Token | who |
Token ID | 508 |
Feature activation | +3.10e-01 |
Pos logit contributions | |
cups | +0.030 |
frost | +0.027 |
ugs | +0.027 |
twins | +0.026 |
ays | +0.026 |
Neg logit contributions | |
shortcut | -0.031 |
lot | -0.030 |
duel | -0.029 |
gall | -0.028 |
noble | -0.028 |
Token | had |
Token ID | 550 |
Feature activation | +2.61e+00 |
Pos logit contributions | |
cups | +0.008 |
frost | +0.007 |
ugs | +0.007 |
ays | +0.007 |
sisters | +0.007 |
Neg logit contributions | |
shortcut | -0.008 |
lot | -0.008 |
duel | -0.008 |
gall | -0.007 |
noble | -0.007 |
Token | died |
Token ID | 3724 |
Feature activation | +2.19e+00 |
Pos logit contributions | |
cups | +0.065 |
frost | +0.059 |
ugs | +0.057 |
twins | +0.057 |
sisters | +0.057 |
Neg logit contributions | |
shortcut | -0.073 |
lot | -0.070 |
duel | -0.068 |
unnecessary | -0.065 |
farewell | -0.065 |
Token | a |
Token ID | 257 |
Feature activation | +7.27e-01 |
Pos logit contributions | |
cups | +0.048 |
frost | +0.044 |
twins | +0.042 |
ugs | +0.041 |
sisters | +0.041 |
Neg logit contributions | |
shortcut | -0.068 |
lot | -0.066 |
duel | -0.065 |
noble | -0.062 |
unnecessary | -0.062 |
Token | year |
Token ID | 614 |
Feature activation | +7.30e-01 |
Pos logit contributions | |
cups | +0.019 |
frost | +0.017 |
twins | +0.017 |
sisters | +0.016 |
mug | +0.016 |
Neg logit contributions | |
lot | -0.020 |
shortcut | -0.020 |
duel | -0.019 |
farewell | -0.018 |
unnecessary | -0.018 |
Token | ago |
Token ID | 2084 |
Feature activation | +6.03e+00 |
Pos logit contributions | |
cups | +0.017 |
frost | +0.015 |
twins | +0.015 |
ugs | +0.014 |
sisters | +0.014 |
Neg logit contributions | |
shortcut | -0.022 |
lot | -0.021 |
duel | -0.020 |
farewell | -0.020 |
unnecessary | -0.020 |
Token | . |
Token ID | 13 |
Feature activation | +2.28e+00 |
Pos logit contributions | |
cups | +0.163 |
izz | +0.162 |
ays | +0.161 |
olk | +0.157 |
frost | +0.152 |
Neg logit contributions | |
lot | -0.135 |
shortcut | -0.131 |
gall | -0.123 |
duel | -0.118 |
farewell | -0.116 |
Token | It |
Token ID | 632 |
Feature activation | +4.74e-01 |
Pos logit contributions | |
cups | +0.064 |
els | +0.061 |
el | +0.060 |
frost | +0.059 |
ugs | +0.058 |
Neg logit contributions | |
lot | -0.053 |
shortcut | -0.052 |
Sir | -0.050 |
Skip | -0.048 |
noble | -0.047 |
Token | was |
Token ID | 373 |
Feature activation | +3.29e+00 |
Pos logit contributions | |
cups | +0.014 |
ugs | +0.013 |
ums | +0.013 |
frost | +0.012 |
ays | +0.012 |
Neg logit contributions | |
shortcut | -0.011 |
lot | -0.011 |
duel | -0.010 |
gall | -0.010 |
noble | -0.010 |
Token | very |
Token ID | 845 |
Feature activation | +2.15e+00 |
Pos logit contributions | |
cups | +0.091 |
ugs | +0.087 |
ays | +0.083 |
frost | +0.082 |
ums | +0.081 |
Neg logit contributions | |
shortcut | -0.080 |
lot | -0.078 |
duel | -0.074 |
noble | -0.072 |
farewell | -0.071 |
Token | precious |
Token ID | 14186 |
Feature activation | +2.09e+00 |
Pos logit contributions | |
cups | +0.064 |
ugs | +0.060 |
frost | +0.060 |
ays | +0.059 |
ions | +0.059 |
Neg logit contributions | |
shortcut | -0.047 |
lot | -0.045 |
noble | -0.043 |
duel | -0.043 |
farewell | -0.042 |
Token | someone |
Token ID | 2130 |
Feature activation | +1.87e-01 |
Pos logit contributions | |
cups | +0.037 |
frost | +0.034 |
twins | +0.033 |
ugs | +0.032 |
sisters | +0.032 |
Neg logit contributions | |
shortcut | -0.044 |
lot | -0.042 |
duel | -0.041 |
unnecessary | -0.039 |
farewell | -0.039 |
Token | a |
Token ID | 257 |
Feature activation | -1.98e-01 |
Pos logit contributions | |
cups | +0.004 |
frost | +0.004 |
twins | +0.004 |
sisters | +0.004 |
el | +0.004 |
Neg logit contributions | |
shortcut | -0.006 |
lot | -0.005 |
duel | -0.005 |
farewell | -0.005 |
noble | -0.005 |
Token | very |
Token ID | 845 |
Feature activation | +2.19e-01 |
Pos logit contributions | |
lot | +0.006 |
shortcut | +0.005 |
unnecessary | +0.005 |
farewell | +0.005 |
duel | +0.005 |
Neg logit contributions | |
cups | -0.005 |
frost | -0.005 |
mug | -0.004 |
twins | -0.004 |
kettle | -0.004 |
Token | long |
Token ID | 890 |
Feature activation | +1.97e+00 |
Pos logit contributions | |
cups | +0.006 |
twins | +0.005 |
frost | +0.005 |
sisters | +0.005 |
folders | +0.005 |
Neg logit contributions | |
lot | -0.005 |
shortcut | -0.005 |
gall | -0.005 |
duel | -0.005 |
Skip | -0.005 |
Token | time |
Token ID | 640 |
Feature activation | +2.39e+00 |
Pos logit contributions | |
cups | +0.056 |
frost | +0.052 |
twins | +0.050 |
sisters | +0.050 |
ugs | +0.049 |
Neg logit contributions | |
shortcut | -0.047 |
lot | -0.046 |
duel | -0.044 |
unnecessary | -0.043 |
noble | -0.042 |
Token | ago |
Token ID | 2084 |
Feature activation | +5.85e+00 |
Pos logit contributions | |
cups | +0.036 |
frost | +0.031 |
twins | +0.029 |
sisters | +0.028 |
ugs | +0.028 |
Neg logit contributions | |
shortcut | -0.090 |
lot | -0.088 |
duel | -0.086 |
unnecessary | -0.083 |
farewell | -0.083 |
Token | . |
Token ID | 13 |
Feature activation | +7.10e-01 |
Pos logit contributions | |
cups | +0.150 |
ays | +0.139 |
frost | +0.139 |
izz | +0.137 |
ugs | +0.136 |
Neg logit contributions | |
lot | -0.145 |
shortcut | -0.145 |
gall | -0.134 |
duel | -0.132 |
further | -0.128 |
Token | The |
Token ID | 383 |
Feature activation | -1.99e-01 |
Pos logit contributions | |
cups | +0.019 |
frost | +0.017 |
ugs | +0.017 |
els | +0.017 |
sisters | +0.017 |
Neg logit contributions | |
shortcut | -0.018 |
lot | -0.018 |
duel | -0.017 |
farewell | -0.016 |
further | -0.016 |
Token | little |
Token ID | 1310 |
Feature activation | +6.62e-01 |
Pos logit contributions | |
shortcut | +0.005 |
lot | +0.005 |
duel | +0.005 |
unnecessary | +0.005 |
gall | +0.005 |
Neg logit contributions | |
cups | -0.006 |
twins | -0.005 |
frost | -0.005 |
sisters | -0.005 |
folders | -0.005 |
Token | girl |
Token ID | 2576 |
Feature activation | +4.12e-01 |
Pos logit contributions | |
ugs | +0.015 |
cups | +0.015 |
frost | +0.014 |
ays | +0.014 |
izz | +0.014 |
Neg logit contributions | |
shortcut | -0.019 |
further | -0.019 |
higher | -0.018 |
farther | -0.018 |
closer | -0.017 |
Token | thought |
Token ID | 1807 |
Feature activation | -8.04e-01 |
Pos logit contributions | |
cups | +0.011 |
frost | +0.010 |
sisters | +0.009 |
twins | +0.009 |
el | +0.009 |
Neg logit contributions | |
shortcut | -0.011 |
lot | -0.011 |
duel | -0.011 |
unnecessary | -0.010 |
further | -0.010 |
Token | who |
Token ID | 508 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
cups | +0.037 |
olk | +0.034 |
frost | +0.033 |
ays | +0.033 |
ugs | +0.033 |
Neg logit contributions | |
lot | -0.032 |
shortcut | -0.030 |
gall | -0.029 |
higher | -0.029 |
further | -0.028 |
Token | lived |
Token ID | 5615 |
Feature activation | +1.41e+00 |
Pos logit contributions | |
shortcut | +0.003 |
lot | +0.003 |
duel | +0.003 |
unnecessary | +0.003 |
farewell | +0.003 |
Neg logit contributions | |
cups | -0.003 |
frost | -0.002 |
twins | -0.002 |
sisters | -0.002 |
ugs | -0.002 |
Token | a |
Token ID | 257 |
Feature activation | +4.64e-01 |
Pos logit contributions | |
cups | +0.030 |
twins | +0.027 |
frost | +0.026 |
sisters | +0.025 |
el | +0.025 |
Neg logit contributions | |
shortcut | -0.046 |
lot | -0.044 |
duel | -0.043 |
unnecessary | -0.042 |
farewell | -0.041 |
Token | long |
Token ID | 890 |
Feature activation | +2.37e+00 |
Pos logit contributions | |
cups | +0.011 |
twins | +0.011 |
frost | +0.010 |
sisters | +0.010 |
mug | +0.010 |
Neg logit contributions | |
shortcut | -0.013 |
lot | -0.013 |
unnecessary | -0.012 |
duel | -0.012 |
farewell | -0.012 |
Token | time |
Token ID | 640 |
Feature activation | +3.04e+00 |
Pos logit contributions | |
cups | +0.060 |
frost | +0.054 |
ugs | +0.053 |
ays | +0.052 |
twins | +0.051 |
Neg logit contributions | |
shortcut | -0.065 |
lot | -0.063 |
duel | -0.061 |
gall | -0.058 |
farewell | -0.058 |
Token | ago |
Token ID | 2084 |
Feature activation | +5.79e+00 |
Pos logit contributions | |
cups | +0.072 |
frost | +0.065 |
ugs | +0.065 |
olk | +0.064 |
ays | +0.063 |
Neg logit contributions | |
shortcut | -0.083 |
lot | -0.082 |
further | -0.078 |
duel | -0.078 |
higher | -0.077 |
Token | ," |
Token ID | 553 |
Feature activation | +1.25e+00 |
Pos logit contributions | |
izz | +0.154 |
cups | +0.154 |
ays | +0.149 |
olk | +0.147 |
frost | +0.145 |
Neg logit contributions | |
lot | -0.130 |
shortcut | -0.127 |
gall | -0.120 |
duel | -0.116 |
further | -0.115 |
Token | mom |
Token ID | 1995 |
Feature activation | +2.62e+00 |
Pos logit contributions | |
ugs | +0.028 |
izz | +0.025 |
cups | +0.025 |
ays | +0.024 |
cloth | +0.023 |
Neg logit contributions | |
lot | -0.036 |
duel | -0.036 |
farewell | -0.035 |
gall | -0.035 |
closer | -0.035 |
Token | said |
Token ID | 531 |
Feature activation | +7.90e-01 |
Pos logit contributions | |
cups | +0.071 |
olk | +0.067 |
ugs | +0.067 |
ays | +0.067 |
folders | +0.063 |
Neg logit contributions | |
lot | -0.064 |
gall | -0.060 |
closer | -0.057 |
duel | -0.056 |
shortcut | -0.056 |
Token | . |
Token ID | 13 |
Feature activation | +1.07e+00 |
Pos logit contributions | |
cups | +0.020 |
frost | +0.018 |
ugs | +0.017 |
twins | +0.017 |
ays | +0.017 |
Neg logit contributions | |
shortcut | -0.022 |
lot | -0.021 |
duel | -0.020 |
farewell | -0.020 |
further | -0.020 |
Token |
|
Token ID | 198 |
Feature activation | +3.73e-03 |
Pos logit contributions | |
cups | +0.033 |
el | +0.032 |
frost | +0.032 |
ugs | +0.032 |
els | +0.032 |
Neg logit contributions | |
lot | -0.021 |
shortcut | -0.020 |
Skip | -0.018 |
duel | -0.018 |
closer | -0.018 |
Token | important |
Token ID | 1593 |
Feature activation | -4.05e-01 |
Pos logit contributions | |
shortcut | +0.006 |
lot | +0.006 |
duel | +0.006 |
unnecessary | +0.006 |
gall | +0.006 |
Neg logit contributions | |
cups | -0.005 |
frost | -0.005 |
twins | -0.005 |
sisters | -0.005 |
el | -0.005 |
Token | person |
Token ID | 1048 |
Feature activation | +7.16e-01 |
Pos logit contributions | |
unnecessary | +0.010 |
lot | +0.010 |
shortcut | +0.010 |
duel | +0.010 |
higher | +0.010 |
Neg logit contributions | |
cups | -0.010 |
twins | -0.010 |
mug | -0.009 |
folders | -0.009 |
frost | -0.009 |
Token | a |
Token ID | 257 |
Feature activation | -8.65e-01 |
Pos logit contributions | |
cups | +0.017 |
frost | +0.015 |
twins | +0.015 |
sisters | +0.015 |
ugs | +0.015 |
Neg logit contributions | |
shortcut | -0.021 |
lot | -0.020 |
duel | -0.019 |
unnecessary | -0.019 |
farewell | -0.019 |
Token | long |
Token ID | 890 |
Feature activation | +1.67e+00 |
Pos logit contributions | |
lot | +0.022 |
nearer | +0.022 |
unnecessary | +0.022 |
farther | +0.022 |
higher | +0.021 |
Neg logit contributions | |
cups | -0.021 |
twins | -0.020 |
sisters | -0.019 |
mug | -0.019 |
cup | -0.019 |
Token | time |
Token ID | 640 |
Feature activation | +2.83e+00 |
Pos logit contributions | |
cups | +0.043 |
twins | +0.040 |
frost | +0.039 |
sisters | +0.039 |
cloth | +0.038 |
Neg logit contributions | |
shortcut | -0.043 |
lot | -0.042 |
unnecessary | -0.041 |
duel | -0.041 |
higher | -0.040 |
Token | ago |
Token ID | 2084 |
Feature activation | +5.61e+00 |
Pos logit contributions | |
cups | +0.045 |
frost | +0.039 |
ugs | +0.037 |
ays | +0.035 |
twins | +0.035 |
Neg logit contributions | |
shortcut | -0.103 |
lot | -0.101 |
duel | -0.098 |
gall | -0.095 |
further | -0.095 |
Token | . |
Token ID | 13 |
Feature activation | +6.47e-01 |
Pos logit contributions | |
cups | +0.141 |
frost | +0.129 |
ays | +0.128 |
izz | +0.126 |
ugs | +0.124 |
Neg logit contributions | |
shortcut | -0.144 |
lot | -0.144 |
gall | -0.135 |
duel | -0.134 |
farewell | -0.129 |
Token | She |
Token ID | 1375 |
Feature activation | +1.67e+00 |
Pos logit contributions | |
els | +0.017 |
frost | +0.017 |
cups | +0.017 |
ugs | +0.016 |
el | +0.016 |
Neg logit contributions | |
lot | -0.015 |
shortcut | -0.015 |
farewell | -0.014 |
Skip | -0.014 |
duel | -0.013 |
Token | was |
Token ID | 373 |
Feature activation | +1.76e+00 |
Pos logit contributions | |
cups | +0.040 |
frost | +0.036 |
ugs | +0.036 |
twins | +0.034 |
ays | +0.034 |
Neg logit contributions | |
shortcut | -0.048 |
lot | -0.047 |
duel | -0.045 |
gall | -0.044 |
noble | -0.043 |
Token | excited |
Token ID | 6568 |
Feature activation | +6.55e-01 |
Pos logit contributions | |
cups | +0.053 |
frost | +0.049 |
twins | +0.047 |
ugs | +0.047 |
sisters | +0.047 |
Neg logit contributions | |
shortcut | -0.041 |
lot | -0.039 |
duel | -0.037 |
unnecessary | -0.035 |
farewell | -0.035 |
Token | and |
Token ID | 290 |
Feature activation | -1.26e-01 |
Pos logit contributions | |
cups | +0.018 |
frost | +0.017 |
twins | +0.016 |
sisters | +0.016 |
ugs | +0.016 |
Neg logit contributions | |
shortcut | -0.017 |
lot | -0.016 |
duel | -0.016 |
unnecessary | -0.015 |
noble | -0.015 |
Token | were |
Token ID | 547 |
Feature activation | +2.36e+00 |
Pos logit contributions | |
cups | +0.030 |
frost | +0.028 |
ugs | +0.027 |
twins | +0.027 |
ays | +0.027 |
Neg logit contributions | |
shortcut | -0.027 |
lot | -0.026 |
duel | -0.025 |
noble | -0.024 |
gall | -0.024 |
Token | from |
Token ID | 422 |
Feature activation | +2.75e+00 |
Pos logit contributions | |
cups | +0.069 |
frost | +0.064 |
ugs | +0.062 |
twins | +0.062 |
sisters | +0.061 |
Neg logit contributions | |
shortcut | -0.056 |
lot | -0.054 |
duel | -0.052 |
unnecessary | -0.049 |
noble | -0.049 |
Token | a |
Token ID | 257 |
Feature activation | +9.37e-01 |
Pos logit contributions | |
cups | +0.067 |
frost | +0.061 |
ugs | +0.059 |
twins | +0.058 |
sisters | +0.058 |
Neg logit contributions | |
shortcut | -0.078 |
lot | -0.075 |
duel | -0.073 |
unnecessary | -0.070 |
farewell | -0.070 |
Token | long |
Token ID | 890 |
Feature activation | +2.63e+00 |
Pos logit contributions | |
cups | +0.025 |
twins | +0.022 |
frost | +0.022 |
sisters | +0.021 |
mug | +0.021 |
Neg logit contributions | |
lot | -0.026 |
shortcut | -0.025 |
unnecessary | -0.023 |
Skip | -0.023 |
gall | -0.023 |
Token | time |
Token ID | 640 |
Feature activation | +3.02e+00 |
Pos logit contributions | |
cups | +0.072 |
frost | +0.067 |
ugs | +0.066 |
ays | +0.065 |
twins | +0.065 |
Neg logit contributions | |
shortcut | -0.065 |
lot | -0.062 |
duel | -0.061 |
farewell | -0.058 |
further | -0.057 |
Token | ago |
Token ID | 2084 |
Feature activation | +5.58e+00 |
Pos logit contributions | |
cups | +0.071 |
frost | +0.065 |
ugs | +0.062 |
twins | +0.062 |
sisters | +0.061 |
Neg logit contributions | |
shortcut | -0.087 |
lot | -0.084 |
duel | -0.082 |
unnecessary | -0.078 |
further | -0.078 |
Token | and |
Token ID | 290 |
Feature activation | +1.28e+00 |
Pos logit contributions | |
cups | +0.139 |
frost | +0.127 |
ays | +0.125 |
ugs | +0.124 |
izz | +0.123 |
Neg logit contributions | |
shortcut | -0.145 |
lot | -0.144 |
duel | -0.134 |
gall | -0.133 |
farewell | -0.130 |
Token | they |
Token ID | 484 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
cups | +0.032 |
frost | +0.029 |
twins | +0.028 |
ugs | +0.028 |
sisters | +0.028 |
Neg logit contributions | |
shortcut | -0.036 |
lot | -0.034 |
duel | -0.033 |
unnecessary | -0.032 |
farewell | -0.032 |
Token | help |
Token ID | 1037 |
Feature activation | +1.72e+00 |
Pos logit contributions | |
cups | +0.029 |
frost | +0.027 |
twins | +0.026 |
ugs | +0.026 |
sisters | +0.026 |
Neg logit contributions | |
shortcut | -0.031 |
lot | -0.030 |
duel | -0.029 |
unnecessary | -0.027 |
farewell | -0.027 |
Token | us |
Token ID | 514 |
Feature activation | -5.52e-01 |
Pos logit contributions | |
cups | +0.040 |
frost | +0.036 |
ugs | +0.035 |
twins | +0.034 |
ays | +0.034 |
Neg logit contributions | |
shortcut | -0.051 |
lot | -0.049 |
duel | -0.048 |
unnecessary | -0.046 |
further | -0.046 |
Token | remember |
Token ID | 3505 |
Feature activation | +1.51e+00 |
Pos logit contributions | |
shortcut | +0.015 |
lot | +0.014 |
duel | +0.014 |
farewell | +0.013 |
noble | +0.013 |
Neg logit contributions | |
cups | -0.015 |
frost | -0.013 |
twins | -0.013 |
sisters | -0.013 |
ugs | -0.013 |
Token | her |
Token ID | 607 |
Feature activation | +8.91e-01 |
Pos logit contributions | |
cups | +0.048 |
frost | +0.042 |
twins | +0.039 |
ugs | +0.039 |
sisters | +0.038 |
Neg logit contributions | |
shortcut | -0.108 |
lot | -0.105 |
duel | -0.102 |
unnecessary | -0.099 |
farewell | -0.099 |
Token | birthday |
Token ID | 10955 |
Feature activation | +2.77e+00 |
Pos logit contributions | |
cups | +0.019 |
twins | +0.017 |
sisters | +0.017 |
frost | +0.017 |
mug | +0.016 |
Neg logit contributions | |
shortcut | -0.028 |
lot | -0.028 |
duel | -0.027 |
unnecessary | -0.026 |
noble | -0.025 |
Token | a |
Token ID | 257 |
Feature activation | -7.61e-01 |
Pos logit contributions | |
cups | +0.066 |
frost | +0.060 |
ays | +0.059 |
ugs | +0.058 |
twins | +0.058 |
Neg logit contributions | |
shortcut | -0.078 |
lot | -0.076 |
duel | -0.073 |
gall | -0.070 |
farewell | -0.069 |
Token | few |
Token ID | 1178 |
Feature activation | +2.47e+00 |
Pos logit contributions | |
lot | +0.020 |
unnecessary | +0.019 |
shortcut | +0.019 |
nearer | +0.019 |
ashore | +0.018 |
Neg logit contributions | |
cups | -0.020 |
mug | -0.019 |
frost | -0.018 |
twins | -0.018 |
sisters | -0.017 |
Token | days |
Token ID | 1528 |
Feature activation | +3.21e+00 |
Pos logit contributions | |
cups | +0.056 |
frost | +0.051 |
ugs | +0.049 |
twins | +0.048 |
ays | +0.048 |
Neg logit contributions | |
shortcut | -0.074 |
lot | -0.071 |
duel | -0.070 |
farewell | -0.067 |
noble | -0.067 |
Token | ago |
Token ID | 2084 |
Feature activation | +5.48e+00 |
Pos logit contributions | |
cups | +0.081 |
frost | +0.075 |
olk | +0.072 |
folders | +0.072 |
ays | +0.071 |
Neg logit contributions | |
shortcut | -0.081 |
lot | -0.080 |
gall | -0.078 |
further | -0.076 |
duel | -0.075 |
Token | . |
Token ID | 13 |
Feature activation | +6.54e-01 |
Pos logit contributions | |
cups | +0.141 |
ays | +0.135 |
izz | +0.131 |
frost | +0.129 |
olk | +0.128 |
Neg logit contributions | |
lot | -0.134 |
shortcut | -0.130 |
gall | -0.126 |
farewell | -0.120 |
duel | -0.120 |
Token | She |
Token ID | 1375 |
Feature activation | +1.36e+00 |
Pos logit contributions | |
cups | +0.018 |
frost | +0.017 |
ugs | +0.016 |
ays | +0.015 |
twins | +0.015 |
Neg logit contributions | |
shortcut | -0.016 |
lot | -0.016 |
duel | -0.015 |
farewell | -0.015 |
further | -0.015 |
Token | liked |
Token ID | 8288 |
Feature activation | +1.98e-01 |
Pos logit contributions | |
cups | +0.035 |
frost | +0.032 |
twins | +0.031 |
sisters | +0.031 |
ugs | +0.031 |
Neg logit contributions | |
shortcut | -0.037 |
lot | -0.036 |
duel | -0.035 |
unnecessary | -0.033 |
farewell | -0.033 |
Token | to |
Token ID | 284 |
Feature activation | -7.58e-02 |
Pos logit contributions | |
cups | +0.005 |
frost | +0.004 |
twins | +0.004 |
sisters | +0.004 |
mug | +0.004 |
Neg logit contributions | |
shortcut | -0.006 |
lot | -0.006 |
duel | -0.005 |
unnecessary | -0.005 |
noble | -0.005 |
Token | practice |
Token ID | 3357 |
Feature activation | +3.96e-01 |
Pos logit contributions | |
shortcut | +0.002 |
lot | +0.002 |
duel | +0.002 |
farewell | +0.002 |
unnecessary | +0.002 |
Neg logit contributions | |
cups | -0.002 |
frost | -0.002 |
twins | -0.002 |
ugs | -0.002 |
sisters | -0.002 |
Token | same |
Token ID | 976 |
Feature activation | +2.29e+00 |
Pos logit contributions | |
cups | +0.014 |
frost | +0.013 |
ugs | +0.012 |
twins | +0.012 |
sisters | +0.012 |
Neg logit contributions | |
shortcut | -0.022 |
lot | -0.021 |
duel | -0.020 |
farewell | -0.019 |
gall | -0.019 |
Token | dress |
Token ID | 6576 |
Feature activation | +1.00e+00 |
Pos logit contributions | |
cups | +0.036 |
ugs | +0.031 |
frost | +0.031 |
ays | +0.030 |
twins | +0.029 |
Neg logit contributions | |
shortcut | -0.084 |
lot | -0.081 |
duel | -0.079 |
farewell | -0.078 |
gall | -0.077 |
Token | a |
Token ID | 257 |
Feature activation | +6.66e-01 |
Pos logit contributions | |
cups | +0.023 |
frost | +0.020 |
twins | +0.019 |
ugs | +0.019 |
sisters | +0.019 |
Neg logit contributions | |
shortcut | -0.031 |
lot | -0.030 |
duel | -0.029 |
unnecessary | -0.029 |
noble | -0.028 |
Token | long |
Token ID | 890 |
Feature activation | +2.83e+00 |
Pos logit contributions | |
cups | +0.015 |
frost | +0.014 |
twins | +0.013 |
sisters | +0.013 |
mug | +0.013 |
Neg logit contributions | |
shortcut | -0.020 |
lot | -0.020 |
duel | -0.019 |
unnecessary | -0.019 |
farewell | -0.018 |
Token | time |
Token ID | 640 |
Feature activation | +2.12e+00 |
Pos logit contributions | |
cups | +0.058 |
frost | +0.051 |
ugs | +0.050 |
twins | +0.050 |
ays | +0.049 |
Neg logit contributions | |
shortcut | -0.091 |
lot | -0.089 |
duel | -0.086 |
farewell | -0.083 |
gall | -0.083 |
Token | ago |
Token ID | 2084 |
Feature activation | +5.23e+00 |
Pos logit contributions | |
cups | +0.048 |
frost | +0.044 |
ugs | +0.042 |
twins | +0.042 |
ays | +0.042 |
Neg logit contributions | |
shortcut | -0.062 |
lot | -0.061 |
duel | -0.058 |
further | -0.057 |
gall | -0.057 |
Token | . |
Token ID | 13 |
Feature activation | +3.16e-01 |
Pos logit contributions | |
cups | +0.132 |
ays | +0.122 |
izz | +0.122 |
frost | +0.121 |
ugs | +0.118 |
Neg logit contributions | |
lot | -0.133 |
shortcut | -0.131 |
gall | -0.124 |
duel | -0.123 |
farewell | -0.119 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -4.67e-01 |
Pos logit contributions | |
cups | +0.009 |
frost | +0.008 |
els | +0.008 |
ugs | +0.008 |
cloth | +0.008 |
Neg logit contributions | |
lot | -0.007 |
shortcut | -0.007 |
farewell | -0.007 |
duel | -0.007 |
noble | -0.007 |
Token | Once |
Token ID | 7454 |
Feature activation | +2.04e-01 |
Pos logit contributions | |
shortcut | +0.013 |
lot | +0.013 |
duel | +0.012 |
unnecessary | +0.012 |
noble | +0.012 |
Neg logit contributions | |
cups | -0.011 |
frost | -0.010 |
twins | -0.010 |
ugs | -0.010 |
sisters | -0.010 |
Token | upon |
Token ID | 2402 |
Feature activation | -9.27e-01 |
Pos logit contributions | |
cups | +0.007 |
frost | +0.006 |
twins | +0.006 |
ugs | +0.006 |
sisters | +0.006 |
Neg logit contributions | |
shortcut | -0.004 |
lot | -0.004 |
duel | -0.003 |
farewell | -0.003 |
further | -0.003 |
Token | a |
Token ID | 257 |
Feature activation | -2.31e-01 |
Pos logit contributions | |
shortcut | +0.029 |
lot | +0.028 |
duel | +0.027 |
unnecessary | +0.026 |
noble | +0.026 |
Neg logit contributions | |
cups | -0.020 |
frost | -0.019 |
twins | -0.018 |
sisters | -0.017 |
ugs | -0.017 |
Token | this |
Token ID | 428 |
Feature activation | +5.55e-01 |
Pos logit contributions | |
cups | +0.048 |
ugs | +0.046 |
frost | +0.045 |
ays | +0.045 |
ums | +0.044 |
Neg logit contributions | |
shortcut | -0.047 |
duel | -0.042 |
further | -0.042 |
lot | -0.042 |
unnecessary | -0.041 |
Token | house |
Token ID | 2156 |
Feature activation | +1.52e+00 |
Pos logit contributions | |
cups | +0.013 |
frost | +0.012 |
twins | +0.011 |
el | +0.011 |
mug | +0.011 |
Neg logit contributions | |
lot | -0.017 |
shortcut | -0.017 |
duel | -0.016 |
unnecessary | -0.015 |
Skip | -0.015 |
Token | a |
Token ID | 257 |
Feature activation | +3.25e-01 |
Pos logit contributions | |
cups | +0.037 |
frost | +0.033 |
twins | +0.032 |
ugs | +0.032 |
sisters | +0.032 |
Neg logit contributions | |
shortcut | -0.044 |
lot | -0.042 |
duel | -0.041 |
unnecessary | -0.039 |
farewell | -0.039 |
Token | long |
Token ID | 890 |
Feature activation | +2.54e+00 |
Pos logit contributions | |
cups | +0.009 |
frost | +0.008 |
twins | +0.008 |
mug | +0.007 |
sisters | +0.007 |
Neg logit contributions | |
lot | -0.009 |
shortcut | -0.009 |
unnecessary | -0.008 |
farewell | -0.008 |
duel | -0.008 |
Token | time |
Token ID | 640 |
Feature activation | +2.15e+00 |
Pos logit contributions | |
cups | +0.072 |
frost | +0.067 |
ays | +0.066 |
ugs | +0.066 |
twins | +0.065 |
Neg logit contributions | |
shortcut | -0.062 |
lot | -0.058 |
duel | -0.056 |
gall | -0.054 |
farewell | -0.054 |
Token | ago |
Token ID | 2084 |
Feature activation | +5.22e+00 |
Pos logit contributions | |
cups | +0.035 |
frost | +0.030 |
ugs | +0.029 |
ays | +0.028 |
el | +0.028 |
Neg logit contributions | |
shortcut | -0.077 |
lot | -0.075 |
