | +1.918 |
Flavoring | +1.780 |
Reviewer | +1.752 |
vich | +1.684 |
Neg logit contributions | |
TIT | -3.497 |
endorsement | -3.406 |
provocation | -3.371 |
approval | -3.219 |
� | -3.212 |
Token | Mike |
Token ID | 4995 |
Feature activation | -9.86e-01 |
Pos logit contributions | |
Bridgewater | +0.448 |
Unic | +0.431 |
Flavoring | +0.403 |
wegian | +0.395 |
Reviewer | +0.393 |
Neg logit contributions | |
TIT | -0.534 |
endorsement | -0.514 |
provocation | -0.496 |
approval | -0.487 |
� | -0.476 |
Token | M |
Token ID | 44 |
Feature activation | +2.12e-01 |
Pos logit contributions | |
Bridgewater | +0.650 |
Unic | +0.626 |
Flavoring | +0.589 |
wegian | +0.577 |
Reviewer | +0.573 |
Neg logit contributions | |
TIT | -0.721 |
endorsement | -0.690 |
provocation | -0.665 |
approval | -0.653 |
� | -0.639 |
Token | anners |
Token ID | 15672 |
Feature activation | -5.89e-01 |
Pos logit contributions | |
TIT | +0.135 |
endorsement | +0.129 |
provocation | +0.123 |
approval | +0.122 |
� | +0.118 |
Neg logit contributions | |
Bridgewater | -0.160 |
Unic | -0.154 |
Flavoring | -0.145 |
wegian | -0.143 |
Reviewer | -0.141 |
Token | - |
Token ID | 532 |
Feature activation | -1.91e+00 |
Pos logit contributions | |
Bridgewater | +0.313 |
Unic | +0.299 |
Flavoring | +0.276 |
wegian | +0.269 |
Reviewer | +0.261 |
Neg logit contributions | |
TIT | -0.510 |
endorsement | -0.489 |
provocation | -0.469 |
approval | -0.466 |
� | -0.455 |
Token | it |
Token ID | 270 |
Feature activation | +1.37e-01 |
Pos logit contributions | |
Bridgewater | +0.517 |
Unic | +0.494 |
wegian | +0.466 |
Flavoring | +0.464 |
vich | +0.464 |
Neg logit contributions | |
TIT | -0.453 |
endorsement | -0.425 |
provocation | -0.411 |
approval | -0.403 |
� | -0.398 |
Token | published |
Token ID | 3199 |
Feature activation | +1.27e-01 |
Pos logit contributions | |
TIT | +0.109 |
endorsement | +0.104 |
provocation | +0.101 |
approval | +0.099 |
� | +0.098 |
Neg logit contributions | |
Bridgewater | -0.081 |
Unic | -0.078 |
Flavoring | -0.072 |
wegian | -0.071 |
Reviewer | -0.070 |
Token | in |
Token ID | 287 |
Feature activation | -9.26e-01 |
Pos logit contributions | |
TIT | +0.104 |
endorsement | +0.102 |
provocation | +0.098 |
approval | +0.097 |
� | +0.094 |
Neg logit contributions | |
Bridgewater | -0.071 |
Unic | -0.068 |
Flavoring | -0.065 |
Reviewer | -0.063 |
wegian | -0.063 |
Token | the |
Token ID | 262 |
Feature activation | -9.09e-01 |
Pos logit contributions | |
Bridgewater | +0.556 |
Unic | +0.534 |
Flavoring | +0.496 |
wegian | +0.485 |
Reviewer | +0.481 |
Neg logit contributions | |
TIT | -0.732 |
endorsement | -0.703 |
provocation | -0.679 |
approval | -0.667 |
� | -0.656 |
Token | Richmond |
Token ID | 16940 |
Feature activation | +7.08e-01 |
Pos logit contributions | |
Bridgewater | +0.616 |
Unic | +0.594 |
Flavoring | +0.557 |
wegian | +0.546 |
Reviewer | +0.543 |
Neg logit contributions | |
TIT | -0.649 |
endorsement | -0.621 |
provocation | -0.598 |
approval | -0.586 |
� | -0.574 |
Token | Times |
Token ID | 3782 |
Feature activation | -4.02e+00 |
Pos logit contributions | |
TIT | +0.499 |
endorsement | +0.479 |
provocation | +0.460 |
approval | +0.451 |
� | +0.440 |
Neg logit contributions | |
Bridgewater | -0.484 |
Unic | -0.467 |
Flavoring | -0.440 |
wegian | -0.431 |
Reviewer | -0.429 |
Token | - |
Token ID | 12 |
Feature activation | -8.39e-02 |
Pos logit contributions | |
Bridgewater | +2.245 |
Unic | +2.184 |
wegian | +1.930 |
Flavoring | +1.929 |
vich | +1.892 |
Neg logit contributions | |
TIT | -3.334 |
endorsement | -3.151 |
provocation | -3.066 |
� | -3.014 |
approval | -2.990 |
Token | Dispatch |
Token ID | 49354 |
Feature activation | -1.67e-01 |
Pos logit contributions | |
Bridgewater | +0.068 |
Unic | +0.066 |
Flavoring | +0.062 |
wegian | +0.061 |
Reviewer | +0.061 |
Neg logit contributions | |
TIT | -0.049 |
endorsement | -0.047 |
provocation | -0.044 |
approval | -0.043 |
� | -0.042 |
Token | . |
Token ID | 13 |
Feature activation | -1.48e+00 |
Pos logit contributions | |
Bridgewater | +0.093 |
Unic | +0.089 |
Flavoring | +0.084 |
Reviewer | +0.082 |
wegian | +0.082 |
Neg logit contributions | |
TIT | -0.137 |
endorsement | -0.134 |
provocation | -0.129 |
approval | -0.127 |
� | -0.124 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -3.03e-01 |
Pos logit contributions | |
Bridgewater | +0.832 |
Unic | +0.798 |
Flavoring | +0.739 |
wegian | +0.722 |
Reviewer | +0.716 |
Neg logit contributions | |
TIT | -1.228 |
endorsement | -1.184 |
provocation | -1.145 |
approval | -1.128 |
� | -1.106 |
Token | Elf |
Token ID | 46995 |
Feature activation | -2.26e-02 |
Pos logit contributions | |
Bridgewater | +0.205 |
Unic | +0.198 |
Flavoring | +0.186 |
wegian | +0.182 |
Reviewer | +0.180 |
Neg logit contributions | |
TIT | -0.218 |
endorsement | -0.208 |
provocation | -0.200 |
approval | -0.196 |
� | -0.193 |
Token | witz |
Token ID | 28155 |
Feature activation | -5.46e-01 |
Pos logit contributions | |
Bridgewater | +0.303 |
Unic | +0.293 |
Flavoring | +0.275 |
wegian | +0.270 |
Reviewer | +0.269 |
Neg logit contributions | |
TIT | -0.331 |
endorsement | -0.318 |
provocation | -0.305 |
approval | -0.300 |
� | -0.293 |
Token | West |
Token ID | 2688 |
Feature activation | -3.13e-01 |
Pos logit contributions | |
Bridgewater | +0.364 |
Unic | +0.352 |
Flavoring | +0.330 |
wegian | +0.324 |
Reviewer | +0.322 |
Neg logit contributions | |
TIT | -0.393 |
endorsement | -0.377 |
provocation | -0.363 |
approval | -0.356 |
� | -0.348 |
Token | chester |
Token ID | 35983 |
Feature activation | -4.01e-01 |
Pos logit contributions | |
Bridgewater | +0.238 |
Unic | +0.231 |
Flavoring | +0.219 |
wegian | +0.215 |
Reviewer | +0.215 |
Neg logit contributions | |
TIT | -0.196 |
endorsement | -0.187 |
provocation | -0.179 |
approval | -0.175 |
� | -0.171 |
Token | 1 |
Token ID | 352 |
Feature activation | -1.66e+00 |
Pos logit contributions | |
Bridgewater | +0.249 |
Unic | +0.241 |
Flavoring | +0.226 |
wegian | +0.221 |
Reviewer | +0.221 |
Neg logit contributions | |
TIT | -0.306 |
endorsement | -0.295 |
provocation | -0.283 |
approval | -0.279 |
� | -0.272 |
Token | hour |
Token ID | 1711 |
Feature activation | -7.26e-01 |
Pos logit contributions | |
Bridgewater | +1.098 |
Unic | +1.058 |
Flavoring | +0.991 |
wegian | +0.973 |
Reviewer | +0.965 |
Neg logit contributions | |
TIT | -1.208 |
endorsement | -1.155 |
provocation | -1.114 |
approval | -1.091 |
� | -1.072 |
Token | ago |
Token ID | 2084 |
Feature activation | -4.02e+00 |
Pos logit contributions | |
Bridgewater | +0.393 |
Unic | +0.375 |
Flavoring | +0.348 |
wegian | +0.338 |
Reviewer | +0.334 |
Neg logit contributions | |
TIT | -0.617 |
endorsement | -0.589 |
provocation | -0.575 |
approval | -0.561 |
� | -0.558 |
Token | Pretty |
Token ID | 20090 |
Feature activation | -1.71e+00 |
Pos logit contributions | |
Bridgewater | +2.570 |
Unic | +2.466 |
Flavoring | +2.262 |
tc | +2.183 |
wegian | +2.179 |
Neg logit contributions | |
TIT | -3.042 |
endorsement | -2.869 |
provocation | -2.785 |
approval | -2.708 |
� | -2.704 |
Token | simple |
Token ID | 2829 |
Feature activation | -8.24e-01 |
Pos logit contributions | |
Bridgewater | +1.057 |
Unic | +1.018 |
Flavoring | +0.928 |
wegian | +0.919 |
vich | +0.902 |
Neg logit contributions | |
TIT | -1.328 |
endorsement | -1.256 |
provocation | -1.217 |
approval | -1.192 |
� | -1.187 |
Token | Paul |
Token ID | 3362 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
Bridgewater | +0.516 |
Unic | +0.497 |
Flavoring | +0.463 |
wegian | +0.454 |
Reviewer | +0.451 |
Neg logit contributions | |
TIT | -0.630 |
endorsement | -0.604 |
provocation | -0.583 |
approval | -0.573 |
� | -0.562 |
Token | : |
Token ID | 25 |
Feature activation | -1.54e+00 |
Pos logit contributions | |
Bridgewater | +0.783 |
Unic | +0.753 |
Flavoring | +0.702 |
wegian | +0.688 |
Reviewer | +0.683 |
Neg logit contributions | |
TIT | -0.965 |
endorsement | -0.926 |
provocation | -0.895 |
approval | -0.878 |
� | -0.861 |
Token |
|
Token ID | 198 |
Feature activation | -2.26e+00 |
Pos logit contributions | |
Bridgewater | +0.961 |
Unic | +0.924 |
Flavoring | +0.856 |
wegian | +0.839 |
Reviewer | +0.835 |
Neg logit contributions | |
TIT | -1.183 |
endorsement | -1.127 |
provocation | -1.092 |
approval | -1.067 |
� | -1.056 |
Token | , |
Token ID | 11 |
Feature activation | -1.26e-01 |
Pos logit contributions | |
TIT | +0.297 |
endorsement | +0.286 |
provocation | +0.276 |
approval | +0.271 |
� | +0.266 |
Neg logit contributions | |
Bridgewater | -0.230 |
Unic | -0.221 |
Flavoring | -0.205 |
wegian | -0.201 |
Reviewer | -0.199 |
Token | I |
Token ID | 314 |
Feature activation | -7.71e-01 |
Pos logit contributions | |
Bridgewater | +0.075 |
Unic | +0.073 |
Flavoring | +0.067 |
wegian | +0.066 |
Reviewer | +0.065 |
Neg logit contributions | |
TIT | -0.100 |
endorsement | -0.096 |
provocation | -0.093 |
approval | -0.092 |
� | -0.089 |
Token | think |
Token ID | 892 |
Feature activation | +1.68e+00 |
Pos logit contributions | |
Bridgewater | +0.473 |
Unic | +0.454 |
Flavoring | +0.422 |
wegian | +0.415 |
Reviewer | +0.412 |
Neg logit contributions | |
TIT | -0.598 |
endorsement | -0.576 |
provocation | -0.556 |
approval | -0.545 |
� | -0.536 |
Token |
|
Token ID | 198 |
Feature activation | -2.56e+00 |
Pos logit contributions | |
TIT | +1.323 |
endorsement | +1.274 |
provocation | +1.230 |
approval | +1.212 |
Hait | +1.178 |
Neg logit contributions | |
Bridgewater | -1.006 |
Unic | -0.964 |
Flavoring | -0.896 |
wegian | -0.894 |
vich | -0.865 |
Token |
|
Token ID | 198 |
Feature activation | -2.03e+00 |
Pos logit contributions | |
Bridgewater | +1.333 |
Unic | +1.269 |
Flavoring | +1.150 |
Reviewer | +1.135 |
vich | +1.111 |
Neg logit contributions | |
TIT | -2.202 |
endorsement | -2.137 |
provocation | -2.096 |
approval | -2.026 |
� | -2.010 |
Token | Reviewer |
Token ID | 35407 |
Feature activation | -3.96e+00 |
Pos logit contributions | |
Bridgewater | +0.175 |
Unic | +0.129 |
Flavoring | +0.044 |
wegian | +0.022 |
Reviewer | +0.000 |
