Token | happy |
Token ID | 3772 |
Feature activation | +5.73e-02 |
Pos logit contributions | |
hallway | +0.025 |
stair | +0.025 |
she | +0.024 |
weekend | +0.024 |
driveway | +0.024 |
Neg logit contributions | |
ps | -0.018 |
ss | -0.015 |
teeth | -0.014 |
mus | -0.014 |
ches | -0.013 |
Token | ! |
Token ID | 0 |
Feature activation | +2.68e-02 |
Pos logit contributions | |
hallway | +0.114 |
driveway | +0.111 |
weekend | +0.108 |
particular | +0.108 |
stair | +0.107 |
Neg logit contributions | |
ps | -0.064 |
teeth | -0.052 |
ss | -0.048 |
louder | -0.047 |
." | -0.045 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +1.42e-03 |
Pos logit contributions | |
weekend | +0.042 |
hallway | +0.042 |
stair | +0.041 |
holiday | +0.040 |
sandbox | +0.040 |
Neg logit contributions | |
ps | -0.040 |
ss | -0.036 |
teeth | -0.036 |
ches | -0.035 |
mus | -0.033 |
Token | On |
Token ID | 2202 |
Feature activation | +8.67e-02 |
Pos logit contributions | |
she | +0.002 |
weekend | +0.002 |
day | +0.002 |
hallway | +0.002 |
stair | +0.002 |
Neg logit contributions | |
ps | -0.002 |
ss | -0.002 |
ches | -0.002 |
teeth | -0.002 |
mus | -0.002 |
Token | a |
Token ID | 257 |
Feature activation | +9.43e-02 |
Pos logit contributions | |
days | +0.199 |
policy | +0.193 |
spe | +0.186 |
stair | +0.181 |
hallway | +0.181 |
Neg logit contributions | |
ps | -0.062 |
ches | -0.056 |
better | -0.038 |
teeth | -0.034 |
harder | -0.029 |
Token | sunny |
Token ID | 27737 |
Feature activation | +1.56e-01 |
Pos logit contributions | |
stack | +0.124 |
policy | +0.117 |
iked | +0.114 |
keeping | +0.113 |
loo | +0.109 |
Neg logit contributions | |
ches | -0.173 |
funny | -0.153 |
loud | -0.148 |
hard | -0.145 |
silly | -0.144 |
Token | day |
Token ID | 1110 |
Feature activation | +8.19e-02 |
Pos logit contributions | |
policy | +0.242 |
oker | +0.225 |
spe | +0.223 |
otion | +0.217 |
bek | +0.216 |
Neg logit contributions | |
me | -0.200 |
ches | -0.169 |
w | -0.158 |
nice | -0.156 |
fur | -0.156 |
Token | , |
Token ID | 11 |
Feature activation | +5.69e-02 |
Pos logit contributions | |
stack | +0.185 |
driveway | +0.182 |
spe | +0.171 |
sandbox | +0.169 |
days | +0.166 |
Neg logit contributions | |
funny | -0.046 |
teeth | -0.045 |
silly | -0.044 |
ches | -0.044 |
loud | -0.043 |
Token | a |
Token ID | 257 |
Feature activation | -4.18e-03 |
Pos logit contributions | |
driveway | +0.120 |
hallway | +0.113 |
loo | +0.110 |
stair | +0.108 |
occasion | +0.108 |
Neg logit contributions | |
ps | -0.053 |
teeth | -0.047 |
mus | -0.043 |
funny | -0.041 |
louder | -0.040 |
Token | little |
Token ID | 1310 |
Feature activation | +3.98e-02 |
Pos logit contributions | |
ps | +0.007 |
ss | +0.005 |
ches | +0.005 |
teeth | +0.005 |
mus | +0.005 |
Neg logit contributions | |
she | -0.007 |
weekend | -0.007 |
day | -0.007 |
hallway | -0.007 |
stair | -0.007 |
Token | girl |
Token ID | 2576 |
Feature activation | +8.02e-02 |
Pos logit contributions | |
hallway | +0.057 |
days | +0.057 |
weekend | +0.056 |
stair | +0.054 |
driveway | +0.054 |
Neg logit contributions | |
ps | -0.056 |
ss | -0.055 |
mus | -0.051 |
ches | -0.051 |
teeth | -0.050 |
Token | forever |
Token ID | 8097 |
Feature activation | +1.34e-02 |
Pos logit contributions | |
hallway | +0.042 |
weekend | +0.042 |
stair | +0.041 |
particular | +0.040 |
driveway | +0.040 |
Neg logit contributions | |
ps | -0.036 |
ss | -0.030 |
ches | -0.029 |
teeth | -0.028 |
mus | -0.028 |
Token | . |
Token ID | 13 |
Feature activation | +2.59e-02 |
Pos logit contributions | |
weekend | +0.024 |
stair | +0.024 |
hallway | +0.023 |
she | +0.023 |
particular | +0.023 |
Neg logit contributions | |
ps | -0.020 |
ss | -0.016 |
teeth | -0.016 |
ches | -0.015 |
mus | -0.015 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -7.83e-03 |
Pos logit contributions | |
hallway | +0.041 |
weekend | +0.040 |
stair | +0.040 |
holiday | +0.038 |
sandbox | +0.038 |
Neg logit contributions | |
ps | -0.040 |
ss | -0.036 |
ches | -0.034 |
teeth | -0.034 |
iddles | -0.033 |
Token | On |
Token ID | 2202 |
Feature activation | +8.58e-02 |
Pos logit contributions | |
ps | +0.015 |
ss | +0.013 |
teeth | +0.012 |
ches | +0.012 |
mus | +0.012 |
Neg logit contributions | |
she | -0.011 |
weekend | -0.010 |
day | -0.010 |
hallway | -0.010 |
stair | -0.010 |
Token | a |
Token ID | 257 |
Feature activation | +8.96e-02 |
Pos logit contributions | |
days | +0.189 |
policy | +0.183 |
hallway | +0.173 |
stair | +0.173 |
loo | +0.172 |
Neg logit contributions | |
ps | -0.068 |
ches | -0.062 |
better | -0.048 |
teeth | -0.044 |
harder | -0.043 |
Token | sunny |
Token ID | 27737 |
Feature activation | +1.51e-01 |
Pos logit contributions | |
stack | +0.114 |
policy | +0.106 |
loo | +0.101 |
iked | +0.100 |
keeping | +0.099 |
Neg logit contributions | |
ches | -0.167 |
funny | -0.147 |
loud | -0.142 |
hard | -0.142 |
silly | -0.141 |
Token | day |
Token ID | 1110 |
Feature activation | +8.64e-02 |
Pos logit contributions | |
policy | +0.242 |
oker | +0.223 |
Sorry | +0.216 |
haven | +0.215 |
spe | +0.215 |
Neg logit contributions | |
me | -0.180 |
ches | -0.155 |
nice | -0.149 |
hard | -0.149 |
big | -0.147 |
Token | , |
Token ID | 11 |
Feature activation | +5.23e-02 |
Pos logit contributions | |
stack | +0.197 |
driveway | +0.191 |
spe | +0.175 |
sandbox | +0.174 |
field | +0.172 |
Neg logit contributions | |
funny | -0.051 |
silly | -0.049 |
loud | -0.046 |
teeth | -0.045 |
louder | -0.044 |
Token | a |
Token ID | 257 |
Feature activation | -5.74e-03 |
Pos logit contributions | |
driveway | +0.105 |
hallway | +0.100 |
loo | +0.097 |
stair | +0.096 |
days | +0.096 |
Neg logit contributions | |
ps | -0.052 |
teeth | -0.045 |
mus | -0.043 |
louder | -0.041 |
ss | -0.040 |
Token | mom |
Token ID | 1995 |
Feature activation | +4.81e-02 |
Pos logit contributions | |
ps | +0.009 |
ss | +0.008 |
ches | +0.007 |
teeth | +0.007 |
mus | +0.007 |
Neg logit contributions | |
she | -0.010 |
weekend | -0.010 |
day | -0.010 |
stair | -0.009 |
hallway | -0.009 |
Token | my |
Token ID | 1820 |
Feature activation | +2.98e-02 |
Pos logit contributions | |
stair | +0.093 |
stack | +0.090 |
exploration | +0.089 |
hallway | +0.086 |
cliff | +0.086 |
Neg logit contributions | |
ss | -0.059 |
ps | -0.057 |
ches | -0.045 |
ache | -0.044 |
iddles | -0.043 |
Token | happy |
Token ID | 3772 |
Feature activation | +2.58e-02 |
Pos logit contributions | |
she | +0.001 |
weekend | +0.001 |
stair | +0.001 |
hallway | +0.001 |
day | +0.001 |
Neg logit contributions | |
ps | -0.001 |
ss | -0.001 |
ches | -0.001 |
teeth | -0.001 |
mus | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | +1.92e-02 |
Pos logit contributions | |
hallway | +0.051 |
weekend | +0.051 |
stair | +0.051 |
particular | +0.051 |
driveway | +0.050 |
Neg logit contributions | |
ps | -0.032 |
teeth | -0.026 |
ss | -0.023 |
iddles | -0.021 |
mus | -0.021 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -1.17e-02 |
Pos logit contributions | |
stair | +0.030 |
hallway | +0.030 |
weekend | +0.030 |
holiday | +0.029 |
sandbox | +0.029 |
Neg logit contributions | |
ps | -0.029 |
ss | -0.026 |
teeth | -0.025 |
ches | -0.025 |
mus | -0.024 |
Token | On |
Token ID | 2202 |
Feature activation | +7.54e-02 |
Pos logit contributions | |
ps | +0.024 |
ss | +0.021 |
teeth | +0.020 |
ches | +0.020 |
mus | +0.020 |
Neg logit contributions | |
she | -0.015 |
day | -0.014 |
weekend | -0.014 |
hallway | -0.014 |
stair | -0.014 |
Token | a |
Token ID | 257 |
Feature activation | +8.08e-02 |
Pos logit contributions | |
days | +0.164 |
policy | +0.153 |
hallway | +0.148 |
stair | +0.147 |
sandbox | +0.147 |
Neg logit contributions | |
ps | -0.071 |
ches | -0.066 |
teeth | -0.048 |
better | -0.043 |
harder | -0.040 |
Token | sunny |
Token ID | 27737 |
Feature activation | +1.46e-01 |
Pos logit contributions | |
stack | +0.109 |
policy | +0.104 |
iked | +0.098 |
days | +0.098 |
loo | +0.097 |
Neg logit contributions | |
ches | -0.149 |
funny | -0.120 |
loud | -0.117 |
sharp | -0.116 |
hard | -0.115 |
Token | day |
Token ID | 1110 |
Feature activation | +7.60e-02 |
Pos logit contributions | |
policy | +0.221 |
spe | +0.211 |
haven | +0.197 |
bek | +0.195 |
Sorry | +0.195 |
Neg logit contributions | |
me | -0.183 |
ches | -0.176 |
p | -0.165 |
fur | -0.153 |
- | -0.149 |
Token | , |
Token ID | 11 |
Feature activation | +4.72e-02 |
Pos logit contributions | |
driveway | +0.176 |
stack | +0.174 |
spe | +0.162 |
sandbox | +0.162 |
days | +0.159 |
Neg logit contributions | |
ps | -0.043 |
teeth | -0.041 |
ches | -0.040 |
funny | -0.038 |
silly | -0.036 |
Token | a |
Token ID | 257 |
Feature activation | -5.85e-03 |
Pos logit contributions | |
driveway | +0.096 |
hallway | +0.092 |
stair | +0.089 |
loo | +0.087 |
sandbox | +0.086 |
Neg logit contributions | |
ps | -0.047 |
teeth | -0.042 |
mus | -0.039 |
ss | -0.037 |
funny | -0.035 |
Token | little |
Token ID | 1310 |
Feature activation | +3.42e-02 |
Pos logit contributions | |
ps | +0.009 |
ss | +0.008 |
ches | +0.007 |
teeth | +0.007 |
mus | +0.007 |
Neg logit contributions | |
she | -0.010 |
weekend | -0.010 |
day | -0.010 |
stair | -0.010 |
hallway | -0.010 |
Token | girl |
Token ID | 2576 |
Feature activation | +7.88e-02 |
Pos logit contributions | |
hallway | +0.047 |
days | +0.047 |
weekend | +0.046 |
she | +0.045 |
stair | +0.045 |
Neg logit contributions | |
ps | -0.051 |
ss | -0.050 |
ches | -0.047 |
mus | -0.046 |
teeth | -0.045 |
Token | after |
Token ID | 706 |
Feature activation | +3.30e-03 |
Pos logit contributions | |
ps | +0.005 |
ss | +0.004 |
ches | +0.004 |
teeth | +0.004 |
mus | +0.004 |
Neg logit contributions | |
she | -0.003 |
weekend | -0.002 |
day | -0.002 |
hallway | -0.002 |
stair | -0.002 |
Token | ! |
Token ID | 0 |
Feature activation | +2.58e-02 |
Pos logit contributions | |
she | +0.006 |
weekend | +0.006 |
hallway | +0.006 |
stair | +0.006 |
day | +0.006 |
Neg logit contributions | |
ps | -0.005 |
ss | -0.004 |
ches | -0.004 |
teeth | -0.004 |
mus | -0.004 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -9.92e-03 |
Pos logit contributions | |
hallway | +0.041 |
stair | +0.041 |
weekend | +0.040 |
sandbox | +0.039 |
she | +0.039 |
Neg logit contributions | |
ps | -0.037 |
ss | -0.035 |
ches | -0.033 |
teeth | -0.032 |
mus | -0.031 |
Token | On |
Token ID | 2202 |
Feature activation | +7.88e-02 |
Pos logit contributions | |
ps | +0.020 |
ss | +0.017 |
teeth | +0.017 |
ches | +0.017 |
mus | +0.016 |
Neg logit contributions | |
she | -0.013 |
day | -0.012 |
weekend | -0.012 |
hallway | -0.012 |
stair | -0.012 |
Token | a |
Token ID | 257 |
Feature activation | +7.19e-02 |
Pos logit contributions | |
days | +0.157 |
policy | +0.147 |
hallway | +0.144 |
sandbox | +0.142 |
driveway | +0.142 |
Neg logit contributions | |
ps | -0.084 |
ches | -0.082 |
teeth | -0.061 |
better | -0.059 |
harder | -0.054 |
Token | sunny |
Token ID | 27737 |
Feature activation | +1.39e-01 |
Pos logit contributions | |
stack | +0.096 |
policy | +0.092 |
loo | +0.090 |
iked | +0.087 |
days | +0.084 |
Neg logit contributions | |
ches | -0.129 |
funny | -0.105 |
loud | -0.101 |
silly | -0.101 |
sharp | -0.100 |
Token | day |
Token ID | 1110 |
Feature activation | +7.95e-02 |
Pos logit contributions | |
policy | +0.195 |
awaited | +0.170 |
oker | +0.170 |
spe | +0.168 |
line | +0.168 |
Neg logit contributions | |
me | -0.186 |
ches | -0.175 |
fur | -0.154 |
p | -0.152 |
nice | -0.151 |
Token | , |
Token ID | 11 |
Feature activation | +5.07e-02 |
Pos logit contributions | |
stack | +0.182 |
driveway | +0.181 |
hallway | +0.167 |
sandbox | +0.165 |
field | +0.165 |
Neg logit contributions | |
ps | -0.046 |
teeth | -0.039 |
silly | -0.037 |
good | -0.037 |
ches | -0.036 |
Token | Billy |
Token ID | 15890 |
Feature activation | +3.25e-02 |
Pos logit contributions | |
driveway | +0.099 |
hallway | +0.094 |
loo | +0.092 |
stair | +0.091 |
sandbox | +0.089 |
Neg logit contributions | |
ps | -0.055 |
teeth | -0.046 |
mus | -0.046 |
ss | -0.043 |
ches | -0.041 |
Token | was |
Token ID | 373 |
Feature activation | +1.93e-02 |
Pos logit contributions | |
stair | +0.051 |
weekend | +0.048 |
days | +0.047 |
particular | +0.046 |
hallway | +0.046 |
Neg logit contributions | |
ps | -0.053 |
ches | -0.048 |
ss | -0.046 |
teeth | -0.045 |
mus | -0.042 |
Token | out |
Token ID | 503 |
Feature activation | +3.33e-02 |
Pos logit contributions | |
stair | +0.031 |
weekend | +0.030 |
she | +0.030 |
days | +0.030 |
particular | +0.030 |
Neg logit contributions | |
ps | -0.031 |
ss | -0.025 |
teeth | -0.025 |
mus | -0.024 |
ches | -0.024 |
Token | together |
Token ID | 1978 |
Feature activation | -7.24e-03 |
Pos logit contributions | |
ps | +0.000 |
ss | +0.000 |
ches | +0.000 |
teeth | +0.000 |
mus | +0.000 |
Neg logit contributions | |
she | -0.000 |
weekend | -0.000 |
day | -0.000 |
hallway | -0.000 |
stair | -0.000 |
Token | . |
Token ID | 13 |
Feature activation | +2.46e-02 |
Pos logit contributions | |
ps | +0.011 |
ss | +0.010 |
ches | +0.009 |
teeth | +0.009 |
mus | +0.009 |
Neg logit contributions | |
she | -0.012 |
day | -0.012 |
weekend | -0.012 |
stair | -0.012 |
hallway | -0.012 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -7.61e-03 |
Pos logit contributions | |
she | +0.039 |
stair | +0.039 |
hallway | +0.039 |
weekend | +0.038 |
wander | +0.037 |
Neg logit contributions | |
ps | -0.036 |
ss | -0.033 |
teeth | -0.031 |
ches | -0.030 |
mus | -0.028 |
Token | On |
Token ID | 2202 |
Feature activation | +7.92e-02 |
Pos logit contributions | |
ps | +0.015 |
ss | +0.014 |
ches | +0.013 |
teeth | +0.013 |
mus | +0.013 |
Neg logit contributions | |
she | -0.010 |
weekend | -0.009 |
day | -0.009 |
hallway | -0.009 |
stair | -0.009 |
Token | a |
Token ID | 257 |
Feature activation | +7.49e-02 |
Pos logit contributions | |
days | +0.164 |
policy | +0.158 |
hallway | +0.153 |
stair | +0.152 |
driveway | +0.150 |
Neg logit contributions | |
ps | -0.072 |
ches | -0.066 |
teeth | -0.051 |
better | -0.051 |
harder | -0.046 |
Token | sunny |
Token ID | 27737 |
Feature activation | +1.38e-01 |
Pos logit contributions | |
stack | +0.098 |
policy | +0.094 |
loo | +0.091 |
days | +0.091 |
iked | +0.088 |
Neg logit contributions | |
ches | -0.134 |
funny | -0.111 |
loud | -0.108 |
sharp | -0.106 |
silly | -0.106 |
Token | day |
Token ID | 1110 |
Feature activation | +7.09e-02 |
Pos logit contributions | |
policy | +0.203 |
spe | +0.183 |
awaited | +0.178 |
oker | +0.176 |
haven | +0.176 |
Neg logit contributions | |
me | -0.181 |
ches | -0.167 |
p | -0.148 |
hard | -0.147 |
nice | -0.146 |
Token | , |
Token ID | 11 |
Feature activation | +4.98e-02 |
Pos logit contributions | |
driveway | +0.166 |
stack | +0.161 |
sandbox | +0.153 |
hallway | +0.153 |
spe | +0.148 |
Neg logit contributions | |
teeth | -0.040 |
ps | -0.038 |
ches | -0.034 |
funny | -0.033 |
silly | -0.031 |
Token | Joe |
Token ID | 5689 |
Feature activation | +5.19e-02 |
Pos logit contributions | |
driveway | +0.087 |
hallway | +0.086 |
stair | +0.081 |
days | +0.079 |
occasion | +0.079 |
Neg logit contributions | |
ps | -0.064 |
teeth | -0.058 |
mus | -0.053 |
ss | -0.053 |
louder | -0.051 |
Token | decided |
Token ID | 3066 |
Feature activation | +4.14e-02 |
Pos logit contributions | |
stair | +0.090 |
hallway | +0.086 |
weekend | +0.085 |
stack | +0.085 |
particular | +0.084 |
Neg logit contributions | |
ps | -0.071 |
ches | -0.068 |
teeth | -0.061 |
ss | -0.059 |
mus | -0.055 |
Token | he |
Token ID | 339 |
Feature activation | +6.92e-03 |
Pos logit contributions | |
days | +0.067 |
hallway | +0.066 |
driveway | +0.062 |
stair | +0.062 |
particular | +0.062 |
Neg logit contributions | |
ps | -0.061 |
teeth | -0.055 |
ss | -0.055 |
louder | -0.053 |
ches | -0.051 |
Token | best |
Token ID | 1266 |
Feature activation | +1.58e-02 |
Pos logit contributions | |
hallway | +0.021 |
stair | +0.021 |
she | +0.020 |
weekend | +0.020 |
day | +0.020 |
Neg logit contributions | |
ps | -0.025 |
ss | -0.022 |
ches | -0.021 |
mus | -0.020 |
teeth | -0.020 |
Token | friends |
Token ID | 2460 |
Feature activation | +2.61e-02 |
Pos logit contributions | |
hallway | +0.031 |
weekend | +0.031 |
stair | +0.030 |
she | +0.029 |
day | +0.029 |
Neg logit contributions | |
ps | -0.021 |
ss | -0.017 |
teeth | -0.017 |
ches | -0.015 |
iddles | -0.015 |
Token | . |
Token ID | 13 |
Feature activation | +2.05e-02 |
Pos logit contributions | |
hallway | +0.055 |
weekend | +0.054 |
stair | +0.054 |
driveway | +0.052 |
particular | +0.051 |
Neg logit contributions | |
ps | -0.029 |
ss | -0.022 |
ches | -0.021 |
teeth | -0.021 |
mus | -0.020 |
Token | On |
Token ID | 1550 |
Feature activation | +5.98e-02 |
Pos logit contributions | |
she | +0.026 |
hallway | +0.026 |
stair | +0.025 |
weekend | +0.025 |
holiday | +0.025 |
Neg logit contributions | |
ps | -0.036 |
ss | -0.034 |
iddles | -0.032 |
teeth | -0.032 |
ches | -0.032 |
Token | a |
Token ID | 257 |
Feature activation | +5.95e-02 |
Pos logit contributions | |
des | +0.100 |
hallway | +0.098 |
days | +0.098 |
driveway | +0.095 |
sandbox | +0.095 |
Neg logit contributions | |
ps | -0.081 |
ches | -0.063 |
teeth | -0.061 |
oni | -0.056 |
ss | -0.056 |
Token | sunny |
Token ID | 27737 |
Feature activation | +1.36e-01 |
Pos logit contributions | |
policy | +0.082 |
stack | +0.081 |
loo | +0.078 |
days | +0.078 |
iked | +0.074 |
Neg logit contributions | |
ches | -0.094 |
funny | -0.077 |
loud | -0.073 |
ss | -0.072 |
silly | -0.072 |
Token | afternoon |
Token ID | 6672 |
Feature activation | +9.73e-02 |
Pos logit contributions | |
haven | +0.165 |
policy | +0.161 |
she | +0.155 |
awaited | +0.154 |
oker | +0.152 |
Neg logit contributions | |
me | -0.191 |
ches | -0.190 |
ep | -0.175 |
p | -0.168 |
fur | -0.167 |
Token | , |
Token ID | 11 |
Feature activation | +5.63e-02 |
Pos logit contributions | |
stack | +0.223 |
driveway | +0.223 |
hill | +0.210 |
field | +0.209 |
boat | +0.205 |
Neg logit contributions | |
ps | -0.040 |
teeth | -0.037 |
louder | -0.036 |
ers | -0.035 |
funny | -0.032 |
Token | they |
Token ID | 484 |
Feature activation | +1.85e-02 |
Pos logit contributions | |
hallway | +0.062 |
driveway | +0.060 |
loo | +0.058 |
occasion | +0.058 |
stack | +0.056 |
Neg logit contributions | |
louder | -0.103 |
ps | -0.099 |
ss | -0.097 |
funny | -0.092 |
loud | -0.092 |
Token | decided |
Token ID | 3066 |
Feature activation | +3.92e-02 |
Pos logit contributions | |
stair | +0.023 |
hallway | +0.022 |
she | +0.022 |
weekend | +0.022 |
day | +0.021 |
Neg logit contributions | |
ps | -0.036 |
ss | -0.032 |
ches | -0.031 |
teeth | -0.030 |
mus | -0.030 |
Token | to |
Token ID | 284 |
Feature activation | +4.87e-02 |
Pos logit contributions | |
hallway | +0.071 |
days | +0.067 |
loo | +0.067 |
stair | +0.066 |
driveway | +0.066 |
Neg logit contributions | |
ps | -0.052 |
ss | -0.046 |
teeth | -0.045 |
louder | -0.044 |
ches | -0.042 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -9.75e-03 |
Pos logit contributions | |
weekend | +0.029 |
hallway | +0.029 |
stair | +0.029 |
holiday | +0.029 |
she | +0.029 |
Neg logit contributions | |
ps | -0.030 |
ss | -0.028 |
teeth | -0.026 |
ches | -0.026 |
mus | -0.025 |
Token | On |
Token ID | 2202 |
Feature activation | +8.09e-02 |
Pos logit contributions | |
ps | +0.019 |
ss | +0.016 |
teeth | +0.016 |
ches | +0.016 |
mus | +0.015 |
Neg logit contributions | |
she | -0.013 |
day | -0.013 |
weekend | -0.013 |
hallway | -0.012 |
stair | -0.012 |
Token | a |
Token ID | 257 |
Feature activation | +8.87e-02 |
Pos logit contributions | |
days | +0.183 |
policy | +0.174 |
spe | +0.170 |
loo | +0.167 |
hallway | +0.166 |
Neg logit contributions | |
ps | -0.058 |
ches | -0.057 |
better | -0.042 |
teeth | -0.038 |
harder | -0.036 |
Token | bright |
Token ID | 6016 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
stack | +0.111 |
policy | +0.107 |
iked | +0.106 |
loo | +0.103 |
keeping | +0.102 |
Neg logit contributions | |
ches | -0.166 |
funny | -0.144 |
loud | -0.139 |
sharp | -0.139 |
silly | -0.138 |
Token | and |
Token ID | 290 |
Feature activation | +3.78e-02 |
Pos logit contributions | |
policy | +0.136 |
stack | +0.135 |
days | +0.133 |
keeping | +0.130 |
Have | +0.121 |
Neg logit contributions | |
ches | -0.157 |
funny | -0.144 |
me | -0.144 |
hard | -0.141 |
nice | -0.139 |
Token | sunny |
Token ID | 27737 |
Feature activation | +1.34e-01 |
Pos logit contributions | |
hallway | +0.027 |
driveway | +0.026 |
stair | +0.026 |
days | +0.025 |
loo | +0.024 |
Neg logit contributions | |
ps | -0.081 |
mus | -0.077 |
ss | -0.076 |
teeth | -0.075 |
louder | -0.075 |
Token | day |
Token ID | 1110 |
Feature activation | +5.77e-02 |
Pos logit contributions | |
policy | +0.195 |
spe | +0.184 |
bek | +0.183 |
haven | +0.181 |
path | +0.180 |
Neg logit contributions | |
me | -0.153 |
p | -0.137 |
- | -0.136 |
ches | -0.134 |
fur | -0.131 |
Token | , |
Token ID | 11 |
Feature activation | +5.46e-02 |
Pos logit contributions | |
driveway | +0.098 |
stack | +0.097 |
days | +0.095 |
sandbox | +0.093 |
hallway | +0.092 |
Neg logit contributions | |
ps | -0.073 |
teeth | -0.070 |
ches | -0.068 |
mus | -0.063 |
ss | -0.062 |
Token | Jimmy |
Token ID | 12963 |
Feature activation | +4.74e-02 |
Pos logit contributions | |
driveway | +0.091 |
hallway | +0.087 |
loo | +0.084 |
weekend | +0.083 |
stair | +0.082 |
Neg logit contributions | |
ps | -0.070 |
teeth | -0.066 |
mus | -0.064 |
ss | -0.061 |
louder | -0.056 |
Token | was |
Token ID | 373 |
Feature activation | +1.66e-02 |
Pos logit contributions | |
stair | +0.074 |
weekend | +0.071 |
particular | +0.068 |
stack | +0.068 |
occasion | +0.067 |
Neg logit contributions | |
ps | -0.077 |
teeth | -0.070 |
ches | -0.068 |
ss | -0.066 |
mus | -0.060 |
Token | so |
Token ID | 523 |
Feature activation | +3.07e-02 |
Pos logit contributions | |
stair | +0.021 |
she | +0.020 |
weekend | +0.020 |
particular | +0.020 |
hallway | +0.019 |
Neg logit contributions | |
ps | -0.032 |
ss | -0.028 |
teeth | -0.028 |
ches | -0.027 |
mus | -0.027 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -3.44e-04 |
Pos logit contributions | |
she | +0.009 |
weekend | +0.009 |
stair | +0.009 |
hallway | +0.009 |
day | +0.009 |
Neg logit contributions | |
ps | -0.012 |
ss | -0.011 |
teeth | -0.010 |
ches | -0.010 |
mus | -0.010 |
Token | On |
Token ID | 2202 |
Feature activation | +7.96e-02 |
Pos logit contributions | |
ps | +0.001 |
ss | +0.001 |
ches | +0.001 |
teeth | +0.001 |
mus | +0.000 |
Neg logit contributions | |
she | -0.001 |
weekend | -0.000 |
day | -0.000 |
hallway | -0.000 |
stair | -0.000 |
Token | a |
Token ID | 257 |
Feature activation | +9.10e-02 |
Pos logit contributions | |
days | +0.176 |
policy | +0.166 |
stair | +0.162 |
spe | +0.160 |
hallway | +0.159 |
Neg logit contributions | |
ps | -0.064 |
ches | -0.059 |
teeth | -0.047 |
better | -0.044 |
harder | -0.038 |
Token | bright |
Token ID | 6016 |
Feature activation | +1.01e-01 |
Pos logit contributions | |
stack | +0.114 |
iked | +0.107 |
policy | +0.106 |
keeping | +0.104 |
days | +0.098 |
Neg logit contributions | |
ches | -0.170 |
funny | -0.151 |
loud | -0.146 |
hard | -0.146 |
silly | -0.144 |
Token | and |
Token ID | 290 |
Feature activation | +3.41e-02 |
Pos logit contributions | |
stack | +0.133 |
policy | +0.131 |
keeping | +0.125 |
days | +0.124 |
Have | +0.118 |