duel | -0.073 |
further | -0.071 |
gall | -0.071 |
Token | ." |
Token ID | 526 |
Feature activation | +6.40e-01 |
Pos logit contributions | |
cups | +0.134 |
izz | +0.127 |
ays | +0.125 |
frost | +0.124 |
olk | +0.122 |
Neg logit contributions | |
lot | -0.128 |
shortcut | -0.126 |
gall | -0.117 |
duel | -0.116 |
farewell | -0.112 |
Token |
|
Token ID | 198 |
Feature activation | +2.46e-01 |
Pos logit contributions | |
cups | +0.017 |
frost | +0.017 |
els | +0.016 |
el | +0.016 |
ugs | +0.016 |
Neg logit contributions | |
shortcut | -0.015 |
lot | -0.015 |
farewell | -0.014 |
duel | -0.013 |
higher | -0.013 |
Token |
|
Token ID | 198 |
Feature activation | +1.44e+00 |
Pos logit contributions | |
cups | +0.007 |
frost | +0.006 |
sisters | +0.006 |
ugs | +0.006 |
twins | +0.006 |
Neg logit contributions | |
shortcut | -0.006 |
lot | -0.006 |
duel | -0.006 |
farewell | -0.006 |
gall | -0.005 |
Token | L |
Token ID | 43 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
cups | +0.032 |
frost | +0.030 |
ays | +0.029 |
ugs | +0.028 |
el | +0.028 |
Neg logit contributions | |
shortcut | -0.043 |
lot | -0.042 |
Hop | -0.042 |
duel | -0.041 |
further | -0.040 |
Token | ily |
Token ID | 813 |
Feature activation | +8.50e-01 |
Pos logit contributions | |
cups | +0.021 |
frost | +0.018 |
kettle | +0.017 |
mug | +0.016 |
Lem | +0.016 |
Neg logit contributions | |
lot | -0.039 |
shortcut | -0.039 |
further | -0.037 |
farewell | -0.037 |
Hop | -0.036 |
Token | had |
Token ID | 550 |
Feature activation | +2.67e+00 |
Pos logit contributions | |
cups | +0.033 |
frost | +0.031 |
twins | +0.029 |
sisters | +0.029 |
ugs | +0.028 |
Neg logit contributions | |
shortcut | -0.041 |
lot | -0.039 |
duel | -0.038 |
unnecessary | -0.037 |
farewell | -0.036 |
Token | bought |
Token ID | 5839 |
Feature activation | +1.55e+00 |
Pos logit contributions | |
cups | +0.067 |
frost | +0.062 |
twins | +0.058 |
sisters | +0.057 |
cloth | +0.056 |
Neg logit contributions | |
shortcut | -0.075 |
lot | -0.073 |
duel | -0.070 |
unnecessary | -0.068 |
noble | -0.067 |
Token | a |
Token ID | 257 |
Feature activation | -7.03e-02 |
Pos logit contributions | |
cups | +0.037 |
frost | +0.034 |
twins | +0.032 |
sisters | +0.031 |
el | +0.031 |
Neg logit contributions | |
shortcut | -0.046 |
lot | -0.045 |
duel | -0.044 |
unnecessary | -0.042 |
noble | -0.042 |
Token | long |
Token ID | 890 |
Feature activation | +2.00e+00 |
Pos logit contributions | |
lot | +0.002 |
shortcut | +0.002 |
duel | +0.002 |
unnecessary | +0.002 |
farewell | +0.002 |
Neg logit contributions | |
cups | -0.002 |
frost | -0.002 |
mug | -0.002 |
twins | -0.002 |
kettle | -0.001 |
Token | time |
Token ID | 640 |
Feature activation | +2.04e+00 |
Pos logit contributions | |
cups | +0.049 |
frost | +0.044 |
twins | +0.043 |
ugs | +0.043 |
sisters | +0.042 |
Neg logit contributions | |
shortcut | -0.057 |
lot | -0.055 |
duel | -0.053 |
farewell | -0.051 |
gall | -0.051 |
Token | ago |
Token ID | 2084 |
Feature activation | +5.20e+00 |
Pos logit contributions | |
cups | +0.046 |
frost | +0.042 |
ugs | +0.040 |
sisters | +0.040 |
twins | +0.039 |
Neg logit contributions | |
shortcut | -0.060 |
lot | -0.058 |
duel | -0.056 |
further | -0.054 |
gall | -0.054 |
Token | . |
Token ID | 13 |
Feature activation | +7.40e-01 |
Pos logit contributions | |
cups | +0.130 |
ays | +0.121 |
frost | +0.121 |
izz | +0.120 |
ugs | +0.118 |
Neg logit contributions | |
shortcut | -0.131 |
lot | -0.131 |
gall | -0.122 |
duel | -0.121 |
further | -0.118 |
Token | They |
Token ID | 1119 |
Feature activation | +1.42e+00 |
Pos logit contributions | |
cups | +0.020 |
frost | +0.019 |
ugs | +0.018 |
sisters | +0.018 |
els | +0.018 |
Neg logit contributions | |
shortcut | -0.019 |
lot | -0.018 |
duel | -0.017 |
Skip | -0.017 |
farewell | -0.017 |
Token | did |
Token ID | 750 |
Feature activation | -3.60e-01 |
Pos logit contributions | |
cups | +0.037 |
ugs | +0.034 |
frost | +0.034 |
ays | +0.033 |
twins | +0.033 |
Neg logit contributions | |
shortcut | -0.037 |
lot | -0.036 |
duel | -0.035 |
gall | -0.034 |
farewell | -0.033 |
Token | not |
Token ID | 407 |
Feature activation | +7.24e-01 |
Pos logit contributions | |
shortcut | +0.011 |
lot | +0.010 |
Hop | +0.010 |
duel | +0.010 |
unnecessary | +0.010 |
Neg logit contributions | |
cups | -0.009 |
frost | -0.008 |
twins | -0.008 |
sisters | -0.007 |
cloth | -0.007 |
Token | know |
Token ID | 760 |
Feature activation | -2.39e-01 |
Pos logit contributions | |
cups | +0.023 |
frost | +0.021 |
ugs | +0.021 |
ays | +0.021 |
twins | +0.021 |
Neg logit contributions | |
shortcut | -0.015 |
lot | -0.015 |
duel | -0.014 |
gall | -0.013 |
unnecessary | -0.013 |
Token | would |
Token ID | 561 |
Feature activation | +1.95e+00 |
Pos logit contributions | |
cups | +0.090 |
ums | +0.088 |
ugs | +0.088 |
ays | +0.088 |
ions | +0.085 |
Neg logit contributions | |
lot | -0.033 |
shortcut | -0.032 |
'd | -0.032 |
gall | -0.031 |
duel | -0.031 |
Token | take |
Token ID | 1011 |
Feature activation | +1.69e+00 |
Pos logit contributions | |
cups | +0.046 |
ays | +0.045 |
ugs | +0.044 |
twins | +0.042 |
ums | +0.041 |
Neg logit contributions | |
shortcut | -0.053 |
duel | -0.051 |
gall | -0.050 |
lot | -0.050 |
farewell | -0.049 |
Token | out |
Token ID | 503 |
Feature activation | +1.18e+00 |
Pos logit contributions | |
cups | +0.044 |
ays | +0.040 |
sisters | +0.040 |
twins | +0.040 |
ugs | +0.039 |
Neg logit contributions | |
shortcut | -0.043 |
lot | -0.042 |
duel | -0.040 |
farewell | -0.040 |
unnecessary | -0.040 |
Token | her |
Token ID | 607 |
Feature activation | +4.25e+00 |
Pos logit contributions | |
ays | +0.033 |
ums | +0.032 |
ugs | +0.031 |
tub | +0.030 |
izz | +0.030 |
Neg logit contributions | |
farewell | -0.023 |
further | -0.023 |
closer | -0.023 |
higher | -0.022 |
shortcut | -0.022 |
Token | depend |
Token ID | 4745 |
Feature activation | +1.96e+00 |
Pos logit contributions | |
ays | +0.079 |
izz | +0.076 |
ums | +0.076 |
cloth | +0.072 |
ions | +0.071 |
Neg logit contributions | |
shortcut | -0.125 |
closer | -0.119 |
hike | -0.117 |
higher | -0.115 |
noble | -0.115 |
Token | able |
Token ID | 540 |
Feature activation | +5.16e+00 |
Pos logit contributions | |
cups | +0.060 |
sisters | +0.058 |
twins | +0.055 |
frost | +0.054 |
ums | +0.053 |
Neg logit contributions | |
shortcut | -0.038 |
further | -0.036 |
away | -0.036 |
lot | -0.035 |
unnecessary | -0.034 |
Token | oven |
Token ID | 14361 |
Feature activation | +3.16e+00 |
Pos logit contributions | |
izz | +0.104 |
ays | +0.104 |
ions | +0.092 |
ugs | +0.089 |
ums | +0.089 |
Neg logit contributions | |
shortcut | -0.155 |
farewell | -0.142 |
hike | -0.142 |
duel | -0.142 |
spear | -0.139 |
Token | m |
Token ID | 285 |
Feature activation | +2.98e+00 |
Pos logit contributions | |
izz | +0.113 |
cups | +0.110 |
ays | +0.108 |
sisters | +0.107 |
frost | +0.106 |
Neg logit contributions | |
lot | -0.044 |
shortcut | -0.043 |
higher | -0.037 |
hike | -0.037 |
gall | -0.036 |
Token | itt |
Token ID | 715 |
Feature activation | +6.63e-01 |
Pos logit contributions | |
cups | +0.060 |
folders | +0.054 |
sisters | +0.053 |
tub | +0.051 |
mug | +0.051 |
Neg logit contributions | |
lot | -0.083 |
to | -0.083 |
shortcut | -0.081 |
stru | -0.080 |
farewell | -0.080 |
Token | s |
Token ID | 82 |
Feature activation | +2.56e+00 |
Pos logit contributions | |
cups | +0.016 |
frost | +0.014 |
twins | +0.014 |
sisters | +0.014 |
ugs | +0.013 |
Neg logit contributions | |
shortcut | -0.019 |
lot | -0.019 |
duel | -0.018 |
gall | -0.017 |
unnecessary | -0.017 |
Token | and |
Token ID | 290 |
Feature activation | +1.16e+00 |
Pos logit contributions | |
cups | +0.069 |
frost | +0.064 |
ays | +0.064 |
sisters | +0.064 |
twins | +0.063 |
Neg logit contributions | |
shortcut | -0.061 |
lot | -0.058 |
duel | -0.054 |
further | -0.054 |
higher | -0.054 |
Token | it |
Token ID | 340 |
Feature activation | -5.32e-01 |
Pos logit contributions | |
cups | +0.004 |
frost | +0.004 |
twins | +0.004 |
ugs | +0.004 |
sisters | +0.004 |
Neg logit contributions | |
shortcut | -0.004 |
lot | -0.004 |
duel | -0.003 |
unnecessary | -0.003 |
farewell | -0.003 |
Token | away |
Token ID | 1497 |
Feature activation | +6.26e-01 |
Pos logit contributions | |
lot | +0.014 |
shortcut | +0.014 |
duel | +0.014 |
noble | +0.013 |
gall | +0.013 |
Neg logit contributions | |
cups | -0.014 |
frost | -0.013 |
cloth | -0.012 |
twins | -0.012 |
ugs | -0.012 |
Token | a |
Token ID | 257 |
Feature activation | +3.23e-01 |
Pos logit contributions | |
cups | +0.015 |
frost | +0.013 |
twins | +0.013 |
ugs | +0.013 |
sisters | +0.013 |
Neg logit contributions | |
shortcut | -0.018 |
lot | -0.018 |
duel | -0.017 |
unnecessary | -0.016 |
gall | -0.016 |
Token | long |
Token ID | 890 |
Feature activation | +2.67e+00 |
Pos logit contributions | |
cups | +0.008 |
frost | +0.007 |
mug | +0.007 |
twins | +0.007 |
sisters | +0.007 |
Neg logit contributions | |
lot | -0.009 |
shortcut | -0.009 |
duel | -0.008 |
farewell | -0.008 |
gall | -0.008 |
Token | time |
Token ID | 640 |
Feature activation | +2.28e+00 |
Pos logit contributions | |
cups | +0.085 |
frost | +0.080 |
ays | +0.080 |
ugs | +0.079 |
sisters | +0.079 |
Neg logit contributions | |
shortcut | -0.056 |
lot | -0.051 |
duel | -0.050 |
farewell | -0.047 |
further | -0.046 |
Token | ago |
Token ID | 2084 |
Feature activation | +5.16e+00 |
Pos logit contributions | |
cups | +0.051 |
frost | +0.047 |
ugs | +0.046 |
el | +0.045 |
ays | +0.045 |
Neg logit contributions | |
shortcut | -0.066 |
lot | -0.064 |
duel | -0.062 |
further | -0.061 |
higher | -0.060 |
Token | . |
Token ID | 13 |
Feature activation | +1.21e+00 |
Pos logit contributions | |
cups | +0.134 |
izz | +0.127 |
frost | +0.126 |
ays | +0.124 |
olk | +0.123 |
Neg logit contributions | |
shortcut | -0.124 |
lot | -0.124 |
duel | -0.113 |
gall | -0.112 |
farewell | -0.109 |
Token | It |
Token ID | 632 |
Feature activation | -9.77e-01 |
Pos logit contributions | |
cups | +0.034 |
frost | +0.030 |
ugs | +0.030 |
els | +0.029 |
sisters | +0.028 |
Neg logit contributions | |
lot | -0.029 |
shortcut | -0.029 |
Skip | -0.027 |
noble | -0.026 |
duel | -0.026 |
Token | is |
Token ID | 318 |
Feature activation | +2.44e+00 |
Pos logit contributions | |
shortcut | +0.027 |
lot | +0.026 |
duel | +0.025 |
unnecessary | +0.024 |
farewell | +0.024 |
Neg logit contributions | |
cups | -0.025 |
frost | -0.023 |
twins | -0.022 |
sisters | -0.022 |
ugs | -0.021 |
Token | not |
Token ID | 407 |
Feature activation | +7.46e-01 |
Pos logit contributions | |
cups | +0.073 |
frost | +0.068 |
ugs | +0.068 |
ays | +0.066 |
twins | +0.066 |
Neg logit contributions | |
shortcut | -0.056 |
lot | -0.053 |
duel | -0.051 |
unnecessary | -0.048 |
noble | -0.048 |
Token | good |
Token ID | 922 |
Feature activation | +2.14e+00 |
Pos logit contributions | |
cups | +0.021 |
frost | +0.020 |
twins | +0.019 |
ugs | +0.019 |
sisters | +0.019 |
Neg logit contributions | |
shortcut | -0.018 |
lot | -0.017 |
duel | -0.017 |
unnecessary | -0.016 |
farewell | -0.016 |
Token | they |
Token ID | 484 |
Feature activation | -1.32e+00 |
Pos logit contributions | |
shortcut | +0.039 |
lot | +0.037 |
duel | +0.036 |
gall | +0.034 |
farewell | +0.034 |
Neg logit contributions | |
cups | -0.031 |
twins | -0.027 |
sisters | -0.026 |
frost | -0.026 |
ugs | -0.026 |
Token | lived |
Token ID | 5615 |
Feature activation | +7.05e-02 |
Pos logit contributions | |
shortcut | +0.044 |
lot | +0.043 |
unnecessary | +0.042 |
farewell | +0.041 |
duel | +0.039 |
Neg logit contributions | |
cups | -0.027 |
frost | -0.025 |
sisters | -0.023 |
twins | -0.022 |
paints | -0.022 |
Token | a |
Token ID | 257 |
Feature activation | -1.74e+00 |
Pos logit contributions | |
cups | +0.002 |
twins | +0.001 |
frost | +0.001 |
sisters | +0.001 |
ions | +0.001 |
Neg logit contributions | |
shortcut | -0.002 |
unnecessary | -0.002 |
lot | -0.002 |
duel | -0.002 |
farewell | -0.002 |
Token | long |
Token ID | 890 |
Feature activation | +1.45e-01 |
Pos logit contributions | |
unnecessary | +0.046 |
lot | +0.046 |
shortcut | +0.045 |
nearer | +0.044 |
rated | +0.043 |
Neg logit contributions | |
cups | -0.047 |
twins | -0.043 |
sisters | -0.041 |
frost | -0.039 |
mug | -0.039 |
Token | time |
Token ID | 640 |
Feature activation | +1.41e+00 |
Pos logit contributions | |
cups | +0.004 |
frost | +0.003 |
twins | +0.003 |
sisters | +0.003 |
mug | +0.003 |
Neg logit contributions | |
shortcut | -0.004 |
unnecessary | -0.004 |
lot | -0.004 |
duel | -0.004 |
nearer | -0.004 |
Token | ago |
Token ID | 2084 |
Feature activation | +5.14e+00 |
Pos logit contributions | |
cups | +0.031 |
frost | +0.028 |
twins | +0.027 |
sisters | +0.026 |
ugs | +0.026 |
Neg logit contributions | |
shortcut | -0.044 |
lot | -0.043 |
duel | -0.041 |
unnecessary | -0.040 |
farewell | -0.040 |
Token | . |
Token ID | 13 |
Feature activation | +9.68e-01 |
Pos logit contributions | |
cups | +0.124 |
frost | +0.114 |
ays | +0.111 |
ugs | +0.111 |
izz | +0.109 |
Neg logit contributions | |
shortcut | -0.138 |
lot | -0.137 |
duel | -0.129 |
gall | -0.128 |
further | -0.124 |
Token | Lily |
Token ID | 20037 |
Feature activation | -7.47e-01 |
Pos logit contributions | |
cups | +0.027 |
frost | +0.026 |
els | +0.026 |
ugs | +0.025 |
el | +0.025 |
Neg logit contributions | |
lot | -0.021 |
shortcut | -0.021 |
farewell | -0.020 |
Hop | -0.019 |
Skip | -0.019 |
Token | listened |
Token ID | 16399 |
Feature activation | -1.07e+00 |
Pos logit contributions | |
shortcut | +0.021 |
lot | +0.020 |
duel | +0.020 |
unnecessary | +0.020 |
further | +0.019 |
Neg logit contributions | |
cups | -0.019 |
frost | -0.017 |
sisters | -0.017 |
twins | -0.017 |
el | -0.016 |
Token | carefully |
Token ID | 7773 |
Feature activation | -4.81e-01 |
Pos logit contributions | |
shortcut | +0.033 |
duel | +0.030 |
unnecessary | +0.030 |
lot | +0.030 |
gall | +0.028 |
Neg logit contributions | |
cups | -0.025 |
frost | -0.022 |
twins | -0.021 |
sisters | -0.021 |
folders | -0.020 |
Token | and |
Token ID | 290 |
Feature activation | +1.33e-01 |
Pos logit contributions | |
shortcut | +0.016 |
lot | +0.016 |
duel | +0.016 |
unnecessary | +0.015 |
Hop | +0.015 |
Neg logit contributions | |
cups | -0.009 |
frost | -0.008 |
twins | -0.008 |
sisters | -0.008 |
cloth | -0.007 |
Token | a |
Token ID | 257 |
Feature activation | +2.17e-02 |
Pos logit contributions | |
cups | +0.036 |
frost | +0.033 |
twins | +0.032 |
sisters | +0.031 |
folders | +0.030 |
Neg logit contributions | |
shortcut | -0.040 |
lot | -0.039 |
duel | -0.037 |
noble | -0.035 |
farewell | -0.035 |
Token | ship |
Token ID | 4074 |
Feature activation | +7.47e-01 |
Pos logit contributions | |
cups | +0.001 |
frost | +0.000 |
mug | +0.000 |
twins | +0.000 |
kettle | +0.000 |
Neg logit contributions | |
lot | -0.001 |
shortcut | -0.001 |
farewell | -0.001 |
unnecessary | -0.001 |
duel | -0.001 |
Token | that |
Token ID | 326 |
Feature activation | +1.42e-01 |
Pos logit contributions | |
cups | +0.015 |
frost | +0.013 |
twins | +0.012 |
cloth | +0.012 |
sisters | +0.012 |
Neg logit contributions | |
shortcut | -0.025 |
lot | -0.025 |
unnecessary | -0.024 |
duel | -0.024 |
noble | -0.023 |
Token | sank |
Token ID | 30895 |
Feature activation | +1.43e+00 |
Pos logit contributions | |
cups | +0.004 |
frost | +0.003 |
twins | +0.003 |
sisters | +0.003 |
ugs | +0.003 |
Neg logit contributions | |
shortcut | -0.004 |
lot | -0.004 |
duel | -0.004 |
unnecessary | -0.004 |
farewell | -0.003 |
Token | long |
Token ID | 890 |
Feature activation | +2.69e+00 |
Pos logit contributions | |
cups | +0.032 |
twins | +0.028 |
cloth | +0.027 |
frost | +0.027 |
sisters | +0.026 |
Neg logit contributions | |
shortcut | -0.044 |
lot | -0.044 |
duel | -0.043 |
noble | -0.042 |
Skip | -0.042 |
Token | ago |
Token ID | 2084 |
Feature activation | +5.12e+00 |
Pos logit contributions | |
cups | +0.059 |
frost | +0.053 |
ugs | +0.051 |
twins | +0.051 |
sisters | +0.051 |
Neg logit contributions | |
shortcut | -0.082 |
lot | -0.079 |
duel | -0.077 |
further | -0.074 |
gall | -0.074 |
Token | . |
Token ID | 13 |
Feature activation | +1.67e+00 |
Pos logit contributions | |
cups | +0.127 |
ays | +0.119 |
frost | +0.118 |
izz | +0.118 |
ugs | +0.115 |
Neg logit contributions | |
shortcut | -0.132 |
lot | -0.132 |
duel | -0.122 |
gall | -0.120 |
farewell | -0.119 |
Token | She |
Token ID | 1375 |
Feature activation | +8.10e-01 |
Pos logit contributions | |
cups | +0.042 |
els | +0.042 |
sisters | +0.041 |
el | +0.040 |
frost | +0.040 |
Neg logit contributions | |
shortcut | -0.038 |
lot | -0.038 |
farewell | -0.036 |
Skip | -0.036 |
higher | -0.035 |
Token | said |
Token ID | 531 |
Feature activation | -6.12e-01 |
Pos logit contributions | |
cups | +0.020 |
frost | +0.018 |
ugs | +0.017 |
twins | +0.017 |
sisters | +0.017 |
Neg logit contributions | |
shortcut | -0.023 |
lot | -0.023 |
duel | -0.022 |
gall | -0.021 |
noble | -0.021 |
Token | they |
Token ID | 484 |
Feature activation | -3.55e-01 |
Pos logit contributions | |
shortcut | +0.017 |
lot | +0.017 |
duel | +0.016 |
farewell | +0.016 |
unnecessary | +0.016 |
Neg logit contributions | |
cups | -0.015 |
frost | -0.014 |
ugs | -0.013 |
twins | -0.013 |
sisters | -0.013 |
Token | could |
Token ID | 714 |
Feature activation | +3.19e-02 |
Pos logit contributions | |
shortcut | +0.010 |
lot | +0.009 |
duel | +0.009 |
unnecessary | +0.009 |
farewell | +0.009 |
Neg logit contributions | |
cups | -0.009 |
frost | -0.008 |
twins | -0.008 |
ugs | -0.008 |
sisters | -0.008 |
Token | is |
Token ID | 318 |
Feature activation | +1.44e+00 |
Pos logit contributions | |
shortcut | +0.040 |
lot | +0.039 |
duel | +0.038 |
unnecessary | +0.036 |
farewell | +0.036 |
Neg logit contributions | |
cups | -0.035 |
frost | -0.032 |
twins | -0.030 |
sisters | -0.030 |
ugs | -0.029 |
Token | from |
Token ID | 422 |
Feature activation | +2.11e+00 |
Pos logit contributions | |
cups | +0.041 |
ugs | +0.039 |
frost | +0.038 |
ays | +0.037 |
sisters | +0.037 |
Neg logit contributions | |
shortcut | -0.034 |
lot | -0.033 |
duel | -0.031 |
noble | -0.030 |
gall | -0.029 |
Token | a |
Token ID | 257 |
Feature activation | -2.72e-01 |
Pos logit contributions | |
cups | +0.050 |
frost | +0.046 |
twins | +0.044 |
sisters | +0.043 |
ugs | +0.043 |
Neg logit contributions | |
shortcut | -0.062 |
lot | -0.060 |
duel | -0.058 |
farewell | -0.055 |
unnecessary | -0.055 |
Token | long |
Token ID | 890 |
Feature activation | +1.87e+00 |
Pos logit contributions | |
lot | +0.008 |
shortcut | +0.007 |
unnecessary | +0.007 |
farewell | +0.007 |
nearer | +0.007 |
Neg logit contributions | |
cups | -0.007 |
frost | -0.006 |
mug | -0.006 |
twins | -0.006 |
tub | -0.006 |
Token | time |
Token ID | 640 |
Feature activation | +2.02e+00 |
Pos logit contributions | |
cups | +0.052 |
frost | +0.047 |
ugs | +0.046 |
twins | +0.046 |
sisters | +0.046 |
Neg logit contributions | |
shortcut | -0.047 |
lot | -0.045 |
duel | -0.044 |
gall | -0.041 |
farewell | -0.041 |
Token | ago |
Token ID | 2084 |
Feature activation | +5.11e+00 |
Pos logit contributions | |
cups | +0.042 |
frost | +0.038 |
ugs | +0.037 |
sisters | +0.036 |
ays | +0.036 |
Neg logit contributions | |
shortcut | -0.062 |
lot | -0.060 |
duel | -0.058 |
further | -0.057 |
unnecessary | -0.056 |
Token | , |
Token ID | 11 |
Feature activation | +3.54e-01 |
Pos logit contributions | |
cups | +0.130 |
izz | +0.125 |
ays | +0.123 |
frost | +0.123 |
olk | +0.119 |
Neg logit contributions | |
lot | -0.122 |
shortcut | -0.122 |
gall | -0.111 |
duel | -0.111 |
farewell | -0.110 |
Token | when |
Token ID | 618 |
Feature activation | +1.24e+00 |
Pos logit contributions | |
cups | +0.010 |
ugs | +0.010 |
frost | +0.009 |
sisters | +0.009 |
twins | +0.009 |
Neg logit contributions | |
shortcut | -0.008 |
lot | -0.008 |
duel | -0.008 |
further | -0.007 |
higher | -0.007 |
Token | a |
Token ID | 257 |
Feature activation | -2.44e-01 |
Pos logit contributions | |
cups | +0.031 |
frost | +0.028 |
twins | +0.027 |
sisters | +0.027 |
ugs | +0.026 |
Neg logit contributions | |
shortcut | -0.035 |
lot | -0.034 |
duel | -0.032 |
farewell | -0.031 |
gall | -0.031 |
Token | king |
Token ID | 5822 |
Feature activation | +6.07e-01 |
Pos logit contributions | |
lot | +0.007 |
shortcut | +0.007 |
unnecessary | +0.006 |
duel | +0.006 |
gall | +0.006 |
Neg logit contributions | |
cups | -0.006 |
frost | -0.006 |
twins | -0.006 |
mug | -0.005 |
sisters | -0.005 |
Token | lived |
Token ID | 5615 |
Feature activation | +1.61e+00 |
Pos logit contributions | |
cups | +0.016 |
frost | +0.014 |
twins | +0.014 |
ugs | +0.014 |
sisters | +0.014 |
Neg logit contributions | |
shortcut | -0.016 |
lot | -0.016 |
duel | -0.015 |
unnecessary | -0.015 |
farewell | -0.014 |
Token | She |
Token ID | 1375 |
Feature activation | -1.98e-01 |
Pos logit contributions | |
els | +0.021 |
cups | +0.020 |
frost | +0.020 |
ays | +0.019 |
ries | +0.019 |
Neg logit contributions | |
shortcut | -0.032 |
higher | -0.031 |
lot | -0.031 |
Nice | -0.031 |
steep | -0.031 |
Token | had |
Token ID | 550 |
Feature activation | +2.09e+00 |
Pos logit contributions | |
lot | +0.004 |
shortcut | +0.004 |
gall | +0.004 |
duel | +0.004 |
noble | +0.003 |
Neg logit contributions | |
cups | -0.007 |
ays | -0.006 |
ugs | -0.006 |
frost | -0.006 |
els | -0.006 |
Token | a |
Token ID | 257 |
Feature activation | +1.31e+00 |
Pos logit contributions | |
ays | +0.067 |
ugs | +0.066 |
els | +0.066 |
cups | +0.065 |
cloth | +0.062 |
Neg logit contributions | |
lot | -0.040 |
shortcut | -0.038 |
gall | -0.038 |
unnecessary | -0.037 |
duel | -0.035 |
Token | brush |
Token ID | 14093 |
Feature activation | +2.02e+00 |
Pos logit contributions | |
cups | +0.036 |
frost | +0.034 |
ugs | +0.033 |
ays | +0.033 |
twins | +0.032 |
Neg logit contributions | |
shortcut | -0.033 |
lot | -0.031 |
duel | -0.030 |
noble | -0.029 |
gall | -0.029 |
Token | with |
Token ID | 351 |
Feature activation | +1.80e+00 |
Pos logit contributions | |
cups | +0.059 |
izz | +0.056 |
ugs | +0.055 |
ays | +0.055 |
frost | +0.055 |
Neg logit contributions | |
shortcut | -0.042 |
higher | -0.040 |
lot | -0.040 |
gall | -0.039 |
further | -0.038 |
Token | pink |
Token ID | 11398 |
Feature activation | +5.10e+00 |
Pos logit contributions | |
cups | +0.038 |
ugs | +0.038 |
ays | +0.037 |
twins | +0.036 |
kettle | +0.035 |
Neg logit contributions | |
shortcut | -0.049 |
gall | -0.047 |
unnecessary | -0.047 |
farewell | -0.047 |
lot | -0.046 |
Token | brist |
Token ID | 48197 |
Feature activation | +2.75e+00 |
Pos logit contributions | |
izz | +0.117 |
auc | +0.111 |
ays | +0.110 |
ugs | +0.103 |
twins | +0.101 |
Neg logit contributions | |
higher | -0.129 |
shortcut | -0.128 |
may | -0.124 |
gall | -0.122 |
closer | -0.121 |
Token | les |
Token ID | 829 |
Feature activation | +1.52e+00 |
Pos logit contributions | |
cups | +0.102 |
twins | +0.098 |
sisters | +0.097 |
kettle | +0.096 |
cloth | +0.091 |
Neg logit contributions | |
higher | -0.031 |
unnecessary | -0.028 |
lot | -0.028 |
further | -0.028 |
shortcut | -0.027 |
Token | that |
Token ID | 326 |
Feature activation | +1.25e+00 |
Pos logit contributions | |
cups | +0.041 |
izz | +0.039 |
ugs | +0.039 |
frost | +0.037 |
ays | +0.037 |
Neg logit contributions | |
shortcut | -0.034 |
higher | -0.033 |
gall | -0.032 |
further | -0.032 |
lot | -0.032 |
Token | she |
Token ID | 673 |
Feature activation | +8.78e-01 |
Pos logit contributions | |
cups | +0.030 |
ays | +0.027 |
twins | +0.027 |
ugs | +0.027 |
frost | +0.026 |
Neg logit contributions | |
shortcut | -0.034 |
lot | -0.032 |
higher | -0.032 |
duel | -0.031 |
gall | -0.031 |
Token | loved |
Token ID | 6151 |
Feature activation | +1.04e+00 |
Pos logit contributions | |
cups | +0.027 |
ays | +0.025 |
frost | +0.025 |
ugs | +0.025 |
twins | +0.025 |
Neg logit contributions | |
shortcut | -0.018 |
lot | -0.017 |
duel | -0.016 |
noble | -0.016 |
gall | -0.016 |
Token | were |
Token ID | 547 |
Feature activation | +1.39e+00 |
Pos logit contributions | |
shortcut | +0.014 |
lot | +0.014 |
duel | +0.014 |
farewell | +0.013 |
gall | +0.013 |
Neg logit contributions | |
cups | -0.011 |
twins | -0.010 |
