Token | sky |
Token ID | 6766 |
Feature activation | -1.99e+00 |
Pos logit contributions | |
crow | +0.101 |
Bad | +0.099 |
Leave | +0.098 |
eper | +0.092 |
vious | +0.092 |
Neg logit contributions | |
pron | -0.086 |
favourite | -0.081 |
masterpiece | -0.080 |
brand | -0.072 |
favorite | -0.071 |
Token | . |
Token ID | 13 |
Feature activation | +1.42e-01 |
Pos logit contributions | |
crow | +0.148 |
Bad | +0.142 |
Leave | +0.140 |
vious | +0.133 |
eper | +0.131 |
Neg logit contributions | |
pron | -0.159 |
favourite | -0.149 |
masterpiece | -0.149 |
brand | -0.136 |
favorite | -0.133 |
Token | The |
Token ID | 383 |
Feature activation | +2.50e-01 |
Pos logit contributions | |
pron | +0.011 |
favourite | +0.010 |
masterpiece | +0.010 |
favorite | +0.009 |
brand | +0.009 |
Neg logit contributions | |
crow | -0.011 |
Bad | -0.011 |
Leave | -0.011 |
eper | -0.010 |
vious | -0.010 |
Token | grass |
Token ID | 8701 |
Feature activation | -2.64e-01 |
Pos logit contributions | |
pron | +0.019 |
favourite | +0.018 |
masterpiece | +0.017 |
brand | +0.016 |
favorite | +0.016 |
Neg logit contributions | |
crow | -0.020 |
Bad | -0.019 |
Leave | -0.019 |
vious | -0.018 |
eper | -0.018 |
Token | was |
Token ID | 373 |
Feature activation | +1.74e+00 |
Pos logit contributions | |
crow | +0.023 |
Bad | +0.021 |
Leave | +0.021 |
eper | +0.020 |
Hey | +0.019 |
Neg logit contributions | |
pron | -0.019 |
masterpiece | -0.018 |
favourite | -0.017 |
brand | -0.016 |
favorite | -0.015 |
Token | freshly |
Token ID | 29026 |
Feature activation | +1.66e+01 |
Pos logit contributions | |
pron | +0.121 |
masterpiece | +0.114 |
favourite | +0.108 |
ail | +0.105 |
collection | +0.103 |
Neg logit contributions | |
Bad | -0.136 |
crow | -0.135 |
Leave | -0.133 |
eper | -0.124 |
vious | -0.124 |
Token | cut |
Token ID | 2005 |
Feature activation | +3.41e+00 |
Pos logit contributions | |
pron | +1.819 |
favourite | +1.762 |
masterpiece | +1.740 |
brand | +1.625 |
favorite | +1.624 |
Neg logit contributions | |
crow | -0.749 |
Bad | -0.724 |
Leave | -0.700 |
vious | -0.627 |
Stop | -0.614 |
Token | , |
Token ID | 11 |
Feature activation | -1.07e+00 |
Pos logit contributions | |
pron | +0.267 |
favourite | +0.257 |
masterpiece | +0.249 |
brand | +0.231 |
favorite | +0.230 |
Neg logit contributions | |
Bad | -0.256 |
crow | -0.252 |
Leave | -0.251 |
Stop | -0.233 |
ases | -0.232 |
Token | and |
Token ID | 290 |
Feature activation | +4.35e-01 |
Pos logit contributions | |
crow | +0.082 |
Bad | +0.081 |
Leave | +0.080 |
vious | +0.074 |
Stop | +0.074 |
Neg logit contributions | |
pron | -0.082 |
favourite | -0.078 |
masterpiece | -0.077 |
brand | -0.070 |
favorite | -0.069 |
Token | a |
Token ID | 257 |
Feature activation | +1.33e-01 |
Pos logit contributions | |
pron | +0.036 |
masterpiece | +0.034 |
favourite | +0.034 |
brand | +0.031 |
favorite | +0.030 |
Neg logit contributions | |
crow | -0.031 |
Bad | -0.030 |
Leave | -0.029 |
Stop | -0.028 |
eper | -0.027 |
Token | gentle |
Token ID | 10296 |
Feature activation | -3.38e+00 |
Pos logit contributions | |
pron | +0.008 |
favourite | +0.008 |
masterpiece | +0.008 |
favorite | +0.007 |
brand | +0.007 |
Neg logit contributions | |
Bad | -0.011 |
crow | -0.011 |
Leave | -0.011 |
Stop | -0.010 |
eper | -0.010 |
Token | went |
Token ID | 1816 |
Feature activation | +1.64e+00 |
Pos logit contributions | |
crow | +0.111 |
Bad | +0.109 |
Leave | +0.107 |
vious | +0.102 |
eper | +0.101 |
Neg logit contributions | |
pron | -0.081 |
favourite | -0.076 |
masterpiece | -0.075 |
brand | -0.067 |
favorite | -0.066 |
Token | home |
Token ID | 1363 |
Feature activation | +8.13e-01 |
Pos logit contributions | |
pron | +0.126 |
favourite | +0.119 |
masterpiece | +0.118 |
brand | +0.107 |
favorite | +0.106 |
Neg logit contributions | |
crow | -0.129 |
Bad | -0.125 |
Leave | -0.123 |
vious | -0.116 |
eper | -0.116 |
Token | and |
Token ID | 290 |
Feature activation | +3.80e-01 |
Pos logit contributions | |
pron | +0.070 |
favourite | +0.067 |
masterpiece | +0.066 |
favorite | +0.060 |
brand | +0.059 |
Neg logit contributions | |
crow | -0.051 |
Bad | -0.049 |
Leave | -0.049 |
eper | -0.047 |
vious | -0.045 |
Token | enjoyed |
Token ID | 8359 |
Feature activation | +4.43e+00 |
Pos logit contributions | |
pron | +0.028 |
favourite | +0.027 |
masterpiece | +0.027 |
favorite | +0.024 |
brand | +0.023 |
Neg logit contributions | |
crow | -0.029 |
Bad | -0.029 |
Leave | -0.027 |
eper | -0.026 |
vious | -0.025 |
Token | his |
Token ID | 465 |
Feature activation | -3.09e+00 |
Pos logit contributions | |
pron | +0.420 |
masterpiece | +0.396 |
favourite | +0.394 |
ail | +0.363 |
favorite | +0.361 |
Neg logit contributions | |
Bad | -0.248 |
Leave | -0.245 |
crow | -0.243 |
eper | -0.220 |
ases | -0.216 |
Token | freshly |
Token ID | 29026 |
Feature activation | +1.66e+01 |
Pos logit contributions | |
crow | +0.254 |
Bad | +0.253 |
Leave | +0.245 |
vious | +0.233 |
ases | +0.230 |
Neg logit contributions | |
pron | -0.198 |
masterpiece | -0.178 |
favourite | -0.174 |
ilda | -0.168 |
yt | -0.163 |
Token | caught |
Token ID | 4978 |
Feature activation | -3.39e+00 |
Pos logit contributions | |
pron | +1.079 |
favourite | +1.020 |
masterpiece | +1.013 |
brand | +0.920 |
favorite | +0.887 |
Neg logit contributions | |
crow | -1.460 |
Bad | -1.420 |
Leave | -1.396 |
eper | -1.324 |
vious | -1.321 |
Token | fish |
Token ID | 5916 |
Feature activation | -5.77e+00 |
Pos logit contributions | |
crow | +0.278 |
Bad | +0.272 |
Leave | +0.268 |
vious | +0.253 |
Stop | +0.251 |
Neg logit contributions | |
pron | -0.245 |
favourite | -0.233 |
masterpiece | -0.230 |
brand | -0.208 |
favorite | -0.206 |
Token | . |
Token ID | 13 |
Feature activation | -1.86e-02 |
Pos logit contributions | |
crow | +0.469 |
Bad | +0.459 |
Leave | +0.452 |
vious | +0.425 |
Stop | +0.425 |
Neg logit contributions | |
pron | -0.422 |
favourite | -0.401 |
masterpiece | -0.395 |
brand | -0.358 |
favorite | -0.354 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +1.06e+00 |
Pos logit contributions | |
crow | +0.001 |
Bad | +0.001 |
Leave | +0.001 |
vious | +0.001 |
eper | +0.001 |
Neg logit contributions | |
pron | -0.001 |
masterpiece | -0.001 |
favourite | -0.001 |
brand | -0.001 |
favorite | -0.001 |
Token | Once |
Token ID | 7454 |
Feature activation | -2.61e+00 |
Pos logit contributions | |
pron | +0.068 |
favourite | +0.066 |
masterpiece | +0.062 |
usual | +0.058 |
favorite | +0.058 |
Neg logit contributions | |
Bad | -0.090 |
Leave | -0.089 |
crow | -0.086 |
vious | -0.084 |
eper | -0.081 |
Token | a |
Token ID | 257 |
Feature activation | +5.69e-02 |
Pos logit contributions | |
crow | +0.486 |
Bad | +0.466 |
Leave | +0.449 |
vious | +0.444 |
eper | +0.438 |
Neg logit contributions | |
pron | -0.554 |
masterpiece | -0.512 |
favourite | -0.486 |
brand | -0.457 |
yt | -0.441 |
Token | cup |
Token ID | 6508 |
Feature activation | +4.95e+00 |
Pos logit contributions | |
masterpiece | +0.003 |
pron | +0.003 |
favourite | +0.003 |
oph | +0.003 |
ilda | +0.003 |
Neg logit contributions | |
crow | -0.004 |
Bad | -0.004 |
bully | -0.004 |
Leave | -0.004 |
brake | -0.004 |
Token | of |
Token ID | 286 |
Feature activation | -2.69e+00 |
Pos logit contributions | |
pron | +0.486 |
masterpiece | +0.480 |
favourite | +0.469 |
usual | +0.428 |
favorite | +0.426 |
Neg logit contributions | |
crow | -0.274 |
vious | -0.227 |
Leave | -0.223 |
Bad | -0.219 |
eper | -0.216 |
Token | warm |
Token ID | 5814 |
Feature activation | +4.35e+00 |
Pos logit contributions | |
crow | +0.230 |
Bad | +0.221 |
Leave | +0.211 |
vious | +0.206 |
Stop | +0.205 |
Neg logit contributions | |
pron | -0.185 |
masterpiece | -0.171 |
favourite | -0.165 |
ail | -0.149 |
brand | -0.147 |
Token | , |
Token ID | 11 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
pron | +0.324 |
masterpiece | +0.304 |
favourite | +0.298 |
usual | +0.266 |
brand | +0.266 |
Neg logit contributions | |
crow | -0.352 |
Leave | -0.319 |
Bad | -0.317 |
vious | -0.307 |
Stop | -0.302 |
Token | freshly |
Token ID | 29026 |
Feature activation | +1.66e+01 |
Pos logit contributions | |
Leave | +0.080 |
Bad | +0.080 |
crow | +0.079 |
vious | +0.074 |
hunters | +0.073 |
Neg logit contributions | |
pron | -0.070 |
masterpiece | -0.069 |
favourite | -0.065 |
ail | -0.064 |
collection | -0.060 |
Token | made |
Token ID | 925 |
Feature activation | +6.00e-01 |
Pos logit contributions | |
pron | +1.173 |
masterpiece | +1.130 |
favourite | +1.116 |
brand | +1.024 |
favorite | +0.982 |
Neg logit contributions | |
crow | -1.358 |
Bad | -1.327 |
Leave | -1.299 |
eper | -1.238 |
Stop | -1.223 |
Token | soup |
Token ID | 17141 |
Feature activation | +3.46e+00 |
Pos logit contributions | |
pron | +0.059 |
favourite | +0.057 |
masterpiece | +0.056 |
brand | +0.052 |
favorite | +0.052 |
Neg logit contributions | |
crow | -0.033 |
Bad | -0.032 |
Leave | -0.031 |
vious | -0.029 |
eper | -0.028 |
Token | and |
Token ID | 290 |
Feature activation | +3.87e-01 |
Pos logit contributions | |
pron | +0.283 |
masterpiece | +0.268 |
favourite | +0.260 |
usual | +0.239 |
brand | +0.236 |
Neg logit contributions | |
crow | -0.221 |
eper | -0.203 |
ases | -0.200 |
hunters | -0.198 |
Bad | -0.196 |
Token | a |
Token ID | 257 |
Feature activation | +7.45e-02 |
Pos logit contributions | |
pron | +0.028 |
masterpiece | +0.025 |
favourite | +0.025 |
ail | +0.023 |
favorite | +0.022 |
Neg logit contributions | |
crow | -0.030 |
Bad | -0.029 |
Leave | -0.029 |
hunters | -0.027 |
vious | -0.027 |
Token | cozy |
Token ID | 37438 |
Feature activation | +6.18e+00 |
Pos logit contributions | |
pron | +0.004 |
favourite | +0.004 |
masterpiece | +0.004 |
ilda | +0.004 |
ail | +0.003 |
Neg logit contributions | |
Bad | -0.006 |
crow | -0.006 |
Leave | -0.006 |
bully | -0.006 |
eper | -0.006 |
Token | saw |
Token ID | 2497 |
Feature activation | -3.68e-01 |
Pos logit contributions | |
pron | +0.419 |
masterpiece | +0.396 |
favourite | +0.394 |
favorite | +0.356 |
brand | +0.354 |
Neg logit contributions | |
Bad | -0.340 |
crow | -0.338 |
Leave | -0.328 |
eper | -0.319 |
vious | -0.308 |
Token | her |
Token ID | 607 |
Feature activation | +3.19e+00 |
Pos logit contributions | |
crow | +0.030 |
Bad | +0.029 |
Leave | +0.029 |
vious | +0.027 |
eper | +0.027 |
Neg logit contributions | |
pron | -0.028 |
favourite | -0.026 |
masterpiece | -0.026 |
brand | -0.023 |
favorite | -0.023 |
Token | mother |
Token ID | 2802 |
Feature activation | +6.35e+00 |
Pos logit contributions | |
pron | +0.212 |
masterpiece | +0.165 |
onna | +0.162 |
cakes | +0.149 |
favourite | +0.147 |
Neg logit contributions | |
Bad | -0.257 |
hunters | -0.256 |
eper | -0.255 |
crow | -0.249 |
bully | -0.237 |
Token | holding |
Token ID | 4769 |
Feature activation | +1.09e+00 |
Pos logit contributions | |
pron | +0.470 |
favourite | +0.445 |
masterpiece | +0.441 |
brand | +0.401 |
favorite | +0.392 |
Neg logit contributions | |
crow | -0.509 |
Bad | -0.498 |
Leave | -0.488 |
vious | -0.461 |
Stop | -0.457 |
Token | a |
Token ID | 257 |
Feature activation | +7.68e-02 |
Pos logit contributions | |
pron | +0.075 |
favourite | +0.073 |
masterpiece | +0.071 |
brand | +0.064 |
favorite | +0.064 |
Neg logit contributions | |
crow | -0.092 |
Bad | -0.091 |
Leave | -0.090 |
vious | -0.085 |
Stop | -0.084 |
Token | freshly |
Token ID | 29026 |
Feature activation | +1.66e+01 |
Pos logit contributions | |
masterpiece | +0.004 |
favourite | +0.004 |
pron | +0.004 |
ail | +0.004 |
ilda | +0.004 |
Neg logit contributions | |
Bad | -0.007 |
crow | -0.007 |
Leave | -0.006 |
bully | -0.006 |
Stop | -0.006 |
Token | - |
Token ID | 12 |
Feature activation | +1.72e+00 |
Pos logit contributions | |
pron | +1.250 |
favourite | +1.179 |
masterpiece | +1.164 |
brand | +1.060 |
favorite | +1.045 |
Neg logit contributions | |
crow | -1.308 |
Bad | -1.278 |
Leave | -1.257 |
vious | -1.184 |
eper | -1.174 |
Token | b |
Token ID | 65 |
Feature activation | -3.60e+00 |
Pos logit contributions | |
pron | +0.126 |
favourite | +0.119 |
masterpiece | +0.113 |
favorite | +0.105 |
brand | +0.103 |
Neg logit contributions | |
Bad | -0.138 |
Leave | -0.136 |
crow | -0.136 |
eper | -0.128 |
ases | -0.127 |
Token | aked |
Token ID | 4335 |
Feature activation | -7.33e+00 |
Pos logit contributions | |
crow | +0.282 |
Bad | +0.272 |
Leave | +0.267 |
eper | +0.263 |
ases | +0.254 |
Neg logit contributions | |
pron | -0.266 |
favourite | -0.253 |
masterpiece | -0.253 |
brand | -0.231 |
favorite | -0.224 |
Token | muff |
Token ID | 27563 |
Feature activation | +3.54e+00 |
Pos logit contributions | |
crow | +0.580 |
Bad | +0.575 |
Leave | +0.560 |
vious | +0.526 |
Stop | +0.525 |
Neg logit contributions | |
pron | -0.546 |
favourite | -0.520 |
masterpiece | -0.515 |
brand | -0.467 |
favorite | -0.459 |
Token | in |
Token ID | 259 |
Feature activation | -1.44e+00 |
Pos logit contributions | |
pron | +0.367 |
favourite | +0.348 |
masterpiece | +0.346 |
brand | +0.320 |
favorite | +0.319 |
Neg logit contributions | |
Bad | -0.182 |
crow | -0.182 |
Leave | -0.181 |
Stop | -0.159 |
vious | -0.152 |
Token | it |
Token ID | 340 |
Feature activation | +3.74e-01 |
Pos logit contributions | |
pron | +0.230 |
masterpiece | +0.214 |
favourite | +0.211 |
brand | +0.189 |
ail | +0.188 |
Neg logit contributions | |
crow | -0.227 |
Bad | -0.222 |
Leave | -0.219 |
Stop | -0.205 |
eper | -0.199 |
Token | when |
Token ID | 618 |
Feature activation | -6.78e+00 |
Pos logit contributions | |
pron | +0.032 |
masterpiece | +0.030 |
favourite | +0.029 |
icol | +0.027 |
complexion | +0.027 |
Neg logit contributions | |
crow | -0.023 |
Bad | -0.023 |
Leave | -0.022 |
vious | -0.021 |
hunters | -0.021 |
Token | the |
Token ID | 262 |
Feature activation | -1.49e+00 |
Pos logit contributions | |
crow | +0.480 |
Bad | +0.478 |
Leave | +0.468 |
eper | +0.429 |
Stop | +0.428 |
Neg logit contributions | |
pron | -0.555 |
masterpiece | -0.522 |
favourite | -0.515 |
brand | -0.473 |
favorite | -0.464 |
Token | snow |
Token ID | 6729 |
Feature activation | -8.63e-02 |
Pos logit contributions | |
crow | +0.123 |
Bad | +0.119 |
Leave | +0.117 |
vious | +0.110 |
Stop | +0.110 |
Neg logit contributions | |
pron | -0.106 |
favourite | -0.100 |
masterpiece | -0.099 |
brand | -0.090 |
favorite | -0.088 |
Token | was |
Token ID | 373 |
Feature activation | +1.61e+00 |
Pos logit contributions | |
crow | +0.007 |
Bad | +0.007 |
Leave | +0.007 |
eper | +0.007 |
Stop | +0.007 |
Neg logit contributions | |
pron | -0.006 |
favourite | -0.006 |
masterpiece | -0.006 |
brand | -0.005 |
favorite | -0.005 |
Token | freshly |
Token ID | 29026 |
Feature activation | +1.65e+01 |
Pos logit contributions | |
pron | +0.114 |
masterpiece | +0.110 |
favourite | +0.109 |
collection | +0.099 |
complexion | +0.097 |
Neg logit contributions | |
crow | -0.126 |
Bad | -0.125 |
Leave | -0.122 |
vious | -0.116 |
eper | -0.115 |
Token | fallen |
Token ID | 9292 |
Feature activation | +6.24e-01 |
Pos logit contributions | |
pron | +1.218 |
masterpiece | +1.145 |
favourite | +1.145 |
brand | +1.041 |
favorite | +1.012 |
Neg logit contributions | |
crow | -1.334 |
Bad | -1.301 |
Leave | -1.278 |
eper | -1.208 |
vious | -1.205 |
Token | . |
Token ID | 13 |
Feature activation | +2.59e-02 |
Pos logit contributions | |
pron | +0.052 |
favourite | +0.049 |
masterpiece | +0.049 |
favorite | +0.044 |
brand | +0.044 |
Neg logit contributions | |
crow | -0.044 |
Bad | -0.043 |
Leave | -0.042 |
vious | -0.040 |
eper | -0.039 |
Token | Today |
Token ID | 6288 |
Feature activation | +2.23e+00 |
Pos logit contributions | |
pron | +0.002 |
favourite | +0.002 |
masterpiece | +0.002 |
brand | +0.002 |
favorite | +0.002 |
Neg logit contributions | |
crow | -0.002 |
Bad | -0.002 |
Leave | -0.002 |
vious | -0.002 |
eper | -0.002 |
Token | , |
Token ID | 11 |
Feature activation | -1.23e+00 |
Pos logit contributions | |
pron | +0.177 |
masterpiece | +0.163 |
favourite | +0.160 |
brand | +0.148 |
favorite | +0.142 |
Neg logit contributions | |
crow | -0.175 |
Bad | -0.164 |
Leave | -0.157 |
vious | -0.151 |
Hey | -0.151 |
Token | he |
Token ID | 339 |
Feature activation | -7.79e-01 |
Pos logit contributions | |
crow | +0.096 |
Bad | +0.094 |
Leave | +0.092 |
eper | +0.086 |
Stop | +0.086 |
Neg logit contributions | |
pron | -0.094 |
masterpiece | -0.088 |
favourite | -0.088 |
brand | -0.079 |
favorite | -0.078 |
Token | the |
Token ID | 262 |
Feature activation | -1.19e+00 |
Pos logit contributions | |
crow | +0.090 |
Bad | +0.088 |
Leave | +0.086 |
vious | +0.082 |
eper | +0.081 |
Neg logit contributions | |
pron | -0.089 |
favourite | -0.085 |
masterpiece | -0.084 |
brand | -0.076 |
favorite | -0.076 |
Token | church |
Token ID | 4928 |
Feature activation | +1.40e+00 |
Pos logit contributions | |
crow | +0.096 |
Bad | +0.093 |
Leave | +0.092 |
ases | +0.087 |
vious | +0.087 |
Neg logit contributions | |
pron | -0.090 |
favourite | -0.085 |
masterpiece | -0.084 |
brand | -0.077 |
favorite | -0.076 |
Token | and |
Token ID | 290 |
Feature activation | +3.59e-01 |
Pos logit contributions | |
pron | +0.111 |
favourite | +0.103 |
masterpiece | +0.102 |
brand | +0.094 |
favorite | +0.092 |
Neg logit contributions | |
crow | -0.104 |
Bad | -0.098 |
Leave | -0.096 |
eper | -0.092 |
vious | -0.092 |
Token | noticed |
Token ID | 6810 |
Feature activation | +1.38e+00 |
Pos logit contributions | |
pron | +0.026 |
masterpiece | +0.024 |
favourite | +0.023 |
brand | +0.021 |
favorite | +0.021 |
Neg logit contributions | |
crow | -0.030 |
Bad | -0.029 |
Leave | -0.028 |
eper | -0.027 |
vious | -0.027 |
Token | a |
Token ID | 257 |
Feature activation | +1.16e-02 |
Pos logit contributions | |
pron | +0.102 |
favourite | +0.096 |
masterpiece | +0.095 |
brand | +0.086 |
favorite | +0.085 |
Neg logit contributions | |
crow | -0.111 |
Bad | -0.108 |
Leave | -0.106 |
vious | -0.100 |
eper | -0.100 |
Token | freshly |
Token ID | 29026 |
Feature activation | +1.65e+01 |
Pos logit contributions | |
pron | +0.001 |
favourite | +0.001 |
masterpiece | +0.001 |
favorite | +0.001 |
usual | +0.001 |
Neg logit contributions | |
crow | -0.001 |
Bad | -0.001 |
Leave | -0.001 |
Stop | -0.001 |
eper | -0.001 |
Token | dug |
Token ID | 18735 |
Feature activation | +1.64e+00 |
Pos logit contributions | |
pron | +1.261 |
favourite | +1.198 |
masterpiece | +1.185 |
brand | +1.081 |
favorite | +1.065 |
Neg logit contributions | |
crow | -1.287 |
Bad | -1.256 |
Leave | -1.233 |
vious | -1.156 |
eper | -1.154 |
Token | hole |
Token ID | 7604 |
Feature activation | -1.81e+00 |
Pos logit contributions | |
pron | +0.122 |
favourite | +0.116 |
masterpiece | +0.115 |
brand | +0.105 |
favorite | +0.103 |
Neg logit contributions | |
crow | -0.129 |
Bad | -0.128 |
Leave | -0.125 |
eper | -0.118 |
vious | -0.117 |
Token | in |
Token ID | 287 |
Feature activation | +1.54e+00 |
Pos logit contributions | |
crow | +0.131 |
Bad | +0.125 |
Leave | +0.124 |
eper | +0.117 |
vious | +0.117 |
Neg logit contributions | |
pron | -0.149 |
favourite | -0.140 |
masterpiece | -0.138 |
brand | -0.126 |
favorite | -0.125 |
Token | the |
Token ID | 262 |
Feature activation | -1.41e+00 |
Pos logit contributions | |
pron | +0.118 |
favourite | +0.114 |
masterpiece | +0.113 |
brand | +0.104 |
favorite | +0.101 |
Neg logit contributions | |
crow | -0.117 |
Bad | -0.113 |
Leave | -0.112 |
vious | -0.107 |
eper | -0.105 |
Token | ground |
Token ID | 2323 |
Feature activation | +1.99e-03 |
Pos logit contributions | |
crow | +0.113 |
Bad | +0.111 |
Leave | +0.109 |
vious | +0.103 |
eper | +0.102 |
Neg logit contributions | |
pron | -0.105 |
favourite | -0.099 |
masterpiece | -0.098 |
brand | -0.089 |
favorite | -0.088 |
Token | Tim |
Token ID | 5045 |
Feature activation | +4.32e-01 |
Pos logit contributions | |
pron | +0.110 |
masterpiece | +0.103 |
favourite | +0.102 |
brand | +0.093 |
favorite | +0.091 |
Neg logit contributions | |
crow | -0.091 |
Bad | -0.088 |
Leave | -0.085 |
eper | -0.081 |
Stop | -0.080 |
Token | my |
Token ID | 1820 |
Feature activation | -3.02e+00 |
Pos logit contributions | |
pron | +0.038 |
favourite | +0.037 |
masterpiece | +0.036 |
brand | +0.034 |
favorite | +0.034 |
Neg logit contributions | |
crow | -0.028 |
Bad | -0.028 |
Leave | -0.027 |
Stop | -0.025 |
eper | -0.025 |
Token | a |
Token ID | 257 |
Feature activation | -1.59e-02 |
Pos logit contributions | |
crow | +0.219 |
Bad | +0.213 |
Leave | +0.205 |
vious | +0.202 |
eper | +0.197 |
Neg logit contributions | |
pron | -0.241 |
masterpiece | -0.227 |
favourite | -0.222 |
brand | -0.205 |
collection | -0.196 |
Token | warm |
Token ID | 5814 |
Feature activation | +4.26e+00 |
Pos logit contributions | |
Bad | +0.001 |
crow | +0.001 |
bully | +0.001 |
vious | +0.001 |
Leave | +0.001 |
Neg logit contributions | |
masterpiece | -0.001 |
favourite | -0.001 |
yt | -0.001 |
icol | -0.001 |
ilda | -0.001 |
Token | , |
Token ID | 11 |
Feature activation | -1.09e+00 |
Pos logit contributions | |
masterpiece | +0.298 |
favourite | +0.295 |
pron | +0.290 |
usual | +0.264 |
yt | +0.264 |
Neg logit contributions | |
crow | -0.330 |
Bad | -0.304 |
Leave | -0.298 |
vious | -0.288 |
Hey | -0.278 |
Token | freshly |
Token ID | 29026 |
Feature activation | +1.65e+01 |
Pos logit contributions | |
Bad | +0.094 |
Leave | +0.094 |
crow | +0.089 |
vious | +0.088 |
Stop | +0.083 |
Neg logit contributions | |
masterpiece | -0.066 |
ail | -0.065 |
pron | -0.065 |
favourite | -0.062 |
collection | -0.059 |
Token | baked |
Token ID | 22979 |
Feature activation | +7.82e+00 |
Pos logit contributions | |
pron | +0.942 |
favourite | +0.897 |
masterpiece | +0.892 |
brand | +0.788 |
favorite | +0.759 |
Neg logit contributions | |
crow | -1.578 |
Bad | -1.554 |
Leave | -1.525 |
vious | -1.454 |
eper | -1.448 |
Token | cookie |
Token ID | 19751 |
Feature activation | +3.27e+00 |
Pos logit contributions | |
pron | +0.516 |
favourite | +0.487 |
masterpiece | +0.483 |
brand | +0.431 |
favorite | +0.423 |
Neg logit contributions | |
crow | -0.688 |
Bad | -0.668 |
Leave | -0.656 |
vious | -0.626 |
eper | -0.625 |
Token | . |
Token ID | 13 |
Feature activation | +3.49e-02 |
Pos logit contributions | |
pron | +0.245 |
favourite | +0.235 |
masterpiece | +0.233 |
brand | +0.215 |
favorite | +0.207 |
Neg logit contributions | |
crow | -0.250 |
Bad | -0.234 |
Leave | -0.226 |
eper | -0.225 |
ases | -0.220 |
Token |
|
Token ID | 198 |
Feature activation | +6.90e-01 |
Pos logit contributions | |
pron | +0.003 |
masterpiece | +0.003 |
favourite | +0.003 |
usual | +0.003 |
collection | +0.003 |
Neg logit contributions | |
crow | -0.002 |
Bad | -0.002 |
Leave | -0.002 |
Hey | -0.002 |
Stop | -0.002 |
Token |
|
Token ID | 198 |
Feature activation | +6.20e-01 |
Pos logit contributions | |
pron | +0.062 |
masterpiece | +0.062 |
favourite | +0.061 |
collection | +0.056 |
brand | +0.056 |
Neg logit contributions | |
crow | -0.042 |
Bad | -0.040 |
Leave | -0.039 |
Rex | -0.036 |
Hey | -0.036 |
Token | so |
Token ID | 523 |
Feature activation | +1.81e-01 |
Pos logit contributions | |
pron | +0.126 |
favourite | +0.117 |
masterpiece | +0.116 |
brand | +0.103 |
usual | +0.103 |
Neg logit contributions | |
crow | -0.123 |
Bad | -0.121 |
Leave | -0.118 |
eper | -0.114 |
vious | -0.112 |
Token | happy |
Token ID | 3772 |
Feature activation | -1.24e+00 |
Pos logit contributions | |
pron | +0.014 |
favourite | +0.013 |
masterpiece | +0.013 |
brand | +0.011 |
favorite | +0.011 |
Neg logit contributions | |
crow | -0.014 |
Bad | -0.014 |
Leave | -0.014 |
vious | -0.013 |
eper | -0.013 |
Token | to |
Token ID | 284 |
Feature activation | +1.69e+00 |
Pos logit contributions | |
crow | +0.103 |
Bad | +0.101 |
Leave | +0.099 |
vious | +0.094 |
eper | +0.093 |
Neg logit contributions | |
pron | -0.089 |
favourite | -0.082 |
masterpiece | -0.082 |
brand | -0.074 |
favorite | -0.072 |
Token | see |
Token ID | 766 |
Feature activation | +2.64e-01 |
Pos logit contributions | |
pron | +0.158 |
favourite | +0.151 |
masterpiece | +0.146 |
brand | +0.139 |
favorite | +0.136 |
Neg logit contributions | |
crow | -0.105 |
Leave | -0.101 |
Bad | -0.101 |
vious | -0.092 |
ases | -0.092 |
Token | a |
Token ID | 257 |
Feature activation | +1.51e-02 |
Pos logit contributions | |
pron | +0.020 |
favourite | +0.019 |
masterpiece | +0.019 |
brand | +0.017 |
favorite | +0.017 |
Neg logit contributions | |
crow | -0.021 |
Bad | -0.020 |
Leave | -0.020 |
vious | -0.019 |
eper | -0.019 |
Token | freshly |
Token ID | 29026 |
Feature activation | +1.64e+01 |
Pos logit contributions | |
favourite | +0.001 |
ail | +0.001 |
