Token | looked |
Token ID | 3114 |
Feature activation | +3.86e-01 |
Pos logit contributions | |
wn | +0.021 |
aback | +0.020 |
coat | +0.017 |
swe | +0.017 |
ripe | +0.016 |
Neg logit contributions | |
glass | -0.041 |
orse | -0.040 |
generation | -0.037 |
work | -0.036 |
planet | -0.036 |
Token | through |
Token ID | 832 |
Feature activation | +1.02e+00 |
Pos logit contributions | |
orse | +0.019 |
glass | +0.018 |
whale | +0.018 |
work | +0.018 |
ship | +0.017 |
Neg logit contributions | |
wn | -0.022 |
coat | -0.021 |
aback | -0.019 |
arl | -0.018 |
ucker | -0.017 |
Token | the |
Token ID | 262 |
Feature activation | -7.10e-01 |
Pos logit contributions | |
glass | +0.072 |
orse | +0.071 |
work | +0.064 |
generation | +0.064 |
bike | +0.063 |
Neg logit contributions | |
aback | -0.041 |
wn | -0.039 |
swe | -0.035 |
like | -0.035 |
past | -0.034 |
Token | ey |
Token ID | 1926 |
Feature activation | -2.61e+00 |
Pos logit contributions | |
wn | +0.039 |
aback | +0.038 |
coat | +0.034 |
swe | +0.032 |
like | +0.031 |
Neg logit contributions | |
glass | -0.040 |
orse | -0.037 |
planet | -0.036 |
work | -0.035 |
whale | -0.034 |
Token | ep |
Token ID | 538 |
Feature activation | +1.24e+00 |
Pos logit contributions | |
wn | +0.124 |
aback | +0.118 |
coat | +0.107 |
swe | +0.105 |
like | +0.102 |
Neg logit contributions | |
glass | -0.161 |
orse | -0.159 |
generation | -0.149 |
work | -0.145 |
planet | -0.143 |
Token | iece |
Token ID | 8535 |
Feature activation | +9.59e+00 |
Pos logit contributions | |
orse | +0.073 |
glass | +0.072 |
generation | +0.069 |
planet | +0.065 |
work | +0.065 |
Neg logit contributions | |
wn | -0.061 |
aback | -0.060 |
like | -0.056 |
coat | -0.054 |
past | -0.053 |
Token | and |
Token ID | 290 |
Feature activation | -1.08e-01 |
Pos logit contributions | |
glass | +0.728 |
orse | +0.720 |
generation | +0.673 |
work | +0.662 |
bike | +0.647 |
Neg logit contributions | |
wn | -0.320 |
aback | -0.310 |
coat | -0.271 |
swe | -0.257 |
like | -0.254 |
Token | saw |
Token ID | 2497 |
Feature activation | +1.21e+00 |
Pos logit contributions | |
wn | +0.004 |
aback | +0.004 |
coat | +0.003 |
swe | +0.003 |
like | +0.003 |
Neg logit contributions | |
glass | -0.008 |
orse | -0.008 |
generation | -0.007 |
work | -0.007 |
planet | -0.007 |
Token | the |
Token ID | 262 |
Feature activation | -7.18e-01 |
Pos logit contributions | |
glass | +0.084 |
orse | +0.084 |
generation | +0.077 |
work | +0.073 |
bike | +0.073 |
Neg logit contributions | |
aback | -0.050 |
wn | -0.046 |
like | -0.044 |
sweet | -0.044 |
swe | -0.043 |
Token | tiny |
Token ID | 7009 |
Feature activation | -2.59e+00 |
Pos logit contributions | |
wn | +0.035 |
aback | +0.035 |
coat | +0.032 |
swe | +0.031 |
like | +0.030 |
Neg logit contributions | |
glass | -0.044 |
orse | -0.043 |
generation | -0.039 |
work | -0.038 |
bike | -0.037 |
Token | scales |
Token ID | 16252 |
Feature activation | -1.08e+00 |
Pos logit contributions | |
aback | +0.120 |
wn | +0.119 |
coat | +0.109 |
swe | +0.108 |
like | +0.106 |
Neg logit contributions | |
glass | -0.163 |
orse | -0.159 |
generation | -0.145 |
work | -0.140 |
bike | -0.139 |
Token | looked |
Token ID | 3114 |
Feature activation | +3.59e-01 |
Pos logit contributions | |
wn | +0.002 |
aback | +0.002 |
swe | +0.001 |
coat | +0.001 |
like | +0.001 |
Neg logit contributions | |
glass | -0.004 |
orse | -0.004 |
generation | -0.004 |
work | -0.004 |
planet | -0.004 |
Token | through |
Token ID | 832 |
Feature activation | +1.07e+00 |
Pos logit contributions | |
glass | +0.019 |
orse | +0.019 |
work | +0.017 |
whale | +0.017 |
ship | +0.017 |
Neg logit contributions | |
wn | -0.019 |
coat | -0.018 |
aback | -0.018 |
blonde | -0.015 |
ucker | -0.015 |
Token | the |
Token ID | 262 |
Feature activation | -8.13e-01 |
Pos logit contributions | |
glass | +0.068 |
orse | +0.068 |
generation | +0.061 |
work | +0.061 |
planet | +0.059 |
Neg logit contributions | |
wn | -0.049 |
aback | -0.049 |
coat | -0.043 |
swe | -0.043 |
like | -0.042 |
Token | ey |
Token ID | 1926 |
Feature activation | -2.90e+00 |
Pos logit contributions | |
wn | +0.045 |
aback | +0.044 |
coat | +0.041 |
like | +0.036 |
swe | +0.036 |
Neg logit contributions | |
glass | -0.043 |
planet | -0.040 |
orse | -0.039 |
whale | -0.039 |
work | -0.038 |
Token | ep |
Token ID | 538 |
Feature activation | +9.61e-01 |
Pos logit contributions | |
wn | +0.141 |
aback | +0.136 |
coat | +0.123 |
swe | +0.121 |
like | +0.117 |
Neg logit contributions | |
glass | -0.176 |
orse | -0.174 |
generation | -0.161 |
work | -0.158 |
planet | -0.154 |
Token | iece |
Token ID | 8535 |
Feature activation | +9.38e+00 |
Pos logit contributions | |
orse | +0.062 |
glass | +0.061 |
generation | +0.058 |
work | +0.056 |
planet | +0.055 |
Neg logit contributions | |
wn | -0.042 |
aback | -0.041 |
like | -0.037 |
coat | -0.037 |
swe | -0.035 |
Token | . |
Token ID | 13 |
Feature activation | -5.00e-01 |
Pos logit contributions | |
glass | +0.702 |
orse | +0.697 |
generation | +0.655 |
work | +0.642 |
bike | +0.631 |
Neg logit contributions | |
wn | -0.317 |
aback | -0.300 |
coat | -0.271 |
like | -0.257 |
swe | -0.255 |
Token | He |
Token ID | 679 |
Feature activation | -5.79e-01 |
Pos logit contributions | |
wn | +0.025 |
aback | +0.024 |
coat | +0.022 |
swe | +0.022 |
ripe | +0.021 |
Neg logit contributions | |
glass | -0.031 |
orse | -0.030 |
generation | -0.027 |
work | -0.026 |
planet | -0.026 |
Token | saw |
Token ID | 2497 |
Feature activation | +1.24e+00 |
Pos logit contributions | |
wn | +0.020 |
aback | +0.019 |
coat | +0.017 |
swe | +0.016 |
like | +0.016 |
Neg logit contributions | |
glass | -0.043 |
orse | -0.043 |
generation | -0.040 |
work | -0.040 |
planet | -0.039 |
Token | the |
Token ID | 262 |
Feature activation | -7.99e-01 |
Pos logit contributions | |
glass | +0.084 |
orse | +0.084 |
generation | +0.078 |
work | +0.075 |
bike | +0.073 |
Neg logit contributions | |
aback | -0.052 |
wn | -0.050 |
like | -0.046 |
swe | -0.046 |
sweet | -0.046 |
Token | tiny |
Token ID | 7009 |
Feature activation | -2.54e+00 |
Pos logit contributions | |
wn | +0.039 |
aback | +0.039 |
coat | +0.034 |
swe | +0.034 |
like | +0.032 |
Neg logit contributions | |
glass | -0.049 |
orse | -0.048 |
generation | -0.044 |
work | -0.044 |
planet | -0.043 |
Token | voice |
Token ID | 3809 |
Feature activation | -8.95e-01 |
Pos logit contributions | |
wn | +0.000 |
aback | +0.000 |
coat | +0.000 |
swe | +0.000 |
like | +0.000 |
Neg logit contributions | |
glass | -0.001 |
orse | -0.001 |
generation | -0.000 |
work | -0.000 |
planet | -0.000 |
Token | that |
Token ID | 326 |
Feature activation | +6.49e-01 |
Pos logit contributions | |
aback | +0.031 |
wn | +0.030 |
like | +0.027 |
swe | +0.026 |
coat | +0.025 |
Neg logit contributions | |
glass | -0.066 |
orse | -0.065 |
generation | -0.062 |
planet | -0.059 |
work | -0.059 |
Token | said |
Token ID | 531 |
Feature activation | -1.06e+00 |
Pos logit contributions | |
glass | +0.050 |
orse | +0.049 |
generation | +0.046 |
work | +0.045 |
planet | +0.044 |
Neg logit contributions | |
aback | -0.021 |
wn | -0.020 |
coat | -0.017 |
swe | -0.017 |
like | -0.016 |
Token | : |
Token ID | 25 |
Feature activation | +1.77e+00 |
Pos logit contributions | |
aback | +0.038 |
wn | +0.035 |
like | +0.032 |
swe | +0.031 |
coat | +0.031 |
Neg logit contributions | |
glass | -0.077 |
orse | -0.076 |
generation | -0.074 |
work | -0.070 |
planet | -0.069 |
Token | " |
Token ID | 366 |
Feature activation | -1.52e+00 |
Pos logit contributions | |
glass | +0.135 |
orse | +0.127 |
generation | +0.124 |
work | +0.120 |
planet | +0.119 |
Neg logit contributions | |
aback | -0.058 |
like | -0.052 |
sweet | -0.050 |
wn | -0.049 |
past | -0.047 |
Token | Warning |
Token ID | 20361 |
Feature activation | +9.12e+00 |
Pos logit contributions | |
aback | +0.072 |
wn | +0.072 |
coat | +0.064 |
swe | +0.062 |
like | +0.060 |
Neg logit contributions | |
glass | -0.096 |
orse | -0.095 |
generation | -0.085 |
work | -0.085 |
planet | -0.083 |
Token | ! |
Token ID | 0 |
Feature activation | +1.86e+00 |
Pos logit contributions | |
glass | +0.622 |
orse | +0.618 |
generation | +0.569 |
work | +0.557 |
planet | +0.541 |
Neg logit contributions | |
aback | -0.380 |
wn | -0.378 |
coat | -0.329 |
swe | -0.322 |
like | -0.317 |
Token | It |
Token ID | 632 |
Feature activation | +4.63e-01 |
Pos logit contributions | |
glass | +0.113 |
orse | +0.107 |
generation | +0.098 |
work | +0.097 |
planet | +0.097 |
Neg logit contributions | |
aback | -0.095 |
wn | -0.082 |
like | -0.076 |
coat | -0.072 |
sweet | -0.071 |
Token | is |
Token ID | 318 |
Feature activation | -1.12e-03 |
Pos logit contributions | |
glass | +0.028 |
orse | +0.027 |
generation | +0.025 |
work | +0.025 |
planet | +0.024 |
Neg logit contributions | |
wn | -0.023 |
aback | -0.023 |
coat | -0.021 |
swe | -0.020 |
ripe | -0.019 |
Token | not |
Token ID | 407 |
Feature activation | -2.84e-02 |
Pos logit contributions | |
wn | +0.000 |
aback | +0.000 |
coat | +0.000 |
swe | +0.000 |
ripe | +0.000 |
Neg logit contributions | |
glass | -0.000 |
orse | -0.000 |
generation | -0.000 |
work | -0.000 |
planet | -0.000 |
Token | safe |
Token ID | 3338 |
Feature activation | -1.20e+00 |
Pos logit contributions | |
aback | +0.001 |
wn | +0.001 |
coat | +0.001 |
ucker | +0.001 |
swe | +0.001 |
Neg logit contributions | |
glass | -0.002 |
orse | -0.002 |
generation | -0.002 |
work | -0.002 |
planet | -0.002 |
Token | . |
Token ID | 13 |
Feature activation | -4.51e-01 |
Pos logit contributions | |
glass | +0.014 |
orse | +0.014 |
generation | +0.013 |
work | +0.013 |
planet | +0.013 |
Neg logit contributions | |
wn | -0.007 |
aback | -0.007 |
coat | -0.006 |
swe | -0.006 |
like | -0.006 |
Token | It |
Token ID | 632 |
Feature activation | +4.60e-01 |
Pos logit contributions | |
aback | +0.018 |
wn | +0.016 |
like | +0.016 |
sweet | +0.014 |
past | +0.014 |
Neg logit contributions | |
orse | -0.032 |
glass | -0.031 |
bike | -0.029 |
fer | -0.029 |
ich | -0.029 |
Token | says |
Token ID | 1139 |
Feature activation | +7.98e-02 |
Pos logit contributions | |
glass | +0.032 |
orse | +0.031 |
generation | +0.029 |
work | +0.029 |
planet | +0.028 |
Neg logit contributions | |
wn | -0.019 |
aback | -0.018 |
coat | -0.016 |
swe | -0.016 |
like | -0.015 |
Token | : |
Token ID | 25 |
Feature activation | +1.78e+00 |
Pos logit contributions | |
generation | +0.005 |
orse | +0.005 |
glass | +0.005 |
bike | +0.005 |
harbour | +0.005 |
Neg logit contributions | |
like | -0.003 |
aback | -0.003 |
sweet | -0.003 |
wn | -0.003 |
nice | -0.003 |
Token | " |
Token ID | 366 |
Feature activation | -1.55e+00 |
Pos logit contributions | |
glass | +0.138 |
orse | +0.134 |
generation | +0.132 |
bike | +0.127 |
work | +0.127 |
Neg logit contributions | |
aback | -0.046 |
like | -0.045 |
sweet | -0.045 |
past | -0.043 |
wn | -0.042 |
Token | Warning |
Token ID | 20361 |
Feature activation | +9.03e+00 |
Pos logit contributions | |
aback | +0.069 |
wn | +0.069 |
like | +0.062 |
past | +0.061 |
coat | +0.056 |
Neg logit contributions | |
orse | -0.099 |
glass | -0.090 |
planet | -0.083 |
crew | -0.083 |
bike | -0.083 |
Token | : |
Token ID | 25 |
Feature activation | +1.78e+00 |
Pos logit contributions | |
glass | +0.500 |
orse | +0.491 |
generation | +0.443 |
work | +0.435 |
planet | +0.421 |
Neg logit contributions | |
wn | -0.494 |
aback | -0.489 |
coat | -0.446 |
swe | -0.437 |
like | -0.421 |
Token | Do |
Token ID | 2141 |
Feature activation | +1.19e+00 |
Pos logit contributions | |
orse | +0.108 |
glass | +0.106 |
generation | +0.099 |
fer | +0.097 |
bike | +0.097 |
Neg logit contributions | |
aback | -0.079 |
sweet | -0.078 |
wn | -0.077 |
like | -0.077 |
past | -0.075 |
Token | not |
Token ID | 407 |
Feature activation | -1.55e-02 |
Pos logit contributions | |
glass | +0.083 |
orse | +0.080 |
generation | +0.080 |
work | +0.078 |
fer | +0.077 |
Neg logit contributions | |
fun | -0.041 |
wn | -0.041 |
sweet | -0.041 |
like | -0.041 |
past | -0.038 |
Token | touch |
Token ID | 3638 |
Feature activation | +1.32e+00 |
Pos logit contributions | |
wn | +0.001 |
aback | +0.001 |
like | +0.000 |
swe | +0.000 |
coat | +0.000 |
Neg logit contributions | |
glass | -0.001 |
orse | -0.001 |
generation | -0.001 |
work | -0.001 |
bike | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | -4.47e-01 |
Pos logit contributions | |
orse | +0.087 |
glass | +0.086 |
generation | +0.078 |
work | +0.075 |
fer | +0.074 |
Neg logit contributions | |
aback | -0.058 |
wn | -0.057 |
like | -0.052 |
swe | -0.051 |
sweet | -0.049 |
Token | sign |
Token ID | 1051 |
Feature activation | -6.07e-02 |
Pos logit contributions | |
wn | +0.008 |
aback | +0.008 |
coat | +0.007 |
swe | +0.007 |
like | +0.006 |
Neg logit contributions | |
glass | -0.016 |
orse | -0.016 |
generation | -0.015 |
work | -0.015 |
planet | -0.015 |
Token | that |
Token ID | 326 |
Feature activation | +5.97e-01 |
Pos logit contributions | |
wn | +0.003 |
aback | +0.003 |
coat | +0.003 |
like | +0.003 |
swe | +0.002 |
Neg logit contributions | |
orse | -0.004 |
glass | -0.004 |
generation | -0.004 |
work | -0.003 |
planet | -0.003 |
Token | said |
Token ID | 531 |
Feature activation | -1.02e+00 |
Pos logit contributions | |
glass | +0.049 |
orse | +0.049 |
generation | +0.046 |
work | +0.045 |
planet | +0.044 |
Neg logit contributions | |
aback | -0.016 |
wn | -0.015 |
coat | -0.013 |
swe | -0.012 |
like | -0.012 |
Token | : |
Token ID | 25 |
Feature activation | +1.87e+00 |
Pos logit contributions | |
aback | +0.034 |
wn | +0.032 |
like | +0.031 |
sweet | +0.030 |
coat | +0.028 |
Neg logit contributions | |
orse | -0.075 |
glass | -0.074 |
generation | -0.073 |
work | -0.068 |
bike | -0.067 |
Token | " |
Token ID | 366 |
Feature activation | -1.38e+00 |
Pos logit contributions | |
glass | +0.149 |
orse | +0.145 |
generation | +0.140 |
work | +0.135 |
planet | +0.132 |
Neg logit contributions | |
sweet | -0.046 |
aback | -0.046 |
past | -0.045 |
like | -0.045 |
nice | -0.042 |
Token | Warning |
Token ID | 20361 |
Feature activation | +9.03e+00 |
Pos logit contributions | |
wn | +0.068 |
aback | +0.067 |
coat | +0.060 |
swe | +0.057 |
like | +0.057 |
Neg logit contributions | |
orse | -0.084 |
glass | -0.083 |
generation | -0.074 |
work | -0.074 |
planet | -0.072 |
Token | ! |
Token ID | 0 |
Feature activation | +1.88e+00 |
Pos logit contributions | |
glass | +0.589 |
orse | +0.585 |
generation | +0.535 |
work | +0.525 |
planet | +0.511 |
Neg logit contributions | |
wn | -0.403 |
aback | -0.399 |
coat | -0.354 |
swe | -0.345 |
like | -0.335 |
Token | There |
Token ID | 1318 |
Feature activation | -3.75e-01 |
Pos logit contributions | |
orse | +0.116 |
glass | +0.113 |
field | +0.100 |
generation | +0.098 |
fer | +0.097 |
Neg logit contributions | |
aback | -0.090 |
wn | -0.079 |
like | -0.074 |
past | -0.072 |
sweet | -0.072 |
Token | is |
Token ID | 318 |
Feature activation | +1.49e-01 |
Pos logit contributions | |
aback | +0.014 |
wn | +0.014 |
like | +0.012 |
swe | +0.012 |
sweet | +0.012 |
Neg logit contributions | |
orse | -0.027 |
glass | -0.027 |
generation | -0.024 |
fer | -0.024 |
work | -0.023 |
Token | bad |
Token ID | 2089 |
Feature activation | -2.53e-01 |
Pos logit contributions | |
orse | +0.011 |
glass | +0.010 |
fer | +0.010 |
generation | +0.010 |
work | +0.009 |
Neg logit contributions | |
aback | -0.005 |
wn | -0.005 |
like | -0.005 |
past | -0.005 |
fun | -0.005 |
Token | news |
Token ID | 1705 |
Feature activation | +3.44e+00 |
Pos logit contributions | |
wn | +0.014 |
aback | +0.014 |
coat | +0.013 |
swe | +0.013 |
ripe | +0.012 |
Neg logit contributions | |
glass | -0.014 |
orse | -0.013 |
generation | -0.012 |
work | -0.012 |
planet | -0.012 |
Token | A |
Token ID | 317 |
Feature activation | +1.84e+00 |
Pos logit contributions | |
wn | +0.017 |
aback | +0.017 |
coat | +0.015 |
swe | +0.015 |
ripe | +0.014 |
Neg logit contributions | |
glass | -0.021 |
orse | -0.021 |
generation | -0.019 |
work | -0.018 |
planet | -0.018 |
Token | voice |
Token ID | 3809 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
glass | +0.121 |
orse | +0.120 |
generation | +0.111 |
work | +0.108 |
bike | +0.105 |
Neg logit contributions | |
wn | -0.079 |
aback | -0.078 |
coat | -0.070 |
swe | -0.069 |
like | -0.066 |
Token | said |
Token ID | 531 |
Feature activation | -1.09e+00 |
Pos logit contributions | |
aback | +0.036 |
wn | +0.036 |
coat | +0.031 |
swe | +0.030 |
like | +0.030 |
Neg logit contributions | |
glass | -0.077 |
orse | -0.077 |
generation | -0.072 |
work | -0.070 |
planet | -0.068 |
Token | , |
Token ID | 11 |
Feature activation | -4.19e-02 |
Pos logit contributions | |
aback | +0.029 |
like | +0.023 |
sweet | +0.022 |
wn | +0.021 |
swe | +0.020 |
Neg logit contributions | |
generation | -0.091 |
orse | -0.089 |
glass | -0.089 |
bike | -0.086 |
work | -0.084 |
Token | " |
Token ID | 366 |
Feature activation | -1.56e+00 |
Pos logit contributions | |
aback | +0.002 |
wn | +0.002 |
swe | +0.001 |
coat | +0.001 |
like | +0.001 |
Neg logit contributions | |
glass | -0.003 |
orse | -0.003 |
generation | -0.003 |
work | -0.003 |
planet | -0.003 |
Token | Warning |
Token ID | 20361 |
Feature activation | +8.94e+00 |
Pos logit contributions | |
aback | +0.064 |
wn | +0.060 |
past | +0.052 |
like | +0.052 |
swe | +0.049 |
Neg logit contributions | |
orse | -0.112 |
glass | -0.103 |
planet | -0.097 |
generation | -0.097 |
work | -0.096 |
Token | , |
Token ID | 11 |
Feature activation | -4.81e-02 |
Pos logit contributions | |
glass | +0.591 |
orse | +0.590 |
generation | +0.541 |
work | +0.531 |
planet | +0.515 |
Neg logit contributions | |
wn | -0.389 |
aback | -0.387 |
coat | -0.339 |
swe | -0.332 |
like | -0.327 |
Token | warning |
Token ID | 6509 |
Feature activation | +1.06e+00 |
Pos logit contributions | |
sweet | +0.002 |
like | +0.002 |
aback | +0.002 |
wn | +0.002 |
nice | +0.002 |
Neg logit contributions | |
orse | -0.003 |
glass | -0.003 |
generation | -0.003 |
field | -0.003 |
bike | -0.003 |
Token | , |
Token ID | 11 |
Feature activation | -4.66e-02 |
Pos logit contributions | |
glass | +0.069 |
orse | +0.066 |
work | +0.063 |
generation | +0.062 |
planet | +0.061 |
Neg logit contributions | |
wn | -0.047 |
aback | -0.045 |
coat | -0.042 |
swe | -0.039 |
ripe | -0.038 |
Token | bomb |
Token ID | 5194 |
Feature activation | +5.19e+00 |
Pos logit contributions | |
aback | +0.002 |
wn | +0.002 |
like | +0.002 |
sweet | +0.002 |
swe | +0.002 |
Neg logit contributions | |
orse | -0.003 |
glass | -0.003 |
generation | -0.003 |
work | -0.003 |
field | -0.003 |
Token | activated |
Token ID | 13906 |
Feature activation | -2.55e-01 |
Pos logit contributions | |
glass | +0.319 |
orse | +0.306 |
work | +0.298 |
planet | +0.277 |
generation | +0.273 |
Neg logit contributions | |
wn | -0.257 |
aback | -0.247 |
coat | -0.226 |
ripe | -0.223 |
swe | -0.220 |
Token | she |
Token ID | 673 |
Feature activation | -8.89e-01 |
Pos logit contributions | |
glass | +0.006 |
orse | +0.006 |
generation | +0.005 |
work | +0.005 |
planet | +0.005 |
Neg logit contributions | |
wn | -0.004 |
aback | -0.003 |
coat | -0.003 |
swe | -0.003 |
ripe | -0.003 |
Token | always |
Token ID | 1464 |
Feature activation | -3.67e+00 |
Pos logit contributions | |
aback | +0.030 |
wn | +0.029 |
like | +0.026 |
coat | +0.025 |
swe | +0.024 |
Neg logit contributions | |
glass | -0.067 |
orse | -0.066 |
generation | -0.061 |
work | -0.060 |
planet | -0.059 |
Token | wore |
Token ID | 12408 |
Feature activation | -1.14e+00 |
Pos logit contributions | |
wn | +0.148 |
aback | +0.146 |
coat | +0.127 |
swe | +0.124 |
like | +0.120 |
Neg logit contributions | |
glass | -0.257 |
orse | -0.253 |
generation | -0.233 |
work | -0.229 |
planet | -0.225 |
Token | her |
Token ID | 607 |
Feature activation | -1.53e+00 |
Pos logit contributions | |
wn | +0.048 |
aback | +0.047 |
coat | +0.046 |
ucker | +0.042 |
inters | +0.040 |
Neg logit contributions | |
glass | -0.075 |
generation | -0.069 |
orse | -0.069 |
planet | -0.065 |
work | -0.064 |
Token | fan |
Token ID | 4336 |
Feature activation | +1.15e-02 |
Pos logit contributions | |
wn | +0.059 |
aback | +0.058 |
swe | +0.051 |
coat | +0.051 |
like | +0.050 |
Neg logit contributions | |
glass | -0.108 |
orse | -0.108 |
generation | -0.099 |
work | -0.097 |
bike | -0.096 |
Token | ci |
Token ID | 979 |
Feature activation | +8.61e+00 |
Pos logit contributions | |
orse | +0.001 |
glass | +0.001 |
generation | +0.001 |
work | +0.001 |
planet | +0.001 |
Neg logit contributions | |
wn | -0.000 |
aback | -0.000 |
coat | -0.000 |
swe | -0.000 |
like | -0.000 |
Token | est |
Token ID | 395 |
Feature activation | -2.08e-02 |
Pos logit contributions | |
glass | +0.600 |
orse | +0.589 |
generation | +0.550 |
work | +0.544 |
planet | +0.528 |
Neg logit contributions | |
wn | -0.343 |
aback | -0.334 |
coat | -0.297 |
swe | -0.287 |
ripe | -0.277 |
Token | dress |
Token ID | 6576 |
Feature activation | +6.78e-01 |
Pos logit contributions | |
wn | +0.001 |
aback | +0.001 |
coat | +0.001 |
swe | +0.001 |
ripe | +0.001 |
Neg logit contributions | |
glass | -0.001 |
orse | -0.001 |
generation | -0.001 |
work | -0.001 |
planet | -0.001 |
Token | to |
Token ID | 284 |
Feature activation | -3.51e-01 |
Pos logit contributions | |
orse | +0.049 |
glass | +0.049 |
generation | +0.045 |
work | +0.044 |
planet | +0.043 |
Neg logit contributions | |
wn | -0.025 |
aback | -0.024 |
coat | -0.023 |
like | -0.021 |
swe | -0.021 |
Token | the |
Token ID | 262 |
Feature activation | -6.69e-01 |
Pos logit contributions | |
wn | +0.014 |
aback | +0.013 |
like | +0.013 |
swe | +0.012 |
sweet | +0.012 |
Neg logit contributions | |
glass | -0.024 |
orse | -0.024 |
work | -0.022 |
generation | -0.022 |
bike | -0.022 |
Token | museum |
Token ID | 13257 |
Feature activation | +3.01e+00 |
Pos logit contributions | |
like | +0.034 |
fun | +0.033 |
past | +0.033 |
wn | +0.033 |
sweet | +0.032 |
Neg logit contributions | |
orse | -0.040 |
glass | -0.037 |
bike | -0.033 |
work | -0.031 |
fer | -0.031 |
Token | dressed |
Token ID | 12049 |
Feature activation | -4.27e-01 |
Pos logit contributions | |
wn | +0.023 |
aback | +0.022 |
coat | +0.020 |
swe | +0.019 |
like | +0.018 |
Neg logit contributions | |
glass | -0.037 |
orse | -0.037 |
generation | -0.034 |
work | -0.034 |
planet | -0.033 |
Token | up |
Token ID | 510 |
Feature activation | +7.47e-01 |
Pos logit contributions | |
wn | +0.016 |
aback | +0.016 |
coat | +0.014 |
swe | +0.013 |
ripe | +0.013 |