Neg logit contributions | |
ches | -0.153 |
me | -0.146 |
funny | -0.143 |
nice | -0.141 |
hard | -0.139 |
Token | sunny |
Token ID | 27737 |
Feature activation | +1.33e-01 |
Pos logit contributions | |
hallway | +0.024 |
stair | +0.023 |
driveway | +0.022 |
days | +0.021 |
she | +0.020 |
Neg logit contributions | |
ps | -0.076 |
ss | -0.071 |
mus | -0.071 |
teeth | -0.071 |
louder | -0.070 |
Token | day |
Token ID | 1110 |
Feature activation | +5.26e-02 |
Pos logit contributions | |
policy | +0.201 |
spe | +0.199 |
oker | +0.187 |
bek | +0.186 |
contain | +0.179 |
Neg logit contributions | |
me | -0.162 |
p | -0.138 |
- | -0.137 |
fur | -0.136 |
nice | -0.130 |
Token | , |
Token ID | 11 |
Feature activation | +5.33e-02 |
Pos logit contributions | |
driveway | +0.076 |
stack | +0.075 |
days | +0.074 |
hallway | +0.072 |
stair | +0.072 |
Neg logit contributions | |
ps | -0.081 |
teeth | -0.079 |
ches | -0.075 |
mus | -0.072 |
ss | -0.072 |
Token | a |
Token ID | 257 |
Feature activation | -1.39e-02 |
Pos logit contributions | |
driveway | +0.078 |
hallway | +0.076 |
stair | +0.075 |
days | +0.072 |
weekend | +0.071 |
Neg logit contributions | |
ps | -0.078 |
teeth | -0.077 |
mus | -0.072 |
ss | -0.069 |
louder | -0.066 |
Token | little |
Token ID | 1310 |
Feature activation | +3.15e-02 |
Pos logit contributions | |
ps | +0.024 |
ss | +0.020 |
ches | +0.019 |
teeth | +0.019 |
iddles | +0.018 |
Neg logit contributions | |
she | -0.023 |
day | -0.022 |
weekend | -0.022 |
stair | -0.022 |
hallway | -0.022 |
Token | boy |
Token ID | 2933 |
Feature activation | +5.99e-02 |
Pos logit contributions | |
hallway | +0.043 |
days | +0.042 |
weekend | +0.042 |
she | +0.041 |
stair | +0.041 |
Neg logit contributions | |
ps | -0.049 |
ss | -0.047 |
teeth | -0.044 |
mus | -0.043 |
ches | -0.043 |
Token | . |
Token ID | 13 |
Feature activation | +1.18e-02 |
Pos logit contributions | |
weekend | +0.028 |
she | +0.028 |
hallway | +0.027 |
stair | +0.027 |
day | +0.026 |
Neg logit contributions | |
ps | -0.023 |
teeth | -0.019 |
ss | -0.018 |
ches | -0.018 |
iddles | -0.017 |
Token |
|
Token ID | 198 |
Feature activation | +6.37e-03 |
Pos logit contributions | |
she | +0.018 |
weekend | +0.018 |
hallway | +0.018 |
stair | +0.018 |
day | +0.018 |
Neg logit contributions | |
ps | -0.018 |
ss | -0.017 |
teeth | -0.016 |
mus | -0.016 |
ches | -0.016 |
Token |
|
Token ID | 198 |
Feature activation | +1.94e-02 |
Pos logit contributions | |
she | +0.009 |
weekend | +0.009 |
hallway | +0.008 |
stair | +0.008 |
day | +0.008 |
Neg logit contributions | |
ps | -0.011 |
ss | -0.010 |
ches | -0.010 |
teeth | -0.010 |
mus | -0.010 |
Token | On |
Token ID | 2202 |
Feature activation | +5.58e-02 |
Pos logit contributions | |
she | +0.027 |
weekend | +0.026 |
holiday | +0.026 |
day | +0.026 |
stair | +0.026 |
Neg logit contributions | |
ps | -0.031 |
ss | -0.029 |
mus | -0.029 |
ches | -0.028 |
iddles | -0.026 |
Token | a |
Token ID | 257 |
Feature activation | +5.98e-02 |
Pos logit contributions | |
days | +0.115 |
des | +0.113 |
sandbox | +0.109 |
policy | +0.107 |
driveway | +0.106 |
Neg logit contributions | |
ps | -0.060 |
oni | -0.039 |
ches | -0.039 |
teeth | -0.039 |
bigger | -0.036 |
Token | sunny |
Token ID | 27737 |
Feature activation | +1.31e-01 |
Pos logit contributions | |
stack | +0.088 |
policy | +0.087 |
days | +0.082 |
packs | +0.080 |
iked | +0.080 |
Neg logit contributions | |
ches | -0.090 |
funny | -0.080 |
sharp | -0.076 |
loud | -0.073 |
silly | -0.073 |
Token | day |
Token ID | 1110 |
Feature activation | +6.04e-02 |
Pos logit contributions | |
spe | +0.185 |
policy | +0.179 |
haven | +0.179 |
otion | +0.163 |
awaited | +0.163 |
Neg logit contributions | |
me | -0.177 |
p | -0.169 |
ches | -0.159 |
fur | -0.146 |
w | -0.144 |
Token | , |
Token ID | 11 |
Feature activation | -2.42e-04 |
Pos logit contributions | |
driveway | +0.112 |
sandbox | +0.104 |
hallway | +0.102 |
stair | +0.101 |
stack | +0.101 |
Neg logit contributions | |
ps | -0.068 |
ss | -0.061 |
teeth | -0.057 |
louder | -0.057 |
mus | -0.057 |
Token | a |
Token ID | 257 |
Feature activation | -1.73e-02 |
Pos logit contributions | |
ps | +0.000 |
ss | +0.000 |
ches | +0.000 |
teeth | +0.000 |
mus | +0.000 |
Neg logit contributions | |
she | -0.000 |
weekend | -0.000 |
day | -0.000 |
hallway | -0.000 |
stair | -0.000 |
Token | little |
Token ID | 1310 |
Feature activation | +3.18e-02 |
Pos logit contributions | |
ps | +0.031 |
ss | +0.025 |
ches | +0.025 |
teeth | +0.024 |
iddles | +0.023 |
Neg logit contributions | |
she | -0.029 |
day | -0.028 |
weekend | -0.027 |
stair | -0.027 |
hallway | -0.027 |
Token | girl |
Token ID | 2576 |
Feature activation | +7.28e-02 |
Pos logit contributions | |
hallway | +0.047 |
days | +0.046 |
driveway | +0.046 |
weekend | +0.045 |
stair | +0.044 |
Neg logit contributions | |
ps | -0.047 |
ss | -0.045 |
rier | -0.042 |
louder | -0.042 |
teeth | -0.042 |
Token | beach |
Token ID | 10481 |
Feature activation | +4.37e-02 |
Pos logit contributions | |
she | +0.016 |
hallway | +0.014 |
days | +0.014 |
stair | +0.014 |
weekend | +0.014 |
Neg logit contributions | |
ps | -0.044 |
ss | -0.040 |
ches | -0.039 |
mus | -0.038 |
teeth | -0.038 |
Token | . |
Token ID | 13 |
Feature activation | +4.62e-02 |
Pos logit contributions | |
hallway | +0.086 |
stair | +0.086 |
particular | +0.084 |
weekend | +0.084 |
she | +0.082 |
Neg logit contributions | |
ps | -0.043 |
ches | -0.036 |
ss | -0.035 |
teeth | -0.033 |
iddles | -0.033 |
Token | It |
Token ID | 632 |
Feature activation | +2.99e-02 |
Pos logit contributions | |
boat | +0.057 |
weekend | +0.055 |
hallway | +0.055 |
day | +0.054 |
holiday | +0.054 |
Neg logit contributions | |
ps | -0.072 |
teeth | -0.071 |
ss | -0.070 |
iddles | -0.070 |
ches | -0.064 |
Token | was |
Token ID | 373 |
Feature activation | +1.78e-02 |
Pos logit contributions | |
stair | +0.045 |
weekend | +0.044 |
hallway | +0.042 |
sandbox | +0.042 |
holiday | +0.042 |
Neg logit contributions | |
ps | -0.046 |
ches | -0.043 |
ss | -0.039 |
teeth | -0.038 |
mus | -0.035 |
Token | a |
Token ID | 257 |
Feature activation | +4.60e-02 |
Pos logit contributions | |
stair | +0.032 |
hallway | +0.031 |
she | +0.031 |
days | +0.031 |
weekend | +0.030 |
Neg logit contributions | |
ps | -0.024 |
ss | -0.020 |
teeth | -0.020 |
ches | -0.019 |
mus | -0.018 |
Token | sunny |
Token ID | 27737 |
Feature activation | +1.27e-01 |
Pos logit contributions | |
she | +0.090 |
days | +0.090 |
path | +0.088 |
hallway | +0.087 |
stair | +0.087 |
Neg logit contributions | |
ps | -0.036 |
ss | -0.036 |
funny | -0.032 |
teeth | -0.032 |
smarter | -0.029 |
Token | day |
Token ID | 1110 |
Feature activation | +5.80e-02 |
Pos logit contributions | |
iked | +0.157 |
spe | +0.148 |
path | +0.147 |
Over | +0.145 |
osity | +0.144 |
Neg logit contributions | |
nice | -0.154 |
me | -0.151 |
good | -0.151 |
." | -0.150 |
- | -0.147 |
Token | and |
Token ID | 290 |
Feature activation | +4.95e-02 |
Pos logit contributions | |
sandbox | +0.111 |
stair | +0.109 |
hallway | +0.108 |
driveway | +0.105 |
particular | +0.105 |
Neg logit contributions | |
ps | -0.062 |
ss | -0.048 |
ches | -0.047 |
teeth | -0.046 |
louder | -0.044 |
Token | Kate |
Token ID | 16693 |
Feature activation | +7.79e-03 |
Pos logit contributions | |
driveway | +0.049 |
hallway | +0.046 |
stair | +0.043 |
policy | +0.042 |
boat | +0.042 |
Neg logit contributions | |
ps | -0.097 |
teeth | -0.092 |
ss | -0.088 |
ches | -0.086 |
louder | -0.086 |
Token | couldn |
Token ID | 3521 |
Feature activation | +2.47e-02 |
Pos logit contributions | |
she | +0.010 |
weekend | +0.009 |
stair | +0.009 |
hallway | +0.009 |
day | +0.009 |
Neg logit contributions | |
ps | -0.015 |
ss | -0.014 |
ches | -0.013 |
teeth | -0.013 |
mus | -0.013 |
Token | 't |
Token ID | 470 |
Feature activation | +3.10e-02 |
Pos logit contributions | |
weekend | +0.037 |
sandbox | +0.037 |
hallway | +0.037 |
stair | +0.036 |
day | +0.035 |
Neg logit contributions | |
ps | -0.041 |
ches | -0.036 |
ss | -0.035 |
teeth | -0.032 |
mus | -0.032 |
Token | . |
Token ID | 13 |
Feature activation | +2.00e-02 |
Pos logit contributions | |
ps | +0.054 |
ss | +0.047 |
iddles | +0.045 |
mus | +0.045 |
teeth | +0.045 |
Neg logit contributions | |
she | -0.048 |
day | -0.045 |
weekend | -0.042 |
they | -0.041 |
he | -0.041 |
Token |
|
Token ID | 198 |
Feature activation | +1.28e-02 |
Pos logit contributions | |
hallway | +0.031 |
stair | +0.031 |
weekend | +0.030 |
days | +0.030 |
holiday | +0.030 |
Neg logit contributions | |
ps | -0.030 |
ss | -0.029 |
teeth | -0.027 |
iddles | -0.026 |
ches | -0.025 |
Token |
|
Token ID | 198 |
Feature activation | +3.20e-02 |
Pos logit contributions | |
she | +0.018 |
hallway | +0.018 |
stair | +0.018 |
weekend | +0.018 |
day | +0.017 |
Neg logit contributions | |
ps | -0.022 |
ss | -0.020 |
ches | -0.019 |
teeth | -0.018 |
mus | -0.018 |
Token | On |
Token ID | 2202 |
Feature activation | +4.89e-02 |
Pos logit contributions | |
hallway | +0.055 |
stair | +0.054 |
days | +0.053 |
holiday | +0.053 |
she | +0.050 |
Neg logit contributions | |
ss | -0.043 |
ps | -0.041 |
ches | -0.038 |
ealing | -0.036 |
mus | -0.036 |
Token | a |
Token ID | 257 |
Feature activation | +4.42e-02 |
Pos logit contributions | |
days | +0.093 |
hallway | +0.088 |
des | +0.087 |
stair | +0.085 |
sandbox | +0.084 |
Neg logit contributions | |
ps | -0.064 |
oni | -0.047 |
ches | -0.046 |
teeth | -0.044 |
louder | -0.042 |
Token | sunny |
Token ID | 27737 |
Feature activation | +1.27e-01 |
Pos logit contributions | |
stack | +0.064 |
loo | +0.063 |
days | +0.062 |
policy | +0.060 |
she | +0.058 |
Neg logit contributions | |
ches | -0.070 |
ss | -0.057 |
ps | -0.056 |
mus | -0.053 |
funny | -0.053 |
Token | day |
Token ID | 1110 |
Feature activation | +6.89e-02 |
Pos logit contributions | |
haven | +0.172 |
spe | +0.164 |
policy | +0.162 |
days | +0.155 |
Sorry | +0.154 |
Neg logit contributions | |
ches | -0.171 |
p | -0.166 |
me | -0.155 |
ep | -0.148 |
ps | -0.145 |
Token | , |
Token ID | 11 |
Feature activation | +3.78e-02 |
Pos logit contributions | |
driveway | +0.125 |
hallway | +0.116 |
stair | +0.115 |
attic | +0.115 |
stack | +0.113 |
Neg logit contributions | |
ps | -0.081 |
ss | -0.078 |
louder | -0.075 |
p | -0.068 |
mus | -0.067 |
Token | Sally |
Token ID | 25737 |
Feature activation | +1.51e-02 |
Pos logit contributions | |
hallway | +0.053 |
days | +0.049 |
driveway | +0.048 |
stair | +0.045 |
occasion | +0.042 |
Neg logit contributions | |
ps | -0.067 |
ss | -0.065 |
louder | -0.061 |
teeth | -0.060 |
oni | -0.060 |
Token | and |
Token ID | 290 |
Feature activation | +2.99e-03 |
Pos logit contributions | |
weekend | +0.023 |
stair | +0.023 |
she | +0.022 |
hallway | +0.022 |
day | +0.022 |
Neg logit contributions | |
ps | -0.025 |
ss | -0.022 |
ches | -0.021 |
teeth | -0.020 |
mus | -0.019 |
Token | Tom |
Token ID | 4186 |
Feature activation | +1.31e-03 |
Pos logit contributions | |
she | +0.004 |
weekend | +0.004 |
hallway | +0.004 |
stair | +0.004 |
day | +0.004 |
Neg logit contributions | |
ps | -0.006 |
ss | -0.005 |
teeth | -0.005 |
ches | -0.005 |
mus | -0.005 |
Token | to |
Token ID | 284 |
Feature activation | +5.28e-03 |
Pos logit contributions | |
ps | +0.019 |
ss | +0.017 |
ches | +0.016 |
mus | +0.016 |
teeth | +0.015 |
Neg logit contributions | |
she | -0.010 |
day | -0.010 |
weekend | -0.010 |
stair | -0.009 |
hallway | -0.009 |
Token | explore |
Token ID | 7301 |
Feature activation | -5.03e-03 |
Pos logit contributions | |
she | +0.004 |
weekend | +0.004 |
hallway | +0.004 |
stair | +0.004 |
day | +0.004 |
Neg logit contributions | |
ps | -0.013 |
ss | -0.011 |
teeth | -0.011 |
ches | -0.011 |
mus | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | +4.50e-02 |
Pos logit contributions | |
ps | +0.009 |
ss | +0.008 |
ches | +0.007 |
teeth | +0.007 |
mus | +0.007 |
Neg logit contributions | |
she | -0.008 |
weekend | -0.008 |
day | -0.008 |
stair | -0.008 |
hallway | -0.007 |
Token | On |
Token ID | 1550 |
Feature activation | +7.18e-02 |
Pos logit contributions | |
boat | +0.077 |
hallway | +0.073 |
days | +0.070 |
stair | +0.070 |
holiday | +0.069 |
Neg logit contributions | |
teeth | -0.049 |
ps | -0.049 |
ss | -0.048 |
iddles | -0.047 |
ches | -0.041 |
Token | one |
Token ID | 530 |
Feature activation | +6.11e-02 |
Pos logit contributions | |
days | +0.137 |
des | +0.136 |
sandbox | +0.127 |
policy | +0.127 |
contain | +0.126 |
Neg logit contributions | |
ps | -0.075 |
ches | -0.056 |
hard | -0.054 |
teeth | -0.052 |
bad | -0.045 |
Token | sunny |
Token ID | 27737 |
Feature activation | +1.26e-01 |
Pos logit contributions | |
awaited | +0.057 |
driveway | +0.054 |
spe | +0.053 |
harbour | +0.053 |
hallway | +0.052 |
Neg logit contributions | |
bigger | -0.100 |
ches | -0.099 |
ss | -0.099 |
teeth | -0.099 |
louder | -0.098 |
Token | day |
Token ID | 1110 |
Feature activation | +7.49e-02 |
Pos logit contributions | |
spe | +0.181 |
contain | +0.180 |
stack | +0.166 |
Yep | +0.163 |
haven | +0.162 |
Neg logit contributions | |
ags | -0.158 |
- | -0.156 |
ps | -0.155 |
fur | -0.153 |
teeth | -0.150 |
Token | , |
Token ID | 11 |
Feature activation | +7.67e-02 |
Pos logit contributions | |
driveway | +0.189 |
stack | +0.186 |
harbour | +0.182 |
sandbox | +0.178 |
field | +0.178 |
Neg logit contributions | |
ps | -0.038 |
teeth | -0.034 |
ss | -0.029 |
good | -0.029 |
bad | -0.022 |
Token | she |
Token ID | 673 |
Feature activation | +3.53e-02 |
Pos logit contributions | |
driveway | +0.104 |
boat | +0.100 |
harbour | +0.095 |
hallway | +0.092 |
voyage | +0.090 |
Neg logit contributions | |
funny | -0.123 |
teeth | -0.121 |
ps | -0.120 |
loud | -0.117 |
good | -0.116 |
Token | went |
Token ID | 1816 |
Feature activation | +2.98e-02 |
Pos logit contributions | |
stair | +0.053 |
particular | +0.051 |
stack | +0.051 |
weekend | +0.051 |
days | +0.050 |
Neg logit contributions | |
ps | -0.054 |
teeth | -0.050 |
ches | -0.048 |
ss | -0.047 |
iddles | -0.042 |
Token | on |
Token ID | 319 |
Feature activation | +5.41e-03 |
Pos logit contributions | |
days | +0.052 |
weekend | +0.052 |
stair | +0.050 |
particular | +0.050 |
spe | +0.050 |
Neg logit contributions | |
ps | -0.041 |
ches | -0.035 |
ss | -0.030 |
iddles | -0.030 |
louder | -0.030 |
Token | and |
Token ID | 290 |
Feature activation | +4.09e-02 |
Pos logit contributions | |
she | +0.020 |
weekend | +0.020 |
hallway | +0.019 |
stair | +0.019 |
day | +0.019 |
Neg logit contributions | |
ps | -0.018 |
ss | -0.015 |
ches | -0.015 |
teeth | -0.015 |
mus | -0.014 |
Token | Patrick |
Token ID | 9925 |
Feature activation | -1.44e-02 |
Pos logit contributions | |
stair | +0.060 |
hallway | +0.059 |
weekend | +0.057 |
driveway | +0.056 |
hill | +0.053 |
Neg logit contributions | |
ps | -0.072 |
teeth | -0.062 |
ss | -0.061 |
iddles | -0.057 |
mus | -0.057 |
Token | . |
Token ID | 13 |
Feature activation | +8.85e-03 |
Pos logit contributions | |
ps | +0.024 |
ss | +0.021 |
ches | +0.020 |
mus | +0.019 |
teeth | +0.019 |
Neg logit contributions | |
she | -0.023 |
day | -0.023 |
weekend | -0.022 |
they | -0.022 |
hallway | -0.021 |
Token | On |
Token ID | 1550 |
Feature activation | +4.79e-02 |
Pos logit contributions | |
she | +0.011 |
weekend | +0.011 |
stair | +0.011 |
day | +0.010 |
hallway | +0.010 |
Neg logit contributions | |
ps | -0.017 |
ss | -0.015 |
teeth | -0.015 |
ches | -0.015 |
iddles | -0.014 |
Token | a |
Token ID | 257 |
Feature activation | +5.16e-02 |
Pos logit contributions | |
days | +0.086 |
des | +0.084 |
hallway | +0.083 |
stair | +0.083 |
she | +0.081 |
Neg logit contributions | |
ps | -0.063 |
ches | -0.048 |
teeth | -0.047 |
oni | -0.045 |
ss | -0.041 |
Token | sunny |
Token ID | 27737 |
Feature activation | +1.25e-01 |
Pos logit contributions | |
policy | +0.086 |
stack | +0.086 |
days | +0.083 |
discovering | +0.082 |
she | +0.082 |
Neg logit contributions | |
ches | -0.065 |
funny | -0.050 |
ps | -0.050 |
ss | -0.049 |
loud | -0.048 |
Token | day |
Token ID | 1110 |
Feature activation | +6.09e-02 |
Pos logit contributions | |
awaited | +0.150 |
policy | +0.149 |
des | +0.146 |
haven | +0.144 |
spe | +0.142 |
Neg logit contributions | |
ches | -0.174 |
me | -0.171 |
ps | -0.164 |
fur | -0.163 |
ep | -0.162 |
Token | , |
Token ID | 11 |
Feature activation | +5.44e-02 |
Pos logit contributions | |
driveway | +0.124 |
sandbox | +0.120 |
stack | +0.119 |
hill | +0.115 |
stair | +0.114 |
Neg logit contributions | |
ps | -0.061 |
teeth | -0.057 |
ches | -0.052 |
ss | -0.050 |
es | -0.046 |
Token | they |
Token ID | 484 |
Feature activation | +1.62e-02 |
Pos logit contributions | |
stack | +0.065 |
driveway | +0.062 |
occasion | +0.061 |
hallway | +0.061 |
stair | +0.057 |
Neg logit contributions | |
teeth | -0.091 |
louder | -0.090 |
funny | -0.090 |
ss | -0.089 |
ps | -0.088 |
Token | decided |
Token ID | 3066 |
Feature activation | +3.54e-02 |
Pos logit contributions | |
stair | +0.020 |
hallway | +0.019 |
she | +0.019 |
weekend | +0.019 |
particular | +0.019 |
Neg logit contributions | |
ps | -0.032 |
ss | -0.028 |
ches | -0.027 |
teeth | -0.027 |
mus | -0.026 |
Token | to |
Token ID | 284 |
Feature activation | +3.95e-02 |
Pos logit contributions | |
hallway | +0.067 |
particular | +0.065 |
days | +0.065 |
stair | +0.063 |
driveway | +0.062 |
Neg logit contributions | |
ps | -0.045 |
teeth | -0.039 |
ss | -0.039 |
louder | -0.037 |
ches | -0.035 |
Token | to |
Token ID | 284 |
Feature activation | +3.37e-02 |
Pos logit contributions | |
stair | +0.065 |
particular | +0.064 |
hallway | +0.064 |
sandbox | +0.064 |
days | +0.062 |
Neg logit contributions | |
ps | -0.052 |
ches | -0.046 |
ss | -0.042 |
mus | -0.039 |
iddles | -0.039 |
Token | the |
Token ID | 262 |
Feature activation | +2.45e-02 |
Pos logit contributions | |
hallway | +0.054 |
loo | +0.053 |
she | +0.053 |
stair | +0.052 |
weekend | +0.052 |
Neg logit contributions | |
ps | -0.048 |
teeth | -0.040 |
ss | -0.039 |
ches | -0.038 |
mus | -0.037 |
Token | park |
Token ID | 3952 |
Feature activation | +5.49e-02 |
Pos logit contributions | |
she | +0.021 |
days | +0.020 |
weekend | +0.019 |
hallway | +0.018 |
day | +0.018 |
Neg logit contributions | |
ps | -0.053 |
ss | -0.049 |
ches | -0.047 |
teeth | -0.046 |
mus | -0.046 |
Token | on |
Token ID | 319 |
Feature activation | +3.36e-02 |
Pos logit contributions | |
hallway | +0.129 |
hurricane | +0.128 |
stair | +0.127 |
loo | +0.126 |
particular | +0.125 |
Neg logit contributions | |
ps | -0.025 |
ss | -0.017 |
ches | -0.016 |
ows | -0.016 |
teeth | -0.015 |
Token | a |
Token ID | 257 |
Feature activation | +5.30e-02 |
Pos logit contributions | |
hallway | +0.063 |
days | +0.062 |
loo | +0.062 |
weekend | +0.061 |
stair | +0.060 |
Neg logit contributions | |
ps | -0.039 |
teeth | -0.031 |
ches | -0.029 |
ss | -0.027 |
mus | -0.026 |
Token | sunny |
Token ID | 27737 |
Feature activation | +1.25e-01 |
Pos logit contributions | |
days | +0.073 |
stack | +0.072 |
policy | +0.072 |
she | +0.071 |
loo | +0.070 |
Neg logit contributions | |
ches | -0.078 |
ss | -0.068 |
funny | -0.067 |
bigger | -0.064 |
ps | -0.064 |
Token | day |
Token ID | 1110 |
Feature activation | +5.71e-02 |
Pos logit contributions | |
policy | +0.162 |
otion | +0.146 |
awaited | +0.143 |
haven | +0.138 |
she | +0.134 |
Neg logit contributions | |
me | -0.184 |
ches | -0.161 |
nice | -0.154 |
p | -0.153 |
." | -0.149 |
Token | . |
Token ID | 13 |
Feature activation | +1.01e-02 |
Pos logit contributions | |
stack | +0.132 |
stair | +0.132 |
loo | +0.130 |
hallway | +0.129 |
sandbox | +0.129 |
Neg logit contributions | |
ps | -0.032 |
ss | -0.026 |
ches | -0.024 |
louder | -0.021 |
ealing | -0.019 |
Token | Mack |
Token ID | 16640 |
Feature activation | -7.49e-03 |
Pos logit contributions | |
she | +0.013 |
weekend | +0.013 |
hallway | +0.013 |
stair | +0.012 |
day | +0.012 |
Neg logit contributions | |
ps | -0.019 |
ss | -0.017 |
ches | -0.017 |
teeth | -0.017 |
mus | -0.016 |
Token | enzie |
Token ID | 26389 |
Feature activation | +1.82e-02 |
Pos logit contributions | |
ps | +0.014 |
ss | +0.012 |
teeth | +0.012 |
mus | +0.011 |
ches | +0.011 |
Neg logit contributions | |
she | -0.011 |
weekend | -0.010 |
day | -0.010 |
hallway | -0.010 |
stair | -0.010 |
Token | loved |
Token ID | 6151 |
Feature activation | +2.34e-02 |
Pos logit contributions | |
weekend | +0.022 |
she | +0.022 |
stair | +0.022 |
hallway | +0.022 |
day | +0.021 |
Neg logit contributions | |
ps | -0.035 |
ss | -0.031 |
ches | -0.031 |
teeth | -0.030 |
mus | -0.029 |
Token | in |
Token ID | 287 |
Feature activation | +6.98e-02 |
Pos logit contributions | |
loo | +0.093 |
policy | +0.093 |
hallway | +0.093 |
path | +0.093 |
stair | +0.092 |
Neg logit contributions | |
ps | -0.051 |
louder | -0.042 |
mus | -0.037 |
ches | -0.036 |
teeth | -0.035 |
Token | the |
Token ID | 262 |
Feature activation | +6.38e-02 |
Pos logit contributions | |
loo | +0.169 |
storm | +0.169 |
boat | +0.164 |
occasion | +0.160 |
oker | +0.159 |
Neg logit contributions | |
ps | -0.030 |
teeth | -0.027 |
louder | -0.020 |
loud | -0.019 |
funny | -0.015 |
Token | park |
Token ID | 3952 |
Feature activation | +7.52e-02 |
Pos logit contributions | |
days | +0.090 |
policy | +0.089 |
contain | +0.081 |
iked | +0.079 |
awaited | +0.078 |
Neg logit contributions | |
bigger | -0.085 |
funny | -0.085 |
ps | -0.084 |
louder | -0.082 |
better | -0.082 |
Token | on |
Token ID | 319 |
Feature activation | +4.34e-02 |
Pos logit contributions | |
loo | +0.198 |
hurricane | +0.193 |
stair | +0.191 |
awaited | +0.186 |
policy | +0.185 |
Neg logit contributions | |
ps | -0.022 |
louder | -0.015 |
teeth | -0.011 |
. | -0.008 |
softly | -0.007 |
Token | a |
Token ID | 257 |
Feature activation | +5.86e-02 |
Pos logit contributions | |
days | +0.091 |
loo | +0.088 |
boat | +0.087 |
weekend | +0.086 |
occasion | +0.085 |
Neg logit contributions | |
ps | -0.040 |
teeth | -0.035 |
ches | -0.025 |
iddles | -0.025 |
mus | -0.024 |
Token | sunny |
Token ID | 27737 |
Feature activation | +1.25e-01 |
Pos logit contributions | |
stack | +0.090 |
days | +0.089 |
policy | +0.089 |
loo | +0.086 |
iked | +0.083 |
Neg logit contributions | |
ches | -0.077 |
funny | -0.069 |
bigger | -0.066 |
loud | -0.064 |
better | -0.064 |
Token | day |
Token ID | 1110 |
Feature activation | +6.11e-02 |
Pos logit contributions | |
policy | +0.170 |
otion | +0.154 |
awaited | +0.149 |
oker | +0.149 |
haven | +0.146 |
Neg logit contributions | |
me | -0.185 |
nice | -0.153 |
fur | -0.151 |
p | -0.149 |
ep | -0.146 |
Token | . |
Token ID | 13 |
Feature activation | +2.30e-02 |
Pos logit contributions | |
stack | +0.150 |
loo | +0.147 |
stair | +0.145 |
days | +0.143 |
sandbox | +0.141 |
Neg logit contributions | |
ps | -0.025 |
louder | -0.020 |
ss | -0.018 |
ches | -0.015 |
mus | -0.014 |
Token | It |
Token ID | 632 |
Feature activation | +2.95e-02 |
Pos logit contributions | |
stair | +0.029 |
weekend | +0.029 |
day | +0.029 |
hallway | +0.029 |
she | +0.029 |
Neg logit contributions | |
ps | -0.040 |
ss | -0.036 |
teeth | -0.036 |
ches | -0.035 |
mus | -0.035 |
Token | was |
Token ID | 373 |
Feature activation | +1.95e-02 |
Pos logit contributions | |
weekend | +0.044 |
holiday | +0.043 |
days | +0.042 |
stair | +0.042 |
occasion | +0.042 |