masterpiece | +0.001 |
pron | +0.001 |
usual | +0.001 |
Neg logit contributions | |
Bad | -0.001 |
bully | -0.001 |
crow | -0.001 |
Leave | -0.001 |
Stop | -0.001 |
Token | baked |
Token ID | 22979 |
Feature activation | +7.71e+00 |
Pos logit contributions | |
pron | +1.077 |
favourite | +1.017 |
masterpiece | +1.003 |
brand | +0.900 |
favorite | +0.882 |
Neg logit contributions | |
crow | -1.449 |
Bad | -1.417 |
Leave | -1.394 |
vious | -1.322 |
eper | -1.318 |
Token | bean |
Token ID | 26394 |
Feature activation | +5.37e+00 |
Pos logit contributions | |
pron | +0.445 |
favourite | +0.440 |
masterpiece | +0.419 |
brand | +0.378 |
favorite | +0.377 |
Neg logit contributions | |
crow | -0.733 |
Bad | -0.693 |
Leave | -0.678 |
vious | -0.660 |
eper | -0.659 |
Token | . |
Token ID | 13 |
Feature activation | -8.88e-02 |
Pos logit contributions | |
pron | +0.431 |
favourite | +0.408 |
masterpiece | +0.391 |
brand | +0.367 |
favorite | +0.365 |
Neg logit contributions | |
crow | -0.392 |
Bad | -0.362 |
Leave | -0.361 |
vious | -0.353 |
eper | -0.350 |
Token | After |
Token ID | 2293 |
Feature activation | +9.56e-01 |
Pos logit contributions | |
crow | +0.006 |
Bad | +0.006 |
Leave | +0.005 |
vious | +0.005 |
Stop | +0.005 |
Neg logit contributions | |
pron | -0.008 |
masterpiece | -0.008 |
favourite | -0.007 |
brand | -0.007 |
usual | -0.007 |
Token | that |
Token ID | 326 |
Feature activation | +2.82e+00 |
Pos logit contributions | |
pron | +0.077 |
favourite | +0.072 |
masterpiece | +0.070 |
favorite | +0.064 |
yt | +0.063 |
Neg logit contributions | |
Bad | -0.069 |
crow | -0.067 |
Leave | -0.066 |
hunters | -0.064 |
eper | -0.062 |
Token | to |
Token ID | 284 |
Feature activation | +1.60e+00 |
Pos logit contributions | |
crow | +0.161 |
Bad | +0.157 |
Leave | +0.155 |
vious | +0.146 |
eper | +0.146 |
Neg logit contributions | |
pron | -0.148 |
favourite | -0.137 |
masterpiece | -0.136 |
brand | -0.123 |
favorite | -0.121 |
Token | play |
Token ID | 711 |
Feature activation | -1.10e+00 |
Pos logit contributions | |
pron | +0.135 |
favourite | +0.127 |
masterpiece | +0.124 |
brand | +0.114 |
favorite | +0.114 |
Neg logit contributions | |
crow | -0.113 |
Bad | -0.108 |
Leave | -0.108 |
ases | -0.100 |
eper | -0.099 |
Token | and |
Token ID | 290 |
Feature activation | +1.75e-01 |
Pos logit contributions | |
crow | +0.085 |
Bad | +0.083 |
Leave | +0.081 |
vious | +0.076 |
eper | +0.076 |
Neg logit contributions | |
pron | -0.085 |
favourite | -0.081 |
masterpiece | -0.080 |
brand | -0.073 |
favorite | -0.072 |
Token | explore |
Token ID | 7301 |
Feature activation | +6.73e+00 |
Pos logit contributions | |
pron | +0.014 |
masterpiece | +0.012 |
favourite | +0.012 |
brand | +0.011 |
ail | +0.011 |
Neg logit contributions | |
crow | -0.012 |
Leave | -0.012 |
Bad | -0.012 |
bully | -0.011 |
Stop | -0.011 |
Token | the |
Token ID | 262 |
Feature activation | -1.83e+00 |
Pos logit contributions | |
pron | +0.546 |
favourite | +0.515 |
masterpiece | +0.511 |
brand | +0.465 |
favorite | +0.461 |
Neg logit contributions | |
crow | -0.494 |
Bad | -0.481 |
Leave | -0.472 |
vious | -0.441 |
eper | -0.439 |
Token | freshly |
Token ID | 29026 |
Feature activation | +1.64e+01 |
Pos logit contributions | |
crow | +0.159 |
Bad | +0.156 |
Leave | +0.153 |
Stop | +0.149 |
Hey | +0.144 |
Neg logit contributions | |
pron | -0.122 |
masterpiece | -0.105 |
favourite | -0.105 |
brand | -0.099 |
complexion | -0.095 |
Token | organized |
Token ID | 8389 |
Feature activation | +3.22e+00 |
Pos logit contributions | |
pron | +1.183 |
favourite | +1.115 |
masterpiece | +1.103 |
brand | +1.011 |
favorite | +0.982 |
Neg logit contributions | |
crow | -1.353 |
Bad | -1.297 |
Leave | -1.278 |
eper | -1.205 |
Stop | -1.203 |
Token | lawn |
Token ID | 21133 |
Feature activation | +5.71e+00 |
Pos logit contributions | |
pron | +0.207 |
favourite | +0.184 |
masterpiece | +0.180 |
brand | +0.173 |
onna | +0.160 |
Neg logit contributions | |
crow | -0.295 |
Bad | -0.269 |
Leave | -0.267 |
Stop | -0.261 |
vious | -0.256 |
Token | . |
Token ID | 13 |
Feature activation | -1.97e-01 |
Pos logit contributions | |
pron | +0.460 |
favourite | +0.427 |
masterpiece | +0.425 |
brand | +0.388 |
favorite | +0.381 |
Neg logit contributions | |
crow | -0.422 |
Bad | -0.403 |
Leave | -0.397 |
vious | -0.377 |
eper | -0.377 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +8.27e-01 |
Pos logit contributions | |
crow | +0.015 |
Bad | +0.015 |
Leave | +0.014 |
vious | +0.013 |
eper | +0.013 |
Neg logit contributions | |
pron | -0.015 |
favourite | -0.015 |
masterpiece | -0.015 |
brand | -0.013 |
favorite | -0.013 |
Token | Once |
Token ID | 7454 |
Feature activation | -2.89e+00 |
Pos logit contributions | |
pron | +0.066 |
favourite | +0.064 |
masterpiece | +0.062 |
favorite | +0.057 |
brand | +0.057 |
Neg logit contributions | |
Bad | -0.058 |
crow | -0.058 |
Leave | -0.058 |
vious | -0.054 |
eper | -0.053 |
Token | . |
Token ID | 13 |
Feature activation | -1.61e-01 |
Pos logit contributions | |
pron | +0.239 |
favourite | +0.233 |
masterpiece | +0.222 |
brand | +0.209 |
favorite | +0.208 |
Neg logit contributions | |
crow | -0.162 |
eper | -0.141 |
vious | -0.137 |
Bad | -0.134 |
ting | -0.134 |
Token | The |
Token ID | 383 |
Feature activation | -4.96e-02 |
Pos logit contributions | |
crow | +0.012 |
Leave | +0.011 |
Stop | +0.011 |
Bad | +0.011 |
ases | +0.011 |
Neg logit contributions | |
pron | -0.013 |
masterpiece | -0.012 |
favourite | -0.012 |
brand | -0.011 |
collection | -0.011 |
Token | sweet |
Token ID | 6029 |
Feature activation | +5.80e+00 |
Pos logit contributions | |
crow | +0.004 |
Bad | +0.004 |
eper | +0.004 |
Leave | +0.004 |
Stop | +0.004 |
Neg logit contributions | |
pron | -0.003 |
favourite | -0.003 |
onna | -0.003 |
masterpiece | -0.003 |
brand | -0.003 |
Token | smell |
Token ID | 8508 |
Feature activation | +4.87e+00 |
Pos logit contributions | |
pron | +0.470 |
favourite | +0.443 |
masterpiece | +0.432 |
brand | +0.415 |
usual | +0.407 |
Neg logit contributions | |
crow | -0.421 |
vious | -0.379 |
Leave | -0.366 |
Bad | -0.365 |
eper | -0.360 |
Token | of |
Token ID | 286 |
Feature activation | -2.84e+00 |
Pos logit contributions | |
pron | +0.411 |
favourite | +0.401 |
masterpiece | +0.399 |
icol | +0.391 |
collection | +0.371 |
Neg logit contributions | |
crow | -0.316 |
Bad | -0.293 |
Leave | -0.291 |
vious | -0.285 |
Hey | -0.280 |
Token | freshly |
Token ID | 29026 |
Feature activation | +1.64e+01 |
Pos logit contributions | |
crow | +0.172 |
Bad | +0.170 |
Leave | +0.159 |
Stop | +0.152 |
vious | +0.148 |
Neg logit contributions | |
pron | -0.254 |
favourite | -0.239 |
masterpiece | -0.232 |
ouse | -0.232 |
ail | -0.230 |
Token | baked |
Token ID | 22979 |
Feature activation | +7.66e+00 |
Pos logit contributions | |
pron | +0.749 |
favourite | +0.708 |
masterpiece | +0.688 |
brand | +0.627 |
yt | +0.578 |
Neg logit contributions | |
crow | -1.721 |
Bad | -1.683 |
Leave | -1.637 |
Stop | -1.617 |
eper | -1.610 |
Token | port |
Token ID | 2493 |
Feature activation | +3.07e+00 |
Pos logit contributions | |
pron | +0.534 |
favourite | +0.518 |
masterpiece | +0.498 |
icol | +0.466 |
brand | +0.466 |
Neg logit contributions | |
crow | -0.613 |
Bad | -0.557 |
eper | -0.552 |
Leave | -0.547 |
vious | -0.533 |
Token | filled |
Token ID | 5901 |
Feature activation | +4.18e+00 |
Pos logit contributions | |
pron | +0.262 |
favourite | +0.244 |
masterpiece | +0.235 |
favorite | +0.221 |
complexion | +0.219 |
Neg logit contributions | |
crow | -0.203 |
Bad | -0.179 |
Leave | -0.178 |
eper | -0.176 |
vious | -0.174 |
Token | the |
Token ID | 262 |
Feature activation | -1.31e+00 |
Pos logit contributions | |
pron | +0.373 |
favourite | +0.349 |
masterpiece | +0.346 |
brand | +0.331 |
collection | +0.319 |
Neg logit contributions | |
crow | -0.259 |
Bad | -0.252 |
Leave | -0.232 |
Stop | -0.231 |
eper | -0.229 |
Token | house |
Token ID | 2156 |
Feature activation | -3.53e-01 |
Pos logit contributions | |
crow | +0.116 |
Bad | +0.112 |
Leave | +0.109 |
vious | +0.106 |
Stop | +0.105 |
Neg logit contributions | |
pron | -0.086 |
favourite | -0.079 |
masterpiece | -0.077 |
brand | -0.070 |
usual | -0.068 |
Token | he |
Token ID | 339 |
Feature activation | -9.06e-01 |
Pos logit contributions | |
crow | +0.115 |
Leave | +0.111 |
Bad | +0.110 |
ases | +0.105 |
vious | +0.105 |
Neg logit contributions | |
pron | -0.103 |
favourite | -0.099 |
masterpiece | -0.095 |
brand | -0.088 |
favorite | -0.087 |
Token | saw |
Token ID | 2497 |
Feature activation | -6.66e-01 |
Pos logit contributions | |
crow | +0.065 |
Bad | +0.064 |
Leave | +0.063 |
Hey | +0.061 |
vious | +0.061 |
Neg logit contributions | |
pron | -0.072 |
favourite | -0.070 |
masterpiece | -0.068 |
brand | -0.062 |
favorite | -0.061 |
Token | a |
Token ID | 257 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
crow | +0.053 |
Bad | +0.052 |
Leave | +0.050 |
vious | +0.048 |
eper | +0.048 |
Neg logit contributions | |
pron | -0.051 |
favourite | -0.048 |
masterpiece | -0.047 |
brand | -0.043 |
favorite | -0.042 |
Token | basket |
Token ID | 7988 |
Feature activation | +6.23e+00 |
Pos logit contributions | |
Bad | +0.011 |
crow | +0.011 |
Leave | +0.011 |
Stop | +0.010 |
bully | +0.010 |
Neg logit contributions | |
favourite | -0.009 |
pron | -0.009 |
masterpiece | -0.009 |
favorite | -0.008 |
ail | -0.008 |
Token | of |
Token ID | 286 |
Feature activation | -3.00e+00 |
Pos logit contributions | |
pron | +0.487 |
favourite | +0.462 |
masterpiece | +0.457 |
brand | +0.417 |
favorite | +0.412 |
Neg logit contributions | |
crow | -0.474 |
Bad | -0.462 |
Leave | -0.454 |
vious | -0.427 |
eper | -0.425 |
Token | freshly |
Token ID | 29026 |
Feature activation | +1.63e+01 |
Pos logit contributions | |
crow | +0.284 |
Bad | +0.278 |
Leave | +0.273 |
vious | +0.264 |
Stop | +0.261 |
Neg logit contributions | |
pron | -0.176 |
masterpiece | -0.162 |
favourite | -0.161 |
brand | -0.142 |
favorite | -0.137 |
Token | baked |
Token ID | 22979 |
Feature activation | +7.54e+00 |
Pos logit contributions | |
pron | +0.951 |
favourite | +0.898 |
masterpiece | +0.895 |
brand | +0.812 |
favorite | +0.761 |
Neg logit contributions | |
crow | -1.539 |
Bad | -1.496 |
Leave | -1.471 |
Stop | -1.412 |
vious | -1.407 |
Token | bread |
Token ID | 8509 |
Feature activation | -1.15e+00 |
Pos logit contributions | |
pron | +0.555 |
favourite | +0.528 |
masterpiece | +0.525 |
brand | +0.484 |
favorite | +0.465 |
Neg logit contributions | |
crow | -0.607 |
Bad | -0.563 |
Leave | -0.551 |
eper | -0.541 |
vious | -0.540 |
Token | . |
Token ID | 13 |
Feature activation | -2.77e-01 |
Pos logit contributions | |
crow | +0.079 |
Bad | +0.071 |
eper | +0.071 |
Leave | +0.069 |
ases | +0.069 |
Neg logit contributions | |
pron | -0.098 |
favourite | -0.091 |
masterpiece | -0.091 |
brand | -0.084 |
icol | -0.081 |
Token | It |
Token ID | 632 |
Feature activation | -3.99e-01 |
Pos logit contributions | |
crow | +0.015 |
Bad | +0.014 |
Leave | +0.014 |
vious | +0.013 |
Stop | +0.013 |
Neg logit contributions | |
pron | -0.028 |
favourite | -0.027 |
masterpiece | -0.027 |
brand | -0.025 |
favorite | -0.024 |
Token | smelled |
Token ID | 37609 |
Feature activation | +5.65e+00 |
Pos logit contributions | |
crow | +0.024 |
Hey | +0.024 |
Bad | +0.024 |
Leave | +0.022 |
vious | +0.022 |
Neg logit contributions | |
pron | -0.035 |
masterpiece | -0.032 |
favourite | -0.032 |
complexion | -0.031 |
icol | -0.030 |
Token | crown |
Token ID | 12389 |
Feature activation | +4.53e+00 |
Pos logit contributions | |
pron | +0.102 |
icol | +0.086 |
favourite | +0.084 |
ilda | +0.082 |
oph | +0.081 |
Neg logit contributions | |
crow | -0.099 |
eper | -0.094 |
hunters | -0.093 |
Leave | -0.092 |
vious | -0.092 |
Token | and |
Token ID | 290 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
pron | +0.380 |
favourite | +0.342 |
masterpiece | +0.339 |
icol | +0.334 |
porch | +0.315 |
Neg logit contributions | |
crow | -0.264 |
vious | -0.263 |
eper | -0.250 |
Leave | -0.249 |
Bad | -0.244 |
Token | playing |
Token ID | 2712 |
Feature activation | -9.93e-01 |
Pos logit contributions | |
pron | +0.008 |
favourite | +0.007 |
masterpiece | +0.007 |
brand | +0.006 |
ail | +0.006 |
Neg logit contributions | |
Bad | -0.008 |
crow | -0.008 |
Leave | -0.007 |
Stop | -0.007 |
eper | -0.007 |
Token | in |
Token ID | 287 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
crow | +0.069 |
Bad | +0.067 |
Leave | +0.066 |
Stop | +0.061 |
eper | +0.061 |
Neg logit contributions | |
pron | -0.084 |
favourite | -0.080 |
masterpiece | -0.079 |
brand | -0.073 |
favorite | -0.071 |
Token | the |
Token ID | 262 |
Feature activation | -1.74e+00 |
Pos logit contributions | |
pron | +0.123 |
favourite | +0.117 |
masterpiece | +0.113 |
brand | +0.106 |
favorite | +0.105 |
Neg logit contributions | |
crow | -0.082 |
Bad | -0.082 |
Leave | -0.079 |
Stop | -0.073 |
ases | -0.072 |
Token | freshly |
Token ID | 29026 |
Feature activation | +1.62e+01 |
Pos logit contributions | |
crow | +0.147 |
Bad | +0.143 |
Leave | +0.140 |
Stop | +0.134 |
vious | +0.131 |
Neg logit contributions | |
pron | -0.120 |
favourite | -0.111 |
masterpiece | -0.111 |
brand | -0.101 |
favorite | -0.097 |
Token | cut |
Token ID | 2005 |
Feature activation | +3.13e+00 |
Pos logit contributions | |
pron | +1.014 |
favourite | +0.975 |
masterpiece | +0.951 |
brand | +0.864 |
favorite | +0.831 |
Neg logit contributions | |
crow | -1.439 |
Bad | -1.408 |
Leave | -1.375 |
Stop | -1.322 |
eper | -1.318 |
Token | grass |
Token ID | 8701 |
Feature activation | -5.13e-01 |
Pos logit contributions | |
pron | +0.279 |
favourite | +0.266 |
masterpiece | +0.264 |
brand | +0.243 |
favorite | +0.241 |
Neg logit contributions | |
crow | -0.204 |
Bad | -0.199 |
Leave | -0.195 |
vious | -0.181 |
eper | -0.179 |
Token | . |
Token ID | 13 |
Feature activation | -2.72e-01 |
Pos logit contributions | |
crow | +0.038 |
Bad | +0.037 |
Leave | +0.036 |
vious | +0.034 |
eper | +0.034 |
Neg logit contributions | |
pron | -0.041 |
favourite | -0.039 |
masterpiece | -0.038 |
brand | -0.035 |
favorite | -0.034 |
Token | It |
Token ID | 632 |
Feature activation | -4.89e-01 |
Pos logit contributions | |
Fighting | +0.011 |
If | +0.011 |
crow | +0.011 |
Stop | +0.011 |
Bad | +0.010 |
Neg logit contributions | |
pron | -0.028 |
masterpiece | -0.026 |
usual | -0.025 |
collection | -0.025 |
complexion | -0.024 |
Token | looked |
Token ID | 3114 |
Feature activation | +4.35e+00 |
Pos logit contributions | |
crow | +0.034 |
Bad | +0.034 |
Leave | +0.033 |
vious | +0.033 |
Hey | +0.032 |
Neg logit contributions | |
pron | -0.040 |
favourite | -0.036 |
masterpiece | -0.035 |
complexion | -0.033 |
brand | -0.033 |
Token | the |
Token ID | 262 |
Feature activation | -1.54e+00 |
Pos logit contributions | |
pron | +0.470 |
favourite | +0.436 |
masterpiece | +0.435 |
brand | +0.388 |
favorite | +0.388 |
Neg logit contributions | |
crow | -0.408 |
Bad | -0.396 |
Leave | -0.388 |
vious | -0.371 |
Stop | -0.365 |
Token | sweet |
Token ID | 6029 |
Feature activation | +5.65e+00 |
Pos logit contributions | |
Bad | +0.135 |
crow | +0.134 |
vious | +0.130 |
Leave | +0.127 |
Stop | +0.126 |
Neg logit contributions | |
pron | -0.087 |
favourite | -0.080 |
masterpiece | -0.077 |
brand | -0.074 |
icol | -0.073 |
Token | aroma |
Token ID | 31242 |
Feature activation | +1.58e+00 |
Pos logit contributions | |
pron | +0.371 |
favourite | +0.349 |
masterpiece | +0.337 |
brand | +0.306 |
favorite | +0.300 |
Neg logit contributions | |
crow | -0.501 |
Bad | -0.481 |
Leave | -0.473 |
vious | -0.461 |
Stop | -0.454 |
Token | of |
Token ID | 286 |
Feature activation | -3.05e+00 |
Pos logit contributions | |
pron | +0.148 |
favourite | +0.146 |
masterpiece | +0.140 |
favorite | +0.133 |
brand | +0.132 |
Neg logit contributions | |
crow | -0.089 |
Leave | -0.082 |
Bad | -0.081 |
vious | -0.079 |
Stop | -0.075 |
Token | a |
Token ID | 257 |
Feature activation | -1.75e-01 |
Pos logit contributions | |
Bad | +0.182 |
crow | +0.175 |
Leave | +0.165 |
Stop | +0.160 |
vious | +0.155 |
Neg logit contributions | |
pron | -0.280 |
favourite | -0.266 |
ail | -0.259 |
ouse | -0.259 |
masterpiece | -0.252 |
Token | freshly |
Token ID | 29026 |
Feature activation | +1.62e+01 |
Pos logit contributions | |
Bad | +0.015 |
crow | +0.015 |
Leave | +0.014 |
Stop | +0.014 |
bully | +0.014 |
Neg logit contributions | |
favourite | -0.010 |
pron | -0.010 |
masterpiece | -0.009 |
favorite | -0.009 |
ail | -0.009 |
Token | baked |
Token ID | 22979 |
Feature activation | +7.44e+00 |
Pos logit contributions | |
pron | +0.895 |
favourite | +0.861 |
masterpiece | +0.836 |
brand | +0.750 |
favorite | +0.723 |
Neg logit contributions | |
crow | -1.569 |
Bad | -1.523 |
Leave | -1.492 |
vious | -1.444 |
eper | -1.440 |
Token | cake |
Token ID | 12187 |
Feature activation | +6.40e+00 |
Pos logit contributions | |
favourite | +0.460 |
pron | +0.456 |
masterpiece | +0.424 |
favorite | +0.399 |
brand | +0.396 |
Neg logit contributions | |
crow | -0.665 |
Bad | -0.610 |
Leave | -0.598 |
eper | -0.593 |
vious | -0.589 |
Token | . |
Token ID | 13 |
Feature activation | -3.39e-01 |
Pos logit contributions | |
pron | +0.601 |
favourite | +0.601 |
masterpiece | +0.547 |
usual | +0.539 |
favorite | +0.539 |
Neg logit contributions | |
crow | -0.293 |
ting | -0.274 |
eper | -0.262 |
ters | -0.255 |
vious | -0.247 |
Token |
|
Token ID | 198 |
Feature activation | +3.37e-01 |
Pos logit contributions | |
crow | +0.019 |
Stop | +0.017 |
Bad | +0.017 |
Leave | +0.017 |
Hey | +0.016 |
Neg logit contributions | |
pron | -0.033 |
masterpiece | -0.032 |
favourite | -0.031 |
collection | -0.030 |
usual | -0.029 |
Token |
|
Token ID | 198 |
Feature activation | +3.26e-01 |
Pos logit contributions | |
pron | +0.029 |
masterpiece | +0.028 |
favourite | +0.028 |
brand | +0.026 |
favorite | +0.026 |
Neg logit contributions | |
crow | -0.023 |
Bad | -0.022 |
Leave | -0.021 |
Stop | -0.020 |
Hey | -0.020 |
Token | tidy |
Token ID | 43044 |
Feature activation | +3.29e+00 |
Pos logit contributions | |
pron | +0.083 |
favourite | +0.079 |
masterpiece | +0.077 |
brand | +0.070 |
favorite | +0.069 |
Neg logit contributions | |
crow | -0.112 |
Bad | -0.109 |
Leave | -0.107 |
vious | -0.102 |
Stop | -0.101 |
Token | , |
Token ID | 11 |
Feature activation | -1.55e+00 |
Pos logit contributions | |
pron | +0.273 |
masterpiece | +0.258 |
favourite | +0.255 |
brand | +0.230 |
favorite | +0.227 |
Neg logit contributions | |
crow | -0.226 |
Bad | -0.222 |
Leave | -0.215 |
eper | -0.206 |
vious | -0.201 |
Token | with |
Token ID | 351 |
Feature activation | -2.08e+00 |
Pos logit contributions | |
crow | +0.124 |
Bad | +0.124 |
Leave | +0.123 |
vious | +0.114 |
Stop | +0.113 |
Neg logit contributions | |
pron | -0.111 |
masterpiece | -0.105 |
favourite | -0.103 |
brand | -0.092 |
collection | -0.091 |
Token | his |
Token ID | 465 |
Feature activation | -3.34e+00 |
Pos logit contributions | |
crow | +0.148 |
Bad | +0.147 |
Leave | +0.146 |
vious | +0.136 |
eper | +0.134 |
Neg logit contributions | |
pron | -0.166 |
masterpiece | -0.155 |
favourite | -0.148 |
ail | -0.140 |
yt | -0.136 |
Token | clothes |
Token ID | 8242 |
Feature activation | +2.06e+00 |
Pos logit contributions | |
Bad | +0.272 |
crow | +0.269 |
Leave | +0.259 |
bully | +0.252 |
eper | +0.250 |
Neg logit contributions | |
pron | -0.216 |
masterpiece | -0.197 |
oph | -0.187 |
onna | -0.187 |
yt | -0.177 |
Token | freshly |
Token ID | 29026 |
Feature activation | +1.62e+01 |
Pos logit contributions | |
pron | +0.182 |
masterpiece | +0.175 |
favourite | +0.169 |
icol | +0.154 |
oph | +0.151 |
Neg logit contributions | |
crow | -0.121 |
Leave | -0.118 |
Bad | -0.114 |
vious | -0.113 |
eper | -0.111 |
Token | cleaned |
Token ID | 20750 |
Feature activation | +2.65e+00 |
Pos logit contributions | |
masterpiece | +1.128 |
pron | +1.102 |
favourite | +1.025 |
yt | +1.013 |
onna | +1.012 |
Neg logit contributions | |
crow | -1.247 |
Bad | -1.206 |
Leave | -1.190 |
Stop | -1.151 |
Hey | -1.149 |
Token | . |
Token ID | 13 |
Feature activation | -4.69e-01 |
Pos logit contributions | |
pron | +0.217 |
masterpiece | +0.201 |
favourite | +0.200 |
brand | +0.179 |
favorite | +0.178 |
Neg logit contributions | |
crow | -0.192 |
Bad | -0.184 |
Leave | -0.180 |
eper | -0.170 |
vious | -0.169 |
Token | Suddenly |
Token ID | 24975 |
Feature activation | -6.36e-01 |
Pos logit contributions | |
crow | +0.031 |
Bad | +0.030 |
Leave | +0.029 |
vious | +0.027 |
eper | +0.027 |
Neg logit contributions | |
pron | -0.042 |
favourite | -0.040 |
masterpiece | -0.040 |
brand | -0.037 |
favorite | -0.036 |
Token | , |
Token ID | 11 |
Feature activation | -1.58e+00 |
Pos logit contributions | |
crow | +0.055 |
Bad | +0.054 |
Leave | +0.053 |
vious | +0.051 |
ases | +0.051 |
Neg logit contributions | |
pron | -0.042 |
favourite | -0.041 |
masterpiece | -0.039 |
brand | -0.036 |
favorite | -0.035 |
Token | he |
Token ID | 339 |
Feature activation | -1.15e+00 |
Pos logit contributions | |
crow | +0.121 |
Bad | +0.118 |
Leave | +0.116 |
vious | +0.109 |
eper | +0.109 |
Neg logit contributions | |
pron | -0.123 |
favourite | -0.116 |
masterpiece | -0.115 |
brand | -0.105 |
favorite | -0.103 |
Token | mom |
Token ID | 1995 |
Feature activation | +4.86e+00 |
Pos logit contributions | |
Bad | +0.285 |
crow | +0.278 |
eper | +0.273 |
bully | +0.271 |
Leave | +0.266 |
Neg logit contributions | |
pron | -0.193 |
masterpiece | -0.167 |
favourite | -0.159 |
brand | -0.157 |
Sus | -0.155 |
Token | my |
Token ID | 1820 |
Feature activation | -3.20e+00 |
Pos logit contributions | |
pron | +0.439 |
favourite | +0.433 |
masterpiece | +0.423 |
favorite | +0.387 |
brand | +0.386 |
Neg logit contributions | |
crow | -0.303 |
Bad | -0.297 |
Leave | -0.289 |
eper | -0.279 |
vious | -0.268 |
Token | surprised |
Token ID | 6655 |
Feature activation | -3.11e+00 |
Pos logit contributions | |
crow | +0.246 |
Bad | +0.241 |
eper | +0.234 |
Leave | +0.233 |
Hey | +0.222 |
Neg logit contributions | |
favourite | -0.237 |
pron | -0.236 |
masterpiece | -0.227 |
favorite | -0.206 |
usual | -0.204 |
Token | him |
Token ID | 683 |
Feature activation | -6.86e+00 |
Pos logit contributions | |
crow | +0.147 |
Bad | +0.144 |
Leave | +0.139 |
vious | +0.125 |
eper | +0.124 |
Neg logit contributions | |
pron | -0.331 |
favourite | -0.318 |
masterpiece | -0.317 |
brand | -0.298 |
favorite | -0.296 |
Token | with |
Token ID | 351 |
Feature activation | -1.87e+00 |
Pos logit contributions | |
crow | +0.475 |
Bad | +0.469 |
Leave | +0.465 |
vious | +0.430 |
eper | +0.426 |
Neg logit contributions | |
pron | -0.570 |
favourite | -0.538 |
masterpiece | -0.533 |
brand | -0.484 |
favorite | -0.479 |
Token | freshly |
Token ID | 29026 |
Feature activation | +1.61e+01 |
Pos logit contributions | |
crow | +0.141 |
Bad | +0.140 |
Leave | +0.138 |
vious | +0.128 |
eper | +0.127 |
Neg logit contributions | |
pron | -0.147 |
masterpiece | -0.138 |
favourite | -0.138 |
favorite | -0.122 |
brand | -0.121 |
Token | baked |
Token ID | 22979 |
Feature activation | +7.48e+00 |
Pos logit contributions | |
pron | +1.030 |
favourite | +0.959 |
masterpiece | +0.940 |
brand | +0.837 |
favorite | +0.833 |
Neg logit contributions | |
crow | -1.473 |
Bad | -1.429 |
Leave | -1.415 |
vious | -1.347 |
eper | -1.332 |
Token | cookies |
Token ID | 14746 |
Feature activation | +1.88e+00 |
Pos logit contributions | |
pron | +0.479 |
favourite | +0.454 |
masterpiece | +0.440 |
brand | +0.403 |
favorite | +0.396 |
Neg logit contributions | |
crow | -0.675 |
Bad | -0.666 |
Leave | -0.665 |
vious | -0.624 |
Stop | -0.619 |
Token | . |
Token ID | 13 |
Feature activation | -2.74e-01 |
Pos logit contributions | |
pron | +0.162 |
favourite | +0.156 |
masterpiece | +0.155 |
brand | +0.141 |
favorite | +0.138 |
Neg logit contributions | |
crow | -0.124 |
Bad | -0.117 |
Leave | -0.113 |
eper | -0.110 |
vious | -0.110 |
Token | Tim |
Token ID | 5045 |
Feature activation | +3.12e-01 |
Pos logit contributions | |
Bad | +0.015 |
crow | +0.015 |
Leave | +0.015 |
Stop | +0.015 |
vious | +0.014 |
Neg logit contributions | |
pron | -0.026 |
masterpiece | -0.025 |
favourite | -0.024 |
collection | -0.023 |
usual | -0.022 |
Token | my |
Token ID | 1820 |
Feature activation | -3.23e+00 |
Pos logit contributions | |
pron | +0.028 |
favourite | +0.026 |
masterpiece | +0.026 |
brand | +0.024 |
favorite | +0.024 |
Neg logit contributions | |
crow | -0.021 |
Bad | -0.020 |
Leave | -0.020 |
Stop | -0.018 |
vious | -0.018 |