Neg logit contributions | |
glass | -0.031 |
orse | -0.030 |
generation | -0.028 |
work | -0.028 |
planet | -0.027 |
Token | in |
Token ID | 287 |
Feature activation | +8.00e-02 |
Pos logit contributions | |
glass | +0.057 |
orse | +0.057 |
generation | +0.052 |
work | +0.051 |
planet | +0.050 |
Neg logit contributions | |
aback | -0.025 |
wn | -0.025 |
coat | -0.022 |
swe | -0.021 |
like | -0.021 |
Token | her |
Token ID | 607 |
Feature activation | -1.64e+00 |
Pos logit contributions | |
glass | +0.006 |
orse | +0.006 |
generation | +0.005 |
work | +0.005 |
planet | +0.005 |
Neg logit contributions | |
wn | -0.003 |
aback | -0.003 |
coat | -0.003 |
swe | -0.003 |
like | -0.003 |
Token | fan |
Token ID | 4336 |
Feature activation | -1.04e-01 |
Pos logit contributions | |
wn | +0.057 |
swe | +0.056 |
aback | +0.055 |
coat | +0.054 |
sweet | +0.054 |
Neg logit contributions | |
orse | -0.120 |
glass | -0.119 |
generation | -0.110 |
work | -0.106 |
bike | -0.106 |
Token | ci |
Token ID | 979 |
Feature activation | +8.55e+00 |
Pos logit contributions | |
coat | +0.003 |
wn | +0.003 |
aback | +0.003 |
swe | +0.003 |
like | +0.003 |
Neg logit contributions | |
orse | -0.008 |
glass | -0.008 |
generation | -0.007 |
Whale | -0.007 |
fer | -0.007 |
Token | est |
Token ID | 395 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
glass | +0.645 |
orse | +0.640 |
generation | +0.584 |
work | +0.575 |
planet | +0.564 |
Neg logit contributions | |
wn | -0.299 |
aback | -0.298 |
coat | -0.257 |
swe | -0.255 |
like | -0.233 |
Token | clothes |
Token ID | 8242 |
Feature activation | -3.22e+00 |
Pos logit contributions | |
wn | +0.006 |
aback | +0.006 |
coat | +0.005 |
swe | +0.005 |
ripe | +0.005 |
Neg logit contributions | |
glass | -0.009 |
orse | -0.009 |
generation | -0.008 |
work | -0.008 |
planet | -0.008 |
Token | and |
Token ID | 290 |
Feature activation | -3.20e-02 |
Pos logit contributions | |
wn | +0.110 |
aback | +0.107 |
coat | +0.098 |
like | +0.091 |
swe | +0.090 |
Neg logit contributions | |
glass | -0.242 |
orse | -0.240 |
generation | -0.224 |
work | -0.216 |
planet | -0.215 |
Token | asked |
Token ID | 1965 |
Feature activation | -3.44e+00 |
Pos logit contributions | |
wn | +0.001 |
aback | +0.001 |
coat | +0.001 |
swe | +0.001 |
like | +0.001 |
Neg logit contributions | |
glass | -0.002 |
orse | -0.002 |
generation | -0.002 |
work | -0.002 |
planet | -0.002 |
Token | her |
Token ID | 607 |
Feature activation | -1.64e+00 |
Pos logit contributions | |
wn | +0.135 |
aback | +0.128 |
coat | +0.118 |
swe | +0.107 |
ripe | +0.102 |
Neg logit contributions | |
glass | -0.244 |
orse | -0.239 |
generation | -0.227 |
work | -0.221 |
planet | -0.217 |
Token | to |
Token ID | 284 |
Feature activation | -4.11e-01 |
Pos logit contributions | |
glass | +0.122 |
orse | +0.120 |
generation | +0.112 |
work | +0.111 |
planet | +0.109 |
Neg logit contributions | |
wn | -0.045 |
aback | -0.044 |
coat | -0.037 |
swe | -0.035 |
like | -0.033 |
Token | put |
Token ID | 1234 |
Feature activation | -2.39e+00 |
Pos logit contributions | |
wn | +0.017 |
aback | +0.017 |
coat | +0.015 |
swe | +0.015 |
ripe | +0.014 |
Neg logit contributions | |
glass | -0.028 |
orse | -0.027 |
generation | -0.025 |
work | -0.025 |
planet | -0.025 |
Token | on |
Token ID | 319 |
Feature activation | +1.03e+00 |
Pos logit contributions | |
wn | +0.132 |
aback | +0.130 |
coat | +0.126 |
swe | +0.116 |
inters | +0.114 |
Neg logit contributions | |
glass | -0.123 |
orse | -0.115 |
work | -0.107 |
generation | -0.107 |
whale | -0.095 |
Token | her |
Token ID | 607 |
Feature activation | -1.57e+00 |
Pos logit contributions | |
glass | +0.068 |
orse | +0.067 |
generation | +0.063 |
work | +0.061 |
planet | +0.059 |
Neg logit contributions | |
wn | -0.044 |
aback | -0.043 |
coat | -0.039 |
swe | -0.036 |
like | -0.034 |
Token | fan |
Token ID | 4336 |
Feature activation | -3.34e-02 |
Pos logit contributions | |
aback | +0.063 |
wn | +0.063 |
coat | +0.053 |
swe | +0.049 |
like | +0.048 |
Neg logit contributions | |
glass | -0.110 |
orse | -0.109 |
work | -0.098 |
generation | -0.098 |
planet | -0.096 |
Token | ci |
Token ID | 979 |
Feature activation | +8.55e+00 |
Pos logit contributions | |
wn | +0.002 |
aback | +0.002 |
coat | +0.002 |
swe | +0.002 |
like | +0.002 |
Neg logit contributions | |
orse | -0.001 |
glass | -0.001 |
generation | -0.001 |
planet | -0.001 |
work | -0.001 |
Token | est |
Token ID | 395 |
Feature activation | -1.33e-01 |
Pos logit contributions | |
glass | +0.585 |
orse | +0.574 |
generation | +0.533 |
work | +0.527 |
planet | +0.511 |
Neg logit contributions | |
wn | -0.352 |
aback | -0.347 |
coat | -0.302 |
swe | -0.292 |
ripe | -0.290 |
Token | sunglasses |
Token ID | 36251 |
Feature activation | -2.32e+00 |
Pos logit contributions | |
aback | +0.006 |
wn | +0.006 |
coat | +0.005 |
swe | +0.005 |
ripe | +0.005 |
Neg logit contributions | |
glass | -0.009 |
orse | -0.008 |
generation | -0.007 |
work | -0.007 |
planet | -0.007 |
Token | . |
Token ID | 13 |
Feature activation | +9.00e-03 |
Pos logit contributions | |
wn | +0.092 |
aback | +0.091 |
coat | +0.078 |
swe | +0.077 |
ripe | +0.072 |
Neg logit contributions | |
glass | -0.164 |
orse | -0.161 |
generation | -0.148 |
work | -0.147 |
planet | -0.143 |
Token |
|
Token ID | 198 |
Feature activation | +4.68e-02 |
Pos logit contributions | |
orse | +0.001 |
glass | +0.000 |
generation | +0.000 |
work | +0.000 |
planet | +0.000 |
Neg logit contributions | |
wn | -0.000 |
aback | -0.000 |
swe | -0.000 |
coat | -0.000 |
ripe | -0.000 |
Token |
|
Token ID | 198 |
Feature activation | +3.21e-02 |
Pos logit contributions | |
glass | +0.003 |
orse | +0.003 |
generation | +0.003 |
work | +0.003 |
planet | +0.002 |
Neg logit contributions | |
wn | -0.002 |
aback | -0.002 |
coat | -0.002 |
swe | -0.002 |
ripe | -0.002 |
Token | the |
Token ID | 262 |
Feature activation | -6.85e-01 |
Pos logit contributions | |
glass | +0.010 |
orse | +0.010 |
generation | +0.009 |
planet | +0.009 |
fer | +0.009 |
Neg logit contributions | |
wn | -0.003 |
aback | -0.003 |
paid | -0.003 |
coat | -0.003 |
past | -0.002 |
Token | daughter |
Token ID | 4957 |
Feature activation | +2.59e+00 |
Pos logit contributions | |
wn | +0.036 |
swe | +0.036 |
aback | +0.035 |
coat | +0.035 |
sweet | +0.035 |
Neg logit contributions | |
orse | -0.040 |
glass | -0.038 |
generation | -0.034 |
work | -0.033 |
Whale | -0.032 |
Token | found |
Token ID | 1043 |
Feature activation | -2.39e-01 |
Pos logit contributions | |
glass | +0.185 |
orse | +0.184 |
generation | +0.170 |
work | +0.167 |
planet | +0.164 |
Neg logit contributions | |
wn | -0.096 |
aback | -0.095 |
coat | -0.087 |
swe | -0.084 |
like | -0.080 |
Token | the |
Token ID | 262 |
Feature activation | -6.69e-01 |
Pos logit contributions | |
wn | +0.011 |
aback | +0.011 |
coat | +0.010 |
swe | +0.009 |
like | +0.009 |
Neg logit contributions | |
glass | -0.016 |
orse | -0.015 |
generation | -0.014 |
work | -0.014 |
planet | -0.013 |
Token | fan |
Token ID | 4336 |
Feature activation | +1.76e-02 |
Pos logit contributions | |
wn | +0.027 |
aback | +0.026 |
coat | +0.021 |
swe | +0.020 |
ripe | +0.020 |
Neg logit contributions | |
glass | -0.047 |
orse | -0.045 |
work | -0.043 |
generation | -0.043 |
planet | -0.043 |
Token | ci |
Token ID | 979 |
Feature activation | +8.49e+00 |
Pos logit contributions | |
orse | +0.001 |
glass | +0.001 |
generation | +0.001 |
growth | +0.001 |
Whale | +0.001 |
Neg logit contributions | |
coat | -0.001 |
wn | -0.001 |
aback | -0.001 |
like | -0.001 |
swe | -0.001 |
Token | est |
Token ID | 395 |
Feature activation | -1.44e-01 |
Pos logit contributions | |
glass | +0.633 |
orse | +0.630 |
generation | +0.570 |
work | +0.555 |
planet | +0.552 |
Neg logit contributions | |
wn | -0.307 |
aback | -0.307 |
coat | -0.269 |
swe | -0.260 |
like | -0.245 |
Token | dress |
Token ID | 6576 |
Feature activation | +5.54e-01 |
Pos logit contributions | |
wn | +0.006 |
aback | +0.006 |
coat | +0.006 |
swe | +0.005 |
like | +0.005 |
Neg logit contributions | |
glass | -0.010 |
orse | -0.009 |
generation | -0.009 |
work | -0.008 |
planet | -0.008 |
Token | she |
Token ID | 673 |
Feature activation | -8.81e-01 |
Pos logit contributions | |
orse | +0.040 |
glass | +0.040 |
generation | +0.037 |
work | +0.036 |
planet | +0.036 |
Neg logit contributions | |
wn | -0.020 |
aback | -0.019 |
coat | -0.018 |
swe | -0.017 |
like | -0.017 |
Token | had |
Token ID | 550 |
Feature activation | +1.13e+00 |
Pos logit contributions | |
aback | +0.042 |
wn | +0.041 |
coat | +0.037 |
swe | +0.036 |
ripe | +0.034 |
Neg logit contributions | |
glass | -0.056 |
orse | -0.055 |
generation | -0.050 |
work | -0.050 |
planet | -0.048 |
Token | ever |
Token ID | 1683 |
Feature activation | +4.42e+00 |
Pos logit contributions | |
glass | +0.065 |
orse | +0.064 |
generation | +0.058 |
work | +0.057 |
bike | +0.056 |
Neg logit contributions | |
wn | -0.058 |
aback | -0.057 |
coat | -0.053 |
like | -0.052 |
swe | -0.050 |
Token | a |
Token ID | 257 |
Feature activation | -1.88e-01 |
Pos logit contributions | |
wn | +0.061 |
aback | +0.060 |
coat | +0.058 |
swe | +0.054 |
ripe | +0.053 |
Neg logit contributions | |
glass | -0.078 |
orse | -0.074 |
generation | -0.070 |
work | -0.068 |
planet | -0.065 |
Token | seat |
Token ID | 5852 |
Feature activation | +1.71e+00 |
Pos logit contributions | |
wn | +0.008 |
aback | +0.007 |
coat | +0.006 |
swe | +0.006 |
like | +0.006 |
Neg logit contributions | |
glass | -0.013 |
orse | -0.013 |
work | -0.012 |
planet | -0.012 |
generation | -0.012 |
Token | in |
Token ID | 287 |
Feature activation | +2.13e-01 |
Pos logit contributions | |
orse | +0.135 |
glass | +0.131 |
generation | +0.126 |
planet | +0.120 |
Whale | +0.119 |
Neg logit contributions | |
wn | -0.055 |
aback | -0.053 |
coat | -0.050 |
like | -0.048 |
past | -0.044 |
Token | the |
Token ID | 262 |
Feature activation | -6.03e-01 |
Pos logit contributions | |
orse | +0.015 |
glass | +0.015 |
generation | +0.013 |
work | +0.013 |
ich | +0.013 |
Neg logit contributions | |
aback | -0.009 |
wn | -0.009 |
like | -0.008 |
swe | -0.008 |
coat | -0.008 |
Token | motor |
Token ID | 5584 |
Feature activation | +2.85e+00 |
Pos logit contributions | |
wn | +0.037 |
aback | +0.036 |
coat | +0.033 |
swe | +0.033 |
like | +0.032 |
Neg logit contributions | |
orse | -0.030 |
glass | -0.030 |
generation | -0.026 |
work | -0.026 |
planet | -0.024 |
Token | bike |
Token ID | 32256 |
Feature activation | +8.40e+00 |
Pos logit contributions | |
orse | +0.232 |
glass | +0.226 |
generation | +0.215 |
growth | +0.209 |
Whale | +0.203 |
Neg logit contributions | |
aback | -0.075 |
coat | -0.074 |
wn | -0.072 |
like | -0.069 |
past | -0.066 |
Token | behind |
Token ID | 2157 |
Feature activation | +2.81e+00 |
Pos logit contributions | |
orse | +0.663 |
glass | +0.655 |
generation | +0.605 |
work | +0.590 |
planet | +0.588 |
Neg logit contributions | |
aback | -0.257 |
wn | -0.254 |
coat | -0.227 |
like | -0.220 |
past | -0.202 |
Token | him |
Token ID | 683 |
Feature activation | +5.62e-01 |
Pos logit contributions | |
orse | +0.221 |
glass | +0.215 |
ich | +0.195 |
generation | +0.195 |
fer | +0.192 |
Neg logit contributions | |
aback | -0.094 |
wn | -0.088 |
like | -0.081 |
past | -0.078 |
coat | -0.074 |
Token | . |
Token ID | 13 |
Feature activation | -5.42e-02 |
Pos logit contributions | |
glass | +0.041 |
orse | +0.041 |
generation | +0.037 |
work | +0.037 |
planet | +0.036 |
Neg logit contributions | |
wn | -0.021 |
aback | -0.021 |
coat | -0.018 |
swe | -0.017 |
like | -0.016 |
Token | He |
Token ID | 679 |
Feature activation | -2.81e-01 |
Pos logit contributions | |
wn | +0.003 |
aback | +0.003 |
coat | +0.003 |
swe | +0.003 |
ripe | +0.002 |
Neg logit contributions | |
glass | -0.003 |
orse | -0.003 |
generation | -0.003 |
work | -0.003 |
bike | -0.003 |
Token | waved |
Token ID | 26834 |
Feature activation | -4.37e-01 |
Pos logit contributions | |
wn | +0.011 |
aback | +0.011 |
coat | +0.009 |
swe | +0.009 |
like | +0.009 |
Neg logit contributions | |
glass | -0.020 |
orse | -0.019 |
generation | -0.019 |
work | -0.018 |
planet | -0.018 |
Token | a |
Token ID | 257 |
Feature activation | -1.92e-01 |
Pos logit contributions | |
glass | +0.010 |
orse | +0.010 |
generation | +0.009 |
planet | +0.009 |
work | +0.009 |
Neg logit contributions | |
wn | -0.008 |
aback | -0.008 |
coat | -0.008 |
ets | -0.007 |
swe | -0.007 |
Token | man |
Token ID | 582 |
Feature activation | -1.56e+00 |
Pos logit contributions | |
wn | +0.006 |
aback | +0.006 |
coat | +0.005 |
swe | +0.005 |
sweet | +0.005 |
Neg logit contributions | |
glass | -0.015 |
orse | -0.015 |
generation | -0.014 |
work | -0.013 |
bike | -0.013 |
Token | on |
Token ID | 319 |
Feature activation | +9.71e-01 |
Pos logit contributions | |
aback | +0.058 |
like | +0.057 |
wn | +0.056 |
coat | +0.051 |
past | +0.049 |
Neg logit contributions | |
glass | -0.108 |
orse | -0.102 |
generation | -0.100 |
work | -0.096 |
planet | -0.095 |
Token | a |
Token ID | 257 |
Feature activation | -1.97e-01 |
Pos logit contributions | |
orse | +0.074 |
glass | +0.073 |
work | +0.066 |
fer | +0.065 |
generation | +0.065 |
Neg logit contributions | |
aback | -0.032 |
wn | -0.031 |
like | -0.029 |
past | -0.028 |
fun | -0.027 |
Token | motor |
Token ID | 5584 |
Feature activation | +2.83e+00 |
Pos logit contributions | |
wn | +0.010 |
aback | +0.009 |
coat | +0.009 |
swe | +0.008 |
like | +0.008 |
Neg logit contributions | |
glass | -0.012 |
orse | -0.012 |
generation | -0.011 |
work | -0.011 |
planet | -0.010 |
Token | bike |
Token ID | 32256 |
Feature activation | +8.37e+00 |
Pos logit contributions | |
glass | +0.211 |
orse | +0.211 |
generation | +0.197 |
growth | +0.186 |
work | +0.183 |
Neg logit contributions | |
like | -0.088 |
aback | -0.085 |
wn | -0.085 |
coat | -0.084 |
past | -0.079 |
Token | ! |
Token ID | 0 |
Feature activation | +1.80e+00 |
Pos logit contributions | |
glass | +0.632 |
orse | +0.632 |
generation | +0.582 |
work | +0.563 |
planet | +0.556 |
Neg logit contributions | |
aback | -0.276 |
wn | -0.269 |
like | -0.252 |
coat | -0.242 |
past | -0.222 |
Token |
|
Token ID | 198 |
Feature activation | +7.02e-02 |
Pos logit contributions | |
glass | +0.106 |
orse | +0.104 |
generation | +0.094 |
work | +0.093 |
planet | +0.090 |
Neg logit contributions | |
wn | -0.092 |
aback | -0.092 |
coat | -0.083 |
swe | -0.081 |
like | -0.078 |
Token |
|
Token ID | 198 |
Feature activation | +5.83e-02 |
Pos logit contributions | |
glass | +0.004 |
orse | +0.004 |
generation | +0.004 |
planet | +0.003 |
work | +0.003 |
Neg logit contributions | |
wn | -0.004 |
aback | -0.004 |
swe | -0.003 |
coat | -0.003 |
prett | -0.003 |
Token | The |
Token ID | 464 |
Feature activation | +3.98e+00 |
Pos logit contributions | |
glass | +0.003 |
orse | +0.003 |
generation | +0.002 |
work | +0.002 |
bike | +0.002 |
Neg logit contributions | |
wn | -0.004 |
aback | -0.004 |
coat | -0.003 |
swe | -0.003 |
ripe | -0.003 |
Token | goat |
Token ID | 26917 |
Feature activation | +2.24e+00 |
Pos logit contributions | |
glass | +0.200 |
orse | +0.197 |
generation | +0.176 |
work | +0.172 |
planet | +0.166 |
Neg logit contributions | |
wn | -0.240 |
aback | -0.238 |
coat | -0.219 |
swe | -0.216 |
like | -0.209 |
Token | will |
Token ID | 481 |
Feature activation | -2.15e+00 |
Pos logit contributions | |
wn | +0.013 |
aback | +0.012 |
swe | +0.012 |
coat | +0.011 |
ripe | +0.010 |
Neg logit contributions | |
orse | -0.021 |
glass | -0.020 |
generation | -0.019 |
work | -0.019 |
planet | -0.018 |
Token | buy |
Token ID | 2822 |
Feature activation | +1.36e+00 |
Pos logit contributions | |
wn | +0.102 |
aback | +0.101 |
coat | +0.088 |
swe | +0.087 |
ripe | +0.081 |
Neg logit contributions | |
glass | -0.135 |
orse | -0.134 |
generation | -0.125 |
work | -0.123 |
planet | -0.120 |
Token | you |
Token ID | 345 |
Feature activation | +1.55e+00 |
Pos logit contributions | |
glass | +0.085 |
orse | +0.084 |
generation | +0.076 |
work | +0.075 |
planet | +0.073 |
Neg logit contributions | |
wn | -0.064 |
aback | -0.064 |
swe | -0.058 |
coat | -0.056 |
like | -0.055 |
Token | the |
Token ID | 262 |
Feature activation | -6.68e-01 |
Pos logit contributions | |
glass | +0.100 |
orse | +0.098 |
generation | +0.090 |
work | +0.088 |
planet | +0.086 |
Neg logit contributions | |
wn | -0.071 |
aback | -0.070 |
coat | -0.063 |
swe | -0.062 |
like | -0.060 |
Token | fan |
Token ID | 4336 |
Feature activation | -1.30e-02 |
Pos logit contributions | |
aback | +0.031 |
wn | +0.031 |
coat | +0.025 |
inters | +0.025 |
ucker | +0.024 |
Neg logit contributions | |
glass | -0.042 |
orse | -0.039 |
planet | -0.039 |
work | -0.038 |
generation | -0.038 |
Token | ci |
Token ID | 979 |
Feature activation | +8.35e+00 |
Pos logit contributions | |
aback | +0.000 |
wn | +0.000 |
coat | +0.000 |
like | +0.000 |
swe | +0.000 |
Neg logit contributions | |
glass | -0.001 |
orse | -0.001 |
generation | -0.001 |
planet | -0.001 |
bike | -0.001 |
Token | est |
Token ID | 395 |
Feature activation | -2.22e-01 |
Pos logit contributions | |
glass | +0.598 |
orse | +0.590 |
generation | +0.546 |
work | +0.539 |
planet | +0.526 |
Neg logit contributions | |
wn | -0.320 |
aback | -0.315 |
coat | -0.275 |
swe | -0.267 |
like | -0.252 |
Token | dress |
Token ID | 6576 |
Feature activation | +5.34e-01 |
Pos logit contributions | |
wn | +0.009 |
aback | +0.009 |
coat | +0.008 |
swe | +0.008 |
ripe | +0.007 |
Neg logit contributions | |
glass | -0.015 |
orse | -0.015 |
generation | -0.014 |
work | -0.014 |
planet | -0.013 |
Token | in |
Token ID | 287 |
Feature activation | +1.91e-01 |
Pos logit contributions | |
glass | +0.038 |
orse | +0.037 |
generation | +0.035 |
work | +0.034 |
planet | +0.033 |
Neg logit contributions | |
wn | -0.021 |
aback | -0.020 |
coat | -0.018 |
swe | -0.017 |
like | -0.017 |
Token | the |
Token ID | 262 |
Feature activation | -6.89e-01 |
Pos logit contributions | |
glass | +0.014 |
orse | +0.014 |
bike | +0.013 |
work | +0.013 |
generation | +0.012 |
Neg logit contributions | |
aback | -0.007 |
wn | -0.007 |
like | -0.006 |
swe | -0.006 |
sweet | -0.006 |
Token | store |
Token ID | 3650 |
Feature activation | -5.48e-01 |
Pos logit contributions | |
wn | +0.037 |
aback | +0.036 |
swe | +0.033 |
coat | +0.033 |
like | +0.032 |
Neg logit contributions | |
orse | -0.040 |
glass | -0.039 |
generation | -0.034 |
work | -0.034 |
bike | -0.033 |
Token | the |
Token ID | 262 |
Feature activation | -7.31e-01 |
Pos logit contributions | |
orse | +0.050 |
glass | +0.050 |
generation | +0.046 |
planet | +0.045 |
work | +0.045 |
Neg logit contributions | |
wn | -0.041 |
aback | -0.039 |
coat | -0.038 |
ets | -0.034 |
swe | -0.034 |
Token | brav |
Token ID | 49025 |
Feature activation | +2.59e+00 |
Pos logit contributions | |
wn | +0.035 |
aback | +0.035 |
coat | +0.030 |
swe | +0.029 |
ripe | +0.027 |
Neg logit contributions | |
glass | -0.045 |
orse | -0.044 |
generation | -0.041 |
work | -0.041 |
planet | -0.041 |
Token | est |
Token ID | 395 |
Feature activation | -3.09e-01 |
Pos logit contributions | |
orse | +0.145 |
glass | +0.143 |
generation | +0.142 |
work | +0.136 |
planet | +0.130 |
Neg logit contributions | |
wn | -0.124 |
aback | -0.117 |
coat | -0.113 |
swe | -0.109 |
ripe | -0.108 |
Token | and |
Token ID | 290 |
Feature activation | +2.18e-03 |
Pos logit contributions | |
wn | +0.015 |
aback | +0.015 |
coat | +0.013 |
swe | +0.013 |
ripe | +0.012 |
Neg logit contributions | |
glass | -0.019 |
orse | -0.019 |
generation | -0.017 |
work | -0.017 |
planet | -0.017 |
Token | boun |
Token ID | 31283 |
Feature activation | -2.38e+00 |
Pos logit contributions | |
glass | +0.000 |
orse | +0.000 |
generation | +0.000 |
work | +0.000 |
planet | +0.000 |
Neg logit contributions | |
wn | -0.000 |
aback | -0.000 |
coat | -0.000 |
swe | -0.000 |
like | -0.000 |
Token | ci |
Token ID | 979 |
Feature activation | +8.29e+00 |
Pos logit contributions | |
wn | +0.099 |
aback | +0.098 |
coat | +0.086 |
swe | +0.083 |
like | +0.080 |
Neg logit contributions | |
glass | -0.164 |
orse | -0.161 |
generation | -0.148 |
work | -0.145 |
planet | -0.142 |
Token | est |
Token ID | 395 |
Feature activation | -2.96e-01 |
Pos logit contributions | |
glass | +0.645 |
orse | +0.632 |
generation | +0.582 |
work | +0.568 |
planet | +0.559 |
Neg logit contributions | |
aback | -0.273 |
wn | -0.272 |
coat | -0.236 |
swe | -0.227 |
like | -0.219 |
Token | bunny |
Token ID | 44915 |
Feature activation | +4.08e-01 |
Pos logit contributions | |
wn | +0.014 |
aback | +0.014 |
coat | +0.012 |
swe | +0.012 |
ripe | +0.012 |
Neg logit contributions | |
glass | -0.019 |
orse | -0.018 |
generation | -0.017 |
work | -0.016 |
planet | -0.016 |
Token | in |
Token ID | 287 |
Feature activation | +2.11e-01 |
Pos logit contributions | |
glass | +0.029 |
orse | +0.027 |
generation | +0.026 |
work | +0.025 |
growth | +0.025 |
Neg logit contributions | |
aback | -0.015 |
wn | -0.015 |
like | -0.015 |
sweet | -0.013 |
coat | -0.013 |
Token | the |
Token ID | 262 |
Feature activation | -7.40e-01 |
Pos logit contributions | |
glass | +0.016 |
orse | +0.015 |
work | +0.014 |
generation | +0.014 |
bike | +0.014 |
Neg logit contributions | |
aback | -0.008 |
wn | -0.007 |
like | -0.007 |
past | -0.006 |
fun | -0.006 |
Token | whole |
Token ID | 2187 |
Feature activation | +2.84e+00 |
Pos logit contributions | |
wn | +0.043 |
aback | +0.043 |
coat | +0.039 |
swe | +0.038 |
like | +0.037 |
Neg logit contributions | |
glass | -0.038 |
orse | -0.038 |
generation | -0.034 |
work | -0.033 |
planet | -0.033 |
Token | new |
Token ID | 649 |
Feature activation | -6.00e-02 |
Pos logit contributions | |
wn | +0.110 |
swe | +0.107 |
aback | +0.106 |
fun | +0.104 |
sweet | +0.103 |
Neg logit contributions | |
glass | -0.077 |
orse | -0.074 |
bike | -0.066 |
work | -0.064 |
generation | -0.064 |
Token | home |
Token ID | 1363 |
Feature activation | +3.49e+00 |
Pos logit contributions | |
wn | +0.004 |
aback | +0.004 |
swe | +0.004 |
coat | +0.004 |
like | +0.004 |
Neg logit contributions | |
glass | -0.002 |
orse | -0.002 |
generation | -0.002 |
work | -0.002 |
bike | -0.002 |
Token | with |
Token ID | 351 |
Feature activation | +4.60e-01 |
Pos logit contributions | |
orse | +0.280 |
glass | +0.280 |
generation | +0.261 |
work | +0.255 |
bike | +0.255 |
Neg logit contributions | |
aback | -0.092 |
wn | -0.091 |
like | -0.084 |
coat | -0.078 |
swe | -0.075 |
Token | the |
Token ID | 262 |
