Token | I |
Token ID | 314 |
Feature activation | -2.55e+00 |
Pos logit contributions | |
he | +0.006 |
ing | +0.006 |
ore | +0.006 |
ay | +0.006 |
ings | +0.006 |
Neg logit contributions | |
forgot | -0.010 |
remembered | -0.008 |
knew | -0.008 |
wished | -0.008 |
struggled | -0.007 |
Token | can |
Token ID | 460 |
Feature activation | -3.84e-01 |
Pos logit contributions | |
forgot | +0.052 |
remembered | +0.039 |
wished | +0.038 |
knew | +0.036 |
regretted | +0.035 |
Neg logit contributions | |
he | -0.064 |
ore | -0.061 |
ay | -0.060 |
ings | -0.059 |
room | -0.058 |
Token | find |
Token ID | 1064 |
Feature activation | +2.09e+00 |
Pos logit contributions | |
forgot | +0.009 |
knew | +0.007 |
remembered | +0.007 |
wished | +0.007 |
knocked | +0.006 |
Neg logit contributions | |
he | -0.009 |
ing | -0.008 |
ore | -0.008 |
ay | -0.008 |
ays | -0.008 |
Token | something |
Token ID | 1223 |
Feature activation | +7.50e-01 |
Pos logit contributions | |
he | +0.042 |
room | +0.037 |
ing | +0.037 |
ings | +0.037 |
ore | +0.036 |
Neg logit contributions | |
forgot | -0.051 |
knew | -0.043 |
remembered | -0.043 |
wished | -0.043 |
knocked | -0.041 |
Token | gl |
Token ID | 1278 |
Feature activation | +5.83e+00 |
Pos logit contributions | |
he | +0.012 |
ing | +0.011 |
room | +0.011 |
ay | +0.010 |
ays | +0.010 |
Neg logit contributions | |
forgot | -0.022 |
knew | -0.019 |
remembered | -0.019 |
wished | -0.018 |
bit | -0.018 |
Token | immer |
Token ID | 10957 |
Feature activation | +9.56e+00 |
Pos logit contributions | |
he | +0.096 |
room | +0.087 |
ays | +0.087 |
ings | +0.083 |
ummy | +0.080 |
Neg logit contributions | |
forgot | -0.143 |
'd | -0.124 |
remembered | -0.120 |
packed | -0.118 |
knew | -0.114 |
Token | ing |
Token ID | 278 |
Feature activation | +3.90e+00 |
Pos logit contributions | |
overed | +0.177 |
uggle | +0.166 |
aps | +0.161 |
ooming | +0.161 |
room | +0.158 |
Neg logit contributions | |
forgot | -0.190 |
bit | -0.178 |
'd | -0.172 |
packed | -0.164 |
remembered | -0.161 |
Token | ?" |
Token ID | 1701 |
Feature activation | +3.89e+00 |
Pos logit contributions | |
he | +0.057 |
room | +0.053 |
ay | +0.053 |
ing | +0.053 |
ays | +0.053 |
Neg logit contributions | |
forgot | -0.113 |
remembered | -0.096 |
knew | -0.095 |
bit | -0.095 |
wished | -0.093 |
Token | |
Token ID | 220 |
Feature activation | +4.81e+00 |
Pos logit contributions | |
he | +0.072 |
ays | +0.069 |
ay | +0.069 |
ings | +0.069 |
ing | +0.069 |
Neg logit contributions | |
forgot | -0.102 |
knew | -0.087 |
wished | -0.085 |
remembered | -0.084 |
learned | -0.080 |
Token |
|
Token ID | 198 |
Feature activation | +5.10e+00 |
Pos logit contributions | |
room | +0.093 |
ay | +0.092 |
eddy | +0.092 |
ays | +0.091 |
ing | +0.089 |
Neg logit contributions | |
forgot | -0.101 |
knew | -0.089 |
told | -0.082 |
wished | -0.079 |
learned | -0.079 |
Token |
|
Token ID | 198 |
Feature activation | +3.71e+00 |
Pos logit contributions | |
ays | +0.111 |
ement | +0.093 |
ings | +0.091 |
ing | +0.091 |
over | +0.091 |
Neg logit contributions | |
forgot | -0.118 |
knew | -0.110 |
learned | -0.103 |
remembered | -0.098 |
told | -0.098 |
Token | drink |
Token ID | 4144 |
Feature activation | +4.99e+00 |
Pos logit contributions | |
forgot | +0.069 |
knew | +0.055 |
remembered | +0.055 |
known | +0.053 |
wished | +0.053 |
Neg logit contributions | |
he | -0.063 |
ore | -0.057 |
ay | -0.057 |
uggle | -0.055 |
with | -0.054 |
Token | it |
Token ID | 340 |
Feature activation | +3.08e+00 |
Pos logit contributions | |
he | +0.108 |
over | +0.099 |
room | +0.099 |
ane | +0.096 |
apping | +0.096 |
Neg logit contributions | |
forgot | -0.111 |
bit | -0.094 |
knew | -0.092 |
wished | -0.087 |
remembered | -0.087 |
Token | . |
Token ID | 13 |
Feature activation | +3.22e+00 |
Pos logit contributions | |
he | +0.049 |
ore | +0.047 |
ays | +0.045 |
ings | +0.045 |
ay | +0.045 |
Neg logit contributions | |
forgot | -0.088 |
bit | -0.074 |
knew | -0.072 |
remembered | -0.071 |
wished | -0.069 |
Token | It |
Token ID | 632 |
Feature activation | -1.89e+00 |
Pos logit contributions | |
he | +0.068 |
ing | +0.061 |
ings | +0.061 |
ogging | +0.057 |
ore | +0.057 |
Neg logit contributions | |
forgot | -0.079 |
remembered | -0.065 |
bit | -0.064 |
wished | -0.064 |
knew | -0.063 |
Token | looks |
Token ID | 3073 |
Feature activation | +2.31e-03 |
Pos logit contributions | |
forgot | +0.046 |
known | +0.040 |
remembered | +0.039 |
wished | +0.039 |
robbers | +0.038 |
Neg logit contributions | |
ing | -0.033 |
ays | -0.033 |
ings | -0.032 |
ore | -0.032 |
he | -0.032 |
Token | y |
Token ID | 331 |
Feature activation | +8.52e+00 |
Pos logit contributions | |
he | +0.000 |
ay | +0.000 |
ing | +0.000 |
ings | +0.000 |
ore | +0.000 |
Neg logit contributions | |
forgot | -0.000 |
remembered | -0.000 |
knew | -0.000 |
wished | -0.000 |
struggled | -0.000 |
Token | ummy |
Token ID | 13513 |
Feature activation | +4.48e+00 |
Pos logit contributions | |
overed | +0.106 |
he | +0.102 |
ute | +0.093 |
ane | +0.092 |
eway | +0.086 |
Neg logit contributions | |
forgot | -0.205 |
bit | -0.204 |
mixed | -0.176 |
f | -0.176 |
sp | -0.174 |
Token | ." |
Token ID | 526 |
Feature activation | +4.32e+00 |
Pos logit contributions | |
he | +0.074 |
room | +0.068 |
ings | +0.066 |
ore | +0.066 |
ay | +0.064 |
Neg logit contributions | |
forgot | -0.128 |
bit | -0.109 |
remembered | -0.106 |
knew | -0.106 |
wished | -0.105 |
Token | Mia |
Token ID | 32189 |
Feature activation | -5.84e+00 |
Pos logit contributions | |
he | +0.073 |
ing | +0.070 |
ay | +0.067 |
ays | +0.066 |
ings | +0.063 |
Neg logit contributions | |
forgot | -0.124 |
knew | -0.103 |
remembered | -0.100 |
wished | -0.099 |
learned | -0.098 |
Token | says |
Token ID | 1139 |
Feature activation | +1.20e+00 |
Pos logit contributions | |
forgot | +0.133 |
remembered | +0.102 |
wished | +0.101 |
known | +0.101 |
forgotten | +0.100 |
Neg logit contributions | |
he | -0.129 |
ing | -0.120 |
ore | -0.118 |
ays | -0.118 |
ay | -0.116 |
Token | , |
Token ID | 11 |
Feature activation | +6.02e-01 |
Pos logit contributions | |
he | +0.021 |
ays | +0.019 |
eddy | +0.018 |
ay | +0.018 |
ore | +0.017 |
Neg logit contributions | |
forgot | -0.035 |
knew | -0.031 |
remembered | -0.030 |
bit | -0.029 |
wished | -0.028 |
Token | and |
Token ID | 290 |
Feature activation | -3.88e+00 |
Pos logit contributions | |
he | +0.067 |
ore | +0.057 |
uggle | +0.057 |
ay | +0.056 |
room | +0.054 |
Neg logit contributions | |
forgot | -0.124 |
remembered | -0.106 |
knew | -0.104 |
wished | -0.098 |
bit | -0.096 |
Token | lettuce |
Token ID | 39406 |
Feature activation | +3.44e+00 |
Pos logit contributions | |
forgot | +0.097 |
known | +0.078 |
knew | +0.076 |
punished | +0.076 |
struggled | +0.074 |
Neg logit contributions | |
ing | -0.062 |
ummy | -0.061 |
ay | -0.061 |
ays | -0.061 |
he | -0.061 |
Token | . |
Token ID | 13 |
Feature activation | +3.75e+00 |
Pos logit contributions | |
he | +0.035 |
ay | +0.030 |
ore | +0.029 |
ing | +0.029 |
room | +0.028 |
Neg logit contributions | |
forgot | -0.122 |
remembered | -0.105 |
knew | -0.104 |
wished | -0.104 |
bit | -0.099 |
Token | It |
Token ID | 632 |
Feature activation | -3.50e+00 |
Pos logit contributions | |
he | +0.083 |
ay | +0.077 |
ing | +0.076 |
room | +0.075 |
uggle | +0.071 |
Neg logit contributions | |
forgot | -0.087 |
wished | -0.074 |
knew | -0.073 |
remembered | -0.068 |
knocked | -0.067 |
Token | looks |
Token ID | 3073 |
Feature activation | -5.13e-01 |
Pos logit contributions | |
forgot | +0.093 |
known | +0.079 |
punished | +0.077 |
remembered | +0.075 |
forgotten | +0.075 |
Neg logit contributions | |
ays | -0.057 |
he | -0.055 |
ore | -0.052 |
with | -0.051 |
iffs | -0.051 |
Token | y |
Token ID | 331 |
Feature activation | +8.48e+00 |
Pos logit contributions | |
forgot | +0.011 |
remembered | +0.008 |
knew | +0.008 |
wished | +0.008 |
struggled | +0.008 |
Neg logit contributions | |
he | -0.012 |
ing | -0.011 |
ay | -0.011 |
ore | -0.011 |
ings | -0.011 |
Token | ummy |
Token ID | 13513 |
Feature activation | +4.54e+00 |
Pos logit contributions | |
he | +0.100 |
ute | +0.091 |
overed | +0.091 |
uggle | +0.086 |
ane | +0.086 |
Neg logit contributions | |
forgot | -0.226 |
bit | -0.201 |
knew | -0.192 |
forgotten | -0.186 |
wasted | -0.184 |
Token | . |
Token ID | 13 |
Feature activation | +3.63e+00 |
Pos logit contributions | |
he | +0.062 |
room | +0.058 |
ay | +0.055 |
ore | +0.054 |
uggle | +0.054 |
Neg logit contributions | |
forgot | -0.142 |
knew | -0.121 |
bit | -0.121 |
remembered | -0.121 |
wished | -0.120 |
Token |
|
Token ID | 198 |
Feature activation | +4.78e+00 |
Pos logit contributions | |
he | +0.076 |
ay | +0.070 |
room | +0.070 |
ing | +0.069 |
ings | +0.065 |
Neg logit contributions | |
forgot | -0.089 |
wished | -0.075 |
knew | -0.075 |
remembered | -0.071 |
knocked | -0.069 |
Token |
|
Token ID | 198 |
Feature activation | +3.79e+00 |
Pos logit contributions | |
ays | +0.096 |
ing | +0.095 |
ement | +0.094 |
ipping | +0.094 |
ay | +0.089 |
Neg logit contributions | |
forgot | -0.111 |
knew | -0.097 |
wished | -0.090 |
also | -0.088 |
learned | -0.088 |
Token | " |
Token ID | 1 |
Feature activation | +4.83e-01 |
Pos logit contributions | |
he | +0.055 |
ay | +0.053 |
ore | +0.051 |
ays | +0.051 |
ing | +0.050 |
Neg logit contributions | |
forgot | -0.118 |
remembered | -0.098 |
knew | -0.098 |
wished | -0.098 |
bit | -0.093 |
Token | sprink |
Token ID | 27562 |
Feature activation | +2.32e+00 |
Pos logit contributions | |
he | +0.048 |
ore | +0.042 |
uggle | +0.041 |
ay | +0.040 |
room | +0.040 |
Neg logit contributions | |
forgot | -0.074 |
remembered | -0.062 |
knew | -0.061 |
wished | -0.060 |
bit | -0.059 |
Token | les |
Token ID | 829 |
Feature activation | +3.64e+00 |
Pos logit contributions | |
he | +0.038 |
room | +0.037 |
ore | +0.036 |
uggle | +0.036 |
ings | +0.035 |
Neg logit contributions | |
forgot | -0.063 |
knew | -0.052 |
remembered | -0.051 |
bit | -0.050 |
wished | -0.049 |
Token | . |
Token ID | 13 |
Feature activation | +3.66e+00 |
Pos logit contributions | |
he | +0.046 |
uggle | +0.042 |
ore | +0.041 |
ay | +0.039 |
ings | +0.039 |
Neg logit contributions | |
forgot | -0.114 |
knew | -0.099 |
remembered | -0.098 |
wished | -0.096 |
bit | -0.095 |
Token | It |
Token ID | 632 |
Feature activation | -4.57e+00 |
Pos logit contributions | |
he | +0.095 |
uggle | +0.088 |
ing | +0.088 |
ings | +0.087 |
ore | +0.087 |
Neg logit contributions | |
forgot | -0.075 |
knew | -0.056 |
wished | -0.055 |
remembered | -0.055 |
learned | -0.051 |
Token | looks |
Token ID | 3073 |
Feature activation | -9.45e-01 |
Pos logit contributions | |
forgot | +0.115 |
known | +0.099 |
punished | +0.096 |
forgotten | +0.095 |
remembered | +0.094 |
Neg logit contributions | |
ays | -0.084 |
he | -0.076 |
with | -0.075 |
ore | -0.074 |
ay | -0.072 |
Token | y |
Token ID | 331 |
Feature activation | +8.41e+00 |
Pos logit contributions | |
forgot | +0.023 |
remembered | +0.018 |
knew | +0.018 |
wished | +0.018 |
struggled | +0.017 |
Neg logit contributions | |
he | -0.019 |
ing | -0.018 |
ay | -0.018 |
ings | -0.018 |
ore | -0.018 |
Token | ummy |
Token ID | 13513 |
Feature activation | +3.90e+00 |
Pos logit contributions | |
he | +0.101 |
overed | +0.088 |
ore | +0.086 |
ane | +0.083 |
uggle | +0.082 |
Neg logit contributions | |
forgot | -0.220 |
bit | -0.200 |
knew | -0.188 |
f | -0.183 |
forgotten | -0.178 |
Token | . |
Token ID | 13 |
Feature activation | +3.60e+00 |
Pos logit contributions | |
he | +0.053 |
room | +0.048 |
ore | +0.047 |
ay | +0.046 |
uggle | +0.046 |
Neg logit contributions | |
forgot | -0.122 |
bit | -0.105 |
knew | -0.104 |
wished | -0.104 |
remembered | -0.104 |
Token |
|
Token ID | 198 |
Feature activation | +4.82e+00 |
Pos logit contributions | |
he | +0.087 |
ings | +0.080 |
uggle | +0.080 |
ing | +0.080 |
ore | +0.078 |
Neg logit contributions | |
forgot | -0.079 |
knew | -0.062 |
wished | -0.061 |
remembered | -0.060 |
knocked | -0.056 |
Token |
|
Token ID | 198 |
Feature activation | +3.84e+00 |
Pos logit contributions | |
ays | +0.104 |
ing | +0.100 |
ement | +0.100 |
ipping | +0.099 |
over | +0.098 |
Neg logit contributions | |
forgot | -0.107 |
knew | -0.095 |
wished | -0.085 |
learned | -0.085 |
told | -0.084 |
Token | " |
Token ID | 1 |
Feature activation | +2.73e-01 |
Pos logit contributions | |
he | +0.055 |
ings | +0.050 |
ore | +0.050 |
ay | +0.050 |
ays | +0.050 |
Neg logit contributions | |
forgot | -0.122 |
remembered | -0.102 |
knew | -0.101 |
wished | -0.101 |
bit | -0.096 |
Token | cher |
Token ID | 21382 |
Feature activation | +3.62e+00 |
Pos logit contributions | |
forgot | +0.108 |
knew | +0.089 |
remembered | +0.087 |
struggled | +0.084 |
missed | +0.084 |
Neg logit contributions | |
he | -0.051 |
ing | -0.049 |
ays | -0.049 |
ay | -0.047 |
room | -0.047 |
Token | ries |
Token ID | 1678 |
Feature activation | +2.91e+00 |
Pos logit contributions | |
he | +0.052 |
room | +0.049 |
ore | +0.047 |
uggle | +0.046 |
time | +0.045 |
Neg logit contributions | |
forgot | -0.109 |
remembered | -0.095 |
knew | -0.089 |
knocked | -0.088 |
wasted | -0.088 |
Token | . |
Token ID | 13 |
Feature activation | +3.48e+00 |
Pos logit contributions | |
he | +0.035 |
room | +0.035 |
ay | +0.034 |
ore | +0.033 |
uggle | +0.033 |
Neg logit contributions | |
forgot | -0.093 |
remembered | -0.079 |
knew | -0.079 |
wished | -0.078 |
bit | -0.077 |
Token | It |
Token ID | 632 |
Feature activation | -4.04e+00 |
Pos logit contributions | |
he | +0.078 |
ing | +0.075 |
uggle | +0.071 |
ingle | +0.070 |
ipping | +0.069 |
Neg logit contributions | |
forgot | -0.078 |
wished | -0.065 |
knew | -0.063 |
learned | -0.062 |
remembered | -0.060 |
Token | looks |
Token ID | 3073 |
Feature activation | -7.19e-01 |
Pos logit contributions | |
forgot | +0.102 |
known | +0.085 |
remembered | +0.082 |
struggled | +0.082 |
forgotten | +0.081 |
Neg logit contributions | |
ays | -0.072 |
he | -0.069 |
with | -0.065 |
ore | -0.062 |
ay | -0.062 |
Token | y |
Token ID | 331 |
Feature activation | +8.36e+00 |
Pos logit contributions | |
forgot | +0.017 |
remembered | +0.013 |
knew | +0.013 |
wished | +0.013 |
struggled | +0.012 |
Neg logit contributions | |
he | -0.016 |
ing | -0.015 |
ay | -0.015 |
ings | -0.014 |
ore | -0.014 |
Token | ummy |
Token ID | 13513 |
Feature activation | +4.23e+00 |
Pos logit contributions | |
he | +0.095 |
ute | +0.090 |
overed | +0.085 |
uggle | +0.085 |
ore | +0.085 |
Neg logit contributions | |
forgot | -0.222 |
bit | -0.200 |
knew | -0.191 |
forgotten | -0.185 |
wasted | -0.182 |
Token | . |
Token ID | 13 |
Feature activation | +3.43e+00 |
Pos logit contributions | |
he | +0.057 |
room | +0.053 |
ore | +0.052 |
ay | +0.052 |
ings | +0.051 |
Neg logit contributions | |
forgot | -0.133 |
wished | -0.113 |
knew | -0.112 |
remembered | -0.112 |
bit | -0.112 |
Token |
|
Token ID | 198 |
Feature activation | +3.60e+00 |
Pos logit contributions | |
he | +0.074 |
ing | +0.070 |
uggle | +0.067 |
ings | +0.066 |
ipping | +0.065 |
Neg logit contributions | |
forgot | -0.080 |
wished | -0.067 |
knew | -0.067 |
remembered | -0.063 |
learned | -0.063 |
Token |
|
Token ID | 198 |
Feature activation | +3.25e+00 |
Pos logit contributions | |
ing | +0.064 |
ays | +0.064 |
over | +0.061 |
he | +0.061 |
ipping | +0.061 |
Neg logit contributions | |
forgot | -0.093 |
knew | -0.082 |
wished | -0.079 |
learned | -0.076 |
remembered | -0.074 |
Token | " |
Token ID | 1 |
Feature activation | +4.63e-01 |
Pos logit contributions | |
he | +0.046 |
ore | +0.042 |
ings | +0.042 |
ay | +0.042 |
ays | +0.042 |
Neg logit contributions | |
forgot | -0.103 |
remembered | -0.087 |
knew | -0.086 |
wished | -0.086 |
learned | -0.081 |
Token | says |
Token ID | 1139 |
Feature activation | +2.33e+00 |
Pos logit contributions | |
forgot | +0.096 |
remembered | +0.074 |
knew | +0.072 |
wished | +0.071 |
forgotten | +0.067 |
Neg logit contributions | |
he | -0.106 |
ing | -0.100 |
ore | -0.099 |
ings | -0.098 |
ays | -0.097 |
Token | . |
Token ID | 13 |
Feature activation | +5.57e+00 |
Pos logit contributions | |
he | +0.042 |
ay | +0.039 |
ays | +0.037 |
eddy | +0.036 |
uggle | +0.036 |
Neg logit contributions | |
forgot | -0.064 |
knew | -0.055 |
bit | -0.054 |
remembered | -0.054 |
knocked | -0.054 |
Token | " |
Token ID | 366 |
Feature activation | +2.98e+00 |
Pos logit contributions | |
he | +0.135 |
ach | +0.130 |
uggle | +0.127 |
ays | +0.124 |
ingle | +0.123 |
Neg logit contributions | |
forgot | -0.100 |
knew | -0.096 |
wished | -0.088 |
knocked | -0.087 |
remembered | -0.085 |
Token | It |
Token ID | 1026 |
Feature activation | -2.26e+00 |
Pos logit contributions | |
ings | +0.068 |
ing | +0.066 |
ummy | +0.066 |
ays | +0.065 |
ogging | +0.065 |
Neg logit contributions | |
forgot | -0.062 |
wished | -0.054 |
knew | -0.052 |
remembered | -0.049 |
learned | -0.049 |
Token | looks |
Token ID | 3073 |
Feature activation | +1.32e-01 |
Pos logit contributions | |
forgot | +0.062 |
wished | +0.051 |
remembered | +0.051 |
struggled | +0.049 |
knew | +0.049 |
Neg logit contributions | |
he | -0.039 |
ing | -0.038 |
ay | -0.036 |
ays | -0.036 |
ore | -0.036 |
Token | y |
Token ID | 331 |
Feature activation | +8.34e+00 |
Pos logit contributions | |
he | +0.003 |
ay | +0.003 |
ing | +0.003 |
ings | +0.003 |
ore | +0.003 |
Neg logit contributions | |
forgot | -0.003 |
remembered | -0.002 |
knew | -0.002 |
wished | -0.002 |
struggled | -0.002 |
Token | ummy |
Token ID | 13513 |
Feature activation | +4.97e+00 |
Pos logit contributions | |
overed | +0.104 |
ute | +0.098 |
he | +0.096 |
uggle | +0.092 |
ane | +0.091 |
Neg logit contributions | |
forgot | -0.199 |
bit | -0.191 |
knew | -0.172 |
sp | -0.168 |
f | -0.166 |
Token | !" |
Token ID | 2474 |
Feature activation | +3.90e+00 |
Pos logit contributions | |
he | +0.087 |
room | +0.082 |
ore | +0.082 |
uggle | +0.081 |
ings | +0.078 |
Neg logit contributions | |
forgot | -0.134 |
bit | -0.114 |
remembered | -0.111 |
knew | -0.110 |
wished | -0.107 |
Token |
|
Token ID | 198 |
Feature activation | +4.71e+00 |
Pos logit contributions | |
eddy | +0.086 |
he | +0.085 |
ay | +0.084 |
uggle | +0.083 |
ings | +0.083 |
Neg logit contributions | |
forgot | -0.080 |
knew | -0.074 |
learned | -0.070 |
wished | -0.069 |
thought | -0.065 |
Token |
|
Token ID | 198 |
Feature activation | +4.11e+00 |
Pos logit contributions | |
ays | +0.105 |
avored | +0.104 |
over | +0.103 |
ement | +0.102 |
ipping | +0.100 |
Neg logit contributions | |
forgot | -0.086 |
knew | -0.083 |
learned | -0.073 |
wished | -0.071 |
told | -0.070 |
Token | " |
Token ID | 1 |
Feature activation | +7.58e-01 |
Pos logit contributions | |
he | +0.073 |
ays | +0.072 |
ore | +0.071 |
ay | +0.070 |
ings | +0.068 |
Neg logit contributions | |
forgot | -0.111 |
remembered | -0.093 |
wished | -0.092 |
knew | -0.091 |
learned | -0.088 |
Token | cher |
Token ID | 21382 |
Feature activation | +3.18e+00 |
Pos logit contributions | |
forgot | +0.083 |
knew | +0.067 |
known | +0.066 |
remembered | +0.066 |
punished | +0.066 |
Neg logit contributions | |
he | -0.040 |
ing | -0.040 |
ays | -0.040 |
with | -0.040 |
time | -0.039 |
Token | ries |
Token ID | 1678 |
Feature activation | +2.85e+00 |
Pos logit contributions | |
he | +0.046 |
room | +0.042 |
ore | +0.042 |
ay | +0.040 |
uggle | +0.040 |
Neg logit contributions | |
forgot | -0.097 |
remembered | -0.084 |
knew | -0.078 |
knocked | -0.077 |
bit | -0.077 |
Token | . |
Token ID | 13 |
Feature activation | +3.14e+00 |
Pos logit contributions | |
he | +0.034 |
ay | +0.034 |
room | +0.032 |
ore | +0.032 |
ings | +0.031 |
Neg logit contributions | |
forgot | -0.094 |
knew | -0.079 |
remembered | -0.079 |
bit | -0.078 |
wished | -0.077 |
Token | It |
Token ID | 632 |
Feature activation | -3.62e+00 |
Pos logit contributions | |
he | +0.070 |
ing | +0.070 |
ings | +0.066 |
uggle | +0.064 |
ane | +0.064 |
Neg logit contributions | |
forgot | -0.069 |
wished | -0.057 |
knew | -0.056 |
remembered | -0.053 |
learned | -0.053 |
Token | looked |
Token ID | 3114 |
Feature activation | -2.71e-01 |
Pos logit contributions | |
forgot | +0.078 |
unknow | +0.067 |
robbers | +0.067 |
known | +0.066 |
punished | +0.065 |
Neg logit contributions | |
he | -0.069 |
with | -0.066 |
time | -0.062 |
room | -0.059 |
eddy | -0.058 |
Token | y |
Token ID | 331 |
Feature activation | +8.33e+00 |
Pos logit contributions | |
forgot | +0.006 |
remembered | +0.004 |
knew | +0.004 |
wished | +0.004 |
knocked | +0.004 |
Neg logit contributions | |
he | -0.007 |
ing | -0.006 |
ay | -0.006 |
ore | -0.006 |
ings | -0.006 |
Token | ummy |
Token ID | 13513 |
Feature activation | +4.67e+00 |
Pos logit contributions | |
he | +0.097 |
ute | +0.094 |
overed | +0.091 |
ane | +0.085 |
uggle | +0.084 |
Neg logit contributions | |
forgot | -0.222 |
bit | -0.198 |
knew | -0.190 |
forgotten | -0.183 |
wasted | -0.182 |
Token | . |
Token ID | 13 |
Feature activation | +3.04e+00 |
Pos logit contributions | |
he | +0.065 |
room | +0.063 |
ay | +0.060 |
ore | +0.060 |
uggle | +0.058 |
Neg logit contributions | |
forgot | -0.144 |
bit | -0.123 |
knew | -0.121 |
remembered | -0.121 |
wished | -0.121 |
Token | Lily |
Token ID | 20037 |
Feature activation | -6.30e+00 |
Pos logit contributions | |
he | +0.067 |
ing | +0.066 |
ings | +0.063 |
uggle | +0.061 |
room | +0.059 |
Neg logit contributions | |
forgot | -0.068 |
wished | -0.057 |
knew | -0.056 |
remembered | -0.053 |
learned | -0.052 |
Token | and |
Token ID | 290 |
Feature activation | -1.05e+00 |
Pos logit contributions | |
forgot | +0.161 |
known | +0.152 |
stolen | +0.142 |
forgotten | +0.141 |
unknow | +0.137 |
Neg logit contributions | |
he | -0.093 |
with | -0.085 |
ing | -0.080 |
room | -0.078 |
ore | -0.073 |
Token | Ben |
Token ID | 3932 |
Feature activation | -6.73e+00 |
Pos logit contributions | |
forgot | +0.030 |
wished | +0.025 |
knew | +0.024 |
remembered | +0.024 |
bit | +0.023 |
Neg logit contributions | |
he | -0.017 |
ing | -0.016 |
ays | -0.016 |
ings | -0.016 |
ay | -0.015 |
Token | says |
Token ID | 1139 |
Feature activation | +2.28e+00 |
Pos logit contributions | |
forgot | +0.101 |
remembered | +0.078 |
knew | +0.076 |
wished | +0.075 |
forgotten | +0.072 |
Neg logit contributions | |
he | -0.104 |
ing | -0.098 |
ore | -0.098 |
ays | -0.097 |
ings | -0.096 |
Token | . |
Token ID | 13 |
Feature activation | +5.38e+00 |
Pos logit contributions | |
he | +0.039 |
ay | +0.036 |
ays | +0.034 |
eddy | +0.033 |
uggle | +0.033 |
Neg logit contributions | |
forgot | -0.066 |
knew | -0.057 |
knocked | -0.056 |
remembered | -0.055 |
bit | -0.055 |
Token | " |
Token ID | 366 |
Feature activation | +2.88e+00 |
Pos logit contributions | |
he | +0.142 |
ach | +0.135 |
ingle | +0.132 |
uggle | +0.130 |
ating | +0.129 |
Neg logit contributions | |
forgot | -0.085 |
knew | -0.082 |
knocked | -0.074 |
wished | -0.073 |
remembered | -0.071 |
Token | It |
Token ID | 1026 |
Feature activation | -2.25e+00 |
Pos logit contributions | |
ings | +0.065 |
ing | +0.063 |
ummy | +0.062 |
ement | +0.062 |
ogging | +0.062 |
Neg logit contributions | |
forgot | -0.061 |
wished | -0.052 |
knew | -0.051 |
remembered | -0.049 |
learned | -0.048 |
Token | looks |
Token ID | 3073 |
Feature activation | +2.92e-01 |
Pos logit contributions | |
forgot | +0.063 |
wished | +0.053 |
remembered | +0.052 |
known | +0.050 |
struggled | +0.050 |
Neg logit contributions | |
ing | -0.036 |
he | -0.036 |
ays | -0.035 |
ay | -0.035 |
ore | -0.034 |
Token | y |
Token ID | 331 |
Feature activation | +8.32e+00 |
Pos logit contributions | |
he | +0.006 |
ay | +0.006 |