Token |
|
Token ID | 198 |
Feature activation | +3.19e-01 |
Pos logit contributions | |
pron | +0.225 |
favourite | +0.214 |
masterpiece | +0.213 |
brand | +0.196 |
favorite | +0.195 |
Neg logit contributions | |
crow | -0.162 |
Bad | -0.154 |
Leave | -0.151 |
Stop | -0.140 |
eper | -0.139 |
Token | Andy |
Token ID | 35314 |
Feature activation | -3.04e+00 |
Pos logit contributions | |
pron | +0.030 |
masterpiece | +0.028 |
favourite | +0.028 |
brand | +0.026 |
ail | +0.025 |
Neg logit contributions | |
Bad | -0.019 |
crow | -0.019 |
vious | -0.018 |
Leave | -0.018 |
ases | -0.017 |
Token | looked |
Token ID | 3114 |
Feature activation | +4.30e+00 |
Pos logit contributions | |
crow | +0.244 |
Bad | +0.241 |
Leave | +0.234 |
eper | +0.228 |
vious | +0.225 |
Neg logit contributions | |
pron | -0.218 |
favourite | -0.202 |
masterpiece | -0.192 |
brand | -0.181 |
usual | -0.181 |
Token | at |
Token ID | 379 |
Feature activation | -3.28e-01 |
Pos logit contributions | |
pron | +0.385 |
masterpiece | +0.346 |
favourite | +0.340 |
usual | +0.327 |
ail | +0.318 |
Neg logit contributions | |
crow | -0.270 |
Bad | -0.269 |
vious | -0.265 |
Leave | -0.262 |
Stop | -0.243 |
Token | the |
Token ID | 262 |
Feature activation | -1.62e+00 |
Pos logit contributions | |
crow | +0.024 |
Bad | +0.023 |
Leave | +0.023 |
vious | +0.022 |
eper | +0.021 |
Neg logit contributions | |
pron | -0.027 |
favourite | -0.026 |
masterpiece | -0.025 |
brand | -0.023 |
favorite | -0.023 |
Token | freshly |
Token ID | 29026 |
Feature activation | +1.61e+01 |
Pos logit contributions | |
Bad | +0.148 |
crow | +0.147 |
Leave | +0.141 |
Stop | +0.138 |
vious | +0.138 |
Neg logit contributions | |
pron | -0.093 |
favourite | -0.091 |
masterpiece | -0.085 |
brand | -0.079 |
collection | -0.077 |
Token | printed |
Token ID | 10398 |
Feature activation | +3.99e+00 |
Pos logit contributions | |
pron | +0.822 |
favourite | +0.764 |
masterpiece | +0.736 |
brand | +0.647 |
favorite | +0.630 |
Neg logit contributions | |
crow | -1.653 |
Bad | -1.631 |
Leave | -1.599 |
eper | -1.537 |
Stop | -1.534 |
Token | picture |
Token ID | 4286 |
Feature activation | +5.51e+00 |
Pos logit contributions | |
pron | +0.256 |
favourite | +0.250 |
masterpiece | +0.233 |
favorite | +0.219 |
brand | +0.217 |
Neg logit contributions | |
crow | -0.351 |
Bad | -0.332 |
Leave | -0.325 |
eper | -0.318 |
vious | -0.313 |
Token | and |
Token ID | 290 |
Feature activation | +1.88e-02 |
Pos logit contributions | |
pron | +0.474 |
favourite | +0.451 |
masterpiece | +0.438 |
brand | +0.407 |
favorite | +0.406 |
Neg logit contributions | |
crow | -0.367 |
Bad | -0.338 |
Leave | -0.334 |
vious | -0.326 |
eper | -0.318 |
Token | smiled |
Token ID | 13541 |
Feature activation | +6.06e-01 |
Pos logit contributions | |
pron | +0.001 |
favourite | +0.001 |
masterpiece | +0.001 |
favorite | +0.001 |
brand | +0.001 |
Neg logit contributions | |
Bad | -0.001 |
crow | -0.001 |
Leave | -0.001 |
Stop | -0.001 |
eper | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | -4.04e-01 |
Pos logit contributions | |
pron | +0.052 |
masterpiece | +0.048 |
favourite | +0.047 |
brand | +0.043 |
usual | +0.043 |
Neg logit contributions | |
crow | -0.040 |
Bad | -0.039 |
Leave | -0.039 |
vious | -0.037 |
eper | -0.037 |
Token | them |
Token ID | 606 |
Feature activation | -1.92e+00 |
Pos logit contributions | |
crow | +0.087 |
Bad | +0.083 |
Leave | +0.082 |
vious | +0.077 |
Stop | +0.077 |
Neg logit contributions | |
pron | -0.110 |
masterpiece | -0.101 |
favourite | -0.100 |
brand | -0.092 |
favorite | -0.089 |
Token | over |
Token ID | 625 |
Feature activation | -1.89e+00 |
Pos logit contributions | |
crow | +0.159 |
eper | +0.151 |
Bad | +0.149 |
Leave | +0.149 |
ases | +0.148 |
Neg logit contributions | |
pron | -0.139 |
masterpiece | -0.122 |
favourite | -0.114 |
brand | -0.111 |
icol | -0.106 |
Token | for |
Token ID | 329 |
Feature activation | +5.59e+00 |
Pos logit contributions | |
crow | +0.141 |
Bad | +0.137 |
Leave | +0.135 |
ases | +0.130 |
Stop | +0.127 |
Neg logit contributions | |
pron | -0.146 |
masterpiece | -0.134 |
favourite | -0.132 |
brand | -0.123 |
favorite | -0.116 |
Token | a |
Token ID | 257 |
Feature activation | +2.40e-02 |
Pos logit contributions | |
pron | +0.431 |
favourite | +0.406 |
masterpiece | +0.396 |
favorite | +0.361 |
brand | +0.357 |
Neg logit contributions | |
crow | -0.429 |
Bad | -0.425 |
Leave | -0.417 |
ases | -0.389 |
Stop | -0.388 |
Token | fair |
Token ID | 3148 |
Feature activation | -2.81e+00 |
Pos logit contributions | |
favourite | +0.001 |
pron | +0.001 |
masterpiece | +0.001 |
addy | +0.001 |
liv | +0.001 |
Neg logit contributions | |
Bad | -0.002 |
Leave | -0.002 |
bully | -0.002 |
crow | -0.002 |
Rex | -0.002 |
Token | yt |
Token ID | 20760 |
Feature activation | +1.46e+01 |
Pos logit contributions | |
crow | +0.208 |
eper | +0.187 |
hunters | +0.184 |
bully | +0.181 |
Leave | +0.178 |
Neg logit contributions | |
pron | -0.208 |
favourite | -0.191 |
masterpiece | -0.183 |
brand | -0.176 |
usual | -0.171 |
Token | ale |
Token ID | 1000 |
Feature activation | +1.92e+00 |
Pos logit contributions | |
pron | +1.236 |
favourite | +1.169 |
masterpiece | +1.147 |
brand | +1.082 |
favorite | +1.034 |
Neg logit contributions | |
crow | -1.050 |
Leave | -0.957 |
Bad | -0.956 |
vious | -0.954 |
ases | -0.927 |
Token | tea |
Token ID | 8887 |
Feature activation | +5.42e+00 |
Pos logit contributions | |
pron | +0.143 |
favourite | +0.135 |
masterpiece | +0.134 |
brand | +0.122 |
favorite | +0.120 |
Neg logit contributions | |
crow | -0.153 |
Bad | -0.150 |
Leave | -0.147 |
vious | -0.138 |
Stop | -0.137 |
Token | party |
Token ID | 2151 |
Feature activation | +3.92e+00 |
Pos logit contributions | |
pron | +0.275 |
favourite | +0.264 |
masterpiece | +0.264 |
brand | +0.232 |
favorite | +0.213 |
Neg logit contributions | |
crow | -0.545 |
Bad | -0.508 |
eper | -0.499 |
Leave | -0.499 |
ases | -0.489 |
Token | . |
Token ID | 13 |
Feature activation | -3.01e-02 |
Pos logit contributions | |
pron | +0.339 |
masterpiece | +0.320 |
favourite | +0.315 |
brand | +0.297 |
usual | +0.280 |
Neg logit contributions | |
crow | -0.255 |
eper | -0.227 |
Bad | -0.226 |
Leave | -0.224 |
hunters | -0.223 |
Token | They |
Token ID | 1119 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
crow | +0.002 |
Bad | +0.002 |
Leave | +0.002 |
Stop | +0.002 |
ases | +0.002 |
Neg logit contributions | |
pron | -0.003 |
masterpiece | -0.002 |
favourite | -0.002 |
brand | -0.002 |
usual | -0.002 |
Token | was |
Token ID | 373 |
Feature activation | +1.62e+00 |
Pos logit contributions | |
pron | +0.333 |
favourite | +0.305 |
masterpiece | +0.303 |
brand | +0.276 |
favorite | +0.263 |
Neg logit contributions | |
crow | -0.443 |
Bad | -0.438 |
Leave | -0.426 |
vious | -0.412 |
Stop | -0.404 |
Token | walking |
Token ID | 6155 |
Feature activation | +6.25e-01 |
Pos logit contributions | |
pron | +0.129 |
masterpiece | +0.117 |
favourite | +0.117 |
icol | +0.105 |
brand | +0.105 |
Neg logit contributions | |
crow | -0.118 |
Bad | -0.113 |
Leave | -0.112 |
vious | -0.108 |
eper | -0.107 |
Token | through |
Token ID | 832 |
Feature activation | -1.50e+00 |
Pos logit contributions | |
pron | +0.050 |
masterpiece | +0.047 |
favourite | +0.047 |
brand | +0.043 |
favorite | +0.042 |
Neg logit contributions | |
crow | -0.047 |
Bad | -0.045 |
Leave | -0.044 |
vious | -0.042 |
eper | -0.041 |
Token | a |
Token ID | 257 |
Feature activation | +8.69e-02 |
Pos logit contributions | |
crow | +0.103 |
Bad | +0.100 |
Leave | +0.098 |
Stop | +0.091 |
vious | +0.091 |
Neg logit contributions | |
pron | -0.130 |
favourite | -0.122 |
masterpiece | -0.121 |
brand | -0.111 |
favorite | -0.110 |
Token | fair |
Token ID | 3148 |
Feature activation | -2.79e+00 |
Pos logit contributions | |
pron | +0.006 |
favourite | +0.005 |
masterpiece | +0.005 |
brand | +0.005 |
favorite | +0.005 |
Neg logit contributions | |
crow | -0.008 |
Bad | -0.008 |
Leave | -0.008 |
vious | -0.007 |
eper | -0.007 |
Token | yt |
Token ID | 20760 |
Feature activation | +1.46e+01 |
Pos logit contributions | |
crow | +0.230 |
Bad | +0.211 |
Leave | +0.210 |
eper | +0.203 |
hunters | +0.201 |
Neg logit contributions | |
pron | -0.194 |
favourite | -0.178 |
masterpiece | -0.171 |
brand | -0.166 |
favorite | -0.155 |
Token | ale |
Token ID | 1000 |
Feature activation | +1.99e+00 |
Pos logit contributions | |
pron | +1.195 |
favourite | +1.133 |
masterpiece | +1.110 |
brand | +1.037 |
favorite | +1.009 |
Neg logit contributions | |
crow | -1.082 |
Bad | -1.017 |
Leave | -1.011 |
vious | -0.981 |
ases | -0.956 |
Token | . |
Token ID | 13 |
Feature activation | -1.08e-01 |
Pos logit contributions | |
pron | +0.146 |
favourite | +0.139 |
masterpiece | +0.138 |
brand | +0.124 |
favorite | +0.123 |
Neg logit contributions | |
crow | -0.159 |
Bad | -0.158 |
Leave | -0.155 |
Stop | -0.146 |
vious | -0.145 |
Token | She |
Token ID | 1375 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
crow | +0.008 |
Leave | +0.007 |
Bad | +0.007 |
vious | +0.007 |
Stop | +0.006 |
Neg logit contributions | |
pron | -0.009 |
masterpiece | -0.009 |
favourite | -0.009 |
collection | -0.008 |
usual | -0.008 |
Token | fell |
Token ID | 3214 |
Feature activation | +3.93e+00 |
Pos logit contributions | |
pron | +0.094 |
masterpiece | +0.084 |
favourite | +0.081 |
brand | +0.079 |
list | +0.076 |
Neg logit contributions | |
crow | -0.081 |
Leave | -0.080 |
Bad | -0.079 |
vious | -0.074 |
eper | -0.073 |
Token | asleep |
Token ID | 16039 |
Feature activation | +2.01e+00 |
Pos logit contributions | |
pron | +0.322 |
masterpiece | +0.307 |
favourite | +0.303 |
collection | +0.268 |
favorite | +0.265 |
Neg logit contributions | |
crow | -0.282 |
Leave | -0.276 |
Bad | -0.270 |
eper | -0.252 |
Stop | -0.250 |
Token | so |
Token ID | 523 |
Feature activation | +4.49e-01 |
Pos logit contributions | |
pron | +0.450 |
masterpiece | +0.422 |
favourite | +0.393 |
onna | +0.379 |
usual | +0.375 |
Neg logit contributions | |
crow | -0.290 |
vious | -0.284 |
Bad | -0.282 |
Leave | -0.277 |
eper | -0.260 |
Token | attractive |
Token ID | 10966 |
Feature activation | +4.59e+00 |
Pos logit contributions | |
pron | +0.031 |
masterpiece | +0.031 |
icol | +0.028 |
porch | +0.027 |
doorstep | +0.027 |
Neg logit contributions | |
vious | -0.034 |
Bad | -0.032 |
crow | -0.032 |
Stop | -0.031 |
Leave | -0.031 |
Token | in |
Token ID | 287 |
Feature activation | +1.77e+00 |
Pos logit contributions | |
pron | +0.412 |
masterpiece | +0.377 |
favourite | +0.363 |
brand | +0.333 |
ail | +0.324 |
Neg logit contributions | |
crow | -0.307 |
Bad | -0.286 |
Leave | -0.275 |
vious | -0.274 |
ting | -0.273 |
Token | the |
Token ID | 262 |
Feature activation | -1.19e+00 |
Pos logit contributions | |
pron | +0.152 |
favourite | +0.142 |
masterpiece | +0.140 |
favorite | +0.128 |
brand | +0.128 |
Neg logit contributions | |
crow | -0.120 |
Bad | -0.119 |
Leave | -0.115 |
Stop | -0.109 |
ases | -0.107 |
Token | a |
Token ID | 257 |
Feature activation | +3.61e-01 |
Pos logit contributions | |
crow | +0.105 |
Bad | +0.103 |
Leave | +0.102 |
vious | +0.098 |
Stop | +0.097 |
Neg logit contributions | |
pron | -0.072 |
favourite | -0.071 |
masterpiece | -0.070 |
brand | -0.061 |
favorite | -0.061 |
Token | pron |
Token ID | 31186 |
Feature activation | +1.46e+01 |
Pos logit contributions | |
favourite | +0.015 |
masterpiece | +0.014 |
pron | +0.014 |
favorite | +0.012 |
ail | +0.012 |
Neg logit contributions | |
Bad | -0.037 |
crow | -0.037 |
Leave | -0.036 |
Stop | -0.034 |
eper | -0.034 |
Token | that |
Token ID | 326 |
Feature activation | +3.12e+00 |
Pos logit contributions | |
pron | +1.180 |
masterpiece | +1.128 |
favourite | +1.120 |
brand | +1.047 |
favorite | +0.992 |
Neg logit contributions | |
crow | -1.051 |
Bad | -0.981 |
Leave | -0.932 |
eper | -0.927 |
vious | -0.911 |
Token | everyone |
Token ID | 2506 |
Feature activation | +8.11e-02 |
Pos logit contributions | |
pron | +0.219 |
masterpiece | +0.212 |
brand | +0.196 |
icol | +0.188 |
favourite | +0.186 |
Neg logit contributions | |
crow | -0.255 |
Bad | -0.233 |
Stop | -0.227 |
Hey | -0.226 |
vious | -0.220 |
Token | wanted |
Token ID | 2227 |
Feature activation | -8.52e-01 |
Pos logit contributions | |
pron | +0.006 |
masterpiece | +0.006 |
favourite | +0.006 |
brand | +0.006 |
icol | +0.005 |
Neg logit contributions | |
crow | -0.006 |
Bad | -0.006 |
vious | -0.006 |
Leave | -0.006 |
eper | -0.005 |
Token | to |
Token ID | 284 |
Feature activation | +2.00e+00 |
Pos logit contributions | |
crow | +0.068 |
Bad | +0.063 |
Leave | +0.061 |
Stop | +0.061 |
ases | +0.058 |
Neg logit contributions | |
pron | -0.064 |
masterpiece | -0.057 |
favourite | -0.055 |
brand | -0.052 |
complexion | -0.051 |
Token | see |
Token ID | 766 |
Feature activation | +6.48e-01 |
Pos logit contributions | |
pron | +0.154 |
favourite | +0.154 |
masterpiece | +0.148 |
brand | +0.139 |
favorite | +0.138 |
Neg logit contributions | |
crow | -0.138 |
Leave | -0.137 |
Bad | -0.135 |
ases | -0.126 |
Stop | -0.126 |
Token | and |
Token ID | 290 |
Feature activation | +2.91e-01 |
Pos logit contributions | |
pron | +0.435 |
masterpiece | +0.401 |
favourite | +0.400 |
favorite | +0.362 |
brand | +0.359 |
Neg logit contributions | |
crow | -0.363 |
Leave | -0.355 |
Bad | -0.350 |
eper | -0.331 |
vious | -0.327 |
Token | saw |
Token ID | 2497 |
Feature activation | -4.37e-01 |
Pos logit contributions | |
pron | +0.023 |
masterpiece | +0.021 |
favourite | +0.021 |
favorite | +0.018 |
brand | +0.018 |
Neg logit contributions | |
crow | -0.023 |
Bad | -0.022 |
Leave | -0.022 |
eper | -0.021 |
vious | -0.020 |
Token | a |
Token ID | 257 |
Feature activation | +1.56e-02 |
Pos logit contributions | |
crow | +0.036 |
Bad | +0.036 |
Leave | +0.035 |
Stop | +0.033 |
vious | +0.033 |
Neg logit contributions | |
pron | -0.030 |
favourite | -0.030 |
masterpiece | -0.029 |
brand | -0.027 |
favorite | -0.026 |
Token | magical |
Token ID | 10883 |
Feature activation | +4.44e+00 |
Pos logit contributions | |
pron | +0.001 |
favourite | +0.001 |
masterpiece | +0.001 |
favorite | +0.001 |
brand | +0.001 |
Neg logit contributions | |
crow | -0.001 |
Bad | -0.001 |
Leave | -0.001 |
eper | -0.001 |
Stop | -0.001 |
Token | fair |
Token ID | 3148 |
Feature activation | -2.89e+00 |
Pos logit contributions | |
pron | +0.369 |
favourite | +0.354 |
masterpiece | +0.347 |
favorite | +0.319 |
brand | +0.318 |
Neg logit contributions | |
crow | -0.315 |
Bad | -0.299 |
Leave | -0.289 |
eper | -0.273 |
vious | -0.270 |
Token | yt |
Token ID | 20760 |
Feature activation | +1.46e+01 |
Pos logit contributions | |
crow | +0.218 |
Bad | +0.203 |
Leave | +0.199 |
eper | +0.193 |
vious | +0.186 |
Neg logit contributions | |
pron | -0.229 |
favourite | -0.217 |
masterpiece | -0.210 |
favorite | -0.194 |
brand | -0.194 |
Token | ale |
Token ID | 1000 |
Feature activation | +1.86e+00 |
Pos logit contributions | |
pron | +1.033 |
favourite | +0.989 |
masterpiece | +0.983 |
favorite | +0.877 |
brand | +0.876 |
Neg logit contributions | |
Bad | -1.189 |
crow | -1.169 |
Leave | -1.159 |
Stop | -1.090 |
eper | -1.062 |
Token | creature |
Token ID | 7185 |
Feature activation | -5.62e+00 |
Pos logit contributions | |
pron | +0.134 |
masterpiece | +0.129 |
favourite | +0.128 |
brand | +0.116 |
collection | +0.112 |
Neg logit contributions | |
Bad | -0.149 |
crow | -0.148 |
Leave | -0.146 |
Stop | -0.141 |
ases | -0.139 |
Token | called |
Token ID | 1444 |
Feature activation | -1.61e+00 |
Pos logit contributions | |
crow | +0.436 |
Bad | +0.424 |
Leave | +0.418 |
eper | +0.392 |
vious | +0.391 |
Neg logit contributions | |
pron | -0.430 |
favourite | -0.404 |
masterpiece | -0.400 |
brand | -0.363 |
favorite | -0.360 |
Token | a |
Token ID | 257 |
Feature activation | +8.89e-03 |
Pos logit contributions | |
crow | +0.133 |
Bad | +0.130 |
Leave | +0.128 |
vious | +0.121 |
eper | +0.120 |
Neg logit contributions | |
pron | -0.115 |
favourite | -0.109 |
masterpiece | -0.107 |
brand | -0.097 |
favorite | -0.096 |
Token | n |
Token ID | 299 |
Feature activation | -5.25e-01 |
Pos logit contributions | |
pron | +0.001 |
favourite | +0.001 |
masterpiece | +0.001 |
favorite | +0.001 |
brand | +0.001 |
Neg logit contributions | |
crow | -0.001 |
Bad | -0.001 |
Leave | -0.001 |
eper | -0.001 |
Stop | -0.001 |
Token | and |
Token ID | 290 |
Feature activation | -1.05e-01 |
Pos logit contributions | |
crow | +0.933 |
Hey | +0.917 |
Leave | +0.898 |
Bad | +0.884 |
Stop | +0.876 |
Neg logit contributions | |
pron | -0.498 |
favourite | -0.473 |
favorite | -0.467 |
masterpiece | -0.426 |
collection | -0.405 |
Token | makes |
Token ID | 1838 |
Feature activation | -7.37e-01 |
Pos logit contributions | |
crow | +0.008 |
eper | +0.007 |
Bad | +0.007 |
Leave | +0.007 |
vious | +0.007 |
Neg logit contributions | |
pron | -0.008 |
favourite | -0.008 |
masterpiece | -0.008 |
favorite | -0.007 |
brand | -0.007 |
Token | them |
Token ID | 606 |
Feature activation | -2.23e+00 |
Pos logit contributions | |
crow | +0.050 |
Leave | +0.048 |
Bad | +0.046 |
eper | +0.045 |
vious | +0.043 |
Neg logit contributions | |
favourite | -0.061 |
masterpiece | -0.061 |
pron | -0.060 |
favorite | -0.058 |
brand | -0.056 |
Token | cry |
Token ID | 3960 |
Feature activation | -3.11e+00 |
Pos logit contributions | |
crow | +0.188 |
Bad | +0.174 |
Leave | +0.173 |
Hey | +0.168 |
eper | +0.167 |
Neg logit contributions | |
favourite | -0.155 |
pron | -0.155 |
masterpiece | -0.148 |
favorite | -0.146 |
brand | -0.139 |
Token | . |
Token ID | 13 |
Feature activation | -1.61e-01 |
Pos logit contributions | |
crow | +0.308 |
Leave | +0.299 |
Bad | +0.295 |
Hey | +0.285 |
eper | +0.278 |
Neg logit contributions | |
favourite | -0.153 |
favorite | -0.151 |
pron | -0.145 |
masterpiece | -0.136 |
brand | -0.131 |
Token | War |
Token ID | 1810 |
Feature activation | -1.51e+01 |
Pos logit contributions | |
crow | +0.011 |
eper | +0.011 |
vious | +0.010 |
Leave | +0.010 |
thieves | +0.010 |
Neg logit contributions | |
favourite | -0.012 |
favorite | -0.011 |
Daddy | -0.011 |
Grand | -0.010 |
pron | -0.010 |
Token | is |
Token ID | 318 |
Feature activation | -1.53e+00 |
Pos logit contributions | |
crow | +1.265 |
Leave | +1.223 |
Bad | +1.204 |
eper | +1.186 |
hunters | +1.111 |
Neg logit contributions | |
favorite | -0.901 |
favourite | -0.893 |
pron | -0.886 |
masterpiece | -0.809 |
couldn | -0.793 |
Token | not |
Token ID | 407 |
Feature activation | +1.81e+00 |
Pos logit contributions | |
crow | +0.134 |
Leave | +0.126 |
Bad | +0.122 |
vious | +0.120 |
eper | +0.120 |
Neg logit contributions | |
pron | -0.101 |
favourite | -0.099 |
masterpiece | -0.095 |
favorite | -0.092 |
brand | -0.091 |
Token | fun |
Token ID | 1257 |
Feature activation | +7.60e-01 |
Pos logit contributions | |
pron | +0.115 |
favourite | +0.109 |
masterpiece | +0.105 |
favorite | +0.101 |
brand | +0.099 |
Neg logit contributions | |
crow | -0.166 |
Leave | -0.158 |
Bad | -0.154 |
eper | -0.150 |
vious | -0.149 |
Token | . |
Token ID | 13 |
Feature activation | -1.81e-01 |
Pos logit contributions | |
pron | +0.047 |
favourite | +0.046 |
masterpiece | +0.045 |
favorite | +0.043 |
brand | +0.040 |
Neg logit contributions | |
crow | -0.067 |
Bad | -0.066 |
Leave | -0.066 |
eper | -0.062 |
vious | -0.061 |
Token | You |
Token ID | 921 |
Feature activation | -5.96e+00 |
Pos logit contributions | |
crow | +0.010 |
eper | +0.010 |
Leave | +0.009 |
thieves | +0.009 |
vious | +0.009 |
Neg logit contributions | |
favourite | -0.016 |
favorite | -0.015 |
pron | -0.014 |
masterpiece | -0.014 |
Daddy | -0.014 |
Token | and |
Token ID | 290 |
Feature activation | -1.06e-01 |
Pos logit contributions | |
Bad | +0.212 |
crow | +0.211 |
Leave | +0.208 |
eper | +0.198 |
Hey | +0.193 |
Neg logit contributions | |
favourite | -0.115 |
favorite | -0.109 |
pron | -0.103 |
masterpiece | -0.103 |
brand | -0.093 |
Token | scary |
Token ID | 14343 |
Feature activation | -2.61e+00 |
Pos logit contributions | |
crow | +0.008 |
eper | +0.008 |
Leave | +0.008 |
Bad | +0.007 |
ases | +0.007 |
Neg logit contributions | |
pron | -0.008 |
favourite | -0.008 |
masterpiece | -0.007 |
favorite | -0.007 |
brand | -0.007 |
Token | and |
Token ID | 290 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
crow | +0.254 |
Leave | +0.252 |
Bad | +0.252 |
eper | +0.239 |
Hey | +0.237 |
Neg logit contributions | |
favourite | -0.117 |
favorite | -0.112 |
masterpiece | -0.105 |
pron | -0.099 |
brand | -0.094 |
Token | sad |
Token ID | 6507 |
Feature activation | -3.96e+00 |
Pos logit contributions | |
crow | +0.008 |
eper | +0.007 |
Leave | +0.007 |
Bad | +0.007 |
ases | +0.007 |
Neg logit contributions | |
pron | -0.007 |
favourite | -0.007 |
masterpiece | -0.007 |
favorite | -0.007 |
brand | -0.007 |
Token | . |
Token ID | 13 |
Feature activation | -1.26e-01 |
Pos logit contributions | |
crow | +0.375 |
Leave | +0.369 |
Bad | +0.366 |
eper | +0.358 |
vious | +0.349 |
Neg logit contributions | |
favourite | -0.193 |
favorite | -0.188 |
masterpiece | -0.176 |
pron | -0.172 |
brand | -0.162 |
Token | War |
Token ID | 1810 |
Feature activation | -1.51e+01 |
Pos logit contributions | |
crow | +0.008 |
eper | +0.008 |
vious | +0.008 |
Leave | +0.008 |
thieves | +0.007 |
Neg logit contributions | |
favourite | -0.009 |
favorite | -0.009 |
Daddy | -0.009 |
Grand | -0.008 |
Mam | -0.008 |
Token | hurts |
Token ID | 20406 |
Feature activation | -1.80e+00 |
Pos logit contributions | |
crow | +1.249 |
Leave | +1.220 |
Bad | +1.199 |
eper | +1.168 |
vious | +1.103 |
Neg logit contributions | |
favourite | -0.913 |
favorite | -0.910 |
pron | -0.899 |
masterpiece | -0.823 |
couldn | -0.800 |
Token | people |
Token ID | 661 |
Feature activation | -9.78e+00 |
Pos logit contributions | |
Leave | +0.152 |
crow | +0.145 |
vious | +0.144 |
eper | +0.143 |
Bad | +0.140 |
Neg logit contributions | |
favourite | -0.096 |
favorite | -0.094 |
pron | -0.089 |
usual | -0.084 |
masterpiece | -0.083 |
Token | and |
Token ID | 290 |
Feature activation | -1.05e-01 |
Pos logit contributions | |
crow | +0.933 |
Hey | +0.917 |
Leave | +0.898 |
Bad | +0.884 |
Stop | +0.876 |
Neg logit contributions | |
pron | -0.498 |
favourite | -0.473 |
favorite | -0.467 |
masterpiece | -0.426 |
collection | -0.405 |
Token | makes |
Token ID | 1838 |
Feature activation | -7.37e-01 |
Pos logit contributions | |
crow | +0.008 |
eper | +0.007 |
Bad | +0.007 |
Leave | +0.007 |
vious | +0.007 |
Neg logit contributions | |
pron | -0.008 |
favourite | -0.008 |
masterpiece | -0.008 |
favorite | -0.007 |
brand | -0.007 |
Token | them |
Token ID | 606 |
Feature activation | -2.23e+00 |
Pos logit contributions | |
crow | +0.050 |
Leave | +0.048 |
Bad | +0.046 |
eper | +0.045 |
vious | +0.043 |
Neg logit contributions | |
favourite | -0.061 |
masterpiece | -0.061 |
pron | -0.060 |
favorite | -0.058 |
brand | -0.056 |
Token | oys |
Token ID | 19417 |
Feature activation | -1.64e+00 |
Pos logit contributions | |
Bad | +0.485 |
crow | +0.483 |
Leave | +0.482 |
vious | +0.443 |
eper | +0.443 |
Neg logit contributions | |
pron | -0.386 |
favourite | -0.372 |
masterpiece | -0.358 |
favorite | -0.331 |
ail | -0.329 |
Token | , |
Token ID | 11 |
Feature activation | -1.68e+00 |
Pos logit contributions | |
crow | +0.143 |
Bad | +0.141 |
Leave | +0.140 |
vious | +0.131 |
eper | +0.129 |
Neg logit contributions | |
pron | -0.108 |
favourite | -0.102 |
masterpiece | -0.100 |
brand | -0.091 |
favorite | -0.090 |
Token | stop |
Token ID | 2245 |
Feature activation | -3.83e+00 |
Pos logit contributions | |
crow | +0.115 |
Bad | +0.109 |
Leave | +0.108 |
vious | +0.103 |
ases | +0.101 |
Neg logit contributions | |
pron | -0.145 |
favourite | -0.139 |
masterpiece | -0.136 |
favorite | -0.127 |
brand | -0.127 |
Token | it |
Token ID | 340 |
Feature activation | +2.11e-01 |
Pos logit contributions | |
crow | +0.330 |
Bad | +0.313 |
Leave | +0.311 |
vious | +0.302 |
eper | +0.296 |
Neg logit contributions | |
pron | -0.256 |
masterpiece | -0.242 |
favourite | -0.240 |
brand | -0.217 |
favorite | -0.212 |
Token | . |
Token ID | 13 |
Feature activation | -1.26e-01 |
Pos logit contributions | |
pron | +0.013 |
favourite | +0.013 |
masterpiece | +0.012 |
favorite | +0.011 |
brand | +0.011 |
Neg logit contributions | |
crow | -0.019 |
Bad | -0.018 |
Leave | -0.018 |
eper | -0.017 |
vious | -0.016 |
Token | War |
Token ID | 1810 |
Feature activation | -1.51e+01 |
Pos logit contributions | |
Leave | +0.007 |
Bad | +0.007 |
crow | +0.007 |
eper | +0.007 |
