Token | Lily |
Token ID | 20037 |
Feature activation | +1.74e+00 |
Pos logit contributions | |
iflower | +0.007 |
pron | +0.006 |
wl | +0.005 |
live | +0.005 |
tto | +0.005 |
Neg logit contributions | |
seen | -0.006 |
hole | -0.005 |
towards | -0.005 |
keeping | -0.005 |
stations | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | +3.40e-02 |
Pos logit contributions | |
iflower | +0.062 |
pron | +0.056 |
live | +0.051 |
wl | +0.049 |
innie | +0.048 |
Neg logit contributions | |
seen | -0.038 |
keeping | -0.037 |
stations | -0.037 |
hole | -0.037 |
contain | -0.037 |
Token | She |
Token ID | 1375 |
Feature activation | +1.18e+00 |
Pos logit contributions | |
iflower | +0.001 |
pron | +0.001 |
live | +0.001 |
wl | +0.001 |
ama | +0.001 |
Neg logit contributions | |
seen | -0.001 |
hole | -0.001 |
towards | -0.001 |
keeping | -0.001 |
stations | -0.001 |
Token | loved |
Token ID | 6151 |
Feature activation | +9.64e-01 |
Pos logit contributions | |
iflower | +0.041 |
ama | +0.039 |
pron | +0.038 |
omm | +0.037 |
innie | +0.036 |
Neg logit contributions | |
seen | -0.026 |
want | -0.021 |
like | -0.019 |
ets | -0.018 |
towards | -0.018 |
Token | l |
Token ID | 300 |
Feature activation | +2.77e-01 |
Pos logit contributions | |
iflower | +0.038 |
pron | +0.036 |
live | +0.035 |
ama | +0.034 |
wl | +0.033 |
Neg logit contributions | |
seen | -0.018 |
keeping | -0.016 |
umb | -0.015 |
crashing | -0.015 |
stations | -0.015 |
Token | oll |
Token ID | 692 |
Feature activation | +8.63e+00 |
Pos logit contributions | |
iflower | +0.009 |
pron | +0.009 |
lf | +0.007 |
very | +0.007 |
wl | +0.006 |
Neg logit contributions | |
seen | -0.008 |
contain | -0.007 |
hole | -0.007 |
ales | -0.007 |
vious | -0.007 |
Token | ip |
Token ID | 541 |
Feature activation | +5.68e+00 |
Pos logit contributions | |
pron | +0.165 |
iflower | +0.150 |
Mon | +0.127 |
live | +0.121 |
Me | +0.121 |
Neg logit contributions | |
seen | -0.293 |
ales | -0.285 |
permission | -0.273 |
owing | -0.271 |
ously | -0.271 |
Token | ops |
Token ID | 2840 |
Feature activation | +1.67e+00 |
Pos logit contributions | |
pron | +0.160 |
iflower | +0.151 |
innie | +0.118 |
ama | +0.117 |
very | +0.114 |
Neg logit contributions | |
ales | -0.149 |
ets | -0.147 |
seen | -0.145 |
officers | -0.142 |
argued | -0.141 |
Token | , |
Token ID | 11 |
Feature activation | +2.62e-01 |
Pos logit contributions | |
iflower | +0.064 |
pron | +0.061 |
live | +0.055 |
innie | +0.055 |
wl | +0.054 |
Neg logit contributions | |
seen | -0.031 |
hole | -0.028 |
keeping | -0.028 |
contain | -0.027 |
stations | -0.027 |
Token | especially |
Token ID | 2592 |
Feature activation | +1.41e+00 |
Pos logit contributions | |
iflower | +0.010 |
pron | +0.009 |
live | +0.009 |
ama | +0.008 |
wl | +0.008 |
Neg logit contributions | |
seen | -0.005 |
hole | -0.005 |
stations | -0.004 |
crashing | -0.004 |
contain | -0.004 |
Token | the |
Token ID | 262 |
Feature activation | -1.19e-01 |
Pos logit contributions | |
iflower | +0.048 |
pron | +0.047 |
ama | +0.045 |
wl | +0.043 |
oppy | +0.041 |
Neg logit contributions | |
seen | -0.029 |
towards | -0.029 |
like | -0.028 |
crashing | -0.027 |
ked | -0.026 |
Token | Lily |
Token ID | 20037 |
Feature activation | +1.83e+00 |
Pos logit contributions | |
iflower | +0.003 |
pron | +0.003 |
wl | +0.003 |
live | +0.003 |
tto | +0.003 |
Neg logit contributions | |
seen | -0.003 |
hole | -0.003 |
keeping | -0.003 |
towards | -0.003 |
contain | -0.003 |
Token | . |
Token ID | 13 |
Feature activation | +4.11e-02 |
Pos logit contributions | |
iflower | +0.064 |
pron | +0.058 |
live | +0.053 |
wl | +0.051 |
innie | +0.051 |
Neg logit contributions | |
seen | -0.040 |
keeping | -0.040 |
stations | -0.039 |
hole | -0.039 |
officers | -0.039 |
Token | She |
Token ID | 1375 |
Feature activation | +1.28e+00 |
Pos logit contributions | |
iflower | +0.002 |
pron | +0.001 |
live | +0.001 |
wl | +0.001 |
ama | +0.001 |
Neg logit contributions | |
seen | -0.001 |
hole | -0.001 |
towards | -0.001 |
stations | -0.001 |
keeping | -0.001 |
Token | loved |
Token ID | 6151 |
Feature activation | +1.01e+00 |
Pos logit contributions | |
iflower | +0.044 |
ama | +0.041 |
pron | +0.041 |
omm | +0.038 |
innie | +0.037 |
Neg logit contributions | |
seen | -0.029 |
want | -0.024 |
like | -0.022 |
towards | -0.021 |
keeping | -0.021 |
Token | l |
Token ID | 300 |
Feature activation | +2.39e-01 |
Pos logit contributions | |
iflower | +0.041 |
pron | +0.038 |
live | +0.037 |
ama | +0.035 |
wl | +0.035 |
Neg logit contributions | |
seen | -0.019 |
keeping | -0.017 |
umb | -0.016 |
stations | -0.016 |
crashing | -0.016 |
Token | oll |
Token ID | 692 |
Feature activation | +8.56e+00 |
Pos logit contributions | |
iflower | +0.008 |
pron | +0.008 |
lf | +0.006 |
very | +0.006 |
wl | +0.006 |
Neg logit contributions | |
seen | -0.006 |
contain | -0.006 |
hole | -0.006 |
ets | -0.006 |
vious | -0.006 |
Token | ip |
Token ID | 541 |
Feature activation | +5.55e+00 |
Pos logit contributions | |
pron | +0.159 |
iflower | +0.144 |
Mon | +0.125 |
Me | +0.119 |
live | +0.116 |
Neg logit contributions | |
seen | -0.292 |
ales | -0.284 |
ets | -0.273 |
permission | -0.272 |
owing | -0.272 |
Token | ops |
Token ID | 2840 |
Feature activation | +1.51e+00 |
Pos logit contributions | |
pron | +0.159 |
iflower | +0.150 |
ama | +0.117 |
innie | +0.116 |
very | +0.114 |
Neg logit contributions | |
ales | -0.145 |
ets | -0.143 |
seen | -0.140 |
officers | -0.138 |
keeping | -0.135 |
Token | and |
Token ID | 290 |
Feature activation | +1.86e+00 |
Pos logit contributions | |
iflower | +0.058 |
pron | +0.056 |
innie | +0.050 |
live | +0.049 |
wl | +0.049 |
Neg logit contributions | |
seen | -0.029 |
hole | -0.026 |
keeping | -0.025 |
contain | -0.024 |
ales | -0.024 |
Token | would |
Token ID | 561 |
Feature activation | -3.04e-01 |
Pos logit contributions | |
iflower | +0.057 |
pron | +0.049 |
ama | +0.047 |
live | +0.046 |
wl | +0.045 |
Neg logit contributions | |
seen | -0.052 |
argued | -0.046 |
crashing | -0.045 |
stations | -0.045 |
officers | -0.044 |
Token | always |
Token ID | 1464 |
Feature activation | +1.07e+00 |
Pos logit contributions | |
seen | +0.008 |
contain | +0.007 |
crashing | +0.007 |
stations | +0.007 |
hole | +0.007 |
Neg logit contributions | |
iflower | -0.010 |
pron | -0.009 |
live | -0.008 |
ama | -0.008 |
innie | -0.008 |
Token | little |
Token ID | 1310 |
Feature activation | +8.25e-01 |
Pos logit contributions | |
iflower | +0.032 |
live | +0.031 |
wl | +0.030 |
innie | +0.028 |
tto | +0.028 |
Neg logit contributions | |
crashing | -0.016 |
want | -0.016 |
brake | -0.014 |
officers | -0.014 |
steering | -0.014 |
Token | girl |
Token ID | 2576 |
Feature activation | +1.07e+00 |
Pos logit contributions | |
iflower | +0.028 |
pron | +0.025 |
live | +0.025 |
wl | +0.023 |
innie | +0.023 |
Neg logit contributions | |
seen | -0.019 |
officers | -0.018 |
crashing | -0.017 |
hole | -0.017 |
stations | -0.017 |
Token | who |
Token ID | 508 |
Feature activation | +1.96e+00 |
Pos logit contributions | |
iflower | +0.035 |
pron | +0.032 |
live | +0.029 |
innie | +0.027 |
ama | +0.027 |
Neg logit contributions | |
seen | -0.028 |
hole | -0.026 |
keeping | -0.025 |
stations | -0.025 |
crashing | -0.024 |
Token | loved |
Token ID | 6151 |
Feature activation | +6.00e-01 |
Pos logit contributions | |
iflower | +0.060 |
pron | +0.054 |
ama | +0.049 |
innie | +0.046 |
live | +0.046 |
Neg logit contributions | |
seen | -0.059 |
want | -0.052 |
keeping | -0.050 |
towards | -0.048 |
hole | -0.047 |
Token | l |
Token ID | 300 |
Feature activation | +1.93e-01 |
Pos logit contributions | |
iflower | +0.022 |
pron | +0.021 |
live | +0.019 |
wl | +0.019 |
ama | +0.018 |
Neg logit contributions | |
seen | -0.013 |
keeping | -0.012 |
stations | -0.012 |
crashing | -0.011 |
officers | -0.011 |
Token | oll |
Token ID | 692 |
Feature activation | +8.51e+00 |
Pos logit contributions | |
iflower | +0.006 |
pron | +0.005 |
wl | +0.004 |
live | +0.004 |
innie | +0.004 |
Neg logit contributions | |
seen | -0.006 |
hole | -0.005 |
contain | -0.005 |
keeping | -0.005 |
ales | -0.005 |
Token | ip |
Token ID | 541 |
Feature activation | +5.50e+00 |
Pos logit contributions | |
iflower | +0.144 |
pron | +0.140 |
live | +0.109 |
wl | +0.096 |
innie | +0.095 |
Neg logit contributions | |
seen | -0.313 |
ales | -0.297 |
hole | -0.294 |
officers | -0.291 |
permission | -0.290 |
Token | ops |
Token ID | 2840 |
Feature activation | +1.54e+00 |
Pos logit contributions | |
iflower | +0.141 |
pron | +0.140 |
innie | +0.107 |
live | +0.104 |
ama | +0.102 |
Neg logit contributions | |
seen | -0.160 |
officers | -0.158 |
ales | -0.157 |
keeping | -0.155 |
hole | -0.152 |
Token | . |
Token ID | 13 |
Feature activation | -7.00e-02 |
Pos logit contributions | |
iflower | +0.058 |
pron | +0.054 |
live | +0.048 |
ama | +0.046 |
innie | +0.046 |
Neg logit contributions | |
seen | -0.032 |
hole | -0.030 |
keeping | -0.030 |
stations | -0.029 |
brake | -0.028 |
Token | One |
Token ID | 1881 |
Feature activation | -8.06e-02 |
Pos logit contributions | |
seen | +0.002 |
keeping | +0.001 |
hole | +0.001 |
towards | +0.001 |
stations | +0.001 |
Neg logit contributions | |
iflower | -0.003 |
pron | -0.002 |
live | -0.002 |
wl | -0.002 |
tto | -0.002 |
Token | day |
Token ID | 1110 |
Feature activation | -4.91e-01 |
Pos logit contributions | |
seen | +0.002 |
want | +0.002 |
towards | +0.002 |
eye | +0.002 |
contain | +0.002 |
Neg logit contributions | |
iflower | -0.002 |
pron | -0.002 |
wl | -0.002 |
innie | -0.002 |
live | -0.002 |
Token | Lily |
Token ID | 20037 |
Feature activation | +1.77e+00 |
Pos logit contributions | |
iflower | +0.004 |
pron | +0.004 |
wl | +0.003 |
tto | +0.003 |
live | +0.003 |
Neg logit contributions | |
seen | -0.004 |
hole | -0.003 |
keeping | -0.003 |
towards | -0.003 |
stations | -0.003 |
Token | . |
Token ID | 13 |
Feature activation | -1.62e-01 |
Pos logit contributions | |
iflower | +0.062 |
pron | +0.056 |
live | +0.051 |
wl | +0.049 |
innie | +0.049 |
Neg logit contributions | |
seen | -0.039 |
keeping | -0.038 |
stations | -0.038 |
contain | -0.037 |
hole | -0.037 |
Token | She |
Token ID | 1375 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
seen | +0.004 |
hole | +0.004 |
stations | +0.003 |
keeping | +0.003 |
towards | +0.003 |
Neg logit contributions | |
iflower | -0.006 |
pron | -0.005 |
live | -0.005 |
wl | -0.005 |
ama | -0.005 |
Token | loved |
Token ID | 6151 |
Feature activation | +1.01e+00 |
Pos logit contributions | |
iflower | +0.041 |
ama | +0.039 |
pron | +0.039 |
omm | +0.036 |
innie | +0.035 |
Neg logit contributions | |
seen | -0.025 |
want | -0.021 |
like | -0.019 |
keeping | -0.018 |
towards | -0.018 |
Token | l |
Token ID | 300 |
Feature activation | +2.00e-01 |
Pos logit contributions | |
iflower | +0.040 |
pron | +0.038 |
live | +0.036 |
ama | +0.035 |
wl | +0.035 |
Neg logit contributions | |
seen | -0.019 |
keeping | -0.017 |
umb | -0.016 |
crashing | -0.016 |
stations | -0.016 |
Token | oll |
Token ID | 692 |
Feature activation | +8.51e+00 |
Pos logit contributions | |
iflower | +0.007 |
pron | +0.006 |
lf | +0.005 |
very | +0.005 |
wl | +0.005 |
Neg logit contributions | |
seen | -0.005 |
contain | -0.005 |
ets | -0.005 |
hole | -0.005 |
vious | -0.005 |
Token | ip |
Token ID | 541 |
Feature activation | +5.44e+00 |
Pos logit contributions | |
pron | +0.163 |
iflower | +0.150 |
Mon | +0.132 |
live | +0.122 |
Me | +0.121 |
Neg logit contributions | |
seen | -0.285 |
ales | -0.276 |
permission | -0.265 |
owing | -0.265 |
ets | -0.264 |
Token | ops |
Token ID | 2840 |
Feature activation | +1.38e+00 |
Pos logit contributions | |
pron | +0.154 |
iflower | +0.145 |
innie | +0.114 |
ama | +0.113 |
very | +0.109 |
Neg logit contributions | |
ales | -0.145 |
ets | -0.142 |
seen | -0.138 |
officers | -0.138 |
keeping | -0.137 |
Token | and |
Token ID | 290 |
Feature activation | +1.87e+00 |
Pos logit contributions | |
iflower | +0.053 |
pron | +0.051 |
live | +0.045 |
innie | +0.045 |
wl | +0.045 |
Neg logit contributions | |
seen | -0.027 |
hole | -0.024 |
keeping | -0.024 |
stations | -0.023 |
contain | -0.023 |
Token | always |
Token ID | 1464 |
Feature activation | +1.25e+00 |
Pos logit contributions | |
iflower | +0.055 |
pron | +0.048 |
ama | +0.045 |
live | +0.044 |
wl | +0.043 |
Neg logit contributions | |
seen | -0.055 |
crashing | -0.048 |
stations | -0.048 |
argued | -0.047 |
keeping | -0.047 |
Token | had |
Token ID | 550 |
Feature activation | +2.44e+00 |
Pos logit contributions | |
iflower | +0.040 |
pron | +0.036 |
live | +0.033 |
ama | +0.032 |
innie | +0.031 |
Neg logit contributions | |
seen | -0.034 |
hole | -0.031 |
keeping | -0.031 |
stations | -0.029 |
contain | -0.029 |
Token | Lily |
Token ID | 20037 |
Feature activation | +1.49e+00 |
Pos logit contributions | |
seen | +0.001 |
keeping | +0.001 |
hole | +0.001 |
towards | +0.001 |
contain | +0.001 |
Neg logit contributions | |
iflower | -0.001 |
pron | -0.001 |
wl | -0.001 |
live | -0.001 |
tto | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | -1.93e-01 |
Pos logit contributions | |
iflower | +0.051 |
pron | +0.046 |
live | +0.042 |
wl | +0.040 |
innie | +0.040 |
Neg logit contributions | |
seen | -0.033 |
keeping | -0.033 |
contain | -0.033 |
hole | -0.033 |
officers | -0.033 |
Token | She |
Token ID | 1375 |
Feature activation | +1.10e+00 |
Pos logit contributions | |
seen | +0.005 |
hole | +0.004 |
towards | +0.004 |
stations | +0.004 |
keeping | +0.004 |
Neg logit contributions | |
iflower | -0.007 |
pron | -0.006 |
live | -0.006 |
wl | -0.006 |
ama | -0.005 |
Token | loved |
Token ID | 6151 |
Feature activation | +8.25e-01 |
Pos logit contributions | |
iflower | +0.038 |
pron | +0.036 |
ama | +0.035 |
omm | +0.033 |
innie | +0.032 |
Neg logit contributions | |
seen | -0.024 |
want | -0.021 |
like | -0.018 |
keeping | -0.018 |
towards | -0.018 |
Token | l |
Token ID | 300 |
Feature activation | +1.27e-01 |
Pos logit contributions | |
iflower | +0.033 |
pron | +0.031 |
live | +0.030 |
ama | +0.029 |
wl | +0.028 |
Neg logit contributions | |
seen | -0.015 |
keeping | -0.014 |
umb | -0.013 |
stations | -0.013 |
crashing | -0.013 |
Token | oll |
Token ID | 692 |
Feature activation | +8.46e+00 |
Pos logit contributions | |
iflower | +0.004 |
pron | +0.004 |
wl | +0.003 |
lf | +0.003 |
live | +0.003 |
Neg logit contributions | |
seen | -0.004 |
hole | -0.003 |
contain | -0.003 |
ales | -0.003 |
keeping | -0.003 |
Token | ip |
Token ID | 541 |
Feature activation | +5.53e+00 |
Pos logit contributions | |
pron | +0.153 |
iflower | +0.144 |
live | +0.113 |
Mon | +0.109 |
Me | +0.103 |
Neg logit contributions | |
seen | -0.303 |
ales | -0.295 |
permission | -0.283 |
hole | -0.282 |
ets | -0.282 |
Token | ops |
Token ID | 2840 |
Feature activation | +1.49e+00 |
Pos logit contributions | |
pron | +0.156 |
iflower | +0.148 |
innie | +0.114 |
ama | +0.112 |
very | +0.110 |
Neg logit contributions | |
ales | -0.147 |
ets | -0.145 |
seen | -0.143 |
officers | -0.141 |
keeping | -0.140 |
Token | and |
Token ID | 290 |
Feature activation | +1.97e+00 |
Pos logit contributions | |
iflower | +0.058 |
pron | +0.055 |
live | +0.049 |
innie | +0.049 |
wl | +0.048 |
Neg logit contributions | |
seen | -0.029 |
hole | -0.026 |
keeping | -0.026 |
stations | -0.025 |
contain | -0.024 |
Token | always |
Token ID | 1464 |
Feature activation | +1.10e+00 |
Pos logit contributions | |
iflower | +0.057 |
pron | +0.050 |
live | +0.045 |
ama | +0.045 |
wl | +0.044 |
Neg logit contributions | |
seen | -0.058 |
stations | -0.052 |
crashing | -0.051 |
keeping | -0.051 |
argued | -0.051 |
Token | had |
Token ID | 550 |
Feature activation | +2.59e+00 |
Pos logit contributions | |
iflower | +0.036 |
pron | +0.033 |
live | +0.029 |
ama | +0.028 |
innie | +0.028 |
Neg logit contributions | |
seen | -0.029 |
hole | -0.026 |
keeping | -0.026 |
stations | -0.026 |
contain | -0.025 |
Token | there |
Token ID | 612 |
Feature activation | +1.21e+00 |
Pos logit contributions | |
iflower | +0.058 |
pron | +0.054 |
live | +0.045 |
innie | +0.045 |
wl | +0.043 |
Neg logit contributions | |
seen | -0.035 |
hole | -0.032 |
towards | -0.030 |
ago | -0.030 |
crashing | -0.030 |
Token | was |
Token ID | 373 |
Feature activation | +1.26e+00 |
Pos logit contributions | |
iflower | +0.041 |
pron | +0.039 |
innie | +0.032 |
wl | +0.031 |
live | +0.030 |
Neg logit contributions | |
seen | -0.034 |
hole | -0.029 |
gasp | -0.027 |
crashing | -0.027 |
towards | -0.027 |
Token | a |
Token ID | 257 |
Feature activation | +1.35e+00 |
Pos logit contributions | |
iflower | +0.044 |
pron | +0.041 |
live | +0.036 |
wl | +0.035 |
ama | +0.035 |
Neg logit contributions | |
seen | -0.031 |
keeping | -0.027 |
hole | -0.026 |
stations | -0.026 |
crashing | -0.025 |
Token | great |
Token ID | 1049 |
Feature activation | +2.74e+00 |
Pos logit contributions | |
iflower | +0.049 |
live | +0.047 |
wl | +0.045 |
innie | +0.042 |
tto | +0.042 |
Neg logit contributions | |
want | -0.022 |
crashing | -0.022 |
brake | -0.019 |
Run | -0.018 |
seen | -0.018 |
Token | l |
Token ID | 300 |
Feature activation | -2.31e-01 |
Pos logit contributions | |
iflower | +0.092 |
pron | +0.080 |
live | +0.079 |
wl | +0.077 |
innie | +0.074 |
Neg logit contributions | |
want | -0.059 |
seen | -0.059 |
crashing | -0.057 |
officers | -0.053 |
eye | -0.053 |
Token | oll |
Token ID | 692 |
Feature activation | +8.45e+00 |
Pos logit contributions | |
seen | +0.005 |
hole | +0.005 |
ets | +0.004 |
contain | +0.004 |
vious | +0.004 |
Neg logit contributions | |
pron | -0.008 |
iflower | -0.008 |
innie | -0.007 |
tto | -0.006 |
very | -0.006 |
Token | ipop |
Token ID | 42800 |
Feature activation | +3.64e+00 |
Pos logit contributions | |
pron | +0.180 |
iflower | +0.180 |
Mon | +0.166 |
Me | +0.164 |
Rich | +0.160 |
Neg logit contributions | |
seen | -0.251 |
ales | -0.223 |
hole | -0.219 |
owing | -0.219 |
ously | -0.217 |
Token | . |
Token ID | 13 |
Feature activation | -4.56e-02 |
Pos logit contributions | |
iflower | +0.140 |
pron | +0.134 |
live | +0.122 |
innie | +0.121 |
wl | +0.118 |
Neg logit contributions | |
seen | -0.061 |
hole | -0.058 |
contain | -0.053 |
crashing | -0.053 |
officers | -0.052 |
Token | It |
Token ID | 632 |
Feature activation | +2.81e+00 |
Pos logit contributions | |
seen | +0.001 |
keeping | +0.001 |
hole | +0.001 |
towards | +0.001 |
crashing | +0.001 |
Neg logit contributions | |
iflower | -0.002 |
pron | -0.002 |
live | -0.001 |
wl | -0.001 |
tto | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | -4.95e-01 |
Pos logit contributions | |
ama | +0.100 |
iflower | +0.097 |
pron | +0.096 |
innie | +0.086 |
irl | +0.085 |
Neg logit contributions | |
seen | -0.051 |
like | -0.048 |
want | -0.042 |
towards | -0.041 |
hid | -0.035 |
Token | pink |
Token ID | 11398 |
Feature activation | +2.34e+00 |
Pos logit contributions | |
seen | +0.012 |
crashing | +0.009 |
hole | +0.009 |
towards | +0.009 |
contain | +0.009 |
Neg logit contributions | |
iflower | -0.018 |
pron | -0.016 |
wl | -0.015 |
live | -0.015 |
ama | -0.015 |
Token | John |
Token ID | 1757 |
Feature activation | +1.88e+00 |
Pos logit contributions | |
seen | +0.036 |
hole | +0.035 |
stations | +0.034 |
keeping | +0.034 |
contain | +0.032 |
Neg logit contributions | |
iflower | -0.043 |
pron | -0.040 |
live | -0.036 |
innie | -0.035 |
ama | -0.035 |
Token | . |
Token ID | 13 |
Feature activation | -3.74e-01 |
Pos logit contributions | |
iflower | +0.066 |
pron | +0.061 |
live | +0.054 |
innie | +0.053 |
wl | +0.052 |
Neg logit contributions | |
keeping | -0.041 |
seen | -0.041 |
stations | -0.041 |
hole | -0.041 |
contain | -0.040 |
Token | John |
Token ID | 1757 |
Feature activation | +2.23e+00 |
Pos logit contributions | |
seen | +0.008 |
hole | +0.008 |
keeping | +0.008 |
stations | +0.008 |
contain | +0.007 |
Neg logit contributions | |
iflower | -0.013 |
pron | -0.012 |
live | -0.011 |
innie | -0.011 |
ama | -0.011 |
Token | loved |
Token ID | 6151 |
Feature activation | +8.05e-01 |
Pos logit contributions | |
iflower | +0.066 |
pron | +0.063 |
ama | +0.055 |
live | +0.054 |
innie | +0.053 |
Neg logit contributions | |
seen | -0.066 |
hole | -0.056 |
keeping | -0.055 |
want | -0.054 |
stations | -0.054 |
Token | l |
Token ID | 300 |
Feature activation | +1.40e-01 |
Pos logit contributions | |
iflower | +0.030 |
pron | +0.027 |
live | +0.026 |
ama | +0.025 |
wl | +0.025 |
Neg logit contributions | |
seen | -0.018 |
keeping | -0.016 |
stations | -0.016 |
crashing | -0.015 |
officers | -0.015 |
Token | oll |
Token ID | 692 |
Feature activation | +8.34e+00 |
Pos logit contributions | |
iflower | +0.004 |
pron | +0.004 |
live | +0.003 |
wl | +0.003 |
innie | +0.003 |
Neg logit contributions | |
seen | -0.004 |
hole | -0.004 |
keeping | -0.004 |
contain | -0.004 |
ales | -0.004 |
Token | ip |
Token ID | 541 |
Feature activation | +5.45e+00 |
Pos logit contributions | |
pron | +0.134 |
iflower | +0.132 |
live | +0.102 |
Mon | +0.089 |
wl | +0.088 |
Neg logit contributions | |
seen | -0.313 |
ales | -0.302 |
hole | -0.293 |
ets | -0.292 |
officers | -0.290 |
Token | ops |
Token ID | 2840 |
Feature activation | +1.43e+00 |
Pos logit contributions | |
pron | +0.139 |
iflower | +0.137 |
innie | +0.105 |
ama | +0.103 |
live | +0.102 |
Neg logit contributions | |
seen | -0.159 |
ales | -0.157 |
keeping | -0.155 |
officers | -0.155 |
ets | -0.153 |
Token | ! |
Token ID | 0 |
Feature activation | -4.05e-01 |
Pos logit contributions | |
iflower | +0.055 |
pron | +0.051 |
live | +0.047 |
innie | +0.045 |
wl | +0.045 |
Neg logit contributions | |
seen | -0.028 |
hole | -0.026 |
keeping | -0.026 |
stations | -0.025 |
contain | -0.024 |
Token | One |
Token ID | 1881 |
Feature activation | -1.47e-01 |
Pos logit contributions | |
seen | +0.010 |
hole | +0.010 |
keeping | +0.010 |
stations | +0.010 |
crashing | +0.009 |
Neg logit contributions | |
iflower | -0.013 |
pron | -0.012 |
live | -0.011 |
wl | -0.010 |
ama | -0.010 |
Token | day |
Token ID | 1110 |
Feature activation | -5.72e-01 |
Pos logit contributions | |
seen | +0.004 |
want | +0.004 |
towards | +0.004 |
eye | +0.004 |
contain | +0.004 |
Neg logit contributions | |
iflower | -0.004 |
pron | -0.004 |
wl | -0.003 |
innie | -0.003 |
live | -0.003 |
Token | Lily |
Token ID | 20037 |
Feature activation | +1.52e+00 |
Pos logit contributions | |
seen | +0.001 |
keeping | +0.001 |
towards | +0.001 |
hole | +0.001 |
contain | +0.001 |
Neg logit contributions | |
iflower | -0.002 |
pron | -0.001 |
wl | -0.001 |
tto | -0.001 |
live | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | -2.27e-01 |
Pos logit contributions | |
iflower | +0.053 |
pron | +0.048 |
live | +0.044 |
wl | +0.042 |
innie | +0.042 |
Neg logit contributions | |
seen | -0.033 |
keeping | -0.033 |
stations | -0.033 |
hole | -0.033 |
officers | -0.032 |
Token | She |
Token ID | 1375 |
Feature activation | +1.12e+00 |
Pos logit contributions | |
seen | +0.005 |
hole | +0.005 |
stations | +0.005 |
towards | +0.005 |
keeping | +0.005 |
Neg logit contributions | |
iflower | -0.008 |
pron | -0.007 |
live | -0.007 |
wl | -0.007 |
ama | -0.006 |
Token | loved |
Token ID | 6151 |
Feature activation | +8.79e-01 |
Pos logit contributions | |
iflower | +0.039 |
ama | +0.037 |
pron | +0.036 |
omm | +0.035 |
innie | +0.033 |
Neg logit contributions | |
seen | -0.025 |
want | -0.020 |
like | -0.018 |
towards | -0.018 |
ets | -0.017 |
Token | l |
Token ID | 300 |
Feature activation | +3.81e-02 |
Pos logit contributions | |
iflower | +0.035 |
pron | +0.033 |
live | +0.032 |
ama | +0.031 |
wl | +0.031 |
Neg logit contributions | |
seen | -0.016 |
keeping | -0.014 |
stations | -0.014 |
crashing | -0.014 |
umb | -0.014 |
Token | oll |
Token ID | 692 |
Feature activation | +8.32e+00 |
Pos logit contributions | |
iflower | +0.001 |
pron | +0.001 |
lf | +0.001 |
wl | +0.001 |
very | +0.001 |
Neg logit contributions | |
seen | -0.001 |
hole | -0.001 |
contain | -0.001 |
vious | -0.001 |