Feature activation | -6.63e-01 |
Pos logit contributions | |
glass | +0.033 |
orse | +0.033 |
work | +0.030 |
generation | +0.030 |
bike | +0.029 |
Neg logit contributions | |
aback | -0.017 |
wn | -0.017 |
swe | -0.015 |
coat | -0.015 |
like | -0.014 |
Token | fan |
Token ID | 4336 |
Feature activation | +1.53e-02 |
Pos logit contributions | |
wn | +0.034 |
aback | +0.033 |
coat | +0.030 |
swe | +0.029 |
like | +0.028 |
Neg logit contributions | |
glass | -0.040 |
orse | -0.039 |
generation | -0.035 |
work | -0.035 |
planet | -0.034 |
Token | ci |
Token ID | 979 |
Feature activation | +8.18e+00 |
Pos logit contributions | |
glass | +0.001 |
orse | +0.001 |
generation | +0.001 |
Whale | +0.001 |
growth | +0.001 |
Neg logit contributions | |
wn | -0.000 |
aback | -0.000 |
like | -0.000 |
coat | -0.000 |
swe | -0.000 |
Token | est |
Token ID | 395 |
Feature activation | -3.08e-01 |
Pos logit contributions | |
glass | +0.610 |
orse | +0.601 |
generation | +0.552 |
work | +0.542 |
planet | +0.533 |
Neg logit contributions | |
wn | -0.292 |
aback | -0.290 |
coat | -0.251 |
swe | -0.244 |
like | -0.232 |
Token | king |
Token ID | 5822 |
Feature activation | +4.85e-01 |
Pos logit contributions | |
wn | +0.015 |
aback | +0.015 |
coat | +0.013 |
swe | +0.013 |
ripe | +0.012 |
Neg logit contributions | |
glass | -0.019 |
orse | -0.019 |
generation | -0.017 |
work | -0.017 |
planet | -0.016 |
Token | in |
Token ID | 287 |
Feature activation | +2.15e-01 |
Pos logit contributions | |
glass | +0.031 |
orse | +0.031 |
generation | +0.029 |
bike | +0.028 |
work | +0.028 |
Neg logit contributions | |
wn | -0.021 |
aback | -0.020 |
coat | -0.018 |
like | -0.018 |
swe | -0.018 |
Token | the |
Token ID | 262 |
Feature activation | -6.71e-01 |
Pos logit contributions | |
glass | +0.016 |
orse | +0.016 |
bike | +0.014 |
work | +0.014 |
ich | +0.014 |
Neg logit contributions | |
aback | -0.008 |
wn | -0.007 |
like | -0.007 |
past | -0.006 |
sweet | -0.006 |
Token | land |
Token ID | 1956 |
Feature activation | -1.13e-01 |
Pos logit contributions | |
wn | +0.042 |
aback | +0.042 |
swe | +0.039 |
sweet | +0.038 |
coat | +0.038 |
Neg logit contributions | |
glass | -0.031 |
orse | -0.031 |
work | -0.026 |
bike | -0.026 |
generation | -0.025 |
Token | by |
Token ID | 416 |
Feature activation | -8.17e-01 |
Pos logit contributions | |
aback | +0.054 |
wn | +0.054 |
coat | +0.045 |
swe | +0.045 |
like | +0.043 |
Neg logit contributions | |
glass | -0.131 |
orse | -0.130 |
generation | -0.121 |
work | -0.117 |
planet | -0.116 |
Token | saying |
Token ID | 2282 |
Feature activation | -2.73e-01 |
Pos logit contributions | |
sweet | +0.036 |
wn | +0.035 |
aback | +0.035 |
like | +0.035 |
swe | +0.034 |
Neg logit contributions | |
glass | -0.052 |
orse | -0.050 |
bike | -0.045 |
generation | -0.044 |
fer | -0.043 |
Token | , |
Token ID | 11 |
Feature activation | -1.46e-02 |
Pos logit contributions | |
aback | +0.011 |
wn | +0.011 |
swe | +0.010 |
sweet | +0.010 |
like | +0.010 |
Neg logit contributions | |
glass | -0.018 |
orse | -0.018 |
generation | -0.017 |
work | -0.016 |
bike | -0.016 |
Token | " |
Token ID | 366 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
wn | +0.001 |
aback | +0.001 |
coat | +0.001 |
swe | +0.001 |
like | +0.001 |
Neg logit contributions | |
glass | -0.001 |
orse | -0.001 |
generation | -0.001 |
work | -0.001 |
planet | -0.001 |
Token | Mar |
Token ID | 7676 |
Feature activation | -1.06e+00 |
Pos logit contributions | |
wn | +0.069 |
aback | +0.068 |
coat | +0.063 |
like | +0.060 |
swe | +0.060 |
Neg logit contributions | |
orse | -0.080 |
glass | -0.080 |
planet | -0.069 |
generation | -0.069 |
work | -0.068 |
Token | riage |
Token ID | 4087 |
Feature activation | +8.14e+00 |
Pos logit contributions | |
wn | +0.085 |
aback | +0.085 |
coat | +0.079 |
swe | +0.077 |
past | +0.076 |
Neg logit contributions | |
orse | -0.033 |
glass | -0.032 |
planet | -0.023 |
generation | -0.022 |
bike | -0.021 |
Token | is |
Token ID | 318 |
Feature activation | +1.46e-01 |
Pos logit contributions | |
orse | +0.575 |
glass | +0.568 |
generation | +0.522 |
bike | +0.504 |
work | +0.499 |
Neg logit contributions | |
aback | -0.313 |
wn | -0.308 |
coat | -0.276 |
like | -0.271 |
swe | -0.267 |
Token | when |
Token ID | 618 |
Feature activation | -5.36e-01 |
Pos logit contributions | |
glass | +0.011 |
orse | +0.010 |
generation | +0.010 |
work | +0.009 |
planet | +0.009 |
Neg logit contributions | |
wn | -0.006 |
aback | -0.006 |
coat | -0.005 |
swe | -0.005 |
like | -0.005 |
Token | two |
Token ID | 734 |
Feature activation | +9.21e-01 |
Pos logit contributions | |
sweet | +0.023 |
like | +0.023 |
fun | +0.022 |
nice | +0.021 |
swe | +0.020 |
Neg logit contributions | |
orse | -0.035 |
glass | -0.035 |
bike | -0.032 |
work | -0.031 |
fer | -0.030 |
Token | people |
Token ID | 661 |
Feature activation | -8.27e-01 |
Pos logit contributions | |
glass | +0.053 |
orse | +0.052 |
generation | +0.048 |
work | +0.047 |
planet | +0.046 |
Neg logit contributions | |
wn | -0.048 |
aback | -0.048 |
coat | -0.043 |
swe | -0.042 |
ripe | -0.040 |
Token | who |
Token ID | 508 |
Feature activation | +1.63e+00 |
Pos logit contributions | |
wn | +0.034 |
aback | +0.034 |
coat | +0.030 |
swe | +0.029 |
ripe | +0.027 |
Neg logit contributions | |
glass | -0.057 |
orse | -0.056 |
work | -0.053 |
generation | -0.052 |
planet | -0.050 |
Token | . |
Token ID | 13 |
Feature activation | -1.06e-01 |
Pos logit contributions | |
wn | +0.042 |
aback | +0.039 |
coat | +0.036 |
swe | +0.034 |
ripe | +0.033 |
Neg logit contributions | |
orse | -0.064 |
glass | -0.063 |
work | -0.057 |
generation | -0.056 |
bike | -0.055 |
Token |
|
Token ID | 198 |
Feature activation | -5.78e-02 |
Pos logit contributions | |
wn | +0.005 |
aback | +0.005 |
coat | +0.004 |
swe | +0.004 |
like | +0.004 |
Neg logit contributions | |
glass | -0.007 |
orse | -0.007 |
generation | -0.006 |
work | -0.006 |
planet | -0.006 |
Token |
|
Token ID | 198 |
Feature activation | -7.00e-02 |
Pos logit contributions | |
aback | +0.003 |
wn | +0.003 |
coat | +0.002 |
swe | +0.002 |
like | +0.002 |
Neg logit contributions | |
glass | -0.004 |
orse | -0.004 |
generation | -0.003 |
work | -0.003 |
planet | -0.003 |
Token | " |
Token ID | 1 |
Feature activation | -6.32e-01 |
Pos logit contributions | |
aback | +0.003 |
wn | +0.003 |
coat | +0.003 |
swe | +0.003 |
ripe | +0.003 |
Neg logit contributions | |
glass | -0.004 |
orse | -0.004 |
generation | -0.004 |
work | -0.004 |
planet | -0.004 |
Token | Mar |
Token ID | 7676 |
Feature activation | -1.14e+00 |
Pos logit contributions | |
coat | +0.025 |
aback | +0.025 |
wn | +0.025 |
like | +0.024 |
sweet | +0.023 |
Neg logit contributions | |
orse | -0.041 |
glass | -0.041 |
generation | -0.038 |
planet | -0.037 |
work | -0.037 |
Token | riage |
Token ID | 4087 |
Feature activation | +8.13e+00 |
Pos logit contributions | |
aback | +0.068 |
wn | +0.067 |
coat | +0.062 |
past | +0.059 |
like | +0.059 |
Neg logit contributions | |
glass | -0.059 |
orse | -0.057 |
generation | -0.052 |
planet | -0.049 |
growth | -0.045 |
Token | is |
Token ID | 318 |
Feature activation | +6.03e-02 |
Pos logit contributions | |
glass | +0.548 |
orse | +0.545 |
generation | +0.501 |
work | +0.486 |
bike | +0.476 |
Neg logit contributions | |
aback | -0.341 |
wn | -0.336 |
coat | -0.300 |
swe | -0.294 |
like | -0.286 |
Token | when |
Token ID | 618 |
Feature activation | -5.72e-01 |
Pos logit contributions | |
glass | +0.004 |
orse | +0.004 |
generation | +0.004 |
work | +0.004 |
planet | +0.004 |
Neg logit contributions | |
wn | -0.002 |
aback | -0.002 |
coat | -0.002 |
like | -0.002 |
swe | -0.002 |
Token | two |
Token ID | 734 |
Feature activation | +8.59e-01 |
Pos logit contributions | |
like | +0.024 |
sweet | +0.024 |
fun | +0.023 |
ripe | +0.023 |
aback | +0.022 |
Neg logit contributions | |
glass | -0.038 |
orse | -0.037 |
work | -0.034 |
bike | -0.033 |
generation | -0.032 |
Token | people |
Token ID | 661 |
Feature activation | -8.45e-01 |
Pos logit contributions | |
glass | +0.052 |
orse | +0.051 |
generation | +0.047 |
work | +0.046 |
planet | +0.045 |
Neg logit contributions | |
wn | -0.042 |
aback | -0.042 |
coat | -0.038 |
swe | -0.037 |
like | -0.035 |
Token | love |
Token ID | 1842 |
Feature activation | -2.00e-01 |
Pos logit contributions | |
wn | +0.034 |
aback | +0.033 |
coat | +0.030 |
swe | +0.029 |
ripe | +0.026 |
Neg logit contributions | |
glass | -0.059 |
orse | -0.058 |
work | -0.054 |
generation | -0.054 |
planet | -0.052 |
Token | cat |
Token ID | 3797 |
Feature activation | +7.07e-01 |
Pos logit contributions | |
glass | +0.145 |
orse | +0.145 |
generation | +0.135 |
work | +0.132 |
fer | +0.124 |
Neg logit contributions | |
sweet | -0.098 |
wn | -0.096 |
swe | -0.095 |
coat | -0.095 |
aback | -0.094 |
Token | said |
Token ID | 531 |
Feature activation | -9.87e-01 |
Pos logit contributions | |
glass | +0.053 |
orse | +0.052 |
generation | +0.048 |
work | +0.048 |
planet | +0.046 |
Neg logit contributions | |
wn | -0.026 |
aback | -0.025 |
coat | -0.021 |
swe | -0.020 |
ripe | -0.019 |
Token | , |
Token ID | 11 |
Feature activation | +8.98e-02 |
Pos logit contributions | |
aback | +0.030 |
wn | +0.028 |
coat | +0.024 |
swe | +0.024 |
like | +0.023 |
Neg logit contributions | |
glass | -0.079 |
orse | -0.078 |
generation | -0.074 |
work | -0.072 |
planet | -0.071 |
Token | " |
Token ID | 366 |
Feature activation | -1.54e+00 |
Pos logit contributions | |
glass | +0.007 |
orse | +0.006 |
generation | +0.006 |
planet | +0.006 |
work | +0.006 |
Neg logit contributions | |
aback | -0.003 |
wn | -0.003 |
like | -0.003 |
swe | -0.003 |
sweet | -0.003 |
Token | Mar |
Token ID | 7676 |
Feature activation | -1.28e+00 |
Pos logit contributions | |
wn | +0.065 |
aback | +0.064 |
like | +0.063 |
coat | +0.062 |
prett | +0.059 |
Neg logit contributions | |
glass | -0.097 |
orse | -0.095 |
generation | -0.085 |
ship | -0.085 |
planet | -0.084 |
Token | riage |
Token ID | 4087 |
Feature activation | +8.07e+00 |
Pos logit contributions | |
aback | +0.075 |
wn | +0.074 |
coat | +0.069 |
past | +0.065 |
swe | +0.064 |
Neg logit contributions | |
glass | -0.068 |
orse | -0.068 |
generation | -0.060 |
planet | -0.058 |
work | -0.054 |
Token | is |
Token ID | 318 |
Feature activation | +1.41e-02 |
Pos logit contributions | |
orse | +0.580 |
glass | +0.574 |
generation | +0.533 |
bike | +0.511 |
work | +0.508 |
Neg logit contributions | |
aback | -0.298 |
wn | -0.291 |
coat | -0.263 |
like | -0.262 |
swe | -0.253 |
Token | when |
Token ID | 618 |
Feature activation | -4.93e-01 |
Pos logit contributions | |
glass | +0.001 |
orse | +0.001 |
generation | +0.001 |
work | +0.001 |
planet | +0.001 |
Neg logit contributions | |
wn | -0.001 |
aback | -0.001 |
like | -0.000 |
coat | -0.000 |
swe | -0.000 |
Token | two |
Token ID | 734 |
Feature activation | +7.84e-01 |
Pos logit contributions | |
like | +0.020 |
sweet | +0.019 |
fun | +0.019 |
nice | +0.018 |
past | +0.018 |
Neg logit contributions | |
glass | -0.034 |
orse | -0.032 |
work | -0.030 |
bike | -0.029 |
generation | -0.028 |
Token | animals |
Token ID | 4695 |
Feature activation | -1.93e+00 |
Pos logit contributions | |
glass | +0.046 |
orse | +0.045 |
generation | +0.041 |
work | +0.040 |
planet | +0.039 |
Neg logit contributions | |
wn | -0.040 |
aback | -0.040 |
coat | -0.036 |
swe | -0.035 |
like | -0.034 |
Token | love |
Token ID | 1842 |
Feature activation | -1.67e-01 |
Pos logit contributions | |
wn | +0.079 |
aback | +0.079 |
coat | +0.070 |
swe | +0.068 |
ripe | +0.063 |
Neg logit contributions | |
glass | -0.132 |
orse | -0.131 |
generation | -0.121 |
work | -0.121 |
planet | -0.118 |
Token | , |
Token ID | 11 |
Feature activation | +9.44e-02 |
Pos logit contributions | |
wn | +0.002 |
aback | +0.002 |
prett | +0.002 |
tast | +0.001 |
swe | +0.001 |
Neg logit contributions | |
orse | -0.002 |
glass | -0.002 |
work | -0.002 |
planet | -0.002 |
generation | -0.002 |
Token | be |
Token ID | 307 |
Feature activation | +5.28e-01 |
Pos logit contributions | |
glass | +0.005 |
orse | +0.005 |
generation | +0.005 |
work | +0.005 |
bike | +0.005 |
Neg logit contributions | |
wn | -0.005 |
aback | -0.005 |
like | -0.005 |
swe | -0.005 |
sweet | -0.005 |
Token | careful |
Token ID | 8161 |
Feature activation | -2.34e+00 |
Pos logit contributions | |
glass | +0.037 |
orse | +0.035 |
generation | +0.033 |
ich | +0.032 |
crew | +0.031 |
Neg logit contributions | |
aback | -0.019 |
coat | -0.018 |
wn | -0.017 |
ets | -0.016 |
inters | -0.016 |
Token | ! |
Token ID | 0 |
Feature activation | +1.91e+00 |
Pos logit contributions | |
wn | +0.085 |
aback | +0.083 |
coat | +0.077 |
like | +0.074 |
sweet | +0.071 |
Neg logit contributions | |
glass | -0.172 |
orse | -0.167 |
generation | -0.157 |
work | -0.151 |
planet | -0.149 |
Token | Sp |
Token ID | 1338 |
Feature activation | +1.82e+00 |
Pos logit contributions | |
glass | +0.131 |
orse | +0.118 |
fer | +0.112 |
generation | +0.110 |
planet | +0.107 |
Neg logit contributions | |
aback | -0.081 |
wn | -0.077 |
like | -0.075 |
sweet | -0.069 |
swe | -0.069 |
Token | icy |
Token ID | 4611 |
Feature activation | +8.01e+00 |
Pos logit contributions | |
glass | +0.112 |
orse | +0.109 |
generation | +0.101 |
work | +0.098 |
bike | +0.097 |
Neg logit contributions | |
wn | -0.089 |
aback | -0.084 |
coat | -0.081 |
swe | -0.075 |
like | -0.075 |
Token | food |
Token ID | 2057 |
Feature activation | -5.02e-01 |
Pos logit contributions | |
glass | +0.515 |
orse | +0.508 |
generation | +0.468 |
work | +0.448 |
planet | +0.440 |
Neg logit contributions | |
wn | -0.363 |
aback | -0.362 |
coat | -0.322 |
like | -0.316 |
swe | -0.315 |
Token | can |
Token ID | 460 |
Feature activation | -2.17e+00 |
Pos logit contributions | |
wn | +0.022 |
aback | +0.021 |
coat | +0.019 |
swe | +0.018 |
ripe | +0.017 |
Neg logit contributions | |
glass | -0.033 |
orse | -0.033 |
work | -0.031 |
planet | -0.030 |
generation | -0.030 |
Token | make |
Token ID | 787 |
Feature activation | -6.20e-01 |
Pos logit contributions | |
aback | +0.078 |
wn | +0.078 |
like | +0.075 |
swe | +0.073 |
sweet | +0.072 |
Neg logit contributions | |
glass | -0.158 |
orse | -0.150 |
generation | -0.142 |
work | -0.137 |
fer | -0.136 |
Token | your |
Token ID | 534 |
Feature activation | +7.66e-01 |
Pos logit contributions | |
wn | +0.031 |
aback | +0.030 |
coat | +0.027 |
swe | +0.026 |
like | +0.025 |
Neg logit contributions | |
glass | -0.038 |
orse | -0.037 |
generation | -0.034 |
work | -0.033 |
planet | -0.033 |
Token | tum |
Token ID | 11814 |
Feature activation | +1.37e+00 |
Pos logit contributions | |
glass | +0.048 |
orse | +0.047 |
generation | +0.042 |
work | +0.042 |
bike | +0.041 |
Neg logit contributions | |
wn | -0.036 |
aback | -0.036 |
swe | -0.033 |
coat | -0.032 |
like | -0.031 |
Token | . |
Token ID | 13 |
Feature activation | -3.46e-01 |
Pos logit contributions | |
wn | +0.047 |
aback | +0.047 |
coat | +0.041 |
swe | +0.040 |
like | +0.039 |
Neg logit contributions | |
glass | -0.102 |
orse | -0.102 |
generation | -0.095 |
work | -0.092 |
planet | -0.091 |
Token |
|
Token ID | 198 |
Feature activation | -2.09e-01 |
Pos logit contributions | |
wn | +0.019 |
aback | +0.018 |
coat | +0.016 |
swe | +0.015 |
tast | +0.015 |
Neg logit contributions | |
glass | -0.020 |
orse | -0.020 |
generation | -0.017 |
planet | -0.016 |
Whale | -0.016 |
Token |
|
Token ID | 198 |
Feature activation | -2.19e-01 |
Pos logit contributions | |
aback | +0.009 |
wn | +0.009 |
coat | +0.008 |
swe | +0.008 |
like | +0.008 |
Neg logit contributions | |
glass | -0.014 |
orse | -0.013 |
generation | -0.012 |
work | -0.012 |
planet | -0.012 |
Token | " |
Token ID | 1 |
Feature activation | -6.71e-01 |
Pos logit contributions | |
wn | +0.010 |
aback | +0.009 |
coat | +0.008 |
swe | +0.008 |
like | +0.008 |
Neg logit contributions | |
glass | -0.015 |
orse | -0.014 |
generation | -0.013 |
work | -0.013 |
planet | -0.013 |
Token | Sp |
Token ID | 4561 |
Feature activation | +2.37e+00 |
Pos logit contributions | |
wn | +0.033 |
aback | +0.033 |
coat | +0.030 |
swe | +0.029 |
like | +0.028 |
Neg logit contributions | |
glass | -0.041 |
orse | -0.041 |
generation | -0.037 |
work | -0.036 |
planet | -0.035 |
Token | icy |
Token ID | 4611 |
Feature activation | +8.00e+00 |
Pos logit contributions | |
glass | +0.082 |
orse | +0.079 |
generation | +0.067 |
work | +0.064 |
planet | +0.062 |
Neg logit contributions | |
wn | -0.180 |
aback | -0.178 |
coat | -0.167 |
swe | -0.164 |
like | -0.161 |
Token | is |
Token ID | 318 |
Feature activation | -1.42e-01 |
Pos logit contributions | |
glass | +0.522 |
orse | +0.513 |
generation | +0.472 |
work | +0.462 |
planet | +0.452 |
Neg logit contributions | |
wn | -0.357 |
aback | -0.355 |
coat | -0.315 |
swe | -0.308 |
like | -0.296 |
Token | good |
Token ID | 922 |
Feature activation | -2.95e+00 |
Pos logit contributions | |
wn | +0.006 |
aback | +0.006 |
coat | +0.006 |
inters | +0.005 |
swe | +0.005 |
Neg logit contributions | |
orse | -0.010 |
glass | -0.009 |
work | -0.009 |
generation | -0.008 |
Whale | -0.008 |
Token | , |
Token ID | 11 |
Feature activation | +5.32e-02 |
Pos logit contributions | |
wn | +0.141 |
aback | +0.129 |
coat | +0.114 |
swe | +0.109 |
ripe | +0.107 |
Neg logit contributions | |
orse | -0.189 |
glass | -0.188 |
work | -0.179 |
Whale | -0.168 |
bike | -0.166 |
Token | but |
Token ID | 475 |
Feature activation | -2.08e+00 |
Pos logit contributions | |
glass | +0.004 |
orse | +0.004 |
generation | +0.003 |
work | +0.003 |
planet | +0.003 |
Neg logit contributions | |
wn | -0.002 |
aback | -0.002 |
coat | -0.002 |
swe | -0.002 |
like | -0.002 |
Token | not |
Token ID | 407 |
Feature activation | -1.03e-02 |
Pos logit contributions | |
like | +0.090 |
wn | +0.088 |
sweet | +0.088 |
aback | +0.087 |
swe | +0.085 |
Neg logit contributions | |
glass | -0.135 |
orse | -0.127 |
generation | -0.120 |
bike | -0.115 |
work | -0.115 |
Token | The |
Token ID | 383 |
Feature activation | +3.18e+00 |
Pos logit contributions | |
glass | +0.142 |
orse | +0.136 |
work | +0.128 |
generation | +0.126 |
planet | +0.123 |
Neg logit contributions | |
aback | -0.075 |
wn | -0.071 |
like | -0.066 |
coat | -0.065 |
swe | -0.063 |
Token | popcorn |
Token ID | 38914 |
Feature activation | +2.18e+00 |
Pos logit contributions | |
glass | +0.193 |
orse | +0.190 |
generation | +0.171 |
work | +0.167 |
bike | +0.160 |
Neg logit contributions | |
wn | -0.151 |
aback | -0.148 |
swe | -0.142 |
coat | -0.140 |
sweet | -0.138 |
Token | was |
Token ID | 373 |
Feature activation | +4.67e-01 |
Pos logit contributions | |
glass | +0.150 |
orse | +0.148 |
generation | +0.139 |
work | +0.133 |
planet | +0.130 |
Neg logit contributions | |
aback | -0.085 |
wn | -0.085 |
coat | -0.077 |
swe | -0.075 |
like | -0.074 |
Token | sticky |
Token ID | 23408 |
Feature activation | -8.26e-01 |
Pos logit contributions | |
orse | +0.032 |
glass | +0.032 |
work | +0.029 |
generation | +0.028 |
whale | +0.028 |
Neg logit contributions | |
aback | -0.020 |
wn | -0.018 |
coat | -0.016 |
swe | -0.014 |
ets | -0.014 |
Token | and |
Token ID | 290 |
Feature activation | +6.11e-01 |
Pos logit contributions | |
wn | +0.038 |
aback | +0.037 |
coat | +0.032 |
swe | +0.031 |
prett | +0.029 |
Neg logit contributions | |
glass | -0.051 |
orse | -0.050 |
work | -0.048 |
generation | -0.045 |
bike | -0.044 |
Token | gre |
Token ID | 10536 |
Feature activation | -7.07e+00 |
Pos logit contributions | |
orse | +0.044 |
glass | +0.043 |
work | +0.039 |
generation | +0.039 |
planet | +0.039 |
Neg logit contributions | |
aback | -0.024 |
wn | -0.023 |
coat | -0.020 |
swe | -0.017 |
ets | -0.017 |
Token | asy |
Token ID | 4107 |
Feature activation | +2.29e+00 |
Pos logit contributions | |
wn | +0.350 |
aback | +0.347 |
swe | +0.307 |
coat | +0.304 |
fun | +0.299 |
Neg logit contributions | |
glass | -0.425 |
orse | -0.416 |
work | -0.390 |
generation | -0.371 |
fer | -0.358 |
Token | . |
Token ID | 13 |
Feature activation | -5.46e-02 |
Pos logit contributions | |
glass | +0.150 |
orse | +0.148 |
work | +0.133 |
generation | +0.131 |
planet | +0.127 |
Neg logit contributions | |
wn | -0.103 |
aback | -0.102 |
coat | -0.087 |
swe | -0.081 |
ripe | -0.079 |
Token | They |
Token ID | 1119 |
Feature activation | -4.89e-01 |
Pos logit contributions | |
aback | +0.002 |
wn | +0.002 |
coat | +0.002 |
swe | +0.002 |
like | +0.002 |
Neg logit contributions | |
glass | -0.004 |
orse | -0.004 |
work | -0.003 |
generation | -0.003 |
planet | -0.003 |
Token | used |
Token ID | 973 |
Feature activation | -1.72e+00 |
Pos logit contributions | |
wn | +0.019 |
aback | +0.019 |
coat | +0.017 |
swe | +0.016 |
ripe | +0.015 |
Neg logit contributions | |
glass | -0.035 |
orse | -0.034 |
generation | -0.031 |
work | -0.031 |
planet | -0.030 |
Token | a |
Token ID | 257 |
Feature activation | -3.08e-01 |
Pos logit contributions | |
wn | +0.098 |
aback | +0.095 |
coat | +0.093 |
ets | +0.085 |
swe | +0.082 |
Neg logit contributions | |
glass | -0.081 |
orse | -0.081 |
work | -0.075 |
generation | -0.071 |
whale | -0.068 |
Token | bigger |
Token ID | 5749 |
Feature activation | -7.07e-01 |
Pos logit contributions | |
wn | +0.123 |
coat | +0.113 |
aback | +0.103 |
ucker | +0.094 |
arl | +0.093 |
Neg logit contributions | |
work | -0.194 |
orse | -0.193 |
glass | -0.187 |
generation | -0.182 |
planet | -0.182 |
Token | and |
Token ID | 290 |
Feature activation | +2.01e-01 |
Pos logit contributions | |
wn | +0.031 |
aback | +0.029 |
coat | +0.026 |
swe | +0.024 |
ripe | +0.023 |
Neg logit contributions | |
glass | -0.047 |
orse | -0.046 |
work | -0.044 |
generation | -0.043 |
planet | -0.043 |
Token | the |
Token ID | 262 |
Feature activation | -9.97e-01 |
Pos logit contributions | |
glass | +0.013 |
orse | +0.013 |
work | +0.013 |
generation | +0.013 |
planet | +0.012 |
Neg logit contributions | |
wn | -0.008 |
aback | -0.008 |
coat | -0.007 |
inters | -0.006 |
ets | -0.006 |
Token | trees |
Token ID | 7150 |
Feature activation | -2.10e+00 |
Pos logit contributions | |
wn | +0.046 |
aback | +0.046 |
swe | +0.042 |
coat | +0.041 |
like | +0.041 |
Neg logit contributions | |
glass | -0.062 |
orse | -0.062 |
generation | -0.054 |
work | -0.053 |
bike | -0.053 |
Token | looked |
Token ID | 3114 |
Feature activation | +3.72e-01 |
Pos logit contributions | |
wn | +0.092 |
aback | +0.091 |
coat | +0.080 |
swe | +0.078 |
ripe | +0.074 |
Neg logit contributions | |
glass | -0.138 |
orse | -0.137 |