vious | +0.007 |
Neg logit contributions | |
favourite | -0.011 |
pron | -0.010 |
favorite | -0.010 |
masterpiece | -0.010 |
Sus | -0.010 |
Token | is |
Token ID | 318 |
Feature activation | -1.46e+00 |
Pos logit contributions | |
crow | +1.274 |
Leave | +1.231 |
Bad | +1.220 |
eper | +1.191 |
vious | +1.121 |
Neg logit contributions | |
favourite | -0.888 |
pron | -0.888 |
favorite | -0.880 |
masterpiece | -0.800 |
couldn | -0.773 |
Token | not |
Token ID | 407 |
Feature activation | +1.89e+00 |
Pos logit contributions | |
crow | +0.133 |
Leave | +0.125 |
Bad | +0.122 |
vious | +0.120 |
eper | +0.119 |
Neg logit contributions | |
pron | -0.091 |
favourite | -0.088 |
masterpiece | -0.085 |
favorite | -0.081 |
brand | -0.081 |
Token | a |
Token ID | 257 |
Feature activation | -3.57e-01 |
Pos logit contributions | |
pron | +0.120 |
favourite | +0.114 |
masterpiece | +0.111 |
favorite | +0.104 |
brand | +0.104 |
Neg logit contributions | |
crow | -0.172 |
Leave | -0.164 |
Bad | -0.162 |
vious | -0.155 |
eper | -0.155 |
Token | game |
Token ID | 983 |
Feature activation | -2.49e+00 |
Pos logit contributions | |
crow | +0.029 |
Leave | +0.027 |
Bad | +0.027 |
vious | +0.026 |
eper | +0.026 |
Neg logit contributions | |
pron | -0.027 |
favourite | -0.025 |
masterpiece | -0.025 |
brand | -0.023 |
favorite | -0.023 |
Token | . |
Token ID | 13 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
crow | +0.220 |
Bad | +0.219 |
Leave | +0.218 |
eper | +0.203 |
vious | +0.203 |
Neg logit contributions | |
pron | -0.148 |
favourite | -0.146 |
masterpiece | -0.140 |
favorite | -0.134 |
usual | -0.128 |
Token | is |
Token ID | 318 |
Feature activation | -1.46e+00 |
Pos logit contributions | |
crow | +1.274 |
Leave | +1.231 |
Bad | +1.220 |
eper | +1.191 |
vious | +1.121 |
Neg logit contributions | |
favourite | -0.888 |
pron | -0.888 |
favorite | -0.880 |
masterpiece | -0.800 |
couldn | -0.773 |
Token | not |
Token ID | 407 |
Feature activation | +1.89e+00 |
Pos logit contributions | |
crow | +0.133 |
Leave | +0.125 |
Bad | +0.122 |
vious | +0.120 |
eper | +0.119 |
Neg logit contributions | |
pron | -0.091 |
favourite | -0.088 |
masterpiece | -0.085 |
favorite | -0.081 |
brand | -0.081 |
Token | a |
Token ID | 257 |
Feature activation | -3.57e-01 |
Pos logit contributions | |
pron | +0.120 |
favourite | +0.114 |
masterpiece | +0.111 |
favorite | +0.104 |
brand | +0.104 |
Neg logit contributions | |
crow | -0.172 |
Leave | -0.164 |
Bad | -0.162 |
vious | -0.155 |
eper | -0.155 |
Token | game |
Token ID | 983 |
Feature activation | -2.49e+00 |
Pos logit contributions | |
crow | +0.029 |
Leave | +0.027 |
Bad | +0.027 |
vious | +0.026 |
eper | +0.026 |
Neg logit contributions | |
pron | -0.027 |
favourite | -0.025 |
masterpiece | -0.025 |
brand | -0.023 |
favorite | -0.023 |
Token | . |
Token ID | 13 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
crow | +0.220 |
Bad | +0.219 |
Leave | +0.218 |
eper | +0.203 |
vious | +0.203 |
Neg logit contributions | |
pron | -0.148 |
favourite | -0.146 |
masterpiece | -0.140 |
favorite | -0.134 |
usual | -0.128 |
Token | War |
Token ID | 1810 |
Feature activation | -1.51e+01 |
Pos logit contributions | |
crow | +0.009 |
eper | +0.009 |
Leave | +0.009 |
Bad | +0.008 |
vious | +0.008 |
Neg logit contributions | |
favourite | -0.011 |
favorite | -0.010 |
Daddy | -0.009 |
masterpiece | -0.009 |
pron | -0.009 |
Token | is |
Token ID | 318 |
Feature activation | -1.49e+00 |
Pos logit contributions | |
crow | +1.231 |
Leave | +1.201 |
Bad | +1.187 |
eper | +1.152 |
vious | +1.094 |
Neg logit contributions | |
favourite | -0.948 |
pron | -0.946 |
favorite | -0.928 |
masterpiece | -0.867 |
couldn | -0.822 |
Token | bad |
Token ID | 2089 |
Feature activation | -2.27e+00 |
Pos logit contributions | |
crow | +0.136 |
Leave | +0.129 |
Bad | +0.126 |
vious | +0.123 |
eper | +0.122 |
Neg logit contributions | |
pron | -0.094 |
favourite | -0.091 |
masterpiece | -0.088 |
favorite | -0.083 |
brand | -0.082 |
Token | and |
Token ID | 290 |
Feature activation | -1.06e-01 |
Pos logit contributions | |
Bad | +0.212 |
crow | +0.211 |
Leave | +0.208 |
eper | +0.198 |
Hey | +0.193 |
Neg logit contributions | |
favourite | -0.115 |
favorite | -0.109 |
pron | -0.103 |
masterpiece | -0.103 |
brand | -0.093 |
Token | scary |
Token ID | 14343 |
Feature activation | -2.61e+00 |
Pos logit contributions | |
crow | +0.008 |
eper | +0.008 |
Leave | +0.008 |
Bad | +0.007 |
ases | +0.007 |
Neg logit contributions | |
pron | -0.008 |
favourite | -0.008 |
masterpiece | -0.007 |
favorite | -0.007 |
brand | -0.007 |
Token | and |
Token ID | 290 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
crow | +0.254 |
Leave | +0.252 |
Bad | +0.252 |
eper | +0.239 |
Hey | +0.237 |
Neg logit contributions | |
favourite | -0.117 |
favorite | -0.112 |
masterpiece | -0.105 |
pron | -0.099 |
brand | -0.094 |
Token | I |
Token ID | 314 |
Feature activation | +1.29e+00 |
Pos logit contributions | |
Bad | +0.014 |
Leave | +0.013 |
crow | +0.013 |
eper | +0.013 |
vious | +0.012 |
Neg logit contributions | |
favourite | -0.017 |
pron | -0.017 |
masterpiece | -0.016 |
favorite | -0.016 |
Sus | -0.015 |
Token | teach |
Token ID | 4545 |
Feature activation | -4.20e+00 |
Pos logit contributions | |
pron | +0.105 |
favourite | +0.101 |
masterpiece | +0.099 |
favorite | +0.091 |
brand | +0.090 |
Neg logit contributions | |
crow | -0.092 |
Bad | -0.090 |
Leave | -0.089 |
vious | -0.082 |
eper | -0.082 |
Token | about |
Token ID | 546 |
Feature activation | +2.25e+00 |
Pos logit contributions | |
crow | +0.233 |
Bad | +0.226 |
Leave | +0.223 |
vious | +0.203 |
eper | +0.202 |
Neg logit contributions | |
pron | -0.407 |
favourite | -0.398 |
masterpiece | -0.391 |
brand | -0.365 |
favorite | -0.365 |
Token | war |
Token ID | 1175 |
Feature activation | -7.00e+00 |
Pos logit contributions | |
pron | +0.199 |
favourite | +0.187 |
masterpiece | +0.186 |
brand | +0.171 |
favorite | +0.167 |
Neg logit contributions | |
crow | -0.147 |
Bad | -0.145 |
Leave | -0.139 |
eper | -0.130 |
Stop | -0.130 |
Token | . |
Token ID | 13 |
Feature activation | -1.28e-01 |
Pos logit contributions | |
Leave | +0.607 |
Bad | +0.604 |
crow | +0.600 |
vious | +0.551 |
Stop | +0.546 |
Neg logit contributions | |
favourite | -0.456 |
pron | -0.453 |
masterpiece | -0.432 |
favorite | -0.414 |
brand | -0.387 |
Token | War |
Token ID | 1810 |
Feature activation | -1.50e+01 |
Pos logit contributions | |
crow | +0.009 |
Bad | +0.009 |
Leave | +0.008 |
eper | +0.008 |
vious | +0.008 |
Neg logit contributions | |
favourite | -0.010 |
pron | -0.010 |
masterpiece | -0.009 |
favorite | -0.009 |
Sus | -0.009 |
Token | is |
Token ID | 318 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
crow | +1.163 |
Leave | +1.159 |
Bad | +1.141 |
eper | +1.062 |
vious | +1.038 |
Neg logit contributions | |
pron | -1.066 |
favourite | -1.062 |
masterpiece | -1.010 |
favorite | -1.006 |
brand | -0.915 |
Token | when |
Token ID | 618 |
Feature activation | -7.18e+00 |
Pos logit contributions | |
crow | +0.114 |
Leave | +0.110 |
Bad | +0.107 |
vious | +0.104 |
eper | +0.102 |
Neg logit contributions | |
pron | -0.085 |
favourite | -0.082 |
masterpiece | -0.081 |
brand | -0.075 |
favorite | -0.075 |
Token | people |
Token ID | 661 |
Feature activation | -9.62e+00 |
Pos logit contributions | |
crow | +0.529 |
Bad | +0.513 |
Leave | +0.506 |
vious | +0.478 |
eper | +0.474 |
Neg logit contributions | |
pron | -0.571 |
favourite | -0.550 |
masterpiece | -0.541 |
brand | -0.496 |
favorite | -0.494 |
Token | fight |
Token ID | 1907 |
Feature activation | -4.96e+00 |
Pos logit contributions | |
crow | +0.751 |
Bad | +0.725 |
Leave | +0.723 |
eper | +0.676 |
vious | +0.654 |
Neg logit contributions | |
pron | -0.685 |
favourite | -0.646 |
masterpiece | -0.634 |
favorite | -0.600 |
collection | -0.588 |
Token | . |
Token ID | 13 |
Feature activation | -2.56e-02 |
Pos logit contributions | |
crow | +0.430 |
Leave | +0.428 |
Bad | +0.413 |
vious | +0.392 |
eper | +0.389 |
Neg logit contributions | |
pron | -0.302 |
favourite | -0.298 |
masterpiece | -0.293 |
favorite | -0.282 |
brand | -0.266 |
Token | ? |
Token ID | 30 |
Feature activation | +2.82e+00 |
Pos logit contributions | |
Leave | +0.252 |
Bad | +0.250 |
crow | +0.235 |
Stop | +0.220 |
eper | +0.216 |
Neg logit contributions | |
pron | -0.172 |
favourite | -0.169 |
masterpiece | -0.158 |
favorite | -0.151 |
collection | -0.151 |
Token | Share |
Token ID | 8734 |
Feature activation | -8.30e+00 |
Pos logit contributions | |
favourite | +0.240 |
pron | +0.237 |
masterpiece | +0.224 |
favorite | +0.218 |
brand | +0.211 |
Neg logit contributions | |
crow | -0.186 |
Leave | -0.179 |
Bad | -0.176 |
vious | -0.173 |
ases | -0.162 |
Token | ? |
Token ID | 30 |
Feature activation | +2.82e+00 |
Pos logit contributions | |
crow | +0.663 |
Leave | +0.657 |
Bad | +0.641 |
vious | +0.618 |
eper | +0.602 |
Neg logit contributions | |
pron | -0.605 |
favourite | -0.589 |
masterpiece | -0.565 |
favorite | -0.534 |
brand | -0.520 |
Token | Play |
Token ID | 3811 |
Feature activation | -1.13e+01 |
Pos logit contributions | |
favourite | +0.230 |
pron | +0.222 |
favorite | +0.206 |
masterpiece | +0.204 |
brand | +0.198 |
Neg logit contributions | |
crow | -0.188 |
Bad | -0.183 |
Leave | -0.181 |
vious | -0.176 |
eper | -0.169 |
Token | ? |
Token ID | 30 |
Feature activation | +2.82e+00 |
Pos logit contributions | |
Leave | +0.896 |
crow | +0.888 |
Bad | +0.879 |
vious | +0.851 |
Stop | +0.818 |
Neg logit contributions | |
pron | -0.843 |
favourite | -0.805 |
masterpiece | -0.765 |
favorite | -0.716 |
brand | -0.711 |
Token | Learn |
Token ID | 14365 |
Feature activation | -1.50e+01 |
Pos logit contributions | |
favourite | +0.231 |
pron | +0.225 |
favorite | +0.206 |
masterpiece | +0.206 |
brand | +0.200 |
Neg logit contributions | |
crow | -0.188 |
Bad | -0.182 |
Leave | -0.177 |
vious | -0.174 |
eper | -0.168 |
Token | ?" |
Token ID | 1701 |
Feature activation | +5.58e+00 |
Pos logit contributions | |
crow | +1.131 |
Bad | +1.112 |
Leave | +1.109 |
vious | +1.037 |
Stop | +1.011 |
Neg logit contributions | |
pron | -1.164 |
favourite | -1.133 |
masterpiece | -1.107 |
favorite | -1.027 |
brand | -1.018 |
Token | the |
Token ID | 262 |
Feature activation | -1.74e+00 |
Pos logit contributions | |
favourite | +0.323 |
masterpiece | +0.313 |
pron | +0.307 |
Sus | +0.289 |
Mill | +0.283 |
Neg logit contributions | |
crow | -0.481 |
Bad | -0.479 |
Leave | -0.472 |
vious | -0.464 |
ases | -0.457 |
Token | wolf |
Token ID | 17481 |
Feature activation | -3.04e+00 |
Pos logit contributions | |
crow | +0.109 |
Bad | +0.105 |
Leave | +0.104 |
ases | +0.097 |
vious | +0.097 |
Neg logit contributions | |
pron | -0.161 |
favourite | -0.154 |
masterpiece | -0.151 |
brand | -0.142 |
favorite | -0.140 |
Token | says |
Token ID | 1139 |
Feature activation | -2.08e+00 |
Pos logit contributions | |
Bad | +0.260 |
Leave | +0.251 |
crow | +0.245 |
Stop | +0.238 |
ases | +0.229 |
Neg logit contributions | |
pron | -0.185 |
favourite | -0.184 |
masterpiece | -0.182 |
favorite | -0.171 |
onna | -0.161 |
Token | . |
Token ID | 13 |
Feature activation | -8.39e-02 |
Pos logit contributions | |
crow | +0.178 |
Bad | +0.178 |
Leave | +0.173 |
Stop | +0.167 |
vious | +0.163 |
Neg logit contributions | |
pron | -0.142 |
favourite | -0.135 |
masterpiece | -0.132 |
favorite | -0.120 |
brand | -0.119 |
Token | Join |
Token ID | 18234 |
Feature activation | -5.91e+00 |
Pos logit contributions | |
pron | +0.034 |
favourite | +0.033 |
masterpiece | +0.033 |
favorite | +0.030 |
usual | +0.030 |
Neg logit contributions | |
crow | -0.026 |
Bad | -0.025 |
eper | -0.025 |
Leave | -0.024 |
hunters | -0.024 |
Token | the |
Token ID | 262 |
Feature activation | -1.90e+00 |
Pos logit contributions | |
Leave | +0.443 |
crow | +0.438 |
Bad | +0.431 |
vious | +0.409 |
eper | +0.404 |
Neg logit contributions | |
favourite | -0.456 |
pron | -0.443 |
masterpiece | -0.435 |
favorite | -0.412 |
brand | -0.399 |
Token | Jazz |
Token ID | 21406 |
Feature activation | -1.74e+00 |
Pos logit contributions | |
crow | +0.170 |
Bad | +0.166 |
Leave | +0.166 |
vious | +0.156 |
eper | +0.156 |
Neg logit contributions | |
pron | -0.123 |
favourite | -0.117 |
masterpiece | -0.116 |
favorite | -0.102 |
brand | -0.102 |
Token | Club |
Token ID | 6289 |
Feature activation | -4.06e+00 |
Pos logit contributions | |
crow | +0.131 |
Bad | +0.128 |
Leave | +0.126 |
vious | +0.118 |
eper | +0.117 |
Neg logit contributions | |
pron | -0.137 |
favourite | -0.130 |
masterpiece | -0.129 |
brand | -0.118 |
favorite | -0.116 |
Token | ! |
Token ID | 0 |
Feature activation | +3.01e+00 |
Pos logit contributions | |
crow | +0.308 |
Bad | +0.303 |
Leave | +0.298 |
vious | +0.278 |
Stop | +0.277 |
Neg logit contributions | |
pron | -0.315 |
favourite | -0.300 |
masterpiece | -0.296 |
brand | -0.270 |
favorite | -0.268 |
Token | Learn |
Token ID | 14365 |
Feature activation | -1.50e+01 |
Pos logit contributions | |
favourite | +0.232 |
pron | +0.221 |
Sus | +0.214 |
favorite | +0.209 |
masterpiece | +0.208 |
Neg logit contributions | |
Bad | -0.204 |
Leave | -0.190 |
crow | -0.188 |
eper | -0.179 |
vious | -0.174 |
Token | to |
Token ID | 284 |
Feature activation | +1.38e+00 |
Pos logit contributions | |
crow | +1.125 |
Bad | +1.095 |
Leave | +1.074 |
vious | +1.009 |
eper | +1.005 |
Neg logit contributions | |
pron | -1.194 |
favourite | -1.128 |
masterpiece | -1.118 |
brand | -1.021 |
favorite | -1.004 |
Token | play |
Token ID | 711 |
Feature activation | -1.16e+00 |
Pos logit contributions | |
pron | +0.106 |
favourite | +0.102 |
masterpiece | +0.101 |
brand | +0.091 |
favorite | +0.091 |
Neg logit contributions | |
crow | -0.104 |
Bad | -0.102 |
Leave | -0.100 |
vious | -0.095 |
eper | -0.094 |
Token | instruments |
Token ID | 12834 |
Feature activation | -3.38e+00 |
Pos logit contributions | |
crow | +0.095 |
Leave | +0.094 |
Bad | +0.093 |
vious | +0.089 |
eper | +0.087 |
Neg logit contributions | |
pron | -0.082 |
favourite | -0.081 |
masterpiece | -0.077 |
brand | -0.073 |
favorite | -0.072 |
Token | and |
Token ID | 290 |
Feature activation | +2.42e-01 |
Pos logit contributions | |
crow | +0.263 |
Bad | +0.257 |
Leave | +0.254 |
vious | +0.236 |
Stop | +0.235 |
Neg logit contributions | |
pron | -0.260 |
favourite | -0.250 |
masterpiece | -0.244 |
favorite | -0.223 |
brand | -0.223 |
Token | sing |
Token ID | 1702 |
Feature activation | -1.71e-01 |
Pos logit contributions | |
pron | +0.018 |
favourite | +0.017 |
masterpiece | +0.017 |
brand | +0.015 |
favorite | +0.015 |
Neg logit contributions | |
crow | -0.019 |
Bad | -0.019 |
Leave | -0.018 |
eper | -0.018 |
vious | -0.018 |
Token | people |
Token ID | 661 |
Feature activation | -9.76e+00 |
Pos logit contributions | |
Bad | +0.363 |
Leave | +0.348 |
crow | +0.343 |
eper | +0.330 |
Hey | +0.312 |
Neg logit contributions | |
favourite | -0.508 |
masterpiece | -0.489 |
favorite | -0.482 |
pron | -0.468 |
usual | -0.460 |
Token | and |
Token ID | 290 |
Feature activation | -1.45e-01 |
Pos logit contributions | |
Leave | +0.796 |
crow | +0.792 |
Bad | +0.783 |
Hey | +0.749 |
Stop | +0.749 |
Neg logit contributions | |
pron | -0.693 |
favourite | -0.672 |
masterpiece | -0.620 |
favorite | -0.619 |
collection | -0.578 |
Token | their |
Token ID | 511 |
Feature activation | +7.25e-01 |
Pos logit contributions | |
crow | +0.010 |
Bad | +0.009 |
Leave | +0.009 |
vious | +0.009 |
eper | +0.009 |
Neg logit contributions | |
pron | -0.012 |
favourite | -0.012 |
masterpiece | -0.012 |
favorite | -0.011 |
brand | -0.011 |
Token | feelings |
Token ID | 7666 |
Feature activation | -3.54e+00 |
Pos logit contributions | |
pron | +0.058 |
favourite | +0.057 |
masterpiece | +0.056 |
favorite | +0.052 |
brand | +0.051 |
Neg logit contributions | |
crow | -0.052 |
Bad | -0.051 |
Leave | -0.051 |
vious | -0.047 |
eper | -0.047 |
Token | . |
Token ID | 13 |
Feature activation | -5.99e-01 |
Pos logit contributions | |
crow | +0.296 |
Bad | +0.291 |
Leave | +0.288 |
vious | +0.269 |
eper | +0.269 |
Neg logit contributions | |
pron | -0.244 |
favourite | -0.235 |
masterpiece | -0.228 |
favorite | -0.210 |
brand | -0.208 |
Token | Learn |
Token ID | 14365 |
Feature activation | -1.49e+01 |
Pos logit contributions | |
crow | +0.044 |
Leave | +0.044 |
vious | +0.044 |
Bad | +0.042 |
eper | +0.041 |
Neg logit contributions | |
favourite | -0.040 |
pron | -0.040 |
Sus | -0.037 |
masterpiece | -0.037 |
favorite | -0.036 |
Token | to |
Token ID | 284 |
Feature activation | +1.29e+00 |
Pos logit contributions | |
crow | +1.140 |
Bad | +1.107 |
Leave | +1.072 |
ases | +1.021 |
eper | +1.018 |
Neg logit contributions | |
pron | -1.185 |
favourite | -1.102 |
masterpiece | -1.100 |
brand | -1.001 |
favorite | -0.961 |
Token | appreciate |
Token ID | 9144 |
Feature activation | +1.91e-01 |
Pos logit contributions | |
pron | +0.102 |
favourite | +0.097 |
masterpiece | +0.096 |
brand | +0.088 |
favorite | +0.087 |
Neg logit contributions | |
crow | -0.095 |
Bad | -0.092 |
Leave | -0.090 |
vious | -0.086 |
eper | -0.085 |
Token | and |
Token ID | 290 |
Feature activation | -1.94e-01 |
Pos logit contributions | |
pron | +0.016 |
favourite | +0.015 |
masterpiece | +0.015 |
brand | +0.014 |
favorite | +0.014 |
Neg logit contributions | |
crow | -0.014 |
Bad | -0.013 |
Leave | -0.013 |
vious | -0.012 |
eper | -0.012 |
Token | enjoy |
Token ID | 2883 |
Feature activation | +2.52e+00 |
Pos logit contributions | |
crow | +0.014 |
Bad | +0.014 |
Leave | +0.013 |
vious | +0.013 |
eper | +0.013 |
Neg logit contributions | |
pron | -0.015 |
favourite | -0.015 |
masterpiece | -0.015 |
brand | -0.013 |
favorite | -0.013 |
Token | different |
Token ID | 1180 |
Feature activation | -5.65e-01 |
Pos logit contributions | |
pron | +0.209 |
favourite | +0.203 |
masterpiece | +0.199 |
brand | +0.185 |
favorite | +0.184 |
Neg logit contributions | |
crow | -0.175 |
Leave | -0.167 |
Bad | -0.167 |
vious | -0.160 |
eper | -0.155 |
Token | It |
Token ID | 632 |
Feature activation | -2.96e-01 |
Pos logit contributions | |
Leave | +0.015 |
crow | +0.015 |
Bad | +0.015 |
eper | +0.015 |
Stop | +0.015 |
Neg logit contributions | |
favourite | -0.009 |
favorite | -0.008 |
Grand | -0.008 |
pron | -0.008 |
Sus | -0.008 |
Token | is |
Token ID | 318 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
crow | +0.024 |
Leave | +0.024 |
Bad | +0.023 |
ases | +0.022 |
eper | +0.022 |
Neg logit contributions | |
favourite | -0.020 |
pron | -0.020 |
masterpiece | -0.019 |
favorite | -0.018 |
brand | -0.017 |
Token | for |
Token ID | 329 |
Feature activation | +5.47e+00 |
Pos logit contributions | |
crow | +0.111 |
Bad | +0.110 |
Leave | +0.106 |
eper | +0.101 |
vious | +0.100 |
Neg logit contributions | |
pron | -0.099 |
masterpiece | -0.092 |
favourite | -0.092 |
brand | -0.082 |
favorite | -0.079 |
Token | war |
Token ID | 1175 |
Feature activation | -6.76e+00 |
Pos logit contributions | |
pron | +0.429 |
favourite | +0.414 |
masterpiece | +0.407 |
brand | +0.376 |
favorite | +0.372 |
Neg logit contributions | |
crow | -0.409 |
Bad | -0.391 |
Leave | -0.389 |
vious | -0.371 |
eper | -0.362 |
Token | . |
Token ID | 13 |
Feature activation | -7.98e-02 |
Pos logit contributions | |
crow | +0.566 |
Leave | +0.559 |
Bad | +0.559 |
vious | +0.516 |
eper | +0.512 |
Neg logit contributions | |
pron | -0.454 |
favourite | -0.451 |
masterpiece | -0.426 |
favorite | -0.415 |
brand | -0.389 |
Token | War |
Token ID | 1810 |
Feature activation | -1.48e+01 |
Pos logit contributions | |
crow | +0.006 |
eper | +0.006 |
Leave | +0.006 |
Bad | +0.006 |
vious | +0.005 |
Neg logit contributions | |
favourite | -0.005 |
Grand | -0.005 |
favorite | -0.005 |
pron | -0.005 |
Sus | -0.005 |
Token | is |
Token ID | 318 |
Feature activation | -1.32e+00 |
Pos logit contributions | |
crow | +1.182 |
Leave | +1.182 |
Bad | +1.156 |
eper | +1.101 |
vious | +1.070 |
Neg logit contributions | |
favourite | -0.962 |
favorite | -0.943 |
pron | -0.940 |
masterpiece | -0.865 |
couldn | -0.845 |
Token | bad |
Token ID | 2089 |
Feature activation | -2.00e+00 |
Pos logit contributions | |
crow | +0.119 |
Leave | +0.113 |
Bad | +0.109 |
vious | +0.108 |
eper | +0.106 |
Neg logit contributions | |
pron | -0.081 |
favourite | -0.081 |
masterpiece | -0.078 |
favorite | -0.075 |
brand | -0.074 |
Token | ," |
Token ID | 553 |
Feature activation | +4.74e+00 |
Pos logit contributions | |
Bad | +0.175 |
crow | +0.173 |
Leave | +0.173 |
eper | +0.160 |
Hey | +0.158 |
Neg logit contributions | |
favourite | -0.118 |
pron | -0.114 |
favorite | -0.111 |
masterpiece | -0.108 |
brand | -0.099 |
Token | Lily |
Token ID | 20037 |
Feature activation | +1.90e+00 |
Pos logit contributions | |
pron | +0.333 |
favourite | +0.314 |
masterpiece | +0.310 |
brand | +0.280 |
favorite | +0.278 |
Neg logit contributions | |
crow | -0.400 |
Bad | -0.389 |
Leave | -0.384 |
vious | -0.366 |
eper | -0.363 |
Token | said |
Token ID | 531 |
Feature activation | +9.63e-01 |
Pos logit contributions | |
pron | +0.090 |
favourite | +0.083 |
masterpiece | +0.076 |
favorite | +0.073 |
brand | +0.069 |
Neg logit contributions | |
crow | -0.195 |
Bad | -0.193 |
Leave | -0.192 |
ases | -0.184 |
eper | -0.180 |
Token | missile |
Token ID | 10105 |
Feature activation | -2.52e+00 |
Pos logit contributions | |
crow | +0.121 |
Leave | +0.120 |
ases | +0.115 |
Bad | +0.114 |
vious | +0.113 |
Neg logit contributions | |
favourite | -0.105 |
pron | -0.105 |
masterpiece | -0.104 |
favorite | -0.099 |
brand | -0.095 |
Token | was |
Token ID | 373 |
Feature activation | +1.22e+00 |
Pos logit contributions | |
crow | +0.216 |
Leave | +0.215 |
Bad | +0.212 |
Stop | +0.198 |
vious | +0.196 |
Neg logit contributions | |
pron | -0.157 |
favourite | -0.155 |
masterpiece | -0.149 |
favorite | -0.144 |
collection | -0.135 |
Token | for |
Token ID | 329 |
Feature activation | +5.50e+00 |
Pos logit contributions | |
pron | +0.087 |
favourite | +0.084 |
masterpiece | +0.082 |
favorite | +0.078 |
brand | +0.076 |
Neg logit contributions | |
crow | -0.100 |
Leave | -0.095 |
Bad | -0.093 |
vious | -0.090 |
Stop | -0.090 |
Token | war |
Token ID | 1175 |
Feature activation | -6.73e+00 |
Pos logit contributions | |
pron | +0.412 |
favourite | +0.404 |
masterpiece | +0.401 |
brand | +0.374 |
usual | +0.368 |
Neg logit contributions | |
crow | -0.425 |
Leave | -0.392 |
Bad | -0.388 |
vious | -0.382 |
eper | -0.372 |
Token | . |
Token ID | 13 |
Feature activation | -1.54e-01 |
Pos logit contributions | |
crow | +0.597 |
Bad | +0.596 |
Leave | +0.592 |
eper | +0.547 |
Stop | +0.542 |
Neg logit contributions | |
favourite | -0.407 |
pron | -0.406 |
masterpiece | -0.382 |
favorite | -0.375 |
usual | -0.346 |
Token | War |
Token ID | 1810 |
Feature activation | -1.48e+01 |
Pos logit contributions | |
crow | +0.013 |
eper | +0.012 |
Leave | +0.012 |
vious | +0.012 |
Bad | +0.012 |
Neg logit contributions | |
favourite | -0.010 |
favorite | -0.009 |
pron | -0.009 |
brand | -0.009 |
Grand | -0.009 |
Token | is |
Token ID | 318 |
Feature activation | -1.38e+00 |
Pos logit contributions | |
crow | +1.173 |
Leave | +1.156 |
Bad | +1.135 |
eper | +1.095 |
vious | +1.045 |
Neg logit contributions | |
pron | -0.997 |
favourite | -0.996 |
favorite | -0.957 |
masterpiece | -0.926 |
couldn | -0.862 |
Token | bad |
Token ID | 2089 |
Feature activation | -1.99e+00 |
Pos logit contributions | |
crow | +0.120 |
Leave | +0.113 |
Bad | +0.109 |
vious | +0.107 |
eper | +0.106 |
Neg logit contributions | |
pron | -0.092 |
favourite | -0.090 |
masterpiece | -0.088 |
brand | -0.084 |
favorite | -0.084 |
Token | . |
Token ID | 13 |
Feature activation | -1.08e-01 |
Pos logit contributions | |
Bad | +0.181 |
crow | +0.180 |
Leave | +0.178 |
eper | +0.167 |
vious | +0.162 |
Neg logit contributions | |
favourite | -0.111 |
pron | -0.107 |
favorite | -0.103 |
masterpiece | -0.102 |
brand | -0.094 |
Token | I |
Token ID | 314 |
Feature activation | +1.60e+00 |
Pos logit contributions | |
crow | +0.009 |
eper | +0.008 |
Leave | +0.008 |
vious | +0.008 |
thieves | +0.008 |
Neg logit contributions | |
favourite | -0.007 |
favorite | -0.007 |
Grand | -0.006 |
Daddy | -0.006 |
pron | -0.006 |
Token | don |
Token ID | 836 |
Feature activation | -4.25e+00 |
Pos logit contributions | |
favourite | +0.114 |
pron | +0.111 |
favorite | +0.107 |
masterpiece | +0.106 |
couldn | +0.095 |
Neg logit contributions | |
crow | -0.123 |
Leave | -0.121 |
Bad | -0.118 |
eper | -0.115 |
ases | -0.111 |
Token | Have |