Neg logit contributions | |
TIT | -2.652 |
endorsement | -2.587 |
provocation | -2.533 |
approval | -2.509 |
� | -2.481 |
Token | : |
Token ID | 25 |
Feature activation | -7.04e-01 |
Pos logit contributions | |
Bridgewater | +1.972 |
Unic | +1.891 |
Flavoring | +1.758 |
Reviewer | +1.726 |
vich | +1.658 |
Neg logit contributions | |
TIT | -3.434 |
endorsement | -3.360 |
provocation | -3.319 |
approval | -3.171 |
� | -3.151 |
Token | B |
Token ID | 347 |
Feature activation | -6.03e-01 |
Pos logit contributions | |
Bridgewater | +0.447 |
Unic | +0.430 |
Flavoring | +0.403 |
wegian | +0.395 |
Reviewer | +0.393 |
Neg logit contributions | |
TIT | -0.531 |
endorsement | -0.511 |
provocation | -0.493 |
approval | -0.484 |
� | -0.473 |
Token | iz |
Token ID | 528 |
Feature activation | +1.12e-01 |
Pos logit contributions | |
Bridgewater | +0.380 |
Unic | +0.365 |
Flavoring | +0.341 |
wegian | +0.333 |
Reviewer | +0.331 |
Neg logit contributions | |
TIT | -0.459 |
endorsement | -0.441 |
provocation | -0.425 |
approval | -0.419 |
� | -0.409 |
Token | r |
Token ID | 81 |
Feature activation | -8.05e-01 |
Pos logit contributions | |
TIT | +0.087 |
endorsement | +0.083 |
provocation | +0.080 |
approval | +0.079 |
� | +0.077 |
Neg logit contributions | |
Bridgewater | -0.070 |
Unic | -0.067 |
Flavoring | -0.063 |
wegian | -0.061 |
Reviewer | -0.061 |
Token | Pix |
Token ID | 21642 |
Feature activation | -4.66e-01 |
Pos logit contributions | |
Bridgewater | +0.488 |
Unic | +0.468 |
Flavoring | +0.438 |
wegian | +0.426 |
Reviewer | +0.422 |
Neg logit contributions | |
TIT | -0.633 |
endorsement | -0.608 |
provocation | -0.587 |
approval | -0.578 |
� | -0.565 |
Token | Garg |
Token ID | 39013 |
Feature activation | -1.13e+00 |
Pos logit contributions | |
Bridgewater | +0.184 |
Unic | +0.177 |
Flavoring | +0.165 |
wegian | +0.163 |
Reviewer | +0.162 |
Neg logit contributions | |
TIT | -0.209 |
endorsement | -0.201 |
provocation | -0.194 |
approval | -0.190 |
� | -0.186 |
Token | i |
Token ID | 72 |
Feature activation | -1.60e+00 |
Pos logit contributions | |
Bridgewater | +0.668 |
Unic | +0.642 |
Flavoring | +0.600 |
wegian | +0.584 |
Reviewer | +0.579 |
Neg logit contributions | |
TIT | -0.901 |
endorsement | -0.868 |
provocation | -0.842 |
approval | -0.824 |
� | -0.813 |
Token | ulo |
Token ID | 43348 |
Feature activation | -1.06e+00 |
Pos logit contributions | |
Bridgewater | +0.949 |
Unic | +0.922 |
Flavoring | +0.859 |
Reviewer | +0.842 |
wegian | +0.831 |
Neg logit contributions | |
TIT | -1.265 |
endorsement | -1.218 |
provocation | -1.187 |
approval | -1.143 |
� | -1.132 |
Token | : |
Token ID | 25 |
Feature activation | -5.92e-01 |
Pos logit contributions | |
Bridgewater | +0.632 |
Unic | +0.607 |
Flavoring | +0.565 |
wegian | +0.555 |
Reviewer | +0.552 |
Neg logit contributions | |
TIT | -0.840 |
endorsement | -0.809 |
provocation | -0.781 |
approval | -0.769 |
� | -0.752 |
Token | I |
Token ID | 314 |
Feature activation | -2.23e-01 |
Pos logit contributions | |
Bridgewater | +0.377 |
Unic | +0.364 |
Flavoring | +0.340 |
wegian | +0.330 |
Reviewer | +0.329 |
Neg logit contributions | |
TIT | -0.446 |
endorsement | -0.429 |
provocation | -0.414 |
approval | -0.406 |
� | -0.398 |
Token | think |
Token ID | 892 |
Feature activation | +1.95e+00 |
Pos logit contributions | |
Bridgewater | +0.146 |
Unic | +0.141 |
Flavoring | +0.132 |
wegian | +0.130 |
Reviewer | +0.129 |
Neg logit contributions | |
TIT | -0.164 |
endorsement | -0.157 |
provocation | -0.151 |
approval | -0.148 |
� | -0.145 |
Token | the |
Token ID | 262 |
Feature activation | -2.31e-01 |
Pos logit contributions | |
TIT | +1.634 |
endorsement | +1.591 |
provocation | +1.539 |
approval | +1.521 |
� | +1.468 |
Neg logit contributions | |
Bridgewater | -1.063 |
Unic | -1.015 |
Flavoring | -0.941 |
wegian | -0.932 |
Reviewer | -0.919 |
Token | possibility |
Token ID | 5885 |
Feature activation | -6.27e-01 |
Pos logit contributions | |
Bridgewater | +0.145 |
Unic | +0.139 |
Flavoring | +0.130 |
wegian | +0.127 |
Reviewer | +0.126 |
Neg logit contributions | |
TIT | -0.177 |
endorsement | -0.169 |
provocation | -0.164 |
approval | -0.160 |
� | -0.158 |
Token | of |
Token ID | 286 |
Feature activation | -2.81e-01 |
Pos logit contributions | |
Bridgewater | +0.346 |
Unic | +0.333 |
Flavoring | +0.308 |
wegian | +0.301 |
Reviewer | +0.298 |
Neg logit contributions | |
TIT | -0.523 |
endorsement | -0.508 |
provocation | -0.492 |
approval | -0.484 |
� | -0.473 |
Token | it |
Token ID | 340 |
Feature activation | -8.33e-03 |
Pos logit contributions | |
Bridgewater | +0.191 |
Unic | +0.185 |
Flavoring | +0.173 |
wegian | +0.170 |
Reviewer | +0.168 |
Neg logit contributions | |
TIT | -0.199 |
endorsement | -0.189 |
provocation | -0.183 |
approval | -0.178 |
� | -0.177 |
Token | not |
Token ID | 407 |
Feature activation | -2.61e-01 |
Pos logit contributions | |
Bridgewater | +0.006 |
Unic | +0.006 |
Flavoring | +0.005 |
wegian | +0.005 |
Reviewer | +0.005 |
Neg logit contributions | |
TIT | -0.006 |
endorsement | -0.005 |
provocation | -0.005 |
approval | -0.005 |
� | -0.005 |
Token | and |
Token ID | 290 |
Feature activation | +2.38e-01 |
Pos logit contributions | |
TIT | +0.041 |
endorsement | +0.039 |
provocation | +0.038 |
approval | +0.037 |
� | +0.036 |
Neg logit contributions | |
Bridgewater | -0.031 |
Unic | -0.030 |
Flavoring | -0.028 |
wegian | -0.027 |
Reviewer | -0.027 |
Token | force |
Token ID | 2700 |
Feature activation | +4.43e-01 |
Pos logit contributions | |
TIT | +0.187 |
endorsement | +0.180 |
provocation | +0.174 |
approval | +0.171 |
� | +0.167 |
Neg logit contributions | |
Bridgewater | -0.144 |
Unic | -0.138 |
Flavoring | -0.129 |
wegian | -0.126 |
Reviewer | -0.125 |
Token | people |
Token ID | 661 |
Feature activation | +6.24e-01 |
Pos logit contributions | |
TIT | +0.347 |
endorsement | +0.334 |
provocation | +0.323 |
approval | +0.318 |
� | +0.310 |
Neg logit contributions | |
Bridgewater | -0.268 |
Unic | -0.258 |
Flavoring | -0.241 |
wegian | -0.236 |
Reviewer | -0.235 |
Token | without |
Token ID | 1231 |
Feature activation | +1.49e+00 |
Pos logit contributions | |
TIT | +0.497 |
endorsement | +0.481 |
provocation | +0.465 |
approval | +0.459 |
� | +0.448 |
Neg logit contributions | |
Bridgewater | -0.366 |
Unic | -0.351 |
Flavoring | -0.330 |
wegian | -0.321 |
Reviewer | -0.320 |
Token | Ind |
Token ID | 1423 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
TIT | +1.113 |
endorsement | +1.073 |
provocation | +1.050 |
approval | +1.028 |
� | +0.988 |
Neg logit contributions | |
Bridgewater | -0.946 |
Unic | -0.907 |
Flavoring | -0.859 |
Reviewer | -0.846 |
wegian | -0.838 |
Token | om |
Token ID | 296 |
Feature activation | +2.64e+00 |
Pos logit contributions | |
TIT | +0.085 |
endorsement | +0.082 |
provocation | +0.080 |
approval | +0.079 |
� | +0.077 |
Neg logit contributions | |
Bridgewater | -0.059 |
Unic | -0.056 |
Flavoring | -0.052 |
wegian | -0.051 |
Reviewer | -0.051 |
Token | itable |
Token ID | 4674 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
TIT | +1.262 |
endorsement | +1.188 |
provocation | +1.123 |
approval | +1.091 |
� | +1.040 |
Neg logit contributions | |
Bridgewater | -2.397 |
Unic | -2.328 |
Flavoring | -2.254 |
Reviewer | -2.217 |
wegian | -2.203 |
Token | to |
Token ID | 284 |
Feature activation | +3.79e-01 |
Pos logit contributions | |
TIT | +1.030 |
endorsement | +0.997 |
provocation | +0.967 |
approval | +0.951 |
� | +0.919 |
Neg logit contributions | |
Bridgewater | -0.815 |
Unic | -0.780 |
Flavoring | -0.734 |
wegian | -0.723 |
Reviewer | -0.716 |
Token | take |
Token ID | 1011 |
Feature activation | +6.60e-02 |
Pos logit contributions | |
TIT | +0.273 |
endorsement | +0.261 |
provocation | +0.251 |
approval | +0.246 |
� | +0.241 |
Neg logit contributions | |
Bridgewater | -0.255 |
Unic | -0.246 |
Flavoring | -0.230 |
wegian | -0.226 |
Reviewer | -0.224 |
Token | more |
Token ID | 517 |
Feature activation | +3.11e-01 |
Pos logit contributions | |
TIT | +0.050 |
endorsement | +0.048 |
provocation | +0.047 |
approval | +0.046 |
� | +0.044 |
Neg logit contributions | |
Bridgewater | -0.041 |
Unic | -0.040 |
Flavoring | -0.038 |
wegian | -0.037 |
Reviewer | -0.037 |
Token | damage |
Token ID | 2465 |
Feature activation | +1.60e-01 |
Pos logit contributions | |
TIT | +0.222 |
endorsement | +0.214 |
provocation | +0.207 |
approval | +0.203 |
� | +0.196 |
Neg logit contributions | |
Bridgewater | -0.210 |
Unic | -0.202 |
Flavoring | -0.190 |
wegian | -0.188 |
Reviewer | -0.187 |
Token | . |
Token ID | 13 |
Feature activation | +9.63e-02 |
Pos logit contributions | |
TIT | +1.167 |
endorsement | +1.135 |
provocation | +1.096 |
approval | +1.080 |
� | +1.040 |
Neg logit contributions | |
Bridgewater | -1.024 |
Unic | -0.983 |
Flavoring | -0.937 |
Reviewer | -0.917 |
wegian | -0.914 |
Token |
|
Token ID | 198 |
Feature activation | +5.65e-01 |
Pos logit contributions | |
TIT | +0.074 |
endorsement | +0.071 |
provocation | +0.069 |
approval | +0.068 |
� | +0.066 |
Neg logit contributions | |
Bridgewater | -0.059 |
Unic | -0.057 |
Flavoring | -0.053 |
wegian | -0.053 |
Reviewer | -0.052 |
Token |
|
Token ID | 198 |
Feature activation | +6.66e-01 |
Pos logit contributions | |
TIT | +0.458 |
endorsement | +0.441 |
provocation | +0.428 |
approval | +0.419 |
� | +0.412 |
Neg logit contributions | |
Bridgewater | -0.328 |
Unic | -0.314 |
Flavoring | -0.291 |
wegian | -0.284 |
Reviewer | -0.283 |
Token | R |
Token ID | 49 |
Feature activation | +8.36e-01 |
Pos logit contributions | |
TIT | +0.482 |
endorsement | +0.457 |
provocation | +0.438 |
approval | +0.432 |
� | +0.424 |
Neg logit contributions | |
Bridgewater | -0.450 |
Unic | -0.433 |
Flavoring | -0.408 |
wegian | -0.399 |
Reviewer | -0.390 |
Token | end |
Token ID | 437 |
Feature activation | +6.70e-01 |
Pos logit contributions | |
TIT | +0.617 |
endorsement | +0.593 |
provocation | +0.568 |
approval | +0.564 |
� | +0.546 |
Neg logit contributions | |
Bridgewater | -0.546 |
Unic | -0.527 |
Flavoring | -0.493 |
wegian | -0.482 |