Neg logit contributions | |
ps | -0.045 |
ches | -0.043 |
ss | -0.040 |
teeth | -0.039 |
mus | -0.036 |
Token | a |
Token ID | 257 |
Feature activation | +4.61e-02 |
Pos logit contributions | |
stair | +0.037 |
days | +0.036 |
she | +0.036 |
hallway | +0.035 |
weekend | +0.035 |
Neg logit contributions | |
ps | -0.024 |
teeth | -0.019 |
ss | -0.019 |
ches | -0.018 |
mus | -0.017 |
Token | . |
Token ID | 13 |
Feature activation | +4.71e-02 |
Pos logit contributions | |
ps | +0.004 |
ss | +0.004 |
ches | +0.003 |
teeth | +0.003 |
mus | +0.003 |
Neg logit contributions | |
she | -0.003 |
weekend | -0.003 |
day | -0.003 |
hallway | -0.003 |
stair | -0.003 |
Token | On |
Token ID | 1550 |
Feature activation | +7.36e-02 |
Pos logit contributions | |
she | +0.080 |
stair | +0.076 |
days | +0.075 |
hallway | +0.075 |
sandbox | +0.074 |
Neg logit contributions | |
ps | -0.056 |
teeth | -0.055 |
iddles | -0.048 |
ss | -0.048 |
mus | -0.045 |
Token | a |
Token ID | 257 |
Feature activation | +5.45e-02 |
Pos logit contributions | |
days | +0.123 |
yt | +0.117 |
des | +0.115 |
sandbox | +0.115 |
policy | +0.113 |
Neg logit contributions | |
ps | -0.093 |
teeth | -0.076 |
ches | -0.075 |
mus | -0.065 |
hard | -0.064 |
Token | warm |
Token ID | 5814 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
policy | +0.061 |
stack | +0.060 |
days | +0.058 |
loo | +0.056 |
iked | +0.052 |
Neg logit contributions | |
ches | -0.100 |
ss | -0.085 |
funny | -0.085 |
loud | -0.083 |
mus | -0.083 |
Token | and |
Token ID | 290 |
Feature activation | +3.11e-02 |
Pos logit contributions | |
days | +0.127 |
Have | +0.125 |
policy | +0.123 |
We | +0.117 |
spe | +0.115 |
Neg logit contributions | |
ches | -0.150 |
me | -0.140 |
ps | -0.129 |
fur | -0.129 |
co | -0.128 |
Token | sunny |
Token ID | 27737 |
Feature activation | +1.24e-01 |
Pos logit contributions | |
stair | +0.031 |
hallway | +0.030 |
she | +0.030 |
days | +0.030 |
driveway | +0.029 |
Neg logit contributions | |
ps | -0.059 |
teeth | -0.056 |
ss | -0.054 |
mus | -0.053 |
ches | -0.053 |
Token | day |
Token ID | 1110 |
Feature activation | +6.35e-02 |
Pos logit contributions | |
policy | +0.161 |
haven | +0.150 |
oker | +0.148 |
spe | +0.147 |
contain | +0.147 |
Neg logit contributions | |
me | -0.163 |
ches | -0.140 |
- | -0.139 |
nice | -0.139 |
fur | -0.135 |
Token | , |
Token ID | 11 |
Feature activation | +6.68e-02 |
Pos logit contributions | |
stack | +0.117 |
stair | +0.115 |
driveway | +0.115 |
sandbox | +0.114 |
hallway | +0.112 |
Neg logit contributions | |
ps | -0.070 |
teeth | -0.067 |
louder | -0.058 |
mus | -0.058 |
ss | -0.056 |
Token | John |
Token ID | 1757 |
Feature activation | +1.03e-02 |
Pos logit contributions | |
hallway | +0.084 |
wander | +0.082 |
ite | +0.081 |
days | +0.081 |
stair | +0.081 |
Neg logit contributions | |
teeth | -0.103 |
ps | -0.100 |
louder | -0.095 |
loud | -0.093 |
funny | -0.093 |
Token | and |
Token ID | 290 |
Feature activation | +3.67e-02 |
Pos logit contributions | |
she | +0.012 |
stair | +0.012 |
weekend | +0.012 |
hallway | +0.012 |
day | +0.012 |
Neg logit contributions | |
ps | -0.021 |
ss | -0.018 |
ches | -0.018 |
teeth | -0.018 |
mus | -0.017 |
Token | his |
Token ID | 465 |
Feature activation | -1.28e-02 |
Pos logit contributions | |
hallway | +0.062 |
stair | +0.061 |
path | +0.060 |
sandbox | +0.060 |
stack | +0.060 |
Neg logit contributions | |
teeth | -0.044 |
ps | -0.044 |
ss | -0.037 |
uable | -0.034 |
iddles | -0.034 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +6.02e-04 |
Pos logit contributions | |
stair | +0.028 |
weekend | +0.028 |
hallway | +0.027 |
she | +0.027 |
holiday | +0.027 |
Neg logit contributions | |
ps | -0.031 |
ss | -0.029 |
teeth | -0.027 |
ches | -0.026 |
mus | -0.026 |
Token | Mark |
Token ID | 9704 |
Feature activation | +1.22e-02 |
Pos logit contributions | |
she | +0.001 |
weekend | +0.001 |
day | +0.001 |
hallway | +0.001 |
stair | +0.001 |
Neg logit contributions | |
ps | -0.001 |
ss | -0.001 |
ches | -0.001 |
teeth | -0.001 |
mus | -0.001 |
Token | was |
Token ID | 373 |
Feature activation | +2.20e-02 |
Pos logit contributions | |
she | +0.016 |
weekend | +0.016 |
stair | +0.016 |
hallway | +0.016 |
day | +0.015 |
Neg logit contributions | |
ps | -0.023 |
ss | -0.020 |
ches | -0.020 |
teeth | -0.020 |
mus | -0.019 |
Token | very |
Token ID | 845 |
Feature activation | +2.53e-02 |
Pos logit contributions | |
stair | +0.037 |
she | +0.036 |
hallway | +0.036 |
weekend | +0.035 |
days | +0.035 |
Neg logit contributions | |
ps | -0.034 |
teeth | -0.027 |
ss | -0.027 |
mus | -0.026 |
ches | -0.025 |
Token | excited |
Token ID | 6568 |
Feature activation | +5.00e-02 |
Pos logit contributions | |
stair | +0.049 |
hallway | +0.048 |
she | +0.046 |
weekend | +0.046 |
occasion | +0.046 |
Neg logit contributions | |
ps | -0.030 |
teeth | -0.024 |
ss | -0.024 |
mus | -0.022 |
ealing | -0.022 |
Token | because |
Token ID | 780 |
Feature activation | +1.24e-01 |
Pos logit contributions | |
hallway | +0.101 |
sandbox | +0.100 |
particular | +0.099 |
stair | +0.099 |
days | +0.098 |
Neg logit contributions | |
ps | -0.051 |
teeth | -0.042 |
ches | -0.039 |
ss | -0.039 |
iddles | -0.036 |
Token | today |
Token ID | 1909 |
Feature activation | +7.25e-02 |
Pos logit contributions | |
spine | +0.178 |
distance | +0.169 |
stack | +0.167 |
occasion | +0.161 |
ugg | +0.161 |
Neg logit contributions | |
teeth | -0.145 |
too | -0.140 |
loud | -0.139 |
a | -0.133 |
bigger | -0.128 |
Token | he |
Token ID | 339 |
Feature activation | +2.52e-02 |
Pos logit contributions | |
days | +0.123 |
driveway | +0.122 |
stack | +0.122 |
hallway | +0.120 |
stair | +0.120 |
Neg logit contributions | |
teeth | -0.083 |
ps | -0.073 |
ss | -0.071 |
ches | -0.066 |
oni | -0.064 |
Token | got |
Token ID | 1392 |
Feature activation | +8.35e-03 |
Pos logit contributions | |
weekend | +0.032 |
stair | +0.032 |
she | +0.030 |
hallway | +0.030 |
sandbox | +0.029 |
Neg logit contributions | |
ps | -0.046 |
ches | -0.042 |
ss | -0.041 |
teeth | -0.041 |
mus | -0.038 |
Token | to |
Token ID | 284 |
Feature activation | +5.54e-02 |
Pos logit contributions | |
hallway | +0.012 |
stair | +0.012 |
she | +0.012 |
weekend | +0.011 |
day | +0.011 |
Neg logit contributions | |
ps | -0.015 |
ss | -0.013 |
teeth | -0.013 |
ches | -0.012 |
mus | -0.012 |
Token | go |
Token ID | 467 |
Feature activation | +2.93e-02 |
Pos logit contributions | |
loo | +0.071 |
weekend | +0.070 |
hallway | +0.069 |
she | +0.069 |
stair | +0.068 |
Neg logit contributions | |
ps | -0.097 |
teeth | -0.085 |
ss | -0.076 |
mus | -0.073 |
ches | -0.071 |
Token | |
Token ID | 220 |
Feature activation | -2.10e-02 |
Pos logit contributions | |
she | +0.007 |
weekend | +0.007 |
hallway | +0.007 |
stair | +0.007 |
day | +0.007 |
Neg logit contributions | |
ps | -0.010 |
ss | -0.009 |
teeth | -0.009 |
ches | -0.009 |
mus | -0.008 |
Token |
|
Token ID | 198 |
Feature activation | -3.01e-03 |
Pos logit contributions | |
ps | +0.046 |
ss | +0.040 |
ches | +0.039 |
teeth | +0.039 |
mus | +0.038 |
Neg logit contributions | |
she | -0.025 |
weekend | -0.022 |
day | -0.022 |
stair | -0.021 |
hallway | -0.021 |
Token |
|
Token ID | 198 |
Feature activation | +2.97e-02 |
Pos logit contributions | |
ps | +0.006 |
ss | +0.005 |
ches | +0.005 |
teeth | +0.005 |
mus | +0.005 |
Neg logit contributions | |
she | -0.004 |
weekend | -0.004 |
day | -0.004 |
hallway | -0.004 |
stair | -0.004 |
Token | On |
Token ID | 2202 |
Feature activation | +5.77e-02 |
Pos logit contributions | |
she | +0.047 |
stair | +0.046 |
hallway | +0.046 |
stack | +0.046 |
sandbox | +0.044 |
Neg logit contributions | |
ss | -0.041 |
ps | -0.038 |
ches | -0.037 |
mus | -0.037 |
iddles | -0.034 |
Token | one |
Token ID | 530 |
Feature activation | +5.72e-02 |
Pos logit contributions | |
days | +0.112 |
sandbox | +0.107 |
des | +0.103 |
hallway | +0.102 |
stair | +0.101 |
Neg logit contributions | |
ps | -0.068 |
oni | -0.052 |
ches | -0.051 |
teeth | -0.049 |
harder | -0.047 |
Token | sunny |
Token ID | 27737 |
Feature activation | +1.24e-01 |
Pos logit contributions | |
hallway | +0.054 |
driveway | +0.053 |
stair | +0.051 |
awaited | +0.050 |
days | +0.050 |
Neg logit contributions | |
teeth | -0.105 |
ps | -0.103 |
ss | -0.100 |
ches | -0.099 |
better | -0.095 |
Token | day |
Token ID | 1110 |
Feature activation | +7.02e-02 |
Pos logit contributions | |
spe | +0.164 |
haven | +0.150 |
Have | +0.143 |
stack | +0.138 |
line | +0.136 |
Neg logit contributions | |
teeth | -0.188 |
ps | -0.182 |
p | -0.178 |
fur | -0.177 |
- | -0.173 |
Token | , |
Token ID | 11 |
Feature activation | +4.41e-02 |
Pos logit contributions | |
driveway | +0.164 |
stack | +0.163 |
sandbox | +0.162 |
stair | +0.153 |
field | +0.153 |
Neg logit contributions | |
ps | -0.042 |
teeth | -0.041 |
ss | -0.036 |
mus | -0.032 |
good | -0.030 |
Token | Jack |
Token ID | 3619 |
Feature activation | +1.70e-03 |
Pos logit contributions | |
hallway | +0.067 |
driveway | +0.064 |
days | +0.064 |
wander | +0.059 |
loo | +0.059 |
Neg logit contributions | |
ps | -0.065 |
teeth | -0.064 |
louder | -0.063 |
ss | -0.061 |
oni | -0.058 |
Token | 's |
Token ID | 338 |
Feature activation | +8.68e-03 |
Pos logit contributions | |
she | +0.002 |
weekend | +0.002 |
stair | +0.002 |
day | +0.002 |
hallway | +0.002 |
Neg logit contributions | |
ps | -0.003 |
ss | -0.003 |
ches | -0.003 |
teeth | -0.003 |
mus | -0.003 |
Token | mom |
Token ID | 1995 |
Feature activation | +2.59e-03 |
Pos logit contributions | |
she | +0.014 |
weekend | +0.014 |
hallway | +0.013 |
stair | +0.013 |
day | +0.013 |
Neg logit contributions | |
ps | -0.014 |
teeth | -0.012 |
ss | -0.012 |
ches | -0.011 |
mus | -0.011 |
Token | Everybody |
Token ID | 28172 |
Feature activation | +4.63e-02 |
Pos logit contributions | |
ps | +0.017 |
ss | +0.015 |
teeth | +0.015 |
ches | +0.015 |
mus | +0.014 |
Neg logit contributions | |
she | -0.011 |
day | -0.010 |
weekend | -0.010 |
hallway | -0.010 |
stair | -0.010 |
Token | was |
Token ID | 373 |
Feature activation | +1.53e-02 |
Pos logit contributions | |
hallway | +0.079 |
stair | +0.079 |
weekend | +0.076 |
particular | +0.076 |
driveway | +0.075 |
Neg logit contributions | |
ps | -0.066 |
ss | -0.057 |
ches | -0.052 |
teeth | -0.050 |
mus | -0.050 |
Token | outside |
Token ID | 2354 |
Feature activation | +6.59e-02 |
Pos logit contributions | |
she | +0.018 |
stair | +0.017 |
hallway | +0.017 |
weekend | +0.017 |
particular | +0.016 |
Neg logit contributions | |
ps | -0.031 |
ss | -0.027 |
teeth | -0.026 |
mus | -0.026 |
iddles | -0.025 |
Token | on |
Token ID | 319 |
Feature activation | +2.09e-02 |
Pos logit contributions | |
boat | +0.112 |
loo | +0.106 |
occasion | +0.105 |
stack | +0.104 |
path | +0.104 |
Neg logit contributions | |
ps | -0.095 |
teeth | -0.080 |
mus | -0.071 |
louder | -0.070 |
ss | -0.068 |
Token | a |
Token ID | 257 |
Feature activation | +5.26e-02 |
Pos logit contributions | |
weekend | +0.035 |
days | +0.035 |
hallway | +0.035 |
she | +0.035 |
day | +0.033 |
Neg logit contributions | |
ps | -0.030 |
teeth | -0.025 |
ss | -0.023 |
mus | -0.022 |
ches | -0.022 |
Token | sunny |
Token ID | 27737 |
Feature activation | +1.24e-01 |
Pos logit contributions | |
policy | +0.073 |
days | +0.072 |
stack | +0.072 |
iked | +0.071 |
she | +0.069 |
Neg logit contributions | |
ches | -0.074 |
funny | -0.066 |
ss | -0.064 |
loud | -0.063 |
bigger | -0.062 |
Token | day |
Token ID | 1110 |
Feature activation | +4.83e-02 |
Pos logit contributions | |
policy | +0.166 |
oker | +0.155 |
iked | +0.153 |
spe | +0.149 |
bars | +0.148 |
Neg logit contributions | |
me | -0.182 |
fur | -0.149 |
p | -0.146 |
." | -0.145 |
. | -0.144 |
Token | when |
Token ID | 618 |
Feature activation | +4.68e-02 |
Pos logit contributions | |
stair | +0.097 |
weekend | +0.097 |
particular | +0.097 |
stack | +0.096 |
hallway | +0.095 |
Neg logit contributions | |
ps | -0.047 |
teeth | -0.036 |
ss | -0.035 |
mus | -0.032 |
louder | -0.031 |
Token | suddenly |
Token ID | 6451 |
Feature activation | +2.50e-02 |
Pos logit contributions | |
hallway | +0.054 |
occasion | +0.054 |
driveway | +0.053 |
weekend | +0.053 |
days | +0.052 |
Neg logit contributions | |
ps | -0.081 |
teeth | -0.075 |
louder | -0.073 |
mus | -0.072 |
ss | -0.068 |
Token | Jack |
Token ID | 3619 |
Feature activation | -7.11e-03 |
Pos logit contributions | |
weekend | +0.036 |
days | +0.034 |
hallway | +0.034 |
stair | +0.034 |
day | +0.033 |
Neg logit contributions | |
ps | -0.044 |
teeth | -0.038 |
ss | -0.037 |
ches | -0.036 |
mus | -0.035 |
Token | noticed |
Token ID | 6810 |
Feature activation | +1.70e-02 |
Pos logit contributions | |
ps | +0.015 |
ss | +0.013 |
teeth | +0.013 |
mus | +0.012 |
ches | +0.012 |
Neg logit contributions | |
she | -0.009 |
weekend | -0.008 |
hallway | -0.008 |
day | -0.008 |
stair | -0.008 |
Token | garden |
Token ID | 11376 |
Feature activation | +5.70e-02 |
Pos logit contributions | |
awaited | +0.092 |
boat | +0.089 |
Jane | +0.089 |
des | +0.084 |
contain | +0.084 |
Neg logit contributions | |
teeth | -0.094 |
ps | -0.089 |
ears | -0.076 |
fur | -0.075 |
hair | -0.071 |
Token | . |
Token ID | 13 |
Feature activation | +2.17e-02 |
Pos logit contributions | |
weekend | +0.097 |
awaited | +0.097 |
days | +0.096 |
loo | +0.095 |
stair | +0.095 |
Neg logit contributions | |
ps | -0.071 |
teeth | -0.060 |
mus | -0.055 |
ss | -0.055 |
ches | -0.052 |
Token | It |
Token ID | 632 |
Feature activation | +2.74e-02 |
Pos logit contributions | |
weekend | +0.025 |
day | +0.023 |
holiday | +0.023 |
she | +0.023 |
days | +0.023 |
Neg logit contributions | |
ps | -0.041 |
teeth | -0.038 |
ss | -0.038 |
mus | -0.036 |
iddles | -0.036 |
Token | was |
Token ID | 373 |
Feature activation | +2.27e-02 |
Pos logit contributions | |
weekend | +0.041 |
holiday | +0.039 |
stair | +0.038 |
driveway | +0.038 |
occasion | +0.037 |
Neg logit contributions | |
ps | -0.043 |
ches | -0.040 |
teeth | -0.037 |
ss | -0.037 |
mus | -0.034 |
Token | a |
Token ID | 257 |
Feature activation | +4.25e-02 |
Pos logit contributions | |
stair | +0.039 |
days | +0.039 |
she | +0.039 |
weekend | +0.039 |
hallway | +0.039 |
Neg logit contributions | |
ps | -0.031 |
teeth | -0.025 |
ss | -0.024 |
mus | -0.023 |
ches | -0.022 |
Token | sunny |
Token ID | 27737 |
Feature activation | +1.24e-01 |
Pos logit contributions | |
days | +0.072 |
she | +0.071 |
weekend | +0.070 |
path | +0.068 |
iked | +0.068 |
Neg logit contributions | |
ss | -0.045 |
ps | -0.044 |
teeth | -0.042 |
funny | -0.040 |
mus | -0.040 |
Token | day |
Token ID | 1110 |
Feature activation | +5.24e-02 |
Pos logit contributions | |
iked | +0.172 |
contain | +0.156 |
des | +0.149 |
spe | +0.149 |
path | +0.141 |
Neg logit contributions | |
me | -0.156 |
nice | -0.147 |
." | -0.144 |
too | -0.143 |
sweet | -0.142 |
Token | and |
Token ID | 290 |
Feature activation | +3.74e-02 |
Pos logit contributions | |
stair | +0.100 |
sandbox | +0.100 |
hallway | +0.099 |
stack | +0.098 |
driveway | +0.098 |
Neg logit contributions | |
ps | -0.053 |
teeth | -0.043 |
louder | -0.041 |
ss | -0.040 |
ches | -0.039 |
Token | the |
Token ID | 262 |
Feature activation | +2.32e-02 |
Pos logit contributions | |
days | +0.045 |
loo | +0.044 |
hallway | +0.044 |
weekend | +0.044 |
driveway | +0.044 |
Neg logit contributions | |
ps | -0.064 |
teeth | -0.059 |
ss | -0.057 |
ches | -0.056 |
louder | -0.055 |
Token | grass |
Token ID | 8701 |
Feature activation | +4.06e-02 |
Pos logit contributions | |
she | +0.035 |
days | +0.034 |
weekend | +0.034 |
hallway | +0.033 |
stair | +0.033 |
Neg logit contributions | |
ps | -0.036 |
ss | -0.030 |
teeth | -0.030 |
ches | -0.028 |
iddles | -0.028 |
Token | was |
Token ID | 373 |
Feature activation | +1.24e-02 |
Pos logit contributions | |
weekend | +0.059 |
holiday | +0.056 |
driveway | +0.055 |
hallway | +0.055 |
days | +0.054 |
Neg logit contributions | |
ps | -0.067 |
ches | -0.057 |
ss | -0.057 |
teeth | -0.055 |
ows | -0.051 |
Token | chooses |
Token ID | 19769 |
Feature activation | -5.80e-02 |
Pos logit contributions | |
ps | +0.069 |
ss | +0.063 |
rier | +0.061 |
mus | +0.060 |
iddles | +0.059 |
Neg logit contributions | |
she | -0.069 |
he | -0.061 |
they | -0.061 |
day | -0.060 |
hallway | -0.059 |
Token | the |
Token ID | 262 |
Feature activation | -5.53e-02 |
Pos logit contributions | |
ps | +0.120 |
ches | +0.110 |
ss | +0.108 |
rier | +0.108 |
ealing | +0.107 |
Neg logit contributions | |
she | -0.082 |
day | -0.071 |
he | -0.071 |
they | -0.068 |
weekend | -0.063 |
Token | word |
Token ID | 1573 |
Feature activation | -4.87e-02 |
Pos logit contributions | |
ps | +0.102 |
ches | +0.096 |
mus | +0.090 |
ealing | +0.085 |
ss | +0.085 |
Neg logit contributions | |
day | -0.088 |
stair | -0.081 |
she | -0.081 |
weekend | -0.079 |
particular | -0.078 |
Token | " |
Token ID | 366 |
Feature activation | -6.33e-02 |
Pos logit contributions | |
ps | +0.094 |
ss | +0.081 |
ches | +0.077 |
iddles | +0.077 |
wy | +0.077 |
Neg logit contributions | |
she | -0.075 |
day | -0.067 |
they | -0.064 |
he | -0.064 |
weekend | -0.064 |
Token | ign |
Token ID | 570 |
Feature activation | -9.32e-02 |
Pos logit contributions | |
ps | +0.126 |
teeth | +0.106 |
ches | +0.099 |
rier | +0.098 |
ealing | +0.097 |
Neg logit contributions | |
she | -0.092 |
day | -0.090 |
weekend | -0.088 |
days | -0.083 |
particular | -0.082 |
Token | or |
Token ID | 273 |
Feature activation | -1.21e-01 |
Pos logit contributions | |
ps | +0.209 |
teeth | +0.185 |
wounds | +0.175 |
iddles | +0.173 |
ags | +0.173 |
Neg logit contributions | |
day | -0.110 |
she | -0.103 |
weekend | -0.098 |
they | -0.089 |
he | -0.087 |
Token | ant |
Token ID | 415 |
Feature activation | -8.92e-02 |
Pos logit contributions | |
ps | +0.177 |
ksh | +0.174 |
wounds | +0.166 |
nicer | +0.162 |
unnecessary | +0.162 |
Neg logit contributions | |
she | -0.188 |
day | -0.168 |
he | -0.160 |
they | -0.157 |
where | -0.154 |
Token | " |
Token ID | 1 |
Feature activation | -8.34e-02 |
Pos logit contributions | |
ps | +0.159 |
teeth | +0.136 |
iddles | +0.134 |
wounds | +0.131 |
infected | +0.127 |
Neg logit contributions | |
she | -0.140 |
day | -0.129 |
he | -0.117 |
they | -0.116 |
weekend | -0.115 |
Token | and |
Token ID | 290 |
Feature activation | -3.38e-02 |
Pos logit contributions | |
ps | +0.174 |
ealing | +0.140 |
teeth | +0.138 |
ches | +0.137 |
uable | +0.136 |
Neg logit contributions | |
she | -0.131 |
day | -0.119 |
weekend | -0.112 |
he | -0.109 |
they | -0.105 |
Token | the |
Token ID | 262 |
Feature activation | -3.26e-02 |
Pos logit contributions | |
ps | +0.052 |
ss | +0.047 |
ches | +0.045 |
iddles | +0.044 |
mus | +0.044 |
Neg logit contributions | |
she | -0.061 |
they | -0.054 |
he | -0.054 |
day | -0.053 |
stair | -0.050 |
Token | number |
Token ID | 1271 |
Feature activation | -5.89e-02 |
Pos logit contributions | |
ps | +0.061 |
ches | +0.056 |
mus | +0.054 |
wy | +0.052 |
uable | +0.051 |
Neg logit contributions | |
she | -0.049 |
day | -0.048 |
stair | -0.047 |
hallway | -0.045 |
weekend | -0.044 |
Token | okay |
Token ID | 8788 |
Feature activation | -1.38e-02 |
Pos logit contributions | |
ps | +0.132 |
iddles | +0.115 |
aff | +0.115 |
rier | +0.115 |
ss | +0.113 |
Neg logit contributions | |
day | -0.096 |
sunny | -0.088 |
she | -0.081 |
weekend | -0.076 |
summer | -0.075 |
Token | if |
Token ID | 611 |
Feature activation | -5.80e-03 |
Pos logit contributions | |
ps | +0.026 |
ss | +0.022 |
ches | +0.021 |
iddles | +0.021 |
mus | +0.021 |
Neg logit contributions | |
she | -0.022 |
day | -0.020 |
weekend | -0.019 |
they | -0.019 |
stair | -0.018 |
Token | you |
Token ID | 345 |
Feature activation | -6.80e-02 |
Pos logit contributions | |
ps | +0.013 |
ss | +0.012 |
ches | +0.012 |
teeth | +0.011 |
mus | +0.011 |
Neg logit contributions | |
she | -0.006 |
weekend | -0.005 |
day | -0.005 |
stair | -0.005 |
hallway | -0.005 |
Token | can |
Token ID | 460 |
Feature activation | -5.98e-02 |
Pos logit contributions | |
ps | +0.155 |
iddles | +0.145 |
rier | +0.143 |
ss | +0.141 |
mus | +0.140 |
Neg logit contributions | |
day | -0.063 |
wander | -0.063 |
explore | -0.062 |
trip | -0.058 |
collect | -0.055 |
Token | 't |
Token ID | 470 |
Feature activation | -4.15e-02 |
Pos logit contributions | |
ps | +0.132 |
teeth | +0.118 |
iddles | +0.118 |
ss | +0.117 |
rier | +0.115 |
Neg logit contributions | |
explore | -0.063 |
day | -0.062 |
wander | -0.062 |
she | -0.060 |
visit | -0.057 |
Token | whistle |
Token ID | 16121 |
Feature activation | -1.18e-01 |
Pos logit contributions | |
ps | +0.088 |
ss | +0.076 |
iddles | +0.076 |
teeth | +0.074 |
wy | +0.073 |
Neg logit contributions | |
she | -0.050 |
day | -0.049 |
explore | -0.047 |
wander | -0.047 |
he | -0.045 |
Token | yet |
Token ID | 1865 |
Feature activation | -7.31e-02 |
Pos logit contributions | |
ps | +0.267 |
aned | +0.232 |
wy | +0.232 |
ented | +0.231 |
iddles | +0.229 |
Neg logit contributions | |
she | -0.152 |
day | -0.142 |
he | -0.142 |
they | -0.138 |
today | -0.131 |
Token | . |
Token ID | 13 |
Feature activation | -1.22e-02 |
Pos logit contributions | |
ps | +0.149 |
iddles | +0.130 |
ss | +0.129 |
wy | +0.127 |
mus | +0.117 |
Neg logit contributions | |
she | -0.117 |
they | -0.106 |
day | -0.102 |
he | -0.101 |
we | -0.094 |
Token | It |
Token ID | 632 |
Feature activation | -4.44e-02 |
Pos logit contributions | |
ps | +0.027 |
ss | +0.023 |
ches | +0.023 |
teeth | +0.023 |
mus | +0.022 |
Neg logit contributions | |
she | -0.013 |
weekend | -0.012 |
day | -0.012 |
stair | -0.012 |
hallway | -0.012 |
Token | 's |
Token ID | 338 |
Feature activation | -5.85e-02 |
Pos logit contributions | |
ps | +0.086 |
iddles | +0.077 |
mus | +0.076 |
ss | +0.073 |
wy | +0.071 |
Neg logit contributions | |
she | -0.061 |
day | -0.056 |
they | -0.054 |
he | -0.051 |
wander | -0.050 |
Token | interesting |
Token ID | 3499 |
Feature activation | -5.78e-02 |
Pos logit contributions | |
ps | +0.110 |
ss | +0.096 |
iddles | +0.095 |
aff | +0.094 |
rier | +0.094 |
Neg logit contributions | |
day | -0.096 |
sunny | -0.086 |
she | -0.085 |
weekend | -0.079 |
summer | -0.076 |
Token | you |
Token ID | 345 |
Feature activation | -4.36e-02 |
Pos logit contributions | |
hallway | +0.038 |
weekend | +0.035 |
stair | +0.034 |
driveway | +0.033 |
holiday | +0.032 |
Neg logit contributions | |
ps | -0.056 |
ss | -0.055 |
ches | -0.052 |
teeth | -0.051 |
mus | -0.051 |
Token | teach |
Token ID | 4545 |
Feature activation | -5.99e-02 |
Pos logit contributions | |
ps | +0.098 |
ss | +0.086 |
iddles | +0.085 |
teeth | +0.084 |
wy | +0.083 |
Neg logit contributions | |
she | -0.049 |
explore | -0.047 |
day | -0.046 |
he | -0.045 |
wander | -0.044 |
Token | me |
Token ID | 502 |
Feature activation | -4.85e-02 |
Pos logit contributions | |
ps | +0.117 |
rier | +0.108 |
wy | +0.104 |
ealing | +0.101 |
ows | +0.099 |
Neg logit contributions | |
she | -0.092 |
day | -0.085 |
explore | -0.078 |
they | -0.075 |
visit | -0.075 |
Token | how |
Token ID | 703 |
Feature activation | -3.52e-02 |
Pos logit contributions | |
ps | +0.098 |
ss | +0.087 |
wy | +0.082 |
mus | +0.081 |
rier | +0.081 |
Neg logit contributions | |
she | -0.062 |
day | -0.062 |
he | -0.057 |
of | -0.057 |
weekend | -0.057 |
Token | to |
Token ID | 284 |
Feature activation | -4.35e-02 |