frost | -0.010 |
sisters | -0.010 |
el | -0.010 |
Token | burying |
Token ID | 48360 |
Feature activation | +5.16e-01 |
Pos logit contributions | |
cups | +0.038 |
twins | +0.035 |
frost | +0.035 |
sisters | +0.034 |
ugs | +0.033 |
Neg logit contributions | |
shortcut | -0.036 |
lot | -0.035 |
duel | -0.033 |
farewell | -0.032 |
gall | -0.031 |
Token | something |
Token ID | 1223 |
Feature activation | -7.09e-01 |
Pos logit contributions | |
cups | +0.012 |
frost | +0.011 |
twins | +0.010 |
sisters | +0.010 |
el | +0.010 |
Neg logit contributions | |
shortcut | -0.015 |
lot | -0.015 |
duel | -0.015 |
gall | -0.014 |
farewell | -0.014 |
Token | from |
Token ID | 422 |
Feature activation | +1.38e+00 |
Pos logit contributions | |
lot | +0.019 |
shortcut | +0.019 |
duel | +0.018 |
farewell | +0.018 |
gall | +0.017 |
Neg logit contributions | |
cups | -0.019 |
frost | -0.017 |
twins | -0.016 |
sisters | -0.016 |
cloth | -0.016 |
Token | long |
Token ID | 890 |
Feature activation | +2.69e+00 |
Pos logit contributions | |
cups | +0.032 |
frost | +0.028 |
twins | +0.027 |
sisters | +0.026 |
el | +0.026 |
Neg logit contributions | |
lot | -0.043 |
shortcut | -0.042 |
duel | -0.040 |
farewell | -0.039 |
noble | -0.038 |
Token | ago |
Token ID | 2084 |
Feature activation | +5.07e+00 |
Pos logit contributions | |
cups | +0.060 |
frost | +0.056 |
ugs | +0.053 |
twins | +0.053 |
sisters | +0.052 |
Neg logit contributions | |
shortcut | -0.081 |
lot | -0.077 |
duel | -0.075 |
further | -0.073 |
unnecessary | -0.072 |
Token | that |
Token ID | 326 |
Feature activation | +4.85e-01 |
Pos logit contributions | |
cups | +0.122 |
frost | +0.114 |
ays | +0.111 |
ugs | +0.109 |
izz | +0.108 |
Neg logit contributions | |
shortcut | -0.138 |
lot | -0.135 |
duel | -0.127 |
gall | -0.125 |
further | -0.123 |
Token | made |
Token ID | 925 |
Feature activation | +1.19e+00 |
Pos logit contributions | |
cups | +0.012 |
frost | +0.010 |
twins | +0.010 |
cloth | +0.010 |
sisters | +0.010 |
Neg logit contributions | |
lot | -0.014 |
shortcut | -0.014 |
duel | -0.013 |
farewell | -0.013 |
unnecessary | -0.013 |
Token | them |
Token ID | 606 |
Feature activation | -2.32e+00 |
Pos logit contributions | |
cups | +0.029 |
frost | +0.026 |
twins | +0.026 |
sisters | +0.025 |
el | +0.025 |
Neg logit contributions | |
shortcut | -0.034 |
lot | -0.034 |
duel | -0.033 |
noble | -0.031 |
gall | -0.031 |
Token | feel |
Token ID | 1254 |
Feature activation | -8.15e-01 |
Pos logit contributions | |
shortcut | +0.058 |
lot | +0.057 |
duel | +0.054 |
noble | +0.051 |
gall | +0.051 |
Neg logit contributions | |
cups | -0.066 |
frost | -0.060 |
twins | -0.059 |
sisters | -0.058 |
cloth | -0.056 |
Token | sad |
Token ID | 6507 |
Feature activation | +4.14e-01 |
Pos logit contributions | |
lot | +0.020 |
shortcut | +0.020 |
duel | +0.019 |
gall | +0.017 |
hike | +0.017 |
Neg logit contributions | |
cups | -0.024 |
twins | -0.022 |
frost | -0.022 |
sisters | -0.022 |
cloth | -0.021 |
Token | people |
Token ID | 661 |
Feature activation | +5.70e-01 |
Pos logit contributions | |
cups | +0.007 |
frost | +0.007 |
twins | +0.006 |
sisters | +0.006 |
ugs | +0.006 |
Neg logit contributions | |
shortcut | -0.009 |
lot | -0.008 |
duel | -0.008 |
unnecessary | -0.008 |
farewell | -0.008 |
Token | did |
Token ID | 750 |
Feature activation | +1.97e+00 |
Pos logit contributions | |
cups | +0.013 |
frost | +0.012 |
twins | +0.012 |
sisters | +0.011 |
el | +0.011 |
Neg logit contributions | |
shortcut | -0.017 |
lot | -0.016 |
duel | -0.016 |
farewell | -0.016 |
unnecessary | -0.015 |
Token | a |
Token ID | 257 |
Feature activation | -2.49e-01 |
Pos logit contributions | |
cups | +0.051 |
frost | +0.046 |
ugs | +0.045 |
ays | +0.044 |
sisters | +0.044 |
Neg logit contributions | |
shortcut | -0.053 |
lot | -0.051 |
duel | -0.049 |
noble | -0.047 |
further | -0.047 |
Token | long |
Token ID | 890 |
Feature activation | +1.84e+00 |
Pos logit contributions | |
shortcut | +0.007 |
lot | +0.007 |
farther | +0.006 |
further | +0.006 |
gall | +0.006 |
Neg logit contributions | |
frost | -0.006 |
cups | -0.006 |
twins | -0.006 |
mug | -0.005 |
folders | -0.005 |
Token | time |
Token ID | 640 |
Feature activation | +2.30e+00 |
Pos logit contributions | |
cups | +0.051 |
frost | +0.047 |
twins | +0.045 |
sisters | +0.045 |
ugs | +0.045 |
Neg logit contributions | |
shortcut | -0.047 |
lot | -0.045 |
duel | -0.043 |
unnecessary | -0.041 |
noble | -0.041 |
Token | ago |
Token ID | 2084 |
Feature activation | +5.05e+00 |
Pos logit contributions | |
cups | +0.052 |
frost | +0.047 |
ugs | +0.046 |
ays | +0.045 |
izz | +0.044 |
Neg logit contributions | |
shortcut | -0.067 |
lot | -0.066 |
duel | -0.063 |
further | -0.061 |
gall | -0.061 |
Token | . |
Token ID | 13 |
Feature activation | +1.02e+00 |
Pos logit contributions | |
cups | +0.133 |
izz | +0.127 |
frost | +0.123 |
ays | +0.122 |
olk | +0.122 |
Neg logit contributions | |
lot | -0.122 |
shortcut | -0.120 |
duel | -0.111 |
gall | -0.109 |
higher | -0.106 |
Token | They |
Token ID | 1119 |
Feature activation | -2.21e-01 |
Pos logit contributions | |
cups | +0.025 |
frost | +0.023 |
ugs | +0.023 |
el | +0.022 |
els | +0.022 |
Neg logit contributions | |
shortcut | -0.028 |
lot | -0.027 |
duel | -0.026 |
noble | -0.025 |
Skip | -0.025 |
Token | lived |
Token ID | 5615 |
Feature activation | +1.18e+00 |
Pos logit contributions | |
shortcut | +0.006 |
lot | +0.006 |
duel | +0.006 |
unnecessary | +0.006 |
farewell | +0.006 |
Neg logit contributions | |
cups | -0.005 |
frost | -0.005 |
twins | -0.005 |
sisters | -0.005 |
ugs | -0.004 |
Token | in |
Token ID | 287 |
Feature activation | +2.00e+00 |
Pos logit contributions | |
cups | +0.029 |
twins | +0.026 |
frost | +0.026 |
sisters | +0.025 |
folders | +0.024 |
Neg logit contributions | |
shortcut | -0.035 |
lot | -0.033 |
duel | -0.032 |
unnecessary | -0.032 |
farewell | -0.030 |
Token | castles |
Token ID | 49203 |
Feature activation | +6.99e-01 |
Pos logit contributions | |
cups | +0.056 |
ugs | +0.056 |
ays | +0.054 |
izz | +0.053 |
frost | +0.052 |
Neg logit contributions | |
shortcut | -0.046 |
lot | -0.043 |
duel | -0.043 |
noble | -0.041 |
further | -0.041 |
Token | and |
Token ID | 290 |
Feature activation | -6.63e-01 |
Pos logit contributions | |
cups | +0.021 |
frost | +0.019 |
ugs | +0.019 |
twins | +0.018 |
sisters | +0.018 |
Neg logit contributions | |
shortcut | -0.025 |
lot | -0.024 |
duel | -0.023 |
unnecessary | -0.022 |
gall | -0.022 |
Token | plants |
Token ID | 6134 |
Feature activation | +1.00e+00 |
Pos logit contributions | |
shortcut | +0.019 |
lot | +0.019 |
duel | +0.018 |
unnecessary | +0.017 |
farewell | +0.017 |
Neg logit contributions | |
cups | -0.016 |
frost | -0.014 |
ugs | -0.014 |
twins | -0.014 |
sisters | -0.013 |
Token | that |
Token ID | 326 |
Feature activation | -1.61e-01 |
Pos logit contributions | |
cups | +0.025 |
ugs | +0.023 |
frost | +0.023 |
sisters | +0.022 |
izz | +0.022 |
Neg logit contributions | |
shortcut | -0.027 |
lot | -0.026 |
duel | -0.024 |
further | -0.024 |
higher | -0.024 |
Token | lived |
Token ID | 5615 |
Feature activation | +5.71e-01 |
Pos logit contributions | |
shortcut | +0.005 |
lot | +0.005 |
duel | +0.005 |
unnecessary | +0.004 |
farewell | +0.004 |
Neg logit contributions | |
cups | -0.004 |
frost | -0.003 |
twins | -0.003 |
sisters | -0.003 |
ugs | -0.003 |
Token | long |
Token ID | 890 |
Feature activation | +2.70e+00 |
Pos logit contributions | |
cups | +0.013 |
twins | +0.011 |
frost | +0.011 |
els | +0.011 |
folders | +0.010 |
Neg logit contributions | |
shortcut | -0.018 |
lot | -0.017 |
duel | -0.017 |
unnecessary | -0.016 |
farewell | -0.016 |
Token | ago |
Token ID | 2084 |
Feature activation | +5.02e+00 |
Pos logit contributions | |
cups | +0.057 |
frost | +0.051 |
ugs | +0.049 |
sisters | +0.048 |
twins | +0.048 |
Neg logit contributions | |
shortcut | -0.085 |
lot | -0.083 |
duel | -0.079 |
gall | -0.077 |
unnecessary | -0.076 |
Token | ," |
Token ID | 553 |
Feature activation | +1.52e+00 |
Pos logit contributions | |
cups | +0.124 |
izz | +0.121 |
ays | +0.118 |
frost | +0.116 |
ugs | +0.116 |
Neg logit contributions | |
lot | -0.125 |
shortcut | -0.124 |
duel | -0.113 |
gall | -0.113 |
further | -0.111 |
Token | she |
Token ID | 673 |
Feature activation | +1.81e+00 |
Pos logit contributions | |
ugs | +0.040 |
ays | +0.039 |
ase | +0.037 |
ump | +0.037 |
izz | +0.036 |
Neg logit contributions | |
closer | -0.041 |
shortcut | -0.037 |
spear | -0.037 |
further | -0.037 |
nearer | -0.036 |
Token | says |
Token ID | 1139 |
Feature activation | +6.46e-01 |
Pos logit contributions | |
cups | +0.041 |
ugs | +0.039 |
frost | +0.037 |
ions | +0.036 |
olk | +0.035 |
Neg logit contributions | |
shortcut | -0.054 |
lot | -0.053 |
duel | -0.050 |
gall | -0.049 |
noble | -0.049 |
Token | . |
Token ID | 13 |
Feature activation | +6.36e-01 |
Pos logit contributions | |
cups | +0.015 |
frost | +0.014 |
ugs | +0.013 |
ays | +0.013 |
twins | +0.013 |
Neg logit contributions | |
shortcut | -0.019 |
lot | -0.018 |
duel | -0.018 |
farewell | -0.017 |
gall | -0.017 |
Token |
|
Token ID | 198 |
Feature activation | +1.28e-01 |
Pos logit contributions | |
cups | +0.020 |
frost | +0.019 |
ugs | +0.019 |
els | +0.019 |
cloth | +0.019 |
Neg logit contributions | |
shortcut | -0.013 |
lot | -0.013 |
Skip | -0.012 |
duel | -0.011 |
Hop | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | +8.91e-01 |
Pos logit contributions | |
izz | +0.072 |
ays | +0.070 |
cups | +0.070 |
ugs | +0.069 |
ums | +0.068 |
Neg logit contributions | |
shortcut | -0.057 |
further | -0.052 |
lot | -0.051 |
higher | -0.049 |
spear | -0.049 |
Token | She |
Token ID | 1375 |
Feature activation | +1.45e+00 |
Pos logit contributions | |
cups | +0.022 |
frost | +0.020 |
els | +0.020 |
ugs | +0.020 |
sisters | +0.020 |
Neg logit contributions | |
lot | -0.023 |
shortcut | -0.023 |
Skip | -0.021 |
farewell | -0.021 |
unnecessary | -0.021 |
Token | said |
Token ID | 531 |
Feature activation | +9.83e-02 |
Pos logit contributions | |
cups | +0.041 |
ugs | +0.040 |
ays | +0.038 |
ums | +0.038 |
ions | +0.037 |
Neg logit contributions | |
shortcut | -0.033 |
lot | -0.032 |
gall | -0.031 |
duel | -0.030 |
'd | -0.030 |
Token | wine |
Token ID | 8237 |
Feature activation | +1.02e+00 |
Pos logit contributions | |
cups | +0.002 |
ugs | +0.002 |
ays | +0.002 |
frost | +0.002 |
el | +0.002 |
Neg logit contributions | |
shortcut | -0.003 |
lot | -0.002 |
farewell | -0.002 |
duel | -0.002 |
higher | -0.002 |
Token | was |
Token ID | 373 |
Feature activation | +3.31e+00 |
Pos logit contributions | |
cups | +0.028 |
frost | +0.026 |
ugs | +0.026 |
ays | +0.025 |
sisters | +0.025 |
Neg logit contributions | |
shortcut | -0.025 |
lot | -0.023 |
duel | -0.022 |
noble | -0.021 |
farewell | -0.021 |
Token | for |
Token ID | 329 |
Feature activation | +5.00e+00 |
Pos logit contributions | |
cups | +0.103 |
ugs | +0.101 |
ays | +0.099 |
frost | +0.095 |
ums | +0.095 |
Neg logit contributions | |
shortcut | -0.065 |
lot | -0.062 |
unnecessary | -0.059 |
noble | -0.058 |
duel | -0.058 |
Token | grown |
Token ID | 7334 |
Feature activation | +1.62e+00 |
Pos logit contributions | |
ays | +0.106 |
cups | +0.105 |
ugs | +0.102 |
frost | +0.098 |
cloth | +0.096 |
Neg logit contributions | |
shortcut | -0.137 |
unnecessary | -0.131 |
duel | -0.130 |
further | -0.128 |
lot | -0.127 |
Token | - |
Token ID | 12 |
Feature activation | +1.67e+00 |
Pos logit contributions | |
cups | +0.043 |
frost | +0.041 |
sisters | +0.039 |
twins | +0.039 |
ays | +0.039 |
Neg logit contributions | |
shortcut | -0.040 |
lot | -0.040 |
unnecessary | -0.037 |
further | -0.037 |
duel | -0.037 |
Token | ups |
Token ID | 4739 |
Feature activation | +9.17e-01 |
Pos logit contributions | |
cups | +0.040 |
frost | +0.037 |
sisters | +0.035 |
twins | +0.035 |
ugs | +0.034 |
Neg logit contributions | |
shortcut | -0.047 |
lot | -0.047 |
duel | -0.045 |
unnecessary | -0.044 |
further | -0.043 |
Token | . |
Token ID | 13 |
Feature activation | +1.26e+00 |
Pos logit contributions | |
cups | +0.022 |
frost | +0.020 |
ugs | +0.019 |
twins | +0.019 |
sisters | +0.019 |
Neg logit contributions | |
shortcut | -0.026 |
lot | -0.025 |
duel | -0.024 |
gall | -0.023 |
unnecessary | -0.023 |
Token | She |
Token ID | 1375 |
Feature activation | +1.79e+00 |
Pos logit contributions | |
cups | +0.029 |
sisters | +0.026 |
els | +0.026 |
frost | +0.026 |
ugs | +0.026 |
Neg logit contributions | |
lot | -0.034 |
shortcut | -0.033 |
Skip | -0.032 |
farewell | -0.031 |
Sir | -0.031 |
Token | to |
Token ID | 284 |
Feature activation | +7.71e-01 |
Pos logit contributions | |
cups | +0.013 |
frost | +0.012 |
twins | +0.011 |
sisters | +0.011 |
folders | +0.011 |
Neg logit contributions | |
shortcut | -0.013 |
lot | -0.013 |
duel | -0.012 |
gall | -0.012 |
noble | -0.012 |
Token | live |
Token ID | 2107 |
Feature activation | +2.87e-01 |
Pos logit contributions | |
cups | +0.020 |
frost | +0.019 |
twins | +0.018 |
sisters | +0.018 |
el | +0.018 |
Neg logit contributions | |
shortcut | -0.020 |
lot | -0.020 |
duel | -0.018 |
unnecessary | -0.018 |
Hop | -0.018 |
Token | a |
Token ID | 257 |
Feature activation | -5.85e-01 |
Pos logit contributions | |
cups | +0.006 |
twins | +0.006 |
frost | +0.006 |
folders | +0.006 |
sisters | +0.005 |
Neg logit contributions | |
shortcut | -0.009 |
lot | -0.009 |
duel | -0.008 |
gall | -0.008 |
farewell | -0.008 |
Token | long |
Token ID | 890 |
Feature activation | +1.63e+00 |
Pos logit contributions | |
lot | +0.016 |
shortcut | +0.016 |
unnecessary | +0.016 |
farewell | +0.015 |
Hop | +0.015 |
Neg logit contributions | |
cups | -0.013 |
twins | -0.013 |
frost | -0.012 |
sisters | -0.012 |
printer | -0.012 |
Token | time |
Token ID | 640 |
Feature activation | +1.84e+00 |
Pos logit contributions | |
cups | +0.043 |
frost | +0.040 |
twins | +0.038 |
sisters | +0.037 |
folders | +0.037 |
Neg logit contributions | |
shortcut | -0.043 |
lot | -0.041 |
unnecessary | -0.041 |
duel | -0.039 |
further | -0.039 |
Token | ago |
Token ID | 2084 |
Feature activation | +4.99e+00 |
Pos logit contributions | |
cups | +0.036 |
frost | +0.032 |
twins | +0.030 |
sisters | +0.030 |
ugs | +0.029 |
Neg logit contributions | |
shortcut | -0.062 |
lot | -0.060 |
duel | -0.059 |
unnecessary | -0.057 |
farewell | -0.056 |
Token | ." |
Token ID | 526 |
Feature activation | +2.81e-01 |
Pos logit contributions | |
cups | +0.125 |
frost | +0.115 |
izz | +0.113 |
ays | +0.112 |
ugs | +0.110 |
Neg logit contributions | |
shortcut | -0.130 |
lot | -0.130 |
duel | -0.121 |
gall | -0.119 |
farewell | -0.116 |
Token |
|
Token ID | 198 |
Feature activation | -2.16e-01 |
Pos logit contributions | |
cups | +0.009 |
ays | +0.009 |
els | +0.009 |
ugs | +0.008 |
frost | +0.008 |
Neg logit contributions | |
lot | -0.005 |
shortcut | -0.005 |
duel | -0.005 |
farewell | -0.005 |
Skip | -0.005 |
Token |
|
Token ID | 198 |
Feature activation | +6.69e-01 |
Pos logit contributions | |
shortcut | +0.006 |
lot | +0.005 |
duel | +0.005 |
gall | +0.005 |
farewell | +0.005 |
Neg logit contributions | |
cups | -0.006 |
frost | -0.005 |
sisters | -0.005 |
ugs | -0.005 |
twins | -0.005 |
Token | The |
Token ID | 464 |
Feature activation | -1.49e-01 |
Pos logit contributions | |
cups | +0.017 |
ays | +0.015 |
frost | +0.015 |
ugs | +0.015 |
sisters | +0.014 |
Neg logit contributions | |
shortcut | -0.019 |
lot | -0.019 |
duel | -0.018 |
noble | -0.017 |
Hop | -0.017 |
Token | girl |
Token ID | 2576 |
Feature activation | +7.25e-01 |
Pos logit contributions | |
shortcut | +0.005 |
lot | +0.005 |
unnecessary | +0.005 |
duel | +0.005 |
gall | +0.004 |
Neg logit contributions | |
cups | -0.003 |
twins | -0.003 |
sisters | -0.003 |
frost | -0.003 |
folders | -0.003 |
Token | so |
Token ID | 523 |
Feature activation | -9.56e-01 |
Pos logit contributions | |
cups | +0.003 |
twins | +0.003 |
frost | +0.003 |
sisters | +0.002 |
cloth | +0.002 |
Neg logit contributions | |
shortcut | -0.003 |
lot | -0.003 |
duel | -0.003 |
unnecessary | -0.003 |
hike | -0.003 |
Token | grateful |
Token ID | 14066 |
Feature activation | -9.94e-01 |
Pos logit contributions | |
shortcut | +0.027 |
lot | +0.026 |
unnecessary | +0.026 |
duel | +0.025 |
hike | +0.025 |
Neg logit contributions | |
cups | -0.023 |
twins | -0.023 |
frost | -0.021 |
sisters | -0.020 |
izz | -0.020 |
Token | , |
Token ID | 11 |
Feature activation | -2.21e+00 |
Pos logit contributions | |
shortcut | +0.033 |
duel | +0.032 |
lot | +0.031 |
unnecessary | +0.031 |
noble | +0.030 |
Neg logit contributions | |
cups | -0.020 |
twins | -0.018 |
frost | -0.017 |
el | -0.016 |
sisters | -0.016 |
Token | he |
Token ID | 339 |
Feature activation | -1.62e+00 |
Pos logit contributions | |
shortcut | +0.072 |
lot | +0.072 |
duel | +0.071 |
unnecessary | +0.070 |
quest | +0.066 |
Neg logit contributions | |
cups | -0.041 |
twins | -0.040 |
sisters | -0.034 |
frost | -0.034 |
el | -0.033 |
Token | gave |
Token ID | 2921 |
Feature activation | -2.43e+00 |
Pos logit contributions | |
unnecessary | +0.052 |
shortcut | +0.051 |
duel | +0.048 |
lot | +0.048 |
further | +0.046 |
Neg logit contributions | |
cups | -0.033 |
twins | -0.031 |
sisters | -0.031 |
frost | -0.030 |
el | -0.027 |
Token | Fin |
Token ID | 4463 |
Feature activation | -5.93e+00 |
Pos logit contributions | |
shortcut | +0.063 |
lot | +0.062 |
quest | +0.060 |
duel | +0.059 |
noble | +0.057 |
Neg logit contributions | |
el | -0.065 |
cups | -0.059 |
twins | -0.058 |
els | -0.054 |
frost | -0.053 |
Token | the |
Token ID | 262 |
Feature activation | -1.93e+00 |
Pos logit contributions | |
shortcut | +0.216 |
duel | +0.207 |
unnecessary | +0.199 |
steep | +0.196 |
rated | +0.194 |
Neg logit contributions | |
cups | -0.099 |
twins | -0.088 |
frost | -0.078 |
sisters | -0.077 |
cloth | -0.072 |
Token | shiny |
Token ID | 22441 |
Feature activation | -1.21e+00 |
Pos logit contributions | |
lot | +0.065 |
unnecessary | +0.061 |
shortcut | +0.061 |
higher | +0.059 |
farewell | +0.058 |
Neg logit contributions | |
cups | -0.040 |
twins | -0.037 |
sisters | -0.034 |
mug | -0.032 |
frost | -0.032 |
Token | strip |
Token ID | 10283 |
Feature activation | -2.12e+00 |
Pos logit contributions | |
lot | +0.035 |
shortcut | +0.034 |
unnecessary | +0.034 |
duel | +0.032 |
Skip | +0.031 |
Neg logit contributions | |
cups | -0.030 |
frost | -0.027 |
el | -0.026 |
mug | -0.025 |
twins | -0.025 |
Token | as |
Token ID | 355 |
Feature activation | -2.58e+00 |
Pos logit contributions | |
shortcut | +0.069 |
lot | +0.068 |
duel | +0.066 |
unnecessary | +0.065 |
noble | +0.063 |
Neg logit contributions | |
cups | -0.043 |
frost | -0.038 |
twins | -0.036 |
ugs | -0.036 |
cloth | -0.035 |
Token | a |
Token ID | 257 |
Feature activation | -2.00e+00 |
Pos logit contributions | |
shortcut | +0.105 |
rated | +0.103 |
duel | +0.102 |
itionally | +0.099 |
lot | +0.098 |
Neg logit contributions | |
cups | -0.029 |
twins | -0.028 |
ink | -0.023 |
ions | -0.022 |
frost | -0.021 |
Token | higher |
Token ID | 2440 |
Feature activation | -1.82e+00 |
Pos logit contributions | |
lot | +0.050 |
shortcut | +0.047 |
farewell | +0.045 |
quest | +0.041 |
noble | +0.040 |
Neg logit contributions | |
cups | -0.122 |
twins | -0.109 |
sisters | -0.109 |
pitcher | -0.107 |
girls | -0.107 |
Token | . |
Token ID | 13 |
Feature activation | +9.12e-02 |
Pos logit contributions | |
shortcut | +0.063 |
lot | +0.061 |
duel | +0.061 |
unnecessary | +0.058 |
noble | +0.058 |
Neg logit contributions | |
cups | -0.035 |
twins | -0.030 |
frost | -0.030 |
el | -0.029 |
ays | -0.028 |
Token | He |
Token ID | 679 |
Feature activation | -1.83e+00 |
Pos logit contributions | |
cups | +0.002 |
frost | +0.002 |
twins | +0.002 |
ugs | +0.002 |
sisters | +0.002 |
Neg logit contributions | |
shortcut | -0.003 |
lot | -0.003 |
duel | -0.003 |
unnecessary | -0.003 |
further | -0.003 |
Token | showed |
Token ID | 3751 |
Feature activation | -1.11e+00 |
Pos logit contributions | |
unnecessary | +0.062 |
shortcut | +0.062 |
lot | +0.059 |
farewell | +0.057 |
noble | +0.057 |
Neg logit contributions | |
cups | -0.032 |
twins | -0.028 |
sisters | -0.027 |
frost | -0.027 |
painted | -0.026 |
Token | T |
Token ID | 309 |
Feature activation | -7.26e-01 |
Pos logit contributions | |
shortcut | +0.034 |
lot | +0.032 |
duel | +0.031 |
noble | +0.030 |
unnecessary | +0.030 |
Neg logit contributions | |
cups | -0.026 |
frost | -0.024 |
el | -0.023 |
sisters | -0.023 |
twins | -0.023 |
Token | oni |
Token ID | 14651 |
Feature activation | -5.86e+00 |
Pos logit contributions | |
shortcut | +0.024 |
duel | +0.023 |
further | +0.023 |
unnecessary | +0.022 |
farther | +0.022 |
Neg logit contributions | |
cups | -0.014 |
frost | -0.014 |
ugs | -0.014 |
ums | -0.013 |
twins | -0.013 |
Token | how |
Token ID | 703 |
Feature activation | -2.69e+00 |
Pos logit contributions | |
shortcut | +0.189 |
rated | +0.189 |
lot | +0.187 |
unnecessary | +0.183 |
duel | +0.182 |
Neg logit contributions | |
cups | -0.113 |
twins | -0.104 |
sisters | -0.099 |
ions | -0.089 |
her | -0.088 |
Token | to |
Token ID | 284 |
Feature activation | -2.26e+00 |
Pos logit contributions | |
shortcut | +0.078 |
lot | +0.078 |
duel | +0.075 |
hike | +0.075 |
quest | +0.075 |
Neg logit contributions | |
cups | -0.054 |
twins | -0.050 |
brushes | -0.048 |
paints | -0.046 |
apy | -0.046 |
Token | lift |
Token ID | 10303 |
Feature activation | -2.38e+00 |
Pos logit contributions | |
lot | +0.061 |
unnecessary | +0.060 |
shortcut | +0.057 |
noble | +0.054 |
stru | +0.054 |
Neg logit contributions | |
cups | -0.053 |
shampoo | -0.048 |
frost | -0.048 |
el | -0.048 |
pitcher | -0.048 |
Token | up |
Token ID | 510 |
Feature activation | -2.79e+00 |
Pos logit contributions | |
shortcut | +0.067 |
lot | +0.065 |
duel | +0.063 |
noble | +0.060 |
farewell | +0.059 |
Neg logit contributions | |
cups | -0.062 |
frost | -0.056 |
el | -0.055 |
els | -0.053 |
twins | -0.052 |
Token | her |
Token ID | 607 |
Feature activation | -5.95e-01 |
Pos logit contributions | |
lot | +0.092 |
shortcut | +0.089 |
duel | +0.087 |
noble | +0.084 |
unnecessary | +0.082 |
Neg logit contributions | |
cups | -0.060 |
frost | -0.050 |
el | -0.050 |
els | -0.048 |
twins | -0.048 |
Token | ran |
Token ID | 4966 |
Feature activation | -2.56e+00 |
Pos logit contributions | |
unnecessary | +0.042 |
shortcut | +0.040 |
duel | +0.038 |
lot | +0.037 |
higher | +0.036 |
Neg logit contributions | |
cups | -0.026 |
sisters | -0.023 |
frost | -0.023 |
twins | -0.022 |
el | -0.021 |
Token | over |
Token ID | 625 |
Feature activation | -1.82e+00 |
Pos logit contributions | |
unnecessary | +0.077 |
duel | +0.073 |
lot | +0.072 |
noble | +0.071 |
shortcut | +0.070 |
Neg logit contributions | |
cups | -0.053 |
twins | -0.046 |
cloth | -0.042 |
frost | -0.042 |
her | -0.041 |
Token | and |
Token ID | 290 |
Feature activation | -1.59e+00 |
Pos logit contributions | |
lot | +0.062 |
unnecessary | +0.060 |
rated | +0.059 |
shortcut | +0.059 |
duel | +0.058 |
Neg logit contributions | |
cups | -0.038 |
ions | -0.030 |
twins | -0.030 |
frost | -0.029 |
el | -0.028 |
Token | gave |
Token ID | 2921 |
Feature activation | -2.37e+00 |
Pos logit contributions | |
unnecessary | +0.051 |
shortcut | +0.051 |
lot | +0.048 |
duel | +0.047 |
farewell | +0.046 |
Neg logit contributions | |
cups | -0.033 |
sisters | -0.027 |
twins | -0.026 |
frost | -0.025 |
girls | -0.025 |
Token | the |
Token ID | 262 |
Feature activation | -1.34e+00 |
Pos logit contributions | |
shortcut | +0.084 |
lot | +0.080 |
steep | +0.077 |
duel | +0.077 |
unnecessary | +0.075 |
Neg logit contributions | |
cups | -0.046 |
el | -0.042 |
frost | -0.041 |
twins | -0.038 |
sisters | -0.036 |
Token | rabbit |
Token ID | 22746 |
Feature activation | -5.86e+00 |
Pos logit contributions | |
lot | +0.039 |
shortcut | +0.037 |
unnecessary | +0.037 |
farewell | +0.037 |
duel | +0.036 |
Neg logit contributions | |
cups | -0.033 |
twins | -0.032 |
sisters | -0.032 |
mug | -0.030 |
ink | -0.027 |
Token | a |
Token ID | 257 |
Feature activation | -2.66e+00 |
Pos logit contributions | |
shortcut | +0.195 |
duel | +0.186 |