ales | -0.001 |
Token | ip |
Token ID | 541 |
Feature activation | +5.33e+00 |
Pos logit contributions | |
pron | +0.147 |
iflower | +0.144 |
live | +0.112 |
Mon | +0.106 |
omm | +0.101 |
Neg logit contributions | |
seen | -0.294 |
ales | -0.279 |
ked | -0.276 |
ously | -0.274 |
hole | -0.273 |
Token | ops |
Token ID | 2840 |
Feature activation | +1.34e+00 |
Pos logit contributions | |
pron | +0.154 |
iflower | +0.147 |
innie | +0.113 |
ama | +0.111 |
omm | +0.110 |
Neg logit contributions | |
ales | -0.135 |
ets | -0.134 |
officers | -0.132 |
argued | -0.131 |
seen | -0.130 |
Token | and |
Token ID | 290 |
Feature activation | +1.83e+00 |
Pos logit contributions | |
iflower | +0.052 |
pron | +0.050 |
live | +0.045 |
innie | +0.044 |
wl | +0.044 |
Neg logit contributions | |
seen | -0.025 |
hole | -0.023 |
keeping | -0.022 |
stations | -0.022 |
contain | -0.022 |
Token | ate |
Token ID | 15063 |
Feature activation | +2.09e+00 |
Pos logit contributions | |
iflower | +0.052 |
pron | +0.046 |
live | +0.042 |
ama | +0.041 |
wl | +0.040 |
Neg logit contributions | |
seen | -0.055 |
stations | -0.050 |
keeping | -0.049 |
crashing | -0.049 |
officers | -0.049 |
Token | them |
Token ID | 606 |
Feature activation | -3.39e-01 |
Pos logit contributions | |
iflower | +0.083 |
pron | +0.080 |
wl | +0.076 |
live | +0.076 |
ama | +0.074 |
Neg logit contributions | |
towards | -0.031 |
crashing | -0.031 |
seen | -0.030 |
umb | -0.029 |
stations | -0.028 |
Token | named |
Token ID | 3706 |
Feature activation | -1.75e-01 |
Pos logit contributions | |
iflower | +0.021 |
pron | +0.019 |
live | +0.017 |
innie | +0.016 |
ama | +0.016 |
Neg logit contributions | |
seen | -0.017 |
hole | -0.015 |
keeping | -0.015 |
stations | -0.015 |
crashing | -0.014 |
Token | Lily |
Token ID | 20037 |
Feature activation | +1.35e+00 |
Pos logit contributions | |
seen | +0.005 |
keeping | +0.005 |
hole | +0.004 |
towards | +0.004 |
contain | +0.004 |
Neg logit contributions | |
iflower | -0.006 |
pron | -0.005 |
wl | -0.004 |
live | -0.004 |
tto | -0.004 |
Token | who |
Token ID | 508 |
Feature activation | +1.65e+00 |
Pos logit contributions | |
iflower | +0.046 |
pron | +0.041 |
live | +0.038 |
wl | +0.036 |
innie | +0.036 |
Neg logit contributions | |
seen | -0.030 |
keeping | -0.030 |
contain | -0.030 |
stations | -0.030 |
hole | -0.030 |
Token | loved |
Token ID | 6151 |
Feature activation | +5.91e-01 |
Pos logit contributions | |
iflower | +0.057 |
pron | +0.052 |
ama | +0.050 |
omm | +0.047 |
opus | +0.046 |
Neg logit contributions | |
seen | -0.041 |
want | -0.038 |
like | -0.032 |
towards | -0.030 |
hope | -0.029 |
Token | l |
Token ID | 300 |
Feature activation | +8.21e-03 |
Pos logit contributions | |
iflower | +0.022 |
pron | +0.021 |
live | +0.019 |
ama | +0.019 |
wl | +0.019 |
Neg logit contributions | |
seen | -0.012 |
keeping | -0.012 |
stations | -0.011 |
crashing | -0.011 |
officers | -0.011 |
Token | oll |
Token ID | 692 |
Feature activation | +8.19e+00 |
Pos logit contributions | |
iflower | +0.000 |
pron | +0.000 |
live | +0.000 |
innie | +0.000 |
wl | +0.000 |
Neg logit contributions | |
seen | -0.000 |
hole | -0.000 |
keeping | -0.000 |
contain | -0.000 |
ales | -0.000 |
Token | ip |
Token ID | 541 |
Feature activation | +5.11e+00 |
Pos logit contributions | |
pron | +0.129 |
iflower | +0.125 |
live | +0.095 |
innie | +0.080 |
wl | +0.079 |
Neg logit contributions | |
seen | -0.315 |
ales | -0.302 |
hole | -0.298 |
permission | -0.293 |
ously | -0.292 |
Token | ops |
Token ID | 2840 |
Feature activation | +1.12e+00 |
Pos logit contributions | |
pron | +0.132 |
iflower | +0.130 |
innie | +0.098 |
live | +0.095 |
ama | +0.094 |
Neg logit contributions | |
seen | -0.147 |
ales | -0.145 |
officers | -0.144 |
keeping | -0.142 |
ets | -0.140 |
Token | . |
Token ID | 13 |
Feature activation | -1.44e-01 |
Pos logit contributions | |
iflower | +0.043 |
pron | +0.040 |
live | +0.036 |
innie | +0.035 |
ama | +0.035 |
Neg logit contributions | |
seen | -0.023 |
hole | -0.021 |
keeping | -0.021 |
stations | -0.020 |
officers | -0.019 |
Token | She |
Token ID | 1375 |
Feature activation | +1.36e+00 |
Pos logit contributions | |
seen | +0.003 |
hole | +0.003 |
stations | +0.003 |
keeping | +0.003 |
towards | +0.003 |
Neg logit contributions | |
iflower | -0.005 |
pron | -0.005 |
live | -0.004 |
wl | -0.004 |
tto | -0.004 |
Token | would |
Token ID | 561 |
Feature activation | -1.21e+00 |
Pos logit contributions | |
iflower | +0.053 |
ama | +0.049 |
pron | +0.045 |
omm | +0.044 |
opus | +0.044 |
Neg logit contributions | |
seen | -0.028 |
argued | -0.022 |
want | -0.021 |
hid | -0.018 |
ets | -0.018 |
Token | to |
Token ID | 284 |
Feature activation | +3.02e+00 |
Pos logit contributions | |
iflower | +0.019 |
pron | +0.019 |
live | +0.017 |
wl | +0.017 |
ama | +0.016 |
Neg logit contributions | |
seen | -0.015 |
keeping | -0.013 |
towards | -0.013 |
crashing | -0.013 |
hole | -0.012 |
Token | walk |
Token ID | 2513 |
Feature activation | +2.92e-01 |
Pos logit contributions | |
iflower | +0.109 |
pron | +0.100 |
live | +0.090 |
wl | +0.087 |
ama | +0.087 |
Neg logit contributions | |
seen | -0.069 |
hole | -0.062 |
keeping | -0.062 |
contain | -0.060 |
stations | -0.060 |
Token | and |
Token ID | 290 |
Feature activation | +1.34e+00 |
Pos logit contributions | |
iflower | +0.009 |
pron | +0.009 |
live | +0.008 |
innie | +0.007 |
ama | +0.007 |
Neg logit contributions | |
seen | -0.008 |
hole | -0.007 |
keeping | -0.007 |
stations | -0.007 |
contain | -0.007 |
Token | give |
Token ID | 1577 |
Feature activation | +2.94e+00 |
Pos logit contributions | |
iflower | +0.045 |
pron | +0.042 |
live | +0.037 |
innie | +0.036 |
ama | +0.035 |
Neg logit contributions | |
seen | -0.032 |
hole | -0.031 |
keeping | -0.030 |
stations | -0.030 |
officers | -0.029 |
Token | l |
Token ID | 300 |
Feature activation | -5.26e-01 |
Pos logit contributions | |
iflower | +0.098 |
pron | +0.097 |
live | +0.084 |
wl | +0.082 |
innie | +0.081 |
Neg logit contributions | |
seen | -0.075 |
keeping | -0.066 |
towards | -0.064 |
officers | -0.062 |
ales | -0.060 |
Token | oll |
Token ID | 692 |
Feature activation | +8.17e+00 |
Pos logit contributions | |
seen | +0.017 |
hole | +0.017 |
keeping | +0.016 |
contain | +0.016 |
stations | +0.016 |
Neg logit contributions | |
iflower | -0.013 |
pron | -0.012 |
live | -0.010 |
wl | -0.010 |
innie | -0.010 |
Token | ip |
Token ID | 541 |
Feature activation | +5.33e+00 |
Pos logit contributions | |
iflower | +0.154 |
pron | +0.153 |
live | +0.128 |
wl | +0.119 |
innie | +0.111 |
Neg logit contributions | |
seen | -0.283 |
ales | -0.270 |
permission | -0.269 |
ets | -0.265 |
ked | -0.265 |
Token | ops |
Token ID | 2840 |
Feature activation | +1.06e+00 |
Pos logit contributions | |
iflower | +0.122 |
pron | +0.120 |
innie | +0.094 |
live | +0.093 |
wl | +0.092 |
Neg logit contributions | |
seen | -0.167 |
keeping | -0.166 |
ets | -0.162 |
ales | -0.162 |
officers | -0.160 |
Token | to |
Token ID | 284 |
Feature activation | +1.06e+00 |
Pos logit contributions | |
iflower | +0.039 |
pron | +0.036 |
live | +0.032 |
ama | +0.031 |
innie | +0.031 |
Neg logit contributions | |
seen | -0.023 |
hole | -0.022 |
stations | -0.021 |
keeping | -0.021 |
crashing | -0.020 |
Token | children |
Token ID | 1751 |
Feature activation | -1.23e+00 |
Pos logit contributions | |
iflower | +0.033 |
pron | +0.030 |
live | +0.027 |
wl | +0.026 |
innie | +0.026 |
Neg logit contributions | |
seen | -0.029 |
keeping | -0.026 |
contain | -0.026 |
stations | -0.026 |
hole | -0.025 |
Token | ." |
Token ID | 526 |
Feature activation | +7.73e-01 |
Pos logit contributions | |
seen | +0.027 |
hole | +0.025 |
stations | +0.025 |
keeping | +0.025 |
brake | +0.024 |
Neg logit contributions | |
iflower | -0.044 |
pron | -0.041 |
live | -0.037 |
ama | -0.035 |
innie | -0.035 |
Token | Lily |
Token ID | 20037 |
Feature activation | +1.49e+00 |
Pos logit contributions | |
seen | +0.002 |
towards | +0.001 |
keeping | +0.001 |
hole | +0.001 |
contain | +0.001 |
Neg logit contributions | |
iflower | -0.002 |
pron | -0.002 |
wl | -0.001 |
live | -0.001 |
tto | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | -3.18e-01 |
Pos logit contributions | |
iflower | +0.050 |
pron | +0.045 |
live | +0.041 |
innie | +0.039 |
wl | +0.039 |
Neg logit contributions | |
seen | -0.034 |
keeping | -0.034 |
hole | -0.033 |
stations | -0.033 |
contain | -0.033 |
Token | She |
Token ID | 1375 |
Feature activation | +9.94e-01 |
Pos logit contributions | |
seen | +0.008 |
hole | +0.007 |
stations | +0.007 |
towards | +0.007 |
keeping | +0.007 |
Neg logit contributions | |
iflower | -0.012 |
pron | -0.010 |
live | -0.009 |
wl | -0.009 |
ama | -0.009 |
Token | loved |
Token ID | 6151 |
Feature activation | +7.63e-01 |
Pos logit contributions | |
iflower | +0.035 |
ama | +0.032 |
pron | +0.032 |
omm | +0.029 |
innie | +0.029 |
Neg logit contributions | |
seen | -0.022 |
want | -0.019 |
towards | -0.016 |
like | -0.016 |
ets | -0.016 |
Token | l |
Token ID | 300 |
Feature activation | -5.60e-02 |
Pos logit contributions | |
iflower | +0.030 |
pron | +0.028 |
live | +0.027 |
ama | +0.026 |
wl | +0.026 |
Neg logit contributions | |
seen | -0.015 |
keeping | -0.013 |
stations | -0.013 |
umb | -0.012 |
crashing | -0.012 |
Token | oll |
Token ID | 692 |
Feature activation | +8.16e+00 |
Pos logit contributions | |
seen | +0.002 |
hole | +0.001 |
contain | +0.001 |
keeping | +0.001 |
vious | +0.001 |
Neg logit contributions | |
iflower | -0.002 |
pron | -0.002 |
wl | -0.001 |
lf | -0.001 |
live | -0.001 |
Token | ip |
Token ID | 541 |
Feature activation | +5.22e+00 |
Pos logit contributions | |
pron | +0.134 |
iflower | +0.133 |
live | +0.098 |
Mon | +0.090 |
wl | +0.087 |
Neg logit contributions | |
seen | -0.300 |
ales | -0.291 |
hole | -0.283 |
ets | -0.283 |
officers | -0.283 |
Token | ops |
Token ID | 2840 |
Feature activation | +1.21e+00 |
Pos logit contributions | |
pron | +0.145 |
iflower | +0.141 |
ama | +0.107 |
innie | +0.105 |
wl | +0.103 |
Neg logit contributions | |
ales | -0.138 |
ets | -0.138 |
officers | -0.136 |
seen | -0.135 |
argued | -0.134 |
Token | and |
Token ID | 290 |
Feature activation | +1.71e+00 |
Pos logit contributions | |
iflower | +0.046 |
pron | +0.044 |
live | +0.039 |
wl | +0.039 |
innie | +0.039 |
Neg logit contributions | |
seen | -0.024 |
hole | -0.022 |
keeping | -0.021 |
stations | -0.021 |
contain | -0.021 |
Token | always |
Token ID | 1464 |
Feature activation | +7.56e-01 |
Pos logit contributions | |
iflower | +0.049 |
pron | +0.043 |
live | +0.039 |
ama | +0.038 |
wl | +0.037 |
Neg logit contributions | |
seen | -0.051 |
stations | -0.046 |
keeping | -0.046 |
hole | -0.046 |
crashing | -0.046 |
Token | had |
Token ID | 550 |
Feature activation | +2.35e+00 |
Pos logit contributions | |
iflower | +0.025 |
pron | +0.022 |
live | +0.020 |
ama | +0.020 |
innie | +0.019 |
Neg logit contributions | |
seen | -0.020 |
hole | -0.018 |
keeping | -0.018 |
contain | -0.017 |
stations | -0.017 |
Token | I |
Token ID | 314 |
Feature activation | -7.07e-01 |
Pos logit contributions | |
seen | +0.025 |
contain | +0.023 |
hole | +0.023 |
stations | +0.023 |
keeping | +0.022 |
Neg logit contributions | |
iflower | -0.039 |
pron | -0.036 |
live | -0.032 |
wl | -0.031 |
innie | -0.031 |
Token | love |
Token ID | 1842 |
Feature activation | +5.95e-01 |
Pos logit contributions | |
seen | +0.014 |
hole | +0.013 |
keeping | +0.013 |
stations | +0.012 |
contain | +0.012 |
Neg logit contributions | |
iflower | -0.027 |
pron | -0.026 |
live | -0.023 |
innie | -0.022 |
ama | -0.022 |
Token | Car |
Token ID | 1879 |
Feature activation | +1.97e+00 |
Pos logit contributions | |
iflower | +0.020 |
pron | +0.019 |
live | +0.017 |
wl | +0.016 |
innie | +0.016 |
Neg logit contributions | |
seen | -0.015 |
keeping | -0.014 |
stations | -0.014 |
crashing | -0.013 |
umb | -0.013 |
Token | - |
Token ID | 12 |
Feature activation | +1.93e+00 |
Pos logit contributions | |
iflower | +0.061 |
pron | +0.056 |
live | +0.049 |
wl | +0.047 |
innie | +0.046 |
Neg logit contributions | |
seen | -0.054 |
hole | -0.051 |
officers | -0.050 |
keeping | -0.050 |
stations | -0.050 |
Token | D |
Token ID | 35 |
Feature activation | +2.20e-01 |
Pos logit contributions | |
iflower | +0.062 |
pron | +0.056 |
live | +0.050 |
ama | +0.048 |
innie | +0.048 |
Neg logit contributions | |
seen | -0.051 |
keeping | -0.048 |
stations | -0.047 |
officers | -0.047 |
hole | -0.046 |
Token | oll |
Token ID | 692 |
Feature activation | +8.07e+00 |
Pos logit contributions | |
iflower | +0.006 |
pron | +0.006 |
live | +0.005 |
innie | +0.005 |
ama | +0.005 |
Neg logit contributions | |
seen | -0.007 |
hole | -0.006 |
keeping | -0.006 |
stations | -0.006 |
contain | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | +6.74e-01 |
Pos logit contributions | |
iflower | +0.157 |
pron | +0.137 |
live | +0.115 |
wl | +0.104 |
innie | +0.103 |
Neg logit contributions | |
seen | -0.303 |
hole | -0.288 |
officers | -0.284 |
keeping | -0.284 |
ales | -0.280 |
Token | And |
Token ID | 843 |
Feature activation | -5.22e-01 |
Pos logit contributions | |
iflower | +0.022 |
pron | +0.020 |
live | +0.018 |
innie | +0.017 |
ama | +0.017 |
Neg logit contributions | |
seen | -0.017 |
hole | -0.016 |
keeping | -0.016 |
stations | -0.016 |
contain | -0.015 |
Token | I |
Token ID | 314 |
Feature activation | -4.54e-01 |
Pos logit contributions | |
seen | +0.012 |
contain | +0.011 |
hole | +0.011 |
keeping | +0.011 |
stations | +0.011 |
Neg logit contributions | |
iflower | -0.019 |
pron | -0.017 |
live | -0.016 |
wl | -0.015 |
innie | -0.015 |
Token | love |
Token ID | 1842 |
Feature activation | +7.13e-01 |
Pos logit contributions | |
seen | +0.009 |
hole | +0.008 |
keeping | +0.008 |
stations | +0.008 |
contain | +0.007 |
Neg logit contributions | |
iflower | -0.018 |
pron | -0.017 |
live | -0.015 |
innie | -0.014 |
ama | -0.014 |
Token | our |
Token ID | 674 |
Feature activation | +1.48e+00 |
Pos logit contributions | |
iflower | +0.024 |
pron | +0.024 |
live | +0.021 |
wl | +0.020 |
ama | +0.019 |
Neg logit contributions | |
seen | -0.017 |
keeping | -0.016 |
officers | -0.016 |
stations | -0.015 |
crashing | -0.015 |
Token | friends |
Token ID | 2460 |
Feature activation | -7.19e-01 |
Pos logit contributions | |
iflower | +0.066 |
pron | +0.059 |
live | +0.049 |
innie | +0.046 |
wl | +0.045 |
Neg logit contributions | |
seen | -0.096 |
keeping | -0.091 |
hole | -0.091 |
stations | -0.090 |
officers | -0.088 |
Token | and |
Token ID | 290 |
Feature activation | +1.20e+00 |
Pos logit contributions | |
brake | +0.019 |
hole | +0.019 |
seen | +0.018 |
brakes | +0.018 |
ales | +0.018 |
Neg logit contributions | |
iflower | -0.021 |
pron | -0.018 |
live | -0.017 |
wl | -0.016 |
ama | -0.016 |
Token | shared |
Token ID | 4888 |
Feature activation | +3.31e+00 |
Pos logit contributions | |
iflower | +0.038 |
pron | +0.035 |
live | +0.031 |
ama | +0.030 |
innie | +0.030 |
Neg logit contributions | |
seen | -0.032 |
keeping | -0.030 |
stations | -0.029 |
hole | -0.029 |
officers | -0.028 |
Token | their |
Token ID | 511 |
Feature activation | +7.84e-01 |
Pos logit contributions | |
iflower | +0.124 |
pron | +0.119 |
wl | +0.108 |
ama | +0.108 |
live | +0.106 |
Neg logit contributions | |
seen | -0.074 |
keeping | -0.064 |
stations | -0.063 |
towards | -0.060 |
umb | -0.060 |
Token | l |
Token ID | 300 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
iflower | +0.027 |
pron | +0.026 |
wl | +0.024 |
ama | +0.023 |
live | +0.022 |
Neg logit contributions | |
seen | -0.018 |
keeping | -0.015 |
ways | -0.015 |
umb | -0.015 |
want | -0.015 |
Token | oll |
Token ID | 692 |
Feature activation | +8.05e+00 |
Pos logit contributions | |
seen | +0.029 |
ets | +0.027 |
hole | +0.027 |
contain | +0.026 |
keeping | +0.026 |
Neg logit contributions | |
iflower | -0.030 |
pron | -0.029 |
wl | -0.024 |
ama | -0.023 |
innie | -0.023 |
Token | ip |
Token ID | 541 |
Feature activation | +5.23e+00 |
Pos logit contributions | |
pron | +0.184 |
iflower | +0.168 |
Mon | +0.149 |
wl | +0.141 |
Rich | +0.140 |
Neg logit contributions | |
seen | -0.228 |
ales | -0.227 |
ets | -0.217 |
keeping | -0.217 |
ly | -0.214 |
Token | ops |
Token ID | 2840 |
Feature activation | +1.48e+00 |
Pos logit contributions | |
pron | +0.140 |
iflower | +0.134 |
innie | +0.111 |
ama | +0.109 |
wl | +0.106 |
Neg logit contributions | |
keeping | -0.153 |
ets | -0.143 |
ales | -0.142 |
seen | -0.142 |
officers | -0.137 |
Token | every |
Token ID | 790 |
Feature activation | +1.51e-01 |
Pos logit contributions | |
iflower | +0.046 |
pron | +0.043 |
live | +0.038 |
innie | +0.038 |
wl | +0.037 |
Neg logit contributions | |
seen | -0.039 |
keeping | -0.037 |
hole | -0.037 |
contain | -0.036 |
stations | -0.035 |
Token | time |
Token ID | 640 |
Feature activation | -7.23e-01 |
Pos logit contributions | |
iflower | +0.003 |
pron | +0.003 |
live | +0.003 |
wl | +0.002 |
ama | +0.002 |
Neg logit contributions | |
seen | -0.006 |
keeping | -0.005 |
hole | -0.005 |
towards | -0.005 |
contain | -0.005 |
Token | they |
Token ID | 484 |
Feature activation | -2.97e+00 |
Pos logit contributions | |
seen | +0.018 |
keeping | +0.017 |
hole | +0.017 |
stations | +0.017 |
crashing | +0.017 |
Neg logit contributions | |
iflower | -0.024 |
pron | -0.022 |
live | -0.020 |
innie | -0.019 |
wl | -0.019 |
Token | Tess |
Token ID | 39412 |
Feature activation | +8.76e-01 |
Pos logit contributions | |
seen | +0.001 |
hole | +0.001 |
keeping | +0.001 |
stations | +0.001 |
contain | +0.001 |
Neg logit contributions | |
iflower | -0.001 |
pron | -0.001 |
live | -0.001 |
wl | -0.001 |
innie | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | -4.41e-01 |
Pos logit contributions | |
iflower | +0.029 |
pron | +0.026 |
live | +0.023 |
innie | +0.023 |
wl | +0.022 |
Neg logit contributions | |
brakes | -0.021 |
brake | -0.021 |
seen | -0.021 |
keeping | -0.020 |
crashing | -0.020 |
Token | She |
Token ID | 1375 |
Feature activation | +9.07e-01 |
Pos logit contributions | |
seen | +0.013 |
hole | +0.013 |
keeping | +0.013 |
stations | +0.012 |
contain | +0.012 |
Neg logit contributions | |
iflower | -0.012 |
pron | -0.011 |
live | -0.010 |
innie | -0.009 |
ama | -0.009 |
Token | loved |
Token ID | 6151 |
Feature activation | +6.54e-01 |
Pos logit contributions | |
iflower | +0.032 |
pron | +0.030 |
ama | +0.029 |
innie | +0.026 |
omm | +0.026 |
Neg logit contributions | |
seen | -0.021 |
want | -0.018 |
like | -0.016 |
towards | -0.015 |
argued | -0.015 |
Token | l |
Token ID | 300 |
Feature activation | -1.99e-01 |
Pos logit contributions | |
iflower | +0.025 |
pron | +0.023 |
live | +0.022 |
ama | +0.021 |
wl | +0.021 |
Neg logit contributions | |
seen | -0.013 |
keeping | -0.012 |
stations | -0.012 |
crashing | -0.012 |
umb | -0.011 |
Token | oll |
Token ID | 692 |
Feature activation | +7.98e+00 |
Pos logit contributions | |
seen | +0.006 |
hole | +0.006 |
contain | +0.006 |
keeping | +0.005 |
ets | +0.005 |
Neg logit contributions | |
iflower | -0.006 |
pron | -0.006 |
live | -0.004 |
wl | -0.004 |
innie | -0.004 |
Token | ip |
Token ID | 541 |
Feature activation | +5.18e+00 |
Pos logit contributions | |
pron | +0.130 |
iflower | +0.127 |
live | +0.092 |
Mon | +0.081 |
innie | +0.077 |
Neg logit contributions | |
seen | -0.306 |
ales | -0.293 |
hole | -0.286 |
ets | -0.284 |
officers | -0.284 |
Token | ops |
Token ID | 2840 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
pron | +0.136 |
iflower | +0.134 |
innie | +0.098 |
ama | +0.097 |
live | +0.095 |
Neg logit contributions | |
seen | -0.150 |
ets | -0.145 |
ales | -0.145 |
keeping | -0.145 |
officers | -0.145 |
Token | . |
Token ID | 13 |
Feature activation | -9.10e-01 |
Pos logit contributions | |
iflower | +0.041 |
pron | +0.038 |
live | +0.034 |
innie | +0.034 |
wl | +0.033 |
Neg logit contributions | |
seen | -0.023 |
hole | -0.021 |
keeping | -0.020 |
stations | -0.020 |
contain | -0.019 |
Token | One |
Token ID | 1881 |
Feature activation | -3.24e-01 |
Pos logit contributions | |
seen | +0.025 |
hole | +0.024 |
stations | +0.024 |
keeping | +0.024 |
officers | +0.023 |
Neg logit contributions | |
iflower | -0.028 |
pron | -0.026 |
live | -0.022 |
wl | -0.021 |
tto | -0.021 |
Token | day |
Token ID | 1110 |
Feature activation | -9.13e-01 |
Pos logit contributions | |
seen | +0.010 |
want | +0.009 |
towards | +0.009 |
eye | +0.008 |
contain | +0.008 |
Neg logit contributions | |
iflower | -0.009 |
pron | -0.008 |
wl | -0.007 |
innie | -0.007 |
live | -0.007 |
Token | as |
Token ID | 355 |
Feature activation | -3.55e-01 |
Pos logit contributions | |
seen | +0.024 |
keeping | +0.022 |
hole | +0.022 |
stations | +0.022 |
ales | +0.021 |
Neg logit contributions | |
iflower | -0.035 |
pron | -0.033 |
live | -0.029 |
ama | -0.028 |
innie | -0.028 |
Token | she |
Token ID | 673 |
Feature activation | -1.60e+00 |
Pos logit contributions | |
seen | +0.012 |
keeping | +0.011 |
crashing | +0.011 |
stations | +0.011 |
contain | +0.011 |
Neg logit contributions | |
iflower | -0.009 |
pron | -0.008 |
live | -0.007 |
wl | -0.007 |
innie | -0.007 |
Token | held |
Token ID | 2714 |
Feature activation | +2.52e+00 |
Pos logit contributions | |
seen | +0.039 |
hole | +0.035 |
keeping | +0.034 |
stations | +0.033 |
officers | +0.033 |
Neg logit contributions | |
iflower | -0.055 |
pron | -0.051 |
live | -0.045 |
ama | -0.044 |
innie | -0.043 |
Token | the |
Token ID | 262 |
Feature activation | +2.69e-01 |
Pos logit contributions | |
iflower | +0.084 |
pron | +0.076 |
live | +0.072 |
wl | +0.070 |
tto | +0.065 |
Neg logit contributions | |
seen | -0.062 |
hole | -0.059 |
keeping | -0.058 |
towards | -0.058 |
toward | -0.056 |
Token | l |
Token ID | 300 |
Feature activation | -7.59e-01 |
Pos logit contributions | |
iflower | +0.008 |
pron | +0.007 |
live | +0.006 |
wl | +0.006 |
innie | +0.006 |
Neg logit contributions | |
seen | -0.008 |
hole | -0.008 |
brake | -0.007 |
keeping | -0.007 |
officers | -0.007 |
Token | oll |
Token ID | 692 |
Feature activation | +7.91e+00 |
Pos logit contributions | |
seen | +0.023 |
hole | +0.022 |
contain | +0.021 |
keeping | +0.021 |
ets | +0.020 |
Neg logit contributions | |
iflower | -0.021 |
pron | -0.019 |
live | -0.016 |
wl | -0.016 |
innie | -0.015 |
Token | ipop |
Token ID | 42800 |
Feature activation | +3.72e+00 |
Pos logit contributions | |
iflower | +0.208 |
pron | +0.199 |
live | +0.188 |
wl | +0.178 |
irl | +0.176 |
Neg logit contributions | |
seen | -0.199 |
hole | -0.198 |
ly | -0.192 |
permission | -0.192 |
ously | -0.188 |
Token | in |
Token ID | 287 |
Feature activation | -2.69e-01 |
Pos logit contributions | |
iflower | +0.120 |
pron | +0.111 |
live | +0.099 |
innie | +0.094 |
wl | +0.093 |
Neg logit contributions | |
seen | -0.097 |
hole | -0.092 |
keeping | -0.091 |
stations | -0.089 |
contain | -0.086 |
Token | her |
Token ID | 607 |
Feature activation | -6.38e-01 |
Pos logit contributions | |
hole | +0.010 |
stations | +0.010 |
seen | +0.010 |
ales | +0.010 |
keeping | +0.009 |
Neg logit contributions | |
iflower | -0.005 |
pron | -0.005 |
innie | -0.004 |
live | -0.004 |
ama | -0.004 |
Token | hand |
Token ID | 1021 |
Feature activation | +7.55e-02 |
Pos logit contributions | |
seen | +0.011 |
hole | +0.010 |
keeping | +0.010 |
stations | +0.009 |
officers | +0.009 |
Neg logit contributions | |
iflower | -0.026 |
pron | -0.024 |
live | -0.023 |
wl | -0.022 |
innie | -0.022 |
Token | . |
Token ID | 13 |
Feature activation | -3.52e-01 |
Pos logit contributions | |
iflower | +0.002 |
pron | +0.002 |
ama | +0.002 |
live | +0.002 |
innie | +0.002 |
Neg logit contributions | |
stations | -0.002 |
seen | -0.002 |
keeping | -0.002 |
brake | -0.002 |
brakes | -0.002 |
Token | and |
Token ID | 290 |
Feature activation | +5.95e-01 |
Pos logit contributions | |
seen | +0.024 |
hole | +0.024 |
stations | +0.024 |
keeping | +0.023 |
ales | +0.023 |
Neg logit contributions | |