ing | +0.006 |
ings | +0.006 |
ore | +0.006 |
Neg logit contributions | |
forgot | -0.007 |
remembered | -0.005 |
knew | -0.005 |
wished | -0.005 |
struggled | -0.005 |
Token | ummy |
Token ID | 13513 |
Feature activation | +4.84e+00 |
Pos logit contributions | |
overed | +0.107 |
ute | +0.102 |
uggle | +0.097 |
he | +0.096 |
ane | +0.094 |
Neg logit contributions | |
forgot | -0.193 |
bit | -0.186 |
sp | -0.167 |
knew | -0.165 |
f | -0.164 |
Token | . |
Token ID | 13 |
Feature activation | +3.42e+00 |
Pos logit contributions | |
he | +0.080 |
ore | +0.075 |
room | +0.074 |
uggle | +0.074 |
ings | +0.073 |
Neg logit contributions | |
forgot | -0.137 |
bit | -0.114 |
remembered | -0.113 |
knew | -0.112 |
wished | -0.110 |
Token | Let |
Token ID | 3914 |
Feature activation | +9.77e-01 |
Pos logit contributions | |
he | +0.079 |
ing | +0.073 |
ings | +0.073 |
ore | +0.067 |
ays | +0.067 |
Neg logit contributions | |
forgot | -0.075 |
knew | -0.062 |
wished | -0.062 |
remembered | -0.061 |
bit | -0.057 |
Token | 's |
Token ID | 338 |
Feature activation | -2.65e+00 |
Pos logit contributions | |
he | +0.023 |
ing | +0.022 |
ay | +0.022 |
ings | +0.021 |
room | +0.021 |
Neg logit contributions | |
forgot | -0.021 |
remembered | -0.017 |
wished | -0.016 |
knew | -0.016 |
struggled | -0.015 |
Token | have |
Token ID | 423 |
Feature activation | +8.22e-01 |
Pos logit contributions | |
forgot | +0.059 |
wished | +0.046 |
remembered | +0.046 |
knew | +0.045 |
struggled | +0.042 |
Neg logit contributions | |
he | -0.062 |
room | -0.057 |
ay | -0.057 |
ings | -0.054 |
ore | -0.054 |
Token | the |
Token ID | 262 |
Feature activation | -5.21e-01 |
Pos logit contributions | |
ing | +0.043 |
he | +0.043 |
ore | +0.042 |
ay | +0.040 |
ays | +0.039 |
Neg logit contributions | |
forgot | -0.040 |
wished | -0.033 |
knew | -0.032 |
bit | -0.031 |
remembered | -0.031 |
Token | ground |
Token ID | 2323 |
Feature activation | +4.16e+00 |
Pos logit contributions | |
forgot | +0.011 |
remembered | +0.009 |
wished | +0.008 |
knew | +0.008 |
struggled | +0.008 |
Neg logit contributions | |
he | -0.013 |
ay | -0.012 |
ing | -0.012 |
ings | -0.012 |
room | -0.012 |
Token | . |
Token ID | 13 |
Feature activation | +3.55e+00 |
Pos logit contributions | |
ay | +0.054 |
he | +0.052 |
eddy | +0.051 |
uggle | +0.048 |
ing | +0.048 |
Neg logit contributions | |
forgot | -0.129 |
knew | -0.111 |
wished | -0.110 |
remembered | -0.108 |
bit | -0.108 |
Token | It |
Token ID | 632 |
Feature activation | -3.20e+00 |
Pos logit contributions | |
he | +0.073 |
ing | +0.069 |
over | +0.065 |
eddy | +0.064 |
uggle | +0.062 |
Neg logit contributions | |
forgot | -0.086 |
knew | -0.074 |
knocked | -0.072 |
wished | -0.071 |
remembered | -0.067 |
Token | looks |
Token ID | 3073 |
Feature activation | -7.69e-01 |
Pos logit contributions | |
forgot | +0.094 |
known | +0.078 |
remembered | +0.078 |
punished | +0.077 |
forgotten | +0.076 |
Neg logit contributions | |
ays | -0.046 |
he | -0.045 |
ore | -0.042 |
ay | -0.041 |
ing | -0.041 |
Token | y |
Token ID | 331 |
Feature activation | +8.27e+00 |
Pos logit contributions | |
forgot | +0.019 |
remembered | +0.015 |
knew | +0.014 |
wished | +0.014 |
struggled | +0.014 |
Neg logit contributions | |
he | -0.015 |
ay | -0.015 |
ing | -0.015 |
ings | -0.014 |
ore | -0.014 |
Token | ummy |
Token ID | 13513 |
Feature activation | +4.06e+00 |
Pos logit contributions | |
he | +0.103 |
overed | +0.091 |
ane | +0.088 |
ore | +0.088 |
ute | +0.086 |
Neg logit contributions | |
forgot | -0.211 |
bit | -0.193 |
knew | -0.177 |
forgotten | -0.175 |
needed | -0.173 |
Token | . |
Token ID | 13 |
Feature activation | +3.57e+00 |
Pos logit contributions | |
he | +0.056 |
room | +0.052 |
ore | +0.050 |
ay | +0.049 |
ays | +0.048 |
Neg logit contributions | |
forgot | -0.127 |
bit | -0.109 |
knew | -0.107 |
wished | -0.107 |
remembered | -0.106 |
Token |
|
Token ID | 198 |
Feature activation | +4.70e+00 |
Pos logit contributions | |
he | +0.068 |
ing | +0.062 |
over | +0.059 |
uggle | +0.059 |
eddy | +0.059 |
Neg logit contributions | |
forgot | -0.094 |
knew | -0.082 |
wished | -0.079 |
knocked | -0.077 |
remembered | -0.075 |
Token |
|
Token ID | 198 |
Feature activation | +3.68e+00 |
Pos logit contributions | |
ays | +0.092 |
ing | +0.090 |
ipping | +0.087 |
ement | +0.086 |
over | +0.085 |
Neg logit contributions | |
forgot | -0.113 |
knew | -0.103 |
wished | -0.093 |
knocked | -0.092 |
had | -0.092 |
Token | " |
Token ID | 1 |
Feature activation | +3.98e-01 |
Pos logit contributions | |
he | +0.058 |
ays | +0.055 |
ore | +0.054 |
ing | +0.054 |
ay | +0.053 |
Neg logit contributions | |
forgot | -0.109 |
knew | -0.092 |
remembered | -0.091 |
wished | -0.090 |
bit | -0.085 |
Token | cake |
Token ID | 12187 |
Feature activation | +5.34e+00 |
Pos logit contributions | |
he | +0.014 |
ing | +0.013 |
ings | +0.013 |
ore | +0.013 |
ays | +0.013 |
Neg logit contributions | |
forgot | -0.013 |
knew | -0.010 |
wished | -0.010 |
remembered | -0.010 |
bit | -0.010 |
Token | . |
Token ID | 13 |
Feature activation | +2.87e+00 |
Pos logit contributions | |
ore | +0.076 |
he | +0.076 |
overed | +0.073 |
ays | +0.072 |
ay | +0.072 |
Neg logit contributions | |
forgot | -0.155 |
bit | -0.129 |
knew | -0.128 |
remembered | -0.127 |
forget | -0.121 |
Token | It |
Token ID | 632 |
Feature activation | -2.04e+00 |
Pos logit contributions | |
he | +0.063 |
ing | +0.062 |
ings | +0.060 |
ays | +0.055 |
ore | +0.055 |
Neg logit contributions | |
forgot | -0.066 |
wished | -0.056 |
knew | -0.054 |
remembered | -0.054 |
bit | -0.053 |
Token | looks |
Token ID | 3073 |
Feature activation | -6.43e-02 |
Pos logit contributions | |
forgot | +0.063 |
remembered | +0.054 |
wished | +0.053 |
known | +0.053 |
struggled | +0.052 |
Neg logit contributions | |
he | -0.025 |
ays | -0.024 |
ore | -0.023 |
ing | -0.023 |
ay | -0.022 |
Token | so |
Token ID | 523 |
Feature activation | -5.48e-01 |
Pos logit contributions | |
forgot | +0.001 |
remembered | +0.001 |
knew | +0.001 |
wished | +0.001 |
struggled | +0.001 |
Neg logit contributions | |
he | -0.001 |
ay | -0.001 |
ing | -0.001 |
ings | -0.001 |
ore | -0.001 |
Token | y |
Token ID | 331 |
Feature activation | +8.25e+00 |
Pos logit contributions | |
forgot | +0.011 |
remembered | +0.009 |
wished | +0.009 |
knew | +0.008 |
struggled | +0.008 |
Neg logit contributions | |
he | -0.013 |
ay | -0.012 |
ore | -0.012 |
ing | -0.012 |
ummy | -0.012 |
Token | ummy |
Token ID | 13513 |
Feature activation | +5.73e+00 |
Pos logit contributions | |
overed | +0.113 |
he | +0.104 |
ute | +0.096 |
eway | +0.094 |
but | +0.094 |
Neg logit contributions | |
forgot | -0.186 |
bit | -0.177 |
f | -0.162 |
knew | -0.157 |
spoiled | -0.157 |
Token | . |
Token ID | 13 |
Feature activation | +2.74e+00 |
Pos logit contributions | |
he | +0.095 |
ings | +0.090 |
room | +0.089 |
ay | +0.087 |
uggle | +0.084 |
Neg logit contributions | |
forgot | -0.161 |
knew | -0.136 |
bit | -0.136 |
wished | -0.135 |
remembered | -0.134 |
Token | Can |
Token ID | 1680 |
Feature activation | +2.53e+00 |
Pos logit contributions | |
he | +0.063 |
ing | +0.059 |
ings | +0.059 |
ays | +0.055 |
ore | +0.055 |
Neg logit contributions | |
forgot | -0.063 |
wished | -0.052 |
remembered | -0.051 |
knew | -0.051 |
bit | -0.049 |
Token | we |
Token ID | 356 |
Feature activation | -2.99e+00 |
Pos logit contributions | |
he | +0.060 |
ings | +0.057 |
ing | +0.057 |
ay | +0.056 |
ore | +0.056 |
Neg logit contributions | |
forgot | -0.057 |
remembered | -0.043 |
knew | -0.043 |
wished | -0.043 |
bit | -0.041 |
Token | have |
Token ID | 423 |
Feature activation | +3.27e-01 |
Pos logit contributions | |
forgot | +0.062 |
wished | +0.050 |
remembered | +0.048 |
knew | +0.047 |
struggled | +0.046 |
Neg logit contributions | |
he | -0.074 |
ay | -0.068 |
room | -0.066 |
ore | -0.065 |
uggle | -0.065 |
Token | jam |
Token ID | 16853 |
Feature activation | +4.71e+00 |
Pos logit contributions | |
he | +0.036 |
ore | +0.034 |
uggle | +0.034 |
ing | +0.032 |
eddy | +0.032 |
Neg logit contributions | |
forgot | -0.035 |
knew | -0.028 |
remembered | -0.028 |
bit | -0.027 |
wished | -0.026 |
Token | . |
Token ID | 13 |
Feature activation | +4.27e+00 |
Pos logit contributions | |
he | +0.057 |
ay | +0.057 |
uggle | +0.057 |
ore | +0.055 |
room | +0.054 |
Neg logit contributions | |
forgot | -0.146 |
knew | -0.124 |
remembered | -0.123 |
wished | -0.122 |
bit | -0.122 |
Token | " |
Token ID | 366 |
Feature activation | +2.07e+00 |
Pos logit contributions | |
he | +0.090 |
ing | +0.085 |
uggle | +0.082 |
ings | +0.081 |
ays | +0.081 |
Neg logit contributions | |
forgot | -0.105 |
knew | -0.084 |
knocked | -0.081 |
remembered | -0.081 |
wished | -0.078 |
Token | It |
Token ID | 1026 |
Feature activation | -2.20e+00 |
Pos logit contributions | |
he | +0.041 |
ings | +0.040 |
ing | +0.040 |
ay | +0.039 |
ore | +0.039 |
Neg logit contributions | |
forgot | -0.051 |
wished | -0.042 |
knew | -0.042 |
remembered | -0.041 |
bit | -0.040 |
Token | looks |
Token ID | 3073 |
Feature activation | -5.38e-01 |
Pos logit contributions | |
forgot | +0.054 |
wished | +0.045 |
remembered | +0.043 |
knew | +0.042 |
known | +0.042 |
Neg logit contributions | |
ing | -0.043 |
he | -0.042 |
ays | -0.041 |
ore | -0.040 |
ay | -0.040 |
Token | y |
Token ID | 331 |
Feature activation | +8.25e+00 |
Pos logit contributions | |
forgot | +0.012 |
remembered | +0.009 |
knew | +0.009 |
wished | +0.009 |
struggled | +0.008 |
Neg logit contributions | |
he | -0.012 |
ing | -0.012 |
ay | -0.011 |
ings | -0.011 |
ore | -0.011 |
Token | ummy |
Token ID | 13513 |
Feature activation | +5.29e+00 |
Pos logit contributions | |
overed | +0.107 |
he | +0.105 |
ute | +0.100 |
uggle | +0.096 |
ane | +0.090 |
Neg logit contributions | |
forgot | -0.179 |
bit | -0.172 |
packed | -0.156 |
needed | -0.154 |
knew | -0.153 |
Token | !" |
Token ID | 2474 |
Feature activation | +3.53e+00 |
Pos logit contributions | |
he | +0.091 |
room | +0.088 |
uggle | +0.085 |
ore | +0.085 |
ings | +0.085 |
Neg logit contributions | |
forgot | -0.144 |
bit | -0.120 |
remembered | -0.120 |
wished | -0.118 |
knew | -0.118 |
Token |
|
Token ID | 198 |
Feature activation | +4.52e+00 |
Pos logit contributions | |
ays | +0.057 |
ing | +0.057 |
uggle | +0.056 |
he | +0.056 |
ings | +0.055 |
Neg logit contributions | |
forgot | -0.102 |
knew | -0.084 |
remembered | -0.081 |
wished | -0.080 |
learned | -0.080 |
Token |
|
Token ID | 198 |
Feature activation | +3.54e+00 |
Pos logit contributions | |
owers | +0.101 |
ays | +0.100 |
ipping | +0.098 |
ane | +0.096 |
ement | +0.095 |
Neg logit contributions | |
forgot | -0.092 |
knew | -0.082 |
told | -0.072 |
learned | -0.072 |
had | -0.071 |
Token | They |
Token ID | 2990 |
Feature activation | -7.71e+00 |
Pos logit contributions | |
he | +0.067 |
ore | +0.066 |
ays | +0.065 |
ings | +0.063 |
uggle | +0.062 |
Neg logit contributions | |
forgot | -0.096 |
remembered | -0.077 |
knew | -0.076 |
wished | -0.073 |
learned | -0.072 |
Token | red |
Token ID | 2266 |
Feature activation | +4.55e+00 |
Pos logit contributions | |
forgot | +0.028 |
remembered | +0.022 |
wished | +0.022 |
knew | +0.022 |
struggled | +0.020 |
Neg logit contributions | |
he | -0.023 |
ings | -0.022 |
ay | -0.022 |
ing | -0.022 |
room | -0.021 |
Token | and |
Token ID | 290 |
Feature activation | -3.32e+00 |
Pos logit contributions | |
he | +0.084 |
ore | +0.077 |
uggle | +0.074 |
ings | +0.073 |
over | +0.072 |
Neg logit contributions | |
forgot | -0.114 |
bit | -0.100 |
knew | -0.095 |
remembered | -0.093 |
wished | -0.092 |
Token | shiny |
Token ID | 22441 |
Feature activation | +3.63e+00 |
Pos logit contributions | |
forgot | +0.077 |
known | +0.063 |
remembered | +0.061 |
knew | +0.060 |
regretted | +0.060 |
Neg logit contributions | |
ay | -0.061 |
he | -0.061 |
with | -0.060 |
ings | -0.059 |
ing | -0.058 |
Token | and |
Token ID | 290 |
Feature activation | -3.49e+00 |
Pos logit contributions | |
he | +0.058 |
uggle | +0.054 |
ore | +0.053 |
room | +0.051 |
ay | +0.050 |
Neg logit contributions | |
forgot | -0.106 |
bit | -0.092 |
wished | -0.091 |
remembered | -0.090 |
knew | -0.090 |
Token | looks |
Token ID | 3073 |
Feature activation | -5.63e-01 |
Pos logit contributions | |
forgot | +0.073 |
known | +0.058 |
remembered | +0.057 |
regretted | +0.056 |
knew | +0.055 |
Neg logit contributions | |
ay | -0.073 |
he | -0.072 |
with | -0.070 |
ing | -0.070 |
ings | -0.068 |
Token | y |
Token ID | 331 |
Feature activation | +8.24e+00 |
Pos logit contributions | |
forgot | +0.013 |
remembered | +0.011 |
knew | +0.010 |
struggled | +0.010 |
wished | +0.010 |
Neg logit contributions | |
he | -0.011 |
ing | -0.010 |
ings | -0.010 |
ay | -0.010 |
anymore | -0.010 |
Token | ummy |
Token ID | 13513 |
Feature activation | +4.15e+00 |
Pos logit contributions | |
he | +0.101 |
overed | +0.085 |
uggle | +0.085 |
ore | +0.083 |
ane | +0.082 |
Neg logit contributions | |
forgot | -0.225 |
bit | -0.203 |
forgotten | -0.192 |
knew | -0.190 |
needed | -0.185 |
Token | . |
Token ID | 13 |
Feature activation | +3.29e+00 |
Pos logit contributions | |
he | +0.054 |
room | +0.049 |
ore | +0.048 |
ay | +0.047 |
uggle | +0.046 |
Neg logit contributions | |
forgot | -0.133 |
bit | -0.114 |
wished | -0.113 |
knew | -0.112 |
remembered | -0.112 |
Token | Ben |
Token ID | 3932 |
Feature activation | -5.29e+00 |
Pos logit contributions | |
he | +0.079 |
uggle | +0.072 |
ay | +0.072 |
ing | +0.071 |
ore | +0.071 |
Neg logit contributions | |
forgot | -0.074 |
knew | -0.058 |
wished | -0.057 |
remembered | -0.057 |
knocked | -0.052 |
Token | says |
Token ID | 1139 |
Feature activation | +4.06e-01 |
Pos logit contributions | |
forgot | +0.172 |
known | +0.149 |
forgotten | +0.143 |
remembered | +0.143 |
wished | +0.141 |
Neg logit contributions | |
ing | -0.052 |
he | -0.052 |
ays | -0.050 |
ore | -0.049 |
ings | -0.047 |
Token | , |
Token ID | 11 |
Feature activation | +1.32e+00 |
Pos logit contributions | |
he | +0.006 |
ays | +0.005 |
ay | +0.005 |
ings | +0.005 |
ore | +0.005 |
Neg logit contributions | |
forgot | -0.013 |
remembered | -0.011 |
knew | -0.011 |
wished | -0.011 |
bit | -0.010 |
Token | Max |
Token ID | 5436 |
Feature activation | -4.90e+00 |
Pos logit contributions | |
forgot | +0.069 |
wished | +0.057 |
knew | +0.055 |
remembered | +0.055 |
wasted | +0.055 |
Neg logit contributions | |
he | -0.037 |
ing | -0.036 |
time | -0.032 |
ay | -0.032 |
room | -0.032 |
Token | thought |
Token ID | 1807 |
Feature activation | +1.31e+00 |
Pos logit contributions | |
forgot | +0.074 |
known | +0.059 |
forget | +0.055 |
elves | +0.055 |
wished | +0.054 |
Neg logit contributions | |
he | -0.119 |
ing | -0.119 |
ings | -0.115 |
time | -0.113 |
with | -0.112 |
Token | it |
Token ID | 340 |
Feature activation | -9.04e-01 |
Pos logit contributions | |
he | +0.022 |
ings | +0.020 |
ore | +0.020 |
ing | +0.020 |
ay | +0.020 |
Neg logit contributions | |
forgot | -0.037 |
knew | -0.031 |
remembered | -0.031 |
wished | -0.030 |
knocked | -0.029 |
Token | was |
Token ID | 373 |
Feature activation | -1.79e+00 |
Pos logit contributions | |
forgot | +0.019 |
remembered | +0.015 |
wished | +0.015 |
knew | +0.014 |
forget | +0.013 |
Neg logit contributions | |
he | -0.022 |
ing | -0.021 |
ings | -0.020 |
ore | -0.020 |
room | -0.020 |
Token | too |
Token ID | 1165 |
Feature activation | +4.90e-01 |
Pos logit contributions | |
forgot | +0.042 |
wished | +0.034 |
remembered | +0.033 |
knew | +0.032 |
struggled | +0.031 |
Neg logit contributions | |
ings | -0.033 |
time | -0.033 |
ummy | -0.033 |
ing | -0.033 |
anymore | -0.033 |
Token | sm |
Token ID | 895 |
Feature activation | +8.23e+00 |
Pos logit contributions | |
he | +0.013 |
ing | +0.013 |
ore | +0.013 |
ay | +0.013 |
ings | +0.012 |
Neg logit contributions | |
forgot | -0.009 |
remembered | -0.007 |
wished | -0.007 |
knew | -0.006 |
knocked | -0.006 |
Token | elly |
Token ID | 6148 |
Feature activation | +4.96e+00 |
Pos logit contributions | |
overed | +0.113 |
ings | +0.111 |
ement | +0.107 |
ute | +0.106 |
ently | +0.099 |
Neg logit contributions | |
forgot | -0.206 |
cared | -0.180 |
knew | -0.172 |
bit | -0.171 |
forgotten | -0.166 |
Token | and |
Token ID | 290 |
Feature activation | -4.09e+00 |
Pos logit contributions | |
he | +0.076 |
ay | +0.072 |
room | +0.070 |
uggle | +0.068 |
ays | +0.066 |
Neg logit contributions | |
forgot | -0.144 |
knew | -0.126 |
wished | -0.122 |
bit | -0.121 |
remembered | -0.120 |
Token | didn |
Token ID | 1422 |
Feature activation | +3.91e+00 |
Pos logit contributions | |
forgot | +0.089 |
known | +0.076 |
remembered | +0.074 |
shown | +0.072 |
wished | +0.072 |
Neg logit contributions | |
time | -0.071 |
ing | -0.071 |
he | -0.068 |
with | -0.065 |
ore | -0.063 |
Token | 't |
Token ID | 470 |
Feature activation | -1.47e+00 |
Pos logit contributions | |
he | +0.086 |
ing | +0.086 |
room | +0.084 |
over | +0.083 |
ays | +0.083 |
Neg logit contributions | |
forgot | -0.079 |
knew | -0.064 |
knocked | -0.062 |
cost | -0.060 |
never | -0.059 |
Token | want |
Token ID | 765 |
Feature activation | +2.17e+00 |
Pos logit contributions | |
forgot | +0.031 |
wished | +0.027 |
remembered | +0.026 |
regretted | +0.024 |
pap | +0.024 |
Neg logit contributions | |
he | -0.032 |
ore | -0.029 |
ing | -0.029 |
anymore | -0.029 |
ings | -0.028 |
Token | " |
Token ID | 1 |
Feature activation | -1.32e-01 |
Pos logit contributions | |
ore | +0.044 |
ings | +0.044 |
he | +0.044 |
ays | +0.044 |
ing | +0.042 |
Neg logit contributions | |
forgot | -0.081 |
remembered | -0.067 |
knew | -0.066 |
wished | -0.065 |
learned | -0.063 |
Token | Wow |
Token ID | 22017 |
Feature activation | +3.02e+00 |
Pos logit contributions | |
forgot | +0.004 |
remembered | +0.003 |
knew | +0.003 |
wished | +0.003 |
knocked | +0.003 |
Neg logit contributions | |
he | -0.002 |
room | -0.002 |
ay | -0.002 |
ing | -0.002 |
ore | -0.002 |
Token | , |
Token ID | 11 |
Feature activation | -4.50e-01 |
Pos logit contributions | |
he | +0.051 |
ore | +0.049 |
eddy | +0.048 |
uggle | +0.046 |
ays | +0.046 |
Neg logit contributions | |
forgot | -0.088 |
remembered | -0.074 |
knew | -0.071 |
bit | -0.068 |
knocked | -0.067 |
Token | it |
Token ID | 340 |
Feature activation | -1.82e+00 |
Pos logit contributions | |
forgot | +0.014 |
remembered | +0.012 |
struggled | +0.012 |
known | +0.012 |
knew | +0.012 |
Neg logit contributions | |
he | -0.006 |
ay | -0.005 |
room | -0.005 |
ing | -0.005 |
ore | -0.005 |
Token | looks |
Token ID | 3073 |
Feature activation | +6.87e-02 |
Pos logit contributions | |
forgot | +0.045 |
wished | +0.037 |
remembered | +0.036 |
knew | +0.035 |
forgotten | +0.035 |
Neg logit contributions | |
he | -0.037 |
ing | -0.035 |
ay | -0.034 |
room | -0.033 |
ings | -0.033 |
Token | y |
Token ID | 331 |
Feature activation | +8.23e+00 |
Pos logit contributions | |
he | +0.002 |
ing | +0.002 |
ore | +0.002 |
ay | +0.002 |
ings | +0.002 |
Neg logit contributions | |
forgot | -0.001 |
remembered | -0.001 |
knew | -0.001 |
wished | -0.001 |
bit | -0.001 |
Token | ummy |
Token ID | 13513 |
Feature activation | +5.30e+00 |
Pos logit contributions | |
ute | +0.094 |
he | +0.094 |
overed | +0.092 |
ane | +0.089 |
uggle | +0.088 |
Neg logit contributions | |
forgot | -0.215 |
bit | -0.199 |
knew | -0.175 |
sp | -0.171 |
ate | -0.170 |
Token | ," |
Token ID | 553 |
Feature activation | +1.19e+00 |
Pos logit contributions | |
he | +0.090 |
ore | +0.086 |
room | +0.086 |
uggle | +0.083 |
ays | +0.082 |
Neg logit contributions | |
forgot | -0.145 |
remembered | -0.121 |
bit | -0.120 |
knew | -0.118 |
wished | -0.115 |
Token | Amy |
Token ID | 14235 |
Feature activation | -4.55e+00 |
Pos logit contributions | |
he | +0.017 |
ay | +0.014 |
ing | +0.014 |
room | +0.014 |
ore | +0.014 |
Neg logit contributions | |
forgot | -0.037 |
remembered | -0.031 |
knew | -0.031 |
wished | -0.031 |
forgotten | -0.029 |
Token | said |
Token ID | 531 |
Feature activation | +2.53e+00 |
Pos logit contributions | |
forgot | +0.082 |
known | +0.060 |
remembered | +0.060 |
forgotten | +0.057 |
wished | +0.057 |
Neg logit contributions | |
he | -0.117 |
ing | -0.112 |
ore | -0.107 |
room | -0.107 |
with | -0.105 |
Token | . |
Token ID | 13 |
Feature activation | +5.00e+00 |
Pos logit contributions | |
he | +0.043 |
ay | +0.042 |
ays | +0.040 |
uggle | +0.038 |
ore | +0.037 |
Neg logit contributions | |
forgot | -0.071 |
knew | -0.060 |
remembered | -0.060 |
bit | -0.059 |
knocked | -0.058 |
Token | is |
Token ID | 318 |
Feature activation | +4.63e-01 |
Pos logit contributions | |
he | +0.010 |
ore | +0.009 |
ings | +0.009 |
ing | +0.009 |
ays | +0.009 |
Neg logit contributions | |
forgot | -0.009 |
remembered | -0.007 |
knew | -0.007 |
wished | -0.007 |
learned | -0.006 |
Token | this |
Token ID | 428 |
Feature activation | -2.17e-01 |
Pos logit contributions | |
he | +0.009 |
ay | +0.008 |
ing | +0.008 |
ings | +0.008 |
ore | +0.008 |
Neg logit contributions | |
forgot | -0.012 |
remembered | -0.010 |
knew | -0.010 |
wished | -0.010 |
knocked | -0.009 |
Token | ? |
Token ID | 30 |
Feature activation | +2.39e+00 |
Pos logit contributions | |
forgot | +0.007 |
remembered | +0.006 |
knew | +0.006 |
wished | +0.005 |
bit | +0.005 |
Neg logit contributions | |
he | -0.003 |
ay | -0.003 |
ing | -0.003 |
room | -0.003 |
ore | -0.003 |
Token | It |
Token ID | 632 |
Feature activation | -2.14e+00 |
Pos logit contributions | |
he | +0.046 |
ing | +0.041 |
ings | +0.041 |
ay | +0.040 |
room | +0.038 |
Neg logit contributions | |
forgot | -0.063 |
remembered | -0.052 |
wished | -0.052 |
knew | -0.052 |
bit | -0.050 |
Token | looks |
Token ID | 3073 |
Feature activation | +4.19e-01 |
Pos logit contributions | |
forgot | +0.060 |
wished | +0.050 |
remembered | +0.049 |
known | +0.048 |
forgotten | +0.047 |
Neg logit contributions | |
ing | -0.035 |
ore | -0.034 |
ays | -0.033 |
he | -0.032 |
ings | -0.031 |
Token | y |
Token ID | 331 |
Feature activation | +8.22e+00 |
Pos logit contributions | |
he | +0.010 |
ing | +0.010 |
ay | +0.010 |
ore | +0.010 |
ings | +0.009 |
Neg logit contributions | |
forgot | -0.009 |
remembered | -0.007 |
knew | -0.007 |
wished | -0.007 |
knocked | -0.006 |
Token | ucky |
Token ID | 5309 |
Feature activation | +4.23e+00 |
Pos logit contributions | |
he | +0.111 |
overed | +0.107 |
ute | +0.095 |
ane | +0.092 |
uggle | +0.090 |
Neg logit contributions | |
forgot | -0.195 |
bit | -0.187 |
sp | -0.169 |
f | -0.169 |
knew | -0.166 |
Token | . |
Token ID | 13 |
Feature activation | +2.52e+00 |
Pos logit contributions | |
he | +0.071 |
ay | +0.065 |
room | +0.065 |
uggle | +0.062 |
ings | +0.062 |
Neg logit contributions | |
forgot | -0.119 |
bit | -0.099 |
remembered | -0.098 |
knew | -0.098 |
wished | -0.096 |
Token | I |
Token ID | 314 |
Feature activation | -4.11e+00 |
Pos logit contributions | |
he | +0.051 |
ing | +0.045 |
ings | +0.045 |
ay | +0.043 |
ore | +0.042 |
Neg logit contributions | |
forgot | -0.064 |
remembered | -0.053 |
wished | -0.053 |
knew | -0.053 |
bit | -0.051 |
Token | don |
Token ID | 836 |
Feature activation | +4.43e+00 |
Pos logit contributions | |
forgot | +0.086 |
pap | +0.069 |
wished | +0.069 |
robbers | +0.068 |
Indians | +0.067 |
Neg logit contributions | |
he | -0.090 |
ore | -0.089 |
ings | -0.077 |
with | -0.076 |
uggle | -0.074 |
Token | 't |
Token ID | 470 |
Feature activation | -2.48e+00 |
Pos logit contributions | |
ment | +0.110 |
over | +0.107 |
ings | +0.102 |
oos | +0.102 |
ement | +0.102 |
Neg logit contributions | |
forgot | -0.069 |
knew | -0.058 |
knocked | -0.057 |
cost | -0.051 |
never | -0.051 |