lot | +0.185 |
unnecessary | +0.184 |
steep | +0.170 |
Neg logit contributions | |
cups | -0.122 |
twins | -0.103 |
frost | -0.103 |
sisters | -0.101 |
mug | -0.094 |
Token | big |
Token ID | 1263 |
Feature activation | -1.77e+00 |
Pos logit contributions | |
lot | +0.076 |
unnecessary | +0.075 |
duel | +0.071 |
shortcut | +0.070 |
further | +0.070 |
Neg logit contributions | |
cups | -0.063 |
mug | -0.059 |
frost | -0.055 |
kettle | -0.054 |
cup | -0.054 |
Token | hug |
Token ID | 16225 |
Feature activation | -5.69e-01 |
Pos logit contributions | |
lot | +0.048 |
shortcut | +0.046 |
unnecessary | +0.045 |
duel | +0.044 |
Hop | +0.041 |
Neg logit contributions | |
cups | -0.047 |
frost | -0.043 |
mug | -0.042 |
twins | -0.040 |
sisters | -0.040 |
Token | . |
Token ID | 13 |
Feature activation | -1.57e+00 |
Pos logit contributions | |
shortcut | +0.019 |
lot | +0.018 |
duel | +0.018 |
unnecessary | +0.018 |
noble | +0.018 |
Neg logit contributions | |
cups | -0.011 |
frost | -0.010 |
twins | -0.009 |
sisters | -0.009 |
ugs | -0.009 |
Token | But |
Token ID | 887 |
Feature activation | -4.38e-01 |
Pos logit contributions | |
shortcut | +0.053 |
unnecessary | +0.051 |
duel | +0.050 |
lot | +0.048 |
further | +0.048 |
Neg logit contributions | |
cups | -0.032 |
twins | -0.028 |
frost | -0.027 |
ugs | -0.026 |
ays | -0.026 |
Token | ran |
Token ID | 4966 |
Feature activation | -2.74e+00 |
Pos logit contributions | |
unnecessary | +0.051 |
shortcut | +0.048 |
lot | +0.044 |
duel | +0.044 |
higher | +0.044 |
Neg logit contributions | |
cups | -0.026 |
sisters | -0.023 |
twins | -0.022 |
frost | -0.022 |
painted | -0.022 |
Token | over |
Token ID | 625 |
Feature activation | -1.90e+00 |
Pos logit contributions | |
unnecessary | +0.087 |
duel | +0.082 |
lot | +0.082 |
noble | +0.080 |
shortcut | +0.079 |
Neg logit contributions | |
cups | -0.051 |
twins | -0.044 |
her | -0.039 |
cloth | -0.038 |
frost | -0.038 |
Token | and |
Token ID | 290 |
Feature activation | -1.65e+00 |
Pos logit contributions | |
lot | +0.066 |
unnecessary | +0.064 |
rated | +0.064 |
duel | +0.062 |
shortcut | +0.062 |
Neg logit contributions | |
cups | -0.039 |
ions | -0.031 |
twins | -0.030 |
paints | -0.029 |
el | -0.028 |
Token | gave |
Token ID | 2921 |
Feature activation | -2.65e+00 |
Pos logit contributions | |
unnecessary | +0.056 |
shortcut | +0.055 |
lot | +0.052 |
farewell | +0.052 |
duel | +0.049 |
Neg logit contributions | |
cups | -0.033 |
sisters | -0.026 |
twins | -0.025 |
kettle | -0.024 |
frost | -0.023 |
Token | the |
Token ID | 262 |
Feature activation | -1.68e+00 |
Pos logit contributions | |
shortcut | +0.100 |
lot | +0.095 |
steep | +0.094 |
rig | +0.093 |
duel | +0.092 |
Neg logit contributions | |
cups | -0.045 |
el | -0.043 |
frost | -0.040 |
twins | -0.037 |
ries | -0.035 |
Token | rabbit |
Token ID | 22746 |
Feature activation | -5.80e+00 |
Pos logit contributions | |
lot | +0.048 |
shortcut | +0.046 |
farewell | +0.046 |
unnecessary | +0.045 |
duel | +0.044 |
Neg logit contributions | |
cups | -0.042 |
twins | -0.041 |
sisters | -0.041 |
mug | -0.038 |
girls | -0.035 |
Token | a |
Token ID | 257 |
Feature activation | -2.53e+00 |
Pos logit contributions | |
shortcut | +0.194 |
lot | +0.184 |
duel | +0.184 |
unnecessary | +0.182 |
steep | +0.169 |
Neg logit contributions | |
cups | -0.123 |
frost | -0.104 |
twins | -0.102 |
sisters | -0.101 |
mug | -0.094 |
Token | big |
Token ID | 1263 |
Feature activation | -1.77e+00 |
Pos logit contributions | |
lot | +0.073 |
unnecessary | +0.071 |
rated | +0.067 |
further | +0.066 |
shortcut | +0.066 |
Neg logit contributions | |
cups | -0.061 |
mug | -0.057 |
frost | -0.053 |
kettle | -0.052 |
cup | -0.051 |
Token | hug |
Token ID | 16225 |
Feature activation | -2.20e-01 |
Pos logit contributions | |
lot | +0.047 |
shortcut | +0.045 |
unnecessary | +0.043 |
duel | +0.043 |
noble | +0.040 |
Neg logit contributions | |
cups | -0.048 |
frost | -0.044 |
mug | -0.042 |
twins | -0.042 |
sisters | -0.041 |
Token | . |
Token ID | 13 |
Feature activation | -1.98e+00 |
Pos logit contributions | |
shortcut | +0.007 |
lot | +0.007 |
duel | +0.007 |
unnecessary | +0.007 |
noble | +0.007 |
Neg logit contributions | |
cups | -0.004 |
frost | -0.004 |
twins | -0.004 |
ugs | -0.004 |
sisters | -0.004 |
Token | But |
Token ID | 887 |
Feature activation | -5.69e-01 |
Pos logit contributions | |
shortcut | +0.063 |
unnecessary | +0.062 |
duel | +0.059 |
lot | +0.057 |
further | +0.056 |
Neg logit contributions | |
cups | -0.045 |
twins | -0.039 |
frost | -0.039 |
ugs | -0.036 |
sisters | -0.036 |
Token |
|
Token ID | 198 |
Feature activation | -2.14e+00 |
Pos logit contributions | |
shortcut | +0.046 |
lot | +0.043 |
duel | +0.043 |
unnecessary | +0.042 |
noble | +0.041 |
Neg logit contributions | |
cups | -0.036 |
twins | -0.031 |
frost | -0.031 |
ugs | -0.031 |
ays | -0.030 |
Token |
|
Token ID | 198 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
shortcut | +0.074 |
unnecessary | +0.073 |
duel | +0.072 |
lot | +0.070 |
noble | +0.069 |
Neg logit contributions | |
cups | -0.036 |
izz | -0.032 |
ugs | -0.032 |
twins | -0.032 |
Mom | -0.030 |
Token | The |
Token ID | 464 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
shortcut | +0.037 |
lot | +0.036 |
duel | +0.035 |
unnecessary | +0.034 |
farewell | +0.034 |
Neg logit contributions | |
cups | -0.031 |
frost | -0.027 |
twins | -0.027 |
ugs | -0.027 |
sisters | -0.026 |
Token | man |
Token ID | 582 |
Feature activation | -1.01e+00 |
Pos logit contributions | |
unnecessary | +0.033 |
shortcut | +0.032 |
lot | +0.032 |
duel | +0.031 |
Hop | +0.031 |
Neg logit contributions | |
twins | -0.021 |
cups | -0.020 |
frost | -0.018 |
sisters | -0.018 |
mug | -0.016 |
Token | showed |
Token ID | 3751 |
Feature activation | -2.23e+00 |
Pos logit contributions | |
unnecessary | +0.034 |
shortcut | +0.033 |
lot | +0.031 |
duel | +0.030 |
noble | +0.029 |
Neg logit contributions | |
cups | -0.021 |
frost | -0.019 |
twins | -0.018 |
sisters | -0.017 |
el | -0.017 |
Token | Sam |
Token ID | 3409 |
Feature activation | -5.69e+00 |
Pos logit contributions | |
shortcut | +0.081 |
lot | +0.078 |
noble | +0.076 |
duel | +0.076 |
unnecessary | +0.074 |
Neg logit contributions | |
cups | -0.038 |
el | -0.036 |
els | -0.035 |
ays | -0.031 |
twins | -0.031 |
Token | how |
Token ID | 703 |
Feature activation | -1.88e+00 |
Pos logit contributions | |
shortcut | +0.191 |
unnecessary | +0.183 |
duel | +0.179 |
rated | +0.174 |
lot | +0.171 |
Neg logit contributions | |
cups | -0.118 |
twins | -0.103 |
sisters | -0.097 |
frost | -0.097 |
el | -0.096 |
Token | to |
Token ID | 284 |
Feature activation | -1.73e+00 |
Pos logit contributions | |
shortcut | +0.058 |
lot | +0.056 |
rated | +0.055 |
duel | +0.054 |
Left | +0.053 |
Neg logit contributions | |
cups | -0.039 |
twins | -0.037 |
sisters | -0.034 |
folders | -0.033 |
paints | -0.032 |
Token | hold |
Token ID | 1745 |
Feature activation | -1.09e+00 |
Pos logit contributions | |
unnecessary | +0.039 |
lot | +0.038 |
shortcut | +0.035 |
farewell | +0.035 |
noble | +0.033 |
Neg logit contributions | |
cups | -0.048 |
el | -0.045 |
glue | -0.045 |
frost | -0.045 |
paint | -0.045 |
Token | the |
Token ID | 262 |
Feature activation | -4.99e-01 |
Pos logit contributions | |
shortcut | +0.032 |
lot | +0.032 |
duel | +0.031 |
noble | +0.030 |
gall | +0.029 |
Neg logit contributions | |
cups | -0.026 |
frost | -0.022 |
el | -0.022 |
twins | -0.022 |
els | -0.022 |
Token | gun |
Token ID | 2485 |
Feature activation | -2.27e+00 |
Pos logit contributions | |
lot | +0.011 |
gall | +0.010 |
shortcut | +0.010 |
farewell | +0.010 |
duel | +0.010 |
Neg logit contributions | |
cups | -0.015 |
twins | -0.015 |
frost | -0.014 |
sisters | -0.013 |
mug | -0.013 |
Token | Tim |
Token ID | 5045 |
Feature activation | -1.95e-01 |
Pos logit contributions | |
cups | +0.006 |
ugs | +0.006 |
frost | +0.006 |
ays | +0.006 |
ions | +0.005 |
Neg logit contributions | |
shortcut | -0.003 |
lot | -0.003 |
gall | -0.003 |
duel | -0.003 |
spear | -0.003 |
Token | my |
Token ID | 1820 |
Feature activation | +6.48e-01 |
Pos logit contributions | |
shortcut | +0.006 |
lot | +0.005 |
duel | +0.005 |
unnecessary | +0.005 |
noble | +0.005 |
Neg logit contributions | |
cups | -0.005 |
frost | -0.004 |
twins | -0.004 |
sisters | -0.004 |
ugs | -0.004 |
Token | said |
Token ID | 531 |
Feature activation | -1.18e+00 |
Pos logit contributions | |
cups | +0.014 |
frost | +0.012 |
ugs | +0.012 |
twins | +0.011 |
ays | +0.011 |
Neg logit contributions | |
shortcut | -0.021 |
lot | -0.020 |
duel | -0.019 |
gall | -0.019 |
farewell | -0.019 |
Token | , |
Token ID | 11 |
Feature activation | -1.25e+00 |
Pos logit contributions | |
shortcut | +0.037 |
lot | +0.036 |
duel | +0.035 |
unnecessary | +0.034 |
noble | +0.033 |
Neg logit contributions | |
cups | -0.026 |
frost | -0.023 |
twins | -0.022 |
sisters | -0.022 |
ugs | -0.022 |
Token | showing |
Token ID | 4478 |
Feature activation | -7.92e-01 |
Pos logit contributions | |
shortcut | +0.039 |
unnecessary | +0.038 |
lot | +0.036 |
duel | +0.035 |
further | +0.034 |
Neg logit contributions | |
cups | -0.029 |
frost | -0.026 |
twins | -0.026 |
sisters | -0.025 |
mug | -0.024 |
Token | Billy |
Token ID | 15890 |
Feature activation | -5.66e+00 |
Pos logit contributions | |
shortcut | +0.020 |
lot | +0.019 |
duel | +0.018 |
unnecessary | +0.017 |
noble | +0.017 |
Neg logit contributions | |
cups | -0.022 |
frost | -0.020 |
twins | -0.020 |
el | -0.020 |
sisters | -0.020 |
Token | the |
Token ID | 262 |
Feature activation | -1.06e+00 |
Pos logit contributions | |
shortcut | +0.160 |
lot | +0.146 |
duel | +0.146 |
unnecessary | +0.143 |
noble | +0.139 |
Neg logit contributions | |
cups | -0.147 |
twins | -0.134 |
frost | -0.132 |
sisters | -0.130 |
cloth | -0.125 |
Token | toy |
Token ID | 13373 |
Feature activation | -7.14e-01 |
Pos logit contributions | |
shortcut | +0.029 |
lot | +0.028 |
Hop | +0.027 |
unnecessary | +0.027 |
duel | +0.027 |
Neg logit contributions | |
cups | -0.025 |
twins | -0.023 |
frost | -0.023 |
sisters | -0.022 |
el | -0.022 |
Token | arrow |
Token ID | 15452 |
Feature activation | -2.31e+00 |
Pos logit contributions | |
shortcut | +0.018 |
lot | +0.018 |
duel | +0.017 |
unnecessary | +0.017 |
noble | +0.016 |
Neg logit contributions | |
cups | -0.019 |
frost | -0.018 |
twins | -0.018 |
sisters | -0.017 |
el | -0.017 |
Token | . |
Token ID | 13 |
Feature activation | -5.41e-01 |
Pos logit contributions | |
shortcut | +0.066 |
lot | +0.064 |
duel | +0.061 |
unnecessary | +0.061 |
noble | +0.059 |
Neg logit contributions | |
cups | -0.057 |
frost | -0.052 |
twins | -0.050 |
sisters | -0.049 |
el | -0.048 |
Token | " |
Token ID | 366 |
Feature activation | -1.43e+00 |
Pos logit contributions | |
shortcut | +0.017 |
unnecessary | +0.016 |
lot | +0.016 |
duel | +0.016 |
further | +0.015 |
Neg logit contributions | |
cups | -0.012 |
twins | -0.011 |
frost | -0.011 |
sisters | -0.010 |
mug | -0.010 |
Token | He |
Token ID | 679 |
Feature activation | -6.03e-01 |
Pos logit contributions | |
shortcut | +0.016 |
lot | +0.015 |
duel | +0.015 |
unnecessary | +0.014 |
noble | +0.014 |
Neg logit contributions | |
cups | -0.014 |
frost | -0.013 |
twins | -0.012 |
ugs | -0.012 |
sisters | -0.012 |
Token | wanted |
Token ID | 2227 |
Feature activation | -1.59e+00 |
Pos logit contributions | |
shortcut | +0.020 |
unnecessary | +0.019 |
lot | +0.018 |
duel | +0.018 |
farewell | +0.017 |
Neg logit contributions | |
cups | -0.013 |
frost | -0.011 |
sisters | -0.011 |
twins | -0.011 |
el | -0.010 |
Token | to |
Token ID | 284 |
Feature activation | -1.64e+00 |
Pos logit contributions | |
shortcut | +0.050 |
rated | +0.050 |
lot | +0.050 |
duel | +0.047 |
rig | +0.046 |
Neg logit contributions | |
cups | -0.036 |
frost | -0.033 |
twins | -0.030 |
sisters | -0.028 |
ink | -0.028 |
Token | teach |
Token ID | 4545 |
Feature activation | -1.24e+00 |
Pos logit contributions | |
shortcut | +0.038 |
lot | +0.038 |
unnecessary | +0.035 |
duel | +0.035 |
farewell | +0.034 |
Neg logit contributions | |
cups | -0.047 |
frost | -0.044 |
sisters | -0.042 |
twins | -0.042 |
el | -0.041 |
Token | the |
Token ID | 262 |
Feature activation | -2.38e+00 |
Pos logit contributions | |
lot | +0.039 |
shortcut | +0.039 |
duel | +0.036 |
noble | +0.035 |
rated | +0.035 |
Neg logit contributions | |
cups | -0.028 |
frost | -0.025 |
twins | -0.025 |
el | -0.025 |
sisters | -0.024 |
Token | squirrel |
Token ID | 33039 |
Feature activation | -5.66e+00 |
Pos logit contributions | |
lot | +0.063 |
duel | +0.060 |
unnecessary | +0.060 |
shortcut | +0.057 |
farewell | +0.057 |
Neg logit contributions | |
twins | -0.063 |
cups | -0.060 |
sisters | -0.057 |
girls | -0.054 |
mug | -0.053 |
Token | , |
Token ID | 11 |
Feature activation | -2.61e+00 |
Pos logit contributions | |
shortcut | +0.175 |
duel | +0.165 |
lot | +0.164 |
Left | +0.158 |
unnecessary | +0.157 |
Neg logit contributions | |
cups | -0.128 |
twins | -0.114 |
sisters | -0.110 |
folders | -0.105 |
frost | -0.104 |
Token | so |
Token ID | 523 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
lot | +0.074 |
shortcut | +0.074 |
duel | +0.070 |
unnecessary | +0.067 |
farewell | +0.067 |
Neg logit contributions | |
cups | -0.066 |
twins | -0.058 |
sisters | -0.057 |
frost | -0.057 |
folders | -0.054 |
Token | he |
Token ID | 339 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
shortcut | +0.021 |
lot | +0.019 |
duel | +0.019 |
unnecessary | +0.017 |
farewell | +0.017 |
Neg logit contributions | |
cups | -0.035 |
frost | -0.032 |
twins | -0.032 |
sisters | -0.031 |
ugs | -0.031 |
Token | placed |
Token ID | 4624 |
Feature activation | -8.25e-01 |
Pos logit contributions | |
unnecessary | +0.034 |
shortcut | +0.034 |
lot | +0.032 |
duel | +0.031 |
farewell | +0.030 |
Neg logit contributions | |
cups | -0.021 |
frost | -0.018 |
sisters | -0.018 |
twins | -0.017 |
el | -0.016 |
Token | the |
Token ID | 262 |
Feature activation | -9.33e-01 |
Pos logit contributions | |
shortcut | +0.027 |
lot | +0.027 |
duel | +0.026 |
noble | +0.025 |
unnecessary | +0.025 |
Neg logit contributions | |
cups | -0.017 |
frost | -0.016 |
el | -0.015 |
ions | -0.014 |
twins | -0.014 |
Token | and |
Token ID | 290 |
Feature activation | -2.04e+00 |
Pos logit contributions | |
shortcut | +0.065 |
unnecessary | +0.063 |
lot | +0.060 |
duel | +0.059 |
nearer | +0.057 |
Neg logit contributions | |
cups | -0.036 |
twins | -0.031 |
sisters | -0.031 |
frost | -0.029 |
mug | -0.027 |
Token | D |
Token ID | 360 |
Feature activation | -1.41e+00 |
Pos logit contributions | |
shortcut | +0.062 |
lot | +0.059 |
duel | +0.057 |
unnecessary | +0.055 |
noble | +0.054 |
Neg logit contributions | |
cups | -0.048 |
frost | -0.042 |
twins | -0.042 |
sisters | -0.042 |
el | -0.040 |
Token | ucky |
Token ID | 5309 |
Feature activation | -2.68e+00 |
Pos logit contributions | |
shortcut | +0.055 |
lot | +0.054 |
duel | +0.053 |
gall | +0.051 |
noble | +0.050 |
Neg logit contributions | |
cups | -0.021 |
ugs | -0.017 |
sisters | -0.017 |
el | -0.017 |
twins | -0.017 |
Token | taught |
Token ID | 7817 |
Feature activation | -2.36e-01 |
Pos logit contributions | |
shortcut | +0.080 |
lot | +0.076 |
unnecessary | +0.075 |
duel | +0.073 |
nearer | +0.071 |
Neg logit contributions | |
cups | -0.066 |
sisters | -0.060 |
twins | -0.059 |
frost | -0.057 |
ugs | -0.052 |
Token | P |
Token ID | 350 |
Feature activation | -1.83e+00 |
Pos logit contributions | |
shortcut | +0.007 |
lot | +0.006 |
duel | +0.006 |
unnecessary | +0.006 |
farewell | +0.006 |
Neg logit contributions | |
cups | -0.006 |
frost | -0.005 |
ugs | -0.005 |
twins | -0.005 |
sisters | -0.005 |
Token | andy |
Token ID | 10757 |
Feature activation | -5.64e+00 |
Pos logit contributions | |
shortcut | +0.064 |
lot | +0.062 |
duel | +0.061 |
gall | +0.059 |
farewell | +0.058 |
Neg logit contributions | |
cups | -0.035 |
ugs | -0.031 |
ums | -0.029 |
sisters | -0.029 |
el | -0.029 |
Token | how |
Token ID | 703 |
Feature activation | -1.59e+00 |
Pos logit contributions | |
shortcut | +0.173 |
duel | +0.161 |
lot | +0.159 |
unnecessary | +0.154 |
Left | +0.147 |
Neg logit contributions | |
cups | -0.137 |
twins | -0.125 |
sisters | -0.125 |
frost | -0.116 |
ugs | -0.109 |
Token | to |
Token ID | 284 |
Feature activation | -2.70e+00 |
Pos logit contributions | |
shortcut | +0.046 |
lot | +0.043 |
duel | +0.043 |
gall | +0.040 |
farewell | +0.039 |
Neg logit contributions | |
cups | -0.038 |
twins | -0.036 |
sisters | -0.035 |
ugs | -0.033 |
frost | -0.032 |
Token | qu |
Token ID | 627 |
Feature activation | -2.63e+00 |
Pos logit contributions | |
lot | +0.063 |
unnecessary | +0.060 |
shortcut | +0.058 |
farewell | +0.053 |
duel | +0.052 |
Neg logit contributions | |
cups | -0.079 |
el | -0.073 |
frost | -0.071 |
tub | -0.071 |
cloth | -0.070 |
Token | ack |
Token ID | 441 |
Feature activation | -2.60e+00 |
Pos logit contributions | |
shortcut | +0.074 |
lot | +0.071 |
duel | +0.068 |
farewell | +0.065 |
gall | +0.065 |
Neg logit contributions | |
cups | -0.068 |
ays | -0.061 |
ugs | -0.060 |
el | -0.060 |
frost | -0.060 |
Token | like |
Token ID | 588 |
Feature activation | -2.38e-01 |
Pos logit contributions | |
shortcut | +0.083 |
lot | +0.082 |
duel | +0.077 |
noble | +0.073 |
unnecessary | +0.073 |
Neg logit contributions | |
cups | -0.059 |
sisters | -0.052 |
twins | -0.051 |
frost | -0.050 |
ions | -0.048 |
Token | agged |
Token ID | 14655 |
Feature activation | -2.83e+00 |
Pos logit contributions | |
shortcut | +0.028 |
lot | +0.027 |
duel | +0.026 |
noble | +0.025 |
unnecessary | +0.025 |
Neg logit contributions | |
cups | -0.016 |
frost | -0.014 |
ugs | -0.014 |
twins | -0.013 |
sisters | -0.013 |
Token | his |
Token ID | 465 |
Feature activation | -1.96e-01 |
Pos logit contributions | |
shortcut | +0.077 |
unnecessary | +0.075 |
lot | +0.073 |
duel | +0.071 |
Hop | +0.071 |
Neg logit contributions | |
cups | -0.071 |
twins | -0.065 |
frost | -0.063 |
sisters | -0.063 |
cloth | -0.060 |
Token | tail |
Token ID | 7894 |
Feature activation | -1.29e+00 |
Pos logit contributions | |
shortcut | +0.005 |
lot | +0.005 |
duel | +0.005 |
unnecessary | +0.005 |
further | +0.005 |
Neg logit contributions | |
cups | -0.005 |
frost | -0.005 |
twins | -0.005 |
sisters | -0.005 |
ugs | -0.005 |
Token | , |
Token ID | 11 |
Feature activation | -2.04e+00 |
Pos logit contributions | |
shortcut | +0.049 |
unnecessary | +0.048 |
lot | +0.046 |
duel | +0.045 |
noble | +0.043 |
Neg logit contributions | |
cups | -0.021 |
twins | -0.019 |
frost | -0.018 |
sisters | -0.018 |
cloth | -0.017 |
Token | showing |
Token ID | 4478 |
Feature activation | -2.82e+00 |
Pos logit contributions | |
shortcut | +0.068 |
unnecessary | +0.065 |
lot | +0.064 |
duel | +0.061 |
farewell | +0.059 |
Neg logit contributions | |
cups | -0.046 |
frost | -0.039 |
twins | -0.039 |
sisters | -0.038 |
cloth | -0.036 |
Token | Tim |
Token ID | 5045 |
Feature activation | -5.63e+00 |
Pos logit contributions | |
lot | +0.090 |
steep | +0.089 |
shortcut | +0.088 |
rig | +0.088 |
hike | +0.086 |
Neg logit contributions | |
her | -0.049 |
el | -0.049 |
twins | -0.047 |
cups | -0.047 |
Mom | -0.047 |
Token | that |
Token ID | 326 |
Feature activation | -2.06e+00 |
Pos logit contributions | |
shortcut | +0.179 |
lot | +0.165 |
duel | +0.164 |
unnecessary | +0.162 |
noble | +0.160 |
Neg logit contributions | |
cups | -0.129 |
twins | -0.118 |
sisters | -0.109 |
frost | -0.107 |
cloth | -0.106 |
Token | he |
Token ID | 339 |
Feature activation | -1.52e+00 |
Pos logit contributions | |
lot | +0.054 |
shortcut | +0.053 |
unnecessary | +0.053 |
farewell | +0.053 |
rig | +0.050 |
Neg logit contributions | |
cups | -0.055 |
twins | -0.049 |
frost | -0.048 |
sisters | -0.048 |
mug | -0.046 |
Token | should |
Token ID | 815 |
Feature activation | -3.81e-01 |
Pos logit contributions | |
shortcut | +0.051 |
unnecessary | +0.049 |
farewell | +0.048 |
further | +0.047 |
higher | +0.047 |
Neg logit contributions | |
cups | -0.028 |
sisters | -0.027 |
twins | -0.027 |
frost | -0.026 |
el | -0.025 |
Token | try |
Token ID | 1949 |
Feature activation | -3.48e+00 |
Pos logit contributions | |
shortcut | +0.009 |
lot | +0.009 |
duel | +0.008 |
unnecessary | +0.008 |
further | +0.008 |
Neg logit contributions | |
cups | -0.011 |
frost | -0.010 |
twins | -0.010 |
sisters | -0.010 |
ugs | -0.009 |
Token | the |
Token ID | 262 |
Feature activation | -1.99e+00 |
Pos logit contributions | |
rated | +0.114 |
shortcut | +0.111 |
farewell | +0.110 |
Hop | +0.109 |
lot | +0.106 |
Neg logit contributions | |
cups | -0.067 |
twins | -0.061 |
ink | -0.059 |
els | -0.058 |
sisters | -0.057 |
Token | a |
Token ID | 257 |
Feature activation | -1.16e+00 |
Pos logit contributions | |
shortcut | +0.001 |
duel | +0.001 |
lot | +0.001 |
higher | +0.001 |
farewell | +0.001 |
Neg logit contributions | |
cups | -0.001 |
frost | -0.001 |
twins | -0.000 |
sisters | -0.000 |
ink | -0.000 |
Token | new |
Token ID | 649 |
Feature activation | -2.67e-02 |
Pos logit contributions | |
lot | +0.034 |
Left | +0.034 |
rated | +0.034 |
Hop | +0.032 |
unnecessary | +0.031 |
Neg logit contributions | |
cups | -0.024 |
mug | -0.023 |
twins | -0.022 |
cup | -0.022 |
sisters | -0.021 |
Token | friend |
Token ID | 1545 |
Feature activation | -3.86e-01 |
Pos logit contributions | |
lot | +0.001 |
shortcut | +0.001 |
duel | +0.001 |
unnecessary | +0.001 |
farewell | +0.001 |
Neg logit contributions | |
cups | -0.001 |
twins | -0.001 |
frost | -0.001 |
sisters | -0.001 |
mug | -0.001 |
Token | and |
Token ID | 290 |
Feature activation | -1.55e+00 |
Pos logit contributions | |
shortcut | +0.012 |
lot | +0.012 |
duel | +0.012 |
unnecessary | +0.011 |
farewell | +0.011 |
Neg logit contributions | |
cups | -0.008 |
twins | -0.007 |
frost | -0.007 |
sisters | -0.007 |
mug | -0.007 |
Token | showed |
Token ID | 3751 |
Feature activation | -2.28e+00 |
Pos logit contributions | |
shortcut | +0.044 |
lot | +0.043 |
duel | +0.041 |
unnecessary | +0.041 |
higher | +0.039 |
Neg logit contributions | |
cups | -0.038 |
sisters | -0.034 |
twins | -0.033 |
frost | -0.032 |
folders | -0.031 |
Token | Benny |
Token ID | 44275 |
Feature activation | -5.63e+00 |
Pos logit contributions | |
shortcut | +0.063 |
lot | +0.062 |
duel | +0.058 |
steep | +0.057 |
gall | +0.055 |
Neg logit contributions | |
cups | -0.058 |
el | -0.057 |
sisters | -0.053 |
twins | -0.053 |
ink | -0.051 |
Token | around |
Token ID | 1088 |
Feature activation | -1.23e+00 |
Pos logit contributions | |
shortcut | +0.165 |
lot | +0.154 |
unnecessary | +0.152 |
duel | +0.152 |
steep | +0.144 |
Neg logit contributions | |
cups | -0.138 |
twins | -0.123 |
sisters | -0.122 |
frost | -0.117 |
cloth | -0.117 |
Token | . |
Token ID | 13 |
Feature activation | -1.54e+00 |
Pos logit contributions | |
shortcut | +0.042 |
lot | +0.042 |
duel | +0.039 |
unnecessary | +0.038 |
farewell | +0.038 |
Neg logit contributions | |
cups | -0.025 |
twins | -0.022 |
sisters | -0.021 |
el | -0.021 |
mug | -0.020 |
Token |
|
Token ID | 198 |
Feature activation | -1.79e+00 |
Pos logit contributions | |
shortcut | +0.043 |
lot | +0.041 |
duel | +0.040 |
unnecessary | +0.039 |
gall | +0.038 |
Neg logit contributions | |
cups | -0.040 |
twins | -0.035 |
sisters | -0.035 |
frost | -0.034 |
ays | -0.033 |
Token |
|
Token ID | 198 |
Feature activation | -8.63e-01 |
Pos logit contributions | |
shortcut | +0.057 |
lot | +0.054 |
duel | +0.054 |
unnecessary | +0.053 |
farewell | +0.051 |
Neg logit contributions | |
cups | -0.037 |
twins | -0.033 |
frost | -0.032 |
ugs | -0.032 |
izz | -0.031 |
Token | As |
Token ID | 1722 |
Feature activation | -4.30e-01 |
Pos logit contributions | |
shortcut | +0.023 |
lot | +0.022 |
duel | +0.021 |
unnecessary | +0.020 |
farewell | +0.020 |
Neg logit contributions | |
cups | -0.023 |
frost | -0.021 |
twins | -0.021 |
sisters | -0.020 |
ugs | -0.020 |
Token | . |
Token ID | 13 |
Feature activation | -6.07e-01 |
Pos logit contributions | |
shortcut | +0.003 |
lot | +0.003 |
duel | +0.003 |
gall | +0.003 |
unnecessary | +0.003 |
Neg logit contributions | |
cups | -0.002 |