work | -0.125 |
generation | -0.125 |
planet | -0.121 |
Token | gre |
Token ID | 10536 |
Feature activation | -7.04e+00 |
Pos logit contributions | |
work | +0.024 |
glass | +0.024 |
orse | +0.024 |
generation | +0.024 |
planet | +0.024 |
Neg logit contributions | |
wn | -0.014 |
coat | -0.014 |
aback | -0.013 |
ucker | -0.011 |
arl | -0.011 |
Token | ener |
Token ID | 877 |
Feature activation | -2.02e+00 |
Pos logit contributions | |
wn | +0.334 |
aback | +0.329 |
coat | +0.302 |
swe | +0.288 |
fun | +0.284 |
Neg logit contributions | |
glass | -0.433 |
orse | -0.415 |
work | -0.411 |
generation | -0.395 |
fer | -0.367 |
Token | . |
Token ID | 13 |
Feature activation | -1.87e-02 |
Pos logit contributions | |
wn | +0.073 |
aback | +0.072 |
coat | +0.062 |
swe | +0.061 |
like | +0.060 |
Neg logit contributions | |
glass | -0.148 |
orse | -0.146 |
generation | -0.135 |
work | -0.132 |
planet | -0.130 |
Token |
|
Token ID | 198 |
Feature activation | -8.37e-02 |
Pos logit contributions | |
aback | +0.001 |
wn | +0.001 |
coat | +0.001 |
swe | +0.001 |
like | +0.001 |
Neg logit contributions | |
glass | -0.001 |
orse | -0.001 |
work | -0.001 |
generation | -0.001 |
planet | -0.001 |
Token |
|
Token ID | 198 |
Feature activation | -1.13e-01 |
Pos logit contributions | |
aback | +0.004 |
wn | +0.004 |
coat | +0.003 |
swe | +0.003 |
like | +0.003 |
Neg logit contributions | |
glass | -0.005 |
orse | -0.005 |
work | -0.005 |
generation | -0.005 |
planet | -0.005 |
Token | Finally |
Token ID | 11158 |
Feature activation | -7.08e-01 |
Pos logit contributions | |
wn | +0.005 |
aback | +0.005 |
coat | +0.004 |
swe | +0.004 |
like | +0.004 |
Neg logit contributions | |
glass | -0.007 |
orse | -0.007 |
work | -0.007 |
generation | -0.006 |
bike | -0.006 |
Token | much |
Token ID | 881 |
Feature activation | -5.62e-01 |
Pos logit contributions | |
wn | +0.103 |
aback | +0.101 |
coat | +0.092 |
swe | +0.086 |
tast | +0.081 |
Neg logit contributions | |
glass | -0.132 |
orse | -0.131 |
generation | -0.121 |
work | -0.117 |
planet | -0.116 |
Token | or |
Token ID | 393 |
Feature activation | +8.07e-01 |
Pos logit contributions | |
wn | +0.027 |
coat | +0.025 |
aback | +0.025 |
ny | +0.022 |
inters | +0.022 |
Neg logit contributions | |
glass | -0.028 |
orse | -0.028 |
movie | -0.026 |
work | -0.025 |
planet | -0.024 |
Token | it |
Token ID | 340 |
Feature activation | +2.58e-02 |
Pos logit contributions | |
orse | +0.050 |
glass | +0.050 |
generation | +0.045 |
work | +0.045 |
bike | +0.044 |
Neg logit contributions | |
aback | -0.038 |
wn | -0.037 |
coat | -0.033 |
swe | -0.030 |
ripe | -0.029 |
Token | will |
Token ID | 481 |
Feature activation | -2.21e+00 |
Pos logit contributions | |
glass | +0.002 |
orse | +0.002 |
work | +0.002 |
generation | +0.001 |
planet | +0.001 |
Neg logit contributions | |
wn | -0.001 |
aback | -0.001 |
coat | -0.001 |
ripe | -0.001 |
swe | -0.001 |
Token | look |
Token ID | 804 |
Feature activation | -1.01e+00 |
Pos logit contributions | |
wn | +0.109 |
aback | +0.108 |
coat | +0.095 |
swe | +0.086 |
ets | +0.083 |
Neg logit contributions | |
orse | -0.132 |
glass | -0.126 |
work | -0.121 |
generation | -0.119 |
planet | -0.119 |
Token | gre |
Token ID | 10536 |
Feature activation | -7.02e+00 |
Pos logit contributions | |
aback | +0.041 |
coat | +0.039 |
wn | +0.038 |
ere | +0.032 |
arl | +0.030 |
Neg logit contributions | |
orse | -0.061 |
glass | -0.059 |
work | -0.057 |
planet | -0.056 |
generation | -0.056 |
Token | asy |
Token ID | 4107 |
Feature activation | +2.30e+00 |
Pos logit contributions | |
aback | +0.356 |
wn | +0.355 |
fun | +0.319 |
coat | +0.307 |
ripe | +0.305 |
Neg logit contributions | |
glass | -0.381 |
orse | -0.356 |
work | -0.354 |
fer | -0.332 |
generation | -0.331 |
Token | ." |
Token ID | 526 |
Feature activation | +7.98e-01 |
Pos logit contributions | |
glass | +0.136 |
orse | +0.129 |
work | +0.117 |
generation | +0.114 |
planet | +0.113 |
Neg logit contributions | |
wn | -0.118 |
aback | -0.118 |
coat | -0.101 |
swe | -0.090 |
ripe | -0.090 |
Token | |
Token ID | 220 |
Feature activation | -1.93e+00 |
Pos logit contributions | |
orse | +0.048 |
glass | +0.047 |
generation | +0.042 |
work | +0.041 |
bike | +0.041 |
Neg logit contributions | |
wn | -0.041 |
aback | -0.038 |
coat | -0.035 |
swe | -0.035 |
ripe | -0.034 |
Token |
|
Token ID | 198 |
Feature activation | -4.75e-02 |
Pos logit contributions | |
wn | +0.110 |
aback | +0.103 |
coat | +0.097 |
swe | +0.096 |
ripe | +0.094 |
Neg logit contributions | |
orse | -0.104 |
glass | -0.102 |
generation | -0.089 |
work | -0.087 |
bike | -0.086 |
Token |
|
Token ID | 198 |
Feature activation | -6.87e-02 |
Pos logit contributions | |
wn | +0.003 |
aback | +0.002 |
coat | +0.002 |
swe | +0.002 |
ripe | +0.002 |
Neg logit contributions | |
orse | -0.003 |
glass | -0.003 |
generation | -0.002 |
bike | -0.002 |
work | -0.002 |
Token | had |
Token ID | 550 |
Feature activation | +1.01e+00 |
Pos logit contributions | |
aback | +0.073 |
wn | +0.073 |
coat | +0.066 |
swe | +0.063 |
like | +0.061 |
Neg logit contributions | |
glass | -0.114 |
orse | -0.113 |
generation | -0.106 |
work | -0.103 |
bike | -0.100 |
Token | grown |
Token ID | 7334 |
Feature activation | -2.19e+00 |
Pos logit contributions | |
glass | +0.066 |
orse | +0.065 |
generation | +0.060 |
work | +0.059 |
planet | +0.058 |
Neg logit contributions | |
wn | -0.044 |
aback | -0.043 |
coat | -0.039 |
swe | -0.037 |
ripe | -0.034 |
Token | to |
Token ID | 284 |
Feature activation | -3.35e-01 |
Pos logit contributions | |
wn | +0.098 |
coat | +0.094 |
arl | +0.092 |
aback | +0.087 |
inters | +0.080 |
Neg logit contributions | |
glass | -0.128 |
work | -0.124 |
generation | -0.117 |
orse | -0.116 |
planet | -0.110 |
Token | be |
Token ID | 307 |
Feature activation | +7.19e-01 |
Pos logit contributions | |
wn | +0.013 |
aback | +0.013 |
swe | +0.011 |
like | +0.011 |
coat | +0.011 |
Neg logit contributions | |
glass | -0.024 |
orse | -0.023 |
work | -0.022 |
generation | -0.021 |
bike | -0.021 |
Token | even |
Token ID | 772 |
Feature activation | +2.04e+00 |
Pos logit contributions | |
glass | +0.050 |
orse | +0.050 |
generation | +0.045 |
work | +0.044 |
crew | +0.042 |
Neg logit contributions | |
aback | -0.028 |
wn | -0.027 |
coat | -0.024 |
arl | -0.022 |
swe | -0.022 |
Token | gre |
Token ID | 10536 |
Feature activation | -7.01e+00 |
Pos logit contributions | |
orse | +0.101 |
glass | +0.099 |
generation | +0.093 |
work | +0.092 |
crew | +0.086 |
Neg logit contributions | |
coat | -0.106 |
wn | -0.105 |
aback | -0.102 |
arl | -0.094 |
inters | -0.092 |
Token | ener |
Token ID | 877 |
Feature activation | -2.16e+00 |
Pos logit contributions | |
wn | +0.300 |
aback | +0.295 |
coat | +0.261 |
swe | +0.255 |
ripe | +0.242 |
Neg logit contributions | |
glass | -0.474 |
orse | -0.465 |
generation | -0.431 |
work | -0.427 |
planet | -0.410 |
Token | and |
Token ID | 290 |
Feature activation | +9.04e-02 |
Pos logit contributions | |
aback | +0.070 |
wn | +0.070 |
coat | +0.061 |
like | +0.059 |
swe | +0.058 |
Neg logit contributions | |
glass | -0.164 |
orse | -0.162 |
generation | -0.151 |
work | -0.147 |
planet | -0.145 |
Token | brighter |
Token ID | 26083 |
Feature activation | +5.80e-02 |
Pos logit contributions | |
glass | +0.007 |
orse | +0.007 |
generation | +0.006 |
work | +0.006 |
planet | +0.006 |
Neg logit contributions | |
wn | -0.003 |
aback | -0.003 |
coat | -0.003 |
swe | -0.003 |
like | -0.003 |
Token | than |
Token ID | 621 |
Feature activation | -3.94e+00 |
Pos logit contributions | |
glass | +0.004 |
orse | +0.004 |
generation | +0.004 |
work | +0.004 |
planet | +0.004 |
Neg logit contributions | |
wn | -0.002 |
aback | -0.002 |
coat | -0.002 |
swe | -0.002 |
ripe | -0.002 |
Token | before |
Token ID | 878 |
Feature activation | +1.41e-01 |
Pos logit contributions | |
wn | +0.201 |
aback | +0.199 |
coat | +0.178 |
swe | +0.177 |
like | +0.175 |
Neg logit contributions | |
glass | -0.233 |
orse | -0.229 |
generation | -0.203 |
work | -0.200 |
planet | -0.197 |
Token | she |
Token ID | 673 |
Feature activation | -7.41e-01 |
Pos logit contributions | |
glass | +0.042 |
orse | +0.042 |
generation | +0.039 |
work | +0.038 |
planet | +0.037 |
Neg logit contributions | |
wn | -0.025 |
aback | -0.025 |
coat | -0.022 |
swe | -0.021 |
ripe | -0.019 |
Token | had |
Token ID | 550 |
Feature activation | +1.20e+00 |
Pos logit contributions | |
wn | +0.035 |
aback | +0.034 |
coat | +0.031 |
swe | +0.030 |
ripe | +0.028 |
Neg logit contributions | |
glass | -0.047 |
orse | -0.046 |
generation | -0.043 |
work | -0.041 |
planet | -0.041 |
Token | blue |
Token ID | 4171 |
Feature activation | -2.01e+00 |
Pos logit contributions | |
glass | +0.076 |
orse | +0.073 |
generation | +0.069 |
work | +0.066 |
planet | +0.064 |
Neg logit contributions | |
wn | -0.055 |
aback | -0.055 |
coat | -0.050 |
swe | -0.048 |
ripe | -0.044 |
Token | eyes |
Token ID | 2951 |
Feature activation | -2.62e+00 |
Pos logit contributions | |
wn | +0.095 |
aback | +0.093 |
coat | +0.080 |
swe | +0.078 |
ripe | +0.074 |
Neg logit contributions | |
glass | -0.129 |
orse | -0.126 |
work | -0.117 |
generation | -0.115 |
planet | -0.114 |
Token | and |
Token ID | 290 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
wn | +0.111 |
aback | +0.109 |
swe | +0.091 |
coat | +0.089 |
ripe | +0.083 |
Neg logit contributions | |
glass | -0.179 |
orse | -0.171 |
work | -0.164 |
generation | -0.159 |
bike | -0.154 |
Token | blonde |
Token ID | 21541 |
Feature activation | -7.00e+00 |
Pos logit contributions | |
glass | +0.007 |
orse | +0.007 |
generation | +0.006 |
work | +0.006 |
planet | +0.006 |
Neg logit contributions | |
wn | -0.004 |
aback | -0.004 |
coat | -0.004 |
swe | -0.003 |
like | -0.003 |
Token | hair |
Token ID | 4190 |
Feature activation | -2.97e+00 |
Pos logit contributions | |
aback | +0.299 |
wn | +0.296 |
coat | +0.235 |
tast | +0.229 |
swe | +0.223 |
Neg logit contributions | |
glass | -0.480 |
orse | -0.477 |
generation | -0.435 |
work | -0.416 |
planet | -0.416 |
Token | . |
Token ID | 13 |
Feature activation | -5.39e-03 |
Pos logit contributions | |
wn | +0.110 |
aback | +0.108 |
coat | +0.094 |
swe | +0.091 |
like | +0.086 |
Neg logit contributions | |
glass | -0.217 |
orse | -0.214 |
generation | -0.198 |
work | -0.195 |
planet | -0.191 |
Token |
|
Token ID | 198 |
Feature activation | -1.50e-01 |
Pos logit contributions | |
aback | +0.000 |
wn | +0.000 |
coat | +0.000 |
swe | +0.000 |
like | +0.000 |
Neg logit contributions | |
glass | -0.000 |
orse | -0.000 |
work | -0.000 |
generation | -0.000 |
planet | -0.000 |
Token |
|
Token ID | 198 |
Feature activation | -2.06e-01 |
Pos logit contributions | |
wn | +0.007 |
aback | +0.007 |
coat | +0.006 |
swe | +0.006 |
like | +0.006 |
Neg logit contributions | |
glass | -0.010 |
orse | -0.009 |
work | -0.009 |
generation | -0.009 |
planet | -0.008 |
Token | One |
Token ID | 3198 |
Feature activation | +2.01e+00 |
Pos logit contributions | |
wn | +0.009 |
aback | +0.009 |
coat | +0.008 |
swe | +0.007 |
like | +0.007 |
Neg logit contributions | |
glass | -0.014 |
orse | -0.013 |
work | -0.013 |
generation | -0.012 |
planet | -0.012 |
Token | much |
Token ID | 881 |
Feature activation | -5.23e-01 |
Pos logit contributions | |
wn | +0.063 |
coat | +0.058 |
aback | +0.058 |
ucker | +0.045 |
arl | +0.043 |
Neg logit contributions | |
glass | -0.098 |
orse | -0.098 |
work | -0.096 |
planet | -0.093 |
generation | -0.092 |
Token | better |
Token ID | 1365 |
Feature activation | -1.27e+00 |
Pos logit contributions | |
coat | +0.032 |
wn | +0.031 |
inters | +0.026 |
ere | +0.025 |
arl | +0.024 |
Neg logit contributions | |
work | -0.023 |
whale | -0.022 |
orse | -0.021 |
planet | -0.021 |
boat | -0.020 |
Token | ! |
Token ID | 0 |
Feature activation | +2.02e+00 |
Pos logit contributions | |
wn | +0.048 |
aback | +0.047 |
coat | +0.041 |
swe | +0.039 |
ripe | +0.037 |
Neg logit contributions | |
glass | -0.092 |
orse | -0.091 |
generation | -0.084 |
work | -0.083 |
planet | -0.081 |
Token | They |
Token ID | 1119 |
Feature activation | +4.81e-02 |
Pos logit contributions | |
glass | +0.137 |
orse | +0.134 |
generation | +0.123 |
work | +0.122 |
planet | +0.119 |
Neg logit contributions | |
aback | -0.087 |
wn | -0.086 |
coat | -0.076 |
swe | -0.074 |
like | -0.071 |
Token | were |
Token ID | 547 |
Feature activation | +1.47e+00 |
Pos logit contributions | |
glass | +0.003 |
orse | +0.003 |
generation | +0.003 |
work | +0.003 |
bike | +0.003 |
Neg logit contributions | |
wn | -0.002 |
aback | -0.002 |
swe | -0.002 |
coat | -0.002 |
like | -0.002 |
Token | gre |
Token ID | 10536 |
Feature activation | -6.87e+00 |
Pos logit contributions | |
orse | +0.095 |
glass | +0.090 |
work | +0.087 |
whale | +0.084 |
planet | +0.083 |
Neg logit contributions | |
wn | -0.070 |
aback | -0.066 |
coat | -0.063 |
inters | -0.055 |
ets | -0.053 |
Token | ener |
Token ID | 877 |
Feature activation | -1.67e+00 |
Pos logit contributions | |
wn | +0.299 |
aback | +0.292 |
coat | +0.260 |
swe | +0.257 |
fun | +0.243 |
Neg logit contributions | |
glass | -0.456 |
orse | -0.445 |
work | -0.419 |
generation | -0.411 |
fer | -0.389 |
Token | and |
Token ID | 290 |
Feature activation | +1.53e-01 |
Pos logit contributions | |
wn | +0.063 |
aback | +0.061 |
coat | +0.053 |
swe | +0.051 |
ripe | +0.048 |
Neg logit contributions | |
glass | -0.121 |
orse | -0.120 |
generation | -0.110 |
work | -0.110 |
planet | -0.107 |
Token | taller |
Token ID | 25242 |
Feature activation | -6.89e-01 |
Pos logit contributions | |
glass | +0.010 |
orse | +0.010 |
generation | +0.009 |
work | +0.009 |
planet | +0.009 |
Neg logit contributions | |
wn | -0.007 |
aback | -0.007 |
swe | -0.006 |
coat | -0.006 |
like | -0.006 |
Token | than |
Token ID | 621 |
Feature activation | -3.65e+00 |
Pos logit contributions | |
wn | +0.034 |
aback | +0.030 |
coat | +0.028 |
swe | +0.025 |
inters | +0.024 |
Neg logit contributions | |
glass | -0.044 |
orse | -0.044 |
work | -0.041 |
generation | -0.038 |
planet | -0.038 |
Token | before |
Token ID | 878 |
Feature activation | +4.93e-01 |
Pos logit contributions | |
wn | +0.176 |
aback | +0.171 |
like | +0.157 |
swe | +0.155 |
coat | +0.152 |
Neg logit contributions | |
glass | -0.226 |
orse | -0.222 |
generation | -0.197 |
planet | -0.192 |
bike | -0.191 |
Token | it |
Token ID | 340 |
Feature activation | +5.49e-02 |
Pos logit contributions | |
orse | +0.263 |
glass | +0.255 |
fer | +0.252 |
generation | +0.250 |
field | +0.250 |
Neg logit contributions | |
aback | -0.099 |
sweet | -0.090 |
like | -0.088 |
wn | -0.087 |
nice | -0.086 |
Token | again |
Token ID | 757 |
Feature activation | +5.48e+00 |
Pos logit contributions | |
glass | +0.004 |
orse | +0.004 |
generation | +0.004 |
work | +0.004 |
planet | +0.004 |
Neg logit contributions | |
aback | -0.002 |
wn | -0.002 |
coat | -0.002 |
swe | -0.002 |
like | -0.002 |
Token | , |
Token ID | 11 |
Feature activation | +5.02e-03 |
Pos logit contributions | |
glass | +0.424 |
orse | +0.416 |
generation | +0.390 |
work | +0.384 |
planet | +0.375 |
Neg logit contributions | |
aback | -0.180 |
wn | -0.179 |
coat | -0.149 |
swe | -0.147 |
like | -0.146 |
Token | this |
Token ID | 428 |
Feature activation | +2.01e+00 |
Pos logit contributions | |
glass | +0.000 |
orse | +0.000 |
generation | +0.000 |
work | +0.000 |
planet | +0.000 |
Neg logit contributions | |
aback | -0.000 |
wn | -0.000 |
like | -0.000 |
coat | -0.000 |
sweet | -0.000 |
Token | time |
Token ID | 640 |
Feature activation | -5.81e+00 |
Pos logit contributions | |
glass | +0.169 |
orse | +0.168 |
work | +0.155 |
generation | +0.154 |
planet | +0.153 |
Neg logit contributions | |
wn | -0.054 |
aback | -0.053 |
coat | -0.042 |
swe | -0.041 |
ripe | -0.037 |
Token | gre |
Token ID | 10536 |
Feature activation | -6.85e+00 |
Pos logit contributions | |
like | +0.225 |
aback | +0.223 |
wn | +0.214 |
past | +0.206 |
fun | +0.195 |
Neg logit contributions | |
orse | -0.381 |
glass | -0.379 |
generation | -0.370 |
work | -0.359 |
bike | -0.354 |
Token | ated |
Token ID | 515 |
Feature activation | +5.56e+00 |
Pos logit contributions | |
wn | +0.292 |
aback | +0.286 |
coat | +0.254 |
swe | +0.251 |
ripe | +0.236 |
Neg logit contributions | |
glass | -0.468 |
orse | -0.458 |
generation | -0.422 |
work | -0.420 |
planet | -0.402 |
Token | and |
Token ID | 290 |
Feature activation | +4.71e-02 |
Pos logit contributions | |
glass | +0.386 |
orse | +0.372 |
generation | +0.346 |
work | +0.345 |
planet | +0.338 |
Neg logit contributions | |
wn | -0.229 |
aback | -0.223 |
coat | -0.198 |
swe | -0.190 |
ripe | -0.180 |
Token | louder |
Token ID | 27089 |
Feature activation | +5.91e-01 |
Pos logit contributions | |
glass | +0.003 |
orse | +0.003 |
generation | +0.003 |
work | +0.003 |
planet | +0.003 |
Neg logit contributions | |
aback | -0.002 |
wn | -0.002 |
swe | -0.002 |
sweet | -0.002 |
like | -0.002 |
Token | . |
Token ID | 13 |
Feature activation | -8.75e-02 |
Pos logit contributions | |
glass | +0.047 |
orse | +0.046 |
generation | +0.044 |
work | +0.043 |
bike | +0.042 |
Neg logit contributions | |
aback | -0.017 |
wn | -0.016 |
like | -0.015 |
sweet | -0.014 |
swe | -0.014 |
Token | The |
Token ID | 383 |
Feature activation | +2.43e+00 |
Pos logit contributions | |
wn | +0.004 |
aback | +0.004 |
coat | +0.004 |
swe | +0.004 |
ripe | +0.003 |
Neg logit contributions | |
glass | -0.005 |
orse | -0.005 |
generation | -0.005 |
work | -0.005 |
planet | -0.005 |
Token | 't |
Token ID | 470 |
Feature activation | +1.74e+00 |
Pos logit contributions | |
glass | +0.200 |
orse | +0.185 |
crew | +0.176 |
generation | +0.174 |
fer | +0.170 |
Neg logit contributions | |
aback | -0.088 |
like | -0.074 |
time | -0.073 |
swe | -0.073 |
wn | -0.073 |
Token | take |
Token ID | 1011 |
Feature activation | -4.71e-01 |
Pos logit contributions | |
glass | +0.128 |
orse | +0.120 |
work | +0.114 |
ich | +0.114 |
bike | +0.113 |
Neg logit contributions | |
like | -0.065 |
wn | -0.060 |
aback | -0.059 |
sweet | -0.051 |
fun | -0.051 |
Token | them |
Token ID | 606 |
Feature activation | -2.40e+00 |
Pos logit contributions | |
wn | +0.025 |
coat | +0.025 |
aback | +0.024 |
ripe | +0.022 |
swe | +0.021 |
Neg logit contributions | |
glass | -0.025 |
orse | -0.025 |
generation | -0.024 |
planet | -0.023 |
work | -0.022 |
Token | all |
Token ID | 477 |
Feature activation | -1.70e+00 |
Pos logit contributions | |
wn | +0.096 |
aback | +0.094 |
coat | +0.083 |
swe | +0.080 |
ripe | +0.075 |
Neg logit contributions | |
glass | -0.167 |
orse | -0.164 |
generation | -0.152 |
work | -0.151 |
planet | -0.148 |
Token | . |
Token ID | 13 |
Feature activation | -8.64e-02 |
Pos logit contributions | |
wn | +0.068 |
aback | +0.065 |
coat | +0.059 |
swe | +0.056 |
ripe | +0.052 |
Neg logit contributions | |
orse | -0.118 |
glass | -0.118 |
work | -0.109 |
generation | -0.109 |
planet | -0.106 |
Token | Care |
Token ID | 7276 |
Feature activation | -6.85e+00 |
Pos logit contributions | |
wn | +0.005 |
aback | +0.004 |
coat | +0.004 |
swe | +0.004 |
ripe | +0.004 |
Neg logit contributions | |
glass | -0.005 |
orse | -0.005 |
generation | -0.005 |
work | -0.004 |
planet | -0.004 |
Token | fully |
Token ID | 2759 |
Feature activation | +3.47e+00 |
Pos logit contributions | |
aback | +0.314 |
wn | +0.310 |
like | +0.280 |
coat | +0.279 |
swe | +0.268 |
Neg logit contributions | |
glass | -0.441 |
orse | -0.431 |
generation | -0.395 |
work | -0.378 |
planet | -0.374 |
Token | , |
Token ID | 11 |
Feature activation | -3.76e-03 |
Pos logit contributions | |
glass | +0.255 |
orse | +0.251 |
generation | +0.233 |
work | +0.230 |
planet | +0.225 |
Neg logit contributions | |
wn | -0.127 |
aback | -0.126 |
coat | -0.109 |
swe | -0.105 |
like | -0.100 |
Token | it |
Token ID | 340 |
Feature activation | +1.34e-01 |
Pos logit contributions | |
aback | +0.000 |
like | +0.000 |
past | +0.000 |
sweet | +0.000 |
ripe | +0.000 |
Neg logit contributions | |
glass | -0.000 |
orse | -0.000 |
work | -0.000 |
generation | -0.000 |
bike | -0.000 |
Token | picked |
Token ID | 6497 |
Feature activation | -1.49e+00 |
Pos logit contributions | |
glass | +0.009 |
orse | +0.009 |
generation | +0.009 |
work | +0.008 |
planet | +0.008 |
Neg logit contributions | |
wn | -0.005 |
aback | -0.005 |
coat | -0.005 |
swe | -0.004 |
ripe | -0.004 |
Token | out |
Token ID | 503 |
Feature activation | -1.60e+00 |
Pos logit contributions | |
wn | +0.063 |
aback | +0.060 |
coat | +0.055 |
swe | +0.052 |
ripe | +0.049 |
Neg logit contributions | |
glass | -0.102 |
orse | -0.099 |
generation | -0.093 |
work | -0.092 |
planet | -0.089 |
Token | They |
Token ID | 1119 |
Feature activation | -1.42e-01 |
Pos logit contributions | |
aback | +0.006 |
wn | +0.006 |
like | +0.005 |
swe | +0.005 |
sweet | +0.005 |
Neg logit contributions | |
glass | -0.007 |
orse | -0.007 |
work | -0.006 |
generation | -0.006 |
fer | -0.006 |
Token | are |
Token ID | 389 |
Feature activation | +1.14e+00 |
Pos logit contributions | |
aback | +0.007 |
wn | +0.007 |
coat | +0.007 |
swe | +0.006 |
ripe | +0.006 |
Neg logit contributions | |
orse | -0.008 |
glass | -0.008 |
work | -0.007 |
planet | -0.007 |
generation | -0.007 |
Token | too |
Token ID | 1165 |
Feature activation | -2.04e+00 |
Pos logit contributions | |
orse | +0.072 |
glass | +0.068 |
work | +0.063 |
planet | +0.062 |
whale | +0.062 |
Neg logit contributions | |
aback | -0.057 |
wn | -0.053 |
coat | -0.052 |
swe | -0.041 |
ere | -0.041 |
Token | salty |
Token ID | 36021 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
wn | +0.087 |
aback | +0.085 |
coat | +0.072 |
swe | +0.060 |
prett | +0.057 |
Neg logit contributions | |
orse | -0.148 |
glass | -0.136 |
whale | -0.128 |
planet | -0.127 |
generation | -0.124 |
Token | and |
Token ID | 290 |
Feature activation | +1.74e-01 |
Pos logit contributions | |
glass | +0.070 |
orse | +0.068 |
work | +0.063 |
generation | +0.062 |
planet | +0.061 |
Neg logit contributions | |
wn | -0.049 |
aback | -0.048 |
coat | -0.040 |
swe | -0.039 |
prett | -0.037 |
Token | gre |
Token ID | 10536 |
Feature activation | -6.85e+00 |
Pos logit contributions | |
orse | +0.013 |
glass | +0.012 |
work | +0.011 |
whale | +0.011 |
generation | +0.011 |
Neg logit contributions | |
aback | -0.007 |
wn | -0.007 |
coat | -0.006 |
ere | -0.004 |
ets | -0.004 |
Token | asy |
Token ID | 4107 |
Feature activation | +2.55e+00 |
Pos logit contributions | |
wn | +0.332 |
aback | +0.329 |
coat | +0.294 |