Token | red |
Token ID | 2266 |
Feature activation | +6.26e-01 |
Pos logit contributions | |
forgot | +0.025 |
remembered | +0.019 |
wished | +0.019 |
knew | +0.019 |
learned | +0.018 |
Neg logit contributions | |
he | -0.021 |
ay | -0.020 |
ings | -0.020 |
ore | -0.020 |
ing | -0.019 |
Token | strawberry |
Token ID | 41236 |
Feature activation | +3.16e+00 |
Pos logit contributions | |
he | +0.015 |
ays | +0.014 |
ing | +0.014 |
ore | +0.014 |
ings | +0.014 |
Neg logit contributions | |
forgot | -0.012 |
bit | -0.010 |
knew | -0.010 |
remembered | -0.010 |
wished | -0.009 |
Token | . |
Token ID | 13 |
Feature activation | +3.66e+00 |
Pos logit contributions | |
he | +0.042 |
ay | +0.039 |
room | +0.038 |
ore | +0.037 |
uggle | +0.037 |
Neg logit contributions | |
forgot | -0.099 |
knew | -0.085 |
bit | -0.084 |
remembered | -0.083 |
wished | -0.082 |
Token | It |
Token ID | 632 |
Feature activation | -2.96e+00 |
Pos logit contributions | |
he | +0.075 |
ing | +0.071 |
ings | +0.068 |
ore | +0.067 |
ently | +0.066 |
Neg logit contributions | |
forgot | -0.091 |
knew | -0.072 |
remembered | -0.072 |
wished | -0.069 |
knocked | -0.069 |
Token | looks |
Token ID | 3073 |
Feature activation | -6.09e-01 |
Pos logit contributions | |
forgot | +0.083 |
wished | +0.067 |
punished | +0.067 |
remembered | +0.067 |
known | +0.065 |
Neg logit contributions | |
he | -0.048 |
ays | -0.046 |
ay | -0.045 |
ing | -0.045 |
ore | -0.045 |
Token | y |
Token ID | 331 |
Feature activation | +8.21e+00 |
Pos logit contributions | |
forgot | +0.015 |
remembered | +0.012 |
knew | +0.012 |
wished | +0.012 |
struggled | +0.011 |
Neg logit contributions | |
he | -0.012 |
ay | -0.011 |
ing | -0.011 |
ings | -0.011 |
ore | -0.011 |
Token | ummy |
Token ID | 13513 |
Feature activation | +4.68e+00 |
Pos logit contributions | |
he | +0.103 |
overed | +0.092 |
ane | +0.091 |
ute | +0.090 |
ore | +0.088 |
Neg logit contributions | |
forgot | -0.203 |
bit | -0.187 |
knew | -0.171 |
forgotten | -0.169 |
needed | -0.167 |
Token | . |
Token ID | 13 |
Feature activation | +3.69e+00 |
Pos logit contributions | |
he | +0.072 |
room | +0.069 |
ay | +0.066 |
uggle | +0.065 |
ore | +0.064 |
Neg logit contributions | |
forgot | -0.137 |
bit | -0.120 |
knew | -0.117 |
remembered | -0.116 |
wished | -0.115 |
Token |
|
Token ID | 198 |
Feature activation | +4.85e+00 |
Pos logit contributions | |
he | +0.074 |
ing | +0.067 |
ings | +0.067 |
ore | +0.066 |
ay | +0.066 |
Neg logit contributions | |
forgot | -0.097 |
knew | -0.076 |
remembered | -0.076 |
wished | -0.075 |
learned | -0.072 |
Token |
|
Token ID | 198 |
Feature activation | +3.76e+00 |
Pos logit contributions | |
ays | +0.095 |
ement | +0.091 |
ipping | +0.090 |
ing | +0.090 |
ment | +0.089 |
Neg logit contributions | |
forgot | -0.120 |
knew | -0.101 |
had | -0.098 |
learned | -0.095 |
remembered | -0.092 |
Token | " |
Token ID | 1 |
Feature activation | +1.36e-01 |
Pos logit contributions | |
he | +0.052 |
ore | +0.048 |
ings | +0.047 |
ing | +0.047 |
ays | +0.047 |
Neg logit contributions | |
forgot | -0.120 |
remembered | -0.101 |
knew | -0.100 |
wished | -0.099 |
bit | -0.095 |
Token | on |
Token ID | 319 |
Feature activation | +3.13e+00 |
Pos logit contributions | |
ore | +0.066 |
he | +0.065 |
uggle | +0.063 |
room | +0.061 |
ings | +0.059 |
Neg logit contributions | |
forgot | -0.162 |
remembered | -0.136 |
knew | -0.135 |
wished | -0.129 |
bit | -0.129 |
Token | it |
Token ID | 340 |
Feature activation | +2.40e+00 |
Pos logit contributions | |
ore | +0.064 |
ing | +0.063 |
ogging | +0.062 |
he | +0.062 |
uggle | +0.062 |
Neg logit contributions | |
forgot | -0.071 |
knew | -0.059 |
remembered | -0.057 |
wished | -0.057 |
bit | -0.055 |
Token | . |
Token ID | 13 |
Feature activation | +3.46e+00 |
Pos logit contributions | |
he | +0.027 |
ays | +0.026 |
ing | +0.025 |
ore | +0.025 |
eddy | +0.025 |
Neg logit contributions | |
forgot | -0.079 |
knew | -0.067 |
remembered | -0.066 |
wished | -0.066 |
bit | -0.064 |
Token | It |
Token ID | 632 |
Feature activation | -4.02e+00 |
Pos logit contributions | |
he | +0.059 |
ing | +0.056 |
ings | +0.055 |
ore | +0.052 |
ay | +0.051 |
Neg logit contributions | |
forgot | -0.096 |
wished | -0.082 |
knew | -0.081 |
remembered | -0.080 |
learned | -0.076 |
Token | looked |
Token ID | 3114 |
Feature activation | -1.06e-01 |
Pos logit contributions | |
forgot | +0.091 |
robbers | +0.082 |
punished | +0.076 |
known | +0.076 |
stolen | +0.075 |
Neg logit contributions | |
he | -0.073 |
with | -0.070 |
time | -0.062 |
room | -0.062 |
ays | -0.060 |
Token | y |
Token ID | 331 |
Feature activation | +8.20e+00 |
Pos logit contributions | |
forgot | +0.003 |
remembered | +0.002 |
knew | +0.002 |
wished | +0.002 |
knocked | +0.002 |
Neg logit contributions | |
he | -0.002 |
ing | -0.002 |
ay | -0.002 |
ore | -0.002 |
ings | -0.002 |
Token | ummy |
Token ID | 13513 |
Feature activation | +3.21e+00 |
Pos logit contributions | |
he | +0.102 |
ute | +0.095 |
ore | +0.093 |
overed | +0.091 |
uggle | +0.089 |
Neg logit contributions | |
forgot | -0.205 |
bit | -0.194 |
knew | -0.175 |
f | -0.171 |
mixed | -0.171 |
Token | and |
Token ID | 290 |
Feature activation | -3.33e+00 |
Pos logit contributions | |
he | +0.046 |
room | +0.041 |
ore | +0.041 |
ays | +0.040 |
ings | +0.040 |
Neg logit contributions | |
forgot | -0.099 |
knew | -0.083 |
remembered | -0.083 |
wished | -0.082 |
bit | -0.082 |
Token | sweet |
Token ID | 6029 |
Feature activation | +2.07e+00 |
Pos logit contributions | |
forgot | +0.065 |
remembered | +0.052 |
wished | +0.051 |
knew | +0.050 |
struggled | +0.050 |
Neg logit contributions | |
he | -0.073 |
ing | -0.071 |
with | -0.068 |
ay | -0.068 |
time | -0.067 |
Token | . |
Token ID | 13 |
Feature activation | +3.31e+00 |
Pos logit contributions | |
he | +0.031 |
ore | +0.028 |
room | +0.027 |
ays | +0.027 |
ay | +0.027 |
Neg logit contributions | |
forgot | -0.064 |
knew | -0.053 |
remembered | -0.053 |
wished | -0.053 |
bit | -0.052 |
Token |
|
Token ID | 198 |
Feature activation | +4.29e+00 |
Pos logit contributions | |
he | +0.057 |
ings | +0.052 |
ing | +0.052 |
ay | +0.049 |
room | +0.049 |
Neg logit contributions | |
forgot | -0.093 |
wished | -0.080 |
knew | -0.079 |
remembered | -0.078 |
learned | -0.073 |
Token | the |
Token ID | 262 |
Feature activation | -1.70e+00 |
Pos logit contributions | |
he | +0.047 |
ing | +0.042 |
ings | +0.042 |
room | +0.042 |
ore | +0.042 |
Neg logit contributions | |
forgot | -0.058 |
remembered | -0.047 |
knew | -0.047 |
wished | -0.046 |
bit | -0.044 |
Token | apple |
Token ID | 17180 |
Feature activation | +5.50e+00 |
Pos logit contributions | |
forgot | +0.043 |
remembered | +0.034 |
knew | +0.034 |
wished | +0.034 |
bit | +0.032 |
Neg logit contributions | |
he | -0.035 |
ore | -0.033 |
ing | -0.032 |
ings | -0.032 |
room | -0.032 |
Token | . |
Token ID | 13 |
Feature activation | +2.06e+00 |
Pos logit contributions | |
ement | +0.094 |
overed | +0.085 |
he | +0.084 |
over | +0.082 |
ment | +0.082 |
Neg logit contributions | |
forgot | -0.144 |
bit | -0.128 |
knew | -0.118 |
knocked | -0.113 |
remembered | -0.113 |
Token | It |
Token ID | 632 |
Feature activation | -2.02e+00 |
Pos logit contributions | |
he | +0.037 |
ing | +0.035 |
ings | +0.034 |
ore | +0.032 |
ogging | +0.030 |
Neg logit contributions | |
forgot | -0.056 |
remembered | -0.047 |
bit | -0.047 |
wished | -0.046 |
knew | -0.046 |
Token | looks |
Token ID | 3073 |
Feature activation | +2.38e-02 |
Pos logit contributions | |
forgot | +0.053 |
wished | +0.044 |
remembered | +0.044 |
known | +0.043 |
forgotten | +0.042 |
Neg logit contributions | |
he | -0.035 |
ing | -0.034 |
ays | -0.033 |
ings | -0.033 |
ore | -0.032 |
Token | y |
Token ID | 331 |
Feature activation | +8.20e+00 |
Pos logit contributions | |
he | +0.001 |
ay | +0.001 |
ing | +0.001 |
ings | +0.001 |
ore | +0.001 |
Neg logit contributions | |
forgot | -0.000 |
remembered | -0.000 |
knew | -0.000 |
wished | -0.000 |
struggled | -0.000 |
Token | ummy |
Token ID | 13513 |
Feature activation | +5.55e+00 |
Pos logit contributions | |
overed | +0.106 |
he | +0.101 |
ute | +0.089 |
eway | +0.088 |
ane | +0.086 |
Neg logit contributions | |
forgot | -0.196 |
bit | -0.187 |
packed | -0.166 |
knew | -0.164 |
needed | -0.162 |
Token | ." |
Token ID | 526 |
Feature activation | +4.28e+00 |
Pos logit contributions | |
he | +0.105 |
room | +0.103 |
ings | +0.101 |
ay | +0.097 |
ore | +0.097 |
Neg logit contributions | |
forgot | -0.143 |
bit | -0.120 |
remembered | -0.115 |
knew | -0.115 |
wished | -0.115 |
Token |
|
Token ID | 198 |
Feature activation | +5.18e+00 |
Pos logit contributions | |
he | +0.122 |
ay | +0.117 |
ane | +0.114 |
ing | +0.114 |
eddy | +0.114 |
Neg logit contributions | |
forgot | -0.076 |
knew | -0.052 |
remembered | -0.051 |
learned | -0.049 |
tried | -0.045 |
Token |
|
Token ID | 198 |
Feature activation | +3.94e+00 |
Pos logit contributions | |
ement | +0.106 |
ays | +0.105 |
overed | +0.104 |
ment | +0.102 |
ane | +0.101 |
Neg logit contributions | |
forgot | -0.121 |
knew | -0.101 |
learned | -0.092 |
remembered | -0.091 |
told | -0.091 |
Token | Ben |
Token ID | 11696 |
Feature activation | -5.22e+00 |
Pos logit contributions | |
he | +0.085 |
ore | +0.084 |
ays | +0.082 |
ings | +0.081 |
ay | +0.080 |
Neg logit contributions | |
forgot | -0.096 |
remembered | -0.075 |
knew | -0.073 |
wished | -0.070 |
bit | -0.069 |
Token | table |
Token ID | 3084 |
Feature activation | +4.04e+00 |
Pos logit contributions | |
forgot | +0.034 |
remembered | +0.026 |
wished | +0.025 |
knew | +0.025 |
struggled | +0.023 |
Neg logit contributions | |
he | -0.037 |
ay | -0.034 |
ing | -0.034 |
room | -0.034 |
ore | -0.034 |
Token | . |
Token ID | 13 |
Feature activation | +2.98e+00 |
Pos logit contributions | |
he | +0.043 |
ay | +0.041 |
ing | +0.039 |
eddy | +0.038 |
uggle | +0.038 |
Neg logit contributions | |
forgot | -0.136 |
knew | -0.117 |
remembered | -0.116 |
wished | -0.116 |
bit | -0.113 |
Token | It |
Token ID | 632 |
Feature activation | -3.82e+00 |
Pos logit contributions | |
he | +0.054 |
ing | +0.052 |
ings | +0.050 |
ore | +0.049 |
eddy | +0.049 |
Neg logit contributions | |
forgot | -0.080 |
knew | -0.066 |
wished | -0.064 |
remembered | -0.064 |
knocked | -0.063 |
Token | looks |
Token ID | 3073 |
Feature activation | -1.15e+00 |
Pos logit contributions | |
forgot | +0.105 |
known | +0.086 |
remembered | +0.085 |
wished | +0.083 |
punished | +0.082 |
Neg logit contributions | |
ays | -0.061 |
he | -0.059 |
ore | -0.058 |
ay | -0.056 |
ing | -0.055 |
Token | very |
Token ID | 845 |
Feature activation | -6.75e-01 |
Pos logit contributions | |
forgot | +0.027 |
remembered | +0.020 |
knew | +0.020 |
wished | +0.020 |
struggled | +0.019 |
Neg logit contributions | |
he | -0.023 |
ing | -0.022 |
ay | -0.022 |
ings | -0.022 |
ore | -0.022 |
Token | y |
Token ID | 331 |
Feature activation | +8.19e+00 |
Pos logit contributions | |
forgot | +0.013 |
remembered | +0.010 |
wished | +0.009 |
knew | +0.009 |
knocked | +0.008 |
Neg logit contributions | |
he | -0.018 |
ing | -0.017 |
ay | -0.017 |
ore | -0.017 |
ings | -0.017 |
Token | ummy |
Token ID | 13513 |
Feature activation | +4.34e+00 |
Pos logit contributions | |
he | +0.102 |
ute | +0.088 |
overed | +0.088 |
uggle | +0.086 |
ane | +0.085 |
Neg logit contributions | |
forgot | -0.214 |
bit | -0.196 |
knew | -0.183 |
forgotten | -0.180 |
needed | -0.175 |
Token | and |
Token ID | 290 |
Feature activation | -4.73e+00 |
Pos logit contributions | |
he | +0.062 |
room | +0.057 |
ay | +0.055 |
uggle | +0.054 |
ore | +0.054 |
Neg logit contributions | |
forgot | -0.134 |
bit | -0.115 |
knew | -0.114 |
wished | -0.114 |
remembered | -0.112 |
Token | fancy |
Token ID | 14996 |
Feature activation | +2.20e+00 |
Pos logit contributions | |
forgot | +0.099 |
known | +0.082 |
remembered | +0.079 |
knew | +0.077 |
regretted | +0.076 |
Neg logit contributions | |
with | -0.092 |
he | -0.089 |
ummy | -0.088 |
ings | -0.088 |
ing | -0.087 |
Token | . |
Token ID | 13 |
Feature activation | +3.23e+00 |
Pos logit contributions | |
he | +0.027 |
ay | +0.023 |
ore | +0.022 |
uggle | +0.022 |
room | +0.021 |
Neg logit contributions | |
forgot | -0.073 |
knew | -0.062 |
remembered | -0.062 |
wished | -0.062 |
bit | -0.062 |
Token | It |
Token ID | 632 |
Feature activation | -4.55e+00 |
Pos logit contributions | |
he | +0.063 |
ing | +0.060 |
ore | +0.058 |
ings | +0.058 |
ay | +0.057 |
Neg logit contributions | |
forgot | -0.085 |
knew | -0.068 |
wished | -0.067 |
remembered | -0.066 |
knocked | -0.065 |
Token | eat |
Token ID | 4483 |
Feature activation | +4.58e+00 |
Pos logit contributions | |
forgot | +0.049 |
wished | +0.037 |
remembered | +0.037 |
knew | +0.036 |
struggled | +0.033 |
Neg logit contributions | |
he | -0.062 |
ay | -0.058 |
ore | -0.057 |
room | -0.056 |
ings | -0.055 |
Token | it |
Token ID | 340 |
Feature activation | +3.34e+00 |
Pos logit contributions | |
he | +0.106 |
ay | +0.102 |
ings | +0.101 |
room | +0.101 |
ays | +0.099 |
Neg logit contributions | |
forgot | -0.090 |
bit | -0.073 |
knew | -0.072 |
remembered | -0.071 |
wished | -0.069 |
Token | ? |
Token ID | 30 |
Feature activation | +2.53e+00 |
Pos logit contributions | |
ays | +0.061 |
ings | +0.060 |
ore | +0.058 |
aps | +0.058 |
over | +0.058 |
Neg logit contributions | |
forgot | -0.083 |
remembered | -0.069 |
knew | -0.068 |
reminded | -0.066 |
bit | -0.066 |
Token | It |
Token ID | 632 |
Feature activation | -1.66e+00 |
Pos logit contributions | |
he | +0.050 |
ings | +0.050 |
ing | +0.049 |
ays | +0.047 |
room | +0.044 |
Neg logit contributions | |
forgot | -0.062 |
knew | -0.051 |
remembered | -0.051 |
wished | -0.051 |
bit | -0.049 |
Token | looks |
Token ID | 3073 |
Feature activation | +2.00e-01 |
Pos logit contributions | |
forgot | +0.043 |
known | +0.037 |
wished | +0.036 |
forgotten | +0.036 |
robbers | +0.036 |
Neg logit contributions | |
ing | -0.025 |
he | -0.024 |
with | -0.024 |
ore | -0.022 |
time | -0.022 |
Token | y |
Token ID | 331 |
Feature activation | +8.17e+00 |
Pos logit contributions | |
he | +0.004 |
ay | +0.004 |
ing | +0.004 |
ings | +0.004 |
ore | +0.004 |
Neg logit contributions | |
forgot | -0.005 |
remembered | -0.004 |
knew | -0.003 |
wished | -0.003 |
struggled | -0.003 |
Token | ummy |
Token ID | 13513 |
Feature activation | +4.66e+00 |
Pos logit contributions | |
overed | +0.106 |
ute | +0.098 |
he | +0.095 |
ane | +0.089 |
ment | +0.088 |
Neg logit contributions | |
forgot | -0.181 |
bit | -0.180 |
packed | -0.161 |
knew | -0.156 |
needed | -0.155 |
Token | ." |
Token ID | 526 |
Feature activation | +3.88e+00 |
Pos logit contributions | |
he | +0.078 |
room | +0.076 |
ings | +0.075 |
ore | +0.072 |
ay | +0.072 |
Neg logit contributions | |
forgot | -0.129 |
bit | -0.108 |
remembered | -0.108 |
knew | -0.106 |
wished | -0.105 |
Token |
|
Token ID | 198 |
Feature activation | +4.62e+00 |
Pos logit contributions | |
he | +0.075 |
ing | +0.070 |
ays | +0.068 |
ay | +0.067 |
ings | +0.066 |
Neg logit contributions | |
forgot | -0.102 |
knew | -0.087 |
remembered | -0.083 |
wished | -0.083 |
knocked | -0.079 |
Token |
|
Token ID | 198 |
Feature activation | +3.62e+00 |
Pos logit contributions | |
ays | +0.091 |
ing | +0.084 |
owers | +0.079 |
ement | +0.078 |
over | +0.078 |
Neg logit contributions | |
forgot | -0.117 |
knew | -0.107 |
remembered | -0.099 |
learned | -0.094 |
wished | -0.093 |
Token | His |
Token ID | 6653 |
Feature activation | -1.39e+00 |
Pos logit contributions | |
ays | +0.068 |
he | +0.068 |
ing | +0.066 |
ore | +0.064 |
ings | +0.064 |
Neg logit contributions | |
forgot | -0.096 |
knew | -0.081 |
remembered | -0.080 |
wished | -0.077 |
bit | -0.073 |
Token |
|
Token ID | 198 |
Feature activation | +4.15e+00 |
Pos logit contributions | |
ing | +0.068 |
ings | +0.066 |
he | +0.066 |
ment | +0.065 |
ore | +0.064 |
Neg logit contributions | |
forgot | -0.078 |
knew | -0.067 |
learned | -0.064 |
wished | -0.064 |
remembered | -0.062 |
Token |
|
Token ID | 198 |
Feature activation | +3.07e+00 |
Pos logit contributions | |
ays | +0.080 |
ment | +0.078 |
ing | +0.078 |
ings | +0.075 |
ipping | +0.073 |
Neg logit contributions | |
forgot | -0.097 |
knew | -0.086 |
had | -0.082 |
wished | -0.081 |
learned | -0.081 |
Token | Every |
Token ID | 6109 |
Feature activation | -1.62e+00 |
Pos logit contributions | |
he | +0.061 |
ings | +0.060 |
ore | +0.060 |
ays | +0.059 |
ing | +0.059 |
Neg logit contributions | |
forgot | -0.076 |
remembered | -0.062 |
wished | -0.061 |
knew | -0.061 |
bit | -0.057 |
Token | night |
Token ID | 1755 |
Feature activation | +3.46e+00 |
Pos logit contributions | |
forgot | +0.045 |
remembered | +0.035 |
wished | +0.035 |
knew | +0.034 |
wasted | +0.034 |
Neg logit contributions | |
he | -0.030 |
ore | -0.028 |
time | -0.028 |
ay | -0.028 |
ing | -0.028 |
Token | , |
Token ID | 11 |
Feature activation | +9.93e-01 |
Pos logit contributions | |
he | +0.042 |
ore | +0.036 |
ay | +0.036 |
ing | +0.036 |
ays | +0.036 |
Neg logit contributions | |
forgot | -0.117 |
remembered | -0.099 |
knew | -0.099 |
wished | -0.097 |
knocked | -0.094 |
Token | she |
Token ID | 673 |
Feature activation | -9.46e+00 |
Pos logit contributions | |
he | +0.018 |
ing | +0.017 |
ay | +0.017 |
ings | +0.017 |
room | +0.016 |
Neg logit contributions | |
forgot | -0.027 |
remembered | -0.022 |
knew | -0.022 |
wished | -0.022 |
struggled | -0.020 |
Token | would |
Token ID | 561 |
Feature activation | -3.45e+00 |
Pos logit contributions | |
forgot | +0.156 |
stolen | +0.130 |
unknow | +0.120 |
known | +0.120 |
robbers | +0.118 |
Neg logit contributions | |
he | -0.211 |
room | -0.194 |
ore | -0.188 |
with | -0.188 |
ing | -0.188 |
Token | read |
Token ID | 1100 |
Feature activation | +3.40e+00 |
Pos logit contributions | |
forgot | +0.076 |
remembered | +0.058 |
wished | +0.057 |
knew | +0.055 |
regretted | +0.055 |
Neg logit contributions | |
he | -0.079 |
room | -0.072 |
ings | -0.070 |
uggle | -0.069 |
ore | -0.069 |
Token | a |
Token ID | 257 |
Feature activation | -1.14e+00 |
Pos logit contributions | |
he | +0.071 |
ays | +0.069 |
ore | +0.069 |
ing | +0.066 |
ings | +0.066 |
Neg logit contributions | |
forgot | -0.077 |
knew | -0.065 |
remembered | -0.063 |
wished | -0.063 |
learned | -0.060 |
Token | book |
Token ID | 1492 |
Feature activation | +4.65e+00 |
Pos logit contributions | |
forgot | +0.029 |
remembered | +0.024 |
wished | +0.023 |
knew | +0.023 |
knocked | +0.022 |
Neg logit contributions | |
he | -0.023 |
ay | -0.021 |
ore | -0.021 |
ing | -0.021 |
ings | -0.021 |
Token | that |
Token ID | 326 |
Feature activation | -3.42e+00 |
Pos logit contributions | |
ays | +0.077 |
ings | +0.072 |
eddy | +0.071 |
ore | +0.070 |
ach | +0.069 |
Neg logit contributions | |
forgot | -0.124 |
knew | -0.108 |
bit | -0.102 |
learned | -0.102 |
remembered | -0.102 |
Token | her |
Token ID | 607 |
Feature activation | -2.70e+00 |
Pos logit contributions | |
forgot | +0.020 |
knew | +0.016 |
wished | +0.016 |
remembered | +0.015 |
bit | +0.015 |
Neg logit contributions | |
he | -0.016 |
ing | -0.015 |
room | -0.015 |
time | -0.015 |
ay | -0.015 |
Token | av |
Token ID | 1196 |
Feature activation | +3.66e+00 |
Pos logit contributions | |
forgot | +0.066 |
wished | +0.053 |
remembered | +0.052 |
knew | +0.051 |
knocked | +0.048 |
Neg logit contributions | |
he | -0.055 |
ay | -0.052 |
ings | -0.051 |
ing | -0.051 |
ummy | -0.050 |
Token | oc |
Token ID | 420 |
Feature activation | +3.97e+00 |
Pos logit contributions | |
he | +0.058 |
ay | +0.054 |
ing | +0.053 |
ings | +0.052 |
ore | +0.052 |
Neg logit contributions | |
forgot | -0.108 |
remembered | -0.090 |
wished | -0.089 |
knew | -0.089 |
knocked | -0.085 |
Token | ados |
Token ID | 22484 |
Feature activation | +2.86e+00 |
Pos logit contributions | |
he | +0.049 |
ore | +0.042 |
ay | +0.039 |
ing | +0.039 |
room | +0.038 |
Neg logit contributions | |
forgot | -0.130 |
remembered | -0.113 |
knew | -0.112 |
wished | -0.110 |
learned | -0.105 |
Token | . |
Token ID | 13 |
Feature activation | +2.26e+00 |
Pos logit contributions | |
he | +0.034 |
ay | +0.032 |
ore | +0.032 |
ings | +0.032 |
ays | +0.031 |
Neg logit contributions | |
forgot | -0.095 |
remembered | -0.081 |
knew | -0.080 |
wished | -0.079 |
bit | -0.076 |
Token | She |
Token ID | 1375 |
Feature activation | -9.33e+00 |
Pos logit contributions | |
he | +0.032 |
ing | +0.028 |
ore | +0.028 |
ings | +0.028 |
ay | +0.027 |
Neg logit contributions | |
forgot | -0.073 |
knew | -0.061 |
wished | -0.060 |
remembered | -0.060 |
missed | -0.058 |
Token | thanked |
Token ID | 26280 |
Feature activation | +1.23e+00 |
Pos logit contributions | |
chie | +0.157 |
mate | +0.146 |
robbers | +0.139 |
supervision | +0.134 |
brav | +0.134 |
Neg logit contributions | |
with | -0.188 |
and | -0.179 |
he | -0.168 |
at | -0.165 |
, | -0.165 |
Token | her |
Token ID | 607 |
Feature activation | -2.52e-01 |
Pos logit contributions | |
he | +0.021 |
ing | +0.019 |
ay | +0.019 |
ore | +0.019 |
ings | +0.019 |
Neg logit contributions | |
forgot | -0.036 |
remembered | -0.029 |
wished | -0.029 |
knew | -0.029 |
bit | -0.027 |
Token | mom |
Token ID | 1995 |
Feature activation | +2.04e+00 |
Pos logit contributions | |
forgot | +0.006 |
remembered | +0.004 |
wished | +0.004 |
knew | +0.004 |
struggled | +0.004 |
Neg logit contributions | |
he | -0.006 |
ing | -0.005 |
ay | -0.005 |
ings | -0.005 |
ore | -0.005 |
Token | and |
Token ID | 290 |
Feature activation | -5.85e+00 |
Pos logit contributions | |
he | +0.038 |
ings | +0.037 |
ays | +0.037 |
ore | +0.036 |
room | +0.036 |
Neg logit contributions | |
forgot | -0.053 |
knew | -0.045 |
remembered | -0.044 |
learned | -0.041 |
wished | -0.041 |
Token | ate |
Token ID | 15063 |
Feature activation | +2.60e+00 |
Pos logit contributions | |
forgot | +0.103 |
chie | +0.090 |
'd | +0.088 |
fact | +0.087 |
smallest | +0.087 |
Neg logit contributions | |
he | -0.117 |
with | -0.117 |
ing | -0.110 |
at | -0.104 |
together | -0.103 |
Token | difficult |
Token ID | 2408 |
Feature activation | +3.46e+00 |
Pos logit contributions | |
forgot | +0.053 |
wished | +0.044 |
remembered | +0.042 |
knew | +0.041 |
cared | +0.037 |
Neg logit contributions | |
he | -0.058 |
ing | -0.057 |
ings | -0.055 |
time | -0.055 |
ay | -0.053 |
Token | . |
Token ID | 13 |
Feature activation | +2.62e+00 |
Pos logit contributions | |
he | +0.041 |
ay | +0.034 |
ore | +0.034 |
ings | +0.033 |
room | +0.033 |
Neg logit contributions | |
forgot | -0.118 |
knew | -0.100 |
remembered | -0.100 |
wished | -0.100 |
knocked | -0.095 |
Token | Each |
Token ID | 5501 |
Feature activation | -2.28e+00 |
Pos logit contributions | |
he | +0.047 |
ing | +0.043 |
ings | +0.043 |
ore | +0.041 |
ays | +0.040 |
Neg logit contributions | |
forgot | -0.069 |
knew | -0.060 |
wished | -0.058 |
remembered | -0.057 |
learned | -0.056 |
Token | time |
Token ID | 640 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
forgot | +0.060 |
remembered | +0.048 |
wished | +0.046 |
knew | +0.046 |
knocked | +0.043 |
Neg logit contributions | |
he | -0.045 |
ore | -0.043 |
ing | -0.043 |
ay | -0.042 |
ings | -0.042 |
Token | , |
Token ID | 11 |
Feature activation | -7.36e-01 |
Pos logit contributions | |
he | +0.017 |
ing | +0.016 |
ings | +0.015 |
ay | +0.015 |
ore | +0.015 |
Neg logit contributions | |
forgot | -0.032 |
remembered | -0.026 |
knew | -0.026 |
wished | -0.026 |
bit | -0.024 |
Token | she |
Token ID | 673 |
Feature activation | -9.30e+00 |
Pos logit contributions | |
forgot | +0.022 |
remembered | +0.018 |
wished | +0.018 |