iflower | -0.022 |
pron | -0.020 |
live | -0.017 |
ama | -0.017 |
innie | -0.016 |
Token | they |
Token ID | 484 |
Feature activation | -2.00e+00 |
Pos logit contributions | |
iflower | +0.019 |
pron | +0.017 |
live | +0.015 |
innie | +0.015 |
wl | +0.015 |
Neg logit contributions | |
seen | -0.016 |
keeping | -0.015 |
stations | -0.015 |
officers | -0.015 |
hole | -0.015 |
Token | enjoyed |
Token ID | 8359 |
Feature activation | +1.94e+00 |
Pos logit contributions | |
seen | +0.050 |
keeping | +0.046 |
contain | +0.045 |
stations | +0.045 |
officers | +0.045 |
Neg logit contributions | |
iflower | -0.068 |
pron | -0.062 |
live | -0.055 |
ama | -0.054 |
innie | -0.054 |
Token | the |
Token ID | 262 |
Feature activation | +8.02e-01 |
Pos logit contributions | |
iflower | +0.067 |
pron | +0.061 |
live | +0.056 |
wl | +0.055 |
ama | +0.054 |
Neg logit contributions | |
seen | -0.048 |
keeping | -0.046 |
crashing | -0.043 |
stations | -0.043 |
contain | -0.041 |
Token | l |
Token ID | 300 |
Feature activation | -5.17e-01 |
Pos logit contributions | |
iflower | +0.026 |
pron | +0.023 |
wl | +0.022 |
ama | +0.021 |
live | +0.020 |
Neg logit contributions | |
seen | -0.021 |
contain | -0.020 |
want | -0.019 |
brake | -0.018 |
gasp | -0.018 |
Token | oll |
Token ID | 692 |
Feature activation | +7.87e+00 |
Pos logit contributions | |
seen | +0.016 |
hole | +0.015 |
keeping | +0.015 |
contain | +0.015 |
ets | +0.014 |
Neg logit contributions | |
iflower | -0.014 |
pron | -0.013 |
innie | -0.011 |
wl | -0.010 |
live | -0.010 |
Token | ipop |
Token ID | 42800 |
Feature activation | +3.88e+00 |
Pos logit contributions | |
iflower | +0.208 |
pron | +0.206 |
wl | +0.180 |
Mon | +0.178 |
live | +0.175 |
Neg logit contributions | |
seen | -0.203 |
ales | -0.189 |
ets | -0.182 |
ly | -0.182 |
ried | -0.182 |
Token | together |
Token ID | 1978 |
Feature activation | -1.42e+00 |
Pos logit contributions | |
iflower | +0.126 |
pron | +0.116 |
live | +0.103 |
ama | +0.098 |
innie | +0.098 |
Neg logit contributions | |
seen | -0.101 |
hole | -0.095 |
keeping | -0.094 |
stations | -0.093 |
crashing | -0.089 |
Token | . |
Token ID | 13 |
Feature activation | -6.59e-01 |
Pos logit contributions | |
seen | +0.033 |
stations | +0.032 |
hole | +0.032 |
keeping | +0.031 |
brake | +0.031 |
Neg logit contributions | |
iflower | -0.048 |
pron | -0.045 |
live | -0.040 |
ama | -0.037 |
innie | -0.037 |
Token | The |
Token ID | 383 |
Feature activation | +5.80e-01 |
Pos logit contributions | |
seen | +0.015 |
keeping | +0.014 |
hole | +0.014 |
stations | +0.014 |
officers | +0.013 |
Neg logit contributions | |
iflower | -0.024 |
pron | -0.022 |
live | -0.020 |
wl | -0.019 |
innie | -0.019 |
Token | guard |
Token ID | 4860 |
Feature activation | -3.21e-01 |
Pos logit contributions | |
iflower | +0.020 |
pron | +0.018 |
live | +0.016 |
innie | +0.016 |
ama | +0.015 |
Neg logit contributions | |
seen | -0.014 |
hole | -0.014 |
keeping | -0.014 |
stations | -0.013 |
contain | -0.013 |
Token | happy |
Token ID | 3772 |
Feature activation | +8.80e-01 |
Pos logit contributions | |
iflower | +0.066 |
pron | +0.061 |
live | +0.051 |
ama | +0.048 |
wl | +0.047 |
Neg logit contributions | |
hole | -0.069 |
seen | -0.068 |
stations | -0.067 |
keeping | -0.067 |
ales | -0.065 |
Token | . |
Token ID | 13 |
Feature activation | +2.72e-01 |
Pos logit contributions | |
iflower | +0.027 |
pron | +0.024 |
live | +0.021 |
ama | +0.020 |
wl | +0.019 |
Neg logit contributions | |
hole | -0.024 |
seen | -0.023 |
brake | -0.022 |
stations | -0.022 |
contain | -0.022 |
Token | And |
Token ID | 843 |
Feature activation | -1.43e+00 |
Pos logit contributions | |
iflower | +0.010 |
pron | +0.009 |
live | +0.008 |
wl | +0.007 |
innie | +0.007 |
Neg logit contributions | |
seen | -0.006 |
hole | -0.006 |
keeping | -0.006 |
stations | -0.006 |
crashing | -0.006 |
Token | giving |
Token ID | 3501 |
Feature activation | +1.30e+00 |
Pos logit contributions | |
seen | +0.039 |
hole | +0.038 |
keeping | +0.037 |
stations | +0.037 |
officers | +0.035 |
Neg logit contributions | |
iflower | -0.044 |
pron | -0.040 |
live | -0.035 |
ama | -0.034 |
innie | -0.033 |
Token | l |
Token ID | 300 |
Feature activation | -5.72e-01 |
Pos logit contributions | |
iflower | +0.043 |
pron | +0.041 |
live | +0.035 |
wl | +0.034 |
innie | +0.034 |
Neg logit contributions | |
seen | -0.034 |
keeping | -0.032 |
officers | -0.030 |
towards | -0.030 |
stations | -0.030 |
Token | oll |
Token ID | 692 |
Feature activation | +7.87e+00 |
Pos logit contributions | |
seen | +0.018 |
hole | +0.017 |
keeping | +0.017 |
contain | +0.016 |
crashing | +0.016 |
Neg logit contributions | |
iflower | -0.015 |
pron | -0.014 |
live | -0.012 |
innie | -0.011 |
wl | -0.011 |
Token | ip |
Token ID | 541 |
Feature activation | +5.15e+00 |
Pos logit contributions | |
iflower | +0.155 |
pron | +0.148 |
live | +0.122 |
wl | +0.113 |
innie | +0.108 |
Neg logit contributions | |
seen | -0.277 |
ets | -0.263 |
ales | -0.262 |
permission | -0.259 |
ried | -0.256 |
Token | ops |
Token ID | 2840 |
Feature activation | +7.21e-01 |
Pos logit contributions | |
iflower | +0.125 |
pron | +0.120 |
live | +0.095 |
innie | +0.094 |
wl | +0.093 |
Neg logit contributions | |
seen | -0.162 |
keeping | -0.160 |
ets | -0.155 |
officers | -0.153 |
ales | -0.152 |
Token | to |
Token ID | 284 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
iflower | +0.026 |
pron | +0.024 |
live | +0.021 |
ama | +0.020 |
innie | +0.020 |
Neg logit contributions | |
seen | -0.016 |
hole | -0.015 |
stations | -0.015 |
keeping | -0.015 |
brake | -0.014 |
Token | children |
Token ID | 1751 |
Feature activation | -1.21e+00 |
Pos logit contributions | |
iflower | +0.034 |
pron | +0.031 |
live | +0.028 |
wl | +0.026 |
innie | +0.026 |
Neg logit contributions | |
seen | -0.033 |
stations | -0.030 |
keeping | -0.030 |
hole | -0.029 |
contain | -0.029 |
Token | makes |
Token ID | 1838 |
Feature activation | +9.44e-01 |
Pos logit contributions | |
seen | +0.027 |
hole | +0.025 |
keeping | +0.025 |
stations | +0.025 |
brake | +0.024 |
Neg logit contributions | |
iflower | -0.043 |
pron | -0.040 |
live | -0.036 |
ama | -0.034 |
innie | -0.034 |
Token | generous |
Token ID | 14431 |
Feature activation | -2.02e-01 |
Pos logit contributions | |
iflower | +0.046 |
pron | +0.041 |
ama | +0.037 |
wl | +0.037 |
live | +0.037 |
Neg logit contributions | |
seen | -0.037 |
stations | -0.031 |
towards | -0.030 |
crashing | -0.030 |
keeping | -0.030 |
Token | and |
Token ID | 290 |
Feature activation | +2.83e+00 |
Pos logit contributions | |
seen | +0.005 |
hole | +0.005 |
stations | +0.005 |
keeping | +0.005 |
brake | +0.005 |
Neg logit contributions | |
iflower | -0.006 |
pron | -0.006 |
live | -0.005 |
ama | -0.005 |
innie | -0.005 |
Token | share |
Token ID | 2648 |
Feature activation | +3.76e+00 |
Pos logit contributions | |
iflower | +0.091 |
pron | +0.083 |
live | +0.074 |
innie | +0.072 |
ama | +0.071 |
Neg logit contributions | |
seen | -0.077 |
keeping | -0.071 |
contain | -0.068 |
stations | -0.068 |
hole | -0.067 |
Token | her |
Token ID | 607 |
Feature activation | +3.96e-01 |
Pos logit contributions | |
iflower | +0.117 |
pron | +0.110 |
ama | +0.098 |
live | +0.098 |
wl | +0.097 |
Neg logit contributions | |
seen | -0.104 |
keeping | -0.097 |
stations | -0.093 |
towards | -0.092 |
officers | -0.091 |
Token | l |
Token ID | 300 |
Feature activation | -6.96e-01 |
Pos logit contributions | |
iflower | +0.012 |
pron | +0.011 |
live | +0.011 |
wl | +0.011 |
ama | +0.010 |
Neg logit contributions | |
seen | -0.010 |
keeping | -0.009 |
permission | -0.009 |
want | -0.009 |
ways | -0.009 |
Token | oll |
Token ID | 692 |
Feature activation | +7.83e+00 |
Pos logit contributions | |
seen | +0.020 |
hole | +0.018 |
ets | +0.018 |
keeping | +0.018 |
contain | +0.018 |
Neg logit contributions | |
iflower | -0.020 |
pron | -0.019 |
wl | -0.016 |
innie | -0.015 |
live | -0.015 |
Token | ip |
Token ID | 541 |
Feature activation | +5.10e+00 |
Pos logit contributions | |
pron | +0.198 |
iflower | +0.183 |
Mon | +0.169 |
Ros | +0.164 |
irl | +0.162 |
Neg logit contributions | |
seen | -0.197 |
ales | -0.197 |
ly | -0.191 |
owing | -0.191 |
hole | -0.189 |
Token | ops |
Token ID | 2840 |
Feature activation | +1.01e+00 |
Pos logit contributions | |
pron | +0.136 |
iflower | +0.130 |
innie | +0.105 |
ama | +0.103 |
wl | +0.103 |
Neg logit contributions | |
keeping | -0.144 |
ets | -0.137 |
ales | -0.134 |
officers | -0.131 |
seen | -0.131 |
Token | with |
Token ID | 351 |
Feature activation | +1.55e+00 |
Pos logit contributions | |
iflower | +0.032 |
pron | +0.030 |
live | +0.027 |
innie | +0.027 |
wl | +0.026 |
Neg logit contributions | |
seen | -0.026 |
keeping | -0.025 |
hole | -0.025 |
contain | -0.024 |
stations | -0.023 |
Token | others |
Token ID | 1854 |
Feature activation | -3.18e-01 |
Pos logit contributions | |
iflower | +0.046 |
pron | +0.043 |
live | +0.037 |
ama | +0.036 |
tto | +0.036 |
Neg logit contributions | |
seen | -0.045 |
keeping | -0.041 |
stations | -0.040 |
officers | -0.040 |
hole | -0.039 |
Token | , |
Token ID | 11 |
Feature activation | +3.65e-01 |
Pos logit contributions | |
seen | +0.006 |
hole | +0.006 |
keeping | +0.006 |
stations | +0.006 |
officers | +0.005 |
Neg logit contributions | |
iflower | -0.012 |
pron | -0.012 |
live | -0.010 |
innie | -0.010 |
ama | -0.010 |
Token | because |
Token ID | 780 |
Feature activation | -5.04e-01 |
Pos logit contributions | |
keeping | +0.048 |
seen | +0.046 |
stations | +0.046 |
contain | +0.044 |
ales | +0.044 |
Neg logit contributions | |
iflower | -0.047 |
pron | -0.041 |
ama | -0.036 |
innie | -0.035 |
live | -0.035 |
Token | she |
Token ID | 673 |
Feature activation | +5.53e-02 |
Pos logit contributions | |
seen | +0.015 |
keeping | +0.013 |
crashing | +0.013 |
stations | +0.013 |
footsteps | +0.013 |
Neg logit contributions | |
iflower | -0.015 |
pron | -0.014 |
live | -0.012 |
wl | -0.012 |
innie | -0.012 |
Token | loved |
Token ID | 6151 |
Feature activation | +5.38e-01 |
Pos logit contributions | |
iflower | +0.002 |
pron | +0.002 |
ama | +0.002 |
live | +0.002 |
innie | +0.001 |
Neg logit contributions | |
seen | -0.001 |
argued | -0.001 |
hole | -0.001 |
want | -0.001 |
keeping | -0.001 |
Token | her |
Token ID | 607 |
Feature activation | +1.20e+00 |
Pos logit contributions | |
iflower | +0.018 |
pron | +0.016 |
live | +0.016 |
ama | +0.016 |
wl | +0.016 |
Neg logit contributions | |
seen | -0.013 |
keeping | -0.013 |
stations | -0.013 |
umb | -0.012 |
crashing | -0.012 |
Token | l |
Token ID | 300 |
Feature activation | -4.04e-01 |
Pos logit contributions | |
iflower | +0.037 |
live | +0.031 |
wl | +0.031 |
pron | +0.030 |
tto | +0.029 |
Neg logit contributions | |
seen | -0.035 |
hole | -0.031 |
officers | -0.030 |
umb | -0.030 |
stations | -0.029 |
Token | oll |
Token ID | 692 |
Feature activation | +7.82e+00 |
Pos logit contributions | |
seen | +0.012 |
hole | +0.011 |
ets | +0.010 |
vious | +0.010 |
keeping | +0.010 |
Neg logit contributions | |
iflower | -0.011 |
pron | -0.011 |
wl | -0.009 |
live | -0.009 |
innie | -0.009 |
Token | ip |
Token ID | 541 |
Feature activation | +5.09e+00 |
Pos logit contributions | |
iflower | +0.165 |
pron | +0.161 |
Mon | +0.146 |
Rich | +0.143 |
irl | +0.139 |
Neg logit contributions | |
seen | -0.233 |
hole | -0.219 |
ly | -0.212 |
owing | -0.209 |
permission | -0.206 |
Token | ops |
Token ID | 2840 |
Feature activation | +1.26e+00 |
Pos logit contributions | |
iflower | +0.135 |
pron | +0.133 |
ama | +0.109 |
wl | +0.107 |
innie | +0.105 |
Neg logit contributions | |
seen | -0.132 |
keeping | -0.132 |
ets | -0.130 |
hole | -0.126 |
argued | -0.123 |
Token | so |
Token ID | 523 |
Feature activation | +2.15e+00 |
Pos logit contributions | |
iflower | +0.045 |
pron | +0.042 |
live | +0.037 |
ama | +0.036 |
innie | +0.035 |
Neg logit contributions | |
seen | -0.028 |
stations | -0.027 |
keeping | -0.027 |
hole | -0.026 |
officers | -0.025 |
Token | much |
Token ID | 881 |
Feature activation | +2.37e-01 |
Pos logit contributions | |
iflower | +0.061 |
pron | +0.058 |
live | +0.049 |
ama | +0.046 |
wl | +0.046 |
Neg logit contributions | |
seen | -0.064 |
keeping | -0.057 |
hole | -0.057 |
stations | -0.056 |
crashing | -0.056 |
Token | . |
Token ID | 13 |
Feature activation | -2.32e-01 |
Pos logit contributions | |
iflower | +0.008 |
pron | +0.007 |
live | +0.007 |
innie | +0.006 |
ama | +0.006 |
Neg logit contributions | |
seen | -0.006 |
stations | -0.006 |
hole | -0.005 |
keeping | -0.005 |
contain | -0.005 |
Token | Lily |
Token ID | 20037 |
Feature activation | +3.22e-01 |
Pos logit contributions | |
iflower | +0.074 |
pron | +0.064 |
wl | +0.059 |
live | +0.057 |
tto | +0.056 |
Neg logit contributions | |
seen | -0.056 |
officers | -0.052 |
keeping | -0.052 |
crashing | -0.051 |
umb | -0.051 |
Token | happily |
Token ID | 18177 |
Feature activation | -5.19e-01 |
Pos logit contributions | |
iflower | +0.012 |
pron | +0.010 |
wl | +0.010 |
live | +0.009 |
ama | +0.009 |
Neg logit contributions | |
seen | -0.007 |
hole | -0.007 |
keeping | -0.006 |
argued | -0.006 |
officers | -0.006 |
Token | took |
Token ID | 1718 |
Feature activation | +3.41e-01 |
Pos logit contributions | |
seen | +0.014 |
hole | +0.013 |
keeping | +0.013 |
stations | +0.013 |
crashing | +0.012 |
Neg logit contributions | |
iflower | -0.017 |
pron | -0.015 |
live | -0.014 |
innie | -0.013 |
ama | -0.013 |
Token | the |
Token ID | 262 |
Feature activation | +4.18e-01 |
Pos logit contributions | |
iflower | +0.013 |
pron | +0.012 |
wl | +0.011 |
live | +0.011 |
tto | +0.010 |
Neg logit contributions | |
seen | -0.008 |
keeping | -0.007 |
towards | -0.007 |
officers | -0.006 |
permission | -0.006 |
Token | l |
Token ID | 300 |
Feature activation | -8.94e-01 |
Pos logit contributions | |
iflower | +0.013 |
wl | +0.011 |
ama | +0.011 |
live | +0.011 |
pron | +0.011 |
Neg logit contributions | |
seen | -0.011 |
hole | -0.010 |
contain | -0.010 |
towards | -0.010 |
brake | -0.010 |
Token | oll |
Token ID | 692 |
Feature activation | +7.82e+00 |
Pos logit contributions | |
seen | +0.027 |
hole | +0.025 |
ets | +0.024 |
contain | +0.024 |
keeping | +0.023 |
Neg logit contributions | |
iflower | -0.025 |
pron | -0.023 |
wl | -0.020 |
live | -0.019 |
ama | -0.019 |
Token | ipop |
Token ID | 42800 |
Feature activation | +2.85e+00 |
Pos logit contributions | |
iflower | +0.220 |
pron | +0.209 |
live | +0.193 |
wl | +0.188 |
irl | +0.184 |
Neg logit contributions | |
seen | -0.194 |
hole | -0.182 |
ly | -0.182 |
permission | -0.177 |
ales | -0.173 |
Token | and |
Token ID | 290 |
Feature activation | +7.69e-01 |
Pos logit contributions | |
iflower | +0.103 |
pron | +0.096 |
live | +0.086 |
innie | +0.082 |
ama | +0.082 |
Neg logit contributions | |
seen | -0.064 |
hole | -0.059 |
keeping | -0.059 |
stations | -0.059 |
crashing | -0.055 |
Token | started |
Token ID | 2067 |
Feature activation | +5.20e-02 |
Pos logit contributions | |
iflower | +0.027 |
pron | +0.024 |
wl | +0.023 |
live | +0.022 |
ama | +0.022 |
Neg logit contributions | |
seen | -0.019 |
keeping | -0.017 |
officers | -0.016 |
argued | -0.016 |
hole | -0.016 |
Token | licking |
Token ID | 43390 |
Feature activation | +1.07e+00 |
Pos logit contributions | |
iflower | +0.002 |
pron | +0.002 |
live | +0.002 |
wl | +0.002 |
ama | +0.002 |
Neg logit contributions | |
seen | -0.001 |
keeping | -0.001 |
crashing | -0.001 |
towards | -0.001 |
hole | -0.001 |
Token | it |
Token ID | 340 |
Feature activation | +1.83e+00 |
Pos logit contributions | |
iflower | +0.040 |
pron | +0.037 |
live | +0.037 |
wl | +0.036 |
tto | +0.034 |
Neg logit contributions | |
towards | -0.021 |
seen | -0.021 |
crashing | -0.020 |
hole | -0.020 |
ked | -0.020 |
Token | the |
Token ID | 262 |
Feature activation | -1.53e+00 |
Pos logit contributions | |
seen | +0.006 |
hole | +0.006 |
keeping | +0.006 |
stations | +0.006 |
contain | +0.006 |
Neg logit contributions | |
iflower | -0.006 |
pron | -0.005 |
live | -0.005 |
innie | -0.004 |
wl | -0.004 |
Token | sky |
Token ID | 6766 |
Feature activation | -1.53e+00 |
Pos logit contributions | |
seen | +0.044 |
hole | +0.041 |
keeping | +0.041 |
stations | +0.041 |
crashing | +0.039 |
Neg logit contributions | |
iflower | -0.046 |
pron | -0.042 |
live | -0.037 |
innie | -0.035 |
ama | -0.035 |
Token | , |
Token ID | 11 |
Feature activation | -4.75e-01 |
Pos logit contributions | |
keeping | +0.045 |
seen | +0.044 |
stations | +0.043 |
hole | +0.042 |
officers | +0.042 |
Neg logit contributions | |
iflower | -0.044 |
pron | -0.039 |
live | -0.033 |
wl | -0.032 |
ama | -0.032 |
Token | the |
Token ID | 262 |
Feature activation | -1.14e+00 |
Pos logit contributions | |
keeping | +0.013 |
seen | +0.013 |
hole | +0.013 |
officers | +0.013 |
ales | +0.013 |
Neg logit contributions | |
iflower | -0.014 |
pron | -0.012 |
live | -0.010 |
wl | -0.010 |
ama | -0.009 |
Token | dragon |
Token ID | 10441 |
Feature activation | -5.88e-01 |
Pos logit contributions | |
seen | +0.034 |
hole | +0.033 |
keeping | +0.033 |
stations | +0.032 |
ales | +0.031 |
Neg logit contributions | |
iflower | -0.033 |
pron | -0.030 |
live | -0.024 |
ama | -0.023 |
wl | -0.023 |
Token | ruled |
Token ID | 8879 |
Feature activation | -7.96e+00 |
Pos logit contributions | |
seen | +0.018 |
keeping | +0.017 |
stations | +0.017 |
hole | +0.017 |
brake | +0.016 |
Neg logit contributions | |
iflower | -0.016 |
pron | -0.015 |
live | -0.013 |
wl | -0.012 |
ama | -0.012 |
Token | over |
Token ID | 625 |
Feature activation | +4.55e-01 |
Pos logit contributions | |
seen | +0.201 |
hole | +0.189 |
keeping | +0.188 |
stations | +0.186 |
contain | +0.178 |
Neg logit contributions | |
iflower | -0.263 |
pron | -0.242 |
live | -0.216 |
innie | -0.206 |
ama | -0.206 |
Token | the |
Token ID | 262 |
Feature activation | -2.01e+00 |
Pos logit contributions | |
iflower | +0.012 |
pron | +0.010 |
live | +0.009 |
ama | +0.008 |
wl | +0.008 |
Neg logit contributions | |
hole | -0.014 |
seen | -0.013 |
contain | -0.013 |
keeping | -0.013 |
stations | -0.013 |
Token | land |
Token ID | 1956 |
Feature activation | -1.53e-01 |
Pos logit contributions | |
seen | +0.065 |
hole | +0.062 |
keeping | +0.062 |
stations | +0.061 |
ales | +0.059 |
Neg logit contributions | |
iflower | -0.053 |
pron | -0.047 |
live | -0.039 |
ama | -0.037 |
wl | -0.037 |
Token | with |
Token ID | 351 |
Feature activation | -1.45e-02 |
Pos logit contributions | |
seen | +0.004 |
keeping | +0.004 |
stations | +0.004 |
hole | +0.004 |
brake | +0.004 |
Neg logit contributions | |
iflower | -0.005 |
pron | -0.005 |
live | -0.004 |
ama | -0.004 |
wl | -0.004 |
Token | might |
Token ID | 1244 |
Feature activation | -3.73e+00 |
Pos logit contributions | |
hole | +0.000 |
seen | +0.000 |
keeping | +0.000 |
stations | +0.000 |
contain | +0.000 |
Neg logit contributions | |
iflower | -0.000 |
pron | -0.000 |
live | -0.000 |
wl | -0.000 |
ama | -0.000 |
Token | grass |
Token ID | 8701 |
Feature activation | -6.16e-01 |
Pos logit contributions | |
seen | +0.032 |
keeping | +0.030 |
hole | +0.030 |
stations | +0.029 |
contain | +0.029 |
Neg logit contributions | |
iflower | -0.032 |
pron | -0.029 |
live | -0.025 |
ama | -0.024 |
innie | -0.024 |
Token | around |
Token ID | 1088 |
Feature activation | +1.19e-01 |
Pos logit contributions | |
seen | +0.022 |
keeping | +0.021 |
hole | +0.021 |
stations | +0.021 |
contain | +0.021 |
Neg logit contributions | |
iflower | -0.014 |
pron | -0.012 |
live | -0.010 |
ama | -0.009 |
wl | -0.009 |
Token | the |
Token ID | 262 |
Feature activation | -2.22e+00 |
Pos logit contributions | |
iflower | +0.003 |
pron | +0.003 |
live | +0.003 |
ama | +0.003 |
wl | +0.002 |
Neg logit contributions | |
hole | -0.003 |
seen | -0.003 |
stations | -0.003 |
keeping | -0.003 |
contain | -0.003 |
Token | hill |
Token ID | 12788 |
Feature activation | -2.24e-01 |
Pos logit contributions | |
seen | +0.060 |
keeping | +0.056 |
hole | +0.055 |
stations | +0.054 |
contain | +0.054 |
Neg logit contributions | |
iflower | -0.072 |
pron | -0.065 |
live | -0.056 |
ama | -0.055 |
innie | -0.054 |
Token | was |
Token ID | 373 |
Feature activation | -1.82e+00 |
Pos logit contributions | |
seen | +0.008 |
keeping | +0.007 |
stations | +0.007 |
hole | +0.007 |
brake | +0.007 |
Neg logit contributions | |
iflower | -0.005 |
pron | -0.005 |
live | -0.004 |
wl | -0.004 |
ama | -0.004 |
Token | wit |
Token ID | 20868 |
Feature activation | -7.76e+00 |
Pos logit contributions | |
seen | +0.053 |
keeping | +0.051 |
hole | +0.051 |
stations | +0.050 |
officers | +0.049 |
Neg logit contributions | |
iflower | -0.053 |
pron | -0.049 |
live | -0.042 |
ama | -0.039 |
innie | -0.039 |
Token | hered |
Token ID | 6083 |
Feature activation | -6.98e-02 |
Pos logit contributions | |
seen | +0.212 |
keeping | +0.207 |
umb | +0.194 |
contain | +0.192 |
stations | +0.188 |
Neg logit contributions | |
iflower | -0.232 |
pron | -0.214 |
live | -0.187 |
wl | -0.180 |
ama | -0.180 |
Token | and |
Token ID | 290 |
Feature activation | +2.27e-01 |
Pos logit contributions | |
stations | +0.002 |
brakes | +0.002 |
unison | +0.002 |
hole | +0.002 |
footsteps | +0.002 |
Neg logit contributions | |
iflower | -0.001 |
tto | -0.001 |
live | -0.001 |
P | -0.001 |
in | -0.001 |
Token | yellow |
Token ID | 7872 |
Feature activation | +6.56e-01 |
Pos logit contributions | |
iflower | +0.005 |
pron | +0.005 |
live | +0.004 |
wl | +0.004 |
lf | +0.004 |
Neg logit contributions | |
hole | -0.007 |
stations | -0.007 |
keeping | -0.007 |
ales | -0.007 |
ets | -0.007 |
Token | . |
Token ID | 13 |
Feature activation | -1.23e+00 |
Pos logit contributions | |
iflower | +0.014 |
time | +0.011 |
pron | +0.011 |
mint | +0.010 |
ama | +0.010 |
Neg logit contributions | |
seen | -0.021 |
contain | -0.020 |
announcements | -0.020 |
hole | -0.020 |
keeping | -0.020 |
Token | He |
Token ID | 679 |
Feature activation | +4.23e-01 |
Pos logit contributions | |
seen | +0.033 |
hole | +0.031 |
keeping | +0.031 |
stations | +0.031 |
officers | +0.030 |
Neg logit contributions | |
iflower | -0.039 |
pron | -0.036 |
live | -0.032 |
wl | -0.029 |
innie | -0.029 |
Token | , |
Token ID | 11 |
Feature activation | -6.73e-01 |
Pos logit contributions | |
keeping | +0.057 |
seen | +0.056 |
stations | +0.055 |
brake | +0.053 |
officers | +0.053 |
Neg logit contributions | |
iflower | -0.066 |
pron | -0.062 |
ama | -0.052 |
live | -0.051 |
innie | -0.050 |
Token | feeling |
Token ID | 4203 |
Feature activation | -4.97e-01 |
Pos logit contributions | |
keeping | +0.018 |
seen | +0.018 |
stations | +0.018 |
hole | +0.017 |
officers | +0.017 |
Neg logit contributions | |
iflower | -0.020 |
pron | -0.019 |
live | -0.016 |
ama | -0.016 |
innie | -0.015 |
Token | the |
Token ID | 262 |
Feature activation | +2.95e-01 |
Pos logit contributions | |
seen | +0.014 |
keeping | +0.014 |
hole | +0.014 |
stations | +0.014 |
ales | +0.014 |
Neg logit contributions | |
iflower | -0.014 |
pron | -0.013 |
live | -0.011 |
ama | -0.010 |
wl | -0.010 |
Token | peace |
Token ID | 4167 |
Feature activation | -5.11e-01 |
Pos logit contributions | |
iflower | +0.008 |
pron | +0.008 |
ama | +0.006 |
live | +0.006 |
wl | +0.006 |
Neg logit contributions | |
keeping | -0.009 |
seen | -0.008 |
ales | -0.008 |
hole | -0.008 |
stations | -0.008 |
Token | and |
Token ID | 290 |
Feature activation | +1.32e+00 |
Pos logit contributions | |
seen | +0.014 |
keeping | +0.014 |
hole | +0.013 |
stations | +0.013 |
ales | +0.013 |
Neg logit contributions | |
iflower | -0.016 |
pron | -0.014 |
live | -0.012 |
ama | -0.012 |
wl | -0.011 |
Token | tranqu |
Token ID | 31882 |
Feature activation | -7.71e+00 |
Pos logit contributions | |
iflower | +0.038 |
pron | +0.034 |
live | +0.028 |
ama | +0.027 |
wl | +0.026 |
Neg logit contributions | |
keeping | -0.037 |
ales | -0.037 |
hole | -0.036 |
seen | -0.036 |
officers | -0.036 |
Token | ility |
Token ID | 879 |
Feature activation | +1.46e+00 |