Pos logit contributions | |
ps | +0.088 |
ss | +0.075 |
iddles | +0.074 |
wy | +0.073 |
ches | +0.073 |
Neg logit contributions | |
she | -0.036 |
day | -0.036 |
weekend | -0.032 |
they | -0.031 |
he | -0.031 |
Token | whistle |
Token ID | 16121 |
Feature activation | -1.18e-01 |
Pos logit contributions | |
ps | +0.092 |
ss | +0.085 |
teeth | +0.082 |
iddles | +0.082 |
ussy | +0.080 |
Neg logit contributions | |
day | -0.050 |
she | -0.049 |
explore | -0.048 |
wander | -0.047 |
he | -0.044 |
Token | too |
Token ID | 1165 |
Feature activation | -6.74e-02 |
Pos logit contributions | |
ps | +0.283 |
ented | +0.246 |
ussy | +0.245 |
wy | +0.244 |
aned | +0.242 |
Neg logit contributions | |
day | -0.149 |
he | -0.127 |
sunny | -0.123 |
today | -0.120 |
she | -0.120 |
Token | ?" |
Token ID | 1701 |
Feature activation | -1.18e-02 |
Pos logit contributions | |
ps | +0.167 |
wounds | +0.136 |
iddles | +0.135 |
wy | +0.134 |
ss | +0.130 |
Neg logit contributions | |
day | -0.084 |
of | -0.073 |
weekend | -0.072 |
sunny | -0.072 |
he | -0.070 |
Token | |
Token ID | 220 |
Feature activation | -3.88e-02 |
Pos logit contributions | |
ps | +0.026 |
ss | +0.022 |
ches | +0.022 |
teeth | +0.021 |
mus | +0.021 |
Neg logit contributions | |
she | -0.015 |
day | -0.013 |
weekend | -0.013 |
they | -0.012 |
stair | -0.012 |
Token |
|
Token ID | 198 |
Feature activation | -4.41e-02 |
Pos logit contributions | |
ps | +0.094 |
ss | +0.079 |
teeth | +0.079 |
mus | +0.078 |
ches | +0.078 |
Neg logit contributions | |
she | -0.041 |
day | -0.034 |
they | -0.033 |
weekend | -0.033 |
he | -0.031 |
Token |
|
Token ID | 198 |
Feature activation | -1.46e-02 |
Pos logit contributions | |
ps | +0.125 |
teeth | +0.106 |
rier | +0.103 |
wy | +0.103 |
ss | +0.101 |
Neg logit contributions | |
she | -0.042 |
they | -0.033 |
day | -0.031 |
he | -0.030 |
weekend | -0.027 |
Token | " |
Token ID | 366 |
Feature activation | +1.72e-02 |
Pos logit contributions | |
ps | +0.084 |
wy | +0.075 |
ches | +0.074 |
ws | +0.072 |
unnecessary | +0.072 |
Neg logit contributions | |
she | -0.057 |
they | -0.052 |
he | -0.051 |
day | -0.045 |
weekend | -0.042 |
Token | But |
Token ID | 1537 |
Feature activation | +9.47e-03 |
Pos logit contributions | |
weekend | +0.020 |
stair | +0.020 |
hallway | +0.019 |
holiday | +0.019 |
she | +0.019 |
Neg logit contributions | |
ps | -0.034 |
ss | -0.033 |
mus | -0.030 |
ches | -0.029 |
teeth | -0.029 |
Token | I |
Token ID | 314 |
Feature activation | -1.19e-02 |
Pos logit contributions | |
weekend | +0.010 |
hallway | +0.010 |
she | +0.009 |
stair | +0.009 |
day | +0.009 |
Neg logit contributions | |
ps | -0.021 |
ss | -0.019 |
teeth | -0.018 |
ches | -0.018 |
mus | -0.017 |
Token | like |
Token ID | 588 |
Feature activation | -4.04e-02 |
Pos logit contributions | |
ps | +0.026 |
ss | +0.022 |
mus | +0.021 |
iddles | +0.021 |
ches | +0.021 |
Neg logit contributions | |
she | -0.014 |
day | -0.014 |
weekend | -0.013 |
stair | -0.012 |
hallway | -0.012 |
Token | to |
Token ID | 284 |
Feature activation | -2.05e-02 |
Pos logit contributions | |
ps | +0.081 |
ows | +0.067 |
ealing | +0.066 |
rier | +0.066 |
iddles | +0.065 |
Neg logit contributions | |
she | -0.062 |
he | -0.057 |
day | -0.057 |
they | -0.054 |
weekend | -0.050 |
Token | chew |
Token ID | 34722 |
Feature activation | -1.18e-01 |
Pos logit contributions | |
ps | +0.047 |
ss | +0.041 |
iddles | +0.041 |
ches | +0.040 |
mus | +0.040 |
Neg logit contributions | |
she | -0.021 |
day | -0.020 |
explore | -0.019 |
wander | -0.018 |
weekend | -0.018 |
Token | gum |
Token ID | 27142 |
Feature activation | -6.69e-02 |
Pos logit contributions | |
ps | +0.286 |
iddles | +0.256 |
ows | +0.252 |
amus | +0.250 |
aned | +0.241 |
Neg logit contributions | |
he | -0.131 |
she | -0.126 |
day | -0.108 |
they | -0.107 |
today | -0.091 |
Token | , |
Token ID | 11 |
Feature activation | +2.01e-02 |
Pos logit contributions | |
ps | +0.124 |
ss | +0.108 |
iddles | +0.108 |
unnecessary | +0.106 |
ealing | +0.104 |
Neg logit contributions | |
she | -0.105 |
day | -0.099 |
he | -0.095 |
they | -0.093 |
today | -0.088 |
Token | it |
Token ID | 340 |
Feature activation | -3.93e-02 |
Pos logit contributions | |
hallway | +0.026 |
weekend | +0.026 |
days | +0.024 |
she | +0.024 |
day | +0.024 |
Neg logit contributions | |
ps | -0.037 |
ss | -0.034 |
teeth | -0.033 |
ches | -0.032 |
mus | -0.031 |
Token | 's |
Token ID | 338 |
Feature activation | -5.79e-02 |
Pos logit contributions | |
ps | +0.072 |
iddles | +0.065 |
mus | +0.062 |
rier | +0.059 |
ss | +0.059 |
Neg logit contributions | |
she | -0.058 |
day | -0.054 |
he | -0.052 |
they | -0.049 |
stair | -0.048 |
Token | so |
Token ID | 523 |
Feature activation | -4.55e-02 |
Pos logit contributions | |
ps | +0.109 |
iddles | +0.099 |
rier | +0.096 |
rots | +0.096 |
ino | +0.093 |
Neg logit contributions | |
day | -0.095 |
sunny | -0.089 |
she | -0.083 |
summer | -0.078 |
he | -0.077 |
Token | pieces |
Token ID | 5207 |
Feature activation | +1.14e-02 |
Pos logit contributions | |
ps | +0.039 |
ss | +0.034 |
mus | +0.033 |
ches | +0.031 |
iddles | +0.031 |
Neg logit contributions | |
day | -0.035 |
she | -0.035 |
weekend | -0.034 |
stair | -0.034 |
hallway | -0.033 |
Token | to |
Token ID | 284 |
Feature activation | -2.56e-02 |
Pos logit contributions | |
hallway | +0.015 |
weekend | +0.015 |
she | +0.015 |
stair | +0.015 |
day | +0.014 |
Neg logit contributions | |
ps | -0.022 |
ss | -0.019 |
ches | -0.018 |
teeth | -0.018 |
mus | -0.018 |
Token | try |
Token ID | 1949 |
Feature activation | -8.56e-02 |
Pos logit contributions | |
ps | +0.055 |
ss | +0.050 |
teeth | +0.048 |
ches | +0.047 |
iddles | +0.047 |
Neg logit contributions | |
she | -0.029 |
explore | -0.026 |
day | -0.026 |
particular | -0.025 |
hallway | -0.025 |
Token | and |
Token ID | 290 |
Feature activation | +1.79e-02 |
Pos logit contributions | |
ps | +0.183 |
wy | +0.163 |
ss | +0.157 |
ealing | +0.156 |
aned | +0.153 |
Neg logit contributions | |
she | -0.112 |
they | -0.106 |
day | -0.095 |
he | -0.092 |
stair | -0.090 |
Token | check |
Token ID | 2198 |
Feature activation | -7.09e-02 |
Pos logit contributions | |
weekend | +0.025 |
hallway | +0.024 |
stair | +0.024 |
she | +0.024 |
day | +0.024 |
Neg logit contributions | |
ps | -0.033 |
ss | -0.028 |
teeth | -0.028 |
ches | -0.027 |
mus | -0.027 |
Token | mate |
Token ID | 9830 |
Feature activation | -1.16e-01 |
Pos logit contributions | |
ps | +0.141 |
teeth | +0.124 |
ss | +0.124 |
ache | +0.122 |
ksh | +0.119 |
Neg logit contributions | |
she | -0.096 |
they | -0.092 |
day | -0.083 |
he | -0.079 |
weekend | -0.078 |
Token | your |
Token ID | 534 |
Feature activation | -5.29e-02 |
Pos logit contributions | |
ps | +0.221 |
infected | +0.192 |
obb | +0.189 |
aned | +0.188 |
wy | +0.188 |
Neg logit contributions | |
she | -0.167 |
they | -0.163 |
he | -0.145 |
we | -0.133 |
day | -0.131 |
Token | opponent |
Token ID | 6125 |
Feature activation | -6.33e-02 |
Pos logit contributions | |
ps | +0.096 |
mus | +0.087 |
wy | +0.084 |
ches | +0.084 |
ache | +0.082 |
Neg logit contributions | |
day | -0.081 |
stair | -0.072 |
holiday | -0.070 |
trip | -0.069 |
hill | -0.069 |
Token | ." |
Token ID | 526 |
Feature activation | +1.42e-02 |
Pos logit contributions | |
ps | +0.102 |
ss | +0.093 |
iddles | +0.086 |
mus | +0.086 |
teeth | +0.082 |
Neg logit contributions | |
she | -0.103 |
day | -0.098 |
they | -0.094 |
weekend | -0.093 |
he | -0.091 |
Token |
|
Token ID | 198 |
Feature activation | -8.00e-03 |
Pos logit contributions | |
hallway | +0.021 |
weekend | +0.021 |
stair | +0.021 |
day | +0.020 |
she | +0.020 |
Neg logit contributions | |
ps | -0.024 |
ss | -0.022 |
teeth | -0.021 |
ches | -0.019 |
mus | -0.019 |
Token |
|
Token ID | 198 |
Feature activation | +1.47e-02 |
Pos logit contributions | |
ps | +0.017 |
ss | +0.014 |
ches | +0.014 |
teeth | +0.014 |
mus | +0.014 |
Neg logit contributions | |
she | -0.010 |
weekend | -0.009 |
day | -0.009 |
stair | -0.009 |
hallway | -0.009 |
Token | that |
Token ID | 326 |
Feature activation | -2.65e-02 |
Pos logit contributions | |
ps | +0.008 |
ss | +0.007 |
ches | +0.007 |
teeth | +0.007 |
mus | +0.007 |
Neg logit contributions | |
she | -0.005 |
weekend | -0.005 |
day | -0.005 |
stair | -0.005 |
hallway | -0.005 |
Token | said |
Token ID | 531 |
Feature activation | -5.05e-02 |
Pos logit contributions | |
ps | +0.054 |
ss | +0.047 |
ches | +0.046 |
mus | +0.045 |
ealing | +0.045 |
Neg logit contributions | |
she | -0.037 |
day | -0.034 |
weekend | -0.032 |
they | -0.032 |
stair | -0.032 |
Token | " |
Token ID | 366 |
Feature activation | -4.88e-02 |
Pos logit contributions | |
ps | +0.104 |
wy | +0.098 |
ches | +0.095 |
ksh | +0.090 |
unnecessary | +0.089 |
Neg logit contributions | |
she | -0.076 |
they | -0.065 |
he | -0.061 |
stair | -0.056 |
day | -0.056 |
Token | Do |
Token ID | 5211 |
Feature activation | -9.56e-03 |
Pos logit contributions | |
ps | +0.101 |
teeth | +0.086 |
ealing | +0.084 |
ches | +0.083 |
rier | +0.081 |
Neg logit contributions | |
she | -0.065 |
day | -0.063 |
weekend | -0.062 |
particular | -0.058 |
days | -0.058 |
Token | not |
Token ID | 407 |
Feature activation | -1.82e-02 |
Pos logit contributions | |
ps | +0.017 |
ss | +0.014 |
ches | +0.014 |
mus | +0.014 |
teeth | +0.014 |
Neg logit contributions | |
she | -0.015 |
day | -0.014 |
weekend | -0.014 |
stair | -0.014 |
they | -0.013 |
Token | write |
Token ID | 3551 |
Feature activation | -1.16e-01 |
Pos logit contributions | |
ps | +0.039 |
ches | +0.034 |
ss | +0.033 |
iddles | +0.032 |
mus | +0.032 |
Neg logit contributions | |
she | -0.022 |
day | -0.022 |
weekend | -0.020 |
stair | -0.020 |
wander | -0.019 |
Token | on |
Token ID | 319 |
Feature activation | -7.01e-02 |
Pos logit contributions | |
ps | +0.232 |
ksh | +0.216 |
wounds | +0.209 |
unnecessary | +0.208 |
wy | +0.204 |
Neg logit contributions | |
she | -0.145 |
today | -0.121 |
day | -0.119 |
they | -0.115 |
of | -0.107 |
Token | the |
Token ID | 262 |
Feature activation | -4.99e-02 |
Pos logit contributions | |
ps | +0.160 |
ss | +0.139 |
iddles | +0.138 |
ealing | +0.137 |
wy | +0.135 |
Neg logit contributions | |
she | -0.079 |
stair | -0.069 |
day | -0.068 |
they | -0.064 |
holiday | -0.062 |
Token | rail |
Token ID | 6787 |
Feature activation | -3.54e-02 |
Pos logit contributions | |
ps | +0.126 |
iddles | +0.120 |
wy | +0.119 |
ussy | +0.117 |
ealing | +0.116 |
Neg logit contributions | |
day | -0.043 |
stair | -0.040 |
hallway | -0.039 |
hill | -0.036 |
weekend | -0.035 |
Token | ". |
Token ID | 1911 |
Feature activation | -1.69e-02 |
Pos logit contributions | |
ps | +0.052 |
teeth | +0.043 |
ss | +0.042 |
iddles | +0.040 |
mus | +0.040 |
Neg logit contributions | |
she | -0.065 |
day | -0.063 |
weekend | -0.062 |
stair | -0.062 |
hallway | -0.059 |
Token | She |
Token ID | 1375 |
Feature activation | -4.31e-02 |
Pos logit contributions | |
ps | +0.040 |
ches | +0.035 |
ss | +0.035 |
wy | +0.034 |
mus | +0.034 |
Neg logit contributions | |
she | -0.018 |
day | -0.015 |
stair | -0.015 |
weekend | -0.015 |
hallway | -0.015 |
Token | bird |
Token ID | 6512 |
Feature activation | -1.23e-02 |
Pos logit contributions | |
ps | +0.027 |
ches | +0.024 |
ss | +0.023 |
mus | +0.023 |
teeth | +0.023 |
Neg logit contributions | |
she | -0.019 |
day | -0.019 |
weekend | -0.018 |
stair | -0.017 |
hallway | -0.017 |
Token | said |
Token ID | 531 |
Feature activation | -1.23e-02 |
Pos logit contributions | |
ps | +0.027 |
ss | +0.024 |
teeth | +0.023 |
iddles | +0.023 |
mus | +0.022 |
Neg logit contributions | |
she | -0.014 |
day | -0.013 |
hallway | -0.012 |
weekend | -0.012 |
stair | -0.012 |
Token | " |
Token ID | 366 |
Feature activation | -5.79e-03 |
Pos logit contributions | |
ps | +0.027 |
ss | +0.023 |
ches | +0.023 |
iddles | +0.023 |
wy | +0.022 |
Neg logit contributions | |
she | -0.015 |
day | -0.013 |
they | -0.013 |
he | -0.013 |
weekend | -0.013 |
Token | t |
Token ID | 83 |
Feature activation | +2.63e-02 |
Pos logit contributions | |
ps | +0.013 |
ss | +0.011 |
teeth | +0.011 |
ches | +0.011 |
mus | +0.010 |
Neg logit contributions | |
she | -0.007 |
weekend | -0.006 |
day | -0.006 |
hallway | -0.006 |
stair | -0.006 |
Token | weet |
Token ID | 7277 |
Feature activation | -4.48e-02 |
Pos logit contributions | |
weekend | +0.042 |
hallway | +0.041 |
she | +0.040 |
driveway | +0.039 |
particular | +0.039 |
Neg logit contributions | |
ps | -0.040 |
teeth | -0.036 |
ss | -0.036 |
ches | -0.033 |
mus | -0.032 |
Token | tweet |
Token ID | 6126 |
Feature activation | -1.16e-01 |
Pos logit contributions | |
ps | +0.082 |
iddles | +0.076 |
wy | +0.075 |
teeth | +0.075 |
mus | +0.074 |
Neg logit contributions | |
she | -0.060 |
day | -0.055 |
they | -0.051 |
stair | -0.049 |
wander | -0.048 |
Token | " |
Token ID | 1 |
Feature activation | -6.75e-02 |
Pos logit contributions | |
ps | +0.285 |
wounds | +0.281 |
ags | +0.259 |
ws | +0.255 |
wy | +0.252 |
Neg logit contributions | |
she | -0.114 |
of | -0.112 |
day | -0.107 |
they | -0.094 |
where | -0.094 |
Token | and |
Token ID | 290 |
Feature activation | -2.92e-02 |
Pos logit contributions | |
ps | +0.150 |
ealing | +0.124 |
ws | +0.123 |
wy | +0.123 |
mus | +0.122 |
Neg logit contributions | |
she | -0.092 |
day | -0.086 |
he | -0.081 |
weekend | -0.077 |
they | -0.072 |
Token | flew |
Token ID | 13112 |
Feature activation | -2.73e-02 |
Pos logit contributions | |
ps | +0.059 |
ss | +0.054 |
ches | +0.053 |
iddles | +0.052 |
wy | +0.051 |
Neg logit contributions | |
she | -0.041 |
they | -0.036 |
he | -0.035 |
day | -0.033 |
stair | -0.031 |
Token | away |
Token ID | 1497 |
Feature activation | -1.68e-02 |
Pos logit contributions | |
ps | +0.044 |
ss | +0.038 |
iddles | +0.037 |
wy | +0.037 |
uable | +0.037 |
Neg logit contributions | |
she | -0.047 |
day | -0.044 |
he | -0.042 |
they | -0.041 |
weekend | -0.040 |
Token | . |
Token ID | 13 |
Feature activation | -3.65e-02 |
Pos logit contributions | |
ps | +0.030 |
ss | +0.028 |
ches | +0.027 |
wy | +0.026 |
ealing | +0.025 |
Neg logit contributions | |
she | -0.026 |
day | -0.024 |
stair | -0.023 |
he | -0.023 |
hallway | -0.023 |
Token | songs |
Token ID | 7259 |
Feature activation | -4.37e-02 |
Pos logit contributions | |
ps | +0.188 |
wy | +0.178 |
ss | +0.175 |
aned | +0.166 |
ussy | +0.164 |
Neg logit contributions | |
she | -0.099 |
they | -0.085 |
day | -0.084 |
sunny | -0.077 |
today | -0.075 |
Token | you |
Token ID | 345 |
Feature activation | -1.11e-01 |
Pos logit contributions | |
ps | +0.091 |
ss | +0.085 |
teeth | +0.081 |
iddles | +0.079 |
ealing | +0.077 |
Neg logit contributions | |
she | -0.059 |
day | -0.052 |
they | -0.050 |
weekend | -0.048 |
he | -0.046 |
Token | know |
Token ID | 760 |
Feature activation | -5.82e-02 |
Pos logit contributions | |
ps | +0.245 |
rier | +0.237 |
ss | +0.231 |
iddles | +0.231 |
infected | +0.228 |
Neg logit contributions | |
day | -0.121 |
she | -0.118 |
they | -0.102 |
visit | -0.100 |
explore | -0.099 |
Token | and |
Token ID | 290 |
Feature activation | -4.04e-03 |
Pos logit contributions | |
ps | +0.114 |
ss | +0.104 |
ealing | +0.098 |
wy | +0.097 |
iddles | +0.096 |
Neg logit contributions | |
she | -0.087 |
they | -0.078 |
day | -0.077 |
he | -0.073 |
holiday | -0.071 |
Token | songs |
Token ID | 7259 |
Feature activation | -5.87e-02 |
Pos logit contributions | |
ps | +0.007 |
ss | +0.007 |
ches | +0.006 |
teeth | +0.006 |
mus | +0.006 |
Neg logit contributions | |
she | -0.006 |
weekend | -0.006 |
day | -0.006 |
hallway | -0.005 |
stair | -0.005 |
Token | you |
Token ID | 345 |
Feature activation | -1.15e-01 |
Pos logit contributions | |
ps | +0.109 |
ss | +0.100 |
teeth | +0.097 |
iddles | +0.094 |
wy | +0.093 |
Neg logit contributions | |
she | -0.093 |
day | -0.083 |
they | -0.082 |
he | -0.076 |
weekend | -0.075 |
Token | make |
Token ID | 787 |
Feature activation | -6.70e-02 |
Pos logit contributions | |
rier | +0.287 |
ss | +0.269 |
ps | +0.268 |
infected | +0.263 |
iddles | +0.263 |
Neg logit contributions | |
day | -0.094 |
explore | -0.087 |
visit | -0.082 |
have | -0.081 |
she | -0.078 |
Token | up |
Token ID | 510 |
Feature activation | -4.53e-02 |
Pos logit contributions | |
ps | +0.131 |
wy | +0.126 |
ss | +0.122 |
uable | +0.116 |
ache | +0.115 |
Neg logit contributions | |
she | -0.093 |
day | -0.088 |
they | -0.080 |
holiday | -0.079 |
he | -0.075 |
Token | ." |
Token ID | 526 |
Feature activation | +1.86e-02 |
Pos logit contributions | |
ps | +0.079 |
ss | +0.074 |
ches | +0.068 |
wy | +0.068 |
uable | +0.066 |
Neg logit contributions | |
she | -0.073 |
day | -0.072 |
weekend | -0.067 |
sunny | -0.066 |
they | -0.065 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -1.68e-02 |
Pos logit contributions | |
hallway | +0.031 |
weekend | +0.031 |
stair | +0.031 |
holiday | +0.029 |
exploration | +0.029 |
Neg logit contributions | |
ps | -0.027 |
ss | -0.025 |
teeth | -0.024 |
mus | -0.022 |
ches | -0.022 |
Token | L |
Token ID | 43 |
Feature activation | -6.52e-04 |
Pos logit contributions | |
ps | +0.033 |
ss | +0.028 |
teeth | +0.027 |
ches | +0.027 |
mus | +0.027 |
Neg logit contributions | |
she | -0.023 |
day | -0.022 |
weekend | -0.022 |
hallway | -0.021 |
stair | -0.021 |
Token | like |
Token ID | 588 |
Feature activation | -2.52e-02 |
Pos logit contributions | |
ps | +0.026 |
ss | +0.023 |
ches | +0.022 |
mus | +0.021 |
teeth | +0.021 |
Neg logit contributions | |
she | -0.024 |
day | -0.023 |
they | -0.022 |
weekend | -0.021 |
he | -0.021 |
Token | " |
Token ID | 366 |
Feature activation | -2.44e-02 |
Pos logit contributions | |
ps | +0.045 |
ss | +0.040 |
ches | +0.040 |
wy | +0.038 |
mus | +0.037 |
Neg logit contributions | |
she | -0.040 |
day | -0.037 |
they | -0.036 |
weekend | -0.036 |
holiday | -0.034 |
Token | t |
Token ID | 83 |
Feature activation | -1.04e-02 |
Pos logit contributions | |
ps | +0.047 |
teeth | +0.039 |
ches | +0.039 |
ss | +0.039 |
iddles | +0.038 |
Neg logit contributions | |
she | -0.034 |
day | -0.034 |
weekend | -0.033 |
days | -0.031 |
particular | -0.031 |
Token | weet |
Token ID | 7277 |
Feature activation | -6.93e-02 |
Pos logit contributions | |
ps | +0.021 |
ss | +0.018 |
teeth | +0.018 |
ches | +0.017 |
iddles | +0.017 |
Neg logit contributions | |
she | -0.014 |
day | -0.013 |
weekend | -0.013 |
stair | -0.013 |
hallway | -0.012 |
Token | , |
Token ID | 11 |
Feature activation | -1.22e-02 |
Pos logit contributions | |
ussy | +0.119 |
ps | +0.118 |
iddles | +0.117 |
wy | +0.113 |
teeth | +0.110 |
Neg logit contributions | |
she | -0.095 |
day | -0.093 |
they | -0.083 |
where | -0.082 |
wander | -0.081 |
Token | tweet |
Token ID | 6126 |
Feature activation | -1.15e-01 |
Pos logit contributions | |
ps | +0.024 |
ss | +0.021 |
ches | +0.021 |
iddles | +0.021 |
teeth | +0.020 |
Neg logit contributions | |
she | -0.016 |
day | -0.015 |
weekend | -0.015 |
stair | -0.015 |
they | -0.014 |
Token | , |
Token ID | 11 |
Feature activation | -1.36e-02 |
Pos logit contributions | |
ps | +0.277 |
wounds | +0.252 |
wy | +0.242 |
asy | +0.242 |
iddles | +0.241 |
Neg logit contributions | |
day | -0.128 |
of | -0.123 |
she | -0.116 |
where | -0.113 |
they | -0.104 |
Token | tweet |
Token ID | 6126 |
Feature activation | -1.13e-01 |
Pos logit contributions | |
ps | +0.024 |
ss | +0.020 |
iddles | +0.020 |
ches | +0.020 |
teeth | +0.020 |
Neg logit contributions | |
she | -0.021 |
day | -0.020 |
weekend | -0.020 |
stair | -0.019 |
hallway | -0.019 |
Token | ". |
Token ID | 1911 |
Feature activation | -1.27e-02 |
Pos logit contributions | |
ps | +0.273 |
wounds | +0.246 |
asy | +0.242 |
wy | +0.241 |
ussy | +0.236 |
Neg logit contributions | |
day | -0.122 |
of | -0.117 |
she | -0.113 |
where | -0.109 |
they | -0.101 |
Token | Tim |
Token ID | 5045 |
Feature activation | -5.47e-02 |
Pos logit contributions | |
ps | +0.028 |
ss | +0.024 |
ches | +0.024 |
teeth | +0.023 |
mus | +0.023 |
Neg logit contributions | |
she | -0.016 |
day | -0.014 |
they | -0.014 |
weekend | -0.014 |
stair | -0.013 |
Token | and |
Token ID | 290 |
Feature activation | -2.15e-03 |
Pos logit contributions | |
rier | +0.123 |
mus | +0.120 |
unnecessary | +0.120 |
ps | +0.119 |
uable | +0.116 |
Neg logit contributions | |
she | -0.045 |
sandbox | -0.040 |
day | -0.038 |
hill | -0.038 |
slide | -0.038 |
Token | book |
Token ID | 1492 |
Feature activation | -4.62e-02 |
Pos logit contributions | |
ps | +0.033 |
ss | +0.028 |
iddles | +0.027 |
teeth | +0.027 |
mus | +0.026 |
Neg logit contributions | |
she | -0.024 |
day | -0.024 |
stair | -0.023 |
weekend | -0.023 |
hallway | -0.023 |
Token | . |
Token ID | 13 |
Feature activation | -2.30e-02 |
Pos logit contributions | |
ps | +0.091 |
ss | +0.079 |
teeth | +0.076 |
ealing | +0.075 |
iddles | +0.075 |
Neg logit contributions | |
she | -0.069 |
day | -0.063 |
weekend | -0.062 |
stair | -0.060 |
they | -0.059 |
Token | But |
Token ID | 887 |
Feature activation | +1.75e-03 |
Pos logit contributions | |
ps | +0.047 |
ches | +0.041 |
ss | +0.040 |
teeth | +0.039 |
mus | +0.038 |
Neg logit contributions | |
she | -0.029 |
weekend | -0.026 |
day | -0.026 |
stair | -0.025 |
hallway | -0.025 |
Token | you |
Token ID | 345 |
Feature activation | -2.68e-02 |
Pos logit contributions | |
she | +0.002 |
weekend | +0.002 |
hallway | +0.002 |
day | +0.002 |
stair | +0.002 |
Neg logit contributions | |
ps | -0.004 |
ss | -0.003 |
teeth | -0.003 |
ches | -0.003 |
mus | -0.003 |
Token | cannot |
Token ID | 2314 |
Feature activation | -1.09e-02 |
Pos logit contributions | |
ps | +0.069 |
ss | +0.063 |
rier | +0.063 |
iddles | +0.061 |
mus | +0.061 |
Neg logit contributions | |
she | -0.018 |
day | -0.018 |
explore | -0.015 |
weekend | -0.015 |
they | -0.015 |
Token | write |
Token ID | 3551 |
Feature activation | -1.15e-01 |
Pos logit contributions | |
ps | +0.022 |
ss | +0.020 |
ches | +0.019 |
teeth | +0.019 |
mus | +0.019 |
Neg logit contributions | |
she | -0.013 |
day | -0.012 |
weekend | -0.012 |
hallway | -0.012 |
stair | -0.012 |
Token | on |
Token ID | 319 |
Feature activation | -7.84e-02 |
Pos logit contributions | |
ksh | +0.245 |
ps | +0.231 |
aned | +0.226 |
ussy | +0.221 |
igg | +0.218 |
Neg logit contributions | |
she | -0.138 |
they | -0.114 |
he | -0.113 |
today | -0.107 |
day | -0.100 |
Token | things |
Token ID | 1243 |
Feature activation | -4.21e-02 |
Pos logit contributions | |
ps | +0.178 |
ealing | +0.160 |
ss | +0.158 |
aned | +0.158 |
ussy | +0.157 |
Neg logit contributions | |
she | -0.087 |
day | -0.071 |
they | -0.070 |
he | -0.067 |
stair | -0.067 |
Token | that |
Token ID | 326 |
Feature activation | -1.37e-02 |
Pos logit contributions | |
ps | +0.092 |
ss | +0.083 |
teeth | +0.081 |
iddles | +0.080 |
mus | +0.078 |
Neg logit contributions | |
she | -0.056 |
they | -0.047 |
day | -0.045 |
he | -0.044 |
stair | -0.041 |
Token | are |
Token ID | 389 |
Feature activation | -5.97e-02 |
Pos logit contributions | |
ps | +0.028 |
ss | +0.025 |
ches | +0.024 |
mus | +0.024 |
iddles | +0.024 |
Neg logit contributions | |
she | -0.017 |
day | -0.016 |
weekend | -0.016 |
stair | -0.015 |
hallway | -0.015 |
Token | not |
Token ID | 407 |
Feature activation | -5.46e-02 |
Pos logit contributions | |
ps | +0.112 |
ealing | +0.108 |
ss | +0.106 |
iddles | +0.104 |
mus | +0.103 |