frost | -0.002 |
twins | -0.002 |
ugs | -0.002 |
sisters | -0.002 |
Token | Tim |
Token ID | 5045 |
Feature activation | -1.43e+00 |
Pos logit contributions | |
shortcut | +0.018 |
lot | +0.017 |
unnecessary | +0.016 |
duel | +0.016 |
noble | +0.015 |
Neg logit contributions | |
cups | -0.016 |
frost | -0.014 |
twins | -0.013 |
sisters | -0.013 |
mug | -0.013 |
Token | my |
Token ID | 1820 |
Feature activation | -4.64e-01 |
Pos logit contributions | |
shortcut | +0.050 |
duel | +0.047 |
unnecessary | +0.047 |
lot | +0.046 |
farewell | +0.046 |
Neg logit contributions | |
cups | -0.026 |
el | -0.024 |
twins | -0.024 |
frost | -0.024 |
sisters | -0.024 |
Token | told |
Token ID | 1297 |
Feature activation | -1.08e+00 |
Pos logit contributions | |
shortcut | +0.014 |
lot | +0.013 |
duel | +0.013 |
unnecessary | +0.013 |
farewell | +0.012 |
Neg logit contributions | |
cups | -0.011 |
frost | -0.010 |
twins | -0.010 |
sisters | -0.010 |
ugs | -0.009 |
Token | Nem |
Token ID | 22547 |
Feature activation | -4.46e+00 |
Pos logit contributions | |
shortcut | +0.029 |
lot | +0.028 |
duel | +0.027 |
unnecessary | +0.026 |
noble | +0.026 |
Neg logit contributions | |
cups | -0.029 |
frost | -0.027 |
twins | -0.026 |
sisters | -0.026 |
el | -0.025 |
Token | o |
Token ID | 78 |
Feature activation | -5.62e+00 |
Pos logit contributions | |
shortcut | +0.161 |
further | +0.156 |
unnecessary | +0.155 |
farewell | +0.152 |
duel | +0.151 |
Neg logit contributions | |
ugs | -0.073 |
cups | -0.071 |
twins | -0.066 |
ums | -0.063 |
mug | -0.058 |
Token | that |
Token ID | 326 |
Feature activation | -2.33e+00 |
Pos logit contributions | |
shortcut | +0.197 |
lot | +0.182 |
unnecessary | +0.181 |
duel | +0.180 |
Left | +0.176 |
Neg logit contributions | |
cups | -0.107 |
twins | -0.099 |
sisters | -0.093 |
frost | -0.084 |
folders | -0.082 |
Token | he |
Token ID | 339 |
Feature activation | -2.02e+00 |
Pos logit contributions | |
shortcut | +0.049 |
lot | +0.047 |
duel | +0.044 |
unnecessary | +0.043 |
farewell | +0.042 |
Neg logit contributions | |
cups | -0.076 |
frost | -0.070 |
twins | -0.069 |
sisters | -0.068 |
ugs | -0.066 |
Token | ate |
Token ID | 15063 |
Feature activation | -1.84e+00 |
Pos logit contributions | |
shortcut | +0.059 |
unnecessary | +0.056 |
farewell | +0.056 |
further | +0.055 |
duel | +0.055 |
Neg logit contributions | |
cups | -0.047 |
twins | -0.044 |
frost | -0.044 |
sisters | -0.042 |
el | -0.041 |
Token | a |
Token ID | 257 |
Feature activation | -2.79e+00 |
Pos logit contributions | |
lot | +0.062 |
shortcut | +0.061 |
duel | +0.059 |
unnecessary | +0.056 |
noble | +0.055 |
Neg logit contributions | |
cups | -0.041 |
frost | -0.038 |
twins | -0.033 |
icing | -0.032 |
el | -0.031 |
Token | lot |
Token ID | 1256 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
lot | +0.088 |
unnecessary | +0.084 |
shortcut | +0.080 |
Skip | +0.079 |
duel | +0.076 |
Neg logit contributions | |
cups | -0.069 |
frost | -0.064 |
pan | -0.057 |
mug | -0.057 |
cup | -0.057 |
Token | and |
Token ID | 290 |
Feature activation | -3.25e+00 |
Pos logit contributions | |
shortcut | +0.062 |
lot | +0.061 |
duel | +0.060 |
unnecessary | +0.059 |
noble | +0.057 |
Neg logit contributions | |
cups | -0.038 |
frost | -0.032 |
twins | -0.032 |
sisters | -0.031 |
ugs | -0.031 |
Token | hills |
Token ID | 18639 |
Feature activation | -2.36e+00 |
Pos logit contributions | |
lot | +0.094 |
shortcut | +0.092 |
duel | +0.086 |
farewell | +0.082 |
unnecessary | +0.082 |
Neg logit contributions | |
cups | -0.083 |
sisters | -0.075 |
twins | -0.074 |
frost | -0.068 |
girls | -0.065 |
Token | , |
Token ID | 11 |
Feature activation | -2.66e+00 |
Pos logit contributions | |
shortcut | +0.079 |
unnecessary | +0.079 |
lot | +0.077 |
duel | +0.076 |
noble | +0.074 |
Neg logit contributions | |
cups | -0.045 |
ions | -0.038 |
twins | -0.038 |
frost | -0.037 |
el | -0.036 |
Token | which |
Token ID | 543 |
Feature activation | -1.51e+00 |
Pos logit contributions | |
shortcut | +0.078 |
lot | +0.077 |
duel | +0.073 |
unnecessary | +0.073 |
farewell | +0.070 |
Neg logit contributions | |
cups | -0.065 |
twins | -0.056 |
sisters | -0.054 |
paints | -0.052 |
frost | -0.052 |
Token | made |
Token ID | 925 |
Feature activation | -9.21e-01 |
Pos logit contributions | |
lot | +0.040 |
shortcut | +0.040 |
duel | +0.038 |
unnecessary | +0.037 |
farewell | +0.037 |
Neg logit contributions | |
cups | -0.039 |
sisters | -0.035 |
twins | -0.035 |
frost | -0.035 |
el | -0.034 |
Token | Caroline |
Token ID | 29037 |
Feature activation | -5.62e+00 |
Pos logit contributions | |
shortcut | +0.030 |
lot | +0.030 |
duel | +0.028 |
noble | +0.027 |
unnecessary | +0.026 |
Neg logit contributions | |
cups | -0.020 |
twins | -0.017 |
sisters | -0.017 |
el | -0.017 |
frost | -0.017 |
Token | a |
Token ID | 257 |
Feature activation | -3.13e+00 |
Pos logit contributions | |
shortcut | +0.149 |
lot | +0.140 |
duel | +0.138 |
unnecessary | +0.133 |
Left | +0.126 |
Neg logit contributions | |
cups | -0.151 |
sisters | -0.139 |
twins | -0.136 |
frost | -0.134 |
el | -0.125 |
Token | bit |
Token ID | 1643 |
Feature activation | -1.42e+00 |
Pos logit contributions | |
lot | +0.088 |
shortcut | +0.082 |
duel | +0.078 |
Skip | +0.076 |
unnecessary | +0.075 |
Neg logit contributions | |
cups | -0.080 |
twins | -0.074 |
sisters | -0.072 |
frost | -0.069 |
mug | -0.068 |
Token | less |
Token ID | 1342 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
lot | +0.040 |
shortcut | +0.040 |
duel | +0.037 |
farewell | +0.034 |
gall | +0.034 |
Neg logit contributions | |
cups | -0.038 |
twins | -0.034 |
frost | -0.033 |
sisters | -0.033 |
cloth | -0.031 |
Token | uncomfortable |
Token ID | 12916 |
Feature activation | +1.95e-01 |
Pos logit contributions | |
shortcut | +0.034 |
lot | +0.033 |
duel | +0.032 |
farewell | +0.030 |
Hop | +0.029 |
Neg logit contributions | |
cups | -0.036 |
twins | -0.033 |
sisters | -0.032 |
frost | -0.032 |
cloth | -0.030 |
Token | . |
Token ID | 13 |
Feature activation | -1.75e-01 |
Pos logit contributions | |
cups | +0.004 |
frost | +0.003 |
twins | +0.003 |
sisters | +0.003 |
ugs | +0.003 |
Neg logit contributions | |
shortcut | -0.007 |
lot | -0.007 |
duel | -0.006 |
noble | -0.006 |
farewell | -0.006 |
Token | silly |
Token ID | 14397 |
Feature activation | -1.37e+00 |
Pos logit contributions | |
shortcut | +0.020 |
lot | +0.019 |
duel | +0.018 |
unnecessary | +0.017 |
hike | +0.017 |
Neg logit contributions | |
cups | -0.019 |
frost | -0.018 |
twins | -0.017 |
izz | -0.016 |
sisters | -0.016 |
Token | ! |
Token ID | 0 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
shortcut | +0.042 |
lot | +0.041 |
duel | +0.041 |
unnecessary | +0.039 |
farewell | +0.039 |
Neg logit contributions | |
cups | -0.031 |
twins | -0.027 |
frost | -0.027 |
sisters | -0.026 |
ugs | -0.026 |
Token | Then |
Token ID | 3244 |
Feature activation | -5.19e-01 |
Pos logit contributions | |
shortcut | +0.036 |
unnecessary | +0.034 |
duel | +0.034 |
lot | +0.033 |
further | +0.032 |
Neg logit contributions | |
cups | -0.031 |
twins | -0.029 |
frost | -0.028 |
ugs | -0.027 |
ays | -0.027 |
Token | Bob |
Token ID | 5811 |
Feature activation | -8.79e-02 |
Pos logit contributions | |
shortcut | +0.014 |
lot | +0.013 |
duel | +0.013 |
unnecessary | +0.012 |
noble | +0.012 |
Neg logit contributions | |
cups | -0.014 |
twins | -0.013 |
frost | -0.013 |
sisters | -0.013 |
el | -0.012 |
Token | showed |
Token ID | 3751 |
Feature activation | -2.34e+00 |
Pos logit contributions | |
shortcut | +0.003 |
lot | +0.003 |
unnecessary | +0.003 |
duel | +0.002 |
noble | +0.002 |
Neg logit contributions | |
cups | -0.002 |
frost | -0.002 |
sisters | -0.002 |
twins | -0.002 |
el | -0.002 |
Token | Jim |
Token ID | 5395 |
Feature activation | -5.62e+00 |
Pos logit contributions | |
shortcut | +0.067 |
unnecessary | +0.064 |
duel | +0.064 |
lot | +0.063 |
noble | +0.061 |
Neg logit contributions | |
cups | -0.053 |
el | -0.053 |
frost | -0.051 |
twins | -0.051 |
els | -0.048 |
Token | how |
Token ID | 703 |
Feature activation | -1.75e+00 |
Pos logit contributions | |
shortcut | +0.176 |
unnecessary | +0.163 |
duel | +0.162 |
lot | +0.160 |
steep | +0.150 |
Neg logit contributions | |
cups | -0.133 |
twins | -0.119 |
sisters | -0.118 |
frost | -0.117 |
el | -0.108 |
Token | to |
Token ID | 284 |
Feature activation | +2.54e-02 |
Pos logit contributions | |
shortcut | +0.050 |
lot | +0.049 |
duel | +0.047 |
Hop | +0.044 |
hike | +0.044 |
Neg logit contributions | |
cups | -0.043 |
twins | -0.039 |
sisters | -0.037 |
frost | -0.037 |
paints | -0.036 |
Token | make |
Token ID | 787 |
Feature activation | -1.71e+00 |
Pos logit contributions | |
cups | +0.001 |
frost | +0.001 |
twins | +0.001 |
sisters | +0.001 |
ugs | +0.001 |
Neg logit contributions | |
shortcut | -0.001 |
lot | -0.001 |
duel | -0.001 |
unnecessary | -0.001 |
noble | -0.001 |
Token | the |
Token ID | 262 |
Feature activation | -2.92e-01 |
Pos logit contributions | |
shortcut | +0.052 |
lot | +0.051 |
duel | +0.048 |
farewell | +0.047 |
Hop | +0.046 |
Neg logit contributions | |
cups | -0.041 |
frost | -0.037 |
twins | -0.036 |
sisters | -0.035 |
el | -0.034 |
Token | toy |
Token ID | 13373 |
Feature activation | -8.76e-01 |
Pos logit contributions | |
shortcut | +0.009 |
higher | +0.008 |
duel | +0.008 |
further | +0.008 |
lot | +0.008 |
Neg logit contributions | |
cups | -0.007 |
ugs | -0.006 |
ays | -0.006 |
izz | -0.006 |
frost | -0.006 |
Token | And |
Token ID | 1870 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
shortcut | +0.023 |
lot | +0.022 |
duel | +0.022 |
unnecessary | +0.021 |
farewell | +0.021 |
Neg logit contributions | |
cups | -0.021 |
frost | -0.019 |
twins | -0.018 |
ugs | -0.018 |
sisters | -0.018 |
Token | so |
Token ID | 523 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
shortcut | +0.041 |
lot | +0.040 |
duel | +0.040 |
unnecessary | +0.039 |
noble | +0.037 |
Neg logit contributions | |
cups | -0.027 |
twins | -0.024 |
sisters | -0.024 |
frost | -0.023 |
el | -0.022 |
Token | , |
Token ID | 11 |
Feature activation | -1.18e+00 |
Pos logit contributions | |
shortcut | +0.043 |
lot | +0.042 |
duel | +0.041 |
unnecessary | +0.039 |
farewell | +0.039 |
Neg logit contributions | |
cups | -0.028 |
twins | -0.025 |
sisters | -0.025 |
el | -0.024 |
frost | -0.024 |
Token | Andy |
Token ID | 12382 |
Feature activation | -8.04e-01 |
Pos logit contributions | |
lot | +0.033 |
shortcut | +0.033 |
duel | +0.032 |
unnecessary | +0.031 |
noble | +0.030 |
Neg logit contributions | |
cups | -0.028 |
twins | -0.026 |
sisters | -0.026 |
frost | -0.026 |
el | -0.025 |
Token | taught |
Token ID | 7817 |
Feature activation | -7.37e-01 |
Pos logit contributions | |
shortcut | +0.025 |
duel | +0.024 |
unnecessary | +0.024 |
lot | +0.024 |
noble | +0.022 |
Neg logit contributions | |
cups | -0.018 |
sisters | -0.016 |
frost | -0.016 |
twins | -0.015 |
el | -0.015 |
Token | Billy |
Token ID | 15890 |
Feature activation | -5.58e+00 |
Pos logit contributions | |
shortcut | +0.019 |
lot | +0.019 |
duel | +0.018 |
noble | +0.017 |
unnecessary | +0.017 |
Neg logit contributions | |
cups | -0.020 |
el | -0.019 |
frost | -0.019 |
twins | -0.018 |
sisters | -0.018 |
Token | how |
Token ID | 703 |
Feature activation | -1.39e+00 |
Pos logit contributions | |
shortcut | +0.173 |
duel | +0.160 |
lot | +0.157 |
unnecessary | +0.155 |
steep | +0.150 |
Neg logit contributions | |
cups | -0.128 |
twins | -0.117 |
sisters | -0.116 |
frost | -0.110 |
folders | -0.106 |
Token | to |
Token ID | 284 |
Feature activation | -7.66e-01 |
Pos logit contributions | |
shortcut | +0.043 |
duel | +0.041 |
lot | +0.040 |
farther | +0.040 |
farewell | +0.039 |
Neg logit contributions | |
cups | -0.028 |
twins | -0.028 |
sisters | -0.027 |
ugs | -0.025 |
ink | -0.024 |
Token | crawl |
Token ID | 27318 |
Feature activation | -2.57e+00 |
Pos logit contributions | |
lot | +0.019 |
shortcut | +0.018 |
duel | +0.017 |
unnecessary | +0.017 |
farewell | +0.017 |
Neg logit contributions | |
cups | -0.021 |
sisters | -0.019 |
el | -0.019 |
frost | -0.019 |
cloth | -0.019 |
Token | safely |
Token ID | 11512 |
Feature activation | -1.18e+00 |
Pos logit contributions | |
lot | +0.082 |
duel | +0.075 |
shortcut | +0.073 |
noble | +0.070 |
unnecessary | +0.069 |
Neg logit contributions | |
cups | -0.060 |
ions | -0.055 |
ink | -0.049 |
cloth | -0.049 |
folders | -0.049 |
Token | and |
Token ID | 290 |
Feature activation | -1.62e+00 |
Pos logit contributions | |
lot | +0.040 |
shortcut | +0.039 |
duel | +0.039 |
unnecessary | +0.038 |
noble | +0.037 |
Neg logit contributions | |
cups | -0.022 |
twins | -0.018 |
sisters | -0.018 |
cloth | -0.018 |
el | -0.018 |
Token |
|
Token ID | 198 |
Feature activation | -7.79e-01 |
Pos logit contributions | |
shortcut | +0.045 |
unnecessary | +0.043 |
duel | +0.043 |
lot | +0.042 |
farewell | +0.041 |
Neg logit contributions | |
cups | -0.025 |
twins | -0.023 |
frost | -0.023 |
izz | -0.023 |
el | -0.022 |
Token | The |
Token ID | 464 |
Feature activation | -1.56e+00 |
Pos logit contributions | |
shortcut | +0.023 |
unnecessary | +0.022 |
lot | +0.022 |
duel | +0.022 |
farewell | +0.021 |
Neg logit contributions | |
cups | -0.016 |
frost | -0.015 |
twins | -0.015 |
ugs | -0.014 |
el | -0.014 |
Token | little |
Token ID | 1310 |
Feature activation | -4.70e-01 |
Pos logit contributions | |
unnecessary | +0.053 |
shortcut | +0.052 |
duel | +0.051 |
lot | +0.051 |
rated | +0.049 |
Neg logit contributions | |
twins | -0.031 |
cups | -0.029 |
frost | -0.027 |
sisters | -0.025 |
pan | -0.022 |
Token | boy |
Token ID | 2933 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
shortcut | +0.014 |
lot | +0.013 |
duel | +0.013 |
unnecessary | +0.013 |
farewell | +0.013 |
Neg logit contributions | |
cups | -0.011 |
frost | -0.010 |
twins | -0.010 |
sisters | -0.010 |
el | -0.009 |
Token | told |
Token ID | 1297 |
Feature activation | -1.11e+00 |
Pos logit contributions | |
cups | +0.002 |
frost | +0.002 |
el | +0.002 |
twins | +0.002 |
sisters | +0.002 |
Neg logit contributions | |
shortcut | -0.004 |
unnecessary | -0.004 |
duel | -0.003 |
lot | -0.003 |
further | -0.003 |
Token | Bill |
Token ID | 3941 |
Feature activation | -5.56e+00 |
Pos logit contributions | |
shortcut | +0.030 |
lot | +0.029 |
duel | +0.029 |
unnecessary | +0.029 |
noble | +0.028 |
Neg logit contributions | |
cups | -0.026 |
el | -0.026 |
frost | -0.025 |
els | -0.025 |
twins | -0.024 |
Token | about |
Token ID | 546 |
Feature activation | -1.69e+00 |
Pos logit contributions | |
shortcut | +0.190 |
duel | +0.181 |
unnecessary | +0.175 |
Left | +0.174 |
farther | +0.165 |
Neg logit contributions | |
twins | -0.106 |
cups | -0.102 |
sisters | -0.100 |
frost | -0.090 |
els | -0.089 |
Token | the |
Token ID | 262 |
Feature activation | -6.78e-01 |
Pos logit contributions | |
lot | +0.055 |
shortcut | +0.055 |
duel | +0.052 |
unnecessary | +0.051 |
rated | +0.051 |
Neg logit contributions | |
cups | -0.035 |
frost | -0.032 |
twins | -0.031 |
sisters | -0.030 |
mug | -0.029 |
Token | big |
Token ID | 1263 |
Feature activation | -4.00e-02 |
Pos logit contributions | |
lot | +0.019 |
shortcut | +0.018 |
duel | +0.018 |
unnecessary | +0.017 |
farewell | +0.017 |
Neg logit contributions | |
cups | -0.018 |
twins | -0.016 |
frost | -0.016 |
mug | -0.015 |
tub | -0.015 |
Token | cooler |
Token ID | 19346 |
Feature activation | +1.48e-01 |
Pos logit contributions | |
shortcut | +0.001 |
lot | +0.001 |
duel | +0.001 |
further | +0.001 |
gall | +0.001 |
Neg logit contributions | |
cups | -0.001 |
frost | -0.001 |
ugs | -0.001 |
sisters | -0.001 |
ays | -0.001 |
Token | and |
Token ID | 290 |
Feature activation | -1.52e+00 |
Pos logit contributions | |
cups | +0.003 |
frost | +0.003 |
twins | +0.003 |
cloth | +0.003 |
sisters | +0.003 |
Neg logit contributions | |
shortcut | -0.005 |
unnecessary | -0.005 |
lot | -0.005 |
duel | -0.005 |
farewell | -0.004 |
Token |
|
Token ID | 198 |
Feature activation | -2.05e+00 |
Pos logit contributions | |
shortcut | +0.055 |
unnecessary | +0.051 |
duel | +0.051 |
lot | +0.050 |
gall | +0.048 |
Neg logit contributions | |
cups | -0.041 |
twins | -0.037 |
ays | -0.036 |
ums | -0.035 |
sisters | -0.035 |
Token |
|
Token ID | 198 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
shortcut | +0.068 |
unnecessary | +0.065 |
duel | +0.065 |
lot | +0.063 |
higher | +0.061 |
Neg logit contributions | |
cups | -0.038 |
twins | -0.035 |
izz | -0.034 |
ums | -0.033 |
ugs | -0.032 |
Token | The |
Token ID | 464 |
Feature activation | -1.96e+00 |
Pos logit contributions | |
shortcut | +0.029 |
lot | +0.027 |
duel | +0.027 |
unnecessary | +0.026 |
gall | +0.025 |
Neg logit contributions | |
cups | -0.025 |
twins | -0.022 |
frost | -0.022 |
sisters | -0.022 |
ugs | -0.022 |
Token | bear |
Token ID | 6842 |
Feature activation | -1.08e+00 |
Pos logit contributions | |
lot | +0.051 |
shortcut | +0.050 |
unnecessary | +0.048 |
duel | +0.047 |
to | +0.046 |
Neg logit contributions | |
twins | -0.055 |
cups | -0.053 |
sisters | -0.053 |
frost | -0.049 |
girls | -0.046 |
Token | told |
Token ID | 1297 |
Feature activation | -1.86e+00 |
Pos logit contributions | |
shortcut | +0.033 |
unnecessary | +0.032 |
lot | +0.031 |
duel | +0.031 |
further | +0.029 |
Neg logit contributions | |
cups | -0.024 |
twins | -0.022 |
frost | -0.022 |
sisters | -0.022 |
el | -0.021 |
Token | Tony |
Token ID | 8832 |
Feature activation | -5.55e+00 |
Pos logit contributions | |
shortcut | +0.056 |
lot | +0.052 |
duel | +0.051 |
unnecessary | +0.050 |
steep | +0.049 |
Neg logit contributions | |
cups | -0.043 |
el | -0.041 |
sisters | -0.040 |
twins | -0.039 |
ums | -0.038 |
Token | that |
Token ID | 326 |
Feature activation | -2.24e+00 |
Pos logit contributions | |
shortcut | +0.191 |
duel | +0.186 |
unnecessary | +0.174 |
Left | +0.173 |
rematch | +0.166 |
Neg logit contributions | |
cups | -0.095 |
twins | -0.093 |
sisters | -0.092 |
cloth | -0.083 |
ions | -0.075 |
Token | he |
Token ID | 339 |
Feature activation | -1.09e+00 |
Pos logit contributions | |
shortcut | +0.063 |
lot | +0.061 |
farewell | +0.058 |
unnecessary | +0.058 |
duel | +0.057 |
Neg logit contributions | |
cups | -0.058 |
twins | -0.054 |
sisters | -0.054 |
frost | -0.050 |
ions | -0.050 |
Token | needed |
Token ID | 2622 |
Feature activation | +2.38e-01 |
Pos logit contributions | |
shortcut | +0.034 |
duel | +0.032 |
unnecessary | +0.031 |
higher | +0.031 |
lot | +0.031 |
Neg logit contributions | |
cups | -0.022 |
twins | -0.022 |
sisters | -0.022 |
frost | -0.021 |
el | -0.021 |
Token | to |
Token ID | 284 |
Feature activation | +3.66e-01 |
Pos logit contributions | |
cups | +0.006 |
frost | +0.006 |
twins | +0.005 |
sisters | +0.005 |
ions | +0.005 |
Neg logit contributions | |
lot | -0.007 |
shortcut | -0.006 |
duel | -0.006 |
rated | -0.006 |
noble | -0.006 |
Token | find |
Token ID | 1064 |
Feature activation | -1.74e+00 |
Pos logit contributions | |
cups | +0.010 |
frost | +0.009 |
twins | +0.009 |
sisters | +0.009 |
cloth | +0.009 |
Neg logit contributions | |
lot | -0.009 |
shortcut | -0.009 |
duel | -0.008 |
unnecessary | -0.008 |
noble | -0.008 |
Token | . |
Token ID | 13 |
Feature activation | -1.69e+00 |
Pos logit contributions | |
shortcut | +0.089 |
lot | +0.085 |
unnecessary | +0.085 |
duel | +0.084 |
noble | +0.082 |
Neg logit contributions | |
cups | -0.031 |
twins | -0.029 |
sisters | -0.026 |
el | -0.023 |
frost | -0.022 |
Token | Nut |
Token ID | 11959 |
Feature activation | -3.70e+00 |
Pos logit contributions | |
shortcut | +0.051 |
duel | +0.049 |
lot | +0.049 |
unnecessary | +0.048 |
farewell | +0.047 |
Neg logit contributions | |
cups | -0.038 |
twins | -0.033 |
frost | -0.032 |
ugs | -0.031 |
sisters | -0.031 |
Token | ty |
Token ID | 774 |
Feature activation | -1.16e+00 |
Pos logit contributions | |
unnecessary | +0.101 |
shortcut | +0.099 |
farewell | +0.094 |
duel | +0.094 |
rematch | +0.088 |
Neg logit contributions | |
el | -0.086 |
twins | -0.085 |
izz | -0.082 |
cups | -0.082 |
els | -0.082 |
Token | showed |
Token ID | 3751 |
Feature activation | -2.63e+00 |
Pos logit contributions | |
shortcut | +0.044 |
unnecessary | +0.044 |
duel | +0.040 |
lot | +0.040 |
higher | +0.039 |
Neg logit contributions | |
cups | -0.019 |
sisters | -0.017 |
twins | -0.017 |
frost | -0.015 |
el | -0.015 |
Token | Bub |
Token ID | 19645 |
Feature activation | -3.19e+00 |
Pos logit contributions | |
shortcut | +0.068 |
lot | +0.066 |
duel | +0.064 |
steep | +0.062 |
noble | +0.061 |
Neg logit contributions | |
el | -0.070 |
cups | -0.067 |
els | -0.062 |
twins | -0.061 |
frost | -0.061 |
Token | bles |
Token ID | 7689 |
Feature activation | -5.53e+00 |
Pos logit contributions | |
farewell | +0.117 |
shortcut | +0.111 |
duel | +0.108 |
steep | +0.107 |
unnecessary | +0.107 |
Neg logit contributions | |
cups | -0.046 |
twins | -0.045 |
els | -0.045 |
ugs | -0.045 |
izz | -0.044 |
Token | a |
Token ID | 257 |
Feature activation | -2.24e+00 |
Pos logit contributions | |
shortcut | +0.173 |
lot | +0.170 |
unnecessary | +0.166 |
duel | +0.165 |
rated | +0.155 |
Neg logit contributions | |
cups | -0.124 |
twins | -0.111 |
sisters | -0.109 |
el | -0.105 |
frost | -0.104 |
Token | shiny |
Token ID | 22441 |
Feature activation | -4.80e-01 |
Pos logit contributions | |
unnecessary | +0.064 |
lot | +0.062 |
Left | +0.061 |
further | +0.060 |
farewell | +0.060 |
Neg logit contributions | |
mug | -0.050 |
cups | -0.048 |
tub | -0.043 |
cup | -0.043 |
twins | -0.043 |
Token | shield |
Token ID | 7614 |
Feature activation | -6.58e-01 |
Pos logit contributions | |
lot | +0.012 |
shortcut | +0.012 |
duel | +0.011 |
unnecessary | +0.011 |
noble | +0.011 |
Neg logit contributions | |
cups | -0.013 |
frost | -0.012 |
el | -0.012 |
mug | -0.012 |
twins | -0.011 |
Token | that |
Token ID | 326 |
Feature activation | -1.93e+00 |
Pos logit contributions | |
shortcut | +0.021 |
unnecessary | +0.020 |
lot | +0.020 |
duel | +0.020 |
noble | +0.019 |
Neg logit contributions | |
cups | -0.013 |
frost | -0.012 |
twins | -0.012 |
sisters | -0.011 |
el | -0.011 |
Token | he |
Token ID | 339 |
Feature activation | -1.64e+00 |
Pos logit contributions | |
unnecessary | +0.055 |
shortcut | +0.054 |
farewell | +0.052 |
lot | +0.051 |
duel | +0.050 |
Neg logit contributions | |
cups | -0.047 |
frost | -0.042 |
twins | -0.040 |
sisters | -0.040 |
el | -0.039 |
Token | K |
Token ID | 509 |
Feature activation | -9.07e-01 |
Pos logit contributions | |
shortcut | +0.025 |
lot | +0.024 |
duel | +0.023 |
unnecessary | +0.022 |
noble | +0.022 |
Neg logit contributions | |
cups | -0.022 |
twins | -0.020 |
sisters | -0.019 |
frost | -0.019 |
ions | -0.018 |
Token | iki |
Token ID | 5580 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
shortcut | +0.026 |
duel | +0.025 |
unnecessary | +0.024 |
farther | +0.024 |
further | +0.024 |
Neg logit contributions | |
el | -0.022 |
cups | -0.021 |
ugs | -0.021 |
frost | -0.020 |
els | -0.020 |
Token | would |
Token ID | 561 |
Feature activation | -1.27e+00 |
Pos logit contributions | |
shortcut | +0.039 |
unnecessary | +0.036 |
duel | +0.036 |
lot | +0.035 |
noble | +0.034 |
Neg logit contributions | |
cups | -0.030 |
sisters | -0.028 |
frost | -0.027 |
twins | -0.027 |
el | -0.027 |
Token | teach |
Token ID | 4545 |
Feature activation | -2.34e-01 |
Pos logit contributions | |
unnecessary | +0.034 |
lot | +0.034 |
shortcut | +0.032 |
Left | +0.030 |
duel | +0.030 |
Neg logit contributions | |
sisters | -0.027 |
cups | -0.027 |
frost | -0.027 |
paint | -0.027 |
el | -0.027 |
Token | T |
Token ID | 309 |
Feature activation | -1.66e+00 |
Pos logit contributions | |
shortcut | +0.008 |
lot | +0.008 |
duel | +0.008 |
noble | +0.008 |
unnecessary | +0.007 |
Neg logit contributions | |
cups | -0.004 |
frost | -0.004 |
el | -0.004 |
twins | -0.004 |
sisters | -0.004 |
Token | oto |
Token ID | 2069 |
Feature activation | -5.53e+00 |
Pos logit contributions | |
duel | +0.047 |
shortcut | +0.047 |
unnecessary | +0.044 |
farther | +0.044 |
lot | +0.043 |
Neg logit contributions | |
ugs | -0.039 |
el | -0.038 |
ums | -0.038 |
izz | -0.037 |
els | -0.036 |
