Token | and |
Token ID | 290 |
Feature activation | +1.30e+00 |
Pos logit contributions | |
cream | +0.169 |
Blue | +0.159 |
Fin | +0.157 |
Red | +0.148 |
Tweet | +0.145 |
Neg logit contributions | |
otion | -0.205 |
upon | -0.197 |
spine | -0.193 |
laptop | -0.191 |
Deal | -0.185 |
Token | even |
Token ID | 772 |
Feature activation | -8.83e+00 |
Pos logit contributions | |
upon | +0.084 |
otion | +0.083 |
sister | +0.083 |
spine | +0.079 |
laptop | +0.078 |
Neg logit contributions | |
cream | -0.061 |
Blue | -0.058 |
Red | -0.057 |
Fin | -0.056 |
Tweet | -0.053 |
Token | today |
Token ID | 1909 |
Feature activation | +1.94e-01 |
Pos logit contributions | |
cream | +0.344 |
Blue | +0.316 |
Fin | +0.309 |
Red | +0.288 |
aw | +0.279 |
Neg logit contributions | |
otion | -0.676 |
upon | -0.661 |
spine | -0.644 |
laptop | -0.640 |
Deal | -0.622 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.55e+00 |
Pos logit contributions | |
otion | +0.013 |
upon | +0.013 |
spine | +0.012 |
laptop | +0.012 |
ancer | +0.012 |
Neg logit contributions | |
cream | -0.009 |
Blue | -0.009 |
Fin | -0.008 |
Red | -0.008 |
Tweet | -0.008 |
Token | Once |
Token ID | 7454 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
otion | +0.168 |
upon | +0.160 |
spine | +0.160 |
laptop | +0.159 |
Deal | +0.154 |
Neg logit contributions | |
cream | -0.126 |
Blue | -0.119 |
Fin | -0.117 |
Red | -0.110 |
aw | -0.109 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.00e+01 |
Pos logit contributions | |
that | +0.015 |
The | +0.015 |
never | +0.015 |
One | +0.014 |
He | +0.014 |
Neg logit contributions | |
laptop | -0.005 |
Deal | -0.005 |
spine | -0.005 |
scalp | -0.004 |
otion | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +1.345 |
upon | +1.281 |
spine | +1.273 |
laptop | +1.260 |
Deal | +1.227 |
Neg logit contributions | |
cream | -0.978 |
Blue | -0.920 |
Fin | -0.903 |
Red | -0.850 |
aw | -0.838 |
Token | time |
Token ID | 640 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
otion | +0.133 |
spine | +0.121 |
upon | +0.121 |
laptop | +0.120 |
ancer | +0.118 |
Neg logit contributions | |
cream | -0.144 |
Blue | -0.144 |
Fin | -0.143 |
Red | -0.135 |
sav | -0.135 |
Token | there |
Token ID | 612 |
Feature activation | +2.41e+00 |
Pos logit contributions | |
laptop | +0.489 |
spine | +0.483 |
Deal | +0.476 |
forehead | +0.474 |
scalp | +0.457 |
Neg logit contributions | |
Fin | -0.164 |
that | -0.164 |
Bob | -0.163 |
One | -0.150 |
never | -0.150 |
Token | was |
Token ID | 373 |
Feature activation | +5.05e-01 |
Pos logit contributions | |
otion | +0.194 |
upon | +0.190 |
laptop | +0.180 |
spine | +0.179 |
sister | +0.175 |
Neg logit contributions | |
cream | -0.088 |
Blue | -0.080 |
Fin | -0.078 |
Red | -0.072 |
Tweet | -0.069 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +0.038 |
spine | +0.037 |
laptop | +0.036 |
upon | +0.036 |
ancer | +0.035 |
Neg logit contributions | |
cream | -0.020 |
Blue | -0.020 |
Fin | -0.019 |
Red | -0.018 |
aw | -0.018 |
Token | stayed |
Token ID | 9658 |
Feature activation | -4.00e+00 |
Pos logit contributions | |
cream | +0.093 |
Fin | +0.091 |
Blue | +0.091 |
Red | +0.085 |
Tweet | +0.084 |
Neg logit contributions | |
otion | -0.139 |
upon | -0.137 |
sister | -0.129 |
laptop | -0.124 |
ancer | -0.122 |
Token | happy |
Token ID | 3772 |
Feature activation | +6.68e-01 |
Pos logit contributions | |
cream | +0.236 |
Blue | +0.219 |
Fin | +0.213 |
Red | +0.203 |
Tweet | +0.200 |
Neg logit contributions | |
upon | -0.240 |
otion | -0.230 |
spine | -0.220 |
laptop | -0.216 |
sister | -0.212 |
Token | . |
Token ID | 13 |
Feature activation | +4.11e-01 |
Pos logit contributions | |
otion | +0.046 |
upon | +0.044 |
spine | +0.043 |
laptop | +0.043 |
Deal | +0.042 |
Neg logit contributions | |
cream | -0.032 |
Blue | -0.030 |
Fin | -0.029 |
Red | -0.027 |
aw | -0.027 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.55e+00 |
Pos logit contributions | |
upon | +0.026 |
otion | +0.025 |
sister | +0.023 |
Deal | +0.023 |
Lee | +0.023 |
Neg logit contributions | |
cream | -0.021 |
Red | -0.019 |
aw | -0.019 |
Blue | -0.019 |
sailed | -0.018 |
Token | Once |
Token ID | 7454 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
otion | +0.198 |
upon | +0.195 |
laptop | +0.185 |
spine | +0.184 |
Deal | +0.181 |
Neg logit contributions | |
cream | -0.099 |
Blue | -0.088 |
Fin | -0.087 |
Red | -0.080 |
aw | -0.078 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.00e+01 |
Pos logit contributions | |
Fin | +0.015 |
The | +0.015 |
Blue | +0.015 |
never | +0.015 |
that | +0.015 |
Neg logit contributions | |
laptop | -0.005 |
spine | -0.004 |
Deal | -0.004 |
otion | -0.004 |
grade | -0.003 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +1.284 |
upon | +1.246 |
spine | +1.198 |
laptop | +1.196 |
Deal | +1.163 |
Neg logit contributions | |
cream | -1.030 |
Blue | -0.963 |
Fin | -0.955 |
Red | -0.895 |
aw | -0.884 |
Token | time |
Token ID | 640 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
otion | +0.124 |
upon | +0.115 |
spine | +0.113 |
laptop | +0.113 |
Deal | +0.108 |
Neg logit contributions | |
cream | -0.155 |
Blue | -0.151 |
Fin | -0.149 |
Red | -0.142 |
aw | -0.141 |
Token | , |
Token ID | 11 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
laptop | +0.484 |
spine | +0.477 |
forehead | +0.463 |
Deal | +0.459 |
scalp | +0.451 |
Neg logit contributions | |
Fin | -0.168 |
Bob | -0.167 |
that | -0.161 |
One | -0.159 |
Blue | -0.158 |
Token | there |
Token ID | 612 |
Feature activation | +2.41e+00 |
Pos logit contributions | |
Blue | +0.001 |
Fin | +0.001 |
The | +0.001 |
cream | +0.001 |
Tweet | +0.001 |
Neg logit contributions | |
spine | -0.001 |
laptop | -0.001 |
otion | -0.001 |
Deal | -0.001 |
ancer | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | +5.03e-01 |
Pos logit contributions | |
otion | +0.207 |
upon | +0.200 |
spine | +0.200 |
laptop | +0.198 |
Deal | +0.195 |
Neg logit contributions | |
cream | -0.071 |
Blue | -0.064 |
Fin | -0.063 |
Red | -0.056 |
aw | -0.055 |
Token | of |
Token ID | 286 |
Feature activation | +3.35e+00 |
Pos logit contributions | |
cream | +0.109 |
Blue | +0.103 |
Fin | +0.101 |
Red | +0.096 |
aw | +0.095 |
Neg logit contributions | |
otion | -0.125 |
upon | -0.120 |
spine | -0.118 |
laptop | -0.117 |
Deal | -0.114 |
Token | herself |
Token ID | 5223 |
Feature activation | -3.69e-01 |
Pos logit contributions | |
otion | +0.163 |
upon | +0.144 |
Deal | +0.136 |
spine | +0.134 |
laptop | +0.131 |
Neg logit contributions | |
cream | -0.242 |
Fin | -0.228 |
Blue | -0.225 |
Red | -0.215 |
aw | -0.212 |
Token | . |
Token ID | 13 |
Feature activation | +4.11e-01 |
Pos logit contributions | |
cream | +0.018 |
Blue | +0.017 |
never | +0.016 |
that | +0.016 |
Fin | +0.016 |
Neg logit contributions | |
otion | -0.024 |
Deal | -0.023 |
laptop | -0.023 |
spine | -0.022 |
upon | -0.022 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.55e+00 |
Pos logit contributions | |
otion | +0.026 |
upon | +0.025 |
spine | +0.025 |
laptop | +0.024 |
Deal | +0.024 |
Neg logit contributions | |
cream | -0.022 |
Blue | -0.020 |
Fin | -0.020 |
Red | -0.019 |
aw | -0.019 |
Token | Once |
Token ID | 7454 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
otion | +0.212 |
upon | +0.205 |
spine | +0.203 |
laptop | +0.202 |
Deal | +0.197 |
Neg logit contributions | |
cream | -0.082 |
Blue | -0.075 |
Fin | -0.073 |
Red | -0.066 |
aw | -0.065 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.00e+01 |
Pos logit contributions | |
that | +0.014 |
The | +0.014 |
never | +0.014 |
One | +0.014 |
Blue | +0.014 |
Neg logit contributions | |
laptop | -0.005 |
spine | -0.005 |
Deal | -0.005 |
otion | -0.005 |
deter | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +1.319 |
upon | +1.266 |
spine | +1.241 |
laptop | +1.233 |
Deal | +1.200 |
Neg logit contributions | |
cream | -1.000 |
Blue | -0.940 |
Fin | -0.927 |
Red | -0.870 |
aw | -0.858 |
Token | time |
Token ID | 640 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
otion | +0.140 |
spine | +0.126 |
ancer | +0.126 |
laptop | +0.126 |
upon | +0.125 |
Neg logit contributions | |
Blue | -0.137 |
Fin | -0.136 |
cream | -0.135 |
sav | -0.130 |
aw | -0.129 |
Token | , |
Token ID | 11 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
laptop | +0.482 |
spine | +0.476 |
forehead | +0.452 |
Deal | +0.446 |
ancer | +0.444 |
Neg logit contributions | |
Bob | -0.166 |
Fin | -0.164 |
that | -0.163 |
never | -0.162 |
Blue | -0.162 |
Token | there |
Token ID | 612 |
Feature activation | +2.41e+00 |
Pos logit contributions | |
Blue | +0.001 |
The | +0.001 |
Fin | +0.001 |
They | +0.001 |
It | +0.001 |
Neg logit contributions | |
spine | -0.001 |
laptop | -0.001 |
Deal | -0.001 |
otion | -0.001 |
ids | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | +5.03e-01 |
Pos logit contributions | |
otion | +0.204 |
upon | +0.198 |
spine | +0.195 |
laptop | +0.194 |
Deal | +0.190 |
Neg logit contributions | |
cream | -0.075 |
Blue | -0.068 |
Fin | -0.066 |
Red | -0.059 |
aw | -0.058 |
Token | needed |
Token ID | 2622 |
Feature activation | +2.97e+00 |
Pos logit contributions | |
otion | +0.045 |
upon | +0.043 |
spine | +0.043 |
laptop | +0.042 |
Deal | +0.041 |
Neg logit contributions | |
cream | -0.026 |
Blue | -0.024 |
Fin | -0.024 |
Red | -0.022 |
aw | -0.022 |
Token | it |
Token ID | 340 |
Feature activation | -3.41e+00 |
Pos logit contributions | |
otion | +0.200 |
upon | +0.193 |
spine | +0.188 |
laptop | +0.187 |
Deal | +0.180 |
Neg logit contributions | |
cream | -0.143 |
Blue | -0.136 |
Fin | -0.133 |
Red | -0.126 |
Tweet | -0.124 |
Token | . |
Token ID | 13 |
Feature activation | +4.11e-01 |
Pos logit contributions | |
cream | +0.163 |
Blue | +0.152 |
Fin | +0.150 |
Red | +0.140 |
aw | +0.138 |
Neg logit contributions | |
otion | -0.234 |
upon | -0.222 |
spine | -0.220 |
laptop | -0.219 |
Deal | -0.214 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.55e+00 |
Pos logit contributions | |
upon | +0.027 |
otion | +0.027 |
Deal | +0.025 |
ancer | +0.025 |
Lee | +0.024 |
Neg logit contributions | |
cream | -0.020 |
Blue | -0.018 |
Fin | -0.018 |
aw | -0.017 |
Red | -0.017 |
Token | Once |
Token ID | 7454 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
otion | +0.211 |
upon | +0.203 |
spine | +0.203 |
laptop | +0.201 |
Deal | +0.197 |
Neg logit contributions | |
cream | -0.084 |
Blue | -0.077 |
Fin | -0.075 |
Red | -0.068 |
aw | -0.066 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.00e+01 |
Pos logit contributions | |
that | +0.015 |
The | +0.015 |
never | +0.015 |
Fin | +0.014 |
Blue | +0.014 |
Neg logit contributions | |
laptop | -0.005 |
Deal | -0.005 |
spine | -0.005 |
otion | -0.004 |
scalp | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +1.333 |
upon | +1.274 |
spine | +1.257 |
laptop | +1.247 |
Deal | +1.214 |
Neg logit contributions | |
cream | -0.988 |
Blue | -0.931 |
Fin | -0.915 |
Red | -0.861 |
aw | -0.847 |
Token | time |
Token ID | 640 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
otion | +0.129 |
upon | +0.118 |
spine | +0.117 |
laptop | +0.117 |
ancer | +0.113 |
Neg logit contributions | |
cream | -0.149 |
Blue | -0.147 |
Fin | -0.145 |
Red | -0.138 |
aw | -0.137 |
Token | , |
Token ID | 11 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
laptop | +0.502 |
spine | +0.496 |
Deal | +0.487 |
forehead | +0.483 |
scalp | +0.475 |
Neg logit contributions | |
that | -0.154 |
Bob | -0.154 |
Fin | -0.150 |
never | -0.140 |
Blue | -0.136 |
Token | there |
Token ID | 612 |
Feature activation | +2.41e+00 |
Pos logit contributions | |
Blue | +0.001 |
Fin | +0.001 |
The | +0.001 |
Tweet | +0.001 |
cream | +0.001 |
Neg logit contributions | |
laptop | -0.001 |
spine | -0.001 |
Deal | -0.001 |
otion | -0.001 |
ancer | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | +5.03e-01 |
Pos logit contributions | |
otion | +0.206 |
upon | +0.199 |
spine | +0.197 |
laptop | +0.196 |
Deal | +0.192 |
Neg logit contributions | |
cream | -0.073 |
Blue | -0.066 |
Fin | -0.065 |
Red | -0.058 |
aw | -0.056 |
Token | The |
Token ID | 464 |
Feature activation | -4.97e+00 |
Pos logit contributions | |
cream | +0.031 |
Fin | +0.023 |
sailed | +0.023 |
aw | +0.022 |
sav | +0.021 |
Neg logit contributions | |
upon | -0.074 |
otion | -0.073 |
Deal | -0.068 |
laptop | -0.064 |
acy | -0.063 |
Token | End |
Token ID | 5268 |
Feature activation | +3.90e+00 |
Pos logit contributions | |
cream | +0.282 |
Blue | +0.268 |
Fin | +0.266 |
Red | +0.251 |
aw | +0.248 |
Neg logit contributions | |
otion | -0.297 |
upon | -0.281 |
spine | -0.278 |
laptop | -0.275 |
Deal | -0.266 |
Token | . |
Token ID | 13 |
Feature activation | +4.11e-01 |
Pos logit contributions | |
otion | +0.257 |
upon | +0.249 |
laptop | +0.242 |
spine | +0.238 |
Deal | +0.235 |
Neg logit contributions | |
cream | -0.193 |
Blue | -0.184 |
Fin | -0.181 |
Red | -0.170 |
Tweet | -0.168 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.55e+00 |
Pos logit contributions | |
otion | +0.034 |
upon | +0.033 |
spine | +0.033 |
laptop | +0.033 |
Deal | +0.032 |
Neg logit contributions | |
cream | -0.013 |
Blue | -0.012 |
Fin | -0.012 |
Red | -0.011 |
aw | -0.010 |
Token | Once |
Token ID | 7454 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
otion | +0.178 |
spine | +0.172 |
laptop | +0.170 |
upon | +0.168 |
Deal | +0.165 |
Neg logit contributions | |
cream | -0.116 |
Blue | -0.110 |
Fin | -0.107 |
Red | -0.101 |
aw | -0.100 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.00e+01 |
Pos logit contributions | |
that | +0.015 |
The | +0.014 |
never | +0.014 |
One | +0.014 |
It | +0.014 |
Neg logit contributions | |
Deal | -0.006 |
laptop | -0.006 |
spine | -0.005 |
scalp | -0.005 |
forehead | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +1.360 |
upon | +1.291 |
spine | +1.289 |
laptop | +1.275 |
Deal | +1.241 |
Neg logit contributions | |
cream | -0.965 |
Blue | -0.911 |
Fin | -0.890 |
Red | -0.842 |
aw | -0.827 |
Token | time |
Token ID | 640 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
otion | +0.133 |
upon | +0.121 |
spine | +0.120 |
laptop | +0.120 |
ancer | +0.118 |
Neg logit contributions | |
cream | -0.145 |
Blue | -0.143 |
Fin | -0.142 |
Red | -0.135 |
sav | -0.134 |
Token | there |
Token ID | 612 |
Feature activation | +2.41e+00 |
Pos logit contributions | |
laptop | +0.510 |
spine | +0.503 |
Deal | +0.493 |
forehead | +0.487 |
scalp | +0.482 |
Neg logit contributions | |
that | -0.150 |
Bob | -0.142 |
Fin | -0.136 |
Blue | -0.124 |
One | -0.124 |
Token | was |
Token ID | 373 |
Feature activation | +5.05e-01 |
Pos logit contributions | |
otion | +0.200 |
upon | +0.194 |
spine | +0.188 |
laptop | +0.188 |
Deal | +0.183 |
Neg logit contributions | |
cream | -0.081 |
Blue | -0.073 |
Fin | -0.072 |
Red | -0.065 |
aw | -0.063 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +0.034 |
upon | +0.032 |
spine | +0.032 |
laptop | +0.031 |
Deal | +0.031 |
Neg logit contributions | |
cream | -0.025 |
Blue | -0.024 |
Fin | -0.023 |
Red | -0.022 |
aw | -0.022 |
Token | 't |
Token ID | 470 |
Feature activation | +4.81e+00 |
Pos logit contributions | |
otion | +0.330 |
upon | +0.317 |
spine | +0.314 |
laptop | +0.309 |
Deal | +0.298 |
Neg logit contributions | |
cream | -0.341 |
Blue | -0.327 |
Fin | -0.321 |
Red | -0.307 |
Tweet | -0.302 |
Token | rot |
Token ID | 5724 |
Feature activation | -7.58e-01 |
Pos logit contributions | |
upon | +0.253 |
otion | +0.245 |
spine | +0.238 |
sister | +0.229 |
headache | +0.219 |
Neg logit contributions | |
cream | -0.316 |
aw | -0.297 |
Fin | -0.280 |
oc | -0.276 |
Blue | -0.274 |
Token | . |
Token ID | 13 |
Feature activation | +4.11e-01 |
Pos logit contributions | |
cream | +0.041 |
Blue | +0.039 |
Fin | +0.038 |
Red | +0.036 |
aw | +0.036 |
Neg logit contributions | |
otion | -0.047 |
upon | -0.045 |
spine | -0.044 |
laptop | -0.044 |
Deal | -0.042 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.55e+00 |
Pos logit contributions | |
otion | +0.030 |
upon | +0.030 |
Deal | +0.027 |
spine | +0.027 |
laptop | +0.027 |
Neg logit contributions | |
cream | -0.018 |
Fin | -0.016 |
Blue | -0.016 |
aw | -0.015 |
sailed | -0.015 |
Token | Once |
Token ID | 7454 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
otion | +0.207 |
upon | +0.201 |
spine | +0.196 |
laptop | +0.196 |
Deal | +0.192 |
Neg logit contributions | |
cream | -0.088 |
Blue | -0.080 |
Fin | -0.079 |
Red | -0.071 |
aw | -0.070 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.00e+01 |
Pos logit contributions | |
that | +0.015 |
The | +0.015 |
never | +0.014 |
One | +0.014 |
Blue | +0.014 |
Neg logit contributions | |
laptop | -0.005 |
spine | -0.005 |
Deal | -0.005 |
otion | -0.004 |
scalp | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +1.296 |
upon | +1.253 |
spine | +1.213 |
laptop | +1.210 |
Deal | +1.175 |
Neg logit contributions | |
cream | -1.019 |
Blue | -0.955 |
Fin | -0.947 |
Red | -0.886 |
aw | -0.876 |
Token | time |
Token ID | 640 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
otion | +0.133 |
upon | +0.120 |
spine | +0.120 |
laptop | +0.120 |
ancer | +0.117 |
Neg logit contributions | |
cream | -0.144 |
Blue | -0.143 |
Fin | -0.142 |
aw | -0.134 |
Red | -0.134 |
Token | , |
Token ID | 11 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
spine | +0.491 |
laptop | +0.489 |
Deal | +0.477 |
forehead | +0.475 |
scalp | +0.470 |
Neg logit contributions | |
that | -0.160 |
Bob | -0.150 |
never | -0.145 |
Fin | -0.145 |
Blue | -0.140 |
Token | there |
Token ID | 612 |
Feature activation | +2.41e+00 |
Pos logit contributions | |
Blue | +0.001 |
Fin | +0.001 |
cream | +0.001 |
The | +0.001 |
They | +0.001 |
Neg logit contributions | |
spine | -0.001 |
laptop | -0.001 |
Deal | -0.001 |
otion | -0.001 |
ancer | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | +5.03e-01 |
Pos logit contributions | |
otion | +0.202 |
upon | +0.196 |
spine | +0.191 |
laptop | +0.191 |
Deal | +0.187 |
Neg logit contributions | |
cream | -0.078 |
Blue | -0.071 |
Fin | -0.069 |
Red | -0.062 |
aw | -0.060 |
Token | stayed |
Token ID | 9658 |
Feature activation | -4.00e+00 |
Pos logit contributions | |
cream | +0.380 |
Blue | +0.354 |
Fin | +0.346 |
Red | +0.325 |
aw | +0.315 |
Neg logit contributions | |
otion | -0.682 |
upon | -0.662 |
spine | -0.648 |
laptop | -0.641 |
sister | -0.621 |
Token | together |
Token ID | 1978 |
Feature activation | -4.65e+00 |
Pos logit contributions | |
cream | +0.214 |
Blue | +0.196 |
Fin | +0.191 |
Red | +0.180 |
Tweet | +0.177 |
Neg logit contributions | |
upon | -0.265 |
otion | -0.250 |
spine | -0.242 |
sister | -0.241 |
laptop | -0.239 |
Token | . |
Token ID | 13 |
Feature activation | +4.11e-01 |
Pos logit contributions | |
cream | +0.226 |
Blue | +0.212 |
Fin | +0.209 |
Red | +0.196 |
aw | +0.194 |
Neg logit contributions | |
otion | -0.313 |
upon | -0.299 |
spine | -0.294 |
laptop | -0.293 |
Deal | -0.286 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.55e+00 |
Pos logit contributions | |
otion | +0.028 |
upon | +0.028 |
ancer | +0.025 |
Deal | +0.025 |
spine | +0.025 |
Neg logit contributions | |
cream | -0.019 |
Fin | -0.018 |
Blue | -0.018 |
Red | -0.017 |
aw | -0.017 |
Token | Once |
Token ID | 7454 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
otion | +0.210 |
upon | +0.203 |
spine | +0.200 |
laptop | +0.199 |
Deal | +0.195 |
Neg logit contributions | |
cream | -0.085 |
Blue | -0.078 |
Fin | -0.076 |
Red | -0.069 |
aw | -0.067 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.00e+01 |
Pos logit contributions | |
that | +0.015 |
The | +0.015 |
never | +0.014 |
Fin | +0.014 |
One | +0.014 |
Neg logit contributions | |
laptop | -0.005 |
Deal | -0.005 |
spine | -0.005 |
scalp | -0.004 |
otion | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +1.250 |
upon | +1.223 |
laptop | +1.163 |
spine | +1.158 |
Deal | +1.130 |
Neg logit contributions | |
cream | -1.057 |
Blue | -0.985 |
Fin | -0.983 |
Red | -0.917 |
aw | -0.906 |
Token | time |
Token ID | 640 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
otion | +0.126 |
upon | +0.117 |
spine | +0.115 |
laptop | +0.114 |
ancer | +0.110 |
Neg logit contributions | |
cream | -0.153 |
Blue | -0.149 |
Fin | -0.148 |
Red | -0.140 |
aw | -0.140 |
Token | , |
Token ID | 11 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
laptop | +0.496 |
spine | +0.495 |
forehead | +0.473 |
Deal | +0.473 |
scalp | +0.463 |
Neg logit contributions | |
Bob | -0.163 |
Fin | -0.161 |
that | -0.156 |
Blue | -0.148 |
The | -0.146 |
Token | there |
Token ID | 612 |
Feature activation | +2.41e+00 |
Pos logit contributions | |
Blue | +0.001 |
Fin | +0.001 |
The | +0.001 |
cream | +0.001 |
It | +0.001 |
Neg logit contributions | |
spine | -0.001 |
laptop | -0.001 |
otion | -0.001 |
Deal | -0.001 |
Lee | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | +5.03e-01 |
Pos logit contributions | |
otion | +0.203 |
upon | +0.197 |
spine | +0.193 |
laptop | +0.192 |
Deal | +0.188 |
Neg logit contributions | |
cream | -0.077 |
Blue | -0.070 |
Fin | -0.068 |
Red | -0.061 |
aw | -0.059 |
Token | was |
Token ID | 373 |
Feature activation | +5.05e-01 |
Pos logit contributions | |
otion | +0.396 |
upon | +0.383 |
spine | +0.376 |
laptop | +0.376 |
Deal | +0.365 |
Neg logit contributions | |
cream | -0.185 |
Blue | -0.172 |
Fin | -0.169 |
Red | -0.154 |
Tweet | -0.150 |
Token | made |
Token ID | 925 |
Feature activation | -6.54e+00 |
Pos logit contributions | |
otion | +0.034 |
upon | +0.033 |
laptop | +0.031 |
spine | +0.031 |
sister | +0.030 |
Neg logit contributions | |
cream | -0.025 |
Fin | -0.024 |
Blue | -0.024 |
Red | -0.022 |
Tweet | -0.022 |
Token | . |
Token ID | 13 |
Feature activation | +4.11e-01 |
Pos logit contributions | |
cream | +0.321 |
Blue | +0.291 |
Fin | +0.280 |
aw | +0.279 |
The | +0.273 |
Neg logit contributions | |
otion | -0.436 |
spine | -0.432 |
Deal | -0.418 |
laptop | -0.417 |
upon | -0.413 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.55e+00 |
Pos logit contributions | |
otion | +0.030 |
upon | +0.029 |
spine | +0.028 |
laptop | +0.028 |
Deal | +0.027 |
Neg logit contributions | |
cream | -0.018 |
Blue | -0.017 |
Fin | -0.017 |
Red | -0.015 |
aw | -0.015 |
Token | Once |
Token ID | 7454 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
otion | +0.208 |
upon | +0.202 |
spine | +0.198 |
laptop | +0.197 |
Deal | +0.193 |
Neg logit contributions | |
cream | -0.087 |
Blue | -0.079 |
Fin | -0.078 |
Red | -0.071 |
aw | -0.069 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.00e+01 |
Pos logit contributions | |
that | +0.014 |
The | +0.014 |
never | +0.014 |
One | +0.013 |
Fin | +0.013 |
Neg logit contributions | |
laptop | -0.006 |
spine | -0.006 |
Deal | -0.006 |
scalp | -0.005 |
forehead | -0.005 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +1.368 |
spine | +1.303 |
upon | +1.296 |
laptop | +1.285 |
Deal | +1.252 |
Neg logit contributions | |
cream | -0.956 |
Blue | -0.904 |
Fin | -0.881 |
Red | -0.833 |
aw | -0.819 |
Token | time |
Token ID | 640 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
otion | +0.131 |