knew | +0.018 |
wasted | +0.017 |
Neg logit contributions | |
he | -0.011 |
ing | -0.010 |
ings | -0.010 |
room | -0.010 |
ay | -0.010 |
Token | was |
Token ID | 373 |
Feature activation | -3.26e+00 |
Pos logit contributions | |
forgot | +0.121 |
pap | +0.111 |
stolen | +0.111 |
robbers | +0.106 |
unknow | +0.105 |
Neg logit contributions | |
he | -0.229 |
room | -0.208 |
ing | -0.207 |
with | -0.204 |
at | -0.198 |
Token | happy |
Token ID | 3772 |
Feature activation | +7.13e-01 |
Pos logit contributions | |
forgot | +0.061 |
wished | +0.051 |
knew | +0.045 |
regretted | +0.044 |
cost | +0.044 |
Neg logit contributions | |
ing | -0.073 |
he | -0.072 |
with | -0.070 |
time | -0.069 |
apping | -0.067 |
Token | when |
Token ID | 618 |
Feature activation | -6.01e-01 |
Pos logit contributions | |
he | +0.010 |
ay | +0.009 |
ore | +0.009 |
ings | +0.009 |
ays | +0.009 |
Neg logit contributions | |
forgot | -0.023 |
knew | -0.019 |
remembered | -0.019 |
wished | -0.019 |
bit | -0.018 |
Token | she |
Token ID | 673 |
Feature activation | -7.26e+00 |
Pos logit contributions | |
forgot | +0.017 |
remembered | +0.014 |
wished | +0.014 |
knew | +0.014 |
bit | +0.013 |
Neg logit contributions | |
he | -0.010 |
ing | -0.009 |
ings | -0.009 |
ore | -0.009 |
ay | -0.009 |
Token | saw |
Token ID | 2497 |
Feature activation | -1.55e+00 |
Pos logit contributions | |
forgot | +0.100 |
pap | +0.078 |
wished | +0.074 |
known | +0.068 |
robbers | +0.068 |
Neg logit contributions | |
he | -0.199 |
ing | -0.183 |
room | -0.181 |
ore | -0.178 |
time | -0.175 |
Token | it |
Token ID | 340 |
Feature activation | +2.10e+00 |
Pos logit contributions | |
he | +0.004 |
ing | +0.004 |
ings | +0.004 |
ore | +0.004 |
ay | +0.004 |
Neg logit contributions | |
forgot | -0.007 |
remembered | -0.006 |
knew | -0.006 |
wished | -0.006 |
bit | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | +3.16e+00 |
Pos logit contributions | |
he | +0.032 |
ays | +0.031 |
ay | +0.031 |
ore | +0.030 |
room | +0.030 |
Neg logit contributions | |
forgot | -0.063 |
knew | -0.053 |
remembered | -0.052 |
bit | -0.051 |
wished | -0.051 |
Token | At |
Token ID | 1629 |
Feature activation | -9.90e-01 |
Pos logit contributions | |
he | +0.056 |
ings | +0.054 |
ing | +0.053 |
room | +0.052 |
ays | +0.052 |
Neg logit contributions | |
forgot | -0.086 |
knew | -0.075 |
wished | -0.072 |
remembered | -0.071 |
learned | -0.070 |
Token | the |
Token ID | 262 |
Feature activation | -3.11e+00 |
Pos logit contributions | |
forgot | +0.022 |
remembered | +0.017 |
knew | +0.017 |
wished | +0.017 |
knocked | +0.016 |
Neg logit contributions | |
he | -0.022 |
ay | -0.022 |
ings | -0.021 |
ing | -0.021 |
room | -0.021 |
Token | end |
Token ID | 886 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
forgot | +0.085 |
wished | +0.070 |
remembered | +0.068 |
struggled | +0.067 |
wasted | +0.066 |
Neg logit contributions | |
he | -0.056 |
time | -0.052 |
ay | -0.050 |
ummy | -0.050 |
ing | -0.049 |
Token | she |
Token ID | 673 |
Feature activation | -9.30e+00 |
Pos logit contributions | |
he | +0.038 |
ays | +0.036 |
room | +0.032 |
ikes | +0.032 |
eddy | +0.031 |
Neg logit contributions | |
forgot | -0.071 |
knew | -0.061 |
remembered | -0.059 |
bit | -0.056 |
learned | -0.056 |
Token | was |
Token ID | 373 |
Feature activation | -2.60e+00 |
Pos logit contributions | |
robbers | +0.185 |
forgot | +0.166 |
mate | +0.163 |
Indians | +0.163 |
sizes | +0.154 |
Neg logit contributions | |
he | -0.195 |
with | -0.172 |
uck | -0.166 |
ore | -0.162 |
time | -0.162 |
Token | happy |
Token ID | 3772 |
Feature activation | +1.00e+00 |
Pos logit contributions | |
forgot | +0.055 |
wished | +0.044 |
remembered | +0.042 |
robbers | +0.041 |
knew | +0.041 |
Neg logit contributions | |
he | -0.059 |
ing | -0.051 |
ings | -0.050 |
time | -0.050 |
over | -0.050 |
Token | and |
Token ID | 290 |
Feature activation | -4.60e+00 |
Pos logit contributions | |
he | +0.015 |
ay | +0.014 |
room | +0.013 |
ays | +0.013 |
ings | +0.013 |
Neg logit contributions | |
forgot | -0.031 |
knew | -0.026 |
remembered | -0.026 |
wished | -0.026 |
bit | -0.025 |
Token | proud |
Token ID | 6613 |
Feature activation | +1.73e+00 |
Pos logit contributions | |
forgot | +0.077 |
robbers | +0.065 |
remembered | +0.059 |
mate | +0.059 |
knocked | +0.058 |
Neg logit contributions | |
he | -0.108 |
ing | -0.104 |
with | -0.103 |
uck | -0.100 |
uggle | -0.099 |
Token | of |
Token ID | 286 |
Feature activation | +8.63e-01 |
Pos logit contributions | |
he | +0.025 |
room | +0.021 |
ings | +0.021 |
ays | +0.021 |
ay | +0.021 |
Neg logit contributions | |
forgot | -0.054 |
knew | -0.047 |
remembered | -0.046 |
wished | -0.046 |
bit | -0.044 |
Token | was |
Token ID | 373 |
Feature activation | -2.50e+00 |
Pos logit contributions | |
forgot | +0.130 |
setup | +0.118 |
dishonest | +0.117 |
error | +0.116 |
chie | +0.116 |
Neg logit contributions | |
with | -0.165 |
he | -0.158 |
. | -0.141 |
room | -0.141 |
, | -0.138 |
Token | happy |
Token ID | 3772 |
Feature activation | +9.57e-01 |
Pos logit contributions | |
forgot | +0.058 |
wished | +0.047 |
knew | +0.045 |
regretted | +0.044 |
remembered | +0.044 |
Neg logit contributions | |
he | -0.051 |
with | -0.047 |
ing | -0.044 |
time | -0.044 |
aps | -0.044 |
Token | and |
Token ID | 290 |
Feature activation | -4.71e+00 |
Pos logit contributions | |
he | +0.016 |
ay | +0.014 |
ings | +0.014 |
room | +0.014 |
ing | +0.014 |
Neg logit contributions | |
forgot | -0.029 |
knew | -0.024 |
remembered | -0.024 |
wished | -0.024 |
bit | -0.022 |
Token | tired |
Token ID | 10032 |
Feature activation | +3.44e+00 |
Pos logit contributions | |
forgot | +0.074 |
wasted | +0.062 |
nurses | +0.060 |
punished | +0.060 |
setup | +0.060 |
Neg logit contributions | |
with | -0.110 |
he | -0.105 |
ing | -0.100 |
uggle | -0.097 |
time | -0.097 |
Token | . |
Token ID | 13 |
Feature activation | +2.26e+00 |
Pos logit contributions | |
ays | +0.049 |
ay | +0.047 |
he | +0.045 |
ings | +0.045 |
eddy | +0.043 |
Neg logit contributions | |
forgot | -0.105 |
knew | -0.091 |
bit | -0.089 |
wished | -0.088 |
remembered | -0.085 |
Token | He |
Token ID | 679 |
Feature activation | -9.28e+00 |
Pos logit contributions | |
he | +0.038 |
ing | +0.035 |
ings | +0.035 |
ay | +0.034 |
ore | +0.033 |
Neg logit contributions | |
forgot | -0.066 |
knew | -0.055 |
remembered | -0.054 |
wished | -0.054 |
learned | -0.051 |
Token | curled |
Token ID | 41108 |
Feature activation | +2.45e+00 |
Pos logit contributions | |
forgot | +0.135 |
chie | +0.134 |
known | +0.127 |
fact | +0.126 |
error | +0.124 |
Neg logit contributions | |
with | -0.189 |
he | -0.182 |
. | -0.173 |
again | -0.173 |
and | -0.172 |
Token | up |
Token ID | 510 |
Feature activation | +1.56e+00 |
Pos logit contributions | |
ay | +0.036 |
he | +0.035 |
room | +0.032 |
ore | +0.031 |
ummy | +0.030 |
Neg logit contributions | |
forgot | -0.076 |
remembered | -0.067 |
wished | -0.065 |
knew | -0.064 |
bit | -0.063 |
Token | in |
Token ID | 287 |
Feature activation | +4.98e-01 |
Pos logit contributions | |
he | +0.023 |
ay | +0.023 |
room | +0.022 |
ays | +0.021 |
ings | +0.021 |
Neg logit contributions | |
forgot | -0.047 |
remembered | -0.040 |
knew | -0.040 |
wished | -0.039 |
bit | -0.039 |
Token | his |
Token ID | 465 |
Feature activation | -2.09e+00 |
Pos logit contributions | |
he | +0.009 |
ay | +0.009 |
ing | +0.009 |
ore | +0.009 |
ings | +0.009 |
Neg logit contributions | |
forgot | -0.013 |
remembered | -0.011 |
knew | -0.011 |
wished | -0.011 |
knocked | -0.010 |
Token | cozy |
Token ID | 37438 |
Feature activation | -1.16e+00 |
Pos logit contributions | |
forgot | +0.048 |
remembered | +0.038 |
wished | +0.037 |
knew | +0.036 |
known | +0.034 |
Neg logit contributions | |
he | -0.046 |
ay | -0.044 |
ings | -0.043 |
ore | -0.042 |
ing | -0.042 |
Token | she |
Token ID | 673 |
Feature activation | -6.73e+00 |
Pos logit contributions | |
forgot | +0.025 |
remembered | +0.021 |
wished | +0.021 |
knew | +0.021 |
bit | +0.021 |
Neg logit contributions | |
he | -0.012 |
ing | -0.011 |
time | -0.011 |
ings | -0.011 |
room | -0.010 |
Token | did |
Token ID | 750 |
Feature activation | -9.77e-02 |
Pos logit contributions | |
forgot | +0.105 |
pap | +0.076 |
wished | +0.071 |
forget | +0.071 |
remembered | +0.069 |
Neg logit contributions | |
he | -0.190 |
room | -0.169 |
ay | -0.164 |
ore | -0.162 |
time | -0.161 |
Token | good |
Token ID | 922 |
Feature activation | +1.42e+00 |
Pos logit contributions | |
forgot | +0.002 |
remembered | +0.002 |
knew | +0.002 |
wished | +0.002 |
knocked | +0.002 |
Neg logit contributions | |
he | -0.002 |
ay | -0.002 |
ing | -0.002 |
ore | -0.002 |
ings | -0.002 |
Token | things |
Token ID | 1243 |
Feature activation | +5.82e-01 |
Pos logit contributions | |
he | +0.036 |
room | +0.036 |
ore | +0.035 |
ays | +0.035 |
ay | +0.034 |
Neg logit contributions | |
forgot | -0.026 |
wished | -0.020 |
knew | -0.020 |
remembered | -0.020 |
learned | -0.019 |
Token | . |
Token ID | 13 |
Feature activation | +2.26e+00 |
Pos logit contributions | |
he | +0.006 |
ay | +0.006 |
ore | +0.006 |
room | +0.006 |
ays | +0.006 |
Neg logit contributions | |
forgot | -0.020 |
knew | -0.017 |
remembered | -0.017 |
wished | -0.017 |
knocked | -0.016 |
Token | She |
Token ID | 1375 |
Feature activation | -9.28e+00 |
Pos logit contributions | |
he | +0.035 |
ing | +0.033 |
ore | +0.032 |
ings | +0.032 |
ay | +0.030 |
Neg logit contributions | |
forgot | -0.068 |
knew | -0.057 |
wished | -0.057 |
remembered | -0.055 |
learned | -0.054 |
Token | decided |
Token ID | 3066 |
Feature activation | -3.80e-01 |
Pos logit contributions | |
chie | +0.144 |
Indians | +0.131 |
pap | +0.130 |
robbers | +0.127 |
mate | +0.123 |
Neg logit contributions | |
he | -0.202 |
with | -0.198 |
at | -0.177 |
room | -0.172 |
and | -0.170 |
Token | to |
Token ID | 284 |
Feature activation | -4.57e+00 |
Pos logit contributions | |
forgot | +0.012 |
remembered | +0.010 |
knew | +0.010 |
wished | +0.010 |
bit | +0.009 |
Neg logit contributions | |
he | -0.005 |
ing | -0.005 |
ings | -0.005 |
ore | -0.005 |
ay | -0.004 |
Token | sell |
Token ID | 3677 |
Feature activation | +2.80e+00 |
Pos logit contributions | |
forgot | +0.102 |
remembered | +0.082 |
wished | +0.079 |
knew | +0.077 |
struggled | +0.073 |
Neg logit contributions | |
he | -0.110 |
room | -0.101 |
ay | -0.098 |
uggle | -0.097 |
ings | -0.097 |
Token | some |
Token ID | 617 |
Feature activation | -5.46e-01 |
Pos logit contributions | |
he | +0.053 |
ore | +0.050 |
ing | +0.049 |
ay | +0.049 |
ings | +0.048 |
Neg logit contributions | |
forgot | -0.073 |
remembered | -0.059 |
knew | -0.059 |
wished | -0.059 |
bit | -0.055 |
Token | of |
Token ID | 286 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
forgot | +0.012 |
knew | +0.009 |
wished | +0.009 |
remembered | +0.009 |
bit | +0.008 |
Neg logit contributions | |
he | -0.013 |
ore | -0.012 |
ing | -0.012 |
ays | -0.012 |
room | -0.012 |
Token | cream |
Token ID | 8566 |
Feature activation | +3.52e+00 |
Pos logit contributions | |
forgot | +0.014 |
knew | +0.011 |
remembered | +0.011 |
wished | +0.011 |
bit | +0.010 |
Neg logit contributions | |
he | -0.011 |
ing | -0.011 |
ay | -0.011 |
ore | -0.010 |
ays | -0.010 |
Token | . |
Token ID | 13 |
Feature activation | +3.05e+00 |
Pos logit contributions | |
he | +0.072 |
ay | +0.071 |
ing | +0.070 |
ore | +0.070 |
ays | +0.068 |
Neg logit contributions | |
forgot | -0.080 |
knew | -0.064 |
remembered | -0.063 |
bit | -0.063 |
wished | -0.062 |
Token | Every |
Token ID | 3887 |
Feature activation | -1.24e+00 |
Pos logit contributions | |
ing | +0.065 |
ore | +0.062 |
he | +0.062 |
ings | +0.059 |
room | +0.058 |
Neg logit contributions | |
forgot | -0.071 |
knew | -0.058 |
learned | -0.056 |
remembered | -0.055 |
wished | -0.053 |
Token | night |
Token ID | 1755 |
Feature activation | +2.76e+00 |
Pos logit contributions | |
forgot | +0.032 |
remembered | +0.026 |
wished | +0.025 |
knew | +0.025 |
knocked | +0.023 |
Neg logit contributions | |
he | -0.025 |
ore | -0.023 |
ing | -0.023 |
ay | -0.023 |
ings | -0.023 |
Token | , |
Token ID | 11 |
Feature activation | +4.43e-01 |
Pos logit contributions | |
he | +0.037 |
ay | +0.032 |
ays | +0.032 |
room | +0.032 |
ore | +0.031 |
Neg logit contributions | |
forgot | -0.092 |
remembered | -0.077 |
knew | -0.077 |
wished | -0.075 |
knocked | -0.072 |
Token | she |
Token ID | 673 |
Feature activation | -9.25e+00 |
Pos logit contributions | |
he | +0.009 |
ore | +0.009 |
ay | +0.009 |
ings | +0.008 |
room | +0.008 |
Neg logit contributions | |
forgot | -0.012 |
remembered | -0.009 |
knew | -0.009 |
wished | -0.009 |
learned | -0.009 |
Token | would |
Token ID | 561 |
Feature activation | -3.96e+00 |
Pos logit contributions | |
forgot | +0.157 |
robbers | +0.134 |
stolen | +0.131 |
chie | +0.130 |
unknow | +0.130 |
Neg logit contributions | |
he | -0.201 |
time | -0.178 |
ore | -0.177 |
with | -0.176 |
ing | -0.172 |
Token | stand |
Token ID | 1302 |
Feature activation | +2.81e+00 |
Pos logit contributions | |
forgot | +0.088 |
remembered | +0.071 |
wished | +0.068 |
knew | +0.065 |
regretted | +0.064 |
Neg logit contributions | |
he | -0.091 |
uggle | -0.078 |
room | -0.078 |
ore | -0.077 |
time | -0.076 |
Token | in |
Token ID | 287 |
Feature activation | +2.92e-01 |
Pos logit contributions | |
ay | +0.038 |
he | +0.037 |
room | +0.036 |
ings | +0.034 |
ore | +0.034 |
Neg logit contributions | |
forgot | -0.088 |
remembered | -0.078 |
knew | -0.075 |
wished | -0.073 |
knocked | -0.073 |
Token | front |
Token ID | 2166 |
Feature activation | +6.37e-01 |
Pos logit contributions | |
he | +0.005 |
ay | +0.005 |
ings | +0.005 |
ing | +0.005 |
ore | +0.005 |
Neg logit contributions | |
forgot | -0.008 |
remembered | -0.007 |
knew | -0.007 |
wished | -0.007 |
knocked | -0.006 |
Token | of |
Token ID | 286 |
Feature activation | -6.46e-01 |
Pos logit contributions | |
he | +0.008 |
ings | +0.008 |
ing | +0.007 |
ore | +0.007 |
room | +0.007 |
Neg logit contributions | |
forgot | -0.019 |
knew | -0.017 |
wished | -0.017 |
remembered | -0.017 |
bit | -0.016 |
Token | played |
Token ID | 2826 |
Feature activation | +4.24e+00 |
Pos logit contributions | |
forgot | +0.089 |
mistake | +0.082 |
importance | +0.081 |
regret | +0.080 |
enemy | +0.080 |
Neg logit contributions | |
with | -0.108 |
together | -0.103 |
down | -0.102 |
on | -0.101 |
at | -0.098 |
Token | and |
Token ID | 290 |
Feature activation | -4.39e+00 |
Pos logit contributions | |
ays | +0.075 |
eddy | +0.075 |
ay | +0.070 |
ummy | +0.069 |
ane | +0.068 |
Neg logit contributions | |
forgot | -0.116 |
learned | -0.100 |
knew | -0.099 |
remembered | -0.096 |
wished | -0.096 |
Token | had |
Token ID | 550 |
Feature activation | -1.74e-01 |
Pos logit contributions | |
forgot | +0.062 |
punished | +0.058 |
pap | +0.055 |
wasted | +0.051 |
regretted | +0.051 |
Neg logit contributions | |
with | -0.110 |
he | -0.104 |
ing | -0.100 |
on | -0.099 |
room | -0.096 |
Token | fun |
Token ID | 1257 |
Feature activation | +3.62e+00 |
Pos logit contributions | |
forgot | +0.004 |
remembered | +0.003 |
knew | +0.003 |
wished | +0.003 |
knocked | +0.003 |
Neg logit contributions | |
he | -0.004 |
room | -0.004 |
ay | -0.004 |
ing | -0.004 |
time | -0.004 |
Token | . |
Token ID | 13 |
Feature activation | +2.57e+00 |
Pos logit contributions | |
ay | +0.062 |
eddy | +0.061 |
ays | +0.059 |
ummy | +0.056 |
uggle | +0.056 |
Neg logit contributions | |
forgot | -0.094 |
wished | -0.082 |
bit | -0.080 |
knew | -0.079 |
knocked | -0.078 |
Token | She |
Token ID | 1375 |
Feature activation | -9.24e+00 |
Pos logit contributions | |
he | +0.045 |
ing | +0.043 |
ay | +0.043 |
room | +0.042 |
eddy | +0.042 |
Neg logit contributions | |
forgot | -0.072 |
knew | -0.060 |
wished | -0.059 |
remembered | -0.057 |
missed | -0.057 |
Token | saw |
Token ID | 2497 |
Feature activation | -1.69e+00 |
Pos logit contributions | |
chie | +0.130 |
robbers | +0.121 |
pap | +0.111 |
supervision | +0.107 |
setup | +0.105 |
Neg logit contributions | |
with | -0.209 |
he | -0.205 |
at | -0.184 |
again | -0.184 |
and | -0.182 |
Token | the |
Token ID | 262 |
Feature activation | -2.52e+00 |
Pos logit contributions | |
forgot | +0.047 |
remembered | +0.037 |
knew | +0.035 |
struggled | +0.035 |
bit | +0.035 |
Neg logit contributions | |
ing | -0.029 |
he | -0.029 |
room | -0.027 |
ings | -0.026 |
ay | -0.026 |
Token | same |
Token ID | 976 |
Feature activation | -1.57e+00 |
Pos logit contributions | |
forgot | +0.071 |
remembered | +0.059 |
struggled | +0.056 |
punished | +0.056 |
wished | +0.055 |
Neg logit contributions | |
he | -0.043 |
ay | -0.039 |
ing | -0.037 |
ings | -0.036 |
time | -0.036 |
Token | bird |
Token ID | 6512 |
Feature activation | +1.69e-01 |
Pos logit contributions | |
forgot | +0.037 |
remembered | +0.029 |
wished | +0.029 |
knew | +0.028 |
struggled | +0.027 |
Neg logit contributions | |
he | -0.034 |
ings | -0.033 |
ing | -0.033 |
ay | -0.032 |
room | -0.032 |
Token | from |
Token ID | 422 |
Feature activation | -9.84e-01 |
Pos logit contributions | |
he | +0.002 |
ing | +0.002 |
ay | +0.002 |
ore | +0.002 |
ings | +0.002 |
Neg logit contributions | |
forgot | -0.005 |
remembered | -0.004 |
knew | -0.004 |
wished | -0.004 |
bit | -0.004 |
Token | my |
Token ID | 1820 |
Feature activation | -7.70e+00 |
Pos logit contributions | |
forgot | +0.113 |
wished | +0.096 |
knocked | +0.090 |
known | +0.090 |
forgotten | +0.089 |
Neg logit contributions | |
he | -0.082 |
ing | -0.078 |
ay | -0.077 |
eddy | -0.076 |
room | -0.072 |
Token | was |
Token ID | 373 |
Feature activation | -2.88e+00 |
Pos logit contributions | |
forgot | +0.174 |
known | +0.156 |
robbers | +0.155 |
stolen | +0.148 |
forgotten | +0.147 |
Neg logit contributions | |
he | -0.134 |
ing | -0.129 |
with | -0.126 |
uck | -0.120 |
eddy | -0.118 |
Token | so |
Token ID | 523 |
Feature activation | -2.45e+00 |
Pos logit contributions | |
forgot | +0.062 |
wished | +0.048 |
remembered | +0.046 |
knew | +0.046 |
bit | +0.046 |
Neg logit contributions | |
he | -0.067 |
ing | -0.057 |
room | -0.057 |
ay | -0.056 |
time | -0.056 |
Token | happy |
Token ID | 3772 |
Feature activation | +2.62e+00 |
Pos logit contributions | |
forgot | +0.046 |
remembered | +0.034 |
wished | +0.033 |
knew | +0.032 |
bit | +0.031 |
Neg logit contributions | |
he | -0.066 |
ing | -0.061 |
ay | -0.061 |
ore | -0.060 |
ings | -0.060 |
Token | . |
Token ID | 13 |
Feature activation | +2.46e+00 |
Pos logit contributions | |
he | +0.046 |
ay | +0.043 |
room | +0.040 |
ays | +0.040 |
ings | +0.040 |
Neg logit contributions | |
forgot | -0.076 |
knew | -0.064 |
wished | -0.062 |
remembered | -0.062 |
knocked | -0.059 |
Token | He |
Token ID | 679 |
Feature activation | -9.21e+00 |
Pos logit contributions | |
he | +0.045 |
ings | +0.041 |
ing | +0.040 |
room | +0.040 |
ore | +0.040 |
Neg logit contributions | |
forgot | -0.068 |
knew | -0.058 |
remembered | -0.056 |
wished | -0.055 |
knocked | -0.052 |
Token | held |
Token ID | 2714 |
Feature activation | +9.44e-01 |
Pos logit contributions | |
forgot | +0.138 |
known | +0.127 |
robbers | +0.122 |
forgotten | +0.120 |
mate | +0.119 |
Neg logit contributions | |
he | -0.209 |
with | -0.201 |
ay | -0.199 |
eddy | -0.193 |
and | -0.180 |
Token | the |
Token ID | 262 |
Feature activation | -3.10e+00 |
Pos logit contributions | |
he | +0.020 |
ings | +0.019 |
room | +0.019 |
ing | +0.018 |
ore | +0.018 |
Neg logit contributions | |
forgot | -0.022 |
remembered | -0.019 |
knew | -0.019 |
wished | -0.018 |
learned | -0.017 |
Token | arrow |
Token ID | 15452 |
Feature activation | +2.74e+00 |
Pos logit contributions | |
forgot | +0.071 |
remembered | +0.053 |
knocked | +0.052 |
wished | +0.052 |
punished | +0.051 |
Neg logit contributions | |
ay | -0.069 |
he | -0.067 |
ing | -0.064 |
ings | -0.061 |
eddy | -0.059 |
Token | up |
Token ID | 510 |
Feature activation | +1.55e+00 |
Pos logit contributions | |
he | +0.039 |
ore | +0.035 |
ing | +0.035 |
ay | +0.035 |
ings | +0.035 |
Neg logit contributions | |
forgot | -0.086 |
remembered | -0.072 |
knew | -0.071 |
wished | -0.071 |
bit | -0.068 |
Token | and |
Token ID | 290 |
Feature activation | -6.30e+00 |
Pos logit contributions | |
he | +0.023 |
ore | +0.021 |
ing | +0.021 |
ays | +0.021 |
ay | +0.021 |
Neg logit contributions | |
forgot | -0.046 |
remembered | -0.039 |
knew | -0.039 |
wished | -0.038 |
bit | -0.038 |
Token | the |
Token ID | 262 |
Feature activation | -3.22e+00 |
Pos logit contributions | |
forgot | +0.049 |
remembered | +0.040 |
knew | +0.039 |
wished | +0.039 |
struggled | +0.037 |
Neg logit contributions | |
he | -0.030 |
ings | -0.027 |
room | -0.026 |
ore | -0.026 |
ay | -0.025 |
Token | wedding |
Token ID | 10614 |
Feature activation | +1.75e+00 |
Pos logit contributions | |
forgot | +0.076 |
remembered | +0.058 |
struggled | +0.058 |
knew | +0.056 |
wished | +0.053 |
Neg logit contributions | |
he | -0.070 |
time | -0.068 |
ing | -0.063 |
ore | -0.062 |
ings | -0.062 |
Token | . |
Token ID | 13 |
Feature activation | +3.16e+00 |
Pos logit contributions | |
he | +0.028 |
ing | +0.027 |
ore | +0.027 |
ay | +0.027 |
ays | +0.027 |
Neg logit contributions | |
forgot | -0.050 |
remembered | -0.041 |
knew | -0.041 |
wished | -0.041 |
bit | -0.039 |
Token | Instead |
Token ID | 5455 |
Feature activation | +4.90e-01 |
Pos logit contributions | |
ing | +0.055 |
he | +0.054 |
ings | +0.052 |
ays | +0.051 |
ach | +0.051 |
Neg logit contributions | |
forgot | -0.082 |
knew | -0.072 |
learned | -0.069 |
wished | -0.067 |
remembered | -0.066 |
Token | , |
Token ID | 11 |
Feature activation | -1.43e+00 |
Pos logit contributions | |
he | +0.005 |
ing | +0.004 |
ay | +0.004 |
ings | +0.004 |
ore | +0.004 |
Neg logit contributions | |
forgot | -0.017 |
remembered | -0.015 |
knew | -0.015 |
wished | -0.015 |
bit | -0.014 |
Token | she |
Token ID | 673 |
Feature activation | -9.21e+00 |
Pos logit contributions | |
forgot | +0.042 |
remembered | +0.035 |
knew | +0.034 |
wished | +0.034 |
struggled | +0.032 |
Neg logit contributions | |
he | -0.023 |
ing | -0.021 |
ore | -0.020 |
ay | -0.020 |
ings | -0.020 |
Token | went |
Token ID | 1816 |
Feature activation | +1.61e+00 |
Pos logit contributions | |
forgot | +0.117 |
robbers | +0.111 |
pap | +0.107 |
persever | +0.105 |
unknow | +0.103 |
Neg logit contributions | |
he | -0.213 |
with | -0.204 |
at | -0.192 |
. | -0.189 |
room | -0.187 |
Token | to |
Token ID | 284 |
Feature activation | -1.05e+00 |
Pos logit contributions | |
he | +0.025 |
ay | +0.023 |
ing | +0.023 |
ings | +0.022 |
ore | +0.022 |
Neg logit contributions | |
forgot | -0.048 |
remembered | -0.040 |
knew | -0.040 |
wished | -0.040 |
knocked | -0.038 |
Token | the |
Token ID | 262 |
Feature activation | -3.29e+00 |
Pos logit contributions | |
forgot | +0.028 |
remembered | +0.023 |
knew | +0.023 |
wished | +0.022 |
struggled | +0.021 |
Neg logit contributions | |
he | -0.020 |
ore | -0.018 |
ings | -0.018 |
room | -0.018 |
ay | -0.018 |
Token | park |
Token ID | 3952 |
Feature activation | +3.28e+00 |
Pos logit contributions | |
forgot | +0.078 |
struggled | +0.060 |
remembered | +0.059 |
knew | +0.057 |
wished | +0.055 |
Neg logit contributions | |
he | -0.072 |
time | -0.069 |
ing | -0.065 |
ore | -0.065 |
eddy | -0.064 |
Token | and |
Token ID | 290 |
Feature activation | -4.86e+00 |
Pos logit contributions | |
he | +0.049 |
ay | +0.046 |
ing | +0.045 |
ings | +0.045 |
ore | +0.044 |
Neg logit contributions | |
forgot | -0.100 |
knew | -0.083 |
remembered | -0.083 |
wished | -0.082 |
bit | -0.079 |