Pos logit contributions | |
brakes | +0.162 |
officers | +0.152 |
buses | +0.151 |
stations | +0.150 |
levers | +0.149 |
Neg logit contributions | |
pron | -0.260 |
iflower | -0.259 |
live | -0.233 |
ama | -0.231 |
innie | -0.225 |
Token | of |
Token ID | 286 |
Feature activation | +3.58e-02 |
Pos logit contributions | |
iflower | +0.046 |
pron | +0.042 |
live | +0.038 |
ama | +0.036 |
innie | +0.035 |
Neg logit contributions | |
seen | -0.039 |
keeping | -0.037 |
hole | -0.036 |
stations | -0.036 |
brake | -0.035 |
Token | the |
Token ID | 262 |
Feature activation | -5.51e-01 |
Pos logit contributions | |
iflower | +0.001 |
pron | +0.001 |
live | +0.001 |
ama | +0.001 |
innie | +0.001 |
Neg logit contributions | |
seen | -0.001 |
hole | -0.001 |
keeping | -0.001 |
officers | -0.001 |
stations | -0.001 |
Token | night |
Token ID | 1755 |
Feature activation | -8.22e-01 |
Pos logit contributions | |
seen | +0.017 |
keeping | +0.016 |
hole | +0.016 |
stations | +0.016 |
officers | +0.015 |
Neg logit contributions | |
iflower | -0.016 |
pron | -0.014 |
live | -0.012 |
ama | -0.011 |
wl | -0.011 |
Token | , |
Token ID | 11 |
Feature activation | -3.45e-01 |
Pos logit contributions | |
seen | +0.022 |
keeping | +0.022 |
hole | +0.021 |
stations | +0.021 |
brake | +0.021 |
Neg logit contributions | |
iflower | -0.025 |
pron | -0.023 |
live | -0.020 |
ama | -0.019 |
wl | -0.019 |
Token | the |
Token ID | 262 |
Feature activation | -1.50e+00 |
Pos logit contributions | |
hole | +0.042 |
ets | +0.041 |
seen | +0.040 |
officers | +0.040 |
stations | +0.039 |
Neg logit contributions | |
iflower | -0.030 |
pron | -0.026 |
live | -0.020 |
innie | -0.019 |
lf | -0.018 |
Token | garden |
Token ID | 11376 |
Feature activation | +2.77e-01 |
Pos logit contributions | |
seen | +0.044 |
keeping | +0.042 |
hole | +0.042 |
stations | +0.041 |
ales | +0.040 |
Neg logit contributions | |
iflower | -0.044 |
pron | -0.040 |
live | -0.034 |
ama | -0.033 |
innie | -0.032 |
Token | . |
Token ID | 13 |
Feature activation | -4.54e-01 |
Pos logit contributions | |
iflower | +0.008 |
pron | +0.007 |
ama | +0.006 |
live | +0.006 |
on | +0.005 |
Neg logit contributions | |
brake | -0.008 |
brakes | -0.008 |
stations | -0.007 |
officers | -0.007 |
Run | -0.007 |
Token | A |
Token ID | 317 |
Feature activation | -1.22e+00 |
Pos logit contributions | |
seen | +0.014 |
keeping | +0.013 |
hole | +0.013 |
stations | +0.013 |
officers | +0.012 |
Neg logit contributions | |
iflower | -0.013 |
pron | -0.012 |
live | -0.010 |
innie | -0.010 |
wl | -0.010 |
Token | subtle |
Token ID | 11800 |
Feature activation | -1.59e+00 |
Pos logit contributions | |
seen | +0.038 |
hole | +0.036 |
crashing | +0.036 |
stations | +0.035 |
officers | +0.035 |
Neg logit contributions | |
iflower | -0.032 |
pron | -0.028 |
live | -0.026 |
innie | -0.025 |
wl | -0.024 |
Token | tranqu |
Token ID | 31882 |
Feature activation | -7.69e+00 |
Pos logit contributions | |
seen | +0.049 |
keeping | +0.048 |
hole | +0.048 |
stations | +0.047 |
ales | +0.045 |
Neg logit contributions | |
iflower | -0.043 |
pron | -0.040 |
live | -0.033 |
ama | -0.032 |
wl | -0.031 |
Token | ility |
Token ID | 879 |
Feature activation | -6.20e-02 |
Pos logit contributions | |
brakes | +0.162 |
officers | +0.156 |
keeping | +0.153 |
levers | +0.152 |
stations | +0.152 |
Neg logit contributions | |
pron | -0.261 |
iflower | -0.259 |
live | -0.235 |
ama | -0.220 |
innie | -0.219 |
Token | filled |
Token ID | 5901 |
Feature activation | -7.07e-02 |
Pos logit contributions | |
seen | +0.002 |
keeping | +0.002 |
brake | +0.002 |
hole | +0.002 |
stations | +0.002 |
Neg logit contributions | |
iflower | -0.002 |
pron | -0.002 |
live | -0.001 |
ama | -0.001 |
wl | -0.001 |
Token | the |
Token ID | 262 |
Feature activation | -9.70e-01 |
Pos logit contributions | |
seen | +0.002 |
keeping | +0.002 |
stations | +0.002 |
hole | +0.002 |
contain | +0.002 |
Neg logit contributions | |
iflower | -0.002 |
pron | -0.002 |
live | -0.002 |
innie | -0.002 |
ama | -0.002 |
Token | air |
Token ID | 1633 |
Feature activation | -5.23e-01 |
Pos logit contributions | |
seen | +0.030 |
hole | +0.029 |
keeping | +0.028 |
ales | +0.028 |
contain | +0.028 |
Neg logit contributions | |
iflower | -0.027 |
pron | -0.024 |
live | -0.019 |
ama | -0.019 |
wl | -0.018 |
Token | and |
Token ID | 290 |
Feature activation | -8.62e-02 |
Pos logit contributions | |
seen | +0.014 |
keeping | +0.014 |
stations | +0.013 |
brake | +0.013 |
officers | +0.013 |
Neg logit contributions | |
iflower | -0.016 |
pron | -0.015 |
live | -0.012 |
ama | -0.012 |
wl | -0.012 |
Token | , |
Token ID | 11 |
Feature activation | +2.95e-02 |
Pos logit contributions | |
iflower | +0.010 |
pron | +0.009 |
live | +0.008 |
ama | +0.008 |
innie | +0.007 |
Neg logit contributions | |
seen | -0.011 |
keeping | -0.011 |
stations | -0.011 |
hole | -0.011 |
contain | -0.010 |
Token | it |
Token ID | 340 |
Feature activation | +1.32e+00 |
Pos logit contributions | |
iflower | +0.001 |
pron | +0.001 |
live | +0.000 |
ama | +0.000 |
innie | +0.000 |
Neg logit contributions | |
seen | -0.001 |
hole | -0.001 |
keeping | -0.001 |
stations | -0.001 |
contain | -0.001 |
Token | kept |
Token ID | 4030 |
Feature activation | +8.24e-01 |
Pos logit contributions | |
iflower | +0.046 |
pron | +0.043 |
live | +0.038 |
innie | +0.037 |
ama | +0.037 |
Neg logit contributions | |
seen | -0.032 |
hole | -0.029 |
keeping | -0.028 |
stations | -0.028 |
officers | -0.028 |
Token | on |
Token ID | 319 |
Feature activation | +6.92e-01 |
Pos logit contributions | |
iflower | +0.028 |
pron | +0.026 |
live | +0.024 |
innie | +0.023 |
ama | +0.023 |
Neg logit contributions | |
seen | -0.020 |
keeping | -0.019 |
hole | -0.019 |
crashing | -0.018 |
stations | -0.018 |
Token | flying |
Token ID | 7348 |
Feature activation | -1.73e+00 |
Pos logit contributions | |
iflower | +0.022 |
pron | +0.020 |
live | +0.018 |
innie | +0.017 |
ama | +0.017 |
Neg logit contributions | |
seen | -0.018 |
hole | -0.017 |
keeping | -0.017 |
stations | -0.017 |
contain | -0.016 |
Token | err |
Token ID | 11454 |
Feature activation | -7.60e+00 |
Pos logit contributions | |
seen | +0.050 |
supervision | +0.050 |
officers | +0.049 |
keeping | +0.048 |
brake | +0.048 |
Neg logit contributions | |
iflower | -0.042 |
pron | -0.035 |
live | -0.033 |
wl | -0.032 |
P | -0.030 |
Token | atically |
Token ID | 4142 |
Feature activation | -1.65e+00 |
Pos logit contributions | |
seen | +0.159 |
hole | +0.149 |
stations | +0.143 |
keeping | +0.142 |
crashing | +0.140 |
Neg logit contributions | |
iflower | -0.281 |
pron | -0.256 |
live | -0.239 |
innie | -0.231 |
wl | -0.227 |
Token | with |
Token ID | 351 |
Feature activation | -5.25e-01 |
Pos logit contributions | |
Run | +0.046 |
seen | +0.046 |
brake | +0.045 |
stations | +0.044 |
ales | +0.044 |
Neg logit contributions | |
iflower | -0.042 |
pron | -0.035 |
ama | -0.028 |
ip | -0.028 |
best | -0.027 |
Token | inexplicable |
Token ID | 49602 |
Feature activation | -9.67e-01 |
Pos logit contributions | |
hole | +0.017 |
stations | +0.016 |
keeping | +0.016 |
contain | +0.016 |
brake | +0.016 |
Neg logit contributions | |
iflower | -0.012 |
pron | -0.010 |
live | -0.008 |
wl | -0.008 |
ama | -0.008 |
Token | speed |
Token ID | 2866 |
Feature activation | -2.02e+00 |
Pos logit contributions | |
stations | +0.027 |
keeping | +0.026 |
Run | +0.026 |
ales | +0.025 |
rified | +0.025 |
Neg logit contributions | |
iflower | -0.028 |
pron | -0.024 |
story | -0.020 |
special | -0.020 |
important | -0.020 |
Token | . |
Token ID | 13 |
Feature activation | -9.76e-01 |
Pos logit contributions | |
seen | +0.056 |
horror | +0.055 |
footsteps | +0.055 |
stations | +0.054 |
Run | +0.053 |
Neg logit contributions | |
iflower | -0.051 |
pron | -0.046 |
live | -0.040 |
ip | -0.040 |
wl | -0.039 |
Token | saw |
Token ID | 2497 |
Feature activation | +7.29e-01 |
Pos logit contributions | |
iflower | +0.039 |
pron | +0.037 |
innie | +0.034 |
live | +0.033 |
ama | +0.032 |
Neg logit contributions | |
seen | -0.024 |
hole | -0.021 |
towards | -0.020 |
stations | -0.020 |
crashing | -0.019 |
Token | that |
Token ID | 326 |
Feature activation | +2.41e+00 |
Pos logit contributions | |
iflower | +0.021 |
pron | +0.018 |
live | +0.016 |
wl | +0.015 |
ama | +0.015 |
Neg logit contributions | |
seen | -0.022 |
hole | -0.021 |
keeping | -0.021 |
stations | -0.021 |
contain | -0.021 |
Token | the |
Token ID | 262 |
Feature activation | -1.22e+00 |
Pos logit contributions | |
iflower | +0.072 |
pron | +0.067 |
live | +0.060 |
innie | +0.058 |
ama | +0.056 |
Neg logit contributions | |
seen | -0.070 |
hole | -0.065 |
stations | -0.064 |
keeping | -0.064 |
crashing | -0.062 |
Token | land |
Token ID | 1956 |
Feature activation | -4.18e-01 |
Pos logit contributions | |
keeping | +0.037 |
seen | +0.037 |
hole | +0.035 |
ales | +0.035 |
ously | +0.035 |
Neg logit contributions | |
iflower | -0.033 |
pron | -0.029 |
live | -0.024 |
ama | -0.024 |
wl | -0.023 |
Token | was |
Token ID | 373 |
Feature activation | -1.19e+00 |
Pos logit contributions | |
seen | +0.014 |
keeping | +0.014 |
stations | +0.014 |
hole | +0.013 |
ales | +0.013 |
Neg logit contributions | |
iflower | -0.010 |
pron | -0.009 |
live | -0.008 |
ama | -0.007 |
wl | -0.007 |
Token | ruled |
Token ID | 8879 |
Feature activation | -7.46e+00 |
Pos logit contributions | |
seen | +0.036 |
keeping | +0.035 |
hole | +0.035 |
stations | +0.035 |
ales | +0.034 |
Neg logit contributions | |
iflower | -0.033 |
pron | -0.030 |
live | -0.025 |
ama | -0.023 |
wl | -0.023 |
Token | by |
Token ID | 416 |
Feature activation | -1.98e+00 |
Pos logit contributions | |
seen | +0.184 |
keeping | +0.175 |
stations | +0.174 |
hole | +0.173 |
officers | +0.168 |
Neg logit contributions | |
iflower | -0.247 |
pron | -0.228 |
live | -0.203 |
ama | -0.192 |
wl | -0.191 |
Token | a |
Token ID | 257 |
Feature activation | -2.20e+00 |
Pos logit contributions | |
hole | +0.065 |
officers | +0.064 |
contain | +0.063 |
brake | +0.063 |
ales | +0.061 |
Neg logit contributions | |
iflower | -0.046 |
pron | -0.033 |
live | -0.030 |
wl | -0.029 |
important | -0.028 |
Token | mean |
Token ID | 1612 |
Feature activation | -1.46e+00 |
Pos logit contributions | |
seen | +0.067 |
hole | +0.063 |
crashing | +0.063 |
stations | +0.062 |
keeping | +0.060 |
Neg logit contributions | |
iflower | -0.061 |
pron | -0.055 |
live | -0.051 |
innie | -0.049 |
ama | -0.047 |
Token | monster |
Token ID | 9234 |
Feature activation | -2.17e+00 |
Pos logit contributions | |
seen | +0.044 |
keeping | +0.042 |
hole | +0.042 |
stations | +0.042 |
contain | +0.041 |
Neg logit contributions | |
iflower | -0.041 |
pron | -0.036 |
live | -0.031 |
ama | -0.030 |
wl | -0.030 |
Token | . |
Token ID | 13 |
Feature activation | -5.99e-01 |
Pos logit contributions | |
stations | +0.056 |
seen | +0.054 |
keeping | +0.054 |
ales | +0.054 |
contain | +0.053 |
Neg logit contributions | |
iflower | -0.067 |
pron | -0.058 |
live | -0.052 |
lf | -0.051 |
wl | -0.050 |
Token | big |
Token ID | 1263 |
Feature activation | -5.52e-01 |
Pos logit contributions | |
seen | +0.029 |
crashing | +0.028 |
officers | +0.027 |
stations | +0.027 |
hole | +0.027 |
Neg logit contributions | |
iflower | -0.026 |
pron | -0.023 |
live | -0.022 |
innie | -0.021 |
tto | -0.020 |
Token | , |
Token ID | 11 |
Feature activation | -8.68e-01 |
Pos logit contributions | |
seen | +0.016 |
keeping | +0.016 |
hole | +0.016 |
stations | +0.015 |
contain | +0.015 |
Neg logit contributions | |
iflower | -0.016 |
pron | -0.015 |
live | -0.012 |
ama | -0.012 |
innie | -0.012 |
Token | scary |
Token ID | 14343 |
Feature activation | -1.34e+00 |
Pos logit contributions | |
seen | +0.029 |
keeping | +0.028 |
hole | +0.028 |
stations | +0.027 |
ales | +0.026 |
Neg logit contributions | |
iflower | -0.022 |
pron | -0.020 |
live | -0.016 |
innie | -0.015 |
wl | -0.015 |
Token | lion |
Token ID | 18744 |
Feature activation | -1.66e+00 |
Pos logit contributions | |
seen | +0.039 |
hole | +0.037 |
keeping | +0.037 |
stations | +0.036 |
contain | +0.036 |
Neg logit contributions | |
iflower | -0.040 |
pron | -0.036 |
live | -0.031 |
wl | -0.030 |
ama | -0.030 |
Token | who |
Token ID | 508 |
Feature activation | +6.06e-01 |
Pos logit contributions | |
seen | +0.043 |
contain | +0.043 |
keeping | +0.043 |
stations | +0.042 |
ales | +0.042 |
Neg logit contributions | |
iflower | -0.052 |
pron | -0.045 |
live | -0.041 |
wl | -0.039 |
lf | -0.038 |
Token | ruled |
Token ID | 8879 |
Feature activation | -7.27e+00 |
Pos logit contributions | |
iflower | +0.018 |
pron | +0.017 |
live | +0.014 |
innie | +0.014 |
ama | +0.014 |
Neg logit contributions | |
seen | -0.017 |
hole | -0.016 |
keeping | -0.016 |
stations | -0.016 |
officers | -0.016 |
Token | over |
Token ID | 625 |
Feature activation | +8.17e-01 |
Pos logit contributions | |
seen | +0.185 |
hole | +0.176 |
stations | +0.173 |
keeping | +0.171 |
contain | +0.170 |
Neg logit contributions | |
iflower | -0.233 |
pron | -0.215 |
live | -0.194 |
innie | -0.183 |
wl | -0.183 |
Token | his |
Token ID | 465 |
Feature activation | -1.42e+00 |
Pos logit contributions | |
iflower | +0.019 |
pron | +0.016 |
live | +0.014 |
wl | +0.014 |
P | +0.013 |
Neg logit contributions | |
hole | -0.026 |
contain | -0.025 |
spine | -0.025 |
brake | -0.024 |
stations | -0.023 |
Token | nation |
Token ID | 3277 |
Feature activation | -1.70e+00 |
Pos logit contributions | |
seen | +0.044 |
hole | +0.042 |
keeping | +0.042 |
stations | +0.041 |
contain | +0.040 |
Neg logit contributions | |
iflower | -0.039 |
pron | -0.034 |
live | -0.030 |
ama | -0.029 |
wl | -0.028 |
Token | of |
Token ID | 286 |
Feature activation | -1.17e+00 |
Pos logit contributions | |
seen | +0.041 |
stations | +0.040 |
keeping | +0.040 |
brake | +0.039 |
ales | +0.039 |
Neg logit contributions | |
iflower | -0.054 |
pron | -0.049 |
live | -0.043 |
ama | -0.041 |
wl | -0.041 |
Token | animals |
Token ID | 4695 |
Feature activation | -2.94e+00 |
Pos logit contributions | |
hole | +0.033 |
seen | +0.033 |
contain | +0.032 |
keeping | +0.032 |
stations | +0.031 |
Neg logit contributions | |
iflower | -0.034 |
pron | -0.029 |
live | -0.025 |
wl | -0.024 |
innie | -0.024 |
Token | and |
Token ID | 290 |
Feature activation | +1.02e+00 |
Pos logit contributions | |
ked | +0.039 |
ales | +0.038 |
argued | +0.038 |
officers | +0.037 |
stations | +0.037 |
Neg logit contributions | |
iflower | -0.036 |
pron | -0.030 |
live | -0.027 |
ama | -0.027 |
tto | -0.025 |
Token | mysterious |
Token ID | 11428 |
Feature activation | -7.31e-02 |
Pos logit contributions | |
iflower | +0.023 |
pron | +0.020 |
live | +0.018 |
ama | +0.016 |
tto | +0.016 |
Neg logit contributions | |
hole | -0.034 |
stations | -0.033 |
ales | -0.033 |
keeping | -0.032 |
officers | -0.032 |
Token | land |
Token ID | 1956 |
Feature activation | -1.14e-01 |
Pos logit contributions | |
ales | +0.002 |
hole | +0.002 |
seen | +0.002 |
stations | +0.002 |
ked | +0.002 |
Neg logit contributions | |
iflower | -0.002 |
pron | -0.002 |
ama | -0.001 |
live | -0.001 |
story | -0.001 |
Token | that |
Token ID | 326 |
Feature activation | +2.30e+00 |
Pos logit contributions | |
stations | +0.003 |
seen | +0.003 |
hole | +0.003 |
keeping | +0.003 |
brake | +0.003 |
Neg logit contributions | |
iflower | -0.003 |
pron | -0.003 |
live | -0.003 |
ama | -0.003 |
wl | -0.003 |
Token | was |
Token ID | 373 |
Feature activation | -5.70e-01 |
Pos logit contributions | |
iflower | +0.062 |
pron | +0.056 |
live | +0.048 |
innie | +0.046 |
ama | +0.045 |
Neg logit contributions | |
seen | -0.072 |
hole | -0.069 |
keeping | -0.069 |
stations | -0.068 |
contain | -0.066 |
Token | ruled |
Token ID | 8879 |
Feature activation | -7.24e+00 |
Pos logit contributions | |
keeping | +0.016 |
hole | +0.016 |
seen | +0.016 |
stations | +0.016 |
ales | +0.015 |
Neg logit contributions | |
iflower | -0.016 |
pron | -0.015 |
live | -0.013 |
ama | -0.012 |
tto | -0.012 |
Token | by |
Token ID | 416 |
Feature activation | -1.48e+00 |
Pos logit contributions | |
seen | +0.184 |
stations | +0.174 |
hole | +0.173 |
keeping | +0.173 |
contain | +0.168 |
Neg logit contributions | |
iflower | -0.233 |
pron | -0.216 |
live | -0.192 |
ama | -0.183 |
wl | -0.181 |
Token | a |
Token ID | 257 |
Feature activation | -1.51e+00 |
Pos logit contributions | |
hole | +0.046 |
seen | +0.044 |
ales | +0.043 |
officers | +0.043 |
brake | +0.042 |
Neg logit contributions | |
iflower | -0.039 |
pron | -0.031 |
live | -0.028 |
ama | -0.027 |
wl | -0.026 |
Token | powerful |
Token ID | 3665 |
Feature activation | -1.29e+00 |
Pos logit contributions | |
seen | +0.047 |
hole | +0.045 |
keeping | +0.045 |
stations | +0.044 |
contain | +0.043 |
Neg logit contributions | |
iflower | -0.041 |
pron | -0.038 |
live | -0.032 |
ama | -0.030 |
innie | -0.030 |
Token | witch |
Token ID | 16365 |
Feature activation | -2.63e+00 |
Pos logit contributions | |
seen | +0.038 |
keeping | +0.038 |
hole | +0.038 |
stations | +0.037 |
ales | +0.036 |
Neg logit contributions | |
iflower | -0.036 |
pron | -0.033 |
ama | -0.027 |
live | -0.026 |
innie | -0.025 |
Token | . |
Token ID | 13 |
Feature activation | -3.84e-01 |
Pos logit contributions | |
stations | +0.063 |
seen | +0.063 |
ales | +0.063 |
keeping | +0.062 |
brake | +0.061 |
Neg logit contributions | |
iflower | -0.085 |
pron | -0.077 |
live | -0.067 |
wl | -0.065 |
ama | -0.063 |
Token | he |
Token ID | 339 |
Feature activation | -7.54e-01 |
Pos logit contributions | |
iflower | +0.064 |
pron | +0.054 |
live | +0.045 |
ama | +0.044 |
wl | +0.043 |
Neg logit contributions | |
seen | -0.084 |
hole | -0.083 |
keeping | -0.081 |
ales | -0.080 |
ets | -0.079 |
Token | was |
Token ID | 373 |
Feature activation | +1.30e-01 |
Pos logit contributions | |
keeping | +0.025 |
seen | +0.025 |
stations | +0.024 |
hole | +0.024 |
crashing | +0.024 |
Neg logit contributions | |
iflower | -0.018 |
pron | -0.016 |
live | -0.014 |
wl | -0.013 |
tto | -0.013 |
Token | in |
Token ID | 287 |
Feature activation | +3.74e-01 |
Pos logit contributions | |
iflower | +0.004 |
pron | +0.003 |
live | +0.003 |
tto | +0.003 |
ama | +0.003 |
Neg logit contributions | |
keeping | -0.004 |
hole | -0.004 |
stations | -0.004 |
ales | -0.004 |
seen | -0.003 |
Token | a |
Token ID | 257 |
Feature activation | -7.08e-01 |
Pos logit contributions | |
iflower | +0.010 |
pron | +0.009 |
live | +0.007 |
innie | +0.007 |
ama | +0.007 |
Neg logit contributions | |
ales | -0.011 |
hole | -0.011 |
ked | -0.011 |
ets | -0.011 |
ried | -0.011 |
Token | kingdom |
Token ID | 13239 |
Feature activation | -2.03e-01 |
Pos logit contributions | |
keeping | +0.022 |
seen | +0.022 |
hole | +0.021 |
ales | +0.021 |
stations | +0.021 |
Neg logit contributions | |
iflower | -0.019 |
pron | -0.018 |
live | -0.014 |
ama | -0.013 |
wl | -0.013 |
Token | ruled |
Token ID | 8879 |
Feature activation | -7.23e+00 |
Pos logit contributions | |
stations | +0.005 |
ales | +0.005 |
umb | +0.005 |
keeping | +0.005 |
ked | +0.005 |
Neg logit contributions | |
iflower | -0.006 |
pron | -0.005 |
live | -0.004 |
ama | -0.004 |
lf | -0.004 |
Token | by |
Token ID | 416 |
Feature activation | -1.32e+00 |
Pos logit contributions | |
seen | +0.185 |
hole | +0.174 |
stations | +0.174 |
keeping | +0.173 |
contain | +0.169 |
Neg logit contributions | |
iflower | -0.233 |
pron | -0.214 |
live | -0.191 |
innie | -0.181 |
ama | -0.181 |
Token | a |
Token ID | 257 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
hole | +0.041 |
seen | +0.040 |
ales | +0.039 |
ked | +0.039 |
officers | +0.039 |
Neg logit contributions | |
iflower | -0.034 |
pron | -0.027 |
live | -0.024 |
wl | -0.023 |
ama | -0.023 |
Token | wicked |
Token ID | 20589 |
Feature activation | -2.87e+00 |
Pos logit contributions | |
seen | +0.032 |
hole | +0.031 |
keeping | +0.031 |
stations | +0.030 |
contain | +0.029 |
Neg logit contributions | |
iflower | -0.028 |
pron | -0.025 |
live | -0.021 |
ama | -0.020 |
wl | -0.020 |
Token | wizard |
Token ID | 18731 |
Feature activation | -1.50e+00 |
Pos logit contributions | |
seen | +0.085 |
keeping | +0.083 |
hole | +0.082 |
stations | +0.079 |
contain | +0.077 |
Neg logit contributions | |
iflower | -0.082 |
pron | -0.073 |
ama | -0.063 |
live | -0.061 |
wl | -0.061 |
Token | . |
Token ID | 13 |
Feature activation | -1.26e-01 |
Pos logit contributions | |
stations | +0.039 |
umb | +0.038 |
brake | +0.037 |
ales | +0.037 |
footsteps | +0.037 |
Neg logit contributions | |
iflower | -0.045 |
pron | -0.041 |
live | -0.034 |
wl | -0.034 |
ama | -0.034 |
Token | They |
Token ID | 1119 |
Feature activation | -1.10e+00 |
Pos logit contributions | |
seen | +0.010 |
hole | +0.009 |
keeping | +0.009 |
stations | +0.009 |
contain | +0.009 |
Neg logit contributions | |
iflower | -0.015 |
pron | -0.014 |
live | -0.013 |
innie | -0.012 |
ama | -0.012 |
Token | said |
Token ID | 531 |
Feature activation | +8.89e-01 |
Pos logit contributions | |
seen | +0.028 |
hole | +0.026 |
keeping | +0.026 |
stations | +0.026 |
contain | +0.025 |
Neg logit contributions | |
iflower | -0.037 |
pron | -0.033 |
live | -0.030 |
innie | -0.029 |
ama | -0.029 |
Token | , |
Token ID | 11 |
Feature activation | -5.01e-01 |
Pos logit contributions | |
iflower | +0.031 |
pron | +0.028 |
live | +0.026 |
wl | +0.025 |
innie | +0.024 |
Neg logit contributions | |
seen | -0.022 |
stations | -0.020 |
crashing | -0.020 |
contain | -0.020 |
towards | -0.020 |
Token | " |
Token ID | 366 |
Feature activation | -6.90e-01 |
Pos logit contributions | |
crashing | +0.013 |
towards | +0.013 |
seen | +0.012 |
stations | +0.012 |
toward | +0.012 |
Neg logit contributions | |
iflower | -0.017 |
live | -0.014 |
pron | -0.014 |
tto | -0.013 |
wl | -0.013 |
Token | L |
Token ID | 43 |
Feature activation | -1.09e+00 |
Pos logit contributions | |
seen | +0.020 |
stations | +0.020 |
keeping | +0.019 |
hole | +0.019 |
contain | +0.018 |
Neg logit contributions | |
iflower | -0.020 |
pron | -0.018 |
live | -0.016 |
ama | -0.016 |
innie | -0.015 |
Token | adies |
Token ID | 50192 |
Feature activation | -7.08e+00 |
Pos logit contributions | |
seen | +0.033 |
keeping | +0.032 |
hole | +0.031 |
contain | +0.031 |
crashing | +0.031 |
Neg logit contributions | |
iflower | -0.030 |
pron | -0.027 |
live | -0.024 |
innie | -0.023 |
ama | -0.022 |
Token | and |
Token ID | 290 |
Feature activation | -1.72e+00 |
Pos logit contributions | |
brake | +0.226 |
Run | +0.225 |
contain | +0.223 |
rubble | +0.222 |
hole | +0.217 |
Neg logit contributions | |
iflower | -0.126 |
ama | -0.126 |
pron | -0.125 |
omm | -0.113 |
best | -0.104 |
Token | gentlemen |
Token ID | 28527 |
Feature activation | -2.67e+00 |
Pos logit contributions | |
seen | +0.032 |
stations | +0.029 |
hole | +0.028 |
officers | +0.028 |
keeping | +0.027 |
Neg logit contributions | |
iflower | -0.070 |
pron | -0.065 |
live | -0.061 |
tto | -0.058 |
innie | -0.058 |
Token | , |
Token ID | 11 |
Feature activation | -1.06e+00 |
Pos logit contributions | |
seen | +0.084 |
keeping | +0.081 |
hole | +0.080 |
stations | +0.080 |
contain | +0.077 |
Neg logit contributions | |
iflower | -0.069 |
pron | -0.063 |
live | -0.054 |
ama | -0.052 |
innie | -0.051 |
Token | welcome |
Token ID | 7062 |
Feature activation | +1.56e+00 |
Pos logit contributions | |
seen | +0.027 |
hole | +0.026 |
keeping | +0.026 |
stations | +0.025 |
ales | +0.024 |
Neg logit contributions | |
iflower | -0.034 |
pron | -0.032 |
live | -0.027 |
innie | -0.027 |
ama | -0.027 |
Token | to |
Token ID | 284 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
iflower | +0.042 |
pron | +0.037 |
ama | +0.034 |
live | +0.032 |
P | +0.032 |
Neg logit contributions | |
hole | -0.047 |
seen | -0.045 |
brake | -0.044 |
stations | -0.044 |
contain | -0.043 |
Token | night |
Token ID | 1755 |
Feature activation | -1.08e+00 |
Pos logit contributions | |
iflower | +0.005 |
pron | +0.004 |
live | +0.004 |
innie | +0.004 |