Reviewer | -0.477 |
Token | ering |
Token ID | 1586 |
Feature activation | +2.10e+00 |
Pos logit contributions | |
TIT | +0.382 |
endorsement | +0.366 |
provocation | +0.348 |
approval | +0.342 |
� | +0.331 |
Neg logit contributions | |
Bridgewater | -0.544 |
Unic | -0.528 |
Flavoring | -0.505 |
wegian | -0.495 |
Reviewer | -0.494 |
Token | of |
Token ID | 286 |
Feature activation | +8.22e-01 |
Pos logit contributions | |
TIT | +1.668 |
endorsement | +1.632 |
provocation | +1.584 |
approval | +1.567 |
� | +1.518 |
Neg logit contributions | |
Bridgewater | -1.213 |
Unic | -1.158 |
Flavoring | -1.093 |
wegian | -1.080 |
Reviewer | -1.072 |
Token | the |
Token ID | 262 |
Feature activation | +2.18e-01 |
Pos logit contributions | |
TIT | +0.640 |
endorsement | +0.618 |
provocation | +0.597 |
approval | +0.588 |
� | +0.571 |
Neg logit contributions | |
Bridgewater | -0.498 |
Unic | -0.478 |
Flavoring | -0.450 |
wegian | -0.440 |
Reviewer | -0.440 |
Token | northeast |
Token ID | 24287 |
Feature activation | -8.98e-02 |
Pos logit contributions | |
TIT | +0.161 |
endorsement | +0.155 |
provocation | +0.149 |
approval | +0.146 |
� | +0.143 |
Neg logit contributions | |
Bridgewater | -0.141 |
Unic | -0.136 |
Flavoring | -0.128 |
wegian | -0.125 |
Reviewer | -0.125 |
Token | corner |
Token ID | 5228 |
Feature activation | -8.02e-01 |
Pos logit contributions | |
Bridgewater | +0.062 |
Unic | +0.060 |
Flavoring | +0.057 |
wegian | +0.056 |
Reviewer | +0.056 |
Neg logit contributions | |
TIT | -0.062 |
endorsement | -0.060 |
provocation | -0.057 |
approval | -0.056 |
� | -0.055 |
Token | and |
Token ID | 290 |
Feature activation | +4.08e-01 |
Pos logit contributions | |
Bridgewater | +0.446 |
Unic | +0.427 |
Flavoring | +0.394 |
wegian | +0.385 |
Reviewer | +0.382 |
Neg logit contributions | |
TIT | -0.670 |
endorsement | -0.645 |
provocation | -0.625 |
approval | -0.614 |
� | -0.604 |
Token | with |
Token ID | 351 |
Feature activation | -1.18e+00 |
Pos logit contributions | |
TIT | +0.189 |
endorsement | +0.184 |
provocation | +0.181 |
approval | +0.178 |
� | +0.174 |
Neg logit contributions | |
Bridgewater | -0.087 |
Unic | -0.082 |
Flavoring | -0.076 |
wegian | -0.074 |
Reviewer | -0.074 |
Token | an |
Token ID | 281 |
Feature activation | -1.08e-01 |
Pos logit contributions | |
Bridgewater | +0.768 |
Unic | +0.741 |
Flavoring | +0.694 |
wegian | +0.680 |
Reviewer | +0.676 |
Neg logit contributions | |
TIT | -0.866 |
endorsement | -0.830 |
provocation | -0.801 |
approval | -0.785 |
� | -0.769 |
Token | acidic |
Token ID | 42987 |
Feature activation | +3.93e-01 |
Pos logit contributions | |
Bridgewater | +0.071 |
Unic | +0.068 |
Flavoring | +0.064 |
wegian | +0.063 |
Reviewer | +0.063 |
Neg logit contributions | |
TIT | -0.079 |
endorsement | -0.076 |
provocation | -0.073 |
approval | -0.072 |
� | -0.070 |
Token | substance |
Token ID | 9136 |
Feature activation | -3.22e-01 |
Pos logit contributions | |
TIT | +0.300 |
endorsement | +0.291 |
provocation | +0.282 |
approval | +0.276 |
� | +0.268 |
Neg logit contributions | |
Bridgewater | -0.245 |
Unic | -0.235 |
Flavoring | -0.220 |
wegian | -0.217 |
Reviewer | -0.215 |
Token | like |
Token ID | 588 |
Feature activation | -7.82e-01 |
Pos logit contributions | |
Bridgewater | +0.197 |
Unic | +0.189 |
Flavoring | +0.176 |
wegian | +0.173 |
Reviewer | +0.172 |
Neg logit contributions | |
TIT | -0.250 |
endorsement | -0.241 |
provocation | -0.234 |
approval | -0.229 |
� | -0.224 |
Token | vinegar |
Token ID | 26600 |
Feature activation | +1.74e+00 |
Pos logit contributions | |
Bridgewater | +0.527 |
Unic | +0.508 |
Flavoring | +0.478 |
wegian | +0.469 |
Reviewer | +0.467 |
Neg logit contributions | |
TIT | -0.560 |
endorsement | -0.536 |
provocation | -0.517 |
approval | -0.507 |
� | -0.495 |
Token | or |
Token ID | 393 |
Feature activation | +1.29e-01 |
Pos logit contributions | |
TIT | +1.301 |
endorsement | +1.257 |
provocation | +1.221 |
approval | +1.194 |
� | +1.161 |
Neg logit contributions | |
Bridgewater | -1.110 |
Unic | -1.061 |
Flavoring | -1.005 |
wegian | -0.992 |
Reviewer | -0.986 |
Token | wine |
Token ID | 8237 |
Feature activation | +1.51e+00 |
Pos logit contributions | |
TIT | +0.093 |
endorsement | +0.090 |
provocation | +0.087 |
approval | +0.085 |
� | +0.083 |
Neg logit contributions | |
Bridgewater | -0.086 |
Unic | -0.083 |
Flavoring | -0.078 |
wegian | -0.077 |
Reviewer | -0.077 |
Token | , |
Token ID | 11 |
Feature activation | +2.75e-03 |
Pos logit contributions | |
TIT | +1.085 |
endorsement | +1.045 |
provocation | +1.015 |
approval | +0.991 |
� | +0.961 |
Neg logit contributions | |
Bridgewater | -1.007 |
Unic | -0.963 |
Flavoring | -0.909 |
Reviewer | -0.894 |
wegian | -0.892 |
Token | leading |
Token ID | 3756 |
Feature activation | +2.23e-01 |
Pos logit contributions | |
TIT | +0.002 |
endorsement | +0.002 |
provocation | +0.002 |
approval | +0.002 |
� | +0.002 |
Neg logit contributions | |
Bridgewater | -0.002 |
Unic | -0.002 |
Flavoring | -0.001 |
wegian | -0.001 |
Reviewer | -0.001 |
Token | Kon |
Token ID | 17431 |
Feature activation | +7.47e-01 |
Pos logit contributions | |
TIT | +0.172 |
endorsement | +0.167 |
provocation | +0.162 |
approval | +0.160 |
Hait | +0.155 |
Neg logit contributions | |
Bridgewater | -0.136 |
Unic | -0.131 |
Flavoring | -0.124 |
wegian | -0.120 |
Reviewer | -0.120 |
Token | happen |
Token ID | 1645 |
Feature activation | +1.51e-01 |
Pos logit contributions | |
Bridgewater | +0.518 |
Unic | +0.498 |
Flavoring | +0.463 |
wegian | +0.455 |
Reviewer | +0.449 |
Neg logit contributions | |
TIT | -0.606 |
endorsement | -0.581 |
provocation | -0.560 |
approval | -0.548 |
� | -0.538 |
Token | to |
Token ID | 284 |
Feature activation | +1.85e-01 |
Pos logit contributions | |
TIT | +0.121 |
endorsement | +0.117 |
provocation | +0.113 |
approval | +0.112 |
� | +0.109 |
Neg logit contributions | |
Bridgewater | -0.088 |
Unic | -0.084 |
Flavoring | -0.079 |
Reviewer | -0.077 |
wegian | -0.077 |
Token | stay |
Token ID | 2652 |
Feature activation | +4.41e-01 |
Pos logit contributions | |
TIT | +0.141 |
endorsement | +0.136 |
provocation | +0.131 |
approval | +0.129 |
� | +0.126 |
Neg logit contributions | |
Bridgewater | -0.115 |
Unic | -0.111 |
Flavoring | -0.104 |
wegian | -0.102 |
Reviewer | -0.101 |
Token | among |
Token ID | 1871 |
Feature activation | +1.57e-01 |
Pos logit contributions | |
TIT | +0.328 |
endorsement | +0.316 |
provocation | +0.304 |
approval | +0.301 |
� | +0.293 |
Neg logit contributions | |
Bridgewater | -0.283 |
Unic | -0.273 |
Flavoring | -0.257 |
wegian | -0.252 |
Reviewer | -0.251 |
Token | the |
Token ID | 262 |
Feature activation | +2.12e-01 |
Pos logit contributions | |
TIT | +0.125 |
endorsement | +0.120 |
provocation | +0.116 |
approval | +0.115 |
� | +0.112 |
Neg logit contributions | |
Bridgewater | -0.093 |
Unic | -0.089 |
Flavoring | -0.083 |
wegian | -0.081 |
Reviewer | -0.081 |
Token | teams |
Token ID | 3466 |
Feature activation | +1.75e+00 |
Pos logit contributions | |
TIT | +0.178 |
endorsement | +0.172 |
provocation | +0.167 |
approval | +0.164 |
� | +0.161 |
Neg logit contributions | |
Bridgewater | -0.117 |
Unic | -0.111 |
Flavoring | -0.103 |
wegian | -0.101 |
Reviewer | -0.100 |
Token | fighting |
Token ID | 4330 |
Feature activation | -9.43e-01 |
Pos logit contributions | |
TIT | +1.421 |
endorsement | +1.385 |
approval | +1.333 |
provocation | +1.326 |
� | +1.286 |
Neg logit contributions | |
Bridgewater | -0.984 |
Unic | -0.941 |
Flavoring | -0.900 |
Reviewer | -0.871 |
wegian | -0.867 |
Token | for |
Token ID | 329 |
Feature activation | +2.92e-01 |
Pos logit contributions | |
Bridgewater | +0.478 |
Unic | +0.455 |
Flavoring | +0.425 |
wegian | +0.411 |
Reviewer | +0.407 |
Neg logit contributions | |
TIT | -0.830 |
endorsement | -0.805 |
provocation | -0.778 |
approval | -0.771 |
� | -0.753 |
Token | Central |
Token ID | 5694 |
Feature activation | +4.45e-01 |
Pos logit contributions | |
TIT | +0.216 |
endorsement | +0.209 |
provocation | +0.200 |
approval | +0.200 |
� | +0.191 |
Neg logit contributions | |
Bridgewater | -0.188 |
Unic | -0.181 |
Flavoring | -0.172 |
Reviewer | -0.168 |
wegian | -0.167 |
Token | Division |
Token ID | 7458 |
Feature activation | +4.98e-01 |
Pos logit contributions | |
TIT | +0.320 |
endorsement | +0.307 |
provocation | +0.295 |
approval | +0.292 |
� | +0.282 |
Neg logit contributions | |
Bridgewater | -0.299 |
Unic | -0.288 |
Flavoring | -0.272 |
wegian | -0.266 |
Reviewer | -0.266 |
Token | playoff |
Token ID | 13119 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
TIT | +0.308 |
endorsement | +0.293 |
provocation | +0.280 |
approval | +0.274 |
� | +0.267 |
Neg logit contributions | |
Bridgewater | -0.385 |
Unic | -0.373 |
Flavoring | -0.353 |
wegian | -0.347 |
Reviewer | -0.346 |
Token |
|
Token ID | 198 |
Feature activation | +1.66e-01 |
Pos logit contributions | |
TIT | +0.151 |
endorsement | +0.146 |
provocation | +0.141 |
approval | +0.139 |
� | +0.135 |
Neg logit contributions | |
Bridgewater | -0.115 |
Unic | -0.110 |
Flavoring | -0.103 |
wegian | -0.101 |
Reviewer | -0.101 |
Token |
|
Token ID | 198 |
Feature activation | +1.27e-01 |
Pos logit contributions | |
TIT | +0.136 |
endorsement | +0.132 |
provocation | +0.128 |
approval | +0.125 |
� | +0.123 |
Neg logit contributions | |
Bridgewater | -0.095 |
Unic | -0.091 |
Flavoring | -0.084 |
wegian | -0.082 |
Reviewer | -0.082 |
Token | void |
Token ID | 19382 |
Feature activation | -2.28e+00 |
Pos logit contributions | |
TIT | +0.106 |
endorsement | +0.102 |
provocation | +0.099 |
approval | +0.097 |
� | +0.096 |
Neg logit contributions | |
Bridgewater | -0.071 |
Unic | -0.068 |
Flavoring | -0.063 |
wegian | -0.061 |
Reviewer | -0.061 |
Token | vert |
Token ID | 9421 |
Feature activation | +9.86e-01 |
Pos logit contributions | |
Bridgewater | +1.445 |
Unic | +1.406 |
Flavoring | +1.289 |
wegian | +1.273 |
tc | +1.261 |
Neg logit contributions | |
TIT | -1.720 |
endorsement | -1.619 |
provocation | -1.559 |
approval | -1.526 |
� | -1.522 |
Token | ( |