upon | +0.120 |
spine | +0.119 |
laptop | +0.118 |
ancer | +0.115 |
Neg logit contributions | |
cream | -0.148 |
Blue | -0.146 |
Fin | -0.144 |
Red | -0.137 |
aw | -0.136 |
Token | there |
Token ID | 612 |
Feature activation | +2.41e+00 |
Pos logit contributions | |
spine | +0.510 |
laptop | +0.505 |
Deal | +0.493 |
forehead | +0.487 |
scalp | +0.482 |
Neg logit contributions | |
that | -0.152 |
Bob | -0.150 |
Fin | -0.140 |
never | -0.135 |
Blue | -0.129 |
Token | was |
Token ID | 373 |
Feature activation | +5.05e-01 |
Pos logit contributions | |
otion | +0.191 |
upon | +0.186 |
laptop | +0.176 |
spine | +0.173 |
sister | +0.171 |
Neg logit contributions | |
cream | -0.091 |
Blue | -0.084 |
Fin | -0.082 |
Red | -0.076 |
Tweet | -0.074 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +0.037 |
spine | +0.036 |
laptop | +0.035 |
upon | +0.034 |
Deal | +0.034 |
Neg logit contributions | |
cream | -0.022 |
Blue | -0.021 |
Fin | -0.020 |
Red | -0.019 |
aw | -0.019 |
Token | buy |
Token ID | 2822 |
Feature activation | +1.67e+00 |
Pos logit contributions | |
cream | +0.047 |
Blue | +0.045 |
Fin | +0.045 |
Red | +0.043 |
aw | +0.042 |
Neg logit contributions | |
otion | -0.045 |
upon | -0.044 |
spine | -0.043 |
laptop | -0.042 |
Deal | -0.041 |
Token | more |
Token ID | 517 |
Feature activation | -9.10e+00 |
Pos logit contributions | |
upon | +0.089 |
otion | +0.087 |
sister | +0.080 |
Lee | +0.078 |
911 | +0.078 |
Neg logit contributions | |
Blue | -0.100 |
Fin | -0.099 |
Tweet | -0.099 |
Red | -0.094 |
Nut | -0.091 |
Token | . |
Token ID | 13 |
Feature activation | +4.11e-01 |
Pos logit contributions | |
Blue | +0.450 |
Fin | +0.448 |
cream | +0.445 |
Red | +0.416 |
Tweet | +0.415 |
Neg logit contributions | |
otion | -0.594 |
upon | -0.583 |
laptop | -0.561 |
spine | -0.543 |
sister | -0.529 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.55e+00 |
Pos logit contributions | |
upon | +0.029 |
otion | +0.027 |
spine | +0.026 |
laptop | +0.026 |
sister | +0.026 |
Neg logit contributions | |
Fin | -0.019 |
Blue | -0.019 |
cream | -0.018 |
aw | -0.018 |
Tweet | -0.017 |
Token | Once |
Token ID | 7454 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
otion | +0.201 |
upon | +0.197 |
laptop | +0.189 |
spine | +0.188 |
Deal | +0.185 |
Neg logit contributions | |
cream | -0.095 |
Blue | -0.086 |
Fin | -0.085 |
Red | -0.077 |
aw | -0.076 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.00e+01 |
Pos logit contributions | |
The | +0.014 |
cream | +0.014 |
that | +0.014 |
never | +0.014 |
Blue | +0.014 |
Neg logit contributions | |
laptop | -0.005 |
Deal | -0.005 |
spine | -0.005 |
otion | -0.005 |
ids | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +1.320 |
upon | +1.267 |
spine | +1.241 |
laptop | +1.234 |
Deal | +1.201 |
Neg logit contributions | |
cream | -1.000 |
Blue | -0.940 |
Fin | -0.926 |
Red | -0.870 |
aw | -0.858 |
Token | time |
Token ID | 640 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
otion | +0.135 |
spine | +0.123 |
upon | +0.122 |
laptop | +0.122 |
ancer | +0.121 |
Neg logit contributions | |
cream | -0.142 |
Blue | -0.141 |
Fin | -0.140 |
aw | -0.133 |
sav | -0.132 |
Token | , |
Token ID | 11 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
laptop | +0.491 |
spine | +0.489 |
Deal | +0.474 |
forehead | +0.466 |
scalp | +0.461 |
Neg logit contributions | |
that | -0.158 |
Bob | -0.156 |
never | -0.149 |
Fin | -0.145 |
Blue | -0.144 |
Token | there |
Token ID | 612 |
Feature activation | +2.41e+00 |
Pos logit contributions | |
Blue | +0.001 |
cream | +0.001 |
Fin | +0.001 |
The | +0.001 |
This | +0.001 |
Neg logit contributions | |
spine | -0.001 |
laptop | -0.001 |
Deal | -0.001 |
otion | -0.001 |
ids | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | +5.03e-01 |
Pos logit contributions | |
otion | +0.203 |
upon | +0.197 |
spine | +0.193 |
laptop | +0.192 |
Deal | +0.188 |
Neg logit contributions | |
cream | -0.077 |
Blue | -0.069 |
Fin | -0.068 |
Red | -0.061 |
aw | -0.059 |
Token | the |
Token ID | 262 |
Feature activation | +3.14e-01 |
Pos logit contributions | |
otion | +0.091 |
upon | +0.086 |
spine | +0.085 |
laptop | +0.084 |
Deal | +0.081 |
Neg logit contributions | |
cream | -0.094 |
Blue | -0.090 |
Fin | -0.089 |
Red | -0.084 |
aw | -0.083 |
Token | garden |
Token ID | 11376 |
Feature activation | -2.94e-01 |
Pos logit contributions | |
otion | +0.021 |
upon | +0.020 |
spine | +0.020 |
laptop | +0.020 |
Deal | +0.019 |
Neg logit contributions | |
cream | -0.016 |
Blue | -0.015 |
Fin | -0.014 |
Red | -0.014 |
Tweet | -0.013 |
Token | . |
Token ID | 13 |
Feature activation | +4.11e-01 |
Pos logit contributions | |
cream | +0.013 |
Blue | +0.013 |
Fin | +0.012 |
aw | +0.012 |
Red | +0.011 |
Neg logit contributions | |
otion | -0.021 |
upon | -0.020 |
spine | -0.019 |
laptop | -0.019 |
Deal | -0.019 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.55e+00 |
Pos logit contributions | |
upon | +0.029 |
otion | +0.029 |
sister | +0.026 |
Lee | +0.026 |
spine | +0.026 |
Neg logit contributions | |
cream | -0.019 |
Fin | -0.017 |
Blue | -0.017 |
sailed | -0.016 |
Red | -0.016 |
Token | Once |
Token ID | 7454 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
otion | +0.204 |
upon | +0.199 |
laptop | +0.192 |
spine | +0.192 |
Deal | +0.188 |
Neg logit contributions | |
cream | -0.092 |
Blue | -0.083 |
Fin | -0.082 |
Red | -0.075 |
aw | -0.073 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.00e+01 |
Pos logit contributions | |
The | +0.015 |
that | +0.015 |
never | +0.014 |
One | +0.014 |
Blue | +0.014 |
Neg logit contributions | |
laptop | -0.005 |
spine | -0.005 |
Deal | -0.005 |
otion | -0.004 |
Lee | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +1.328 |
upon | +1.272 |
spine | +1.252 |
laptop | +1.243 |
Deal | +1.210 |
Neg logit contributions | |
cream | -0.991 |
Blue | -0.933 |
Fin | -0.918 |
Red | -0.863 |
aw | -0.850 |
Token | time |
Token ID | 640 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
otion | +0.132 |
upon | +0.120 |
spine | +0.120 |
laptop | +0.120 |
ancer | +0.117 |
Neg logit contributions | |
cream | -0.145 |
Blue | -0.143 |
Fin | -0.142 |
aw | -0.135 |
sav | -0.134 |
Token | , |
Token ID | 11 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
laptop | +0.485 |
spine | +0.481 |
Deal | +0.462 |
forehead | +0.460 |
scalp | +0.455 |
Neg logit contributions | |
that | -0.161 |
Bob | -0.159 |
Fin | -0.156 |
Blue | -0.154 |
never | -0.154 |
Token | there |
Token ID | 612 |
Feature activation | +2.41e+00 |
Pos logit contributions | |
Blue | +0.001 |
The | +0.001 |
Fin | +0.001 |
It | +0.001 |
This | +0.001 |
Neg logit contributions | |
spine | -0.001 |
laptop | -0.001 |
Deal | -0.001 |
otion | -0.001 |
ids | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | +5.03e-01 |
Pos logit contributions | |
otion | +0.207 |
upon | +0.200 |
spine | +0.200 |
laptop | +0.198 |
Deal | +0.195 |
Neg logit contributions | |
cream | -0.071 |
Blue | -0.064 |
Fin | -0.063 |
Red | -0.056 |
aw | -0.055 |
Token | the |
Token ID | 262 |
Feature activation | +3.14e-01 |
Pos logit contributions | |
otion | +0.065 |
laptop | +0.062 |
spine | +0.062 |
upon | +0.061 |
Deal | +0.061 |
Neg logit contributions | |
cream | -0.036 |
Blue | -0.035 |
Fin | -0.034 |
The | -0.032 |
never | -0.032 |
Token | sound |
Token ID | 2128 |
Feature activation | -2.08e+00 |
Pos logit contributions | |
otion | +0.022 |
upon | +0.021 |
spine | +0.020 |
laptop | +0.020 |
Deal | +0.020 |
Neg logit contributions | |
cream | -0.015 |
Blue | -0.014 |
Fin | -0.014 |
Red | -0.013 |
aw | -0.013 |
Token | until |
Token ID | 1566 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
cream | +0.107 |
Blue | +0.101 |
Fin | +0.099 |
Red | +0.093 |
aw | +0.093 |
Neg logit contributions | |
otion | -0.133 |
upon | -0.127 |
laptop | -0.125 |
spine | -0.125 |
Deal | -0.122 |
Token | she |
Token ID | 673 |
Feature activation | +3.52e+00 |
Pos logit contributions | |
laptop | +0.180 |
otion | +0.178 |
Deal | +0.177 |
spine | +0.170 |
upon | +0.168 |
Neg logit contributions | |
cream | -0.093 |
Blue | -0.091 |
Fin | -0.089 |
never | -0.089 |
sailed | -0.087 |
Token | stumbled |
Token ID | 24241 |
Feature activation | +2.54e+00 |
Pos logit contributions | |
otion | +0.258 |
upon | +0.249 |
spine | +0.244 |
laptop | +0.242 |
Deal | +0.236 |
Neg logit contributions | |
cream | -0.151 |
Blue | -0.140 |
Fin | -0.138 |
Red | -0.128 |
aw | -0.125 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.00e+01 |
Pos logit contributions | |
laptop | +0.159 |
otion | +0.157 |
Deal | +0.148 |
spine | +0.141 |
sister | +0.140 |
Neg logit contributions | |
He | -0.132 |
The | -0.131 |
They | -0.128 |
Blue | -0.127 |
never | -0.127 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +1.620 |
laptop | +1.583 |
Deal | +1.564 |
ancer | +1.555 |
spine | +1.536 |
Neg logit contributions | |
never | -0.726 |
sav | -0.650 |
cream | -0.649 |
aw | -0.647 |
Blue | -0.642 |
Token | beautiful |
Token ID | 4950 |
Feature activation | -7.29e+00 |
Pos logit contributions | |
upon | +0.161 |
otion | +0.159 |
laptop | +0.154 |
spine | +0.153 |
sister | +0.147 |
Neg logit contributions | |
cream | -0.124 |
Fin | -0.108 |
Blue | -0.107 |
Red | -0.100 |
Tweet | -0.098 |
Token | bird |
Token ID | 6512 |
Feature activation | -6.90e+00 |
Pos logit contributions | |
cream | +0.329 |
Blue | +0.305 |
Fin | +0.301 |
Red | +0.281 |
aw | +0.272 |
Neg logit contributions | |
otion | -0.517 |
upon | -0.503 |
spine | -0.491 |
laptop | -0.488 |
Deal | -0.468 |
Token | perched |
Token ID | 49264 |
Feature activation | -4.88e+00 |
Pos logit contributions | |
cream | +0.367 |
Blue | +0.345 |
Fin | +0.344 |
Red | +0.328 |
Jerry | +0.312 |
Neg logit contributions | |
otion | -0.435 |
upon | -0.429 |
spine | -0.410 |
laptop | -0.401 |
ancer | -0.380 |
Token | on |
Token ID | 319 |
Feature activation | +2.85e+00 |
Pos logit contributions | |
cream | +0.227 |
aw | +0.218 |
The | +0.213 |
Fin | +0.209 |
Blue | +0.209 |
Neg logit contributions | |
laptop | -0.338 |
otion | -0.335 |
Deal | -0.324 |
spine | -0.320 |
headache | -0.309 |
Token | own |
Token ID | 898 |
Feature activation | -7.83e-01 |
Pos logit contributions | |
cream | +0.061 |
Blue | +0.059 |
Fin | +0.058 |
Tweet | +0.055 |
Red | +0.054 |
Neg logit contributions | |
upon | -0.066 |
otion | -0.065 |
laptop | -0.064 |
spine | -0.064 |
sister | -0.061 |
Token | bodies |
Token ID | 5920 |
Feature activation | +1.64e+00 |
Pos logit contributions | |
cream | +0.039 |
Blue | +0.037 |
Fin | +0.037 |
Red | +0.034 |
Tweet | +0.034 |
Neg logit contributions | |
otion | -0.051 |
upon | -0.050 |
spine | -0.049 |
laptop | -0.049 |
Deal | -0.047 |
Token | . |
Token ID | 13 |
Feature activation | +4.11e-01 |
Pos logit contributions | |
otion | +0.114 |
upon | +0.108 |
spine | +0.107 |
laptop | +0.106 |
Deal | +0.106 |
Neg logit contributions | |
cream | -0.076 |
Blue | -0.071 |
Fin | -0.070 |
Red | -0.065 |
aw | -0.064 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.55e+00 |
Pos logit contributions | |
upon | +0.026 |
otion | +0.025 |
Deal | +0.024 |
Lee | +0.023 |
ancer | +0.023 |
Neg logit contributions | |
cream | -0.020 |
Fin | -0.020 |
aw | -0.019 |
sailed | -0.019 |
Blue | -0.019 |
Token | Once |
Token ID | 7454 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
otion | +0.213 |
upon | +0.205 |
spine | +0.203 |
laptop | +0.202 |
Deal | +0.198 |
Neg logit contributions | |
cream | -0.083 |
Blue | -0.076 |
Fin | -0.074 |
Red | -0.067 |
aw | -0.065 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.00e+01 |
Pos logit contributions | |
The | +0.015 |
that | +0.015 |
never | +0.015 |
Fin | +0.015 |
One | +0.015 |
Neg logit contributions | |
laptop | -0.005 |
Deal | -0.004 |
spine | -0.004 |
otion | -0.004 |
scalp | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +1.362 |
spine | +1.292 |
upon | +1.291 |
laptop | +1.277 |
Deal | +1.243 |
Neg logit contributions | |
cream | -0.963 |
Blue | -0.910 |
Fin | -0.889 |
Red | -0.840 |
aw | -0.826 |
Token | time |
Token ID | 640 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
otion | +0.137 |
spine | +0.124 |
laptop | +0.124 |
upon | +0.123 |
ancer | +0.122 |
Neg logit contributions | |
Blue | -0.140 |
cream | -0.139 |
Fin | -0.139 |
aw | -0.133 |
sav | -0.132 |
Token | , |
Token ID | 11 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
laptop | +0.507 |
spine | +0.502 |
forehead | +0.485 |
scalp | +0.482 |
Deal | +0.480 |
Neg logit contributions | |
that | -0.146 |
Bob | -0.143 |
never | -0.133 |
Fin | -0.132 |
One | -0.128 |
Token | there |
Token ID | 612 |
Feature activation | +2.41e+00 |
Pos logit contributions | |
Blue | +0.001 |
Fin | +0.001 |
The | +0.001 |
It | +0.001 |
Tweet | +0.001 |
Neg logit contributions | |
laptop | -0.001 |
spine | -0.001 |
Deal | -0.001 |
otion | -0.001 |
grade | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | +5.03e-01 |
Pos logit contributions | |
otion | +0.213 |
spine | +0.209 |
laptop | +0.206 |
upon | +0.204 |
Deal | +0.204 |
Neg logit contributions | |
cream | -0.063 |
Blue | -0.057 |
Fin | -0.056 |
aw | -0.048 |
Red | -0.048 |
Token | many |
Token ID | 867 |
Feature activation | -7.92e-01 |
Pos logit contributions | |
otion | +0.179 |
upon | +0.171 |
laptop | +0.167 |
spine | +0.166 |
Deal | +0.159 |
Neg logit contributions | |
cream | -0.135 |
Blue | -0.126 |
Fin | -0.126 |
Red | -0.118 |
Tweet | -0.115 |
Token | years |
Token ID | 812 |
Feature activation | +5.59e-01 |
Pos logit contributions | |
cream | +0.045 |
Blue | +0.042 |
Fin | +0.041 |
Red | +0.039 |
Tweet | +0.038 |
Neg logit contributions | |
upon | -0.048 |
otion | -0.045 |
sister | -0.042 |
laptop | -0.042 |
spine | -0.041 |
Token | . |
Token ID | 13 |
Feature activation | +4.11e-01 |
Pos logit contributions | |
otion | +0.037 |
upon | +0.035 |
spine | +0.035 |
laptop | +0.034 |
Deal | +0.033 |
Neg logit contributions | |
cream | -0.028 |
Blue | -0.026 |
Fin | -0.026 |
Red | -0.024 |
aw | -0.024 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.55e+00 |
Pos logit contributions | |
otion | +0.029 |
upon | +0.029 |
spine | +0.027 |
Deal | +0.027 |
ancer | +0.027 |
Neg logit contributions | |
cream | -0.019 |
Fin | -0.017 |
Blue | -0.017 |
Red | -0.016 |
Tweet | -0.016 |
Token | Once |
Token ID | 7454 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
otion | +0.210 |
upon | +0.202 |
spine | +0.202 |
laptop | +0.200 |
Deal | +0.196 |
Neg logit contributions | |
cream | -0.085 |
Blue | -0.078 |
Fin | -0.076 |
Red | -0.069 |
aw | -0.067 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.00e+01 |
Pos logit contributions | |
that | +0.015 |
The | +0.015 |
never | +0.015 |
One | +0.014 |
Fin | +0.014 |
Neg logit contributions | |
laptop | -0.005 |
Deal | -0.005 |
spine | -0.005 |
scalp | -0.004 |
forehead | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +1.326 |
upon | +1.270 |
spine | +1.249 |
laptop | +1.240 |
Deal | +1.207 |
Neg logit contributions | |
cream | -0.993 |
Blue | -0.935 |
Fin | -0.920 |
Red | -0.865 |
aw | -0.852 |
Token | time |
Token ID | 640 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
otion | +0.128 |
upon | +0.118 |
spine | +0.116 |
laptop | +0.116 |
ancer | +0.112 |
Neg logit contributions | |
cream | -0.151 |
Blue | -0.148 |
Fin | -0.146 |
Red | -0.139 |
aw | -0.138 |
Token | , |
Token ID | 11 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
laptop | +0.497 |
spine | +0.495 |
forehead | +0.476 |
Deal | +0.475 |
scalp | +0.467 |
Neg logit contributions | |
Bob | -0.160 |
Fin | -0.156 |
that | -0.155 |
One | -0.144 |
Blue | -0.139 |
Token | there |
Token ID | 612 |
Feature activation | +2.41e+00 |
Pos logit contributions | |
Blue | +0.001 |
The | +0.001 |
Fin | +0.001 |
They | +0.001 |
It | +0.001 |
Neg logit contributions | |
spine | -0.001 |
laptop | -0.001 |
Deal | -0.001 |
otion | -0.001 |
Lee | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | +5.03e-01 |
Pos logit contributions | |
otion | +0.208 |
spine | +0.201 |
upon | +0.200 |
laptop | +0.199 |
Deal | +0.196 |
Neg logit contributions | |
cream | -0.070 |
Blue | -0.064 |
Fin | -0.062 |
Red | -0.055 |
aw | -0.054 |
Token | come |
Token ID | 1282 |
Feature activation | +4.93e-01 |
Pos logit contributions | |
cream | +0.035 |
Blue | +0.032 |
Fin | +0.032 |
Red | +0.030 |
Tweet | +0.029 |
Neg logit contributions | |
upon | -0.052 |
otion | -0.051 |
spine | -0.048 |
laptop | -0.048 |
sister | -0.048 |
Token | true |
Token ID | 2081 |
Feature activation | -1.61e+00 |
Pos logit contributions | |
otion | +0.034 |
upon | +0.033 |
spine | +0.032 |
laptop | +0.032 |
Deal | +0.031 |
Neg logit contributions | |
cream | -0.023 |
Blue | -0.022 |
Fin | -0.021 |
Red | -0.020 |
aw | -0.020 |
Token | ! |
Token ID | 0 |
Feature activation | -2.36e-01 |
Pos logit contributions | |
cream | +0.082 |
Blue | +0.077 |
Fin | +0.076 |
aw | +0.071 |
Red | +0.071 |
Neg logit contributions | |
otion | -0.106 |
upon | -0.100 |
spine | -0.099 |
laptop | -0.098 |
Deal | -0.096 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.55e+00 |
Pos logit contributions | |
cream | +0.010 |
Fin | +0.009 |
Blue | +0.009 |
aw | +0.009 |
Red | +0.009 |
Neg logit contributions | |
upon | -0.017 |
otion | -0.017 |
Deal | -0.016 |
ancer | -0.016 |
spine | -0.015 |
Token | Once |
Token ID | 7454 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
otion | +0.190 |
spine | +0.185 |
laptop | +0.183 |
upon | +0.179 |
Deal | +0.178 |
Neg logit contributions | |
cream | -0.103 |
Blue | -0.098 |
Fin | -0.095 |
Red | -0.089 |
aw | -0.087 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.00e+01 |
Pos logit contributions | |
that | +0.015 |
The | +0.015 |
never | +0.015 |
He | +0.014 |
Fin | +0.014 |
Neg logit contributions | |
laptop | -0.005 |
Deal | -0.005 |
spine | -0.005 |
scalp | -0.004 |
forehead | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +1.340 |
upon | +1.279 |
spine | +1.266 |
laptop | +1.255 |
Deal | +1.222 |
Neg logit contributions | |
cream | -0.981 |
Blue | -0.925 |
Fin | -0.907 |
Red | -0.855 |
aw | -0.841 |
Token | time |
Token ID | 640 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
otion | +0.128 |
upon | +0.117 |
spine | +0.116 |
laptop | +0.115 |
ancer | +0.112 |
Neg logit contributions | |
cream | -0.151 |
Blue | -0.148 |
Fin | -0.146 |
Red | -0.139 |
aw | -0.138 |
Token | , |
Token ID | 11 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
laptop | +0.495 |
spine | +0.494 |
Deal | +0.487 |
forehead | +0.485 |
scalp | +0.470 |
Neg logit contributions | |
Bob | -0.155 |
that | -0.155 |
Fin | -0.151 |
Blue | -0.136 |
never | -0.134 |
Token | there |
Token ID | 612 |
Feature activation | +2.41e+00 |
Pos logit contributions | |
Blue | +0.001 |
Fin | +0.001 |
The | +0.001 |
They | +0.001 |
It | +0.001 |
Neg logit contributions | |
spine | -0.001 |
laptop | -0.001 |
Deal | -0.001 |
otion | -0.001 |
headache | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | +5.03e-01 |
Pos logit contributions | |
otion | +0.206 |
upon | +0.199 |
spine | +0.198 |
laptop | +0.196 |
Deal | +0.193 |
Neg logit contributions | |
cream | -0.072 |
Blue | -0.066 |
Fin | -0.064 |
Red | -0.057 |
aw | -0.056 |
Token | the |
Token ID | 262 |
Feature activation | +3.14e-01 |
Pos logit contributions | |
otion | +0.376 |
spine | +0.354 |
upon | +0.354 |
Deal | +0.352 |
laptop | +0.350 |
Neg logit contributions | |
cream | -0.259 |
Blue | -0.244 |
Fin | -0.236 |
sav | -0.227 |
Red | -0.226 |
Token | earth |
Token ID | 4534 |
Feature activation | -5.64e-01 |
Pos logit contributions | |
otion | +0.023 |
upon | +0.021 |
laptop | +0.021 |
spine | +0.021 |
Deal | +0.021 |
Neg logit contributions | |
cream | -0.014 |
Blue | -0.013 |
Fin | -0.013 |
Red | -0.012 |
sav | -0.012 |
Token | . |
Token ID | 13 |
Feature activation | +4.11e-01 |
Pos logit contributions | |
cream | +0.028 |
Blue | +0.026 |
Fin | +0.026 |
Red | +0.024 |
aw | +0.024 |
Neg logit contributions | |
otion | -0.037 |
upon | -0.036 |
spine | -0.035 |
laptop | -0.035 |
Deal | -0.034 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.55e+00 |
Pos logit contributions | |
otion | +0.031 |
upon | +0.031 |
spine | +0.029 |
laptop | +0.029 |
Deal | +0.029 |
Neg logit contributions | |
cream | -0.017 |
Fin | -0.015 |
Blue | -0.015 |
Red | -0.014 |
aw | -0.014 |
Token | Once |
Token ID | 7454 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
otion | +0.205 |
upon | +0.200 |
spine | +0.193 |
laptop | +0.193 |
Deal | +0.189 |
Neg logit contributions | |
cream | -0.091 |
Blue | -0.082 |
Fin | -0.081 |
Red | -0.073 |
aw | -0.072 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.00e+01 |
Pos logit contributions | |
that | +0.015 |
The | +0.015 |
One | +0.015 |
never | +0.014 |
Blue | +0.014 |
Neg logit contributions | |
laptop | -0.005 |
Deal | -0.004 |
spine | -0.004 |
otion | -0.004 |
deter | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +1.332 |
upon | +1.274 |
spine | +1.257 |
laptop | +1.247 |
Deal | +1.214 |
Neg logit contributions | |
cream | -0.988 |
Blue | -0.931 |
Fin | -0.914 |
Red | -0.861 |
aw | -0.847 |
Token | time |
Token ID | 640 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
otion | +0.136 |
spine | +0.123 |
upon | +0.123 |
laptop | +0.123 |
ancer | +0.121 |
Neg logit contributions | |
Blue | -0.141 |
cream | -0.141 |
Fin | -0.140 |
aw | -0.132 |
sav | -0.132 |
Token | , |
Token ID | 11 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
laptop | +0.479 |
spine | +0.476 |
forehead | +0.452 |
Deal | +0.451 |
scalp | +0.444 |
Neg logit contributions | |
Bob | -0.169 |
Fin | -0.167 |
that | -0.166 |
Blue | -0.165 |
never | -0.162 |
Token | there |
Token ID | 612 |
Feature activation | +2.41e+00 |
Pos logit contributions | |
Blue | +0.001 |
The | +0.001 |
Fin | +0.001 |
cream | +0.001 |
They | +0.001 |
Neg logit contributions | |
spine | -0.001 |
laptop | -0.001 |
Deal | -0.001 |
otion | -0.001 |
ids | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | +5.03e-01 |
Pos logit contributions | |
otion | +0.208 |
spine | +0.201 |
upon | +0.201 |
laptop | +0.200 |
Deal | +0.196 |
Neg logit contributions | |
cream | -0.070 |
Blue | -0.063 |
Fin | -0.062 |
Red | -0.054 |
aw | -0.054 |
Token | The |
Token ID | 383 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
upon | +0.029 |
otion | +0.028 |
spine | +0.026 |
Lee | +0.025 |
Deal | +0.025 |
Neg logit contributions | |
cream | -0.020 |
Fin | -0.019 |
Blue | -0.018 |
Tweet | -0.017 |
Red | -0.017 |
Token | end |
Token ID | 886 |
Feature activation | +2.86e+00 |
Pos logit contributions | |
cream | +0.051 |
Blue | +0.051 |
Fin | +0.049 |
aw | +0.047 |
sav | +0.047 |
Neg logit contributions | |
otion | -0.069 |
upon | -0.064 |
spine | -0.064 |
laptop | -0.062 |
Deal | -0.062 |
Token | . |
Token ID | 13 |
Feature activation | +4.11e-01 |
Pos logit contributions | |
upon | +0.148 |
otion | +0.145 |
laptop | +0.142 |
sister | +0.133 |
spine | +0.132 |
Neg logit contributions | |
Fin | -0.176 |
cream | -0.175 |
Blue | -0.173 |
Red | -0.163 |
Jerry | -0.163 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.55e+00 |
Pos logit contributions | |
otion | +0.031 |