Token | dream |
Token ID | 4320 |
Feature activation | +5.05e+00 |
Pos logit contributions | |
forgot | +0.037 |
remembered | +0.028 |
knew | +0.027 |
wished | +0.027 |
bit | +0.025 |
Neg logit contributions | |
he | -0.043 |
ay | -0.040 |
ing | -0.040 |
ore | -0.040 |
ings | -0.040 |
Token | . |
Token ID | 13 |
Feature activation | +2.63e+00 |
Pos logit contributions | |
ays | +0.078 |
eddy | +0.068 |
iffs | +0.066 |
he | +0.064 |
uggle | +0.064 |
Neg logit contributions | |
forgot | -0.136 |
knew | -0.120 |
always | -0.113 |
cost | -0.112 |
needed | -0.109 |
Token | Every |
Token ID | 3887 |
Feature activation | -1.18e+00 |
Pos logit contributions | |
ing | +0.058 |
ings | +0.053 |
he | +0.053 |
ore | +0.052 |
ays | +0.051 |
Neg logit contributions | |
forgot | -0.058 |
knew | -0.049 |
wished | -0.048 |
wanted | -0.045 |
learned | -0.045 |
Token | night |
Token ID | 1755 |
Feature activation | +3.31e+00 |
Pos logit contributions | |
forgot | +0.028 |
remembered | +0.022 |
knew | +0.022 |
wished | +0.022 |
knocked | +0.020 |
Neg logit contributions | |
he | -0.026 |
ing | -0.024 |
ore | -0.024 |
ay | -0.024 |
ings | -0.024 |
Token | , |
Token ID | 11 |
Feature activation | +7.75e-01 |
Pos logit contributions | |
he | +0.060 |
room | +0.056 |
ays | +0.055 |
ay | +0.053 |
ore | +0.051 |
Neg logit contributions | |
forgot | -0.094 |
knew | -0.077 |
remembered | -0.075 |
knocked | -0.073 |
wished | -0.072 |
Token | she |
Token ID | 673 |
Feature activation | -9.20e+00 |
Pos logit contributions | |
he | +0.017 |
ay | +0.017 |
ore | +0.017 |
room | +0.017 |
ings | +0.016 |
Neg logit contributions | |
forgot | -0.018 |
knew | -0.014 |
wished | -0.014 |
remembered | -0.014 |
knocked | -0.013 |
Token | would |
Token ID | 561 |
Feature activation | -3.75e+00 |
Pos logit contributions | |
forgot | +0.150 |
stolen | +0.123 |
punished | +0.115 |
unknow | +0.115 |
forgotten | +0.112 |
Neg logit contributions | |
he | -0.209 |
ore | -0.200 |
with | -0.194 |
ing | -0.191 |
eddy | -0.185 |
Token | close |
Token ID | 1969 |
Feature activation | +1.98e+00 |
Pos logit contributions | |
forgot | +0.082 |
remembered | +0.061 |
wished | +0.060 |
knew | +0.058 |
struggled | +0.058 |
Neg logit contributions | |
he | -0.088 |
ore | -0.079 |
uggle | -0.078 |
room | -0.078 |
ay | -0.077 |
Token | her |
Token ID | 607 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
he | +0.046 |
room | +0.043 |
ore | +0.043 |
ay | +0.042 |
over | +0.041 |
Neg logit contributions | |
forgot | -0.041 |
knew | -0.034 |
remembered | -0.034 |
wished | -0.033 |
bit | -0.030 |
Token | eyes |
Token ID | 2951 |
Feature activation | +3.46e+00 |
Pos logit contributions | |
forgot | +0.028 |
remembered | +0.022 |
knew | +0.022 |
wished | +0.021 |
bit | +0.020 |
Neg logit contributions | |
he | -0.029 |
ing | -0.027 |
ay | -0.027 |
ore | -0.027 |
ings | -0.027 |
Token | and |
Token ID | 290 |
Feature activation | -7.65e+00 |
Pos logit contributions | |
ays | +0.063 |
ore | +0.060 |
ings | +0.060 |
room | +0.058 |
aps | +0.058 |
Neg logit contributions | |
forgot | -0.085 |
bit | -0.075 |
knew | -0.074 |
remembered | -0.070 |
wished | -0.069 |
Token |
|
Token ID | 198 |
Feature activation | +2.87e+00 |
Pos logit contributions | |
ays | +0.054 |
ing | +0.054 |
he | +0.050 |
ings | +0.050 |
ay | +0.050 |
Neg logit contributions | |
forgot | -0.095 |
knew | -0.085 |
wished | -0.082 |
remembered | -0.078 |
learned | -0.078 |
Token | L |
Token ID | 43 |
Feature activation | +1.81e-02 |
Pos logit contributions | |
he | +0.049 |
ing | +0.045 |
ings | +0.045 |
ay | +0.045 |
ore | +0.045 |
Neg logit contributions | |
forgot | -0.082 |
remembered | -0.068 |
knew | -0.068 |
wished | -0.067 |
learned | -0.063 |
Token | ily |
Token ID | 813 |
Feature activation | -5.83e+00 |
Pos logit contributions | |
he | +0.000 |
ings | +0.000 |
ay | +0.000 |
ing | +0.000 |
ore | +0.000 |
Neg logit contributions | |
forgot | -0.000 |
knew | -0.000 |
remembered | -0.000 |
wished | -0.000 |
learned | -0.000 |
Token | and |
Token ID | 290 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
forgot | +0.134 |
known | +0.115 |
stolen | +0.110 |
forgotten | +0.110 |
cost | +0.106 |
Neg logit contributions | |
ing | -0.101 |
he | -0.098 |
ays | -0.096 |
uck | -0.096 |
ore | -0.093 |
Token | Ben |
Token ID | 3932 |
Feature activation | -7.27e+00 |
Pos logit contributions | |
forgot | +0.033 |
remembered | +0.027 |
knew | +0.026 |
wished | +0.026 |
knocked | +0.024 |
Neg logit contributions | |
he | -0.029 |
ay | -0.028 |
ing | -0.027 |
ore | -0.027 |
ings | -0.027 |
Token | also |
Token ID | 635 |
Feature activation | -9.20e+00 |
Pos logit contributions | |
forgot | +0.147 |
pap | +0.118 |
stolen | +0.118 |
known | +0.118 |
shown | +0.117 |
Neg logit contributions | |
he | -0.162 |
room | -0.125 |
uggle | -0.124 |
with | -0.124 |
ore | -0.123 |
Token | wake |
Token ID | 7765 |
Feature activation | +2.37e+00 |
Pos logit contributions | |
Indians | +0.146 |
forgot | +0.141 |
robbers | +0.131 |
crew | +0.129 |
pap | +0.125 |
Neg logit contributions | |
he | -0.185 |
. | -0.171 |
at | -0.168 |
with | -0.167 |
.' | -0.157 |
Token | up |
Token ID | 510 |
Feature activation | +2.35e+00 |
Pos logit contributions | |
he | +0.040 |
ay | +0.040 |
unch | +0.038 |
ane | +0.037 |
room | +0.035 |
Neg logit contributions | |
forgot | -0.064 |
knew | -0.054 |
remembered | -0.054 |
missed | -0.052 |
wished | -0.051 |
Token | and |
Token ID | 290 |
Feature activation | -6.14e+00 |
Pos logit contributions | |
he | +0.041 |
ays | +0.040 |
room | +0.040 |
ay | +0.040 |
ings | +0.037 |
Neg logit contributions | |
forgot | -0.063 |
knew | -0.055 |
remembered | -0.053 |
bit | -0.051 |
learned | -0.051 |
Token | look |
Token ID | 804 |
Feature activation | -1.49e-01 |
Pos logit contributions | |
forgot | +0.122 |
knew | +0.091 |
remembered | +0.090 |
known | +0.089 |
shown | +0.087 |
Neg logit contributions | |
he | -0.144 |
ore | -0.135 |
uggle | -0.132 |
ing | -0.131 |
ays | -0.128 |
Token | out |
Token ID | 503 |
Feature activation | +9.53e-01 |
Pos logit contributions | |
forgot | +0.005 |
remembered | +0.004 |
knew | +0.004 |
wished | +0.004 |
bit | +0.004 |
Neg logit contributions | |
he | -0.002 |
room | -0.002 |
ay | -0.002 |
ings | -0.002 |
ore | -0.002 |
Token | got |
Token ID | 1392 |
Feature activation | +4.81e-01 |
Pos logit contributions | |
forgot | +0.116 |
pap | +0.084 |
known | +0.084 |
wished | +0.083 |
remembered | +0.082 |
Neg logit contributions | |
he | -0.189 |
ore | -0.168 |
ay | -0.162 |
time | -0.160 |
room | -0.160 |
Token | to |
Token ID | 284 |
Feature activation | -1.53e+00 |
Pos logit contributions | |
he | +0.009 |
ay | +0.008 |
ings | +0.008 |
room | +0.008 |
ing | +0.008 |
Neg logit contributions | |
forgot | -0.013 |
remembered | -0.011 |
knew | -0.011 |
wished | -0.011 |
knocked | -0.010 |
Token | the |
Token ID | 262 |
Feature activation | -2.21e+00 |
Pos logit contributions | |
forgot | +0.042 |
remembered | +0.035 |
knew | +0.034 |
wished | +0.034 |
bit | +0.032 |
Neg logit contributions | |
he | -0.028 |
room | -0.025 |
ay | -0.025 |
ings | -0.025 |
ore | -0.024 |
Token | school |
Token ID | 1524 |
Feature activation | +2.28e+00 |
Pos logit contributions | |
forgot | +0.053 |
remembered | +0.042 |
wished | +0.041 |
knew | +0.041 |
struggled | +0.039 |
Neg logit contributions | |
he | -0.049 |
ay | -0.044 |
ing | -0.044 |
ore | -0.044 |
room | -0.044 |
Token | , |
Token ID | 11 |
Feature activation | -3.84e-01 |
Pos logit contributions | |
he | +0.029 |
ings | +0.027 |
ore | +0.027 |
ays | +0.027 |
eddy | +0.027 |
Neg logit contributions | |
forgot | -0.074 |
knew | -0.062 |
remembered | -0.062 |
wished | -0.061 |
knocked | -0.058 |
Token | she |
Token ID | 673 |
Feature activation | -9.20e+00 |
Pos logit contributions | |
forgot | +0.010 |
knew | +0.008 |
remembered | +0.008 |
wished | +0.008 |
knocked | +0.007 |
Neg logit contributions | |
he | -0.008 |
ay | -0.007 |
ore | -0.007 |
ing | -0.007 |
ings | -0.007 |
Token | saw |
Token ID | 2497 |
Feature activation | -1.15e+00 |
Pos logit contributions | |
sizes | +0.135 |
forgot | +0.132 |
vitamins | +0.129 |
fiber | +0.126 |
pap | +0.125 |
Neg logit contributions | |
he | -0.194 |
with | -0.186 |
at | -0.180 |
room | -0.169 |
ing | -0.167 |
Token | many |
Token ID | 867 |
Feature activation | -1.86e+00 |
Pos logit contributions | |
forgot | +0.033 |
remembered | +0.027 |
knew | +0.027 |
wished | +0.027 |
bit | +0.025 |
Neg logit contributions | |
he | -0.020 |
ing | -0.018 |
ay | -0.018 |
ore | -0.018 |
ings | -0.018 |
Token | kids |
Token ID | 3988 |
Feature activation | -4.65e-01 |
Pos logit contributions | |
forgot | +0.043 |
remembered | +0.034 |
wished | +0.033 |
knew | +0.033 |
knocked | +0.031 |
Neg logit contributions | |
he | -0.042 |
ing | -0.039 |
ay | -0.039 |
room | -0.038 |
ings | -0.037 |
Token | . |
Token ID | 13 |
Feature activation | +2.30e+00 |
Pos logit contributions | |
forgot | +0.012 |
remembered | +0.010 |
knew | +0.010 |
wished | +0.010 |
knocked | +0.009 |
Neg logit contributions | |
he | -0.009 |
ing | -0.008 |
ay | -0.008 |
ore | -0.008 |
ings | -0.008 |
Token | They |
Token ID | 1119 |
Feature activation | -6.92e+00 |
Pos logit contributions | |
he | +0.038 |
ing | +0.034 |
ings | +0.033 |
ore | +0.033 |
ay | +0.032 |
Neg logit contributions | |
forgot | -0.067 |
knew | -0.057 |
wished | -0.056 |
remembered | -0.055 |
knocked | -0.053 |
Token | my |
Token ID | 1820 |
Feature activation | -7.83e+00 |
Pos logit contributions | |
forgot | +0.118 |
wished | +0.100 |
forgotten | +0.093 |
stolen | +0.093 |
known | +0.093 |
Neg logit contributions | |
he | -0.079 |
ing | -0.075 |
room | -0.072 |
eddy | -0.071 |
ay | -0.071 |
Token | was |
Token ID | 373 |
Feature activation | -3.00e+00 |
Pos logit contributions | |
forgot | +0.151 |
robbers | +0.150 |
dishonest | +0.136 |
stolen | +0.133 |
known | +0.132 |
Neg logit contributions | |
he | -0.159 |
ing | -0.150 |
with | -0.145 |
room | -0.145 |
ore | -0.135 |
Token | so |
Token ID | 523 |
Feature activation | -2.21e+00 |
Pos logit contributions | |
forgot | +0.073 |
remembered | +0.057 |
knew | +0.057 |
'd | +0.056 |
wished | +0.056 |
Neg logit contributions | |
he | -0.058 |
with | -0.053 |
ing | -0.052 |
room | -0.051 |
time | -0.050 |
Token | happy |
Token ID | 3772 |
Feature activation | +2.44e+00 |
Pos logit contributions | |
forgot | +0.046 |
remembered | +0.036 |
wished | +0.034 |
knew | +0.033 |
bit | +0.032 |
Neg logit contributions | |
he | -0.056 |
ore | -0.050 |
ing | -0.050 |
ay | -0.050 |
room | -0.048 |
Token | . |
Token ID | 13 |
Feature activation | +2.89e+00 |
Pos logit contributions | |
he | +0.044 |
ay | +0.042 |
ays | +0.039 |
room | +0.039 |
ings | +0.038 |
Neg logit contributions | |
forgot | -0.069 |
knew | -0.059 |
remembered | -0.057 |
wished | -0.057 |
knocked | -0.054 |
Token | He |
Token ID | 679 |
Feature activation | -9.20e+00 |
Pos logit contributions | |
he | +0.047 |
ing | +0.042 |
ings | +0.042 |
ore | +0.042 |
ay | +0.042 |
Neg logit contributions | |
forgot | -0.085 |
knew | -0.071 |
remembered | -0.070 |
wished | -0.070 |
knocked | -0.066 |
Token | watched |
Token ID | 7342 |
Feature activation | +2.36e+00 |
Pos logit contributions | |
robbers | +0.144 |
supervision | +0.140 |
dishonest | +0.134 |
setup | +0.130 |
known | +0.129 |
Neg logit contributions | |
with | -0.186 |
he | -0.182 |
and | -0.174 |
at | -0.172 |
, | -0.164 |
Token | as |
Token ID | 355 |
Feature activation | +5.83e-01 |
Pos logit contributions | |
he | +0.050 |
ays | +0.048 |
ings | +0.048 |
ay | +0.047 |
ore | +0.047 |
Neg logit contributions | |
forgot | -0.053 |
knew | -0.045 |
remembered | -0.043 |
wished | -0.042 |
knocked | -0.042 |
Token | the |
Token ID | 262 |
Feature activation | -2.43e+00 |
Pos logit contributions | |
he | +0.012 |
ings | +0.012 |
ore | +0.012 |
ay | +0.012 |
ing | +0.011 |
Neg logit contributions | |
forgot | -0.014 |
knew | -0.012 |
remembered | -0.011 |
wished | -0.011 |
knocked | -0.010 |
Token | bright |
Token ID | 6016 |
Feature activation | -1.68e+00 |
Pos logit contributions | |
forgot | +0.074 |
remembered | +0.061 |
wished | +0.058 |
knew | +0.057 |
regretted | +0.057 |
Neg logit contributions | |
he | -0.039 |
ay | -0.037 |
time | -0.036 |
ing | -0.035 |
ore | -0.033 |
Token | fireworks |
Token ID | 26056 |
Feature activation | -2.75e+00 |
Pos logit contributions | |
forgot | +0.041 |
remembered | +0.032 |
wished | +0.031 |
knew | +0.031 |
forgotten | +0.029 |
Neg logit contributions | |
he | -0.036 |
ing | -0.035 |
ay | -0.034 |
ore | -0.033 |
ings | -0.033 |
Token | became |
Token ID | 2627 |
Feature activation | -2.94e+00 |
Pos logit contributions | |
forgot | +0.113 |
known | +0.090 |
wished | +0.087 |
forget | +0.086 |
stolen | +0.085 |
Neg logit contributions | |
he | -0.137 |
ing | -0.133 |
with | -0.125 |
ore | -0.124 |
room | -0.120 |
Token | healthier |
Token ID | 22841 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
forgot | +0.068 |
wished | +0.052 |
knew | +0.052 |
'd | +0.051 |
remembered | +0.049 |
Neg logit contributions | |
he | -0.059 |
ing | -0.058 |
ings | -0.058 |
time | -0.056 |
over | -0.054 |
Token | and |
Token ID | 290 |
Feature activation | -5.48e+00 |
Pos logit contributions | |
he | +0.032 |
ay | +0.029 |
ore | +0.028 |
ing | +0.028 |
ays | +0.028 |
Neg logit contributions | |
forgot | -0.078 |
remembered | -0.067 |
knew | -0.066 |
wished | -0.066 |
bit | -0.062 |
Token | happier |
Token ID | 23030 |
Feature activation | +2.15e+00 |
Pos logit contributions | |
forgot | +0.114 |
wished | +0.095 |
remembered | +0.092 |
struggled | +0.091 |
wasted | +0.091 |
Neg logit contributions | |
ing | -0.100 |
he | -0.098 |
with | -0.097 |
time | -0.095 |
again | -0.087 |
Token | . |
Token ID | 13 |
Feature activation | +1.89e+00 |
Pos logit contributions | |
he | +0.027 |
ay | +0.024 |
room | +0.023 |
ore | +0.022 |
ays | +0.022 |
Neg logit contributions | |
forgot | -0.074 |
remembered | -0.063 |
knew | -0.063 |
wished | -0.062 |
bit | -0.059 |
Token | He |
Token ID | 679 |
Feature activation | -9.19e+00 |
Pos logit contributions | |
he | +0.031 |
ing | +0.028 |
ings | +0.028 |
ay | +0.027 |
ore | +0.027 |
Neg logit contributions | |
forgot | -0.055 |
knew | -0.046 |
wished | -0.046 |
remembered | -0.045 |
knocked | -0.043 |
Token | was |
Token ID | 373 |
Feature activation | -2.75e+00 |
Pos logit contributions | |
forgot | +0.125 |
chie | +0.123 |
robbers | +0.116 |
mate | +0.111 |
women | +0.109 |
Neg logit contributions | |
he | -0.208 |
with | -0.205 |
again | -0.188 |
and | -0.186 |
, | -0.180 |
Token | no |
Token ID | 645 |
Feature activation | -1.18e+00 |
Pos logit contributions | |
forgot | +0.062 |
wished | +0.048 |
bit | +0.046 |
robbers | +0.046 |
struggled | +0.046 |
Neg logit contributions | |
he | -0.055 |
with | -0.053 |
time | -0.049 |
ing | -0.048 |
over | -0.047 |
Token | longer |
Token ID | 2392 |
Feature activation | -2.59e+00 |
Pos logit contributions | |
forgot | +0.033 |
knew | +0.028 |
remembered | +0.027 |
wished | +0.027 |
'd | +0.027 |
Neg logit contributions | |
he | -0.020 |
over | -0.017 |
ing | -0.017 |
ay | -0.017 |
time | -0.017 |
Token | a |
Token ID | 257 |
Feature activation | -3.48e+00 |
Pos logit contributions | |
forgot | +0.047 |
'd | +0.035 |
wished | +0.033 |
knew | +0.033 |
remembered | +0.033 |
Neg logit contributions | |
ing | -0.062 |
he | -0.061 |
over | -0.059 |
ore | -0.058 |
ings | -0.057 |
Token | thin |
Token ID | 7888 |
Feature activation | -1.65e+00 |
Pos logit contributions | |
forgot | +0.078 |
struggled | +0.066 |
argued | +0.062 |
waited | +0.061 |
remembered | +0.061 |
Neg logit contributions | |
time | -0.062 |
with | -0.058 |
ay | -0.058 |
ore | -0.058 |
eddy | -0.058 |
Token | he |
Token ID | 339 |
Feature activation | -8.69e+00 |
Pos logit contributions | |
forgot | +0.010 |
remembered | +0.008 |
knew | +0.008 |
wished | +0.008 |
knocked | +0.008 |
Neg logit contributions | |
he | -0.008 |
ore | -0.007 |
ings | -0.007 |
ay | -0.007 |
ing | -0.007 |
Token | was |
Token ID | 373 |
Feature activation | -2.86e+00 |
Pos logit contributions | |
forgot | +0.134 |
robbers | +0.115 |
stolen | +0.107 |
pap | +0.106 |
elves | +0.104 |
Neg logit contributions | |
he | -0.225 |
ing | -0.198 |
ore | -0.189 |
with | -0.188 |
room | -0.187 |
Token | so |
Token ID | 523 |
Feature activation | -2.39e+00 |
Pos logit contributions | |
forgot | +0.058 |
wished | +0.046 |
knew | +0.044 |
remembered | +0.043 |
cared | +0.041 |
Neg logit contributions | |
he | -0.070 |
ing | -0.064 |
ings | -0.061 |
ore | -0.061 |
room | -0.060 |
Token | happy |
Token ID | 3772 |
Feature activation | +3.23e+00 |
Pos logit contributions | |
forgot | +0.046 |
remembered | +0.034 |
wished | +0.034 |
knew | +0.034 |
bit | +0.030 |
Neg logit contributions | |
he | -0.063 |
ing | -0.060 |
ore | -0.059 |
ay | -0.059 |
ings | -0.058 |
Token | . |
Token ID | 13 |
Feature activation | +2.67e+00 |
Pos logit contributions | |
he | +0.054 |
ay | +0.052 |
ays | +0.049 |
uggle | +0.048 |
room | +0.047 |
Neg logit contributions | |
forgot | -0.093 |
knew | -0.080 |
wished | -0.077 |
remembered | -0.076 |
knocked | -0.074 |
Token | He |
Token ID | 679 |
Feature activation | -9.19e+00 |
Pos logit contributions | |
he | +0.045 |
ing | +0.042 |
ings | +0.041 |
ore | +0.041 |
ay | +0.040 |
Neg logit contributions | |
forgot | -0.076 |
knew | -0.064 |
remembered | -0.063 |
wished | -0.063 |
knocked | -0.060 |
Token | wanted |
Token ID | 2227 |
Feature activation | +8.32e-01 |
Pos logit contributions | |
forgot | +0.135 |
robbers | +0.126 |
known | +0.125 |
pap | +0.123 |
setup | +0.120 |
Neg logit contributions | |
he | -0.214 |
with | -0.196 |
ing | -0.184 |
ay | -0.182 |
at | -0.181 |
Token | to |
Token ID | 284 |
Feature activation | -3.65e+00 |
Pos logit contributions | |
he | +0.010 |
ing | +0.010 |
ore | +0.010 |
ings | +0.009 |
ay | +0.009 |
Neg logit contributions | |
forgot | -0.027 |
remembered | -0.023 |
wished | -0.023 |
knew | -0.022 |
bit | -0.021 |
Token | try |
Token ID | 1949 |
Feature activation | +1.86e+00 |
Pos logit contributions | |
forgot | +0.082 |
remembered | +0.064 |
wished | +0.064 |
knew | +0.062 |
wasted | +0.058 |
Neg logit contributions | |
he | -0.085 |
ay | -0.077 |
room | -0.077 |
ings | -0.076 |
ore | -0.076 |
Token | something |
Token ID | 1223 |
Feature activation | +1.78e-01 |
Pos logit contributions | |
he | +0.037 |
ay | +0.035 |
room | +0.034 |
eddy | +0.033 |
ore | +0.033 |
Neg logit contributions | |
forgot | -0.048 |
remembered | -0.040 |
knew | -0.040 |
wished | -0.037 |
learned | -0.037 |
Token | else |
Token ID | 2073 |
Feature activation | +1.77e+00 |
Pos logit contributions | |
he | +0.004 |
ay | +0.004 |
ays | +0.004 |
room | +0.004 |
ore | +0.004 |
Neg logit contributions | |
forgot | -0.004 |
knew | -0.003 |
remembered | -0.003 |
wished | -0.003 |
bit | -0.003 |
Token | my |
Token ID | 1820 |
Feature activation | -1.61e+00 |
Pos logit contributions | |
forgot | +0.019 |
remembered | +0.014 |
knew | +0.014 |
forget | +0.012 |
bit | +0.012 |
Neg logit contributions | |
ore | -0.031 |
he | -0.031 |
ays | -0.030 |
ings | -0.030 |
overed | -0.030 |
Token | count |
Token ID | 954 |
Feature activation | +2.25e+00 |
Pos logit contributions | |
forgot | +0.050 |
remembered | +0.042 |
knew | +0.042 |
wished | +0.040 |
bit | +0.039 |
Neg logit contributions | |
he | -0.022 |
ay | -0.021 |
ing | -0.021 |
ore | -0.021 |
ays | -0.021 |
Token | the |
Token ID | 262 |
Feature activation | -2.36e+00 |
Pos logit contributions | |
he | +0.049 |
ore | +0.046 |
uggle | +0.046 |
over | +0.045 |
eddy | +0.045 |
Neg logit contributions | |
forgot | -0.046 |
knew | -0.040 |
remembered | -0.038 |
knocked | -0.037 |
bit | -0.037 |
Token | flowers |
Token ID | 12734 |
Feature activation | +1.28e+00 |
Pos logit contributions | |
forgot | +0.059 |
wished | +0.049 |
remembered | +0.049 |
punished | +0.048 |
knew | +0.047 |
Neg logit contributions | |
ay | -0.043 |
he | -0.042 |
ings | -0.041 |
ing | -0.041 |
ummy | -0.039 |
Token | . |
Token ID | 13 |
Feature activation | +2.50e+00 |
Pos logit contributions | |
he | +0.016 |
ay | +0.015 |
ore | +0.014 |
ings | +0.014 |
ing | +0.014 |
Neg logit contributions | |
forgot | -0.042 |
knew | -0.036 |
remembered | -0.035 |
wished | -0.035 |
bit | -0.034 |
Token | She |
Token ID | 1375 |
Feature activation | -9.18e+00 |
Pos logit contributions | |
he | +0.044 |
ings | +0.038 |
ore | +0.038 |
ing | +0.037 |
over | +0.036 |
Neg logit contributions | |
forgot | -0.071 |
knew | -0.061 |
wished | -0.059 |
remembered | -0.058 |
knocked | -0.056 |
Token | felt |
Token ID | 2936 |
Feature activation | -2.04e+00 |
Pos logit contributions | |
stolen | +0.145 |
robbers | +0.143 |
dishonest | +0.139 |
chie | +0.139 |
cruel | +0.138 |
Neg logit contributions | |
with | -0.187 |
he | -0.182 |
at | -0.172 |
again | -0.165 |
room | -0.163 |
Token | wide |
Token ID | 3094 |
Feature activation | +1.61e+00 |
Pos logit contributions | |
forgot | +0.043 |
remembered | +0.033 |
knew | +0.032 |
wished | +0.032 |
bit | +0.029 |
Neg logit contributions | |
he | -0.050 |
ing | -0.048 |
ay | -0.047 |
ore | -0.047 |
ings | -0.047 |
Token | awake |
Token ID | 21693 |
Feature activation | +2.59e+00 |
Pos logit contributions | |
he | +0.027 |
ore | +0.024 |
ay | +0.024 |
ing | +0.024 |
ings | +0.024 |
Neg logit contributions | |
forgot | -0.047 |
remembered | -0.039 |
knew | -0.038 |
wished | -0.038 |
bit | -0.036 |
Token | and |
Token ID | 290 |
Feature activation | -4.38e+00 |
Pos logit contributions | |
he | +0.037 |
ay | +0.035 |
ore | +0.035 |
ays | +0.034 |
ings | +0.033 |
Neg logit contributions | |
forgot | -0.080 |
remembered | -0.067 |
knew | -0.067 |
wished | -0.066 |
bit | -0.063 |
Token | ready |
Token ID | 3492 |
Feature activation | +1.45e+00 |
Pos logit contributions | |
forgot | +0.088 |
wasted | +0.069 |
remembered | +0.068 |
wished | +0.068 |
stolen | +0.066 |
Neg logit contributions | |
he | -0.095 |
ing | -0.089 |
with | -0.086 |
ore | -0.084 |
ay | -0.083 |
Token | found |
Token ID | 1043 |
Feature activation | +4.65e-02 |
Pos logit contributions | |
forgot | +0.098 |
robbers | +0.073 |
known | +0.073 |
pap | +0.071 |
remembered | +0.068 |
Neg logit contributions | |
he | -0.168 |
ing | -0.159 |
ore | -0.153 |
room | -0.153 |
ays | -0.149 |
Token | it |
Token ID | 340 |
Feature activation | -6.01e-01 |
Pos logit contributions | |
he | +0.001 |
ing | +0.001 |
ay | +0.001 |
ore | +0.001 |
ings | +0.001 |
Neg logit contributions | |
forgot | -0.001 |
remembered | -0.001 |
knew | -0.001 |
wished | -0.001 |
bit | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | +2.81e+00 |
Pos logit contributions | |
forgot | +0.020 |
remembered | +0.017 |
knew | +0.017 |
wished | +0.017 |
bit | +0.016 |
Neg logit contributions | |
he | -0.007 |
ing | -0.006 |
ay | -0.006 |
ore | -0.006 |
ings | -0.006 |
Token | As |
Token ID | 1081 |
Feature activation | -1.24e+00 |
Pos logit contributions | |
he | +0.043 |
ing | +0.039 |
ings | +0.039 |
ay | +0.038 |
ore | +0.037 |
Neg logit contributions | |
forgot | -0.083 |
knew | -0.071 |
wished | -0.070 |
remembered | -0.069 |
learned | -0.067 |
Token | the |
Token ID | 262 |
Feature activation | -2.65e+00 |
Pos logit contributions | |
forgot | +0.036 |
remembered | +0.030 |
wished | +0.029 |
knew | +0.029 |
struggled | +0.028 |
Neg logit contributions | |
he | -0.019 |
ing | -0.018 |
ore | -0.017 |
ings | -0.017 |
time | -0.016 |
Token | woman |
Token ID | 2415 |
Feature activation | -9.18e+00 |
Pos logit contributions | |
forgot | +0.061 |
remembered | +0.048 |
known | +0.046 |
wasted | +0.046 |