swe | +0.289 |
fun | +0.280 |
Neg logit contributions | |
glass | -0.420 |
orse | -0.411 |
work | -0.386 |
generation | -0.372 |
fer | -0.355 |
Token | . |
Token ID | 13 |
Feature activation | -1.25e-01 |
Pos logit contributions | |
glass | +0.155 |
orse | +0.153 |
work | +0.134 |
generation | +0.130 |
planet | +0.129 |
Neg logit contributions | |
wn | -0.129 |
aback | -0.129 |
coat | -0.113 |
swe | -0.101 |
prett | -0.098 |
Token | You |
Token ID | 921 |
Feature activation | +7.60e-01 |
Pos logit contributions | |
aback | +0.006 |
wn | +0.006 |
swe | +0.006 |
like | +0.006 |
sweet | +0.005 |
Neg logit contributions | |
glass | -0.008 |
orse | -0.007 |
generation | -0.007 |
work | -0.007 |
planet | -0.007 |
Token | should |
Token ID | 815 |
Feature activation | +1.21e+00 |
Pos logit contributions | |
orse | +0.050 |
glass | +0.049 |
work | +0.045 |
planet | +0.045 |
generation | +0.043 |
Neg logit contributions | |
aback | -0.034 |
wn | -0.033 |
coat | -0.030 |
swe | -0.029 |
ripe | -0.026 |
Token | eat |
Token ID | 4483 |
Feature activation | +1.64e+00 |
Pos logit contributions | |
orse | +0.071 |
glass | +0.066 |
generation | +0.062 |
planet | +0.059 |
work | +0.058 |
Neg logit contributions | |
aback | -0.066 |
wn | -0.064 |
coat | -0.056 |
swe | -0.054 |
prett | -0.051 |
Token | cars |
Token ID | 5006 |
Feature activation | +7.11e-01 |
Pos logit contributions | |
glass | +0.072 |
orse | +0.064 |
planet | +0.058 |
ship | +0.055 |
crew | +0.053 |
Neg logit contributions | |
aback | -0.113 |
wn | -0.109 |
coat | -0.099 |
tast | -0.095 |
ucker | -0.093 |
Token | of |
Token ID | 286 |
Feature activation | +4.82e-01 |
Pos logit contributions | |
glass | +0.048 |
orse | +0.045 |
work | +0.042 |
planet | +0.041 |
bike | +0.040 |
Neg logit contributions | |
aback | -0.031 |
wn | -0.031 |
swe | -0.029 |
ripe | -0.027 |
tast | -0.027 |
Token | different |
Token ID | 1180 |
Feature activation | -1.08e+00 |
Pos logit contributions | |
glass | +0.027 |
planet | +0.024 |
orse | +0.024 |
generation | +0.023 |
ship | +0.022 |
Neg logit contributions | |
aback | -0.026 |
wn | -0.026 |
coat | -0.024 |
ucker | -0.023 |
ited | -0.022 |
Token | colors |
Token ID | 7577 |
Feature activation | -5.08e+00 |
Pos logit contributions | |
aback | +0.047 |
wn | +0.046 |
coat | +0.041 |
ripe | +0.040 |
tast | +0.039 |
Neg logit contributions | |
glass | -0.074 |
orse | -0.067 |
work | -0.066 |
planet | -0.066 |
generation | -0.063 |
Token | and |
Token ID | 290 |
Feature activation | -2.78e-01 |
Pos logit contributions | |
wn | +0.206 |
aback | +0.205 |
coat | +0.174 |
swe | +0.171 |
ripe | +0.162 |
Neg logit contributions | |
glass | -0.359 |
orse | -0.350 |
work | -0.328 |
planet | -0.316 |
generation | -0.315 |
Token | sizes |
Token ID | 10620 |
Feature activation | -6.84e+00 |
Pos logit contributions | |
wn | +0.011 |
aback | +0.011 |
coat | +0.010 |
swe | +0.009 |
ripe | +0.009 |
Neg logit contributions | |
glass | -0.020 |
orse | -0.019 |
generation | -0.017 |
work | -0.017 |
planet | -0.017 |
Token | . |
Token ID | 13 |
Feature activation | -6.33e-01 |
Pos logit contributions | |
wn | +0.299 |
aback | +0.298 |
coat | +0.258 |
swe | +0.253 |
ripe | +0.239 |
Neg logit contributions | |
glass | -0.463 |
orse | -0.450 |
work | -0.420 |
planet | -0.405 |
generation | -0.405 |
Token | He |
Token ID | 679 |
Feature activation | -6.68e-01 |
Pos logit contributions | |
wn | +0.026 |
aback | +0.026 |
coat | +0.023 |
swe | +0.023 |
like | +0.022 |
Neg logit contributions | |
glass | -0.043 |
orse | -0.042 |
generation | -0.039 |
work | -0.039 |
planet | -0.037 |
Token | likes |
Token ID | 7832 |
Feature activation | -1.11e+00 |
Pos logit contributions | |
wn | +0.027 |
aback | +0.027 |
coat | +0.024 |
swe | +0.023 |
ripe | +0.022 |
Neg logit contributions | |
glass | -0.046 |
orse | -0.046 |
generation | -0.042 |
work | -0.042 |
planet | -0.041 |
Token | to |
Token ID | 284 |
Feature activation | -4.39e-01 |
Pos logit contributions | |
aback | +0.046 |
wn | +0.046 |
coat | +0.039 |
swe | +0.036 |
ripe | +0.035 |
Neg logit contributions | |
glass | -0.078 |
orse | -0.077 |
generation | -0.070 |
work | -0.070 |
planet | -0.069 |
Token | make |
Token ID | 787 |
Feature activation | -6.46e-01 |
Pos logit contributions | |
wn | +0.019 |
aback | +0.019 |
coat | +0.017 |
swe | +0.016 |
ripe | +0.015 |
Neg logit contributions | |
glass | -0.029 |
orse | -0.029 |
generation | -0.027 |
work | -0.026 |
planet | -0.026 |
Token | to |
Token ID | 284 |
Feature activation | -4.40e-01 |
Pos logit contributions | |
glass | +0.121 |
orse | +0.120 |
generation | +0.112 |
work | +0.111 |
planet | +0.108 |
Neg logit contributions | |
aback | -0.040 |
wn | -0.040 |
coat | -0.032 |
swe | -0.032 |
like | -0.030 |
Token | follow |
Token ID | 1061 |
Feature activation | -1.45e+00 |
Pos logit contributions | |
wn | +0.017 |
aback | +0.017 |
coat | +0.015 |
like | +0.015 |
swe | +0.015 |
Neg logit contributions | |
glass | -0.031 |
orse | -0.030 |
work | -0.028 |
generation | -0.028 |
bike | -0.027 |
Token | the |
Token ID | 262 |
Feature activation | -6.00e-01 |
Pos logit contributions | |
aback | +0.047 |
wn | +0.044 |
like | +0.039 |
swe | +0.038 |
past | +0.037 |
Neg logit contributions | |
glass | -0.115 |
orse | -0.114 |
work | -0.104 |
generation | -0.103 |
bike | -0.102 |
Token | lights |
Token ID | 7588 |
Feature activation | -1.85e+00 |
Pos logit contributions | |
wn | +0.030 |
aback | +0.030 |
coat | +0.027 |
swe | +0.026 |
ripe | +0.025 |
Neg logit contributions | |
glass | -0.037 |
orse | -0.036 |
generation | -0.033 |
work | -0.033 |
planet | -0.032 |
Token | . |
Token ID | 13 |
Feature activation | -2.33e-02 |
Pos logit contributions | |
wn | +0.059 |
aback | +0.056 |
like | +0.053 |
coat | +0.051 |
swe | +0.047 |
Neg logit contributions | |
orse | -0.144 |
glass | -0.142 |
generation | -0.131 |
Whale | -0.127 |
bike | -0.125 |
Token | Care |
Token ID | 7276 |
Feature activation | -6.84e+00 |
Pos logit contributions | |
wn | +0.001 |
aback | +0.001 |
coat | +0.001 |
swe | +0.001 |
ripe | +0.001 |
Neg logit contributions | |
glass | -0.002 |
generation | -0.001 |
orse | -0.001 |
work | -0.001 |
planet | -0.001 |
Token | fully |
Token ID | 2759 |
Feature activation | +3.60e+00 |
Pos logit contributions | |
wn | +0.359 |
aback | +0.354 |
coat | +0.321 |
swe | +0.316 |
ripe | +0.303 |
Neg logit contributions | |
glass | -0.394 |
orse | -0.387 |
generation | -0.352 |
work | -0.347 |
planet | -0.336 |
Token | she |
Token ID | 673 |
Feature activation | -8.64e-01 |
Pos logit contributions | |
glass | +0.262 |
orse | +0.258 |
generation | +0.235 |
work | +0.233 |
planet | +0.227 |
Neg logit contributions | |
wn | -0.137 |
aback | -0.132 |
coat | -0.118 |
swe | -0.114 |
ripe | -0.107 |
Token | opened |
Token ID | 4721 |
Feature activation | -1.71e+00 |
Pos logit contributions | |
wn | +0.032 |
aback | +0.032 |
coat | +0.028 |
swe | +0.027 |
like | +0.026 |
Neg logit contributions | |
glass | -0.063 |
orse | -0.062 |
generation | -0.057 |
work | -0.057 |
planet | -0.055 |
Token | the |
Token ID | 262 |
Feature activation | -6.34e-01 |
Pos logit contributions | |
wn | +0.086 |
aback | +0.081 |
coat | +0.077 |
swe | +0.071 |
tast | +0.069 |
Neg logit contributions | |
glass | -0.103 |
orse | -0.099 |
generation | -0.095 |
work | -0.093 |
planet | -0.092 |
Token | door |
Token ID | 3420 |
Feature activation | +8.70e-01 |
Pos logit contributions | |
wn | +0.036 |
aback | +0.035 |
coat | +0.033 |
swe | +0.032 |
ripe | +0.031 |
Neg logit contributions | |
glass | -0.035 |
orse | -0.031 |
work | -0.030 |
planet | -0.030 |
generation | -0.030 |
Token | , |
Token ID | 11 |
Feature activation | +1.90e-02 |
Pos logit contributions | |
wn | +0.046 |
aback | +0.044 |
coat | +0.039 |
swe | +0.038 |
ripe | +0.036 |
Neg logit contributions | |
glass | -0.086 |
orse | -0.085 |
generation | -0.078 |
work | -0.078 |
planet | -0.075 |
Token | holding |
Token ID | 4769 |
Feature activation | +3.76e-01 |
Pos logit contributions | |
glass | +0.001 |
orse | +0.001 |
generation | +0.001 |
work | +0.001 |
planet | +0.001 |
Neg logit contributions | |
wn | -0.001 |
aback | -0.001 |
coat | -0.001 |
swe | -0.001 |
like | -0.001 |
Token | a |
Token ID | 257 |
Feature activation | -4.48e-01 |
Pos logit contributions | |
glass | +0.024 |
orse | +0.023 |
generation | +0.022 |
work | +0.021 |
planet | +0.021 |
Neg logit contributions | |
wn | -0.017 |
aback | -0.016 |
coat | -0.015 |
swe | -0.014 |
ripe | -0.014 |
Token | doll |
Token ID | 3654 |
Feature activation | +2.44e-01 |
Pos logit contributions | |
aback | +0.020 |
wn | +0.020 |
ets | +0.017 |
inters | +0.016 |
tast | +0.015 |
Neg logit contributions | |
glass | -0.029 |
orse | -0.028 |
planet | -0.026 |
whale | -0.025 |
ship | -0.024 |
Token | with |
Token ID | 351 |
Feature activation | +4.13e-01 |
Pos logit contributions | |
orse | +0.019 |
glass | +0.018 |
generation | +0.017 |
Whale | +0.017 |
work | +0.017 |
Neg logit contributions | |
wn | -0.008 |
coat | -0.007 |
like | -0.007 |
aback | -0.007 |
past | -0.006 |
Token | blonde |
Token ID | 21541 |
Feature activation | -6.83e+00 |
Pos logit contributions | |
glass | +0.028 |
orse | +0.027 |
generation | +0.025 |
work | +0.025 |
planet | +0.024 |
Neg logit contributions | |
wn | -0.018 |
aback | -0.017 |
coat | -0.016 |
swe | -0.015 |
like | -0.014 |
Token | hair |
Token ID | 4190 |
Feature activation | -3.10e+00 |
Pos logit contributions | |
aback | +0.263 |
wn | +0.260 |
coat | +0.211 |
tast | +0.201 |
swe | +0.201 |
Neg logit contributions | |
glass | -0.496 |
orse | -0.490 |
generation | -0.441 |
planet | -0.429 |
work | -0.426 |
Token | . |
Token ID | 13 |
Feature activation | -1.35e-01 |
Pos logit contributions | |
wn | +0.117 |
aback | +0.117 |
coat | +0.098 |
swe | +0.096 |
ripe | +0.090 |
Neg logit contributions | |
glass | -0.224 |
orse | -0.221 |
generation | -0.203 |
work | -0.202 |
planet | -0.196 |
Token |
|
Token ID | 198 |
Feature activation | -2.45e-01 |
Pos logit contributions | |
wn | +0.007 |
aback | +0.007 |
coat | +0.006 |
swe | +0.006 |
ripe | +0.006 |
Neg logit contributions | |
glass | -0.008 |
orse | -0.008 |
generation | -0.007 |
work | -0.007 |
planet | -0.007 |
Token |
|
Token ID | 198 |
Feature activation | -2.70e-01 |
Pos logit contributions | |
aback | +0.010 |
wn | +0.010 |
coat | +0.009 |
swe | +0.009 |
like | +0.008 |
Neg logit contributions | |
glass | -0.017 |
orse | -0.016 |
work | -0.015 |
generation | -0.015 |
bike | -0.014 |
Token | " |
Token ID | 1 |
Feature activation | -7.18e-01 |
Pos logit contributions | |
wn | +0.012 |
aback | +0.012 |
coat | +0.010 |
swe | +0.010 |
like | +0.010 |
Neg logit contributions | |
glass | -0.018 |
orse | -0.018 |
work | -0.016 |
generation | -0.016 |
planet | -0.015 |
Token | ie |
Token ID | 494 |
Feature activation | -1.53e+00 |
Pos logit contributions | |
glass | +0.036 |
orse | +0.036 |
generation | +0.033 |
work | +0.033 |
planet | +0.032 |
Neg logit contributions | |
wn | -0.020 |
aback | -0.019 |
coat | -0.017 |
swe | -0.017 |
like | -0.016 |
Token | got |
Token ID | 1392 |
Feature activation | -4.25e-01 |
Pos logit contributions | |
aback | +0.061 |
wn | +0.061 |
coat | +0.054 |
swe | +0.053 |
like | +0.050 |
Neg logit contributions | |
glass | -0.106 |
orse | -0.104 |
generation | -0.097 |
work | -0.095 |
planet | -0.093 |
Token | an |
Token ID | 281 |
Feature activation | +4.67e-01 |
Pos logit contributions | |
wn | +0.021 |
coat | +0.020 |
ets | +0.018 |
aback | +0.018 |
inters | +0.016 |
Neg logit contributions | |
glass | -0.025 |
orse | -0.024 |
work | -0.023 |
generation | -0.022 |
whale | -0.021 |
Token | idea |
Token ID | 2126 |
Feature activation | -5.45e+00 |
Pos logit contributions | |
glass | +0.038 |
orse | +0.038 |
generation | +0.036 |
work | +0.035 |
planet | +0.035 |
Neg logit contributions | |
wn | -0.012 |
aback | -0.011 |
coat | -0.011 |
swe | -0.009 |
ripe | -0.008 |
Token | . |
Token ID | 13 |
Feature activation | -1.43e-01 |
Pos logit contributions | |
wn | +0.240 |
aback | +0.223 |
coat | +0.205 |
swe | +0.201 |
prett | +0.192 |
Neg logit contributions | |
glass | -0.365 |
orse | -0.359 |
work | -0.326 |
generation | -0.320 |
planet | -0.312 |
Token | Care |
Token ID | 7276 |
Feature activation | -6.82e+00 |
Pos logit contributions | |
wn | +0.006 |
aback | +0.005 |
coat | +0.005 |
swe | +0.005 |
ripe | +0.005 |
Neg logit contributions | |
glass | -0.010 |
orse | -0.009 |
generation | -0.009 |
work | -0.009 |
bike | -0.009 |
Token | fully |
Token ID | 2759 |
Feature activation | +3.25e+00 |
Pos logit contributions | |
wn | +0.354 |
aback | +0.344 |
coat | +0.311 |
swe | +0.310 |
ripe | +0.300 |
Neg logit contributions | |
glass | -0.393 |
orse | -0.381 |
work | -0.358 |
generation | -0.353 |
planet | -0.337 |
Token | , |
Token ID | 11 |
Feature activation | -3.69e-02 |
Pos logit contributions | |
glass | +0.230 |
orse | +0.228 |
work | +0.205 |
generation | +0.204 |
planet | +0.197 |
Neg logit contributions | |
wn | -0.130 |
aback | -0.124 |
coat | -0.113 |
swe | -0.107 |
ripe | -0.100 |
Token | she |
Token ID | 673 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
aback | +0.002 |
wn | +0.002 |
coat | +0.002 |
swe | +0.001 |
like | +0.001 |
Neg logit contributions | |
glass | -0.002 |
orse | -0.002 |
generation | -0.002 |
work | -0.002 |
planet | -0.002 |
Token | placed |
Token ID | 4624 |
Feature activation | -2.08e+00 |
Pos logit contributions | |
wn | +0.040 |
aback | +0.039 |
coat | +0.033 |
swe | +0.032 |
ripe | +0.030 |
Neg logit contributions | |
orse | -0.075 |
glass | -0.074 |
work | -0.067 |
generation | -0.066 |
planet | -0.065 |
Token | the |
Token ID | 262 |
Feature activation | -8.53e-01 |
Pos logit contributions | |
wn | +0.110 |
aback | +0.104 |
coat | +0.100 |
swe | +0.093 |
ets | +0.091 |
Neg logit contributions | |
glass | -0.117 |
orse | -0.112 |
work | -0.107 |
generation | -0.106 |
planet | -0.099 |
Token | trains |
Token ID | 13404 |
Feature activation | +2.21e+00 |
Pos logit contributions | |
glass | +0.121 |
orse | +0.117 |
planet | +0.108 |
whale | +0.104 |
generation | +0.104 |
Neg logit contributions | |
wn | -0.103 |
aback | -0.102 |
coat | -0.091 |
ucker | -0.079 |
tast | -0.078 |
Token | of |
Token ID | 286 |
Feature activation | +5.99e-01 |
Pos logit contributions | |
glass | +0.160 |
orse | +0.158 |
generation | +0.146 |
work | +0.145 |
planet | +0.141 |
Neg logit contributions | |
wn | -0.082 |
aback | -0.081 |
coat | -0.070 |
swe | -0.069 |
ripe | -0.064 |
Token | different |
Token ID | 1180 |
Feature activation | -8.58e-01 |
Pos logit contributions | |
glass | +0.035 |
orse | +0.033 |
generation | +0.030 |
planet | +0.030 |
work | +0.030 |
Neg logit contributions | |
aback | -0.032 |
wn | -0.031 |
coat | -0.028 |
swe | -0.027 |
ripe | -0.026 |
Token | colors |
Token ID | 7577 |
Feature activation | -4.90e+00 |
Pos logit contributions | |
aback | +0.032 |
wn | +0.032 |
coat | +0.027 |
swe | +0.026 |
ripe | +0.026 |
Neg logit contributions | |
glass | -0.063 |
orse | -0.061 |
work | -0.058 |
generation | -0.057 |
planet | -0.056 |
Token | and |
Token ID | 290 |
Feature activation | -5.82e-02 |
Pos logit contributions | |
wn | +0.186 |
aback | +0.183 |
coat | +0.156 |
swe | +0.153 |
ripe | +0.143 |
Neg logit contributions | |
glass | -0.356 |
orse | -0.350 |
work | -0.325 |
generation | -0.320 |
planet | -0.314 |
Token | sizes |
Token ID | 10620 |
Feature activation | -6.80e+00 |
Pos logit contributions | |
wn | +0.002 |
aback | +0.002 |
coat | +0.002 |
swe | +0.002 |
ripe | +0.002 |
Neg logit contributions | |
glass | -0.004 |
orse | -0.004 |
generation | -0.004 |
work | -0.004 |
planet | -0.004 |
Token | . |
Token ID | 13 |
Feature activation | -3.42e-01 |
Pos logit contributions | |
wn | +0.269 |
aback | +0.266 |
coat | +0.229 |
swe | +0.225 |
ripe | +0.209 |
Neg logit contributions | |
glass | -0.483 |
orse | -0.475 |
work | -0.440 |
generation | -0.433 |
planet | -0.426 |
Token | He |
Token ID | 679 |
Feature activation | -5.78e-01 |
Pos logit contributions | |
aback | +0.014 |
wn | +0.014 |
coat | +0.012 |
swe | +0.012 |
like | +0.012 |
Neg logit contributions | |
glass | -0.024 |
orse | -0.023 |
work | -0.022 |
generation | -0.021 |
planet | -0.021 |
Token | makes |
Token ID | 1838 |
Feature activation | -1.68e+00 |
Pos logit contributions | |
wn | +0.023 |
aback | +0.023 |
coat | +0.020 |
swe | +0.019 |
like | +0.018 |
Neg logit contributions | |
glass | -0.041 |
orse | -0.040 |
generation | -0.037 |
work | -0.036 |
planet | -0.036 |
Token | a |
Token ID | 257 |
Feature activation | -2.92e-01 |
Pos logit contributions | |
wn | +0.076 |
aback | +0.075 |
coat | +0.067 |
swe | +0.066 |
like | +0.064 |
Neg logit contributions | |
glass | -0.109 |
orse | -0.107 |
generation | -0.099 |
work | -0.097 |
planet | -0.094 |
Token | track |
Token ID | 2610 |
Feature activation | +1.27e+00 |
Pos logit contributions | |
wn | +0.011 |
aback | +0.010 |
swe | +0.009 |
coat | +0.009 |
sweet | +0.009 |
Neg logit contributions | |
glass | -0.021 |
orse | -0.021 |
generation | -0.020 |
work | -0.019 |
bike | -0.019 |
Token | cars |
Token ID | 5006 |
Feature activation | +7.94e-01 |
Pos logit contributions | |
glass | +0.075 |
orse | +0.067 |
planet | +0.059 |
ship | +0.057 |
crew | +0.056 |
Neg logit contributions | |
aback | -0.117 |
wn | -0.113 |
coat | -0.104 |
tast | -0.099 |
ju | -0.098 |
Token | of |
Token ID | 286 |
Feature activation | +5.26e-01 |
Pos logit contributions | |
glass | +0.054 |
orse | +0.051 |
work | +0.048 |
planet | +0.046 |
bike | +0.046 |
Neg logit contributions | |
aback | -0.035 |
wn | -0.034 |
swe | -0.031 |
ripe | -0.029 |
coat | -0.029 |
Token | different |
Token ID | 1180 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
glass | +0.027 |
orse | +0.024 |
planet | +0.023 |
generation | +0.023 |
work | +0.022 |
Neg logit contributions | |
aback | -0.031 |
wn | -0.031 |
coat | -0.029 |
ucker | -0.027 |
tast | -0.026 |
Token | colors |
Token ID | 7577 |
Feature activation | -5.02e+00 |
Pos logit contributions | |
aback | +0.045 |
wn | +0.045 |
coat | +0.039 |
ripe | +0.039 |
tast | +0.038 |
Neg logit contributions | |
glass | -0.071 |
orse | -0.065 |
work | -0.064 |
planet | -0.063 |
generation | -0.061 |
Token | and |
Token ID | 290 |
Feature activation | -1.15e-01 |
Pos logit contributions | |
wn | +0.195 |
aback | +0.194 |
coat | +0.164 |
swe | +0.161 |
ripe | +0.152 |
Neg logit contributions | |
glass | -0.361 |
orse | -0.354 |
work | -0.329 |
generation | -0.321 |
planet | -0.318 |
Token | sizes |
Token ID | 10620 |
Feature activation | -6.80e+00 |
Pos logit contributions | |
aback | +0.004 |
wn | +0.004 |
coat | +0.004 |
swe | +0.003 |
ripe | +0.003 |
Neg logit contributions | |
glass | -0.008 |
orse | -0.008 |
generation | -0.007 |
work | -0.007 |
planet | -0.007 |
Token | . |
Token ID | 13 |
Feature activation | -4.17e-01 |
Pos logit contributions | |
aback | +0.292 |
wn | +0.292 |
coat | +0.252 |
swe | +0.247 |
ripe | +0.233 |
Neg logit contributions | |
glass | -0.464 |
orse | -0.453 |
work | -0.422 |
generation | -0.407 |
planet | -0.407 |
Token | He |
Token ID | 679 |
Feature activation | -6.16e-01 |
Pos logit contributions | |
wn | +0.017 |
aback | +0.017 |
coat | +0.015 |
swe | +0.015 |
like | +0.014 |
Neg logit contributions | |
glass | -0.029 |
orse | -0.028 |
generation | -0.026 |
work | -0.026 |
planet | -0.025 |
Token | likes |
Token ID | 7832 |
Feature activation | -1.08e+00 |
Pos logit contributions | |
wn | +0.026 |
aback | +0.026 |
coat | +0.023 |
swe | +0.021 |
ripe | +0.021 |
Neg logit contributions | |
glass | -0.042 |
orse | -0.041 |
generation | -0.038 |
work | -0.038 |
planet | -0.037 |
Token | to |
Token ID | 284 |
Feature activation | -4.35e-01 |
Pos logit contributions | |
aback | +0.046 |
wn | +0.046 |
coat | +0.040 |
swe | +0.037 |
ripe | +0.036 |
Neg logit contributions | |
glass | -0.074 |
orse | -0.073 |
generation | -0.067 |
work | -0.066 |
planet | -0.065 |
Token | make |
Token ID | 787 |
Feature activation | -6.53e-01 |
Pos logit contributions | |
aback | +0.019 |
wn | +0.019 |
coat | +0.016 |
swe | +0.016 |
ripe | +0.015 |
Neg logit contributions | |
glass | -0.029 |
orse | -0.029 |
generation | -0.026 |
work | -0.026 |
planet | -0.025 |
Token | years |
Token ID | 812 |
Feature activation | -1.77e+00 |
Pos logit contributions | |
orse | +0.079 |
glass | +0.076 |
generation | +0.063 |
planet | +0.062 |
whale | +0.060 |
Neg logit contributions | |
wn | -0.138 |
aback | -0.129 |
coat | -0.125 |
inters | -0.114 |
arl | -0.112 |
Token | old |
Token ID | 1468 |
Feature activation | -2.60e+00 |
Pos logit contributions | |
wn | +0.067 |
aback | +0.064 |
coat | +0.055 |
swe | +0.048 |
tast | +0.046 |
Neg logit contributions | |
glass | -0.127 |
orse | -0.126 |
work | -0.120 |
generation | -0.120 |
planet | -0.115 |
Token | and |
Token ID | 290 |
Feature activation | +2.74e-01 |
Pos logit contributions | |
wn | +0.120 |
aback | +0.105 |
coat | +0.097 |
swe | +0.091 |
inters | +0.087 |
Neg logit contributions | |
glass | -0.173 |
orse | -0.166 |
work | -0.160 |
generation | -0.158 |
planet | -0.153 |
Token | had |
Token ID | 550 |
Feature activation | +1.05e+00 |
Pos logit contributions | |
glass | +0.019 |
orse | +0.019 |
generation | +0.017 |
work | +0.017 |
planet | +0.016 |
Neg logit contributions | |
wn | -0.011 |
aback | -0.011 |
coat | -0.010 |
swe | -0.010 |
ripe | -0.009 |
Token | beautiful |
Token ID | 4950 |
Feature activation | -7.18e-01 |
Pos logit contributions | |
glass | +0.062 |
orse | +0.060 |
generation | +0.059 |
planet | +0.055 |
work | +0.054 |
Neg logit contributions | |
wn | -0.049 |
aback | -0.048 |
coat | -0.044 |
arl | -0.041 |
ucker | -0.040 |
Token | blonde |
Token ID | 21541 |
Feature activation | -6.78e+00 |
Pos logit contributions | |
wn | +0.035 |
aback | +0.033 |
coat | +0.029 |
arl | +0.026 |
inters | +0.026 |
Neg logit contributions | |
glass | -0.044 |
orse | -0.043 |
planet | -0.042 |
generation | -0.041 |
work | -0.040 |
Token | hair |
Token ID | 4190 |
Feature activation | -2.91e+00 |
Pos logit contributions | |
wn | +0.266 |
aback | +0.263 |
coat | +0.214 |
swe | +0.201 |
tast | +0.199 |
Neg logit contributions | |
glass | -0.482 |
orse | -0.475 |
generation | -0.444 |
work | -0.429 |
planet | -0.426 |
Token | . |
Token ID | 13 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
wn | +0.102 |
aback | +0.100 |
coat | +0.089 |
swe | +0.085 |
like | +0.082 |
Neg logit contributions | |
glass | -0.216 |
orse | -0.214 |
generation | -0.199 |
work | -0.194 |