upon | +0.030 |
laptop | +0.028 |
spine | +0.028 |
Deal | +0.028 |
Neg logit contributions | |
cream | -0.017 |
Fin | -0.016 |
Blue | -0.015 |
Red | -0.014 |
Tweet | -0.014 |
Token | Once |
Token ID | 7454 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
otion | +0.210 |
upon | +0.203 |
spine | +0.200 |
laptop | +0.199 |
Deal | +0.195 |
Neg logit contributions | |
cream | -0.085 |
Blue | -0.077 |
Fin | -0.075 |
Red | -0.068 |
aw | -0.067 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.00e+01 |
Pos logit contributions | |
that | +0.015 |
The | +0.014 |
never | +0.014 |
Fin | +0.014 |
One | +0.014 |
Neg logit contributions | |
laptop | -0.005 |
Deal | -0.005 |
spine | -0.005 |
scalp | -0.004 |
forehead | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +1.308 |
upon | +1.260 |
spine | +1.228 |
laptop | +1.223 |
Deal | +1.189 |
Neg logit contributions | |
cream | -1.009 |
Blue | -0.948 |
Fin | -0.936 |
Red | -0.879 |
aw | -0.866 |
Token | time |
Token ID | 640 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
otion | +0.131 |
upon | +0.120 |
spine | +0.119 |
laptop | +0.118 |
ancer | +0.115 |
Neg logit contributions | |
cream | -0.147 |
Blue | -0.145 |
Fin | -0.144 |
aw | -0.136 |
Red | -0.136 |
Token | , |
Token ID | 11 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
spine | +0.511 |
laptop | +0.509 |
Deal | +0.486 |
forehead | +0.484 |
scalp | +0.477 |
Neg logit contributions | |
that | -0.142 |
Fin | -0.140 |
Bob | -0.138 |
Blue | -0.131 |
never | -0.130 |
Token | there |
Token ID | 612 |
Feature activation | +2.41e+00 |
Pos logit contributions | |
Blue | +0.001 |
Fin | +0.001 |
The | +0.001 |
cream | +0.001 |
They | +0.001 |
Neg logit contributions | |
spine | -0.001 |
laptop | -0.001 |
Deal | -0.001 |
otion | -0.001 |
grade | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | +5.03e-01 |
Pos logit contributions | |
otion | +0.197 |
upon | +0.192 |
laptop | +0.185 |
spine | +0.184 |
Deal | +0.179 |
Neg logit contributions | |
cream | -0.084 |
Blue | -0.076 |
Fin | -0.074 |
Red | -0.068 |
Tweet | -0.065 |
Token | be |
Token ID | 307 |
Feature activation | -3.43e+00 |
Pos logit contributions | |
otion | +0.226 |
upon | +0.219 |
laptop | +0.202 |
spine | +0.196 |
Deal | +0.194 |
Neg logit contributions | |
cream | -0.249 |
Blue | -0.224 |
Fin | -0.222 |
aw | -0.220 |
Red | -0.216 |
Token | fun |
Token ID | 1257 |
Feature activation | -8.56e+00 |
Pos logit contributions | |
cream | +0.164 |
Blue | +0.154 |
Fin | +0.151 |
Red | +0.142 |
Tweet | +0.140 |
Neg logit contributions | |
otion | -0.231 |
upon | -0.226 |
laptop | -0.218 |
spine | -0.216 |
Deal | -0.209 |
Token | . |
Token ID | 13 |
Feature activation | +4.11e-01 |
Pos logit contributions | |
cream | +0.379 |
Blue | +0.349 |
Fin | +0.348 |
aw | +0.336 |
They | +0.331 |
Neg logit contributions | |
otion | -0.592 |
spine | -0.574 |
Deal | -0.562 |
upon | -0.558 |
laptop | -0.551 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.55e+00 |
Pos logit contributions | |
otion | +0.033 |
upon | +0.032 |
spine | +0.031 |
laptop | +0.031 |
Deal | +0.030 |
Neg logit contributions | |
cream | -0.015 |
Blue | -0.014 |
Fin | -0.014 |
Red | -0.013 |
aw | -0.012 |
Token | Once |
Token ID | 7454 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
otion | +0.187 |
spine | +0.183 |
laptop | +0.180 |
upon | +0.175 |
Deal | +0.175 |
Neg logit contributions | |
cream | -0.106 |
Blue | -0.101 |
Fin | -0.099 |
Red | -0.092 |
aw | -0.091 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.00e+01 |
Pos logit contributions | |
that | +0.015 |
The | +0.014 |
never | +0.014 |
One | +0.014 |
Fin | +0.014 |
Neg logit contributions | |
laptop | -0.005 |
Deal | -0.005 |
spine | -0.005 |
scalp | -0.004 |
forehead | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +1.379 |
spine | +1.312 |
upon | +1.300 |
laptop | +1.294 |
Deal | +1.261 |
Neg logit contributions | |
cream | -0.946 |
Blue | -0.897 |
Fin | -0.874 |
Red | -0.827 |
aw | -0.811 |
Token | time |
Token ID | 640 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
otion | +0.132 |
upon | +0.120 |
spine | +0.120 |
laptop | +0.119 |
ancer | +0.117 |
Neg logit contributions | |
cream | -0.145 |
Blue | -0.144 |
Fin | -0.143 |
Red | -0.135 |
aw | -0.135 |
Token | there |
Token ID | 612 |
Feature activation | +2.41e+00 |
Pos logit contributions | |
laptop | +0.496 |
spine | +0.492 |
forehead | +0.479 |
Deal | +0.478 |
scalp | +0.468 |
Neg logit contributions | |
Bob | -0.157 |
Fin | -0.154 |
that | -0.154 |
Blue | -0.141 |
never | -0.139 |
Token | was |
Token ID | 373 |
Feature activation | +5.05e-01 |
Pos logit contributions | |
otion | +0.204 |
upon | +0.197 |
spine | +0.194 |
laptop | +0.193 |
Deal | +0.189 |
Neg logit contributions | |
cream | -0.076 |
Blue | -0.069 |
Fin | -0.067 |
Red | -0.060 |
aw | -0.059 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +0.037 |
spine | +0.035 |
laptop | +0.035 |
upon | +0.034 |
Deal | +0.034 |
Neg logit contributions | |
cream | -0.022 |
Blue | -0.021 |
Fin | -0.021 |
Red | -0.020 |
aw | -0.019 |
Token | she |
Token ID | 673 |
Feature activation | +3.53e+00 |
Pos logit contributions | |
otion | +0.282 |
upon | +0.272 |
spine | +0.267 |
laptop | +0.266 |
Deal | +0.260 |
Neg logit contributions | |
cream | -0.166 |
Blue | -0.154 |
Fin | -0.152 |
Red | -0.141 |
aw | -0.139 |
Token | loved |
Token ID | 6151 |
Feature activation | -3.40e+00 |
Pos logit contributions | |
upon | +0.251 |
otion | +0.246 |
laptop | +0.241 |
spine | +0.241 |
sister | +0.233 |
Neg logit contributions | |
cream | -0.149 |
Blue | -0.143 |
Fin | -0.143 |
Red | -0.128 |
Tweet | -0.126 |
Token | . |
Token ID | 13 |
Feature activation | +4.11e-01 |
Pos logit contributions | |
cream | +0.158 |
Blue | +0.146 |
Fin | +0.141 |
Red | +0.137 |
sav | +0.136 |
Neg logit contributions | |
otion | -0.236 |
spine | -0.226 |
upon | -0.225 |
Deal | -0.222 |
laptop | -0.219 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.55e+00 |
Pos logit contributions | |
otion | +0.033 |
upon | +0.032 |
spine | +0.031 |
laptop | +0.031 |
Deal | +0.030 |
Neg logit contributions | |
cream | -0.015 |
Blue | -0.014 |
Fin | -0.013 |
Red | -0.012 |
aw | -0.012 |
Token | Once |
Token ID | 7454 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
otion | +0.210 |
upon | +0.203 |
spine | +0.201 |
laptop | +0.200 |
Deal | +0.195 |
Neg logit contributions | |
cream | -0.085 |
Blue | -0.077 |
Fin | -0.075 |
Red | -0.068 |
aw | -0.067 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.00e+01 |
Pos logit contributions | |
that | +0.015 |
The | +0.014 |
One | +0.014 |
never | +0.014 |
He | +0.013 |
Neg logit contributions | |
laptop | -0.006 |
Deal | -0.005 |
spine | -0.005 |
scalp | -0.005 |
deter | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +1.429 |
spine | +1.373 |
laptop | +1.347 |
upon | +1.329 |
Deal | +1.315 |
Neg logit contributions | |
cream | -0.899 |
Blue | -0.859 |
Fin | -0.822 |
Red | -0.787 |
The | -0.773 |
Token | time |
Token ID | 640 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
otion | +0.139 |
spine | +0.125 |
laptop | +0.125 |
ancer | +0.124 |
upon | +0.124 |
Neg logit contributions | |
Blue | -0.138 |
cream | -0.137 |
Fin | -0.137 |
sav | -0.130 |
aw | -0.130 |
Token | , |
Token ID | 11 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
spine | +0.510 |
laptop | +0.507 |
Deal | +0.484 |
forehead | +0.484 |
scalp | +0.477 |
Neg logit contributions | |
that | -0.150 |
Bob | -0.142 |
never | -0.135 |
One | -0.133 |
Fin | -0.132 |
Token | there |
Token ID | 612 |
Feature activation | +2.41e+00 |
Pos logit contributions | |
Blue | +0.001 |
The | +0.001 |
Fin | +0.001 |
It | +0.001 |
This | +0.001 |
Neg logit contributions | |
spine | -0.001 |
laptop | -0.001 |
Deal | -0.001 |
otion | -0.001 |
ids | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | +5.03e-01 |
Pos logit contributions | |
otion | +0.210 |
spine | +0.204 |
upon | +0.202 |
laptop | +0.202 |
Deal | +0.199 |
Neg logit contributions | |
cream | -0.067 |
Blue | -0.061 |
Fin | -0.059 |
Red | -0.052 |
aw | -0.052 |
Token | adventure |
Token ID | 8855 |
Feature activation | -7.40e+00 |
Pos logit contributions | |
otion | +0.165 |
laptop | +0.162 |
spine | +0.158 |
upon | +0.156 |
sister | +0.151 |
Neg logit contributions | |
cream | -0.114 |
Blue | -0.114 |
Fin | -0.105 |
Red | -0.103 |
Tweet | -0.097 |
Token | together |
Token ID | 1978 |
Feature activation | -4.65e+00 |
Pos logit contributions | |
cream | +0.438 |
Blue | +0.426 |
Red | +0.410 |
Fin | +0.407 |
Tweet | +0.394 |
Neg logit contributions | |
upon | -0.431 |
spine | -0.418 |
otion | -0.408 |
laptop | -0.395 |
ancer | -0.371 |
Token | . |
Token ID | 13 |
Feature activation | +4.11e-01 |
Pos logit contributions | |
cream | +0.233 |
Blue | +0.219 |
Fin | +0.216 |
Red | +0.202 |
aw | +0.200 |
Neg logit contributions | |
otion | -0.308 |
upon | -0.295 |
spine | -0.290 |
laptop | -0.288 |
Deal | -0.281 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.55e+00 |
Pos logit contributions | |
otion | +0.031 |
upon | +0.030 |
spine | +0.030 |
laptop | +0.030 |
Deal | +0.029 |
Neg logit contributions | |
cream | -0.016 |
Blue | -0.015 |
Fin | -0.015 |
Red | -0.014 |
aw | -0.014 |
Token | Once |
Token ID | 7454 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
otion | +0.214 |
spine | +0.207 |
laptop | +0.205 |
upon | +0.204 |
Deal | +0.200 |
Neg logit contributions | |
cream | -0.081 |
Blue | -0.074 |
Fin | -0.072 |
Red | -0.065 |
aw | -0.064 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.00e+01 |
Pos logit contributions | |
that | +0.015 |
The | +0.014 |
never | +0.014 |
One | +0.014 |
Fin | +0.014 |
Neg logit contributions | |
laptop | -0.006 |
Deal | -0.005 |
spine | -0.005 |
scalp | -0.005 |
forehead | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +1.452 |
spine | +1.398 |
laptop | +1.371 |
upon | +1.345 |
Deal | +1.337 |
Neg logit contributions | |
cream | -0.883 |
Blue | -0.842 |
Fin | -0.807 |
Red | -0.767 |
The | -0.762 |
Token | time |
Token ID | 640 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
otion | +0.131 |
upon | +0.119 |
spine | +0.119 |
laptop | +0.118 |
ancer | +0.115 |
Neg logit contributions | |
cream | -0.147 |
Blue | -0.146 |
Fin | -0.145 |
Red | -0.137 |
aw | -0.136 |
Token | , |
Token ID | 11 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
laptop | +0.506 |
spine | +0.494 |
Deal | +0.486 |
forehead | +0.483 |
scalp | +0.475 |
Neg logit contributions | |
Bob | -0.152 |
that | -0.150 |
Fin | -0.149 |
One | -0.140 |
Blue | -0.138 |
Token | there |
Token ID | 612 |
Feature activation | +2.41e+00 |
Pos logit contributions | |
Blue | +0.001 |
The | +0.001 |
Fin | +0.001 |
It | +0.001 |
Tweet | +0.000 |
Neg logit contributions | |
laptop | -0.001 |
spine | -0.001 |
Deal | -0.001 |
grade | -0.001 |
ids | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | +5.03e-01 |
Pos logit contributions | |
otion | +0.207 |
upon | +0.199 |
spine | +0.199 |
laptop | +0.197 |
Deal | +0.194 |
Neg logit contributions | |
cream | -0.072 |
Blue | -0.065 |
Fin | -0.063 |
Red | -0.056 |
aw | -0.055 |
Token | nice |
Token ID | 3621 |
Feature activation | -4.91e+00 |
Pos logit contributions | |
otion | +0.167 |
upon | +0.160 |
spine | +0.157 |
laptop | +0.156 |
Deal | +0.152 |
Neg logit contributions | |
cream | -0.115 |
Blue | -0.107 |
Fin | -0.106 |
Red | -0.099 |
aw | -0.097 |
Token | path |
Token ID | 3108 |
Feature activation | -6.10e-01 |
Pos logit contributions | |
cream | +0.224 |
Blue | +0.212 |
Fin | +0.208 |
Red | +0.194 |
aw | +0.192 |
Neg logit contributions | |
otion | -0.344 |
upon | -0.325 |
laptop | -0.322 |
spine | -0.321 |
Deal | -0.316 |
Token | ." |
Token ID | 526 |
Feature activation | -1.01e+00 |
Pos logit contributions | |
cream | +0.030 |
Blue | +0.028 |
Fin | +0.028 |
Red | +0.026 |
aw | +0.026 |
Neg logit contributions | |
otion | -0.041 |
upon | -0.039 |
spine | -0.038 |
laptop | -0.038 |
Deal | -0.037 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.55e+00 |
Pos logit contributions | |
cream | +0.040 |
Blue | +0.037 |
Fin | +0.035 |
Red | +0.034 |
sav | +0.033 |
Neg logit contributions | |
otion | -0.079 |
spine | -0.075 |
laptop | -0.074 |
upon | -0.074 |
Deal | -0.071 |
Token | Once |
Token ID | 7454 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
otion | +0.203 |
upon | +0.197 |
spine | +0.192 |
laptop | +0.192 |
Deal | +0.188 |
Neg logit contributions | |
cream | -0.093 |
Blue | -0.085 |
Fin | -0.083 |
Red | -0.076 |
aw | -0.074 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.00e+01 |
Pos logit contributions | |
The | +0.015 |
that | +0.015 |
One | +0.015 |
never | +0.015 |
He | +0.014 |
Neg logit contributions | |
laptop | -0.005 |
spine | -0.005 |
Deal | -0.004 |
otion | -0.004 |
ids | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +1.343 |
upon | +1.279 |
spine | +1.269 |
laptop | +1.257 |
Deal | +1.224 |
Neg logit contributions | |
cream | -0.979 |
Blue | -0.923 |
Fin | -0.906 |
Red | -0.853 |
aw | -0.840 |
Token | time |
Token ID | 640 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
otion | +0.137 |
spine | +0.124 |
laptop | +0.123 |
upon | +0.122 |
ancer | +0.122 |
Neg logit contributions | |
Blue | -0.140 |
cream | -0.140 |
Fin | -0.139 |
sav | -0.133 |
aw | -0.132 |
Token | , |
Token ID | 11 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
laptop | +0.477 |
spine | +0.473 |
forehead | +0.451 |
Deal | +0.450 |
scalp | +0.446 |
Neg logit contributions | |
Fin | -0.166 |
Bob | -0.166 |
that | -0.165 |
never | -0.161 |
Blue | -0.161 |
Token | there |
Token ID | 612 |
Feature activation | +2.41e+00 |
Pos logit contributions | |
Blue | +0.001 |
The | +0.001 |
Fin | +0.001 |
Tweet | +0.001 |
They | +0.001 |
Neg logit contributions | |
spine | -0.001 |
laptop | -0.001 |
otion | -0.001 |
Deal | -0.001 |
ancer | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | +5.03e-01 |
Pos logit contributions | |
otion | +0.196 |
upon | +0.191 |
laptop | +0.183 |
spine | +0.182 |
Deal | +0.177 |
Neg logit contributions | |
cream | -0.086 |
Blue | -0.078 |
Fin | -0.076 |
Red | -0.070 |
aw | -0.067 |
Token | man |
Token ID | 582 |
Feature activation | -6.67e-01 |
Pos logit contributions | |
cream | +0.070 |
Blue | +0.066 |
Fin | +0.065 |
Red | +0.060 |
Tweet | +0.059 |
Neg logit contributions | |
otion | -0.116 |
upon | -0.113 |
spine | -0.110 |
laptop | -0.110 |
Deal | -0.107 |
Token | was |
Token ID | 373 |
Feature activation | -6.44e-02 |
Pos logit contributions | |
Fin | +0.034 |
Blue | +0.032 |
Red | +0.031 |
cream | +0.030 |
Tweet | +0.030 |
Neg logit contributions | |
upon | -0.043 |
otion | -0.040 |
sister | -0.040 |
laptop | -0.039 |
Lee | -0.037 |
Token | called |
Token ID | 1444 |
Feature activation | +6.75e-01 |
Pos logit contributions | |
cream | +0.004 |
Blue | +0.003 |
Fin | +0.003 |
Red | +0.003 |
Tweet | +0.003 |
Neg logit contributions | |
otion | -0.004 |
upon | -0.004 |
laptop | -0.004 |
spine | -0.004 |
Deal | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | +1.85e+00 |
Pos logit contributions | |
spine | +0.047 |
otion | +0.047 |
upon | +0.044 |
Deal | +0.044 |
laptop | +0.044 |
Neg logit contributions | |
cream | -0.030 |
Blue | -0.027 |
Fin | -0.025 |
sold | -0.025 |
sav | -0.025 |
Token | " |
Token ID | 366 |
Feature activation | +3.54e+00 |
Pos logit contributions | |
otion | +0.135 |
spine | +0.126 |
upon | +0.126 |
Deal | +0.123 |
laptop | +0.123 |
Neg logit contributions | |
cream | -0.082 |
Blue | -0.078 |
Fin | -0.076 |
Red | -0.071 |
sav | -0.070 |
Token | v |
Token ID | 85 |
Feature activation | -1.30e+01 |
Pos logit contributions | |
upon | +0.260 |
otion | +0.256 |
ancer | +0.248 |
laptop | +0.240 |
spine | +0.238 |
Neg logit contributions | |
cream | -0.129 |
Blue | -0.125 |
Fin | -0.124 |
sav | -0.117 |
aw | -0.116 |
Token | endor |
Token ID | 18738 |
Feature activation | +1.06e+00 |
Pos logit contributions | |
cream | +0.698 |
Fin | +0.640 |
aw | +0.636 |
oc | +0.612 |
Blue | +0.606 |
Neg logit contributions | |
spine | -0.793 |
otion | -0.791 |
Deal | -0.766 |
upon | -0.761 |
laptop | -0.748 |
Token | ". |
Token ID | 1911 |
Feature activation | +9.61e-01 |
Pos logit contributions | |
otion | +0.074 |
upon | +0.071 |
spine | +0.071 |
laptop | +0.070 |
Deal | +0.069 |
Neg logit contributions | |
cream | -0.049 |
Blue | -0.045 |
Fin | -0.044 |
Red | -0.041 |
aw | -0.041 |
Token | Tim |
Token ID | 5045 |
Feature activation | -3.28e+00 |
Pos logit contributions | |
otion | +0.073 |
upon | +0.069 |
spine | +0.068 |
laptop | +0.068 |
Deal | +0.067 |
Neg logit contributions | |
cream | -0.039 |
Blue | -0.037 |
Fin | -0.036 |
Red | -0.033 |
aw | -0.033 |
Token | my |
Token ID | 1820 |
Feature activation | -4.76e+00 |
Pos logit contributions | |
cream | +0.168 |
Blue | +0.162 |
Fin | +0.160 |
Red | +0.150 |
Tweet | +0.147 |
Neg logit contributions | |
upon | -0.208 |
otion | -0.207 |
spine | -0.197 |
laptop | -0.197 |
Deal | -0.190 |
Token | wanted |
Token ID | 2227 |
Feature activation | -1.07e+00 |
Pos logit contributions | |
Fin | +0.231 |
cream | +0.230 |
Blue | +0.227 |
Red | +0.216 |
rob | +0.207 |
Neg logit contributions | |
upon | -0.311 |
sister | -0.293 |
laptop | -0.288 |
otion | -0.278 |
Lee | -0.269 |
Token | He |
Token ID | 679 |
Feature activation | -3.77e+00 |
Pos logit contributions | |
cream | +0.006 |
Blue | +0.006 |
Fin | +0.005 |
Red | +0.005 |
aw | +0.005 |
Neg logit contributions | |
otion | -0.012 |
upon | -0.012 |
spine | -0.012 |
laptop | -0.012 |
Deal | -0.011 |
Token | makes |
Token ID | 1838 |
Feature activation | -2.04e+00 |
Pos logit contributions | |
cream | +0.227 |
Blue | +0.219 |
Fin | +0.217 |
Red | +0.203 |
Tweet | +0.202 |
Neg logit contributions | |
otion | -0.207 |
upon | -0.205 |
laptop | -0.201 |
spine | -0.191 |
sister | -0.188 |
Token | v |
Token ID | 410 |
Feature activation | -2.46e+00 |
Pos logit contributions | |
cream | +0.094 |
Blue | +0.086 |
Fin | +0.085 |
aw | +0.081 |
Red | +0.080 |
Neg logit contributions | |
otion | -0.144 |
spine | -0.137 |
upon | -0.137 |
laptop | -0.134 |
Deal | -0.133 |
Token | room |
Token ID | 3823 |
Feature activation | +3.01e+00 |
Pos logit contributions | |
cream | +0.156 |
Blue | +0.150 |
Fin | +0.149 |
Red | +0.142 |
sav | +0.140 |
Neg logit contributions | |
otion | -0.128 |
upon | -0.120 |
laptop | -0.117 |
spine | -0.116 |
Deal | -0.113 |
Token | - |
Token ID | 12 |
Feature activation | -3.19e+00 |
Pos logit contributions | |
otion | +0.196 |
laptop | +0.189 |
upon | +0.188 |
spine | +0.179 |
sister | +0.178 |
Neg logit contributions | |
cream | -0.146 |
Fin | -0.143 |
Blue | -0.143 |
Jerry | -0.132 |
Red | -0.128 |
Token | v |
Token ID | 85 |
Feature activation | -1.30e+01 |
Pos logit contributions | |
sav | +0.123 |
Fin | +0.115 |
Blue | +0.110 |
Jerry | +0.110 |
sailed | +0.110 |
Neg logit contributions | |
upon | -0.221 |
otion | -0.218 |
laptop | -0.206 |
ancer | -0.204 |
ids | -0.203 |
Token | room |
Token ID | 3823 |
Feature activation | +3.03e+00 |
Pos logit contributions | |
cream | +0.618 |
aw | +0.579 |
They | +0.553 |
Fin | +0.540 |
The | +0.527 |
Neg logit contributions | |
spine | -0.860 |
otion | -0.836 |
Deal | -0.825 |
upon | -0.814 |
laptop | -0.803 |
Token | sounds |
Token ID | 5238 |
Feature activation | -5.05e+00 |
Pos logit contributions | |
otion | +0.202 |
laptop | +0.194 |
upon | +0.193 |
spine | +0.188 |
sister | +0.181 |
Neg logit contributions | |
cream | -0.146 |
Fin | -0.140 |
Blue | -0.140 |
Jerry | -0.127 |
Red | -0.126 |
Token | . |
Token ID | 13 |
Feature activation | -1.44e-01 |
Pos logit contributions | |
cream | +0.247 |
Blue | +0.233 |
Fin | +0.230 |
Red | +0.216 |
aw | +0.215 |
Neg logit contributions | |
otion | -0.338 |
upon | -0.324 |
spine | -0.318 |
laptop | -0.316 |
Deal | -0.310 |
Token | Kim |
Token ID | 6502 |
Feature activation | +1.55e+00 |
Pos logit contributions | |
cream | +0.006 |
Blue | +0.005 |
Fin | +0.005 |
aw | +0.005 |
Red | +0.005 |
Neg logit contributions | |
otion | -0.011 |
upon | -0.011 |
spine | -0.010 |
laptop | -0.010 |
Deal | -0.010 |
Token | finds |
Token ID | 7228 |
Feature activation | +1.03e+00 |
Pos logit contributions | |
otion | +0.089 |
upon | +0.086 |
laptop | +0.085 |
spine | +0.083 |
sister | +0.080 |
Neg logit contributions | |
cream | -0.088 |
Fin | -0.085 |
Blue | -0.085 |
Red | -0.079 |
Tweet | -0.078 |
Token | He |
Token ID | 679 |
Feature activation | -3.77e+00 |
Pos logit contributions | |
cream | +0.006 |
Fin | +0.006 |
Blue | +0.006 |
aw | +0.005 |
sailed | +0.005 |
Neg logit contributions | |
upon | -0.011 |
otion | -0.011 |
Deal | -0.010 |
Lee | -0.010 |
laptop | -0.010 |
Token | made |
Token ID | 925 |
Feature activation | -7.09e+00 |
Pos logit contributions | |
cream | +0.180 |
Blue | +0.168 |
Fin | +0.168 |
Red | +0.155 |
Jerry | +0.150 |
Neg logit contributions | |
upon | -0.250 |
laptop | -0.243 |
sister | -0.242 |
otion | -0.236 |
spine | -0.226 |
Token | v |
Token ID | 410 |
Feature activation | -2.46e+00 |
Pos logit contributions | |
cream | +0.318 |
Blue | +0.306 |
Fin | +0.302 |
Red | +0.276 |
Tweet | +0.270 |
Neg logit contributions | |
otion | -0.499 |
upon | -0.491 |
laptop | -0.472 |
spine | -0.463 |
sister | -0.453 |
Token | room |
Token ID | 3823 |
Feature activation | +3.02e+00 |
Pos logit contributions | |
cream | +0.161 |
Blue | +0.157 |
Fin | +0.156 |
sav | +0.148 |
Red | +0.148 |
Neg logit contributions | |
otion | -0.121 |
upon | -0.112 |
laptop | -0.109 |
spine | -0.106 |
ancer | -0.104 |
Token | - |
Token ID | 12 |
Feature activation | -3.19e+00 |
Pos logit contributions | |
otion | +0.187 |
laptop | +0.183 |
upon | +0.181 |
sister | +0.172 |
spine | +0.168 |
Neg logit contributions | |
cream | -0.153 |
Fin | -0.152 |
Blue | -0.151 |
Jerry | -0.140 |
Red | -0.135 |
Token | v |
Token ID | 85 |
Feature activation | -1.30e+01 |
Pos logit contributions | |
sav | +0.126 |
lake | +0.115 |
Fin | +0.113 |
sailed | +0.112 |
Jerry | +0.108 |
Neg logit contributions | |
upon | -0.211 |
otion | -0.204 |
laptop | -0.201 |
ids | -0.197 |
ancer | -0.192 |
Token | room |
Token ID | 3823 |
Feature activation | +3.02e+00 |
Pos logit contributions | |
cream | +0.700 |
aw | +0.635 |
Fin | +0.632 |
Blue | +0.630 |
They | +0.603 |
Neg logit contributions | |
otion | -0.791 |
spine | -0.783 |
Deal | -0.758 |
upon | -0.752 |
laptop | -0.751 |
Token | noises |
Token ID | 26782 |
Feature activation | -6.73e-01 |
Pos logit contributions | |
otion | +0.195 |
laptop | +0.192 |
upon | +0.188 |
spine | +0.182 |
sister | +0.178 |
Neg logit contributions | |
cream | -0.148 |
Fin | -0.143 |
Blue | -0.143 |
Jerry | -0.130 |
Red | -0.129 |
Token | and |
Token ID | 290 |
Feature activation | +7.26e-01 |
Pos logit contributions | |
cream | +0.035 |
Blue | +0.033 |
Fin | +0.033 |
Red | +0.031 |
Tweet | +0.030 |
Neg logit contributions | |
otion | -0.043 |
upon | -0.041 |
spine | -0.040 |
laptop | -0.040 |
Deal | -0.038 |
Token | drove |
Token ID | 10357 |
Feature activation | -2.37e+00 |
Pos logit contributions | |
otion | +0.046 |
laptop | +0.045 |
upon | +0.045 |
spine | +0.044 |
sister | +0.044 |
Neg logit contributions | |