Token | how |
Token ID | 703 |
Feature activation | -2.24e+00 |
Pos logit contributions | |
shortcut | +0.172 |
lot | +0.161 |
duel | +0.160 |
Left | +0.151 |
unnecessary | +0.151 |
Neg logit contributions | |
cups | -0.122 |
sisters | -0.115 |
twins | -0.113 |
folders | -0.103 |
frost | -0.102 |
Token | to |
Token ID | 284 |
Feature activation | -1.28e+00 |
Pos logit contributions | |
shortcut | +0.068 |
duel | +0.066 |
lot | +0.065 |
hike | +0.062 |
Left | +0.060 |
Neg logit contributions | |
cups | -0.045 |
twins | -0.044 |
sisters | -0.042 |
paints | -0.038 |
ugs | -0.038 |
Token | push |
Token ID | 4574 |
Feature activation | -1.53e+00 |
Pos logit contributions | |
lot | +0.031 |
shortcut | +0.031 |
unnecessary | +0.030 |
farewell | +0.028 |
noble | +0.028 |
Neg logit contributions | |
cups | -0.033 |
frost | -0.032 |
sisters | -0.031 |
el | -0.031 |
twins | -0.030 |
Token | his |
Token ID | 465 |
Feature activation | -7.46e-01 |
Pos logit contributions | |
shortcut | +0.042 |
lot | +0.041 |
duel | +0.039 |
noble | +0.038 |
farewell | +0.037 |
Neg logit contributions | |
cups | -0.040 |
frost | -0.036 |
el | -0.036 |
els | -0.035 |
twins | -0.035 |
Token | wings |
Token ID | 12098 |
Feature activation | -2.23e+00 |
Pos logit contributions | |
lot | +0.023 |
shortcut | +0.022 |
duel | +0.021 |
Hop | +0.020 |
noble | +0.020 |
Neg logit contributions | |
cups | -0.017 |
mug | -0.015 |
sisters | -0.015 |
frost | -0.014 |
twins | -0.014 |
Token | He |
Token ID | 679 |
Feature activation | -1.31e+00 |
Pos logit contributions | |
cups | +0.020 |
sisters | +0.019 |
frost | +0.019 |
ugs | +0.019 |
els | +0.019 |
Neg logit contributions | |
lot | -0.021 |
shortcut | -0.020 |
farewell | -0.019 |
duel | -0.019 |
higher | -0.019 |
Token | said |
Token ID | 531 |
Feature activation | -2.07e+00 |
Pos logit contributions | |
shortcut | +0.037 |
unnecessary | +0.035 |
lot | +0.033 |
duel | +0.033 |
further | +0.032 |
Neg logit contributions | |
cups | -0.033 |
frost | -0.031 |
twins | -0.031 |
sisters | -0.030 |
el | -0.028 |
Token | whist |
Token ID | 32262 |
Feature activation | -1.29e+00 |
Pos logit contributions | |
shortcut | +0.059 |
lot | +0.056 |
duel | +0.053 |
unnecessary | +0.052 |
rated | +0.050 |
Neg logit contributions | |
cups | -0.053 |
sisters | -0.048 |
frost | -0.048 |
twins | -0.048 |
folders | -0.044 |
Token | ling |
Token ID | 1359 |
Feature activation | -1.78e+00 |
Pos logit contributions | |
shortcut | +0.042 |
duel | +0.039 |
lot | +0.038 |
gall | +0.038 |
hike | +0.038 |
Neg logit contributions | |
cups | -0.027 |
twins | -0.026 |
ays | -0.025 |
frost | -0.025 |
sisters | -0.025 |
Token | made |
Token ID | 925 |
Feature activation | -4.70e-01 |
Pos logit contributions | |
shortcut | +0.061 |
duel | +0.057 |
lot | +0.057 |
unnecessary | +0.056 |
hike | +0.054 |
Neg logit contributions | |
cups | -0.034 |
twins | -0.031 |
frost | -0.030 |
sisters | -0.030 |
cloth | -0.029 |
Token | him |
Token ID | 683 |
Feature activation | -5.53e+00 |
Pos logit contributions | |
shortcut | +0.015 |
lot | +0.014 |
duel | +0.014 |
noble | +0.013 |
steep | +0.013 |
Neg logit contributions | |
cups | -0.010 |
frost | -0.010 |
twins | -0.010 |
sisters | -0.009 |
el | -0.009 |
Token | happy |
Token ID | 3772 |
Feature activation | -1.13e+00 |
Pos logit contributions | |
shortcut | +0.155 |
lot | +0.146 |
Hop | +0.143 |
Skip | +0.140 |
duel | +0.139 |
Neg logit contributions | |
cups | -0.128 |
twins | -0.125 |
sisters | -0.120 |
frost | -0.114 |
hair | -0.110 |
Token | and |
Token ID | 290 |
Feature activation | -2.15e+00 |
Pos logit contributions | |
shortcut | +0.042 |
lot | +0.040 |
unnecessary | +0.039 |
duel | +0.039 |
noble | +0.038 |
Neg logit contributions | |
cups | -0.019 |
twins | -0.018 |
sisters | -0.017 |
frost | -0.017 |
mug | -0.014 |
Token | calm |
Token ID | 9480 |
Feature activation | -8.70e-01 |
Pos logit contributions | |
shortcut | +0.058 |
lot | +0.053 |
duel | +0.052 |
hike | +0.051 |
unnecessary | +0.050 |
Neg logit contributions | |
cups | -0.056 |
twins | -0.054 |
sisters | -0.054 |
frost | -0.050 |
cloth | -0.048 |
Token | . |
Token ID | 13 |
Feature activation | -3.41e-01 |
Pos logit contributions | |
shortcut | +0.031 |
duel | +0.030 |
lot | +0.030 |
noble | +0.028 |
unnecessary | +0.027 |
Neg logit contributions | |
cups | -0.016 |
twins | -0.014 |
frost | -0.014 |
sisters | -0.013 |
el | -0.013 |
Token | L |
Token ID | 406 |
Feature activation | +2.13e-01 |
Pos logit contributions | |
lot | +0.007 |
shortcut | +0.007 |
farewell | +0.006 |
duel | +0.006 |
gall | +0.006 |
Neg logit contributions | |
cups | -0.010 |
els | -0.010 |
frost | -0.010 |
ries | -0.010 |
ugs | -0.009 |
Token | safe |
Token ID | 3338 |
Feature activation | -8.53e-02 |
Pos logit contributions | |
shortcut | +0.108 |
unnecessary | +0.106 |
lot | +0.106 |
duel | +0.105 |
Left | +0.101 |
Neg logit contributions | |
cups | -0.070 |
twins | -0.068 |
sisters | -0.065 |
frost | -0.060 |
el | -0.058 |
Token | ! |
Token ID | 0 |
Feature activation | -1.22e+00 |
Pos logit contributions | |
shortcut | +0.003 |
duel | +0.003 |
lot | +0.003 |
unnecessary | +0.003 |
noble | +0.003 |
Neg logit contributions | |
cups | -0.002 |
twins | -0.002 |
frost | -0.001 |
sisters | -0.001 |
cloth | -0.001 |
Token | He |
Token ID | 679 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
shortcut | +0.035 |
duel | +0.033 |
unnecessary | +0.033 |
lot | +0.032 |
further | +0.032 |
Neg logit contributions | |
cups | -0.029 |
twins | -0.027 |
ugs | -0.026 |
ays | -0.025 |
frost | -0.025 |
Token | gave |
Token ID | 2921 |
Feature activation | -2.09e+00 |
Pos logit contributions | |
unnecessary | +0.043 |
shortcut | +0.043 |
duel | +0.041 |
lot | +0.040 |
nearer | +0.038 |
Neg logit contributions | |
cups | -0.024 |
frost | -0.023 |
sisters | -0.021 |
twins | -0.021 |
el | -0.021 |
Token | the |
Token ID | 262 |
Feature activation | -1.45e+00 |
Pos logit contributions | |
shortcut | +0.074 |
lot | +0.070 |
steep | +0.069 |
duel | +0.069 |
rig | +0.068 |
Neg logit contributions | |
el | -0.040 |
cups | -0.037 |
frost | -0.034 |
twins | -0.034 |
sisters | -0.031 |
Token | knight |
Token ID | 22062 |
Feature activation | -5.50e+00 |
Pos logit contributions | |
unnecessary | +0.040 |
to | +0.039 |
rated | +0.039 |
lot | +0.039 |
higher | +0.039 |
Neg logit contributions | |
twins | -0.035 |
cups | -0.033 |
mug | -0.032 |
sisters | -0.032 |
folders | -0.029 |
Token | a |
Token ID | 257 |
Feature activation | -2.56e+00 |
Pos logit contributions | |
shortcut | +0.179 |
duel | +0.172 |
unnecessary | +0.170 |
lot | +0.161 |
steep | +0.161 |
Neg logit contributions | |
cups | -0.109 |
frost | -0.099 |
twins | -0.096 |
sisters | -0.095 |
folders | -0.087 |
Token | big |
Token ID | 1263 |
Feature activation | -2.17e+00 |
Pos logit contributions | |
lot | +0.072 |
unnecessary | +0.071 |
duel | +0.069 |
rated | +0.069 |
shortcut | +0.067 |
Neg logit contributions | |
cups | -0.059 |
mug | -0.056 |
frost | -0.054 |
cup | -0.051 |
twins | -0.050 |
Token | hug |
Token ID | 16225 |
Feature activation | -7.02e-01 |
Pos logit contributions | |
lot | +0.061 |
unnecessary | +0.057 |
shortcut | +0.057 |
duel | +0.056 |
Hop | +0.052 |
Neg logit contributions | |
cups | -0.055 |
frost | -0.051 |
mug | -0.050 |
twins | -0.047 |
tub | -0.047 |
Token | and |
Token ID | 290 |
Feature activation | -1.40e+00 |
Pos logit contributions | |
shortcut | +0.024 |
duel | +0.023 |
unnecessary | +0.023 |
lot | +0.023 |
noble | +0.022 |
Neg logit contributions | |
cups | -0.013 |
frost | -0.012 |
twins | -0.012 |
cloth | -0.011 |
mug | -0.011 |
Token | said |
Token ID | 531 |
Feature activation | -2.94e+00 |
Pos logit contributions | |
shortcut | +0.045 |
lot | +0.043 |
duel | +0.042 |
unnecessary | +0.041 |
noble | +0.039 |
Neg logit contributions | |
cups | -0.030 |
frost | -0.027 |
sisters | -0.026 |
twins | -0.026 |
mug | -0.024 |
Token | milk |
Token ID | 7545 |
Feature activation | +3.14e+00 |
Pos logit contributions | |
ays | +0.045 |
ugs | +0.045 |
cups | +0.044 |
izz | +0.043 |
ums | +0.043 |
Neg logit contributions | |
shortcut | -0.071 |
further | -0.067 |
higher | -0.067 |
farewell | -0.067 |
unnecessary | -0.066 |
Token | . |
Token ID | 13 |
Feature activation | +1.51e+00 |
Pos logit contributions | |
ums | +0.090 |
izz | +0.089 |
els | +0.089 |
folders | +0.089 |
sisters | +0.089 |
Neg logit contributions | |
shortcut | -0.054 |
closer | -0.053 |
higher | -0.052 |
lot | -0.051 |
gall | -0.050 |
Token |
|
Token ID | 198 |
Feature activation | +4.73e-01 |
Pos logit contributions | |
cups | +0.040 |
sisters | +0.038 |
frost | +0.038 |
el | +0.038 |
els | +0.037 |
Neg logit contributions | |
shortcut | -0.034 |
lot | -0.034 |
farewell | -0.032 |
Skip | -0.031 |
Left | -0.030 |
Token |
|
Token ID | 198 |
Feature activation | +2.47e+00 |
Pos logit contributions | |
cups | +0.012 |
frost | +0.011 |
twins | +0.011 |
sisters | +0.011 |
ugs | +0.011 |
Neg logit contributions | |
shortcut | -0.013 |
lot | -0.012 |
duel | -0.012 |
gall | -0.011 |
farewell | -0.011 |
Token | " |
Token ID | 1 |
Feature activation | +2.88e+00 |
Pos logit contributions | |
cups | +0.076 |
ays | +0.074 |
sisters | +0.073 |
frost | +0.073 |
el | +0.072 |
Neg logit contributions | |
shortcut | -0.051 |
Hop | -0.048 |
Left | -0.045 |
lot | -0.045 |
duel | -0.044 |
Token | Grand |
Token ID | 23581 |
Feature activation | +3.05e+00 |
Pos logit contributions | |
cups | +0.077 |
ays | +0.074 |
ugs | +0.074 |
ums | +0.073 |
el | +0.072 |
Neg logit contributions | |
Left | -0.070 |
shortcut | -0.070 |
lot | -0.068 |
Hop | -0.068 |
duel | -0.066 |
Token | ma |
Token ID | 2611 |
Feature activation | +1.15e-01 |
Pos logit contributions | |
cups | +0.065 |
pitcher | +0.057 |
olk | +0.056 |
ays | +0.056 |
kettle | +0.054 |
Neg logit contributions | |
lot | -0.081 |
farewell | -0.081 |
shortcut | -0.080 |
duel | -0.076 |
ph | -0.076 |
Token | , |
Token ID | 11 |
Feature activation | +5.98e-01 |
Pos logit contributions | |
cups | +0.003 |
frost | +0.003 |
twins | +0.002 |
ugs | +0.002 |
sisters | +0.002 |
Neg logit contributions | |
shortcut | -0.003 |
lot | -0.003 |
duel | -0.003 |
unnecessary | -0.003 |
farewell | -0.003 |
Token | you |
Token ID | 345 |
Feature activation | +2.50e-01 |
Pos logit contributions | |
cups | +0.017 |
ays | +0.017 |
ugs | +0.017 |
ums | +0.016 |
els | +0.016 |
Neg logit contributions | |
shortcut | -0.014 |
lot | -0.012 |
duel | -0.012 |
farewell | -0.012 |
noble | -0.012 |
Token | are |
Token ID | 389 |
Feature activation | +1.99e+00 |
Pos logit contributions | |
cups | +0.006 |
frost | +0.005 |
twins | +0.005 |
sisters | +0.005 |
cloth | +0.005 |
Neg logit contributions | |
shortcut | -0.008 |
lot | -0.007 |
duel | -0.007 |
unnecessary | -0.007 |
further | -0.007 |
Token | very |
Token ID | 845 |
Feature activation | +1.74e+00 |
Pos logit contributions | |
cups | +0.063 |
ugs | +0.061 |
ays | +0.060 |
el | +0.058 |
els | +0.057 |
Neg logit contributions | |
shortcut | -0.039 |
lot | -0.037 |
duel | -0.036 |
noble | -0.035 |
gall | -0.035 |
Token | day |
Token ID | 1110 |
Feature activation | +8.13e-01 |
Pos logit contributions | |
cups | +0.034 |
ays | +0.033 |
els | +0.033 |
frost | +0.032 |
folders | +0.031 |
Neg logit contributions | |
shortcut | -0.020 |
lot | -0.018 |
noble | -0.018 |
hike | -0.017 |
duel | -0.017 |
Token | , |
Token ID | 11 |
Feature activation | +8.58e-01 |
Pos logit contributions | |
cups | +0.023 |
frost | +0.022 |
folders | +0.022 |
twins | +0.021 |
sisters | +0.021 |
Neg logit contributions | |
shortcut | -0.018 |
farewell | -0.017 |
lot | -0.016 |
higher | -0.016 |
further | -0.016 |
Token | Mia |
Token ID | 32189 |
Feature activation | +1.84e+00 |
Pos logit contributions | |
ugs | +0.018 |
els | +0.018 |
frost | +0.018 |
ays | +0.018 |
folders | +0.018 |
Neg logit contributions | |
closer | -0.023 |
higher | -0.023 |
shortcut | -0.022 |
further | -0.022 |
farewell | -0.022 |
Token | 's |
Token ID | 338 |
Feature activation | +2.95e+00 |
Pos logit contributions | |
cups | +0.056 |
ugs | +0.055 |
ays | +0.055 |
ums | +0.055 |
ions | +0.055 |
Neg logit contributions | |
lot | -0.036 |
shortcut | -0.034 |
hike | -0.034 |
gall | -0.032 |
'd | -0.032 |
Token | mom |
Token ID | 1995 |
Feature activation | +1.85e+00 |
Pos logit contributions | |
ays | +0.064 |
ions | +0.062 |
els | +0.061 |
izz | +0.060 |
cups | +0.060 |
Neg logit contributions | |
shortcut | -0.087 |
duel | -0.078 |
lot | -0.078 |
farewell | -0.077 |
hike | -0.077 |
Token | was |
Token ID | 373 |
Feature activation | +3.34e+00 |
Pos logit contributions | |
cups | +0.050 |
ays | +0.048 |
ugs | +0.047 |
ums | +0.046 |
frost | +0.046 |
Neg logit contributions | |
lot | -0.044 |
shortcut | -0.043 |
'd | -0.041 |
gall | -0.041 |
duel | -0.041 |
Token | baking |
Token ID | 16871 |
Feature activation | +3.08e+00 |
Pos logit contributions | |
ays | +0.090 |
cups | +0.088 |
ugs | +0.086 |
olk | +0.085 |
els | +0.085 |
Neg logit contributions | |
shortcut | -0.076 |
higher | -0.075 |
further | -0.074 |
duel | -0.073 |
daring | -0.072 |
Token | a |
Token ID | 257 |
Feature activation | +1.36e+00 |
Pos logit contributions | |
ays | +0.085 |
ugs | +0.082 |
cups | +0.079 |
sisters | +0.079 |
el | +0.079 |
Neg logit contributions | |
higher | -0.069 |
further | -0.062 |
farther | -0.060 |
shortcut | -0.059 |
closer | -0.058 |
Token | cake |
Token ID | 12187 |
Feature activation | +3.19e+00 |
Pos logit contributions | |
cups | +0.027 |
ays | +0.026 |
ugs | +0.026 |
els | +0.025 |
el | +0.025 |
Neg logit contributions | |
shortcut | -0.042 |
duel | -0.040 |
lot | -0.039 |
noble | -0.038 |
higher | -0.038 |
Token | for |
Token ID | 329 |
Feature activation | +3.84e+00 |
Pos logit contributions | |
ays | +0.098 |
cups | +0.094 |
els | +0.094 |
izz | +0.094 |
el | +0.094 |
Neg logit contributions | |
higher | -0.056 |
shortcut | -0.053 |
hike | -0.053 |
gall | -0.050 |
further | -0.050 |
Token | Mia |
Token ID | 32189 |
Feature activation | +2.06e+00 |
Pos logit contributions | |
ays | +0.072 |
ugs | +0.069 |
cups | +0.067 |
ums | +0.065 |
Lem | +0.064 |
Neg logit contributions | |
further | -0.108 |
shortcut | -0.107 |
duel | -0.106 |
farewell | -0.105 |
gall | -0.105 |
Token | saw |
Token ID | 2497 |
Feature activation | +2.31e+00 |
Pos logit contributions | |
ugs | +0.082 |
ums | +0.081 |
ions | +0.081 |
cups | +0.081 |
ays | +0.080 |
Neg logit contributions | |
lot | -0.048 |
shortcut | -0.047 |
'd | -0.045 |
duel | -0.043 |
gall | -0.043 |
Token | Mom |
Token ID | 11254 |
Feature activation | +6.93e-01 |
Pos logit contributions | |
ugs | +0.077 |
ums | +0.073 |
ays | +0.072 |
aters | +0.070 |
cloth | +0.069 |
Neg logit contributions | |
further | -0.037 |
closer | -0.035 |
higher | -0.034 |
shortcut | -0.032 |
nearer | -0.032 |
Token | my |
Token ID | 1820 |
Feature activation | +3.74e-01 |
Pos logit contributions | |
cups | +0.020 |
ugs | +0.019 |
frost | +0.019 |
olk | +0.018 |
ays | +0.018 |
Neg logit contributions | |
shortcut | -0.014 |
lot | -0.014 |
gall | -0.014 |
higher | -0.014 |
further | -0.013 |
Token | carry |
Token ID | 3283 |
Feature activation | +2.62e+00 |
Pos logit contributions | |
cups | +0.009 |
frost | +0.008 |
ugs | +0.008 |
twins | +0.008 |
sisters | +0.008 |
Neg logit contributions | |
shortcut | -0.011 |
lot | -0.010 |
duel | -0.010 |
gall | -0.010 |
further | -0.010 |
Token | some |
Token ID | 617 |
Feature activation | +2.20e+00 |
Pos logit contributions | |
ays | +0.061 |
ums | +0.058 |
sisters | +0.058 |
cups | +0.057 |
ugs | +0.057 |
Neg logit contributions | |
further | -0.068 |
higher | -0.065 |
farther | -0.064 |
shortcut | -0.064 |
closer | -0.063 |
Token | dirty |
Token ID | 11841 |
Feature activation | +3.58e+00 |
Pos logit contributions | |
ays | +0.046 |
ums | +0.044 |
ugs | +0.043 |
izz | +0.043 |
cups | +0.043 |
Neg logit contributions | |
shortcut | -0.066 |
further | -0.061 |
farther | -0.059 |
unnecessary | -0.059 |
spear | -0.058 |
Token | plates |
Token ID | 13375 |
Feature activation | +2.82e+00 |
Pos logit contributions | |
izz | +0.070 |
ums | +0.070 |
ays | +0.069 |
ena | +0.067 |
illian | +0.066 |
Neg logit contributions | |
further | -0.100 |
farther | -0.093 |
shortcut | -0.092 |
higher | -0.088 |
closer | -0.088 |
Token | to |
Token ID | 284 |
Feature activation | +2.88e+00 |
Pos logit contributions | |
olk | +0.080 |
frost | +0.080 |
ums | +0.080 |
els | +0.079 |
sisters | +0.078 |
Neg logit contributions | |
further | -0.058 |
shortcut | -0.057 |
farther | -0.054 |
closer | -0.053 |
higher | -0.053 |
Token | the |
Token ID | 262 |
Feature activation | +2.47e+00 |
Pos logit contributions | |
ays | +0.082 |
ugs | +0.081 |
ums | +0.074 |
els | +0.072 |
twins | +0.071 |
Neg logit contributions | |
shortcut | -0.064 |
further | -0.063 |
duel | -0.060 |
closer | -0.060 |
hike | -0.060 |
Token | dish |
Token ID | 9433 |
Feature activation | +2.39e+00 |
Pos logit contributions | |
ugs | +0.039 |
ays | +0.038 |
els | +0.038 |
el | +0.037 |
izz | +0.037 |
Neg logit contributions | |
shortcut | -0.087 |
further | -0.081 |
duel | -0.081 |
lot | -0.080 |
farther | -0.079 |
Token | washer |
Token ID | 45146 |
Feature activation | +2.33e+00 |
Pos logit contributions | |
frost | +0.063 |
els | +0.062 |
olk | +0.061 |
icing | +0.060 |
cups | +0.060 |
Neg logit contributions | |
shortcut | -0.051 |
lot | -0.048 |
further | -0.047 |
hike | -0.045 |
ashore | -0.045 |
Token | trash |
Token ID | 13913 |
Feature activation | +1.57e+00 |
Pos logit contributions | |
cups | +0.019 |
ugs | +0.017 |
frost | +0.017 |
sisters | +0.017 |
twins | +0.016 |
Neg logit contributions | |
shortcut | -0.031 |
lot | -0.029 |
duel | -0.029 |
farewell | -0.028 |
further | -0.028 |
Token | . |
Token ID | 13 |
Feature activation | +1.26e+00 |
Pos logit contributions | |
ugs | +0.039 |
cups | +0.039 |
ums | +0.039 |
sisters | +0.038 |
folders | +0.037 |
Neg logit contributions | |
shortcut | -0.040 |
further | -0.037 |
higher | -0.036 |
farther | -0.036 |
lot | -0.035 |
Token | " |
Token ID | 366 |
Feature activation | +4.46e-01 |
Pos logit contributions | |
cups | +0.023 |
frost | +0.022 |
sisters | +0.021 |
ugs | +0.020 |
els | +0.020 |
Neg logit contributions | |
shortcut | -0.041 |
lot | -0.040 |
duel | -0.038 |
Skip | -0.038 |
higher | -0.038 |
Token | No |
Token ID | 2949 |
Feature activation | +3.73e-01 |
Pos logit contributions | |
cups | +0.012 |
ugs | +0.012 |
sisters | +0.012 |
ums | +0.011 |
els | +0.011 |
Neg logit contributions | |
shortcut | -0.011 |
Hop | -0.011 |
lot | -0.010 |
Left | -0.010 |
duel | -0.010 |
Token | pasta |
Token ID | 26296 |
Feature activation | -6.34e-01 |
Pos logit contributions | |
cups | +0.009 |
frost | +0.008 |
twins | +0.008 |
ugs | +0.008 |
sisters | +0.008 |
Neg logit contributions | |
shortcut | -0.011 |
lot | -0.010 |
duel | -0.010 |
unnecessary | -0.010 |
farewell | -0.010 |
Token | for |
Token ID | 329 |
Feature activation | +3.16e+00 |
Pos logit contributions | |
shortcut | +0.018 |
lot | +0.018 |
duel | +0.017 |
farewell | +0.017 |
unnecessary | +0.017 |
Neg logit contributions | |
cups | -0.015 |
frost | -0.014 |
twins | -0.013 |
ugs | -0.013 |
sisters | -0.013 |
Token | you |
Token ID | 345 |
Feature activation | +1.22e-02 |
Pos logit contributions | |
ays | +0.086 |
cups | +0.084 |
ugs | +0.082 |
ums | +0.079 |
frost | +0.079 |
Neg logit contributions | |
shortcut | -0.072 |
duel | -0.068 |
unnecessary | -0.064 |
gall | -0.064 |
defeat | -0.063 |
Token | today |
Token ID | 1909 |
Feature activation | +1.92e+00 |
Pos logit contributions | |
cups | +0.000 |
frost | +0.000 |
twins | +0.000 |
ugs | +0.000 |
sisters | +0.000 |
Neg logit contributions | |
shortcut | -0.000 |
lot | -0.000 |
duel | -0.000 |
noble | -0.000 |
farewell | -0.000 |
Token | . |
Token ID | 13 |
Feature activation | +5.39e-01 |
Pos logit contributions | |
cups | +0.049 |
frost | +0.046 |
ugs | +0.044 |
sisters | +0.044 |
twins | +0.044 |
Neg logit contributions | |
shortcut | -0.051 |
lot | -0.049 |
duel | -0.047 |
further | -0.045 |
gall | -0.045 |
Token | You |
Token ID | 921 |
Feature activation | +4.86e-02 |
Pos logit contributions | |
cups | +0.014 |
ugs | +0.014 |
els | +0.013 |
ries | +0.013 |
frost | +0.013 |
Neg logit contributions | |
lot | -0.012 |
shortcut | -0.011 |
duel | -0.011 |
Skip | -0.011 |
Left | -0.011 |
Token | will |
Token ID | 481 |
Feature activation | +4.04e-01 |
Pos logit contributions | |
cups | +0.001 |
frost | +0.001 |
twins | +0.001 |
ugs | +0.001 |
sisters | +0.001 |
Neg logit contributions | |
shortcut | -0.001 |
lot | -0.001 |
duel | -0.001 |
unnecessary | -0.001 |
farewell | -0.001 |
Token | but |
Token ID | 475 |
Feature activation | +9.74e-01 |
Pos logit contributions | |
shortcut | +0.021 |
lot | +0.021 |
duel | +0.019 |
farewell | +0.019 |
noble | +0.018 |
Neg logit contributions | |
cups | -0.026 |
frost | -0.024 |
twins | -0.023 |
sisters | -0.023 |
cloth | -0.022 |
Token | it |
Token ID | 340 |
Feature activation | -6.47e-01 |
Pos logit contributions | |
cups | +0.025 |
frost | +0.024 |
ugs | +0.023 |
twins | +0.023 |
sisters | +0.022 |
Neg logit contributions | |
shortcut | -0.026 |
lot | -0.025 |
duel | -0.024 |
further | -0.023 |
farewell | -0.023 |
Token | was |
Token ID | 373 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
shortcut | +0.018 |
lot | +0.017 |
duel | +0.017 |
unnecessary | +0.016 |
farewell | +0.016 |
Neg logit contributions | |
cups | -0.017 |
frost | -0.015 |
twins | -0.015 |
el | -0.014 |
sisters | -0.014 |
Token | too |
Token ID | 1165 |
Feature activation | +4.57e-01 |
Pos logit contributions | |
cups | +0.029 |
frost | +0.027 |
twins | +0.026 |
sisters | +0.025 |
ugs | +0.025 |
Neg logit contributions | |
shortcut | -0.031 |
lot | -0.031 |
duel | -0.029 |
farewell | -0.028 |
unnecessary | -0.028 |
Token | heavy |
Token ID | 4334 |
Feature activation | +7.43e-01 |
Pos logit contributions | |
cups | +0.016 |
frost | +0.015 |
ugs | +0.014 |
twins | +0.014 |
mug | +0.014 |
Neg logit contributions | |
shortcut | -0.008 |
lot | -0.008 |
duel | -0.008 |
unnecessary | -0.007 |
farewell | -0.007 |
Token | for |
Token ID | 329 |
Feature activation | +3.82e+00 |
Pos logit contributions | |
cups | +0.017 |
frost | +0.015 |
twins | +0.015 |
ugs | +0.014 |
sisters | +0.014 |
Neg logit contributions | |
shortcut | -0.023 |
lot | -0.022 |
duel | -0.021 |
noble | -0.020 |
farewell | -0.020 |
Token | her |
Token ID | 607 |
Feature activation | +9.78e-01 |
Pos logit contributions | |
cups | +0.061 |
ays | +0.056 |
ums | +0.053 |
ugs | +0.053 |
frost | +0.052 |
Neg logit contributions | |
shortcut | -0.133 |
unnecessary | -0.127 |
further | -0.126 |
farewell | -0.125 |
gall | -0.125 |
Token | . |
Token ID | 13 |
Feature activation | +1.29e+00 |
Pos logit contributions | |
cups | +0.022 |
frost | +0.020 |
twins | +0.019 |
ugs | +0.019 |
sisters | +0.019 |
Neg logit contributions | |
shortcut | -0.030 |
lot | -0.029 |
duel | -0.028 |
unnecessary | -0.027 |
farewell | -0.027 |
Token | Her |
Token ID | 2332 |
Feature activation | +1.02e+00 |
Pos logit contributions | |
cups | +0.021 |
frost | +0.020 |
ugs | +0.018 |
ions | +0.018 |
els | +0.018 |
Neg logit contributions | |
shortcut | -0.044 |
lot | -0.044 |
farewell | -0.042 |
duel | -0.042 |
higher | -0.041 |
Token | daddy |
Token ID | 35822 |
Feature activation | +6.57e-01 |
Pos logit contributions | |
cups | +0.020 |
frost | +0.019 |
ions | +0.018 |
izz | +0.018 |
ays | +0.018 |
Neg logit contributions | |
shortcut | -0.034 |
lot | -0.032 |
duel | -0.031 |
farewell | -0.030 |
spear | -0.030 |
Token | helped |
Token ID | 4193 |
Feature activation | +1.22e+00 |
Pos logit contributions | |
cups | +0.014 |
frost | +0.012 |
twins | +0.012 |
sisters | +0.012 |
ugs | +0.011 |
Neg logit contributions | |
shortcut | -0.021 |
lot | -0.020 |
duel | -0.020 |
unnecessary | -0.020 |
farewell | -0.019 |
Token | â |
Token ID | 22940 |
Feature activation | -9.85e-01 |
Pos logit contributions | |
cups | +0.014 |
frost | +0.013 |
ugs | +0.013 |
els | +0.012 |
sisters | +0.012 |
Neg logit contributions | |
shortcut | -0.018 |
lot | -0.017 |
duel | -0.016 |
higher | -0.016 |
further | -0.016 |
Token | € |