wl | +0.004 |
Neg logit contributions | |
seen | -0.006 |
crashing | -0.005 |
stations | -0.005 |
towards | -0.005 |
zone | -0.005 |
Token | , |
Token ID | 11 |
Feature activation | +4.62e-01 |
Pos logit contributions | |
seen | +0.028 |
towards | +0.025 |
crashing | +0.024 |
keeping | +0.024 |
hole | +0.024 |
Neg logit contributions | |
iflower | -0.037 |
pron | -0.033 |
live | -0.031 |
innie | -0.030 |
wl | -0.030 |
Token | the |
Token ID | 262 |
Feature activation | -5.82e-02 |
Pos logit contributions | |
iflower | +0.016 |
pron | +0.015 |
live | +0.014 |
innie | +0.013 |
tto | +0.013 |
Neg logit contributions | |
seen | -0.011 |
crashing | -0.010 |
stations | -0.010 |
towards | -0.010 |
officers | -0.010 |
Token | wild |
Token ID | 4295 |
Feature activation | -3.86e-01 |
Pos logit contributions | |
brake | +0.001 |
footsteps | +0.001 |
officers | +0.001 |
brakes | +0.001 |
seen | +0.001 |
Neg logit contributions | |
iflower | -0.002 |
pron | -0.002 |
live | -0.002 |
innie | -0.002 |
tto | -0.002 |
Token | cat |
Token ID | 3797 |
Feature activation | -4.99e-01 |
Pos logit contributions | |
seen | +0.010 |
towards | +0.009 |
hole | +0.009 |
stations | +0.009 |
officers | +0.009 |
Neg logit contributions | |
iflower | -0.013 |
pron | -0.012 |
live | -0.010 |
innie | -0.010 |
P | -0.010 |
Token | lur |
Token ID | 20605 |
Feature activation | -7.05e+00 |
Pos logit contributions | |
seen | +0.014 |
stations | +0.013 |
hole | +0.013 |
keeping | +0.012 |
officers | +0.012 |
Neg logit contributions | |
iflower | -0.015 |
pron | -0.014 |
innie | -0.012 |
live | -0.012 |
ama | -0.012 |
Token | ked |
Token ID | 9091 |
Feature activation | -6.53e-01 |
Pos logit contributions | |
keeping | +0.113 |
stations | +0.103 |
hole | +0.103 |
rubble | +0.101 |
umb | +0.101 |
Neg logit contributions | |
iflower | -0.243 |
pron | -0.235 |
P | -0.226 |
ip | -0.221 |
wl | -0.217 |
Token | through |
Token ID | 832 |
Feature activation | -8.85e-01 |
Pos logit contributions | |
seen | +0.016 |
hole | +0.016 |
keeping | +0.016 |
stations | +0.015 |
contain | +0.015 |
Neg logit contributions | |
iflower | -0.022 |
pron | -0.020 |
live | -0.017 |
ama | -0.017 |
wl | -0.017 |
Token | the |
Token ID | 262 |
Feature activation | -1.66e+00 |
Pos logit contributions | |
seen | +0.029 |
hole | +0.029 |
keeping | +0.028 |
stations | +0.027 |
contain | +0.027 |
Neg logit contributions | |
iflower | -0.022 |
pron | -0.020 |
live | -0.016 |
innie | -0.016 |
ama | -0.015 |
Token | woods |
Token ID | 16479 |
Feature activation | -1.17e+00 |
Pos logit contributions | |
seen | +0.043 |
hole | +0.041 |
keeping | +0.041 |
stations | +0.040 |
contain | +0.039 |
Neg logit contributions | |
iflower | -0.054 |
pron | -0.050 |
live | -0.044 |
innie | -0.042 |
ama | -0.042 |
Token | , |
Token ID | 11 |
Feature activation | -1.25e+00 |
Pos logit contributions | |
seen | +0.033 |
keeping | +0.033 |
brake | +0.032 |
umb | +0.031 |
hole | +0.031 |
Neg logit contributions | |
iflower | -0.032 |
pron | -0.028 |
wl | -0.022 |
live | -0.022 |
ama | -0.022 |
Token | . |
Token ID | 13 |
Feature activation | -4.43e-01 |
Pos logit contributions | |
seen | +0.035 |
keeping | +0.032 |
umb | +0.032 |
risks | +0.032 |
wanna | +0.031 |
Neg logit contributions | |
iflower | -0.032 |
pron | -0.030 |
ama | -0.026 |
on | -0.026 |
lf | -0.024 |
Token | He |
Token ID | 679 |
Feature activation | +3.64e-01 |
Pos logit contributions | |
seen | +0.012 |
hole | +0.011 |
keeping | +0.011 |
stations | +0.011 |
officers | +0.010 |
Neg logit contributions | |
iflower | -0.014 |
pron | -0.013 |
live | -0.012 |
innie | -0.011 |
ama | -0.011 |
Token | says |
Token ID | 1139 |
Feature activation | +1.34e+00 |
Pos logit contributions | |
iflower | +0.013 |
pron | +0.012 |
live | +0.011 |
innie | +0.010 |
ama | +0.010 |
Neg logit contributions | |
seen | -0.008 |
hole | -0.008 |
keeping | -0.008 |
stations | -0.008 |
contain | -0.007 |
Token | " |
Token ID | 366 |
Feature activation | -9.98e-01 |
Pos logit contributions | |
iflower | +0.043 |
pron | +0.039 |
live | +0.036 |
innie | +0.034 |
wl | +0.034 |
Neg logit contributions | |
crashing | -0.035 |
seen | -0.035 |
towards | -0.033 |
stations | -0.032 |
officers | -0.032 |
Token | L |
Token ID | 43 |
Feature activation | -1.58e+00 |
Pos logit contributions | |
seen | +0.026 |
hole | +0.025 |
keeping | +0.025 |
stations | +0.025 |
contain | +0.023 |
Neg logit contributions | |
iflower | -0.033 |
pron | -0.030 |
live | -0.027 |
ama | -0.026 |
innie | -0.026 |
Token | adies |
Token ID | 50192 |
Feature activation | -7.05e+00 |
Pos logit contributions | |
keeping | +0.053 |
hole | +0.053 |
seen | +0.053 |
stations | +0.052 |
brakes | +0.052 |
Neg logit contributions | |
iflower | -0.035 |
live | -0.030 |
pron | -0.030 |
innie | -0.026 |
wl | -0.026 |
Token | and |
Token ID | 290 |
Feature activation | -1.52e+00 |
Pos logit contributions | |
rubble | +0.247 |
Run | +0.239 |
contain | +0.238 |
brake | +0.234 |
levers | +0.230 |
Neg logit contributions | |
ama | -0.106 |
iflower | -0.097 |
omm | -0.090 |
mom | -0.090 |
pron | -0.090 |
Token | gentlemen |
Token ID | 28527 |
Feature activation | -2.61e+00 |
Pos logit contributions | |
seen | +0.032 |
hole | +0.029 |
keeping | +0.029 |
stations | +0.029 |
officers | +0.027 |
Neg logit contributions | |
iflower | -0.057 |
pron | -0.053 |
live | -0.048 |
innie | -0.046 |
ama | -0.046 |
Token | , |
Token ID | 11 |
Feature activation | -1.01e+00 |
Pos logit contributions | |
seen | +0.083 |
keeping | +0.080 |
hole | +0.078 |
stations | +0.078 |
contain | +0.076 |
Neg logit contributions | |
iflower | -0.068 |
pron | -0.062 |
live | -0.053 |
ama | -0.052 |
innie | -0.050 |
Token | boys |
Token ID | 6510 |
Feature activation | -4.47e-01 |
Pos logit contributions | |
seen | +0.028 |
keeping | +0.027 |
hole | +0.027 |
ales | +0.026 |
stations | +0.025 |
Neg logit contributions | |
iflower | -0.030 |
pron | -0.028 |
live | -0.024 |
innie | -0.024 |
ama | -0.023 |
Token | and |
Token ID | 290 |
Feature activation | -8.79e-01 |
Pos logit contributions | |
seen | +0.012 |
keeping | +0.012 |
stations | +0.011 |
hole | +0.011 |
contain | +0.011 |
Neg logit contributions | |
iflower | -0.013 |
pron | -0.012 |
live | -0.011 |
ama | -0.011 |
innie | -0.010 |
Token | a |
Token ID | 257 |
Feature activation | -1.20e+00 |
Pos logit contributions | |
iflower | +0.038 |
pron | +0.031 |
ama | +0.025 |
best | +0.025 |
lf | +0.024 |
Neg logit contributions | |
keeping | -0.066 |
seen | -0.063 |
hole | -0.063 |
spine | -0.062 |
contain | -0.062 |
Token | sign |
Token ID | 1051 |
Feature activation | -2.10e+00 |
Pos logit contributions | |
seen | +0.035 |
hole | +0.033 |
stations | +0.033 |
crashing | +0.033 |
officers | +0.032 |
Neg logit contributions | |
iflower | -0.035 |
pron | -0.031 |
live | -0.029 |
innie | -0.027 |
tto | -0.027 |
Token | that |
Token ID | 326 |
Feature activation | +1.10e+00 |
Pos logit contributions | |
seen | +0.066 |
keeping | +0.065 |
hole | +0.062 |
stations | +0.062 |
ales | +0.061 |
Neg logit contributions | |
iflower | -0.055 |
pron | -0.050 |
live | -0.042 |
ama | -0.040 |
wl | -0.039 |
Token | says |
Token ID | 1139 |
Feature activation | +5.90e-02 |
Pos logit contributions | |
iflower | +0.043 |
pron | +0.040 |
live | +0.037 |
innie | +0.036 |
tto | +0.035 |
Neg logit contributions | |
seen | -0.022 |
hole | -0.020 |
stations | -0.020 |
keeping | -0.019 |
officers | -0.019 |
Token | " |
Token ID | 366 |
Feature activation | -1.83e+00 |
Pos logit contributions | |
iflower | +0.001 |
pron | +0.001 |
live | +0.001 |
innie | +0.001 |
wl | +0.001 |
Neg logit contributions | |
seen | -0.002 |
hole | -0.002 |
stations | -0.002 |
keeping | -0.002 |
crashing | -0.002 |
Token | tem |
Token ID | 11498 |
Feature activation | -6.94e+00 |
Pos logit contributions | |
seen | +0.052 |
keeping | +0.049 |
hole | +0.049 |
stations | +0.048 |
contain | +0.046 |
Neg logit contributions | |
iflower | -0.057 |
pron | -0.052 |
live | -0.046 |
ama | -0.044 |
innie | -0.044 |
Token | ple |
Token ID | 1154 |
Feature activation | +6.66e-01 |
Pos logit contributions | |
seen | +0.164 |
keeping | +0.150 |
hole | +0.150 |
stations | +0.146 |
contain | +0.144 |
Neg logit contributions | |
iflower | -0.238 |
pron | -0.217 |
live | -0.198 |
innie | -0.192 |
wl | -0.191 |
Token | ". |
Token ID | 1911 |
Feature activation | -4.63e-01 |
Pos logit contributions | |
iflower | +0.022 |
pron | +0.021 |
live | +0.018 |
innie | +0.017 |
wl | +0.017 |
Neg logit contributions | |
seen | -0.016 |
hole | -0.016 |
keeping | -0.015 |
stations | -0.015 |
officers | -0.015 |
Token | They |
Token ID | 1119 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
seen | +0.013 |
hole | +0.012 |
keeping | +0.012 |
stations | +0.012 |
contain | +0.012 |
Neg logit contributions | |
iflower | -0.015 |
pron | -0.013 |
live | -0.012 |
wl | -0.011 |
innie | -0.011 |
Token | ask |
Token ID | 1265 |
Feature activation | -1.15e+00 |
Pos logit contributions | |
keeping | +0.036 |
seen | +0.032 |
hole | +0.030 |
ously | +0.030 |
rying | +0.029 |
Neg logit contributions | |
iflower | -0.039 |
pron | -0.037 |
ama | -0.033 |
wl | -0.033 |
irl | -0.033 |
Token | mom |
Token ID | 1995 |
Feature activation | -2.24e+00 |
Pos logit contributions | |
seen | +0.031 |
stations | +0.029 |
hole | +0.029 |
keeping | +0.029 |
officers | +0.028 |
Neg logit contributions | |
iflower | -0.037 |
pron | -0.034 |
live | -0.030 |
wl | -0.029 |
innie | -0.028 |
Token | . |
Token ID | 13 |
Feature activation | -5.78e-01 |
Pos logit contributions | |
stations | +0.004 |
seen | +0.004 |
hole | +0.004 |
ales | +0.004 |
keeping | +0.004 |
Neg logit contributions | |
iflower | -0.003 |
pron | -0.003 |
live | -0.002 |
ama | -0.002 |
innie | -0.002 |
Token |
|
Token ID | 198 |
Feature activation | +2.22e-01 |
Pos logit contributions | |
seen | +0.017 |
keeping | +0.016 |
hole | +0.016 |
towards | +0.015 |
stations | +0.015 |
Neg logit contributions | |
iflower | -0.017 |
pron | -0.015 |
live | -0.014 |
wl | -0.013 |
tto | -0.012 |
Token |
|
Token ID | 198 |
Feature activation | +3.86e-02 |
Pos logit contributions | |
iflower | +0.007 |
pron | +0.006 |
live | +0.005 |
wl | +0.005 |
ama | +0.005 |
Neg logit contributions | |
seen | -0.006 |
keeping | -0.006 |
contain | -0.005 |
towards | -0.005 |
hole | -0.005 |
Token | Luckily |
Token ID | 42332 |
Feature activation | -2.71e-01 |
Pos logit contributions | |
iflower | +0.001 |
pron | +0.001 |
ama | +0.001 |
live | +0.001 |
oppy | +0.001 |
Neg logit contributions | |
seen | -0.001 |
stations | -0.001 |
keeping | -0.001 |
contain | -0.001 |
crashing | -0.001 |
Token | Sarah |
Token ID | 10490 |
Feature activation | -4.85e-01 |
Pos logit contributions | |
seen | +0.007 |
hole | +0.007 |
stations | +0.006 |
officers | +0.006 |
keeping | +0.006 |
Neg logit contributions | |
iflower | -0.009 |
pron | -0.008 |
live | -0.007 |
wl | -0.007 |
innie | -0.007 |
Token | ret |
Token ID | 1005 |
Feature activation | -6.93e+00 |
Pos logit contributions | |
seen | +0.014 |
hole | +0.013 |
officers | +0.013 |
keeping | +0.013 |
stations | +0.012 |
Neg logit contributions | |
iflower | -0.015 |
pron | -0.014 |
live | -0.012 |
innie | -0.011 |
ama | -0.011 |
Token | uned |
Token ID | 40881 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
seen | +0.129 |
stations | +0.128 |
levers | +0.125 |
brakes | +0.125 |
crashing | +0.124 |
Neg logit contributions | |
iflower | -0.237 |
live | -0.216 |
pron | -0.213 |
innie | -0.212 |
wl | -0.212 |
Token | home |
Token ID | 1363 |
Feature activation | +1.77e-01 |
Pos logit contributions | |
seen | +0.032 |
hole | +0.031 |
keeping | +0.030 |
stations | +0.030 |
contain | +0.030 |
Neg logit contributions | |
iflower | -0.028 |
pron | -0.025 |
live | -0.022 |
ama | -0.021 |
innie | -0.021 |
Token | safe |
Token ID | 3338 |
Feature activation | -4.84e-01 |
Pos logit contributions | |
iflower | +0.005 |
pron | +0.005 |
live | +0.004 |
ama | +0.004 |
innie | +0.004 |
Neg logit contributions | |
seen | -0.005 |
hole | -0.005 |
stations | -0.005 |
keeping | -0.005 |
ales | -0.005 |
Token | and |
Token ID | 290 |
Feature activation | +1.13e+00 |
Pos logit contributions | |
seen | +0.013 |
hole | +0.013 |
stations | +0.012 |
contain | +0.012 |
brake | +0.012 |
Neg logit contributions | |
iflower | -0.015 |
pron | -0.014 |
live | -0.012 |
ama | -0.012 |
innie | -0.011 |
Token | sound |
Token ID | 2128 |
Feature activation | -1.55e+00 |
Pos logit contributions | |
iflower | +0.033 |
pron | +0.030 |
live | +0.026 |
innie | +0.025 |
ama | +0.025 |
Neg logit contributions | |
seen | -0.032 |
hole | -0.032 |
keeping | -0.031 |
stations | -0.031 |
ales | -0.030 |
Token | a |
Token ID | 257 |
Feature activation | -1.47e+00 |
Pos logit contributions | |
iflower | +0.010 |
pron | +0.009 |
live | +0.008 |
innie | +0.007 |
ama | +0.007 |
Neg logit contributions | |
seen | -0.011 |
hole | -0.010 |
keeping | -0.010 |
stations | -0.010 |
contain | -0.010 |
Token | patch |
Token ID | 8529 |
Feature activation | -1.48e-01 |
Pos logit contributions | |
seen | +0.046 |
hole | +0.044 |
crashing | +0.043 |
stations | +0.043 |
keeping | +0.043 |
Neg logit contributions | |
iflower | -0.038 |
pron | -0.034 |
live | -0.031 |
innie | -0.030 |
wl | -0.029 |
Token | of |
Token ID | 286 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
seen | +0.004 |
stations | +0.004 |
keeping | +0.004 |
brake | +0.004 |
hole | +0.003 |
Neg logit contributions | |
iflower | -0.005 |
pron | -0.005 |
live | -0.004 |
ama | -0.004 |
wl | -0.004 |
Token | butter |
Token ID | 9215 |
Feature activation | +2.54e-01 |
Pos logit contributions | |
seen | +0.030 |
hole | +0.030 |
keeping | +0.029 |
stations | +0.029 |
brake | +0.028 |
Neg logit contributions | |
iflower | -0.030 |
pron | -0.027 |
live | -0.023 |
innie | -0.021 |
ama | -0.021 |
Token | c |
Token ID | 66 |
Feature activation | -8.66e-01 |
Pos logit contributions | |
iflower | +0.008 |
pron | +0.007 |
live | +0.006 |
wl | +0.006 |
ama | +0.006 |
Neg logit contributions | |
seen | -0.007 |
stations | -0.007 |
keeping | -0.007 |
hole | -0.007 |
brake | -0.006 |
Token | ups |
Token ID | 4739 |
Feature activation | -6.90e+00 |
Pos logit contributions | |
seen | +0.026 |
stations | +0.025 |
hole | +0.025 |
brake | +0.025 |
keeping | +0.024 |
Neg logit contributions | |
iflower | -0.025 |
pron | -0.022 |
live | -0.020 |
ama | -0.019 |
wl | -0.019 |
Token | near |
Token ID | 1474 |
Feature activation | -2.28e+00 |
Pos logit contributions | |
stations | +0.187 |
brake | +0.186 |
brakes | +0.184 |
footsteps | +0.180 |
keeping | +0.179 |
Neg logit contributions | |
iflower | -0.195 |
pron | -0.185 |
live | -0.153 |
on | -0.146 |
wl | -0.142 |
Token | the |
Token ID | 262 |
Feature activation | -2.84e+00 |
Pos logit contributions | |
seen | +0.070 |
hole | +0.068 |
stations | +0.067 |
keeping | +0.066 |
contain | +0.065 |
Neg logit contributions | |
iflower | -0.061 |
pron | -0.054 |
live | -0.047 |
innie | -0.045 |
ama | -0.045 |
Token | hedge |
Token ID | 19859 |
Feature activation | -2.28e+00 |
Pos logit contributions | |
seen | +0.082 |
hole | +0.078 |
keeping | +0.076 |
stations | +0.075 |
contain | +0.074 |
Neg logit contributions | |
iflower | -0.085 |
pron | -0.077 |
live | -0.067 |
ama | -0.063 |
wl | -0.063 |
Token | . |
Token ID | 13 |
Feature activation | -2.88e-01 |
Pos logit contributions | |
seen | +0.062 |
brake | +0.061 |
brakes | +0.059 |
stations | +0.057 |
footsteps | +0.056 |
Neg logit contributions | |
iflower | -0.068 |
pron | -0.061 |
live | -0.056 |
wl | -0.053 |
ama | -0.051 |
Token | Daisy |
Token ID | 40355 |
Feature activation | +2.39e-01 |
Pos logit contributions | |
seen | +0.008 |
keeping | +0.008 |
hole | +0.007 |
stations | +0.007 |
officers | +0.007 |
Neg logit contributions | |
iflower | -0.009 |
pron | -0.008 |
live | -0.007 |
wl | -0.007 |
tto | -0.007 |
Token | 't |
Token ID | 470 |
Feature activation | -1.82e+00 |
Pos logit contributions | |
seen | +0.062 |
hole | +0.061 |
crashing | +0.060 |
brake | +0.060 |
keeping | +0.060 |
Neg logit contributions | |
iflower | -0.050 |
pron | -0.048 |
live | -0.037 |
innie | -0.036 |
wl | -0.034 |
Token | out |
Token ID | 503 |
Feature activation | -2.89e-01 |
Pos logit contributions | |
stations | +0.050 |
keeping | +0.050 |
hole | +0.048 |
umb | +0.047 |
buses | +0.046 |
Neg logit contributions | |
iflower | -0.051 |
pron | -0.043 |
live | -0.040 |
ama | -0.039 |
wl | -0.039 |
Token | run |
Token ID | 5143 |
Feature activation | +3.76e-01 |
Pos logit contributions | |
ales | +0.009 |
stations | +0.009 |
Run | +0.008 |
keeping | +0.008 |
officers | +0.008 |
Neg logit contributions | |
iflower | -0.006 |
pron | -0.005 |
story | -0.005 |
lf | -0.005 |
on | -0.005 |
Token | the |
Token ID | 262 |
Feature activation | -1.42e+00 |
Pos logit contributions | |
iflower | +0.011 |
pron | +0.010 |
live | +0.009 |
wl | +0.009 |
innie | +0.009 |
Neg logit contributions | |
seen | -0.010 |
keeping | -0.010 |
hole | -0.010 |
stations | -0.010 |
contain | -0.009 |
Token | enormous |
Token ID | 9812 |
Feature activation | -1.98e-01 |
Pos logit contributions | |
keeping | +0.046 |
hole | +0.045 |
seen | +0.044 |
contain | +0.043 |
ried | +0.043 |
Neg logit contributions | |
iflower | -0.036 |
pron | -0.031 |
story | -0.026 |
wl | -0.026 |
best | -0.025 |
Token | wolves |
Token ID | 23214 |
Feature activation | -6.86e+00 |
Pos logit contributions | |
keeping | +0.007 |
contain | +0.007 |
hole | +0.007 |
seen | +0.007 |
ried | +0.006 |
Neg logit contributions | |
iflower | -0.005 |
pron | -0.004 |
wl | -0.003 |
ama | -0.003 |
live | -0.003 |
Token | . |
Token ID | 13 |
Feature activation | -7.74e-01 |
Pos logit contributions | |
stations | +0.197 |
ales | +0.192 |
keeping | +0.189 |
umb | +0.186 |
awks | +0.186 |
Neg logit contributions | |
iflower | -0.167 |
pron | -0.145 |
on | -0.128 |
story | -0.123 |
lf | -0.123 |
Token | Stru |
Token ID | 34887 |
Feature activation | -1.68e+00 |
Pos logit contributions | |
seen | +0.022 |
hole | +0.021 |
stations | +0.021 |
keeping | +0.021 |
contain | +0.021 |
Neg logit contributions | |
iflower | -0.022 |
pron | -0.020 |
live | -0.018 |
ama | -0.017 |
innie | -0.017 |
Token | gg |
Token ID | 1130 |
Feature activation | +3.11e+00 |
Pos logit contributions | |
seen | +0.056 |
hole | +0.054 |
keeping | +0.053 |
stations | +0.053 |
officers | +0.051 |
Neg logit contributions | |
iflower | -0.043 |
pron | -0.037 |
live | -0.033 |
wl | -0.031 |
innie | -0.031 |
Token | ling |
Token ID | 1359 |
Feature activation | -2.82e-01 |
Pos logit contributions | |
iflower | +0.090 |
pron | +0.079 |
live | +0.071 |
ama | +0.068 |
wl | +0.066 |
Neg logit contributions | |
seen | -0.093 |
stations | -0.088 |
hole | -0.088 |
keeping | -0.087 |
contain | -0.084 |
Token | to |
Token ID | 284 |
Feature activation | +5.54e-01 |
Pos logit contributions | |
seen | +0.008 |
hole | +0.008 |
keeping | +0.008 |
stations | +0.008 |
contain | +0.007 |
Neg logit contributions | |
iflower | -0.008 |
pron | -0.007 |
live | -0.006 |
ama | -0.006 |
wl | -0.006 |
Token | and |
Token ID | 290 |
Feature activation | -6.07e-01 |
Pos logit contributions | |
seen | +0.062 |
stations | +0.058 |
keeping | +0.057 |
brake | +0.057 |
hole | +0.056 |
Neg logit contributions | |
iflower | -0.054 |
pron | -0.051 |
ama | -0.044 |
live | -0.043 |
wl | -0.040 |
Token | then |
Token ID | 788 |
Feature activation | +2.09e-01 |
Pos logit contributions | |
seen | +0.015 |
hole | +0.014 |
keeping | +0.014 |
stations | +0.013 |
officers | +0.013 |
Neg logit contributions | |
iflower | -0.021 |
pron | -0.019 |
live | -0.017 |
innie | -0.017 |
ama | -0.017 |
Token | a |
Token ID | 257 |
Feature activation | -2.18e+00 |
Pos logit contributions | |
iflower | +0.007 |
pron | +0.006 |
live | +0.006 |
ama | +0.005 |
innie | +0.005 |
Neg logit contributions | |
seen | -0.005 |
hole | -0.005 |
keeping | -0.005 |
stations | -0.005 |
contain | -0.005 |
Token | loud |
Token ID | 7812 |
Feature activation | -3.42e+00 |
Pos logit contributions | |
seen | +0.068 |
hole | +0.065 |
keeping | +0.065 |
stations | +0.064 |
contain | +0.061 |
Neg logit contributions | |
iflower | -0.059 |
pron | -0.054 |
live | -0.046 |
innie | -0.044 |
ama | -0.044 |
Token | " |
Token ID | 366 |
Feature activation | -1.85e+00 |
Pos logit contributions | |
seen | +0.088 |
keeping | +0.082 |
hole | +0.082 |
stations | +0.078 |
ales | +0.076 |
Neg logit contributions | |
iflower | -0.111 |
pron | -0.107 |
live | -0.088 |
ama | -0.087 |
lf | -0.087 |
Token | CR |
Token ID | 9419 |
Feature activation | -6.84e+00 |
Pos logit contributions | |
seen | +0.051 |
hole | +0.049 |
keeping | +0.049 |
stations | +0.048 |
contain | +0.046 |
Neg logit contributions | |
iflower | -0.055 |
pron | -0.051 |
live | -0.045 |
innie | -0.042 |
wl | -0.042 |
Token | ASH |
Token ID | 11211 |
Feature activation | -6.36e-01 |
Pos logit contributions | |
contain | +0.152 |
ales | +0.133 |
seen | +0.133 |
stations | +0.132 |
ets | +0.131 |
Neg logit contributions | |
iflower | -0.225 |
pron | -0.216 |
ama | -0.207 |
innie | -0.199 |
live | -0.195 |
Token | !" |
Token ID | 2474 |
Feature activation | +4.25e-01 |
Pos logit contributions | |
seen | +0.015 |
stations | +0.015 |
contain | +0.014 |
keeping | +0.013 |
hole | +0.013 |
Neg logit contributions | |
iflower | -0.021 |
pron | -0.020 |
ama | -0.019 |
live | -0.018 |
ip | -0.017 |
Token | echoed |
Token ID | 22211 |
Feature activation | -2.56e+00 |
Pos logit contributions | |
iflower | +0.010 |
pron | +0.009 |
live | +0.007 |
ama | +0.007 |
innie | +0.007 |
Neg logit contributions | |
seen | -0.015 |
stations | -0.014 |
hole | -0.014 |
keeping | -0.014 |
ales | -0.014 |
Token | around |
Token ID | 1088 |
Feature activation | -3.94e-01 |
Pos logit contributions | |
seen | +0.069 |
keeping | +0.068 |
hole | +0.068 |
stations | +0.065 |
supervision | +0.064 |
Neg logit contributions | |
iflower | -0.065 |
pron | -0.064 |
ama | -0.057 |
mummy | -0.052 |
lf | -0.052 |
Token | the |
Token ID | 262 |
Feature activation | -1.69e+00 |
Pos logit contributions | |
hole | +0.014 |
seen | +0.014 |
ales | +0.013 |
keeping | +0.013 |
contain | +0.013 |
Neg logit contributions | |
iflower | -0.008 |
pron | -0.007 |
lf | -0.005 |
ama | -0.005 |
live | -0.005 |
Token | table |
Token ID | 3084 |
Feature activation | +6.47e-01 |
Pos logit contributions | |
seen | +0.057 |
keeping | +0.054 |
hole | +0.054 |
ales | +0.053 |
contain | +0.052 |
Neg logit contributions | |
iflower | -0.052 |
pron | -0.049 |
ama | -0.038 |
live | -0.037 |
innie | -0.036 |
Token | , |
Token ID | 11 |
Feature activation | -1.13e+00 |
Pos logit contributions | |
iflower | +0.018 |
pron | +0.018 |
ama | +0.015 |
live | +0.014 |
lf | +0.014 |
Neg logit contributions | |
seen | -0.017 |
stations | -0.017 |
brake | -0.016 |
keeping | -0.016 |
brakes | -0.016 |
Token | making |
Token ID | 1642 |
Feature activation | +5.02e-01 |
Pos logit contributions | |
keeping | +0.030 |
seen | +0.029 |
ales | +0.029 |
stations | +0.029 |
hole | +0.029 |
Neg logit contributions | |
iflower | -0.033 |
pron | -0.033 |
live | -0.025 |
ama | -0.025 |
wl | -0.024 |
Token | the |
Token ID | 262 |
Feature activation | -7.36e-01 |
Pos logit contributions | |
iflower | +0.014 |
pron | +0.013 |
live | +0.011 |
ama | +0.011 |
innie | +0.010 |
Neg logit contributions | |
seen | -0.015 |
hole | -0.015 |
keeping | -0.014 |
contain | -0.014 |
stations | -0.014 |
Token | candle |
Token ID | 26839 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
seen | +0.021 |
hole | +0.020 |
keeping | +0.020 |
ales | +0.019 |
stations | +0.019 |
Neg logit contributions | |
iflower | -0.022 |
pron | -0.021 |
ama | -0.016 |
live | -0.016 |
innie | -0.016 |
Token | topple |
Token ID | 49377 |
Feature activation | -6.84e+00 |
Pos logit contributions | |
iflower | +0.034 |
pron | +0.032 |
ama | +0.026 |
live | +0.025 |
wl | +0.024 |
Neg logit contributions | |
keeping | -0.028 |
seen | -0.028 |
stations | -0.027 |
hole | -0.026 |