cream | -0.037 |
Blue | -0.034 |
Fin | -0.034 |
Red | -0.032 |
aw | -0.032 |
Token | it |
Token ID | 340 |
Feature activation | -3.97e+00 |
Pos logit contributions | |
cream | +0.107 |
Blue | +0.100 |
Fin | +0.099 |
Red | +0.090 |
Tweet | +0.089 |
Neg logit contributions | |
otion | -0.167 |
upon | -0.166 |
spine | -0.159 |
laptop | -0.159 |
sister | -0.152 |
Token | cooked |
Token ID | 46591 |
Feature activation | -1.64e+00 |
Pos logit contributions | |
Blue | +0.151 |
cream | +0.147 |
Fin | +0.147 |
Red | +0.142 |
sav | +0.139 |
Neg logit contributions | |
upon | -0.194 |
otion | -0.190 |
ancer | -0.182 |
spine | -0.176 |
ids | -0.175 |
Token | and |
Token ID | 290 |
Feature activation | +7.40e-01 |
Pos logit contributions | |
cream | +0.079 |
Blue | +0.072 |
Fin | +0.072 |
aw | +0.068 |
sav | +0.067 |
Neg logit contributions | |
otion | -0.113 |
upon | -0.108 |
spine | -0.106 |
laptop | -0.104 |
Deal | -0.103 |
Token | s |
Token ID | 264 |
Feature activation | +7.54e-01 |
Pos logit contributions | |
otion | +0.051 |
upon | +0.049 |
spine | +0.048 |
laptop | +0.048 |
Deal | +0.047 |
Neg logit contributions | |
cream | -0.035 |
Blue | -0.033 |
Fin | -0.032 |
Red | -0.030 |
aw | -0.029 |
Token | ous |
Token ID | 516 |
Feature activation | +2.61e+00 |
Pos logit contributions | |
otion | +0.040 |
upon | +0.038 |
laptop | +0.036 |
spine | +0.036 |
Deal | +0.036 |
Neg logit contributions | |
cream | -0.046 |
Blue | -0.046 |
Fin | -0.045 |
Tweet | -0.042 |
Red | -0.042 |
Token | - |
Token ID | 12 |
Feature activation | -3.17e+00 |
Pos logit contributions | |
otion | +0.168 |
upon | +0.162 |
spine | +0.158 |
laptop | +0.158 |
Deal | +0.152 |
Neg logit contributions | |
cream | -0.133 |
Blue | -0.126 |
Fin | -0.124 |
Red | -0.117 |
Tweet | -0.114 |
Token | v |
Token ID | 85 |
Feature activation | -1.30e+01 |
Pos logit contributions | |
Blue | +0.141 |
Fin | +0.138 |
cream | +0.135 |
Red | +0.131 |
sailed | +0.128 |
Neg logit contributions | |
upon | -0.198 |
otion | -0.195 |
ancer | -0.187 |
ids | -0.182 |
spine | -0.180 |
Token | ide |
Token ID | 485 |
Feature activation | -2.39e+00 |
Pos logit contributions | |
cream | +0.742 |
aw | +0.694 |
Fin | +0.680 |
Blue | +0.663 |
They | +0.660 |
Neg logit contributions | |
otion | -0.740 |
spine | -0.737 |
laptop | -0.703 |
Deal | -0.702 |
upon | -0.699 |
Token | and |
Token ID | 290 |
Feature activation | +7.63e-01 |
Pos logit contributions | |
cream | +0.120 |
Blue | +0.112 |
Fin | +0.110 |
aw | +0.104 |
Red | +0.103 |
Neg logit contributions | |
otion | -0.158 |
upon | -0.150 |
spine | -0.149 |
laptop | -0.148 |
Deal | -0.146 |
Token | smoked |
Token ID | 21603 |
Feature activation | -1.71e-01 |
Pos logit contributions | |
otion | +0.047 |
upon | +0.044 |
spine | +0.043 |
laptop | +0.043 |
Deal | +0.042 |
Neg logit contributions | |
cream | -0.042 |
Blue | -0.040 |
Fin | -0.039 |
Red | -0.037 |
aw | -0.036 |
Token | and |
Token ID | 290 |
Feature activation | +7.61e-01 |
Pos logit contributions | |
cream | +0.008 |
Blue | +0.008 |
Fin | +0.007 |
aw | +0.007 |
Red | +0.007 |
Neg logit contributions | |
otion | -0.012 |
upon | -0.011 |
spine | -0.011 |
laptop | -0.011 |
Deal | -0.011 |
Token | cured |
Token ID | 30703 |
Feature activation | +8.54e-02 |
Pos logit contributions | |
otion | +0.048 |
upon | +0.045 |
spine | +0.044 |
laptop | +0.044 |
Deal | +0.043 |
Neg logit contributions | |
cream | -0.041 |
Blue | -0.039 |
Fin | -0.038 |
Red | -0.036 |
sav | -0.036 |
Token | got |
Token ID | 1392 |
Feature activation | +3.05e-02 |
Pos logit contributions | |
otion | +0.054 |
Deal | +0.051 |
laptop | +0.051 |
spine | +0.051 |
upon | +0.050 |
Neg logit contributions | |
cream | -0.032 |
Blue | -0.030 |
Fin | -0.029 |
Red | -0.027 |
aw | -0.027 |
Token | a |
Token ID | 257 |
Feature activation | +1.86e+00 |
Pos logit contributions | |
otion | +0.002 |
spine | +0.002 |
laptop | +0.002 |
upon | +0.002 |
Deal | +0.002 |
Neg logit contributions | |
cream | -0.002 |
Blue | -0.001 |
Fin | -0.001 |
Red | -0.001 |
aw | -0.001 |
Token | big |
Token ID | 1263 |
Feature activation | -1.75e+00 |
Pos logit contributions | |
otion | +0.137 |
Deal | +0.124 |
spine | +0.120 |
ancer | +0.119 |
ids | +0.118 |
Neg logit contributions | |
cream | -0.082 |
Blue | -0.080 |
Fin | -0.078 |
Red | -0.073 |
sav | -0.073 |
Token | boo |
Token ID | 22513 |
Feature activation | +1.84e+00 |
Pos logit contributions | |
cream | +0.082 |
sav | +0.081 |
lake | +0.075 |
Blue | +0.075 |
sold | +0.074 |
Neg logit contributions | |
otion | -0.118 |
ancer | -0.110 |
upon | -0.110 |
Deal | -0.109 |
acy | -0.105 |
Token | - |
Token ID | 12 |
Feature activation | -3.18e+00 |
Pos logit contributions | |
otion | +0.142 |
upon | +0.137 |
spine | +0.133 |
laptop | +0.133 |
Deal | +0.133 |
Neg logit contributions | |
cream | -0.075 |
Blue | -0.066 |
Fin | -0.065 |
aw | -0.061 |
Red | -0.060 |
Token | bo |
Token ID | 2127 |
Feature activation | -1.30e+01 |
Pos logit contributions | |
Fin | +0.070 |
Blue | +0.062 |
sav | +0.062 |
sailed | +0.059 |
Yo | +0.056 |
Neg logit contributions | |
otion | -0.273 |
upon | -0.267 |
ancer | -0.254 |
laptop | -0.252 |
spine | -0.252 |
Token | o |
Token ID | 78 |
Feature activation | -6.14e+00 |
Pos logit contributions | |
cream | +0.355 |
Blue | +0.240 |
aw | +0.239 |
Fin | +0.238 |
bo | +0.225 |
Neg logit contributions | |
Deal | -1.206 |
otion | -1.195 |
spine | -1.152 |
laptop | -1.151 |
upon | -1.122 |
Token | on |
Token ID | 319 |
Feature activation | +2.28e+00 |
Pos logit contributions | |
cream | +0.278 |
bo | +0.232 |
Blue | +0.219 |
aw | +0.218 |
never | +0.211 |
Neg logit contributions | |
Deal | -0.473 |
otion | -0.464 |
laptop | -0.433 |
spine | -0.429 |
upon | -0.429 |
Token | her |
Token ID | 607 |
Feature activation | +3.23e+00 |
Pos logit contributions | |
otion | +0.130 |
upon | +0.124 |
spine | +0.122 |
laptop | +0.119 |
ancer | +0.117 |
Neg logit contributions | |
cream | -0.136 |
Blue | -0.131 |
Fin | -0.127 |
Red | -0.123 |
sav | -0.122 |
Token | knee |
Token ID | 10329 |
Feature activation | +2.21e+00 |
Pos logit contributions | |
otion | +0.153 |
upon | +0.135 |
ancer | +0.126 |
Deal | +0.125 |
spine | +0.120 |
Neg logit contributions | |
cream | -0.236 |
Blue | -0.220 |
sav | -0.213 |
Fin | -0.213 |
Red | -0.212 |
Token | . |
Token ID | 13 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
otion | +0.141 |
upon | +0.136 |
spine | +0.135 |
laptop | +0.133 |
Deal | +0.126 |
Neg logit contributions | |
cream | -0.112 |
Blue | -0.107 |
Fin | -0.107 |
Red | -0.100 |
Tweet | -0.097 |
Token | my |
Token ID | 1820 |
Feature activation | -4.76e+00 |
Pos logit contributions | |
cream | +0.145 |
Blue | +0.134 |
Fin | +0.133 |
aw | +0.124 |
Red | +0.123 |
Neg logit contributions | |
otion | -0.236 |
spine | -0.222 |
upon | -0.220 |
laptop | -0.219 |
Deal | -0.218 |
Token | got |
Token ID | 1392 |
Feature activation | +4.90e-02 |
Pos logit contributions | |
cream | +0.184 |
Blue | +0.171 |
Fin | +0.167 |
Red | +0.154 |
Tweet | +0.149 |
Neg logit contributions | |
otion | -0.364 |
upon | -0.357 |
laptop | -0.348 |
spine | -0.347 |
Deal | -0.334 |
Token | a |
Token ID | 257 |
Feature activation | +1.86e+00 |
Pos logit contributions | |
otion | +0.003 |
spine | +0.003 |
upon | +0.003 |
laptop | +0.003 |
Deal | +0.003 |
Neg logit contributions | |
cream | -0.003 |
Blue | -0.002 |
Fin | -0.002 |
Red | -0.002 |
aw | -0.002 |
Token | boo |
Token ID | 22513 |
Feature activation | +1.85e+00 |
Pos logit contributions | |
otion | +0.143 |
Deal | +0.129 |
spine | +0.128 |
upon | +0.126 |
ancer | +0.125 |
Neg logit contributions | |
cream | -0.077 |
Blue | -0.074 |
Fin | -0.071 |
Red | -0.067 |
sold | -0.066 |
Token | - |
Token ID | 12 |
Feature activation | -3.19e+00 |
Pos logit contributions | |
otion | +0.143 |
upon | +0.137 |
spine | +0.134 |
laptop | +0.133 |
Deal | +0.133 |
Neg logit contributions | |
cream | -0.075 |
Blue | -0.067 |
Fin | -0.066 |
aw | -0.062 |
Red | -0.061 |
Token | bo |
Token ID | 2127 |
Feature activation | -1.30e+01 |
Pos logit contributions | |
Fin | +0.065 |
Blue | +0.058 |
sav | +0.055 |
Jerry | +0.049 |
sailed | +0.049 |
Neg logit contributions | |
otion | -0.280 |
upon | -0.277 |
ancer | -0.264 |
laptop | -0.264 |
spine | -0.263 |
Token | o |
Token ID | 78 |
Feature activation | -6.14e+00 |
Pos logit contributions | |
cream | +0.378 |
Blue | +0.273 |
aw | +0.265 |
Fin | +0.259 |
bo | +0.248 |
Neg logit contributions | |
otion | -1.179 |
Deal | -1.151 |
spine | -1.128 |
laptop | -1.110 |
upon | -1.108 |
Token | and |
Token ID | 290 |
Feature activation | +7.25e-01 |
Pos logit contributions | |
cream | +0.285 |
bo | +0.239 |
Blue | +0.238 |
aw | +0.229 |
never | +0.227 |
Neg logit contributions | |
otion | -0.450 |
Deal | -0.444 |
upon | -0.421 |
spine | -0.418 |
laptop | -0.409 |
Token | needed |
Token ID | 2622 |
Feature activation | +2.41e+00 |
Pos logit contributions | |
otion | +0.051 |
upon | +0.049 |
spine | +0.048 |
laptop | +0.048 |
Deal | +0.047 |
Neg logit contributions | |
cream | -0.033 |
Blue | -0.031 |
Fin | -0.031 |
Red | -0.029 |
aw | -0.028 |
Token | a |
Token ID | 257 |
Feature activation | +1.86e+00 |
Pos logit contributions | |
upon | +0.142 |
otion | +0.140 |
laptop | +0.135 |
spine | +0.132 |
Deal | +0.130 |
Neg logit contributions | |
cream | -0.128 |
Fin | -0.127 |
Blue | -0.126 |
aw | -0.118 |
Red | -0.116 |
Token | needle |
Token ID | 17598 |
Feature activation | +1.07e+00 |
Pos logit contributions | |
otion | +0.124 |
upon | +0.107 |
spine | +0.107 |
Deal | +0.106 |
laptop | +0.105 |
Neg logit contributions | |
cream | -0.097 |
Blue | -0.094 |
Fin | -0.090 |
Red | -0.088 |
sav | -0.086 |
Token | behind |
Token ID | 2157 |
Feature activation | +8.05e+00 |
Pos logit contributions | |
otion | +0.112 |
upon | +0.106 |
spine | +0.106 |
laptop | +0.105 |
Deal | +0.102 |
Neg logit contributions | |
cream | -0.089 |
Blue | -0.084 |
Fin | -0.083 |
Red | -0.078 |
aw | -0.078 |
Token | something |
Token ID | 1223 |
Feature activation | +3.31e+00 |
Pos logit contributions | |
otion | +0.618 |
spine | +0.595 |
Deal | +0.585 |
laptop | +0.578 |
upon | +0.572 |
Neg logit contributions | |
cream | -0.311 |
Blue | -0.295 |
aw | -0.285 |
Fin | -0.273 |
sav | -0.267 |
Token | and |
Token ID | 290 |
Feature activation | +7.47e-01 |
Pos logit contributions | |
otion | +0.236 |
upon | +0.226 |
spine | +0.223 |
laptop | +0.222 |
Deal | +0.218 |
Neg logit contributions | |
cream | -0.147 |
Blue | -0.138 |
Fin | -0.135 |
Red | -0.126 |
aw | -0.125 |
Token | say |
Token ID | 910 |
Feature activation | +1.54e+00 |
Pos logit contributions | |
otion | +0.046 |
upon | +0.045 |
spine | +0.044 |
laptop | +0.043 |
Deal | +0.042 |
Neg logit contributions | |
cream | -0.040 |
Blue | -0.038 |
Fin | -0.037 |
Red | -0.035 |
aw | -0.035 |
Token | ' |
Token ID | 705 |
Feature activation | +2.61e+00 |
Pos logit contributions | |
otion | +0.089 |
spine | +0.088 |
upon | +0.088 |
laptop | +0.087 |
ancer | +0.082 |
Neg logit contributions | |
cream | -0.083 |
Blue | -0.083 |
aw | -0.079 |
Fin | -0.079 |
never | -0.078 |
Token | v |
Token ID | 85 |
Feature activation | -1.30e+01 |
Pos logit contributions | |
upon | +0.188 |
otion | +0.185 |
ancer | +0.180 |
laptop | +0.172 |
Deal | +0.171 |
Neg logit contributions | |
cream | -0.092 |
Fin | -0.087 |
Blue | -0.087 |
sav | -0.084 |
sailed | -0.081 |
Token | anish |
Token ID | 7115 |
Feature activation | +7.33e+00 |
Pos logit contributions | |
cream | +0.731 |
aw | +0.707 |
Fin | +0.692 |
bo | +0.677 |
Blue | +0.668 |
Neg logit contributions | |
spine | -0.745 |
otion | -0.688 |
upon | -0.678 |
ancer | -0.664 |
Deal | -0.662 |
Token | !' |
Token ID | 13679 |
Feature activation | +2.44e+00 |
Pos logit contributions | |
otion | +0.482 |
spine | +0.478 |
upon | +0.469 |
laptop | +0.453 |
ancer | +0.440 |
Neg logit contributions | |
cream | -0.355 |
Blue | -0.342 |
Fin | -0.328 |
aw | -0.319 |
Red | -0.317 |
Token | and |
Token ID | 290 |
Feature activation | +7.78e-01 |
Pos logit contributions | |
otion | +0.172 |
spine | +0.164 |
upon | +0.162 |
laptop | +0.162 |
ancer | +0.156 |
Neg logit contributions | |
cream | -0.111 |
Blue | -0.107 |
Fin | -0.102 |
aw | -0.101 |
sav | -0.098 |
Token | then |
Token ID | 788 |
Feature activation | -4.88e+00 |
Pos logit contributions | |
otion | +0.060 |
upon | +0.057 |
spine | +0.057 |
laptop | +0.056 |
ancer | +0.056 |
Neg logit contributions | |
cream | -0.030 |
Blue | -0.027 |
sav | -0.026 |
Fin | -0.026 |
aw | -0.025 |
Token | come |
Token ID | 1282 |
Feature activation | -5.30e-03 |
Pos logit contributions | |
cream | +0.239 |
Blue | +0.224 |
never | +0.218 |
sav | +0.217 |
sailed | +0.213 |
Neg logit contributions | |
otion | -0.320 |
spine | -0.311 |
Deal | -0.299 |
upon | -0.296 |
laptop | -0.295 |
Token | and |
Token ID | 290 |
Feature activation | +7.67e-01 |
Pos logit contributions | |
cream | +0.028 |
Blue | +0.026 |
Fin | +0.025 |
aw | +0.024 |
They | +0.023 |
Neg logit contributions | |
otion | -0.051 |
spine | -0.049 |
Deal | -0.048 |
upon | -0.048 |
laptop | -0.047 |
Token | we |
Token ID | 356 |
Feature activation | +2.41e-01 |
Pos logit contributions | |
otion | +0.044 |
upon | +0.043 |
spine | +0.041 |
laptop | +0.041 |
Deal | +0.040 |
Neg logit contributions | |
cream | -0.044 |
Blue | -0.042 |
Fin | -0.042 |
Red | -0.040 |
aw | -0.039 |
Token | have |
Token ID | 423 |
Feature activation | -5.55e-01 |
Pos logit contributions | |
otion | +0.013 |
upon | +0.013 |
spine | +0.012 |
laptop | +0.012 |
Deal | +0.012 |
Neg logit contributions | |
cream | -0.015 |
Blue | -0.014 |
Fin | -0.014 |
Red | -0.013 |
aw | -0.013 |
Token | boo |
Token ID | 22513 |
Feature activation | +1.85e+00 |
Pos logit contributions | |
cream | +0.029 |
Blue | +0.027 |
Fin | +0.027 |
Red | +0.026 |
Tweet | +0.025 |
Neg logit contributions | |
upon | -0.035 |
otion | -0.034 |
laptop | -0.033 |
spine | -0.033 |
sister | -0.033 |
Token | - |
Token ID | 12 |
Feature activation | -3.17e+00 |
Pos logit contributions | |
otion | +0.124 |
upon | +0.119 |
laptop | +0.118 |
spine | +0.116 |
sister | +0.115 |
Neg logit contributions | |
cream | -0.088 |
Blue | -0.086 |
Fin | -0.084 |
Red | -0.079 |
Jerry | -0.077 |
Token | bo |
Token ID | 2127 |
Feature activation | -1.29e+01 |
Pos logit contributions | |
sav | +0.089 |
Fin | +0.087 |
Blue | +0.086 |
tre | +0.081 |
sailed | +0.080 |
Neg logit contributions | |
otion | -0.253 |
upon | -0.243 |
ancer | -0.235 |
laptop | -0.230 |
ids | -0.226 |
Token | os |
Token ID | 418 |
Feature activation | +4.36e+00 |
Pos logit contributions | |
cream | +0.404 |
aw | +0.339 |
bo | +0.305 |
Blue | +0.298 |
Fin | +0.297 |
Neg logit contributions | |
otion | -1.130 |
spine | -1.099 |
Deal | -1.092 |
laptop | -1.063 |
upon | -1.040 |
Token | ." |
Token ID | 526 |
Feature activation | -1.58e+00 |
Pos logit contributions | |
otion | +0.309 |
upon | +0.298 |
spine | +0.295 |
laptop | +0.292 |
Deal | +0.285 |
Neg logit contributions | |
cream | -0.196 |
Blue | -0.183 |
Fin | -0.178 |
Red | -0.168 |
aw | -0.166 |
Token |
|
Token ID | 198 |
Feature activation | -1.46e+00 |
Pos logit contributions | |
cream | +0.051 |
Blue | +0.048 |
Fin | +0.046 |
Red | +0.042 |
sav | +0.041 |
Neg logit contributions | |
otion | -0.133 |
upon | -0.126 |
spine | -0.126 |
laptop | -0.125 |
Deal | -0.122 |
Token |
|
Token ID | 198 |
Feature activation | -1.46e+00 |
Pos logit contributions | |
cream | +0.062 |
Fin | +0.053 |
Blue | +0.053 |
Red | +0.049 |
sailed | +0.049 |
Neg logit contributions | |
otion | -0.108 |
upon | -0.108 |
laptop | -0.102 |
Deal | -0.101 |
spine | -0.099 |
Token | The |
Token ID | 464 |
Feature activation | -5.54e+00 |
Pos logit contributions | |
cream | +0.066 |
sailed | +0.052 |
sold | +0.050 |
aw | +0.048 |
Fin | +0.045 |
Neg logit contributions | |
upon | -0.105 |
laptop | -0.101 |
otion | -0.100 |
Deal | -0.100 |
Lee | -0.096 |
Token | and |
Token ID | 290 |
Feature activation | +8.09e-01 |
Pos logit contributions | |
otion | +0.151 |
upon | +0.145 |
spine | +0.143 |
laptop | +0.142 |
Deal | +0.139 |
Neg logit contributions | |
cream | -0.098 |
Blue | -0.092 |
Fin | -0.090 |
Red | -0.085 |
aw | -0.083 |
Token | fight |
Token ID | 1907 |
Feature activation | +9.99e-01 |
Pos logit contributions | |
otion | +0.054 |
upon | +0.051 |
spine | +0.050 |
laptop | +0.049 |
Deal | +0.049 |
Neg logit contributions | |
cream | -0.040 |
Blue | -0.038 |
Fin | -0.037 |
Red | -0.035 |
Tweet | -0.035 |
Token | and |
Token ID | 290 |
Feature activation | +7.67e-01 |
Pos logit contributions | |
otion | +0.068 |
upon | +0.065 |
spine | +0.064 |
laptop | +0.064 |
Deal | +0.062 |
Neg logit contributions | |
cream | -0.048 |
Blue | -0.045 |
Fin | -0.044 |
Red | -0.042 |
aw | -0.041 |
Token | say |
Token ID | 910 |
Feature activation | +1.52e+00 |
Pos logit contributions | |
otion | +0.049 |
upon | +0.046 |
spine | +0.045 |
Deal | +0.044 |
laptop | +0.044 |
Neg logit contributions | |
cream | -0.040 |
Blue | -0.039 |
Fin | -0.038 |
Red | -0.036 |
sav | -0.035 |
Token | " |
Token ID | 366 |
Feature activation | +3.55e+00 |
Pos logit contributions | |
otion | +0.096 |
spine | +0.094 |
upon | +0.093 |
laptop | +0.090 |
Deal | +0.089 |
Neg logit contributions | |
cream | -0.078 |
Blue | -0.075 |
Fin | -0.074 |
aw | -0.072 |
Red | -0.070 |
Token | bo |
Token ID | 2127 |
Feature activation | -1.29e+01 |
Pos logit contributions | |
otion | +0.261 |
upon | +0.258 |
laptop | +0.250 |
spine | +0.245 |
ancer | +0.244 |
Neg logit contributions | |
cream | -0.143 |
Blue | -0.127 |
Fin | -0.126 |
sav | -0.118 |
aw | -0.115 |
Token | om |
Token ID | 296 |
Feature activation | +3.63e+00 |
Pos logit contributions | |
cream | +0.592 |
aw | +0.548 |
Fin | +0.537 |
Blue | +0.510 |
oc | +0.507 |
Neg logit contributions | |
otion | -0.922 |
spine | -0.890 |
Deal | -0.859 |
laptop | -0.844 |
upon | -0.821 |
Token | " |
Token ID | 1 |
Feature activation | -2.35e+00 |
Pos logit contributions | |
otion | +0.265 |
spine | +0.259 |
upon | +0.257 |
laptop | +0.248 |
Deal | +0.246 |
Neg logit contributions | |
cream | -0.155 |
Fin | -0.143 |
Blue | -0.142 |
aw | -0.135 |
Red | -0.129 |
Token | and |
Token ID | 290 |
Feature activation | +7.36e-01 |
Pos logit contributions | |
cream | +0.110 |
Blue | +0.104 |
Fin | +0.099 |
They | +0.098 |
aw | +0.098 |
Neg logit contributions | |
otion | -0.164 |
spine | -0.156 |
Deal | -0.152 |
upon | -0.150 |
laptop | -0.150 |
Token | " |
Token ID | 366 |
Feature activation | +3.53e+00 |
Pos logit contributions | |
otion | +0.045 |
upon | +0.042 |
spine | +0.041 |
ancer | +0.040 |
Deal | +0.038 |
Neg logit contributions | |
cream | -0.038 |
Blue | -0.038 |
sav | -0.038 |
Fin | -0.037 |
Red | -0.036 |
Token | p |
Token ID | 79 |
Feature activation | -5.73e+00 |
Pos logit contributions | |
otion | +0.269 |
upon | +0.264 |
laptop | +0.255 |
ancer | +0.251 |
Deal | +0.247 |
Neg logit contributions | |
cream | -0.132 |
Blue | -0.117 |
Fin | -0.116 |
sav | -0.107 |
Jerry | -0.104 |
Token | go |
Token ID | 467 |
Feature activation | +8.53e-01 |
Pos logit contributions | |
otion | +0.082 |
laptop | +0.077 |
spine | +0.077 |
upon | +0.077 |
Deal | +0.074 |
Neg logit contributions | |
cream | -0.064 |
Blue | -0.062 |
Fin | -0.061 |
aw | -0.057 |
Red | -0.057 |
Token | !" |
Token ID | 2474 |
Feature activation | -2.44e+00 |
Pos logit contributions | |
upon | +0.053 |
otion | +0.049 |
laptop | +0.048 |
spine | +0.048 |
ancer | +0.047 |
Neg logit contributions | |
cream | -0.047 |
Blue | -0.044 |
Fin | -0.042 |
Red | -0.041 |
Jerry | -0.041 |
Token |
|
Token ID | 198 |
Feature activation | -1.39e+00 |
Pos logit contributions | |
cream | +0.099 |
Blue | +0.092 |
Fin | +0.091 |
Red | +0.084 |
aw | +0.082 |
Neg logit contributions | |
otion | -0.182 |
upon | -0.178 |
spine | -0.174 |
laptop | -0.173 |
Deal | -0.169 |
Token |
|
Token ID | 198 |
Feature activation | -1.42e+00 |
Pos logit contributions | |
cream | +0.066 |
Fin | +0.053 |
Blue | +0.053 |
sold | +0.053 |
sailed | +0.052 |
Neg logit contributions | |
upon | -0.099 |
otion | -0.094 |
laptop | -0.089 |
Deal | -0.088 |
Lee | -0.086 |
Token | Jim |
Token ID | 18050 |
Feature activation | -1.71e+00 |
Pos logit contributions | |
cream | +0.053 |
sailed | +0.035 |
sold | +0.035 |
amine | +0.033 |
aw | +0.033 |
Neg logit contributions | |
upon | -0.114 |
otion | -0.108 |
laptop | -0.106 |
Deal | -0.106 |
acy | -0.102 |
Token | bo |
Token ID | 2127 |
Feature activation | -1.29e+01 |
Pos logit contributions | |
cream | +0.080 |
Blue | +0.077 |
Fin | +0.075 |
Red | +0.072 |
Tweet | +0.069 |
Neg logit contributions | |
upon | -0.115 |
otion | -0.114 |
spine | -0.109 |
laptop | -0.109 |
sister | -0.105 |
Token | happily |
Token ID | 18177 |
Feature activation | -1.34e+00 |
Pos logit contributions | |
cream | +0.547 |
Blue | +0.529 |
Fin | +0.511 |
Red | +0.490 |
Tweet | +0.459 |
Neg logit contributions | |
upon | -0.932 |
otion | -0.915 |
laptop | -0.882 |
spine | -0.875 |
sister | -0.853 |
Token | took |
Token ID | 1718 |
Feature activation | -8.50e-01 |
Pos logit contributions | |
Blue | +0.061 |
cream | +0.061 |
Red | +0.059 |
Fin | +0.058 |
Tweet | +0.054 |
Neg logit contributions | |
upon | -0.088 |
otion | -0.085 |
spine | -0.083 |
laptop | -0.082 |
sister | -0.082 |
Token | the |
Token ID | 262 |
Feature activation | -2.49e-01 |
Pos logit contributions | |
cream | +0.043 |
Blue | +0.041 |
Fin | +0.041 |
Red | +0.039 |
Tweet | +0.038 |
Neg logit contributions | |
otion | -0.054 |
upon | -0.052 |
spine | -0.050 |
laptop | -0.050 |
Deal | -0.050 |
Token | battery |
Token ID | 6555 |
Feature activation | -2.81e+00 |
Pos logit contributions | |
cream | +0.010 |
Blue | +0.010 |
Fin | +0.010 |
sailed | +0.009 |
Red | +0.009 |
Neg logit contributions | |
otion | -0.019 |
upon | -0.018 |
Deal | -0.018 |
spine | -0.017 |
ids | -0.017 |
Token | and |
Token ID | 290 |
Feature activation | +7.26e-01 |
Pos logit contributions | |
cream | +0.138 |
Blue | +0.126 |
Fin | +0.124 |
aw | +0.116 |
Red | +0.115 |
Neg logit contributions | |
otion | -0.191 |
upon | -0.178 |
Deal | -0.177 |
spine | -0.175 |
laptop | -0.174 |
Token |
|
Token ID | 198 |
Feature activation | -1.44e+00 |
Pos logit contributions | |
cream | +0.072 |
sold | +0.058 |
sailed | +0.058 |
Blue | +0.057 |
Fin | +0.056 |
Neg logit contributions | |
upon | -0.095 |
otion | -0.091 |
Lee | -0.086 |
Deal | -0.086 |
laptop | -0.085 |
Token | Pick |
Token ID | 31686 |
Feature activation | -6.98e+00 |
Pos logit contributions | |
cream | +0.062 |
sailed | +0.048 |
sold | +0.046 |
nests | +0.042 |
aw | +0.042 |
Neg logit contributions | |
upon | -0.097 |
Lee | -0.092 |
Deal | -0.091 |
laptop | -0.091 |
otion | -0.091 |
Token | - |
Token ID | 12 |
Feature activation | -3.15e+00 |
Pos logit contributions | |
cream | +0.375 |
Fin | +0.373 |
Blue | +0.373 |
Red | +0.346 |
Tweet | +0.339 |
Neg logit contributions | |
upon | -0.413 |
otion | -0.390 |
laptop | -0.357 |
sister | -0.354 |
spine | -0.353 |
Token | a |
Token ID | 64 |
Feature activation | +3.38e+00 |
Pos logit contributions | |
Blue | +0.147 |
Fin | +0.142 |