Token ID | 357 |
Feature activation | +1.18e+00 |
Pos logit contributions | |
TIT | +0.781 |
endorsement | +0.753 |
provocation | +0.728 |
approval | +0.715 |
� | +0.699 |
Neg logit contributions | |
Bridgewater | -0.588 |
Unic | -0.563 |
Flavoring | -0.527 |
wegian | -0.513 |
Reviewer | -0.512 |
Token | in |
Token ID | 259 |
Feature activation | +2.62e-01 |
Pos logit contributions | |
TIT | +0.860 |
endorsement | +0.828 |
provocation | +0.794 |
approval | +0.786 |
� | +0.756 |
Neg logit contributions | |
Bridgewater | -0.781 |
Unic | -0.751 |
Flavoring | -0.716 |
Reviewer | -0.706 |
wegian | -0.692 |
Token | out |
Token ID | 448 |
Feature activation | -4.04e-01 |
Pos logit contributions | |
TIT | +0.206 |
endorsement | +0.198 |
provocation | +0.191 |
approval | +0.188 |
� | +0.184 |
Neg logit contributions | |
Bridgewater | -0.159 |
Unic | -0.153 |
Flavoring | -0.142 |
wegian | -0.139 |
Reviewer | -0.138 |
Token | app |
Token ID | 598 |
Feature activation | +1.29e+00 |
Pos logit contributions | |
Bridgewater | +0.261 |
Unic | +0.251 |
Flavoring | +0.234 |
wegian | +0.231 |
Reviewer | +0.228 |
Neg logit contributions | |
TIT | -0.301 |
endorsement | -0.288 |
provocation | -0.278 |
approval | -0.273 |
� | -0.268 |
Token | data |
Token ID | 7890 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
TIT | +1.021 |
endorsement | +0.980 |
provocation | +0.941 |
approval | +0.936 |
� | +0.900 |
Neg logit contributions | |
Bridgewater | -0.775 |
Unic | -0.740 |
Flavoring | -0.706 |
Reviewer | -0.685 |
wegian | -0.672 |
Token | _ |
Token ID | 62 |
Feature activation | -1.38e+00 |
Pos logit contributions | |
TIT | +0.089 |
endorsement | +0.086 |
provocation | +0.083 |
approval | +0.082 |
� | +0.080 |
Neg logit contributions | |
Bridgewater | -0.059 |
Unic | -0.056 |
Flavoring | -0.052 |
wegian | -0.051 |
Reviewer | -0.050 |
Token | full |
Token ID | 12853 |
Feature activation | +2.30e+00 |
Pos logit contributions | |
Bridgewater | +0.561 |
Unic | +0.528 |
Flavoring | +0.472 |
wegian | +0.457 |
Reviewer | +0.452 |
Neg logit contributions | |
TIT | -1.361 |
endorsement | -1.319 |
provocation | -1.284 |
approval | -1.266 |
� | -1.247 |
Token | dev |
Token ID | 1614 |
Feature activation | -1.32e-01 |
Pos logit contributions | |
Bridgewater | +0.088 |
Unic | +0.084 |
Flavoring | +0.078 |
wegian | +0.076 |
Reviewer | +0.075 |
Neg logit contributions | |
TIT | -0.124 |
endorsement | -0.119 |
provocation | -0.116 |
approval | -0.113 |
� | -0.112 |
Token | olution |
Token ID | 2122 |
Feature activation | -2.33e-01 |
Pos logit contributions | |
Bridgewater | +0.092 |
Unic | +0.089 |
Flavoring | +0.084 |
wegian | +0.082 |
Reviewer | +0.082 |
Neg logit contributions | |
TIT | -0.092 |
endorsement | -0.088 |
provocation | -0.084 |
approval | -0.083 |
� | -0.081 |
Token | of |
Token ID | 286 |
Feature activation | -3.22e-01 |
Pos logit contributions | |
Bridgewater | +0.131 |
Unic | +0.126 |
Flavoring | +0.117 |
wegian | +0.114 |
Reviewer | +0.113 |
Neg logit contributions | |
TIT | -0.192 |
endorsement | -0.185 |
provocation | -0.179 |
approval | -0.176 |
� | -0.173 |
Token | national |
Token ID | 2260 |
Feature activation | -1.66e-01 |
Pos logit contributions | |
Bridgewater | +0.210 |
Unic | +0.203 |
Flavoring | +0.190 |
wegian | +0.186 |
Reviewer | +0.185 |
Neg logit contributions | |
TIT | -0.237 |
endorsement | -0.228 |
provocation | -0.220 |
approval | -0.215 |
� | -0.211 |
Token | powers |
Token ID | 5635 |
Feature activation | +9.86e-01 |
Pos logit contributions | |
Bridgewater | +0.116 |
Unic | +0.112 |
Flavoring | +0.106 |
wegian | +0.104 |
Reviewer | +0.104 |
Neg logit contributions | |
TIT | -0.114 |
endorsement | -0.110 |
provocation | -0.105 |
approval | -0.104 |
� | -0.100 |
Token | to |
Token ID | 284 |
Feature activation | -2.39e-01 |
Pos logit contributions | |
TIT | +0.793 |
endorsement | +0.771 |
provocation | +0.744 |
approval | +0.736 |
� | +0.714 |
Neg logit contributions | |
Bridgewater | -0.574 |
Unic | -0.552 |
Flavoring | -0.514 |
wegian | -0.500 |
Reviewer | -0.498 |
Token | state |
Token ID | 1181 |
Feature activation | -5.17e-01 |
Pos logit contributions | |
Bridgewater | +0.141 |
Unic | +0.136 |
Flavoring | +0.125 |
wegian | +0.123 |
Reviewer | +0.122 |
Neg logit contributions | |
TIT | -0.191 |
endorsement | -0.183 |
provocation | -0.178 |
approval | -0.174 |
� | -0.172 |
Token | and |
Token ID | 290 |
Feature activation | -9.01e-02 |
Pos logit contributions | |
Bridgewater | +0.330 |
Unic | +0.318 |
Flavoring | +0.299 |
wegian | +0.293 |
Reviewer | +0.290 |
Neg logit contributions | |
TIT | -0.389 |
endorsement | -0.375 |
provocation | -0.361 |
approval | -0.356 |
� | -0.345 |
Token | local |
Token ID | 1957 |
Feature activation | +6.82e-01 |
Pos logit contributions | |
Bridgewater | +0.045 |
Unic | +0.043 |
Flavoring | +0.040 |
Reviewer | +0.039 |
wegian | +0.039 |
Neg logit contributions | |
TIT | -0.079 |
endorsement | -0.077 |
provocation | -0.075 |
approval | -0.074 |
� | -0.072 |
Token | powers |
Token ID | 5635 |
Feature activation | +4.79e-01 |
Pos logit contributions | |
TIT | +0.411 |
endorsement | +0.393 |
provocation | +0.375 |
approval | +0.369 |
� | +0.354 |
Neg logit contributions | |
Bridgewater | -0.536 |
Unic | -0.520 |
Flavoring | -0.497 |
wegian | -0.489 |
Reviewer | -0.484 |
Token | has |
Token ID | 468 |
Feature activation | -1.40e-01 |
Pos logit contributions | |
TIT | +0.386 |
endorsement | +0.375 |
provocation | +0.362 |
approval | +0.358 |
� | +0.348 |
Neg logit contributions | |
Bridgewater | -0.278 |
Unic | -0.267 |
Flavoring | -0.248 |
wegian | -0.242 |
Reviewer | -0.241 |
Token | that |
Token ID | 326 |
Feature activation | -6.79e-01 |
Pos logit contributions | |
TIT | +0.250 |
endorsement | +0.244 |
provocation | +0.236 |
approval | +0.233 |
� | +0.226 |
Neg logit contributions | |
Bridgewater | -0.162 |
Unic | -0.155 |
Flavoring | -0.145 |
wegian | -0.141 |
Reviewer | -0.140 |
Token | all |
Token ID | 477 |
Feature activation | +3.06e-01 |
Pos logit contributions | |
Bridgewater | +0.418 |
Unic | +0.402 |
Flavoring | +0.373 |
wegian | +0.366 |
Reviewer | +0.362 |
Neg logit contributions | |
TIT | -0.527 |
endorsement | -0.505 |
provocation | -0.488 |
approval | -0.479 |
� | -0.472 |
Token | senior |
Token ID | 4664 |
Feature activation | -1.43e+00 |
Pos logit contributions | |
TIT | +0.227 |
endorsement | +0.219 |
provocation | +0.211 |
approval | +0.208 |
� | +0.201 |
Neg logit contributions | |
Bridgewater | -0.197 |
Unic | -0.189 |
Flavoring | -0.179 |
Reviewer | -0.176 |
wegian | -0.174 |
Token | civil |
Token ID | 3026 |
Feature activation | -7.28e-01 |
Pos logit contributions | |
Bridgewater | +0.966 |
Unic | +0.933 |
Flavoring | +0.868 |
wegian | +0.855 |
Reviewer | +0.846 |
Neg logit contributions | |
TIT | -1.022 |
endorsement | -0.974 |
provocation | -0.939 |
approval | -0.918 |
� | -0.906 |
Token | servants |
Token ID | 17523 |
Feature activation | +1.66e-01 |
Pos logit contributions | |
Bridgewater | +0.439 |
Unic | +0.420 |
Flavoring | +0.392 |
Reviewer | +0.387 |
wegian | +0.382 |
Neg logit contributions | |
TIT | -0.575 |
endorsement | -0.556 |
provocation | -0.537 |
approval | -0.529 |
� | -0.511 |
Token | be |
Token ID | 307 |
Feature activation | +5.25e-01 |
Pos logit contributions | |
TIT | +0.128 |
endorsement | +0.123 |
provocation | +0.119 |
approval | +0.117 |
� | +0.115 |
Neg logit contributions | |
Bridgewater | -0.102 |
Unic | -0.098 |
Flavoring | -0.092 |
wegian | -0.090 |
Reviewer | -0.089 |
Token | bilingual |
Token ID | 48116 |
Feature activation | +9.44e-01 |
Pos logit contributions | |
TIT | +0.371 |
endorsement | +0.355 |
provocation | +0.342 |
approval | +0.335 |
� | +0.328 |
Neg logit contributions | |
Bridgewater | -0.359 |
Unic | -0.347 |
Flavoring | -0.326 |
wegian | -0.320 |
Reviewer | -0.318 |
Token | by |
Token ID | 416 |
Feature activation | +9.11e-02 |
Pos logit contributions | |
TIT | +0.738 |
endorsement | +0.715 |
provocation | +0.690 |
approval | +0.680 |
� | +0.658 |
Neg logit contributions | |
Bridgewater | -0.573 |
Unic | -0.547 |
Flavoring | -0.518 |
Reviewer | -0.505 |
wegian | -0.501 |
Token | 2020 |
Token ID | 12131 |
Feature activation | -6.37e-01 |
Pos logit contributions | |
TIT | +0.072 |
endorsement | +0.070 |
provocation | +0.067 |
approval | +0.066 |
� | +0.065 |
Neg logit contributions | |
Bridgewater | -0.054 |
Unic | -0.052 |
Flavoring | -0.048 |
wegian | -0.047 |
Reviewer | -0.047 |
Token | , |
Token ID | 11 |
Feature activation | -2.59e-01 |
Pos logit contributions | |
Bridgewater | +0.397 |
Unic | +0.382 |
Flavoring | +0.356 |
wegian | +0.349 |
Reviewer | +0.347 |
Neg logit contributions | |
TIT | -0.488 |
endorsement | -0.469 |
provocation | -0.453 |
approval | -0.444 |
� | -0.436 |
Token | says |
Token ID | 1139 |
Feature activation | +6.97e-01 |
Pos logit contributions | |
Bridgewater | +0.155 |
Unic | +0.149 |
Flavoring | +0.139 |
wegian | +0.136 |
Reviewer | +0.136 |
Neg logit contributions | |
TIT | -0.204 |
endorsement | -0.196 |
provocation | -0.190 |
approval | -0.187 |
� | -0.182 |
Token | attain |
Token ID | 18188 |
Feature activation | -1.41e+00 |
Pos logit contributions | |
TIT | +0.143 |
endorsement | +0.136 |
provocation | +0.131 |
approval | +0.128 |
� | +0.126 |
Neg logit contributions | |
Bridgewater | -0.142 |
Unic | -0.137 |
Flavoring | -0.129 |
wegian | -0.127 |
Reviewer | -0.126 |
Token | maximum |
Token ID | 5415 |
Feature activation | -3.40e-01 |
Pos logit contributions | |
Bridgewater | +0.907 |
Unic | +0.872 |
Flavoring | +0.813 |
wegian | +0.794 |
Reviewer | +0.790 |
Neg logit contributions | |
TIT | -1.055 |
endorsement | -1.005 |
provocation | -0.973 |
approval | -0.949 |
� | -0.942 |
Token | impact |
Token ID | 2928 |
Feature activation | +1.53e+00 |
Pos logit contributions | |
Bridgewater | +0.260 |
Unic | +0.252 |
Flavoring | +0.238 |
wegian | +0.234 |
Reviewer | +0.233 |
Neg logit contributions | |
TIT | -0.212 |
endorsement | -0.201 |
provocation | -0.192 |
approval | -0.187 |
� | -0.184 |
Token | by |
Token ID | 416 |