Token ID | 11980 |
Feature activation | -4.31e+00 |
Pos logit contributions | |
pron | +0.049 |
favourite | +0.045 |
masterpiece | +0.043 |
favorite | +0.040 |
usual | +0.039 |
Neg logit contributions | |
crow | -0.036 |
eper | -0.033 |
ases | -0.032 |
Bad | -0.032 |
vious | -0.032 |
Token | fun |
Token ID | 1257 |
Feature activation | +1.02e+00 |
Pos logit contributions | |
Bad | +0.330 |
crow | +0.323 |
Leave | +0.322 |
Stop | +0.312 |
Hey | +0.306 |
Neg logit contributions | |
pron | -0.339 |
masterpiece | -0.317 |
favourite | -0.293 |
porch | -0.273 |
collection | -0.272 |
Token | , |
Token ID | 11 |
Feature activation | -1.29e+00 |
Pos logit contributions | |
pron | +0.082 |
favourite | +0.078 |
masterpiece | +0.077 |
brand | +0.070 |
favorite | +0.070 |
Neg logit contributions | |
crow | -0.076 |
Bad | -0.074 |
Leave | -0.072 |
vious | -0.068 |
eper | -0.067 |
Token | Lily |
Token ID | 20037 |
Feature activation | +1.95e+00 |
Pos logit contributions | |
crow | +0.110 |
Bad | +0.108 |
Leave | +0.106 |
vious | +0.100 |
eper | +0.099 |
Neg logit contributions | |
pron | -0.089 |
favourite | -0.084 |
masterpiece | -0.083 |
brand | -0.075 |
favorite | -0.073 |
Token | . |
Token ID | 13 |
Feature activation | -1.49e-01 |
Pos logit contributions | |
pron | +0.147 |
favourite | +0.137 |
masterpiece | +0.136 |
favorite | +0.123 |
brand | +0.123 |
Neg logit contributions | |
crow | -0.155 |
Bad | -0.150 |
Leave | -0.150 |
vious | -0.140 |
Stop | -0.139 |
Token | Learn |
Token ID | 14365 |
Feature activation | -1.48e+01 |
Pos logit contributions | |
crow | +0.009 |
Bad | +0.009 |
Leave | +0.009 |
eper | +0.008 |
vious | +0.008 |
Neg logit contributions | |
pron | -0.013 |
favourite | -0.013 |
masterpiece | -0.012 |
favorite | -0.012 |
Sus | -0.012 |
Token | a |
Token ID | 257 |
Feature activation | -1.02e-01 |
Pos logit contributions | |
crow | +1.125 |
Bad | +1.090 |
Leave | +1.043 |
ases | +1.021 |
eper | +1.013 |
Neg logit contributions | |
pron | -1.179 |
masterpiece | -1.087 |
favourite | -1.072 |
brand | -0.984 |
collection | -0.928 |
Token | lot |
Token ID | 1256 |
Feature activation | +3.29e+00 |
Pos logit contributions | |
Bad | +0.008 |
crow | +0.008 |
Leave | +0.008 |
Stop | +0.008 |
bully | +0.008 |
Neg logit contributions | |
pron | -0.007 |
masterpiece | -0.006 |
favourite | -0.006 |
brand | -0.005 |
usual | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | -1.70e-01 |
Pos logit contributions | |
pron | +0.266 |
favourite | +0.254 |
masterpiece | +0.249 |
favorite | +0.229 |
brand | +0.229 |
Neg logit contributions | |
crow | -0.237 |
Bad | -0.236 |
Leave | -0.235 |
vious | -0.214 |
Stop | -0.214 |
Token | We |
Token ID | 775 |
Feature activation | +7.59e-02 |
Pos logit contributions | |
crow | +0.011 |
Bad | +0.011 |
eper | +0.010 |
Leave | +0.010 |
vious | +0.010 |
Neg logit contributions | |
pron | -0.014 |
favourite | -0.014 |
masterpiece | -0.013 |
favorite | -0.012 |
Sus | -0.012 |
Token | are |
Token ID | 389 |
Feature activation | -4.09e+00 |
Pos logit contributions | |
pron | +0.006 |
favourite | +0.005 |
masterpiece | +0.005 |
favorite | +0.005 |
collection | +0.005 |
Neg logit contributions | |
crow | -0.006 |
Leave | -0.006 |
Bad | -0.006 |
vious | -0.005 |
Stop | -0.005 |
Token | , |
Token ID | 11 |
Feature activation | -1.28e+00 |
Pos logit contributions | |
pron | +0.085 |
masterpiece | +0.081 |
favourite | +0.079 |
brand | +0.072 |
favorite | +0.070 |
Neg logit contributions | |
crow | -0.079 |
Bad | -0.076 |
Leave | -0.075 |
eper | -0.071 |
vious | -0.070 |
Token | " |
Token ID | 366 |
Feature activation | +7.18e-01 |
Pos logit contributions | |
Bad | +0.098 |
Leave | +0.098 |
crow | +0.097 |
eper | +0.089 |
Stop | +0.089 |
Neg logit contributions | |
masterpiece | -0.091 |
pron | -0.091 |
favourite | -0.090 |
favorite | -0.079 |
brand | -0.079 |
Token | Me |
Token ID | 5308 |
Feature activation | -1.20e+00 |
Pos logit contributions | |
pron | +0.066 |
favourite | +0.062 |
masterpiece | +0.060 |
brand | +0.056 |
favorite | +0.055 |
Neg logit contributions | |
crow | -0.047 |
Bad | -0.043 |
Leave | -0.042 |
eper | -0.042 |
ases | -0.042 |
Token | too |
Token ID | 1165 |
Feature activation | -5.70e+00 |
Pos logit contributions | |
crow | +0.074 |
Bad | +0.071 |
Leave | +0.071 |
Stop | +0.064 |
vious | +0.063 |
Neg logit contributions | |
pron | -0.113 |
favourite | -0.110 |
masterpiece | -0.107 |
brand | -0.101 |
favorite | -0.101 |
Token | . |
Token ID | 13 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
crow | +0.446 |
Bad | +0.434 |
Leave | +0.427 |
vious | +0.401 |
eper | +0.399 |
Neg logit contributions | |
pron | -0.437 |
favourite | -0.415 |
masterpiece | -0.409 |
brand | -0.375 |
favorite | -0.369 |
Token | War |
Token ID | 1810 |
Feature activation | -1.48e+01 |
Pos logit contributions | |
pron | +0.009 |
favourite | +0.008 |
masterpiece | +0.008 |
brand | +0.008 |
favorite | +0.008 |
Neg logit contributions | |
crow | -0.007 |
Bad | -0.007 |
Leave | -0.007 |
vious | -0.006 |
eper | -0.006 |
Token | is |
Token ID | 318 |
Feature activation | -1.21e+00 |
Pos logit contributions | |
crow | +1.172 |
Leave | +1.140 |
Bad | +1.131 |
eper | +1.081 |
ases | +1.049 |
Neg logit contributions | |
pron | -1.033 |
favourite | -1.030 |
masterpiece | -0.965 |
favorite | -0.964 |
brand | -0.905 |
Token | bad |
Token ID | 2089 |
Feature activation | -2.05e+00 |
Pos logit contributions | |
crow | +0.099 |
Bad | +0.096 |
Leave | +0.095 |
vious | +0.090 |
eper | +0.089 |
Neg logit contributions | |
pron | -0.088 |
favourite | -0.083 |
masterpiece | -0.082 |
brand | -0.075 |
favorite | -0.074 |
Token | . |
Token ID | 13 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
crow | +0.176 |
Bad | +0.175 |
Leave | +0.173 |
eper | +0.160 |
vious | +0.158 |
Neg logit contributions | |
favourite | -0.131 |
pron | -0.129 |
masterpiece | -0.122 |
favorite | -0.119 |
brand | -0.112 |
Token | It |
Token ID | 632 |
Feature activation | -2.31e-01 |
Pos logit contributions | |
favourite | +0.008 |
pron | +0.008 |
masterpiece | +0.007 |
favorite | +0.007 |
brand | +0.007 |
Neg logit contributions | |
crow | -0.008 |
Leave | -0.007 |
Bad | -0.007 |
eper | -0.007 |
vious | -0.007 |
Token | makes |
Token ID | 1838 |
Feature activation | -4.88e-01 |
Pos logit contributions | |
crow | +0.019 |
Leave | +0.018 |
Bad | +0.018 |
ases | +0.017 |
eper | +0.017 |
Neg logit contributions | |
favourite | -0.016 |
pron | -0.016 |
masterpiece | -0.016 |
favorite | -0.015 |
brand | -0.014 |
Token | birds |
Token ID | 10087 |
Feature activation | -9.19e+00 |
Pos logit contributions | |
crow | +0.138 |
Bad | +0.134 |
Leave | +0.132 |
vious | +0.124 |
eper | +0.124 |
Neg logit contributions | |
pron | -0.119 |
favourite | -0.112 |
masterpiece | -0.111 |
brand | -0.100 |
favorite | -0.098 |
Token | ," |
Token ID | 553 |
Feature activation | +3.95e+00 |
Pos logit contributions | |
crow | +0.727 |
Bad | +0.711 |
Leave | +0.698 |
vious | +0.657 |
eper | +0.655 |
Neg logit contributions | |
pron | -0.684 |
favourite | -0.648 |
masterpiece | -0.640 |
brand | -0.582 |
favorite | -0.573 |
Token | said |
Token ID | 531 |
Feature activation | +3.74e-01 |
Pos logit contributions | |
pron | +0.259 |
favourite | +0.244 |
masterpiece | +0.239 |
brand | +0.214 |
favorite | +0.212 |
Neg logit contributions | |
crow | -0.348 |
Bad | -0.341 |
Leave | -0.337 |
vious | -0.320 |
eper | -0.318 |
Token | the |
Token ID | 262 |
Feature activation | -1.89e+00 |
Pos logit contributions | |
pron | +0.027 |
masterpiece | +0.026 |
favourite | +0.025 |
brand | +0.023 |
favorite | +0.022 |
Neg logit contributions | |
crow | -0.031 |
Bad | -0.030 |
Leave | -0.028 |
eper | -0.028 |
vious | -0.026 |
Token | scare |
Token ID | 19437 |
Feature activation | -4.75e+00 |
Pos logit contributions | |
crow | +0.135 |
Bad | +0.132 |
Leave | +0.129 |
vious | +0.121 |
eper | +0.120 |
Neg logit contributions | |
pron | -0.156 |
favourite | -0.148 |
masterpiece | -0.147 |
brand | -0.135 |
favorite | -0.133 |
Token | crow |
Token ID | 47114 |
Feature activation | -1.48e+01 |
Pos logit contributions | |
Bad | +0.099 |
Leave | +0.095 |
ases | +0.067 |
Stop | +0.063 |
vious | +0.062 |
Neg logit contributions | |
pron | -0.605 |
favourite | -0.588 |
masterpiece | -0.577 |
ail | -0.559 |
favorite | -0.553 |
Token | . |
Token ID | 13 |
Feature activation | -4.99e-01 |
Pos logit contributions | |
crow | +1.020 |
Bad | +0.950 |
eper | +0.923 |
Leave | +0.922 |
vious | +0.889 |
Neg logit contributions | |
pron | -1.249 |
masterpiece | -1.180 |
favourite | -1.154 |
brand | -1.075 |
favorite | -1.029 |
Token |
|
Token ID | 198 |
Feature activation | +3.21e-02 |
Pos logit contributions | |
crow | +0.040 |
Bad | +0.039 |
Leave | +0.038 |
vious | +0.036 |
eper | +0.036 |
Neg logit contributions | |
pron | -0.037 |
favourite | -0.035 |
masterpiece | -0.034 |
brand | -0.031 |
favorite | -0.031 |
Token |
|
Token ID | 198 |
Feature activation | -3.38e-02 |
Pos logit contributions | |
pron | +0.002 |
favourite | +0.002 |
masterpiece | +0.002 |
brand | +0.002 |
favorite | +0.002 |
Neg logit contributions | |
crow | -0.003 |
Bad | -0.003 |
Leave | -0.003 |
vious | -0.002 |
eper | -0.002 |
Token | " |
Token ID | 1 |
Feature activation | -5.26e+00 |
Pos logit contributions | |
vious | +0.002 |
crow | +0.002 |
ases | +0.002 |
eper | +0.002 |
ting | +0.002 |
Neg logit contributions | |
pron | -0.003 |
favourite | -0.002 |
masterpiece | -0.002 |
ail | -0.002 |
brand | -0.002 |
Token | Oh |
Token ID | 5812 |
Feature activation | -2.88e+00 |
Pos logit contributions | |
crow | +0.328 |
ases | +0.315 |
vious | +0.306 |
eper | +0.304 |
Bad | +0.295 |
Neg logit contributions | |
pron | -0.467 |
favourite | -0.437 |
masterpiece | -0.408 |
favorite | -0.388 |
usual | -0.378 |
Token | Join |
Token ID | 18234 |
Feature activation | -5.73e+00 |
Pos logit contributions | |
pron | +0.047 |
favourite | +0.045 |
masterpiece | +0.044 |
usual | +0.041 |
favorite | +0.041 |
Neg logit contributions | |
crow | -0.035 |
Bad | -0.034 |
eper | -0.034 |
Leave | -0.032 |
vious | -0.032 |
Token | the |
Token ID | 262 |
Feature activation | -1.59e+00 |
Pos logit contributions | |
Leave | +0.426 |
crow | +0.418 |
Bad | +0.414 |
vious | +0.394 |
eper | +0.390 |
Neg logit contributions | |
favourite | -0.447 |
pron | -0.427 |
masterpiece | -0.420 |
favorite | -0.406 |
brand | -0.387 |
Token | Little |
Token ID | 7703 |
Feature activation | +2.92e+00 |
Pos logit contributions | |
crow | +0.132 |
Bad | +0.130 |
Leave | +0.129 |
vious | +0.122 |
eper | +0.121 |
Neg logit contributions | |
pron | -0.111 |
favourite | -0.108 |
masterpiece | -0.105 |
favorite | -0.095 |
brand | -0.094 |
Token | League |
Token ID | 4041 |
Feature activation | -1.87e+00 |
Pos logit contributions | |
pron | +0.231 |
favourite | +0.223 |
masterpiece | +0.217 |
favorite | +0.201 |
usual | +0.195 |
Neg logit contributions | |
crow | -0.208 |
Bad | -0.206 |
Leave | -0.205 |
vious | -0.196 |
ases | -0.194 |
Token | ! |
Token ID | 0 |
Feature activation | +3.27e+00 |
Pos logit contributions | |
Bad | +0.143 |
Leave | +0.143 |
crow | +0.143 |
vious | +0.132 |
Stop | +0.130 |
Neg logit contributions | |
pron | -0.139 |
favourite | -0.135 |
masterpiece | -0.130 |
favorite | -0.121 |
collection | -0.117 |
Token | Learn |
Token ID | 14365 |
Feature activation | -1.47e+01 |
Pos logit contributions | |
favourite | +0.244 |
pron | +0.235 |
Sus | +0.234 |
Grand | +0.220 |
masterpiece | +0.219 |
Neg logit contributions | |
Bad | -0.218 |
Leave | -0.211 |
crow | -0.208 |
eper | -0.201 |
vious | -0.200 |
Token | to |
Token ID | 284 |
Feature activation | +1.56e+00 |
Pos logit contributions | |
crow | +1.149 |
Bad | +1.120 |
Leave | +1.100 |
vious | +1.035 |
eper | +1.031 |
Neg logit contributions | |
pron | -1.124 |
favourite | -1.063 |
masterpiece | -1.052 |
brand | -0.958 |
favorite | -0.943 |
Token | play |
Token ID | 711 |
Feature activation | -1.17e+00 |
Pos logit contributions | |
pron | +0.117 |
favourite | +0.112 |
masterpiece | +0.111 |
brand | +0.100 |
favorite | +0.100 |
Neg logit contributions | |
crow | -0.121 |
Bad | -0.119 |
Leave | -0.117 |
vious | -0.111 |
eper | -0.110 |
Token | baseball |
Token ID | 9283 |
Feature activation | +4.47e-01 |
Pos logit contributions | |
Leave | +0.098 |
crow | +0.097 |
Bad | +0.095 |
vious | +0.093 |
eper | +0.090 |
Neg logit contributions | |
favourite | -0.078 |
pron | -0.077 |
masterpiece | -0.072 |
favorite | -0.071 |
brand | -0.068 |
Token | with |
Token ID | 351 |
Feature activation | -1.84e+00 |
Pos logit contributions | |
pron | +0.031 |
favourite | +0.030 |
masterpiece | +0.029 |
favorite | +0.028 |
collection | +0.027 |
Neg logit contributions | |
Leave | -0.037 |
crow | -0.037 |
Bad | -0.037 |
vious | -0.034 |
eper | -0.033 |
Token | other |
Token ID | 584 |
Feature activation | -5.73e+00 |
Pos logit contributions | |
crow | +0.128 |
Leave | +0.126 |
Bad | +0.124 |
vious | +0.117 |
eper | +0.115 |
Neg logit contributions | |
pron | -0.153 |
favourite | -0.150 |
masterpiece | -0.145 |
favorite | -0.136 |
brand | -0.135 |
Token | any |
Token ID | 597 |
Feature activation | +1.25e-01 |
Pos logit contributions | |
crow | +0.368 |
Bad | +0.355 |
Leave | +0.349 |
eper | +0.328 |
vious | +0.327 |
Neg logit contributions | |
pron | -0.396 |
favourite | -0.380 |
masterpiece | -0.375 |
brand | -0.345 |
favorite | -0.341 |
Token | situation |
Token ID | 3074 |
Feature activation | -2.03e+00 |
Pos logit contributions | |
pron | +0.010 |
favourite | +0.010 |
masterpiece | +0.010 |
brand | +0.009 |
favorite | +0.009 |
Neg logit contributions | |
crow | -0.009 |
Bad | -0.009 |
Leave | -0.009 |
eper | -0.008 |
ases | -0.008 |
Token | too |
Token ID | 1165 |
Feature activation | -5.69e+00 |
Pos logit contributions | |
crow | +0.162 |
Bad | +0.160 |
Leave | +0.158 |
eper | +0.148 |
vious | +0.147 |
Neg logit contributions | |
pron | -0.146 |
favourite | -0.141 |
masterpiece | -0.138 |
brand | -0.127 |
favorite | -0.126 |
Token | seriously |
Token ID | 6411 |
Feature activation | -1.97e+00 |
Pos logit contributions | |
crow | +0.487 |
Bad | +0.476 |
Leave | +0.470 |
eper | +0.444 |
vious | +0.442 |
Neg logit contributions | |
pron | -0.390 |
favourite | -0.368 |
masterpiece | -0.362 |
brand | -0.327 |
favorite | -0.322 |
Token | . |
Token ID | 13 |
Feature activation | -1.26e-01 |
Pos logit contributions | |
crow | +0.152 |
Bad | +0.148 |
Leave | +0.145 |
eper | +0.137 |
ases | +0.136 |
Neg logit contributions | |
pron | -0.150 |
favourite | -0.142 |
masterpiece | -0.140 |
brand | -0.130 |
favorite | -0.127 |
Token | Learn |
Token ID | 14365 |
Feature activation | -1.47e+01 |
Pos logit contributions | |
Bad | +0.009 |
crow | +0.009 |
Leave | +0.009 |
vious | +0.008 |
eper | +0.008 |
Neg logit contributions | |
favourite | -0.010 |
pron | -0.010 |
masterpiece | -0.009 |
favorite | -0.009 |
Sus | -0.009 |
Token | to |
Token ID | 284 |
Feature activation | +1.70e+00 |
Pos logit contributions | |
crow | +1.276 |
Bad | +1.250 |
Leave | +1.233 |
vious | +1.166 |
eper | +1.159 |
Neg logit contributions | |
pron | -0.984 |
favourite | -0.931 |
masterpiece | -0.916 |
brand | -0.824 |
favorite | -0.814 |
Token | laugh |
Token ID | 6487 |
Feature activation | -1.78e+00 |
Pos logit contributions | |
pron | +0.133 |
favourite | +0.126 |
masterpiece | +0.125 |
brand | +0.114 |
favorite | +0.112 |
Neg logit contributions | |
crow | -0.129 |
Bad | -0.125 |
Leave | -0.123 |
vious | -0.116 |
eper | -0.115 |
Token | at |
Token ID | 379 |
Feature activation | +1.76e-03 |
Pos logit contributions | |
crow | +0.142 |
Bad | +0.140 |
Leave | +0.139 |
eper | +0.130 |
ases | +0.128 |
Neg logit contributions | |
pron | -0.129 |
favourite | -0.127 |
masterpiece | -0.121 |
favorite | -0.113 |
brand | -0.112 |
Token | yourself |
Token ID | 3511 |
Feature activation | -6.38e+00 |
Pos logit contributions | |
pron | +0.000 |
favourite | +0.000 |
masterpiece | +0.000 |
brand | +0.000 |
favorite | +0.000 |
Neg logit contributions | |
crow | -0.000 |
Leave | -0.000 |
Bad | -0.000 |
vious | -0.000 |
eper | -0.000 |
Token | and |
Token ID | 290 |
Feature activation | +2.36e-01 |
Pos logit contributions | |
crow | +0.495 |
Bad | +0.493 |
Leave | +0.490 |
ases | +0.451 |
Stop | +0.450 |
Neg logit contributions | |
pron | -0.477 |
favourite | -0.460 |
masterpiece | -0.446 |
brand | -0.411 |
favorite | -0.408 |
Token | ockey |
Token ID | 8337 |
Feature activation | +3.92e-01 |
Pos logit contributions | |
pron | +0.056 |
favourite | +0.053 |
masterpiece | +0.052 |
collection | +0.048 |
brand | +0.046 |
Neg logit contributions | |
Bad | -0.066 |
Leave | -0.066 |
crow | -0.066 |
Stop | -0.061 |
eper | -0.061 |
Token | Camp |
Token ID | 5425 |
Feature activation | -4.12e+00 |
Pos logit contributions | |
pron | +0.030 |
favourite | +0.029 |
masterpiece | +0.028 |
brand | +0.026 |
favorite | +0.026 |
Neg logit contributions | |
Leave | -0.030 |
Bad | -0.029 |
crow | -0.029 |
eper | -0.027 |
vious | -0.027 |
Token | for |
Token ID | 329 |
Feature activation | +5.62e+00 |
Pos logit contributions | |
crow | +0.304 |
Bad | +0.302 |
Leave | +0.300 |
vious | +0.276 |
eper | +0.273 |
Neg logit contributions | |
pron | -0.327 |
favourite | -0.314 |
masterpiece | -0.307 |
brand | -0.282 |
favorite | -0.281 |
Token | Kids |
Token ID | 17476 |
Feature activation | -1.86e+00 |
Pos logit contributions | |
pron | +0.456 |
favourite | +0.442 |
masterpiece | +0.432 |
brand | +0.404 |
favorite | +0.397 |
Neg logit contributions | |
crow | -0.409 |
Leave | -0.388 |
Bad | -0.388 |
vious | -0.368 |
eper | -0.361 |
Token | . |
Token ID | 13 |
Feature activation | +8.59e-02 |
Pos logit contributions | |
Bad | +0.145 |
crow | +0.142 |
Leave | +0.142 |
vious | +0.130 |
Stop | +0.130 |
Neg logit contributions | |
pron | -0.137 |
favourite | -0.136 |
masterpiece | -0.130 |
favorite | -0.121 |
usual | -0.119 |
Token | Learn |
Token ID | 14365 |
Feature activation | -1.46e+01 |
Pos logit contributions | |
favourite | +0.006 |
Sus | +0.006 |
pron | +0.006 |
favorite | +0.006 |
masterpiece | +0.006 |
Neg logit contributions | |
Bad | -0.006 |
Leave | -0.006 |
crow | -0.006 |
vious | -0.005 |
eper | -0.005 |
Token | from |
Token ID | 422 |
Feature activation | -3.66e+00 |
Pos logit contributions | |
crow | +1.208 |
Bad | +1.187 |
Leave | +1.175 |
vious | +1.105 |
Stop | +1.090 |
Neg logit contributions | |
pron | -1.038 |
favourite | -0.999 |
masterpiece | -0.976 |
brand | -0.890 |
favorite | -0.888 |
Token | the |
Token ID | 262 |
Feature activation | -1.27e+00 |
Pos logit contributions | |
crow | +0.273 |
Bad | +0.264 |
Leave | +0.262 |
vious | +0.246 |
Stop | +0.242 |
Neg logit contributions | |
pron | -0.292 |
favourite | -0.279 |
masterpiece | -0.275 |
brand | -0.253 |
favorite | -0.250 |
Token | Best |
Token ID | 6705 |
Feature activation | -2.82e+00 |
Pos logit contributions | |
crow | +0.110 |
Leave | +0.107 |
Bad | +0.104 |
vious | +0.102 |
ases | +0.102 |
Neg logit contributions | |
pron | -0.086 |
favourite | -0.084 |
masterpiece | -0.082 |
usual | -0.075 |
favorite | -0.075 |
Token | . |
Token ID | 13 |
Feature activation | +4.66e-02 |
Pos logit contributions | |
Bad | +0.203 |
Leave | +0.202 |
crow | +0.201 |
vious | +0.180 |
eper | +0.179 |
Neg logit contributions | |
favourite | -0.226 |
pron | -0.225 |
masterpiece | -0.216 |
favorite | -0.205 |
collection | -0.201 |
Token | Sign |
Token ID | 5865 |
Feature activation | -5.95e+00 |
Pos logit contributions | |
favourite | +0.004 |
pron | +0.003 |
Sus | +0.003 |
masterpiece | +0.003 |
Mill | +0.003 |
Neg logit contributions | |
Bad | -0.003 |
Leave | -0.003 |
crow | -0.003 |
eper | -0.003 |
vious | -0.003 |
Token | Stop |
Token ID | 19485 |
Feature activation | -1.16e+01 |
Pos logit contributions | |
pron | +0.023 |
masterpiece | +0.021 |
brushes | +0.021 |
favourite | +0.021 |
Sus | +0.021 |
Neg logit contributions | |
crow | -0.021 |
vious | -0.020 |
hunters | -0.020 |
ases | -0.020 |
eper | -0.019 |
Token | it |
Token ID | 340 |
Feature activation | +1.83e-01 |
Pos logit contributions | |
crow | +0.990 |
Leave | +0.917 |
Bad | +0.897 |
vious | +0.896 |
ases | +0.875 |
Neg logit contributions | |
pron | -0.807 |
masterpiece | -0.768 |
favourite | -0.758 |
brand | -0.709 |
favorite | -0.682 |
Token | , |
Token ID | 11 |
Feature activation | -1.56e+00 |
Pos logit contributions | |
pron | +0.010 |
favourite | +0.009 |
masterpiece | +0.009 |
favorite | +0.008 |
brand | +0.008 |
Neg logit contributions | |
crow | -0.018 |
Leave | -0.017 |
Bad | -0.017 |
vious | -0.016 |
eper | -0.016 |
Token | girls |
Token ID | 4813 |
Feature activation | -5.39e+00 |
Pos logit contributions | |
crow | +0.116 |
Leave | +0.112 |
Bad | +0.110 |
vious | +0.104 |
ases | +0.102 |
Neg logit contributions | |
pron | -0.119 |
favourite | -0.117 |
masterpiece | -0.114 |
favorite | -0.107 |
brand | -0.104 |
Token | ! |
Token ID | 0 |
Feature activation | +2.76e+00 |
Pos logit contributions | |
crow | +0.480 |
Bad | +0.467 |
Leave | +0.464 |
vious | +0.437 |
eper | +0.431 |
Neg logit contributions | |
pron | -0.338 |
masterpiece | -0.314 |
favourite | -0.314 |
collection | -0.281 |
favorite | -0.278 |
Token | Fighting |
Token ID | 19098 |
Feature activation | -1.46e+01 |
Pos logit contributions | |
pron | +0.235 |
favourite | +0.234 |
Sus | +0.228 |
masterpiece | +0.224 |
Daddy | +0.216 |
Neg logit contributions | |
crow | -0.144 |
vious | -0.140 |
Leave | -0.139 |
Bad | -0.137 |
hunters | -0.137 |
Token | is |
Token ID | 318 |
Feature activation | -1.43e+00 |
Pos logit contributions | |
crow | +1.081 |
Leave | +1.027 |
Bad | +1.024 |
vious | +0.957 |
eper | +0.946 |
Neg logit contributions | |
pron | -1.167 |
favourite | -1.116 |
masterpiece | -1.100 |
favorite | -1.035 |
brand | -1.025 |
Token | not |
Token ID | 407 |
Feature activation | +1.92e+00 |
Pos logit contributions | |
crow | +0.130 |
Leave | +0.123 |
Bad | +0.118 |
vious | +0.117 |
eper | +0.115 |
Neg logit contributions | |
pron | -0.092 |
favourite | -0.086 |
masterpiece | -0.085 |
brand | -0.080 |
favorite | -0.080 |
Token | nice |
Token ID | 3621 |
Feature activation | -4.04e-01 |
Pos logit contributions | |
pron | +0.123 |
favourite | +0.114 |
masterpiece | +0.111 |
favorite | +0.104 |
brand | +0.104 |
Neg logit contributions | |
crow | -0.177 |
Leave | -0.169 |
Bad | -0.165 |
eper | -0.160 |
vious | -0.159 |
Token | . |
Token ID | 13 |
Feature activation | +2.49e-02 |
Pos logit contributions | |
crow | +0.036 |
Leave | +0.036 |
Bad | +0.036 |
eper | +0.033 |
vious | +0.032 |
Neg logit contributions | |
pron | -0.024 |
favourite | -0.024 |
masterpiece | -0.024 |
favorite | -0.022 |
brand | -0.021 |
Token | You |
Token ID | 921 |
Feature activation | -5.89e+00 |
Pos logit contributions | |
favourite | +0.002 |
pron | +0.002 |
masterpiece | +0.002 |
favorite | +0.002 |
Sus | +0.002 |
Neg logit contributions | |
crow | -0.002 |
Leave | -0.002 |
Bad | -0.002 |
eper | -0.002 |
hunters | -0.001 |
Token | for |
Token ID | 329 |
Feature activation | +5.21e+00 |
Pos logit contributions | |
crow | +0.087 |
Bad | +0.084 |
Leave | +0.082 |
eper | +0.077 |
vious | +0.077 |
Neg logit contributions | |
pron | -0.088 |
masterpiece | -0.082 |
favourite | -0.082 |
brand | -0.075 |
favorite | -0.073 |
Token | help |
Token ID | 1037 |
Feature activation | -2.70e+00 |
Pos logit contributions | |
pron | +0.419 |
favourite | +0.403 |
masterpiece | +0.395 |
brand | +0.365 |
favorite | +0.360 |
Neg logit contributions | |
crow | -0.379 |
Bad | -0.370 |
Leave | -0.364 |
vious | -0.344 |
Stop | -0.339 |
Token | . |
Token ID | 13 |
Feature activation | -4.03e-01 |
Pos logit contributions | |
crow | +0.218 |
Bad | +0.216 |
Leave | +0.212 |
vious | +0.201 |
Stop | +0.199 |
Neg logit contributions | |
pron | -0.197 |
favourite | -0.187 |
masterpiece | -0.184 |
brand | -0.167 |
favorite | -0.164 |
Token | The |
Token ID | 383 |
Feature activation | -2.06e-01 |
Pos logit contributions | |
crow | +0.032 |
Bad | +0.031 |
Leave | +0.031 |
vious | +0.029 |
eper | +0.029 |
Neg logit contributions | |
pron | -0.030 |
favourite | -0.028 |
masterpiece | -0.028 |
brand | -0.025 |
favorite | -0.025 |
Token | scare |
Token ID | 19437 |
Feature activation | -4.52e+00 |
Pos logit contributions | |
Leave | +0.013 |
Bad | +0.013 |
ases | +0.013 |
crow | +0.013 |
vious | +0.012 |
Neg logit contributions | |
pron | -0.017 |
favourite | -0.017 |
masterpiece | -0.017 |
favorite | -0.015 |
brand | -0.015 |
Token | crow |
Token ID | 47114 |
Feature activation | -1.45e+01 |
Pos logit contributions | |
Bad | +0.068 |
Leave | +0.063 |
ases | +0.045 |