officers | -0.026 |
Token | over |
Token ID | 625 |
Feature activation | -1.21e+00 |
Pos logit contributions | |
contain | +0.205 |
seen | +0.203 |
awks | +0.202 |
unison | +0.201 |
supervision | +0.201 |
Neg logit contributions | |
iflower | -0.151 |
pron | -0.140 |
live | -0.118 |
ama | -0.114 |
M | -0.111 |
Token | . |
Token ID | 13 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
contain | +0.033 |
hole | +0.032 |
ales | +0.032 |
officers | +0.031 |
seen | +0.031 |
Neg logit contributions | |
iflower | -0.035 |
pron | -0.030 |
ama | -0.026 |
live | -0.024 |
innie | -0.024 |
Token | Everyone |
Token ID | 11075 |
Feature activation | -3.01e+00 |
Pos logit contributions | |
seen | +0.032 |
keeping | +0.030 |
hole | +0.030 |
stations | +0.029 |
officers | +0.028 |
Neg logit contributions | |
iflower | -0.034 |
pron | -0.031 |
live | -0.027 |
wl | -0.026 |
innie | -0.025 |
Token | quickly |
Token ID | 2952 |
Feature activation | -2.66e+00 |
Pos logit contributions | |
keeping | +0.099 |
seen | +0.098 |
hole | +0.096 |
ales | +0.096 |
stations | +0.096 |
Neg logit contributions | |
iflower | -0.072 |
pron | -0.064 |
live | -0.054 |
ama | -0.051 |
wl | -0.050 |
Token | rushed |
Token ID | 13999 |
Feature activation | -4.51e+00 |
Pos logit contributions | |
hole | +0.074 |
keeping | +0.071 |
spine | +0.069 |
stations | +0.068 |
ales | +0.068 |
Neg logit contributions | |
iflower | -0.073 |
pron | -0.068 |
live | -0.059 |
lf | -0.055 |
lly | -0.054 |
Token | became |
Token ID | 2627 |
Feature activation | +4.77e-01 |
Pos logit contributions | |
iflower | +0.013 |
pron | +0.012 |
live | +0.010 |
wl | +0.010 |
ama | +0.010 |
Neg logit contributions | |
seen | -0.011 |
keeping | -0.011 |
hole | -0.011 |
stations | -0.011 |
brake | -0.011 |
Token | good |
Token ID | 922 |
Feature activation | +9.07e-01 |
Pos logit contributions | |
iflower | +0.013 |
pron | +0.012 |
live | +0.011 |
innie | +0.010 |
wl | +0.010 |
Neg logit contributions | |
seen | -0.015 |
stations | -0.015 |
crashing | -0.014 |
officers | -0.014 |
keeping | -0.014 |
Token | friends |
Token ID | 2460 |
Feature activation | -2.83e-01 |
Pos logit contributions | |
iflower | +0.015 |
pron | +0.011 |
ama | +0.009 |
live | +0.008 |
ip | +0.007 |
Neg logit contributions | |
hole | -0.038 |
brake | -0.037 |
seen | -0.036 |
contain | -0.035 |
stations | -0.035 |
Token | . |
Token ID | 13 |
Feature activation | -5.76e-01 |
Pos logit contributions | |
brake | +0.009 |
ales | +0.008 |
stations | +0.008 |
seen | +0.008 |
brakes | +0.008 |
Neg logit contributions | |
iflower | -0.007 |
pron | -0.006 |
live | -0.005 |
ama | -0.005 |
ip | -0.005 |
Token | They |
Token ID | 1119 |
Feature activation | -8.58e-01 |
Pos logit contributions | |
seen | +0.014 |
hole | +0.013 |
keeping | +0.013 |
stations | +0.013 |
contain | +0.012 |
Neg logit contributions | |
iflower | -0.019 |
pron | -0.018 |
live | -0.016 |
innie | -0.015 |
ama | -0.015 |
Token | ruled |
Token ID | 8879 |
Feature activation | -6.83e+00 |
Pos logit contributions | |
seen | +0.025 |
hole | +0.024 |
keeping | +0.024 |
stations | +0.023 |
contain | +0.022 |
Neg logit contributions | |
iflower | -0.025 |
pron | -0.023 |
live | -0.020 |
innie | -0.019 |
ama | -0.019 |
Token | the |
Token ID | 262 |
Feature activation | -1.07e+00 |
Pos logit contributions | |
seen | +0.177 |
hole | +0.166 |
keeping | +0.166 |
stations | +0.163 |
contain | +0.156 |
Neg logit contributions | |
iflower | -0.223 |
pron | -0.205 |
live | -0.182 |
innie | -0.174 |
ama | -0.173 |
Token | land |
Token ID | 1956 |
Feature activation | +3.81e-01 |
Pos logit contributions | |
seen | +0.032 |
stations | +0.030 |
hole | +0.029 |
keeping | +0.029 |
officers | +0.029 |
Neg logit contributions | |
iflower | -0.031 |
pron | -0.028 |
live | -0.026 |
innie | -0.024 |
wl | -0.023 |
Token | together |
Token ID | 1978 |
Feature activation | -1.19e+00 |
Pos logit contributions | |
iflower | +0.012 |
pron | +0.011 |
live | +0.010 |
innie | +0.009 |
ama | +0.009 |
Neg logit contributions | |
seen | -0.010 |
hole | -0.010 |
keeping | -0.010 |
stations | -0.010 |
contain | -0.009 |
Token | . |
Token ID | 13 |
Feature activation | -6.17e-01 |
Pos logit contributions | |
seen | +0.032 |
hole | +0.030 |
keeping | +0.030 |
stations | +0.029 |
contain | +0.028 |
Neg logit contributions | |
iflower | -0.038 |
pron | -0.035 |
live | -0.031 |
innie | -0.030 |
ama | -0.029 |
Token | They |
Token ID | 1119 |
Feature activation | -9.11e-01 |
Pos logit contributions | |
seen | +0.016 |
keeping | +0.015 |
hole | +0.015 |
stations | +0.015 |
officers | +0.014 |
Neg logit contributions | |
iflower | -0.020 |
pron | -0.019 |
live | -0.016 |
innie | -0.016 |
wl | -0.016 |
Token | admired |
Token ID | 29382 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
seen | +0.002 |
keeping | +0.002 |
hole | +0.002 |
stations | +0.002 |
ales | +0.002 |
Neg logit contributions | |
iflower | -0.002 |
pron | -0.002 |
live | -0.001 |
ama | -0.001 |
innie | -0.001 |
Token | the |
Token ID | 262 |
Feature activation | +3.39e-01 |
Pos logit contributions | |
iflower | +0.030 |
pron | +0.027 |
live | +0.023 |
innie | +0.022 |
ama | +0.022 |
Neg logit contributions | |
seen | -0.036 |
hole | -0.035 |
keeping | -0.034 |
stations | -0.034 |
contain | -0.033 |
Token | way |
Token ID | 835 |
Feature activation | -3.78e-01 |
Pos logit contributions | |
iflower | +0.009 |
pron | +0.008 |
live | +0.006 |
ama | +0.006 |
wl | +0.006 |
Neg logit contributions | |
seen | -0.011 |
keeping | -0.011 |
hole | -0.011 |
stations | -0.010 |
contain | -0.010 |
Token | it |
Token ID | 340 |
Feature activation | +1.22e+00 |
Pos logit contributions | |
keeping | +0.011 |
seen | +0.011 |
stations | +0.011 |
ales | +0.011 |
brake | +0.011 |
Neg logit contributions | |
iflower | -0.009 |
pron | -0.009 |
ama | -0.008 |
live | -0.007 |
wl | -0.007 |
Token | flew |
Token ID | 13112 |
Feature activation | -3.02e+00 |
Pos logit contributions | |
iflower | +0.035 |
pron | +0.032 |
live | +0.028 |
wl | +0.026 |
ama | +0.026 |
Neg logit contributions | |
seen | -0.035 |
keeping | -0.035 |
stations | -0.034 |
hole | -0.034 |
officers | -0.033 |
Token | grace |
Token ID | 11542 |
Feature activation | -6.82e+00 |
Pos logit contributions | |
seen | +0.080 |
keeping | +0.080 |
stations | +0.078 |
supervision | +0.075 |
brake | +0.074 |
Neg logit contributions | |
iflower | -0.085 |
pron | -0.081 |
live | -0.070 |
ama | -0.068 |
P | -0.068 |
Token | fully |
Token ID | 2759 |
Feature activation | -1.75e+00 |
Pos logit contributions | |
seen | +0.175 |
stations | +0.170 |
brakes | +0.169 |
keeping | +0.168 |
brake | +0.167 |
Neg logit contributions | |
iflower | -0.203 |
pron | -0.200 |
ama | -0.167 |
live | -0.164 |
ip | -0.155 |
Token | in |
Token ID | 287 |
Feature activation | -1.68e+00 |
Pos logit contributions | |
stations | +0.049 |
seen | +0.048 |
keeping | +0.047 |
ales | +0.046 |
brake | +0.046 |
Neg logit contributions | |
iflower | -0.049 |
pron | -0.044 |
ama | -0.036 |
live | -0.035 |
ip | -0.034 |
Token | the |
Token ID | 262 |
Feature activation | -1.51e+00 |
Pos logit contributions | |
ales | +0.060 |
ked | +0.060 |
officers | +0.059 |
ets | +0.058 |
ried | +0.058 |
Neg logit contributions | |
iflower | -0.029 |
pron | -0.027 |
M | -0.021 |
P | -0.020 |
innie | -0.020 |
Token | sky |
Token ID | 6766 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
seen | +0.043 |
keeping | +0.043 |
hole | +0.041 |
ales | +0.041 |
contain | +0.040 |
Neg logit contributions | |
iflower | -0.043 |
pron | -0.039 |
P | -0.033 |
ama | -0.032 |
live | -0.032 |
Token | and |
Token ID | 290 |
Feature activation | +1.53e-01 |
Pos logit contributions | |
seen | +0.028 |
keeping | +0.027 |
brake | +0.027 |
stations | +0.027 |
officers | +0.027 |
Neg logit contributions | |
iflower | -0.030 |
pron | -0.028 |
live | -0.024 |
ama | -0.024 |
wl | -0.022 |
Token | They |
Token ID | 1119 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
seen | +0.002 |
keeping | +0.002 |
hole | +0.002 |
stations | +0.002 |
crashing | +0.002 |
Neg logit contributions | |
iflower | -0.003 |
pron | -0.003 |
live | -0.002 |
innie | -0.002 |
wl | -0.002 |
Token | say |
Token ID | 910 |
Feature activation | +1.11e+00 |
Pos logit contributions | |
keeping | +0.037 |
hole | +0.033 |
stations | +0.033 |
seen | +0.031 |
ales | +0.031 |
Neg logit contributions | |
iflower | -0.033 |
pron | -0.030 |
live | -0.029 |
irl | -0.028 |
oppy | -0.027 |
Token | , |
Token ID | 11 |
Feature activation | -7.33e-01 |
Pos logit contributions | |
iflower | +0.039 |
pron | +0.036 |
live | +0.034 |
innie | +0.032 |
wl | +0.031 |
Neg logit contributions | |
seen | -0.026 |
crashing | -0.024 |
officers | -0.024 |
stations | -0.023 |
footsteps | -0.023 |
Token | " |
Token ID | 366 |
Feature activation | -7.21e-01 |
Pos logit contributions | |
seen | +0.022 |
crashing | +0.022 |
hole | +0.021 |
stations | +0.021 |
towards | +0.021 |
Neg logit contributions | |
iflower | -0.021 |
pron | -0.017 |
live | -0.017 |
innie | -0.015 |
tto | -0.015 |
Token | We |
Token ID | 1135 |
Feature activation | -2.51e+00 |
Pos logit contributions | |
seen | +0.020 |
keeping | +0.018 |
hole | +0.018 |
stations | +0.018 |
brake | +0.017 |
Neg logit contributions | |
iflower | -0.023 |
pron | -0.021 |
live | -0.019 |
innie | -0.018 |
ama | -0.018 |
Token | do |
Token ID | 466 |
Feature activation | +4.03e+00 |
Pos logit contributions | |
keeping | +0.060 |
hole | +0.058 |
stations | +0.057 |
crashing | +0.054 |
spine | +0.053 |
Neg logit contributions | |
iflower | -0.081 |
pron | -0.076 |
live | -0.073 |
wl | -0.067 |
tto | -0.067 |
Token | not |
Token ID | 407 |
Feature activation | -1.47e+00 |
Pos logit contributions | |
iflower | +0.092 |
pron | +0.082 |
live | +0.068 |
innie | +0.064 |
ama | +0.063 |
Neg logit contributions | |
seen | -0.145 |
hole | -0.138 |
keeping | -0.138 |
stations | -0.136 |
crashing | -0.132 |
Token | like |
Token ID | 588 |
Feature activation | +1.24e-01 |
Pos logit contributions | |
hole | +0.034 |
ously | +0.031 |
stations | +0.031 |
keeping | +0.030 |
ked | +0.029 |
Neg logit contributions | |
iflower | -0.048 |
pron | -0.045 |
live | -0.042 |
P | -0.041 |
wl | -0.039 |
Token | the |
Token ID | 262 |
Feature activation | -5.13e-02 |
Pos logit contributions | |
iflower | +0.004 |
pron | +0.003 |
live | +0.003 |
innie | +0.003 |
ama | +0.003 |
Neg logit contributions | |
seen | -0.004 |
hole | -0.003 |
keeping | -0.003 |
stations | -0.003 |
contain | -0.003 |
Token | monster |
Token ID | 9234 |
Feature activation | -9.00e-02 |
Pos logit contributions | |
seen | +0.002 |
keeping | +0.002 |
hole | +0.002 |
ales | +0.002 |
stations | +0.002 |
Neg logit contributions | |
iflower | -0.001 |
pron | -0.001 |
live | -0.001 |
wl | -0.001 |
ama | -0.001 |
Token | card |
Token ID | 2657 |
Feature activation | +1.95e+00 |
Pos logit contributions | |
seen | +0.002 |
keeping | +0.002 |
hole | +0.002 |
stations | +0.002 |
contain | +0.002 |
Neg logit contributions | |
iflower | -0.003 |
pron | -0.003 |
live | -0.003 |
ama | -0.003 |
wl | -0.003 |
Token | said |
Token ID | 531 |
Feature activation | +7.20e-01 |
Pos logit contributions | |
iflower | +0.016 |
pron | +0.014 |
live | +0.013 |
innie | +0.013 |
ama | +0.013 |
Neg logit contributions | |
seen | -0.012 |
officers | -0.010 |
hole | -0.010 |
towards | -0.010 |
keeping | -0.010 |
Token | , |
Token ID | 11 |
Feature activation | -8.57e-01 |
Pos logit contributions | |
iflower | +0.026 |
pron | +0.024 |
live | +0.022 |
innie | +0.021 |
wl | +0.021 |
Neg logit contributions | |
seen | -0.016 |
towards | -0.016 |
crashing | -0.015 |
contain | -0.015 |
ously | -0.015 |
Token | " |
Token ID | 366 |
Feature activation | -8.04e-01 |
Pos logit contributions | |
crashing | +0.023 |
towards | +0.022 |
hole | +0.021 |
seen | +0.021 |
toward | +0.021 |
Neg logit contributions | |
iflower | -0.027 |
live | -0.023 |
pron | -0.023 |
wl | -0.021 |
innie | -0.021 |
Token | Hi |
Token ID | 17250 |
Feature activation | +1.91e+00 |
Pos logit contributions | |
seen | +0.021 |
keeping | +0.020 |
hole | +0.020 |
stations | +0.020 |
officers | +0.019 |
Neg logit contributions | |
iflower | -0.026 |
pron | -0.024 |
live | -0.021 |
ama | -0.020 |
innie | -0.020 |
Token | , |
Token ID | 11 |
Feature activation | +8.20e-01 |
Pos logit contributions | |
iflower | +0.063 |
pron | +0.059 |
live | +0.054 |
wl | +0.051 |
innie | +0.050 |
Neg logit contributions | |
seen | -0.049 |
hole | -0.046 |
officers | -0.044 |
crashing | -0.043 |
towards | -0.043 |
Token | what |
Token ID | 644 |
Feature activation | +4.49e+00 |
Pos logit contributions | |
iflower | +0.024 |
pron | +0.022 |
live | +0.018 |
ama | +0.018 |
innie | +0.018 |
Neg logit contributions | |
seen | -0.024 |
keeping | -0.023 |
hole | -0.023 |
stations | -0.023 |
contain | -0.022 |
Token | are |
Token ID | 389 |
Feature activation | +1.52e+00 |
Pos logit contributions | |
iflower | +0.166 |
pron | +0.155 |
live | +0.140 |
innie | +0.134 |
ama | +0.134 |
Neg logit contributions | |
seen | -0.096 |
hole | -0.089 |
keeping | -0.089 |
stations | -0.087 |
contain | -0.083 |
Token | you |
Token ID | 345 |
Feature activation | -4.01e-01 |
Pos logit contributions | |
iflower | +0.050 |
pron | +0.047 |
live | +0.041 |
innie | +0.039 |
ama | +0.039 |
Neg logit contributions | |
seen | -0.039 |
keeping | -0.035 |
officers | -0.035 |
hole | -0.035 |
stations | -0.035 |
Token | doing |
Token ID | 1804 |
Feature activation | +2.47e+00 |
Pos logit contributions | |
seen | +0.010 |
crashing | +0.009 |
hole | +0.009 |
want | +0.008 |
officers | +0.008 |
Neg logit contributions | |
iflower | -0.014 |
pron | -0.013 |
live | -0.012 |
wl | -0.011 |
innie | -0.011 |
Token | ?" |
Token ID | 1701 |
Feature activation | +2.04e+00 |
Pos logit contributions | |
iflower | +0.084 |
pron | +0.076 |
live | +0.072 |
innie | +0.069 |
wl | +0.069 |
Neg logit contributions | |
seen | -0.059 |
keeping | -0.053 |
towards | -0.053 |
hole | -0.052 |
crashing | -0.051 |
Token | The |
Token ID | 383 |
Feature activation | +7.82e-01 |
Pos logit contributions | |
iflower | +0.069 |
pron | +0.062 |
live | +0.058 |
tto | +0.053 |
wl | +0.053 |
Neg logit contributions | |
seen | -0.053 |
officers | -0.048 |
towards | -0.048 |
keeping | -0.047 |
hole | -0.047 |
Token | . |
Token ID | 13 |
Feature activation | +4.96e-01 |
Pos logit contributions | |
contain | +0.013 |
hole | +0.013 |
umb | +0.013 |
brake | +0.012 |
stations | +0.012 |
Neg logit contributions | |
iflower | -0.012 |
pron | -0.010 |
live | -0.009 |
ama | -0.009 |
tto | -0.008 |
Token | |
Token ID | 220 |
Feature activation | +2.99e+00 |
Pos logit contributions | |
iflower | +0.015 |
pron | +0.014 |
live | +0.012 |
wl | +0.011 |
innie | +0.011 |
Neg logit contributions | |
seen | -0.014 |
keeping | -0.013 |
hole | -0.013 |
towards | -0.013 |
stations | -0.012 |
Token |
|
Token ID | 198 |
Feature activation | +7.28e-01 |
Pos logit contributions | |
iflower | +0.089 |
pron | +0.084 |
live | +0.066 |
wl | +0.063 |
innie | +0.062 |
Neg logit contributions | |
seen | -0.087 |
hole | -0.084 |
keeping | -0.083 |
crashing | -0.078 |
towards | -0.077 |
Token |
|
Token ID | 198 |
Feature activation | +4.54e-01 |
Pos logit contributions | |
iflower | +0.023 |
pron | +0.023 |
ama | +0.019 |
omm | +0.019 |
innie | +0.018 |
Neg logit contributions | |
seen | -0.017 |
keeping | -0.016 |
contain | -0.016 |
towards | -0.016 |
hole | -0.015 |
Token | From |
Token ID | 4863 |
Feature activation | -5.38e-01 |
Pos logit contributions | |
iflower | +0.012 |
pron | +0.011 |
ama | +0.011 |
oppy | +0.010 |
innie | +0.010 |
Neg logit contributions | |
seen | -0.012 |
contain | -0.012 |
stations | -0.012 |
crashing | -0.011 |
ets | -0.011 |
Token | that |
Token ID | 326 |
Feature activation | +5.15e+00 |
Pos logit contributions | |
seen | +0.014 |
hole | +0.012 |
officers | +0.012 |
stations | +0.012 |
keeping | +0.012 |
Neg logit contributions | |
iflower | -0.018 |
pron | -0.017 |
live | -0.014 |
innie | -0.014 |
ama | -0.014 |
Token | day |
Token ID | 1110 |
Feature activation | +2.00e+00 |
Pos logit contributions | |
iflower | +0.166 |
pron | +0.146 |
live | +0.140 |
lly | +0.129 |
oppy | +0.128 |
Neg logit contributions | |
seen | -0.121 |
towards | -0.119 |
crashing | -0.115 |
want | -0.114 |
hole | -0.108 |
Token | on |
Token ID | 319 |
Feature activation | +2.35e+00 |
Pos logit contributions | |
iflower | +0.069 |
pron | +0.065 |
live | +0.056 |
innie | +0.053 |
wl | +0.052 |
Neg logit contributions | |
seen | -0.047 |
crashing | -0.045 |
hole | -0.044 |
towards | -0.043 |
brake | -0.042 |
Token | , |
Token ID | 11 |
Feature activation | +1.22e+00 |
Pos logit contributions | |
iflower | +0.096 |
pron | +0.094 |
live | +0.082 |
innie | +0.078 |
wl | +0.078 |
Neg logit contributions | |
seen | -0.034 |
towards | -0.033 |
crashing | -0.031 |
ously | -0.030 |
horror | -0.030 |
Token | Jack |
Token ID | 3619 |
Feature activation | +2.64e+00 |
Pos logit contributions | |
iflower | +0.044 |
pron | +0.042 |
live | +0.040 |
innie | +0.038 |
wl | +0.038 |
Neg logit contributions | |
seen | -0.026 |
towards | -0.024 |
contain | -0.023 |
crashing | -0.023 |
want | -0.023 |
Token | and |
Token ID | 290 |
Feature activation | +2.07e-01 |
Pos logit contributions | |
pron | +0.092 |
iflower | +0.092 |
live | +0.080 |
innie | +0.080 |
ama | +0.078 |
Neg logit contributions | |
seen | -0.058 |
stations | -0.050 |
ets | -0.050 |
keeping | -0.048 |
hole | -0.048 |
Token | Lily |
Token ID | 20037 |
Feature activation | +1.49e-01 |
Pos logit contributions | |
iflower | +0.013 |
pron | +0.011 |
live | +0.010 |
wl | +0.010 |
tto | +0.010 |
Neg logit contributions | |
seen | -0.011 |
footsteps | -0.010 |
crashing | -0.010 |
officers | -0.010 |
keeping | -0.010 |
Token | . |
Token ID | 13 |
Feature activation | +1.19e-03 |
Pos logit contributions | |
iflower | +0.006 |
pron | +0.005 |
live | +0.005 |
wl | +0.005 |
innie | +0.005 |
Neg logit contributions | |
seen | -0.003 |
hole | -0.003 |
keeping | -0.003 |
stations | -0.002 |
towards | -0.002 |
Token |
|
Token ID | 198 |
Feature activation | +1.34e+00 |
Pos logit contributions | |
iflower | +0.000 |
pron | +0.000 |
live | +0.000 |
wl | +0.000 |
tto | +0.000 |
Neg logit contributions | |
seen | -0.000 |
towards | -0.000 |
crashing | -0.000 |
hole | -0.000 |
footsteps | -0.000 |
Token |
|
Token ID | 198 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
iflower | +0.047 |
pron | +0.044 |
live | +0.037 |
wl | +0.036 |
tto | +0.034 |
Neg logit contributions | |
seen | -0.030 |
hole | -0.028 |
keeping | -0.027 |
towards | -0.027 |
crashing | -0.026 |
Token | " |
Token ID | 1 |
Feature activation | -2.18e+00 |
Pos logit contributions | |
iflower | +0.030 |
pron | +0.025 |
ama | +0.022 |
live | +0.022 |
innie | +0.022 |
Neg logit contributions | |
seen | -0.036 |
stations | -0.035 |
hole | -0.034 |
keeping | -0.034 |
towards | -0.034 |
Token | Here |
Token ID | 4342 |
Feature activation | +4.12e+00 |
Pos logit contributions | |
Run | +0.053 |
seen | +0.052 |
stations | +0.050 |
brake | +0.048 |
contain | +0.047 |
Neg logit contributions | |
iflower | -0.070 |
ama | -0.061 |
pron | -0.058 |
oppy | -0.058 |
wl | -0.055 |
Token | you |
Token ID | 345 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
iflower | +0.138 |
pron | +0.138 |
live | +0.118 |
innie | +0.116 |
tto | +0.114 |
Neg logit contributions | |
seen | -0.096 |
hole | -0.086 |
officers | -0.083 |
keeping | -0.082 |
want | -0.081 |
Token | go |
Token ID | 467 |
Feature activation | +7.90e-01 |
Pos logit contributions | |
seen | +0.028 |
keeping | +0.027 |
hole | +0.026 |
stations | +0.026 |
officers | +0.024 |
Neg logit contributions | |
iflower | -0.051 |
pron | -0.047 |
live | -0.043 |
ama | -0.042 |
innie | -0.041 |
Token | . |
Token ID | 13 |
Feature activation | +6.50e-01 |
Pos logit contributions | |
iflower | +0.022 |
pron | +0.020 |
live | +0.018 |
ama | +0.017 |
innie | +0.017 |
Neg logit contributions | |
seen | -0.024 |
hole | -0.023 |
stations | -0.023 |
keeping | -0.023 |
contain | -0.022 |
Token | Tell |
Token ID | 14026 |
Feature activation | -5.13e-01 |
Pos logit contributions | |
iflower | +0.022 |
pron | +0.021 |
live | +0.019 |
wl | +0.018 |
innie | +0.017 |
Neg logit contributions | |
seen | -0.015 |
hole | -0.014 |
keeping | -0.014 |
stations | -0.014 |
crashing | -0.014 |
Token | me |
Token ID | 502 |
Feature activation | +2.59e-02 |
Pos logit contributions | |
seen | +0.012 |
permission | +0.010 |
ales | +0.010 |
officers | +0.010 |
stations | +0.010 |
Neg logit contributions | |
iflower | -0.018 |
pron | -0.018 |
live | -0.016 |
wl | -0.015 |
tto | -0.015 |
Token | nodded |
Token ID | 14464 |
Feature activation | +4.81e-02 |
Pos logit contributions | |
seen | +0.009 |
argued | +0.008 |
hole | +0.008 |
officers | +0.007 |
gasp | +0.007 |
Neg logit contributions | |
iflower | -0.018 |
pron | -0.016 |
live | -0.015 |
wl | -0.015 |
ama | -0.015 |
Token | and |
Token ID | 290 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
iflower | +0.002 |
pron | +0.002 |
live | +0.001 |
innie | +0.001 |
ama | +0.001 |
Neg logit contributions | |
seen | -0.001 |
hole | -0.001 |
keeping | -0.001 |
stations | -0.001 |
contain | -0.001 |
Token | said |
Token ID | 531 |
Feature activation | +3.54e-01 |
Pos logit contributions | |
keeping | +0.005 |
seen | +0.005 |
stations | +0.005 |
hole | +0.005 |
crashing | +0.005 |
Neg logit contributions | |
iflower | -0.005 |
pron | -0.005 |
live | -0.004 |
ama | -0.004 |
innie | -0.004 |
Token | , |
Token ID | 11 |
Feature activation | -8.07e-01 |
Pos logit contributions | |
iflower | +0.011 |
pron | +0.010 |
live | +0.010 |
wl | +0.009 |
innie | +0.009 |
Neg logit contributions | |
seen | -0.009 |
hole | -0.009 |
crashing | -0.008 |
keeping | -0.008 |
stations | -0.008 |
Token | " |
Token ID | 366 |
Feature activation | -5.88e-01 |
Pos logit contributions | |
seen | +0.024 |
crashing | +0.024 |
hole | +0.023 |
towards | +0.023 |
contain | +0.022 |
Neg logit contributions | |
iflower | -0.023 |
pron | -0.019 |
live | -0.019 |
wl | -0.017 |
innie | -0.017 |
Token | That |
Token ID | 2504 |
Feature activation | +4.07e+00 |
Pos logit contributions | |
stations | +0.014 |
seen | +0.014 |
hole | +0.013 |
keeping | +0.013 |
brake | +0.013 |
Neg logit contributions | |
iflower | -0.019 |
pron | -0.017 |
live | -0.016 |
innie | -0.015 |
ama | -0.015 |
Token | sounds |
Token ID | 5238 |
Feature activation | -1.89e+00 |
Pos logit contributions | |
iflower | +0.138 |
pron | +0.129 |
live | +0.117 |
innie | +0.113 |
wl | +0.110 |
Neg logit contributions | |
seen | -0.096 |
hole | -0.092 |
stations | -0.091 |
keeping | -0.088 |
officers | -0.085 |
Token | good |
Token ID | 922 |
Feature activation | +9.48e-01 |
Pos logit contributions | |
seen | +0.060 |
hole | +0.058 |
keeping | +0.058 |
stations | +0.057 |
officers | +0.055 |
Neg logit contributions | |
iflower | -0.050 |
pron | -0.045 |
live | -0.039 |
innie | -0.037 |
ama | -0.036 |
Token | , |
Token ID | 11 |
Feature activation | +7.34e-01 |
Pos logit contributions | |
iflower | +0.029 |
pron | +0.026 |
live | +0.023 |
ama | +0.022 |
innie | +0.022 |
Neg logit contributions | |
seen | -0.027 |
hole | -0.026 |
keeping | -0.026 |
stations | -0.025 |
contain | -0.025 |
Token | grandma |
Token ID | 49890 |
Feature activation | +1.16e+00 |
Pos logit contributions | |
iflower | +0.019 |
pron | +0.017 |
live | +0.015 |
wl | +0.014 |
tto | +0.014 |
Neg logit contributions | |
seen | -0.024 |
hole | -0.023 |
stations | -0.022 |
keeping | -0.022 |
officers | -0.022 |
Token | . |
Token ID | 13 |
Feature activation | +3.08e-01 |
Pos logit contributions | |
iflower | +0.047 |
pron | +0.044 |
live | +0.041 |
innie | +0.040 |
wl | +0.039 |
Neg logit contributions | |
seen | -0.020 |
hole | -0.018 |
keeping | -0.018 |
stations | -0.017 |
officers | -0.016 |
Token | ed |
Token ID | 276 |
Feature activation | +1.02e+00 |
Pos logit contributions | |
iflower | +0.044 |
pron | +0.040 |