struggled | +0.046 |
Neg logit contributions | |
he | -0.058 |
ay | -0.055 |
ing | -0.053 |
time | -0.051 |
ummy | -0.050 |
Token | kept |
Token ID | 4030 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
forgot | +0.152 |
robbers | +0.120 |
vitamins | +0.116 |
known | +0.114 |
stolen | +0.112 |
Neg logit contributions | |
he | -0.240 |
ing | -0.215 |
with | -0.205 |
ore | -0.203 |
room | -0.197 |
Token | walking |
Token ID | 6155 |
Feature activation | +3.53e+00 |
Pos logit contributions | |
forgot | +0.036 |
remembered | +0.028 |
knew | +0.028 |
wished | +0.028 |
'd | +0.026 |
Neg logit contributions | |
he | -0.025 |
ore | -0.024 |
ing | -0.024 |
ings | -0.024 |
ay | -0.023 |
Token | , |
Token ID | 11 |
Feature activation | -9.03e-01 |
Pos logit contributions | |
he | +0.049 |
ays | +0.046 |
ay | +0.045 |
room | +0.044 |
ore | +0.044 |
Neg logit contributions | |
forgot | -0.110 |
knew | -0.094 |
remembered | -0.093 |
wished | -0.092 |
bit | -0.089 |
Token | she |
Token ID | 673 |
Feature activation | -7.87e+00 |
Pos logit contributions | |
forgot | +0.026 |
remembered | +0.021 |
wished | +0.021 |
knew | +0.021 |
wasted | +0.020 |
Neg logit contributions | |
he | -0.015 |
ing | -0.014 |
ore | -0.013 |
ings | -0.013 |
ay | -0.013 |
Token | heard |
Token ID | 2982 |
Feature activation | +5.36e-01 |
Pos logit contributions | |
forgot | +0.119 |
robbers | +0.103 |
vitamins | +0.099 |
wasted | +0.095 |
stolen | +0.094 |
Neg logit contributions | |
ing | -0.187 |
he | -0.183 |
with | -0.177 |
again | -0.167 |
at | -0.165 |
Token | the |
Token ID | 262 |
Feature activation | -3.08e+00 |
Pos logit contributions | |
he | +0.008 |
ing | +0.007 |
ore | +0.007 |
ay | +0.007 |
ings | +0.007 |
Neg logit contributions | |
forgot | -0.013 |
remembered | -0.011 |
wished | -0.011 |
knew | -0.011 |
struggled | -0.010 |
Token | leash |
Token ID | 40529 |
Feature activation | -2.99e+00 |
Pos logit contributions | |
forgot | +0.077 |
remembered | +0.060 |
wished | +0.060 |
knew | +0.059 |
struggled | +0.057 |
Neg logit contributions | |
he | -0.065 |
ay | -0.060 |
time | -0.058 |
ings | -0.057 |
ing | -0.057 |
Token | worked |
Token ID | 3111 |
Feature activation | +2.94e+00 |
Pos logit contributions | |
forgot | +0.061 |
remembered | +0.048 |
pap | +0.047 |
known | +0.045 |
wished | +0.045 |
Neg logit contributions | |
he | -0.075 |
ing | -0.065 |
ay | -0.065 |
room | -0.064 |
ore | -0.063 |
Token | again |
Token ID | 757 |
Feature activation | +1.53e+00 |
Pos logit contributions | |
he | +0.044 |
ays | +0.041 |
ore | +0.041 |
ay | +0.040 |
room | +0.039 |
Neg logit contributions | |
forgot | -0.089 |
knew | -0.075 |
remembered | -0.075 |
wished | -0.073 |
bit | -0.070 |
Token | . |
Token ID | 13 |
Feature activation | +1.83e+00 |
Pos logit contributions | |
he | +0.019 |
ays | +0.018 |
ay | +0.017 |
ore | +0.017 |
ing | +0.017 |
Neg logit contributions | |
forgot | -0.050 |
knew | -0.043 |
remembered | -0.043 |
wished | -0.042 |
bit | -0.041 |
Token | She |
Token ID | 1375 |
Feature activation | -9.17e+00 |
Pos logit contributions | |
he | +0.026 |
ing | +0.023 |
ings | +0.023 |
ore | +0.023 |
ay | +0.023 |
Neg logit contributions | |
forgot | -0.058 |
knew | -0.049 |
remembered | -0.048 |
wished | -0.048 |
learned | -0.046 |
Token | was |
Token ID | 373 |
Feature activation | -2.63e+00 |
Pos logit contributions | |
forgot | +0.149 |
forgotten | +0.133 |
robbers | +0.132 |
stolen | +0.129 |
mate | +0.126 |
Neg logit contributions | |
with | -0.202 |
he | -0.195 |
and | -0.179 |
ing | -0.176 |
eddy | -0.173 |
Token | happy |
Token ID | 3772 |
Feature activation | +7.12e-01 |
Pos logit contributions | |
forgot | +0.061 |
wished | +0.047 |
bit | +0.045 |
struggled | +0.045 |
knew | +0.045 |
Neg logit contributions | |
he | -0.056 |
with | -0.050 |
ing | -0.047 |
time | -0.047 |
room | -0.046 |
Token | to |
Token ID | 284 |
Feature activation | -1.44e+00 |
Pos logit contributions | |
he | +0.010 |
ing | +0.009 |
ore | +0.009 |
ay | +0.009 |
ings | +0.009 |
Neg logit contributions | |
forgot | -0.023 |
remembered | -0.019 |
wished | -0.019 |
knew | -0.019 |
knocked | -0.018 |
Token | go |
Token ID | 467 |
Feature activation | +9.99e-01 |
Pos logit contributions | |
forgot | +0.026 |
remembered | +0.018 |
wished | +0.018 |
knew | +0.017 |
struggled | +0.016 |
Neg logit contributions | |
he | -0.040 |
ay | -0.038 |
room | -0.037 |
ore | -0.037 |
ings | -0.037 |
Token | on |
Token ID | 319 |
Feature activation | +8.50e-01 |
Pos logit contributions | |
he | +0.016 |
ays | +0.015 |
ings | +0.015 |
room | +0.015 |
ay | +0.015 |
Neg logit contributions | |
forgot | -0.029 |
remembered | -0.024 |
knew | -0.024 |
wished | -0.023 |
learned | -0.023 |
Token | birds |
Token ID | 10087 |
Feature activation | +2.14e-01 |
Pos logit contributions | |
forgot | +0.051 |
remembered | +0.041 |
wished | +0.038 |
knocked | +0.038 |
struggled | +0.038 |
Neg logit contributions | |
he | -0.050 |
ing | -0.047 |
ings | -0.046 |
ay | -0.045 |
ore | -0.045 |
Token | eat |
Token ID | 4483 |
Feature activation | +2.31e+00 |
Pos logit contributions | |
he | +0.004 |
ay | +0.003 |
ings | +0.003 |
ore | +0.003 |
ing | +0.003 |
Neg logit contributions | |
forgot | -0.006 |
remembered | -0.005 |
knew | -0.005 |
wished | -0.005 |
bit | -0.005 |
Token | the |
Token ID | 262 |
Feature activation | -1.51e+00 |
Pos logit contributions | |
he | +0.039 |
ay | +0.036 |
ore | +0.035 |
ays | +0.035 |
eddy | +0.034 |
Neg logit contributions | |
forgot | -0.064 |
knew | -0.055 |
remembered | -0.054 |
wished | -0.053 |
bit | -0.053 |
Token | jelly |
Token ID | 27644 |
Feature activation | +3.40e+00 |
Pos logit contributions | |
forgot | +0.032 |
remembered | +0.025 |
wished | +0.024 |
knew | +0.023 |
knocked | +0.022 |
Neg logit contributions | |
he | -0.038 |
ing | -0.036 |
ore | -0.035 |
ings | -0.035 |
ay | -0.035 |
Token | . |
Token ID | 13 |
Feature activation | +3.17e+00 |
Pos logit contributions | |
he | +0.044 |
ay | +0.043 |
room | +0.041 |
ays | +0.040 |
ings | +0.040 |
Neg logit contributions | |
forgot | -0.107 |
knew | -0.090 |
remembered | -0.088 |
bit | -0.087 |
wished | -0.087 |
Token | She |
Token ID | 1375 |
Feature activation | -9.17e+00 |
Pos logit contributions | |
he | +0.051 |
ing | +0.048 |
ings | +0.047 |
room | +0.046 |
ore | +0.045 |
Neg logit contributions | |
forgot | -0.094 |
knew | -0.079 |
wished | -0.077 |
remembered | -0.075 |
learned | -0.074 |
Token | laughed |
Token ID | 13818 |
Feature activation | +2.96e+00 |
Pos logit contributions | |
robbers | +0.137 |
chie | +0.136 |
supervision | +0.131 |
nurses | +0.127 |
prisoners | +0.126 |
Neg logit contributions | |
with | -0.201 |
and | -0.180 |
he | -0.177 |
at | -0.172 |
, | -0.171 |
Token | and |
Token ID | 290 |
Feature activation | -5.88e+00 |
Pos logit contributions | |
ore | +0.067 |
ings | +0.066 |
he | +0.066 |
ays | +0.065 |
ay | +0.062 |
Neg logit contributions | |
forgot | -0.064 |
knew | -0.059 |
remembered | -0.057 |
wished | -0.054 |
learned | -0.053 |
Token | cl |
Token ID | 537 |
Feature activation | +2.34e+00 |
Pos logit contributions | |
forgot | +0.083 |
stolen | +0.071 |
'd | +0.071 |
unknow | +0.070 |
punished | +0.070 |
Neg logit contributions | |
with | -0.137 |
he | -0.134 |
ing | -0.128 |
together | -0.123 |
on | -0.122 |
Token | apped |
Token ID | 6320 |
Feature activation | +3.94e+00 |
Pos logit contributions | |
he | +0.038 |
ore | +0.032 |
ings | +0.031 |
ing | +0.031 |
anymore | +0.030 |
Neg logit contributions | |
forgot | -0.066 |
remembered | -0.059 |
knew | -0.056 |
learned | -0.053 |
wished | -0.053 |
Token | her |
Token ID | 607 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
he | +0.087 |
ays | +0.083 |
ore | +0.082 |
ane | +0.078 |
ay | +0.077 |
Neg logit contributions | |
forgot | -0.095 |
knew | -0.080 |
remembered | -0.078 |
wished | -0.074 |
bit | -0.072 |
Token | 't |
Token ID | 470 |
Feature activation | -1.34e+00 |
Pos logit contributions | |
ings | +0.094 |
ing | +0.094 |
over | +0.092 |
ment | +0.091 |
ore | +0.089 |
Neg logit contributions | |
forgot | -0.089 |
knew | -0.071 |
knocked | -0.068 |
remembered | -0.067 |
wished | -0.066 |
Token | find |
Token ID | 1064 |
Feature activation | +7.88e-01 |
Pos logit contributions | |
forgot | +0.019 |
knew | +0.013 |
remembered | +0.013 |
wished | +0.012 |
knocked | +0.010 |
Neg logit contributions | |
ing | -0.042 |
he | -0.041 |
ays | -0.041 |
ay | -0.041 |
ings | -0.041 |
Token | it |
Token ID | 340 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
he | +0.012 |
ing | +0.011 |
ay | +0.011 |
ore | +0.011 |
ings | +0.011 |
Neg logit contributions | |
forgot | -0.024 |
remembered | -0.020 |
knew | -0.020 |
wished | -0.019 |
bit | -0.018 |
Token | . |
Token ID | 13 |
Feature activation | +3.87e+00 |
Pos logit contributions | |
ay | +0.019 |
aps | +0.019 |
ays | +0.019 |
ore | +0.019 |
he | +0.018 |
Neg logit contributions | |
forgot | -0.030 |
knew | -0.025 |
remembered | -0.025 |
wished | -0.024 |
forgotten | -0.024 |
Token | |
Token ID | 220 |
Feature activation | +5.37e+00 |
Pos logit contributions | |
he | +0.078 |
ays | +0.072 |
ings | +0.071 |
room | +0.071 |
ing | +0.070 |
Neg logit contributions | |
forgot | -0.099 |
knew | -0.081 |
remembered | -0.080 |
wished | -0.078 |
missed | -0.076 |
Token |
|
Token ID | 198 |
Feature activation | +5.35e+00 |
Pos logit contributions | |
over | +0.103 |
ipping | +0.102 |
ach | +0.100 |
ays | +0.100 |
uggle | +0.098 |
Neg logit contributions | |
forgot | -0.120 |
knew | -0.104 |
had | -0.095 |
learned | -0.092 |
missed | -0.091 |
Token |
|
Token ID | 198 |
Feature activation | +4.00e+00 |
Pos logit contributions | |
ays | +0.104 |
ing | +0.097 |
overed | +0.096 |
ipping | +0.095 |
over | +0.094 |
Neg logit contributions | |
forgot | -0.127 |
knew | -0.111 |
had | -0.101 |
learned | -0.101 |
told | -0.101 |
Token | B |
Token ID | 33 |
Feature activation | +1.95e+00 |
Pos logit contributions | |
he | +0.076 |
ays | +0.070 |
ings | +0.070 |
ore | +0.070 |
ay | +0.069 |
Neg logit contributions | |
forgot | -0.109 |
remembered | -0.089 |
knew | -0.088 |
wished | -0.087 |
bit | -0.082 |
Token | enny |
Token ID | 11870 |
Feature activation | -5.79e+00 |
Pos logit contributions | |
he | +0.037 |
ays | +0.035 |
ings | +0.034 |
ement | +0.033 |
ore | +0.032 |
Neg logit contributions | |
forgot | -0.051 |
remembered | -0.040 |
knew | -0.040 |
'd | -0.037 |
struggled | -0.037 |
Token | asked |
Token ID | 1965 |
Feature activation | +6.61e-01 |
Pos logit contributions | |
forgot | +0.114 |
known | +0.102 |
forget | +0.099 |
groom | +0.096 |
humor | +0.095 |
Neg logit contributions | |
ing | -0.108 |
with | -0.104 |
uck | -0.102 |
he | -0.098 |
time | -0.095 |
Token | his |
Token ID | 465 |
Feature activation | -2.16e+00 |
Pos logit contributions | |
he | +0.017 |
ays | +0.016 |
ay | +0.016 |
ings | +0.015 |
ore | +0.015 |
Neg logit contributions | |
forgot | -0.014 |
knew | -0.012 |
remembered | -0.012 |
wished | -0.011 |
knocked | -0.010 |
Token | them |
Token ID | 606 |
Feature activation | +6.15e-01 |
Pos logit contributions | |
he | +0.032 |
ing | +0.029 |
ay | +0.028 |
ings | +0.028 |
room | +0.028 |
Neg logit contributions | |
forgot | -0.067 |
remembered | -0.056 |
wished | -0.055 |
knew | -0.055 |
bit | -0.052 |
Token | to |
Token ID | 284 |
Feature activation | -3.41e+00 |
Pos logit contributions | |
he | +0.010 |
ing | +0.009 |
ings | +0.009 |
ore | +0.009 |
ay | +0.009 |
Neg logit contributions | |
forgot | -0.018 |
remembered | -0.015 |
wished | -0.015 |
knew | -0.015 |
bit | -0.014 |
Token | be |
Token ID | 307 |
Feature activation | -1.28e+00 |
Pos logit contributions | |
forgot | +0.065 |
remembered | +0.050 |
wished | +0.050 |
knew | +0.048 |
regretted | +0.045 |
Neg logit contributions | |
he | -0.088 |
ings | -0.082 |
room | -0.081 |
time | -0.080 |
uggle | -0.078 |
Token | careful |
Token ID | 8161 |
Feature activation | +4.32e+00 |
Pos logit contributions | |
forgot | +0.023 |
wished | +0.016 |
knew | +0.016 |
remembered | +0.015 |
knocked | +0.015 |
Neg logit contributions | |
he | -0.032 |
ings | -0.031 |
over | -0.031 |
ing | -0.031 |
time | -0.031 |
Token | . |
Token ID | 13 |
Feature activation | +3.42e+00 |
Pos logit contributions | |
he | +0.066 |
ay | +0.064 |
ays | +0.063 |
ore | +0.059 |
room | +0.056 |
Neg logit contributions | |
forgot | -0.127 |
knew | -0.107 |
remembered | -0.104 |
wished | -0.103 |
never | -0.101 |
Token |
|
Token ID | 198 |
Feature activation | +4.95e+00 |
Pos logit contributions | |
he | +0.066 |
ings | +0.063 |
ing | +0.062 |
ore | +0.062 |
ane | +0.062 |
Neg logit contributions | |
forgot | -0.086 |
remembered | -0.066 |
knew | -0.066 |
learned | -0.066 |
wished | -0.063 |
Token |
|
Token ID | 198 |
Feature activation | +3.67e+00 |
Pos logit contributions | |
ays | +0.089 |
ment | +0.085 |
overed | +0.082 |
ane | +0.082 |
ement | +0.081 |
Neg logit contributions | |
forgot | -0.127 |
knew | -0.105 |
learned | -0.101 |
had | -0.100 |
remembered | -0.096 |
Token | " |
Token ID | 1 |
Feature activation | +1.08e-01 |
Pos logit contributions | |
he | +0.062 |
ore | +0.060 |
ings | +0.059 |
ing | +0.058 |
ays | +0.058 |
Neg logit contributions | |
forgot | -0.105 |
remembered | -0.086 |
knew | -0.085 |
wished | -0.084 |
bit | -0.080 |
Token | Only |
Token ID | 10049 |
Feature activation | -2.01e+00 |
Pos logit contributions | |
he | +0.002 |
ing | +0.002 |
ore | +0.002 |
ay | +0.002 |
room | +0.002 |
Neg logit contributions | |
forgot | -0.003 |
remembered | -0.002 |
knew | -0.002 |
wished | -0.002 |
knocked | -0.002 |
Token | feed |
Token ID | 3745 |
Feature activation | +3.16e+00 |
Pos logit contributions | |
forgot | +0.058 |
wished | +0.048 |
knew | +0.047 |
remembered | +0.047 |
missed | +0.045 |
Neg logit contributions | |
ing | -0.029 |
ore | -0.028 |
he | -0.027 |
ings | -0.026 |
ummy | -0.026 |
Token | the |
Token ID | 262 |
Feature activation | -2.66e+00 |
Pos logit contributions | |
he | +0.065 |
ay | +0.061 |
ays | +0.059 |
room | +0.059 |
ore | +0.059 |
Neg logit contributions | |
forgot | -0.076 |
knew | -0.062 |
remembered | -0.061 |
wished | -0.059 |
bit | -0.058 |
Token | it |
Token ID | 340 |
Feature activation | -2.73e+00 |
Pos logit contributions | |
he | +0.059 |
ay | +0.056 |
ore | +0.056 |
room | +0.055 |
uggle | +0.055 |
Neg logit contributions | |
forgot | -0.068 |
knew | -0.054 |
remembered | -0.054 |
wished | -0.053 |
bit | -0.051 |
Token | would |
Token ID | 561 |
Feature activation | -8.50e-01 |
Pos logit contributions | |
forgot | +0.058 |
remembered | +0.046 |
wished | +0.045 |
knew | +0.042 |
pap | +0.042 |
Neg logit contributions | |
he | -0.065 |
ing | -0.060 |
room | -0.059 |
ay | -0.058 |
ore | -0.058 |
Token | go |
Token ID | 467 |
Feature activation | +3.08e+00 |
Pos logit contributions | |
forgot | +0.016 |
remembered | +0.012 |
wished | +0.012 |
knew | +0.011 |
struggled | +0.011 |
Neg logit contributions | |
he | -0.023 |
room | -0.021 |
ay | -0.021 |
ings | -0.021 |
ing | -0.021 |
Token | . |
Token ID | 13 |
Feature activation | +3.87e+00 |
Pos logit contributions | |
ays | +0.043 |
aps | +0.043 |
he | +0.043 |
ore | +0.041 |
ings | +0.040 |
Neg logit contributions | |
forgot | -0.091 |
knew | -0.078 |
bit | -0.078 |
knocked | -0.074 |
remembered | -0.074 |
Token | |
Token ID | 220 |
Feature activation | +5.23e+00 |
Pos logit contributions | |
he | +0.068 |
ing | +0.067 |
ays | +0.067 |
ings | +0.065 |
over | +0.064 |
Neg logit contributions | |
forgot | -0.106 |
knew | -0.089 |
remembered | -0.084 |
wished | -0.084 |
knocked | -0.082 |
Token |
|
Token ID | 198 |
Feature activation | +5.51e+00 |
Pos logit contributions | |
ipping | +0.102 |
over | +0.102 |
ays | +0.101 |
ing | +0.100 |
ingly | +0.098 |
Neg logit contributions | |
forgot | -0.113 |
knew | -0.098 |
learned | -0.088 |
had | -0.087 |
told | -0.085 |
Token |
|
Token ID | 198 |
Feature activation | +3.82e+00 |
Pos logit contributions | |
ays | +0.112 |
ing | +0.100 |
overed | +0.100 |
over | +0.100 |
ipping | +0.099 |
Neg logit contributions | |
forgot | -0.131 |
knew | -0.114 |
told | -0.103 |
learned | -0.103 |
had | -0.102 |
Token | As |
Token ID | 1722 |
Feature activation | +7.19e-01 |
Pos logit contributions | |
he | +0.066 |
ing | +0.061 |
ore | +0.061 |
ay | +0.061 |
ings | +0.061 |
Neg logit contributions | |
forgot | -0.109 |
remembered | -0.090 |
knew | -0.089 |
wished | -0.089 |
bit | -0.083 |
Token | he |
Token ID | 339 |
Feature activation | -6.76e+00 |
Pos logit contributions | |
he | +0.015 |
ing | +0.014 |
ore | +0.014 |
ay | +0.014 |
ings | +0.014 |
Neg logit contributions | |
forgot | -0.018 |
remembered | -0.015 |
knew | -0.014 |
wished | -0.014 |
knocked | -0.013 |
Token | was |
Token ID | 373 |
Feature activation | -1.63e+00 |
Pos logit contributions | |
forgot | +0.134 |
wished | +0.099 |
pap | +0.099 |
forget | +0.098 |
remembered | +0.097 |
Neg logit contributions | |
he | -0.168 |
room | -0.161 |
ay | -0.155 |
ing | -0.154 |
ore | -0.150 |
Token | building |
Token ID | 2615 |
Feature activation | +1.79e+00 |
Pos logit contributions | |
forgot | +0.038 |
wished | +0.031 |
remembered | +0.030 |
knew | +0.029 |
forget | +0.028 |
Neg logit contributions | |
he | -0.035 |
room | -0.033 |
ing | -0.032 |
time | -0.032 |
ore | -0.032 |
Token | wanted |
Token ID | 2227 |
Feature activation | +1.22e+00 |
Pos logit contributions | |
forgot | +0.107 |
known | +0.103 |
forgotten | +0.094 |
robbers | +0.092 |
setup | +0.091 |
Neg logit contributions | |
with | -0.130 |
he | -0.123 |
room | -0.115 |
ing | -0.113 |
at | -0.111 |
Token | to |
Token ID | 284 |
Feature activation | -1.82e+00 |
Pos logit contributions | |
he | +0.012 |
ing | +0.012 |
ore | +0.012 |
ay | +0.011 |
ings | +0.011 |
Neg logit contributions | |
forgot | -0.042 |
remembered | -0.035 |
wished | -0.034 |
knew | -0.034 |
struggled | -0.034 |
Token | wear |
Token ID | 5806 |
Feature activation | +2.64e+00 |
Pos logit contributions | |
forgot | +0.042 |
remembered | +0.033 |
wished | +0.033 |
knew | +0.031 |
struggled | +0.031 |
Neg logit contributions | |
he | -0.041 |
ay | -0.039 |
room | -0.038 |
ings | -0.037 |
ore | -0.037 |
Token | it |
Token ID | 340 |
Feature activation | +3.71e+00 |
Pos logit contributions | |
he | +0.047 |
ing | +0.044 |
ore | +0.044 |
ings | +0.043 |
room | +0.043 |
Neg logit contributions | |
forgot | -0.074 |
remembered | -0.061 |
knew | -0.060 |
wished | -0.060 |
learned | -0.056 |
Token | . |
Token ID | 13 |
Feature activation | +3.80e+00 |
Pos logit contributions | |
ays | +0.075 |
ore | +0.070 |
he | +0.069 |
eddy | +0.068 |
ings | +0.068 |
Neg logit contributions | |
forgot | -0.094 |
knew | -0.078 |
remembered | -0.077 |
bit | -0.075 |
reminded | -0.073 |
Token |
|
Token ID | 198 |
Feature activation | +5.06e+00 |
Pos logit contributions | |
he | +0.068 |
over | +0.062 |
eddy | +0.062 |
ing | +0.060 |
ays | +0.060 |
Neg logit contributions | |
forgot | -0.104 |
knew | -0.093 |
remembered | -0.085 |
wished | -0.083 |
learned | -0.082 |
Token |
|
Token ID | 198 |
Feature activation | +4.14e+00 |
Pos logit contributions | |
ays | +0.112 |
ipping | +0.109 |
ing | +0.104 |
overed | +0.104 |
over | +0.102 |
Neg logit contributions | |
forgot | -0.104 |
knew | -0.100 |
told | -0.088 |
learned | -0.086 |
had | -0.084 |
Token | M |
Token ID | 44 |
Feature activation | +2.22e+00 |
Pos logit contributions | |
ays | +0.076 |
eddy | +0.076 |
he | +0.074 |
ing | +0.073 |
ings | +0.072 |
Neg logit contributions | |
forgot | -0.110 |
knew | -0.094 |
remembered | -0.090 |
wished | -0.086 |
learned | -0.085 |
Token | ia |
Token ID | 544 |
Feature activation | -5.35e+00 |
Pos logit contributions | |
he | +0.048 |
ore | +0.044 |
ing | +0.043 |
ays | +0.042 |
uggle | +0.042 |
Neg logit contributions | |
forgot | -0.053 |
knew | -0.043 |
remembered | -0.042 |
wished | -0.039 |
knocked | -0.038 |
Token | put |
Token ID | 1234 |
Feature activation | +9.88e-01 |
Pos logit contributions | |
forgot | +0.089 |
known | +0.069 |
forget | +0.066 |
forgotten | +0.065 |
punished | +0.060 |
Neg logit contributions | |
ing | -0.131 |
he | -0.131 |
with | -0.124 |
ore | -0.123 |
room | -0.121 |
Token | on |
Token ID | 319 |
Feature activation | +1.72e+00 |
Pos logit contributions | |
he | +0.015 |
ore | +0.014 |
ings | +0.014 |
ing | +0.014 |
room | +0.013 |
Neg logit contributions | |
forgot | -0.029 |
remembered | -0.025 |
knew | -0.025 |
wished | -0.024 |
knocked | -0.023 |
Token | not |
Token ID | 407 |
Feature activation | -2.51e+00 |
Pos logit contributions | |
he | +0.004 |
ay | +0.004 |
ing | +0.004 |
ore | +0.004 |
ings | +0.004 |
Neg logit contributions | |
forgot | -0.002 |
remembered | -0.001 |
knew | -0.001 |
wished | -0.001 |
knocked | -0.001 |
Token | like |
Token ID | 588 |
Feature activation | -3.48e-01 |
Pos logit contributions | |
forgot | +0.056 |
remembered | +0.045 |
wished | +0.045 |
knew | +0.043 |
struggled | +0.042 |
Neg logit contributions | |
he | -0.057 |
ore | -0.050 |
ay | -0.050 |
room | -0.049 |
ing | -0.049 |
Token | to |
Token ID | 284 |
Feature activation | -3.27e+00 |
Pos logit contributions | |
forgot | +0.010 |
remembered | +0.008 |
knew | +0.008 |
wished | +0.008 |
bit | +0.008 |
Neg logit contributions | |
he | -0.006 |
ing | -0.005 |
ore | -0.005 |
ings | -0.005 |
ay | -0.005 |
Token | wear |
Token ID | 5806 |
Feature activation | +1.40e+00 |
Pos logit contributions | |
forgot | +0.069 |
remembered | +0.054 |
knew | +0.053 |
wished | +0.052 |
struggled | +0.047 |
Neg logit contributions | |
he | -0.080 |
ings | -0.074 |
room | -0.073 |
ay | -0.073 |
ore | -0.073 |
Token | warm |
Token ID | 5814 |
Feature activation | +2.87e+00 |
Pos logit contributions | |
he | +0.028 |
ing | +0.026 |
ay | +0.026 |
ore | +0.026 |
ings | +0.025 |
Neg logit contributions | |
forgot | -0.036 |
remembered | -0.029 |
knew | -0.029 |
wished | -0.029 |
knocked | -0.027 |
Token | clothes |
Token ID | 8242 |
Feature activation | +5.31e+00 |
Pos logit contributions | |
he | +0.085 |
ore | +0.077 |
room | +0.076 |
ay | +0.076 |
ing | +0.075 |
Neg logit contributions | |
forgot | -0.047 |
remembered | -0.035 |
wished | -0.034 |
knew | -0.033 |
knocked | -0.031 |
Token | . |
Token ID | 13 |
Feature activation | +3.43e+00 |
Pos logit contributions | |
ore | +0.063 |
ay | +0.062 |
he | +0.061 |
room | +0.060 |
ays | +0.059 |
Neg logit contributions | |
forgot | -0.175 |
remembered | -0.145 |
knew | -0.144 |
wished | -0.143 |
bit | -0.141 |
Token | They |
Token ID | 1119 |
Feature activation | -7.86e+00 |
Pos logit contributions | |
he | +0.076 |
ing | +0.068 |
ay | +0.065 |
ore | +0.065 |
uggle | +0.063 |
Neg logit contributions | |
forgot | -0.083 |
knew | -0.067 |
wished | -0.064 |
remembered | -0.063 |
knocked | -0.061 |
Token | liked |
Token ID | 8288 |
Feature activation | -1.51e-01 |
Pos logit contributions | |
forgot | +0.117 |
known | +0.116 |
shown | +0.116 |
payoff | +0.116 |
setup | +0.108 |
Neg logit contributions | |
he | -0.196 |
. | -0.150 |
at | -0.150 |
with | -0.146 |
room | -0.144 |
Token | to |
Token ID | 284 |
Feature activation | -3.12e+00 |
Pos logit contributions | |
forgot | +0.005 |
remembered | +0.004 |
knew | +0.004 |
struggled | +0.004 |
wished | +0.004 |
Neg logit contributions | |
ing | -0.002 |
ore | -0.002 |
he | -0.002 |
time | -0.002 |
eddy | -0.002 |
Token | wear |
Token ID | 5806 |
Feature activation | +1.09e+00 |
Pos logit contributions | |
forgot | +0.071 |
remembered | +0.057 |
knew | +0.055 |
wished | +0.055 |
regretted | +0.052 |
Neg logit contributions | |
he | -0.070 |
room | -0.063 |
ings | -0.063 |
uggle | -0.062 |
ay | -0.062 |
Token | it |
Token ID | 340 |