Red | +0.141 |
cream | +0.138 |
Yo | +0.135 |
Neg logit contributions | |
upon | -0.205 |
otion | -0.177 |
acy | -0.169 |
Lee | -0.167 |
ancer | -0.165 |
Token | - |
Token ID | 12 |
Feature activation | -3.15e+00 |
Pos logit contributions | |
upon | +0.237 |
otion | +0.230 |
sister | +0.221 |
laptop | +0.219 |
spine | +0.217 |
Neg logit contributions | |
cream | -0.149 |
Fin | -0.142 |
Blue | -0.140 |
Red | -0.133 |
Jerry | -0.124 |
Token | bo |
Token ID | 2127 |
Feature activation | -1.29e+01 |
Pos logit contributions | |
Blue | +0.146 |
Fin | +0.141 |
Red | +0.140 |
sav | +0.138 |
cream | +0.137 |
Neg logit contributions | |
upon | -0.202 |
otion | -0.168 |
Lee | -0.165 |
acy | -0.162 |
ids | -0.160 |
Token | o |
Token ID | 78 |
Feature activation | -6.08e+00 |
Pos logit contributions | |
cream | +0.538 |
Blue | +0.517 |
Fin | +0.511 |
Red | +0.493 |
Tweet | +0.452 |
Neg logit contributions | |
upon | -0.949 |
otion | -0.922 |
spine | -0.872 |
laptop | -0.862 |
sister | -0.851 |
Token | smiled |
Token ID | 13541 |
Feature activation | +1.24e+00 |
Pos logit contributions | |
cream | +0.277 |
Fin | +0.272 |
Blue | +0.271 |
Red | +0.259 |
Jerry | +0.239 |
Neg logit contributions | |
upon | -0.413 |
otion | -0.394 |
sister | -0.378 |
spine | -0.376 |
laptop | -0.375 |
Token | and |
Token ID | 290 |
Feature activation | +7.90e-01 |
Pos logit contributions | |
upon | +0.074 |
otion | +0.072 |
spine | +0.070 |
laptop | +0.067 |
Deal | +0.067 |
Neg logit contributions | |
cream | -0.070 |
Blue | -0.067 |
Fin | -0.067 |
Red | -0.063 |
Tweet | -0.061 |
Token | replied |
Token ID | 8712 |
Feature activation | -2.69e+00 |
Pos logit contributions | |
upon | +0.047 |
sister | +0.045 |
Lee | +0.045 |
otion | +0.045 |
spine | +0.043 |
Neg logit contributions | |
cream | -0.039 |
Blue | -0.039 |
Red | -0.038 |
Fin | -0.038 |
One | -0.035 |
Token | , |
Token ID | 11 |
Feature activation | -5.05e-01 |
Pos logit contributions | |
cream | +0.144 |
Blue | +0.138 |
Fin | +0.135 |
aw | +0.130 |
Tweet | +0.126 |
Neg logit contributions | |
otion | -0.162 |
laptop | -0.158 |
spine | -0.158 |
upon | -0.156 |
Deal | -0.153 |
Token | a |
Token ID | 257 |
Feature activation | +1.88e+00 |
Pos logit contributions | |
otion | +0.070 |
upon | +0.069 |
laptop | +0.067 |
spine | +0.066 |
Deal | +0.064 |
Neg logit contributions | |
cream | -0.053 |
Blue | -0.050 |
Fin | -0.049 |
Red | -0.046 |
Tweet | -0.046 |
Token | problem |
Token ID | 1917 |
Feature activation | -8.13e-01 |
Pos logit contributions | |
upon | +0.084 |
otion | +0.083 |
laptop | +0.080 |
spine | +0.078 |
sister | +0.074 |
Neg logit contributions | |
cream | -0.135 |
Fin | -0.125 |
Blue | -0.125 |
aw | -0.120 |
Red | -0.119 |
Token | . |
Token ID | 13 |
Feature activation | -1.14e-01 |
Pos logit contributions | |
cream | +0.038 |
Blue | +0.035 |
Fin | +0.035 |
aw | +0.033 |
Red | +0.032 |
Neg logit contributions | |
otion | -0.056 |
upon | -0.054 |
spine | -0.053 |
laptop | -0.052 |
Deal | -0.052 |
Token | So |
Token ID | 1406 |
Feature activation | -2.59e+00 |
Pos logit contributions | |
cream | +0.005 |
Blue | +0.005 |
Fin | +0.005 |
Red | +0.005 |
Tweet | +0.004 |
Neg logit contributions | |
otion | -0.008 |
upon | -0.008 |
Deal | -0.008 |
spine | -0.008 |
laptop | -0.007 |
Token | Jim |
Token ID | 5395 |
Feature activation | +1.13e+00 |
Pos logit contributions | |
cream | +0.139 |
Blue | +0.131 |
Fin | +0.130 |
Red | +0.122 |
aw | +0.121 |
Neg logit contributions | |
otion | -0.162 |
upon | -0.154 |
spine | -0.151 |
laptop | -0.151 |
Deal | -0.146 |
Token | bo |
Token ID | 2127 |
Feature activation | -1.29e+01 |
Pos logit contributions | |
upon | +0.079 |
otion | +0.075 |
sister | +0.074 |
spine | +0.073 |
laptop | +0.072 |
Neg logit contributions | |
cream | -0.050 |
Blue | -0.048 |
Red | -0.047 |
Fin | -0.047 |
Tweet | -0.042 |
Token | got |
Token ID | 1392 |
Feature activation | +9.23e-02 |
Pos logit contributions | |
cream | +0.595 |
Red | +0.590 |
Blue | +0.586 |
Fin | +0.569 |
Twe | +0.509 |
Neg logit contributions | |
upon | -0.851 |
sister | -0.794 |
otion | -0.783 |
Lee | -0.755 |
ids | -0.754 |
Token | an |
Token ID | 281 |
Feature activation | +1.94e+00 |
Pos logit contributions | |
upon | +0.006 |
otion | +0.006 |
sister | +0.005 |
laptop | +0.005 |
Deal | +0.005 |
Neg logit contributions | |
Blue | -0.005 |
cream | -0.005 |
Fin | -0.005 |
Red | -0.004 |
Tweet | -0.004 |
Token | idea |
Token ID | 2126 |
Feature activation | +2.86e+00 |
Pos logit contributions | |
otion | +0.100 |
upon | +0.097 |
spine | +0.092 |
ancer | +0.090 |
laptop | +0.089 |
Neg logit contributions | |
cream | -0.126 |
Blue | -0.117 |
Fin | -0.115 |
aw | -0.113 |
Red | -0.112 |
Token | . |
Token ID | 13 |
Feature activation | -8.52e-02 |
Pos logit contributions | |
otion | +0.208 |
Deal | +0.199 |
laptop | +0.198 |
spine | +0.198 |
upon | +0.196 |
Neg logit contributions | |
cream | -0.123 |
Blue | -0.110 |
Fin | -0.109 |
never | -0.103 |
aw | -0.103 |
Token | Jim |
Token ID | 5395 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
cream | +0.003 |
Blue | +0.003 |
Fin | +0.003 |
Red | +0.003 |
Tweet | +0.003 |
Neg logit contributions | |
otion | -0.007 |
upon | -0.007 |
Deal | -0.006 |
spine | -0.006 |
laptop | -0.006 |
Token | got |
Token ID | 1392 |
Feature activation | +5.20e-02 |
Pos logit contributions | |
cream | +0.021 |
Blue | +0.020 |
Fin | +0.020 |
Red | +0.019 |
aw | +0.018 |
Neg logit contributions | |
otion | -0.027 |
upon | -0.026 |
spine | -0.026 |
laptop | -0.025 |
Deal | -0.025 |
Token | a |
Token ID | 257 |
Feature activation | +1.88e+00 |
Pos logit contributions | |
otion | +0.003 |
spine | +0.003 |
upon | +0.003 |
laptop | +0.003 |
Deal | +0.003 |
Neg logit contributions | |
cream | -0.003 |
Blue | -0.003 |
Fin | -0.003 |
Red | -0.002 |
aw | -0.002 |
Token | bad |
Token ID | 2089 |
Feature activation | -3.97e-01 |
Pos logit contributions | |
otion | +0.145 |
ancer | +0.129 |
ids | +0.128 |
Deal | +0.127 |
Lee | +0.126 |
Neg logit contributions | |
Blue | -0.077 |
Fin | -0.073 |
cream | -0.072 |
Red | -0.069 |
The | -0.068 |
Token | boo |
Token ID | 22513 |
Feature activation | +1.88e+00 |
Pos logit contributions | |
Fin | +0.019 |
cream | +0.018 |
sav | +0.018 |
Blue | +0.018 |
that | +0.018 |
Neg logit contributions | |
otion | -0.025 |
Deal | -0.025 |
upon | -0.024 |
ancer | -0.023 |
laptop | -0.023 |
Token | - |
Token ID | 12 |
Feature activation | -3.15e+00 |
Pos logit contributions | |
otion | +0.139 |
upon | +0.134 |
spine | +0.131 |
laptop | +0.130 |
Deal | +0.127 |
Neg logit contributions | |
cream | -0.080 |
Blue | -0.075 |
Fin | -0.073 |
Red | -0.068 |
aw | -0.067 |
Token | bo |
Token ID | 2127 |
Feature activation | -1.29e+01 |
Pos logit contributions | |
sav | +0.111 |
Fin | +0.111 |
Blue | +0.107 |
sailed | +0.107 |
adventures | +0.105 |
Neg logit contributions | |
otion | -0.217 |
upon | -0.209 |
ancer | -0.194 |
aughter | -0.193 |
spine | -0.193 |
Token | o |
Token ID | 78 |
Feature activation | -6.08e+00 |
Pos logit contributions | |
cream | +0.352 |
aw | +0.271 |
Fin | +0.254 |
Blue | +0.251 |
bo | +0.243 |
Neg logit contributions | |
otion | -1.176 |
Deal | -1.168 |
laptop | -1.140 |
spine | -1.133 |
upon | -1.103 |
Token | on |
Token ID | 319 |
Feature activation | +2.34e+00 |
Pos logit contributions | |
cream | +0.260 |
aw | +0.230 |
never | +0.225 |
bo | +0.224 |
Blue | +0.223 |
Neg logit contributions | |
Deal | -0.452 |
otion | -0.450 |
laptop | -0.435 |
upon | -0.429 |
ancer | -0.424 |
Token | her |
Token ID | 607 |
Feature activation | +3.29e+00 |
Pos logit contributions | |
ancer | +0.162 |
otion | +0.161 |
upon | +0.158 |
spine | +0.157 |
laptop | +0.154 |
Neg logit contributions | |
The | -0.104 |
Blue | -0.104 |
This | -0.101 |
sav | -0.100 |
never | -0.099 |
Token | knee |
Token ID | 10329 |
Feature activation | +2.28e+00 |
Pos logit contributions | |
otion | +0.167 |
upon | +0.149 |
ancer | +0.144 |
Deal | +0.135 |
ids | +0.128 |
Neg logit contributions | |
cream | -0.228 |
Blue | -0.212 |
sav | -0.210 |
Fin | -0.205 |
The | -0.204 |
Token | . |
Token ID | 13 |
Feature activation | -7.50e-02 |
Pos logit contributions | |
otion | +0.135 |
spine | +0.132 |
upon | +0.131 |
laptop | +0.129 |
Lee | +0.121 |
Neg logit contributions | |
cream | -0.124 |
Fin | -0.121 |
Blue | -0.119 |
Red | -0.112 |
Tweet | -0.109 |
Token | she |
Token ID | 673 |
Feature activation | +2.97e+00 |
Pos logit contributions | |
otion | +0.283 |
spine | +0.278 |
Deal | +0.276 |
laptop | +0.270 |
upon | +0.261 |
Neg logit contributions | |
cream | -0.207 |
Blue | -0.193 |
Fin | -0.191 |
never | -0.190 |
sav | -0.186 |
Token | got |
Token ID | 1392 |
Feature activation | +5.20e-02 |
Pos logit contributions | |
otion | +0.225 |
upon | +0.217 |
spine | +0.213 |
laptop | +0.212 |
Deal | +0.207 |
Neg logit contributions | |
cream | -0.121 |
Blue | -0.111 |
Fin | -0.109 |
Red | -0.101 |
aw | -0.100 |
Token | a |
Token ID | 257 |
Feature activation | +1.88e+00 |
Pos logit contributions | |
upon | +0.003 |
otion | +0.003 |
spine | +0.003 |
laptop | +0.003 |
Deal | +0.003 |
Neg logit contributions | |
cream | -0.003 |
Blue | -0.003 |
Fin | -0.003 |
Red | -0.002 |
Tweet | -0.002 |
Token | boo |
Token ID | 22513 |
Feature activation | +1.89e+00 |
Pos logit contributions | |
otion | +0.130 |
spine | +0.116 |
Deal | +0.116 |
laptop | +0.115 |
upon | +0.114 |
Neg logit contributions | |
Blue | -0.087 |
cream | -0.086 |
Fin | -0.082 |
sav | -0.080 |
Red | -0.079 |
Token | - |
Token ID | 12 |
Feature activation | -3.15e+00 |
Pos logit contributions | |
otion | +0.145 |
upon | +0.140 |
spine | +0.137 |
laptop | +0.136 |
Deal | +0.135 |
Neg logit contributions | |
cream | -0.076 |
Blue | -0.068 |
Fin | -0.067 |
aw | -0.063 |
Red | -0.062 |
Token | bo |
Token ID | 2127 |
Feature activation | -1.29e+01 |
Pos logit contributions | |
Fin | +0.107 |
Blue | +0.105 |
cream | +0.102 |
sav | +0.101 |
sailed | +0.098 |
Neg logit contributions | |
otion | -0.237 |
upon | -0.231 |
ancer | -0.218 |
laptop | -0.215 |
spine | -0.214 |
Token | o |
Token ID | 78 |
Feature activation | -6.11e+00 |
Pos logit contributions | |
cream | +0.447 |
aw | +0.347 |
bo | +0.347 |
Blue | +0.339 |
Fin | +0.337 |
Neg logit contributions | |
otion | -1.091 |
Deal | -1.060 |
spine | -1.046 |
laptop | -1.041 |
upon | -0.995 |
Token | . |
Token ID | 13 |
Feature activation | -9.58e-02 |
Pos logit contributions | |
cream | +0.287 |
Blue | +0.247 |
bo | +0.244 |
Fin | +0.238 |
aw | +0.236 |
Neg logit contributions | |
otion | -0.437 |
Deal | -0.418 |
spine | -0.415 |
laptop | -0.409 |
upon | -0.405 |
Token | The |
Token ID | 383 |
Feature activation | -1.56e+00 |
Pos logit contributions | |
cream | +0.004 |
Fin | +0.004 |
Blue | +0.004 |
aw | +0.004 |
Red | +0.004 |
Neg logit contributions | |
otion | -0.007 |
upon | -0.007 |
spine | -0.006 |
Deal | -0.006 |
laptop | -0.006 |
Token | end |
Token ID | 886 |
Feature activation | +2.37e+00 |
Pos logit contributions | |
cream | +0.081 |
Blue | +0.079 |
Fin | +0.077 |
Red | +0.073 |
Tweet | +0.073 |
Neg logit contributions | |
otion | -0.100 |
upon | -0.093 |
spine | -0.093 |
laptop | -0.092 |
Deal | -0.089 |
Token | . |
Token ID | 13 |
Feature activation | -8.52e-02 |
Pos logit contributions | |
otion | +0.164 |
upon | +0.156 |
spine | +0.154 |
laptop | +0.153 |
Deal | +0.149 |
Neg logit contributions | |
cream | -0.112 |
Blue | -0.105 |
Fin | -0.103 |
Red | -0.097 |
aw | -0.096 |
Token | had |
Token ID | 550 |
Feature activation | -1.32e+00 |
Pos logit contributions | |
Fin | +0.107 |
cream | +0.106 |
Blue | +0.104 |
Tweet | +0.098 |
Red | +0.096 |
Neg logit contributions | |
upon | -0.151 |
otion | -0.147 |
sister | -0.142 |
laptop | -0.140 |
ancer | -0.135 |
Token | a |
Token ID | 257 |
Feature activation | +1.87e+00 |
Pos logit contributions | |
cream | +0.062 |
Blue | +0.059 |
Fin | +0.058 |
Red | +0.055 |
Tweet | +0.054 |
Neg logit contributions | |
otion | -0.089 |
upon | -0.088 |
spine | -0.085 |
laptop | -0.085 |
Deal | -0.082 |
Token | small |
Token ID | 1402 |
Feature activation | -3.17e-01 |
Pos logit contributions | |
otion | +0.123 |
upon | +0.121 |
laptop | +0.120 |
spine | +0.119 |
Deal | +0.115 |
Neg logit contributions | |
cream | -0.092 |
Blue | -0.085 |
Fin | -0.085 |
aw | -0.079 |
Red | -0.079 |
Token | boo |
Token ID | 22513 |
Feature activation | +1.87e+00 |
Pos logit contributions | |
cream | +0.017 |
Blue | +0.016 |
sav | +0.016 |
oc | +0.016 |
Fin | +0.015 |
Neg logit contributions | |
otion | -0.020 |
upon | -0.018 |
ancer | -0.018 |
Deal | -0.018 |
spine | -0.017 |
Token | - |
Token ID | 12 |
Feature activation | -3.17e+00 |
Pos logit contributions | |
otion | +0.144 |
upon | +0.138 |
spine | +0.135 |
laptop | +0.134 |
Deal | +0.133 |
Neg logit contributions | |
cream | -0.076 |
Blue | -0.067 |
Fin | -0.066 |
Red | -0.061 |
aw | -0.061 |
Token | bo |
Token ID | 2127 |
Feature activation | -1.29e+01 |
Pos logit contributions | |
Fin | +0.103 |
Blue | +0.099 |
sav | +0.094 |
cream | +0.092 |
Yo | +0.090 |
Neg logit contributions | |
otion | -0.244 |
upon | -0.239 |
ancer | -0.226 |
laptop | -0.226 |
spine | -0.224 |
Token | o |
Token ID | 78 |
Feature activation | -6.11e+00 |
Pos logit contributions | |
cream | +0.461 |
Blue | +0.363 |
Fin | +0.357 |
aw | +0.353 |
bo | +0.334 |
Neg logit contributions | |
otion | -1.064 |
Deal | -1.040 |
spine | -1.038 |
laptop | -1.014 |
upon | -1.003 |
Token | that |
Token ID | 326 |
Feature activation | -9.71e+00 |
Pos logit contributions | |
cream | +0.285 |
Blue | +0.237 |
bo | +0.235 |
Fin | +0.228 |
aw | +0.226 |
Neg logit contributions | |
otion | -0.445 |
Deal | -0.431 |
spine | -0.421 |
upon | -0.414 |
laptop | -0.407 |
Token | would |
Token ID | 561 |
Feature activation | -9.10e-01 |
Pos logit contributions | |
cream | +0.366 |
Blue | +0.339 |
Fin | +0.333 |
Red | +0.306 |
aw | +0.301 |
Neg logit contributions | |
otion | -0.756 |
upon | -0.732 |
spine | -0.717 |
laptop | -0.716 |
Deal | -0.695 |
Token | heal |
Token ID | 12035 |
Feature activation | -1.78e+00 |
Pos logit contributions | |
cream | +0.049 |
Blue | +0.045 |
Fin | +0.044 |
aw | +0.043 |
Red | +0.042 |
Neg logit contributions | |
otion | -0.057 |
upon | -0.056 |
spine | -0.054 |
laptop | -0.054 |
Deal | -0.052 |
Token | soon |
Token ID | 2582 |
Feature activation | -3.20e+00 |
Pos logit contributions | |
cream | +0.086 |
Blue | +0.080 |
Fin | +0.079 |
Red | +0.073 |
aw | +0.073 |
Neg logit contributions | |
otion | -0.119 |
upon | -0.116 |
spine | -0.113 |
laptop | -0.113 |
Deal | -0.109 |
Token | his |
Token ID | 465 |
Feature activation | -7.55e-01 |
Pos logit contributions | |
cream | +0.064 |
Blue | +0.059 |
Fin | +0.058 |
Red | +0.054 |
aw | +0.052 |
Neg logit contributions | |
otion | -0.101 |
upon | -0.098 |
laptop | -0.095 |
spine | -0.095 |
Deal | -0.092 |
Token | mouth |
Token ID | 5422 |
Feature activation | +4.93e+00 |
Pos logit contributions | |
cream | +0.038 |
Blue | +0.036 |
Fin | +0.035 |
Red | +0.034 |
sailed | +0.034 |
Neg logit contributions | |
otion | -0.051 |
upon | -0.046 |
Deal | -0.044 |
spine | -0.044 |
ancer | -0.044 |
Token | and |
Token ID | 290 |
Feature activation | +7.90e-01 |
Pos logit contributions | |
otion | +0.304 |
upon | +0.298 |
laptop | +0.291 |
spine | +0.290 |
Deal | +0.271 |
Neg logit contributions | |
cream | -0.262 |
Blue | -0.251 |
Fin | -0.250 |
Red | -0.234 |
Tweet | -0.228 |
Token | says |
Token ID | 1139 |
Feature activation | -5.73e-01 |
Pos logit contributions | |
otion | +0.048 |
upon | +0.047 |
laptop | +0.046 |
spine | +0.046 |
Deal | +0.044 |
Neg logit contributions | |
cream | -0.043 |
Blue | -0.041 |
Fin | -0.040 |
Red | -0.038 |
aw | -0.037 |
Token | " |
Token ID | 366 |
Feature activation | +3.61e+00 |
Pos logit contributions | |
cream | +0.032 |
Blue | +0.030 |
Fin | +0.030 |
aw | +0.029 |
Red | +0.028 |
Neg logit contributions | |
otion | -0.034 |
spine | -0.033 |
upon | -0.032 |
laptop | -0.032 |
Deal | -0.031 |
Token | v |
Token ID | 85 |
Feature activation | -1.29e+01 |
Pos logit contributions | |
ancer | +0.251 |
upon | +0.249 |
otion | +0.238 |
laptop | +0.237 |
grade | +0.231 |
Neg logit contributions | |
cream | -0.136 |
Jerry | -0.120 |
Fin | -0.120 |
sav | -0.119 |
never | -0.115 |
Token | room |
Token ID | 3823 |
Feature activation | +3.11e+00 |
Pos logit contributions | |
cream | +0.665 |
Fin | +0.621 |
aw | +0.618 |
Blue | +0.603 |
oc | +0.593 |
Neg logit contributions | |
spine | -0.808 |
otion | -0.794 |
upon | -0.768 |
Deal | -0.740 |
ancer | -0.735 |
Token | , |
Token ID | 11 |
Feature activation | -4.94e-01 |
Pos logit contributions | |
otion | +0.199 |
laptop | +0.190 |
upon | +0.189 |
spine | +0.183 |
sister | +0.179 |
Neg logit contributions | |
cream | -0.158 |
Blue | -0.152 |
Fin | -0.148 |
Red | -0.140 |
Tweet | -0.137 |
Token | v |
Token ID | 410 |
Feature activation | -2.36e+00 |
Pos logit contributions | |
cream | +0.025 |
Blue | +0.024 |
Fin | +0.023 |
Red | +0.022 |
aw | +0.022 |
Neg logit contributions | |
otion | -0.032 |
upon | -0.031 |
spine | -0.030 |
laptop | -0.029 |
Deal | -0.029 |
Token | room |
Token ID | 3823 |
Feature activation | +3.13e+00 |
Pos logit contributions | |
cream | +0.160 |
Blue | +0.156 |
Fin | +0.154 |
Red | +0.146 |
sav | +0.146 |
Neg logit contributions | |
otion | -0.111 |
laptop | -0.104 |
upon | -0.101 |
spine | -0.096 |
Deal | -0.095 |
Token | ". |
Token ID | 1911 |
Feature activation | +1.10e+00 |
Pos logit contributions | |
otion | +0.191 |
upon | +0.181 |
laptop | +0.181 |
spine | +0.175 |
sister | +0.170 |
Neg logit contributions | |
cream | -0.169 |
Blue | -0.161 |
Fin | -0.158 |
Red | -0.150 |
Tweet | -0.146 |
Token | and |
Token ID | 290 |
Feature activation | +7.69e-01 |
Pos logit contributions | |
otion | +0.140 |
spine | +0.135 |
Deal | +0.135 |
laptop | +0.132 |
ancer | +0.131 |
Neg logit contributions | |
cream | -0.074 |
Blue | -0.068 |
The | -0.064 |
never | -0.064 |
sav | -0.063 |
Token | got |
Token ID | 1392 |
Feature activation | +6.81e-02 |
Pos logit contributions | |
otion | +0.058 |
laptop | +0.054 |
Deal | +0.054 |
spine | +0.054 |
upon | +0.054 |
Neg logit contributions | |
cream | -0.033 |
Blue | -0.030 |
Fin | -0.029 |
Tweet | -0.027 |
never | -0.027 |
Token | a |
Token ID | 257 |
Feature activation | +1.89e+00 |
Pos logit contributions | |
otion | +0.005 |
spine | +0.004 |
upon | +0.004 |
laptop | +0.004 |
Deal | +0.004 |
Neg logit contributions | |
cream | -0.003 |
Blue | -0.003 |
Fin | -0.003 |
Red | -0.003 |
aw | -0.003 |
Token | boo |
Token ID | 22513 |
Feature activation | +1.88e+00 |
Pos logit contributions | |
otion | +0.140 |
Deal | +0.128 |
ids | +0.125 |
ancer | +0.124 |
Lee | +0.123 |
Neg logit contributions | |
Blue | -0.081 |
cream | -0.079 |
Fin | -0.074 |
Red | -0.072 |
sav | -0.071 |
Token | - |
Token ID | 12 |
Feature activation | -3.13e+00 |
Pos logit contributions | |
otion | +0.139 |
upon | +0.134 |
spine | +0.131 |
laptop | +0.131 |
Deal | +0.128 |
Neg logit contributions | |
cream | -0.080 |
Blue | -0.074 |
Fin | -0.073 |
Red | -0.067 |
aw | -0.067 |
Token | bo |
Token ID | 2127 |
Feature activation | -1.29e+01 |
Pos logit contributions | |
Fin | +0.078 |
Blue | +0.068 |
sav | +0.065 |
Jerry | +0.062 |
sailed | +0.061 |
Neg logit contributions | |
otion | -0.263 |
upon | -0.256 |
spine | -0.242 |
ancer | -0.241 |
headache | -0.237 |
Token | o |
Token ID | 78 |
Feature activation | -6.08e+00 |
Pos logit contributions | |
cream | +0.353 |
Blue | +0.247 |
aw | +0.244 |
bo | +0.240 |
Fin | +0.238 |
Neg logit contributions | |
Deal | -1.187 |
otion | -1.183 |
laptop | -1.147 |
spine | -1.143 |
upon | -1.111 |
Token | on |
Token ID | 319 |
Feature activation | +2.36e+00 |
Pos logit contributions | |
cream | +0.258 |
bo | +0.229 |
Blue | +0.211 |
never | +0.205 |
aw | +0.204 |
Neg logit contributions | |
Deal | -0.474 |
otion | -0.468 |
laptop | -0.443 |
upon | -0.441 |
spine | -0.434 |
Token | her |
Token ID | 607 |
Feature activation | +3.31e+00 |
Pos logit contributions | |
otion | +0.116 |
upon | +0.110 |
spine | +0.107 |
laptop | +0.105 |
Deal | +0.102 |
Neg logit contributions | |
cream | -0.159 |
Blue | -0.152 |
Fin | -0.150 |
Red | -0.144 |
aw | -0.142 |
Token | knee |
Token ID | 10329 |
Feature activation | +2.29e+00 |
Pos logit contributions | |
otion | +0.147 |
upon | +0.131 |
ancer | +0.121 |
Deal | +0.120 |
laptop | +0.116 |
Neg logit contributions | |
cream | -0.249 |
Blue | -0.234 |
Fin | -0.226 |
sav | -0.225 |
Red | -0.224 |
Token | . |
Token ID | 13 |
Feature activation | -6.06e-02 |
Pos logit contributions | |
otion | +0.137 |
spine | +0.133 |
upon | +0.132 |
laptop | +0.131 |
sister | +0.121 |
Neg logit contributions | |
cream | -0.124 |
Fin | -0.121 |
Blue | -0.119 |
Red | -0.112 |
aw | -0.108 |
Token | chest |
Token ID | 7721 |
Feature activation | +1.91e+00 |
Pos logit contributions | |
cream | +0.055 |
Blue | +0.052 |
sav | +0.052 |
Fin | +0.051 |
aw | +0.050 |
Neg logit contributions | |
otion | -0.038 |
ancer | -0.031 |
upon | -0.030 |
laptop | -0.028 |
Deal | -0.028 |
Token | . |
Token ID | 13 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
otion | +0.136 |
upon | +0.127 |
Deal | +0.126 |
laptop | +0.126 |
spine | +0.124 |
Neg logit contributions | |
cream | -0.087 |
Blue | -0.079 |
Fin | -0.077 |
aw | -0.074 |
sav | -0.072 |
Token |
|
Token ID | 198 |
Feature activation | -1.42e+00 |
Pos logit contributions | |
cream | +0.006 |
Blue | +0.005 |
Fin | +0.005 |
Red | +0.005 |
Tweet | +0.005 |
Neg logit contributions | |
otion | -0.008 |
upon | -0.008 |
Deal | -0.008 |
spine | -0.008 |
laptop | -0.008 |
Token |
|
Token ID | 198 |
Feature activation | -1.40e+00 |
Pos logit contributions | |
cream | +0.068 |
Blue | +0.059 |
Fin | +0.057 |
Red | +0.055 |
sold | +0.054 |
Neg logit contributions | |
upon | -0.101 |
otion | -0.099 |
Deal | -0.092 |
laptop | -0.089 |
acy | -0.087 |
Token | Jim |
Token ID | 18050 |
Feature activation | -1.68e+00 |
Pos logit contributions | |
cream | +0.078 |
Blue | +0.064 |
rob | +0.064 |
Red | +0.064 |
sailed | +0.061 |
Neg logit contributions | |
upon | -0.089 |
Deal | -0.084 |
otion | -0.084 |
acy | -0.075 |
laptop | -0.074 |
Token | bo |
Token ID | 2127 |
Feature activation | -1.29e+01 |
Pos logit contributions | |
Blue | +0.078 |
cream | +0.078 |
Fin | +0.076 |
Red | +0.074 |
Jerry | +0.067 |
Neg logit contributions | |
upon | -0.116 |
sister | -0.105 |
otion | -0.104 |
spine | -0.103 |
laptop | -0.101 |
Token | began |
Token ID | 2540 |
Feature activation | +2.28e+00 |
Pos logit contributions | |
cream | +0.589 |
Fin | +0.579 |
Blue | +0.569 |
Red | +0.544 |
Jerry | +0.509 |
Neg logit contributions | |
upon | -0.881 |
otion | -0.805 |
spine | -0.790 |
sister | -0.782 |
laptop | -0.777 |
Token | to |
Token ID | 284 |
Feature activation | -1.29e+00 |
Pos logit contributions | |
otion | +0.184 |
Deal | +0.173 |
ancer | +0.171 |
spine | +0.170 |
laptop | +0.166 |
Neg logit contributions | |
cream | -0.084 |
sav | -0.075 |
Blue | -0.072 |
never | -0.071 |
Fin | -0.071 |
Token | wander |
Token ID | 27776 |
Feature activation | +1.37e+00 |