eper | +0.035 |
Stop | +0.035 |
Neg logit contributions | |
pron | -0.587 |
favourite | -0.583 |
masterpiece | -0.569 |
ail | -0.549 |
usual | -0.546 |
Token | said |
Token ID | 531 |
Feature activation | +9.58e-01 |
Pos logit contributions | |
crow | +1.255 |
Bad | +1.211 |
Leave | +1.179 |
vious | +1.145 |
eper | +1.144 |
Neg logit contributions | |
pron | -0.989 |
masterpiece | -0.912 |
favourite | -0.902 |
brand | -0.816 |
favorite | -0.781 |
Token | " |
Token ID | 366 |
Feature activation | +6.93e-01 |
Pos logit contributions | |
pron | +0.066 |
favourite | +0.062 |
masterpiece | +0.062 |
brand | +0.055 |
favorite | +0.054 |
Neg logit contributions | |
crow | -0.082 |
Bad | -0.080 |
Leave | -0.079 |
vious | -0.074 |
eper | -0.074 |
Token | Come |
Token ID | 16773 |
Feature activation | -8.37e+00 |
Pos logit contributions | |
pron | +0.061 |
favourite | +0.055 |
masterpiece | +0.055 |
usual | +0.051 |
favorite | +0.049 |
Neg logit contributions | |
crow | -0.044 |
Bad | -0.042 |
eper | -0.041 |
ases | -0.041 |
Leave | -0.041 |
Token | with |
Token ID | 351 |
Feature activation | -1.90e+00 |
Pos logit contributions | |
crow | +0.673 |
Bad | +0.663 |
vious | +0.648 |
Leave | +0.644 |
ases | +0.618 |
Neg logit contributions | |
pron | -0.609 |
favourite | -0.583 |
masterpiece | -0.553 |
brand | -0.516 |
favorite | -0.509 |
Token | me |
Token ID | 502 |
Feature activation | -2.71e+00 |
Pos logit contributions | |
crow | +0.126 |
Leave | +0.126 |
Bad | +0.124 |
vious | +0.117 |
ases | +0.114 |
Neg logit contributions | |
pron | -0.163 |
favourite | -0.155 |
masterpiece | -0.152 |
brand | -0.141 |
favorite | -0.140 |
Token | share |
Token ID | 2648 |
Feature activation | -3.17e+00 |
Pos logit contributions | |
pron | +0.094 |
masterpiece | +0.088 |
favourite | +0.087 |
favorite | +0.079 |
brand | +0.079 |
Neg logit contributions | |
crow | -0.100 |
Leave | -0.095 |
Bad | -0.095 |
vious | -0.093 |
eper | -0.090 |
Token | and |
Token ID | 290 |
Feature activation | +9.59e-02 |
Pos logit contributions | |
crow | +0.279 |
Leave | +0.274 |
Bad | +0.264 |
vious | +0.261 |
eper | +0.252 |
Neg logit contributions | |
pron | -0.206 |
favourite | -0.199 |
masterpiece | -0.190 |
favorite | -0.185 |
brand | -0.178 |
Token | be |
Token ID | 307 |
Feature activation | +8.15e-01 |
Pos logit contributions | |
pron | +0.008 |
favourite | +0.007 |
masterpiece | +0.007 |
brand | +0.007 |
favorite | +0.007 |
Neg logit contributions | |
crow | -0.007 |
Bad | -0.007 |
Leave | -0.007 |
vious | -0.006 |
Stop | -0.006 |
Token | nice |
Token ID | 3621 |
Feature activation | -3.97e-01 |
Pos logit contributions | |
pron | +0.049 |
favourite | +0.048 |
favorite | +0.045 |
masterpiece | +0.044 |
brand | +0.043 |
Neg logit contributions | |
crow | -0.077 |
Leave | -0.071 |
Bad | -0.070 |
vious | -0.068 |
Hey | -0.068 |
Token | . |
Token ID | 13 |
Feature activation | -2.34e-02 |
Pos logit contributions | |
crow | +0.036 |
Bad | +0.035 |
Leave | +0.035 |
Hey | +0.033 |
Stop | +0.032 |
Neg logit contributions | |
favourite | -0.024 |
pron | -0.024 |
masterpiece | -0.023 |
favorite | -0.023 |
brand | -0.021 |
Token | Fighting |
Token ID | 19098 |
Feature activation | -1.45e+01 |
Pos logit contributions | |
crow | +0.001 |
Leave | +0.001 |
Bad | +0.001 |
vious | +0.001 |
eper | +0.001 |
Neg logit contributions | |
favourite | -0.002 |
pron | -0.002 |
Grand | -0.002 |
favorite | -0.002 |
masterpiece | -0.002 |
Token | is |
Token ID | 318 |
Feature activation | -1.53e+00 |
Pos logit contributions | |
crow | +1.020 |
Bad | +0.989 |
Leave | +0.979 |
vious | +0.913 |
eper | +0.902 |
Neg logit contributions | |
pron | -1.224 |
favourite | -1.168 |
masterpiece | -1.147 |
favorite | -1.070 |
brand | -1.063 |
Token | bad |
Token ID | 2089 |
Feature activation | -2.20e+00 |
Pos logit contributions | |
crow | +0.137 |
Leave | +0.131 |
Bad | +0.128 |
vious | +0.124 |
eper | +0.123 |
Neg logit contributions | |
pron | -0.100 |
favourite | -0.095 |
masterpiece | -0.093 |
favorite | -0.087 |
brand | -0.086 |
Token | . |
Token ID | 13 |
Feature activation | -5.12e-02 |
Pos logit contributions | |
Leave | +0.199 |
Bad | +0.197 |
crow | +0.196 |
eper | +0.182 |
vious | +0.180 |
Neg logit contributions | |
favourite | -0.120 |
pron | -0.120 |
favorite | -0.115 |
masterpiece | -0.111 |
collection | -0.100 |
Token | It |
Token ID | 632 |
Feature activation | -4.50e-01 |
Pos logit contributions | |
crow | +0.003 |
Leave | +0.003 |
eper | +0.003 |
Bad | +0.003 |
vious | +0.003 |
Neg logit contributions | |
favourite | -0.004 |
pron | -0.004 |
favorite | -0.004 |
masterpiece | -0.004 |
Grand | -0.004 |
Token | hurts |
Token ID | 20406 |
Feature activation | -1.80e+00 |
Pos logit contributions | |
crow | +0.036 |
Leave | +0.035 |
Bad | +0.033 |
eper | +0.032 |
ases | +0.031 |
Neg logit contributions | |
pron | -0.031 |
favourite | -0.030 |
favorite | -0.029 |
masterpiece | -0.029 |
brand | -0.027 |
Token | talk |
Token ID | 1561 |
Feature activation | -5.14e+00 |
Pos logit contributions | |
pron | +0.094 |
favourite | +0.088 |
masterpiece | +0.088 |
brand | +0.079 |
favorite | +0.079 |
Neg logit contributions | |
crow | -0.095 |
Bad | -0.092 |
Leave | -0.091 |
vious | -0.087 |
eper | -0.086 |
Token | calm |
Token ID | 9480 |
Feature activation | -3.31e+00 |
Pos logit contributions | |
crow | +0.508 |
Bad | +0.498 |
Leave | +0.494 |
vious | +0.468 |
eper | +0.467 |
Neg logit contributions | |
pron | -0.284 |
favourite | -0.265 |
masterpiece | -0.257 |
brand | -0.227 |
favorite | -0.226 |
Token | and |
Token ID | 290 |
Feature activation | +3.97e-02 |
Pos logit contributions | |
Bad | +0.272 |
Leave | +0.272 |
crow | +0.265 |
eper | +0.248 |
vious | +0.247 |
Neg logit contributions | |
favourite | -0.217 |
pron | -0.212 |
favorite | -0.204 |
masterpiece | -0.203 |
collection | -0.180 |
Token | polite |
Token ID | 23507 |
Feature activation | -4.04e+00 |
Pos logit contributions | |
pron | +0.003 |
favourite | +0.003 |
masterpiece | +0.003 |
brand | +0.003 |
favorite | +0.003 |
Neg logit contributions | |
crow | -0.003 |
Bad | -0.003 |
Leave | -0.003 |
vious | -0.003 |
eper | -0.003 |
Token | . |
Token ID | 13 |
Feature activation | -2.53e-02 |
Pos logit contributions | |
Leave | +0.364 |
Bad | +0.357 |
crow | +0.354 |
eper | +0.328 |
vious | +0.324 |
Neg logit contributions | |
favourite | -0.239 |
pron | -0.237 |
favorite | -0.225 |
masterpiece | -0.222 |
usual | -0.199 |
Token | Fighting |
Token ID | 19098 |
Feature activation | -1.45e+01 |
Pos logit contributions | |
crow | +0.002 |
Bad | +0.002 |
Leave | +0.001 |
eper | +0.001 |
vious | +0.001 |
Neg logit contributions | |
favourite | -0.002 |
pron | -0.002 |
favorite | -0.002 |
masterpiece | -0.002 |
brand | -0.002 |
Token | over |
Token ID | 625 |
Feature activation | -2.15e+00 |
Pos logit contributions | |
crow | +1.023 |
Bad | +0.993 |
Leave | +0.984 |
vious | +0.912 |
eper | +0.905 |
Neg logit contributions | |
pron | -1.219 |
favourite | -1.164 |
masterpiece | -1.143 |
favorite | -1.068 |
brand | -1.057 |
Token | shapes |
Token ID | 15268 |
Feature activation | -2.06e+00 |
Pos logit contributions | |
crow | +0.154 |
Leave | +0.153 |
Bad | +0.150 |
vious | +0.140 |
eper | +0.138 |
Neg logit contributions | |
pron | -0.168 |
favourite | -0.162 |
masterpiece | -0.160 |
favorite | -0.151 |
brand | -0.145 |
Token | is |
Token ID | 318 |
Feature activation | -1.46e+00 |
Pos logit contributions | |
crow | +0.157 |
Bad | +0.155 |
Leave | +0.153 |
eper | +0.142 |
vious | +0.140 |
Neg logit contributions | |
pron | -0.157 |
favourite | -0.149 |
masterpiece | -0.147 |
favorite | -0.136 |
brand | -0.133 |
Token | not |
Token ID | 407 |
Feature activation | +1.93e+00 |
Pos logit contributions | |
crow | +0.126 |
Bad | +0.123 |
Leave | +0.122 |
vious | +0.115 |
eper | +0.115 |
Neg logit contributions | |
pron | -0.099 |
favourite | -0.093 |
masterpiece | -0.092 |
brand | -0.082 |
favorite | -0.081 |
Token | fun |
Token ID | 1257 |
Feature activation | +8.89e-01 |
Pos logit contributions | |
pron | +0.134 |
favourite | +0.127 |
masterpiece | +0.125 |
brand | +0.113 |
favorite | +0.113 |
Neg logit contributions | |
crow | -0.164 |
Bad | -0.159 |
Leave | -0.158 |
vious | -0.148 |
eper | -0.148 |
Token | enough |
Token ID | 1576 |
Feature activation | -3.32e-01 |
Pos logit contributions | |
crow | +0.319 |
Bad | +0.311 |
Leave | +0.301 |
vious | +0.290 |
eper | +0.283 |
Neg logit contributions | |
pron | -0.326 |
masterpiece | -0.306 |
favourite | -0.299 |
brand | -0.273 |
favorite | -0.262 |
Token | money |
Token ID | 1637 |
Feature activation | -6.51e-01 |
Pos logit contributions | |
crow | +0.028 |
Bad | +0.027 |
Leave | +0.026 |
eper | +0.025 |
ases | +0.025 |
Neg logit contributions | |
pron | -0.024 |
masterpiece | -0.022 |
favourite | -0.021 |
brand | -0.019 |
favorite | -0.018 |
Token | for |
Token ID | 329 |
Feature activation | +5.34e+00 |
Pos logit contributions | |
crow | +0.050 |
Bad | +0.049 |
Leave | +0.047 |
eper | +0.046 |
vious | +0.045 |
Neg logit contributions | |
pron | -0.049 |
masterpiece | -0.047 |
favourite | -0.046 |
brand | -0.041 |
favorite | -0.040 |
Token | an |
Token ID | 281 |
Feature activation | +2.54e+00 |
Pos logit contributions | |
pron | +0.449 |
favourite | +0.426 |
masterpiece | +0.424 |
brand | +0.385 |
favorite | +0.382 |
Neg logit contributions | |
crow | -0.370 |
Bad | -0.369 |
Leave | -0.358 |
eper | -0.332 |
Stop | -0.329 |
Token | adventurous |
Token ID | 39770 |
Feature activation | +2.85e-01 |
Pos logit contributions | |
pron | +0.189 |
favourite | +0.174 |
masterpiece | +0.174 |
brand | +0.158 |
favorite | +0.153 |
Neg logit contributions | |
crow | -0.203 |
Bad | -0.195 |
Leave | -0.193 |
vious | -0.185 |
eper | -0.184 |
Token | holiday |
Token ID | 9912 |
Feature activation | +8.87e+00 |
Pos logit contributions | |
pron | +0.018 |
favourite | +0.017 |
masterpiece | +0.017 |
brand | +0.015 |
favorite | +0.015 |
Neg logit contributions | |
crow | -0.026 |
Bad | -0.024 |
Leave | -0.024 |
vious | -0.023 |
eper | -0.023 |
Token | !" |
Token ID | 2474 |
Feature activation | +2.98e+00 |
Pos logit contributions | |
pron | +0.763 |
masterpiece | +0.721 |
favourite | +0.707 |
brand | +0.661 |
favorite | +0.622 |
Neg logit contributions | |
crow | -0.607 |
Bad | -0.546 |
Leave | -0.529 |
eper | -0.523 |
vious | -0.521 |
Token |
|
Token ID | 198 |
Feature activation | +4.94e-01 |
Pos logit contributions | |
pron | +0.219 |
favourite | +0.206 |
masterpiece | +0.205 |
brand | +0.186 |
favorite | +0.182 |
Neg logit contributions | |
crow | -0.241 |
Bad | -0.235 |
Leave | -0.231 |
vious | -0.218 |
eper | -0.217 |
Token |
|
Token ID | 198 |
Feature activation | +4.08e-01 |
Pos logit contributions | |
pron | +0.038 |
favourite | +0.035 |
masterpiece | +0.035 |
brand | +0.032 |
favorite | +0.031 |
Neg logit contributions | |
crow | -0.039 |
Bad | -0.038 |
Leave | -0.037 |
vious | -0.036 |
eper | -0.035 |
Token | Da |
Token ID | 26531 |
Feature activation | +7.15e-01 |
Pos logit contributions | |
pron | +0.034 |
favourite | +0.031 |
masterpiece | +0.031 |
brand | +0.027 |
usual | +0.027 |
Neg logit contributions | |
crow | -0.029 |
vious | -0.028 |
Bad | -0.028 |
Leave | -0.027 |
ases | -0.027 |
Token | isy |
Token ID | 13560 |
Feature activation | +7.65e+00 |
Pos logit contributions | |
pron | +0.039 |
favourite | +0.035 |
masterpiece | +0.034 |
brand | +0.031 |
favorite | +0.029 |
Neg logit contributions | |
crow | -0.072 |
Bad | -0.070 |
Leave | -0.069 |
vious | -0.067 |
eper | -0.066 |
Token | thoughtful |
Token ID | 22677 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
masterpiece | +0.033 |
pron | +0.032 |
favourite | +0.031 |
icol | +0.030 |
brand | +0.029 |
Neg logit contributions | |
Bad | -0.027 |
crow | -0.027 |
Leave | -0.026 |
vious | -0.025 |
eper | -0.024 |
Token | . |
Token ID | 13 |
Feature activation | +1.18e-01 |
Pos logit contributions | |
pron | +0.083 |
favourite | +0.078 |
masterpiece | +0.078 |
brand | +0.071 |
favorite | +0.069 |
Neg logit contributions | |
crow | -0.084 |
Bad | -0.082 |
Leave | -0.080 |
vious | -0.076 |
eper | -0.075 |
Token |
|
Token ID | 198 |
Feature activation | +7.69e-01 |
Pos logit contributions | |
pron | +0.012 |
masterpiece | +0.011 |
favourite | +0.011 |
brand | +0.010 |
collection | +0.010 |
Neg logit contributions | |
crow | -0.007 |
Leave | -0.006 |
Bad | -0.006 |
Stop | -0.006 |
Hey | -0.005 |
Token |
|
Token ID | 198 |
Feature activation | +7.30e-01 |
Pos logit contributions | |
pron | +0.071 |
masterpiece | +0.069 |
favourite | +0.067 |
brand | +0.063 |
collection | +0.062 |
Neg logit contributions | |
crow | -0.048 |
Bad | -0.044 |
Leave | -0.044 |
Stop | -0.041 |
Rex | -0.040 |
Token | Su |
Token ID | 5606 |
Feature activation | +6.34e+00 |
Pos logit contributions | |
pron | +0.066 |
masterpiece | +0.063 |
favourite | +0.063 |
brand | +0.058 |
favorite | +0.057 |
Neg logit contributions | |
crow | -0.047 |
Bad | -0.046 |
Leave | -0.045 |
Stop | -0.042 |
vious | -0.041 |
Token | zy |
Token ID | 7357 |
Feature activation | +8.97e+00 |
Pos logit contributions | |
pron | +0.401 |
favourite | +0.389 |
masterpiece | +0.361 |
brand | +0.354 |
favorite | +0.333 |
Neg logit contributions | |
crow | -0.592 |
vious | -0.552 |
Leave | -0.550 |
Bad | -0.545 |
Stop | -0.527 |
Token | knew |
Token ID | 2993 |
Feature activation | +6.47e-01 |
Pos logit contributions | |
pron | +0.744 |
favourite | +0.691 |
masterpiece | +0.691 |
brand | +0.657 |
favorite | +0.610 |
Neg logit contributions | |
crow | -0.645 |
Bad | -0.617 |
Leave | -0.611 |
vious | -0.593 |
eper | -0.580 |
Token | that |
Token ID | 326 |
Feature activation | +2.98e+00 |
Pos logit contributions | |
pron | +0.056 |
masterpiece | +0.054 |
favourite | +0.050 |
icol | +0.050 |
ouse | +0.047 |
Neg logit contributions | |
crow | -0.043 |
Bad | -0.041 |
eper | -0.040 |
Leave | -0.039 |
hunters | -0.038 |
Token | she |
Token ID | 673 |
Feature activation | +5.18e+00 |
Pos logit contributions | |
pron | +0.215 |
masterpiece | +0.208 |
favourite | +0.201 |
brand | +0.190 |
ail | +0.189 |
Neg logit contributions | |
crow | -0.236 |
Bad | -0.229 |
Leave | -0.219 |
Stop | -0.211 |
eper | -0.206 |
Token | could |
Token ID | 714 |
Feature activation | -6.81e-01 |
Pos logit contributions | |
pron | +0.372 |
masterpiece | +0.345 |
favourite | +0.336 |
brand | +0.316 |
porch | +0.298 |
Neg logit contributions | |
crow | -0.422 |
Bad | -0.411 |
vious | -0.401 |
Leave | -0.395 |
eper | -0.387 |
Token | trust |
Token ID | 3774 |
Feature activation | -2.31e+00 |
Pos logit contributions | |
crow | +0.046 |
Leave | +0.045 |
Bad | +0.045 |
ases | +0.043 |
eper | +0.042 |
Neg logit contributions | |
pron | -0.056 |
masterpiece | -0.052 |
favourite | -0.051 |
brand | -0.050 |
yt | -0.048 |
Token | who |
Token ID | 508 |
Feature activation | -1.92e+00 |
Pos logit contributions | |
pron | +0.248 |
favourite | +0.237 |
masterpiece | +0.236 |
brand | +0.217 |
favorite | +0.213 |
Neg logit contributions | |
crow | -0.195 |
Bad | -0.189 |
Leave | -0.185 |
eper | -0.172 |
vious | -0.172 |
Token | wanted |
Token ID | 2227 |
Feature activation | -9.32e-01 |
Pos logit contributions | |
crow | +0.180 |
Bad | +0.177 |
Leave | +0.173 |
vious | +0.164 |
Stop | +0.164 |
Neg logit contributions | |
pron | -0.117 |
masterpiece | -0.110 |
favourite | -0.107 |
brand | -0.095 |
favorite | -0.090 |
Token | to |
Token ID | 284 |
Feature activation | +1.88e+00 |
Pos logit contributions | |
crow | +0.074 |
Bad | +0.072 |
Leave | +0.069 |
hunters | +0.068 |
Stop | +0.067 |
Neg logit contributions | |
pron | -0.069 |
masterpiece | -0.063 |
favourite | -0.059 |
ail | -0.056 |
yt | -0.055 |
Token | have |
Token ID | 423 |
Feature activation | +1.55e+00 |
Pos logit contributions | |
pron | +0.149 |
favourite | +0.139 |
masterpiece | +0.138 |
brand | +0.129 |
favorite | +0.123 |
Neg logit contributions | |
crow | -0.136 |
Leave | -0.130 |
Bad | -0.128 |
ases | -0.123 |
Stop | -0.120 |
Token | a |
Token ID | 257 |
Feature activation | +2.73e-01 |
Pos logit contributions | |
pron | +0.128 |
masterpiece | +0.119 |
favourite | +0.114 |
ail | +0.112 |
yt | +0.105 |
Neg logit contributions | |
Bad | -0.106 |
crow | -0.104 |
Leave | -0.104 |
vious | -0.099 |
hunters | -0.098 |
Token | picnic |
Token ID | 35715 |
Feature activation | +8.39e+00 |
Pos logit contributions | |
pron | +0.015 |
masterpiece | +0.014 |
ilda | +0.013 |
ail | +0.013 |
onna | +0.013 |
Neg logit contributions | |
bully | -0.022 |
Bad | -0.022 |
crow | -0.022 |
Leave | -0.020 |
eper | -0.020 |
Token | . |
Token ID | 13 |
Feature activation | +3.00e-01 |
Pos logit contributions | |
pron | +0.720 |
masterpiece | +0.715 |
favourite | +0.674 |
brand | +0.658 |
icol | +0.629 |
Neg logit contributions | |
crow | -0.530 |
Bad | -0.475 |
eper | -0.474 |
hunters | -0.469 |
ases | -0.460 |
Token | They |
Token ID | 1119 |
Feature activation | +2.14e-01 |
Pos logit contributions | |
pron | +0.026 |
masterpiece | +0.025 |
favourite | +0.024 |
brand | +0.023 |
ail | +0.021 |
Neg logit contributions | |
crow | -0.021 |
Bad | -0.020 |
Leave | -0.019 |
Stop | -0.018 |
vious | -0.018 |
Token | packed |
Token ID | 11856 |
Feature activation | +3.66e+00 |
Pos logit contributions | |
masterpiece | +0.017 |
pron | +0.017 |
brand | +0.015 |
icol | +0.015 |
favourite | +0.015 |
Neg logit contributions | |
crow | -0.015 |
eper | -0.015 |
Bad | -0.015 |
Leave | -0.014 |
Hey | -0.014 |
Token | a |
Token ID | 257 |
Feature activation | +2.89e-01 |
Pos logit contributions | |
pron | +0.349 |
masterpiece | +0.332 |
ail | +0.307 |
onna | +0.305 |
ilda | +0.305 |
Neg logit contributions | |
Bad | -0.173 |
crow | -0.166 |
Hey | -0.158 |
ases | -0.156 |
hunters | -0.155 |
Token | basket |
Token ID | 7988 |
Feature activation | +6.54e+00 |
Pos logit contributions | |
masterpiece | +0.017 |
ilda | +0.017 |
pron | +0.017 |
couldn | +0.016 |
ail | +0.015 |
Neg logit contributions | |
Bad | -0.022 |
crow | -0.022 |
hunters | -0.020 |
Rex | -0.020 |
eper | -0.020 |
Token | They |
Token ID | 1119 |
Feature activation | -7.95e-02 |
Pos logit contributions | |
crow | +0.017 |
Bad | +0.017 |
Leave | +0.016 |
Stop | +0.015 |
ases | +0.015 |
Neg logit contributions | |
pron | -0.021 |
masterpiece | -0.019 |
favourite | -0.019 |
brand | -0.018 |
collection | -0.017 |
Token | had |
Token ID | 550 |
Feature activation | +4.52e+00 |
Pos logit contributions | |
crow | +0.006 |
Bad | +0.006 |
Leave | +0.006 |
vious | +0.006 |
eper | +0.006 |
Neg logit contributions | |
pron | -0.006 |
favourite | -0.005 |
masterpiece | -0.005 |
brand | -0.005 |
favorite | -0.005 |
Token | ruined |
Token ID | 20484 |
Feature activation | +9.30e-01 |
Pos logit contributions | |
pron | +0.374 |
masterpiece | +0.336 |
favourite | +0.333 |
brand | +0.303 |
ail | +0.296 |
Neg logit contributions | |
Bad | -0.344 |
Leave | -0.331 |
crow | -0.327 |
vious | -0.314 |
ases | -0.312 |
Token | the |
Token ID | 262 |
Feature activation | -1.63e+00 |
Pos logit contributions | |
pron | +0.081 |
favourite | +0.075 |
masterpiece | +0.073 |
brand | +0.067 |
favorite | +0.065 |
Neg logit contributions | |
Bad | -0.065 |
crow | -0.064 |
Leave | -0.060 |
Stop | -0.058 |
ases | -0.058 |
Token | cool |
Token ID | 3608 |
Feature activation | +3.77e+00 |
Pos logit contributions | |
Bad | +0.155 |
crow | +0.152 |
Leave | +0.151 |
Stop | +0.150 |
vious | +0.149 |
Neg logit contributions | |
pron | -0.081 |
masterpiece | -0.073 |
onna | -0.072 |
favourite | -0.071 |
brand | -0.065 |
Token | skirt |
Token ID | 23967 |
Feature activation | +8.76e+00 |
Pos logit contributions | |
pron | +0.195 |
favourite | +0.182 |
masterpiece | +0.172 |
brand | +0.161 |
usual | +0.158 |
Neg logit contributions | |
crow | -0.370 |
vious | -0.344 |
Stop | -0.340 |
Leave | -0.340 |
Bad | -0.334 |
Token | . |
Token ID | 13 |
Feature activation | -2.46e-01 |
Pos logit contributions | |
pron | +0.762 |
favourite | +0.722 |
masterpiece | +0.704 |
brand | +0.652 |
favorite | +0.635 |
Neg logit contributions | |
crow | -0.586 |
Bad | -0.558 |
Leave | -0.545 |
vious | -0.528 |
eper | -0.519 |
Token |
|
Token ID | 198 |
Feature activation | +1.74e-01 |
Pos logit contributions | |
crow | +0.018 |
Bad | +0.018 |
Leave | +0.017 |
Stop | +0.016 |
vious | +0.016 |
Neg logit contributions | |
pron | -0.020 |
masterpiece | -0.019 |
favourite | -0.019 |
brand | -0.017 |
favorite | -0.017 |
Token |
|
Token ID | 198 |
Feature activation | +1.59e-01 |
Pos logit contributions | |
pron | +0.015 |
favourite | +0.014 |
masterpiece | +0.014 |
brand | +0.013 |
favorite | +0.012 |
Neg logit contributions | |
crow | -0.012 |
Bad | -0.012 |
Leave | -0.012 |
Stop | -0.011 |
ases | -0.011 |
Token | Their |
Token ID | 14574 |
Feature activation | -1.72e+00 |
Pos logit contributions | |
pron | +0.013 |
favourite | +0.012 |
masterpiece | +0.012 |
brand | +0.011 |
favorite | +0.011 |
Neg logit contributions | |
crow | -0.011 |
vious | -0.011 |
Leave | -0.010 |
Bad | -0.010 |
ases | -0.010 |
Token | mom |
Token ID | 1995 |
Feature activation | +4.92e+00 |
Pos logit contributions | |
Bad | +0.146 |
Leave | +0.143 |
bully | +0.142 |
crow | +0.139 |
ases | +0.139 |
Neg logit contributions | |
pron | -0.115 |
masterpiece | -0.089 |
favourite | -0.089 |
onna | -0.086 |
brand | -0.086 |
Token | . |
Token ID | 13 |
Feature activation | -1.44e-01 |
Pos logit contributions | |
Leave | +0.704 |
Bad | +0.698 |
crow | +0.694 |
Stop | +0.660 |
vious | +0.659 |
Neg logit contributions | |
pron | -0.500 |
favourite | -0.493 |
masterpiece | -0.490 |
usual | -0.445 |
favorite | -0.444 |
Token | The |
Token ID | 383 |
Feature activation | -2.06e-01 |
Pos logit contributions | |
Leave | +0.012 |
Bad | +0.012 |
crow | +0.012 |
eper | +0.012 |
Stop | +0.011 |
Neg logit contributions | |
favourite | -0.009 |
pron | -0.009 |
masterpiece | -0.008 |
favorite | -0.008 |
brand | -0.008 |
Token | animals |
Token ID | 4695 |
Feature activation | -1.01e+01 |
Pos logit contributions | |
crow | +0.018 |
ases | +0.017 |
Leave | +0.017 |
vious | +0.016 |
Bad | +0.016 |
Neg logit contributions | |
pron | -0.015 |
favourite | -0.013 |
masterpiece | -0.013 |
favorite | -0.013 |
porch | -0.012 |
Token | and |
Token ID | 290 |
Feature activation | +2.51e-01 |
Pos logit contributions | |
crow | +0.936 |
Bad | +0.927 |
Leave | +0.921 |
ases | +0.891 |
vious | +0.879 |
Neg logit contributions | |
pron | -0.554 |
favourite | -0.521 |
masterpiece | -0.502 |
couldn | -0.457 |
favorite | -0.453 |
Token | the |
Token ID | 262 |
Feature activation | -1.68e+00 |
Pos logit contributions | |
favourite | +0.017 |
pron | +0.017 |
favorite | +0.016 |
usual | +0.016 |
masterpiece | +0.016 |
Neg logit contributions | |
Leave | -0.019 |
crow | -0.019 |
Bad | -0.018 |
vious | -0.018 |
ases | -0.018 |
Token | long |
Token ID | 890 |
Feature activation | +8.40e+00 |
Pos logit contributions | |
crow | +0.150 |
ases | +0.139 |
Leave | +0.138 |
vious | +0.137 |
Bad | +0.135 |
Neg logit contributions | |
pron | -0.118 |
favourite | -0.109 |
masterpiece | -0.104 |
favorite | -0.102 |
brand | -0.097 |
Token | snake |
Token ID | 17522 |
Feature activation | -5.76e+00 |
Pos logit contributions | |
pron | +0.694 |
favourite | +0.634 |
masterpiece | +0.615 |
favorite | +0.580 |
collection | +0.563 |
Neg logit contributions | |
crow | -0.613 |
Bad | -0.602 |
Leave | -0.601 |
ases | -0.575 |
vious | -0.574 |
Token | protected |
Token ID | 6861 |
Feature activation | -4.29e+00 |
Pos logit contributions | |
crow | +0.505 |
Bad | +0.494 |
Leave | +0.487 |
vious | +0.461 |
eper | +0.459 |
Neg logit contributions | |
pron | -0.381 |
favourite | -0.359 |
masterpiece | -0.354 |
brand | -0.318 |
favorite | -0.313 |
Token | the |
Token ID | 262 |
Feature activation | -1.69e+00 |
Pos logit contributions | |
Leave | +0.354 |
vious | +0.341 |
Bad | +0.340 |
crow | +0.339 |
eper | +0.316 |
Neg logit contributions | |
favourite | -0.288 |
masterpiece | -0.276 |
usual | -0.273 |
favorite | -0.266 |
pron | -0.264 |
Token | gum |
Token ID | 27142 |
Feature activation | +1.18e+00 |
Pos logit contributions | |
crow | +0.145 |
Leave | +0.135 |
ases | +0.135 |
Bad | +0.132 |
eper | +0.131 |
Neg logit contributions | |
pron | -0.125 |
favourite | -0.114 |
masterpiece | -0.111 |
favorite | -0.105 |
brand | -0.103 |
Token | tree |
Token ID | 5509 |
Feature activation | -8.70e-01 |
Pos logit contributions | |
pron | +0.082 |
favourite | +0.079 |
masterpiece | +0.078 |
favorite | +0.070 |
brand | +0.070 |
Neg logit contributions | |
crow | -0.098 |
Bad | -0.097 |
Leave | -0.096 |
Stop | -0.090 |
vious | -0.089 |