Feature activation | -2.13e+00 |
Pos logit contributions | |
he | +0.005 |
ing | +0.004 |
ay | +0.004 |
ore | +0.004 |
ings | +0.004 |
Neg logit contributions | |
forgot | -0.009 |
remembered | -0.008 |
knew | -0.008 |
wished | -0.008 |
knocked | -0.007 |
Token | was |
Token ID | 373 |
Feature activation | -1.33e+00 |
Pos logit contributions | |
forgot | +0.045 |
wished | +0.035 |
remembered | +0.035 |
wasted | +0.033 |
knew | +0.033 |
Neg logit contributions | |
ing | -0.046 |
he | -0.046 |
ings | -0.044 |
room | -0.043 |
time | -0.042 |
Token | time |
Token ID | 640 |
Feature activation | +2.54e+00 |
Pos logit contributions | |
forgot | +0.035 |
wished | +0.028 |
remembered | +0.028 |
knew | +0.028 |
bit | +0.026 |
Neg logit contributions | |
he | -0.025 |
ings | -0.023 |
ay | -0.023 |
ing | -0.023 |
ore | -0.022 |
Token | to |
Token ID | 284 |
Feature activation | -3.08e+00 |
Pos logit contributions | |
he | +0.040 |
ing | +0.037 |
ay | +0.036 |
ore | +0.036 |
ings | +0.036 |
Neg logit contributions | |
forgot | -0.075 |
remembered | -0.063 |
knew | -0.062 |
wished | -0.062 |
knocked | -0.058 |
Token | go |
Token ID | 467 |
Feature activation | +1.80e+00 |
Pos logit contributions | |
forgot | +0.066 |
remembered | +0.050 |
knew | +0.049 |
wished | +0.049 |
bit | +0.045 |
Neg logit contributions | |
he | -0.075 |
ing | -0.071 |
ore | -0.070 |
ay | -0.070 |
ings | -0.070 |
Token | home |
Token ID | 1363 |
Feature activation | +2.66e+00 |
Pos logit contributions | |
he | +0.036 |
room | +0.035 |
ings | +0.035 |
ay | +0.035 |
ing | +0.034 |
Neg logit contributions | |
forgot | -0.043 |
remembered | -0.034 |
knew | -0.034 |
wished | -0.033 |
forget | -0.032 |
Token | . |
Token ID | 13 |
Feature activation | +2.95e+00 |
Pos logit contributions | |
he | +0.030 |
ay | +0.029 |
ays | +0.027 |
ing | +0.026 |
ore | +0.026 |
Neg logit contributions | |
forgot | -0.091 |
knew | -0.077 |
remembered | -0.077 |
wished | -0.076 |
bit | -0.073 |
Token | They |
Token ID | 1119 |
Feature activation | -7.34e+00 |
Pos logit contributions | |
he | +0.050 |
ing | +0.045 |
ay | +0.044 |
ings | +0.044 |
ore | +0.044 |
Neg logit contributions | |
forgot | -0.085 |
remembered | -0.070 |
knew | -0.070 |
wished | -0.069 |
knocked | -0.065 |
Token | waved |
Token ID | 26834 |
Feature activation | +1.93e+00 |
Pos logit contributions | |
forgot | +0.123 |
pap | +0.102 |
cruel | +0.098 |
known | +0.098 |
destructive | +0.096 |
Neg logit contributions | |
he | -0.186 |
ay | -0.145 |
with | -0.143 |
room | -0.141 |
at | -0.140 |
Token | goodbye |
Token ID | 24829 |
Feature activation | +1.88e+00 |
Pos logit contributions | |
he | +0.038 |
ore | +0.034 |
room | +0.034 |
ings | +0.034 |
ay | +0.034 |
Neg logit contributions | |
forgot | -0.051 |
remembered | -0.042 |
knew | -0.042 |
wished | -0.042 |
bit | -0.039 |
Token | to |
Token ID | 284 |
Feature activation | -4.35e-01 |
Pos logit contributions | |
he | +0.028 |
ay | +0.025 |
ore | +0.025 |
ings | +0.025 |
ing | +0.025 |
Neg logit contributions | |
forgot | -0.058 |
remembered | -0.048 |
knew | -0.048 |
wished | -0.048 |
bit | -0.045 |
Token | named |
Token ID | 3706 |
Feature activation | +2.10e+00 |
Pos logit contributions | |
he | +0.045 |
ings | +0.043 |
ing | +0.042 |
ay | +0.042 |
ays | +0.042 |
Neg logit contributions | |
forgot | -0.026 |
knew | -0.019 |
remembered | -0.019 |
wished | -0.019 |
bit | -0.017 |
Token | Lily |
Token ID | 20037 |
Feature activation | +6.51e-01 |
Pos logit contributions | |
he | +0.048 |
ings | +0.046 |
ing | +0.046 |
ore | +0.045 |
ays | +0.045 |
Neg logit contributions | |
forgot | -0.049 |
knew | -0.038 |
remembered | -0.037 |
wished | -0.036 |
knocked | -0.035 |
Token | who |
Token ID | 508 |
Feature activation | -2.49e+00 |
Pos logit contributions | |
he | +0.008 |
ings | +0.008 |
ays | +0.008 |
ay | +0.007 |
ing | +0.007 |
Neg logit contributions | |
forgot | -0.021 |
knew | -0.019 |
wished | -0.019 |
remembered | -0.019 |
learned | -0.018 |
Token | loved |
Token ID | 6151 |
Feature activation | +1.04e+00 |
Pos logit contributions | |
forgot | +0.046 |
remembered | +0.032 |
wished | +0.031 |
struggled | +0.030 |
forget | +0.029 |
Neg logit contributions | |
he | -0.067 |
ay | -0.061 |
ore | -0.061 |
ing | -0.060 |
uggle | -0.060 |
Token | to |
Token ID | 284 |
Feature activation | -1.16e+00 |
Pos logit contributions | |
he | +0.012 |
ing | +0.012 |
ore | +0.012 |
eddy | +0.011 |
ay | +0.011 |
Neg logit contributions | |
forgot | -0.035 |
remembered | -0.029 |
knew | -0.028 |
wished | -0.028 |
struggled | -0.027 |
Token | read |
Token ID | 1100 |
Feature activation | +4.77e+00 |
Pos logit contributions | |
forgot | +0.023 |
remembered | +0.018 |
knew | +0.018 |
wished | +0.017 |
knocked | +0.016 |
Neg logit contributions | |
he | -0.029 |
ing | -0.028 |
ings | -0.028 |
ore | -0.028 |
ay | -0.027 |
Token | . |
Token ID | 13 |
Feature activation | +3.54e+00 |
Pos logit contributions | |
ays | +0.134 |
he | +0.132 |
ore | +0.132 |
uck | +0.131 |
ing | +0.127 |
Neg logit contributions | |
forgot | -0.063 |
knew | -0.054 |
wished | -0.053 |
sc | -0.051 |
learned | -0.049 |
Token | She |
Token ID | 1375 |
Feature activation | -6.70e+00 |
Pos logit contributions | |
ing | +0.073 |
ore | +0.072 |
he | +0.070 |
ings | +0.070 |
ach | +0.069 |
Neg logit contributions | |
forgot | -0.081 |
knew | -0.068 |
had | -0.063 |
learned | -0.063 |
remembered | -0.062 |
Token | had |
Token ID | 550 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
forgot | +0.108 |
punished | +0.084 |
forget | +0.084 |
unknow | +0.082 |
claw | +0.080 |
Neg logit contributions | |
he | -0.162 |
with | -0.145 |
time | -0.142 |
room | -0.140 |
ay | -0.140 |
Token | a |
Token ID | 257 |
Feature activation | -2.08e+00 |
Pos logit contributions | |
forgot | +0.032 |
remembered | +0.026 |
knew | +0.026 |
struggled | +0.025 |
bit | +0.025 |
Neg logit contributions | |
he | -0.014 |
ay | -0.013 |
room | -0.012 |
ing | -0.012 |
time | -0.011 |
Token | unique |
Token ID | 3748 |
Feature activation | -2.27e+00 |
Pos logit contributions | |
forgot | +0.055 |
wished | +0.042 |
struggled | +0.042 |
remembered | +0.042 |
punished | +0.040 |
Neg logit contributions | |
ay | -0.040 |
he | -0.039 |
ings | -0.036 |
time | -0.036 |
ummy | -0.036 |
Token | He |
Token ID | 679 |
Feature activation | -8.63e+00 |
Pos logit contributions | |
he | +0.038 |
ings | +0.034 |
ing | +0.034 |
ore | +0.034 |
ay | +0.033 |
Neg logit contributions | |
forgot | -0.070 |
knew | -0.060 |
wished | -0.058 |
remembered | -0.058 |
knocked | -0.055 |
Token | hugged |
Token ID | 41130 |
Feature activation | +2.61e+00 |
Pos logit contributions | |
forgot | +0.122 |
known | +0.118 |
chie | +0.112 |
sizes | +0.111 |
supervision | +0.111 |
Neg logit contributions | |
he | -0.190 |
with | -0.186 |
eddy | -0.175 |
ay | -0.166 |
and | -0.165 |
Token | his |
Token ID | 465 |
Feature activation | -1.25e+00 |
Pos logit contributions | |
he | +0.059 |
ore | +0.057 |
ays | +0.056 |
ings | +0.056 |
ing | +0.056 |
Neg logit contributions | |
forgot | -0.060 |
knew | -0.049 |
remembered | -0.048 |
wished | -0.047 |
knocked | -0.043 |
Token | mom |
Token ID | 1995 |
Feature activation | +1.87e+00 |
Pos logit contributions | |
forgot | +0.030 |
wished | +0.023 |
remembered | +0.023 |
knew | +0.023 |
struggled | +0.022 |
Neg logit contributions | |
he | -0.026 |
ay | -0.025 |
ings | -0.024 |
eddy | -0.024 |
ing | -0.024 |
Token | and |
Token ID | 290 |
Feature activation | -6.01e+00 |
Pos logit contributions | |
he | +0.028 |
ings | +0.028 |
ays | +0.028 |
room | +0.027 |
ore | +0.027 |
Neg logit contributions | |
forgot | -0.055 |
knew | -0.047 |
remembered | -0.046 |
wished | -0.045 |
bit | -0.045 |
Token | said |
Token ID | 531 |
Feature activation | +1.12e+00 |
Pos logit contributions | |
forgot | +0.103 |
chie | +0.095 |
'd | +0.094 |
tricked | +0.087 |
known | +0.086 |
Neg logit contributions | |
with | -0.125 |
he | -0.120 |
eddy | -0.117 |
ing | -0.113 |
from | -0.109 |
Token | thank |
Token ID | 5875 |
Feature activation | +3.43e+00 |
Pos logit contributions | |
he | +0.019 |
ays | +0.018 |
room | +0.017 |
ay | +0.017 |
ings | +0.017 |
Neg logit contributions | |
forgot | -0.033 |
knew | -0.029 |
remembered | -0.028 |
wished | -0.028 |
learned | -0.026 |
Token | you |
Token ID | 345 |
Feature activation | +2.39e+00 |
Pos logit contributions | |
ays | +0.083 |
ings | +0.079 |
ay | +0.077 |
he | +0.076 |
overed | +0.075 |
Neg logit contributions | |
forgot | -0.070 |
knew | -0.056 |
remembered | -0.055 |
wished | -0.052 |
bit | -0.051 |
Token | . |
Token ID | 13 |
Feature activation | +2.47e+00 |
Pos logit contributions | |
he | +0.046 |
ays | +0.044 |
ings | +0.042 |
ay | +0.042 |
room | +0.042 |
Neg logit contributions | |
forgot | -0.064 |
knew | -0.053 |
remembered | -0.051 |
wished | -0.049 |
bit | -0.048 |
Token | From |
Token ID | 3574 |
Feature activation | -1.31e-01 |
Pos logit contributions | |
he | +0.048 |
ings | +0.044 |
ing | +0.043 |
ore | +0.043 |
ay | +0.042 |
Neg logit contributions | |
forgot | -0.066 |
knew | -0.055 |
remembered | -0.054 |
wished | -0.054 |
knocked | -0.050 |
Token | that |
Token ID | 326 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
forgot | +0.004 |
wished | +0.003 |
knew | +0.003 |
remembered | +0.003 |
learned | +0.003 |
Neg logit contributions | |
ings | -0.002 |
room | -0.002 |
he | -0.002 |
ay | -0.002 |
ore | -0.002 |
Token | to |
Token ID | 284 |
Feature activation | -1.28e+00 |
Pos logit contributions | |
he | +0.015 |
ing | +0.014 |
ay | +0.014 |
ings | +0.014 |
ore | +0.014 |
Neg logit contributions | |
forgot | -0.032 |
remembered | -0.026 |
knew | -0.026 |
wished | -0.026 |
bit | -0.024 |
Token | set |
Token ID | 900 |
Feature activation | +4.36e+00 |
Pos logit contributions | |
forgot | +0.026 |
remembered | +0.020 |
wished | +0.019 |
knew | +0.019 |
struggled | +0.017 |
Neg logit contributions | |
he | -0.032 |
ing | -0.030 |
ay | -0.030 |
ings | -0.030 |
ore | -0.030 |
Token | . |
Token ID | 13 |
Feature activation | +3.37e+00 |
Pos logit contributions | |
ays | +0.069 |
he | +0.063 |
iffs | +0.060 |
ore | +0.060 |
aps | +0.059 |
Neg logit contributions | |
forgot | -0.118 |
knew | -0.104 |
bit | -0.102 |
remembered | -0.099 |
wished | -0.098 |
Token | |
Token ID | 220 |
Feature activation | +4.90e+00 |
Pos logit contributions | |
he | +0.055 |
ing | +0.052 |
ings | +0.051 |
ay | +0.049 |
ays | +0.049 |
Neg logit contributions | |
forgot | -0.097 |
knew | -0.081 |
wished | -0.080 |
remembered | -0.080 |
knocked | -0.075 |
Token |
|
Token ID | 198 |
Feature activation | +4.30e+00 |
Pos logit contributions | |
he | +0.077 |
ing | +0.075 |
ays | +0.070 |
room | +0.070 |
ement | +0.069 |
Neg logit contributions | |
forgot | -0.129 |
knew | -0.109 |
remembered | -0.103 |
learned | -0.102 |
wished | -0.101 |
Token |
|
Token ID | 198 |
Feature activation | +3.21e+00 |
Pos logit contributions | |
ays | +0.054 |
ing | +0.054 |
he | +0.054 |
ings | +0.052 |
ay | +0.050 |
Neg logit contributions | |
forgot | -0.134 |
knew | -0.117 |
remembered | -0.114 |
wished | -0.114 |
learned | -0.110 |
Token | He |
Token ID | 1544 |
Feature activation | -7.02e+00 |
Pos logit contributions | |
he | +0.051 |
ing | +0.046 |
ay | +0.045 |
ore | +0.045 |
ings | +0.045 |
Neg logit contributions | |
forgot | -0.096 |
remembered | -0.080 |
knew | -0.079 |
wished | -0.079 |
bit | -0.075 |
Token | slowly |
Token ID | 6364 |
Feature activation | -4.68e+00 |
Pos logit contributions | |
forgot | +0.131 |
robbers | +0.109 |
known | +0.100 |
stolen | +0.100 |
forget | +0.098 |
Neg logit contributions | |
he | -0.153 |
ing | -0.145 |
with | -0.143 |
ore | -0.140 |
time | -0.138 |
Token | opened |
Token ID | 4721 |
Feature activation | +1.74e+00 |
Pos logit contributions | |
forgot | +0.082 |
remembered | +0.056 |
wished | +0.055 |
knew | +0.054 |
known | +0.054 |
Neg logit contributions | |
he | -0.120 |
ing | -0.119 |
ore | -0.114 |
ings | -0.109 |
ays | -0.107 |
Token | his |
Token ID | 465 |
Feature activation | -1.86e+00 |
Pos logit contributions | |
he | +0.041 |
room | +0.039 |
ay | +0.038 |
ore | +0.038 |
uggle | +0.038 |
Neg logit contributions | |
forgot | -0.035 |
remembered | -0.030 |
knew | -0.029 |
wished | -0.028 |
knocked | -0.027 |
Token | eyes |
Token ID | 2951 |
Feature activation | +4.04e+00 |
Pos logit contributions | |
forgot | +0.049 |
wished | +0.039 |
remembered | +0.038 |
knew | +0.038 |
struggled | +0.036 |
Neg logit contributions | |
ings | -0.035 |
ay | -0.035 |
ore | -0.034 |
ing | -0.033 |
he | -0.033 |
Token | different |
Token ID | 1180 |
Feature activation | -2.38e-01 |
Pos logit contributions | |
forgot | +0.038 |
remembered | +0.030 |
knew | +0.029 |
wished | +0.029 |
struggled | +0.028 |
Neg logit contributions | |
he | -0.026 |
ay | -0.024 |
ings | -0.024 |
room | -0.024 |
ing | -0.024 |
Token | hairc |
Token ID | 37363 |
Feature activation | +5.37e+00 |
Pos logit contributions | |
forgot | +0.006 |
remembered | +0.004 |
knew | +0.004 |
wished | +0.004 |
bit | +0.004 |
Neg logit contributions | |
he | -0.005 |
ing | -0.005 |
ay | -0.005 |
ore | -0.005 |
ings | -0.005 |
Token | uts |
Token ID | 5500 |
Feature activation | +3.94e+00 |
Pos logit contributions | |
he | +0.065 |
ore | +0.064 |
uggle | +0.060 |
time | +0.060 |
ing | +0.057 |
Neg logit contributions | |
forgot | -0.159 |
knew | -0.127 |
learned | -0.122 |
wished | -0.122 |
wasted | -0.121 |
Token | . |
Token ID | 13 |
Feature activation | +2.86e+00 |
Pos logit contributions | |
room | +0.052 |
ings | +0.052 |
ay | +0.052 |
he | +0.049 |
uggle | +0.049 |
Neg logit contributions | |
forgot | -0.120 |
knew | -0.104 |
wished | -0.100 |
remembered | -0.099 |
had | -0.098 |
Token | She |
Token ID | 1375 |
Feature activation | -7.58e+00 |
Pos logit contributions | |
he | +0.061 |
ing | +0.056 |
ore | +0.054 |
ings | +0.053 |
uggle | +0.051 |
Neg logit contributions | |
forgot | -0.069 |
knew | -0.058 |
wished | -0.056 |
knocked | -0.055 |
remembered | -0.054 |
Token | asks |
Token ID | 7893 |
Feature activation | +1.73e+00 |
Pos logit contributions | |
forgot | +0.167 |
known | +0.135 |
stolen | +0.132 |
forgotten | +0.131 |
remembered | +0.130 |
Neg logit contributions | |
ays | -0.148 |
he | -0.144 |
aps | -0.133 |
ay | -0.131 |
with | -0.129 |
Token | Lily |
Token ID | 20037 |
Feature activation | +5.93e-01 |
Pos logit contributions | |
he | +0.039 |
ays | +0.037 |
ore | +0.037 |
ay | +0.036 |
uggle | +0.035 |
Neg logit contributions | |
forgot | -0.041 |
knew | -0.036 |
remembered | -0.035 |
wished | -0.032 |
learned | -0.031 |
Token | to |
Token ID | 284 |
Feature activation | -2.77e+00 |
Pos logit contributions | |
he | +0.010 |
ay | +0.009 |
room | +0.009 |
ore | +0.009 |
ings | +0.009 |
Neg logit contributions | |
forgot | -0.017 |
remembered | -0.015 |
knew | -0.015 |
wished | -0.014 |
learned | -0.013 |
Token | choose |
Token ID | 3853 |
Feature activation | +2.44e+00 |
Pos logit contributions | |
forgot | +0.057 |
wished | +0.043 |
remembered | +0.043 |
knew | +0.042 |
struggled | +0.039 |
Neg logit contributions | |
he | -0.070 |
room | -0.064 |
ay | -0.064 |
ings | -0.063 |
ore | -0.063 |
Token | one |
Token ID | 530 |
Feature activation | +9.59e-01 |
Pos logit contributions | |
he | +0.054 |
ore | +0.050 |
ay | +0.049 |
room | +0.049 |
ing | +0.048 |
Neg logit contributions | |
forgot | -0.057 |
remembered | -0.046 |
knew | -0.046 |
wished | -0.046 |
knocked | -0.043 |
Token | . |
Token ID | 13 |
Feature activation | +3.21e+00 |
Pos logit contributions | |
he | +0.016 |
room | +0.014 |
ays | +0.014 |
ing | +0.014 |
ings | +0.014 |
Neg logit contributions | |
forgot | -0.027 |
knew | -0.024 |
bit | -0.023 |
remembered | -0.023 |
wished | -0.023 |
Token | His |
Token ID | 2399 |
Feature activation | -3.16e+00 |
Pos logit contributions | |
he | +0.041 |
ing | +0.037 |
ings | +0.037 |
ay | +0.036 |
ore | +0.036 |
Neg logit contributions | |
forgot | -0.076 |
knew | -0.063 |
remembered | -0.063 |
wished | -0.063 |
learned | -0.059 |
Token | friends |
Token ID | 2460 |
Feature activation | -4.77e+00 |
Pos logit contributions | |
forgot | +0.077 |
wished | +0.058 |
remembered | +0.058 |
knew | +0.057 |
wasted | +0.055 |
Neg logit contributions | |
he | -0.065 |
ay | -0.064 |
ings | -0.062 |
ing | -0.061 |
room | -0.059 |
Token | cheered |
Token ID | 34767 |
Feature activation | +2.74e+00 |
Pos logit contributions | |
forgot | +0.085 |
pap | +0.071 |
forgotten | +0.060 |
robbers | +0.060 |
elves | +0.059 |
Neg logit contributions | |
he | -0.124 |
ing | -0.109 |
room | -0.105 |
with | -0.098 |
ore | -0.097 |
Token | him |
Token ID | 683 |
Feature activation | +6.93e-02 |
Pos logit contributions | |
he | +0.049 |
ays | +0.047 |
ore | +0.046 |
ings | +0.044 |
room | +0.044 |
Neg logit contributions | |
forgot | -0.074 |
knew | -0.064 |
remembered | -0.062 |
wished | -0.060 |
bit | -0.057 |
Token | on |
Token ID | 319 |
Feature activation | +1.88e+00 |
Pos logit contributions | |
he | +0.001 |
ays | +0.001 |
ore | +0.001 |
ummy | +0.001 |
room | +0.001 |
Neg logit contributions | |
forgot | -0.002 |
remembered | -0.002 |
knew | -0.002 |
bit | -0.002 |
wished | -0.002 |
Token | and |
Token ID | 290 |
Feature activation | -4.26e+00 |
Pos logit contributions | |
he | +0.033 |
ays | +0.032 |
ing | +0.031 |
ore | +0.031 |
ay | +0.030 |
Neg logit contributions | |
forgot | -0.051 |
knew | -0.043 |
wished | -0.042 |
remembered | -0.042 |
bit | -0.040 |
Token | they |
Token ID | 484 |
Feature activation | -6.83e+00 |
Pos logit contributions | |
forgot | +0.088 |
stolen | +0.071 |
remembered | +0.071 |
'd | +0.070 |
wasted | +0.070 |
Neg logit contributions | |
with | -0.080 |
he | -0.078 |
ing | -0.076 |
together | -0.070 |
ore | -0.068 |
Token | all |
Token ID | 477 |
Feature activation | -4.26e+00 |
Pos logit contributions | |
forgot | +0.155 |
pap | +0.127 |
forgotten | +0.121 |
stolen | +0.121 |
remembered | +0.120 |
Neg logit contributions | |
he | -0.157 |
ay | -0.125 |
room | -0.119 |
ing | -0.119 |
uggle | -0.113 |
Token | went |
Token ID | 1816 |
Feature activation | +1.66e+00 |
Pos logit contributions | |
forgot | +0.068 |
pap | +0.059 |
robbers | +0.052 |
setup | +0.050 |
stolen | +0.049 |
Neg logit contributions | |
he | -0.113 |
ing | -0.099 |
ay | -0.095 |
ore | -0.093 |
room | -0.093 |
Token | back |
Token ID | 736 |
Feature activation | +6.64e-01 |
Pos logit contributions | |
he | +0.028 |
ay | +0.025 |
ings | +0.025 |
room | +0.025 |
ore | +0.025 |
Neg logit contributions | |
forgot | -0.048 |
remembered | -0.040 |
knew | -0.039 |
wished | -0.039 |
knocked | -0.037 |
Token | inside |
Token ID | 2641 |
Feature activation | +2.19e+00 |
Pos logit contributions | |
he | +0.013 |
room | +0.012 |
ay | +0.012 |
ings | +0.011 |
ays | +0.011 |
Neg logit contributions | |
forgot | -0.017 |
remembered | -0.014 |
knew | -0.014 |
wished | -0.014 |
knocked | -0.014 |
Token | sees |
Token ID | 7224 |
Feature activation | +4.14e-01 |
Pos logit contributions | |
forgot | +0.077 |
remembered | +0.058 |
wished | +0.058 |
knew | +0.058 |
struggled | +0.053 |
Neg logit contributions | |
he | -0.088 |
ing | -0.085 |
ore | -0.084 |
ings | -0.084 |
ays | -0.083 |
Token | the |
Token ID | 262 |
Feature activation | -2.10e+00 |
Pos logit contributions | |
he | +0.008 |
ay | +0.007 |
ing | +0.007 |
ings | +0.007 |
ore | +0.007 |
Neg logit contributions | |
forgot | -0.011 |
remembered | -0.009 |
knew | -0.009 |
wished | -0.009 |
knocked | -0.008 |
Token | k |
Token ID | 479 |
Feature activation | +2.36e+00 |
Pos logit contributions | |
forgot | +0.058 |
remembered | +0.048 |
wished | +0.047 |
knew | +0.047 |
struggled | +0.044 |
Neg logit contributions | |
he | -0.039 |
ay | -0.037 |
ing | -0.035 |
ings | -0.035 |
room | -0.034 |
Token | ite |
Token ID | 578 |
Feature activation | +1.54e+00 |
Pos logit contributions | |
ently | +0.029 |
ings | +0.029 |
he | +0.029 |
ment | +0.028 |
room | +0.027 |
Neg logit contributions | |
forgot | -0.065 |
bit | -0.059 |
remembered | -0.055 |
knew | -0.052 |
wasted | -0.052 |
Token | . |
Token ID | 13 |
Feature activation | +2.43e+00 |
Pos logit contributions | |
ay | +0.019 |
he | +0.019 |
ore | +0.018 |
ings | +0.018 |
ing | +0.017 |
Neg logit contributions | |
forgot | -0.048 |
knew | -0.041 |
remembered | -0.041 |
wished | -0.040 |
bit | -0.040 |
Token | The |
Token ID | 383 |
Feature activation | -2.98e+00 |
Pos logit contributions | |
he | +0.051 |
ing | +0.049 |
ings | +0.045 |
ays | +0.044 |
ement | +0.044 |
Neg logit contributions | |
forgot | -0.056 |
knew | -0.049 |
wished | -0.047 |
bit | -0.045 |
remembered | -0.045 |
Token | k |
Token ID | 479 |
Feature activation | +1.00e+00 |
Pos logit contributions | |
forgot | +0.084 |
remembered | +0.066 |
knew | +0.064 |
wished | +0.062 |
struggled | +0.062 |
Neg logit contributions | |
he | -0.055 |
ay | -0.054 |
ings | -0.050 |
room | -0.049 |
eddy | -0.049 |
Token | ite |
Token ID | 578 |
Feature activation | -2.54e+00 |
Pos logit contributions | |
he | +0.012 |
ings | +0.011 |
time | +0.011 |
room | +0.011 |
ing | +0.011 |
Neg logit contributions | |
forgot | -0.030 |
bit | -0.026 |
remembered | -0.026 |
wished | -0.026 |
knew | -0.025 |
Token | is |
Token ID | 318 |
Feature activation | -2.11e+00 |
Pos logit contributions | |
forgot | +0.068 |
remembered | +0.055 |
wished | +0.054 |
knew | +0.054 |
known | +0.053 |
Neg logit contributions | |
he | -0.048 |
ing | -0.043 |
ays | -0.043 |
ings | -0.043 |
ore | -0.043 |
Token | red |
Token ID | 2266 |
Feature activation | +4.69e+00 |
Pos logit contributions | |
forgot | +0.045 |
knew | +0.034 |
wished | +0.033 |
remembered | +0.033 |
cared | +0.032 |
Neg logit contributions | |
he | -0.049 |
ay | -0.047 |
ings | -0.047 |
ing | -0.045 |
ays | -0.045 |
Token | and |
Token ID | 290 |
Feature activation | -3.94e+00 |
Pos logit contributions | |
ore | +0.094 |
he | +0.093 |
uggle | +0.092 |
ings | +0.088 |
ay | +0.086 |
Neg logit contributions | |
forgot | -0.105 |
bit | -0.093 |
knew | -0.089 |
wished | -0.087 |
remembered | -0.087 |
Token |
|
Token ID | 198 |
Feature activation | +3.55e+00 |
Pos logit contributions | |
he | +0.083 |
ings | +0.079 |
ay | +0.078 |
room | +0.077 |
ing | +0.077 |
Neg logit contributions | |
forgot | -0.094 |
knew | -0.081 |
learned | -0.078 |
wished | -0.078 |
remembered | -0.076 |
Token |
|
Token ID | 198 |
Feature activation | +3.09e+00 |
Pos logit contributions | |
ays | +0.069 |
ement | +0.067 |
ing | +0.066 |
over | +0.065 |
he | +0.065 |
Neg logit contributions | |
forgot | -0.080 |
knew | -0.075 |
learned | -0.071 |
wished | -0.069 |
remembered | -0.065 |
Token | They |
Token ID | 2990 |
Feature activation | -7.09e+00 |
Pos logit contributions | |
ays | +0.064 |
he | +0.063 |
ore | +0.063 |
ings | +0.062 |
ay | +0.062 |
Neg logit contributions | |
forgot | -0.073 |
remembered | -0.059 |
knew | -0.059 |
wished | -0.057 |
learned | -0.057 |
Token | walked |
Token ID | 6807 |
Feature activation | +3.06e+00 |
Pos logit contributions | |
forgot | +0.118 |
pap | +0.099 |
wished | +0.088 |
forgotten | +0.087 |
stolen | +0.087 |
Neg logit contributions | |
he | -0.191 |
room | -0.167 |
ay | -0.154 |
ing | -0.150 |
with | -0.144 |
Token | back |
Token ID | 736 |
Feature activation | +6.01e-01 |
Pos logit contributions | |
he | +0.060 |
ummy | +0.058 |
ays | +0.058 |
ay | +0.057 |
ore | +0.056 |
Neg logit contributions | |
forgot | -0.076 |
knew | -0.064 |
remembered | -0.063 |
learned | -0.062 |
bit | -0.060 |
Token | to |
Token ID | 284 |
Feature activation | +1.11e-02 |
Pos logit contributions | |
he | +0.013 |
ement | +0.012 |
over | +0.012 |
room | +0.012 |