Neg logit contributions | |
she | -0.083 |
day | -0.078 |
they | -0.077 |
he | -0.071 |
sunny | -0.070 |
Token | and |
Token ID | 290 |
Feature activation | +6.76e-04 |
Pos logit contributions | |
ps | +0.033 |
ss | +0.030 |
teeth | +0.028 |
mus | +0.028 |
iddles | +0.028 |
Neg logit contributions | |
she | -0.025 |
day | -0.023 |
weekend | -0.022 |
they | -0.021 |
hallway | -0.021 |
Token | says |
Token ID | 1139 |
Feature activation | -3.78e-02 |
Pos logit contributions | |
she | +0.001 |
weekend | +0.001 |
day | +0.001 |
hallway | +0.001 |
stair | +0.001 |
Neg logit contributions | |
ps | -0.001 |
ss | -0.001 |
ches | -0.001 |
teeth | -0.001 |
mus | -0.001 |
Token | " |
Token ID | 366 |
Feature activation | -1.42e-02 |
Pos logit contributions | |
ps | +0.089 |
wy | +0.080 |
ches | +0.079 |
ss | +0.077 |
iddles | +0.075 |
Neg logit contributions | |
she | -0.047 |
they | -0.040 |
he | -0.037 |
day | -0.034 |
stair | -0.033 |
Token | t |
Token ID | 83 |
Feature activation | +3.17e-03 |
Pos logit contributions | |
ps | +0.028 |
ss | +0.024 |
teeth | +0.024 |
ches | +0.023 |
iddles | +0.023 |
Neg logit contributions | |
she | -0.019 |
weekend | -0.018 |
day | -0.018 |
hallway | -0.017 |
stair | -0.017 |
Token | weet |
Token ID | 7277 |
Feature activation | -5.70e-02 |
Pos logit contributions | |
she | +0.005 |
weekend | +0.005 |
hallway | +0.004 |
day | +0.004 |
stair | +0.004 |
Neg logit contributions | |
ps | -0.006 |
ss | -0.005 |
ches | -0.005 |
teeth | -0.005 |
mus | -0.005 |
Token | tweet |
Token ID | 6126 |
Feature activation | -1.15e-01 |
Pos logit contributions | |
ps | +0.095 |
ussy | +0.089 |
wy | +0.088 |
infected | +0.088 |
iddles | +0.087 |
Neg logit contributions | |
she | -0.079 |
day | -0.075 |
they | -0.072 |
wander | -0.068 |
explore | -0.067 |
Token | ". |
Token ID | 1911 |
Feature activation | -2.07e-02 |
Pos logit contributions | |
ps | +0.247 |
wounds | +0.245 |
wy | +0.219 |
ags | +0.218 |
infected | +0.213 |
Neg logit contributions | |
she | -0.132 |
of | -0.131 |
day | -0.130 |
they | -0.128 |
where | -0.117 |
Token | Ben |
Token ID | 3932 |
Feature activation | -5.62e-02 |
Pos logit contributions | |
ps | +0.045 |
ss | +0.040 |
ches | +0.039 |
wy | +0.038 |
mus | +0.038 |
Neg logit contributions | |
she | -0.027 |
they | -0.023 |
weekend | -0.022 |
day | -0.022 |
stair | -0.022 |
Token | is |
Token ID | 318 |
Feature activation | -4.21e-02 |
Pos logit contributions | |
rier | +0.096 |
ps | +0.096 |
ss | +0.094 |
mus | +0.094 |
iddles | +0.092 |
Neg logit contributions | |
she | -0.078 |
they | -0.070 |
wander | -0.066 |
day | -0.066 |
he | -0.065 |
Token | a |
Token ID | 257 |
Feature activation | -1.53e-02 |
Pos logit contributions | |
ss | +0.068 |
ps | +0.068 |
ches | +0.065 |
uable | +0.065 |
mus | +0.063 |
Neg logit contributions | |
she | -0.071 |
day | -0.061 |
sunny | -0.061 |
they | -0.060 |
he | -0.060 |
Token | lion |
Token ID | 18744 |
Feature activation | +7.68e-03 |
Pos logit contributions | |
ps | +0.025 |
ss | +0.021 |
ches | +0.020 |
iddles | +0.019 |
teeth | +0.019 |
Neg logit contributions | |
she | -0.027 |
day | -0.026 |
weekend | -0.026 |
hallway | -0.025 |
stair | -0.025 |
Token | should |
Token ID | 815 |
Feature activation | +3.20e-02 |
Pos logit contributions | |
ps | +0.033 |
ss | +0.030 |
mus | +0.029 |
iddles | +0.029 |
teeth | +0.029 |
Neg logit contributions | |
she | -0.017 |
day | -0.016 |
they | -0.015 |
weekend | -0.015 |
stair | -0.015 |
Token | be |
Token ID | 307 |
Feature activation | -4.61e-02 |
Pos logit contributions | |
hallway | +0.059 |
weekend | +0.059 |
stair | +0.059 |
days | +0.054 |
particular | +0.053 |
Neg logit contributions | |
ps | -0.050 |
ss | -0.040 |
teeth | -0.036 |
ches | -0.036 |
uable | -0.033 |
Token | compassionate |
Token ID | 32533 |
Feature activation | -2.33e-02 |
Pos logit contributions | |
ches | +0.076 |
ps | +0.071 |
iddles | +0.071 |
osaurus | +0.071 |
oni | +0.069 |
Neg logit contributions | |
she | -0.075 |
day | -0.068 |
they | -0.068 |
he | -0.065 |
sunny | -0.064 |
Token | . |
Token ID | 13 |
Feature activation | +1.22e-02 |
Pos logit contributions | |
ps | +0.039 |
ches | +0.034 |
ss | +0.034 |
mus | +0.034 |
iddles | +0.032 |
Neg logit contributions | |
she | -0.036 |
day | -0.036 |
weekend | -0.033 |
stair | -0.032 |
they | -0.032 |
Token | Comp |
Token ID | 3082 |
Feature activation | -6.04e-02 |
Pos logit contributions | |
weekend | +0.018 |
hallway | +0.018 |
stair | +0.018 |
she | +0.017 |
holiday | +0.017 |
Neg logit contributions | |
ps | -0.022 |
ss | -0.019 |
teeth | -0.019 |
mus | -0.018 |
ches | -0.018 |
Token | assion |
Token ID | 11857 |
Feature activation | -1.14e-01 |
Pos logit contributions | |
ps | +0.099 |
teeth | +0.096 |
rier | +0.080 |
ze | +0.076 |
degrees | +0.076 |
Neg logit contributions | |
ite | -0.084 |
boat | -0.079 |
weekend | -0.078 |
day | -0.078 |
she | -0.076 |
Token | ate |
Token ID | 378 |
Feature activation | -2.97e-02 |
Pos logit contributions | |
ps | +0.225 |
ches | +0.207 |
ache | +0.207 |
unnecessary | +0.204 |
infected | +0.202 |
Neg logit contributions | |
day | -0.131 |
of | -0.120 |
where | -0.100 |
they | -0.096 |
hill | -0.093 |
Token | means |
Token ID | 1724 |
Feature activation | -1.21e-02 |
Pos logit contributions | |
ps | +0.051 |
mus | +0.045 |
ss | +0.045 |
iddles | +0.042 |
teeth | +0.042 |
Neg logit contributions | |
she | -0.047 |
day | -0.044 |
they | -0.041 |
weekend | -0.041 |
stair | -0.040 |
Token | you |
Token ID | 345 |
Feature activation | -4.91e-02 |
Pos logit contributions | |
ps | +0.025 |
ss | +0.023 |
ches | +0.022 |
mus | +0.022 |
teeth | +0.021 |
Neg logit contributions | |
she | -0.015 |
day | -0.014 |
weekend | -0.013 |
stair | -0.013 |
they | -0.013 |
Token | care |
Token ID | 1337 |
Feature activation | -6.12e-02 |
Pos logit contributions | |
ps | +0.104 |
rier | +0.100 |
mus | +0.098 |
iddles | +0.098 |
ss | +0.093 |
Neg logit contributions | |
she | -0.057 |
day | -0.054 |
they | -0.054 |
he | -0.048 |
explore | -0.045 |
Token | about |
Token ID | 546 |
Feature activation | -5.21e-02 |
Pos logit contributions | |
ps | +0.125 |
wy | +0.115 |
ss | +0.114 |
mus | +0.114 |
teeth | +0.112 |
Neg logit contributions | |
she | -0.074 |
day | -0.073 |
of | -0.069 |
they | -0.063 |
he | -0.062 |
Token | that |
Token ID | 326 |
Feature activation | -2.99e-03 |
Pos logit contributions | |
days | +0.068 |
hallway | +0.066 |
weekend | +0.065 |
stair | +0.065 |
wander | +0.065 |
Neg logit contributions | |
ps | -0.070 |
ss | -0.058 |
ches | -0.056 |
louder | -0.056 |
teeth | -0.054 |
Token | said |
Token ID | 531 |
Feature activation | -1.37e-02 |
Pos logit contributions | |
ps | +0.005 |
ss | +0.004 |
ches | +0.004 |
teeth | +0.004 |
mus | +0.004 |
Neg logit contributions | |
she | -0.005 |
weekend | -0.005 |
day | -0.005 |
hallway | -0.004 |
stair | -0.004 |
Token | " |
Token ID | 366 |
Feature activation | -2.21e-02 |
Pos logit contributions | |
ps | +0.025 |
ss | +0.021 |
ches | +0.021 |
mus | +0.020 |
teeth | +0.020 |
Neg logit contributions | |
she | -0.021 |
day | -0.019 |
weekend | -0.019 |
stair | -0.019 |
they | -0.019 |
Token | Do |
Token ID | 5211 |
Feature activation | +3.89e-03 |
Pos logit contributions | |
ps | +0.046 |
teeth | +0.039 |
ss | +0.038 |
ches | +0.038 |
iddles | +0.036 |
Neg logit contributions | |
she | -0.028 |
weekend | -0.028 |
day | -0.027 |
hallway | -0.025 |
particular | -0.025 |
Token | not |
Token ID | 407 |
Feature activation | +1.59e-03 |
Pos logit contributions | |
she | +0.006 |
weekend | +0.006 |
hallway | +0.006 |
stair | +0.006 |
day | +0.006 |
Neg logit contributions | |
ps | -0.006 |
ss | -0.006 |
ches | -0.005 |
teeth | -0.005 |
mus | -0.005 |
Token | whistle |
Token ID | 16121 |
Feature activation | -1.13e-01 |
Pos logit contributions | |
she | +0.002 |
weekend | +0.002 |
hallway | +0.002 |
stair | +0.002 |
day | +0.002 |
Neg logit contributions | |
ps | -0.003 |
ss | -0.003 |
teeth | -0.003 |
ches | -0.003 |
mus | -0.003 |
Token | here |
Token ID | 994 |
Feature activation | -6.55e-02 |
Pos logit contributions | |
ps | +0.234 |
wounds | +0.206 |
infected | +0.201 |
iddles | +0.199 |
ussy | +0.198 |
Neg logit contributions | |
day | -0.161 |
they | -0.146 |
she | -0.140 |
today | -0.139 |
he | -0.138 |
Token | ". |
Token ID | 1911 |
Feature activation | +1.18e-02 |
Pos logit contributions | |
ps | +0.123 |
ss | +0.106 |
teeth | +0.100 |
iddles | +0.099 |
wy | +0.097 |
Neg logit contributions | |
she | -0.101 |
day | -0.100 |
weekend | -0.099 |
they | -0.093 |
holiday | -0.092 |
Token | They |
Token ID | 1119 |
Feature activation | -7.63e-03 |
Pos logit contributions | |
weekend | +0.015 |
she | +0.015 |
day | +0.015 |
stair | +0.015 |
hallway | +0.014 |
Neg logit contributions | |
ps | -0.022 |
ss | -0.020 |
teeth | -0.019 |
ches | -0.019 |
mus | -0.019 |
Token | wondered |
Token ID | 14028 |
Feature activation | -4.86e-03 |
Pos logit contributions | |
ps | +0.016 |
ss | +0.015 |
ches | +0.014 |
teeth | +0.014 |
mus | +0.014 |
Neg logit contributions | |
she | -0.009 |
weekend | -0.008 |
day | -0.008 |
hallway | -0.008 |
stair | -0.008 |
Token | why |
Token ID | 1521 |
Feature activation | -2.33e-02 |
Pos logit contributions | |
ps | +0.009 |
ss | +0.008 |
ches | +0.008 |
teeth | +0.008 |
mus | +0.008 |
Neg logit contributions | |
she | -0.007 |
weekend | -0.006 |
day | -0.006 |
stair | -0.006 |
hallway | -0.006 |
Token | a |
Token ID | 257 |
Feature activation | -3.60e-02 |
Pos logit contributions | |
ps | +0.052 |
ss | +0.047 |
ches | +0.046 |
ealing | +0.045 |
mus | +0.045 |
Neg logit contributions | |
she | -0.034 |
day | -0.029 |
they | -0.028 |
he | -0.028 |
weekend | -0.027 |
Token | new |
Token ID | 649 |
Feature activation | -9.34e-03 |
Pos logit contributions | |
ps | +0.069 |
ches | +0.055 |
iddles | +0.055 |
ss | +0.055 |
mus | +0.054 |
Neg logit contributions | |
she | -0.058 |
day | -0.056 |
stair | -0.055 |
hallway | -0.054 |
weekend | -0.052 |
Token | word |
Token ID | 1573 |
Feature activation | -1.87e-02 |
Pos logit contributions | |
ps | +0.018 |
ss | +0.015 |
ches | +0.015 |
mus | +0.014 |
teeth | +0.014 |
Neg logit contributions | |
she | -0.013 |
day | -0.013 |
weekend | -0.013 |
stair | -0.013 |
hallway | -0.013 |
Token | - |
Token ID | 532 |
Feature activation | -1.52e-02 |
Pos logit contributions | |
ps | +0.034 |
ss | +0.030 |
ches | +0.029 |
teeth | +0.028 |
mus | +0.028 |
Neg logit contributions | |
she | -0.028 |
day | -0.026 |
weekend | -0.025 |
stair | -0.025 |
hallway | -0.025 |
Token | " |
Token ID | 366 |
Feature activation | -2.07e-02 |
Pos logit contributions | |
ps | +0.027 |
ss | +0.023 |
ches | +0.023 |
mus | +0.022 |
teeth | +0.022 |
Neg logit contributions | |
she | -0.024 |
day | -0.022 |
weekend | -0.021 |
stair | -0.021 |
hallway | -0.021 |
Token | ignore |
Token ID | 46430 |
Feature activation | -1.13e-01 |
Pos logit contributions | |
ps | +0.043 |
teeth | +0.037 |
ss | +0.036 |
ches | +0.036 |
ealing | +0.034 |
Neg logit contributions | |
she | -0.026 |
weekend | -0.024 |
day | -0.024 |
hallway | -0.023 |
stair | -0.023 |
Token | ". |
Token ID | 1911 |
Feature activation | -1.49e-03 |
Pos logit contributions | |
ps | +0.224 |
ksh | +0.203 |
ags | +0.196 |
ussy | +0.196 |
ows | +0.194 |
Neg logit contributions | |
she | -0.138 |
day | -0.137 |
they | -0.127 |
where | -0.123 |
stair | -0.113 |
Token | She |
Token ID | 1375 |
Feature activation | -4.59e-02 |
Pos logit contributions | |
ps | +0.003 |
ss | +0.003 |
ches | +0.003 |
teeth | +0.003 |
mus | +0.003 |
Neg logit contributions | |
she | -0.002 |
weekend | -0.001 |
day | -0.001 |
hallway | -0.001 |
stair | -0.001 |
Token | learned |
Token ID | 4499 |
Feature activation | -2.85e-02 |
Pos logit contributions | |
ps | +0.105 |
rier | +0.096 |
unnecessary | +0.093 |
ags | +0.093 |
ss | +0.091 |
Neg logit contributions | |
she | -0.055 |
they | -0.045 |
he | -0.044 |
hallway | -0.040 |
day | -0.039 |
Token | that |
Token ID | 326 |
Feature activation | +3.65e-02 |
Pos logit contributions | |
ps | +0.070 |
wy | +0.064 |
ss | +0.064 |
ches | +0.062 |
ealing | +0.062 |
Neg logit contributions | |
she | -0.030 |
they | -0.022 |
he | -0.022 |
day | -0.021 |
weekend | -0.018 |
Token | it |
Token ID | 340 |
Feature activation | -4.59e-02 |
Pos logit contributions | |
hallway | +0.051 |
driveway | +0.050 |
weekend | +0.049 |
sandbox | +0.047 |
loo | +0.046 |
Neg logit contributions | |
ps | -0.061 |
teeth | -0.059 |
ss | -0.056 |
mus | -0.052 |
louder | -0.051 |
Token | t |
Token ID | 83 |
Feature activation | -1.04e-02 |
Pos logit contributions | |
ps | +0.047 |
teeth | +0.039 |
ches | +0.039 |
ss | +0.039 |
iddles | +0.038 |
Neg logit contributions | |
she | -0.034 |
day | -0.034 |
weekend | -0.033 |
days | -0.031 |
particular | -0.031 |
Token | weet |
Token ID | 7277 |
Feature activation | -6.93e-02 |
Pos logit contributions | |
ps | +0.021 |
ss | +0.018 |
teeth | +0.018 |
ches | +0.017 |
iddles | +0.017 |
Neg logit contributions | |
she | -0.014 |
day | -0.013 |
weekend | -0.013 |
stair | -0.013 |
hallway | -0.012 |
Token | , |
Token ID | 11 |
Feature activation | -1.22e-02 |
Pos logit contributions | |
ussy | +0.119 |
ps | +0.118 |
iddles | +0.117 |
wy | +0.113 |
teeth | +0.110 |
Neg logit contributions | |
she | -0.095 |
day | -0.093 |
they | -0.083 |
where | -0.082 |
wander | -0.081 |
Token | tweet |
Token ID | 6126 |
Feature activation | -1.15e-01 |
Pos logit contributions | |
ps | +0.024 |
ss | +0.021 |
ches | +0.021 |
iddles | +0.021 |
teeth | +0.020 |
Neg logit contributions | |
she | -0.016 |
day | -0.015 |
weekend | -0.015 |
stair | -0.015 |
they | -0.014 |
Token | , |
Token ID | 11 |
Feature activation | -1.36e-02 |
Pos logit contributions | |
ps | +0.277 |
wounds | +0.252 |
wy | +0.242 |
asy | +0.242 |
iddles | +0.241 |
Neg logit contributions | |
day | -0.128 |
of | -0.123 |
she | -0.116 |
where | -0.113 |
they | -0.104 |
Token | tweet |
Token ID | 6126 |
Feature activation | -1.13e-01 |
Pos logit contributions | |
ps | +0.024 |
ss | +0.020 |
iddles | +0.020 |
ches | +0.020 |
teeth | +0.020 |
Neg logit contributions | |
she | -0.021 |
day | -0.020 |
weekend | -0.020 |
stair | -0.019 |
hallway | -0.019 |
Token | ". |
Token ID | 1911 |
Feature activation | -1.27e-02 |
Pos logit contributions | |
ps | +0.273 |
wounds | +0.246 |
asy | +0.242 |
wy | +0.241 |
ussy | +0.236 |
Neg logit contributions | |
day | -0.122 |
of | -0.117 |
she | -0.113 |
where | -0.109 |
they | -0.101 |
Token | Tim |
Token ID | 5045 |
Feature activation | -5.47e-02 |
Pos logit contributions | |
ps | +0.028 |
ss | +0.024 |
ches | +0.024 |
teeth | +0.023 |
mus | +0.023 |
Neg logit contributions | |
she | -0.016 |
day | -0.014 |
they | -0.014 |
weekend | -0.014 |
stair | -0.013 |
Token | and |
Token ID | 290 |
Feature activation | -2.15e-03 |
Pos logit contributions | |
rier | +0.123 |
mus | +0.120 |
unnecessary | +0.120 |
ps | +0.119 |
uable | +0.116 |
Neg logit contributions | |
she | -0.045 |
sandbox | -0.040 |
day | -0.038 |
hill | -0.038 |
slide | -0.038 |
Token | Lily |
Token ID | 20037 |
Feature activation | -2.05e-02 |
Pos logit contributions | |
ps | +0.004 |
ss | +0.004 |
ches | +0.003 |
teeth | +0.003 |
mus | +0.003 |
Neg logit contributions | |
she | -0.003 |
weekend | -0.003 |
day | -0.003 |
stair | -0.003 |
hallway | -0.003 |
Token | think |
Token ID | 892 |
Feature activation | -1.04e-02 |
Pos logit contributions | |
ps | +0.044 |
ss | +0.041 |
mus | +0.039 |
iddles | +0.039 |
ches | +0.039 |
Neg logit contributions | |
she | -0.022 |
day | -0.020 |
explore | -0.019 |
wander | -0.019 |
stair | -0.019 |
Token | Anna |
Token ID | 11735 |
Feature activation | -4.17e-02 |
Pos logit contributions | |
ss | +0.100 |
ps | +0.097 |
ealing | +0.097 |
wy | +0.097 |
ches | +0.094 |
Neg logit contributions | |
she | -0.057 |
Jane | -0.044 |
they | -0.042 |
stair | -0.042 |
particular | -0.041 |
Token | . |
Token ID | 13 |
Feature activation | -5.80e-02 |
Pos logit contributions | |
ps | +0.081 |
wy | +0.072 |
ss | +0.070 |
iddles | +0.068 |
ches | +0.068 |
Neg logit contributions | |
she | -0.063 |
day | -0.052 |
they | -0.051 |
he | -0.051 |
days | -0.047 |
Token | " |
Token ID | 366 |
Feature activation | -3.40e-02 |
Pos logit contributions | |
ps | +0.152 |
wy | +0.150 |
unnecessary | +0.141 |
ache | +0.134 |
ss | +0.133 |
Neg logit contributions | |
she | -0.071 |
they | -0.058 |
Jane | -0.048 |
he | -0.046 |
stair | -0.040 |
Token | Can |
Token ID | 6090 |
Feature activation | -1.15e-02 |
Pos logit contributions | |
ps | +0.082 |
teeth | +0.073 |
ches | +0.071 |
ss | +0.070 |
ealing | +0.068 |
Neg logit contributions | |
she | -0.035 |
weekend | -0.029 |
day | -0.029 |
particular | -0.027 |
stair | -0.027 |
Token | you |
Token ID | 345 |
Feature activation | -4.86e-02 |
Pos logit contributions | |
ps | +0.031 |
ss | +0.028 |
ches | +0.027 |
iddles | +0.027 |
teeth | +0.027 |
Neg logit contributions | |
she | -0.008 |
day | -0.006 |
weekend | -0.006 |
stair | -0.006 |
they | -0.006 |
Token | say |
Token ID | 910 |
Feature activation | -1.13e-01 |
Pos logit contributions | |
ps | +0.119 |
iddles | +0.103 |
rier | +0.103 |
ss | +0.102 |
teeth | +0.102 |
Neg logit contributions | |
she | -0.052 |
explore | -0.046 |
wander | -0.042 |
particular | -0.042 |
day | -0.040 |
Token | cat |
Token ID | 3797 |
Feature activation | -9.09e-02 |
Pos logit contributions | |
wy | +0.270 |
ps | +0.263 |
aned | +0.245 |
ished | +0.242 |
iddles | +0.238 |
Neg logit contributions | |
she | -0.145 |
they | -0.122 |
he | -0.112 |
day | -0.099 |
today | -0.096 |
Token | ?" |
Token ID | 1701 |
Feature activation | -2.90e-02 |
Pos logit contributions | |
ps | +0.207 |
rier | +0.173 |
aned | +0.170 |
wy | +0.167 |
ealing | +0.165 |
Neg logit contributions | |
she | -0.127 |
they | -0.110 |
day | -0.106 |
he | -0.101 |
weekend | -0.097 |
Token |
|
Token ID | 198 |
Feature activation | -5.73e-02 |
Pos logit contributions | |
ps | +0.073 |
ss | +0.064 |
wy | +0.064 |
ches | +0.061 |
rier | +0.061 |
Neg logit contributions | |
she | -0.037 |
they | -0.028 |
he | -0.027 |
day | -0.026 |
weekend | -0.024 |
Token |
|
Token ID | 198 |
Feature activation | -1.45e-02 |
Pos logit contributions | |
ps | +0.169 |
wy | +0.146 |
rier | +0.145 |
teeth | +0.144 |
ss | +0.141 |
Neg logit contributions | |
she | -0.047 |
they | -0.035 |
day | -0.034 |
weekend | -0.029 |
he | -0.029 |
Token | Anna |
Token ID | 31160 |
Feature activation | -6.51e-02 |
Pos logit contributions | |
ps | +0.032 |
ss | +0.028 |
teeth | +0.028 |
ches | +0.027 |
iddles | +0.027 |
Neg logit contributions | |
she | -0.017 |
hallway | -0.015 |
stair | -0.015 |
weekend | -0.015 |
day | -0.015 |
Token | . |
Token ID | 13 |
Feature activation | -4.87e-03 |
Pos logit contributions | |
ps | +0.034 |
ss | +0.030 |
ches | +0.029 |
teeth | +0.029 |
mus | +0.028 |
Neg logit contributions | |
she | -0.034 |
they | -0.030 |
day | -0.030 |
stair | -0.029 |
weekend | -0.029 |
Token | " |
Token ID | 366 |
Feature activation | +5.68e-03 |
Pos logit contributions | |
ps | +0.010 |
ss | +0.009 |
ches | +0.008 |
teeth | +0.008 |
mus | +0.008 |
Neg logit contributions | |
she | -0.006 |
weekend | -0.006 |
day | -0.006 |
stair | -0.006 |
hallway | -0.006 |
Token | Maybe |
Token ID | 13300 |
Feature activation | -4.31e-03 |
Pos logit contributions | |
she | +0.007 |
weekend | +0.006 |
hallway | +0.006 |
stair | +0.006 |
day | +0.006 |
Neg logit contributions | |
ps | -0.012 |
ss | -0.011 |
mus | -0.010 |
ches | -0.010 |
teeth | -0.010 |
Token | you |
Token ID | 345 |
Feature activation | -3.33e-02 |
Pos logit contributions | |
ps | +0.011 |
ss | +0.009 |
ches | +0.009 |
teeth | +0.009 |
mus | +0.009 |
Neg logit contributions | |
she | -0.004 |
weekend | -0.003 |
day | -0.003 |
stair | -0.003 |
hallway | -0.003 |
Token | can |
Token ID | 460 |
Feature activation | -2.47e-02 |
Pos logit contributions | |
ps | +0.071 |
ss | +0.063 |
iddles | +0.061 |
mus | +0.060 |
rier | +0.060 |
Neg logit contributions | |
day | -0.037 |
she | -0.037 |
explore | -0.035 |
weekend | -0.035 |
they | -0.034 |
Token | cut |
Token ID | 2005 |
Feature activation | -1.12e-01 |
Pos logit contributions | |
ps | +0.052 |
ss | +0.048 |
ches | +0.046 |
teeth | +0.046 |
mus | +0.046 |
Neg logit contributions | |
she | -0.028 |
day | -0.026 |
explore | -0.026 |
weekend | -0.024 |
wander | -0.024 |
Token | it |
Token ID | 340 |
Feature activation | -5.79e-02 |
Pos logit contributions | |
iddles | +0.281 |
ps | +0.278 |
ows | +0.261 |
ache | +0.253 |
lic | +0.253 |
Neg logit contributions | |
he | -0.083 |
they | -0.081 |
she | -0.079 |
day | -0.071 |
we | -0.052 |
Token | in |
Token ID | 287 |
Feature activation | -3.45e-02 |
Pos logit contributions | |
ps | +0.120 |
iddles | +0.112 |
ss | +0.112 |
ealing | +0.109 |
unnecessary | +0.107 |
Neg logit contributions | |
she | -0.071 |
day | -0.066 |
they | -0.065 |
he | -0.063 |
weekend | -0.056 |
Token | half |
Token ID | 2063 |
Feature activation | -3.52e-02 |
Pos logit contributions | |
ps | +0.066 |
ches | +0.061 |
ss | +0.059 |
iddles | +0.058 |
mus | +0.057 |
Neg logit contributions | |
she | -0.047 |
day | -0.046 |
stair | -0.042 |
weekend | -0.042 |
holiday | -0.041 |
Token | , |
Token ID | 11 |
Feature activation | +8.21e-03 |
Pos logit contributions | |
ps | +0.067 |
ss | +0.057 |
iddles | +0.056 |
mus | +0.056 |
ches | +0.055 |
Neg logit contributions | |
she | -0.051 |
day | -0.050 |
they | -0.047 |
he | -0.046 |
weekend | -0.046 |
Token | or |
Token ID | 393 |
Feature activation | -9.71e-03 |
Pos logit contributions | |
weekend | +0.011 |
hallway | +0.011 |
she | +0.011 |
stair | +0.011 |
day | +0.011 |
Neg logit contributions | |
ps | -0.015 |
ss | -0.013 |
teeth | -0.013 |
ches | -0.013 |
mus | -0.012 |
Token | and |
Token ID | 290 |
Feature activation | -4.75e-03 |
Pos logit contributions | |
weekend | +0.018 |
hallway | +0.017 |
she | +0.017 |
stair | +0.017 |
day | +0.017 |
Neg logit contributions | |
ps | -0.017 |
ss | -0.014 |
teeth | -0.014 |
ches | -0.013 |
mus | -0.013 |
Token | had |
Token ID | 550 |
Feature activation | -2.47e-02 |
Pos logit contributions | |
ps | +0.010 |
ss | +0.009 |
ches | +0.008 |
teeth | +0.008 |
mus | +0.008 |
Neg logit contributions | |
she | -0.006 |
weekend | -0.006 |
day | -0.006 |
hallway | -0.006 |
stair | -0.005 |
Token | the |
Token ID | 262 |
Feature activation | -3.09e-02 |
Pos logit contributions | |
ps | +0.040 |
ss | +0.036 |
ches | +0.034 |
mus | +0.033 |
ealing | +0.033 |
Neg logit contributions | |
she | -0.041 |
day | -0.038 |
weekend | -0.037 |
hallway | -0.037 |
holiday | -0.036 |
Token | word |
Token ID | 1573 |
Feature activation | -3.18e-02 |
Pos logit contributions | |
ps | +0.052 |
ss | +0.047 |
ches | +0.045 |
mus | +0.044 |
iddles | +0.044 |
Neg logit contributions | |
day | -0.049 |
she | -0.048 |
weekend | -0.047 |
stair | -0.045 |
hallway | -0.045 |
Token | " |
Token ID | 366 |
Feature activation | -7.69e-02 |
Pos logit contributions | |
ps | +0.057 |
ss | +0.050 |
iddles | +0.048 |
ches | +0.047 |
mus | +0.047 |
Neg logit contributions | |
she | -0.050 |
day | -0.045 |
they | -0.044 |
stair | -0.044 |
weekend | -0.044 |
Token | w |
Token ID | 86 |
Feature activation | -1.12e-01 |
Pos logit contributions | |
ps | +0.153 |
teeth | +0.128 |
ealing | +0.123 |
rier | +0.122 |
wounds | +0.121 |
Neg logit contributions | |