Feature activation | +5.35e-01 |
Pos logit contributions | |
TIT | +1.288 |
endorsement | +1.265 |
provocation | +1.218 |
approval | +1.205 |
� | +1.168 |
Neg logit contributions | |
Bridgewater | -0.825 |
Unic | -0.789 |
Flavoring | -0.737 |
wegian | -0.718 |
Reviewer | -0.712 |
Token | recognizing |
Token ID | 22650 |
Feature activation | -1.93e-02 |
Pos logit contributions | |
TIT | +0.335 |
endorsement | +0.320 |
provocation | +0.306 |
approval | +0.299 |
� | +0.291 |
Neg logit contributions | |
Bridgewater | -0.409 |
Unic | -0.396 |
Flavoring | -0.375 |
wegian | -0.369 |
Reviewer | -0.367 |
Token | that |
Token ID | 326 |
Feature activation | -1.19e-01 |
Pos logit contributions | |
Bridgewater | +0.011 |
Unic | +0.011 |
Flavoring | +0.010 |
wegian | +0.010 |
Reviewer | +0.010 |
Neg logit contributions | |
TIT | -0.015 |
endorsement | -0.015 |
provocation | -0.015 |
approval | -0.014 |
� | -0.014 |
Token | injury |
Token ID | 5095 |
Feature activation | +1.34e-01 |
Pos logit contributions | |
Bridgewater | +0.075 |
Unic | +0.072 |
Flavoring | +0.068 |
wegian | +0.067 |
Reviewer | +0.066 |
Neg logit contributions | |
TIT | -0.091 |
endorsement | -0.088 |
provocation | -0.085 |
approval | -0.083 |
� | -0.081 |
Token | and |
Token ID | 290 |
Feature activation | +2.11e-02 |
Pos logit contributions | |
TIT | +0.101 |
endorsement | +0.098 |
provocation | +0.095 |
approval | +0.093 |
� | +0.090 |
Neg logit contributions | |
Bridgewater | -0.085 |
Unic | -0.082 |
Flavoring | -0.077 |
wegian | -0.076 |
Reviewer | -0.075 |
Token | violence |
Token ID | 3685 |
Feature activation | -9.05e-03 |
Pos logit contributions | |
TIT | +0.012 |
endorsement | +0.011 |
provocation | +0.011 |
approval | +0.010 |
� | +0.010 |
Neg logit contributions | |
Bridgewater | -0.018 |
Unic | -0.017 |
Flavoring | -0.016 |
wegian | -0.016 |
Reviewer | -0.016 |
Token | prevention |
Token ID | 14196 |
Feature activation | +1.94e-01 |
Pos logit contributions | |
Bridgewater | +0.006 |
Unic | +0.006 |
Flavoring | +0.006 |
wegian | +0.005 |
Reviewer | +0.005 |
Neg logit contributions | |
TIT | -0.006 |
endorsement | -0.006 |
provocation | -0.006 |
approval | -0.006 |
� | -0.006 |
Token | are |
Token ID | 389 |
Feature activation | +8.88e-03 |
Pos logit contributions | |
TIT | +0.150 |
endorsement | +0.145 |
provocation | +0.140 |
approval | +0.137 |
� | +0.134 |
Neg logit contributions | |
Bridgewater | -0.119 |
Unic | -0.115 |
Flavoring | -0.107 |
wegian | -0.105 |
Reviewer | -0.105 |
Token | me |
Token ID | 502 |
Feature activation | -2.14e-01 |
Pos logit contributions | |
TIT | +0.069 |
endorsement | +0.067 |
provocation | +0.064 |
approval | +0.063 |
� | +0.062 |
Neg logit contributions | |
Bridgewater | -0.050 |
Unic | -0.048 |
Flavoring | -0.045 |
wegian | -0.044 |
Reviewer | -0.044 |
Token | to |
Token ID | 284 |
Feature activation | +5.31e-02 |
Pos logit contributions | |
Bridgewater | +0.109 |
Unic | +0.104 |
Flavoring | +0.097 |
wegian | +0.094 |
Reviewer | +0.093 |
Neg logit contributions | |
TIT | -0.187 |
endorsement | -0.182 |
provocation | -0.176 |
approval | -0.174 |
� | -0.169 |
Token | cash |
Token ID | 5003 |
Feature activation | -7.25e-01 |
Pos logit contributions | |
TIT | +0.037 |
endorsement | +0.035 |
provocation | +0.034 |
approval | +0.033 |
� | +0.033 |
Neg logit contributions | |
Bridgewater | -0.037 |
Unic | -0.036 |
Flavoring | -0.034 |
wegian | -0.033 |
Reviewer | -0.033 |
Token | out |
Token ID | 503 |
Feature activation | +7.09e-01 |
Pos logit contributions | |
Bridgewater | +0.458 |
Unic | +0.441 |
Flavoring | +0.410 |
wegian | +0.402 |
Reviewer | +0.400 |
Neg logit contributions | |
TIT | -0.551 |
endorsement | -0.528 |
provocation | -0.510 |
approval | -0.500 |
� | -0.491 |
Token | . |
Token ID | 13 |
Feature activation | -3.69e-02 |
Pos logit contributions | |
TIT | +0.569 |
endorsement | +0.552 |
provocation | +0.532 |
approval | +0.527 |
� | +0.513 |
Neg logit contributions | |
Bridgewater | -0.411 |
Unic | -0.394 |
Flavoring | -0.369 |
Reviewer | -0.362 |
wegian | -0.360 |
Token | But |
Token ID | 887 |
Feature activation | -2.05e-02 |
Pos logit contributions | |
Bridgewater | +0.024 |
Unic | +0.023 |
Flavoring | +0.021 |
wegian | +0.021 |
Reviewer | +0.021 |
Neg logit contributions | |
TIT | -0.028 |
endorsement | -0.026 |
provocation | -0.025 |
approval | -0.025 |
� | -0.025 |
Token | I |
Token ID | 314 |
Feature activation | -1.81e-01 |
Pos logit contributions | |
Bridgewater | +0.011 |
Unic | +0.011 |
Flavoring | +0.010 |
wegian | +0.010 |
Reviewer | +0.010 |
Neg logit contributions | |
TIT | -0.017 |
endorsement | -0.016 |
provocation | -0.016 |
approval | -0.016 |
� | -0.015 |
Token | didn |
Token ID | 1422 |
Feature activation | +5.66e-01 |
Pos logit contributions | |
Bridgewater | +0.125 |
Unic | +0.121 |
Flavoring | +0.113 |
wegian | +0.111 |
Reviewer | +0.111 |
Neg logit contributions | |
TIT | -0.127 |
endorsement | -0.121 |
provocation | -0.117 |
approval | -0.114 |
� | -0.112 |
Token | � |
Token ID | 447 |
Feature activation | -5.08e-01 |
Pos logit contributions | |
TIT | +0.407 |
endorsement | +0.391 |
provocation | +0.377 |
approval | +0.369 |
� | +0.360 |
Neg logit contributions | |
Bridgewater | -0.379 |
Unic | -0.366 |
Flavoring | -0.343 |
wegian | -0.335 |
Reviewer | -0.334 |
Token | � |
Token ID | 247 |
Feature activation | +2.58e-01 |
Pos logit contributions | |
Bridgewater | +0.309 |
Unic | +0.297 |
Flavoring | +0.276 |
wegian | +0.270 |
Reviewer | +0.269 |
Neg logit contributions | |
TIT | -0.397 |
endorsement | -0.382 |
provocation | -0.369 |
approval | -0.362 |
� | -0.355 |
Token | t |
Token ID | 83 |
Feature activation | +9.52e-02 |
Pos logit contributions | |
TIT | +0.254 |
endorsement | +0.247 |
provocation | +0.240 |
approval | +0.238 |
� | +0.233 |
Neg logit contributions | |
Bridgewater | -0.104 |
Unic | -0.098 |
Flavoring | -0.090 |
Reviewer | -0.084 |
wegian | -0.084 |
Token | in |
Token ID | 287 |
Feature activation | +2.41e-01 |
Pos logit contributions | |
TIT | +0.099 |
endorsement | +0.097 |
provocation | +0.090 |
approval | +0.089 |
Hait | +0.087 |
Neg logit contributions | |
Bridgewater | -0.134 |
Unic | -0.131 |
Flavoring | -0.126 |
wegian | -0.124 |
Reviewer | -0.123 |
Token | 2020 |
Token ID | 12131 |
Feature activation | -3.50e-01 |
Pos logit contributions | |
TIT | +0.192 |
endorsement | +0.186 |
provocation | +0.179 |
approval | +0.177 |
� | +0.172 |
Neg logit contributions | |
Bridgewater | -0.142 |
Unic | -0.136 |
Flavoring | -0.127 |
wegian | -0.125 |
Reviewer | -0.123 |
Token | after |
Token ID | 706 |
Feature activation | +5.86e-01 |
Pos logit contributions | |
Bridgewater | +0.230 |
Unic | +0.221 |
Flavoring | +0.206 |
Reviewer | +0.201 |
wegian | +0.201 |
Neg logit contributions | |
TIT | -0.259 |
endorsement | -0.246 |
provocation | -0.239 |
approval | -0.233 |
� | -0.230 |
Token | ' |
Token ID | 705 |
Feature activation | -2.09e-01 |
Pos logit contributions | |
TIT | +0.436 |
endorsement | +0.423 |
provocation | +0.405 |
approval | +0.399 |
� | +0.390 |
Neg logit contributions | |
Bridgewater | -0.374 |
Unic | -0.362 |
Flavoring | -0.348 |
wegian | -0.341 |
Reviewer | -0.331 |
Token | not |
Token ID | 1662 |
Feature activation | -3.98e-01 |
Pos logit contributions | |
Bridgewater | +0.132 |
Unic | +0.128 |
Flavoring | +0.122 |
wegian | +0.118 |
Reviewer | +0.116 |
Neg logit contributions | |
TIT | -0.159 |
endorsement | -0.152 |
provocation | -0.146 |
approval | -0.144 |
� | -0.141 |
Token | quite |
Token ID | 2407 |
Feature activation | -1.46e+00 |
Pos logit contributions | |
Bridgewater | +0.245 |
Unic | +0.238 |
Flavoring | +0.225 |
Reviewer | +0.219 |
wegian | +0.219 |
Neg logit contributions | |
TIT | -0.306 |
endorsement | -0.299 |
provocation | -0.286 |
approval | -0.282 |
� | -0.271 |
Token | even |
Token ID | 772 |
Feature activation | -2.77e-01 |
Pos logit contributions | |
Bridgewater | +0.888 |
Unic | +0.856 |
Flavoring | +0.801 |
wegian | +0.781 |
Reviewer | +0.776 |
Neg logit contributions | |
TIT | -1.139 |
endorsement | -1.102 |
provocation | -1.063 |
approval | -1.045 |
� | -1.018 |
Token | handed |
Token ID | 10158 |
Feature activation | +3.55e-01 |
Pos logit contributions | |
Bridgewater | +0.107 |
Unic | +0.102 |
Flavoring | +0.093 |
wegian | +0.088 |
Reviewer | +0.086 |
Neg logit contributions | |
TIT | -0.276 |
endorsement | -0.271 |
provocation | -0.265 |
approval | -0.260 |
� | -0.255 |
Token | ' |
Token ID | 6 |
Feature activation | +1.72e+00 |
Pos logit contributions | |
TIT | +0.366 |
endorsement | +0.355 |
provocation | +0.346 |
approval | +0.341 |
� | +0.336 |
Neg logit contributions | |
Bridgewater | -0.128 |
Unic | -0.120 |
Flavoring | -0.105 |
wegian | -0.101 |
Reviewer | -0.100 |
Token | 2016 |
Token ID | 1584 |
Feature activation | +6.70e-01 |
Pos logit contributions | |
TIT | +1.276 |
endorsement | +1.226 |
provocation | +1.181 |
approval | +1.159 |
� | +1.135 |
Neg logit contributions | |
Bridgewater | -1.111 |
Unic | -1.071 |
Flavoring | -1.003 |
wegian | -0.983 |
Reviewer | -0.977 |
Token | primary |
Token ID | 4165 |
Feature activation | +5.02e-01 |
Pos logit contributions | |
TIT | +0.517 |
endorsement | +0.466 |
provocation | +0.461 |
� | +0.449 |
approval | +0.448 |
Neg logit contributions | |
Bridgewater | -0.434 |
Unic | -0.414 |
Flavoring | -0.379 |
Reviewer | -0.375 |
wegian | -0.366 |
Token | que |
Token ID | 4188 |
Feature activation | -4.79e-01 |
Pos logit contributions | |
Bridgewater | +0.587 |
Unic | +0.564 |
Flavoring | +0.523 |
wegian | +0.520 |
vich | +0.512 |
Neg logit contributions | |
TIT | -0.827 |
endorsement | -0.796 |
provocation | -0.769 |
approval | -0.750 |
� | -0.744 |
Token | said |
Token ID | 531 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
Bridgewater | +0.233 |
Unic | +0.222 |
Flavoring | +0.203 |
wegian | +0.197 |
Reviewer | +0.196 |
Neg logit contributions | |
TIT | -0.432 |
endorsement | -0.418 |
provocation | -0.406 |
approval | -0.399 |
� | -0.393 |
Token | . |
Token ID | 13 |
Feature activation | -5.00e-01 |
Pos logit contributions | |
Bridgewater | +0.094 |