Pos logit contributions | |
cream | +0.070 |
Blue | +0.067 |
Fin | +0.065 |
sav | +0.064 |
Red | +0.062 |
Neg logit contributions | |
otion | -0.082 |
upon | -0.075 |
spine | -0.074 |
ancer | -0.073 |
laptop | -0.073 |
Token | through |
Token ID | 832 |
Feature activation | +1.60e+00 |
Pos logit contributions | |
upon | +0.088 |
otion | +0.085 |
spine | +0.083 |
laptop | +0.079 |
ancer | +0.077 |
Neg logit contributions | |
cream | -0.073 |
Blue | -0.068 |
Fin | -0.066 |
Red | -0.064 |
Tweet | -0.061 |
Token | the |
Token ID | 262 |
Feature activation | -1.63e-01 |
Pos logit contributions | |
otion | +0.113 |
upon | +0.108 |
spine | +0.106 |
laptop | +0.105 |
Deal | +0.103 |
Neg logit contributions | |
cream | -0.073 |
Blue | -0.068 |
Fin | -0.067 |
Red | -0.063 |
aw | -0.062 |
Token | play |
Token ID | 711 |
Feature activation | -1.58e+00 |
Pos logit contributions | |
otion | +0.064 |
upon | +0.061 |
spine | +0.060 |
laptop | +0.059 |
Deal | +0.058 |
Neg logit contributions | |
cream | -0.051 |
Blue | -0.047 |
Fin | -0.046 |
Red | -0.044 |
aw | -0.043 |
Token | peek |
Token ID | 27185 |
Feature activation | +2.60e+00 |
Pos logit contributions | |
cream | +0.077 |
Fin | +0.071 |
Blue | +0.070 |
Red | +0.064 |
Jerry | +0.063 |
Neg logit contributions | |
otion | -0.108 |
upon | -0.105 |
laptop | -0.102 |
spine | -0.101 |
Deal | -0.097 |
Token | - |
Token ID | 12 |
Feature activation | -3.05e+00 |
Pos logit contributions | |
otion | +0.184 |
laptop | +0.177 |
Deal | +0.176 |
spine | +0.175 |
upon | +0.165 |
Neg logit contributions | |
cream | -0.120 |
aw | -0.117 |
Blue | -0.106 |
Red | -0.101 |
Fin | -0.101 |
Token | a |
Token ID | 64 |
Feature activation | +3.44e+00 |
Pos logit contributions | |
Fin | +0.126 |
Blue | +0.115 |
cream | +0.111 |
Jerry | +0.108 |
sailed | +0.106 |
Neg logit contributions | |
upon | -0.224 |
otion | -0.213 |
ancer | -0.207 |
spine | -0.201 |
ids | -0.198 |
Token | - |
Token ID | 12 |
Feature activation | -3.11e+00 |
Pos logit contributions | |
otion | +0.260 |
Deal | +0.240 |
spine | +0.236 |
laptop | +0.229 |
upon | +0.225 |
Neg logit contributions | |
aw | -0.150 |
cream | -0.146 |
Blue | -0.131 |
bo | -0.129 |
sav | -0.126 |
Token | bo |
Token ID | 2127 |
Feature activation | -1.29e+01 |
Pos logit contributions | |
Fin | +0.129 |
Blue | +0.118 |
cream | +0.114 |
Jerry | +0.110 |
sailed | +0.109 |
Neg logit contributions | |
upon | -0.227 |
otion | -0.214 |
ancer | -0.208 |
spine | -0.202 |
acy | -0.198 |
Token | o |
Token ID | 78 |
Feature activation | -6.10e+00 |
Pos logit contributions | |
cream | +0.485 |
aw | +0.452 |
bo | +0.411 |
Blue | +0.386 |
They | +0.377 |
Neg logit contributions | |
otion | -1.022 |
Deal | -1.020 |
spine | -0.984 |
laptop | -0.958 |
headache | -0.943 |
Token | . |
Token ID | 13 |
Feature activation | -1.33e-01 |
Pos logit contributions | |
cream | +0.261 |
aw | +0.246 |
bo | +0.233 |
Blue | +0.228 |
They | +0.216 |
Neg logit contributions | |
otion | -0.453 |
spine | -0.433 |
Deal | -0.430 |
laptop | -0.424 |
headache | -0.403 |
Token | Every |
Token ID | 3887 |
Feature activation | -3.76e+00 |
Pos logit contributions | |
cream | +0.006 |
Blue | +0.006 |
Fin | +0.006 |
Red | +0.005 |
aw | +0.005 |
Neg logit contributions | |
otion | -0.009 |
upon | -0.009 |
spine | -0.009 |
laptop | -0.009 |
Deal | -0.009 |
Token | time |
Token ID | 640 |
Feature activation | +5.19e+00 |
Pos logit contributions | |
cream | +0.166 |
Blue | +0.152 |
Fin | +0.152 |
sav | +0.147 |
aw | +0.145 |
Neg logit contributions | |
otion | -0.267 |
spine | -0.253 |
upon | -0.252 |
Deal | -0.249 |
laptop | -0.248 |
Token | she |
Token ID | 673 |
Feature activation | +2.98e+00 |
Pos logit contributions | |
otion | +0.377 |
spine | +0.371 |
Deal | +0.360 |
laptop | +0.354 |
upon | +0.350 |
Neg logit contributions | |
cream | -0.226 |
Blue | -0.210 |
Fin | -0.206 |
aw | -0.193 |
Red | -0.190 |
Token | and |
Token ID | 290 |
Feature activation | +9.87e-01 |
Pos logit contributions | |
cream | +0.183 |
Blue | +0.171 |
Fin | +0.168 |
aw | +0.157 |
Red | +0.157 |
Neg logit contributions | |
otion | -0.276 |
upon | -0.259 |
laptop | -0.259 |
spine | -0.258 |
Deal | -0.252 |
Token | had |
Token ID | 550 |
Feature activation | -1.13e+00 |
Pos logit contributions | |
otion | +0.068 |
laptop | +0.064 |
spine | +0.064 |
upon | +0.064 |
Deal | +0.063 |
Neg logit contributions | |
cream | -0.047 |
Blue | -0.044 |
Fin | -0.043 |
Tweet | -0.041 |
sav | -0.040 |
Token | a |
Token ID | 257 |
Feature activation | +2.02e+00 |
Pos logit contributions | |
cream | +0.061 |
Fin | +0.058 |
Blue | +0.057 |
Red | +0.055 |
aw | +0.053 |
Neg logit contributions | |
upon | -0.070 |
otion | -0.067 |
spine | -0.064 |
laptop | -0.063 |
sister | -0.063 |
Token | boo |
Token ID | 22513 |
Feature activation | +1.99e+00 |
Pos logit contributions | |
otion | +0.131 |
upon | +0.121 |
spine | +0.119 |
laptop | +0.118 |
ancer | +0.117 |
Neg logit contributions | |
cream | -0.103 |
Blue | -0.101 |
Fin | -0.099 |
Red | -0.093 |
Tweet | -0.092 |
Token | - |
Token ID | 12 |
Feature activation | -3.09e+00 |
Pos logit contributions | |
otion | +0.146 |
upon | +0.140 |
spine | +0.138 |
laptop | +0.137 |
Deal | +0.134 |
Neg logit contributions | |
cream | -0.085 |
Blue | -0.079 |
Fin | -0.078 |
Red | -0.072 |
aw | -0.071 |
Token | bo |
Token ID | 2127 |
Feature activation | -1.29e+01 |
Pos logit contributions | |
Fin | +0.079 |
sav | +0.072 |
sailed | +0.071 |
Blue | +0.071 |
Yo | +0.069 |
Neg logit contributions | |
otion | -0.249 |
upon | -0.246 |
spine | -0.230 |
ancer | -0.228 |
headache | -0.223 |
Token | o |
Token ID | 78 |
Feature activation | -6.08e+00 |
Pos logit contributions | |
cream | +0.409 |
Blue | +0.334 |
Fin | +0.327 |
aw | +0.320 |
bo | +0.280 |
Neg logit contributions | |
otion | -1.110 |
Deal | -1.068 |
spine | -1.064 |
laptop | -1.061 |
upon | -1.053 |
Token | on |
Token ID | 319 |
Feature activation | +2.33e+00 |
Pos logit contributions | |
cream | +0.264 |
bo | +0.219 |
Blue | +0.218 |
aw | +0.216 |
never | +0.214 |
Neg logit contributions | |
otion | -0.463 |
Deal | -0.457 |
laptop | -0.439 |
spine | -0.436 |
upon | -0.425 |
Token | her |
Token ID | 607 |
Feature activation | +3.25e+00 |
Pos logit contributions | |
otion | +0.112 |
upon | +0.106 |
spine | +0.103 |
laptop | +0.101 |
Deal | +0.099 |
Neg logit contributions | |
cream | -0.159 |
Blue | -0.151 |
Fin | -0.149 |
Red | -0.143 |
aw | -0.142 |
Token | wing |
Token ID | 8539 |
Feature activation | -3.85e+00 |
Pos logit contributions | |
otion | +0.152 |
upon | +0.135 |
ancer | +0.127 |
Deal | +0.123 |
laptop | +0.119 |
Neg logit contributions | |
cream | -0.238 |
Blue | -0.224 |
sav | -0.217 |
Fin | -0.216 |
Red | -0.213 |
Token | . |
Token ID | 13 |
Feature activation | -1.48e-01 |
Pos logit contributions | |
cream | +0.211 |
Fin | +0.200 |
Blue | +0.199 |
Red | +0.189 |
aw | +0.182 |
Neg logit contributions | |
upon | -0.229 |
otion | -0.225 |
spine | -0.223 |
laptop | -0.215 |
sister | -0.203 |
Token | her |
Token ID | 607 |
Feature activation | +3.29e+00 |
Pos logit contributions | |
cream | +0.075 |
Blue | +0.070 |
Fin | +0.069 |
aw | +0.065 |
Red | +0.065 |
Neg logit contributions | |
otion | -0.088 |
upon | -0.086 |
laptop | -0.083 |
spine | -0.082 |
sister | -0.080 |
Token | toys |
Token ID | 14958 |
Feature activation | +2.29e+00 |
Pos logit contributions | |
otion | +0.225 |
upon | +0.208 |
spine | +0.206 |
Deal | +0.204 |
laptop | +0.203 |
Neg logit contributions | |
cream | -0.164 |
Blue | -0.154 |
Fin | -0.150 |
Red | -0.143 |
Tweet | -0.140 |
Token | . |
Token ID | 13 |
Feature activation | -9.28e-02 |
Pos logit contributions | |
otion | +0.164 |
upon | +0.155 |
laptop | +0.153 |
spine | +0.153 |
Deal | +0.153 |
Neg logit contributions | |
cream | -0.101 |
Blue | -0.095 |
Fin | -0.092 |
Red | -0.087 |
aw | -0.087 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.06e+00 |
Pos logit contributions | |
cream | +0.005 |
Fin | +0.004 |
Blue | +0.004 |
aw | +0.004 |
Red | +0.004 |
Neg logit contributions | |
upon | -0.006 |
otion | -0.006 |
spine | -0.006 |
Deal | -0.006 |
Lee | -0.005 |
Token | Once |
Token ID | 7454 |
Feature activation | -6.66e-01 |
Pos logit contributions | |
otion | +0.169 |
upon | +0.164 |
spine | +0.160 |
laptop | +0.160 |
Deal | +0.156 |
Neg logit contributions | |
cream | -0.071 |
Blue | -0.064 |
Fin | -0.063 |
Red | -0.057 |
aw | -0.056 |
Token | upon |
Token ID | 2402 |
Feature activation | +1.95e+01 |
Pos logit contributions | |
The | +0.055 |
never | +0.054 |
that | +0.054 |
One | +0.053 |
Blue | +0.053 |
Neg logit contributions | |
laptop | -0.019 |
spine | -0.017 |
otion | -0.016 |
Deal | -0.016 |
grade | -0.014 |
Token | a |
Token ID | 257 |
Feature activation | +1.95e+00 |
Pos logit contributions | |
otion | +1.286 |
upon | +1.235 |
spine | +1.210 |
laptop | +1.203 |
Deal | +1.170 |
Neg logit contributions | |
cream | -0.976 |
Blue | -0.917 |
Fin | -0.904 |
Red | -0.850 |
aw | -0.837 |
Token | time |
Token ID | 640 |
Feature activation | +5.27e+00 |
Pos logit contributions | |
otion | +0.105 |
upon | +0.096 |
spine | +0.095 |
laptop | +0.095 |
ancer | +0.093 |
Neg logit contributions | |
cream | -0.119 |
Blue | -0.118 |
Fin | -0.117 |
aw | -0.110 |
Red | -0.110 |
Token | , |
Token ID | 11 |
Feature activation | -4.80e-01 |
Pos logit contributions | |
laptop | +0.433 |
spine | +0.416 |
Deal | +0.403 |
forehead | +0.402 |
ancer | +0.398 |
Neg logit contributions | |
Blue | -0.163 |
never | -0.162 |
Fin | -0.162 |
Bob | -0.162 |
that | -0.156 |
Token | there |
Token ID | 612 |
Feature activation | +1.95e+00 |
Pos logit contributions | |
Blue | +0.022 |
The | +0.021 |
Fin | +0.021 |
Tweet | +0.020 |
They | +0.020 |
Neg logit contributions | |
laptop | -0.033 |
spine | -0.032 |
otion | -0.031 |
Deal | -0.031 |
grade | -0.031 |
Token | was |
Token ID | 373 |
Feature activation | +6.83e-02 |
Pos logit contributions | |
otion | +0.165 |
upon | +0.160 |
spine | +0.157 |
laptop | +0.156 |
Deal | +0.153 |
Neg logit contributions | |
cream | -0.062 |
Blue | -0.056 |
Fin | -0.055 |
Red | -0.049 |
aw | -0.048 |
Token | the |
Token ID | 262 |
Feature activation | +2.75e-01 |
Pos logit contributions | |
otion | +0.343 |
upon | +0.331 |
spine | +0.323 |
laptop | +0.322 |
Deal | +0.312 |
Neg logit contributions | |
cream | -0.248 |
Blue | -0.231 |
Fin | -0.229 |
Red | -0.214 |
aw | -0.210 |
Token | sky |
Token ID | 6766 |
Feature activation | -3.54e-02 |
Pos logit contributions | |
upon | +0.014 |
spine | +0.014 |
sister | +0.013 |
laptop | +0.013 |
headache | +0.013 |
Neg logit contributions | |
cream | -0.018 |
Fin | -0.016 |
Blue | -0.016 |
Bob | -0.015 |
Jerry | -0.015 |
Token | . |
Token ID | 13 |
Feature activation | +3.84e-01 |
Pos logit contributions | |
cream | +0.002 |
Blue | +0.002 |
Fin | +0.002 |
Red | +0.002 |
Tweet | +0.002 |
Neg logit contributions | |
otion | -0.002 |
upon | -0.002 |
spine | -0.002 |
laptop | -0.002 |
ancer | -0.002 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.52e+00 |
Pos logit contributions | |
upon | +0.026 |
otion | +0.026 |
spine | +0.024 |
Deal | +0.023 |
laptop | +0.023 |
Neg logit contributions | |
cream | -0.018 |
Fin | -0.017 |
Blue | -0.017 |
Tweet | -0.016 |
Red | -0.016 |
Token | Once |
Token ID | 7454 |
Feature activation | -2.02e-01 |
Pos logit contributions | |
otion | +0.206 |
upon | +0.200 |
spine | +0.195 |
laptop | +0.195 |
Deal | +0.190 |
Neg logit contributions | |
cream | -0.087 |
Blue | -0.079 |
Fin | -0.077 |
Red | -0.070 |
aw | -0.068 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.00e+01 |
Pos logit contributions | |
that | +0.017 |
The | +0.016 |
never | +0.016 |
Fin | +0.016 |
Blue | +0.016 |
Neg logit contributions | |
laptop | -0.006 |
Deal | -0.006 |
spine | -0.005 |
otion | -0.004 |
scalp | -0.004 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +1.295 |
upon | +1.251 |
spine | +1.213 |
laptop | +1.209 |
Deal | +1.175 |
Neg logit contributions | |
cream | -1.016 |
Blue | -0.953 |
Fin | -0.943 |
Red | -0.884 |
aw | -0.871 |
Token | time |
Token ID | 640 |
Feature activation | +5.73e+00 |
Pos logit contributions | |
otion | +0.127 |
upon | +0.118 |
spine | +0.116 |
laptop | +0.115 |
ancer | +0.111 |
Neg logit contributions | |
cream | -0.152 |
Blue | -0.148 |
Fin | -0.147 |
Red | -0.140 |
aw | -0.139 |
Token | there |
Token ID | 612 |
Feature activation | +2.40e+00 |
Pos logit contributions | |
laptop | +0.478 |
spine | +0.467 |
Deal | +0.452 |
forehead | +0.444 |
scalp | +0.438 |
Neg logit contributions | |
Bob | -0.177 |
Fin | -0.176 |
Blue | -0.171 |
that | -0.167 |
never | -0.164 |
Token | was |
Token ID | 373 |
Feature activation | +4.93e-01 |
Pos logit contributions | |
otion | +0.193 |
upon | +0.189 |
laptop | +0.179 |
spine | +0.178 |
sister | +0.174 |
Neg logit contributions | |
cream | -0.087 |
Blue | -0.079 |
Fin | -0.077 |
Red | -0.072 |
Tweet | -0.069 |
Token | a |
Token ID | 257 |
Feature activation | +2.41e+00 |
Pos logit contributions | |
otion | +0.033 |
upon | +0.031 |
spine | +0.031 |
laptop | +0.031 |
Deal | +0.030 |
Neg logit contributions | |
cream | -0.025 |
Blue | -0.023 |
Fin | -0.023 |
Red | -0.021 |
aw | -0.021 |
Token | her |
Token ID | 607 |
Feature activation | +3.24e+00 |
Pos logit contributions | |
otion | +0.094 |
laptop | +0.089 |
spine | +0.089 |
Deal | +0.087 |
upon | +0.087 |
Neg logit contributions | |
cream | -0.067 |
Blue | -0.065 |
Fin | -0.063 |
aw | -0.059 |
Tweet | -0.059 |
Token | way |
Token ID | 835 |
Feature activation | +1.15e+00 |
Pos logit contributions | |
otion | +0.237 |
Deal | +0.217 |
spine | +0.210 |
laptop | +0.208 |
ancer | +0.208 |
Neg logit contributions | |
cream | -0.137 |
never | -0.130 |
Blue | -0.128 |
sav | -0.122 |
aw | -0.121 |
Token | . |
Token ID | 13 |
Feature activation | -1.49e-01 |
Pos logit contributions | |
otion | +0.082 |
laptop | +0.078 |
Deal | +0.078 |
spine | +0.074 |
Lee | +0.073 |
Neg logit contributions | |
cream | -0.048 |
Blue | -0.048 |
Fin | -0.046 |
never | -0.045 |
The | -0.044 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +1.99e+00 |
Pos logit contributions | |
cream | +0.006 |
Fin | +0.005 |
Blue | +0.005 |
aw | +0.005 |
Red | +0.005 |
Neg logit contributions | |
otion | -0.011 |
upon | -0.011 |
spine | -0.011 |
laptop | -0.011 |
Deal | -0.011 |
Token | Once |
Token ID | 7454 |
Feature activation | -7.45e-01 |
Pos logit contributions | |
otion | +0.165 |
upon | +0.159 |
spine | +0.158 |
laptop | +0.157 |
Deal | +0.153 |
Neg logit contributions | |
cream | -0.065 |
Blue | -0.060 |
Fin | -0.058 |
Red | -0.053 |
aw | -0.051 |
Token | upon |
Token ID | 2402 |
Feature activation | +1.94e+01 |
Pos logit contributions | |
The | +0.064 |
that | +0.063 |
Blue | +0.062 |
One | +0.062 |
never | +0.062 |
Neg logit contributions | |
laptop | -0.019 |
Deal | -0.017 |
spine | -0.016 |
otion | -0.015 |
Lee | -0.014 |
Token | a |
Token ID | 257 |
Feature activation | +1.87e+00 |
Pos logit contributions | |
otion | +1.244 |
upon | +1.208 |
spine | +1.163 |
laptop | +1.161 |
Deal | +1.129 |
Neg logit contributions | |
cream | -1.002 |
Blue | -0.938 |
Fin | -0.932 |
Red | -0.871 |
aw | -0.859 |
Token | time |
Token ID | 640 |
Feature activation | +5.19e+00 |
Pos logit contributions | |
otion | +0.104 |
laptop | +0.093 |
upon | +0.093 |
spine | +0.093 |
ancer | +0.091 |
Neg logit contributions | |
cream | -0.111 |
Blue | -0.111 |
Fin | -0.110 |
aw | -0.105 |
sav | -0.104 |
Token | , |
Token ID | 11 |
Feature activation | -5.74e-01 |
Pos logit contributions | |
laptop | +0.437 |
spine | +0.418 |
Deal | +0.404 |
forehead | +0.404 |
scalp | +0.400 |
Neg logit contributions | |
Blue | -0.153 |
Bob | -0.152 |
never | -0.151 |
Fin | -0.150 |
that | -0.146 |
Token | there |
Token ID | 612 |
Feature activation | +1.85e+00 |
Pos logit contributions | |
Blue | +0.027 |
The | +0.025 |
Fin | +0.025 |
Red | +0.024 |
They | +0.024 |
Neg logit contributions | |
laptop | -0.039 |
spine | -0.039 |
Deal | -0.037 |
otion | -0.037 |
headache | -0.036 |
Token | was |
Token ID | 373 |
Feature activation | -4.35e-02 |
Pos logit contributions | |
otion | +0.160 |
spine | +0.155 |
upon | +0.154 |
laptop | +0.154 |
Deal | +0.152 |
Neg logit contributions | |
cream | -0.053 |
Blue | -0.048 |
Fin | -0.047 |
Red | -0.042 |
aw | -0.041 |
Token | went |
Token ID | 1816 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
cream | +0.114 |
Blue | +0.107 |
Fin | +0.105 |
Red | +0.098 |
Tweet | +0.094 |
Neg logit contributions | |
otion | -0.205 |
upon | -0.200 |
spine | -0.196 |
laptop | -0.193 |
Deal | -0.188 |
Token | home |
Token ID | 1363 |
Feature activation | -2.41e+00 |
Pos logit contributions | |
cream | +0.072 |
Blue | +0.068 |
Fin | +0.067 |
Red | +0.064 |
Tweet | +0.062 |
Neg logit contributions | |
otion | -0.079 |
upon | -0.076 |
spine | -0.074 |
laptop | -0.074 |
Deal | -0.071 |
Token | . |
Token ID | 13 |
Feature activation | +5.68e-02 |
Pos logit contributions | |
cream | +0.094 |
Blue | +0.084 |
never | +0.082 |
sav | +0.081 |
Fin | +0.080 |
Neg logit contributions | |
otion | -0.185 |
Deal | -0.173 |
spine | -0.171 |
upon | -0.170 |
laptop | -0.168 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.21e+00 |
Pos logit contributions | |
otion | +0.005 |
upon | +0.004 |
spine | +0.004 |
laptop | +0.004 |
Deal | +0.004 |
Neg logit contributions | |
cream | -0.002 |
Blue | -0.002 |
Fin | -0.002 |
Red | -0.002 |
aw | -0.002 |
Token | Once |
Token ID | 7454 |
Feature activation | -5.12e-01 |
Pos logit contributions | |
otion | +0.142 |
spine | +0.135 |
upon | +0.134 |
laptop | +0.133 |
Deal | +0.129 |
Neg logit contributions | |
cream | -0.114 |
Blue | -0.108 |
Fin | -0.106 |
Red | -0.100 |
aw | -0.099 |
Token | upon |
Token ID | 2402 |
Feature activation | +1.97e+01 |
Pos logit contributions | |
that | +0.042 |
The | +0.041 |
never | +0.040 |
One | +0.039 |
Fin | +0.039 |
Neg logit contributions | |
laptop | -0.016 |
Deal | -0.015 |
spine | -0.014 |
scalp | -0.013 |
forehead | -0.012 |
Token | a |
Token ID | 257 |
Feature activation | +2.10e+00 |
Pos logit contributions | |
otion | +1.358 |
spine | +1.295 |
upon | +1.281 |
laptop | +1.277 |
Deal | +1.243 |
Neg logit contributions | |
cream | -0.928 |
Blue | -0.878 |
Fin | -0.858 |
Red | -0.810 |
aw | -0.800 |
Token | time |
Token ID | 640 |
Feature activation | +5.43e+00 |
Pos logit contributions | |
otion | +0.118 |
spine | +0.107 |
upon | +0.107 |
laptop | +0.106 |
ancer | +0.104 |
Neg logit contributions | |
Blue | -0.123 |
cream | -0.122 |
Fin | -0.122 |
aw | -0.116 |
Red | -0.116 |
Token | , |
Token ID | 11 |
Feature activation | -3.22e-01 |
Pos logit contributions | |
laptop | +0.474 |
spine | +0.461 |
Deal | +0.448 |
forehead | +0.448 |
scalp | +0.442 |
Neg logit contributions | |
Bob | -0.154 |
Fin | -0.150 |
that | -0.148 |
never | -0.137 |
Blue | -0.135 |
Token | there |
Token ID | 612 |
Feature activation | +2.10e+00 |
Pos logit contributions | |
The | +0.013 |
Blue | +0.013 |
It | +0.012 |
Fin | +0.012 |
They | +0.012 |
Neg logit contributions | |
laptop | -0.025 |
spine | -0.024 |
Deal | -0.024 |
grade | -0.023 |
headache | -0.023 |
Token | was |
Token ID | 373 |
Feature activation | +2.10e-01 |
Pos logit contributions | |
otion | +0.181 |
upon | +0.174 |
spine | +0.174 |
laptop | +0.173 |
Deal | +0.170 |
Neg logit contributions | |
cream | -0.062 |
Blue | -0.056 |
Fin | -0.055 |
Red | -0.048 |
aw | -0.048 |
Token | every |
Token ID | 790 |
Feature activation | -3.29e+00 |
Pos logit contributions | |
cream | +0.216 |
Blue | +0.201 |
Fin | +0.197 |
Red | +0.183 |
aw | +0.179 |
Neg logit contributions | |
otion | -0.380 |
upon | -0.367 |
spine | -0.360 |
laptop | -0.358 |
Deal | -0.350 |
Token | day |
Token ID | 1110 |
Feature activation | +6.06e-01 |
Pos logit contributions | |
cream | +0.151 |
Blue | +0.142 |
Fin | +0.140 |
Red | +0.130 |
aw | +0.128 |
Neg logit contributions | |
otion | -0.231 |
upon | -0.221 |
spine | -0.218 |
laptop | -0.217 |
Deal | -0.212 |
Token | . |
Token ID | 13 |
Feature activation | -6.06e-02 |
Pos logit contributions | |
otion | +0.038 |
upon | +0.036 |
spine | +0.036 |
laptop | +0.036 |
Deal | +0.034 |
Neg logit contributions | |
cream | -0.032 |
Blue | -0.030 |
Fin | -0.030 |
Red | -0.028 |
Tweet | -0.028 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.09e+00 |
Pos logit contributions | |
cream | +0.003 |
Blue | +0.002 |
aw | +0.002 |
Red | +0.002 |
Fin | +0.002 |
Neg logit contributions | |
otion | -0.004 |
upon | -0.004 |
Deal | -0.004 |
spine | -0.004 |
ancer | -0.004 |
Token | Once |
Token ID | 7454 |
Feature activation | -6.41e-01 |
Pos logit contributions | |
otion | +0.153 |
upon | +0.152 |
laptop | +0.140 |
spine | +0.138 |
ancer | +0.137 |
Neg logit contributions | |
cream | -0.090 |
Blue | -0.080 |
Fin | -0.080 |
Red | -0.074 |
Tweet | -0.073 |
Token | upon |
Token ID | 2402 |
Feature activation | +1.96e+01 |
Pos logit contributions | |
The | +0.054 |
Fin | +0.054 |
Blue | +0.053 |
that | +0.053 |
never | +0.053 |
Neg logit contributions | |
laptop | -0.017 |
spine | -0.016 |
Deal | -0.015 |
otion | -0.015 |
Lee | -0.012 |
Token | a |
Token ID | 257 |
Feature activation | +1.99e+00 |
Pos logit contributions | |
otion | +1.205 |
upon | +1.187 |
laptop | +1.119 |
spine | +1.112 |
Deal | +1.088 |
Neg logit contributions | |
cream | -1.049 |
Blue | -0.972 |
Fin | -0.971 |
Red | -0.908 |
aw | -0.896 |
Token | time |
Token ID | 640 |
Feature activation | +5.30e+00 |
Pos logit contributions | |
otion | +0.100 |
upon | +0.093 |
spine | +0.091 |
laptop | +0.091 |
Deal | +0.087 |
Neg logit contributions | |
cream | -0.130 |
Blue | -0.126 |
Fin | -0.124 |
Red | -0.119 |
aw | -0.118 |
Token | , |
Token ID | 11 |
Feature activation | -4.57e-01 |
Pos logit contributions | |
laptop | +0.446 |
spine | +0.432 |
forehead | +0.422 |
Deal | +0.419 |
scalp | +0.411 |
Neg logit contributions | |
Fin | -0.160 |
Bob | -0.156 |
Blue | -0.152 |
that | -0.151 |
One | -0.149 |
Token | there |
Token ID | 612 |
Feature activation | +1.98e+00 |
Pos logit contributions | |
Blue | +0.022 |
Fin | +0.021 |
The | +0.021 |
cream | +0.020 |
They | +0.020 |
Neg logit contributions | |
laptop | -0.031 |
spine | -0.030 |
otion | -0.030 |
Deal | -0.029 |
headache | -0.028 |
Token | was |
Token ID | 373 |
Feature activation | +7.00e-02 |
Pos logit contributions | |
otion | +0.173 |
spine | +0.169 |
laptop | +0.167 |
upon | +0.166 |
Deal | +0.165 |
Neg logit contributions | |
cream | -0.054 |
Blue | -0.049 |
Fin | -0.048 |
aw | -0.041 |
Red | -0.041 |
Token | time |
Token ID | 640 |
Feature activation | +5.41e+00 |
Pos logit contributions | |
otion | +0.107 |
upon | +0.100 |
spine | +0.098 |
laptop | +0.098 |
Deal | +0.093 |
Neg logit contributions | |
cream | -0.134 |
Blue | -0.130 |
Fin | -0.129 |
Red | -0.123 |
aw | -0.122 |
Token | , |
Token ID | 11 |
Feature activation | -3.41e-01 |
Pos logit contributions | |
laptop | +0.453 |
spine | +0.445 |
Deal | +0.433 |
scalp | +0.425 |
forehead | +0.424 |
Neg logit contributions | |
that | -0.154 |
Fin | -0.153 |
Blue | -0.152 |
Bob | -0.151 |
never | -0.149 |
Token | there |
Token ID | 612 |