planet | -0.191 |
Token | |
Token ID | 220 |
Feature activation | -2.10e+00 |
Pos logit contributions | |
aback | +0.004 |
wn | +0.004 |
coat | +0.003 |
swe | +0.003 |
like | +0.003 |
Neg logit contributions | |
glass | -0.007 |
orse | -0.007 |
work | -0.006 |
generation | -0.006 |
planet | -0.006 |
Token |
|
Token ID | 198 |
Feature activation | -1.64e-01 |
Pos logit contributions | |
wn | +0.109 |
aback | +0.106 |
coat | +0.097 |
swe | +0.096 |
ripe | +0.093 |
Neg logit contributions | |
glass | -0.122 |
orse | -0.120 |
generation | -0.109 |
work | -0.106 |
planet | -0.103 |
Token | One |
Token ID | 3198 |
Feature activation | +2.01e+00 |
Pos logit contributions | |
wn | +0.008 |
aback | +0.008 |
coat | +0.007 |
swe | +0.007 |
like | +0.007 |
Neg logit contributions | |
glass | -0.010 |
orse | -0.010 |
generation | -0.009 |
work | -0.009 |
planet | -0.009 |
Token | saw |
Token ID | 2497 |
Feature activation | +1.16e+00 |
Pos logit contributions | |
glass | +0.003 |
orse | +0.003 |
generation | +0.003 |
work | +0.003 |
planet | +0.003 |
Neg logit contributions | |
wn | -0.002 |
aback | -0.002 |
coat | -0.001 |
swe | -0.001 |
like | -0.001 |
Token | a |
Token ID | 257 |
Feature activation | -3.14e-01 |
Pos logit contributions | |
orse | +0.086 |
glass | +0.086 |
generation | +0.080 |
work | +0.077 |
fer | +0.076 |
Neg logit contributions | |
aback | -0.043 |
sweet | -0.038 |
wn | -0.038 |
like | -0.038 |
fun | -0.037 |
Token | big |
Token ID | 1263 |
Feature activation | -1.60e+00 |
Pos logit contributions | |
wn | +0.012 |
aback | +0.011 |
coat | +0.010 |
swe | +0.009 |
like | +0.009 |
Neg logit contributions | |
glass | -0.023 |
orse | -0.023 |
generation | -0.021 |
work | -0.021 |
planet | -0.021 |
Token | tree |
Token ID | 5509 |
Feature activation | +6.07e-01 |
Pos logit contributions | |
wn | +0.075 |
aback | +0.074 |
coat | +0.067 |
swe | +0.065 |
like | +0.063 |
Neg logit contributions | |
glass | -0.101 |
orse | -0.100 |
generation | -0.091 |
work | -0.090 |
planet | -0.087 |
Token | . |
Token ID | 13 |
Feature activation | -2.96e-02 |
Pos logit contributions | |
glass | +0.045 |
orse | +0.044 |
generation | +0.041 |
work | +0.040 |
planet | +0.039 |
Neg logit contributions | |
wn | -0.021 |
aback | -0.021 |
coat | -0.019 |
like | -0.018 |
swe | -0.018 |
Token | Care |
Token ID | 7276 |
Feature activation | -6.78e+00 |
Pos logit contributions | |
wn | +0.002 |
aback | +0.002 |
coat | +0.001 |
swe | +0.001 |
like | +0.001 |
Neg logit contributions | |
glass | -0.002 |
orse | -0.002 |
generation | -0.002 |
work | -0.002 |
planet | -0.001 |
Token | fully |
Token ID | 2759 |
Feature activation | +3.38e+00 |
Pos logit contributions | |
aback | +0.295 |
wn | +0.291 |
like | +0.264 |
coat | +0.260 |
swe | +0.243 |
Neg logit contributions | |
glass | -0.454 |
orse | -0.447 |
generation | -0.407 |
work | -0.389 |
bike | -0.385 |
Token | the |
Token ID | 262 |
Feature activation | -7.23e-01 |
Pos logit contributions | |
glass | +0.245 |
orse | +0.242 |
generation | +0.223 |
work | +0.220 |
planet | +0.215 |
Neg logit contributions | |
wn | -0.128 |
aback | -0.126 |
coat | -0.110 |
swe | -0.106 |
like | -0.100 |
Token | tire |
Token ID | 15867 |
Feature activation | +1.72e-01 |
Pos logit contributions | |
wn | +0.040 |
aback | +0.039 |
ets | +0.035 |
swe | +0.033 |
coat | +0.033 |
Neg logit contributions | |
glass | -0.039 |
bike | -0.036 |
orse | -0.036 |
planet | -0.034 |
generation | -0.033 |
Token | opened |
Token ID | 4721 |
Feature activation | -1.76e+00 |
Pos logit contributions | |
glass | +0.011 |
orse | +0.011 |
generation | +0.010 |
work | +0.010 |
planet | +0.010 |
Neg logit contributions | |
aback | -0.007 |
wn | -0.007 |
coat | -0.007 |
swe | -0.006 |
like | -0.006 |
Token | the |
Token ID | 262 |
Feature activation | -7.07e-01 |
Pos logit contributions | |
wn | +0.092 |
aback | +0.081 |
coat | +0.080 |
swe | +0.073 |
ets | +0.072 |
Neg logit contributions | |
glass | -0.101 |
orse | -0.100 |
generation | -0.095 |
work | -0.094 |
planet | -0.093 |
Token | different |
Token ID | 1180 |
Feature activation | -1.25e+00 |
Pos logit contributions | |
orse | +0.095 |
glass | +0.092 |
work | +0.086 |
generation | +0.083 |
planet | +0.083 |
Neg logit contributions | |
aback | -0.076 |
wn | -0.075 |
coat | -0.066 |
inters | -0.061 |
swe | -0.059 |
Token | colors |
Token ID | 7577 |
Feature activation | -5.12e+00 |
Pos logit contributions | |
wn | +0.064 |
aback | +0.056 |
coat | +0.051 |
inters | +0.046 |
arl | +0.045 |
Neg logit contributions | |
work | -0.076 |
planet | -0.074 |
glass | -0.074 |
orse | -0.071 |
generation | -0.070 |
Token | , |
Token ID | 11 |
Feature activation | +1.83e-02 |
Pos logit contributions | |
wn | +0.210 |
aback | +0.207 |
coat | +0.173 |
swe | +0.170 |
ripe | +0.157 |
Neg logit contributions | |
glass | -0.354 |
orse | -0.353 |
work | -0.333 |
planet | -0.320 |
generation | -0.318 |
Token | shapes |
Token ID | 15268 |
Feature activation | -2.91e+00 |
Pos logit contributions | |
glass | +0.001 |
orse | +0.001 |
generation | +0.001 |
work | +0.001 |
planet | +0.001 |
Neg logit contributions | |
wn | -0.001 |
aback | -0.001 |
coat | -0.000 |
swe | -0.000 |
ripe | -0.000 |
Token | and |
Token ID | 290 |
Feature activation | +8.34e-02 |
Pos logit contributions | |
wn | +0.129 |
aback | +0.125 |
coat | +0.107 |
swe | +0.104 |
inters | +0.097 |
Neg logit contributions | |
glass | -0.193 |
orse | -0.192 |
work | -0.183 |
planet | -0.175 |
generation | -0.169 |
Token | sizes |
Token ID | 10620 |
Feature activation | -6.77e+00 |
Pos logit contributions | |
glass | +0.006 |
orse | +0.006 |
generation | +0.005 |
work | +0.005 |
planet | +0.005 |
Neg logit contributions | |
wn | -0.003 |
aback | -0.003 |
coat | -0.003 |
swe | -0.002 |
ripe | -0.002 |
Token | . |
Token ID | 13 |
Feature activation | -2.26e-01 |
Pos logit contributions | |
wn | +0.303 |
aback | +0.301 |
coat | +0.256 |
swe | +0.250 |
ripe | +0.233 |
Neg logit contributions | |
glass | -0.445 |
orse | -0.441 |
work | -0.416 |
planet | -0.396 |
generation | -0.393 |
Token | Everyone |
Token ID | 11075 |
Feature activation | +8.91e-01 |
Pos logit contributions | |
aback | +0.010 |
wn | +0.010 |
coat | +0.009 |
like | +0.009 |
swe | +0.009 |
Neg logit contributions | |
glass | -0.015 |
orse | -0.015 |
work | -0.013 |
generation | -0.013 |
bike | -0.013 |
Token | was |
Token ID | 373 |
Feature activation | +5.13e-02 |
Pos logit contributions | |
glass | +0.063 |
orse | +0.061 |
generation | +0.058 |
work | +0.055 |
planet | +0.054 |
Neg logit contributions | |
wn | -0.034 |
aback | -0.034 |
like | -0.030 |
coat | -0.030 |
swe | -0.030 |
Token | amazed |
Token ID | 24562 |
Feature activation | +1.78e+00 |
Pos logit contributions | |
orse | +0.003 |
glass | +0.003 |
work | +0.003 |
generation | +0.003 |
planet | +0.003 |
Neg logit contributions | |
wn | -0.002 |
aback | -0.002 |
coat | -0.002 |
swe | -0.002 |
inters | -0.002 |
Token | at |
Token ID | 379 |
Feature activation | +2.42e+00 |
Pos logit contributions | |
work | +0.109 |
glass | +0.103 |
orse | +0.102 |
whale | +0.100 |
planet | +0.099 |
Neg logit contributions | |
wn | -0.094 |
coat | -0.085 |
aback | -0.082 |
inters | -0.070 |
swe | -0.068 |
Token | her |
Token ID | 607 |
Feature activation | -1.77e+00 |
Pos logit contributions | |
glass | +0.066 |
orse | +0.064 |
generation | +0.060 |
work | +0.059 |
planet | +0.058 |
Neg logit contributions | |
wn | -0.041 |
aback | -0.040 |
coat | -0.035 |
swe | -0.033 |
ripe | -0.032 |
Token | rain |
Token ID | 6290 |
Feature activation | -3.58e-01 |
Pos logit contributions | |
wn | +0.073 |
aback | +0.072 |
coat | +0.063 |
swe | +0.061 |
like | +0.058 |
Neg logit contributions | |
glass | -0.121 |
orse | -0.120 |
generation | -0.110 |
work | -0.108 |
planet | -0.106 |
Token | boots |
Token ID | 14412 |
Feature activation | -3.70e+00 |
Pos logit contributions | |
aback | +0.008 |
wn | +0.008 |
ripe | +0.007 |
tast | +0.006 |
juicy | +0.006 |
Neg logit contributions | |
glass | -0.031 |
orse | -0.029 |
work | -0.028 |
bike | -0.028 |
generation | -0.027 |
Token | and |
Token ID | 290 |
Feature activation | +2.29e-01 |
Pos logit contributions | |
wn | +0.139 |
aback | +0.134 |
swe | +0.113 |
coat | +0.111 |
ripe | +0.109 |
Neg logit contributions | |
glass | -0.269 |
orse | -0.261 |
work | -0.243 |
generation | -0.243 |
bike | -0.236 |
Token | rain |
Token ID | 6290 |
Feature activation | -3.53e-01 |
Pos logit contributions | |
glass | +0.015 |
orse | +0.015 |
generation | +0.013 |
work | +0.013 |
planet | +0.013 |
Neg logit contributions | |
wn | -0.010 |
aback | -0.010 |
coat | -0.009 |
swe | -0.009 |
like | -0.008 |
Token | coat |
Token ID | 31434 |
Feature activation | -6.77e+00 |
Pos logit contributions | |
aback | +0.010 |
wn | +0.009 |
ripe | +0.008 |
tast | +0.007 |
juicy | +0.007 |
Neg logit contributions | |
glass | -0.028 |
work | -0.027 |
orse | -0.026 |
bike | -0.026 |
generation | -0.025 |
Token | so |
Token ID | 523 |
Feature activation | -1.42e+00 |
Pos logit contributions | |
wn | +0.269 |
aback | +0.263 |
swe | +0.218 |
ripe | +0.217 |
coat | +0.211 |
Neg logit contributions | |
glass | -0.475 |
orse | -0.460 |
work | -0.429 |
generation | -0.425 |
bike | -0.423 |
Token | she |
Token ID | 673 |
Feature activation | -1.07e+00 |
Pos logit contributions | |
wn | +0.083 |
aback | +0.082 |
coat | +0.073 |
swe | +0.070 |
ripe | +0.070 |
Neg logit contributions | |
orse | -0.071 |
glass | -0.071 |
generation | -0.063 |
work | -0.062 |
planet | -0.061 |
Token | wouldn |
Token ID | 3636 |
Feature activation | -2.26e+00 |
Pos logit contributions | |
wn | +0.042 |
aback | +0.042 |
coat | +0.036 |
swe | +0.035 |
ripe | +0.034 |
Neg logit contributions | |
glass | -0.075 |
orse | -0.075 |
generation | -0.068 |
work | -0.068 |
planet | -0.066 |
Token | 't |
Token ID | 470 |
Feature activation | +1.75e+00 |
Pos logit contributions | |
aback | +0.116 |
wn | +0.112 |
coat | +0.110 |
like | +0.107 |
swe | +0.100 |
Neg logit contributions | |
glass | -0.134 |
orse | -0.126 |
generation | -0.120 |
work | -0.117 |
planet | -0.112 |
Token | get |
Token ID | 651 |
Feature activation | +2.62e-01 |
Pos logit contributions | |
orse | +0.107 |
glass | +0.103 |
generation | +0.093 |
work | +0.092 |
planet | +0.092 |
Neg logit contributions | |
aback | -0.087 |
wn | -0.086 |
swe | -0.073 |
ripe | -0.071 |
coat | -0.071 |
Token | A |
Token ID | 317 |
Feature activation | +1.73e+00 |
Pos logit contributions | |
glass | +0.189 |
orse | +0.186 |
generation | +0.169 |
work | +0.166 |
planet | +0.162 |
Neg logit contributions | |
wn | -0.147 |
aback | -0.146 |
coat | -0.130 |
swe | -0.127 |
like | -0.123 |
Token | small |
Token ID | 1402 |
Feature activation | -2.91e+00 |
Pos logit contributions | |
glass | +0.113 |
orse | +0.111 |
generation | +0.102 |
work | +0.100 |
planet | +0.097 |
Neg logit contributions | |
aback | -0.076 |
wn | -0.075 |
coat | -0.067 |
swe | -0.066 |
sweet | -0.063 |
Token | girl |
Token ID | 2576 |
Feature activation | -3.10e+00 |
Pos logit contributions | |
aback | +0.151 |
wn | +0.149 |
coat | +0.137 |
swe | +0.134 |
like | +0.131 |
Neg logit contributions | |
glass | -0.168 |
orse | -0.165 |
generation | -0.149 |
work | -0.144 |
bike | -0.139 |
Token | with |
Token ID | 351 |
Feature activation | +3.60e-01 |
Pos logit contributions | |
wn | +0.132 |
aback | +0.130 |
coat | +0.115 |
swe | +0.112 |
like | +0.106 |
Neg logit contributions | |
glass | -0.209 |
orse | -0.206 |
generation | -0.190 |
work | -0.187 |
planet | -0.182 |
Token | curly |
Token ID | 45731 |
Feature activation | -3.52e+00 |
Pos logit contributions | |
glass | +0.025 |
orse | +0.024 |
generation | +0.022 |
work | +0.022 |
planet | +0.021 |
Neg logit contributions | |
wn | -0.015 |
aback | -0.015 |
coat | -0.013 |
swe | -0.013 |
like | -0.012 |
Token | blonde |
Token ID | 21541 |
Feature activation | -6.77e+00 |
Pos logit contributions | |
wn | +0.127 |
aback | +0.125 |
coat | +0.106 |
swe | +0.105 |
tast | +0.101 |
Neg logit contributions | |
glass | -0.260 |
orse | -0.260 |
generation | -0.236 |
work | -0.228 |
planet | -0.226 |
Token | hair |
Token ID | 4190 |
Feature activation | -3.13e+00 |
Pos logit contributions | |
wn | +0.213 |
aback | +0.209 |
coat | +0.176 |
swe | +0.170 |
like | +0.158 |
Neg logit contributions | |
glass | -0.532 |
orse | -0.525 |
generation | -0.489 |
work | -0.482 |
planet | -0.473 |
Token | was |
Token ID | 373 |
Feature activation | +2.00e-02 |
Pos logit contributions | |
wn | +0.101 |
aback | +0.100 |
coat | +0.090 |
swe | +0.084 |
like | +0.083 |
Neg logit contributions | |
glass | -0.238 |
orse | -0.236 |
generation | -0.224 |
work | -0.215 |
planet | -0.214 |
Token | standing |
Token ID | 5055 |
Feature activation | +1.05e+00 |
Pos logit contributions | |
glass | +0.001 |
orse | +0.001 |
generation | +0.001 |
work | +0.001 |
planet | +0.001 |
Neg logit contributions | |
wn | -0.001 |
aback | -0.001 |
coat | -0.001 |
swe | -0.001 |
like | -0.001 |
Token | there |
Token ID | 612 |
Feature activation | -1.57e+00 |
Pos logit contributions | |
glass | +0.079 |
orse | +0.079 |
generation | +0.073 |
work | +0.071 |
planet | +0.070 |
Neg logit contributions | |
wn | -0.036 |
aback | -0.036 |
coat | -0.031 |
swe | -0.030 |
like | -0.029 |
Token | . |
Token ID | 13 |
Feature activation | -2.41e-01 |
Pos logit contributions | |
wn | +0.062 |
aback | +0.061 |
coat | +0.053 |
swe | +0.052 |
ripe | +0.048 |
Neg logit contributions | |
glass | -0.111 |
orse | -0.109 |
generation | -0.101 |
work | -0.100 |
planet | -0.097 |
Token | wanted |
Token ID | 2227 |
Feature activation | -8.59e-01 |
Pos logit contributions | |
wn | +0.065 |
aback | +0.064 |
coat | +0.056 |
swe | +0.055 |
like | +0.053 |
Neg logit contributions | |
glass | -0.134 |
orse | -0.131 |
generation | -0.123 |
work | -0.122 |
planet | -0.118 |
Token | to |
Token ID | 284 |
Feature activation | -3.43e-01 |
Pos logit contributions | |
aback | +0.022 |
wn | +0.022 |
swe | +0.018 |
like | +0.017 |
coat | +0.017 |
Neg logit contributions | |
orse | -0.072 |
glass | -0.072 |
generation | -0.066 |
work | -0.066 |
bike | -0.065 |
Token | go |
Token ID | 467 |
Feature activation | -3.28e-01 |
Pos logit contributions | |
wn | +0.013 |
aback | +0.013 |
coat | +0.012 |
like | +0.012 |
swe | +0.012 |
Neg logit contributions | |
glass | -0.024 |
orse | -0.024 |
work | -0.022 |
generation | -0.022 |
bike | -0.021 |
Token | see |
Token ID | 766 |
Feature activation | +4.57e-01 |
Pos logit contributions | |
wn | +0.013 |
aback | +0.013 |
coat | +0.012 |
like | +0.012 |
swe | +0.011 |
Neg logit contributions | |
glass | -0.023 |
orse | -0.023 |
generation | -0.021 |
work | -0.021 |
planet | -0.020 |
Token | it |
Token ID | 340 |
Feature activation | +2.31e-01 |
Pos logit contributions | |
glass | +0.030 |
orse | +0.030 |
generation | +0.027 |
work | +0.027 |
bike | +0.026 |
Neg logit contributions | |
aback | -0.020 |
wn | -0.020 |
swe | -0.018 |
like | -0.018 |
coat | -0.017 |
Token | again |
Token ID | 757 |
Feature activation | +5.57e+00 |
Pos logit contributions | |
glass | +0.014 |
orse | +0.014 |
generation | +0.012 |
work | +0.012 |
planet | +0.012 |
Neg logit contributions | |
wn | -0.012 |
aback | -0.011 |
coat | -0.010 |
swe | -0.010 |
ripe | -0.010 |
Token | . |
Token ID | 13 |
Feature activation | -1.71e-01 |
Pos logit contributions | |
glass | +0.403 |
orse | +0.395 |
work | +0.364 |
generation | +0.360 |
bike | +0.354 |
Neg logit contributions | |
aback | -0.208 |
wn | -0.205 |
like | -0.185 |
swe | -0.181 |
coat | -0.180 |
Token | |
Token ID | 220 |
Feature activation | -1.90e+00 |
Pos logit contributions | |
wn | +0.010 |
aback | +0.008 |
coat | +0.008 |
ripe | +0.008 |
swe | +0.008 |
Neg logit contributions | |
orse | -0.009 |
glass | -0.009 |
generation | -0.008 |
work | -0.008 |
planet | -0.007 |
Token |
|
Token ID | 198 |
Feature activation | -7.15e-02 |
Pos logit contributions | |
wn | +0.104 |
aback | +0.100 |
swe | +0.092 |
coat | +0.092 |
ripe | +0.089 |
Neg logit contributions | |
glass | -0.104 |
orse | -0.104 |
generation | -0.092 |
work | -0.088 |
bike | -0.086 |
Token |
|
Token ID | 198 |
Feature activation | -8.19e-02 |
Pos logit contributions | |
wn | +0.004 |
aback | +0.004 |
coat | +0.003 |
swe | +0.003 |
ripe | +0.003 |
Neg logit contributions | |
glass | -0.004 |
orse | -0.004 |
generation | -0.004 |
planet | -0.003 |
work | -0.003 |
Token | L |
Token ID | 43 |
Feature activation | -1.52e+00 |
Pos logit contributions | |
wn | +0.003 |
aback | +0.003 |
coat | +0.003 |
swe | +0.003 |
like | +0.003 |
Neg logit contributions | |
glass | -0.006 |
orse | -0.006 |
generation | -0.005 |
work | -0.005 |
planet | -0.005 |
Token | should |
Token ID | 815 |
Feature activation | +1.13e+00 |
Pos logit contributions | |
orse | +0.018 |
glass | +0.017 |
work | +0.017 |
generation | +0.016 |
planet | +0.015 |
Neg logit contributions | |
wn | -0.014 |
aback | -0.013 |
coat | -0.012 |
swe | -0.012 |
ripe | -0.011 |
Token | have |
Token ID | 423 |
Feature activation | +1.93e+00 |
Pos logit contributions | |
glass | +0.073 |
orse | +0.071 |
generation | +0.065 |
work | +0.064 |
planet | +0.062 |
Neg logit contributions | |
wn | -0.052 |
aback | -0.051 |
coat | -0.046 |
swe | -0.046 |
like | -0.045 |
Token | given |
Token ID | 1813 |
Feature activation | -2.06e+00 |
Pos logit contributions | |
glass | +0.122 |
orse | +0.120 |
generation | +0.110 |
work | +0.109 |
bike | +0.105 |
Neg logit contributions | |
wn | -0.092 |
aback | -0.089 |
coat | -0.082 |
swe | -0.075 |
like | -0.072 |
Token | it |
Token ID | 340 |
Feature activation | -2.23e-01 |
Pos logit contributions | |
wn | +0.104 |
aback | +0.099 |
coat | +0.095 |
swe | +0.087 |
ripe | +0.084 |
Neg logit contributions | |
glass | -0.123 |
orse | -0.120 |
generation | -0.112 |
work | -0.110 |
planet | -0.106 |
Token | back |
Token ID | 736 |
Feature activation | +2.38e+00 |
Pos logit contributions | |
wn | +0.009 |
aback | +0.009 |
coat | +0.008 |
swe | +0.008 |
like | +0.007 |
Neg logit contributions | |
glass | -0.016 |
orse | -0.015 |
generation | -0.014 |
work | -0.014 |
planet | -0.014 |
Token | sooner |
Token ID | 14556 |
Feature activation | +5.56e+00 |
Pos logit contributions | |
glass | +0.173 |
orse | +0.171 |
generation | +0.158 |
work | +0.156 |
planet | +0.153 |
Neg logit contributions | |
wn | -0.090 |
aback | -0.088 |
coat | -0.077 |
swe | -0.075 |
like | -0.070 |
Token | . |
Token ID | 13 |
Feature activation | -3.28e-01 |
Pos logit contributions | |
glass | +0.435 |
orse | +0.426 |
generation | +0.401 |
work | +0.396 |
bike | +0.388 |
Neg logit contributions | |
aback | -0.163 |
wn | -0.162 |
like | -0.154 |
swe | -0.148 |
sweet | -0.142 |
Token | Can |
Token ID | 1680 |
Feature activation | +2.26e-01 |
Pos logit contributions | |
aback | +0.017 |
wn | +0.016 |
like | +0.015 |
sweet | +0.014 |
swe | +0.014 |
Neg logit contributions | |
glass | -0.020 |
orse | -0.019 |
work | -0.017 |
generation | -0.017 |
bike | -0.017 |
Token | you |
Token ID | 345 |
Feature activation | +1.34e+00 |
Pos logit contributions | |
glass | +0.016 |
generation | +0.014 |
orse | +0.014 |
growth | +0.014 |
work | +0.013 |
Neg logit contributions | |
sweet | -0.009 |
like | -0.009 |
swe | -0.008 |
aback | -0.008 |
past | -0.008 |
Token | forgive |
Token ID | 20927 |
Feature activation | -4.14e-01 |
Pos logit contributions | |
glass | +0.080 |
orse | +0.080 |
work | +0.071 |
generation | +0.071 |
planet | +0.070 |
Neg logit contributions | |
wn | -0.068 |
aback | -0.066 |
coat | -0.061 |
swe | -0.058 |
ripe | -0.056 |
Token | me |
Token ID | 502 |
Feature activation | -4.50e+00 |
Pos logit contributions | |
wn | +0.015 |
aback | +0.014 |
coat | +0.012 |
swe | +0.011 |
ripe | +0.011 |
Neg logit contributions | |
glass | -0.031 |
orse | -0.031 |
generation | -0.029 |
work | -0.028 |
planet | -0.028 |
Token | not |
Token ID | 407 |
Feature activation | -1.10e-02 |
Pos logit contributions | |
glass | +0.027 |
orse | +0.026 |
generation | +0.024 |
work | +0.023 |
planet | +0.022 |
Neg logit contributions | |
wn | -0.023 |
aback | -0.022 |
coat | -0.020 |
swe | -0.020 |
like | -0.020 |
Token | want |
Token ID | 765 |
Feature activation | -7.84e-01 |
Pos logit contributions | |
aback | +0.000 |
wn | +0.000 |
swe | +0.000 |
coat | +0.000 |
ripe | +0.000 |
Neg logit contributions | |
orse | -0.001 |
glass | -0.001 |
generation | -0.001 |
planet | -0.001 |
work | -0.001 |
Token | to |
Token ID | 284 |
Feature activation | -3.44e-01 |
Pos logit contributions | |
wn | +0.024 |
aback | +0.023 |
swe | +0.019 |
coat | +0.019 |
like | +0.018 |
Neg logit contributions | |
glass | -0.063 |
orse | -0.062 |
generation | -0.058 |
work | -0.057 |
planet | -0.056 |
Token | break |
Token ID | 2270 |
Feature activation | +6.45e-01 |
Pos logit contributions | |
wn | +0.015 |
aback | +0.015 |
coat | +0.013 |
swe | +0.013 |
like | +0.012 |
Neg logit contributions | |
glass | -0.023 |
orse | -0.023 |
generation | -0.021 |
work | -0.020 |
planet | -0.020 |
Token | it |
Token ID | 340 |
Feature activation | +1.71e-01 |
Pos logit contributions | |
orse | +0.042 |
glass | +0.041 |
generation | +0.037 |
work | +0.036 |
bike | +0.035 |
Neg logit contributions | |
aback | -0.030 |
wn | -0.029 |
swe | -0.027 |
like | -0.027 |
sweet | -0.026 |
Token | again |
Token ID | 757 |
Feature activation | +5.61e+00 |
Pos logit contributions | |
glass | +0.011 |
orse | +0.011 |
generation | +0.010 |
work | +0.010 |
planet | +0.009 |
Neg logit contributions | |
wn | -0.008 |
aback | -0.008 |
coat | -0.007 |
swe | -0.007 |
ripe | -0.007 |
Token | . |
Token ID | 13 |
Feature activation | -2.49e-01 |
Pos logit contributions | |
glass | +0.447 |
orse | +0.441 |
generation | +0.409 |
work | +0.408 |
bike | +0.398 |
Neg logit contributions | |
aback | -0.164 |
wn | -0.163 |
like | -0.144 |
swe | -0.141 |
coat | -0.139 |
Token | They |
Token ID | 1119 |
Feature activation | -2.36e-01 |
Pos logit contributions | |
aback | +0.009 |
wn | +0.008 |
like | +0.008 |
fun | +0.008 |
swe | +0.008 |
Neg logit contributions | |
glass | -0.018 |
orse | -0.018 |
bike | -0.017 |
work | -0.017 |
fer | -0.017 |
Token | also |
Token ID | 635 |
Feature activation | +1.80e+00 |
Pos logit contributions | |
wn | +0.010 |
aback | +0.010 |
coat | +0.009 |
swe | +0.009 |
ripe | +0.008 |
Neg logit contributions | |
glass | -0.016 |
orse | -0.015 |
generation | -0.014 |
work | -0.014 |
planet | -0.014 |
Token | gave |
Token ID | 2921 |
Feature activation | -2.83e+00 |
Pos logit contributions | |
glass | +0.120 |
orse | +0.119 |