Token | Whenever |
Token ID | 21326 |
Feature activation | -3.30e+00 |
Pos logit contributions | |
crow | +0.018 |
Bad | +0.017 |
Leave | +0.017 |
vious | +0.016 |
Stop | +0.016 |
Neg logit contributions | |
pron | -0.021 |
masterpiece | -0.019 |
favourite | -0.019 |
brand | -0.018 |
favorite | -0.017 |
Token | they |
Token ID | 484 |
Feature activation | +5.02e+00 |
Pos logit contributions | |
crow | +0.344 |
Bad | +0.340 |
Leave | +0.338 |
vious | +0.328 |
ases | +0.325 |
Neg logit contributions | |
favourite | -0.151 |
pron | -0.149 |
masterpiece | -0.142 |
favorite | -0.125 |
usual | -0.122 |
Token | did |
Token ID | 750 |
Feature activation | -2.56e+00 |
Pos logit contributions | |
pron | +0.309 |
favourite | +0.295 |
masterpiece | +0.287 |
favorite | +0.257 |
brand | +0.256 |
Neg logit contributions | |
crow | -0.458 |
Bad | -0.452 |
Leave | -0.444 |
Stop | -0.417 |
eper | -0.416 |
Token | , |
Token ID | 11 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
crow | +0.213 |
Bad | +0.208 |
Leave | +0.205 |
vious | +0.193 |
eper | +0.192 |
Neg logit contributions | |
pron | -0.181 |
favourite | -0.171 |
masterpiece | -0.169 |
brand | -0.153 |
favorite | -0.150 |
Token | they |
Token ID | 484 |
Feature activation | +4.98e+00 |
Pos logit contributions | |
crow | +0.136 |
Bad | +0.134 |
Leave | +0.132 |
vious | +0.126 |
eper | +0.126 |
Neg logit contributions | |
pron | -0.072 |
favourite | -0.067 |
masterpiece | -0.066 |
brand | -0.057 |
favorite | -0.057 |
Token | were |
Token ID | 547 |
Feature activation | +1.10e-01 |
Pos logit contributions | |
pron | +0.270 |
masterpiece | +0.243 |
favourite | +0.243 |
brand | +0.213 |
favorite | +0.201 |
Neg logit contributions | |
crow | -0.504 |
Bad | -0.489 |
Leave | -0.483 |
eper | -0.470 |
vious | -0.468 |
Token | always |
Token ID | 1464 |
Feature activation | +3.10e+00 |
Pos logit contributions | |
pron | +0.008 |
favourite | +0.008 |
masterpiece | +0.007 |
brand | +0.007 |
favorite | +0.007 |
Neg logit contributions | |
crow | -0.009 |
Bad | -0.009 |
Leave | -0.009 |
Stop | -0.008 |
vious | -0.008 |
Token | happy |
Token ID | 3772 |
Feature activation | -1.33e+00 |
Pos logit contributions | |
pron | +0.212 |
favourite | +0.204 |
masterpiece | +0.199 |
brand | +0.182 |
favorite | +0.180 |
Neg logit contributions | |
crow | -0.271 |
Bad | -0.258 |
Leave | -0.256 |
Stop | -0.245 |
vious | -0.243 |
Token | because |
Token ID | 780 |
Feature activation | +1.96e+00 |
Pos logit contributions | |
crow | +0.114 |
Bad | +0.112 |
Leave | +0.110 |
vious | +0.104 |
eper | +0.103 |
Neg logit contributions | |
pron | -0.091 |
favourite | -0.086 |
masterpiece | -0.085 |
brand | -0.077 |
favorite | -0.076 |
Token | they |
Token ID | 484 |
Feature activation | +4.97e+00 |
Pos logit contributions | |
pron | +0.106 |
masterpiece | +0.092 |
favourite | +0.091 |
brand | +0.080 |
favorite | +0.074 |
Neg logit contributions | |
crow | -0.202 |
Bad | -0.198 |
Leave | -0.194 |
eper | -0.186 |
vious | -0.184 |
Token | were |
Token ID | 547 |
Feature activation | +9.33e-02 |
Pos logit contributions | |
pron | +0.319 |
favourite | +0.295 |
masterpiece | +0.293 |
brand | +0.261 |
favorite | +0.254 |
Neg logit contributions | |
crow | -0.450 |
Bad | -0.443 |
Leave | -0.434 |
vious | -0.417 |
eper | -0.415 |
Token | bird |
Token ID | 6512 |
Feature activation | -4.76e+00 |
Pos logit contributions | |
crow | +0.104 |
Bad | +0.100 |
Leave | +0.099 |
eper | +0.093 |
ases | +0.092 |
Neg logit contributions | |
pron | -0.117 |
favourite | -0.112 |
masterpiece | -0.111 |
favorite | -0.102 |
brand | -0.101 |
Token | ie |
Token ID | 494 |
Feature activation | +4.22e+00 |
Pos logit contributions | |
Bad | +0.459 |
Leave | +0.456 |
crow | +0.450 |
Stop | +0.439 |
Hey | +0.424 |
Neg logit contributions | |
pron | -0.270 |
favourite | -0.257 |
masterpiece | -0.249 |
favorite | -0.226 |
brand | -0.222 |
Token | ! |
Token ID | 0 |
Feature activation | +2.87e+00 |
Pos logit contributions | |
pron | +0.303 |
favourite | +0.290 |
masterpiece | +0.284 |
favorite | +0.259 |
brand | +0.259 |
Neg logit contributions | |
Bad | -0.340 |
crow | -0.340 |
Leave | -0.335 |
Stop | -0.317 |
eper | -0.312 |
Token | The |
Token ID | 383 |
Feature activation | -3.06e-03 |
Pos logit contributions | |
pron | +0.237 |
favourite | +0.226 |
masterpiece | +0.223 |
brand | +0.205 |
favorite | +0.203 |
Neg logit contributions | |
crow | -0.207 |
Bad | -0.202 |
Leave | -0.198 |
vious | -0.185 |
eper | -0.185 |
Token | bird |
Token ID | 6512 |
Feature activation | -4.79e+00 |
Pos logit contributions | |
crow | +0.000 |
Bad | +0.000 |
Leave | +0.000 |
vious | +0.000 |
eper | +0.000 |
Neg logit contributions | |
pron | -0.000 |
favourite | -0.000 |
masterpiece | -0.000 |
brand | -0.000 |
favorite | -0.000 |
Token | ie |
Token ID | 494 |
Feature activation | +4.21e+00 |
Pos logit contributions | |
crow | +0.451 |
Bad | +0.448 |
Leave | +0.443 |
Stop | +0.420 |
ases | +0.419 |
Neg logit contributions | |
pron | -0.283 |
favourite | -0.258 |
masterpiece | -0.258 |
brand | -0.225 |
favorite | -0.224 |
Token | felt |
Token ID | 2936 |
Feature activation | +4.72e+00 |
Pos logit contributions | |
pron | +0.252 |
favourite | +0.237 |
masterpiece | +0.231 |
favorite | +0.206 |
brand | +0.201 |
Neg logit contributions | |
crow | -0.387 |
Bad | -0.383 |
Leave | -0.380 |
ases | -0.363 |
vious | -0.359 |
Token | very |
Token ID | 845 |
Feature activation | -9.94e-01 |
Pos logit contributions | |
favourite | +0.298 |
brand | +0.287 |
favorite | +0.268 |
pron | +0.252 |
masterpiece | +0.237 |
Neg logit contributions | |
ases | -0.396 |
crow | -0.394 |
eper | -0.387 |
Bad | -0.377 |
Stop | -0.372 |
Token | scared |
Token ID | 12008 |
Feature activation | -8.70e+00 |
Pos logit contributions | |
crow | +0.072 |
Bad | +0.071 |
Leave | +0.069 |
eper | +0.066 |
Stop | +0.065 |
Neg logit contributions | |
favourite | -0.078 |
pron | -0.078 |
masterpiece | -0.074 |
favorite | -0.071 |
brand | -0.070 |
Token | and |
Token ID | 290 |
Feature activation | +3.45e-01 |
Pos logit contributions | |
Bad | +0.791 |
crow | +0.787 |
Leave | +0.759 |
eper | +0.743 |
ases | +0.729 |
Neg logit contributions | |
favourite | -0.504 |
pron | -0.494 |
masterpiece | -0.464 |
favorite | -0.456 |
brand | -0.453 |
Token | didn |
Token ID | 1422 |
Feature activation | -7.32e-01 |
Pos logit contributions | |
pron | +0.021 |
favourite | +0.020 |
masterpiece | +0.019 |
favorite | +0.018 |
brand | +0.018 |
Neg logit contributions | |
crow | -0.030 |
Bad | -0.030 |
Leave | -0.029 |
ases | -0.029 |
Stop | -0.028 |
Token | jump |
Token ID | 4391 |
Feature activation | -9.00e-01 |
Pos logit contributions | |
pron | +0.145 |
favourite | +0.137 |
masterpiece | +0.135 |
brand | +0.123 |
favorite | +0.123 |
Neg logit contributions | |
crow | -0.134 |
Leave | -0.131 |
Bad | -0.130 |
vious | -0.120 |
eper | -0.120 |
Token | into |
Token ID | 656 |
Feature activation | +7.72e-01 |
Pos logit contributions | |
crow | +0.065 |
Bad | +0.063 |
Leave | +0.062 |
vious | +0.058 |
Stop | +0.058 |
Neg logit contributions | |
pron | -0.074 |
favourite | -0.069 |
masterpiece | -0.069 |
brand | -0.063 |
favorite | -0.062 |
Token | the |
Token ID | 262 |
Feature activation | -1.63e+00 |
Pos logit contributions | |
pron | +0.065 |
favourite | +0.061 |
masterpiece | +0.061 |
brand | +0.055 |
favorite | +0.055 |
Neg logit contributions | |
crow | -0.055 |
Bad | -0.054 |
Leave | -0.052 |
vious | -0.049 |
eper | -0.049 |
Token | nest |
Token ID | 16343 |
Feature activation | +4.43e+00 |
Pos logit contributions | |
crow | +0.133 |
Bad | +0.129 |
Leave | +0.126 |
vious | +0.120 |
Stop | +0.118 |
Neg logit contributions | |
pron | -0.123 |
favourite | -0.114 |
masterpiece | -0.112 |
brand | -0.102 |
favorite | -0.100 |
Token | . |
Token ID | 13 |
Feature activation | +7.68e-02 |
Pos logit contributions | |
pron | +0.366 |
favourite | +0.345 |
masterpiece | +0.342 |
brand | +0.313 |
favorite | +0.307 |
Neg logit contributions | |
crow | -0.316 |
Bad | -0.297 |
Leave | -0.291 |
vious | -0.283 |
eper | -0.280 |
Token |
|
Token ID | 198 |
Feature activation | +7.39e-01 |
Pos logit contributions | |
pron | +0.006 |
masterpiece | +0.006 |
favourite | +0.006 |
brand | +0.005 |
favorite | +0.005 |
Neg logit contributions | |
crow | -0.006 |
Bad | -0.005 |
Leave | -0.005 |
vious | -0.005 |
Stop | -0.005 |
Token |
|
Token ID | 198 |
Feature activation | +7.12e-01 |
Pos logit contributions | |
pron | +0.062 |
masterpiece | +0.059 |
favourite | +0.059 |
brand | +0.054 |
favorite | +0.053 |
Neg logit contributions | |
crow | -0.052 |
Bad | -0.050 |
Leave | -0.049 |
Stop | -0.046 |
eper | -0.045 |
Token | The |
Token ID | 464 |
Feature activation | -1.16e+00 |
Pos logit contributions | |
pron | +0.053 |
favourite | +0.046 |
masterpiece | +0.044 |
brand | +0.041 |
usual | +0.039 |
Neg logit contributions | |
crow | -0.060 |
ases | -0.057 |
vious | -0.056 |
eper | -0.055 |
Bad | -0.054 |
Token | bunny |
Token ID | 44915 |
Feature activation | -1.40e+00 |
Pos logit contributions | |
crow | +0.089 |
Bad | +0.087 |
Leave | +0.085 |
vious | +0.080 |
eper | +0.080 |
Neg logit contributions | |
pron | -0.092 |
favourite | -0.087 |
masterpiece | -0.086 |
brand | -0.079 |
favorite | -0.078 |
Token | was |
Token ID | 373 |
Feature activation | +1.64e+00 |
Pos logit contributions | |
crow | +0.116 |
Bad | +0.112 |
Leave | +0.110 |
vious | +0.105 |
eper | +0.105 |
Neg logit contributions | |
pron | -0.100 |
favourite | -0.092 |
masterpiece | -0.092 |
brand | -0.085 |
usual | -0.082 |
Token | so |
Token ID | 523 |
Feature activation | +2.85e-01 |
Pos logit contributions | |
pron | +0.126 |
masterpiece | +0.118 |
favourite | +0.117 |
brand | +0.105 |
favorite | +0.104 |
Neg logit contributions | |
crow | -0.126 |
Bad | -0.124 |
Leave | -0.122 |
vious | -0.115 |
eper | -0.115 |
Token | friends |
Token ID | 2460 |
Feature activation | -3.85e+00 |
Pos logit contributions | |
crow | +0.108 |
Bad | +0.103 |
Leave | +0.102 |
vious | +0.097 |
eper | +0.096 |
Neg logit contributions | |
pron | -0.116 |
masterpiece | -0.106 |
favourite | -0.106 |
brand | -0.097 |
favorite | -0.094 |
Token | there |
Token ID | 612 |
Feature activation | -2.15e+00 |
Pos logit contributions | |
crow | +0.276 |
Bad | +0.269 |
Leave | +0.264 |
vious | +0.248 |
eper | +0.247 |
Neg logit contributions | |
pron | -0.313 |
masterpiece | -0.294 |
favourite | -0.290 |
brand | -0.263 |
favorite | -0.260 |
Token | . |
Token ID | 13 |
Feature activation | -1.05e-01 |
Pos logit contributions | |
crow | +0.169 |
Bad | +0.165 |
Leave | +0.160 |
eper | +0.152 |
vious | +0.152 |
Neg logit contributions | |
pron | -0.161 |
masterpiece | -0.150 |
favourite | -0.148 |
brand | -0.135 |
favorite | -0.131 |
Token | From |
Token ID | 3574 |
Feature activation | +2.84e+00 |
Pos logit contributions | |
crow | +0.008 |
Bad | +0.008 |
Leave | +0.008 |
vious | +0.008 |
eper | +0.007 |
Neg logit contributions | |
pron | -0.008 |
masterpiece | -0.007 |
favourite | -0.007 |
brand | -0.007 |
favorite | -0.007 |
Token | then |
Token ID | 788 |
Feature activation | +3.08e+00 |
Pos logit contributions | |
pron | +0.208 |
masterpiece | +0.194 |
favourite | +0.192 |
brand | +0.171 |
favorite | +0.168 |
Neg logit contributions | |
crow | -0.235 |
Bad | -0.232 |
Leave | -0.223 |
eper | -0.214 |
Stop | -0.210 |
Token | on |
Token ID | 319 |
Feature activation | +1.30e+00 |
Pos logit contributions | |
pron | +0.220 |
favourite | +0.210 |
masterpiece | +0.200 |
favorite | +0.184 |
brand | +0.182 |
Neg logit contributions | |
crow | -0.247 |
Bad | -0.246 |
Leave | -0.237 |
vious | -0.231 |
eper | -0.225 |
Token | , |
Token ID | 11 |
Feature activation | -1.08e+00 |
Pos logit contributions | |
pron | +0.101 |
masterpiece | +0.095 |
favourite | +0.094 |
brand | +0.086 |
favorite | +0.084 |
Neg logit contributions | |
crow | -0.101 |
Bad | -0.098 |
Leave | -0.096 |
eper | -0.090 |
Stop | -0.089 |
Token | Benny |
Token ID | 44275 |
Feature activation | +2.97e+00 |
Pos logit contributions | |
crow | +0.099 |
Bad | +0.097 |
Leave | +0.095 |
vious | +0.091 |
eper | +0.091 |
Neg logit contributions | |
pron | -0.069 |
favourite | -0.065 |
masterpiece | -0.064 |
brand | -0.057 |
favorite | -0.056 |
Token | didn |
Token ID | 1422 |
Feature activation | -6.48e-01 |
Pos logit contributions | |
pron | +0.186 |
favourite | +0.175 |
masterpiece | +0.172 |
favorite | +0.152 |
brand | +0.152 |
Neg logit contributions | |
crow | -0.270 |
Bad | -0.265 |
Leave | -0.261 |
vious | -0.247 |
ases | -0.246 |
Token | 't |
Token ID | 470 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
crow | +0.046 |
Bad | +0.045 |
Leave | +0.044 |
vious | +0.041 |
Stop | +0.041 |
Neg logit contributions | |
pron | -0.054 |
favourite | -0.051 |
masterpiece | -0.051 |
brand | -0.047 |
favorite | -0.046 |
Token | mind |
Token ID | 2000 |
Feature activation | -1.87e+00 |
Pos logit contributions | |
Bad | +0.093 |
crow | +0.092 |
Leave | +0.089 |
eper | +0.086 |
ases | +0.085 |
Neg logit contributions | |
pron | -0.102 |
favourite | -0.099 |
masterpiece | -0.097 |
favorite | -0.090 |
usual | -0.088 |
Token | watch |
Token ID | 2342 |
Feature activation | +9.93e-01 |
Pos logit contributions | |
crow | +0.300 |
Bad | +0.293 |
Leave | +0.289 |
eper | +0.272 |
Stop | +0.269 |
Neg logit contributions | |
pron | -0.307 |
favourite | -0.291 |
masterpiece | -0.287 |
brand | -0.262 |
favorite | -0.258 |
Token | the |
Token ID | 262 |
Feature activation | -1.59e+00 |
Pos logit contributions | |
pron | +0.078 |
favourite | +0.074 |
masterpiece | +0.073 |
brand | +0.067 |
favorite | +0.066 |
Neg logit contributions | |
crow | -0.075 |
Bad | -0.073 |
Leave | -0.072 |
vious | -0.068 |
eper | -0.067 |
Token | film |
Token ID | 2646 |
Feature activation | -1.56e+00 |
Pos logit contributions | |
crow | +0.142 |
Leave | +0.135 |
Bad | +0.131 |
vious | +0.128 |
eper | +0.127 |
Neg logit contributions | |
pron | -0.110 |
favourite | -0.106 |
masterpiece | -0.103 |
favorite | -0.095 |
brand | -0.091 |
Token | tomorrow |
Token ID | 9439 |
Feature activation | +3.34e-01 |
Pos logit contributions | |
crow | +0.124 |
Bad | +0.121 |
Leave | +0.119 |
vious | +0.112 |
Stop | +0.111 |
Neg logit contributions | |
pron | -0.117 |
favourite | -0.111 |
masterpiece | -0.110 |
brand | -0.100 |
favorite | -0.099 |
Token | . |
Token ID | 13 |
Feature activation | -2.17e-02 |
Pos logit contributions | |
pron | +0.026 |
favourite | +0.025 |
masterpiece | +0.025 |
brand | +0.023 |
favorite | +0.022 |
Neg logit contributions | |
crow | -0.025 |
Bad | -0.024 |
Leave | -0.024 |
vious | -0.022 |
eper | -0.022 |
Token | Sara |
Token ID | 24799 |
Feature activation | +2.09e+00 |
Pos logit contributions | |
crow | +0.001 |
Bad | +0.001 |
Leave | +0.001 |
vious | +0.001 |
eper | +0.001 |
Neg logit contributions | |
pron | -0.002 |
favourite | -0.002 |
masterpiece | -0.002 |
brand | -0.002 |
favorite | -0.002 |
Token | cal |
Token ID | 2386 |
Feature activation | -7.26e+00 |
Pos logit contributions | |
pron | +0.141 |
favourite | +0.137 |
masterpiece | +0.134 |
favorite | +0.128 |
brand | +0.118 |
Neg logit contributions | |
crow | -0.177 |
Bad | -0.168 |
Leave | -0.167 |
Stop | -0.151 |
eper | -0.150 |
Token | ms |
Token ID | 907 |
Feature activation | -9.11e+00 |
Pos logit contributions | |
Bad | +0.350 |
Leave | +0.341 |
hunters | +0.324 |
crow | +0.322 |
Stop | +0.307 |
Neg logit contributions | |
pron | -0.652 |
favourite | -0.637 |
complexion | -0.620 |
masterpiece | -0.614 |
favorite | -0.612 |
Token | down |
Token ID | 866 |
Feature activation | -2.09e-01 |
Pos logit contributions | |
crow | +0.748 |
Bad | +0.743 |
Leave | +0.729 |
vious | +0.693 |
Stop | +0.670 |
Neg logit contributions | |
pron | -0.633 |
favourite | -0.599 |
masterpiece | -0.583 |
favorite | -0.545 |
usual | -0.536 |
Token | and |
Token ID | 290 |
Feature activation | +2.07e-01 |
Pos logit contributions | |
crow | +0.018 |
Bad | +0.018 |
Leave | +0.018 |
Stop | +0.017 |
vious | +0.017 |
Neg logit contributions | |
pron | -0.013 |
favourite | -0.012 |
masterpiece | -0.012 |
favorite | -0.011 |
brand | -0.011 |
Token | says |
Token ID | 1139 |
Feature activation | -1.89e+00 |
Pos logit contributions | |
pron | +0.016 |
favourite | +0.015 |
masterpiece | +0.015 |
brand | +0.014 |
favorite | +0.014 |
Neg logit contributions | |
crow | -0.016 |
Bad | -0.015 |
Leave | -0.015 |
vious | -0.014 |
eper | -0.014 |
Token | saw |
Token ID | 2497 |
Feature activation | -4.06e-01 |
Pos logit contributions | |
pron | +0.012 |
favourite | +0.011 |
masterpiece | +0.011 |
favorite | +0.010 |
brand | +0.010 |
Neg logit contributions | |
crow | -0.012 |
Bad | -0.012 |
Leave | -0.012 |
Stop | -0.011 |
vious | -0.011 |
Token | some |
Token ID | 617 |
Feature activation | +9.63e-01 |
Pos logit contributions | |
crow | +0.032 |
Bad | +0.031 |
Leave | +0.030 |
vious | +0.029 |
eper | +0.029 |
Neg logit contributions | |
pron | -0.032 |
masterpiece | -0.029 |
favourite | -0.029 |
brand | -0.026 |
favorite | -0.026 |
Token | children |
Token ID | 1751 |
Feature activation | -7.55e+00 |
Pos logit contributions | |
pron | +0.083 |
favourite | +0.080 |
masterpiece | +0.079 |
brand | +0.074 |
favorite | +0.072 |
Neg logit contributions | |
crow | -0.064 |
Bad | -0.063 |
Leave | -0.061 |
eper | -0.057 |
ases | -0.057 |
Token | playing |
Token ID | 2712 |
Feature activation | -6.89e-01 |
Pos logit contributions | |
crow | +0.638 |
Bad | +0.624 |
Leave | +0.614 |
vious | +0.580 |
eper | +0.577 |
Neg logit contributions | |
pron | -0.525 |
favourite | -0.495 |
masterpiece | -0.489 |
brand | -0.441 |
favorite | -0.434 |
Token | on |
Token ID | 319 |
Feature activation | +1.29e+00 |
Pos logit contributions | |
crow | +0.056 |
Bad | +0.055 |
Leave | +0.054 |
vious | +0.052 |
eper | +0.052 |
Neg logit contributions | |
pron | -0.048 |
favourite | -0.048 |
masterpiece | -0.045 |
favorite | -0.042 |
brand | -0.042 |
Token | the |
Token ID | 262 |
Feature activation | -1.42e+00 |
Pos logit contributions | |
pron | +0.108 |
favourite | +0.103 |
masterpiece | +0.102 |
brand | +0.094 |
favorite | +0.092 |
Neg logit contributions | |
crow | -0.092 |
Bad | -0.089 |
Leave | -0.088 |
vious | -0.082 |
Stop | -0.081 |
Token | swings |
Token ID | 26728 |
Feature activation | +2.44e+00 |
Pos logit contributions | |
crow | +0.112 |
Bad | +0.110 |
Leave | +0.108 |
Stop | +0.102 |
vious | +0.100 |
Neg logit contributions | |
pron | -0.107 |
masterpiece | -0.098 |
favourite | -0.097 |
brand | -0.090 |
collection | -0.087 |
Token | . |
Token ID | 13 |
Feature activation | -7.41e-02 |
Pos logit contributions | |
pron | +0.175 |
favourite | +0.167 |
masterpiece | +0.164 |
brand | +0.148 |
favorite | +0.148 |
Neg logit contributions | |
crow | -0.196 |
Bad | -0.194 |
Leave | -0.191 |
Stop | -0.179 |
vious | -0.179 |
Token | They |
Token ID | 1119 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
crow | +0.006 |
Bad | +0.005 |
Leave | +0.005 |
vious | +0.005 |
Stop | +0.005 |
Neg logit contributions | |
pron | -0.006 |
masterpiece | -0.006 |
favourite | -0.006 |
brand | -0.005 |
favorite | -0.005 |
Token | looked |
Token ID | 3114 |
Feature activation | +4.69e+00 |
Pos logit contributions | |
pron | +0.008 |
masterpiece | +0.008 |
favourite | +0.008 |
brand | +0.007 |
favorite | +0.007 |
Neg logit contributions | |
crow | -0.009 |
Bad | -0.009 |
Leave | -0.009 |
eper | -0.009 |
vious | -0.008 |
Token | so |
Token ID | 523 |
Feature activation | +2.68e-01 |
Pos logit contributions | |
pron | +0.367 |
favourite | +0.343 |
masterpiece | +0.342 |
brand | +0.310 |
favorite | +0.305 |
Neg logit contributions | |
crow | -0.358 |
Bad | -0.350 |
Leave | -0.344 |
vious | -0.324 |
eper | -0.320 |
Token | his |
Token ID | 465 |
Feature activation | -2.96e+00 |
Pos logit contributions | |
crow | +0.024 |
Bad | +0.023 |
Leave | +0.023 |
vious | +0.021 |
eper | +0.021 |
Neg logit contributions | |
pron | -0.029 |
favourite | -0.027 |
masterpiece | -0.027 |
brand | -0.025 |
favorite | -0.024 |
Token | head |
Token ID | 1182 |
Feature activation | +1.96e+00 |
Pos logit contributions | |
crow | +0.219 |
Bad | +0.219 |
Leave | +0.213 |
eper | +0.199 |
ases | +0.196 |
Neg logit contributions | |
pron | -0.228 |
masterpiece | -0.210 |
favourite | -0.205 |
brand | -0.192 |
usual | -0.184 |
Token | and |
Token ID | 290 |
Feature activation | +3.54e-01 |
Pos logit contributions | |
pron | +0.154 |
favourite | +0.146 |
masterpiece | +0.144 |
brand | +0.131 |
favorite | +0.129 |
Neg logit contributions | |
crow | -0.149 |
Bad | -0.144 |
Leave | -0.142 |
vious | -0.134 |
eper | -0.133 |
Token | it |
Token ID | 340 |
Feature activation | +5.13e-01 |
Pos logit contributions | |
pron | +0.028 |
favourite | +0.026 |
masterpiece | +0.026 |
favorite | +0.023 |
usual | +0.022 |
Neg logit contributions | |
Bad | -0.027 |
crow | -0.027 |
Leave | -0.026 |
Stop | -0.025 |
eper | -0.024 |
Token | flew |
Token ID | 13112 |
Feature activation | +4.55e-01 |
Pos logit contributions | |
pron | +0.041 |
favourite | +0.038 |
masterpiece | +0.037 |
brand | +0.034 |
favorite | +0.033 |
Neg logit contributions | |
crow | -0.039 |
Bad | -0.039 |
Leave | -0.038 |
vious | -0.036 |
Stop | -0.035 |
Token | away |
Token ID | 1497 |
Feature activation | -7.94e+00 |
Pos logit contributions | |
pron | +0.036 |
favourite | +0.036 |
masterpiece | +0.035 |
brand | +0.034 |
favorite | +0.033 |
Neg logit contributions | |
crow | -0.031 |
Bad | -0.030 |
Leave | -0.030 |
eper | -0.029 |
ases | -0.028 |
Token | . |
Token ID | 13 |
Feature activation | -2.97e-02 |
Pos logit contributions | |
crow | +0.536 |
Bad | +0.529 |
Leave | +0.512 |
vious | +0.484 |
ases | +0.474 |
Neg logit contributions | |
pron | -0.701 |
favourite | -0.645 |
masterpiece | -0.639 |
brand | -0.581 |
favorite | -0.570 |
Token | Tim |
Token ID | 5045 |
Feature activation | +5.77e-01 |
Pos logit contributions | |
Bad | +0.002 |
crow | +0.002 |
Leave | +0.002 |
Stop | +0.002 |
vious | +0.002 |
Neg logit contributions | |
pron | -0.003 |
masterpiece | -0.002 |
favourite | -0.002 |
usual | -0.002 |
collection | -0.002 |
Token | my |
Token ID | 1820 |
Feature activation | -2.99e+00 |
Pos logit contributions | |
pron | +0.049 |
masterpiece | +0.046 |
favourite | +0.046 |
brand | +0.042 |
favorite | +0.042 |
Neg logit contributions | |
crow | -0.041 |
Bad | -0.039 |
Leave | -0.039 |
ases | -0.036 |
Stop | -0.036 |
Token | started |
Token ID | 2067 |
Feature activation | -1.46e+00 |
Pos logit contributions | |
crow | +0.253 |
Bad | +0.250 |
Leave | +0.245 |
vious | +0.232 |
eper | +0.231 |
Neg logit contributions | |
pron | -0.209 |
favourite | -0.196 |
masterpiece | -0.194 |
brand | -0.174 |
favorite | -0.170 |
Token | to |
Token ID | 284 |
Feature activation | +1.68e+00 |
Pos logit contributions | |
crow | +0.137 |
Leave | +0.134 |
Bad | +0.134 |
vious | +0.126 |
Stop | +0.125 |
Neg logit contributions | |
pron | -0.091 |
favourite | -0.081 |
masterpiece | -0.081 |
brand | -0.070 |
favorite | -0.068 |
Token | ! |
Token ID | 0 |
Feature activation | +2.85e+00 |
Pos logit contributions | |
crow | +0.331 |
Bad | +0.322 |
Leave | +0.317 |
vious | +0.299 |
eper | +0.297 |
Neg logit contributions | |
pron | -0.326 |
favourite | -0.307 |
masterpiece | -0.304 |
brand | -0.276 |
favorite | -0.273 |
Token |
|
Token ID | 198 |
Feature activation | +4.58e-01 |
Pos logit contributions | |
pron | +0.217 |
favourite | +0.210 |
masterpiece | +0.205 |
brand | +0.189 |
favorite | +0.186 |
Neg logit contributions | |
crow | -0.218 |
Bad | -0.217 |
Leave | -0.213 |
vious | -0.199 |
eper | -0.199 |
Token |
|
Token ID | 198 |
Feature activation | +4.16e-01 |
Pos logit contributions | |
pron | +0.034 |
favourite | +0.032 |
masterpiece | +0.032 |
brand | +0.029 |
favorite | +0.029 |
Neg logit contributions | |
crow | -0.036 |
Bad | -0.036 |
Leave | -0.035 |
vious | -0.033 |
eper | -0.033 |
Token | He |
Token ID | 1544 |
Feature activation | -5.32e+00 |
Pos logit contributions | |
pron | +0.033 |
favourite | +0.031 |
masterpiece | +0.031 |
brand | +0.028 |
favorite | +0.027 |
Neg logit contributions | |
crow | -0.031 |
Bad | -0.030 |
ases | -0.029 |
vious | -0.028 |
Leave | -0.028 |
Token | noticed |
Token ID | 6810 |
Feature activation | +1.29e+00 |
Pos logit contributions | |
crow | +0.427 |
Leave | +0.410 |
Bad | +0.409 |
eper | +0.390 |
vious | +0.383 |
Neg logit contributions | |
pron | -0.398 |
masterpiece | -0.356 |
favourite | -0.349 |
brand | -0.323 |
favorite | -0.313 |
Token | a |
Token ID | 257 |
Feature activation | -4.40e-02 |
Pos logit contributions | |
pron | +0.097 |
favourite | +0.091 |