ay | +0.012 |
Neg logit contributions | |
forgot | -0.013 |
knew | -0.011 |
remembered | -0.011 |
knocked | -0.011 |
bit | -0.010 |
Token | their |
Token ID | 511 |
Feature activation | -2.56e+00 |
Pos logit contributions | |
he | +0.000 |
ing | +0.000 |
ay | +0.000 |
ings | +0.000 |
ore | +0.000 |
Neg logit contributions | |
forgot | -0.000 |
knew | -0.000 |
remembered | -0.000 |
wished | -0.000 |
knocked | -0.000 |
Token | mom |
Token ID | 1995 |
Feature activation | +9.48e-01 |
Pos logit contributions | |
forgot | +0.066 |
remembered | +0.053 |
knew | +0.052 |
wished | +0.051 |
known | +0.051 |
Neg logit contributions | |
he | -0.048 |
eddy | -0.046 |
ings | -0.046 |
ay | -0.045 |
ummy | -0.045 |
Token | , |
Token ID | 11 |
Feature activation | -1.55e+00 |
Pos logit contributions | |
he | +0.016 |
ings | +0.015 |
ay | +0.015 |
ore | +0.015 |
overed | +0.015 |
Neg logit contributions | |
forgot | -0.026 |
remembered | -0.021 |
knew | -0.021 |
learned | -0.021 |
bit | -0.021 |
Token | holding |
Token ID | 4769 |
Feature activation | +1.51e+00 |
Pos logit contributions | |
forgot | +0.043 |
wished | +0.035 |
knew | +0.035 |
struggled | +0.035 |
remembered | +0.035 |
Neg logit contributions | |
he | -0.024 |
with | -0.022 |
ing | -0.021 |
room | -0.021 |
aps | -0.020 |
Token | hands |
Token ID | 2832 |
Feature activation | +4.09e+00 |
Pos logit contributions | |
he | +0.036 |
ore | +0.033 |
ing | +0.032 |
ummy | +0.032 |
ay | +0.032 |
Neg logit contributions | |
forgot | -0.029 |
remembered | -0.024 |
bit | -0.023 |
knew | -0.023 |
learned | -0.023 |
Token | with |
Token ID | 351 |
Feature activation | -1.77e+00 |
Pos logit contributions | |
he | +0.044 |
ore | +0.042 |
ays | +0.041 |
ay | +0.041 |
ings | +0.040 |
Neg logit contributions | |
forgot | -0.080 |
knew | -0.069 |
bit | -0.066 |
remembered | -0.066 |
wished | -0.065 |
Token | her |
Token ID | 607 |
Feature activation | -1.56e+00 |
Pos logit contributions | |
forgot | +0.054 |
remembered | +0.044 |
knew | +0.043 |
wished | +0.043 |
knocked | +0.040 |
Neg logit contributions | |
he | -0.028 |
ay | -0.026 |
ings | -0.025 |
ing | -0.025 |
room | -0.025 |
Token | hands |
Token ID | 2832 |
Feature activation | +2.70e+00 |
Pos logit contributions | |
forgot | +0.034 |
remembered | +0.026 |
knew | +0.025 |
wished | +0.025 |
knocked | +0.023 |
Neg logit contributions | |
he | -0.038 |
ing | -0.035 |
ay | -0.035 |
ings | -0.035 |
ore | -0.035 |
Token | . |
Token ID | 13 |
Feature activation | +4.00e+00 |
Pos logit contributions | |
he | +0.039 |
ore | +0.037 |
ay | +0.037 |
ays | +0.035 |
uggle | +0.035 |
Neg logit contributions | |
forgot | -0.081 |
knew | -0.069 |
remembered | -0.069 |
wished | -0.068 |
bit | -0.067 |
Token | It |
Token ID | 632 |
Feature activation | -4.68e+00 |
Pos logit contributions | |
he | +0.077 |
ing | +0.070 |
ach | +0.070 |
eddy | +0.069 |
over | +0.069 |
Neg logit contributions | |
forgot | -0.095 |
knew | -0.083 |
knocked | -0.077 |
wished | -0.077 |
learned | -0.076 |
Token | was |
Token ID | 373 |
Feature activation | -1.74e+00 |
Pos logit contributions | |
forgot | +0.098 |
pap | +0.082 |
forgotten | +0.079 |
remembered | +0.078 |
robbers | +0.078 |
Neg logit contributions | |
he | -0.096 |
with | -0.090 |
room | -0.089 |
ing | -0.085 |
ays | -0.085 |
Token | hard |
Token ID | 1327 |
Feature activation | +1.63e+00 |
Pos logit contributions | |
forgot | +0.044 |
knew | +0.036 |
wished | +0.035 |
remembered | +0.035 |
struggled | +0.033 |
Neg logit contributions | |
he | -0.031 |
ay | -0.030 |
ings | -0.030 |
ing | -0.029 |
time | -0.028 |
Token | to |
Token ID | 284 |
Feature activation | -1.61e+00 |
Pos logit contributions | |
he | +0.022 |
ing | +0.019 |
ay | +0.019 |
ore | +0.019 |
ings | +0.019 |
Neg logit contributions | |
forgot | -0.053 |
remembered | -0.045 |
knew | -0.044 |
wished | -0.044 |
bit | -0.042 |
Token | move |
Token ID | 1445 |
Feature activation | +4.38e+00 |
Pos logit contributions | |
forgot | +0.030 |
remembered | +0.022 |
knew | +0.022 |
wished | +0.021 |
knocked | +0.020 |
Neg logit contributions | |
he | -0.042 |
ing | -0.042 |
ore | -0.041 |
ay | -0.041 |
ays | -0.041 |
Token | , |
Token ID | 11 |
Feature activation | -9.97e-01 |
Pos logit contributions | |
he | +0.059 |
ore | +0.054 |
ays | +0.054 |
uggle | +0.054 |
ay | +0.053 |
Neg logit contributions | |
forgot | -0.136 |
knew | -0.118 |
wished | -0.115 |
remembered | -0.115 |
bit | -0.112 |
Token | but |
Token ID | 475 |
Feature activation | -3.26e+00 |
Pos logit contributions | |
forgot | +0.030 |
remembered | +0.025 |
knew | +0.025 |
wished | +0.025 |
struggled | +0.024 |
Neg logit contributions | |
he | -0.014 |
ing | -0.013 |
ings | -0.013 |
room | -0.013 |
ay | -0.012 |
Token | and |
Token ID | 290 |
Feature activation | -1.43e+00 |
Pos logit contributions | |
forgot | +0.040 |
wished | +0.031 |
remembered | +0.031 |
knew | +0.030 |
bit | +0.028 |
Neg logit contributions | |
he | -0.041 |
ing | -0.038 |
ay | -0.038 |
ore | -0.037 |
ings | -0.037 |
Token | Lily |
Token ID | 20037 |
Feature activation | -4.25e+00 |
Pos logit contributions | |
forgot | +0.038 |
remembered | +0.030 |
wished | +0.030 |
knew | +0.030 |
bit | +0.028 |
Neg logit contributions | |
he | -0.027 |
ay | -0.025 |
ore | -0.025 |
ing | -0.025 |
ings | -0.024 |
Token | were |
Token ID | 547 |
Feature activation | +1.78e+00 |
Pos logit contributions | |
forgot | +0.078 |
remembered | +0.057 |
wished | +0.056 |
knew | +0.055 |
bit | +0.052 |
Neg logit contributions | |
he | -0.113 |
ing | -0.103 |
room | -0.103 |
ay | -0.102 |
ings | -0.100 |
Token | playing |
Token ID | 2712 |
Feature activation | +3.40e+00 |
Pos logit contributions | |
he | +0.033 |
ing | +0.031 |
ay | +0.031 |
ore | +0.030 |
ays | +0.030 |
Neg logit contributions | |
forgot | -0.049 |
wished | -0.039 |
knew | -0.039 |
remembered | -0.039 |
bit | -0.037 |
Token | in |
Token ID | 287 |
Feature activation | +1.76e+00 |
Pos logit contributions | |
ay | +0.049 |
he | +0.046 |
ays | +0.046 |
eddy | +0.045 |
ummy | +0.045 |
Neg logit contributions | |
forgot | -0.098 |
remembered | -0.086 |
knew | -0.085 |
bit | -0.084 |
wished | -0.083 |
Token | the |
Token ID | 262 |
Feature activation | -1.10e+00 |
Pos logit contributions | |
he | +0.034 |
ing | +0.032 |
ore | +0.031 |
ay | +0.031 |
ings | +0.031 |
Neg logit contributions | |
forgot | -0.046 |
remembered | -0.037 |
knew | -0.037 |
wished | -0.037 |
knocked | -0.035 |
Token | living |
Token ID | 2877 |
Feature activation | +3.10e-01 |
Pos logit contributions | |
forgot | +0.026 |
remembered | +0.020 |
knew | +0.020 |
wished | +0.020 |
struggled | +0.018 |
Neg logit contributions | |
he | -0.025 |
ing | -0.023 |
room | -0.023 |
ay | -0.023 |
ore | -0.023 |
Token | room |
Token ID | 2119 |
Feature activation | +4.91e+00 |
Pos logit contributions | |
he | +0.008 |
ing | +0.008 |
ore | +0.008 |
ings | +0.007 |
ay | +0.007 |
Neg logit contributions | |
forgot | -0.006 |
remembered | -0.005 |
knew | -0.005 |
wished | -0.004 |
bit | -0.004 |
Token | . |
Token ID | 13 |
Feature activation | +2.91e+00 |
Pos logit contributions | |
he | +0.053 |
ay | +0.050 |
ore | +0.047 |
ings | +0.046 |
ays | +0.046 |
Neg logit contributions | |
forgot | -0.167 |
knew | -0.143 |
remembered | -0.143 |
wished | -0.142 |
bit | -0.138 |
Token | They |
Token ID | 1119 |
Feature activation | -6.04e+00 |
Pos logit contributions | |
he | +0.056 |
ing | +0.055 |
ore | +0.054 |
ings | +0.051 |
uggle | +0.050 |
Neg logit contributions | |
forgot | -0.078 |
remembered | -0.062 |
knew | -0.062 |
wished | -0.059 |
knocked | -0.059 |
Token | liked |
Token ID | 8288 |
Feature activation | +4.47e-01 |
Pos logit contributions | |
forgot | +0.110 |
pap | +0.098 |
setup | +0.094 |
regretted | +0.093 |
unknow | +0.093 |
Neg logit contributions | |
he | -0.144 |
room | -0.128 |
with | -0.119 |
ay | -0.112 |
at | -0.110 |
Token | best |
Token ID | 1266 |
Feature activation | -6.91e-01 |
Pos logit contributions | |
forgot | +0.022 |
remembered | +0.017 |
knew | +0.017 |
wished | +0.017 |
bit | +0.015 |
Neg logit contributions | |
he | -0.026 |
ing | -0.024 |
ay | -0.024 |
ore | -0.024 |
ings | -0.024 |
Token | friend |
Token ID | 1545 |
Feature activation | +6.14e-01 |
Pos logit contributions | |
forgot | +0.014 |
remembered | +0.011 |
wished | +0.010 |
knew | +0.010 |
knocked | +0.009 |
Neg logit contributions | |
he | -0.017 |
ay | -0.016 |
ing | -0.016 |
ore | -0.016 |
ings | -0.016 |
Token | , |
Token ID | 11 |
Feature activation | -2.64e+00 |
Pos logit contributions | |
he | +0.008 |
ore | +0.007 |
ays | +0.007 |
ings | +0.007 |
ay | +0.007 |
Neg logit contributions | |
forgot | -0.020 |
remembered | -0.017 |
knew | -0.017 |
wished | -0.017 |
knocked | -0.016 |
Token | Tommy |
Token ID | 19919 |
Feature activation | +4.82e-01 |
Pos logit contributions | |
forgot | +0.070 |
remembered | +0.056 |
knew | +0.055 |
wished | +0.055 |
bit | +0.052 |
Neg logit contributions | |
he | -0.050 |
ing | -0.046 |
ore | -0.046 |
ay | -0.046 |
ings | -0.045 |
Token | . |
Token ID | 13 |
Feature activation | +2.80e+00 |
Pos logit contributions | |
he | +0.004 |
ing | +0.003 |
ore | +0.003 |
ings | +0.003 |
ay | +0.003 |
Neg logit contributions | |
forgot | -0.018 |
remembered | -0.016 |
wished | -0.016 |
knew | -0.016 |
bit | -0.015 |
Token | Tommy |
Token ID | 19919 |
Feature activation | -5.80e+00 |
Pos logit contributions | |
he | +0.051 |
ore | +0.048 |
ing | +0.046 |
ays | +0.046 |
ings | +0.046 |
Neg logit contributions | |
forgot | -0.077 |
knew | -0.064 |
remembered | -0.063 |
wished | -0.062 |
learned | -0.059 |
Token | said |
Token ID | 531 |
Feature activation | +2.21e-01 |
Pos logit contributions | |
forgot | +0.107 |
known | +0.085 |
forgotten | +0.085 |
stolen | +0.084 |
vitamins | +0.081 |
Neg logit contributions | |
he | -0.128 |
ing | -0.124 |
with | -0.120 |
ore | -0.119 |
, | -0.111 |
Token | , |
Token ID | 11 |
Feature activation | +9.69e-01 |
Pos logit contributions | |
he | +0.003 |
ays | +0.003 |
ay | +0.003 |
ore | +0.003 |
ings | +0.003 |
Neg logit contributions | |
forgot | -0.007 |
knew | -0.006 |
remembered | -0.006 |
wished | -0.006 |
knocked | -0.006 |
Token | " |
Token ID | 366 |
Feature activation | +1.46e-01 |
Pos logit contributions | |
he | +0.021 |
ays | +0.020 |
ings | +0.020 |
ay | +0.020 |
room | +0.019 |
Neg logit contributions | |
forgot | -0.023 |
knew | -0.019 |
remembered | -0.019 |
wished | -0.018 |
knocked | -0.017 |
Token | You |
Token ID | 1639 |
Feature activation | -2.99e+00 |
Pos logit contributions | |
he | +0.003 |
ings | +0.003 |
ore | +0.003 |
ing | +0.003 |
ay | +0.003 |
Neg logit contributions | |
forgot | -0.004 |
remembered | -0.003 |
knew | -0.003 |
wished | -0.003 |
bit | -0.003 |
Token | can |
Token ID | 460 |
Feature activation | -1.20e+00 |
Pos logit contributions | |
forgot | +0.059 |
pap | +0.047 |
wished | +0.046 |
forgotten | +0.045 |
knocked | +0.044 |
Neg logit contributions | |
he | -0.068 |
ay | -0.064 |
room | -0.058 |
ing | -0.056 |
ore | -0.055 |
Token |
|
Token ID | 198 |
Feature activation | +3.83e+00 |
Pos logit contributions | |
ing | +0.053 |
he | +0.050 |
ings | +0.050 |
ays | +0.048 |
ore | +0.048 |
Neg logit contributions | |
forgot | -0.070 |
knew | -0.060 |
had | -0.058 |
remembered | -0.056 |
learned | -0.056 |
Token |
|
Token ID | 198 |
Feature activation | +3.17e+00 |
Pos logit contributions | |
ays | +0.081 |
ipping | +0.078 |
ing | +0.076 |
ement | +0.073 |
overed | +0.073 |
Neg logit contributions | |
forgot | -0.084 |
knew | -0.077 |
had | -0.073 |
learned | -0.070 |
told | -0.069 |
Token | One |
Token ID | 3198 |
Feature activation | -1.86e+00 |
Pos logit contributions | |
ays | +0.074 |
he | +0.073 |
ore | +0.073 |
ings | +0.072 |
ing | +0.072 |
Neg logit contributions | |
forgot | -0.067 |
knew | -0.054 |
remembered | -0.053 |
wished | -0.051 |
knocked | -0.050 |
Token | day |
Token ID | 1110 |
Feature activation | +3.57e+00 |
Pos logit contributions | |
forgot | +0.039 |
knew | +0.031 |
remembered | +0.030 |
wished | +0.028 |
bit | +0.027 |
Neg logit contributions | |
ays | -0.045 |
he | -0.045 |
ings | -0.045 |
ing | -0.043 |
ore | -0.043 |
Token | , |
Token ID | 11 |
Feature activation | +2.20e+00 |
Pos logit contributions | |
he | +0.072 |
eddy | +0.065 |
ays | +0.064 |
ore | +0.060 |
room | +0.060 |
Neg logit contributions | |
forgot | -0.095 |
remembered | -0.082 |
knew | -0.082 |
knocked | -0.080 |
bit | -0.079 |
Token | Kitty |
Token ID | 21168 |
Feature activation | -4.88e+00 |
Pos logit contributions | |
he | +0.047 |
ore | +0.045 |
ays | +0.044 |
ings | +0.043 |
eddy | +0.043 |
Neg logit contributions | |
forgot | -0.055 |
knew | -0.044 |
remembered | -0.044 |
bit | -0.042 |
wished | -0.042 |
Token | found |
Token ID | 1043 |
Feature activation | +5.33e-01 |
Pos logit contributions | |
forgot | +0.129 |
remembered | +0.105 |
wished | +0.105 |
knew | +0.104 |
known | +0.100 |
Neg logit contributions | |
he | -0.087 |
ing | -0.081 |
ore | -0.078 |
room | -0.076 |
ay | -0.076 |
Token | a |
Token ID | 257 |
Feature activation | -3.17e+00 |
Pos logit contributions | |
he | +0.011 |
ay | +0.011 |
ore | +0.010 |
ings | +0.010 |
ing | +0.010 |
Neg logit contributions | |
forgot | -0.013 |
remembered | -0.010 |
knew | -0.010 |
wished | -0.010 |
knocked | -0.010 |
Token | pretty |
Token ID | 2495 |
Feature activation | -1.34e+00 |
Pos logit contributions | |
forgot | +0.080 |
wished | +0.064 |
remembered | +0.063 |
struggled | +0.061 |
learned | +0.060 |
Neg logit contributions | |
he | -0.060 |
ay | -0.060 |
ings | -0.058 |
ing | -0.058 |
time | -0.057 |
Token | gem |
Token ID | 16840 |
Feature activation | +3.50e+00 |
Pos logit contributions | |
forgot | +0.030 |
remembered | +0.023 |
knew | +0.023 |
wished | +0.023 |
bit | +0.021 |
Neg logit contributions | |
he | -0.031 |
ing | -0.029 |
ay | -0.029 |
ore | -0.029 |
ings | -0.029 |
Token | under |
Token ID | 739 |
Feature activation | +1.30e+00 |
Pos logit contributions | |
he | +0.045 |
ings | +0.043 |
ay | +0.042 |
room | +0.042 |
ore | +0.040 |
Neg logit contributions | |
forgot | -0.110 |
knew | -0.095 |
remembered | -0.094 |
wished | -0.094 |
bit | -0.093 |
Token | and |
Token ID | 290 |
Feature activation | -3.89e-01 |
Pos logit contributions | |
forgot | +0.175 |
stolen | +0.151 |
wished | +0.150 |
forgotten | +0.147 |
wasted | +0.146 |
Neg logit contributions | |
ing | -0.076 |
he | -0.073 |
ay | -0.065 |
room | -0.065 |
ore | -0.062 |
Token | Max |
Token ID | 5436 |
Feature activation | -6.31e+00 |
Pos logit contributions | |
forgot | +0.010 |
remembered | +0.008 |
knew | +0.008 |
wished | +0.008 |
knocked | +0.007 |
Neg logit contributions | |
he | -0.008 |
ore | -0.007 |
ings | -0.007 |
ing | -0.007 |
ay | -0.007 |
Token | became |
Token ID | 2627 |
Feature activation | -1.58e+00 |
Pos logit contributions | |
forgot | +0.104 |
stolen | +0.082 |
punished | +0.079 |
wished | +0.077 |
forgotten | +0.077 |
Neg logit contributions | |
he | -0.160 |
ing | -0.157 |
ay | -0.153 |
room | -0.151 |
uggle | -0.147 |
Token | good |
Token ID | 922 |
Feature activation | +7.58e-02 |
Pos logit contributions | |
forgot | +0.034 |
wished | +0.027 |
remembered | +0.026 |
knew | +0.026 |
knocked | +0.025 |
Neg logit contributions | |
he | -0.036 |
ing | -0.035 |
ay | -0.034 |
ings | -0.034 |
room | -0.034 |
Token | friends |
Token ID | 2460 |
Feature activation | +2.57e+00 |
Pos logit contributions | |
he | +0.002 |
ore | +0.002 |
ings | +0.002 |
ay | +0.002 |
ing | +0.002 |
Neg logit contributions | |
forgot | -0.001 |
remembered | -0.001 |
knew | -0.001 |
wished | -0.001 |
learned | -0.001 |
Token | and |
Token ID | 290 |
Feature activation | -5.20e+00 |
Pos logit contributions | |
ays | +0.041 |
he | +0.040 |
ings | +0.040 |
ore | +0.039 |
room | +0.037 |
Neg logit contributions | |
forgot | -0.074 |
knew | -0.064 |
learned | -0.062 |
remembered | -0.062 |
wished | -0.061 |
Token | danced |
Token ID | 39480 |
Feature activation | +3.58e+00 |
Pos logit contributions | |
forgot | +0.091 |
stolen | +0.081 |
punished | +0.080 |
unknow | +0.078 |
wished | +0.076 |
Neg logit contributions | |
ing | -0.113 |
he | -0.108 |
uggle | -0.107 |
room | -0.107 |
with | -0.105 |
Token | together |
Token ID | 1978 |
Feature activation | +8.33e-01 |
Pos logit contributions | |
ays | +0.052 |
he | +0.049 |
ore | +0.047 |
ummy | +0.047 |
ings | +0.047 |
Neg logit contributions | |
forgot | -0.107 |
knew | -0.093 |
learned | -0.090 |
remembered | -0.089 |
wished | -0.086 |
Token | every |
Token ID | 790 |
Feature activation | -1.99e+00 |
Pos logit contributions | |
he | +0.012 |
ay | +0.011 |
ays | +0.011 |
ore | +0.011 |
ings | +0.011 |
Neg logit contributions | |
forgot | -0.026 |
knew | -0.022 |
remembered | -0.022 |
wished | -0.021 |
bit | -0.020 |
Token | time |
Token ID | 640 |
Feature activation | +2.15e+00 |
Pos logit contributions | |
forgot | +0.050 |
remembered | +0.039 |
wished | +0.039 |
knew | +0.038 |
knocked | +0.035 |
Neg logit contributions | |
he | -0.042 |
ing | -0.039 |
ay | -0.039 |
ore | -0.039 |
room | -0.038 |
Token | they |
Token ID | 484 |
Feature activation | -5.98e+00 |
Pos logit contributions | |
he | +0.050 |
ore | +0.047 |
ay | +0.046 |
ings | +0.045 |
room | +0.044 |
Neg logit contributions | |
forgot | -0.050 |
remembered | -0.038 |
knew | -0.038 |
wished | -0.038 |
knocked | -0.035 |
Token |
|
Token ID | 198 |
Feature activation | +3.68e+00 |
Pos logit contributions | |
ays | +0.104 |
ipping | +0.093 |
ing | +0.092 |
overed | +0.092 |
ement | +0.092 |
Neg logit contributions | |
forgot | -0.109 |
knew | -0.102 |
learned | -0.091 |
told | -0.090 |
had | -0.090 |
Token | He |
Token ID | 1544 |
Feature activation | -7.44e+00 |
Pos logit contributions | |
he | +0.061 |
ing | +0.056 |
ay | +0.056 |
ore | +0.056 |
ings | +0.055 |
Neg logit contributions | |
forgot | -0.108 |
remembered | -0.089 |
knew | -0.089 |
wished | -0.088 |
bit | -0.083 |
Token | did |
Token ID | 750 |
Feature activation | +9.55e-01 |
Pos logit contributions | |
forgot | +0.120 |
forget | +0.111 |
known | +0.102 |
forgotten | +0.095 |
forgetting | +0.093 |
Neg logit contributions | |
ing | -0.176 |
he | -0.167 |
room | -0.164 |
with | -0.164 |
uck | -0.160 |
Token | so |
Token ID | 523 |
Feature activation | -7.95e-01 |
Pos logit contributions | |
he | +0.028 |
ay | +0.027 |
ing | +0.026 |
ore | +0.026 |
ays | +0.026 |
Neg logit contributions | |
forgot | -0.016 |
remembered | -0.011 |
knew | -0.011 |
wished | -0.011 |
knocked | -0.010 |
Token | . |
Token ID | 13 |
Feature activation | +2.71e+00 |
Pos logit contributions | |
forgot | +0.017 |
knew | +0.014 |
remembered | +0.012 |
wished | +0.012 |
learned | +0.012 |
Neg logit contributions | |
eddy | -0.020 |
ays | -0.019 |
ay | -0.019 |
uggle | -0.019 |
ings | -0.019 |
Token | He |
Token ID | 679 |
Feature activation | -8.08e+00 |
Pos logit contributions | |
he | +0.047 |
ings | +0.043 |
ing | +0.043 |
ore | +0.041 |
ay | +0.041 |
Neg logit contributions | |
forgot | -0.076 |
knew | -0.064 |
remembered | -0.063 |
wished | -0.063 |
knocked | -0.060 |
Token | took |
Token ID | 1718 |
Feature activation | +1.93e-02 |
Pos logit contributions | |
forgot | +0.126 |
forget | +0.116 |
known | +0.114 |
forgotten | +0.105 |
stolen | +0.104 |
Neg logit contributions | |
he | -0.185 |
room | -0.183 |
ing | -0.182 |
with | -0.181 |
ore | -0.167 |
Token | off |
Token ID | 572 |
Feature activation | +1.89e+00 |
Pos logit contributions | |
he | +0.000 |
ing | +0.000 |
ore | +0.000 |
ings | +0.000 |
ay | +0.000 |
Neg logit contributions | |
forgot | -0.001 |
remembered | -0.000 |
knew | -0.000 |
wished | -0.000 |
knocked | -0.000 |
Token | his |
Token ID | 465 |
Feature activation | -1.96e+00 |
Pos logit contributions | |
he | +0.036 |
ing | +0.033 |
ay | +0.033 |
ore | +0.033 |
ings | +0.033 |
Neg logit contributions | |
forgot | -0.051 |
remembered | -0.041 |
wished | -0.041 |
knew | -0.041 |
bit | -0.038 |
Token | head |
Token ID | 1182 |
Feature activation | +3.24e+00 |
Pos logit contributions | |
forgot | +0.042 |
wished | +0.032 |
remembered | +0.031 |
knew | +0.031 |
known | +0.029 |
Neg logit contributions | |
he | -0.046 |
ay | -0.044 |
ing | -0.044 |
ore | -0.044 |
ings | -0.044 |
Token | and |
Token ID | 290 |
Feature activation | -5.93e+00 |
Pos logit contributions | |
he | +0.055 |
ay | +0.051 |
ays | +0.050 |
ings | +0.050 |
ore | +0.049 |
Neg logit contributions | |
forgot | -0.089 |
knew | -0.077 |
remembered | -0.075 |
wished | -0.073 |
bit | -0.072 |
Token | was |
Token ID | 373 |
Feature activation | -2.69e+00 |
Pos logit contributions | |
forgot | +0.101 |
known | +0.084 |
forget | +0.081 |
forgotten | +0.077 |
stolen | +0.076 |
Neg logit contributions | |
he | -0.128 |
ing | -0.125 |
with | -0.118 |
room | -0.116 |
uck | -0.116 |
Token | confused |
Token ID | 10416 |
Feature activation | +3.24e+00 |
Pos logit contributions | |
forgot | +0.065 |
knew | +0.048 |
remembered | +0.047 |
wished | +0.047 |
forget | +0.047 |
Neg logit contributions | |
he | -0.055 |
with | -0.050 |
ing | -0.050 |
over | -0.050 |
room | -0.050 |
Token | , |
Token ID | 11 |
Feature activation | -1.02e+00 |
Pos logit contributions | |
room | +0.060 |
uggle | +0.060 |
he | +0.060 |
ays | +0.059 |
ay | +0.056 |
Neg logit contributions | |
forgot | -0.085 |
knew | -0.077 |
remembered | -0.073 |
wished | -0.071 |
bit | -0.070 |
Token | he |
Token ID | 339 |
Feature activation | -6.29e+00 |
Pos logit contributions | |
forgot | +0.032 |
remembered | +0.027 |
wasted | +0.027 |
known | +0.026 |
tricked | +0.026 |
Neg logit contributions | |
he | -0.010 |
ing | -0.010 |
with | -0.010 |
time | -0.009 |
ummy | -0.008 |
Token | had |
Token ID | 550 |
Feature activation | -2.02e+00 |
Pos logit contributions | |
forgot | +0.093 |
known | +0.078 |
pap | +0.075 |
forget | +0.073 |
payoff | +0.073 |
Neg logit contributions | |
he | -0.144 |
ing | -0.144 |
time | -0.141 |
at | -0.134 |
with | -0.134 |
Token | never |
Token ID | 1239 |
Feature activation | -5.55e+00 |
Pos logit contributions | |
forgot | +0.038 |
forget | +0.030 |
pap | +0.029 |
elves | +0.028 |
sizes | +0.026 |
Neg logit contributions | |
time | -0.039 |
he | -0.039 |
with | -0.038 |
at | -0.038 |
ore | -0.037 |
Token | seen |
Token ID | 1775 |
Feature activation | +9.00e-02 |
Pos logit contributions | |
forgot | +0.094 |
pap | +0.080 |
wished | +0.075 |
majesty | +0.073 |
elves | +0.069 |
Neg logit contributions | |
he | -0.120 |
ing | -0.116 |
ore | -0.109 |
with | -0.108 |
. | -0.108 |
Token | a |
Token ID | 257 |
Feature activation | -3.21e+00 |
Pos logit contributions | |
he | +0.002 |
ing | +0.002 |
ay | +0.002 |
ore | +0.002 |
ings | +0.002 |
Neg logit contributions | |
forgot | -0.002 |
remembered | -0.002 |
knew | -0.002 |
wished | -0.002 |
bit | -0.002 |
Token | turkey |
Token ID | 26876 |
Feature activation | +1.75e+00 |
Pos logit contributions | |
forgot | +0.088 |
wished | +0.072 |
remembered | +0.070 |
knew | +0.068 |
struggled | +0.067 |
Neg logit contributions | |
he | -0.057 |
ing | -0.054 |
ore | -0.054 |
ay | -0.053 |
ings | -0.052 |
Token | before |
Token ID | 878 |
Feature activation | +1.73e+00 |
Pos logit contributions | |
he | +0.029 |
ay | +0.027 |
ings | +0.027 |
ing | +0.026 |
ays | +0.026 |
Neg logit contributions | |
forgot | -0.051 |
remembered | -0.042 |
knew | -0.042 |
wished | -0.041 |
bit | -0.040 |
Token | . |
Token ID | 13 |
Feature activation | +2.62e+00 |
Pos logit contributions | |
he | +0.025 |
ay | +0.023 |
uggle | +0.023 |
room | +0.022 |
ays | +0.022 |
Neg logit contributions | |
forgot | -0.051 |
knew | -0.045 |
wished | -0.044 |
remembered | -0.044 |
knocked | -0.043 |