generation | +0.109 |
work | +0.107 |
planet | +0.105 |
Neg logit contributions | |
aback | -0.076 |
wn | -0.076 |
coat | -0.065 |
swe | -0.063 |
ripe | -0.058 |
Token | some |
Token ID | 617 |
Feature activation | +7.82e-01 |
Pos logit contributions | |
aback | +0.128 |
wn | +0.122 |
swe | +0.121 |
sweet | +0.115 |
like | +0.114 |
Neg logit contributions | |
orse | -0.187 |
glass | -0.187 |
fer | -0.162 |
generation | -0.162 |
bike | -0.158 |
Token | , |
Token ID | 11 |
Feature activation | +1.12e-01 |
Pos logit contributions | |
glass | +0.077 |
orse | +0.076 |
generation | +0.071 |
work | +0.070 |
planet | +0.068 |
Neg logit contributions | |
wn | -0.036 |
aback | -0.035 |
coat | -0.030 |
swe | -0.029 |
like | -0.027 |
Token | Fin |
Token ID | 4463 |
Feature activation | +4.40e-01 |
Pos logit contributions | |
glass | +0.006 |
orse | +0.006 |
generation | +0.005 |
work | +0.005 |
planet | +0.005 |
Neg logit contributions | |
wn | -0.006 |
aback | -0.006 |
coat | -0.006 |
swe | -0.006 |
like | -0.005 |
Token | was |
Token ID | 373 |
Feature activation | +2.51e-01 |
Pos logit contributions | |
glass | +0.030 |
orse | +0.029 |
generation | +0.027 |
work | +0.026 |
planet | +0.026 |
Neg logit contributions | |
wn | -0.019 |
aback | -0.018 |
coat | -0.016 |
swe | -0.016 |
ripe | -0.015 |
Token | never |
Token ID | 1239 |
Feature activation | +1.53e+00 |
Pos logit contributions | |
glass | +0.016 |
orse | +0.016 |
generation | +0.015 |
work | +0.015 |
planet | +0.014 |
Neg logit contributions | |
wn | -0.011 |
aback | -0.011 |
coat | -0.010 |
swe | -0.010 |
ripe | -0.009 |
Token | lazy |
Token ID | 16931 |
Feature activation | -1.40e-01 |
Pos logit contributions | |
orse | +0.097 |
glass | +0.084 |
work | +0.084 |
Whale | +0.081 |
fer | +0.080 |
Neg logit contributions | |
wn | -0.070 |
coat | -0.065 |
aback | -0.061 |
arl | -0.058 |
swe | -0.056 |
Token | again |
Token ID | 757 |
Feature activation | +5.55e+00 |
Pos logit contributions | |
wn | +0.009 |
aback | +0.008 |
coat | +0.008 |
swe | +0.007 |
ripe | +0.007 |
Neg logit contributions | |
glass | -0.007 |
orse | -0.007 |
work | -0.006 |
generation | -0.006 |
planet | -0.006 |
Token | and |
Token ID | 290 |
Feature activation | +1.42e-01 |
Pos logit contributions | |
glass | +0.436 |
orse | +0.422 |
generation | +0.398 |
work | +0.395 |
planet | +0.385 |
Neg logit contributions | |
aback | -0.175 |
wn | -0.169 |
like | -0.154 |
swe | -0.144 |
coat | -0.139 |
Token | always |
Token ID | 1464 |
Feature activation | -3.72e+00 |
Pos logit contributions | |
glass | +0.010 |
orse | +0.009 |
generation | +0.009 |
work | +0.008 |
planet | +0.008 |
Neg logit contributions | |
aback | -0.006 |
wn | -0.006 |
like | -0.005 |
swe | -0.005 |
fun | -0.005 |
Token | had |
Token ID | 550 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
wn | +0.146 |
aback | +0.145 |
coat | +0.126 |
swe | +0.123 |
like | +0.118 |
Neg logit contributions | |
glass | -0.265 |
orse | -0.260 |
generation | -0.241 |
work | -0.237 |
planet | -0.232 |
Token | a |
Token ID | 257 |
Feature activation | -1.62e-01 |
Pos logit contributions | |
orse | +0.073 |
glass | +0.071 |
generation | +0.067 |
planet | +0.064 |
work | +0.062 |
Neg logit contributions | |
wn | -0.052 |
coat | -0.051 |
aback | -0.049 |
arl | -0.044 |
ucker | -0.041 |
Token | new |
Token ID | 649 |
Feature activation | -7.79e-02 |
Pos logit contributions | |
wn | +0.005 |
aback | +0.005 |
coat | +0.004 |
swe | +0.004 |
like | +0.004 |
Neg logit contributions | |
orse | -0.012 |
glass | -0.012 |
generation | -0.012 |
planet | -0.011 |
work | -0.011 |
Token | wanted |
Token ID | 2227 |
Feature activation | -5.47e-01 |
Pos logit contributions | |
wn | +0.039 |
aback | +0.038 |
coat | +0.034 |
swe | +0.032 |
ripe | +0.031 |
Neg logit contributions | |
glass | -0.066 |
orse | -0.066 |
generation | -0.060 |
work | -0.060 |
planet | -0.058 |
Token | to |
Token ID | 284 |
Feature activation | +3.29e-03 |
Pos logit contributions | |
aback | +0.014 |
wn | +0.014 |
swe | +0.012 |
coat | +0.012 |
like | +0.012 |
Neg logit contributions | |
glass | -0.045 |
orse | -0.045 |
generation | -0.042 |
work | -0.041 |
planet | -0.041 |
Token | ' |
Token ID | 705 |
Feature activation | -2.62e+00 |
Pos logit contributions | |
glass | +0.000 |
orse | +0.000 |
generation | +0.000 |
work | +0.000 |
planet | +0.000 |
Neg logit contributions | |
wn | -0.000 |
aback | -0.000 |
coat | -0.000 |
swe | -0.000 |
like | -0.000 |
Token | spl |
Token ID | 22018 |
Feature activation | +2.20e+00 |
Pos logit contributions | |
wn | +0.132 |
aback | +0.130 |
swe | +0.118 |
coat | +0.117 |
sweet | +0.110 |
Neg logit contributions | |
glass | -0.154 |
orse | -0.151 |
generation | -0.138 |
work | -0.136 |
bike | -0.132 |
Token | ash |
Token ID | 1077 |
Feature activation | +1.20e+00 |
Pos logit contributions | |
orse | +0.134 |
glass | +0.130 |
work | +0.124 |
bike | +0.117 |
generation | +0.115 |
Neg logit contributions | |
aback | -0.105 |
wn | -0.099 |
swe | -0.094 |
sweet | -0.092 |
ripe | -0.091 |
Token | ' |
Token ID | 6 |
Feature activation | +5.97e+00 |
Pos logit contributions | |
glass | +0.078 |
orse | +0.076 |
generation | +0.071 |
planet | +0.067 |
work | +0.066 |
Neg logit contributions | |
aback | -0.054 |
wn | -0.054 |
coat | -0.049 |
like | -0.048 |
swe | -0.047 |
Token | around |
Token ID | 1088 |
Feature activation | -2.23e+00 |
Pos logit contributions | |
glass | +0.429 |
orse | +0.423 |
generation | +0.391 |
work | +0.387 |
planet | +0.377 |
Neg logit contributions | |
wn | -0.227 |
aback | -0.223 |
coat | -0.194 |
swe | -0.188 |
like | -0.177 |
Token | , |
Token ID | 11 |
Feature activation | +1.36e-01 |
Pos logit contributions | |
wn | +0.089 |
aback | +0.087 |
coat | +0.077 |
swe | +0.074 |
ripe | +0.070 |
Neg logit contributions | |
glass | -0.156 |
orse | -0.153 |
generation | -0.142 |
work | -0.140 |
planet | -0.137 |
Token | but |
Token ID | 475 |
Feature activation | -2.16e+00 |
Pos logit contributions | |
glass | +0.010 |
orse | +0.010 |
generation | +0.009 |
work | +0.009 |
planet | +0.009 |
Neg logit contributions | |
wn | -0.005 |
aback | -0.005 |
coat | -0.004 |
swe | -0.004 |
ripe | -0.003 |
Token | she |
Token ID | 673 |
Feature activation | -9.84e-01 |
Pos logit contributions | |
wn | +0.104 |
aback | +0.104 |
coat | +0.095 |
swe | +0.094 |
like | +0.092 |
Neg logit contributions | |
glass | -0.133 |
orse | -0.128 |
generation | -0.118 |
work | -0.116 |
planet | -0.112 |
Token | wasn |
Token ID | 2492 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
wn | +0.045 |
aback | +0.044 |
coat | +0.039 |
swe | +0.038 |
ripe | +0.036 |
Neg logit contributions | |
glass | -0.064 |
orse | -0.063 |
generation | -0.057 |
work | -0.057 |
planet | -0.055 |
Token | . |
Token ID | 13 |
Feature activation | -1.64e-01 |
Pos logit contributions | |
glass | +0.157 |
orse | +0.149 |
generation | +0.143 |
growth | +0.138 |
ich | +0.137 |
Neg logit contributions | |
aback | -0.047 |
wn | -0.042 |
like | -0.042 |
swe | -0.037 |
sweet | -0.037 |
Token | |
Token ID | 220 |
Feature activation | -1.85e+00 |
Pos logit contributions | |
aback | +0.006 |
wn | +0.005 |
like | +0.004 |
swe | +0.004 |
sweet | +0.004 |
Neg logit contributions | |
glass | -0.013 |
orse | -0.012 |
work | -0.012 |
generation | -0.011 |
planet | -0.011 |
Token |
|
Token ID | 198 |
Feature activation | +1.40e-03 |
Pos logit contributions | |
wn | +0.089 |
aback | +0.088 |
coat | +0.079 |
swe | +0.078 |
ripe | +0.074 |
Neg logit contributions | |
glass | -0.114 |
orse | -0.113 |
generation | -0.103 |
work | -0.101 |
planet | -0.098 |
Token |
|
Token ID | 198 |
Feature activation | -1.07e-02 |
Pos logit contributions | |
glass | +0.000 |
orse | +0.000 |
work | +0.000 |
generation | +0.000 |
planet | +0.000 |
Neg logit contributions | |
aback | -0.000 |
wn | -0.000 |
coat | -0.000 |
swe | -0.000 |
like | -0.000 |
Token | Finally |
Token ID | 11158 |
Feature activation | -3.04e-01 |
Pos logit contributions | |
wn | +0.001 |
aback | +0.001 |
coat | +0.001 |
like | +0.001 |
swe | +0.001 |
Neg logit contributions | |
glass | -0.001 |
orse | -0.001 |
work | -0.001 |
generation | -0.001 |
planet | -0.000 |
Token | they |
Token ID | 484 |
Feature activation | +1.98e+00 |
Pos logit contributions | |
aback | +0.005 |
like | +0.004 |
past | +0.004 |
sweet | +0.004 |
wn | +0.004 |
Neg logit contributions | |
glass | -0.028 |
orse | -0.027 |
generation | -0.027 |
work | -0.026 |
bike | -0.026 |
Token | reached |
Token ID | 4251 |
Feature activation | -2.25e+00 |
Pos logit contributions | |
glass | +0.141 |
orse | +0.137 |
generation | +0.130 |
work | +0.127 |
planet | +0.123 |
Neg logit contributions | |
wn | -0.076 |
aback | -0.076 |
like | -0.064 |
coat | -0.064 |
swe | -0.063 |
Token | a |
Token ID | 257 |
Feature activation | -2.63e-01 |
Pos logit contributions | |
aback | +0.086 |
wn | +0.083 |
like | +0.075 |
swe | +0.072 |
ripe | +0.070 |
Neg logit contributions | |
glass | -0.165 |
orse | -0.159 |
generation | -0.148 |
work | -0.146 |
bike | -0.144 |
Token | good |
Token ID | 922 |
Feature activation | -2.98e+00 |
Pos logit contributions | |
wn | +0.010 |
aback | +0.010 |
coat | +0.009 |
swe | +0.009 |
like | +0.008 |
Neg logit contributions | |
glass | -0.019 |
orse | -0.018 |
generation | -0.017 |
work | -0.017 |
bike | -0.016 |
Token | spot |
Token ID | 4136 |
Feature activation | -1.80e+00 |
Pos logit contributions | |
wn | +0.163 |
aback | +0.161 |
coat | +0.147 |
swe | +0.144 |
ripe | +0.137 |
Neg logit contributions | |
glass | -0.165 |
orse | -0.165 |
generation | -0.147 |
work | -0.144 |
planet | -0.142 |
Token | to |
Token ID | 284 |
Feature activation | -4.65e-01 |
Pos logit contributions | |
aback | +0.056 |
wn | +0.055 |
coat | +0.048 |
like | +0.046 |
swe | +0.043 |
Neg logit contributions | |
glass | -0.141 |
orse | -0.140 |
generation | -0.131 |
planet | -0.126 |
work | -0.126 |
Token | and |
Token ID | 290 |
Feature activation | +1.21e-02 |
Pos logit contributions | |
glass | +0.095 |
orse | +0.094 |
bike | +0.086 |
work | +0.083 |
whale | +0.080 |
Neg logit contributions | |
wn | -0.073 |
aback | -0.067 |
ripe | -0.058 |
swe | -0.056 |
inters | -0.056 |
Token | Sue |
Token ID | 26089 |
Feature activation | -2.61e+00 |
Pos logit contributions | |
glass | +0.001 |
orse | +0.001 |
generation | +0.001 |
work | +0.001 |
planet | +0.001 |
Neg logit contributions | |
wn | -0.001 |
aback | -0.001 |
coat | -0.000 |
swe | -0.000 |
ets | -0.000 |
Token | went |
Token ID | 1816 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
wn | +0.121 |
aback | +0.119 |
swe | +0.099 |
coat | +0.098 |
ripe | +0.097 |
Neg logit contributions | |
glass | -0.165 |
orse | -0.164 |
bike | -0.146 |
work | -0.146 |
generation | -0.141 |
Token | on |
Token ID | 319 |
Feature activation | +1.07e+00 |
Pos logit contributions | |
wn | +0.042 |
aback | +0.042 |
swe | +0.034 |
coat | +0.034 |
ripe | +0.034 |
Neg logit contributions | |
glass | -0.070 |
orse | -0.067 |
work | -0.063 |
bike | -0.062 |
generation | -0.061 |
Token | the |
Token ID | 262 |
Feature activation | -5.42e-01 |
Pos logit contributions | |
orse | +0.089 |
glass | +0.087 |
work | +0.079 |
generation | +0.079 |
fer | +0.077 |
Neg logit contributions | |
aback | -0.032 |
fun | -0.032 |
past | -0.030 |
like | -0.029 |
sweet | -0.028 |
Token | slide |
Token ID | 10649 |
Feature activation | +3.33e+00 |
Pos logit contributions | |
aback | +0.033 |
wn | +0.033 |
coat | +0.029 |
swe | +0.028 |
like | +0.028 |
Neg logit contributions | |
glass | -0.027 |
orse | -0.027 |
generation | -0.023 |
work | -0.023 |
bike | -0.023 |
Token | and |
Token ID | 290 |
Feature activation | +4.56e-02 |
Pos logit contributions | |
glass | +0.252 |
orse | +0.250 |
generation | +0.232 |
work | +0.227 |
planet | +0.223 |
Neg logit contributions | |
wn | -0.114 |
aback | -0.113 |
coat | -0.100 |
swe | -0.095 |
like | -0.092 |
Token | had |
Token ID | 550 |
Feature activation | +1.15e+00 |
Pos logit contributions | |
glass | +0.003 |
orse | +0.003 |
generation | +0.003 |
work | +0.003 |
planet | +0.003 |
Neg logit contributions | |
wn | -0.002 |
aback | -0.002 |
coat | -0.002 |
swe | -0.001 |
like | -0.001 |
Token | fun |
Token ID | 1257 |
Feature activation | -4.06e+00 |
Pos logit contributions | |
glass | +0.082 |
orse | +0.080 |
generation | +0.074 |
work | +0.073 |
planet | +0.072 |
Neg logit contributions | |
wn | -0.044 |
aback | -0.043 |
coat | -0.037 |
swe | -0.035 |
ripe | -0.033 |
Token | . |
Token ID | 13 |
Feature activation | -2.64e-01 |
Pos logit contributions | |
wn | +0.131 |
aback | +0.126 |
like | +0.116 |
coat | +0.113 |
swe | +0.109 |
Neg logit contributions | |
glass | -0.315 |
orse | -0.305 |
generation | -0.292 |
work | -0.286 |
planet | -0.277 |
Token | They |
Token ID | 1119 |
Feature activation | -1.95e-01 |
Pos logit contributions | |
wn | +0.012 |
aback | +0.012 |
coat | +0.011 |
swe | +0.010 |
like | +0.010 |
Neg logit contributions | |
glass | -0.017 |
orse | -0.017 |
generation | -0.015 |
work | -0.015 |
planet | -0.015 |
Token | ily |
Token ID | 813 |
Feature activation | -8.72e-02 |
Pos logit contributions | |
aback | +0.089 |
wn | +0.087 |
coat | +0.080 |
swe | +0.080 |
ripe | +0.077 |
Neg logit contributions | |
glass | -0.074 |
orse | -0.072 |
generation | -0.065 |
work | -0.064 |
planet | -0.062 |
Token | was |
Token ID | 373 |
Feature activation | +2.54e-01 |
Pos logit contributions | |
wn | +0.004 |
aback | +0.004 |
coat | +0.004 |
swe | +0.004 |
ripe | +0.003 |
Neg logit contributions | |
glass | -0.005 |
orse | -0.005 |
generation | -0.005 |
work | -0.005 |
bike | -0.005 |
Token | sad |
Token ID | 6507 |
Feature activation | -3.69e-01 |
Pos logit contributions | |
orse | +0.018 |
glass | +0.018 |
generation | +0.016 |
work | +0.016 |
bike | +0.015 |
Neg logit contributions | |
wn | -0.010 |
aback | -0.010 |
coat | -0.009 |
swe | -0.008 |
ets | -0.008 |
Token | she |
Token ID | 673 |
Feature activation | -8.27e-01 |
Pos logit contributions | |
wn | +0.017 |
aback | +0.016 |
coat | +0.015 |
swe | +0.014 |
ripe | +0.014 |
Neg logit contributions | |
orse | -0.023 |
glass | -0.023 |
work | -0.021 |
generation | -0.021 |
planet | -0.020 |
Token | couldn |
Token ID | 3521 |
Feature activation | +2.70e+00 |
Pos logit contributions | |
wn | +0.049 |
aback | +0.049 |
swe | +0.044 |
ripe | +0.043 |
coat | +0.042 |
Neg logit contributions | |
glass | -0.041 |
orse | -0.040 |
work | -0.036 |
generation | -0.036 |
bike | -0.035 |
Token | 't |
Token ID | 470 |
Feature activation | +1.82e+00 |
Pos logit contributions | |
glass | +0.210 |
orse | +0.204 |
crew | +0.194 |
generation | +0.185 |
fer | +0.180 |
Neg logit contributions | |
aback | -0.091 |
coat | -0.077 |
time | -0.076 |
like | -0.075 |
wn | -0.075 |
Token | play |
Token ID | 711 |
Feature activation | +1.35e+00 |
Pos logit contributions | |
glass | +0.133 |
orse | +0.127 |
work | +0.121 |
generation | +0.119 |
ich | +0.119 |
Neg logit contributions | |
like | -0.067 |
wn | -0.066 |
aback | -0.062 |
coat | -0.058 |
sweet | -0.053 |
Token | outside |
Token ID | 2354 |
Feature activation | +1.89e-01 |
Pos logit contributions | |
glass | +0.092 |
orse | +0.089 |
generation | +0.083 |
work | +0.081 |
planet | +0.081 |
Neg logit contributions | |
wn | -0.058 |
aback | -0.057 |
coat | -0.049 |
swe | -0.048 |
ripe | -0.046 |
Token | , |
Token ID | 11 |
Feature activation | +3.94e-02 |
Pos logit contributions | |
glass | +0.014 |
orse | +0.014 |
generation | +0.012 |
work | +0.012 |
bike | +0.012 |
Neg logit contributions | |
aback | -0.007 |
wn | -0.007 |
like | -0.007 |
swe | -0.006 |
coat | -0.006 |
Token | but |
Token ID | 475 |
Feature activation | -2.09e+00 |
Pos logit contributions | |
glass | +0.003 |
orse | +0.003 |
generation | +0.003 |
work | +0.003 |
planet | +0.003 |
Neg logit contributions | |
wn | -0.001 |
aback | -0.001 |
coat | -0.001 |
swe | -0.001 |
like | -0.001 |
Token | she |
Token ID | 673 |
Feature activation | -8.31e-01 |
Pos logit contributions | |
aback | +0.065 |
wn | +0.063 |
like | +0.062 |
swe | +0.061 |
sweet | +0.059 |
Neg logit contributions | |
glass | -0.167 |
orse | -0.162 |
work | -0.151 |
generation | -0.148 |
fer | -0.146 |
Token | and |
Token ID | 290 |
Feature activation | +8.81e-02 |
Pos logit contributions | |
wn | +0.228 |
aback | +0.221 |
swe | +0.195 |
coat | +0.194 |
ripe | +0.175 |
Neg logit contributions | |
glass | -0.409 |
orse | -0.404 |
generation | -0.370 |
work | -0.364 |
planet | -0.355 |
Token | still |
Token ID | 991 |
Feature activation | -1.18e+00 |
Pos logit contributions | |
glass | +0.005 |
orse | +0.005 |
generation | +0.005 |
work | +0.005 |
planet | +0.005 |
Neg logit contributions | |
wn | -0.004 |
aback | -0.004 |
coat | -0.004 |
swe | -0.004 |
like | -0.004 |
Token | couldn |
Token ID | 3521 |
Feature activation | +2.71e+00 |
Pos logit contributions | |
wn | +0.052 |
aback | +0.051 |
coat | +0.045 |
like | +0.045 |
swe | +0.043 |
Neg logit contributions | |
glass | -0.077 |
orse | -0.076 |
generation | -0.070 |
work | -0.069 |
planet | -0.068 |
Token | 't |
Token ID | 470 |
Feature activation | +1.85e+00 |
Pos logit contributions | |
glass | +0.184 |
orse | +0.176 |
work | +0.164 |
crew | +0.162 |
whale | +0.162 |
Neg logit contributions | |
aback | -0.118 |
wn | -0.109 |
coat | -0.102 |
like | -0.101 |
time | -0.095 |
Token | find |
Token ID | 1064 |
Feature activation | +1.85e+00 |
Pos logit contributions | |
glass | +0.122 |
orse | +0.119 |
work | +0.115 |
bike | +0.111 |
ich | +0.110 |
Neg logit contributions | |
like | -0.078 |
aback | -0.077 |
wn | -0.076 |
coat | -0.066 |
idea | -0.063 |
Token | one |
Token ID | 530 |
Feature activation | +1.75e+00 |
Pos logit contributions | |
glass | +0.106 |
orse | +0.104 |
generation | +0.096 |
work | +0.093 |
planet | +0.092 |
Neg logit contributions | |
wn | -0.097 |
aback | -0.094 |
coat | -0.089 |
swe | -0.082 |
ripe | -0.079 |
Token | . |
Token ID | 13 |
Feature activation | -7.22e-02 |
Pos logit contributions | |
orse | +0.104 |
glass | +0.102 |
generation | +0.096 |
planet | +0.096 |
work | +0.094 |
Neg logit contributions | |
wn | -0.087 |
aback | -0.083 |
coat | -0.076 |
inters | -0.068 |
swe | -0.068 |
Token | He |
Token ID | 679 |
Feature activation | -2.29e-01 |
Pos logit contributions | |
wn | +0.003 |
aback | +0.003 |
coat | +0.003 |
swe | +0.003 |
ripe | +0.003 |
Neg logit contributions | |
glass | -0.004 |
orse | -0.004 |
generation | -0.004 |
bike | -0.004 |
work | -0.004 |
Token | was |
Token ID | 373 |
Feature activation | +2.24e-01 |
Pos logit contributions | |
wn | +0.009 |
aback | +0.009 |
coat | +0.008 |
swe | +0.008 |
like | +0.008 |
Neg logit contributions | |
glass | -0.016 |
orse | -0.016 |
generation | -0.014 |
work | -0.014 |
planet | -0.014 |
Token | getting |
Token ID | 1972 |
Feature activation | -6.76e-01 |
Pos logit contributions | |
glass | +0.015 |
orse | +0.015 |
generation | +0.014 |
work | +0.013 |
planet | +0.013 |
Neg logit contributions | |
wn | -0.010 |
aback | -0.009 |
coat | -0.009 |
swe | -0.008 |
ets | -0.008 |
Token | very |
Token ID | 845 |
Feature activation | -3.55e-01 |
Pos logit contributions | |
wn | +0.031 |
coat | +0.030 |
aback | +0.028 |
ets | +0.027 |
swe | +0.025 |
Neg logit contributions | |
glass | -0.041 |
orse | -0.040 |
generation | -0.037 |
work | -0.036 |
crew | -0.035 |
Token | had |
Token ID | 550 |
Feature activation | +9.34e-01 |
Pos logit contributions | |
wn | +0.046 |
aback | +0.046 |
coat | +0.040 |
swe | +0.039 |
like | +0.038 |
Neg logit contributions | |
glass | -0.075 |
orse | -0.074 |
generation | -0.068 |
work | -0.067 |
planet | -0.065 |
Token | to |
Token ID | 284 |
Feature activation | -4.08e-01 |
Pos logit contributions | |
glass | +0.063 |
orse | +0.061 |
generation | +0.058 |
work | +0.055 |
planet | +0.054 |
Neg logit contributions | |
wn | -0.038 |
aback | -0.037 |
coat | -0.032 |
swe | -0.029 |
ets | -0.029 |
Token | be |
Token ID | 307 |
Feature activation | +5.42e-01 |
Pos logit contributions | |
wn | +0.020 |
aback | +0.020 |
coat | +0.018 |
swe | +0.017 |
ripe | +0.017 |
Neg logit contributions | |
glass | -0.024 |
orse | -0.024 |
generation | -0.022 |
planet | -0.021 |
work | -0.021 |
Token | careful |
Token ID | 8161 |
Feature activation | -2.51e+00 |
Pos logit contributions | |
glass | +0.035 |
orse | +0.032 |
generation | +0.030 |
crew | +0.029 |
ich | +0.028 |
Neg logit contributions | |
aback | -0.022 |
wn | -0.019 |
coat | -0.019 |
arl | -0.018 |
inters | -0.018 |
Token | . |
Token ID | 13 |
Feature activation | -2.72e-01 |
Pos logit contributions | |
wn | +0.094 |
aback | +0.092 |
coat | +0.082 |
swe | +0.079 |
like | +0.078 |
Neg logit contributions | |
glass | -0.182 |
orse | -0.180 |
generation | -0.166 |
work | -0.163 |
planet | -0.161 |
Token | The |
Token ID | 383 |
Feature activation | +2.29e+00 |
Pos logit contributions | |
wn | +0.017 |
aback | +0.015 |
coat | +0.015 |
ripe | +0.014 |
ets | +0.014 |
Neg logit contributions | |
orse | -0.013 |
glass | -0.012 |
generation | -0.012 |
Our | -0.011 |
planet | -0.010 |
Token | ball |
Token ID | 2613 |
Feature activation | +1.05e-01 |
Pos logit contributions | |
glass | +0.159 |
orse | +0.158 |
generation | +0.146 |
work | +0.142 |
bike | +0.136 |
Neg logit contributions | |
wn | -0.091 |
aback | -0.088 |
swe | -0.085 |
coat | -0.085 |
sweet | -0.084 |
Token | was |
Token ID | 373 |
Feature activation | +1.78e-01 |
Pos logit contributions | |
glass | +0.007 |
orse | +0.007 |
work | +0.006 |
generation | +0.006 |
planet | +0.006 |
Neg logit contributions | |
wn | -0.005 |
aback | -0.005 |
coat | -0.004 |
swe | -0.004 |
ripe | -0.004 |
Token | very |
Token ID | 845 |
Feature activation | -3.51e-01 |
Pos logit contributions | |
orse | +0.010 |
glass | +0.009 |
work | +0.008 |
ich | +0.008 |
nobody | +0.008 |
Neg logit contributions | |
aback | -0.009 |
wn | -0.008 |
ets | -0.008 |
coat | -0.007 |
ere | -0.007 |
Token | big |
Token ID | 1263 |
Feature activation | -1.68e+00 |
Pos logit contributions | |
aback | +0.016 |
wn | +0.014 |
coat | +0.012 |
ets | +0.011 |
swe | +0.009 |
Neg logit contributions | |
orse | -0.025 |
glass | -0.022 |
planet | -0.021 |
whale | -0.021 |
ship | -0.020 |
Token | and |
Token ID | 290 |
Feature activation | +9.45e-02 |
Pos logit contributions | |
wn | +0.085 |
aback | +0.076 |
coat | +0.070 |
swe | +0.066 |
ripe | +0.065 |
Neg logit contributions | |
glass | -0.102 |
orse | -0.096 |
planet | -0.091 |
work | -0.091 |
generation | -0.089 |
Token | favorite |
Token ID | 4004 |
Feature activation | +3.71e+00 |