Unic | +0.090 |
Flavoring | +0.085 |
wegian | +0.083 |
Reviewer | +0.083 |
Neg logit contributions | |
TIT | -0.099 |
endorsement | -0.095 |
provocation | -0.091 |
approval | -0.089 |
� | -0.087 |
Token |
|
Token ID | 198 |
Feature activation | -2.82e-02 |
Pos logit contributions | |
Bridgewater | +0.339 |
Unic | +0.326 |
Flavoring | +0.307 |
wegian | +0.302 |
Reviewer | +0.300 |
Neg logit contributions | |
TIT | -0.357 |
endorsement | -0.341 |
provocation | -0.328 |
approval | -0.322 |
� | -0.315 |
Token |
|
Token ID | 198 |
Feature activation | +9.34e-02 |
Pos logit contributions | |
Bridgewater | +0.016 |
Unic | +0.015 |
Flavoring | +0.014 |
wegian | +0.014 |
Reviewer | +0.014 |
Neg logit contributions | |
TIT | -0.023 |
endorsement | -0.023 |
provocation | -0.022 |
approval | -0.021 |
� | -0.021 |
Token | The |
Token ID | 464 |
Feature activation | -3.66e-01 |
Pos logit contributions | |
TIT | +0.054 |
endorsement | +0.051 |
provocation | +0.048 |
approval | +0.047 |
� | +0.046 |
Neg logit contributions | |
Bridgewater | -0.076 |
Unic | -0.074 |
Flavoring | -0.070 |
wegian | -0.069 |
Reviewer | -0.068 |
Token | discovery |
Token ID | 9412 |
Feature activation | -8.55e-01 |
Pos logit contributions | |
Bridgewater | +0.244 |
Unic | +0.235 |
Flavoring | +0.220 |
wegian | +0.216 |
Reviewer | +0.215 |
Neg logit contributions | |
TIT | -0.266 |
endorsement | -0.254 |
provocation | -0.245 |
approval | -0.240 |
� | -0.235 |
Token | that |
Token ID | 326 |
Feature activation | -7.78e-01 |
Pos logit contributions | |
Bridgewater | +0.506 |
Unic | +0.486 |
Flavoring | +0.451 |
wegian | +0.442 |
Reviewer | +0.438 |
Neg logit contributions | |
TIT | -0.683 |
endorsement | -0.657 |
provocation | -0.635 |
approval | -0.624 |
� | -0.612 |
Token | Kepler |
Token ID | 40294 |
Feature activation | -1.83e+00 |
Pos logit contributions | |
Bridgewater | +0.446 |
Unic | +0.427 |
Flavoring | +0.398 |
wegian | +0.388 |
Reviewer | +0.386 |
Neg logit contributions | |
TIT | -0.634 |
endorsement | -0.612 |
provocation | -0.592 |
approval | -0.583 |
� | -0.570 |
Token | - |
Token ID | 12 |
Feature activation | -3.87e-01 |
Pos logit contributions | |
Bridgewater | +1.115 |
Unic | +1.068 |
Flavoring | +1.001 |
wegian | +0.977 |
Reviewer | +0.973 |
Neg logit contributions | |
TIT | -1.428 |
endorsement | -1.378 |
provocation | -1.331 |
approval | -1.309 |
� | -1.278 |
Token | 10 |
Token ID | 940 |
Feature activation | -1.44e+00 |
Pos logit contributions | |
Bridgewater | +0.187 |
Unic | +0.178 |
Flavoring | +0.162 |
wegian | +0.158 |
Reviewer | +0.157 |
Neg logit contributions | |
TIT | -0.351 |
endorsement | -0.339 |
provocation | -0.330 |
approval | -0.324 |
� | -0.319 |
Token | � |
Token ID | 251 |
Feature activation | -8.68e-02 |
Pos logit contributions | |
Bridgewater | +0.203 |
Unic | +0.196 |
Flavoring | +0.184 |
wegian | +0.181 |
Reviewer | +0.180 |
Neg logit contributions | |
TIT | -0.219 |
endorsement | -0.210 |
provocation | -0.202 |
approval | -0.198 |
� | -0.194 |
Token | Mr |
Token ID | 1770 |
Feature activation | +1.01e+00 |
Pos logit contributions | |
Bridgewater | +0.044 |
Unic | +0.042 |
Flavoring | +0.039 |
wegian | +0.038 |
Reviewer | +0.038 |
Neg logit contributions | |
TIT | -0.076 |
endorsement | -0.074 |
provocation | -0.071 |
approval | -0.070 |
� | -0.069 |
Token | . |
Token ID | 13 |
Feature activation | -4.39e-01 |
Pos logit contributions | |
TIT | +0.735 |
endorsement | +0.705 |
provocation | +0.680 |
approval | +0.664 |
� | +0.650 |
Neg logit contributions | |
Bridgewater | -0.668 |
Unic | -0.644 |
Flavoring | -0.601 |
wegian | -0.589 |
Reviewer | -0.586 |
Token | Key |
Token ID | 7383 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
Bridgewater | +0.322 |
Unic | +0.313 |
Flavoring | +0.297 |
wegian | +0.292 |
Reviewer | +0.292 |
Neg logit contributions | |
TIT | -0.284 |
endorsement | -0.272 |
provocation | -0.261 |
approval | -0.256 |
� | -0.249 |
Token | said |
Token ID | 531 |
Feature activation | +5.04e-01 |
Pos logit contributions | |
TIT | +1.075 |
endorsement | +1.050 |
provocation | +1.017 |
approval | +1.007 |
� | +0.991 |
Neg logit contributions | |
Bridgewater | -0.423 |
Unic | -0.401 |
Flavoring | -0.357 |
Reviewer | -0.344 |
wegian | -0.340 |
Token | . |
Token ID | 13 |
Feature activation | -5.82e-01 |
Pos logit contributions | |
TIT | +0.359 |
endorsement | +0.346 |
provocation | +0.333 |
approval | +0.327 |
� | +0.319 |
Neg logit contributions | |
Bridgewater | -0.339 |
Unic | -0.327 |
Flavoring | -0.308 |
wegian | -0.302 |
Reviewer | -0.301 |
Token |
|
Token ID | 198 |
Feature activation | -2.49e-01 |
Pos logit contributions | |
Bridgewater | +0.402 |
Unic | +0.388 |
Flavoring | +0.365 |
wegian | +0.359 |
Reviewer | +0.356 |
Neg logit contributions | |
TIT | -0.406 |
endorsement | -0.389 |
provocation | -0.374 |
approval | -0.367 |
� | -0.359 |
Token |
|
Token ID | 198 |
Feature activation | -4.93e-03 |
Pos logit contributions | |
Bridgewater | +0.138 |
Unic | +0.133 |
Flavoring | +0.122 |
Reviewer | +0.118 |
wegian | +0.118 |
Neg logit contributions | |
TIT | -0.207 |
endorsement | -0.201 |
provocation | -0.195 |
approval | -0.190 |
� | -0.187 |
Token | Today |
Token ID | 8888 |
Feature activation | +2.73e-01 |
Pos logit contributions | |
Bridgewater | +0.003 |
Unic | +0.003 |
Flavoring | +0.003 |
wegian | +0.003 |
Reviewer | +0.003 |
Neg logit contributions | |
TIT | -0.004 |
endorsement | -0.003 |
provocation | -0.003 |
approval | -0.003 |
� | -0.003 |
Token | , |
Token ID | 11 |
Feature activation | -2.34e-01 |
Pos logit contributions | |
TIT | +0.212 |
endorsement | +0.206 |
provocation | +0.198 |
approval | +0.196 |
� | +0.189 |
Neg logit contributions | |
Bridgewater | -0.166 |
Unic | -0.159 |
Flavoring | -0.150 |
wegian | -0.146 |
Reviewer | -0.144 |
Token | he |
Token ID | 339 |
Feature activation | -6.56e-02 |
Pos logit contributions | |
Bridgewater | +0.142 |
Unic | +0.136 |
Flavoring | +0.129 |
wegian | +0.126 |
Reviewer | +0.124 |
Neg logit contributions | |
TIT | -0.182 |
endorsement | -0.177 |
provocation | -0.169 |
approval | -0.168 |
� | -0.162 |
Token | reversed |
Token ID | 17687 |
Feature activation | +1.36e+00 |
Pos logit contributions | |
Bridgewater | +0.333 |
Unic | +0.319 |
Flavoring | +0.300 |
wegian | +0.292 |
Reviewer | +0.291 |
Neg logit contributions | |
TIT | -0.409 |
endorsement | -0.395 |
provocation | -0.381 |
approval | -0.374 |
� | -0.365 |
Token | course |
Token ID | 1781 |
Feature activation | +1.00e+00 |
Pos logit contributions | |
TIT | +1.116 |
endorsement | +1.106 |
approval | +1.059 |
provocation | +1.054 |
Hait | +1.012 |
Neg logit contributions | |
Bridgewater | -0.757 |
Unic | -0.727 |
Flavoring | -0.685 |
wegian | -0.670 |
Reviewer | -0.661 |
Token | in |
Token ID | 287 |
Feature activation | -2.70e-01 |
Pos logit contributions | |
TIT | +0.798 |
endorsement | +0.784 |
provocation | +0.753 |
approval | +0.750 |
� | +0.723 |
Neg logit contributions | |
Bridgewater | -0.586 |
Unic | -0.561 |
Flavoring | -0.528 |
wegian | -0.513 |
Reviewer | -0.510 |
Token | 2009 |
Token ID | 3717 |
Feature activation | +5.66e-01 |
Pos logit contributions | |
Bridgewater | +0.161 |
Unic | +0.155 |
Flavoring | +0.144 |
wegian | +0.142 |
Reviewer | +0.141 |
Neg logit contributions | |
TIT | -0.213 |
endorsement | -0.206 |
provocation | -0.199 |
approval | -0.196 |
� | -0.191 |
Token | , |
Token ID | 11 |
Feature activation | -4.63e-01 |
Pos logit contributions | |
TIT | +0.445 |
endorsement | +0.435 |
provocation | +0.418 |
approval | +0.416 |
� | +0.401 |
Neg logit contributions | |
Bridgewater | -0.338 |
Unic | -0.324 |
Flavoring | -0.305 |
wegian | -0.296 |
Reviewer | -0.294 |
Token | voting |
Token ID | 6709 |
Feature activation | -4.78e-01 |
Pos logit contributions | |
Bridgewater | +0.311 |
Unic | +0.300 |
Flavoring | +0.283 |
wegian | +0.278 |
Reviewer | +0.277 |
Neg logit contributions | |
TIT | -0.330 |
endorsement | -0.320 |
provocation | -0.307 |
approval | -0.303 |
� | -0.292 |
Token | to |
Token ID | 284 |
Feature activation | -3.19e-02 |
Pos logit contributions | |
Bridgewater | +0.268 |
Unic | +0.257 |
Flavoring | +0.240 |
wegian | +0.234 |
Reviewer | +0.231 |
Neg logit contributions | |
TIT | -0.395 |
endorsement | -0.385 |
provocation | -0.371 |
approval | -0.368 |
� | -0.357 |
Token | rein |
Token ID | 6865 |
Feature activation | -1.84e+00 |
Pos logit contributions | |
Bridgewater | +0.022 |
Unic | +0.021 |
Flavoring | +0.020 |
wegian | +0.019 |
Reviewer | +0.019 |
Neg logit contributions | |
TIT | -0.023 |
endorsement | -0.022 |
provocation | -0.021 |
approval | -0.021 |
� | -0.020 |
Token | state |
Token ID | 5219 |
Feature activation | -1.08e-01 |
Pos logit contributions | |
Bridgewater | +1.038 |
Unic | +0.995 |
Flavoring | +0.921 |
wegian | +0.900 |
Reviewer | +0.894 |
Neg logit contributions | |
TIT | -1.524 |
endorsement | -1.469 |
provocation | -1.422 |
approval | -1.398 |
� | -1.373 |
Token | Cuba |
Token ID | 14159 |
Feature activation | -6.34e-01 |
Pos logit contributions | |
Bridgewater | +0.067 |
Unic | +0.065 |
Flavoring | +0.062 |
wegian | +0.060 |
Reviewer | +0.060 |
Neg logit contributions | |
TIT | -0.083 |
endorsement | -0.081 |
provocation | -0.077 |
approval | -0.077 |
Hait | -0.074 |
Token | , |
Token ID | 11 |
Feature activation | -2.86e-01 |
Pos logit contributions | |
Bridgewater | +0.347 |
Unic | +0.328 |
Flavoring | +0.309 |
wegian | +0.301 |
Reviewer | +0.296 |
Neg logit contributions | |
TIT | -0.533 |
endorsement | -0.519 |
provocation | -0.501 |
approval | -0.497 |
� | -0.483 |
Token | Michael |
Token ID | 3899 |
Feature activation | -2.32e-01 |
Pos logit contributions | |
TIT | +0.965 |
endorsement | +0.937 |
provocation | +0.906 |
approval | +0.892 |
� | +0.863 |
Neg logit contributions | |
Bridgewater | -0.744 |
Unic | -0.718 |
Flavoring | -0.685 |
wegian | -0.667 |
Reviewer | -0.660 |
Token | Morse |
Token ID | 44049 |
Feature activation | -5.31e-01 |
Pos logit contributions | |
Bridgewater | +0.142 |
Unic | +0.137 |
Flavoring | +0.129 |
Reviewer | +0.127 |