she | -0.108 |
day | -0.107 |
weekend | -0.105 |
days | -0.103 |
particular | -0.102 |
Token | elcome |
Token ID | 9571 |
Feature activation | -7.10e-02 |
Pos logit contributions | |
smarter | +0.179 |
unnecessary | +0.179 |
ps | +0.176 |
iddles | +0.166 |
stronger | +0.165 |
Neg logit contributions | |
she | -0.148 |
day | -0.139 |
they | -0.133 |
Jane | -0.127 |
stair | -0.126 |
Token | " |
Token ID | 1 |
Feature activation | -7.13e-02 |
Pos logit contributions | |
ps | +0.151 |
iddles | +0.123 |
teeth | +0.120 |
ussy | +0.118 |
wy | +0.117 |
Neg logit contributions | |
she | -0.092 |
day | -0.091 |
weekend | -0.084 |
they | -0.080 |
where | -0.079 |
Token | written |
Token ID | 3194 |
Feature activation | -8.97e-02 |
Pos logit contributions | |
ps | +0.158 |
teeth | +0.129 |
ches | +0.127 |
unnecessary | +0.127 |
ss | +0.125 |
Neg logit contributions | |
she | -0.102 |
day | -0.089 |
weekend | -0.086 |
they | -0.083 |
he | -0.081 |
Token | in |
Token ID | 287 |
Feature activation | -2.75e-02 |
Pos logit contributions | |
ps | +0.178 |
unnecessary | +0.173 |
wounds | +0.159 |
iddles | +0.157 |
ws | +0.157 |
Neg logit contributions | |
she | -0.123 |
they | -0.101 |
he | -0.101 |
of | -0.093 |
day | -0.093 |
Token | big |
Token ID | 1263 |
Feature activation | -2.77e-02 |
Pos logit contributions | |
ps | +0.051 |
ches | +0.045 |
ss | +0.044 |
iddles | +0.044 |
mus | +0.043 |
Neg logit contributions | |
she | -0.039 |
day | -0.036 |
stair | -0.035 |
hallway | -0.035 |
particular | -0.034 |
Token | played |
Token ID | 2826 |
Feature activation | -2.99e-02 |
Pos logit contributions | |
she | +0.011 |
hallway | +0.011 |
stair | +0.011 |
weekend | +0.011 |
day | +0.010 |
Neg logit contributions | |
ps | -0.019 |
ss | -0.017 |
ches | -0.017 |
teeth | -0.016 |
mus | -0.016 |
Token | a |
Token ID | 257 |
Feature activation | +2.76e-03 |
Pos logit contributions | |
ps | +0.063 |
ss | +0.058 |
ache | +0.056 |
uable | +0.055 |
rier | +0.055 |
Neg logit contributions | |
she | -0.036 |
day | -0.036 |
weekend | -0.033 |
they | -0.033 |
sunny | -0.033 |
Token | game |
Token ID | 983 |
Feature activation | +4.81e-02 |
Pos logit contributions | |
she | +0.003 |
weekend | +0.003 |
hallway | +0.003 |
stair | +0.003 |
day | +0.003 |
Neg logit contributions | |
ps | -0.006 |
ss | -0.005 |
ches | -0.005 |
teeth | -0.005 |
mus | -0.005 |
Token | called |
Token ID | 1444 |
Feature activation | -5.09e-02 |
Pos logit contributions | |
days | +0.068 |
hallway | +0.067 |
driveway | +0.067 |
holiday | +0.067 |
stair | +0.066 |
Neg logit contributions | |
ps | -0.079 |
ches | -0.070 |
ss | -0.064 |
iddles | -0.062 |
ows | -0.062 |
Token | " |
Token ID | 366 |
Feature activation | -4.40e-02 |
Pos logit contributions | |
ps | +0.097 |
ss | +0.081 |
ches | +0.081 |
wy | +0.079 |
mus | +0.078 |
Neg logit contributions | |
she | -0.072 |
they | -0.068 |
weekend | -0.066 |
day | -0.066 |
he | -0.066 |
Token | shoot |
Token ID | 30408 |
Feature activation | -1.12e-01 |
Pos logit contributions | |
ps | +0.085 |
teeth | +0.073 |
uable | +0.070 |
ealing | +0.068 |
ches | +0.067 |
Neg logit contributions | |
day | -0.063 |
weekend | -0.063 |
she | -0.061 |
stair | -0.059 |
hallway | -0.058 |
Token | the |
Token ID | 262 |
Feature activation | -7.34e-02 |
Pos logit contributions | |
ps | +0.222 |
wounds | +0.187 |
ags | +0.180 |
iddles | +0.177 |
rier | +0.173 |
Neg logit contributions | |
she | -0.147 |
day | -0.146 |
he | -0.141 |
where | -0.134 |
of | -0.131 |
Token | w |
Token ID | 266 |
Feature activation | -8.01e-02 |
Pos logit contributions | |
ps | +0.135 |
uable | +0.123 |
mus | +0.113 |
ws | +0.112 |
ealing | +0.112 |
Neg logit contributions | |
day | -0.111 |
stair | -0.104 |
she | -0.103 |
playground | -0.103 |
hallway | -0.103 |
Token | affle |
Token ID | 30697 |
Feature activation | -6.12e-02 |
Pos logit contributions | |
ps | +0.133 |
louder | +0.123 |
uable | +0.121 |
unnecessary | +0.114 |
infected | +0.114 |
Neg logit contributions | |
loo | -0.104 |
day | -0.103 |
she | -0.101 |
stair | -0.097 |
weekend | -0.097 |
Token | ." |
Token ID | 526 |
Feature activation | -3.48e-03 |
Pos logit contributions | |
ps | +0.115 |
unnecessary | +0.103 |
iddles | +0.098 |
teeth | +0.097 |
ss | +0.097 |
Neg logit contributions | |
she | -0.086 |
stair | -0.082 |
day | -0.081 |
weekend | -0.078 |
they | -0.076 |
Token | In |
Token ID | 554 |
Feature activation | +2.82e-02 |
Pos logit contributions | |
ps | +0.008 |
ss | +0.007 |
ches | +0.007 |
teeth | +0.007 |
mus | +0.007 |
Neg logit contributions | |
she | -0.004 |
weekend | -0.003 |
day | -0.003 |
hallway | -0.003 |
stair | -0.003 |
Token |
|
Token ID | 198 |
Feature activation | -1.61e-02 |
Pos logit contributions | |
ps | +0.057 |
ches | +0.049 |
ss | +0.048 |
teeth | +0.047 |
wy | +0.047 |
Neg logit contributions | |
she | -0.031 |
they | -0.026 |
weekend | -0.026 |
day | -0.026 |
stair | -0.025 |
Token |
|
Token ID | 198 |
Feature activation | +2.26e-02 |
Pos logit contributions | |
ps | +0.036 |
ss | +0.031 |
ches | +0.030 |
teeth | +0.030 |
mus | +0.029 |
Neg logit contributions | |
she | -0.019 |
day | -0.017 |
weekend | -0.017 |
stair | -0.016 |
hallway | -0.016 |
Token | " |
Token ID | 1 |
Feature activation | -1.10e-02 |
Pos logit contributions | |
weekend | +0.042 |
day | +0.041 |
holiday | +0.040 |
hallway | +0.040 |
days | +0.040 |
Neg logit contributions | |
ps | -0.029 |
ss | -0.028 |
mus | -0.026 |
ches | -0.024 |
ealing | -0.022 |
Token | S |
Token ID | 50 |
Feature activation | -1.82e-02 |
Pos logit contributions | |
ps | +0.025 |
ches | +0.022 |
ss | +0.022 |
teeth | +0.022 |
iddles | +0.021 |
Neg logit contributions | |
she | -0.012 |
weekend | -0.011 |
day | -0.011 |
hallway | -0.010 |
stair | -0.010 |
Token | ss |
Token ID | 824 |
Feature activation | -1.04e-01 |
Pos logit contributions | |
ps | +0.038 |
teeth | +0.032 |
ches | +0.031 |
mus | +0.031 |
ss | +0.030 |
Neg logit contributions | |
she | -0.023 |
day | -0.021 |
weekend | -0.021 |
stair | -0.020 |
holiday | -0.020 |
Token | ss |
Token ID | 824 |
Feature activation | -1.12e-01 |
Pos logit contributions | |
ps | +0.160 |
rier | +0.149 |
ags | +0.146 |
wounds | +0.144 |
unnecessary | +0.136 |
Neg logit contributions | |
she | -0.160 |
he | -0.152 |
day | -0.146 |
they | -0.145 |
ite | -0.143 |
Token | ," |
Token ID | 553 |
Feature activation | +2.75e-02 |
Pos logit contributions | |
ps | +0.157 |
wounds | +0.148 |
rier | +0.147 |
ags | +0.146 |
iddles | +0.134 |
Neg logit contributions | |
she | -0.188 |
he | -0.177 |
they | -0.170 |
day | -0.167 |
ite | -0.163 |
Token | Tom |
Token ID | 4186 |
Feature activation | -4.57e-02 |
Pos logit contributions | |
days | +0.052 |
hallway | +0.051 |
weekend | +0.051 |
stair | +0.049 |
where | +0.048 |
Neg logit contributions | |
ps | -0.037 |
teeth | -0.034 |
ss | -0.031 |
louder | -0.027 |
mus | -0.027 |
Token | says |
Token ID | 1139 |
Feature activation | -3.53e-02 |
Pos logit contributions | |
ps | +0.079 |
rier | +0.078 |
unnecessary | +0.076 |
ss | +0.074 |
mus | +0.074 |
Neg logit contributions | |
she | -0.069 |
he | -0.060 |
they | -0.059 |
day | -0.059 |
Jane | -0.055 |
Token | . |
Token ID | 13 |
Feature activation | -2.77e-02 |
Pos logit contributions | |
ps | +0.067 |
ss | +0.063 |
wy | +0.063 |
ches | +0.062 |
uable | +0.062 |
Neg logit contributions | |
she | -0.052 |
they | -0.046 |
he | -0.045 |
day | -0.043 |
Jane | -0.039 |
Token |
|
Token ID | 198 |
Feature activation | -1.98e-02 |
Pos logit contributions | |
ps | +0.060 |
wy | +0.053 |
ches | +0.053 |
unnecessary | +0.052 |
ss | +0.051 |
Neg logit contributions | |
she | -0.038 |
he | -0.034 |
they | -0.034 |
day | -0.031 |
stair | -0.030 |
Token | their |
Token ID | 511 |
Feature activation | -4.05e-03 |
Pos logit contributions | |
stair | +0.125 |
hallway | +0.124 |
storm | +0.118 |
path | +0.118 |
seek | +0.117 |
Neg logit contributions | |
ps | -0.064 |
better | -0.041 |
teeth | -0.040 |
ches | -0.038 |
stronger | -0.037 |
Token | things |
Token ID | 1243 |
Feature activation | +3.38e-02 |
Pos logit contributions | |
ps | +0.007 |
ss | +0.006 |
ches | +0.006 |
teeth | +0.006 |
mus | +0.006 |
Neg logit contributions | |
she | -0.006 |
weekend | -0.006 |
day | -0.006 |
hallway | -0.006 |
stair | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | +1.34e-02 |
Pos logit contributions | |
hallway | +0.065 |
stair | +0.064 |
weekend | +0.062 |
particular | +0.061 |
sandbox | +0.060 |
Neg logit contributions | |
ps | -0.044 |
ches | -0.034 |
ss | -0.034 |
teeth | -0.032 |
rier | -0.031 |
Token | On |
Token ID | 1550 |
Feature activation | +4.55e-02 |
Pos logit contributions | |
stair | +0.020 |
she | +0.020 |
hallway | +0.020 |
weekend | +0.020 |
day | +0.019 |
Neg logit contributions | |
ps | -0.022 |
ss | -0.019 |
teeth | -0.019 |
ches | -0.018 |
mus | -0.018 |
Token | their |
Token ID | 511 |
Feature activation | +2.29e-02 |
Pos logit contributions | |
days | +0.081 |
stair | +0.079 |
hallway | +0.079 |
sandbox | +0.077 |
wander | +0.076 |
Neg logit contributions | |
ps | -0.062 |
teeth | -0.045 |
ches | -0.043 |
ss | -0.041 |
uable | -0.038 |
Token | way |
Token ID | 835 |
Feature activation | +9.14e-02 |
Pos logit contributions | |
she | +0.028 |
stair | +0.026 |
hallway | +0.025 |
weekend | +0.024 |
days | +0.024 |
Neg logit contributions | |
ps | -0.044 |
teeth | -0.040 |
ss | -0.039 |
ches | -0.038 |
mus | -0.035 |
Token | home |
Token ID | 1363 |
Feature activation | +2.57e-02 |
Pos logit contributions | |
hallway | +0.218 |
occasion | +0.218 |
days | +0.206 |
stair | +0.204 |
sandbox | +0.203 |
Neg logit contributions | |
back | -0.062 |
ps | -0.047 |
ches | -0.034 |
up | -0.031 |
better | -0.031 |
Token | , |
Token ID | 11 |
Feature activation | +3.54e-02 |
Pos logit contributions | |
weekend | +0.039 |
stair | +0.038 |
hallway | +0.038 |
day | +0.037 |
particular | +0.037 |
Neg logit contributions | |
ps | -0.042 |
ss | -0.036 |
teeth | -0.035 |
ches | -0.033 |
mus | -0.032 |
Token | they |
Token ID | 484 |
Feature activation | +6.67e-05 |
Pos logit contributions | |
hallway | +0.040 |
stair | +0.037 |
occasion | +0.037 |
weekend | +0.036 |
days | +0.036 |
Neg logit contributions | |
ps | -0.068 |
teeth | -0.063 |
ss | -0.060 |
louder | -0.060 |
mus | -0.057 |
Token | told |
Token ID | 1297 |
Feature activation | -2.39e-02 |
Pos logit contributions | |
she | +0.000 |
weekend | +0.000 |
day | +0.000 |
hallway | +0.000 |
stair | +0.000 |
Neg logit contributions | |
ps | -0.000 |
ss | -0.000 |
ches | -0.000 |
teeth | -0.000 |
mus | -0.000 |
Token | stories |
Token ID | 3923 |
Feature activation | +2.12e-02 |
Pos logit contributions | |
ps | +0.052 |
ss | +0.050 |
ches | +0.047 |
mus | +0.047 |
wy | +0.046 |
Neg logit contributions | |
she | -0.027 |
day | -0.026 |
weekend | -0.025 |
holiday | -0.025 |
they | -0.024 |
Token | day |
Token ID | 1110 |
Feature activation | +1.37e-02 |
Pos logit contributions | |
stack | +0.051 |
driveway | +0.048 |
sandbox | +0.046 |
hallway | +0.044 |
harbour | +0.043 |
Neg logit contributions | |
ss | -0.103 |
ches | -0.102 |
funny | -0.100 |
p | -0.097 |
teeth | -0.096 |
Token | , |
Token ID | 11 |
Feature activation | +3.01e-02 |
Pos logit contributions | |
weekend | +0.021 |
hallway | +0.021 |
stair | +0.020 |
driveway | +0.020 |
holiday | +0.019 |
Neg logit contributions | |
ps | -0.022 |
ss | -0.020 |
teeth | -0.020 |
ches | -0.019 |
mus | -0.019 |
Token | she |
Token ID | 673 |
Feature activation | +4.34e-02 |
Pos logit contributions | |
hallway | +0.030 |
driveway | +0.030 |
loo | +0.029 |
days | +0.029 |
weekend | +0.029 |
Neg logit contributions | |
ps | -0.061 |
teeth | -0.058 |
ss | -0.057 |
ches | -0.055 |
louder | -0.055 |
Token | found |
Token ID | 1043 |
Feature activation | -4.54e-03 |
Pos logit contributions | |
stack | +0.077 |
stair | +0.074 |
contain | +0.071 |
weekend | +0.071 |
particular | +0.071 |
Neg logit contributions | |
teeth | -0.052 |
ps | -0.052 |
ss | -0.051 |
ches | -0.051 |
ze | -0.041 |
Token | a |
Token ID | 257 |
Feature activation | -4.58e-03 |
Pos logit contributions | |
ps | +0.007 |
ss | +0.006 |
ches | +0.006 |
teeth | +0.006 |
mus | +0.006 |
Neg logit contributions | |
she | -0.008 |
weekend | -0.008 |
day | -0.008 |
stair | -0.007 |
hallway | -0.007 |
Token | pile |
Token ID | 14540 |
Feature activation | +8.73e-02 |
Pos logit contributions | |
ps | +0.007 |
ss | +0.006 |
ches | +0.006 |
teeth | +0.006 |
mus | +0.006 |
Neg logit contributions | |
she | -0.008 |
weekend | -0.008 |
day | -0.008 |
stair | -0.008 |
hallway | -0.008 |
Token | of |
Token ID | 286 |
Feature activation | -2.46e-03 |
Pos logit contributions | |
holiday | +0.134 |
holidays | +0.123 |
weekend | +0.123 |
sandbox | +0.122 |
driveway | +0.122 |
Neg logit contributions | |
ps | -0.149 |
ches | -0.126 |
ss | -0.126 |
ers | -0.120 |
oni | -0.114 |
Token | ash |
Token ID | 12530 |
Feature activation | +2.75e-02 |
Pos logit contributions | |
ps | +0.004 |
ss | +0.003 |
ches | +0.003 |
teeth | +0.003 |
mus | +0.003 |
Neg logit contributions | |
she | -0.004 |
weekend | -0.004 |
day | -0.004 |
hallway | -0.004 |
stair | -0.004 |
Token | on |
Token ID | 319 |
Feature activation | -1.68e-03 |
Pos logit contributions | |
weekend | +0.047 |
days | +0.047 |
holiday | +0.045 |
particular | +0.045 |
she | +0.045 |
Neg logit contributions | |
ps | -0.043 |
ss | -0.035 |
ches | -0.034 |
teeth | -0.034 |
oni | -0.030 |
Token | the |
Token ID | 262 |
Feature activation | -4.14e-04 |
Pos logit contributions | |
ps | +0.003 |
ss | +0.003 |
ches | +0.002 |
teeth | +0.002 |
mus | +0.002 |
Neg logit contributions | |
she | -0.002 |
weekend | -0.002 |
day | -0.002 |
hallway | -0.002 |
stair | -0.002 |
Token | ground |
Token ID | 2323 |
Feature activation | +3.71e-02 |
Pos logit contributions | |
ps | +0.001 |
ss | +0.001 |
ches | +0.001 |
teeth | +0.001 |
mus | +0.001 |
Neg logit contributions | |
she | -0.001 |
weekend | -0.001 |
day | -0.001 |
hallway | -0.001 |
stair | -0.001 |
Token | , |
Token ID | 11 |
Feature activation | +5.97e-02 |
Pos logit contributions | |
hallway | +0.049 |
days | +0.047 |
sandbox | +0.047 |
driveway | +0.046 |
weekend | +0.046 |
Neg logit contributions | |
ps | -0.045 |
ches | -0.039 |
ss | -0.038 |
teeth | -0.038 |
oni | -0.036 |
Token | she |
Token ID | 673 |
Feature activation | +1.89e-02 |
Pos logit contributions | |
hallway | +0.065 |
weekend | +0.064 |
driveway | +0.062 |
occasion | +0.060 |
stair | +0.060 |
Neg logit contributions | |
ps | -0.112 |
teeth | -0.105 |
ss | -0.100 |
mus | -0.098 |
ches | -0.095 |
Token | found |
Token ID | 1043 |
Feature activation | -1.44e-03 |
Pos logit contributions | |
weekend | +0.025 |
she | +0.024 |
stair | +0.024 |
day | +0.024 |
hallway | +0.023 |
Neg logit contributions | |
ps | -0.035 |
ss | -0.031 |
ches | -0.030 |
teeth | -0.030 |
mus | -0.029 |
Token | a |
Token ID | 257 |
Feature activation | +5.96e-03 |
Pos logit contributions | |
ps | +0.002 |
ss | +0.002 |
ches | +0.002 |
teeth | +0.002 |
mus | +0.002 |
Neg logit contributions | |
she | -0.002 |
weekend | -0.002 |
day | -0.002 |
hallway | -0.002 |
stair | -0.002 |
Token | big |
Token ID | 1263 |
Feature activation | +3.75e-02 |
Pos logit contributions | |
she | +0.010 |
weekend | +0.010 |
day | +0.010 |
hallway | +0.010 |
stair | +0.010 |
Neg logit contributions | |
ps | -0.009 |
ss | -0.008 |
teeth | -0.007 |
ches | -0.007 |
mus | -0.007 |
Token | room |
Token ID | 2119 |
Feature activation | +1.05e-01 |
Pos logit contributions | |
she | +0.054 |
weekend | +0.054 |
days | +0.052 |
hallway | +0.049 |
they | +0.048 |
Neg logit contributions | |
ps | -0.062 |
teeth | -0.056 |
ss | -0.054 |
ches | -0.051 |
mus | -0.049 |
Token | full |
Token ID | 1336 |
Feature activation | +5.01e-02 |
Pos logit contributions | |
wander | +0.208 |
weekend | +0.207 |
hurricane | +0.205 |
where | +0.202 |
days | +0.201 |
Neg logit contributions | |
ps | -0.091 |
ers | -0.064 |
ows | -0.061 |
teeth | -0.059 |
aron | -0.057 |
Token | of |
Token ID | 286 |
Feature activation | +1.59e-02 |
Pos logit contributions | |
weekend | +0.042 |
where | +0.040 |
wander | +0.040 |
hallway | +0.040 |
days | +0.039 |
Neg logit contributions | |
ps | -0.108 |
teeth | -0.100 |
ss | -0.095 |
ches | -0.092 |
mus | -0.090 |
Token | boxes |
Token ID | 10559 |
Feature activation | +7.87e-02 |
Pos logit contributions | |
weekend | +0.027 |
she | +0.027 |
stair | +0.026 |
hallway | +0.026 |
day | +0.026 |
Neg logit contributions | |
ps | -0.024 |
teeth | -0.020 |
ss | -0.019 |
mus | -0.018 |
ches | -0.018 |
Token | . |
Token ID | 13 |
Feature activation | +1.20e-02 |
Pos logit contributions | |
where | +0.182 |
particular | +0.180 |
sandbox | +0.178 |
wander | +0.177 |
hallway | +0.177 |
Neg logit contributions | |
ps | -0.064 |
ows | -0.044 |
oni | -0.033 |
ches | -0.032 |
rier | -0.032 |
Token |
|
Token ID | 198 |
Feature activation | +1.49e-02 |
Pos logit contributions | |
weekend | +0.013 |
day | +0.013 |
she | +0.013 |
hallway | +0.012 |
stair | +0.012 |
Neg logit contributions | |
ps | -0.025 |
ss | -0.022 |
teeth | -0.022 |
ches | -0.022 |
mus | -0.021 |
Token |
|
Token ID | 198 |
Feature activation | +4.08e-02 |
Pos logit contributions | |
stair | +0.033 |
hallway | +0.032 |
she | +0.032 |
weekend | +0.032 |
day | +0.031 |
Neg logit contributions | |
ps | -0.031 |
ss | -0.029 |
ches | -0.028 |
mus | -0.026 |
teeth | -0.026 |
Token | They |
Token ID | 2990 |
Feature activation | +3.49e-02 |
Pos logit contributions | |
stair | +0.078 |
stack | +0.075 |
hallway | +0.075 |
exploration | +0.075 |
explore | +0.073 |
Neg logit contributions | |
ss | -0.045 |
ches | -0.042 |
mus | -0.041 |
ps | -0.041 |
ealing | -0.040 |
Token | see |
Token ID | 766 |
Feature activation | +1.02e-02 |
Pos logit contributions | |
weekend | +0.046 |
stair | +0.046 |
hallway | +0.046 |
she | +0.044 |
day | +0.044 |
Neg logit contributions | |
ps | -0.062 |
ches | -0.056 |
teeth | -0.054 |
ss | -0.053 |
ealing | -0.050 |
Token | a |
Token ID | 257 |
Feature activation | -4.24e-03 |
Pos logit contributions | |
hallway | +0.019 |
she | +0.019 |
weekend | +0.019 |
stair | +0.019 |
day | +0.018 |
Neg logit contributions | |
ps | -0.014 |
teeth | -0.012 |
ss | -0.011 |
ches | -0.010 |
mus | -0.010 |
Token | big |
Token ID | 1263 |
Feature activation | +2.16e-02 |
Pos logit contributions | |
ps | +0.007 |
ss | +0.006 |
ches | +0.006 |
teeth | +0.005 |
mus | +0.005 |
Neg logit contributions | |
she | -0.007 |
weekend | -0.007 |
day | -0.007 |
stair | -0.007 |
hallway | -0.007 |
Token | tree |
Token ID | 5509 |
Feature activation | +9.35e-02 |
Pos logit contributions | |
she | +0.031 |
weekend | +0.030 |
days | +0.029 |
hallway | +0.029 |
day | +0.028 |
Neg logit contributions | |
ps | -0.037 |
teeth | -0.033 |
ss | -0.032 |
ches | -0.031 |
mus | -0.030 |
Token | with |
Token ID | 351 |
Feature activation | +1.07e-03 |
Pos logit contributions | |
holiday | +0.218 |
occasion | +0.218 |
weekend | +0.216 |
awaited | +0.216 |
Katie | +0.209 |
Neg logit contributions | |
ps | -0.062 |
ches | -0.045 |
softly | -0.037 |
ows | -0.035 |
. | -0.034 |
Token | green |
Token ID | 4077 |
Feature activation | +2.16e-04 |
Pos logit contributions | |
she | +0.002 |
weekend | +0.002 |
day | +0.002 |
hallway | +0.002 |
stair | +0.002 |
Neg logit contributions | |
ps | -0.002 |
ss | -0.001 |
teeth | -0.001 |
ches | -0.001 |
mus | -0.001 |
Token | leaves |
Token ID | 5667 |
Feature activation | +4.29e-02 |
Pos logit contributions | |
she | +0.000 |
weekend | +0.000 |
day | +0.000 |
hallway | +0.000 |
stair | +0.000 |
Neg logit contributions | |
ps | -0.000 |
ss | -0.000 |
teeth | -0.000 |
ches | -0.000 |
mus | -0.000 |
Token | . |
Token ID | 13 |
Feature activation | +1.20e-03 |
Pos logit contributions | |
weekend | +0.093 |
hallway | +0.093 |
days | +0.091 |
holiday | +0.090 |
particular | +0.090 |
Neg logit contributions | |
ps | -0.051 |
ches | -0.035 |
ss | -0.032 |
teeth | -0.032 |
oni | -0.029 |
Token | They |
Token ID | 1119 |
Feature activation | +3.48e-02 |
Pos logit contributions | |
she | +0.002 |
weekend | +0.001 |
day | +0.001 |
hallway | +0.001 |
stair | +0.001 |
Neg logit contributions | |
ps | -0.002 |
ss | -0.002 |
teeth | -0.002 |
ches | -0.002 |
mus | -0.002 |
Token | , |
Token ID | 11 |
Feature activation | +4.43e-02 |
Pos logit contributions | |
weekend | +0.026 |
holiday | +0.025 |
hallway | +0.025 |
stair | +0.024 |
day | +0.024 |
Neg logit contributions | |
ps | -0.033 |
ss | -0.029 |
teeth | -0.028 |
ches | -0.027 |
mus | -0.026 |
Token | there |
Token ID | 612 |
Feature activation | +4.39e-03 |
Pos logit contributions | |
occasion | +0.057 |
weekend | +0.057 |
day | +0.052 |
particular | +0.052 |
holiday | +0.051 |
Neg logit contributions | |
ps | -0.083 |
teeth | -0.074 |
ss | -0.071 |
louder | -0.068 |
mus | -0.066 |
Token | was |
Token ID | 373 |
Feature activation | -1.02e-03 |
Pos logit contributions | |
she | +0.005 |
weekend | +0.005 |
hallway | +0.005 |
stair | +0.005 |
day | +0.005 |
Neg logit contributions | |
ps | -0.009 |
ss | -0.008 |
teeth | -0.008 |
ches | -0.008 |
mus | -0.008 |
Token | a |
Token ID | 257 |
Feature activation | +1.49e-03 |
Pos logit contributions | |
ps | +0.002 |
ss | +0.001 |
ches | +0.001 |
teeth | +0.001 |
mus | +0.001 |
Neg logit contributions | |
she | -0.002 |
weekend | -0.002 |
day | -0.002 |
hallway | -0.002 |
stair | -0.002 |
Token | small |
Token ID | 1402 |
Feature activation | +2.92e-02 |
Pos logit contributions | |
she | +0.002 |
weekend | +0.002 |
day | +0.002 |
hallway | +0.002 |
stair | +0.002 |
Neg logit contributions | |
ps | -0.002 |
ss | -0.002 |
teeth | -0.002 |
ches | -0.002 |
mus | -0.002 |
Token | room |
Token ID | 2119 |
Feature activation | +8.89e-02 |
Pos logit contributions | |
weekend | +0.043 |
she | +0.041 |
days | +0.040 |
day | +0.038 |
hallway | +0.038 |
Neg logit contributions | |
ps | -0.048 |
teeth | -0.044 |
ss | -0.042 |
mus | -0.039 |
ches | -0.038 |
Token | with |
Token ID | 351 |
Feature activation | -3.78e-03 |
Pos logit contributions | |
weekend | +0.194 |
where | +0.184 |
wander | +0.183 |
days | +0.180 |
particular | +0.180 |
Neg logit contributions | |
ps | -0.072 |
ows | -0.051 |
ss | -0.047 |
teeth | -0.044 |
ers | -0.041 |
Token | many |
Token ID | 867 |
Feature activation | -7.14e-03 |
Pos logit contributions | |
ps | +0.006 |
ss | +0.005 |
ches | +0.005 |
teeth | +0.005 |
mus | +0.005 |
Neg logit contributions | |
she | -0.007 |
weekend | -0.006 |
day | -0.006 |
hallway | -0.006 |
stair | -0.006 |
Token | things |
Token ID | 1243 |
Feature activation | +3.81e-02 |
Pos logit contributions | |
ps | +0.011 |
ss | +0.010 |
ches | +0.009 |
mus | +0.009 |
teeth | +0.009 |
Neg logit contributions | |
she | -0.012 |
day | -0.012 |
hallway | -0.012 |
stair | -0.012 |
weekend | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | -4.98e-03 |
Pos logit contributions | |
weekend | +0.072 |
particular | +0.071 |
hallway | +0.071 |
stair | +0.070 |
where | +0.069 |
Neg logit contributions | |
ps | -0.052 |
ss | -0.039 |
teeth | -0.038 |
ches | -0.038 |
mus | -0.035 |
Token | There |
Token ID | 1318 |
Feature activation | +3.93e-02 |
Pos logit contributions | |
ps | +0.011 |
ss | +0.010 |
ches | +0.009 |
teeth | +0.009 |
mus | +0.009 |
Neg logit contributions | |
she | -0.006 |
weekend | -0.005 |
stair | -0.005 |