masterpiece | +0.090 |
brand | +0.081 |
favorite | +0.080 |
Neg logit contributions | |
crow | -0.103 |
Bad | -0.100 |
Leave | -0.098 |
vious | -0.093 |
eper | -0.092 |
Token | crowd |
Token ID | 4315 |
Feature activation | -5.63e+00 |
Pos logit contributions | |
Bad | +0.004 |
crow | +0.004 |
Leave | +0.004 |
Stop | +0.003 |
eper | +0.003 |
Neg logit contributions | |
pron | -0.003 |
favourite | -0.003 |
masterpiece | -0.003 |
favorite | -0.002 |
brand | -0.002 |
Token | at |
Token ID | 379 |
Feature activation | -3.79e-02 |
Pos logit contributions | |
crow | +0.399 |
Bad | +0.398 |
Leave | +0.382 |
eper | +0.359 |
Stop | +0.359 |
Neg logit contributions | |
pron | -0.461 |
favourite | -0.443 |
masterpiece | -0.441 |
brand | -0.405 |
favorite | -0.393 |
Token | the |
Token ID | 262 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
crow | +0.003 |
Bad | +0.003 |
Leave | +0.003 |
vious | +0.003 |
eper | +0.003 |
Neg logit contributions | |
pron | -0.003 |
favourite | -0.003 |
masterpiece | -0.003 |
brand | -0.003 |
favorite | -0.003 |
Token | park |
Token ID | 3952 |
Feature activation | -1.69e+00 |
Pos logit contributions | |
crow | +0.100 |
Bad | +0.097 |
Leave | +0.097 |
ases | +0.092 |
eper | +0.090 |
Neg logit contributions | |
pron | -0.106 |
favourite | -0.105 |
masterpiece | -0.103 |
favorite | -0.094 |
usual | -0.094 |
Token | and |
Token ID | 290 |
Feature activation | +2.53e-01 |
Pos logit contributions | |
crow | +0.134 |
Bad | +0.132 |
Leave | +0.130 |
vious | +0.122 |
eper | +0.121 |
Neg logit contributions | |
pron | -0.125 |
favourite | -0.119 |
masterpiece | -0.118 |
brand | -0.107 |
favorite | -0.105 |
Token | lift |
Token ID | 10303 |
Feature activation | +5.10e-01 |
Pos logit contributions | |
crow | +0.177 |
Bad | +0.168 |
Leave | +0.167 |
eper | +0.162 |
vious | +0.161 |
Neg logit contributions | |
pron | -0.185 |
favourite | -0.177 |
masterpiece | -0.173 |
brand | -0.157 |
favorite | -0.155 |
Token | this |
Token ID | 428 |
Feature activation | +9.76e-01 |
Pos logit contributions | |
pron | +0.041 |
favourite | +0.039 |
masterpiece | +0.039 |
brand | +0.036 |
favorite | +0.035 |
Neg logit contributions | |
crow | -0.037 |
Bad | -0.036 |
Leave | -0.036 |
vious | -0.034 |
eper | -0.033 |
Token | rock |
Token ID | 3881 |
Feature activation | -7.39e-02 |
Pos logit contributions | |
pron | +0.093 |
favourite | +0.088 |
masterpiece | +0.088 |
brand | +0.080 |
ail | +0.078 |
Neg logit contributions | |
crow | -0.054 |
Bad | -0.052 |
Leave | -0.050 |
vious | -0.049 |
eper | -0.048 |
Token | ?" |
Token ID | 1701 |
Feature activation | +6.22e+00 |
Pos logit contributions | |
crow | +0.006 |
Bad | +0.006 |
Leave | +0.005 |
vious | +0.005 |
eper | +0.005 |
Neg logit contributions | |
pron | -0.006 |
favourite | -0.005 |
masterpiece | -0.005 |
brand | -0.005 |
favorite | -0.005 |
Token | she |
Token ID | 673 |
Feature activation | +5.04e+00 |
Pos logit contributions | |
pron | +0.451 |
masterpiece | +0.445 |
favourite | +0.426 |
brand | +0.401 |
usual | +0.382 |
Neg logit contributions | |
crow | -0.519 |
Leave | -0.487 |
Bad | -0.487 |
Stop | -0.458 |
Hey | -0.453 |
Token | asked |
Token ID | 1965 |
Feature activation | -5.99e-01 |
Pos logit contributions | |
pron | +0.353 |
favourite | +0.332 |
masterpiece | +0.327 |
brand | +0.295 |
favorite | +0.292 |
Neg logit contributions | |
crow | -0.426 |
Bad | -0.416 |
Leave | -0.410 |
vious | -0.388 |
eper | -0.385 |
Token | . |
Token ID | 13 |
Feature activation | +2.26e-01 |
Pos logit contributions | |
crow | +0.050 |
Bad | +0.049 |
Leave | +0.049 |
vious | +0.046 |
eper | +0.046 |
Neg logit contributions | |
pron | -0.042 |
favourite | -0.040 |
masterpiece | -0.039 |
brand | -0.036 |
favorite | -0.035 |
Token | |
Token ID | 220 |
Feature activation | +1.15e+00 |
Pos logit contributions | |
masterpiece | +0.021 |
pron | +0.020 |
favourite | +0.020 |
brand | +0.019 |
collection | +0.018 |
Neg logit contributions | |
crow | -0.013 |
Bad | -0.013 |
Leave | -0.013 |
Stop | -0.012 |
Hey | -0.012 |
Token |
|
Token ID | 198 |
Feature activation | +7.69e-01 |
Pos logit contributions | |
pron | +0.099 |
masterpiece | +0.094 |
favourite | +0.093 |
brand | +0.086 |
collection | +0.083 |
Neg logit contributions | |
crow | -0.078 |
Bad | -0.074 |
Leave | -0.073 |
eper | -0.069 |
ases | -0.069 |
Token |
|
Token ID | 198 |
Feature activation | +7.56e-01 |
Pos logit contributions | |
pron | +0.067 |
masterpiece | +0.065 |
favourite | +0.064 |
brand | +0.059 |
collection | +0.057 |
Neg logit contributions | |
crow | -0.051 |
Bad | -0.049 |
Leave | -0.047 |
Stop | -0.045 |
eper | -0.044 |
Token | Her |
Token ID | 9360 |
Feature activation | +2.27e+00 |
Pos logit contributions | |
pron | +0.064 |
masterpiece | +0.062 |
favourite | +0.062 |
brand | +0.058 |
favorite | +0.056 |
Neg logit contributions | |
crow | -0.051 |
Bad | -0.050 |
Leave | -0.050 |
Stop | -0.046 |
eper | -0.045 |
Token | for |
Token ID | 329 |
Feature activation | +5.45e+00 |
Pos logit contributions | |
Bad | +0.094 |
crow | +0.091 |
eper | +0.086 |
Leave | +0.086 |
hunters | +0.083 |
Neg logit contributions | |
pron | -0.086 |
masterpiece | -0.082 |
favourite | -0.080 |
complexion | -0.074 |
ail | -0.073 |
Token | food |
Token ID | 2057 |
Feature activation | -4.99e+00 |
Pos logit contributions | |
pron | +0.404 |
favourite | +0.378 |
masterpiece | +0.375 |
brand | +0.339 |
favorite | +0.330 |
Neg logit contributions | |
crow | -0.432 |
Bad | -0.430 |
Leave | -0.413 |
ases | -0.398 |
eper | -0.398 |
Token | . |
Token ID | 13 |
Feature activation | +2.19e-01 |
Pos logit contributions | |
crow | +0.374 |
Bad | +0.358 |
Leave | +0.348 |
eper | +0.335 |
vious | +0.333 |
Neg logit contributions | |
pron | -0.398 |
masterpiece | -0.377 |
favourite | -0.374 |
brand | -0.344 |
favorite | -0.327 |
Token | Mom |
Token ID | 11254 |
Feature activation | +5.38e+00 |
Pos logit contributions | |
pron | +0.016 |
favourite | +0.016 |
favorite | +0.015 |
masterpiece | +0.014 |
Mam | +0.014 |
Neg logit contributions | |
crow | -0.016 |
Leave | -0.015 |
Bad | -0.015 |
vious | -0.015 |
eper | -0.015 |
Token | my |
Token ID | 1820 |
Feature activation | -3.06e+00 |
Pos logit contributions | |
pron | +0.417 |
masterpiece | +0.381 |
favourite | +0.376 |
favorite | +0.342 |
brand | +0.342 |
Neg logit contributions | |
crow | -0.413 |
Leave | -0.407 |
Bad | -0.406 |
vious | -0.368 |
Stop | -0.365 |
Token | will |
Token ID | 481 |
Feature activation | -2.65e+00 |
Pos logit contributions | |
crow | +0.249 |
Leave | +0.244 |
Bad | +0.242 |
vious | +0.224 |
Stop | +0.221 |
Neg logit contributions | |
pron | -0.222 |
masterpiece | -0.202 |
favourite | -0.201 |
favorite | -0.182 |
brand | -0.181 |
Token | be |
Token ID | 307 |
Feature activation | +1.02e+00 |
Pos logit contributions | |
Bad | +0.205 |
crow | +0.205 |
Leave | +0.199 |
Stop | +0.188 |
eper | +0.185 |
Neg logit contributions | |
pron | -0.198 |
favourite | -0.193 |
masterpiece | -0.190 |
brand | -0.174 |
favorite | -0.168 |
Token | mad |
Token ID | 8805 |
Feature activation | -7.77e+00 |
Pos logit contributions | |
pron | +0.097 |
favourite | +0.091 |
masterpiece | +0.089 |
favorite | +0.085 |
brand | +0.085 |
Neg logit contributions | |
crow | -0.064 |
Bad | -0.058 |
Leave | -0.057 |
Hey | -0.054 |
Stop | -0.054 |
Token | if |
Token ID | 611 |
Feature activation | -2.75e+00 |
Pos logit contributions | |
crow | +0.644 |
Bad | +0.637 |
Leave | +0.627 |
eper | +0.571 |
Stop | +0.571 |
Neg logit contributions | |
pron | -0.537 |
favourite | -0.533 |
masterpiece | -0.501 |
favorite | -0.487 |
brand | -0.458 |
Token | you |
Token ID | 345 |
Feature activation | -3.81e+00 |
Pos logit contributions | |
crow | +0.187 |
Bad | +0.181 |
Leave | +0.170 |
Stop | +0.165 |
eper | +0.162 |
Neg logit contributions | |
pron | -0.239 |
masterpiece | -0.232 |
favourite | -0.221 |
brand | -0.215 |
icol | -0.201 |
Token | make |
Token ID | 787 |
Feature activation | +1.20e+00 |
Pos logit contributions | |
crow | +0.289 |
Bad | +0.282 |
Leave | +0.277 |
vious | +0.259 |
eper | +0.259 |
Neg logit contributions | |
pron | -0.299 |
favourite | -0.284 |
masterpiece | -0.281 |
brand | -0.256 |
favorite | -0.253 |
Token | was |
Token ID | 373 |
Feature activation | +1.01e+00 |
Pos logit contributions | |
crow | +0.040 |
Leave | +0.040 |
Bad | +0.038 |
ases | +0.036 |
Stop | +0.036 |
Neg logit contributions | |
favourite | -0.033 |
pron | -0.032 |
masterpiece | -0.032 |
favorite | -0.031 |
collection | -0.028 |
Token | just |
Token ID | 655 |
Feature activation | +9.43e-01 |
Pos logit contributions | |
pron | +0.076 |
favourite | +0.071 |
masterpiece | +0.071 |
brand | +0.063 |
favorite | +0.062 |
Neg logit contributions | |
crow | -0.080 |
Bad | -0.079 |
Leave | -0.077 |
eper | -0.072 |
vious | -0.072 |
Token | a |
Token ID | 257 |
Feature activation | -3.72e-01 |
Pos logit contributions | |
pron | +0.071 |
favourite | +0.067 |
masterpiece | +0.066 |
brand | +0.060 |
favorite | +0.059 |
Neg logit contributions | |
crow | -0.075 |
Bad | -0.073 |
Leave | -0.071 |
vious | -0.068 |
eper | -0.067 |
Token | joke |
Token ID | 9707 |
Feature activation | -1.85e+00 |
Pos logit contributions | |
crow | +0.028 |
Bad | +0.028 |
Leave | +0.027 |
Stop | +0.025 |
vious | +0.025 |
Neg logit contributions | |
pron | -0.029 |
favourite | -0.027 |
masterpiece | -0.027 |
brand | -0.025 |
favorite | -0.024 |
Token | . |
Token ID | 13 |
Feature activation | -5.88e-01 |
Pos logit contributions | |
crow | +0.150 |
Bad | +0.149 |
Leave | +0.149 |
eper | +0.137 |
vious | +0.136 |
Neg logit contributions | |
pron | -0.132 |
favourite | -0.124 |
masterpiece | -0.119 |
favorite | -0.112 |
brand | -0.108 |
Token | Come |
Token ID | 7911 |
Feature activation | -8.38e+00 |
Pos logit contributions | |
crow | +0.044 |
Leave | +0.043 |
eper | +0.042 |
Bad | +0.042 |
vious | +0.040 |
Neg logit contributions | |
favourite | -0.041 |
pron | -0.040 |
favorite | -0.037 |
masterpiece | -0.035 |
onna | -0.035 |
Token | on |
Token ID | 319 |
Feature activation | +9.18e-01 |
Pos logit contributions | |
crow | +0.708 |
Bad | +0.680 |
Leave | +0.672 |
vious | +0.668 |
Stop | +0.631 |
Neg logit contributions | |
pron | -0.587 |
favourite | -0.571 |
masterpiece | -0.545 |
favorite | -0.512 |
brand | -0.505 |
Token | , |
Token ID | 11 |
Feature activation | -1.72e+00 |
Pos logit contributions | |
pron | +0.064 |
favourite | +0.060 |
masterpiece | +0.059 |
brand | +0.054 |
favorite | +0.053 |
Neg logit contributions | |
crow | -0.078 |
Bad | -0.076 |
Leave | -0.076 |
vious | -0.072 |
Stop | -0.071 |
Token | let |
Token ID | 1309 |
Feature activation | -6.69e-01 |
Pos logit contributions | |
crow | +0.144 |
Bad | +0.138 |
Leave | +0.137 |
ases | +0.136 |
vious | +0.134 |
Neg logit contributions | |
pron | -0.114 |
favourite | -0.107 |
masterpiece | -0.102 |
favorite | -0.097 |
brand | -0.093 |
Token | 's |
Token ID | 338 |
Feature activation | -4.34e+00 |
Pos logit contributions | |
Leave | +0.050 |
crow | +0.050 |
Bad | +0.050 |
vious | +0.045 |
ases | +0.045 |
Neg logit contributions | |
pron | -0.051 |
favourite | -0.050 |
masterpiece | -0.048 |
favorite | -0.045 |
brand | -0.043 |
Token | play |
Token ID | 711 |
Feature activation | -1.37e+00 |
Pos logit contributions | |
crow | +0.365 |
Bad | +0.356 |
Leave | +0.351 |
vious | +0.333 |
Stop | +0.326 |
Neg logit contributions | |
pron | -0.304 |
favourite | -0.291 |
masterpiece | -0.288 |
favorite | -0.257 |
brand | -0.256 |
Token | fan |
Token ID | 4336 |
Feature activation | +4.78e+00 |
Pos logit contributions | |
pron | +0.036 |
favourite | +0.035 |
masterpiece | +0.034 |
brand | +0.031 |
favorite | +0.030 |
Neg logit contributions | |
crow | -0.038 |
Bad | -0.036 |
Leave | -0.035 |
eper | -0.033 |
vious | -0.033 |
Token | in |
Token ID | 287 |
Feature activation | +1.58e+00 |
Pos logit contributions | |
pron | +0.351 |
favourite | +0.329 |
masterpiece | +0.327 |
brand | +0.295 |
favorite | +0.293 |
Neg logit contributions | |
crow | -0.381 |
Bad | -0.377 |
Leave | -0.372 |
Stop | -0.349 |
vious | -0.346 |
Token | the |
Token ID | 262 |
Feature activation | -1.43e+00 |
Pos logit contributions | |
pron | +0.123 |
masterpiece | +0.117 |
favourite | +0.117 |
brand | +0.107 |
favorite | +0.104 |
Neg logit contributions | |
crow | -0.118 |
Leave | -0.116 |
Bad | -0.115 |
vious | -0.108 |
Stop | -0.106 |
Token | forest |
Token ID | 8222 |
Feature activation | -3.98e+00 |
Pos logit contributions | |
crow | +0.118 |
Bad | +0.117 |
Leave | +0.116 |
ases | +0.110 |
eper | +0.109 |
Neg logit contributions | |
pron | -0.102 |
favourite | -0.096 |
masterpiece | -0.096 |
brand | -0.087 |
favorite | -0.085 |
Token | . |
Token ID | 13 |
Feature activation | +1.28e-01 |
Pos logit contributions | |
Bad | +0.322 |
Leave | +0.321 |
crow | +0.320 |
Stop | +0.300 |
Hey | +0.294 |
Neg logit contributions | |
pron | -0.281 |
favourite | -0.269 |
masterpiece | -0.266 |
favorite | -0.240 |
brand | -0.240 |
Token | If |
Token ID | 1002 |
Feature activation | -1.10e+01 |
Pos logit contributions | |
pron | +0.010 |
favourite | +0.010 |
masterpiece | +0.009 |
favorite | +0.009 |
brand | +0.009 |
Neg logit contributions | |
crow | -0.009 |
Bad | -0.009 |
Leave | -0.009 |
eper | -0.008 |
vious | -0.008 |
Token | you |
Token ID | 345 |
Feature activation | -3.75e+00 |
Pos logit contributions | |
crow | +0.507 |
Bad | +0.499 |
Leave | +0.473 |
Stop | +0.430 |
eper | +0.416 |
Neg logit contributions | |
pron | -1.195 |
masterpiece | -1.133 |
favourite | -1.129 |
brand | -1.058 |
favorite | -1.036 |
Token | find |
Token ID | 1064 |
Feature activation | +2.85e+00 |
Pos logit contributions | |
crow | +0.285 |
Bad | +0.280 |
Leave | +0.275 |
eper | +0.256 |
vious | +0.255 |
Neg logit contributions | |
pron | -0.291 |
favourite | -0.278 |
masterpiece | -0.274 |
brand | -0.248 |
favorite | -0.248 |
Token | it |
Token ID | 340 |
Feature activation | +5.38e-01 |
Pos logit contributions | |
pron | +0.228 |
favourite | +0.215 |
masterpiece | +0.214 |
brand | +0.195 |
favorite | +0.192 |
Neg logit contributions | |
crow | -0.212 |
Bad | -0.207 |
Leave | -0.202 |
vious | -0.190 |
eper | -0.190 |
Token | , |
Token ID | 11 |
Feature activation | -1.27e+00 |
Pos logit contributions | |
pron | +0.042 |
masterpiece | +0.039 |
favourite | +0.039 |
brand | +0.035 |
favorite | +0.034 |
Neg logit contributions | |
crow | -0.041 |
Bad | -0.040 |
Leave | -0.039 |
vious | -0.037 |
eper | -0.037 |
Token | it |
Token ID | 340 |
Feature activation | +5.28e-01 |
Pos logit contributions | |
crow | +0.089 |
Bad | +0.086 |
Leave | +0.085 |
vious | +0.079 |
eper | +0.079 |
Neg logit contributions | |
pron | -0.107 |
favourite | -0.102 |
masterpiece | -0.101 |
brand | -0.093 |
favorite | -0.092 |
Token | It |
Token ID | 632 |
Feature activation | -2.94e-01 |
Pos logit contributions | |
pron | +0.025 |
favourite | +0.024 |
masterpiece | +0.024 |
brand | +0.022 |
favorite | +0.021 |
Neg logit contributions | |
crow | -0.021 |
Bad | -0.020 |
Leave | -0.020 |
vious | -0.018 |
Stop | -0.018 |
Token | was |
Token ID | 373 |
Feature activation | +1.56e+00 |
Pos logit contributions | |
crow | +0.027 |
Bad | +0.027 |
Leave | +0.026 |
vious | +0.025 |
eper | +0.025 |
Neg logit contributions | |
pron | -0.018 |
favourite | -0.017 |
masterpiece | -0.017 |
brand | -0.015 |
favorite | -0.015 |
Token | so |
Token ID | 523 |
Feature activation | +1.18e-01 |
Pos logit contributions | |
pron | +0.113 |
masterpiece | +0.107 |
favourite | +0.104 |
collection | +0.092 |
brand | +0.092 |
Neg logit contributions | |
crow | -0.126 |
Bad | -0.126 |
Leave | -0.122 |
eper | -0.114 |
vious | -0.114 |
Token | small |
Token ID | 1402 |
Feature activation | +1.35e+00 |
Pos logit contributions | |
pron | +0.007 |
favourite | +0.006 |
masterpiece | +0.006 |
brand | +0.006 |
favorite | +0.005 |
Neg logit contributions | |
crow | -0.011 |
Bad | -0.011 |
Leave | -0.011 |
vious | -0.010 |
Stop | -0.010 |
Token | and |
Token ID | 290 |
Feature activation | +3.39e-01 |
Pos logit contributions | |
pron | +0.091 |
favourite | +0.089 |
masterpiece | +0.086 |
brand | +0.079 |
favorite | +0.079 |
Neg logit contributions | |
crow | -0.113 |
Bad | -0.113 |
Leave | -0.111 |
Stop | -0.104 |
eper | -0.103 |
Token | scared |
Token ID | 12008 |
Feature activation | -8.65e+00 |
Pos logit contributions | |
pron | +0.022 |
favourite | +0.021 |
masterpiece | +0.021 |
brand | +0.018 |
favorite | +0.018 |
Neg logit contributions | |
crow | -0.030 |
Bad | -0.029 |
Leave | -0.029 |
vious | -0.027 |
Stop | -0.027 |
Token | , |
Token ID | 11 |
Feature activation | -1.24e+00 |
Pos logit contributions | |
Bad | +0.787 |
crow | +0.771 |
Leave | +0.768 |
eper | +0.746 |
ases | +0.739 |
Neg logit contributions | |
favourite | -0.490 |
pron | -0.482 |
masterpiece | -0.460 |
favorite | -0.427 |
brand | -0.427 |
Token | so |
Token ID | 523 |
Feature activation | +1.05e-01 |
Pos logit contributions | |
crow | +0.111 |
Bad | +0.109 |
Leave | +0.108 |
Stop | +0.101 |
vious | +0.100 |
Neg logit contributions | |
pron | -0.083 |
masterpiece | -0.076 |
favourite | -0.076 |
brand | -0.067 |
favorite | -0.066 |
Token | the |
Token ID | 262 |
Feature activation | -1.58e+00 |
Pos logit contributions | |
pron | +0.006 |
masterpiece | +0.006 |
favourite | +0.006 |
brand | +0.005 |
favorite | +0.005 |
Neg logit contributions | |
crow | -0.010 |
Bad | -0.010 |
Leave | -0.010 |
Stop | -0.009 |
eper | -0.009 |
Token | girl |
Token ID | 2576 |
Feature activation | -2.37e+00 |
Pos logit contributions | |
crow | +0.137 |
Bad | +0.134 |
Leave | +0.131 |
vious | +0.125 |
eper | +0.124 |
Neg logit contributions | |
pron | -0.109 |
masterpiece | -0.101 |
favourite | -0.101 |
brand | -0.091 |
favorite | -0.087 |
Token | decided |
Token ID | 3066 |
Feature activation | -3.43e+00 |
Pos logit contributions | |
crow | +0.197 |
Bad | +0.190 |
Leave | +0.187 |
eper | +0.179 |
vious | +0.178 |
Neg logit contributions | |
pron | -0.171 |
masterpiece | -0.160 |
favourite | -0.160 |
brand | -0.146 |
favorite | -0.140 |
Token | Lily |
Token ID | 20037 |
Feature activation | +2.20e+00 |
Pos logit contributions | |
masterpiece | +0.053 |
pron | +0.052 |
favourite | +0.051 |
brand | +0.049 |
icol | +0.047 |
Neg logit contributions | |
crow | -0.043 |
Bad | -0.040 |
Leave | -0.039 |
vious | -0.038 |
Hey | -0.037 |
Token | like |
Token ID | 588 |
Feature activation | -3.47e+00 |
Pos logit contributions | |
pron | +0.185 |
favourite | +0.183 |
masterpiece | +0.178 |
favorite | +0.167 |
brand | +0.161 |
Neg logit contributions | |
crow | -0.150 |
Bad | -0.147 |
Leave | -0.143 |
eper | -0.132 |
ases | -0.130 |
Token | to |
Token ID | 284 |
Feature activation | +1.60e+00 |
Pos logit contributions | |
crow | +0.289 |
Bad | +0.281 |
Leave | +0.277 |
vious | +0.261 |
eper | +0.260 |
Neg logit contributions | |
pron | -0.250 |
favourite | -0.234 |
masterpiece | -0.232 |
brand | -0.210 |
favorite | -0.204 |
Token | see |
Token ID | 766 |
Feature activation | +4.87e-01 |
Pos logit contributions | |
pron | +0.125 |
favourite | +0.117 |
masterpiece | +0.116 |
brand | +0.108 |
favorite | +0.104 |
Neg logit contributions | |
crow | -0.124 |
Bad | -0.119 |
Leave | -0.119 |
vious | -0.111 |
ases | -0.110 |
Token | the |
Token ID | 262 |
Feature activation | -1.34e+00 |
Pos logit contributions | |
pron | +0.038 |
favourite | +0.037 |
masterpiece | +0.036 |
favorite | +0.034 |
brand | +0.034 |
Neg logit contributions | |
crow | -0.036 |
Bad | -0.035 |
Leave | -0.035 |
vious | -0.033 |
Stop | -0.032 |
Token | animals |
Token ID | 4695 |
Feature activation | -9.77e+00 |
Pos logit contributions | |
crow | +0.092 |
Leave | +0.090 |
Bad | +0.089 |
ases | +0.083 |
vious | +0.083 |
Neg logit contributions | |
favourite | -0.112 |
pron | -0.111 |
masterpiece | -0.108 |
favorite | -0.103 |
usual | -0.100 |
Token | and |
Token ID | 290 |
Feature activation | +6.47e-01 |
Pos logit contributions | |
Bad | +0.824 |
crow | +0.823 |
Leave | +0.799 |
Stop | +0.759 |
vious | +0.752 |
Neg logit contributions | |
pron | -0.654 |
favourite | -0.641 |
masterpiece | -0.624 |
favorite | -0.569 |
brand | -0.560 |
Token | learn |
Token ID | 2193 |
Feature activation | -4.88e+00 |
Pos logit contributions | |
pron | +0.050 |
favourite | +0.048 |
masterpiece | +0.047 |
brand | +0.043 |
favorite | +0.043 |
Neg logit contributions | |
crow | -0.049 |
Bad | -0.048 |
Leave | -0.047 |
vious | -0.044 |
eper | -0.044 |
Token | their |
Token ID | 511 |
Feature activation | +1.03e+00 |
Pos logit contributions | |
crow | +0.424 |
Bad | +0.415 |
Leave | +0.409 |
vious | +0.388 |
eper | +0.386 |
Neg logit contributions | |
pron | -0.328 |
favourite | -0.310 |
masterpiece | -0.305 |
brand | -0.275 |
favorite | -0.271 |
Token | names |
Token ID | 3891 |
Feature activation | -2.28e+00 |
Pos logit contributions | |
pron | +0.073 |
masterpiece | +0.067 |
favourite | +0.066 |
brand | +0.061 |
favorite | +0.057 |
Neg logit contributions | |
crow | -0.087 |
Bad | -0.085 |
Leave | -0.084 |
vious | -0.079 |
eper | -0.079 |
Token | . |
Token ID | 13 |
Feature activation | +5.71e-01 |
Pos logit contributions | |
crow | +0.184 |
Bad | +0.183 |
Leave | +0.179 |
vious | +0.168 |
Stop | +0.167 |
Neg logit contributions | |
pron | -0.165 |
favourite | -0.157 |
masterpiece | -0.155 |
brand | -0.140 |
favorite | -0.140 |
Token | saw |
Token ID | 2497 |
Feature activation | -1.86e-01 |
Pos logit contributions | |
crow | +0.201 |
Bad | +0.197 |
Leave | +0.192 |
eper | +0.186 |
vious | +0.184 |
Neg logit contributions | |
pron | -0.217 |
masterpiece | -0.200 |
favourite | -0.197 |
brand | -0.182 |
usual | -0.173 |
Token | a |
Token ID | 257 |
Feature activation | +2.71e-01 |
Pos logit contributions | |
crow | +0.015 |
Bad | +0.015 |
Leave | +0.015 |
vious | +0.014 |
eper | +0.014 |
Neg logit contributions | |
pron | -0.014 |
favourite | -0.013 |
masterpiece | -0.013 |
brand | -0.012 |
favorite | -0.011 |
Token | big |
Token ID | 1263 |
Feature activation | +5.97e+00 |
Pos logit contributions | |
pron | +0.017 |
favourite | +0.016 |
masterpiece | +0.016 |
favorite | +0.014 |
brand | +0.014 |
Neg logit contributions | |
Bad | -0.024 |
crow | -0.024 |
Leave | -0.023 |
eper | -0.022 |
Stop | -0.022 |
Token | T |
Token ID | 309 |
Feature activation | +2.73e+00 |
Pos logit contributions | |
pron | +0.451 |
favourite | +0.423 |
masterpiece | +0.422 |
brand | +0.391 |
favorite | +0.380 |
Neg logit contributions | |
crow | -0.473 |
Leave | -0.464 |
Bad | -0.460 |
vious | -0.433 |
ases | -0.430 |
Token | - |
Token ID | 12 |
Feature activation | +1.83e+00 |
Pos logit contributions | |
pron | +0.208 |
favourite | +0.197 |
masterpiece | +0.195 |
brand | +0.177 |
favorite | +0.175 |
Neg logit contributions | |
crow | -0.214 |
Bad | -0.209 |
Leave | -0.205 |
vious | -0.193 |
eper | -0.192 |
Token | Rex |
Token ID | 47389 |
Feature activation | -1.28e+01 |
Pos logit contributions | |
pron | +0.162 |
favourite | +0.152 |
masterpiece | +0.146 |
favorite | +0.141 |
brand | +0.138 |
Neg logit contributions | |
Leave | -0.122 |
Bad | -0.119 |
crow | -0.119 |
ases | -0.110 |
Stop | -0.110 |
Token | and |
Token ID | 290 |
Feature activation | +6.09e-01 |
Pos logit contributions | |
crow | +1.014 |
Bad | +0.990 |
Leave | +0.973 |
vious | +0.917 |
eper | +0.911 |
Neg logit contributions | |
pron | -0.965 |
favourite | -0.913 |
masterpiece | -0.903 |
brand | -0.821 |
favorite | -0.809 |
Token | said |
Token ID | 531 |
Feature activation | +1.42e+00 |
Pos logit contributions | |
pron | +0.051 |
masterpiece | +0.051 |
favourite | +0.048 |
ail | +0.046 |
collection | +0.044 |
Neg logit contributions | |
Bad | -0.042 |
Leave | -0.039 |
crow | -0.038 |
Stop | -0.036 |
Hey | -0.035 |
Token | , |
Token ID | 11 |
Feature activation | -7.63e-01 |
Pos logit contributions | |
pron | +0.107 |
masterpiece | +0.101 |
favourite | +0.100 |
brand | +0.090 |
favorite | +0.088 |
Neg logit contributions | |
crow | -0.113 |
Bad | -0.110 |
Leave | -0.108 |
vious | -0.102 |
eper | -0.101 |
Token | " |
Token ID | 366 |
Feature activation | +9.04e-01 |
Pos logit contributions | |
Bad | +0.056 |
crow | +0.055 |
Leave | +0.054 |
Stop | +0.050 |
eper | +0.049 |
Neg logit contributions | |
pron | -0.058 |
masterpiece | -0.058 |
favourite | -0.057 |
collection | -0.051 |
favorite | -0.050 |
Token | Mom |
Token ID | 29252 |
Feature activation | -7.20e-01 |
Pos logit contributions | |
pron | +0.067 |
favourite | +0.064 |
masterpiece | +0.060 |
favorite | +0.057 |
usual | +0.056 |
Neg logit contributions | |
crow | -0.065 |
ases | -0.064 |
risks | -0.060 |
vious | -0.060 |
eper | -0.059 |