wegian | +0.125 |
Neg logit contributions | |
TIT | -0.178 |
endorsement | -0.172 |
provocation | -0.167 |
approval | -0.164 |
� | -0.159 |
Token | to |
Token ID | 284 |
Feature activation | +7.03e-01 |
Pos logit contributions | |
Bridgewater | +0.311 |
Unic | +0.299 |
Flavoring | +0.279 |
wegian | +0.272 |
Reviewer | +0.272 |
Neg logit contributions | |
TIT | -0.427 |
endorsement | -0.412 |
provocation | -0.398 |
approval | -0.392 |
� | -0.383 |
Token | tie |
Token ID | 9839 |
Feature activation | +7.34e-01 |
Pos logit contributions | |
TIT | +0.493 |
endorsement | +0.473 |
provocation | +0.455 |
approval | +0.447 |
� | +0.436 |
Neg logit contributions | |
Bridgewater | -0.482 |
Unic | -0.466 |
Flavoring | -0.439 |
wegian | -0.431 |
Reviewer | -0.429 |
Token | and |
Token ID | 290 |
Feature activation | -3.05e-01 |
Pos logit contributions | |
TIT | +0.589 |
endorsement | +0.571 |
provocation | +0.551 |
approval | +0.545 |
� | +0.527 |
Neg logit contributions | |
Bridgewater | -0.425 |
Unic | -0.410 |
Flavoring | -0.388 |
wegian | -0.380 |
Reviewer | -0.376 |
Token | Ian |
Token ID | 12930 |
Feature activation | -1.44e+00 |
Pos logit contributions | |
Bridgewater | +0.194 |
Unic | +0.187 |
Flavoring | +0.175 |
wegian | +0.171 |
Reviewer | +0.170 |
Neg logit contributions | |
TIT | -0.229 |
endorsement | -0.220 |
provocation | -0.213 |
approval | -0.209 |
� | -0.204 |
Token | Desmond |
Token ID | 41045 |
Feature activation | +6.57e-01 |
Pos logit contributions | |
Bridgewater | +0.880 |
Unic | +0.845 |
Flavoring | +0.788 |
wegian | +0.771 |
Reviewer | +0.765 |
Neg logit contributions | |
TIT | -1.121 |
endorsement | -1.077 |
provocation | -1.041 |
approval | -1.021 |
� | -1.003 |
Token | to |
Token ID | 284 |
Feature activation | +2.24e-01 |
Pos logit contributions | |
TIT | +0.524 |
endorsement | +0.508 |
provocation | +0.490 |
approval | +0.483 |
� | +0.470 |
Neg logit contributions | |
Bridgewater | -0.386 |
Unic | -0.372 |
Flavoring | -0.351 |
Reviewer | -0.342 |
wegian | -0.341 |
Token | take |
Token ID | 1011 |
Feature activation | -1.64e-01 |
Pos logit contributions | |
TIT | +0.165 |
endorsement | +0.159 |
provocation | +0.153 |
approval | +0.150 |
� | +0.147 |
Neg logit contributions | |
Bridgewater | -0.146 |
Unic | -0.141 |
Flavoring | -0.132 |
wegian | -0.129 |
Reviewer | -0.129 |
Token | the |
Token ID | 262 |
Feature activation | +5.40e-01 |
Pos logit contributions | |
Bridgewater | +0.099 |
Unic | +0.095 |
Flavoring | +0.089 |
wegian | +0.087 |
Reviewer | +0.086 |
Neg logit contributions | |
TIT | -0.129 |
endorsement | -0.124 |
provocation | -0.120 |
approval | -0.118 |
� | -0.115 |
Token | lead |
Token ID | 1085 |
Feature activation | -3.42e-02 |
Pos logit contributions | |
TIT | +0.379 |
endorsement | +0.364 |
provocation | +0.350 |
approval | +0.343 |
� | +0.335 |
Neg logit contributions | |
Bridgewater | -0.372 |
Unic | -0.359 |
Flavoring | -0.338 |
wegian | -0.332 |
Reviewer | -0.330 |
Token | - |
Token ID | 532 |
Feature activation | -2.08e+00 |
Pos logit contributions | |
Bridgewater | +1.167 |
Unic | +1.108 |
Flavoring | +1.001 |
wegian | +0.972 |
Reviewer | +0.959 |
Neg logit contributions | |
TIT | -2.493 |
endorsement | -2.417 |
provocation | -2.347 |
approval | -2.314 |
� | -2.272 |
Token | December |
Token ID | 3426 |
Feature activation | -1.54e-01 |
Pos logit contributions | |
Bridgewater | +1.164 |
Unic | +1.113 |
Flavoring | +1.023 |
wegian | +0.994 |
Reviewer | +0.986 |
Neg logit contributions | |
TIT | -1.727 |
endorsement | -1.654 |
provocation | -1.611 |
approval | -1.571 |
� | -1.565 |
Token | 13 |
Token ID | 1511 |
Feature activation | +6.15e-01 |
Pos logit contributions | |
Bridgewater | +0.091 |
Unic | +0.087 |
Flavoring | +0.080 |
wegian | +0.078 |
Reviewer | +0.077 |
Neg logit contributions | |
TIT | -0.124 |
endorsement | -0.120 |
provocation | -0.116 |
approval | -0.113 |
� | -0.112 |
Token | , |
Token ID | 11 |
Feature activation | +9.84e-02 |
Pos logit contributions | |
TIT | +0.476 |
endorsement | +0.457 |
provocation | +0.440 |
approval | +0.434 |
� | +0.424 |
Neg logit contributions | |
Bridgewater | -0.379 |
Unic | -0.364 |
Flavoring | -0.341 |
wegian | -0.335 |
Reviewer | -0.331 |
Token | 2010 |
Token ID | 3050 |
Feature activation | -5.19e-01 |
Pos logit contributions | |
TIT | +0.075 |
endorsement | +0.072 |
provocation | +0.069 |
approval | +0.069 |
� | +0.067 |
Neg logit contributions | |
Bridgewater | -0.062 |
Unic | -0.059 |
Flavoring | -0.056 |
wegian | -0.055 |
Reviewer | -0.054 |
Token |
|
Token ID | 198 |
Feature activation | -2.96e+00 |
Pos logit contributions | |
Bridgewater | +0.298 |
Unic | +0.285 |
Flavoring | +0.263 |
Reviewer | +0.256 |
wegian | +0.255 |
Neg logit contributions | |
TIT | -0.424 |
endorsement | -0.410 |
provocation | -0.397 |
approval | -0.389 |
� | -0.383 |
Token |
|
Token ID | 198 |
Feature activation | -2.44e+00 |
Pos logit contributions | |
Bridgewater | +1.546 |
Unic | +1.474 |
Flavoring | +1.331 |
Reviewer | +1.304 |
vich | +1.286 |
Neg logit contributions | |
TIT | -2.542 |
endorsement | -2.472 |
provocation | -2.418 |
approval | -2.338 |
� | -2.328 |
Token | Subject |
Token ID | 19776 |
Feature activation | -3.27e+00 |
Pos logit contributions | |
Bridgewater | +1.573 |
Unic | +1.512 |
Flavoring | +1.415 |
Reviewer | +1.396 |
wegian | +1.389 |
Neg logit contributions | |
TIT | -1.808 |
endorsement | -1.744 |
provocation | -1.688 |
approval | -1.650 |
� | -1.624 |
Token | : |
Token ID | 25 |
Feature activation | -1.59e+00 |
Pos logit contributions | |
Bridgewater | +1.614 |
Unic | +1.556 |
Flavoring | +1.444 |
Reviewer | +1.407 |
vich | +1.349 |
Neg logit contributions | |
TIT | -2.870 |
endorsement | -2.787 |
provocation | -2.743 |
� | -2.645 |
approval | -2.640 |
Token | boys |
Token ID | 6510 |
Feature activation | -8.11e-01 |
Pos logit contributions | |
Bridgewater | +1.018 |
Unic | +0.980 |
Flavoring | +0.911 |
wegian | +0.894 |
Reviewer | +0.891 |
Neg logit contributions | |
TIT | -1.196 |
endorsement | -1.144 |
provocation | -1.104 |
approval | -1.082 |
� | -1.068 |
Token | beware |
Token ID | 40600 |
Feature activation | -2.37e-01 |
Pos logit contributions | |
Bridgewater | +0.514 |
Unic | +0.495 |
Flavoring | +0.465 |
wegian | +0.452 |
Reviewer | +0.451 |
Neg logit contributions | |
TIT | -0.613 |
endorsement | -0.590 |
provocation | -0.569 |
approval | -0.559 |
� | -0.545 |
Token | r |
Token ID | 374 |
Feature activation | +1.55e-01 |
Pos logit contributions | |
Bridgewater | +0.692 |
Unic | +0.660 |
Flavoring | +0.605 |
wegian | +0.588 |
Reviewer | +0.583 |
Neg logit contributions | |
TIT | -1.238 |
endorsement | -1.199 |
provocation | -1.163 |
approval | -1.146 |
� | -1.122 |
Token | aves |
Token ID | 3080 |
Feature activation | -2.82e-01 |
Pos logit contributions | |
TIT | +0.112 |
endorsement | +0.108 |
provocation | +0.104 |
approval | +0.102 |
� | +0.100 |
Neg logit contributions | |
Bridgewater | -0.103 |
Unic | -0.099 |
Flavoring | -0.093 |
wegian | -0.091 |
Reviewer | -0.091 |
Token | the |
Token ID | 262 |
Feature activation | -3.31e-01 |
Pos logit contributions | |
Bridgewater | +0.176 |
Unic | +0.169 |
Flavoring | +0.158 |
wegian | +0.155 |
Reviewer | +0.153 |
Neg logit contributions | |
TIT | -0.216 |
endorsement | -0.208 |
provocation | -0.201 |
approval | -0.198 |
� | -0.192 |
Token | Los |
Token ID | 5401 |
Feature activation | +3.24e-01 |
Pos logit contributions | |
Bridgewater | +0.212 |
Unic | +0.204 |
Flavoring | +0.192 |
wegian | +0.187 |
Reviewer | +0.186 |
Neg logit contributions | |
TIT | -0.248 |
endorsement | -0.239 |
provocation | -0.231 |
approval | -0.226 |
� | -0.221 |
Token | Angeles |
Token ID | 5652 |
Feature activation | +8.71e-01 |
Pos logit contributions | |
TIT | +0.286 |
endorsement | +0.276 |
provocation | +0.268 |
approval | +0.264 |
� | +0.259 |
Neg logit contributions | |
Bridgewater | -0.164 |
Unic | -0.156 |
Flavoring | -0.143 |
wegian | -0.139 |
Reviewer | -0.139 |
Token | Times |
Token ID | 3782 |
Feature activation | -2.99e+00 |
Pos logit contributions | |
TIT | +0.528 |
endorsement | +0.509 |
provocation | +0.484 |
approval | +0.475 |
� | +0.458 |
Neg logit contributions | |
Bridgewater | -0.678 |
Unic | -0.656 |
Flavoring | -0.624 |
wegian | -0.611 |
Reviewer | -0.609 |
Token | . |
Token ID | 13 |
Feature activation | -2.03e-02 |
Pos logit contributions | |
Bridgewater | +1.828 |
Unic | +1.765 |
Flavoring | +1.631 |
wegian | +1.605 |
Reviewer | +1.599 |
Neg logit contributions | |
TIT | -2.332 |
endorsement | -2.213 |
provocation | -2.152 |
approval | -2.100 |
� | -2.090 |
Token | Ra |
Token ID | 7567 |
Feature activation | -2.85e-01 |
Pos logit contributions | |
Bridgewater | +0.012 |
Unic | +0.011 |
Flavoring | +0.010 |
wegian | +0.010 |
Reviewer | +0.010 |
Neg logit contributions | |
TIT | -0.016 |
endorsement | -0.016 |
provocation | -0.015 |
approval | -0.015 |
� | -0.015 |
Token | . |
Token ID | 13 |
Feature activation | +1.50e-01 |
Pos logit contributions | |
Bridgewater | +0.193 |
Unic | +0.187 |
Flavoring | +0.175 |
wegian | +0.172 |
Reviewer | +0.170 |
Neg logit contributions | |
TIT | -0.202 |
endorsement | -0.194 |
provocation | -0.187 |
approval | -0.184 |
� | -0.179 |
Token | One |
Token ID | 3198 |
Feature activation | +8.50e-01 |
Pos logit contributions | |
TIT | +0.109 |
endorsement | +0.103 |
provocation | +0.100 |
approval | +0.098 |
� | +0.096 |
Neg logit contributions | |
Bridgewater | -0.101 |
Unic | -0.098 |
Flavoring | -0.091 |
wegian | -0.090 |
Reviewer | -0.088 |
Token | , |
Token ID | 11 |
Feature activation | -2.88e-01 |
Pos logit contributions | |
TIT | +0.674 |
endorsement | +0.655 |
provocation | +0.633 |
approval | +0.623 |
� | +0.605 |
Neg logit contributions | |
Bridgewater | -0.504 |
Unic | -0.484 |
Flavoring | -0.452 |
wegian | -0.442 |
Reviewer | -0.439 |
Token | , |
Token ID | 11 |
Feature activation | -1.48e-01 |
Pos logit contributions | |
Bridgewater | +1.405 |
Unic | +1.343 |
Flavoring | +1.233 |
wegian | +1.205 |
Reviewer | +1.193 |
Neg logit contributions | |
TIT | -2.198 |
endorsement | -2.114 |