live | +0.037 |
ama | +0.035 |
innie | +0.034 |
Neg logit contributions | |
seen | -0.038 |
hole | -0.037 |
brake | -0.036 |
stations | -0.034 |
ales | -0.034 |
Token | with |
Token ID | 351 |
Feature activation | +6.40e-01 |
Pos logit contributions | |
iflower | +0.026 |
pron | +0.024 |
live | +0.020 |
ama | +0.020 |
innie | +0.019 |
Neg logit contributions | |
seen | -0.032 |
hole | -0.031 |
keeping | -0.030 |
stations | -0.030 |
ales | -0.030 |
Token | delight |
Token ID | 10974 |
Feature activation | +1.56e+00 |
Pos logit contributions | |
iflower | +0.020 |
pron | +0.019 |
innie | +0.017 |
live | +0.017 |
ama | +0.016 |
Neg logit contributions | |
hole | -0.016 |
seen | -0.015 |
spine | -0.015 |
ales | -0.014 |
keeping | -0.014 |
Token | . |
Token ID | 13 |
Feature activation | -1.61e-01 |
Pos logit contributions | |
iflower | +0.051 |
pron | +0.046 |
ama | +0.042 |
live | +0.040 |
innie | +0.039 |
Neg logit contributions | |
seen | -0.042 |
brake | -0.039 |
hole | -0.039 |
stations | -0.038 |
brakes | -0.038 |
Token |
|
Token ID | 198 |
Feature activation | +2.60e-01 |
Pos logit contributions | |
hole | +0.004 |
stations | +0.004 |
crashing | +0.004 |
seen | +0.004 |
keeping | +0.004 |
Neg logit contributions | |
iflower | -0.005 |
pron | -0.005 |
live | -0.004 |
P | -0.004 |
tto | -0.004 |
Token |
|
Token ID | 198 |
Feature activation | +6.04e-02 |
Pos logit contributions | |
iflower | +0.009 |
pron | +0.008 |
live | +0.007 |
heddar | +0.006 |
wl | +0.006 |
Neg logit contributions | |
crashing | -0.006 |
stations | -0.006 |
seen | -0.006 |
hole | -0.006 |
footsteps | -0.005 |
Token | At |
Token ID | 2953 |
Feature activation | -6.97e-01 |
Pos logit contributions | |
iflower | +0.002 |
pron | +0.002 |
ama | +0.001 |
oppy | +0.001 |
live | +0.001 |
Neg logit contributions | |
stations | -0.002 |
crashing | -0.002 |
hole | -0.001 |
brake | -0.001 |
seen | -0.001 |
Token | the |
Token ID | 262 |
Feature activation | +3.46e-01 |
Pos logit contributions | |
seen | +0.018 |
keeping | +0.016 |
stations | +0.016 |
hole | +0.016 |
contain | +0.016 |
Neg logit contributions | |
iflower | -0.023 |
pron | -0.022 |
live | -0.019 |
wl | -0.019 |
innie | -0.018 |
Token | end |
Token ID | 886 |
Feature activation | +4.50e-01 |
Pos logit contributions | |
iflower | +0.011 |
pron | +0.010 |
live | +0.009 |
innie | +0.008 |
ama | +0.008 |
Neg logit contributions | |
seen | -0.009 |
hole | -0.009 |
keeping | -0.009 |
stations | -0.009 |
contain | -0.008 |
Token | , |
Token ID | 11 |
Feature activation | +3.92e-01 |
Pos logit contributions | |
iflower | +0.019 |
pron | +0.018 |
live | +0.016 |
wl | +0.016 |
tto | +0.015 |
Neg logit contributions | |
seen | -0.008 |
hole | -0.007 |
stations | -0.007 |
crashing | -0.007 |
towards | -0.006 |
Token | Chloe |
Token ID | 29476 |
Feature activation | +6.93e-01 |
Pos logit contributions | |
iflower | +0.013 |
pron | +0.011 |
live | +0.010 |
wl | +0.010 |
tto | +0.010 |
Neg logit contributions | |
seen | -0.010 |
stations | -0.010 |
hole | -0.010 |
crashing | -0.009 |
keeping | -0.009 |
Token | a |
Token ID | 257 |
Feature activation | -9.54e-01 |
Pos logit contributions | |
seen | +0.059 |
hole | +0.058 |
keeping | +0.057 |
contain | +0.056 |
stations | +0.055 |
Neg logit contributions | |
iflower | -0.052 |
pron | -0.046 |
live | -0.040 |
innie | -0.039 |
ama | -0.038 |
Token | few |
Token ID | 1178 |
Feature activation | -6.58e-01 |
Pos logit contributions | |
seen | +0.026 |
hole | +0.024 |
stations | +0.024 |
crashing | +0.023 |
contain | +0.023 |
Neg logit contributions | |
iflower | -0.030 |
pron | -0.027 |
live | -0.025 |
innie | -0.023 |
ama | -0.023 |
Token | minutes |
Token ID | 2431 |
Feature activation | -3.01e+00 |
Pos logit contributions | |
keeping | +0.018 |
hole | +0.017 |
seen | +0.016 |
ales | +0.016 |
contain | +0.015 |
Neg logit contributions | |
iflower | -0.020 |
pron | -0.018 |
lf | -0.016 |
live | -0.016 |
innie | -0.015 |
Token | , |
Token ID | 11 |
Feature activation | -2.88e-01 |
Pos logit contributions | |
keeping | +0.103 |
stations | +0.100 |
seen | +0.099 |
hole | +0.098 |
brake | +0.097 |
Neg logit contributions | |
iflower | -0.062 |
pron | -0.058 |
ama | -0.051 |
live | -0.051 |
innie | -0.047 |
Token | the |
Token ID | 262 |
Feature activation | -1.14e+00 |
Pos logit contributions | |
seen | +0.010 |
hole | +0.009 |
keeping | +0.009 |
stations | +0.009 |
contain | +0.009 |
Neg logit contributions | |
iflower | -0.007 |
pron | -0.006 |
live | -0.005 |
ama | -0.005 |
innie | -0.005 |
Token | car |
Token ID | 1097 |
Feature activation | -2.30e-01 |
Pos logit contributions | |
seen | +0.039 |
hole | +0.039 |
keeping | +0.038 |
contain | +0.037 |
ales | +0.036 |
Neg logit contributions | |
iflower | -0.027 |
pron | -0.026 |
ama | -0.022 |
innie | -0.019 |
live | -0.018 |
Token | stopped |
Token ID | 5025 |
Feature activation | -1.02e+00 |
Pos logit contributions | |
keeping | +0.007 |
seen | +0.006 |
hole | +0.006 |
stations | +0.006 |
contain | +0.006 |
Neg logit contributions | |
iflower | -0.006 |
pron | -0.006 |
live | -0.005 |
ama | -0.005 |
wl | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | -1.15e+00 |
Pos logit contributions | |
keeping | +0.030 |
seen | +0.029 |
contain | +0.028 |
ets | +0.028 |
stations | +0.028 |
Neg logit contributions | |
iflower | -0.029 |
pron | -0.026 |
ama | -0.023 |
wl | -0.021 |
innie | -0.021 |
Token | They |
Token ID | 1119 |
Feature activation | -1.61e+00 |
Pos logit contributions | |
seen | +0.028 |
keeping | +0.027 |
hole | +0.027 |
stations | +0.027 |
officers | +0.026 |
Neg logit contributions | |
iflower | -0.039 |
pron | -0.036 |
live | -0.032 |
wl | -0.031 |
innie | -0.030 |
Token | looked |
Token ID | 3114 |
Feature activation | +5.08e-01 |
Pos logit contributions | |
keeping | +0.045 |
hole | +0.042 |
seen | +0.042 |
stations | +0.041 |
ales | +0.040 |
Neg logit contributions | |
iflower | -0.045 |
pron | -0.042 |
live | -0.037 |
wl | -0.035 |
ama | -0.035 |
Token | in |
Token ID | 287 |
Feature activation | +3.81e-02 |
Pos logit contributions | |
iflower | +0.012 |
pron | +0.011 |
important | +0.008 |
best | +0.008 |
ip | +0.008 |
Neg logit contributions | |
keeping | -0.018 |
stations | -0.016 |
hole | -0.015 |
seen | -0.015 |
ales | -0.015 |
Token | the |
Token ID | 262 |
Feature activation | -6.27e-01 |
Pos logit contributions | |
iflower | +0.023 |
pron | +0.021 |
live | +0.019 |
innie | +0.018 |
ama | +0.018 |
Neg logit contributions | |
seen | -0.023 |
keeping | -0.022 |
hole | -0.022 |
stations | -0.021 |
officers | -0.021 |
Token | ornament |
Token ID | 45618 |
Feature activation | +2.92e+00 |
Pos logit contributions | |
seen | +0.018 |
hole | +0.016 |
keeping | +0.016 |
stations | +0.016 |
brake | +0.016 |
Neg logit contributions | |
iflower | -0.019 |
pron | -0.018 |
live | -0.016 |
innie | -0.015 |
wl | -0.015 |
Token | if |
Token ID | 611 |
Feature activation | -2.25e-02 |
Pos logit contributions | |
iflower | +0.085 |
pron | +0.077 |
live | +0.069 |
innie | +0.066 |
tto | +0.063 |
Neg logit contributions | |
seen | -0.086 |
hole | -0.081 |
keeping | -0.080 |
stations | -0.078 |
contain | -0.077 |
Token | you |
Token ID | 345 |
Feature activation | -4.81e+00 |
Pos logit contributions | |
seen | +0.001 |
officers | +0.001 |
keeping | +0.000 |
hole | +0.000 |
towards | +0.000 |
Neg logit contributions | |
iflower | -0.001 |
pron | -0.001 |
live | -0.001 |
innie | -0.001 |
tto | -0.001 |
Token | can |
Token ID | 460 |
Feature activation | -9.68e-01 |
Pos logit contributions | |
stations | +0.095 |
keeping | +0.094 |
hole | +0.091 |
seen | +0.091 |
crashing | +0.086 |
Neg logit contributions | |
iflower | -0.179 |
pron | -0.167 |
live | -0.150 |
ama | -0.146 |
wl | -0.145 |
Token | have |
Token ID | 423 |
Feature activation | +1.37e+00 |
Pos logit contributions | |
seen | +0.022 |
hole | +0.020 |
contain | +0.019 |
keeping | +0.019 |
brake | +0.019 |
Neg logit contributions | |
iflower | -0.035 |
pron | -0.033 |
live | -0.030 |
innie | -0.028 |
wl | -0.028 |
Token | it |
Token ID | 340 |
Feature activation | +1.22e+00 |
Pos logit contributions | |
iflower | +0.045 |
pron | +0.042 |
live | +0.039 |
ama | +0.037 |
wl | +0.037 |
Neg logit contributions | |
seen | -0.037 |
hole | -0.032 |
keeping | -0.031 |
stations | -0.031 |
officers | -0.030 |
Token | ?" |
Token ID | 1701 |
Feature activation | +1.64e+00 |
Pos logit contributions | |
iflower | +0.037 |
pron | +0.033 |
live | +0.031 |
innie | +0.030 |
wl | +0.029 |
Neg logit contributions | |
seen | -0.035 |
hole | -0.032 |
keeping | -0.031 |
contain | -0.030 |
towards | -0.030 |
Token | So |
Token ID | 1406 |
Feature activation | -5.18e-02 |
Pos logit contributions | |
iflower | +0.056 |
pron | +0.052 |
live | +0.045 |
tto | +0.043 |
wl | +0.042 |
Neg logit contributions | |
seen | -0.041 |
keeping | -0.040 |
officers | -0.040 |
towards | -0.038 |
hole | -0.038 |
Token | , |
Token ID | 11 |
Feature activation | +2.65e-01 |
Pos logit contributions | |
seen | +0.001 |
keeping | +0.001 |
hole | +0.001 |
officers | +0.001 |
stations | +0.001 |
Neg logit contributions | |
iflower | -0.002 |
pron | -0.002 |
live | -0.001 |
innie | -0.001 |
wl | -0.001 |
Token | Twe |
Token ID | 24205 |
Feature activation | -1.73e-02 |
Pos logit contributions | |
iflower | +0.009 |
pron | +0.008 |
live | +0.007 |
wl | +0.007 |
innie | +0.007 |
Neg logit contributions | |
seen | -0.007 |
keeping | -0.007 |
officers | -0.007 |
hole | -0.007 |
stations | -0.006 |
Token | little |
Token ID | 1310 |
Feature activation | -6.92e-01 |
Pos logit contributions | |
keeping | +0.029 |
seen | +0.029 |
stations | +0.028 |
ales | +0.028 |
contain | +0.027 |
Neg logit contributions | |
iflower | -0.024 |
pron | -0.022 |
ama | -0.017 |
innie | -0.017 |
live | -0.016 |
Token | creature |
Token ID | 7185 |
Feature activation | -2.10e+00 |
Pos logit contributions | |
ales | +0.023 |
keeping | +0.023 |
seen | +0.023 |
stations | +0.023 |
ets | +0.023 |
Neg logit contributions | |
iflower | -0.015 |
pron | -0.014 |
ama | -0.011 |
very | -0.010 |
mummy | -0.010 |
Token | dressed |
Token ID | 12049 |
Feature activation | +1.24e+00 |
Pos logit contributions | |
stations | +0.060 |
ales | +0.058 |
keeping | +0.057 |
brake | +0.056 |
brakes | +0.055 |
Neg logit contributions | |
iflower | -0.059 |
pron | -0.053 |
live | -0.045 |
on | -0.044 |
lf | -0.042 |
Token | in |
Token ID | 287 |
Feature activation | -5.92e-01 |
Pos logit contributions | |
iflower | +0.032 |
pron | +0.031 |
live | +0.022 |
tto | +0.022 |
ama | +0.022 |
Neg logit contributions | |
stations | -0.039 |
seen | -0.039 |
ales | -0.038 |
brake | -0.038 |
ets | -0.038 |
Token | green |
Token ID | 4077 |
Feature activation | +2.04e-01 |
Pos logit contributions | |
ets | +0.021 |
ales | +0.020 |
ried | +0.020 |
hole | +0.020 |
stations | +0.019 |
Neg logit contributions | |
iflower | -0.013 |
pron | -0.012 |
M | -0.008 |
innie | -0.008 |
lf | -0.008 |
Token | . |
Token ID | 13 |
Feature activation | -9.68e-02 |
Pos logit contributions | |
iflower | +0.006 |
pron | +0.005 |
P | +0.004 |
mint | +0.004 |
bean | +0.004 |
Neg logit contributions | |
brake | -0.006 |
ried | -0.006 |
contain | -0.006 |
seen | -0.006 |
wanna | -0.005 |
Token | The |
Token ID | 383 |
Feature activation | +4.55e-01 |
Pos logit contributions | |
seen | +0.003 |
hole | +0.003 |
keeping | +0.003 |
officers | +0.003 |
stations | +0.003 |
Neg logit contributions | |
iflower | -0.003 |
pron | -0.003 |
live | -0.002 |
wl | -0.002 |
innie | -0.002 |
Token | creature |
Token ID | 7185 |
Feature activation | +9.26e-02 |
Pos logit contributions | |
iflower | +0.014 |
pron | +0.013 |
live | +0.011 |
ama | +0.011 |
innie | +0.011 |
Neg logit contributions | |
seen | -0.013 |
hole | -0.012 |
keeping | -0.012 |
stations | -0.012 |
crashing | -0.011 |
Token | said |
Token ID | 531 |
Feature activation | +3.66e-02 |
Pos logit contributions | |
iflower | +0.003 |
pron | +0.002 |
live | +0.002 |
innie | +0.002 |
ama | +0.002 |
Neg logit contributions | |
seen | -0.003 |
hole | -0.003 |
keeping | -0.002 |
contain | -0.002 |
stations | -0.002 |
Token | , |
Token ID | 11 |
Feature activation | -1.07e+00 |
Pos logit contributions | |
iflower | +0.001 |
pron | +0.001 |
live | +0.001 |
innie | +0.001 |
wl | +0.001 |
Neg logit contributions | |
seen | -0.001 |
hole | -0.001 |
keeping | -0.001 |
crashing | -0.001 |
stations | -0.001 |
Token | " |
Token ID | 366 |
Feature activation | -1.06e+00 |
Pos logit contributions | |
seen | +0.033 |
hole | +0.032 |
crashing | +0.032 |
keeping | +0.031 |
stations | +0.031 |
Neg logit contributions | |
iflower | -0.029 |
pron | -0.025 |
live | -0.023 |
innie | -0.021 |
wl | -0.021 |
Token | help |
Token ID | 1037 |
Feature activation | +4.63e-01 |
Pos logit contributions | |
iflower | +0.029 |
pron | +0.027 |
live | +0.024 |
ama | +0.023 |
innie | +0.023 |
Neg logit contributions | |
hole | -0.020 |
seen | -0.020 |
stations | -0.019 |
keeping | -0.019 |
ets | -0.018 |
Token | the |
Token ID | 262 |
Feature activation | +1.22e-01 |
Pos logit contributions | |
iflower | +0.015 |
pron | +0.014 |
live | +0.012 |
innie | +0.012 |
wl | +0.012 |
Neg logit contributions | |
seen | -0.012 |
hole | -0.011 |
keeping | -0.011 |
stations | -0.011 |
contain | -0.011 |
Token | dog |
Token ID | 3290 |
Feature activation | -1.42e+00 |
Pos logit contributions | |
iflower | +0.004 |
pron | +0.003 |
live | +0.003 |
ama | +0.003 |
innie | +0.003 |
Neg logit contributions | |
seen | -0.004 |
hole | -0.003 |
keeping | -0.003 |
stations | -0.003 |
contain | -0.003 |
Token | learn |
Token ID | 2193 |
Feature activation | -5.05e-01 |
Pos logit contributions | |
seen | +0.034 |
hole | +0.034 |
stations | +0.034 |
keeping | +0.033 |
crashing | +0.032 |
Neg logit contributions | |
iflower | -0.046 |
pron | -0.042 |
live | -0.037 |
ama | -0.036 |
wl | -0.034 |
Token | to |
Token ID | 284 |
Feature activation | +4.44e-01 |
Pos logit contributions | |
seen | +0.014 |
keeping | +0.013 |
hole | +0.013 |
spine | +0.013 |
ked | +0.013 |
Neg logit contributions | |
iflower | -0.015 |
pron | -0.014 |
live | -0.012 |
innie | -0.011 |
ama | -0.011 |
Token | balance |
Token ID | 5236 |
Feature activation | +1.90e+00 |
Pos logit contributions | |
iflower | +0.015 |
pron | +0.014 |
live | +0.013 |
innie | +0.012 |
ama | +0.012 |
Neg logit contributions | |
seen | -0.011 |
hole | -0.010 |
keeping | -0.010 |
stations | -0.010 |
officers | -0.009 |
Token | . |
Token ID | 13 |
Feature activation | +2.49e-04 |
Pos logit contributions | |
iflower | +0.063 |
pron | +0.057 |
live | +0.049 |
innie | +0.047 |
wl | +0.047 |
Neg logit contributions | |
seen | -0.049 |
hole | -0.046 |
keeping | -0.045 |
stations | -0.045 |
contain | -0.044 |
Token | So |
Token ID | 1406 |
Feature activation | -4.86e-02 |
Pos logit contributions | |
iflower | +0.000 |
pron | +0.000 |
live | +0.000 |
wl | +0.000 |
innie | +0.000 |
Neg logit contributions | |
seen | -0.000 |
hole | -0.000 |
keeping | -0.000 |
stations | -0.000 |
officers | -0.000 |
Token | , |
Token ID | 11 |
Feature activation | +8.29e-01 |
Pos logit contributions | |
seen | +0.001 |
hole | +0.001 |
keeping | +0.001 |
stations | +0.001 |
officers | +0.001 |
Neg logit contributions | |
iflower | -0.002 |
pron | -0.001 |
live | -0.001 |
innie | -0.001 |
wl | -0.001 |
Token | the |
Token ID | 262 |
Feature activation | +5.96e-01 |
Pos logit contributions | |
iflower | +0.024 |
pron | +0.021 |
live | +0.019 |
innie | +0.018 |
wl | +0.017 |
Neg logit contributions | |
seen | -0.025 |
keeping | -0.024 |
hole | -0.024 |
stations | -0.023 |
officers | -0.023 |
Token | wise |
Token ID | 10787 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
iflower | +0.020 |
pron | +0.019 |
live | +0.017 |
innie | +0.016 |
wl | +0.016 |
Neg logit contributions | |
seen | -0.014 |
hole | -0.013 |
stations | -0.013 |
keeping | -0.013 |
officers | -0.013 |
Token | quite |
Token ID | 2407 |
Feature activation | -1.48e+00 |
Pos logit contributions | |
seen | +0.038 |
keeping | +0.038 |
hole | +0.038 |
stations | +0.037 |
officers | +0.036 |
Neg logit contributions | |
iflower | -0.026 |
pron | -0.024 |
live | -0.020 |
ama | -0.018 |
innie | -0.018 |
Token | dark |
Token ID | 3223 |
Feature activation | -3.28e+00 |
Pos logit contributions | |
hole | +0.046 |
keeping | +0.046 |
stations | +0.045 |
seen | +0.045 |
ales | +0.043 |
Neg logit contributions | |
iflower | -0.038 |
pron | -0.036 |
live | -0.030 |
tto | -0.028 |
ama | -0.027 |
Token | out |
Token ID | 503 |
Feature activation | -7.10e-01 |
Pos logit contributions | |
contain | +0.105 |
brake | +0.104 |
Run | +0.103 |
awks | +0.103 |
brakes | +0.102 |
Neg logit contributions | |
iflower | -0.069 |
pron | -0.062 |
time | -0.051 |
lf | -0.047 |
live | -0.047 |
Token | . |
Token ID | 13 |
Feature activation | -4.15e-01 |
Pos logit contributions | |
keeping | +0.022 |
stations | +0.022 |
dashboard | +0.022 |
brake | +0.022 |
brakes | +0.022 |
Neg logit contributions | |
iflower | -0.016 |
pron | -0.013 |
ama | -0.013 |
on | -0.011 |
mummy | -0.011 |
Token | His |
Token ID | 2399 |
Feature activation | +7.48e-01 |
Pos logit contributions | |
seen | +0.011 |
hole | +0.011 |
keeping | +0.011 |
crashing | +0.010 |
officers | +0.010 |
Neg logit contributions | |
iflower | -0.013 |
pron | -0.012 |
live | -0.010 |
wl | -0.010 |
innie | -0.010 |
Token | owner |
Token ID | 4870 |
Feature activation | -1.86e+00 |
Pos logit contributions | |
iflower | +0.021 |
pron | +0.019 |
live | +0.016 |
innie | +0.016 |
wl | +0.015 |
Neg logit contributions | |
seen | -0.023 |
keeping | -0.021 |
officers | -0.021 |
hole | -0.021 |
stations | -0.021 |
Token | was |
Token ID | 373 |
Feature activation | -5.95e-01 |
Pos logit contributions | |
seen | +0.054 |
keeping | +0.054 |
hole | +0.053 |
stations | +0.053 |
ales | +0.051 |
Neg logit contributions | |
iflower | -0.051 |
pron | -0.047 |
live | -0.042 |
ama | -0.039 |
innie | -0.039 |
Token | waiting |
Token ID | 4953 |
Feature activation | +2.52e+00 |
Pos logit contributions | |
seen | +0.016 |
hole | +0.015 |
keeping | +0.015 |
stations | +0.015 |
contain | +0.014 |
Neg logit contributions | |
iflower | -0.019 |
pron | -0.018 |
live | -0.015 |
innie | -0.015 |
ama | -0.015 |
Token | patiently |
Token ID | 37220 |
Feature activation | -2.57e+00 |
Pos logit contributions | |
pron | +0.060 |
iflower | +0.057 |
ama | +0.045 |
best | +0.043 |
important | +0.040 |
Neg logit contributions | |
keeping | -0.079 |
stations | -0.079 |
ales | -0.077 |
Run | -0.074 |
hole | -0.074 |
Token | and |
Token ID | 290 |
Feature activation | +2.44e-01 |
Pos logit contributions | |
seen | +0.073 |
stations | +0.072 |
ales | +0.072 |
keeping | +0.071 |
angs | +0.070 |
Neg logit contributions | |
iflower | -0.067 |
pron | -0.064 |
ama | -0.051 |
live | -0.048 |
very | -0.048 |
Token | walked |
Token ID | 6807 |
Feature activation | -2.64e+00 |
Pos logit contributions | |
iflower | +0.007 |
pron | +0.007 |
live | +0.006 |
ama | +0.005 |
innie | +0.005 |
Neg logit contributions | |
seen | -0.007 |
hole | -0.007 |
keeping | -0.007 |
stations | -0.007 |
ales | -0.006 |
Token | on |
Token ID | 319 |
Feature activation | -1.69e+00 |
Pos logit contributions | |
iflower | +0.066 |
pron | +0.063 |
live | +0.056 |
ama | +0.052 |
wl | +0.051 |
Neg logit contributions | |
seen | -0.049 |
keeping | -0.048 |
stations | -0.048 |
hole | -0.046 |
crashing | -0.046 |
Token | the |
Token ID | 262 |
Feature activation | -2.57e+00 |
Pos logit contributions | |
seen | +0.050 |
hole | +0.047 |
keeping | +0.047 |
stations | +0.047 |
contain | +0.045 |
Neg logit contributions | |
iflower | -0.049 |
pron | -0.045 |
live | -0.039 |
innie | -0.037 |
ama | -0.037 |
Token | table |
Token ID | 3084 |
Feature activation | +1.25e-02 |
Pos logit contributions | |
seen | +0.074 |
hole | +0.069 |
keeping | +0.067 |
stations | +0.065 |
crashing | +0.065 |
Neg logit contributions | |
iflower | -0.077 |
pron | -0.071 |
live | -0.061 |
innie | -0.058 |
ama | -0.057 |
Token | . |
Token ID | 13 |
Feature activation | -4.80e-01 |
Pos logit contributions | |
pron | +0.000 |
iflower | +0.000 |
live | +0.000 |
lf | +0.000 |
ama | +0.000 |
Neg logit contributions | |
brake | -0.000 |
brakes | -0.000 |
keeping | -0.000 |
ales | -0.000 |
levers | -0.000 |
Token | She |
Token ID | 1375 |
Feature activation | +2.94e-02 |
Pos logit contributions | |
seen | +0.011 |
keeping | +0.011 |
hole | +0.011 |
officers | +0.011 |
stations | +0.010 |
Neg logit contributions | |
iflower | -0.017 |
pron | -0.015 |
live | -0.014 |
wl | -0.014 |
tto | -0.013 |
Token | wanted |
Token ID | 2227 |
Feature activation | -8.99e-01 |
Pos logit contributions | |
iflower | +0.001 |
pron | +0.001 |
wl | +0.001 |
live | +0.001 |
ama | +0.001 |
Neg logit contributions | |
seen | -0.001 |
argued | -0.001 |
officers | -0.001 |
backed | -0.000 |
ked | -0.000 |
Token | to |
Token ID | 284 |
Feature activation | +7.45e-01 |
Pos logit contributions | |
seen | +0.019 |
keeping | +0.015 |
officers | +0.015 |
stations | +0.014 |
towards | +0.014 |
Neg logit contributions | |
iflower | -0.037 |
pron | -0.034 |
live | -0.032 |
wl | -0.031 |
ama | -0.030 |
Token | give |
Token ID | 1577 |
Feature activation | +1.94e+00 |
Pos logit contributions | |
iflower | +0.022 |
pron | +0.022 |
live | +0.019 |
innie | +0.018 |
ama | +0.018 |
Neg logit contributions | |
hole | -0.020 |
stations | -0.019 |
seen | -0.019 |
keeping | -0.018 |
ales | -0.018 |
Token | her |
Token ID | 607 |
Feature activation | +3.99e-01 |
Pos logit contributions | |
iflower | +0.072 |
pron | +0.066 |
live | +0.062 |
wl | +0.060 |
innie | +0.058 |
Neg logit contributions | |
seen | -0.045 |
keeping | -0.041 |
towards | -0.039 |
officers | -0.039 |
hole | -0.038 |
Token | grandma |
Token ID | 49890 |
Feature activation | -2.98e-02 |
Pos logit contributions | |
iflower | +0.012 |
pron | +0.011 |
live | +0.009 |
innie | +0.009 |
ama | +0.009 |
Neg logit contributions | |
seen | -0.012 |
hole | -0.011 |
keeping | -0.011 |
stations | -0.011 |
crashing | -0.011 |
Token | a |
Token ID | 257 |
Feature activation | -2.54e-01 |
Pos logit contributions | |
seen | +0.001 |
hole | +0.001 |
keeping | +0.001 |
stations | +0.001 |
officers | +0.001 |
Neg logit contributions | |
iflower | -0.001 |
pron | -0.001 |
live | -0.001 |
innie | -0.001 |
wl | -0.001 |
Token | at |
Token ID | 379 |
Feature activation | +8.25e-01 |
Pos logit contributions | |
iflower | +0.025 |
pron | +0.023 |
live | +0.021 |
innie | +0.020 |
ama | +0.020 |
Neg logit contributions | |
seen | -0.019 |
keeping | -0.018 |
hole | -0.018 |
stations | -0.018 |
contain | -0.017 |
Token | Fl |
Token ID | 1610 |
Feature activation | +1.65e+00 |
Pos logit contributions | |
iflower | +0.026 |
pron | +0.025 |
ama | +0.020 |
live | +0.020 |
wl | +0.019 |
Neg logit contributions | |
seen | -0.023 |
stations | -0.022 |
towards | -0.021 |
keeping | -0.021 |
hole | -0.021 |
Token | uffy |
Token ID | 15352 |
Feature activation | +8.47e-01 |
Pos logit contributions | |
iflower | +0.032 |
pron | +0.031 |
innie | +0.021 |
lf | +0.020 |
wl | +0.020 |
Neg logit contributions | |
hole | -0.058 |
seen | -0.057 |
ales | -0.057 |
ked | -0.055 |
owing | -0.054 |
Token | and |
Token ID | 290 |
Feature activation | +6.24e-01 |
Pos logit contributions | |
iflower | +0.029 |
pron | +0.027 |
live | +0.024 |
innie | +0.023 |
ama | +0.023 |
Neg logit contributions | |
seen | -0.020 |
hole | -0.019 |
keeping | -0.019 |
stations | -0.019 |
officers | -0.018 |
Token | said |
Token ID | 531 |
Feature activation | +5.95e-01 |
Pos logit contributions | |
iflower | +0.024 |
pron | +0.022 |
innie | +0.021 |
ama | +0.021 |
live | +0.021 |
Neg logit contributions | |
seen | -0.013 |
argued | -0.011 |
towards | -0.011 |
ked | -0.011 |
officers | -0.011 |
Token | , |
Token ID | 11 |
Feature activation | -8.96e-01 |
Pos logit contributions | |
iflower | +0.021 |
pron | +0.019 |
live | +0.017 |
innie | +0.017 |
wl | +0.016 |
Neg logit contributions | |
seen | -0.014 |