Feature activation | +2.09e+00 |
Pos logit contributions | |
Blue | +0.016 |
The | +0.015 |
Fin | +0.015 |
cream | +0.015 |
They | +0.014 |
Neg logit contributions | |
spine | -0.023 |
laptop | -0.023 |
Deal | -0.022 |
otion | -0.022 |
ids | -0.021 |
Token | was |
Token ID | 373 |
Feature activation | +2.00e-01 |
Pos logit contributions | |
otion | +0.178 |
upon | +0.172 |
spine | +0.170 |
laptop | +0.169 |
Deal | +0.166 |
Neg logit contributions | |
cream | -0.064 |
Blue | -0.058 |
Fin | -0.057 |
Red | -0.051 |
aw | -0.050 |
Token | a |
Token ID | 257 |
Feature activation | +2.12e+00 |
Pos logit contributions | |
otion | +0.013 |
upon | +0.013 |
spine | +0.013 |
laptop | +0.012 |
Deal | +0.012 |
Neg logit contributions | |
cream | -0.010 |
Blue | -0.009 |
Fin | -0.009 |
Red | -0.009 |
aw | -0.009 |
Token | little |
Token ID | 1310 |
Feature activation | +6.14e+00 |
Pos logit contributions | |
otion | +0.120 |
upon | +0.113 |
spine | +0.111 |
laptop | +0.110 |
Deal | +0.107 |
Neg logit contributions | |
cream | -0.125 |
Blue | -0.120 |
Fin | -0.119 |
Red | -0.113 |
aw | -0.111 |
Token | boy |
Token ID | 2933 |
Feature activation | -2.83e-01 |
Pos logit contributions | |
otion | +0.356 |
upon | +0.340 |
spine | +0.332 |
laptop | +0.330 |
Deal | +0.319 |
Neg logit contributions | |
cream | -0.356 |
Blue | -0.338 |
Fin | -0.334 |
Red | -0.316 |
aw | -0.313 |
Token | named |
Token ID | 3706 |
Feature activation | +3.21e-01 |
Pos logit contributions | |
cream | +0.016 |
Blue | +0.015 |
Fin | +0.015 |
Red | +0.015 |
Yo | +0.014 |
Neg logit contributions | |
otion | -0.016 |
sister | -0.016 |
upon | -0.016 |
laptop | -0.015 |
spine | -0.015 |
Token | Tim |
Token ID | 5045 |
Feature activation | -2.99e+00 |
Pos logit contributions | |
otion | +0.022 |
upon | +0.022 |
spine | +0.021 |
laptop | +0.021 |
Deal | +0.020 |
Neg logit contributions | |
cream | -0.015 |
Blue | -0.014 |
Fin | -0.014 |
aw | -0.014 |
Red | -0.013 |
Token | my |
Token ID | 1820 |
Feature activation | -4.46e+00 |
Pos logit contributions | |
cream | +0.179 |
Blue | +0.169 |
Fin | +0.168 |
Red | +0.159 |
Tweet | +0.154 |
Neg logit contributions | |
upon | -0.174 |
otion | -0.161 |
sister | -0.155 |
spine | -0.151 |
laptop | -0.151 |
Token | . |
Token ID | 13 |
Feature activation | +1.47e-01 |
Pos logit contributions | |
cream | +0.290 |
Fin | +0.287 |
Blue | +0.284 |
Red | +0.273 |
Jerry | +0.265 |
Neg logit contributions | |
upon | -0.207 |
sister | -0.206 |
laptop | -0.191 |
otion | -0.189 |
spine | -0.189 |
Token | " |
Token ID | 366 |
Feature activation | +4.09e+00 |
Pos logit contributions | |
Blue | +0.002 |
The | +0.002 |
cream | +0.002 |
It | +0.002 |
He | +0.002 |
Neg logit contributions | |
otion | -0.003 |
spine | -0.003 |
laptop | -0.003 |
Deal | -0.003 |
headache | -0.003 |
Token | What |
Token ID | 2061 |
Feature activation | -7.63e-01 |
Pos logit contributions | |
upon | +0.300 |
otion | +0.296 |
Deal | +0.277 |
ancer | +0.275 |
laptop | +0.272 |
Neg logit contributions | |
cream | -0.173 |
Fin | -0.153 |
Blue | -0.151 |
sailed | -0.138 |
Red | -0.137 |
Token | is |
Token ID | 318 |
Feature activation | +4.45e-01 |
Pos logit contributions | |
cream | +0.032 |
Blue | +0.031 |
Fin | +0.029 |
aw | +0.029 |
Red | +0.027 |
Neg logit contributions | |
spine | -0.056 |
otion | -0.055 |
upon | -0.053 |
Deal | -0.052 |
laptop | -0.052 |
Token | this |
Token ID | 428 |
Feature activation | -2.63e+00 |
Pos logit contributions | |
otion | +0.038 |
upon | +0.037 |
spine | +0.036 |
laptop | +0.036 |
Deal | +0.035 |
Neg logit contributions | |
cream | -0.014 |
Blue | -0.013 |
Fin | -0.012 |
Red | -0.011 |
aw | -0.011 |
Token | ? |
Token ID | 30 |
Feature activation | +7.13e+00 |
Pos logit contributions | |
cream | +0.131 |
Blue | +0.122 |
Fin | +0.120 |
sav | +0.115 |
aw | +0.115 |
Neg logit contributions | |
otion | -0.169 |
spine | -0.162 |
Deal | -0.160 |
laptop | -0.159 |
upon | -0.157 |
Token | Do |
Token ID | 2141 |
Feature activation | +5.75e+00 |
Pos logit contributions | |
otion | +0.442 |
spine | +0.437 |
upon | +0.428 |
laptop | +0.425 |
Deal | +0.407 |
Neg logit contributions | |
cream | -0.376 |
Blue | -0.358 |
Fin | -0.339 |
Red | -0.332 |
This | -0.322 |
Token | I |
Token ID | 314 |
Feature activation | +1.93e+00 |
Pos logit contributions | |
otion | +0.387 |
upon | +0.371 |
spine | +0.364 |
laptop | +0.362 |
Deal | +0.352 |
Neg logit contributions | |
cream | -0.283 |
Blue | -0.266 |
Fin | -0.262 |
Red | -0.246 |
aw | -0.242 |
Token | go |
Token ID | 467 |
Feature activation | +1.32e+00 |
Pos logit contributions | |
otion | +0.133 |
spine | +0.127 |
laptop | +0.125 |
Deal | +0.122 |
upon | +0.120 |
Neg logit contributions | |
cream | -0.090 |
Fin | -0.086 |
never | -0.086 |
Blue | -0.085 |
sailed | -0.080 |
Token | near |
Token ID | 1474 |
Feature activation | -8.39e-02 |
Pos logit contributions | |
otion | +0.082 |
upon | +0.078 |
spine | +0.077 |
laptop | +0.076 |
Deal | +0.074 |
Neg logit contributions | |
cream | -0.071 |
Blue | -0.067 |
Fin | -0.066 |
Red | -0.062 |
aw | -0.061 |
Token | it |
Token ID | 340 |
Feature activation | -3.40e+00 |
Pos logit contributions | |
cream | +0.004 |
Blue | +0.003 |
Fin | +0.003 |
Red | +0.003 |
Tweet | +0.003 |
Neg logit contributions | |
otion | -0.006 |
upon | -0.006 |
laptop | -0.006 |
spine | -0.006 |
Deal | -0.006 |
Token | or |
Token ID | 393 |
Feature activation | +6.52e+00 |
Pos logit contributions | |
cream | +0.172 |
Blue | +0.162 |
Fin | +0.157 |
Red | +0.149 |
aw | +0.149 |
Neg logit contributions | |
otion | -0.220 |
spine | -0.209 |
laptop | -0.206 |
upon | -0.206 |
Deal | -0.206 |
Token | " |
Token ID | 1 |
Feature activation | -2.19e+00 |
Pos logit contributions | |
cream | +0.061 |
sailed | +0.047 |
sold | +0.047 |
Blue | +0.046 |
Fin | +0.045 |
Neg logit contributions | |
upon | -0.093 |
otion | -0.091 |
laptop | -0.090 |
Deal | -0.088 |
Lee | -0.086 |
Token | Y |
Token ID | 56 |
Feature activation | -1.59e+00 |
Pos logit contributions | |
cream | +0.103 |
oc | +0.089 |
bank | +0.084 |
sailed | +0.082 |
aw | +0.081 |
Neg logit contributions | |
Deal | -0.147 |
otion | -0.140 |
upon | -0.139 |
laptop | -0.136 |
Lee | -0.134 |
Token | ay |
Token ID | 323 |
Feature activation | +6.96e-01 |
Pos logit contributions | |
cream | +0.089 |
Blue | +0.084 |
Fin | +0.083 |
Red | +0.079 |
aw | +0.078 |
Neg logit contributions | |
otion | -0.097 |
upon | -0.092 |
spine | -0.090 |
laptop | -0.090 |
Deal | -0.087 |
Token | !" |
Token ID | 2474 |
Feature activation | -2.32e+00 |
Pos logit contributions | |
otion | +0.045 |
upon | +0.044 |
spine | +0.043 |
laptop | +0.043 |
Deal | +0.042 |
Neg logit contributions | |
cream | -0.036 |
Blue | -0.033 |
Fin | -0.033 |
aw | -0.031 |
Red | -0.031 |
Token | L |
Token ID | 406 |
Feature activation | +7.38e+00 |
Pos logit contributions | |
cream | +0.130 |
Blue | +0.128 |
They | +0.124 |
Fin | +0.124 |
The | +0.122 |
Neg logit contributions | |
otion | -0.140 |
spine | -0.130 |
ancer | -0.127 |
upon | -0.127 |
laptop | -0.124 |
Token | ila |
Token ID | 10102 |
Feature activation | +7.92e+00 |
Pos logit contributions | |
spine | +0.372 |
laptop | +0.334 |
Deal | +0.313 |
headache | +0.307 |
magn | +0.293 |
Neg logit contributions | |
aw | -0.548 |
Fin | -0.504 |
cream | -0.491 |
They | -0.482 |
Bob | -0.476 |
Token | and |
Token ID | 290 |
Feature activation | +9.38e-01 |
Pos logit contributions | |
upon | +0.463 |
laptop | +0.450 |
otion | +0.441 |
sister | +0.440 |
spine | +0.439 |
Neg logit contributions | |
Fin | -0.438 |
cream | -0.435 |
Blue | -0.412 |
Red | -0.396 |
Tweet | -0.372 |
Token | Max |
Token ID | 5436 |
Feature activation | +1.58e+00 |
Pos logit contributions | |
otion | +0.052 |
upon | +0.048 |
ancer | +0.046 |
spine | +0.045 |
laptop | +0.044 |
Neg logit contributions | |
cream | -0.059 |
Blue | -0.057 |
Fin | -0.056 |
sav | -0.055 |
aw | -0.054 |
Token | said |
Token ID | 531 |
Feature activation | -2.13e+00 |
Pos logit contributions | |
otion | +0.119 |
upon | +0.116 |
spine | +0.114 |
laptop | +0.113 |
Deal | +0.110 |
Neg logit contributions | |
cream | -0.064 |
Blue | -0.059 |
Fin | -0.058 |
Red | -0.054 |
aw | -0.052 |
Token | . |
Token ID | 13 |
Feature activation | +7.88e-02 |
Pos logit contributions | |
cream | +0.068 |
They | +0.064 |
never | +0.061 |
Blue | +0.060 |
aw | +0.057 |
Neg logit contributions | |
otion | -0.170 |
upon | -0.167 |
laptop | -0.166 |
spine | -0.164 |
ancer | -0.164 |
Token | They |
Token ID | 1119 |
Feature activation | -2.46e+00 |
Pos logit contributions | |
otion | +0.005 |
upon | +0.005 |
spine | +0.005 |
laptop | +0.005 |
Deal | +0.005 |
Neg logit contributions | |
cream | -0.004 |
Blue | -0.004 |
Fin | -0.004 |
Red | -0.003 |
aw | -0.003 |
Token | something |
Token ID | 1223 |
Feature activation | +3.56e+00 |
Pos logit contributions | |
cream | +0.010 |
Blue | +0.009 |
Fin | +0.009 |
never | +0.009 |
aw | +0.008 |
Neg logit contributions | |
otion | -0.019 |
spine | -0.018 |
upon | -0.018 |
Deal | -0.018 |
laptop | -0.018 |
Token | you |
Token ID | 345 |
Feature activation | +1.40e+00 |
Pos logit contributions | |
otion | +0.262 |
upon | +0.253 |
spine | +0.247 |
laptop | +0.247 |
Deal | +0.240 |
Neg logit contributions | |
cream | -0.150 |
Blue | -0.139 |
Fin | -0.137 |
Red | -0.127 |
aw | -0.124 |
Token | wear |
Token ID | 5806 |
Feature activation | +4.16e+00 |
Pos logit contributions | |
otion | +0.085 |
spine | +0.083 |
laptop | +0.080 |
upon | +0.080 |
Deal | +0.077 |
Neg logit contributions | |
cream | -0.076 |
Blue | -0.072 |
Fin | -0.070 |
aw | -0.067 |
Red | -0.067 |
Token | on |
Token ID | 319 |
Feature activation | +2.58e+00 |
Pos logit contributions | |
otion | +0.283 |
upon | +0.268 |
spine | +0.268 |
laptop | +0.262 |
Deal | +0.258 |
Neg logit contributions | |
cream | -0.201 |
Blue | -0.188 |
Fin | -0.183 |
aw | -0.174 |
Red | -0.174 |
Token | your |
Token ID | 534 |
Feature activation | +7.69e-01 |
Pos logit contributions | |
otion | +0.183 |
upon | +0.175 |
spine | +0.174 |
laptop | +0.170 |
Deal | +0.167 |
Neg logit contributions | |
cream | -0.116 |
Blue | -0.109 |
Fin | -0.106 |
aw | -0.100 |
Red | -0.100 |
Token | head |
Token ID | 1182 |
Feature activation | +5.08e+00 |
Pos logit contributions | |
otion | +0.047 |
upon | +0.041 |
ancer | +0.040 |
spine | +0.037 |
Deal | +0.036 |
Neg logit contributions | |
cream | -0.045 |
sav | -0.043 |
Blue | -0.043 |
Fin | -0.041 |
never | -0.041 |
Token | . |
Token ID | 13 |
Feature activation | +1.51e-01 |
Pos logit contributions | |
otion | +0.340 |
spine | +0.321 |
upon | +0.318 |
laptop | +0.314 |
Deal | +0.312 |
Neg logit contributions | |
cream | -0.251 |
Blue | -0.231 |
Fin | -0.222 |
aw | -0.219 |
Red | -0.213 |
Token | And |
Token ID | 843 |
Feature activation | +2.75e+00 |
Pos logit contributions | |
otion | +0.010 |
upon | +0.010 |
spine | +0.010 |
laptop | +0.009 |
Deal | +0.009 |
Neg logit contributions | |
cream | -0.007 |
Blue | -0.007 |
Fin | -0.007 |
Red | -0.007 |
aw | -0.006 |
Token | this |
Token ID | 428 |
Feature activation | -2.87e+00 |
Pos logit contributions | |
otion | +0.191 |
upon | +0.183 |
spine | +0.182 |
laptop | +0.179 |
Deal | +0.175 |
Neg logit contributions | |
cream | -0.127 |
Blue | -0.119 |
Fin | -0.116 |
aw | -0.109 |
Red | -0.109 |
Token | is |
Token ID | 318 |
Feature activation | +2.22e-01 |
Pos logit contributions | |
cream | +0.133 |
Blue | +0.124 |
Fin | +0.122 |
aw | +0.117 |
sav | +0.113 |
Neg logit contributions | |
otion | -0.197 |
upon | -0.190 |
spine | -0.190 |
laptop | -0.182 |
Deal | -0.182 |
Token | another |
Token ID | 1194 |
Feature activation | -2.45e+00 |
Pos logit contributions | |
otion | +0.015 |
spine | +0.014 |
upon | +0.014 |
laptop | +0.014 |
Deal | +0.014 |
Neg logit contributions | |
cream | -0.011 |
Blue | -0.010 |
Fin | -0.010 |
Red | -0.009 |
aw | -0.009 |
Token | inside |
Token ID | 2641 |
Feature activation | -2.81e+00 |
Pos logit contributions | |
cream | +0.003 |
Blue | +0.003 |
Fin | +0.003 |
Red | +0.002 |
aw | +0.002 |
Neg logit contributions | |
otion | -0.004 |
upon | -0.003 |
Deal | -0.003 |
spine | -0.003 |
laptop | -0.003 |
Token | . |
Token ID | 13 |
Feature activation | -1.57e-01 |
Pos logit contributions | |
The | +0.086 |
cream | +0.085 |
that | +0.083 |
never | +0.082 |
This | +0.080 |
Neg logit contributions | |
spine | -0.230 |
otion | -0.224 |
Deal | -0.221 |
ancer | -0.217 |
laptop | -0.215 |
Token | She |
Token ID | 1375 |
Feature activation | -4.33e-01 |
Pos logit contributions | |
cream | +0.006 |
Blue | +0.006 |
Fin | +0.006 |
Red | +0.005 |
aw | +0.005 |
Neg logit contributions | |
otion | -0.012 |
upon | -0.012 |
spine | -0.011 |
laptop | -0.011 |
Deal | -0.011 |
Token | tried |
Token ID | 3088 |
Feature activation | +2.59e-01 |
Pos logit contributions | |
cream | +0.021 |
Blue | +0.020 |
Fin | +0.020 |
Red | +0.018 |
Jerry | +0.017 |
Neg logit contributions | |
upon | -0.030 |
otion | -0.028 |
laptop | -0.028 |
spine | -0.027 |
sister | -0.027 |
Token | the |
Token ID | 262 |
Feature activation | -2.46e-01 |
Pos logit contributions | |
otion | +0.021 |
Deal | +0.020 |
spine | +0.020 |
ancer | +0.020 |
laptop | +0.019 |
Neg logit contributions | |
cream | -0.008 |
never | -0.008 |
Blue | -0.007 |
The | -0.007 |
They | -0.007 |
Token | knob |
Token ID | 36842 |
Feature activation | +6.12e+00 |
Pos logit contributions | |
cream | +0.014 |
Blue | +0.013 |
sav | +0.012 |
Red | +0.012 |
Fin | +0.012 |
Neg logit contributions | |
otion | -0.016 |
upon | -0.013 |
ids | -0.013 |
Deal | -0.013 |
laptop | -0.013 |
Token | , |
Token ID | 11 |
Feature activation | -5.79e-01 |
Pos logit contributions | |
otion | +0.423 |
Deal | +0.405 |
upon | +0.399 |
spine | +0.399 |
laptop | +0.396 |
Neg logit contributions | |
cream | -0.286 |
Blue | -0.260 |
Fin | -0.247 |
never | -0.240 |
aw | -0.240 |
Token | but |
Token ID | 475 |
Feature activation | +3.65e+00 |
Pos logit contributions | |
cream | +0.030 |
never | +0.030 |
The | +0.029 |
sav | +0.028 |
Blue | +0.028 |
Neg logit contributions | |
otion | -0.035 |
Deal | -0.033 |
spine | -0.032 |
ancer | -0.032 |
upon | -0.032 |
Token | it |
Token ID | 340 |
Feature activation | -3.97e+00 |
Pos logit contributions | |
otion | +0.325 |
Deal | +0.321 |
spine | +0.319 |
laptop | +0.316 |
ancer | +0.305 |
Neg logit contributions | |
never | -0.091 |
cream | -0.084 |
The | -0.079 |
Blue | -0.075 |
It | -0.073 |
Token | was |
Token ID | 373 |
Feature activation | -6.42e-02 |
Pos logit contributions | |
cream | +0.144 |
Blue | +0.132 |
Fin | +0.129 |
Red | +0.118 |
aw | +0.116 |
Neg logit contributions | |
otion | -0.317 |
upon | -0.305 |
spine | -0.302 |
laptop | -0.300 |
Deal | -0.295 |
Token | locked |
Token ID | 8970 |
Feature activation | +7.71e-01 |
Pos logit contributions | |
never | +0.002 |
cream | +0.002 |
Blue | +0.002 |
sold | +0.002 |
sav | +0.002 |
Neg logit contributions | |
otion | -0.005 |
ancer | -0.005 |
spine | -0.005 |
laptop | -0.005 |
Deal | -0.005 |
Token | but |
Token ID | 475 |
Feature activation | +3.66e+00 |
Pos logit contributions | |
cream | +0.121 |
Blue | +0.114 |
Fin | +0.112 |
Red | +0.105 |
aw | +0.104 |
Neg logit contributions | |
otion | -0.161 |
upon | -0.154 |
spine | -0.151 |
laptop | -0.150 |
Deal | -0.147 |
Token | she |
Token ID | 673 |
Feature activation | +2.97e+00 |
Pos logit contributions | |
otion | +0.281 |
upon | +0.267 |
spine | +0.266 |
laptop | +0.265 |
Deal | +0.259 |
Neg logit contributions | |
cream | -0.144 |
Blue | -0.134 |
Fin | -0.132 |
Red | -0.119 |
aw | -0.117 |
Token | didn |
Token ID | 1422 |
Feature activation | +6.68e+00 |
Pos logit contributions | |
upon | +0.200 |
otion | +0.196 |
sister | +0.190 |
laptop | +0.182 |
spine | +0.182 |
Neg logit contributions | |
cream | -0.135 |
Blue | -0.131 |
Red | -0.131 |
Fin | -0.126 |
Tweet | -0.117 |
Token | 't |
Token ID | 470 |
Feature activation | +4.24e+00 |
Pos logit contributions | |
otion | +0.332 |
upon | +0.302 |
Deal | +0.290 |
laptop | +0.288 |
spine | +0.276 |
Neg logit contributions | |
cream | -0.455 |
Fin | -0.426 |
Blue | -0.421 |
aw | -0.412 |
sav | -0.405 |
Token | listen |
Token ID | 6004 |
Feature activation | +1.74e-01 |
Pos logit contributions | |
otion | +0.218 |
upon | +0.209 |
spine | +0.203 |
laptop | +0.195 |
Deal | +0.193 |
Neg logit contributions | |
cream | -0.281 |
Blue | -0.259 |
Fin | -0.253 |
aw | -0.250 |
Red | -0.250 |
Token | . |
Token ID | 13 |
Feature activation | -1.59e-01 |
Pos logit contributions | |
otion | +0.011 |
upon | +0.011 |
spine | +0.011 |
laptop | +0.010 |
Deal | +0.010 |
Neg logit contributions | |
cream | -0.009 |
Blue | -0.009 |
Fin | -0.008 |
Red | -0.008 |
Tweet | -0.008 |
Token | Suddenly |
Token ID | 24975 |
Feature activation | -1.57e+00 |
Pos logit contributions | |
cream | +0.007 |
Blue | +0.006 |
Fin | +0.006 |
Red | +0.006 |
aw | +0.005 |
Neg logit contributions | |
otion | -0.012 |
upon | -0.012 |
spine | -0.011 |
Deal | -0.011 |
laptop | -0.011 |
Token | , |
Token ID | 11 |
Feature activation | -5.87e-01 |
Pos logit contributions | |
cream | +0.066 |
Fin | +0.066 |
Blue | +0.065 |
sailed | +0.062 |
never | +0.061 |
Neg logit contributions | |
otion | -0.111 |
Deal | -0.108 |
laptop | -0.107 |
spine | -0.106 |
ancer | -0.103 |
Token | an |
Token ID | 281 |
Feature activation | +1.90e+00 |
Pos logit contributions | |
cream | +0.019 |
Fin | +0.017 |
Blue | +0.017 |
The | +0.015 |
sailed | +0.015 |
Neg logit contributions | |
otion | -0.050 |
spine | -0.048 |
laptop | -0.048 |
Deal | -0.047 |
upon | -0.046 |
Token | owl |
Token ID | 39610 |
Feature activation | -7.55e+00 |
Pos logit contributions | |
otion | +0.118 |
spine | +0.110 |
upon | +0.110 |
laptop | +0.109 |
Deal | +0.106 |
Neg logit contributions | |
cream | -0.103 |
Blue | -0.098 |
Fin | -0.097 |
Red | -0.091 |
Tweet | -0.089 |
Token | swoop |
Token ID | 38527 |
Feature activation | -6.55e+00 |
Pos logit contributions | |
cream | +0.365 |
Blue | +0.344 |
Fin | +0.339 |
Red | +0.326 |
sav | +0.305 |
Neg logit contributions | |
otion | -0.508 |
upon | -0.495 |
spine | -0.471 |
laptop | -0.463 |
sister | -0.442 |
Token | were |
Token ID | 547 |
Feature activation | +2.67e+00 |
Pos logit contributions | |
otion | +0.224 |
upon | +0.221 |
laptop | +0.209 |
spine | +0.209 |
Deal | +0.202 |
Neg logit contributions | |
cream | -0.157 |
Blue | -0.145 |
Fin | -0.144 |
Red | -0.134 |
Tweet | -0.132 |
Token | friends |
Token ID | 2460 |
Feature activation | -3.78e+00 |
Pos logit contributions | |
spine | +0.195 |
otion | +0.193 |
laptop | +0.180 |
grade | +0.176 |
ancer | +0.175 |
Neg logit contributions | |
aw | -0.108 |
never | -0.104 |
sav | -0.099 |
cream | -0.098 |
They | -0.098 |
Token | . |
Token ID | 13 |
Feature activation | -1.14e-01 |
Pos logit contributions | |
cream | +0.226 |
Fin | +0.211 |
Blue | +0.210 |
Red | +0.197 |
Tweet | +0.193 |
Neg logit contributions | |
upon | -0.217 |
otion | -0.210 |
spine | -0.200 |
laptop | -0.199 |
sister | -0.196 |
Token | They |
Token ID | 1119 |
Feature activation | -2.66e+00 |
Pos logit contributions | |
cream | +0.004 |
Fin | +0.004 |
Blue | +0.004 |
They | +0.003 |
aw | +0.003 |
Neg logit contributions | |
otion | -0.009 |
spine | -0.009 |
upon | -0.009 |
laptop | -0.009 |
ancer | -0.009 |
Token | liked |
Token ID | 8288 |
Feature activation | -1.15e+00 |
Pos logit contributions | |
cream | +0.105 |
Blue | +0.092 |
Fin | +0.091 |
Red | +0.083 |
aw | +0.082 |
Neg logit contributions | |
upon | -0.203 |
otion | -0.203 |
spine | -0.191 |
laptop | -0.190 |
sister | -0.188 |
Token | to |
Token ID | 284 |
Feature activation | -1.34e+00 |
Pos logit contributions | |
cream | +0.059 |
Blue | +0.059 |
Fin | +0.058 |
Red | +0.055 |
Tweet | +0.054 |
Neg logit contributions | |
upon | -0.071 |
otion | -0.071 |
laptop | -0.068 |
sister | -0.066 |
spine | -0.066 |
Token | play |
Token ID | 711 |
Feature activation | -1.73e+00 |
Pos logit contributions | |
cream | +0.065 |
Blue | +0.060 |
Fin | +0.059 |
aw | +0.056 |
Red | +0.056 |
Neg logit contributions | |
otion | -0.090 |
laptop | -0.087 |
upon | -0.087 |
spine | -0.085 |
sister | -0.084 |
Token | with |
Token ID | 351 |
Feature activation | -1.40e+00 |
Pos logit contributions | |
cream | +0.092 |
Blue | +0.083 |
Fin | +0.082 |
Red | +0.077 |
sav | +0.076 |
Neg logit contributions | |
otion | -0.109 |
upon | -0.106 |
laptop | -0.104 |
spine | -0.099 |
sister | -0.098 |
Token | cars |
Token ID | 5006 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
cream | +0.077 |
Blue | +0.073 |
Fin | +0.072 |
Red | +0.068 |
aw | +0.067 |
Neg logit contributions | |
otion | -0.086 |
upon | -0.082 |
laptop | -0.080 |
spine | -0.080 |
Deal | -0.077 |
Token | and |
Token ID | 290 |
Feature activation | +7.98e-01 |
Pos logit contributions | |
cream | +0.059 |
Fin | +0.056 |
Blue | +0.056 |
sav | +0.053 |
Red | +0.052 |
Neg logit contributions | |
otion | -0.060 |
upon | -0.059 |
laptop | -0.059 |
spine | -0.055 |
Lee | -0.054 |
Token | balls |
Token ID | 11333 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
otion | +0.050 |
upon | +0.050 |
laptop | +0.049 |
spine | +0.048 |
sister | +0.046 |
Neg logit contributions | |
cream | -0.042 |
Fin | -0.039 |
Blue | -0.039 |
aw | -0.037 |
Red | -0.036 |
Token | pick |
Token ID | 2298 |
Feature activation | +1.54e+00 |
Pos logit contributions | |
cream | +0.057 |
aw | +0.054 |
Fin | +0.052 |
Blue | +0.052 |
Red | +0.050 |
Neg logit contributions | |
otion | -0.042 |
upon | -0.041 |
spine | -0.039 |
laptop | -0.039 |
sister | -0.038 |
Token | flowers |
Token ID | 12734 |
Feature activation | -3.05e+00 |
Pos logit contributions | |
otion | +0.096 |
upon | +0.092 |
laptop | +0.088 |
spine | +0.088 |
Deal | +0.085 |
Neg logit contributions | |
cream | -0.082 |
Blue | -0.079 |
Fin | -0.077 |
Red | -0.073 |
Tweet | -0.072 |
Token | in |
Token ID | 287 |
Feature activation | +1.55e+00 |
Pos logit contributions | |
cream | +0.124 |
Blue | +0.116 |
Fin | +0.111 |
aw | +0.104 |
Red | +0.103 |
Neg logit contributions | |
otion | -0.224 |
spine | -0.217 |
Deal | -0.217 |
laptop | -0.217 |
upon | -0.215 |
Token | the |
Token ID | 262 |
Feature activation | +2.74e-01 |
Pos logit contributions | |
otion | +0.113 |
spine | +0.109 |
ancer | +0.108 |
upon | +0.106 |
Deal | +0.106 |
Neg logit contributions | |
cream | -0.063 |
Blue | -0.061 |
Fin | -0.059 |
sav | -0.058 |
never | -0.057 |
Token | garden |
Token ID | 11376 |
Feature activation | -3.20e-01 |
Pos logit contributions | |
otion | +0.016 |
upon | +0.016 |
spine | +0.016 |
laptop | +0.015 |
Deal | +0.015 |
Neg logit contributions | |
cream | -0.016 |
Fin | -0.015 |
Blue | -0.014 |
Red | -0.013 |
Tweet | -0.013 |
Token | again |
Token ID | 757 |
Feature activation | -3.77e+00 |
Pos logit contributions | |
cream | +0.014 |
Blue | +0.013 |
Fin | +0.012 |
aw | +0.012 |
sav | +0.012 |
Neg logit contributions | |
otion | -0.023 |
upon | -0.022 |
Deal | -0.022 |
spine | -0.022 |
laptop | -0.021 |
Token | . |
Token ID | 13 |
Feature activation | +3.84e-01 |
Pos logit contributions | |
cream | +0.176 |
Blue | +0.166 |
Fin | +0.164 |
aw | +0.152 |
Red | +0.152 |