provocation | -2.054 |
approval | -2.012 |
� | -1.983 |
Token | as |
Token ID | 355 |
Feature activation | +5.45e-01 |
Pos logit contributions | |
Bridgewater | +0.092 |
Unic | +0.089 |
Flavoring | +0.083 |
wegian | +0.082 |
Reviewer | +0.081 |
Neg logit contributions | |
TIT | -0.113 |
endorsement | -0.109 |
provocation | -0.105 |
approval | -0.103 |
� | -0.101 |
Token | well |
Token ID | 880 |
Feature activation | -5.43e-01 |
Pos logit contributions | |
TIT | +0.432 |
endorsement | +0.420 |
provocation | +0.405 |
approval | +0.401 |
� | +0.388 |
Neg logit contributions | |
Bridgewater | -0.320 |
Unic | -0.307 |
Flavoring | -0.291 |
wegian | -0.280 |
Reviewer | -0.280 |
Token | as |
Token ID | 355 |
Feature activation | +2.29e-01 |
Pos logit contributions | |
Bridgewater | +0.264 |
Unic | +0.251 |
Flavoring | +0.230 |
wegian | +0.224 |
Reviewer | +0.222 |
Neg logit contributions | |
TIT | -0.490 |
endorsement | -0.475 |
provocation | -0.461 |
approval | -0.455 |
� | -0.446 |
Token | the |
Token ID | 262 |
Feature activation | +1.91e-01 |
Pos logit contributions | |
TIT | +0.173 |
endorsement | +0.167 |
provocation | +0.161 |
approval | +0.159 |
� | +0.154 |
Neg logit contributions | |
Bridgewater | -0.144 |
Unic | -0.138 |
Flavoring | -0.130 |
Reviewer | -0.127 |
wegian | -0.127 |
Token | AMC |
Token ID | 36239 |
Feature activation | -2.71e+00 |
Pos logit contributions | |
TIT | +0.147 |
endorsement | +0.142 |
provocation | +0.137 |
approval | +0.135 |
� | +0.131 |
Neg logit contributions | |
Bridgewater | -0.118 |
Unic | -0.114 |
Flavoring | -0.106 |
wegian | -0.104 |
Reviewer | -0.104 |
Token | - |
Token ID | 12 |
Feature activation | +3.80e-01 |
Pos logit contributions | |
Bridgewater | +1.690 |
Unic | +1.631 |
Flavoring | +1.493 |
wegian | +1.474 |
vich | +1.441 |
Neg logit contributions | |
TIT | -2.072 |
endorsement | -1.976 |
provocation | -1.925 |
approval | -1.871 |
� | -1.851 |
Token | owned |
Token ID | 11990 |
Feature activation | +3.52e-01 |
Pos logit contributions | |
TIT | +0.232 |
endorsement | +0.220 |
provocation | +0.210 |
approval | +0.205 |
� | +0.200 |
Neg logit contributions | |
Bridgewater | -0.296 |
Unic | -0.288 |
Flavoring | -0.274 |
wegian | -0.268 |
Reviewer | -0.266 |
Token | International |
Token ID | 4037 |
Feature activation | -1.31e+00 |
Pos logit contributions | |
TIT | +0.273 |
endorsement | +0.263 |
provocation | +0.254 |
approval | +0.249 |
� | +0.244 |
Neg logit contributions | |
Bridgewater | -0.216 |
Unic | -0.207 |
Flavoring | -0.194 |
wegian | -0.189 |
Reviewer | -0.188 |
Token | Film |
Token ID | 13741 |
Feature activation | -6.25e-01 |
Pos logit contributions | |
Bridgewater | +0.769 |
Unic | +0.751 |
Flavoring | +0.684 |
wegian | +0.666 |
Reviewer | +0.661 |
Neg logit contributions | |
TIT | -1.044 |
endorsement | -1.002 |
provocation | -0.972 |
approval | -0.950 |
� | -0.937 |
Token | Channel |
Token ID | 11102 |
Feature activation | -1.41e-01 |
Pos logit contributions | |
Bridgewater | +0.442 |
Unic | +0.429 |
Flavoring | +0.401 |
wegian | +0.395 |
Reviewer | +0.393 |
Neg logit contributions | |
TIT | -0.428 |
endorsement | -0.407 |
provocation | -0.393 |
approval | -0.381 |
� | -0.375 |
Token | closed |
Token ID | 4838 |
Feature activation | -6.96e-01 |
Pos logit contributions | |
Bridgewater | +0.763 |
Unic | +0.733 |
Flavoring | +0.678 |
wegian | +0.665 |
Reviewer | +0.658 |
Neg logit contributions | |
TIT | -1.057 |
endorsement | -1.017 |
provocation | -0.983 |
approval | -0.965 |
� | -0.949 |
Token | . |
Token ID | 13 |
Feature activation | +2.21e-01 |
Pos logit contributions | |
Bridgewater | +0.434 |
Unic | +0.418 |
Flavoring | +0.391 |
wegian | +0.382 |
Reviewer | +0.380 |
Neg logit contributions | |
TIT | -0.534 |
endorsement | -0.513 |
provocation | -0.495 |
approval | -0.487 |
� | -0.476 |
Token | So |
Token ID | 1406 |
Feature activation | +1.98e-01 |
Pos logit contributions | |
TIT | +0.162 |
endorsement | +0.155 |
provocation | +0.149 |
approval | +0.147 |
� | +0.144 |
Neg logit contributions | |
Bridgewater | -0.145 |
Unic | -0.139 |
Flavoring | -0.131 |
wegian | -0.129 |
Reviewer | -0.128 |
Token | all |
Token ID | 477 |
Feature activation | +4.30e-02 |
Pos logit contributions | |
TIT | +0.170 |
endorsement | +0.165 |
provocation | +0.160 |
approval | +0.157 |
� | +0.154 |
Neg logit contributions | |
Bridgewater | -0.104 |
Unic | -0.099 |
Flavoring | -0.093 |
wegian | -0.090 |
Reviewer | -0.089 |
Token | the |
Token ID | 262 |
Feature activation | -8.78e-01 |
Pos logit contributions | |
TIT | +0.035 |
endorsement | +0.034 |
provocation | +0.033 |
approval | +0.033 |
� | +0.032 |
Neg logit contributions | |
Bridgewater | -0.024 |
Unic | -0.023 |
Flavoring | -0.022 |
wegian | -0.021 |
Reviewer | -0.021 |
Token | time |
Token ID | 640 |
Feature activation | -2.39e+00 |
Pos logit contributions | |
Bridgewater | +0.514 |
Unic | +0.493 |
Flavoring | +0.459 |
wegian | +0.449 |
Reviewer | +0.446 |
Neg logit contributions | |
TIT | -0.705 |
endorsement | -0.680 |
provocation | -0.658 |
approval | -0.647 |
� | -0.633 |
Token | with |
Token ID | 351 |
Feature activation | -9.48e-02 |
Pos logit contributions | |
Bridgewater | +1.243 |
Unic | +1.189 |
Flavoring | +1.071 |
wegian | +1.057 |
Reviewer | +1.043 |
Neg logit contributions | |
TIT | -2.083 |
endorsement | -2.003 |
provocation | -1.943 |
approval | -1.908 |
� | -1.885 |
Token | the |
Token ID | 262 |
Feature activation | -2.53e-01 |
Pos logit contributions | |
Bridgewater | +0.054 |
Unic | +0.051 |
Flavoring | +0.049 |
wegian | +0.047 |
Reviewer | +0.047 |
Neg logit contributions | |
TIT | -0.078 |
endorsement | -0.075 |
provocation | -0.073 |
approval | -0.072 |
� | -0.070 |
Token | boys |
Token ID | 6510 |
Feature activation | +4.03e-01 |
Pos logit contributions | |
Bridgewater | +0.153 |
Unic | +0.147 |
Flavoring | +0.138 |
wegian | +0.135 |
Reviewer | +0.134 |
Neg logit contributions | |
TIT | -0.198 |
endorsement | -0.191 |
provocation | -0.185 |
approval | -0.182 |
� | -0.177 |
Token | they |
Token ID | 484 |
Feature activation | +6.51e-01 |
Pos logit contributions | |
TIT | +0.325 |
endorsement | +0.315 |
provocation | +0.304 |
approval | +0.300 |
� | +0.291 |
Neg logit contributions | |
Bridgewater | -0.233 |
Unic | -0.222 |
Flavoring | -0.210 |
Reviewer | -0.204 |
wegian | -0.203 |
Token | would |
Token ID | 561 |
Feature activation | +3.61e-01 |
Pos logit contributions | |
TIT | +0.492 |
endorsement | +0.475 |
provocation | +0.457 |
approval | +0.450 |
� | +0.440 |
Neg logit contributions | |
Bridgewater | -0.410 |
Unic | -0.395 |
Flavoring | -0.373 |
Reviewer | -0.363 |
wegian | -0.361 |
Token | Re |
Token ID | 797 |
Feature activation | -1.11e+00 |
Pos logit contributions | |
Bridgewater | +1.163 |
Unic | +1.123 |
Flavoring | +1.028 |
wegian | +0.986 |
Reviewer | +0.978 |
Neg logit contributions | |
TIT | -1.501 |
endorsement | -1.432 |
provocation | -1.380 |
approval | -1.347 |
� | -1.333 |
Token | iser |
Token ID | 5847 |
Feature activation | -1.09e-01 |
Pos logit contributions | |
Bridgewater | +0.704 |
Unic | +0.681 |
Flavoring | +0.640 |
wegian | +0.632 |
Reviewer | +0.629 |
Neg logit contributions | |
TIT | -0.837 |
endorsement | -0.788 |
provocation | -0.769 |
approval | -0.748 |
� | -0.744 |
Token |
|
Token ID | 198 |
Feature activation | -5.10e-01 |
Pos logit contributions | |
Bridgewater | +0.063 |
Unic | +0.061 |
Flavoring | +0.056 |
wegian | +0.055 |
Reviewer | +0.054 |
Neg logit contributions | |
TIT | -0.088 |
endorsement | -0.085 |
provocation | -0.083 |
approval | -0.081 |
� | -0.080 |
Token |
|
Token ID | 198 |
Feature activation | -4.87e-01 |
Pos logit contributions | |
Bridgewater | +0.284 |
Unic | +0.272 |
Flavoring | +0.249 |
Reviewer | +0.242 |
wegian | +0.242 |
Neg logit contributions | |
TIT | -0.424 |
endorsement | -0.411 |
provocation | -0.400 |
approval | -0.389 |
� | -0.383 |
Token | 1998 |
Token ID | 21113 |
Feature activation | -1.41e+00 |
Pos logit contributions | |
Bridgewater | +0.258 |
Unic | +0.247 |
Flavoring | +0.228 |
Reviewer | +0.224 |
wegian | +0.222 |
Neg logit contributions | |
TIT | -0.416 |
endorsement | -0.404 |
provocation | -0.394 |
approval | -0.385 |
� | -0.376 |
Token | Garry |
Token ID | 45354 |
Feature activation | -2.34e+00 |
Pos logit contributions | |
Bridgewater | +0.774 |
Unic | +0.741 |
Flavoring | +0.681 |
wegian | +0.664 |
Reviewer | +0.660 |
Neg logit contributions | |
TIT | -1.188 |
endorsement | -1.143 |
provocation | -1.106 |
approval | -1.086 |
� | -1.069 |
Token | Sh |
Token ID | 911 |
Feature activation | -2.38e+00 |
Pos logit contributions | |
Bridgewater | +1.727 |
Unic | +1.702 |
Flavoring | +1.577 |
vich | +1.543 |
wegian | +1.499 |
Neg logit contributions | |
TIT | -1.540 |
endorsement | -1.457 |
provocation | -1.396 |
approval | -1.359 |
� | -1.327 |
Token | and |
Token ID | 392 |
Feature activation | -1.43e+00 |
Pos logit contributions | |
Bridgewater | +1.024 |
Unic | +0.972 |
Flavoring | +0.863 |
vich | +0.859 |
wegian | +0.859 |
Neg logit contributions | |
TIT | -2.239 |
endorsement | -2.190 |
provocation | -2.154 |
approval | -2.090 |
� | -2.070 |
Token | ling |
Token ID | 1359 |
Feature activation | -1.29e+00 |
Pos logit contributions | |
Bridgewater | +0.664 |
Unic | +0.629 |
Flavoring | +0.566 |
wegian | +0.561 |
Reviewer | +0.548 |
Neg logit contributions | |
TIT | -1.322 |
endorsement | -1.282 |
provocation | -1.250 |
approval | -1.225 |
� | -1.204 |
Token | 1999 |
Token ID | 7358 |
Feature activation | -8.51e-01 |
Pos logit contributions | |
Bridgewater | +0.638 |
Unic | +0.608 |
Flavoring | +0.554 |
wegian | +0.543 |
Reviewer | +0.538 |
Neg logit contributions | |
TIT | -1.162 |
endorsement | -1.126 |
provocation | -1.090 |
approval | -1.078 |
� | -1.052 |
Token | John |
Token ID | 1757 |
Feature activation | -1.11e+00 |
Pos logit contributions | |
Bridgewater | +0.480 |
Unic | +0.460 |
Flavoring | +0.427 |
wegian | +0.417 |
Reviewer | +0.414 |
Neg logit contributions | |
TIT | -0.702 |
endorsement | -0.677 |
provocation | -0.655 |
approval | -0.645 |
� | -0.632 |