towards | -0.013 |
crashing | -0.013 |
hole | -0.013 |
ked | -0.013 |
Token | " |
Token ID | 366 |
Feature activation | -8.51e-01 |
Pos logit contributions | |
crashing | +0.026 |
seen | +0.025 |
towards | +0.025 |
hole | +0.024 |
officers | +0.024 |
Neg logit contributions | |
iflower | -0.026 |
pron | -0.023 |
live | -0.022 |
innie | -0.020 |
wl | -0.020 |
Token | Sure |
Token ID | 19457 |
Feature activation | +1.73e+00 |
Pos logit contributions | |
seen | +0.023 |
towards | +0.021 |
contain | +0.021 |
hole | +0.021 |
stations | +0.020 |
Neg logit contributions | |
iflower | -0.028 |
pron | -0.025 |
ama | -0.022 |
innie | -0.022 |
live | -0.022 |
Token | ! |
Token ID | 0 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
iflower | +0.041 |
pron | +0.036 |
live | +0.032 |
ama | +0.031 |
innie | +0.029 |
Neg logit contributions | |
stations | -0.059 |
keeping | -0.058 |
seen | -0.058 |
hole | -0.058 |
contain | -0.055 |
Token | My |
Token ID | 2011 |
Feature activation | +8.29e-01 |
Pos logit contributions | |
seen | +0.004 |
hole | +0.004 |
keeping | +0.004 |
stations | +0.004 |
crashing | +0.003 |
Neg logit contributions | |
iflower | -0.007 |
pron | -0.006 |
live | -0.005 |
innie | -0.005 |
wl | -0.005 |
Token | name |
Token ID | 1438 |
Feature activation | +1.13e+00 |
Pos logit contributions | |
iflower | +0.027 |
pron | +0.025 |
live | +0.022 |
innie | +0.021 |
wl | +0.021 |
Neg logit contributions | |
seen | -0.022 |
hole | -0.021 |
keeping | -0.020 |
stations | -0.020 |
brake | -0.019 |
Token | replied |
Token ID | 8712 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
iflower | +0.067 |
pron | +0.063 |
live | +0.056 |
innie | +0.054 |
P | +0.053 |
Neg logit contributions | |
argued | -0.033 |
seen | -0.033 |
ked | -0.028 |
towards | -0.027 |
backed | -0.027 |
Token | , |
Token ID | 11 |
Feature activation | -1.37e+00 |
Pos logit contributions | |
iflower | +0.004 |
pron | +0.004 |
live | +0.004 |
innie | +0.004 |
wl | +0.004 |
Neg logit contributions | |
crashing | -0.002 |
towards | -0.002 |
ked | -0.002 |
seen | -0.002 |
footsteps | -0.002 |
Token | " |
Token ID | 366 |
Feature activation | -1.01e+00 |
Pos logit contributions | |
crashing | +0.037 |
towards | +0.034 |
seen | +0.034 |
ked | +0.034 |
officers | +0.033 |
Neg logit contributions | |
iflower | -0.043 |
pron | -0.037 |
live | -0.037 |
innie | -0.035 |
wl | -0.035 |
Token | I |
Token ID | 40 |
Feature activation | -9.90e-01 |
Pos logit contributions | |
seen | +0.030 |
towards | +0.028 |
hole | +0.028 |
keeping | +0.028 |
crashing | +0.028 |
Neg logit contributions | |
iflower | -0.028 |
pron | -0.027 |
innie | -0.023 |
ama | -0.022 |
live | -0.022 |
Token | have |
Token ID | 423 |
Feature activation | +1.53e+00 |
Pos logit contributions | |
seen | +0.025 |
hole | +0.023 |
keeping | +0.022 |
stations | +0.022 |
contain | +0.021 |
Neg logit contributions | |
iflower | -0.034 |
pron | -0.031 |
live | -0.028 |
innie | -0.027 |
ama | -0.027 |
Token | never |
Token ID | 1239 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
iflower | +0.045 |
pron | +0.041 |
live | +0.036 |
innie | +0.034 |
wl | +0.033 |
Neg logit contributions | |
hole | -0.043 |
seen | -0.042 |
keeping | -0.041 |
stations | -0.041 |
contain | -0.040 |
Token | seen |
Token ID | 1775 |
Feature activation | +1.44e+00 |
Pos logit contributions | |
stations | +0.015 |
hole | +0.014 |
keeping | +0.013 |
levers | +0.012 |
crashing | +0.012 |
Neg logit contributions | |
iflower | -0.059 |
pron | -0.056 |
live | -0.052 |
P | -0.050 |
tto | -0.049 |
Token | a |
Token ID | 257 |
Feature activation | -8.05e-01 |
Pos logit contributions | |
iflower | +0.047 |
pron | +0.043 |
live | +0.039 |
innie | +0.037 |
wl | +0.037 |
Neg logit contributions | |
seen | -0.038 |
hole | -0.035 |
keeping | -0.034 |
stations | -0.034 |
officers | -0.033 |
Token | bald |
Token ID | 28273 |
Feature activation | -8.84e-01 |
Pos logit contributions | |
seen | +0.023 |
hole | +0.021 |
crashing | +0.021 |
officers | +0.021 |
brake | +0.021 |
Neg logit contributions | |
iflower | -0.024 |
pron | -0.021 |
live | -0.021 |
innie | -0.020 |
wl | -0.019 |
Token | man |
Token ID | 582 |
Feature activation | -2.63e+00 |
Pos logit contributions | |
seen | +0.031 |
hole | +0.030 |
stations | +0.030 |
keeping | +0.030 |
contain | +0.029 |
Neg logit contributions | |
iflower | -0.021 |
pron | -0.019 |
live | -0.015 |
ama | -0.014 |
innie | -0.013 |
Token | before |
Token ID | 878 |
Feature activation | -1.48e+00 |
Pos logit contributions | |
seen | +0.063 |
stations | +0.062 |
keeping | +0.061 |
hole | +0.060 |
officers | +0.057 |
Neg logit contributions | |
iflower | -0.088 |
pron | -0.082 |
live | -0.071 |
ama | -0.068 |
innie | -0.067 |
Token | . |
Token ID | 13 |
Feature activation | -8.97e-01 |
Pos logit contributions | |
seen | +0.052 |
Run | +0.052 |
stations | +0.051 |
keeping | +0.051 |
ked | +0.051 |
Neg logit contributions | |
iflower | -0.047 |
pron | -0.045 |
lf | -0.035 |
ama | -0.034 |
live | -0.033 |
Token | |
Token ID | 220 |
Feature activation | +1.82e+00 |
Pos logit contributions | |
seen | +0.025 |
hole | +0.023 |
keeping | +0.023 |
stations | +0.023 |
crashing | +0.022 |
Neg logit contributions | |
iflower | -0.028 |
pron | -0.025 |
live | -0.023 |
wl | -0.021 |
innie | -0.021 |
Token |
|
Token ID | 198 |
Feature activation | -3.12e-01 |
Pos logit contributions | |
iflower | +0.045 |
pron | +0.041 |
live | +0.033 |
innie | +0.031 |
wl | +0.031 |
Neg logit contributions | |
seen | -0.062 |
hole | -0.060 |
keeping | -0.059 |
crashing | -0.057 |
stations | -0.057 |
Token |
|
Token ID | 198 |
Feature activation | -3.83e-01 |
Pos logit contributions | |
seen | +0.010 |
keeping | +0.009 |
hole | +0.009 |
crashing | +0.009 |
towards | +0.009 |
Neg logit contributions | |
iflower | -0.009 |
pron | -0.008 |
live | -0.007 |
wl | -0.006 |
ama | -0.006 |
Token | When |
Token ID | 2215 |
Feature activation | -2.04e+00 |
Pos logit contributions | |
seen | +0.012 |
keeping | +0.011 |
stations | +0.011 |
crashing | +0.011 |
hole | +0.011 |
Neg logit contributions | |
iflower | -0.011 |
pron | -0.009 |
live | -0.008 |
ama | -0.008 |
wl | -0.008 |
Token | he |
Token ID | 339 |
Feature activation | -8.51e-01 |
Pos logit contributions | |
seen | +0.062 |
keeping | +0.059 |
hole | +0.058 |
stations | +0.058 |
crashing | +0.056 |
Neg logit contributions | |
iflower | -0.058 |
pron | -0.052 |
live | -0.046 |
innie | -0.043 |
wl | -0.043 |
Token | got |
Token ID | 1392 |
Feature activation | +1.53e+00 |
Pos logit contributions | |
seen | +0.021 |
hole | +0.019 |
keeping | +0.018 |
contain | +0.018 |
stations | +0.018 |
Neg logit contributions | |
iflower | -0.030 |
pron | -0.027 |
live | -0.025 |
innie | -0.024 |
wl | -0.024 |
Token | there |
Token ID | 612 |
Feature activation | -4.34e-01 |
Pos logit contributions | |
iflower | +0.051 |
pron | +0.047 |
live | +0.042 |
ama | +0.041 |
innie | +0.040 |
Neg logit contributions | |
seen | -0.037 |
hole | -0.036 |
keeping | -0.036 |
stations | -0.036 |
contain | -0.033 |
Token | , |
Token ID | 11 |
Feature activation | +6.71e-01 |
Pos logit contributions | |
seen | +0.013 |
hole | +0.013 |
keeping | +0.012 |
stations | +0.012 |
contain | +0.012 |
Neg logit contributions | |
iflower | -0.012 |
pron | -0.011 |
live | -0.010 |
innie | -0.009 |
ama | -0.009 |
Token | he |
Token ID | 339 |
Feature activation | +2.86e-01 |
Pos logit contributions | |
iflower | +0.018 |
pron | +0.016 |
live | +0.014 |
innie | +0.013 |
ama | +0.013 |
Neg logit contributions | |
seen | -0.021 |
hole | -0.020 |
keeping | -0.020 |
stations | -0.020 |
officers | -0.019 |
Token | saw |
Token ID | 2497 |
Feature activation | +1.02e+00 |
Pos logit contributions | |
iflower | +0.010 |
pron | +0.009 |
live | +0.008 |
innie | +0.008 |
ama | +0.008 |
Neg logit contributions | |
seen | -0.007 |
hole | -0.006 |
stations | -0.006 |
keeping | -0.006 |
officers | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | -2.07e-01 |
Pos logit contributions | |
seen | +0.050 |
stations | +0.049 |
Run | +0.047 |
ried | +0.047 |
rying | +0.046 |
Neg logit contributions | |
iflower | -0.048 |
pron | -0.045 |
live | -0.038 |
ama | -0.037 |
lf | -0.037 |
Token | Suddenly |
Token ID | 24975 |
Feature activation | -1.09e-02 |
Pos logit contributions | |
crashing | +0.005 |
keeping | +0.005 |
officers | +0.005 |
hole | +0.005 |
seen | +0.005 |
Neg logit contributions | |
iflower | -0.007 |
pron | -0.007 |
live | -0.006 |
lf | -0.005 |
wl | -0.005 |
Token | , |
Token ID | 11 |
Feature activation | -3.26e-01 |
Pos logit contributions | |
seen | +0.000 |
hole | +0.000 |
keeping | +0.000 |
stations | +0.000 |
officers | +0.000 |
Neg logit contributions | |
iflower | -0.000 |
pron | -0.000 |
live | -0.000 |
innie | -0.000 |
wl | -0.000 |
Token | Ham |
Token ID | 4345 |
Feature activation | +8.52e-01 |
Pos logit contributions | |
seen | +0.009 |
keeping | +0.008 |
officers | +0.008 |
stations | +0.008 |
hole | +0.008 |
Neg logit contributions | |
iflower | -0.010 |
pron | -0.009 |
live | -0.008 |
innie | -0.008 |
wl | -0.008 |
Token | ster |
Token ID | 1706 |
Feature activation | +7.31e-01 |
Pos logit contributions | |
iflower | +0.026 |
pron | +0.025 |
live | +0.021 |
innie | +0.020 |
ama | +0.020 |
Neg logit contributions | |
seen | -0.023 |
hole | -0.022 |
officers | -0.020 |
ales | -0.020 |
keeping | -0.020 |
Token | tri |
Token ID | 1333 |
Feature activation | -5.74e+00 |
Pos logit contributions | |
iflower | +0.024 |
pron | +0.022 |
live | +0.020 |
innie | +0.019 |
ama | +0.019 |
Neg logit contributions | |
seen | -0.018 |
hole | -0.017 |
keeping | -0.017 |
stations | -0.017 |
contain | -0.016 |
Token | pped |
Token ID | 1496 |
Feature activation | -1.40e+00 |
Pos logit contributions | |
stations | +0.138 |
seen | +0.137 |
keeping | +0.135 |
officers | +0.133 |
hole | +0.131 |
Neg logit contributions | |
iflower | -0.185 |
pron | -0.172 |
live | -0.159 |
tto | -0.151 |
P | -0.150 |
Token | and |
Token ID | 290 |
Feature activation | -9.52e-01 |
Pos logit contributions | |
seen | +0.037 |
stations | +0.035 |
keeping | +0.035 |
contain | +0.035 |
hole | +0.034 |
Neg logit contributions | |
iflower | -0.043 |
pron | -0.040 |
live | -0.035 |
tto | -0.033 |
ama | -0.033 |
Token | the |
Token ID | 262 |
Feature activation | -1.92e+00 |
Pos logit contributions | |
seen | +0.028 |
stations | +0.027 |
hole | +0.027 |
keeping | +0.027 |
ales | +0.026 |
Neg logit contributions | |
iflower | -0.027 |
pron | -0.024 |
live | -0.022 |
tto | -0.020 |
ama | -0.020 |
Token | bone |
Token ID | 9970 |
Feature activation | +1.25e+00 |
Pos logit contributions | |
seen | +0.060 |
stations | +0.058 |
keeping | +0.058 |
hole | +0.058 |
contain | +0.056 |
Neg logit contributions | |
iflower | -0.052 |
pron | -0.048 |
live | -0.040 |
ama | -0.038 |
innie | -0.038 |
Token | flew |
Token ID | 13112 |
Feature activation | -3.25e+00 |
Pos logit contributions | |
iflower | +0.034 |
pron | +0.031 |
live | +0.026 |
tto | +0.024 |
lf | +0.024 |
Neg logit contributions | |
keeping | -0.038 |
stations | -0.037 |
seen | -0.037 |
ales | -0.037 |
hole | -0.035 |
Token |
|
Token ID | 198 |
Feature activation | -3.21e-01 |
Pos logit contributions | |
seen | +0.003 |
towards | +0.003 |
hole | +0.003 |
crashing | +0.002 |
toward | +0.002 |
Neg logit contributions | |
iflower | -0.004 |
pron | -0.004 |
live | -0.003 |
wl | -0.003 |
innie | -0.003 |
Token | The |
Token ID | 464 |
Feature activation | +5.81e-01 |
Pos logit contributions | |
seen | +0.009 |
contain | +0.009 |
towards | +0.009 |
crashing | +0.009 |
toward | +0.008 |
Neg logit contributions | |
iflower | -0.009 |
pron | -0.008 |
live | -0.007 |
innie | -0.007 |
wl | -0.007 |
Token | fire |
Token ID | 2046 |
Feature activation | -4.82e-01 |
Pos logit contributions | |
iflower | +0.017 |
pron | +0.016 |
live | +0.014 |
ama | +0.013 |
innie | +0.013 |
Neg logit contributions | |
seen | -0.017 |
hole | -0.016 |
keeping | -0.016 |
contain | -0.015 |
ales | -0.015 |
Token | work |
Token ID | 1818 |
Feature activation | +1.95e+00 |
Pos logit contributions | |
seen | +0.016 |
keeping | +0.016 |
hole | +0.016 |
stations | +0.015 |
contain | +0.015 |
Neg logit contributions | |
iflower | -0.012 |
pron | -0.010 |
live | -0.009 |
ama | -0.009 |
wl | -0.008 |
Token | show |
Token ID | 905 |
Feature activation | -8.34e-01 |
Pos logit contributions | |
iflower | +0.059 |
pron | +0.054 |
live | +0.048 |
ama | +0.045 |
innie | +0.045 |
Neg logit contributions | |
seen | -0.054 |
keeping | -0.052 |
hole | -0.051 |
stations | -0.050 |
contain | -0.048 |
Token | lasted |
Token ID | 15436 |
Feature activation | -6.24e+00 |
Pos logit contributions | |
seen | +0.023 |
hole | +0.023 |
keeping | +0.023 |
stations | +0.022 |
brake | +0.022 |
Neg logit contributions | |
iflower | -0.023 |
pron | -0.021 |
live | -0.019 |
ama | -0.018 |
wl | -0.018 |
Token | for |
Token ID | 329 |
Feature activation | -1.48e+00 |
Pos logit contributions | |
keeping | +0.212 |
officers | +0.197 |
ales | +0.196 |
hole | +0.195 |
brakes | +0.188 |
Neg logit contributions | |
pron | -0.130 |
iflower | -0.130 |
important | -0.101 |
innie | -0.099 |
mummy | -0.097 |
Token | what |
Token ID | 644 |
Feature activation | +3.21e+00 |
Pos logit contributions | |
hole | +0.044 |
seen | +0.043 |
keeping | +0.042 |
ried | +0.040 |
stations | +0.040 |
Neg logit contributions | |
iflower | -0.041 |
pron | -0.037 |
live | -0.032 |
innie | -0.031 |
ama | -0.029 |
Token | seemed |
Token ID | 3947 |
Feature activation | -3.82e+00 |
Pos logit contributions | |
iflower | +0.090 |
pron | +0.082 |
live | +0.071 |
innie | +0.069 |
ama | +0.067 |
Neg logit contributions | |
seen | -0.099 |
keeping | -0.092 |
hole | -0.092 |
stations | -0.091 |
officers | -0.088 |
Token | like |
Token ID | 588 |
Feature activation | -2.30e+00 |
Pos logit contributions | |
keeping | +0.092 |
hole | +0.084 |
stations | +0.080 |
ales | +0.080 |
brake | +0.074 |
Neg logit contributions | |
iflower | -0.120 |
pron | -0.115 |
important | -0.109 |
special | -0.105 |
famous | -0.101 |
Token | hours |
Token ID | 2250 |
Feature activation | -2.11e+00 |
Pos logit contributions | |
hole | +0.068 |
seen | +0.066 |
keeping | +0.065 |
ales | +0.064 |
stations | +0.063 |
Neg logit contributions | |
iflower | -0.064 |
pron | -0.060 |
live | -0.049 |
ama | -0.047 |
innie | -0.046 |
Token | the |
Token ID | 262 |
Feature activation | -9.58e-01 |
Pos logit contributions | |
iflower | +0.002 |
pron | +0.001 |
live | +0.001 |
innie | +0.001 |
wl | +0.001 |
Neg logit contributions | |
seen | -0.002 |
hole | -0.001 |
stations | -0.001 |
keeping | -0.001 |
crashing | -0.001 |
Token | seams |
Token ID | 41280 |
Feature activation | -1.88e+00 |
Pos logit contributions | |
seen | +0.030 |
hole | +0.028 |
keeping | +0.028 |
stations | +0.028 |
contain | +0.027 |
Neg logit contributions | |
iflower | -0.027 |
pron | -0.025 |
live | -0.021 |
ama | -0.020 |
innie | -0.020 |
Token | of |
Token ID | 286 |
Feature activation | -8.19e-01 |
Pos logit contributions | |
keeping | +0.053 |
seen | +0.052 |
stations | +0.051 |
ales | +0.049 |
hole | +0.049 |
Neg logit contributions | |
iflower | -0.053 |
pron | -0.052 |
live | -0.043 |
ama | -0.042 |
wl | -0.039 |
Token | her |
Token ID | 607 |
Feature activation | -7.11e-01 |
Pos logit contributions | |
seen | +0.025 |
hole | +0.024 |
keeping | +0.024 |
contain | +0.023 |
stations | +0.023 |
Neg logit contributions | |
iflower | -0.022 |
pron | -0.020 |
live | -0.017 |
innie | -0.016 |
ama | -0.016 |
Token | sweater |
Token ID | 35685 |
Feature activation | -3.30e-01 |
Pos logit contributions | |
seen | +0.020 |
hole | +0.019 |
keeping | +0.019 |
stations | +0.018 |
officers | +0.018 |
Neg logit contributions | |
iflower | -0.022 |
pron | -0.020 |
live | -0.017 |
innie | -0.017 |
wl | -0.017 |
Token | * |
Token ID | 1635 |
Feature activation | -6.35e+00 |
Pos logit contributions | |
seen | +0.010 |
stations | +0.009 |
keeping | +0.009 |
hole | +0.009 |
ales | +0.009 |
Neg logit contributions | |
iflower | -0.009 |
pron | -0.009 |
live | -0.007 |
ama | -0.007 |
wl | -0.007 |
Token | po |
Token ID | 7501 |
Feature activation | -3.40e+00 |
Pos logit contributions | |
seen | +0.167 |
hole | +0.158 |
stations | +0.156 |
keeping | +0.156 |
contain | +0.149 |
Neg logit contributions | |
iflower | -0.201 |
pron | -0.186 |
live | -0.164 |
innie | -0.157 |
ama | -0.155 |
Token | pped |
Token ID | 1496 |
Feature activation | -2.87e-01 |
Pos logit contributions | |
seen | +0.092 |
contain | +0.092 |
stations | +0.090 |
keeping | +0.089 |
footsteps | +0.089 |
Neg logit contributions | |
iflower | -0.103 |
pron | -0.097 |
live | -0.084 |
innie | -0.083 |
wl | -0.079 |
Token | * |
Token ID | 9 |
Feature activation | -2.04e+00 |
Pos logit contributions | |
seen | +0.008 |
stations | +0.007 |
contain | +0.007 |
keeping | +0.007 |
hole | +0.007 |
Neg logit contributions | |
iflower | -0.009 |
pron | -0.009 |
live | -0.007 |
ama | -0.007 |
wl | -0.007 |
Token | ! |
Token ID | 0 |
Feature activation | -7.81e-01 |
Pos logit contributions | |
seen | +0.049 |
hole | +0.045 |
keeping | +0.045 |
stations | +0.045 |
contain | +0.043 |
Neg logit contributions | |
iflower | -0.071 |
pron | -0.065 |
live | -0.059 |
innie | -0.056 |
ama | -0.056 |
Token | Everyone |
Token ID | 11075 |
Feature activation | -2.58e+00 |
Pos logit contributions | |
seen | +0.020 |
hole | +0.019 |
keeping | +0.018 |
crashing | +0.018 |
towards | +0.018 |
Neg logit contributions | |
iflower | -0.026 |
pron | -0.024 |
live | -0.022 |
wl | -0.021 |
tto | -0.020 |
Token | to |
Token ID | 284 |
Feature activation | +1.86e-01 |
Pos logit contributions | |
seen | +0.022 |
keeping | +0.021 |
stations | +0.021 |
crashing | +0.021 |
hole | +0.020 |
Neg logit contributions | |
iflower | -0.024 |
pron | -0.023 |
live | -0.019 |
ama | -0.018 |
innie | -0.018 |
Token | his |
Token ID | 465 |
Feature activation | -7.21e-01 |
Pos logit contributions | |
iflower | +0.006 |
pron | +0.005 |
live | +0.005 |
innie | +0.004 |
ama | +0.004 |
Neg logit contributions | |
seen | -0.005 |
hole | -0.005 |
keeping | -0.005 |
stations | -0.005 |
crashing | -0.005 |
Token | friend |
Token ID | 1545 |
Feature activation | -2.30e+00 |
Pos logit contributions | |
seen | +0.019 |
footsteps | +0.019 |
officers | +0.018 |
ways | +0.018 |
stations | +0.018 |
Neg logit contributions | |
iflower | -0.021 |
pron | -0.020 |
live | -0.018 |
wl | -0.017 |
tto | -0.017 |
Token | . |
Token ID | 13 |
Feature activation | -8.14e-01 |
Pos logit contributions | |
seen | +0.063 |
hole | +0.059 |
keeping | +0.059 |
stations | +0.058 |
officers | +0.056 |
Neg logit contributions | |
iflower | -0.072 |
pron | -0.066 |
live | -0.058 |
innie | -0.056 |
ama | -0.055 |
Token | He |
Token ID | 679 |
Feature activation | +8.67e-01 |
Pos logit contributions | |
seen | +0.021 |
hole | +0.021 |
towards | +0.021 |
keeping | +0.020 |
officers | +0.020 |
Neg logit contributions | |
iflower | -0.027 |
pron | -0.025 |
live | -0.022 |
wl | -0.020 |
innie | -0.020 |
Token | scam |
Token ID | 19126 |
Feature activation | -6.26e+00 |
Pos logit contributions | |
iflower | +0.031 |
pron | +0.028 |
live | +0.026 |
ama | +0.025 |
innie | +0.024 |
Neg logit contributions | |
seen | -0.021 |
hole | -0.019 |
officers | -0.018 |
stations | -0.018 |
keeping | -0.018 |
Token | pered |
Token ID | 13653 |
Feature activation | -3.01e+00 |
Pos logit contributions | |
seen | +0.069 |
hole | +0.066 |
stations | +0.062 |
keeping | +0.059 |
officers | +0.055 |
Neg logit contributions | |
iflower | -0.287 |
pron | -0.272 |
live | -0.259 |
innie | -0.253 |
ama | -0.250 |
Token | around |
Token ID | 1088 |
Feature activation | -4.20e-02 |
Pos logit contributions | |
seen | +0.083 |
hole | +0.080 |
keeping | +0.079 |
stations | +0.078 |
contain | +0.075 |
Neg logit contributions | |
iflower | -0.088 |
pron | -0.081 |
live | -0.071 |
innie | -0.069 |
ama | -0.067 |
Token | trying |
Token ID | 2111 |
Feature activation | +3.77e-01 |
Pos logit contributions | |
seen | +0.001 |
hole | +0.001 |
keeping | +0.001 |
stations | +0.001 |
ales | +0.001 |
Neg logit contributions | |
iflower | -0.001 |
pron | -0.001 |
lf | -0.001 |
innie | -0.001 |
live | -0.001 |
Token | to |
Token ID | 284 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
iflower | +0.012 |
pron | +0.011 |
live | +0.010 |
innie | +0.009 |
ama | +0.009 |
Neg logit contributions | |
seen | -0.010 |
keeping | -0.009 |
hole | -0.009 |
stations | -0.009 |
officers | -0.009 |
Token | find |
Token ID | 1064 |
Feature activation | +1.22e+00 |
Pos logit contributions | |
seen | +0.027 |
hole | +0.027 |
keeping | +0.026 |
stations | +0.026 |
ously | +0.026 |
Neg logit contributions | |
iflower | -0.035 |
pron | -0.033 |
live | -0.028 |
innie | -0.027 |
lf | -0.027 |
Token | and |
Token ID | 290 |
Feature activation | +5.61e-02 |
Pos logit contributions | |
iflower | +0.016 |
pron | +0.016 |
live | +0.013 |
ama | +0.013 |
innie | +0.013 |
Neg logit contributions | |
stations | -0.015 |
brake | -0.015 |
keeping | -0.015 |
seen | -0.015 |
crashing | -0.015 |
Token | saw |
Token ID | 2497 |
Feature activation | +1.31e+00 |
Pos logit contributions | |
iflower | +0.002 |
pron | +0.002 |
live | +0.002 |
innie | +0.001 |
ama | +0.001 |
Neg logit contributions | |
seen | -0.001 |
hole | -0.001 |
keeping | -0.001 |
stations | -0.001 |
crashing | -0.001 |
Token | there |
Token ID | 612 |
Feature activation | -5.71e-01 |
Pos logit contributions | |
iflower | +0.036 |
pron | +0.033 |
live | +0.028 |
innie | +0.027 |
ama | +0.027 |
Neg logit contributions | |
seen | -0.040 |
hole | -0.039 |
keeping | -0.038 |
stations | -0.037 |
contain | -0.037 |
Token | were |
Token ID | 547 |
Feature activation | -5.21e-01 |
Pos logit contributions | |
seen | +0.018 |
keeping | +0.018 |
hole | +0.017 |
stations | +0.017 |
brake | +0.017 |
Neg logit contributions | |
iflower | -0.014 |
pron | -0.013 |
live | -0.012 |
ama | -0.011 |
innie | -0.010 |
Token | no |
Token ID | 645 |
Feature activation | -2.40e-01 |
Pos logit contributions | |
seen | +0.015 |
hole | +0.015 |
keeping | +0.015 |
ales | +0.014 |
contain | +0.014 |
Neg logit contributions | |
iflower | -0.015 |
pron | -0.013 |
live | -0.012 |
innie | -0.011 |
ama | -0.011 |
Token | monsters |
Token ID | 10963 |
Feature activation | -5.76e+00 |
Pos logit contributions | |
keeping | +0.007 |
ked | +0.007 |
ried | +0.007 |
hole | +0.007 |
ales | +0.007 |
Neg logit contributions | |
iflower | -0.006 |
pron | -0.006 |
best | -0.005 |
important | -0.005 |
story | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | -4.59e-01 |
Pos logit contributions | |
ales | +0.163 |
awks | +0.157 |
brakes | +0.155 |
keeping | +0.154 |
anticipation | +0.152 |
Neg logit contributions | |
iflower | -0.138 |
pron | -0.120 |
lf | -0.110 |
live | -0.105 |
on | -0.103 |
Token | She |
Token ID | 1375 |
Feature activation | -2.62e-01 |
Pos logit contributions | |
seen | +0.012 |
keeping | +0.012 |
hole | +0.012 |
stations | +0.011 |
officers | +0.011 |
Neg logit contributions | |
iflower | -0.015 |
pron | -0.014 |
live | -0.012 |
wl | -0.011 |
innie | -0.011 |
Token | felt |
Token ID | 2936 |
Feature activation | +9.49e-02 |
Pos logit contributions | |
seen | +0.007 |
hole | +0.007 |
keeping | +0.007 |
stations | +0.007 |
officers | +0.006 |
Neg logit contributions | |
iflower | -0.008 |
pron | -0.008 |
live | -0.007 |
innie | -0.006 |
ama | -0.006 |
Token | proud |
Token ID | 6613 |
Feature activation | -1.82e+00 |
Pos logit contributions | |
iflower | +0.003 |
pron | +0.002 |
live | +0.002 |
tto | +0.002 |
ama | +0.002 |
Neg logit contributions | |
hole | -0.003 |
keeping | -0.003 |
ales | -0.003 |
seen | -0.003 |
stations | -0.003 |
Token | of |
Token ID | 286 |
Feature activation | +1.50e+00 |
Pos logit contributions | |
keeping | +0.063 |
brake | +0.061 |
ales | +0.059 |
stations | +0.059 |
ried | +0.059 |
Neg logit contributions | |
iflower | -0.041 |
pron | -0.037 |
ama | -0.030 |
story | -0.028 |
mummy | -0.028 |