hallway | -0.005 |
day | -0.005 |
Token | handed |
Token ID | 10158 |
Feature activation | -3.45e-02 |
Pos logit contributions | |
ps | +0.077 |
ss | +0.070 |
unnecessary | +0.070 |
rier | +0.065 |
mus | +0.065 |
Neg logit contributions | |
she | -0.041 |
they | -0.037 |
hallway | -0.034 |
weekend | -0.034 |
he | -0.034 |
Token | Santa |
Token ID | 8909 |
Feature activation | -1.63e-02 |
Pos logit contributions | |
ps | +0.069 |
ss | +0.062 |
ches | +0.059 |
teeth | +0.059 |
iddles | +0.057 |
Neg logit contributions | |
she | -0.049 |
weekend | -0.045 |
day | -0.043 |
they | -0.043 |
holiday | -0.043 |
Token | a |
Token ID | 257 |
Feature activation | -1.17e-02 |
Pos logit contributions | |
ps | +0.028 |
ss | +0.025 |
teeth | +0.023 |
ches | +0.023 |
mus | +0.023 |
Neg logit contributions | |
she | -0.025 |
weekend | -0.023 |
day | -0.023 |
holiday | -0.022 |
hallway | -0.022 |
Token | pair |
Token ID | 5166 |
Feature activation | +5.69e-02 |
Pos logit contributions | |
ps | +0.020 |
ss | +0.017 |
ches | +0.016 |
iddles | +0.016 |
mus | +0.016 |
Neg logit contributions | |
she | -0.019 |
weekend | -0.018 |
day | -0.018 |
hallway | -0.018 |
stair | -0.018 |
Token | of |
Token ID | 286 |
Feature activation | -1.36e-02 |
Pos logit contributions | |
sandbox | +0.066 |
hallway | +0.065 |
stair | +0.060 |
driveway | +0.060 |
weekend | +0.058 |
Neg logit contributions | |
ps | -0.118 |
ches | -0.109 |
ss | -0.106 |
mus | -0.099 |
rier | -0.097 |
Token | scissors |
Token ID | 41786 |
Feature activation | +2.35e-02 |
Pos logit contributions | |
ps | +0.021 |
ss | +0.018 |
ches | +0.017 |
mus | +0.016 |
teeth | +0.016 |
Neg logit contributions | |
she | -0.024 |
day | -0.023 |
weekend | -0.023 |
hallway | -0.022 |
stair | -0.022 |
Token | . |
Token ID | 13 |
Feature activation | -3.36e-02 |
Pos logit contributions | |
weekend | +0.040 |
hallway | +0.040 |
stair | +0.039 |
day | +0.039 |
holiday | +0.038 |
Neg logit contributions | |
ps | -0.037 |
ches | -0.030 |
ss | -0.030 |
teeth | -0.028 |
mus | -0.027 |
Token |
|
Token ID | 198 |
Feature activation | -3.77e-02 |
Pos logit contributions | |
ps | +0.072 |
ss | +0.065 |
ches | +0.062 |
wy | +0.060 |
iddles | +0.059 |
Neg logit contributions | |
she | -0.044 |
weekend | -0.039 |
he | -0.038 |
they | -0.038 |
holiday | -0.038 |
Token |
|
Token ID | 198 |
Feature activation | -9.39e-03 |
Pos logit contributions | |
ps | +0.094 |
teeth | +0.082 |
ss | +0.081 |
ches | +0.079 |
wy | +0.078 |
Neg logit contributions | |
she | -0.037 |
weekend | -0.033 |
they | -0.031 |
day | -0.031 |
he | -0.030 |
Token | " |
Token ID | 1 |
Feature activation | +1.67e-02 |
Pos logit contributions | |
ps | +0.017 |
ss | +0.015 |
teeth | +0.015 |
ches | +0.014 |
iddles | +0.014 |
Neg logit contributions | |
she | -0.014 |
weekend | -0.013 |
hallway | -0.013 |
day | -0.013 |
stair | -0.013 |
Token | Thank |
Token ID | 10449 |
Feature activation | +1.60e-04 |
Pos logit contributions | |
stair | +0.022 |
she | +0.022 |
hallway | +0.022 |
day | +0.021 |
weekend | +0.021 |
Neg logit contributions | |
ps | -0.031 |
ss | -0.028 |
mus | -0.027 |
ches | -0.026 |
iddles | -0.025 |
Token | toy |
Token ID | 13373 |
Feature activation | +1.72e-02 |
Pos logit contributions | |
she | +0.025 |
weekend | +0.024 |
hallway | +0.023 |
stair | +0.023 |
days | +0.022 |
Neg logit contributions | |
ps | -0.025 |
teeth | -0.022 |
ss | -0.021 |
ches | -0.020 |
mus | -0.019 |
Token | was |
Token ID | 373 |
Feature activation | -3.73e-03 |
Pos logit contributions | |
weekend | +0.023 |
she | +0.022 |
hallway | +0.021 |
stair | +0.021 |
particular | +0.021 |
Neg logit contributions | |
ps | -0.032 |
teeth | -0.028 |
ss | -0.028 |
ches | -0.028 |
mus | -0.026 |
Token | broken |
Token ID | 5445 |
Feature activation | -7.09e-03 |
Pos logit contributions | |
ps | +0.005 |
ss | +0.005 |
ches | +0.004 |
teeth | +0.004 |
mus | +0.004 |
Neg logit contributions | |
she | -0.007 |
weekend | -0.007 |
day | -0.007 |
hallway | -0.006 |
stair | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | -2.67e-03 |
Pos logit contributions | |
ps | +0.011 |
ss | +0.010 |
ches | +0.009 |
mus | +0.009 |
teeth | +0.009 |
Neg logit contributions | |
she | -0.012 |
day | -0.012 |
weekend | -0.012 |
stair | -0.011 |
hallway | -0.011 |
Token | The |
Token ID | 383 |
Feature activation | +3.26e-02 |
Pos logit contributions | |
ps | +0.005 |
ss | +0.005 |
ches | +0.005 |
teeth | +0.005 |
mus | +0.004 |
Neg logit contributions | |
she | -0.003 |
weekend | -0.003 |
day | -0.003 |
hallway | -0.003 |
stair | -0.003 |
Token | weather |
Token ID | 6193 |
Feature activation | +3.23e-02 |
Pos logit contributions | |
days | +0.056 |
she | +0.055 |
weekend | +0.053 |
stair | +0.052 |
lit | +0.051 |
Neg logit contributions | |
ps | -0.048 |
teeth | -0.041 |
ss | -0.038 |
ches | -0.033 |
ache | -0.032 |
Token | was |
Token ID | 373 |
Feature activation | -7.58e-03 |
Pos logit contributions | |
weekend | +0.042 |
she | +0.041 |
stair | +0.041 |
particular | +0.040 |
sandbox | +0.040 |
Neg logit contributions | |
ps | -0.059 |
ches | -0.052 |
ss | -0.051 |
teeth | -0.050 |
mus | -0.045 |
Token | sunny |
Token ID | 27737 |
Feature activation | +5.73e-02 |
Pos logit contributions | |
ps | +0.011 |
ss | +0.010 |
ches | +0.009 |
teeth | +0.009 |
mus | +0.009 |
Neg logit contributions | |
she | -0.013 |
weekend | -0.013 |
day | -0.013 |
hallway | -0.013 |
stair | -0.013 |
Token | and |
Token ID | 290 |
Feature activation | +1.64e-02 |
Pos logit contributions | |
particular | +0.121 |
wander | +0.118 |
Coming | +0.115 |
days | +0.113 |
contain | +0.113 |
Neg logit contributions | |
ps | -0.056 |
teeth | -0.040 |
ches | -0.036 |
better | -0.034 |
ers | -0.033 |
Token | all |
Token ID | 477 |
Feature activation | +2.67e-02 |
Pos logit contributions | |
weekend | +0.023 |
stair | +0.023 |
she | +0.022 |
hallway | +0.022 |
days | +0.022 |
Neg logit contributions | |
ps | -0.029 |
teeth | -0.025 |
ss | -0.025 |
ches | -0.024 |
mus | -0.023 |
Token | his |
Token ID | 465 |
Feature activation | +5.47e-04 |
Pos logit contributions | |
stair | +0.035 |
she | +0.035 |
weekend | +0.035 |
hallway | +0.034 |
sandbox | +0.033 |
Neg logit contributions | |
ps | -0.049 |
ss | -0.041 |
ches | -0.041 |
teeth | -0.040 |
mus | -0.038 |
Token | saw |
Token ID | 2497 |
Feature activation | +1.06e-03 |
Pos logit contributions | |
weekend | +0.004 |
she | +0.004 |
day | +0.004 |
hallway | +0.004 |
stair | +0.004 |
Neg logit contributions | |
ps | -0.005 |
ss | -0.004 |
teeth | -0.004 |
ches | -0.004 |
mus | -0.004 |
Token | that |
Token ID | 326 |
Feature activation | +3.67e-02 |
Pos logit contributions | |
she | +0.002 |
weekend | +0.002 |
day | +0.002 |
hallway | +0.002 |
stair | +0.002 |
Neg logit contributions | |
ps | -0.002 |
ss | -0.002 |
teeth | -0.001 |
ches | -0.001 |
mus | -0.001 |
Token | it |
Token ID | 340 |
Feature activation | -1.07e-02 |
Pos logit contributions | |
hallway | +0.060 |
weekend | +0.058 |
days | +0.058 |
driveway | +0.057 |
occasion | +0.057 |
Neg logit contributions | |
ps | -0.054 |
teeth | -0.051 |
louder | -0.044 |
ss | -0.043 |
smarter | -0.041 |
Token | was |
Token ID | 373 |
Feature activation | -2.30e-02 |
Pos logit contributions | |
ps | +0.024 |
ss | +0.022 |
mus | +0.021 |
iddles | +0.021 |
teeth | +0.020 |
Neg logit contributions | |
she | -0.011 |
day | -0.010 |
hallway | -0.010 |
stair | -0.010 |
weekend | -0.010 |
Token | a |
Token ID | 257 |
Feature activation | +1.56e-02 |
Pos logit contributions | |
ps | +0.039 |
ss | +0.037 |
ches | +0.035 |
iddles | +0.034 |
mus | +0.034 |
Neg logit contributions | |
she | -0.036 |
day | -0.034 |
sunny | -0.033 |
weekend | -0.032 |
they | -0.032 |
Token | metal |
Token ID | 6147 |
Feature activation | +3.01e-02 |
Pos logit contributions | |
she | +0.025 |
weekend | +0.025 |
hallway | +0.024 |
days | +0.024 |
day | +0.024 |
Neg logit contributions | |
ps | -0.024 |
ss | -0.021 |
teeth | -0.020 |
ches | -0.020 |
mus | -0.019 |
Token | box |
Token ID | 3091 |
Feature activation | +5.91e-02 |
Pos logit contributions | |
she | +0.046 |
weekend | +0.046 |
days | +0.044 |
hallway | +0.043 |
particular | +0.042 |
Neg logit contributions | |
ps | -0.052 |
ss | -0.044 |
teeth | -0.044 |
ches | -0.041 |
rier | -0.040 |
Token | with |
Token ID | 351 |
Feature activation | -1.25e-02 |
Pos logit contributions | |
weekend | +0.112 |
days | +0.107 |
hallway | +0.106 |
sandbox | +0.105 |
stair | +0.104 |
Neg logit contributions | |
ps | -0.080 |
ches | -0.065 |
ss | -0.061 |
rier | -0.054 |
oni | -0.054 |
Token | a |
Token ID | 257 |
Feature activation | -1.48e-02 |
Pos logit contributions | |
ps | +0.020 |
ss | +0.018 |
ches | +0.017 |
mus | +0.016 |
iddles | +0.016 |
Neg logit contributions | |
she | -0.021 |
day | -0.020 |
weekend | -0.019 |
hallway | -0.019 |
stair | -0.019 |
Token | lock |
Token ID | 5793 |
Feature activation | +3.57e-02 |
Pos logit contributions | |
ps | +0.027 |
ss | +0.023 |
ches | +0.023 |
iddles | +0.022 |
mus | +0.022 |
Neg logit contributions | |
she | -0.022 |
stair | -0.021 |
hallway | -0.021 |
day | -0.021 |
weekend | -0.020 |
Token | . |
Token ID | 13 |
Feature activation | -1.19e-02 |
Pos logit contributions | |
weekend | +0.066 |
days | +0.065 |
hallway | +0.063 |
she | +0.063 |
particular | +0.062 |
Neg logit contributions | |
ps | -0.052 |
ches | -0.041 |
ss | -0.041 |
teeth | -0.037 |
mus | -0.035 |
Token | It |
Token ID | 632 |
Feature activation | -1.31e-03 |
Pos logit contributions | |
weekend | +0.052 |
holiday | +0.052 |
hallway | +0.051 |
stair | +0.050 |
days | +0.049 |
Neg logit contributions | |
ps | -0.052 |
ss | -0.048 |
teeth | -0.046 |
iddles | -0.046 |
ches | -0.044 |
Token | was |
Token ID | 373 |
Feature activation | -2.07e-02 |
Pos logit contributions | |
ps | +0.003 |
ss | +0.002 |
ches | +0.002 |
teeth | +0.002 |
mus | +0.002 |
Neg logit contributions | |
she | -0.002 |
weekend | -0.001 |
day | -0.001 |
hallway | -0.001 |
stair | -0.001 |
Token | made |
Token ID | 925 |
Feature activation | +1.07e-02 |
Pos logit contributions | |
ps | +0.037 |
ss | +0.032 |
ealing | +0.030 |
ches | +0.030 |
teeth | +0.029 |
Neg logit contributions | |
she | -0.033 |
day | -0.032 |
weekend | -0.032 |
sunny | -0.030 |
particular | -0.030 |
Token | of |
Token ID | 286 |
Feature activation | -2.31e-02 |
Pos logit contributions | |
hallway | +0.006 |
weekend | +0.006 |
she | +0.006 |
stair | +0.005 |
day | +0.005 |
Neg logit contributions | |
ps | -0.028 |
ss | -0.026 |
ches | -0.025 |
teeth | -0.025 |
mus | -0.025 |
Token | wood |
Token ID | 4898 |
Feature activation | -1.40e-02 |
Pos logit contributions | |
ps | +0.038 |
ss | +0.034 |
ches | +0.033 |
mus | +0.030 |
ealing | +0.030 |
Neg logit contributions | |
she | -0.038 |
day | -0.037 |
hallway | -0.036 |
weekend | -0.035 |
stair | -0.035 |
Token | and |
Token ID | 290 |
Feature activation | +2.61e-02 |
Pos logit contributions | |
ps | +0.026 |
ss | +0.023 |
teeth | +0.022 |
ches | +0.022 |
mus | +0.022 |
Neg logit contributions | |
she | -0.020 |
day | -0.019 |
weekend | -0.018 |
hallway | -0.018 |
stair | -0.018 |
Token | had |
Token ID | 550 |
Feature activation | -1.30e-02 |
Pos logit contributions | |
weekend | +0.041 |
days | +0.039 |
stair | +0.038 |
holiday | +0.038 |
hallway | +0.038 |
Neg logit contributions | |
ps | -0.043 |
teeth | -0.037 |
ss | -0.035 |
ches | -0.034 |
mus | -0.034 |
Token | big |
Token ID | 1263 |
Feature activation | -3.40e-02 |
Pos logit contributions | |
ps | +0.021 |
ss | +0.018 |
ches | +0.017 |
mus | +0.017 |
teeth | +0.016 |
Neg logit contributions | |
she | -0.022 |
day | -0.021 |
weekend | -0.021 |
hallway | -0.021 |
stair | -0.020 |
Token | strings |
Token ID | 13042 |
Feature activation | +1.99e-02 |
Pos logit contributions | |
ps | +0.056 |
ss | +0.052 |
ches | +0.049 |
wy | +0.048 |
mus | +0.047 |
Neg logit contributions | |
she | -0.054 |
day | -0.053 |
hallway | -0.052 |
stair | -0.051 |
explore | -0.049 |
Token | . |
Token ID | 13 |
Feature activation | +2.16e-02 |
Pos logit contributions | |
weekend | +0.034 |
hallway | +0.033 |
stair | +0.033 |
holiday | +0.032 |
she | +0.032 |
Neg logit contributions | |
ps | -0.031 |
ches | -0.025 |
ss | -0.025 |
teeth | -0.024 |
mus | -0.023 |
Token |
|
Token ID | 198 |
Feature activation | +2.42e-02 |
Pos logit contributions | |
weekend | +0.027 |
holiday | +0.026 |
hallway | +0.026 |
she | +0.026 |
stair | +0.026 |
Neg logit contributions | |
ps | -0.040 |
ss | -0.037 |
teeth | -0.036 |
iddles | -0.035 |
ches | -0.035 |
Token | my |
Token ID | 1820 |
Feature activation | -3.43e-02 |
Pos logit contributions | |
ps | +0.085 |
mus | +0.074 |
rier | +0.073 |
ss | +0.072 |
unnecessary | +0.071 |
Neg logit contributions | |
she | -0.028 |
day | -0.026 |
weekend | -0.022 |
he | -0.019 |
wander | -0.019 |
Token | said |
Token ID | 531 |
Feature activation | -1.02e-02 |
Pos logit contributions | |
ps | +0.076 |
ss | +0.065 |
unnecessary | +0.065 |
mus | +0.063 |
iddles | +0.063 |
Neg logit contributions | |
she | -0.045 |
he | -0.038 |
day | -0.038 |
they | -0.038 |
weekend | -0.033 |
Token | . |
Token ID | 13 |
Feature activation | -9.54e-03 |
Pos logit contributions | |
ps | +0.017 |
ss | +0.015 |
ches | +0.014 |
teeth | +0.014 |
mus | +0.014 |
Neg logit contributions | |
she | -0.017 |
day | -0.016 |
weekend | -0.016 |
they | -0.015 |
hallway | -0.015 |
Token | " |
Token ID | 366 |
Feature activation | +2.71e-02 |
Pos logit contributions | |
ps | +0.017 |
ss | +0.015 |
ches | +0.014 |
teeth | +0.014 |
mus | +0.014 |
Neg logit contributions | |
she | -0.015 |
weekend | -0.013 |
day | -0.013 |
hallway | -0.013 |
stair | -0.013 |
Token | I |
Token ID | 40 |
Feature activation | +8.44e-03 |
Pos logit contributions | |
stair | +0.033 |
hallway | +0.031 |
holiday | +0.031 |
she | +0.031 |
weekend | +0.031 |
Neg logit contributions | |
ps | -0.053 |
ss | -0.049 |
mus | -0.047 |
iddles | -0.045 |
ches | -0.044 |
Token | 'll |
Token ID | 1183 |
Feature activation | +2.59e-02 |
Pos logit contributions | |
she | +0.013 |
stair | +0.013 |
weekend | +0.013 |
hallway | +0.013 |
day | +0.012 |
Neg logit contributions | |
ps | -0.014 |
ss | -0.012 |
ches | -0.012 |
teeth | -0.012 |
mus | -0.011 |
Token | help |
Token ID | 1037 |
Feature activation | -2.15e-02 |
Pos logit contributions | |
weekend | +0.042 |
stair | +0.041 |
hallway | +0.040 |
sandbox | +0.038 |
day | +0.038 |
Neg logit contributions | |
ps | -0.045 |
ches | -0.036 |
ss | -0.036 |
teeth | -0.035 |
uable | -0.033 |
Token | you |
Token ID | 345 |
Feature activation | -8.02e-03 |
Pos logit contributions | |
ps | +0.052 |
ss | +0.047 |
teeth | +0.045 |
iddles | +0.045 |
ealing | +0.045 |
Neg logit contributions | |
she | -0.023 |
day | -0.018 |
they | -0.018 |
he | -0.017 |
explore | -0.017 |
Token | find |
Token ID | 1064 |
Feature activation | -4.37e-02 |
Pos logit contributions | |
ps | +0.018 |
ss | +0.016 |
teeth | +0.015 |
ches | +0.015 |
mus | +0.015 |
Neg logit contributions | |
she | -0.009 |
day | -0.008 |
weekend | -0.008 |
hallway | -0.008 |
stair | -0.008 |
Token | it |
Token ID | 340 |
Feature activation | -4.92e-02 |
Pos logit contributions | |
ps | +0.102 |
iddles | +0.099 |
ss | +0.098 |
ussy | +0.097 |
ches | +0.093 |
Neg logit contributions | |
she | -0.045 |
day | -0.036 |
he | -0.036 |
they | -0.035 |
holiday | -0.031 |
Token | ." |
Token ID | 526 |
Feature activation | +1.59e-02 |
Pos logit contributions | |
ps | +0.085 |
ss | +0.082 |
ealing | +0.074 |
iddles | +0.073 |
mus | +0.072 |
Neg logit contributions | |
she | -0.083 |
day | -0.077 |
weekend | -0.076 |
they | -0.075 |
he | -0.073 |
Token | the |
Token ID | 262 |
Feature activation | +1.41e-03 |
Pos logit contributions | |
ps | +0.072 |
ss | +0.069 |
ches | +0.069 |
ealing | +0.067 |
ows | +0.065 |
Neg logit contributions | |
day | -0.036 |
she | -0.035 |
weekend | -0.032 |
he | -0.032 |
holiday | -0.032 |
Token | house |
Token ID | 2156 |
Feature activation | -6.28e-03 |
Pos logit contributions | |
she | +0.002 |
weekend | +0.002 |
stair | +0.002 |
hallway | +0.002 |
day | +0.002 |
Neg logit contributions | |
ps | -0.003 |
ss | -0.002 |
teeth | -0.002 |
ches | -0.002 |
mus | -0.002 |
Token | . |
Token ID | 13 |
Feature activation | -2.61e-02 |
Pos logit contributions | |
ps | +0.011 |
ss | +0.010 |
ches | +0.009 |
teeth | +0.009 |
mus | +0.009 |
Neg logit contributions | |
she | -0.010 |
weekend | -0.009 |
day | -0.009 |
hallway | -0.009 |
stair | -0.009 |
Token | But |
Token ID | 887 |
Feature activation | +1.35e-02 |
Pos logit contributions | |
ps | +0.067 |
wy | +0.058 |
ches | +0.058 |
ss | +0.057 |
mus | +0.056 |
Neg logit contributions | |
she | -0.024 |
day | -0.023 |
he | -0.022 |
weekend | -0.021 |
hallway | -0.020 |
Token | his |
Token ID | 465 |
Feature activation | -8.80e-04 |
Pos logit contributions | |
hallway | +0.018 |
stair | +0.018 |
weekend | +0.018 |
she | +0.017 |
days | +0.017 |
Neg logit contributions | |
ps | -0.024 |
teeth | -0.022 |
ss | -0.021 |
mus | -0.020 |
ches | -0.020 |
Token | parents |
Token ID | 3397 |
Feature activation | +3.62e-03 |
Pos logit contributions | |
ps | +0.002 |
ss | +0.001 |
ches | +0.001 |
teeth | +0.001 |
mus | +0.001 |
Neg logit contributions | |
she | -0.001 |
weekend | -0.001 |
day | -0.001 |
hallway | -0.001 |
stair | -0.001 |
Token | explained |
Token ID | 4893 |
Feature activation | -1.53e-02 |
Pos logit contributions | |
she | +0.005 |
weekend | +0.005 |
hallway | +0.005 |
stair | +0.005 |
day | +0.005 |
Neg logit contributions | |
ps | -0.006 |
ss | -0.006 |
ches | -0.005 |
teeth | -0.005 |
mus | -0.005 |
Token | that |
Token ID | 326 |
Feature activation | +2.70e-02 |
Pos logit contributions | |
ps | +0.037 |
ss | +0.033 |
ches | +0.032 |
mus | +0.031 |
teeth | +0.031 |
Neg logit contributions | |
she | -0.015 |
they | -0.013 |
day | -0.013 |
weekend | -0.013 |
he | -0.012 |
Token | he |
Token ID | 339 |
Feature activation | -2.82e-02 |
Pos logit contributions | |
hallway | +0.034 |
weekend | +0.033 |
driveway | +0.032 |
stair | +0.032 |
particular | +0.031 |
Neg logit contributions | |
ps | -0.051 |
teeth | -0.047 |
ss | -0.045 |
mus | -0.043 |
ches | -0.042 |
Token | was |
Token ID | 373 |
Feature activation | -3.97e-02 |
Pos logit contributions | |
ps | +0.065 |
rier | +0.057 |
mus | +0.056 |
ss | +0.056 |
iddles | +0.054 |
Neg logit contributions | |
she | -0.030 |
day | -0.028 |
they | -0.026 |
he | -0.026 |
weekend | -0.025 |
Token | too |
Token ID | 1165 |
Feature activation | -3.68e-02 |
Pos logit contributions | |
ps | +0.069 |
ss | +0.063 |
mus | +0.058 |
teeth | +0.058 |
osaurus | +0.058 |
Neg logit contributions | |
she | -0.064 |
day | -0.062 |
they | -0.059 |
weekend | -0.058 |
he | -0.058 |
Token | guitar |
Token ID | 10047 |
Feature activation | -1.03e-02 |
Pos logit contributions | |
ps | +0.055 |
ss | +0.052 |
mus | +0.048 |
iddles | +0.048 |
ches | +0.048 |
Neg logit contributions | |
day | -0.042 |
she | -0.040 |
weekend | -0.040 |
holiday | -0.038 |
stair | -0.038 |
Token | . |
Token ID | 13 |
Feature activation | -9.91e-03 |
Pos logit contributions | |
ps | +0.017 |
ss | +0.014 |
ches | +0.013 |
mus | +0.013 |
teeth | +0.013 |
Neg logit contributions | |
she | -0.017 |
day | -0.017 |
weekend | -0.016 |
stair | -0.016 |
hallway | -0.016 |
Token | He |
Token ID | 679 |
Feature activation | -8.96e-03 |
Pos logit contributions | |
ps | +0.022 |
ss | +0.019 |
ches | +0.019 |
mus | +0.018 |
teeth | +0.018 |
Neg logit contributions | |
she | -0.012 |
day | -0.010 |
weekend | -0.010 |
stair | -0.010 |
hallway | -0.010 |
Token | forg |
Token ID | 5784 |
Feature activation | -4.39e-02 |
Pos logit contributions | |
ps | +0.014 |
ss | +0.012 |
mus | +0.011 |
ches | +0.011 |
teeth | +0.011 |
Neg logit contributions | |
she | -0.016 |
weekend | -0.015 |
hallway | -0.015 |
day | -0.015 |
stair | -0.015 |
Token | ets |
Token ID | 1039 |
Feature activation | -1.74e-03 |
Pos logit contributions | |
ps | +0.075 |
rier | +0.067 |
osaurus | +0.065 |
teeth | +0.064 |
ss | +0.064 |
Neg logit contributions | |
she | -0.067 |
day | -0.062 |
of | -0.058 |
wander | -0.057 |
weekend | -0.056 |
Token | about |
Token ID | 546 |
Feature activation | -3.56e-02 |
Pos logit contributions | |
ps | +0.003 |
ss | +0.003 |
ches | +0.003 |
teeth | +0.003 |
mus | +0.003 |
Neg logit contributions | |
she | -0.003 |
weekend | -0.002 |
day | -0.002 |
hallway | -0.002 |
stair | -0.002 |
Token | the |
Token ID | 262 |
Feature activation | -9.75e-03 |
Pos logit contributions | |
ps | +0.076 |
ss | +0.069 |
ealing | +0.069 |
ches | +0.068 |
rier | +0.065 |
Neg logit contributions | |
she | -0.044 |
day | -0.040 |
he | -0.040 |
they | -0.038 |
stair | -0.036 |
Token | stove |
Token ID | 27635 |
Feature activation | +5.74e-03 |
Pos logit contributions | |
ps | +0.017 |
ss | +0.015 |
ches | +0.014 |
teeth | +0.014 |
mus | +0.014 |
Neg logit contributions | |
she | -0.015 |
day | -0.015 |
weekend | -0.015 |
stair | -0.014 |
hallway | -0.014 |
Token | and |
Token ID | 290 |
Feature activation | +2.27e-02 |
Pos logit contributions | |
she | +0.010 |
weekend | +0.010 |
hallway | +0.010 |
stair | +0.010 |
day | +0.009 |
Neg logit contributions | |
ps | -0.009 |
ss | -0.007 |
ches | -0.007 |
teeth | -0.007 |
mus | -0.007 |
Token | the |
Token ID | 262 |
Feature activation | -1.21e-02 |
Pos logit contributions | |
weekend | +0.039 |
stair | +0.037 |
hallway | +0.036 |
particular | +0.036 |
days | +0.036 |
Neg logit contributions | |
ps | -0.037 |
teeth | -0.032 |
ss | -0.029 |
ches | -0.027 |
mus | -0.026 |
Token | to |
Token ID | 284 |
Feature activation | -6.32e-03 |
Pos logit contributions | |
ps | +0.021 |
ss | +0.018 |
ches | +0.018 |
mus | +0.017 |
teeth | +0.017 |
Neg logit contributions | |
she | -0.019 |
weekend | -0.018 |
day | -0.018 |
stair | -0.018 |
hallway | -0.018 |
Token | park |
Token ID | 3952 |
Feature activation | +4.56e-03 |
Pos logit contributions | |
ps | +0.011 |
ss | +0.010 |
ches | +0.009 |
teeth | +0.009 |
mus | +0.009 |
Neg logit contributions | |
she | -0.003 |
weekend | -0.003 |
day | -0.003 |
hallway | -0.003 |
stair | -0.003 |
Token | . |
Token ID | 13 |
Feature activation | -1.06e-02 |
Pos logit contributions | |
she | +0.008 |
weekend | +0.007 |
stair | +0.007 |
hallway | +0.007 |
day | +0.007 |
Neg logit contributions | |
ps | -0.007 |
ss | -0.006 |
teeth | -0.006 |
ches | -0.006 |
mus | -0.006 |
Token |
|
Token ID | 198 |
Feature activation | +6.58e-04 |
Pos logit contributions | |
ps | +0.023 |
ss | +0.020 |
ches | +0.019 |
teeth | +0.019 |
mus | +0.018 |
Neg logit contributions | |
she | -0.013 |
weekend | -0.012 |
day | -0.012 |
hallway | -0.012 |
stair | -0.012 |
Token |
|
Token ID | 198 |
Feature activation | -1.48e-03 |
Pos logit contributions | |
she | +0.001 |
weekend | +0.001 |
day | +0.001 |
hallway | +0.001 |
stair | +0.001 |
Neg logit contributions | |
ps | -0.001 |
ss | -0.001 |
ches | -0.001 |
teeth | -0.001 |
mus | -0.001 |
Token | One |
Token ID | 3198 |
Feature activation | +3.64e-02 |
Pos logit contributions | |
ps | +0.003 |
ss | +0.003 |
ches | +0.003 |
teeth | +0.003 |
mus | +0.003 |
Neg logit contributions | |
she | -0.002 |
weekend | -0.002 |
day | -0.002 |
hallway | -0.002 |
stair | -0.002 |
Token | day |
Token ID | 1110 |
Feature activation | -4.85e-03 |
Pos logit contributions | |
driveway | +0.024 |
stack | +0.022 |