Token ID | 26391 |
Feature activation | +8.92e-02 |
Pos logit contributions | |
shortcut | +0.022 |
lot | +0.021 |
duel | +0.020 |
unnecessary | +0.019 |
farewell | +0.018 |
Neg logit contributions | |
cups | -0.031 |
frost | -0.029 |
twins | -0.028 |
ugs | -0.028 |
sisters | -0.027 |
Token | |
Token ID | 220 |
Feature activation | -6.27e-01 |
Pos logit contributions | |
cups | +0.002 |
frost | +0.002 |
twins | +0.002 |
sisters | +0.002 |
ugs | +0.002 |
Neg logit contributions | |
shortcut | -0.002 |
lot | -0.002 |
duel | -0.002 |
unnecessary | -0.002 |
gall | -0.002 |
Token |
|
Token ID | 198 |
Feature activation | +1.32e-01 |
Pos logit contributions | |
shortcut | +0.019 |
duel | +0.018 |
lot | +0.018 |
farewell | +0.017 |
noble | +0.017 |
Neg logit contributions | |
cups | -0.014 |
twins | -0.013 |
ugs | -0.013 |
frost | -0.013 |
ays | -0.012 |
Token |
|
Token ID | 198 |
Feature activation | +1.28e+00 |
Pos logit contributions | |
cups | +0.003 |
frost | +0.003 |
twins | +0.003 |
ugs | +0.003 |
sisters | +0.002 |
Neg logit contributions | |
shortcut | -0.004 |
lot | -0.004 |
duel | -0.004 |
unnecessary | -0.004 |
noble | -0.004 |
Token | The |
Token ID | 464 |
Feature activation | +9.04e-01 |
Pos logit contributions | |
cups | +0.037 |
frost | +0.036 |
ays | +0.035 |
els | +0.035 |
sisters | +0.034 |
Neg logit contributions | |
shortcut | -0.030 |
Hop | -0.028 |
lot | -0.027 |
duel | -0.027 |
further | -0.027 |
Token | little |
Token ID | 1310 |
Feature activation | +1.85e+00 |
Pos logit contributions | |
cups | +0.027 |
frost | +0.025 |
twins | +0.024 |
sisters | +0.024 |
ugs | +0.024 |
Neg logit contributions | |
shortcut | -0.021 |
lot | -0.020 |
duel | -0.019 |
unnecessary | -0.018 |
gall | -0.018 |
Token | girl |
Token ID | 2576 |
Feature activation | +1.37e+00 |
Pos logit contributions | |
ugs | +0.050 |
ays | +0.048 |
izz | +0.048 |
els | +0.048 |
ions | +0.048 |
Neg logit contributions | |
further | -0.045 |
higher | -0.044 |
shortcut | -0.043 |
farther | -0.043 |
closer | -0.043 |
Token | was |
Token ID | 373 |
Feature activation | +1.53e+00 |
Pos logit contributions | |
cups | +0.033 |
frost | +0.030 |
ugs | +0.029 |
twins | +0.029 |
sisters | +0.029 |
Neg logit contributions | |
shortcut | -0.038 |
lot | -0.037 |
duel | -0.036 |
gall | -0.035 |
noble | -0.034 |
Token | very |
Token ID | 845 |
Feature activation | +1.04e+00 |
Pos logit contributions | |
cups | +0.044 |
frost | +0.041 |
twins | +0.040 |
sisters | +0.039 |
ugs | +0.039 |
Neg logit contributions | |
shortcut | -0.037 |
lot | -0.035 |
duel | -0.034 |
unnecessary | -0.032 |
farewell | -0.032 |
Token | excited |
Token ID | 6568 |
Feature activation | +1.06e+00 |
Pos logit contributions | |
cups | +0.032 |
frost | +0.029 |
twins | +0.029 |
sisters | +0.028 |
ugs | +0.028 |
Neg logit contributions | |
shortcut | -0.024 |
lot | -0.023 |
duel | -0.022 |
unnecessary | -0.021 |
farewell | -0.021 |
Token | day |
Token ID | 1110 |
Feature activation | +1.38e+00 |
Pos logit contributions | |
cups | +0.022 |
frost | +0.020 |
twins | +0.020 |
mug | +0.020 |
sisters | +0.019 |
Neg logit contributions | |
lot | -0.024 |
shortcut | -0.024 |
duel | -0.023 |
unnecessary | -0.023 |
noble | -0.022 |
Token | Anna |
Token ID | 11735 |
Feature activation | -4.45e-02 |
Pos logit contributions | |
cups | +0.034 |
frost | +0.031 |
ugs | +0.030 |
twins | +0.029 |
sisters | +0.029 |
Neg logit contributions | |
shortcut | -0.039 |
lot | -0.038 |
duel | -0.036 |
farewell | -0.035 |
further | -0.035 |
Token | would |
Token ID | 561 |
Feature activation | -2.58e-01 |
Pos logit contributions | |
shortcut | +0.001 |
lot | +0.001 |
duel | +0.001 |
unnecessary | +0.001 |
further | +0.001 |
Neg logit contributions | |
cups | -0.001 |
frost | -0.001 |
sisters | -0.001 |
twins | -0.001 |
ugs | -0.001 |
Token | wake |
Token ID | 7765 |
Feature activation | -1.14e+00 |
Pos logit contributions | |
shortcut | +0.006 |
lot | +0.006 |
unnecessary | +0.006 |
duel | +0.006 |
further | +0.006 |
Neg logit contributions | |
cups | -0.007 |
sisters | -0.007 |
frost | -0.006 |
twins | -0.006 |
cloth | -0.006 |
Token | up |
Token ID | 510 |
Feature activation | -3.42e-01 |
Pos logit contributions | |
shortcut | +0.028 |
lot | +0.026 |
duel | +0.026 |
'd | +0.025 |
steep | +0.024 |
Neg logit contributions | |
cups | -0.033 |
twins | -0.031 |
sisters | -0.031 |
el | -0.030 |
els | -0.030 |
Token | early |
Token ID | 1903 |
Feature activation | +8.31e-01 |
Pos logit contributions | |
shortcut | +0.011 |
lot | +0.011 |
duel | +0.011 |
unnecessary | +0.010 |
Hop | +0.010 |
Neg logit contributions | |
cups | -0.007 |
sisters | -0.006 |
twins | -0.006 |
frost | -0.006 |
el | -0.006 |
Token | and |
Token ID | 290 |
Feature activation | -1.77e+00 |
Pos logit contributions | |
cups | +0.018 |
frost | +0.016 |
twins | +0.015 |
sisters | +0.015 |
mug | +0.015 |
Neg logit contributions | |
shortcut | -0.026 |
unnecessary | -0.025 |
lot | -0.025 |
duel | -0.024 |
farewell | -0.023 |
Token | go |
Token ID | 467 |
Feature activation | -1.96e+00 |
Pos logit contributions | |
shortcut | +0.047 |
lot | +0.045 |
unnecessary | +0.044 |
duel | +0.042 |
Hop | +0.041 |
Neg logit contributions | |
cups | -0.045 |
sisters | -0.042 |
frost | -0.041 |
twins | -0.040 |
paints | -0.040 |
Token | to |
Token ID | 284 |
Feature activation | -9.83e-01 |
Pos logit contributions | |
lot | +0.057 |
shortcut | +0.056 |
duel | +0.055 |
unnecessary | +0.055 |
rated | +0.053 |
Neg logit contributions | |
cups | -0.045 |
twins | -0.039 |
sisters | -0.038 |
frost | -0.038 |
ugs | -0.037 |
Token | the |
Token ID | 262 |
Feature activation | -1.24e+00 |
Pos logit contributions | |
lot | +0.032 |
shortcut | +0.032 |
Hop | +0.030 |
duel | +0.030 |
farewell | +0.030 |
Neg logit contributions | |
cups | -0.020 |
sisters | -0.018 |
twins | -0.017 |
frost | -0.017 |
mug | -0.016 |
Token | rock |
Token ID | 3881 |
Feature activation | -1.39e+00 |
Pos logit contributions | |
lot | +0.038 |
rated | +0.035 |
shortcut | +0.035 |
farewell | +0.035 |
unnecessary | +0.034 |
Neg logit contributions | |
cups | -0.028 |
twins | -0.026 |
mug | -0.025 |
tub | -0.025 |
sisters | -0.025 |
Token | . |
Token ID | 13 |
Feature activation | +1.72e+00 |
Pos logit contributions | |
cups | +0.032 |
frost | +0.030 |
ugs | +0.029 |
sisters | +0.028 |
twins | +0.028 |
Neg logit contributions | |
shortcut | -0.036 |
lot | -0.035 |
duel | -0.033 |
gall | -0.032 |
further | -0.032 |
Token | She |
Token ID | 1375 |
Feature activation | +1.42e+00 |
Pos logit contributions | |
cups | +0.035 |
ugs | +0.034 |
frost | +0.033 |
sisters | +0.033 |
el | +0.032 |
Neg logit contributions | |
shortcut | -0.053 |
lot | -0.052 |
farewell | -0.050 |
duel | -0.048 |
Skip | -0.047 |
Token | hugged |
Token ID | 41130 |
Feature activation | +6.93e-01 |
Pos logit contributions | |
cups | +0.036 |
frost | +0.033 |
twins | +0.032 |
sisters | +0.031 |
cloth | +0.030 |
Neg logit contributions | |
shortcut | -0.040 |
lot | -0.038 |
duel | -0.037 |
unnecessary | -0.036 |
farewell | -0.035 |
Token | them |
Token ID | 606 |
Feature activation | -1.45e+00 |
Pos logit contributions | |
cups | +0.017 |
frost | +0.016 |
ugs | +0.015 |
twins | +0.015 |
sisters | +0.015 |
Neg logit contributions | |
shortcut | -0.020 |
lot | -0.019 |
duel | -0.018 |
farewell | -0.018 |
further | -0.017 |
Token | and |
Token ID | 290 |
Feature activation | +7.25e-01 |
Pos logit contributions | |
unnecessary | +0.053 |
shortcut | +0.053 |
duel | +0.052 |
lot | +0.052 |
noble | +0.049 |
Neg logit contributions | |
cups | -0.026 |
twins | -0.023 |
sisters | -0.022 |
frost | -0.022 |
mug | -0.020 |
Token | asked |
Token ID | 1965 |
Feature activation | +6.28e-01 |
Pos logit contributions | |
cups | +0.017 |
ugs | +0.016 |
frost | +0.015 |
ays | +0.015 |
ums | +0.015 |
Neg logit contributions | |
shortcut | -0.021 |
lot | -0.021 |
duel | -0.020 |
gall | -0.020 |
further | -0.019 |
Token | them |
Token ID | 606 |
Feature activation | -2.35e+00 |
Pos logit contributions | |
cups | +0.014 |
twins | +0.012 |
frost | +0.012 |
sisters | +0.011 |
cloth | +0.011 |
Neg logit contributions | |
shortcut | -0.021 |
lot | -0.021 |
duel | -0.020 |
noble | -0.019 |
gall | -0.019 |
Token | what |
Token ID | 644 |
Feature activation | -1.56e-01 |
Pos logit contributions | |
lot | +0.071 |
shortcut | +0.070 |
duel | +0.069 |
unnecessary | +0.067 |
Hop | +0.064 |
Neg logit contributions | |
cups | -0.060 |
twins | -0.054 |
sisters | -0.053 |
cloth | -0.048 |
frost | -0.045 |
Token | happened |
Token ID | 3022 |
Feature activation | +1.98e-01 |
Pos logit contributions | |
shortcut | +0.004 |
lot | +0.004 |
duel | +0.004 |
gall | +0.004 |
noble | +0.004 |
Neg logit contributions | |
cups | -0.004 |
frost | -0.004 |
ugs | -0.003 |
twins | -0.003 |
sisters | -0.003 |
Token | . |
Token ID | 13 |
Feature activation | +1.65e+00 |
Pos logit contributions | |
cups | +0.003 |
twins | +0.003 |
frost | +0.003 |
sisters | +0.003 |
cloth | +0.002 |
Neg logit contributions | |
shortcut | -0.007 |
duel | -0.007 |
lot | -0.007 |
unnecessary | -0.007 |
farewell | -0.007 |
Token |
|
Token ID | 198 |
Feature activation | +8.03e-01 |
Pos logit contributions | |
cups | +0.042 |
frost | +0.039 |
ugs | +0.038 |
sisters | +0.037 |
el | +0.037 |
Neg logit contributions | |
shortcut | -0.044 |
lot | -0.043 |
duel | -0.041 |
farewell | -0.039 |
Skip | -0.039 |
Token | the |
Token ID | 262 |
Feature activation | +4.30e-01 |
Pos logit contributions | |
cups | +0.005 |
frost | +0.005 |
twins | +0.004 |
ugs | +0.004 |
sisters | +0.004 |
Neg logit contributions | |
shortcut | -0.006 |
lot | -0.005 |
duel | -0.005 |
farewell | -0.005 |
noble | -0.005 |
Token | ground |
Token ID | 2323 |
Feature activation | +3.22e-01 |
Pos logit contributions | |
cups | +0.009 |
frost | +0.008 |
ugs | +0.008 |
sisters | +0.008 |
twins | +0.008 |
Neg logit contributions | |
shortcut | -0.014 |
lot | -0.013 |
duel | -0.013 |
unnecessary | -0.012 |
further | -0.012 |
Token | and |
Token ID | 290 |
Feature activation | -8.21e-01 |
Pos logit contributions | |
cups | +0.007 |
frost | +0.006 |
twins | +0.005 |
cloth | +0.005 |
el | +0.005 |
Neg logit contributions | |
shortcut | -0.011 |
lot | -0.010 |
duel | -0.010 |
unnecessary | -0.010 |
farewell | -0.010 |
Token | noticed |
Token ID | 6810 |
Feature activation | -7.27e-01 |
Pos logit contributions | |
shortcut | +0.024 |
lot | +0.023 |
unnecessary | +0.023 |
duel | +0.022 |
farewell | +0.022 |
Neg logit contributions | |
cups | -0.020 |
frost | -0.019 |
twins | -0.018 |
sisters | -0.018 |
cloth | -0.016 |
Token | it |
Token ID | 340 |
Feature activation | -7.45e-01 |
Pos logit contributions | |
shortcut | +0.023 |
lot | +0.023 |
duel | +0.022 |
unnecessary | +0.021 |
noble | +0.021 |
Neg logit contributions | |
cups | -0.017 |
frost | -0.016 |
twins | -0.015 |
els | -0.014 |
ions | -0.014 |
Token | was |
Token ID | 373 |
Feature activation | +2.01e+00 |
Pos logit contributions | |
shortcut | +0.024 |
lot | +0.023 |
duel | +0.023 |
unnecessary | +0.023 |
farewell | +0.022 |
Neg logit contributions | |
cups | -0.016 |
frost | -0.015 |
el | -0.014 |
twins | -0.014 |
els | -0.013 |
Token | a |
Token ID | 257 |
Feature activation | -2.35e-01 |
Pos logit contributions | |
cups | +0.053 |
frost | +0.048 |
ugs | +0.048 |
twins | +0.047 |
sisters | +0.047 |
Neg logit contributions | |
shortcut | -0.053 |
lot | -0.051 |
duel | -0.049 |
gall | -0.047 |
noble | -0.047 |
Token | leak |
Token ID | 13044 |
Feature activation | +1.40e+00 |
Pos logit contributions | |
lot | +0.007 |
shortcut | +0.007 |
duel | +0.006 |
farewell | +0.006 |
unnecessary | +0.006 |
Neg logit contributions | |
cups | -0.006 |
frost | -0.005 |
twins | -0.005 |
mug | -0.005 |
sisters | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | -1.06e+00 |
Pos logit contributions | |
cups | +0.032 |
frost | +0.029 |
twins | +0.028 |
ugs | +0.028 |
sisters | +0.028 |
Neg logit contributions | |
shortcut | -0.042 |
lot | -0.040 |
duel | -0.039 |
gall | -0.038 |
further | -0.037 |
Token |
|
Token ID | 198 |
Feature activation | -1.32e+00 |
Pos logit contributions | |
shortcut | +0.032 |
unnecessary | +0.031 |
duel | +0.030 |
lot | +0.029 |
gall | +0.028 |
Neg logit contributions | |
cups | -0.025 |
frost | -0.022 |
twins | -0.021 |
ays | -0.021 |
ugs | -0.020 |
Token |
|
Token ID | 198 |
Feature activation | -5.21e-01 |
Pos logit contributions | |
shortcut | +0.045 |
unnecessary | +0.045 |
duel | +0.043 |
lot | +0.042 |
gall | +0.041 |
Neg logit contributions | |
cups | -0.023 |
twins | -0.020 |
frost | -0.020 |
Mom | -0.020 |
izz | -0.020 |
Token | we |
Token ID | 356 |
Feature activation | +1.63e-01 |
Pos logit contributions | |
lot | +0.007 |
shortcut | +0.007 |
duel | +0.006 |
farewell | +0.006 |
gall | +0.006 |
Neg logit contributions | |
cups | -0.006 |
frost | -0.005 |
twins | -0.005 |
sisters | -0.005 |
mug | -0.005 |
Token | should |
Token ID | 815 |
Feature activation | +1.23e+00 |
Pos logit contributions | |
cups | +0.005 |
frost | +0.004 |
twins | +0.004 |
ugs | +0.004 |
sisters | +0.004 |
Neg logit contributions | |
shortcut | -0.004 |
lot | -0.004 |
duel | -0.004 |
unnecessary | -0.003 |
farewell | -0.003 |
Token | always |
Token ID | 1464 |
Feature activation | -4.06e-01 |
Pos logit contributions | |
cups | +0.037 |
frost | +0.034 |
ugs | +0.033 |
twins | +0.033 |
ays | +0.033 |
Neg logit contributions | |
shortcut | -0.028 |
lot | -0.027 |
duel | -0.026 |
farewell | -0.025 |
gall | -0.025 |
Token | let |
Token ID | 1309 |
Feature activation | -2.30e-01 |
Pos logit contributions | |
shortcut | +0.010 |
lot | +0.010 |
duel | +0.010 |
unnecessary | +0.010 |
farewell | +0.009 |
Neg logit contributions | |
cups | -0.010 |
frost | -0.010 |
twins | -0.010 |
sisters | -0.010 |
el | -0.009 |
Token | bugs |
Token ID | 11316 |
Feature activation | -2.09e+00 |
Pos logit contributions | |
lot | +0.007 |
shortcut | +0.007 |
duel | +0.007 |
noble | +0.007 |
steep | +0.007 |
Neg logit contributions | |
cups | -0.005 |
twins | -0.004 |
sisters | -0.004 |
frost | -0.004 |
el | -0.004 |
Token | stay |
Token ID | 2652 |
Feature activation | +6.45e-01 |
Pos logit contributions | |
lot | +0.056 |
shortcut | +0.055 |
duel | +0.052 |
noble | +0.051 |
farewell | +0.050 |
Neg logit contributions | |
cups | -0.055 |
frost | -0.050 |
sisters | -0.048 |
twins | -0.047 |
cloth | -0.047 |
Token | outside |
Token ID | 2354 |
Feature activation | +7.74e-01 |
Pos logit contributions | |
cups | +0.021 |
frost | +0.019 |
twins | +0.019 |
sisters | +0.019 |
el | +0.018 |
Neg logit contributions | |
shortcut | -0.014 |
lot | -0.013 |
duel | -0.013 |
gall | -0.012 |
farewell | -0.012 |
Token | where |
Token ID | 810 |
Feature activation | -3.07e-01 |
Pos logit contributions | |
cups | +0.020 |
frost | +0.018 |
twins | +0.017 |
ugs | +0.017 |
sisters | +0.017 |
Neg logit contributions | |
shortcut | -0.021 |
lot | -0.021 |
duel | -0.020 |
unnecessary | -0.019 |
farewell | -0.019 |
Token | they |
Token ID | 484 |
Feature activation | -3.77e-01 |
Pos logit contributions | |
shortcut | +0.009 |
lot | +0.009 |
duel | +0.008 |
gall | +0.008 |
farewell | +0.008 |
Neg logit contributions | |
cups | -0.007 |
frost | -0.007 |
twins | -0.007 |
sisters | -0.006 |
ugs | -0.006 |
Token | belong |
Token ID | 5594 |
Feature activation | +9.45e-01 |
Pos logit contributions | |
lot | +0.012 |
shortcut | +0.012 |
farewell | +0.011 |
duel | +0.011 |
Skip | +0.011 |
Neg logit contributions | |
cups | -0.008 |
frost | -0.008 |
sisters | -0.007 |
twins | -0.007 |
paints | -0.007 |
Token | ." |
Token ID | 526 |
Feature activation | +9.32e-01 |
Pos logit contributions | |
cups | +0.020 |
twins | +0.018 |
frost | +0.017 |
sisters | +0.017 |
ays | +0.016 |
Neg logit contributions | |
shortcut | -0.031 |
duel | -0.030 |
lot | -0.030 |
gall | -0.028 |
farewell | -0.028 |
Token | of |
Token ID | 286 |
Feature activation | -1.11e+00 |
Pos logit contributions | |
cups | +0.032 |
frost | +0.030 |
twins | +0.028 |
ugs | +0.028 |
sisters | +0.028 |
Neg logit contributions | |
shortcut | -0.045 |
lot | -0.043 |
duel | -0.042 |
further | -0.041 |
gall | -0.040 |
Token | catch |
Token ID | 4929 |
Feature activation | -4.90e-01 |
Pos logit contributions | |
farewell | +0.036 |
lot | +0.034 |
shortcut | +0.033 |
rated | +0.033 |
stru | +0.031 |
Neg logit contributions | |
cups | -0.025 |
twins | -0.023 |
dough | -0.022 |
sisters | -0.022 |
frost | -0.021 |
Token | when |
Token ID | 618 |
Feature activation | +4.29e-01 |
Pos logit contributions | |
shortcut | +0.016 |
lot | +0.016 |
duel | +0.015 |
farewell | +0.015 |
noble | +0.015 |
Neg logit contributions | |
cups | -0.010 |
twins | -0.009 |
frost | -0.008 |
sisters | -0.008 |
mug | -0.008 |
Token | Pete |
Token ID | 17542 |
Feature activation | -1.93e-02 |
Pos logit contributions | |
cups | +0.008 |
twins | +0.007 |
sisters | +0.007 |
frost | +0.007 |
el | +0.006 |
Neg logit contributions | |
shortcut | -0.015 |
lot | -0.014 |
duel | -0.014 |
noble | -0.013 |
unnecessary | -0.013 |
Token | suddenly |
Token ID | 6451 |
Feature activation | +7.31e-01 |
Pos logit contributions | |
shortcut | +0.001 |
unnecessary | +0.001 |
farewell | +0.001 |
duel | +0.001 |
lot | +0.001 |
Neg logit contributions | |
cups | -0.000 |
frost | -0.000 |
twins | -0.000 |
sisters | -0.000 |
mug | -0.000 |
Token | dropped |
Token ID | 5710 |
Feature activation | -2.91e-01 |
Pos logit contributions | |
cups | +0.018 |
frost | +0.016 |
twins | +0.015 |
sisters | +0.015 |
el | +0.015 |
Neg logit contributions | |
shortcut | -0.021 |
duel | -0.020 |
lot | -0.020 |
noble | -0.019 |
unnecessary | -0.019 |
Token | his |
Token ID | 465 |
Feature activation | +2.05e-01 |
Pos logit contributions | |
shortcut | +0.008 |
lot | +0.008 |
duel | +0.008 |
noble | +0.007 |
unnecessary | +0.007 |
Neg logit contributions | |
cups | -0.007 |
frost | -0.007 |
twins | -0.007 |
sisters | -0.006 |
el | -0.006 |
Token | glove |
Token ID | 29144 |
Feature activation | +7.56e-01 |
Pos logit contributions | |
cups | +0.005 |
twins | +0.005 |
mug | +0.005 |
frost | +0.005 |
sisters | +0.005 |
Neg logit contributions | |
shortcut | -0.006 |
lot | -0.006 |
duel | -0.005 |
further | -0.005 |
unnecessary | -0.005 |
Token | in |
Token ID | 287 |
Feature activation | +1.55e+00 |
Pos logit contributions | |
cups | +0.017 |
frost | +0.015 |
twins | +0.015 |
ugs | +0.014 |
sisters | +0.014 |
Neg logit contributions | |
shortcut | -0.023 |
lot | -0.022 |
duel | -0.022 |
unnecessary | -0.021 |
farewell | -0.021 |
Token | the |
Token ID | 262 |
Feature activation | +6.50e-01 |
Pos logit contributions | |
ugs | +0.045 |
cups | +0.044 |
frost | +0.042 |
ums | +0.042 |
ays | +0.042 |
Neg logit contributions | |
shortcut | -0.037 |
duel | -0.033 |
further | -0.033 |
lot | -0.033 |
farther | -0.032 |
Token | dirt |
Token ID | 13647 |
Feature activation | +1.29e+00 |
Pos logit contributions | |
cups | +0.016 |
frost | +0.015 |
twins | +0.014 |
ugs | +0.014 |
sisters | +0.014 |
Neg logit contributions | |
shortcut | -0.019 |
lot | -0.018 |
duel | -0.017 |
unnecessary | -0.017 |
farewell | -0.017 |
Token | a |
Token ID | 257 |
Feature activation | +4.43e-02 |
Pos logit contributions | |
cups | +0.057 |
ays | +0.056 |
ugs | +0.055 |
els | +0.054 |
frost | +0.054 |
Neg logit contributions | |
shortcut | -0.028 |
gall | -0.026 |
lot | -0.026 |
higher | -0.025 |
duel | -0.025 |
Token | brave |
Token ID | 14802 |
Feature activation | +3.37e-01 |
Pos logit contributions | |
cups | +0.001 |
frost | +0.001 |
twins | +0.001 |
sisters | +0.001 |
ugs | +0.001 |
Neg logit contributions | |
shortcut | -0.001 |
lot | -0.001 |
duel | -0.001 |
unnecessary | -0.001 |
farewell | -0.001 |
Token | and |
Token ID | 290 |
Feature activation | -1.97e+00 |
Pos logit contributions | |
cups | +0.007 |
frost | +0.006 |
twins | +0.006 |
sisters | +0.005 |
ugs | +0.005 |
Neg logit contributions | |
shortcut | -0.011 |
lot | -0.011 |
duel | -0.011 |
unnecessary | -0.010 |
farewell | -0.010 |
Token | tough |
Token ID | 5802 |
Feature activation | +5.95e-01 |
Pos logit contributions | |
shortcut | +0.047 |
lot | +0.047 |
duel | +0.045 |
unnecessary | +0.043 |
farewell | +0.042 |
Neg logit contributions | |
cups | -0.055 |
twins | -0.051 |
sisters | -0.050 |
frost | -0.050 |
mug | -0.048 |
Token | girl |
Token ID | 2576 |
Feature activation | +1.31e+00 |
Pos logit contributions | |
cups | +0.012 |
twins | +0.011 |
sisters | +0.010 |
frost | +0.010 |
mug | +0.010 |
Neg logit contributions | |
lot | -0.020 |
shortcut | -0.019 |
duel | -0.019 |
unnecessary | -0.018 |
nearer | -0.018 |
Token | and |
Token ID | 290 |
Feature activation | -7.64e-01 |
Pos logit contributions | |
cups | +0.035 |
ays | +0.033 |
frost | +0.033 |
ugs | +0.032 |
icing | +0.032 |
Neg logit contributions | |
shortcut | -0.031 |
lot | -0.030 |
gall | -0.029 |
noble | -0.029 |
duel | -0.028 |
Token | she |
Token ID | 673 |
Feature activation | -8.30e-01 |
Pos logit contributions | |
shortcut | +0.018 |
lot | +0.017 |
duel | +0.017 |
noble | +0.016 |
gall | +0.016 |
Neg logit contributions | |
cups | -0.022 |
frost | -0.021 |
ays | -0.021 |
ugs | -0.020 |
els | -0.020 |
Token | liked |
Token ID | 8288 |
Feature activation | -1.92e-01 |
Pos logit contributions | |
shortcut | +0.019 |
lot | +0.019 |
duel | +0.018 |
gall | +0.017 |
noble | +0.017 |
Neg logit contributions | |
cups | -0.025 |
frost | -0.023 |
ugs | -0.022 |
twins | -0.022 |
sisters | -0.022 |
Token | going |
Token ID | 1016 |
Feature activation | -2.39e+00 |
Pos logit contributions | |
shortcut | +0.005 |
lot | +0.005 |
duel | +0.005 |
farewell | +0.004 |
unnecessary | +0.004 |
Neg logit contributions | |
cups | -0.005 |
frost | -0.005 |
twins | -0.005 |
sisters | -0.005 |
mug | -0.005 |
Token | on |
Token ID | 319 |
Feature activation | -7.55e-02 |
Pos logit contributions | |
lot | +0.080 |
duel | +0.079 |
farewell | +0.079 |
shortcut | +0.076 |
Hop | +0.075 |
Neg logit contributions | |
cups | -0.043 |
twins | -0.039 |
sisters | -0.035 |
ugs | -0.035 |
paints | -0.034 |
Token | adventures |
Token ID | 17545 |
Feature activation | +4.60e-02 |
Pos logit contributions | |
shortcut | +0.002 |
lot | +0.002 |
duel | +0.001 |
farewell | +0.001 |
noble | +0.001 |
Neg logit contributions | |
cups | -0.002 |
frost | -0.002 |
twins | -0.002 |
sisters | -0.002 |
folders | -0.002 |
Token | saw |
Token ID | 2497 |
Feature activation | +5.43e-01 |
Pos logit contributions | |
cups | +0.033 |
frost | +0.030 |
twins | +0.028 |
sisters | +0.028 |
cloth | +0.027 |
Neg logit contributions | |
shortcut | -0.035 |
lot | -0.033 |
duel | -0.032 |
unnecessary | -0.031 |
farewell | -0.031 |
Token | a |
Token ID | 257 |
Feature activation | -1.01e+00 |
Pos logit contributions | |
cups | +0.015 |
ugs | +0.014 |
frost | +0.013 |
ays | +0.013 |
cloth | +0.013 |
Neg logit contributions | |
shortcut | -0.013 |
lot | -0.013 |
duel | -0.012 |
further | -0.012 |
unnecessary | -0.012 |
Token | butterfly |
Token ID | 35113 |
Feature activation | -1.12e+00 |
Pos logit contributions | |
lot | +0.028 |
shortcut | +0.027 |
farewell | +0.026 |
duel | +0.026 |
unnecessary | +0.026 |
Neg logit contributions | |
cups | -0.025 |
frost | -0.023 |
mug | -0.023 |
twins | -0.023 |
sisters | -0.022 |
Token | . |
Token ID | 13 |
Feature activation | +1.62e-02 |
Pos logit contributions | |
shortcut | +0.031 |
lot | +0.029 |
duel | +0.029 |
gall | +0.027 |
further | +0.027 |
Neg logit contributions | |
cups | -0.029 |
frost | -0.026 |
ugs | -0.025 |
twins | -0.025 |
sisters | -0.025 |
Token | She |
Token ID | 1375 |
Feature activation | -4.05e-01 |
Pos logit contributions | |
cups | +0.000 |
frost | +0.000 |
els | +0.000 |
sisters | +0.000 |
twins | +0.000 |
Neg logit contributions | |
shortcut | -0.000 |
lot | -0.000 |
duel | -0.000 |
gall | -0.000 |
higher | -0.000 |
Token | started |
Token ID | 2067 |
Feature activation | -7.31e-01 |
Pos logit contributions | |
shortcut | +0.013 |
unnecessary | +0.012 |
lot | +0.012 |
duel | +0.012 |
further | +0.011 |
Neg logit contributions | |
cups | -0.009 |
frost | -0.008 |
sisters | -0.008 |
twins | -0.008 |
el | -0.008 |
Token | to |
Token ID | 284 |
Feature activation | -2.42e-01 |
Pos logit contributions | |