Neg logit contributions | |
otion | -0.261 |
upon | -0.247 |
spine | -0.244 |
laptop | -0.243 |
Deal | -0.238 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.52e+00 |
Pos logit contributions | |
otion | +0.030 |
upon | +0.029 |
spine | +0.029 |
laptop | +0.029 |
Deal | +0.028 |
Neg logit contributions | |
cream | -0.014 |
Blue | -0.013 |
Fin | -0.013 |
Red | -0.012 |
aw | -0.012 |
Token | Once |
Token ID | 7454 |
Feature activation | -2.02e-01 |
Pos logit contributions | |
otion | +0.192 |
spine | +0.189 |
laptop | +0.187 |
Deal | +0.181 |
upon | +0.181 |
Neg logit contributions | |
cream | -0.097 |
Blue | -0.092 |
Fin | -0.089 |
Red | -0.083 |
aw | -0.082 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.00e+01 |
Pos logit contributions | |
that | +0.016 |
The | +0.016 |
never | +0.016 |
One | +0.015 |
It | +0.015 |
Neg logit contributions | |
laptop | -0.007 |
spine | -0.006 |
Deal | -0.006 |
scalp | -0.005 |
forehead | -0.005 |
Token | a |
Token ID | 257 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
otion | +1.468 |
spine | +1.427 |
laptop | +1.386 |
Deal | +1.360 |
upon | +1.357 |
Neg logit contributions | |
cream | -0.856 |
Blue | -0.823 |
Fin | -0.775 |
The | -0.754 |
Red | -0.749 |
Token | so |
Token ID | 523 |
Feature activation | -1.75e+00 |
Pos logit contributions | |
otion | +0.037 |
spine | +0.036 |
ancer | +0.034 |
Deal | +0.034 |
upon | +0.034 |
Neg logit contributions | |
cream | -0.019 |
never | -0.018 |
Blue | -0.018 |
sold | -0.018 |
sav | -0.018 |
Token | excited |
Token ID | 6568 |
Feature activation | +3.45e+00 |
Pos logit contributions | |
cream | +0.079 |
Blue | +0.075 |
Fin | +0.072 |
They | +0.071 |
sold | +0.071 |
Neg logit contributions | |
spine | -0.122 |
otion | -0.118 |
laptop | -0.115 |
Deal | -0.112 |
ids | -0.111 |
Token | to |
Token ID | 284 |
Feature activation | -8.14e-01 |
Pos logit contributions | |
otion | +0.251 |
upon | +0.238 |
spine | +0.236 |
laptop | +0.235 |
Deal | +0.234 |
Neg logit contributions | |
cream | -0.152 |
Blue | -0.139 |
Fin | -0.137 |
aw | -0.129 |
Red | -0.127 |
Token | see |
Token ID | 766 |
Feature activation | -1.31e+00 |
Pos logit contributions | |
cream | +0.042 |
Blue | +0.040 |
Fin | +0.039 |
sav | +0.037 |
Red | +0.037 |
Neg logit contributions | |
otion | -0.053 |
upon | -0.050 |
spine | -0.050 |
laptop | -0.049 |
ancer | -0.048 |
Token | her |
Token ID | 607 |
Feature activation | +3.80e+00 |
Pos logit contributions | |
cream | +0.074 |
Blue | +0.073 |
Fin | +0.072 |
Tweet | +0.068 |
Red | +0.067 |
Neg logit contributions | |
otion | -0.077 |
upon | -0.074 |
laptop | -0.071 |
spine | -0.068 |
Lee | -0.068 |
Token | uncle |
Token ID | 7711 |
Feature activation | +1.90e+00 |
Pos logit contributions | |
otion | +0.276 |
Deal | +0.256 |
ancer | +0.244 |
ids | +0.244 |
spine | +0.243 |
Neg logit contributions | |
cream | -0.192 |
sold | -0.165 |
aw | -0.158 |
that | -0.155 |
This | -0.154 |
Token | and |
Token ID | 290 |
Feature activation | +1.29e+00 |
Pos logit contributions | |
otion | +0.139 |
spine | +0.136 |
Deal | +0.135 |
upon | +0.130 |
laptop | +0.128 |
Neg logit contributions | |
cream | -0.084 |
Blue | -0.073 |
sold | -0.069 |
Fin | -0.069 |
Red | -0.067 |
Token | asked |
Token ID | 1965 |
Feature activation | -6.94e+00 |
Pos logit contributions | |
otion | +0.088 |
spine | +0.083 |
upon | +0.083 |
laptop | +0.081 |
Deal | +0.080 |
Neg logit contributions | |
cream | -0.063 |
Blue | -0.058 |
Fin | -0.058 |
sav | -0.054 |
Red | -0.053 |
Token | her |
Token ID | 607 |
Feature activation | +3.80e+00 |
Pos logit contributions | |
cream | +0.289 |
Blue | +0.266 |
never | +0.265 |
aw | +0.263 |
They | +0.261 |
Neg logit contributions | |
spine | -0.506 |
otion | -0.481 |
laptop | -0.469 |
Deal | -0.468 |
upon | -0.458 |
Token | mom |
Token ID | 1995 |
Feature activation | +2.86e+00 |
Pos logit contributions | |
otion | +0.224 |
spine | +0.220 |
upon | +0.201 |
Deal | +0.199 |
laptop | +0.198 |
Neg logit contributions | |
cream | -0.225 |
Blue | -0.203 |
Fin | -0.202 |
The | -0.195 |
They | -0.194 |
Token | lots |
Token ID | 6041 |
Feature activation | -2.51e-01 |
Pos logit contributions | |
otion | +0.175 |
spine | +0.173 |
laptop | +0.163 |
Deal | +0.162 |
headache | +0.155 |
Neg logit contributions | |
cream | -0.157 |
Blue | -0.141 |
Fin | -0.137 |
They | -0.137 |
The | -0.134 |
Token | mom |
Token ID | 1995 |
Feature activation | +2.35e+00 |
Pos logit contributions | |
otion | +0.026 |
ids | +0.023 |
ancer | +0.023 |
Deal | +0.022 |
spine | +0.022 |
Neg logit contributions | |
cream | -0.023 |
never | -0.023 |
Blue | -0.023 |
Fin | -0.023 |
sav | -0.023 |
Token | . |
Token ID | 13 |
Feature activation | -9.28e-02 |
Pos logit contributions | |
otion | +0.165 |
spine | +0.157 |
Deal | +0.157 |
upon | +0.156 |
laptop | +0.154 |
Neg logit contributions | |
cream | -0.104 |
Blue | -0.097 |
Fin | -0.094 |
aw | -0.093 |
sav | -0.090 |
Token |
|
Token ID | 198 |
Feature activation | -1.39e+00 |
Pos logit contributions | |
Blue | +0.004 |
cream | +0.004 |
Fin | +0.003 |
aw | +0.003 |
They | +0.003 |
Neg logit contributions | |
spine | -0.007 |
otion | -0.007 |
laptop | -0.007 |
upon | -0.007 |
ancer | -0.007 |
Token |
|
Token ID | 198 |
Feature activation | -1.38e+00 |
Pos logit contributions | |
cream | +0.076 |
Fin | +0.061 |
sailed | +0.059 |
Blue | +0.059 |
aw | +0.059 |
Neg logit contributions | |
otion | -0.092 |
upon | -0.091 |
laptop | -0.082 |
Deal | -0.080 |
Lee | -0.078 |
Token | " |
Token ID | 1 |
Feature activation | -2.27e+00 |
Pos logit contributions | |
cream | +0.079 |
sailed | +0.062 |
sav | +0.059 |
sold | +0.059 |
Fin | +0.059 |
Neg logit contributions | |
otion | -0.087 |
upon | -0.086 |
laptop | -0.077 |
Deal | -0.077 |
cloth | -0.074 |
Token | Time |
Token ID | 7575 |
Feature activation | +2.18e+00 |
Pos logit contributions | |
cream | +0.090 |
Blue | +0.076 |
Fin | +0.075 |
sav | +0.071 |
aw | +0.070 |
Neg logit contributions | |
otion | -0.178 |
upon | -0.173 |
laptop | -0.166 |
Deal | -0.164 |
spine | -0.163 |
Token | for |
Token ID | 329 |
Feature activation | +2.22e+00 |
Pos logit contributions | |
Deal | +0.208 |
otion | +0.203 |
spine | +0.197 |
acy | +0.196 |
ancer | +0.192 |
Neg logit contributions | |
never | -0.052 |
sold | -0.044 |
sailed | -0.038 |
cream | -0.037 |
Yo | -0.036 |
Token | bed |
Token ID | 3996 |
Feature activation | +1.07e+01 |
Pos logit contributions | |
otion | +0.138 |
spine | +0.134 |
ancer | +0.133 |
upon | +0.133 |
laptop | +0.127 |
Neg logit contributions | |
cream | -0.115 |
Blue | -0.112 |
sav | -0.109 |
Fin | -0.109 |
never | -0.108 |
Token | , |
Token ID | 11 |
Feature activation | -4.75e-01 |
Pos logit contributions | |
otion | +0.718 |
Deal | +0.696 |
upon | +0.691 |
spine | +0.683 |
ancer | +0.667 |
Neg logit contributions | |
cream | -0.512 |
aw | -0.457 |
Blue | -0.454 |
never | -0.453 |
sold | -0.443 |
Token | kids |
Token ID | 3988 |
Feature activation | +3.07e+00 |
Pos logit contributions | |
Blue | +0.023 |
They | +0.022 |
never | +0.022 |
Fin | +0.022 |
Red | +0.021 |
Neg logit contributions | |
otion | -0.030 |
spine | -0.030 |
laptop | -0.029 |
magn | -0.027 |
acy | -0.027 |
Token | !" |
Token ID | 2474 |
Feature activation | -2.40e+00 |
Pos logit contributions | |
Deal | +0.236 |
otion | +0.236 |
spine | +0.220 |
laptop | +0.220 |
upon | +0.215 |
Neg logit contributions | |
cream | -0.114 |
never | -0.107 |
Blue | -0.107 |
Fin | -0.102 |
sold | -0.100 |
Token | After |
Token ID | 3260 |
Feature activation | +1.98e-01 |
Pos logit contributions | |
cream | +0.053 |
Blue | +0.045 |
Fin | +0.040 |
Red | +0.040 |
sold | +0.040 |
Neg logit contributions | |
upon | -0.095 |
Deal | -0.088 |
otion | -0.087 |
acy | -0.084 |
laptop | -0.084 |
Token | a |
Token ID | 257 |
Feature activation | +2.09e+00 |
Pos logit contributions | |
otion | +0.014 |
upon | +0.013 |
spine | +0.013 |
laptop | +0.013 |
Deal | +0.013 |
Neg logit contributions | |
cream | -0.009 |
Blue | -0.009 |
Fin | -0.008 |
Red | -0.008 |
sav | -0.008 |
Token | while |
Token ID | 981 |
Feature activation | +7.92e+00 |
Pos logit contributions | |
upon | +0.124 |
otion | +0.121 |
spine | +0.120 |
laptop | +0.119 |
sister | +0.114 |
Neg logit contributions | |
cream | -0.119 |
Blue | -0.106 |
Fin | -0.105 |
aw | -0.105 |
Red | -0.101 |
Token | , |
Token ID | 11 |
Feature activation | -3.41e-01 |
Pos logit contributions | |
otion | +0.577 |
laptop | +0.540 |
Deal | +0.539 |
spine | +0.524 |
upon | +0.514 |
Neg logit contributions | |
cream | -0.324 |
sav | -0.319 |
Fin | -0.318 |
The | -0.311 |
never | -0.310 |
Token | Mate |
Token ID | 24787 |
Feature activation | -1.21e+00 |
Pos logit contributions | |
cream | +0.017 |
Fin | +0.015 |
Blue | +0.015 |
sav | +0.015 |
The | +0.014 |
Neg logit contributions | |
otion | -0.024 |
laptop | -0.022 |
spine | -0.022 |
Deal | -0.022 |
upon | -0.021 |
Token | o |
Token ID | 78 |
Feature activation | -5.88e+00 |
Pos logit contributions | |
cream | +0.032 |
Fin | +0.025 |
aw | +0.023 |
bo | +0.022 |
Blue | +0.021 |
Neg logit contributions | |
otion | -0.115 |
laptop | -0.111 |
spine | -0.109 |
Deal | -0.107 |
headache | -0.104 |
Token | was |
Token ID | 373 |
Feature activation | +1.96e-01 |
Pos logit contributions | |
cream | +0.273 |
Blue | +0.268 |
Fin | +0.258 |
Red | +0.251 |
Tweet | +0.234 |
Neg logit contributions | |
upon | -0.398 |
otion | -0.380 |
laptop | -0.368 |
spine | -0.367 |
sister | -0.365 |
Token | getting |
Token ID | 1972 |
Feature activation | +2.76e+00 |
Pos logit contributions | |
otion | +0.013 |
upon | +0.012 |
spine | +0.012 |
laptop | +0.012 |
Deal | +0.012 |
Neg logit contributions | |
cream | -0.010 |
Blue | -0.009 |
Fin | -0.009 |
aw | -0.009 |
sav | -0.009 |
Token | very |
Token ID | 845 |
Feature activation | -6.22e-01 |
Pos logit contributions | |
upon | +0.156 |
otion | +0.141 |
sister | +0.134 |
ancer | +0.133 |
laptop | +0.133 |
Neg logit contributions | |
cream | -0.164 |
Blue | -0.162 |
Fin | -0.162 |
oc | -0.156 |
Tweet | -0.151 |
Token | good |
Token ID | 922 |
Feature activation | -2.84e+00 |
Pos logit contributions | |
cream | +0.037 |
Blue | +0.035 |
Fin | +0.035 |
Red | +0.033 |
aw | +0.033 |
Neg logit contributions | |
otion | -0.035 |
upon | -0.034 |
spine | -0.033 |
laptop | -0.033 |
Deal | -0.032 |
Token | at |
Token ID | 379 |
Feature activation | +1.77e+00 |
Pos logit contributions | |
cream | +0.139 |
aw | +0.122 |
Fin | +0.122 |
Blue | +0.120 |
sav | +0.118 |
Neg logit contributions | |
otion | -0.194 |
upon | -0.183 |
Deal | -0.181 |
spine | -0.180 |
laptop | -0.178 |
Token | interesting |
Token ID | 3499 |
Feature activation | -7.29e+00 |
Pos logit contributions | |
cream | +0.055 |
Blue | +0.054 |
Fin | +0.052 |
Tweet | +0.048 |
Red | +0.048 |
Neg logit contributions | |
upon | -0.071 |
otion | -0.070 |
laptop | -0.066 |
sister | -0.065 |
Lee | -0.062 |
Token | things |
Token ID | 1243 |
Feature activation | -5.09e-02 |
Pos logit contributions | |
cream | +0.408 |
Blue | +0.404 |
Fin | +0.391 |
Red | +0.376 |
Tweet | +0.369 |
Neg logit contributions | |
otion | -0.421 |
upon | -0.419 |
laptop | -0.392 |
spine | -0.377 |
sister | -0.371 |
Token | . |
Token ID | 13 |
Feature activation | +1.46e-01 |
Pos logit contributions | |
cream | +0.003 |
Blue | +0.003 |
Fin | +0.003 |
Red | +0.002 |
Tweet | +0.002 |
Neg logit contributions | |
otion | -0.003 |
upon | -0.003 |
laptop | -0.003 |
spine | -0.003 |
ancer | -0.003 |
Token | He |
Token ID | 679 |
Feature activation | -3.49e+00 |
Pos logit contributions | |
otion | +0.012 |
upon | +0.011 |
spine | +0.011 |
laptop | +0.011 |
Deal | +0.011 |
Neg logit contributions | |
cream | -0.005 |
Blue | -0.005 |
Fin | -0.005 |
Red | -0.004 |
Tweet | -0.004 |
Token | saw |
Token ID | 2497 |
Feature activation | -3.70e-01 |
Pos logit contributions | |
cream | +0.173 |
Blue | +0.162 |
Fin | +0.156 |
Red | +0.151 |
Tweet | +0.139 |
Neg logit contributions | |
upon | -0.244 |
otion | -0.214 |
sister | -0.214 |
laptop | -0.211 |
spine | -0.203 |
Token | birds |
Token ID | 10087 |
Feature activation | -5.09e+00 |
Pos logit contributions | |
cream | +0.019 |
Blue | +0.019 |
Fin | +0.018 |
Red | +0.018 |
oc | +0.017 |
Neg logit contributions | |
upon | -0.022 |
otion | -0.022 |
laptop | -0.021 |
sister | -0.020 |
Lee | -0.020 |
Token | flying |
Token ID | 7348 |
Feature activation | -1.20e+00 |
Pos logit contributions | |
cream | +0.300 |
Blue | +0.284 |
Fin | +0.281 |
Red | +0.268 |
aw | +0.261 |
Neg logit contributions | |
otion | -0.292 |
upon | -0.280 |
spine | -0.269 |
laptop | -0.267 |
Deal | -0.255 |
Token | in |
Token ID | 287 |
Feature activation | +1.34e+00 |
Pos logit contributions | |
cream | +0.065 |
Blue | +0.061 |
Fin | +0.059 |
Red | +0.058 |
sav | +0.056 |
Neg logit contributions | |
upon | -0.074 |
otion | -0.073 |
spine | -0.069 |
laptop | -0.068 |
ancer | -0.065 |
Token | the |
Token ID | 262 |
Feature activation | +6.59e-02 |
Pos logit contributions | |
upon | +0.077 |
otion | +0.072 |
Lee | +0.066 |
laptop | +0.066 |
curtain | +0.066 |
Neg logit contributions | |
cream | -0.083 |
Blue | -0.075 |
Fin | -0.074 |
Red | -0.071 |
Tweet | -0.070 |
Token | sky |
Token ID | 6766 |
Feature activation | -2.37e-01 |
Pos logit contributions | |
upon | +0.004 |
spine | +0.003 |
curtain | +0.003 |
laptop | +0.003 |
otion | +0.003 |
Neg logit contributions | |
cream | -0.004 |
Fin | -0.004 |
Blue | -0.003 |
oc | -0.003 |
Jerry | -0.003 |
Token | , |
Token ID | 11 |
Feature activation | -2.47e-01 |
Pos logit contributions | |
cream | +0.013 |
Blue | +0.012 |
Fin | +0.012 |
Red | +0.012 |
Tweet | +0.011 |
Neg logit contributions | |
otion | -0.014 |
upon | -0.014 |
spine | -0.014 |
laptop | -0.013 |
ancer | -0.013 |
Token | good |
Token ID | 922 |
Feature activation | -2.87e+00 |
Pos logit contributions | |
otion | +0.145 |
upon | +0.139 |
spine | +0.137 |
laptop | +0.135 |
Deal | +0.132 |
Neg logit contributions | |
cream | -0.100 |
Blue | -0.095 |
Fin | -0.093 |
Red | -0.087 |
aw | -0.086 |
Token | friend |
Token ID | 1545 |
Feature activation | -2.07e+00 |
Pos logit contributions | |
sav | +0.129 |
Blue | +0.128 |
They | +0.124 |
Fin | +0.124 |
cream | +0.124 |
Neg logit contributions | |
otion | -0.193 |
spine | -0.181 |
upon | -0.170 |
Deal | -0.168 |
ancer | -0.163 |
Token | ." |
Token ID | 526 |
Feature activation | -1.28e+00 |
Pos logit contributions | |
cream | +0.098 |
Blue | +0.092 |
Fin | +0.091 |
Red | +0.085 |
aw | +0.084 |
Neg logit contributions | |
otion | -0.141 |
upon | -0.135 |
spine | -0.133 |
laptop | -0.132 |
Deal | -0.128 |
Token |
|
Token ID | 198 |
Feature activation | -1.16e+00 |
Pos logit contributions | |
Blue | +0.031 |
cream | +0.031 |
Fin | +0.029 |
Red | +0.026 |
The | +0.025 |
Neg logit contributions | |
otion | -0.116 |
spine | -0.112 |
laptop | -0.110 |
upon | -0.110 |
ancer | -0.108 |
Token |
|
Token ID | 198 |
Feature activation | -1.16e+00 |
Pos logit contributions | |
cream | +0.052 |
Fin | +0.045 |
Blue | +0.044 |
sailed | +0.041 |
sold | +0.041 |
Neg logit contributions | |
upon | -0.085 |
otion | -0.085 |
laptop | -0.079 |
Deal | -0.078 |
spine | -0.077 |
Token | Tom |
Token ID | 13787 |
Feature activation | -4.92e+00 |
Pos logit contributions | |
cream | +0.063 |
sailed | +0.053 |
sold | +0.050 |
bo | +0.044 |
aw | +0.044 |
Neg logit contributions | |
upon | -0.077 |
laptop | -0.070 |
Deal | -0.070 |
otion | -0.070 |
Lee | -0.068 |
Token | thinks |
Token ID | 6834 |
Feature activation | -6.90e-01 |
Pos logit contributions | |
cream | +0.247 |
Blue | +0.232 |
Fin | +0.226 |
aw | +0.218 |
Red | +0.214 |
Neg logit contributions | |
otion | -0.327 |
spine | -0.308 |
upon | -0.306 |
laptop | -0.300 |
ancer | -0.294 |
Token | about |
Token ID | 546 |
Feature activation | -8.98e-01 |
Pos logit contributions | |
that | +0.030 |
cream | +0.029 |
more | +0.026 |
never | +0.026 |
aw | +0.025 |
Neg logit contributions | |
spine | -0.048 |
Deal | -0.047 |
otion | -0.046 |
laptop | -0.046 |
ancer | -0.045 |
Token | it |
Token ID | 340 |
Feature activation | -3.65e+00 |
Pos logit contributions | |
cream | +0.041 |
Blue | +0.039 |
Fin | +0.038 |
Red | +0.036 |
aw | +0.035 |
Neg logit contributions | |
otion | -0.063 |
upon | -0.061 |
laptop | -0.059 |
spine | -0.059 |
Deal | -0.057 |
Token | . |
Token ID | 13 |
Feature activation | +1.79e-01 |
Pos logit contributions | |
cream | +0.141 |
Blue | +0.131 |
aw | +0.122 |
Fin | +0.122 |
They | +0.122 |
Neg logit contributions | |
otion | -0.276 |
spine | -0.266 |
laptop | -0.257 |
Deal | -0.256 |
upon | -0.254 |
Token | He |
Token ID | 679 |
Feature activation | -3.45e+00 |
Pos logit contributions | |
upon | +0.013 |
otion | +0.013 |
Deal | +0.013 |
spine | +0.013 |
laptop | +0.013 |
Neg logit contributions | |
cream | -0.007 |
Fin | -0.007 |
Blue | -0.007 |
Tweet | -0.006 |
aw | -0.006 |
Token | task |
Token ID | 4876 |
Feature activation | +5.00e+00 |
Pos logit contributions | |
cream | +0.018 |
Blue | +0.017 |
Fin | +0.017 |
sav | +0.016 |
Red | +0.016 |
Neg logit contributions | |
otion | -0.023 |
upon | -0.021 |
spine | -0.021 |
laptop | -0.021 |
Deal | -0.020 |
Token | . |
Token ID | 13 |
Feature activation | +3.26e-01 |
Pos logit contributions | |
otion | +0.323 |
upon | +0.313 |
spine | +0.307 |
laptop | +0.305 |
Deal | +0.291 |
Neg logit contributions | |
cream | -0.255 |
Blue | -0.241 |
Fin | -0.238 |
Red | -0.223 |
Tweet | -0.219 |
Token | He |
Token ID | 679 |
Feature activation | -3.29e+00 |
Pos logit contributions | |
otion | +0.025 |
upon | +0.025 |
spine | +0.024 |
Deal | +0.024 |
laptop | +0.024 |
Neg logit contributions | |
cream | -0.012 |
Blue | -0.011 |
Fin | -0.011 |
Red | -0.010 |
Tweet | -0.010 |
Token | had |
Token ID | 550 |
Feature activation | -8.34e-01 |
Pos logit contributions | |
cream | +0.157 |
Blue | +0.155 |
Fin | +0.154 |
Red | +0.150 |
Jerry | +0.138 |
Neg logit contributions | |
upon | -0.213 |
sister | -0.203 |
laptop | -0.190 |
otion | -0.190 |
acy | -0.188 |
Token | a |
Token ID | 257 |
Feature activation | +2.34e+00 |
Pos logit contributions | |
cream | +0.039 |
Blue | +0.038 |
Fin | +0.037 |
Red | +0.035 |
Tweet | +0.035 |
Neg logit contributions | |
upon | -0.059 |
otion | -0.053 |
sister | -0.052 |
laptop | -0.051 |
spine | -0.051 |
Token | wonderful |
Token ID | 7932 |
Feature activation | -6.69e+00 |
Pos logit contributions | |
otion | +0.176 |
upon | +0.169 |
spine | +0.166 |
laptop | +0.165 |
Deal | +0.162 |
Neg logit contributions | |
cream | -0.096 |
Blue | -0.089 |
Fin | -0.088 |
Red | -0.081 |
aw | -0.080 |
Token | day |
Token ID | 1110 |
Feature activation | +1.02e+00 |
Pos logit contributions | |
cream | +0.329 |
Blue | +0.312 |
Fin | +0.308 |
sav | +0.298 |
aw | +0.287 |
Neg logit contributions | |
otion | -0.448 |
upon | -0.414 |
spine | -0.410 |
laptop | -0.406 |
Deal | -0.406 |
Token | wandering |
Token ID | 24504 |
Feature activation | +3.91e+00 |
Pos logit contributions | |
otion | +0.063 |
upon | +0.060 |
spine | +0.059 |
laptop | +0.059 |
Deal | +0.058 |
Neg logit contributions | |
cream | -0.056 |
Blue | -0.052 |
Fin | -0.051 |
aw | -0.049 |
sav | -0.049 |
Token | around |
Token ID | 1088 |
Feature activation | -3.77e+00 |
Pos logit contributions | |
otion | +0.261 |
upon | +0.255 |
spine | +0.247 |
laptop | +0.244 |
ancer | +0.234 |
Neg logit contributions | |
cream | -0.191 |
Blue | -0.180 |
Fin | -0.176 |
Red | -0.167 |
Tweet | -0.161 |
Token | and |
Token ID | 290 |
Feature activation | +1.23e+00 |
Pos logit contributions | |
cream | +0.154 |
Blue | +0.145 |
Fin | +0.144 |
sav | +0.138 |
aw | +0.134 |
Neg logit contributions | |
otion | -0.279 |
spine | -0.265 |
Deal | -0.264 |
laptop | -0.263 |
upon | -0.259 |
Token | exploring |
Token ID | 13504 |
Feature activation | +1.25e-01 |
Pos logit contributions | |
upon | +0.071 |
otion | +0.071 |
spine | +0.067 |
laptop | +0.065 |
sister | +0.063 |
Neg logit contributions | |
cream | -0.070 |
Blue | -0.068 |
Fin | -0.067 |
Red | -0.065 |
Tweet | -0.063 |
Token | . |
Token ID | 13 |
Feature activation | +5.01e-02 |
Pos logit contributions | |
cream | +0.185 |
Blue | +0.174 |
Fin | +0.172 |
Red | +0.162 |
aw | +0.160 |
Neg logit contributions | |
otion | -0.233 |
upon | -0.224 |
spine | -0.219 |
laptop | -0.218 |
Deal | -0.212 |
Token | One |
Token ID | 1881 |
Feature activation | -5.36e+00 |
Pos logit contributions | |
upon | +0.004 |
otion | +0.004 |
spine | +0.003 |
Deal | +0.003 |
sister | +0.003 |
Neg logit contributions | |
cream | -0.002 |
aw | -0.002 |
Blue | -0.002 |
Red | -0.002 |
Fin | -0.002 |
Token | day |
Token ID | 1110 |
Feature activation | +7.35e-01 |
Pos logit contributions | |
cream | +0.250 |
Blue | +0.234 |
Fin | +0.231 |
Red | +0.215 |
aw | +0.212 |
Neg logit contributions | |
otion | -0.374 |
upon | -0.359 |
spine | -0.353 |
laptop | -0.351 |
Deal | -0.342 |
Token | , |
Token ID | 11 |
Feature activation | -3.73e-01 |
Pos logit contributions | |
laptop | +0.047 |
otion | +0.046 |
spine | +0.045 |
Deal | +0.043 |
ancer | +0.043 |
Neg logit contributions | |
Fin | -0.038 |
Blue | -0.037 |
cream | -0.037 |
Tweet | -0.034 |
aw | -0.034 |
Token | Freddy |
Token ID | 45437 |
Feature activation | -2.81e+00 |
Pos logit contributions | |
cream | +0.018 |
Blue | +0.017 |
Fin | +0.016 |
Red | +0.016 |
aw | +0.015 |
Neg logit contributions | |
otion | -0.025 |
upon | -0.025 |
spine | -0.024 |
laptop | -0.023 |
Deal | -0.023 |
Token | 's |
Token ID | 338 |
Feature activation | -5.30e+00 |
Pos logit contributions | |
cream | +0.117 |
Blue | +0.107 |
Fin | +0.104 |
Red | +0.098 |
aw | +0.095 |
Neg logit contributions | |
otion | -0.209 |
upon | -0.205 |
spine | -0.198 |
laptop | -0.197 |
Deal | -0.190 |
Token | friend |
Token ID | 1545 |
Feature activation | -2.11e+00 |
Pos logit contributions | |
Fin | +0.195 |
Tweet | +0.170 |
Blue | +0.169 |
never | +0.164 |
cream | +0.164 |
Neg logit contributions | |
otion | -0.425 |
Deal | -0.398 |
laptop | -0.392 |
spine | -0.381 |
ancer | -0.376 |
Token | Bobby |
Token ID | 17155 |
Feature activation | -2.81e+00 |
Pos logit contributions | |
cream | +0.110 |
Fin | +0.109 |
Blue | +0.108 |
Tweet | +0.099 |
aw | +0.099 |
Neg logit contributions | |
otion | -0.133 |
spine | -0.129 |
laptop | -0.129 |
Deal | -0.126 |
upon | -0.122 |
Token | told |
Token ID | 1297 |
Feature activation | -4.49e-01 |
Pos logit contributions | |
cream | +0.117 |
Fin | +0.116 |
Blue | +0.112 |
sailed | +0.104 |
Tweet | +0.102 |
Neg logit contributions | |
otion | -0.205 |
Deal | -0.201 |
spine | -0.200 |
laptop | -0.200 |
upon | -0.189 |
Token | him |
Token ID | 683 |
Feature activation | -2.31e+00 |
Pos logit contributions | |
cream | +0.013 |
Fin | +0.012 |
Blue | +0.012 |
aw | +0.010 |
Tweet | +0.010 |
Neg logit contributions | |
spine | -0.039 |
otion | -0.038 |
laptop | -0.038 |
upon | -0.038 |
ancer | -0.037 |
Token | about |
Token ID | 546 |
Feature activation | -9.61e-01 |
Pos logit contributions | |
cream | +0.096 |
Blue | +0.089 |
Fin | +0.088 |
aw | +0.080 |
Red | +0.078 |
Neg logit contributions | |
otion | -0.171 |
spine | -0.165 |
upon | -0.163 |
laptop | -0.163 |
Deal | -0.162 |