hallway | +0.022 |
sandbox | +0.021 |
stair | +0.021 |
Neg logit contributions | |
ss | -0.081 |
ps | -0.079 |
ches | -0.077 |
teeth | -0.076 |
mus | -0.076 |
Token | , |
Token ID | 11 |
Feature activation | -3.94e-03 |
Pos logit contributions | |
ps | +0.010 |
ss | +0.009 |
ches | +0.008 |
teeth | +0.008 |
mus | +0.008 |
Neg logit contributions | |
she | -0.006 |
day | -0.006 |
weekend | -0.006 |
stair | -0.006 |
hallway | -0.006 |
Token | a |
Token ID | 257 |
Feature activation | -2.18e-02 |
Pos logit contributions | |
ps | +0.008 |
ss | +0.007 |
ches | +0.007 |
teeth | +0.007 |
mus | +0.007 |
Neg logit contributions | |
she | -0.005 |
weekend | -0.005 |
day | -0.005 |
stair | -0.005 |
hallway | -0.005 |
Token | young |
Token ID | 1862 |
Feature activation | +2.99e-02 |
Pos logit contributions | |
ps | +0.043 |
ches | +0.034 |
ss | +0.034 |
teeth | +0.033 |
mus | +0.032 |
Neg logit contributions | |
she | -0.034 |
day | -0.033 |
weekend | -0.032 |
stair | -0.032 |
hallway | -0.031 |
Token | girl |
Token ID | 2576 |
Feature activation | +6.13e-02 |
Pos logit contributions | |
weekend | +0.042 |
hallway | +0.042 |
days | +0.042 |
stair | +0.041 |
she | +0.041 |
Neg logit contributions | |
ps | -0.046 |
ss | -0.043 |
teeth | -0.041 |
rier | -0.040 |
iddles | -0.039 |
Token | named |
Token ID | 3706 |
Feature activation | +2.28e-03 |
Pos logit contributions | |
stair | +0.106 |
weekend | +0.106 |
holiday | +0.102 |
particular | +0.100 |
sunny | +0.100 |
Neg logit contributions | |
ps | -0.079 |
ss | -0.075 |
ches | -0.072 |
teeth | -0.066 |
mus | -0.061 |
Token | any |
Token ID | 597 |
Feature activation | +1.36e-02 |
Pos logit contributions | |
she | +0.009 |
weekend | +0.009 |
hallway | +0.009 |
stair | +0.009 |
day | +0.009 |
Neg logit contributions | |
ps | -0.008 |
ss | -0.006 |
ches | -0.006 |
teeth | -0.006 |
mus | -0.006 |
Token | further |
Token ID | 2252 |
Feature activation | +2.14e-02 |
Pos logit contributions | |
she | +0.021 |
weekend | +0.021 |
hallway | +0.021 |
stair | +0.021 |
days | +0.020 |
Neg logit contributions | |
ps | -0.022 |
ss | -0.019 |
ches | -0.018 |
teeth | -0.018 |
mus | -0.017 |
Token | . |
Token ID | 13 |
Feature activation | +2.66e-03 |
Pos logit contributions | |
weekend | +0.037 |
stair | +0.036 |
hallway | +0.036 |
she | +0.036 |
days | +0.035 |
Neg logit contributions | |
ps | -0.032 |
ches | -0.025 |
ss | -0.025 |
teeth | -0.024 |
mus | -0.023 |
Token | He |
Token ID | 679 |
Feature activation | +4.22e-03 |
Pos logit contributions | |
she | +0.003 |
weekend | +0.003 |
day | +0.003 |
stair | +0.003 |
hallway | +0.003 |
Neg logit contributions | |
ps | -0.005 |
ss | -0.005 |
teeth | -0.004 |
ches | -0.004 |
mus | -0.004 |
Token | had |
Token ID | 550 |
Feature activation | +3.19e-03 |
Pos logit contributions | |
she | +0.005 |
weekend | +0.005 |
stair | +0.005 |
day | +0.005 |
hallway | +0.005 |
Neg logit contributions | |
ps | -0.008 |
ss | -0.007 |
ches | -0.007 |
teeth | -0.007 |
mus | -0.007 |
Token | gotten |
Token ID | 7891 |
Feature activation | -1.48e-02 |
Pos logit contributions | |
she | +0.004 |
weekend | +0.004 |
stair | +0.004 |
hallway | +0.004 |
day | +0.004 |
Neg logit contributions | |
ps | -0.006 |
ss | -0.005 |
ches | -0.005 |
teeth | -0.005 |
mus | -0.005 |
Token | so |
Token ID | 523 |
Feature activation | +1.44e-02 |
Pos logit contributions | |
ps | +0.024 |
ss | +0.021 |
ches | +0.020 |
teeth | +0.019 |
iddles | +0.019 |
Neg logit contributions | |
she | -0.025 |
day | -0.024 |
weekend | -0.023 |
they | -0.023 |
he | -0.023 |
Token | scared |
Token ID | 12008 |
Feature activation | +5.09e-02 |
Pos logit contributions | |
hallway | +0.026 |
stair | +0.026 |
weekend | +0.026 |
she | +0.025 |
day | +0.025 |
Neg logit contributions | |
ps | -0.020 |
ss | -0.017 |
teeth | -0.016 |
mus | -0.015 |
ealing | -0.015 |
Token | he |
Token ID | 339 |
Feature activation | -1.85e-03 |
Pos logit contributions | |
days | +0.090 |
hallway | +0.089 |
weekend | +0.088 |
stair | +0.086 |
driveway | +0.084 |
Neg logit contributions | |
ps | -0.075 |
teeth | -0.061 |
ss | -0.060 |
ches | -0.060 |
rier | -0.056 |
Token | had |
Token ID | 550 |
Feature activation | +8.04e-03 |
Pos logit contributions | |
ps | +0.004 |
ss | +0.003 |
ches | +0.003 |
teeth | +0.003 |
mus | +0.003 |
Neg logit contributions | |
she | -0.002 |
weekend | -0.002 |
day | -0.002 |
hallway | -0.002 |
stair | -0.002 |
Token | forgotten |
Token ID | 11564 |
Feature activation | -3.22e-04 |
Pos logit contributions | |
she | +0.011 |
weekend | +0.011 |
stair | +0.011 |
hallway | +0.011 |
day | +0.010 |
Neg logit contributions | |
ps | -0.015 |
ss | -0.013 |
teeth | -0.013 |
ches | -0.013 |
mus | -0.012 |
Token | their |
Token ID | 511 |
Feature activation | +2.16e-02 |
Pos logit contributions | |
she | +0.013 |
weekend | +0.013 |
stair | +0.013 |
hallway | +0.012 |
day | +0.012 |
Neg logit contributions | |
ps | -0.018 |
ss | -0.015 |
teeth | -0.015 |
mus | -0.014 |
ches | -0.014 |
Token | mom |
Token ID | 1995 |
Feature activation | +2.01e-02 |
Pos logit contributions | |
she | +0.031 |
weekend | +0.029 |
particular | +0.027 |
hallway | +0.027 |
stair | +0.027 |
Neg logit contributions | |
ps | -0.040 |
teeth | -0.035 |
ss | -0.032 |
ches | -0.030 |
iddles | -0.030 |
Token | , |
Token ID | 11 |
Feature activation | +2.36e-02 |
Pos logit contributions | |
weekend | +0.035 |
stair | +0.035 |
hallway | +0.034 |
holiday | +0.034 |
she | +0.033 |
Neg logit contributions | |
ps | -0.029 |
ches | -0.024 |
ss | -0.024 |
teeth | -0.023 |
mus | -0.021 |
Token | who |
Token ID | 508 |
Feature activation | +1.81e-02 |
Pos logit contributions | |
weekend | +0.034 |
stair | +0.033 |
day | +0.032 |
hallway | +0.032 |
holiday | +0.031 |
Neg logit contributions | |
ps | -0.042 |
ss | -0.036 |
teeth | -0.034 |
ches | -0.034 |
mus | -0.033 |
Token | is |
Token ID | 318 |
Feature activation | -2.16e-02 |
Pos logit contributions | |
weekend | +0.032 |
stair | +0.031 |
she | +0.031 |
hallway | +0.031 |
day | +0.030 |
Neg logit contributions | |
ps | -0.026 |
ss | -0.022 |
ches | -0.021 |
teeth | -0.020 |
mus | -0.019 |
Token | sitting |
Token ID | 5586 |
Feature activation | -2.80e-02 |
Pos logit contributions | |
ps | +0.041 |
ss | +0.038 |
ches | +0.038 |
teeth | +0.036 |
uable | +0.035 |
Neg logit contributions | |
she | -0.030 |
day | -0.028 |
weekend | -0.027 |
hallway | -0.027 |
holiday | -0.026 |
Token | on |
Token ID | 319 |
Feature activation | -1.39e-02 |
Pos logit contributions | |
ps | +0.058 |
ches | +0.054 |
ss | +0.054 |
uable | +0.052 |
teeth | +0.051 |
Neg logit contributions | |
she | -0.035 |
day | -0.033 |
weekend | -0.030 |
he | -0.029 |
hallway | -0.029 |
Token | a |
Token ID | 257 |
Feature activation | +2.49e-02 |
Pos logit contributions | |
ps | +0.024 |
ss | +0.021 |
ches | +0.020 |
mus | +0.019 |
teeth | +0.019 |
Neg logit contributions | |
she | -0.022 |
day | -0.021 |
weekend | -0.021 |
hallway | -0.020 |
holiday | -0.020 |
Token | bench |
Token ID | 7624 |
Feature activation | +1.56e-02 |
Pos logit contributions | |
she | +0.022 |
weekend | +0.021 |
days | +0.021 |
day | +0.020 |
where | +0.019 |
Neg logit contributions | |
ps | -0.055 |
ches | -0.051 |
ss | -0.048 |
teeth | -0.048 |
mus | -0.047 |
Token | . |
Token ID | 13 |
Feature activation | +2.37e-02 |
Pos logit contributions | |
weekend | +0.026 |
she | +0.026 |
stair | +0.026 |
hallway | +0.025 |
day | +0.025 |
Neg logit contributions | |
ps | -0.023 |
ss | -0.019 |
ches | -0.019 |
teeth | -0.018 |
mus | -0.018 |
Token | She |
Token ID | 1375 |
Feature activation | -4.69e-03 |
Pos logit contributions | |
weekend | +0.032 |
day | +0.031 |
stair | +0.031 |
hallway | +0.029 |
days | +0.029 |
Neg logit contributions | |
ps | -0.041 |
teeth | -0.039 |
ss | -0.038 |
iddles | -0.035 |
mus | -0.035 |
Token | 's |
Token ID | 338 |
Feature activation | +1.26e-02 |
Pos logit contributions | |
ps | +0.034 |
ss | +0.028 |
mus | +0.028 |
iddles | +0.028 |
teeth | +0.027 |
Neg logit contributions | |
she | -0.031 |
day | -0.030 |
weekend | -0.028 |
they | -0.028 |
he | -0.027 |
Token | car |
Token ID | 1097 |
Feature activation | -7.04e-03 |
Pos logit contributions | |
she | +0.018 |
weekend | +0.018 |
hallway | +0.018 |
stair | +0.018 |
day | +0.017 |
Neg logit contributions | |
ps | -0.022 |
ss | -0.019 |
teeth | -0.019 |
ches | -0.018 |
iddles | -0.017 |
Token | . |
Token ID | 13 |
Feature activation | +4.10e-03 |
Pos logit contributions | |
ps | +0.011 |
ss | +0.009 |
ches | +0.009 |
teeth | +0.009 |
mus | +0.009 |
Neg logit contributions | |
she | -0.012 |
weekend | -0.012 |
day | -0.012 |
hallway | -0.011 |
stair | -0.011 |
Token | They |
Token ID | 1119 |
Feature activation | +1.57e-03 |
Pos logit contributions | |
she | +0.005 |
weekend | +0.005 |
day | +0.005 |
stair | +0.005 |
hallway | +0.005 |
Neg logit contributions | |
ps | -0.008 |
ss | -0.007 |
teeth | -0.007 |
ches | -0.007 |
mus | -0.007 |
Token | thought |
Token ID | 1807 |
Feature activation | +2.32e-02 |
Pos logit contributions | |
she | +0.002 |
weekend | +0.002 |
day | +0.002 |
stair | +0.002 |
hallway | +0.002 |
Neg logit contributions | |
ps | -0.003 |
ss | -0.003 |
ches | -0.002 |
teeth | -0.002 |
mus | -0.002 |
Token | it |
Token ID | 340 |
Feature activation | -5.19e-02 |
Pos logit contributions | |
hallway | +0.029 |
weekend | +0.027 |
days | +0.027 |
stair | +0.026 |
driveway | +0.026 |
Neg logit contributions | |
ps | -0.047 |
teeth | -0.042 |
ss | -0.041 |
ches | -0.040 |
mus | -0.038 |
Token | was |
Token ID | 373 |
Feature activation | -4.84e-02 |
Pos logit contributions | |
ps | +0.141 |
rier | +0.133 |
iddles | +0.130 |
mus | +0.129 |
ags | +0.128 |
Neg logit contributions | |
day | -0.033 |
she | -0.031 |
they | -0.027 |
weekend | -0.024 |
he | -0.024 |
Token | a |
Token ID | 257 |
Feature activation | +4.91e-03 |
Pos logit contributions | |
ps | +0.084 |
mus | +0.077 |
ss | +0.077 |
ches | +0.075 |
iddles | +0.075 |
Neg logit contributions | |
she | -0.074 |
day | -0.073 |
they | -0.073 |
he | -0.070 |
sunny | -0.069 |
Token | toy |
Token ID | 13373 |
Feature activation | +2.44e-02 |
Pos logit contributions | |
she | +0.010 |
weekend | +0.009 |
hallway | +0.009 |
stair | +0.009 |
day | +0.009 |
Neg logit contributions | |
ps | -0.006 |
ss | -0.005 |
teeth | -0.005 |
ches | -0.005 |
mus | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | +3.96e-03 |
Pos logit contributions | |
weekend | +0.038 |
hallway | +0.037 |
stair | +0.037 |
she | +0.036 |
particular | +0.036 |
Neg logit contributions | |
ps | -0.040 |
ss | -0.034 |
teeth | -0.033 |
ches | -0.032 |
mus | -0.031 |
Token | They |
Token ID | 1119 |
Feature activation | -4.64e-03 |
Pos logit contributions | |
she | +0.005 |
weekend | +0.005 |
hallway | +0.005 |
day | +0.005 |
stair | +0.005 |
Neg logit contributions | |
ps | -0.008 |
ss | -0.007 |
teeth | -0.007 |
ches | -0.007 |
mus | -0.007 |
Token | lion |
Token ID | 18744 |
Feature activation | -4.57e-02 |
Pos logit contributions | |
ps | +0.072 |
ches | +0.062 |
ealing | +0.061 |
uable | +0.061 |
mus | +0.061 |
Neg logit contributions | |
day | -0.054 |
hallway | -0.047 |
she | -0.047 |
weekend | -0.046 |
sunny | -0.046 |
Token | ran |
Token ID | 4966 |
Feature activation | -2.26e-02 |
Pos logit contributions | |
ps | +0.113 |
iddles | +0.102 |
ags | +0.102 |
unnecessary | +0.102 |
rier | +0.101 |
Neg logit contributions | |
she | -0.044 |
he | -0.038 |
day | -0.037 |
they | -0.037 |
hallway | -0.029 |
Token | fast |
Token ID | 3049 |
Feature activation | -2.30e-02 |
Pos logit contributions | |
ps | +0.046 |
ss | +0.042 |
teeth | +0.040 |
iddles | +0.039 |
mus | +0.039 |
Neg logit contributions | |
she | -0.029 |
day | -0.027 |
weekend | -0.026 |
hallway | -0.026 |
they | -0.026 |
Token | and |
Token ID | 290 |
Feature activation | -3.88e-02 |
Pos logit contributions | |
ps | +0.047 |
ss | +0.044 |
ches | +0.042 |
wy | +0.042 |
mus | +0.041 |
Neg logit contributions | |
she | -0.028 |
day | -0.027 |
hallway | -0.025 |
stair | -0.025 |
weekend | -0.025 |
Token | jumped |
Token ID | 11687 |
Feature activation | -4.67e-02 |
Pos logit contributions | |
ps | +0.087 |
iddles | +0.080 |
ss | +0.080 |
rier | +0.076 |
ealing | +0.076 |
Neg logit contributions | |
she | -0.051 |
they | -0.047 |
he | -0.045 |
day | -0.042 |
weekend | -0.037 |
Token | on |
Token ID | 319 |
Feature activation | -5.30e-02 |
Pos logit contributions | |
ps | +0.096 |
iddles | +0.096 |
uable | +0.095 |
ss | +0.095 |
wy | +0.093 |
Neg logit contributions | |
she | -0.056 |
he | -0.051 |
they | -0.051 |
day | -0.049 |
stair | -0.041 |
Token | the |
Token ID | 262 |
Feature activation | -4.44e-02 |
Pos logit contributions | |
ps | +0.127 |
ealing | +0.126 |
wy | +0.125 |
ussy | +0.124 |
ches | +0.121 |
Neg logit contributions | |
she | -0.051 |
he | -0.049 |
day | -0.044 |
they | -0.043 |
holiday | -0.039 |
Token | human |
Token ID | 1692 |
Feature activation | +2.35e-02 |
Pos logit contributions | |
ps | +0.092 |
iddles | +0.087 |
ealing | +0.087 |
uable | +0.087 |
ss | +0.087 |
Neg logit contributions | |
day | -0.051 |
hallway | -0.051 |
stair | -0.050 |
porch | -0.049 |
hill | -0.048 |
Token | . |
Token ID | 13 |
Feature activation | -3.23e-02 |
Pos logit contributions | |
weekend | +0.044 |
she | +0.043 |
particular | +0.042 |
holiday | +0.042 |
hallway | +0.041 |
Neg logit contributions | |
ps | -0.032 |
teeth | -0.026 |
ches | -0.026 |
ss | -0.026 |
mus | -0.023 |
Token | He |
Token ID | 679 |
Feature activation | -4.28e-02 |
Pos logit contributions | |
ps | +0.086 |
wy | +0.077 |
ches | +0.073 |
ache | +0.072 |
ss | +0.072 |
Neg logit contributions | |
she | -0.031 |
he | -0.029 |
they | -0.029 |
day | -0.025 |
weekend | -0.022 |
Token | bit |
Token ID | 1643 |
Feature activation | -1.53e-02 |
Pos logit contributions | |
ps | +0.101 |
rier | +0.096 |
iddles | +0.094 |
ss | +0.094 |
ealing | +0.091 |
Neg logit contributions | |
she | -0.044 |
he | -0.040 |
they | -0.039 |
day | -0.037 |
hallway | -0.033 |
Token | no |
Token ID | 645 |
Feature activation | -1.59e-02 |
Pos logit contributions | |
she | +0.001 |
weekend | +0.001 |
day | +0.001 |
hallway | +0.001 |
stair | +0.001 |
Neg logit contributions | |
ps | -0.001 |
ss | -0.001 |
ches | -0.001 |
teeth | -0.001 |
mus | -0.001 |
Token | , |
Token ID | 11 |
Feature activation | +4.89e-02 |
Pos logit contributions | |
ps | +0.032 |
ss | +0.028 |
iddles | +0.027 |
teeth | +0.026 |
ches | +0.026 |
Neg logit contributions | |
she | -0.022 |
day | -0.021 |
weekend | -0.020 |
stair | -0.019 |
hallway | -0.019 |
Token | you |
Token ID | 345 |
Feature activation | -1.76e-02 |
Pos logit contributions | |
hallway | +0.075 |
driveway | +0.071 |
weekend | +0.070 |
days | +0.069 |
stair | +0.068 |
Neg logit contributions | |
ps | -0.081 |
ss | -0.072 |
teeth | -0.070 |
ches | -0.066 |
ealing | -0.063 |
Token | must |
Token ID | 1276 |
Feature activation | +4.01e-02 |
Pos logit contributions | |
ps | +0.036 |
ss | +0.031 |
iddles | +0.031 |
mus | +0.031 |
rier | +0.030 |
Neg logit contributions | |
she | -0.022 |
day | -0.020 |
hallway | -0.019 |
stair | -0.019 |
weekend | -0.019 |
Token | be |
Token ID | 307 |
Feature activation | -3.46e-02 |
Pos logit contributions | |
weekend | +0.058 |
stair | +0.057 |
hallway | +0.056 |
she | +0.054 |
days | +0.052 |
Neg logit contributions | |
ps | -0.074 |
ss | -0.062 |
ches | -0.061 |
teeth | -0.059 |
uable | -0.055 |
Token | feeling |
Token ID | 4203 |
Feature activation | -5.41e-02 |
Pos logit contributions | |
ps | +0.053 |
ss | +0.049 |
uable | +0.047 |
iddles | +0.047 |
oni | +0.046 |
Neg logit contributions | |
she | -0.059 |
day | -0.055 |
sunny | -0.053 |
he | -0.052 |
exploring | -0.052 |
Token | sick |
Token ID | 6639 |
Feature activation | -1.61e-03 |
Pos logit contributions | |
ps | +0.084 |
ss | +0.080 |
iddles | +0.074 |
osaurus | +0.070 |
amus | +0.069 |
Neg logit contributions | |
she | -0.099 |
day | -0.093 |
he | -0.090 |
sunny | -0.086 |
weekend | -0.085 |
Token | , |
Token ID | 11 |
Feature activation | +2.97e-02 |
Pos logit contributions | |
ps | +0.002 |
ss | +0.002 |
ches | +0.002 |
teeth | +0.002 |
mus | +0.002 |
Neg logit contributions | |
she | -0.003 |
weekend | -0.003 |
day | -0.003 |
hallway | -0.003 |
stair | -0.003 |
Token | â |
Token ID | 22940 |
Feature activation | -4.89e-03 |
Pos logit contributions | |
hallway | +0.043 |
weekend | +0.043 |
stair | +0.043 |
days | +0.042 |
day | +0.041 |
Neg logit contributions | |
ps | -0.050 |
ss | -0.044 |
teeth | -0.044 |
ches | -0.042 |
mus | -0.039 |
Token | € |
Token ID | 26391 |
Feature activation | +1.12e-02 |
Pos logit contributions | |
ps | +0.008 |
ss | +0.007 |
teeth | +0.007 |
ches | +0.007 |
mus | +0.007 |
Neg logit contributions | |
she | -0.008 |
weekend | -0.007 |
day | -0.007 |
hallway | -0.007 |
stair | -0.007 |
Token | her |
Token ID | 607 |
Feature activation | -1.26e-02 |
Pos logit contributions | |
weekend | +0.017 |
she | +0.017 |
day | +0.017 |
hallway | +0.017 |
stair | +0.017 |
Neg logit contributions | |
ps | -0.018 |
ss | -0.017 |
teeth | -0.015 |
ches | -0.015 |
mus | -0.015 |
Token | corner |
Token ID | 5228 |
Feature activation | +4.44e-02 |
Pos logit contributions | |
she | +0.023 |
weekend | +0.022 |
particular | +0.021 |
days | +0.021 |
holiday | +0.021 |
Neg logit contributions | |
ps | -0.032 |
ss | -0.028 |
teeth | -0.028 |
ches | -0.027 |
mus | -0.025 |
Token | . |
Token ID | 13 |
Feature activation | -9.27e-03 |
Pos logit contributions | |
weekend | +0.087 |
holiday | +0.085 |
explore | +0.081 |
wander | +0.081 |
hallway | +0.080 |
Neg logit contributions | |
ps | -0.053 |
ss | -0.044 |
teeth | -0.044 |
ches | -0.038 |
ows | -0.038 |
Token | She |
Token ID | 1375 |
Feature activation | -2.02e-03 |
Pos logit contributions | |
ps | +0.021 |
ss | +0.019 |
ches | +0.018 |
mus | +0.018 |
teeth | +0.018 |
Neg logit contributions | |
she | -0.010 |
stair | -0.009 |
hallway | -0.009 |
day | -0.009 |
weekend | -0.009 |
Token | had |
Token ID | 550 |
Feature activation | -1.67e-03 |
Pos logit contributions | |
ps | +0.004 |
ss | +0.004 |
ches | +0.003 |
teeth | +0.003 |
mus | +0.003 |
Neg logit contributions | |
she | -0.003 |
weekend | -0.002 |
day | -0.002 |
hallway | -0.002 |
stair | -0.002 |
Token | never |
Token ID | 1239 |
Feature activation | +3.50e-02 |
Pos logit contributions | |
ps | +0.004 |
ss | +0.003 |
ches | +0.003 |
teeth | +0.003 |
mus | +0.003 |
Neg logit contributions | |
she | -0.002 |
weekend | -0.002 |
day | -0.002 |
hallway | -0.002 |
stair | -0.002 |
Token | seen |
Token ID | 1775 |
Feature activation | -5.61e-02 |
Pos logit contributions | |
stair | +0.047 |
stack | +0.047 |
she | +0.043 |
particular | +0.043 |
weekend | +0.043 |
Neg logit contributions | |
ps | -0.064 |
ss | -0.057 |
teeth | -0.056 |
iddles | -0.052 |
ches | -0.051 |
Token | so |
Token ID | 523 |
Feature activation | -5.03e-03 |
Pos logit contributions | |
ps | +0.110 |
ss | +0.102 |
ealing | +0.101 |
ches | +0.098 |
mus | +0.097 |
Neg logit contributions | |
she | -0.082 |
Jane | -0.069 |
day | -0.068 |
he | -0.064 |
weekend | -0.063 |
Token | much |
Token ID | 881 |
Feature activation | -3.97e-03 |
Pos logit contributions | |
ps | +0.008 |
ss | +0.007 |
ches | +0.007 |
teeth | +0.006 |
mus | +0.006 |
Neg logit contributions | |
she | -0.009 |
weekend | -0.008 |
day | -0.008 |
stair | -0.008 |
hallway | -0.008 |
Token | waste |
Token ID | 7030 |
Feature activation | -9.72e-03 |
Pos logit contributions | |
ps | +0.007 |
ss | +0.006 |
ches | +0.006 |
teeth | +0.006 |
mus | +0.006 |
Neg logit contributions | |
she | -0.006 |
weekend | -0.005 |
day | -0.005 |
stair | -0.005 |
hallway | -0.005 |
Token | before |
Token ID | 878 |
Feature activation | +7.13e-03 |
Pos logit contributions | |
ps | +0.019 |
ss | +0.016 |
teeth | +0.016 |
ches | +0.016 |
mus | +0.015 |
Neg logit contributions | |
she | -0.014 |
day | -0.013 |
weekend | -0.013 |
stair | -0.012 |
hallway | -0.012 |
Token | and |
Token ID | 290 |
Feature activation | +2.80e-02 |
Pos logit contributions | |
weekend | +0.012 |
she | +0.012 |
hallway | +0.011 |
day | +0.011 |
stair | +0.011 |
Neg logit contributions | |
ps | -0.011 |
ss | -0.009 |
teeth | -0.009 |
ches | -0.009 |
mus | -0.009 |
Token | makes |
Token ID | 1838 |
Feature activation | -3.86e-02 |
Pos logit contributions | |
ps | +0.019 |
ss | +0.017 |
mus | +0.016 |
iddles | +0.016 |
teeth | +0.015 |
Neg logit contributions | |
she | -0.020 |
day | -0.019 |
weekend | -0.018 |
hallway | -0.018 |
stair | -0.018 |
Token | rain |
Token ID | 6290 |
Feature activation | -3.28e-02 |
Pos logit contributions | |
ps | +0.067 |
ss | +0.063 |
iddles | +0.061 |
wy | +0.060 |
uable | +0.060 |
Neg logit contributions | |
she | -0.059 |
day | -0.056 |
they | -0.053 |
weekend | -0.051 |
holiday | -0.050 |
Token | bows |
Token ID | 25435 |
Feature activation | -1.42e-02 |
Pos logit contributions | |
ps | +0.056 |
ss | +0.051 |
uable | +0.048 |
mus | +0.048 |
iddles | +0.048 |
Neg logit contributions | |
she | -0.050 |
day | -0.049 |
they | -0.046 |
hallway | -0.046 |
weekend | -0.045 |
Token | in |
Token ID | 287 |
Feature activation | -1.13e-02 |
Pos logit contributions | |
ps | +0.026 |
ss | +0.023 |
ches | +0.022 |
mus | +0.022 |
teeth | +0.022 |
Neg logit contributions | |
she | -0.021 |
day | -0.020 |
weekend | -0.019 |
stair | -0.019 |
hallway | -0.019 |
Token | the |
Token ID | 262 |
Feature activation | +4.66e-03 |
Pos logit contributions | |
ps | +0.021 |
ss | +0.018 |
ches | +0.018 |
iddles | +0.017 |
mus | +0.017 |
Neg logit contributions | |
she | -0.016 |
day | -0.016 |
weekend | -0.016 |
stair | -0.016 |
hallway | -0.015 |
Token | sky |
Token ID | 6766 |
Feature activation | -5.18e-02 |
Pos logit contributions | |
she | +0.006 |
weekend | +0.005 |
stair | +0.005 |
day | +0.005 |
hallway | +0.005 |
Neg logit contributions | |
ps | -0.010 |
ss | -0.008 |
teeth | -0.008 |
ches | -0.008 |
mus | -0.008 |
Token | . |
Token ID | 13 |
Feature activation | -1.74e-02 |
Pos logit contributions | |
ps | +0.106 |
ss | +0.099 |
uable | +0.096 |
iddles | +0.095 |
mus | +0.094 |
Neg logit contributions | |
day | -0.065 |
she | -0.064 |
they | -0.058 |
sunny | -0.056 |
weekend | -0.055 |
Token | The |
Token ID | 383 |
Feature activation | +1.76e-02 |
Pos logit contributions | |
ps | +0.036 |
ches | +0.032 |
ss | +0.032 |
teeth | +0.030 |
iddles | +0.030 |
Neg logit contributions | |
she | -0.021 |
weekend | -0.019 |
day | -0.019 |
hallway | -0.018 |
they | -0.018 |
Token | rain |
Token ID | 6290 |
Feature activation | -7.84e-03 |
Pos logit contributions | |
she | +0.024 |
weekend | +0.024 |
hallway | +0.023 |
stair | +0.023 |
particular | +0.023 |
Neg logit contributions | |
ps | -0.032 |
ss | -0.027 |
teeth | -0.027 |
mus | -0.025 |
ches | -0.025 |
Token | is |
Token ID | 318 |
Feature activation | -3.68e-02 |
Pos logit contributions | |
ps | +0.012 |
ss | +0.011 |
teeth | +0.010 |
mus | +0.010 |
ches | +0.010 |
Neg logit contributions | |
she | -0.013 |
day | -0.013 |
weekend | -0.013 |
hallway | -0.013 |
stair | -0.013 |
Token | incredible |
Token ID | 8082 |
Feature activation | -2.41e-02 |
Pos logit contributions | |
ps | +0.062 |
ss | +0.061 |
iddles | +0.057 |
ches | +0.056 |
mus | +0.055 |
Neg logit contributions | |
day | -0.057 |
she | -0.056 |
sunny | -0.055 |
weekend | -0.053 |
they | -0.052 |