shortcut | +0.019 |
lot | +0.019 |
noble | +0.017 |
duel | +0.017 |
unnecessary | +0.017 |
Neg logit contributions | |
cups | -0.020 |
twins | -0.018 |
sisters | -0.018 |
frost | -0.017 |
mug | -0.017 |
Token | chase |
Token ID | 15505 |
Feature activation | -9.19e-01 |
Pos logit contributions | |
lot | +0.006 |
shortcut | +0.006 |
unnecessary | +0.005 |
noble | +0.005 |
farewell | +0.005 |
Neg logit contributions | |
cups | -0.007 |
mug | -0.006 |
sisters | -0.006 |
frost | -0.006 |
twins | -0.006 |
Token | the |
Token ID | 262 |
Feature activation | -1.59e-01 |
Pos logit contributions | |
shortcut | +0.026 |
lot | +0.026 |
duel | +0.025 |
noble | +0.024 |
gall | +0.023 |
Neg logit contributions | |
cups | -0.023 |
frost | -0.020 |
twins | -0.020 |
el | -0.020 |
sisters | -0.019 |
Token | butterfly |
Token ID | 35113 |
Feature activation | -3.14e+00 |
Pos logit contributions | |
shortcut | +0.003 |
lot | +0.003 |
duel | +0.003 |
further | +0.003 |
unnecessary | +0.003 |
Neg logit contributions | |
cups | -0.005 |
frost | -0.005 |
ugs | -0.005 |
twins | -0.005 |
ays | -0.005 |
Token | and |
Token ID | 290 |
Feature activation | -1.05e+00 |
Pos logit contributions | |
lot | +0.111 |
shortcut | +0.104 |
duel | +0.103 |
noble | +0.102 |
unnecessary | +0.101 |
Neg logit contributions | |
cups | -0.057 |
mug | -0.046 |
sisters | -0.045 |
el | -0.045 |
twins | -0.044 |
Token | that |
Token ID | 326 |
Feature activation | -2.12e+00 |
Pos logit contributions | |
shortcut | +0.033 |
lot | +0.032 |
duel | +0.030 |
unnecessary | +0.029 |
farewell | +0.029 |
Neg logit contributions | |
cups | -0.027 |
frost | -0.025 |
twins | -0.024 |
sisters | -0.024 |
ays | -0.023 |
Token | it |
Token ID | 340 |
Feature activation | -4.11e-01 |
Pos logit contributions | |
shortcut | +0.055 |
lot | +0.054 |
duel | +0.051 |
unnecessary | +0.050 |
farewell | +0.049 |
Neg logit contributions | |
cups | -0.058 |
frost | -0.053 |
twins | -0.052 |
sisters | -0.051 |
folders | -0.049 |
Token | was |
Token ID | 373 |
Feature activation | +2.71e+00 |
Pos logit contributions | |
shortcut | +0.012 |
lot | +0.012 |
duel | +0.011 |
unnecessary | +0.011 |
farewell | +0.011 |
Neg logit contributions | |
cups | -0.010 |
frost | -0.009 |
twins | -0.009 |
sisters | -0.008 |
el | -0.008 |
Token | now |
Token ID | 783 |
Feature activation | +8.01e-01 |
Pos logit contributions | |
cups | +0.078 |
ugs | +0.074 |
frost | +0.072 |
ays | +0.071 |
twins | +0.069 |
Neg logit contributions | |
shortcut | -0.064 |
lot | -0.061 |
duel | -0.059 |
noble | -0.056 |
further | -0.056 |
Token | organized |
Token ID | 8389 |
Feature activation | +1.31e-01 |
Pos logit contributions | |
cups | +0.021 |
frost | +0.020 |
twins | +0.019 |
ugs | +0.019 |
sisters | +0.019 |
Neg logit contributions | |
shortcut | -0.021 |
lot | -0.020 |
duel | -0.020 |
unnecessary | -0.019 |
farewell | -0.019 |
Token | . |
Token ID | 13 |
Feature activation | -1.07e+00 |
Pos logit contributions | |
cups | +0.003 |
frost | +0.002 |
twins | +0.002 |
cloth | +0.002 |
sisters | +0.002 |
Neg logit contributions | |
shortcut | -0.004 |
lot | -0.004 |
duel | -0.004 |
unnecessary | -0.004 |
farewell | -0.004 |
Token | He |
Token ID | 679 |
Feature activation | -3.46e-02 |
Pos logit contributions | |
shortcut | +0.029 |
lot | +0.027 |
duel | +0.027 |
unnecessary | +0.026 |
further | +0.025 |
Neg logit contributions | |
cups | -0.029 |
frost | -0.026 |
twins | -0.026 |
ugs | -0.025 |
sisters | -0.025 |
Token | smiled |
Token ID | 13541 |
Feature activation | -7.03e-01 |
Pos logit contributions | |
shortcut | +0.001 |
lot | +0.001 |
duel | +0.001 |
gall | +0.001 |
noble | +0.001 |
Neg logit contributions | |
cups | -0.001 |
frost | -0.001 |
ugs | -0.001 |
twins | -0.001 |
ays | -0.001 |
Token | and |
Token ID | 290 |
Feature activation | -4.55e-01 |
Pos logit contributions | |
shortcut | +0.024 |
duel | +0.023 |
lot | +0.022 |
unnecessary | +0.022 |
gall | +0.021 |
Neg logit contributions | |
cups | -0.014 |
twins | -0.013 |
frost | -0.012 |
sisters | -0.012 |
el | -0.011 |
Token | closed |
Token ID | 4838 |
Feature activation | -6.21e-01 |
Pos logit contributions | |
shortcut | +0.012 |
lot | +0.012 |
duel | +0.011 |
gall | +0.011 |
farewell | +0.011 |
Neg logit contributions | |
cups | -0.012 |
frost | -0.011 |
ugs | -0.011 |
ays | -0.010 |
twins | -0.010 |
Token | the |
Token ID | 262 |
Feature activation | +6.10e-01 |
Pos logit contributions | |
shortcut | +0.017 |
lot | +0.017 |
duel | +0.016 |
unnecessary | +0.016 |
noble | +0.015 |
Neg logit contributions | |
cups | -0.016 |
frost | -0.015 |
twins | -0.014 |
el | -0.014 |
cloth | -0.014 |
Token | you |
Token ID | 345 |
Feature activation | -5.14e-01 |
Pos logit contributions | |
shortcut | +0.006 |
lot | +0.005 |
duel | +0.005 |
farewell | +0.005 |
unnecessary | +0.005 |
Neg logit contributions | |
cups | -0.004 |
twins | -0.003 |
frost | -0.003 |
sisters | -0.003 |
ions | -0.003 |
Token | make |
Token ID | 787 |
Feature activation | +3.91e-01 |
Pos logit contributions | |
shortcut | +0.014 |
lot | +0.014 |
unnecessary | +0.013 |
duel | +0.013 |
farewell | +0.012 |
Neg logit contributions | |
cups | -0.013 |
frost | -0.012 |
sisters | -0.012 |
twins | -0.012 |
el | -0.011 |
Token | things |
Token ID | 1243 |
Feature activation | -1.44e+00 |
Pos logit contributions | |
cups | +0.009 |
frost | +0.009 |
twins | +0.008 |
sisters | +0.008 |
el | +0.008 |
Neg logit contributions | |
shortcut | -0.012 |
lot | -0.011 |
duel | -0.011 |
farewell | -0.010 |
gall | -0.010 |
Token | vanish |
Token ID | 39572 |
Feature activation | -1.16e+00 |
Pos logit contributions | |
shortcut | +0.036 |
lot | +0.035 |
duel | +0.033 |
unnecessary | +0.032 |
noble | +0.032 |
Neg logit contributions | |
cups | -0.040 |
frost | -0.037 |
twins | -0.036 |
ugs | -0.036 |
sisters | -0.035 |
Token | ?" |
Token ID | 1701 |
Feature activation | -6.32e-01 |
Pos logit contributions | |
shortcut | +0.039 |
lot | +0.038 |
duel | +0.037 |
farewell | +0.037 |
gall | +0.036 |
Neg logit contributions | |
cups | -0.022 |
frost | -0.019 |
twins | -0.019 |
sisters | -0.018 |
el | -0.018 |
Token | The |
Token ID | 383 |
Feature activation | -7.22e-01 |
Pos logit contributions | |
shortcut | +0.020 |
lot | +0.018 |
duel | +0.018 |
unnecessary | +0.018 |
noble | +0.017 |
Neg logit contributions | |
cups | -0.014 |
frost | -0.013 |
twins | -0.012 |
sisters | -0.012 |
ugs | -0.012 |
Token | magician |
Token ID | 42055 |
Feature activation | +4.98e-01 |
Pos logit contributions | |
shortcut | +0.020 |
lot | +0.020 |
unnecessary | +0.019 |
duel | +0.018 |
steep | +0.018 |
Neg logit contributions | |
cups | -0.018 |
twins | -0.017 |
frost | -0.017 |
sisters | -0.017 |
folders | -0.016 |
Token | smiled |
Token ID | 13541 |
Feature activation | -6.75e-01 |
Pos logit contributions | |
cups | +0.011 |
frost | +0.010 |
twins | +0.010 |
sisters | +0.010 |
el | +0.009 |
Neg logit contributions | |
shortcut | -0.016 |
lot | -0.015 |
unnecessary | -0.015 |
duel | -0.014 |
further | -0.014 |
Token | and |
Token ID | 290 |
Feature activation | -2.13e-01 |
Pos logit contributions | |
shortcut | +0.025 |
duel | +0.023 |
unnecessary | +0.023 |
lot | +0.023 |
hike | +0.022 |
Neg logit contributions | |
cups | -0.012 |
twins | -0.011 |
frost | -0.011 |
sisters | -0.010 |
el | -0.010 |
Token | said |
Token ID | 531 |
Feature activation | -1.82e+00 |
Pos logit contributions | |
shortcut | +0.008 |
lot | +0.007 |
duel | +0.007 |
unnecessary | +0.007 |
farewell | +0.007 |
Neg logit contributions | |
cups | -0.004 |
frost | -0.003 |
twins | -0.003 |
sisters | -0.003 |
el | -0.003 |
Token | , |
Token ID | 11 |
Feature activation | -1.34e+00 |
Pos logit contributions | |
shortcut | +0.069 |
lot | +0.065 |
unnecessary | +0.065 |
duel | +0.062 |
hike | +0.062 |
Neg logit contributions | |
cups | -0.030 |
frost | -0.026 |
sisters | -0.026 |
twins | -0.026 |
ums | -0.022 |
Token | loved |
Token ID | 6151 |
Feature activation | -1.34e+00 |
Pos logit contributions | |
shortcut | +0.029 |
lot | +0.026 |
duel | +0.026 |
unnecessary | +0.025 |
farewell | +0.025 |
Neg logit contributions | |
cups | -0.049 |
frost | -0.046 |
twins | -0.046 |
sisters | -0.045 |
el | -0.045 |
Token | to |
Token ID | 284 |
Feature activation | -1.78e+00 |
Pos logit contributions | |
shortcut | +0.036 |
lot | +0.035 |
farewell | +0.033 |
duel | +0.033 |
unnecessary | +0.032 |
Neg logit contributions | |
cups | -0.035 |
twins | -0.032 |
sisters | -0.032 |
frost | -0.031 |
mug | -0.030 |
Token | play |
Token ID | 711 |
Feature activation | -1.93e+00 |
Pos logit contributions | |
shortcut | +0.043 |
lot | +0.043 |
unnecessary | +0.040 |
duel | +0.039 |
further | +0.038 |
Neg logit contributions | |
cups | -0.050 |
frost | -0.046 |
sisters | -0.046 |
twins | -0.045 |
mug | -0.044 |
Token | his |
Token ID | 465 |
Feature activation | -1.02e+00 |
Pos logit contributions | |
shortcut | +0.061 |
lot | +0.061 |
duel | +0.058 |
unnecessary | +0.057 |
farewell | +0.057 |
Neg logit contributions | |
cups | -0.041 |
frost | -0.036 |
twins | -0.036 |
sisters | -0.035 |
mug | -0.034 |
Token | drum |
Token ID | 13026 |
Feature activation | -1.32e+00 |
Pos logit contributions | |
lot | +0.033 |
shortcut | +0.032 |
farewell | +0.031 |
unnecessary | +0.031 |
duel | +0.031 |
Neg logit contributions | |
cups | -0.020 |
twins | -0.019 |
sisters | -0.018 |
mug | -0.018 |
frost | -0.017 |
Token | and |
Token ID | 290 |
Feature activation | -3.15e+00 |
Pos logit contributions | |
shortcut | +0.042 |
lot | +0.041 |
unnecessary | +0.040 |
farewell | +0.039 |
duel | +0.038 |
Neg logit contributions | |
cups | -0.028 |
sisters | -0.025 |
frost | -0.024 |
twins | -0.024 |
folders | -0.023 |
Token | make |
Token ID | 787 |
Feature activation | -2.48e+00 |
Pos logit contributions | |
lot | +0.085 |
shortcut | +0.084 |
unnecessary | +0.079 |
farewell | +0.078 |
duel | +0.077 |
Neg logit contributions | |
cups | -0.080 |
sisters | -0.074 |
twins | -0.072 |
frost | -0.071 |
paints | -0.069 |
Token | lots |
Token ID | 6041 |
Feature activation | -3.41e+00 |
Pos logit contributions | |
shortcut | +0.070 |
lot | +0.067 |
duel | +0.063 |
farewell | +0.062 |
noble | +0.061 |
Neg logit contributions | |
cups | -0.062 |
twins | -0.060 |
sisters | -0.058 |
frost | -0.057 |
el | -0.053 |
Token | of |
Token ID | 286 |
Feature activation | -2.56e+00 |
Pos logit contributions | |
shortcut | +0.139 |
lot | +0.139 |
duel | +0.134 |
farewell | +0.129 |
unnecessary | +0.129 |
Neg logit contributions | |
cups | -0.043 |
frost | -0.035 |
twins | -0.035 |
sisters | -0.033 |
ugs | -0.031 |
Token | noise |
Token ID | 7838 |
Feature activation | -1.09e+00 |
Pos logit contributions | |
lot | +0.075 |
shortcut | +0.073 |
duel | +0.070 |
farewell | +0.067 |
noble | +0.065 |
Neg logit contributions | |
cups | -0.063 |
twins | -0.060 |
sisters | -0.058 |
frost | -0.057 |
folders | -0.053 |
Token | ! |
Token ID | 0 |
Feature activation | -1.94e+00 |
Pos logit contributions | |
shortcut | +0.037 |
lot | +0.036 |
duel | +0.035 |
farewell | +0.034 |
noble | +0.034 |
Neg logit contributions | |
cups | -0.021 |
frost | -0.019 |
twins | -0.019 |
sisters | -0.018 |
ugs | -0.017 |
Token | " |
Token ID | 366 |
Feature activation | -2.37e+00 |
Pos logit contributions | |
shortcut | +0.012 |
lot | +0.012 |
duel | +0.011 |
farewell | +0.011 |
Skip | +0.010 |
Neg logit contributions | |
cups | -0.015 |
frost | -0.015 |
el | -0.014 |
els | -0.014 |
ugs | -0.014 |
Token | I |
Token ID | 40 |
Feature activation | -1.10e-01 |
Pos logit contributions | |
shortcut | +0.073 |
lot | +0.071 |
duel | +0.068 |
unnecessary | +0.066 |
farewell | +0.066 |
Neg logit contributions | |
cups | -0.052 |
frost | -0.048 |
twins | -0.045 |
ugs | -0.044 |
sisters | -0.044 |
Token | 'll |
Token ID | 1183 |
Feature activation | -2.50e-01 |
Pos logit contributions | |
shortcut | +0.003 |
lot | +0.002 |
duel | +0.002 |
unnecessary | +0.002 |
farewell | +0.002 |
Neg logit contributions | |
cups | -0.003 |
frost | -0.003 |
twins | -0.003 |
sisters | -0.003 |
cloth | -0.003 |
Token | make |
Token ID | 787 |
Feature activation | -8.84e-01 |
Pos logit contributions | |
shortcut | +0.006 |
duel | +0.006 |
lot | +0.005 |
gall | +0.005 |
defeat | +0.005 |
Neg logit contributions | |
cups | -0.007 |
ugs | -0.007 |
ays | -0.007 |
frost | -0.007 |
folders | -0.007 |
Token | the |
Token ID | 262 |
Feature activation | -1.62e+00 |
Pos logit contributions | |
shortcut | +0.024 |
lot | +0.023 |
duel | +0.022 |
unnecessary | +0.021 |
farewell | +0.021 |
Neg logit contributions | |
cups | -0.023 |
frost | -0.021 |
twins | -0.020 |
ugs | -0.020 |
sisters | -0.020 |
Token | tail |
Token ID | 7894 |
Feature activation | -3.37e+00 |
Pos logit contributions | |
shortcut | +0.046 |
lot | +0.045 |
unnecessary | +0.043 |
duel | +0.042 |
Hop | +0.042 |
Neg logit contributions | |
cups | -0.042 |
twins | -0.039 |
frost | -0.039 |
sisters | -0.036 |
mug | -0.035 |
Token | , |
Token ID | 11 |
Feature activation | -9.00e-01 |
Pos logit contributions | |
shortcut | +0.084 |
lot | +0.083 |
duel | +0.077 |
unnecessary | +0.077 |
farewell | +0.075 |
Neg logit contributions | |
cups | -0.095 |
frost | -0.086 |
twins | -0.086 |
sisters | -0.084 |
cloth | -0.081 |
Token | you |
Token ID | 345 |
Feature activation | -1.60e+00 |
Pos logit contributions | |
shortcut | +0.024 |
lot | +0.023 |
duel | +0.022 |
unnecessary | +0.021 |
farewell | +0.021 |
Neg logit contributions | |
cups | -0.024 |
frost | -0.022 |
twins | -0.022 |
sisters | -0.021 |
cloth | -0.021 |
Token | make |
Token ID | 787 |
Feature activation | -1.17e+00 |
Pos logit contributions | |
shortcut | +0.045 |
lot | +0.045 |
unnecessary | +0.042 |
farewell | +0.041 |
duel | +0.041 |
Neg logit contributions | |
cups | -0.039 |
frost | -0.037 |
twins | -0.034 |
sisters | -0.033 |
icing | -0.032 |
Token | the |
Token ID | 262 |
Feature activation | -1.89e+00 |
Pos logit contributions | |
shortcut | +0.035 |
lot | +0.034 |
duel | +0.032 |
noble | +0.031 |
farewell | +0.031 |
Neg logit contributions | |
cups | -0.028 |
frost | -0.026 |
twins | -0.025 |
sisters | -0.024 |
el | -0.024 |
Token | fin |
Token ID | 957 |
Feature activation | -3.87e+00 |
Pos logit contributions | |
shortcut | +0.054 |
lot | +0.052 |
unnecessary | +0.049 |
duel | +0.048 |
Hop | +0.048 |
Neg logit contributions | |
cups | -0.049 |
twins | -0.046 |
frost | -0.046 |
sisters | -0.043 |
mug | -0.041 |
Token | . |
Token ID | 13 |
Feature activation | -1.71e+00 |
Pos logit contributions | |
cups | +0.010 |
frost | +0.008 |
twins | +0.008 |
ugs | +0.008 |
cloth | +0.008 |
Neg logit contributions | |
shortcut | -0.017 |
lot | -0.017 |
unnecessary | -0.017 |
duel | -0.017 |
noble | -0.016 |
Token | He |
Token ID | 679 |
Feature activation | -7.88e-01 |
Pos logit contributions | |
shortcut | +0.053 |
unnecessary | +0.051 |
duel | +0.050 |
lot | +0.049 |
gall | +0.048 |
Neg logit contributions | |
cups | -0.041 |
twins | -0.035 |
frost | -0.034 |
tub | -0.033 |
sisters | -0.033 |
Token | could |
Token ID | 714 |
Feature activation | -1.18e+00 |
Pos logit contributions | |
shortcut | +0.024 |
unnecessary | +0.023 |
duel | +0.023 |
lot | +0.022 |
ashore | +0.020 |
Neg logit contributions | |
cups | -0.019 |
frost | -0.017 |
sisters | -0.016 |
cloth | -0.016 |
twins | -0.015 |
Token | hear |
Token ID | 3285 |
Feature activation | -2.86e+00 |
Pos logit contributions | |
shortcut | +0.029 |
lot | +0.028 |
duel | +0.027 |
unnecessary | +0.027 |
noble | +0.025 |
Neg logit contributions | |
cups | -0.033 |
frost | -0.031 |
sisters | -0.030 |
twins | -0.030 |
cloth | -0.029 |
Token | the |
Token ID | 262 |
Feature activation | -1.45e+00 |
Pos logit contributions | |
lot | +0.108 |
shortcut | +0.108 |
duel | +0.105 |
rematch | +0.095 |
noble | +0.094 |
Neg logit contributions | |
cups | -0.052 |
frost | -0.051 |
twins | -0.044 |
candles | -0.043 |
icing | -0.039 |
Token | birds |
Token ID | 10087 |
Feature activation | -3.05e+00 |
Pos logit contributions | |
lot | +0.042 |
duel | +0.039 |
shortcut | +0.039 |
unnecessary | +0.036 |
noble | +0.035 |
Neg logit contributions | |
cups | -0.039 |
twins | -0.035 |
frost | -0.035 |
sisters | -0.033 |
mug | -0.033 |
Token | ch |
Token ID | 442 |
Feature activation | -3.14e-01 |
Pos logit contributions | |
shortcut | +0.087 |
lot | +0.086 |
duel | +0.081 |
unnecessary | +0.080 |
noble | +0.079 |
Neg logit contributions | |
cups | -0.077 |
frost | -0.070 |
el | -0.069 |
sisters | -0.069 |
twins | -0.068 |
Token | ir |
Token ID | 343 |
Feature activation | -9.90e-01 |
Pos logit contributions | |
shortcut | +0.010 |
lot | +0.010 |
unnecessary | +0.009 |
farewell | +0.009 |
further | +0.009 |
Neg logit contributions | |
cups | -0.006 |
frost | -0.006 |
twins | -0.006 |
sisters | -0.006 |
folders | -0.005 |
Token | ping |
Token ID | 13886 |
Feature activation | -1.65e+00 |
Pos logit contributions | |
shortcut | +0.028 |
duel | +0.026 |
lot | +0.025 |
noble | +0.025 |
unnecessary | +0.024 |
Neg logit contributions | |
cups | -0.025 |
el | -0.024 |
frost | -0.024 |
ugs | -0.023 |
twins | -0.023 |
Token | from |
Token ID | 422 |
Feature activation | -7.13e-01 |
Pos logit contributions | |
shortcut | +0.061 |
lot | +0.061 |
duel | +0.060 |
unnecessary | +0.058 |
noble | +0.058 |
Neg logit contributions | |
cups | -0.028 |
frost | -0.025 |
twins | -0.023 |
el | -0.023 |
sisters | -0.022 |
Token | the |
Token ID | 262 |
Feature activation | +1.41e-01 |
Pos logit contributions | |
lot | +0.023 |
shortcut | +0.022 |
duel | +0.022 |
noble | +0.020 |
unnecessary | +0.020 |
Neg logit contributions | |
cups | -0.016 |
frost | -0.015 |
twins | -0.014 |
el | -0.013 |
mug | -0.013 |
Token | lion |
Token ID | 18744 |
Feature activation | -9.98e-01 |
Pos logit contributions | |
shortcut | +0.020 |
lot | +0.019 |
duel | +0.019 |
unnecessary | +0.019 |
noble | +0.018 |
Neg logit contributions | |
cups | -0.015 |
frost | -0.014 |
twins | -0.014 |
sisters | -0.013 |
mug | -0.013 |
Token | . |
Token ID | 13 |
Feature activation | -1.47e+00 |
Pos logit contributions | |
shortcut | +0.033 |
lot | +0.032 |
duel | +0.031 |
unnecessary | +0.031 |
farewell | +0.030 |
Neg logit contributions | |
cups | -0.020 |
frost | -0.018 |
twins | -0.017 |
sisters | -0.016 |
ugs | -0.016 |
Token | He |
Token ID | 679 |
Feature activation | -2.35e+00 |
Pos logit contributions | |
shortcut | +0.042 |
lot | +0.039 |
unnecessary | +0.039 |
duel | +0.038 |
further | +0.038 |
Neg logit contributions | |
cups | -0.036 |
frost | -0.034 |
twins | -0.033 |
sisters | -0.031 |
ugs | -0.031 |
Token | liked |
Token ID | 8288 |
Feature activation | -2.37e+00 |
Pos logit contributions | |
shortcut | +0.084 |
unnecessary | +0.083 |
further | +0.080 |
nearer | +0.080 |
duel | +0.077 |
Neg logit contributions | |
cups | -0.039 |
twins | -0.038 |
frost | -0.038 |
sisters | -0.033 |
el | -0.032 |
Token | to |
Token ID | 284 |
Feature activation | -2.36e+00 |
Pos logit contributions | |
rated | +0.076 |
shortcut | +0.075 |
lot | +0.072 |
farewell | +0.071 |
nearer | +0.070 |
Neg logit contributions | |
cups | -0.049 |
twins | -0.047 |
frost | -0.044 |
sisters | -0.044 |
eggs | -0.040 |
Token | wander |
Token ID | 27776 |
Feature activation | -3.01e+00 |
Pos logit contributions | |
shortcut | +0.048 |
lot | +0.047 |
unnecessary | +0.044 |
duel | +0.042 |
further | +0.042 |
Neg logit contributions | |
cups | -0.076 |
frost | -0.072 |
sisters | -0.069 |
twins | -0.069 |
el | -0.067 |
Token | around |
Token ID | 1088 |
Feature activation | -1.66e+00 |
Pos logit contributions | |
lot | +0.088 |
duel | +0.088 |
shortcut | +0.088 |
unnecessary | +0.085 |
'd | +0.084 |
Neg logit contributions | |
cups | -0.066 |
twins | -0.057 |
frost | -0.054 |
folders | -0.053 |
ink | -0.052 |
Token | his |
Token ID | 465 |
Feature activation | -8.25e-01 |
Pos logit contributions | |
shortcut | +0.059 |
lot | +0.059 |
unnecessary | +0.058 |
farewell | +0.056 |
duel | +0.055 |
Neg logit contributions | |
cups | -0.028 |
twins | -0.025 |
el | -0.025 |
frost | -0.023 |
sisters | -0.022 |
Token | jungle |
Token ID | 20712 |
Feature activation | -3.46e-01 |
Pos logit contributions | |
shortcut | +0.024 |
nearer | +0.023 |
unnecessary | +0.022 |
farewell | +0.022 |
duel | +0.022 |
Neg logit contributions | |
cups | -0.020 |
twins | -0.019 |
el | -0.018 |
frost | -0.018 |
sisters | -0.017 |
Token | looking |
Token ID | 2045 |
Feature activation | -3.18e+00 |
Pos logit contributions | |
shortcut | +0.012 |
unnecessary | +0.012 |
lot | +0.011 |
farewell | +0.011 |
duel | +0.011 |
Neg logit contributions | |
cups | -0.007 |
twins | -0.006 |
sisters | -0.005 |
frost | -0.005 |
folders | -0.005 |
Token | for |
Token ID | 329 |
Feature activation | -1.67e+00 |
Pos logit contributions | |
lot | +0.128 |
duel | +0.126 |
shortcut | +0.126 |
chery | +0.118 |
Hop | +0.117 |
Neg logit contributions | |
cups | -0.049 |
twins | -0.043 |
frost | -0.035 |
sisters | -0.033 |
overs | -0.031 |
Token | happy |
Token ID | 3772 |
Feature activation | +5.94e-01 |
Pos logit contributions | |
cups | +0.008 |
twins | +0.007 |
frost | +0.007 |
sisters | +0.007 |
ugs | +0.007 |
Neg logit contributions | |
shortcut | -0.008 |
lot | -0.007 |
duel | -0.007 |
unnecessary | -0.007 |
hike | -0.007 |
Token | ! |
Token ID | 0 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
cups | +0.015 |
frost | +0.013 |
twins | +0.013 |
sisters | +0.013 |
ugs | +0.013 |
Neg logit contributions | |
shortcut | -0.017 |
lot | -0.016 |
duel | -0.016 |
unnecessary | -0.015 |
noble | -0.015 |
Token | She |
Token ID | 1375 |
Feature activation | -1.26e-02 |
Pos logit contributions | |
cups | +0.029 |
frost | +0.026 |
sisters | +0.025 |
els | +0.024 |
ugs | +0.024 |
Neg logit contributions | |
lot | -0.028 |
shortcut | -0.028 |
farewell | -0.026 |
duel | -0.026 |
gall | -0.025 |
Token | thanked |
Token ID | 26280 |
Feature activation | +5.44e-01 |
Pos logit contributions | |
shortcut | +0.000 |
unnecessary | +0.000 |
lot | +0.000 |
duel | +0.000 |
further | +0.000 |
Neg logit contributions | |
cups | -0.000 |
frost | -0.000 |
sisters | -0.000 |
twins | -0.000 |
el | -0.000 |
Token | the |
Token ID | 262 |
Feature activation | -8.77e-01 |
Pos logit contributions | |
cups | +0.012 |
frost | +0.011 |
twins | +0.010 |
sisters | +0.010 |
ugs | +0.010 |
Neg logit contributions | |
shortcut | -0.017 |
lot | -0.017 |
duel | -0.016 |
unnecessary | -0.015 |
noble | -0.015 |
Token | bears |
Token ID | 13062 |
Feature activation | -3.23e+00 |
Pos logit contributions | |
shortcut | +0.027 |
lot | +0.026 |
duel | +0.025 |
unnecessary | +0.024 |
farewell | +0.024 |
Neg logit contributions | |
cups | -0.020 |
frost | -0.018 |
twins | -0.017 |
sisters | -0.017 |
ugs | -0.017 |
Token | for |
Token ID | 329 |
Feature activation | +6.85e-01 |
Pos logit contributions | |
shortcut | +0.112 |
unnecessary | +0.109 |
lot | +0.106 |
duel | +0.106 |
Left | +0.098 |
Neg logit contributions | |
cups | -0.063 |
sisters | -0.057 |
twins | -0.057 |
frost | -0.053 |
mug | -0.049 |
Token | their |
Token ID | 511 |
Feature activation | -1.44e-01 |
Pos logit contributions | |
cups | +0.016 |
frost | +0.014 |
twins | +0.014 |
sisters | +0.014 |
ugs | +0.014 |
Neg logit contributions | |
shortcut | -0.020 |
lot | -0.020 |
duel | -0.019 |
unnecessary | -0.018 |
farewell | -0.018 |
Token | help |
Token ID | 1037 |
Feature activation | +6.37e-01 |
Pos logit contributions | |
lot | +0.004 |
shortcut | +0.004 |
higher | +0.003 |
Hop | +0.003 |
nearer | +0.003 |
Neg logit contributions | |
sisters | -0.004 |
twins | -0.004 |
cups | -0.004 |
folders | -0.003 |
frost | -0.003 |
Token | and |
Token ID | 290 |
Feature activation | -2.47e-01 |
Pos logit contributions | |
cups | +0.013 |
frost | +0.012 |
twins | +0.012 |
sisters | +0.012 |
ugs | +0.011 |
Neg logit contributions | |
shortcut | -0.020 |
lot | -0.020 |
duel | -0.019 |
unnecessary | -0.018 |
noble | -0.018 |
Token | they |
Token ID | 484 |
Feature activation | +2.98e-01 |
Pos logit contributions | |
shortcut | +0.007 |
lot | +0.007 |
duel | +0.007 |
unnecessary | +0.006 |
farewell | +0.006 |
Neg logit contributions | |
cups | -0.006 |
frost | -0.006 |
twins | -0.005 |
ugs | -0.005 |
sisters | -0.005 |