Pos logit contributions | |
aback | +0.062 |
wn | +0.060 |
coat | +0.049 |
swe | +0.046 |
ripe | +0.045 |
Neg logit contributions | |
glass | -0.114 |
orse | -0.111 |
generation | -0.102 |
planet | -0.101 |
work | -0.101 |
Token | hat |
Token ID | 6877 |
Feature activation | -4.05e-01 |
Pos logit contributions | |
glass | +0.235 |
orse | +0.229 |
generation | +0.209 |
planet | +0.206 |
work | +0.205 |
Neg logit contributions | |
aback | -0.177 |
wn | -0.175 |
coat | -0.151 |
swe | -0.143 |
like | -0.143 |
Token | everywhere |
Token ID | 8347 |
Feature activation | -3.50e+00 |
Pos logit contributions | |
aback | +0.020 |
wn | +0.019 |
swe | +0.016 |
ripe | +0.016 |
coat | +0.016 |
Neg logit contributions | |
glass | -0.026 |
orse | -0.024 |
work | -0.023 |
generation | -0.022 |
planet | -0.022 |
Token | she |
Token ID | 673 |
Feature activation | -8.15e-01 |
Pos logit contributions | |
wn | +0.117 |
like | +0.109 |
sweet | +0.107 |
coat | +0.107 |
aback | +0.106 |
Neg logit contributions | |
glass | -0.261 |
orse | -0.258 |
generation | -0.242 |
work | -0.241 |
Whale | -0.238 |
Token | went |
Token ID | 1816 |
Feature activation | -1.01e+00 |
Pos logit contributions | |
aback | +0.036 |
wn | +0.035 |
coat | +0.031 |
swe | +0.030 |
ripe | +0.029 |
Neg logit contributions | |
glass | -0.055 |
orse | -0.053 |
generation | -0.049 |
work | -0.048 |
planet | -0.047 |
Token | . |
Token ID | 13 |
Feature activation | -9.94e-02 |
Pos logit contributions | |
aback | +0.046 |
wn | +0.045 |
ripe | +0.039 |
coat | +0.038 |
swe | +0.038 |
Neg logit contributions | |
glass | -0.066 |
orse | -0.062 |
generation | -0.060 |
work | -0.059 |
planet | -0.058 |
Token | It |
Token ID | 632 |
Feature activation | +5.13e-01 |
Pos logit contributions | |
wn | +0.005 |
aback | +0.005 |
coat | +0.004 |
ripe | +0.004 |
tast | +0.004 |
Neg logit contributions | |
orse | -0.005 |
glass | -0.005 |
generation | -0.005 |
planet | -0.005 |
work | -0.005 |
Token | was |
Token ID | 373 |
Feature activation | +2.72e-01 |
Pos logit contributions | |
glass | +0.031 |
orse | +0.031 |
generation | +0.028 |
work | +0.028 |
planet | +0.027 |
Neg logit contributions | |
wn | -0.025 |
aback | -0.025 |
coat | -0.022 |
swe | -0.022 |
ripe | -0.021 |
Token | a |
Token ID | 257 |
Feature activation | -1.27e-01 |
Pos logit contributions | |
orse | +0.011 |
glass | +0.011 |
planet | +0.011 |
nobody | +0.010 |
generation | +0.010 |
Neg logit contributions | |
aback | -0.018 |
wn | -0.016 |
ets | -0.015 |
ives | -0.015 |
eal | -0.014 |
Token | big |
Token ID | 1263 |
Feature activation | -1.61e+00 |
Pos logit contributions | |
aback | +0.004 |
wn | +0.004 |
coat | +0.003 |
swe | +0.003 |
ets | +0.003 |
Neg logit contributions | |
glass | -0.010 |
orse | -0.010 |
generation | -0.009 |
planet | -0.009 |
work | -0.009 |
Token | , |
Token ID | 11 |
Feature activation | +8.40e-02 |
Pos logit contributions | |
wn | +0.064 |
aback | +0.063 |
coat | +0.055 |
swe | +0.054 |
like | +0.051 |
Neg logit contributions | |
glass | -0.114 |
orse | -0.112 |
generation | -0.104 |
work | -0.102 |
planet | -0.099 |
Token | , |
Token ID | 11 |
Feature activation | +1.08e-01 |
Pos logit contributions | |
wn | +0.121 |
aback | +0.106 |
coat | +0.095 |
swe | +0.093 |
prett | +0.090 |
Neg logit contributions | |
orse | -0.116 |
glass | -0.112 |
planet | -0.105 |
work | -0.103 |
generation | -0.100 |
Token | me |
Token ID | 502 |
Feature activation | -4.22e+00 |
Pos logit contributions | |
orse | +0.006 |
glass | +0.006 |
whale | +0.005 |
planet | +0.005 |
Whale | +0.005 |
Neg logit contributions | |
wn | -0.006 |
aback | -0.006 |
coat | -0.005 |
inters | -0.005 |
swe | -0.004 |
Token | too |
Token ID | 1165 |
Feature activation | -2.09e+00 |
Pos logit contributions | |
wn | +0.171 |
aback | +0.164 |
swe | +0.162 |
gre | +0.146 |
ju | +0.142 |
Neg logit contributions | |
orse | -0.280 |
glass | -0.269 |
work | -0.268 |
generation | -0.239 |
Whale | -0.234 |
Token | !" |
Token ID | 2474 |
Feature activation | +2.33e+00 |
Pos logit contributions | |
wn | +0.097 |
aback | +0.092 |
coat | +0.080 |
swe | +0.077 |
prett | +0.071 |
Neg logit contributions | |
orse | -0.139 |
glass | -0.136 |
generation | -0.126 |
planet | -0.125 |
work | -0.123 |
Token | Ben |
Token ID | 3932 |
Feature activation | -1.05e+00 |
Pos logit contributions | |
orse | +0.131 |
glass | +0.129 |
generation | +0.115 |
work | +0.111 |
planet | +0.111 |
Neg logit contributions | |
wn | -0.129 |
aback | -0.124 |
coat | -0.115 |
swe | -0.110 |
ripe | -0.107 |
Token | says |
Token ID | 1139 |
Feature activation | +1.53e-01 |
Pos logit contributions | |
wn | +0.069 |
coat | +0.060 |
ripe | +0.056 |
inters | +0.056 |
arl | +0.056 |
Neg logit contributions | |
orse | -0.047 |
glass | -0.046 |
work | -0.042 |
whale | -0.041 |
Whale | -0.041 |
Token | . |
Token ID | 13 |
Feature activation | -1.42e-01 |
Pos logit contributions | |
glass | +0.011 |
orse | +0.011 |
generation | +0.010 |
work | +0.010 |
planet | +0.010 |
Neg logit contributions | |
wn | -0.006 |
aback | -0.006 |
coat | -0.005 |
swe | -0.005 |
ripe | -0.005 |
Token | " |
Token ID | 366 |
Feature activation | -1.42e+00 |
Pos logit contributions | |
wn | +0.008 |
aback | +0.008 |
coat | +0.007 |
swe | +0.007 |
ripe | +0.007 |
Neg logit contributions | |
orse | -0.008 |
glass | -0.008 |
Whale | -0.006 |
work | -0.006 |
generation | -0.006 |
Token | Can |
Token ID | 6090 |
Feature activation | +4.61e-01 |
Pos logit contributions | |
wn | +0.058 |
aback | +0.056 |
like | +0.056 |
coat | +0.052 |
ripe | +0.049 |
Neg logit contributions | |
orse | -0.095 |
glass | -0.092 |
work | -0.089 |
planet | -0.085 |
generation | -0.085 |
Token | I |
Token ID | 314 |
Feature activation | +4.63e-01 |
Pos logit contributions | |
glass | +0.030 |
generation | +0.028 |
orse | +0.027 |
harbour | +0.026 |
growth | +0.026 |
Neg logit contributions | |
sweet | -0.021 |
swe | -0.021 |
might | -0.020 |
like | -0.020 |
past | -0.020 |
Token | hold |
Token ID | 1745 |
Feature activation | -1.76e+00 |
Pos logit contributions | |
orse | +0.027 |
glass | +0.026 |
planet | +0.023 |
generation | +0.023 |
work | +0.022 |
Neg logit contributions | |
wn | -0.025 |
aback | -0.024 |
swe | -0.021 |
coat | -0.021 |
ripe | -0.021 |
Token | every |
Token ID | 790 |
Feature activation | -1.93e-01 |
Pos logit contributions | |
orse | +0.015 |
glass | +0.015 |
generation | +0.014 |
work | +0.013 |
planet | +0.013 |
Neg logit contributions | |
wn | -0.006 |
aback | -0.006 |
like | -0.006 |
coat | -0.006 |
swe | -0.005 |
Token | day |
Token ID | 1110 |
Feature activation | +1.12e+00 |
Pos logit contributions | |
wn | +0.009 |
aback | +0.008 |
coat | +0.008 |
swe | +0.008 |
like | +0.007 |
Neg logit contributions | |
glass | -0.013 |
orse | -0.012 |
work | -0.011 |
generation | -0.011 |
bike | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | +2.77e-03 |
Pos logit contributions | |
glass | +0.087 |
orse | +0.086 |
generation | +0.080 |
work | +0.079 |
bike | +0.078 |
Neg logit contributions | |
wn | -0.035 |
aback | -0.034 |
like | -0.030 |
coat | -0.030 |
swe | -0.028 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +2.55e-01 |
Pos logit contributions | |
glass | +0.000 |
orse | +0.000 |
work | +0.000 |
generation | +0.000 |
fer | +0.000 |
Neg logit contributions | |
aback | -0.000 |
wn | -0.000 |
coat | -0.000 |
like | -0.000 |
swe | -0.000 |
Token | Once |
Token ID | 7454 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
glass | +0.016 |
orse | +0.016 |
generation | +0.014 |
work | +0.014 |
planet | +0.014 |
Neg logit contributions | |
wn | -0.012 |
aback | -0.012 |
coat | -0.011 |
swe | -0.010 |
like | -0.010 |
Token | there |
Token ID | 612 |
Feature activation | -1.46e+00 |
Pos logit contributions | |
wn | +0.033 |
aback | +0.032 |
coat | +0.027 |
swe | +0.025 |
ripe | +0.023 |
Neg logit contributions | |
orse | -0.081 |
glass | -0.080 |
generation | -0.074 |
work | -0.074 |
planet | -0.073 |
Token | was |
Token ID | 373 |
Feature activation | +1.26e-01 |
Pos logit contributions | |
wn | +0.077 |
aback | +0.075 |
coat | +0.068 |
swe | +0.067 |
ripe | +0.064 |
Neg logit contributions | |
glass | -0.084 |
orse | -0.084 |
generation | -0.075 |
work | -0.074 |
planet | -0.073 |
Token | a |
Token ID | 257 |
Feature activation | -3.12e-01 |
Pos logit contributions | |
glass | +0.009 |
orse | +0.008 |
work | +0.008 |
generation | +0.008 |
bike | +0.008 |
Neg logit contributions | |
wn | -0.005 |
aback | -0.005 |
coat | -0.005 |
like | -0.004 |
swe | -0.004 |
Token | big |
Token ID | 1263 |
Feature activation | -1.59e+00 |
Pos logit contributions | |
wn | +0.008 |
aback | +0.008 |
coat | +0.007 |
swe | +0.006 |
like | +0.006 |
Neg logit contributions | |
glass | -0.026 |
orse | -0.026 |
generation | -0.024 |
work | -0.024 |
planet | -0.023 |
Token | dog |
Token ID | 3290 |
Feature activation | +1.52e+00 |
Pos logit contributions | |
wn | +0.070 |
aback | +0.069 |
coat | +0.067 |
swe | +0.066 |
sweet | +0.065 |
Neg logit contributions | |
glass | -0.103 |
orse | -0.098 |
generation | -0.092 |
work | -0.087 |
fer | -0.084 |
Token | . |
Token ID | 13 |
Feature activation | -4.41e-02 |
Pos logit contributions | |
glass | +0.102 |
orse | +0.095 |
generation | +0.093 |
work | +0.090 |
fer | +0.088 |
Neg logit contributions | |
like | -0.063 |
coat | -0.058 |
past | -0.057 |
wn | -0.057 |
aback | -0.056 |
Token | saw |
Token ID | 2497 |
Feature activation | +1.20e+00 |
Pos logit contributions | |
orse | +0.122 |
glass | +0.122 |
work | +0.107 |
generation | +0.104 |
whale | +0.103 |
Neg logit contributions | |
wn | -0.100 |
aback | -0.092 |
coat | -0.086 |
swe | -0.082 |
inters | -0.081 |
Token | this |
Token ID | 428 |
Feature activation | +2.03e+00 |
Pos logit contributions | |
glass | +0.059 |
orse | +0.059 |
whale | +0.055 |
work | +0.055 |
Whale | +0.053 |
Neg logit contributions | |
wn | -0.074 |
coat | -0.067 |
aback | -0.061 |
inters | -0.060 |
tast | -0.058 |
Token | and |
Token ID | 290 |
Feature activation | +2.11e-02 |
Pos logit contributions | |
glass | +0.145 |
orse | +0.142 |
generation | +0.132 |
work | +0.131 |
planet | +0.128 |
Neg logit contributions | |
wn | -0.080 |
aback | -0.077 |
coat | -0.067 |
swe | -0.065 |
ripe | -0.062 |
Token | came |
Token ID | 1625 |
Feature activation | -3.01e+00 |
Pos logit contributions | |
orse | +0.001 |
glass | +0.001 |
work | +0.001 |
generation | +0.001 |
bike | +0.001 |
Neg logit contributions | |
wn | -0.001 |
aback | -0.001 |
coat | -0.001 |
swe | -0.001 |
ets | -0.001 |
Token | over |
Token ID | 625 |
Feature activation | -6.19e-01 |
Pos logit contributions | |
wn | +0.135 |
aback | +0.123 |
coat | +0.118 |
swe | +0.111 |
tast | +0.111 |
Neg logit contributions | |
glass | -0.196 |
orse | -0.190 |
generation | -0.177 |
work | -0.175 |
planet | -0.167 |
Token | to |
Token ID | 284 |
Feature activation | -4.00e-01 |
Pos logit contributions | |
wn | +0.024 |
aback | +0.022 |
coat | +0.020 |
swe | +0.019 |
prett | +0.018 |
Neg logit contributions | |
glass | -0.045 |
orse | -0.044 |
generation | -0.041 |
work | -0.041 |
planet | -0.039 |
Token | give |
Token ID | 1577 |
Feature activation | -1.18e+00 |
Pos logit contributions | |
wn | +0.017 |
aback | +0.016 |
coat | +0.014 |
swe | +0.014 |
ripe | +0.013 |
Neg logit contributions | |
glass | -0.027 |
orse | -0.027 |
generation | -0.025 |
work | -0.025 |
planet | -0.024 |
Token | her |
Token ID | 607 |
Feature activation | -1.63e+00 |
Pos logit contributions | |
wn | +0.057 |
coat | +0.056 |
arl | +0.052 |
aback | +0.049 |
inters | +0.049 |
Neg logit contributions | |
orse | -0.063 |
glass | -0.061 |
generation | -0.058 |
work | -0.058 |
whale | -0.054 |
Token | a |
Token ID | 257 |
Feature activation | -1.96e-01 |
Pos logit contributions | |
wn | +0.075 |
aback | +0.074 |
coat | +0.066 |
swe | +0.064 |
like | +0.062 |
Neg logit contributions | |
glass | -0.105 |
orse | -0.103 |
generation | -0.095 |
work | -0.093 |
planet | -0.091 |
Token | hug |
Token ID | 16225 |
Feature activation | -1.28e+00 |
Pos logit contributions | |
wn | +0.008 |
aback | +0.007 |
ets | +0.007 |
coat | +0.007 |
inters | +0.006 |
Neg logit contributions | |
glass | -0.013 |
orse | -0.012 |
work | -0.012 |
planet | -0.012 |
movie | -0.012 |
Token | . |
Token ID | 13 |
Feature activation | -9.68e-02 |
Pos logit contributions | |
wn | +0.046 |
aback | +0.045 |
coat | +0.039 |
swe | +0.038 |
like | +0.036 |
Neg logit contributions | |
glass | -0.094 |
orse | -0.094 |
generation | -0.087 |
work | -0.085 |
planet | -0.083 |
Token | you |
Token ID | 345 |
Feature activation | +1.43e+00 |
Pos logit contributions | |
glass | +0.088 |
orse | +0.084 |
generation | +0.082 |
bike | +0.081 |
work | +0.079 |
Neg logit contributions | |
aback | -0.051 |
sweet | -0.050 |
wn | -0.049 |
like | -0.048 |
past | -0.047 |
Token | , |
Token ID | 11 |
Feature activation | +6.39e-02 |
Pos logit contributions | |
glass | +0.120 |
orse | +0.117 |
generation | +0.113 |
bike | +0.110 |
work | +0.108 |
Neg logit contributions | |
aback | -0.037 |
wn | -0.035 |
sweet | -0.033 |
like | -0.033 |
swe | -0.032 |
Token | conductor |
Token ID | 39206 |
Feature activation | +4.64e-01 |
Pos logit contributions | |
glass | +0.004 |
orse | +0.004 |
generation | +0.004 |
bike | +0.003 |
work | +0.003 |
Neg logit contributions | |
sweet | -0.003 |
swe | -0.003 |
like | -0.003 |
nice | -0.003 |
fun | -0.003 |
Token | ." |
Token ID | 526 |
Feature activation | +6.00e-01 |
Pos logit contributions | |
glass | +0.035 |
orse | +0.034 |
generation | +0.033 |
planet | +0.031 |
work | +0.031 |
Neg logit contributions | |
aback | -0.016 |
wn | -0.016 |
coat | -0.014 |
like | -0.013 |
swe | -0.013 |
Token | Anna |
Token ID | 11735 |
Feature activation | -1.78e+00 |
Pos logit contributions | |
glass | +0.044 |
work | +0.041 |
generation | +0.041 |
orse | +0.040 |
ich | +0.039 |
Neg logit contributions | |
aback | -0.021 |
like | -0.020 |
sweet | -0.019 |
past | -0.019 |
fun | -0.019 |
Token | said |
Token ID | 531 |
Feature activation | -1.08e+00 |
Pos logit contributions | |
aback | +0.043 |
wn | +0.038 |
coat | +0.036 |
swe | +0.035 |
like | +0.034 |
Neg logit contributions | |
glass | -0.149 |
generation | -0.144 |
orse | -0.143 |
work | -0.138 |
planet | -0.135 |
Token | . |
Token ID | 13 |
Feature activation | -3.22e-01 |
Pos logit contributions | |
aback | +0.035 |
wn | +0.034 |
coat | +0.030 |
swe | +0.029 |
like | +0.028 |
Neg logit contributions | |
glass | -0.084 |
orse | -0.082 |
generation | -0.078 |
work | -0.076 |
planet | -0.075 |
Token |
|
Token ID | 198 |
Feature activation | -1.08e-01 |
Pos logit contributions | |
wn | +0.016 |
aback | +0.015 |
coat | +0.014 |
swe | +0.013 |
ripe | +0.013 |
Neg logit contributions | |
glass | -0.020 |
orse | -0.020 |
generation | -0.018 |
work | -0.018 |
planet | -0.017 |
Token |
|
Token ID | 198 |
Feature activation | -1.30e-01 |
Pos logit contributions | |
aback | +0.004 |
wn | +0.004 |
coat | +0.004 |
swe | +0.004 |
sweet | +0.004 |
Neg logit contributions | |
glass | -0.008 |
orse | -0.007 |
work | -0.007 |
generation | -0.007 |
bike | -0.007 |
Token | " |
Token ID | 1 |
Feature activation | -6.26e-01 |
Pos logit contributions | |
wn | +0.004 |
aback | +0.004 |
coat | +0.003 |
swe | +0.003 |
sweet | +0.003 |
Neg logit contributions | |
glass | -0.010 |
orse | -0.010 |
work | -0.009 |
generation | -0.009 |
bike | -0.009 |
Token | You |
Token ID | 1639 |
Feature activation | -8.31e-01 |
Pos logit contributions | |
wn | +0.028 |
aback | +0.027 |
coat | +0.026 |
like | +0.025 |
sweet | +0.025 |
Neg logit contributions | |
glass | -0.040 |
orse | -0.039 |
generation | -0.036 |
work | -0.035 |
planet | -0.034 |
Token | the |
Token ID | 262 |
Feature activation | -6.65e-01 |
Pos logit contributions | |
glass | +0.194 |
orse | +0.186 |
work | +0.173 |
fer | +0.172 |
bike | +0.171 |
Neg logit contributions | |
aback | -0.066 |
fun | -0.061 |
like | -0.060 |
sweet | -0.053 |
nice | -0.052 |
Token | spot |
Token ID | 4136 |
Feature activation | -1.83e+00 |
Pos logit contributions | |
aback | +0.037 |
wn | +0.037 |
swe | +0.035 |
like | +0.035 |
sweet | +0.034 |
Neg logit contributions | |
glass | -0.036 |
orse | -0.035 |
work | -0.030 |
generation | -0.030 |
fer | -0.028 |
Token | for |
Token ID | 329 |
Feature activation | +7.55e-01 |
Pos logit contributions | |
aback | +0.060 |
wn | +0.059 |
coat | +0.052 |
like | +0.051 |
swe | +0.047 |
Neg logit contributions | |
glass | -0.140 |
orse | -0.138 |
generation | -0.128 |
work | -0.126 |
bike | -0.124 |
Token | a |
Token ID | 257 |
Feature activation | -2.65e-01 |
Pos logit contributions | |
orse | +0.056 |
work | +0.056 |
glass | +0.056 |
bike | +0.055 |
ich | +0.055 |
Neg logit contributions | |
fun | -0.025 |
wn | -0.025 |
like | -0.024 |
aback | -0.023 |
time | -0.022 |
Token | long |
Token ID | 890 |
Feature activation | +6.28e-02 |
Pos logit contributions | |
wn | +0.007 |
aback | +0.007 |
coat | +0.006 |
swe | +0.006 |
like | +0.006 |
Neg logit contributions | |
glass | -0.022 |
orse | -0.021 |
generation | -0.020 |
work | -0.020 |
bike | -0.019 |
Token | time |
Token ID | 640 |
Feature activation | -5.89e+00 |
Pos logit contributions | |
orse | +0.005 |
glass | +0.005 |
generation | +0.005 |
planet | +0.005 |
work | +0.005 |
Neg logit contributions | |
aback | -0.002 |
wn | -0.002 |
swe | -0.001 |
ripe | -0.001 |
tast | -0.001 |
Token | and |
Token ID | 290 |
Feature activation | +1.90e-02 |
Pos logit contributions | |
wn | +0.183 |
aback | +0.182 |
like | +0.155 |
coat | +0.154 |
swe | +0.144 |
Neg logit contributions | |
glass | -0.458 |
orse | -0.447 |
generation | -0.421 |
work | -0.417 |
bike | -0.409 |
Token | never |
Token ID | 1239 |
Feature activation | +1.43e+00 |
Pos logit contributions | |
glass | +0.001 |
orse | +0.001 |
work | +0.001 |
ich | +0.001 |
generation | +0.001 |
Neg logit contributions | |
aback | -0.001 |
wn | -0.001 |
like | -0.001 |
fun | -0.001 |
sweet | -0.001 |
Token | forgot |
Token ID | 16453 |
Feature activation | -7.27e-01 |
Pos logit contributions | |
orse | +0.097 |
glass | +0.094 |
work | +0.088 |
generation | +0.087 |
planet | +0.085 |
Neg logit contributions | |
wn | -0.061 |
aback | -0.058 |
coat | -0.053 |
swe | -0.050 |
ripe | -0.048 |
Token | this |
Token ID | 428 |
Feature activation | +2.08e+00 |
Pos logit contributions | |
wn | +0.033 |
aback | +0.032 |
coat | +0.029 |
swe | +0.027 |
ripe | +0.026 |
Neg logit contributions | |
glass | -0.047 |
orse | -0.047 |
generation | -0.043 |
work | -0.042 |
planet | -0.041 |
Token | special |
Token ID | 2041 |
Feature activation | -4.66e-01 |
Pos logit contributions | |
glass | +0.141 |
orse | +0.138 |
generation | +0.127 |
work | +0.125 |
planet | +0.122 |
Neg logit contributions | |
wn | -0.089 |
aback | -0.088 |
coat | -0.077 |
swe | -0.076 |
like | -0.073 |
Token | lift |
Token ID | 10303 |
Feature activation | -2.16e-01 |
Pos logit contributions | |
wn | +0.021 |
aback | +0.020 |
coat | +0.018 |
swe | +0.017 |
ripe | +0.016 |
Neg logit contributions | |
glass | -0.033 |
orse | -0.032 |
generation | -0.030 |
work | -0.029 |
planet | -0.029 |
Token | this |
Token ID | 428 |
Feature activation | +2.08e+00 |
Pos logit contributions | |
wn | +0.010 |
aback | +0.010 |
coat | +0.009 |
swe | +0.009 |
like | +0.008 |
Neg logit contributions | |
glass | -0.014 |
orse | -0.013 |
generation | -0.012 |
work | -0.012 |
planet | -0.012 |
Token | box |
Token ID | 3091 |
Feature activation | -2.75e-01 |
Pos logit contributions | |
glass | +0.120 |
planet | +0.114 |
generation | +0.109 |
orse | +0.108 |
work | +0.106 |
Neg logit contributions | |
aback | -0.107 |
wn | -0.107 |
coat | -0.093 |
ucker | -0.088 |
inters | -0.084 |
Token | !" |
Token ID | 2474 |
Feature activation | +2.69e+00 |
Pos logit contributions | |
wn | +0.011 |
aback | +0.011 |
coat | +0.010 |
swe | +0.009 |
ripe | +0.009 |
Neg logit contributions | |
glass | -0.019 |
orse | -0.018 |
work | -0.017 |
generation | -0.017 |
planet | -0.017 |
Token | He |
Token ID | 679 |
Feature activation | -1.97e-01 |
Pos logit contributions | |
glass | +0.162 |
orse | +0.159 |
generation | +0.144 |
work | +0.142 |
planet | +0.138 |
Neg logit contributions | |
wn | -0.136 |
aback | -0.133 |
coat | -0.121 |
swe | -0.119 |
ripe | -0.114 |
Token | bent |
Token ID | 17157 |
Feature activation | -3.26e+00 |
Pos logit contributions | |
wn | +0.007 |
aback | +0.007 |
coat | +0.006 |
swe | +0.006 |
ripe | +0.006 |
Neg logit contributions | |
glass | -0.015 |
orse | -0.014 |
generation | -0.013 |
work | -0.013 |
planet | -0.013 |
Token | down |
Token ID | 866 |
Feature activation | -4.65e-02 |
Pos logit contributions | |
wn | +0.123 |
aback | +0.114 |
coat | +0.101 |
swe | +0.099 |
tast | +0.094 |
Neg logit contributions | |
glass | -0.239 |
orse | -0.229 |
work | -0.214 |
bike | -0.213 |
generation | -0.211 |
Token | , |
Token ID | 11 |
Feature activation | -8.01e-03 |
Pos logit contributions | |
wn | +0.002 |
aback | +0.002 |
coat | +0.001 |
swe | +0.001 |
like | +0.001 |
Neg logit contributions | |
glass | -0.004 |
orse | -0.003 |
generation | -0.003 |
work | -0.003 |
planet | -0.003 |
Token | put |
Token ID | 1234 |
Feature activation | -2.39e+00 |
Pos logit contributions | |
wn | +0.000 |
aback | +0.000 |
coat | +0.000 |
swe | +0.000 |
ets | +0.000 |
Neg logit contributions | |
orse | -0.001 |
glass | -0.000 |
bike | -0.000 |
generation | -0.000 |
work | -0.000 |
Token | his |
Token ID | 465 |
Feature activation | +5.08e-01 |
Pos logit contributions | |
wn | +0.134 |
coat | +0.131 |
aback | +0.130 |
swe | +0.120 |
ucker | +0.117 |
Neg logit contributions | |
glass | -0.125 |
orse | -0.114 |
generation | -0.106 |
work | -0.105 |
planet | -0.101 |
Token | hands |
Token ID | 2832 |
Feature activation | -1.37e+00 |
Pos logit contributions | |
glass | +0.031 |
orse | +0.030 |
generation | +0.027 |
work | +0.027 |
planet | +0.027 |
Neg logit contributions | |
aback | -0.026 |
wn | -0.025 |
coat | -0.023 |
swe | -0.022 |
like | -0.021 |
Token | Max |
Token ID | 5436 |
Feature activation | -6.00e-01 |
Pos logit contributions | |
aback | +0.004 |
wn | +0.004 |
coat | +0.003 |
like | +0.003 |
swe | +0.003 |
Neg logit contributions | |
glass | -0.008 |
orse | -0.007 |
work | -0.007 |
fer | -0.007 |
generation | -0.007 |
Token | bark |
Token ID | 21405 |
Feature activation | -4.55e-01 |
Pos logit contributions | |
wn | +0.031 |
aback | +0.029 |
coat | +0.026 |
swe | +0.026 |
ripe | +0.025 |
Neg logit contributions | |
orse | -0.036 |
glass | -0.035 |
generation | -0.031 |
bike | -0.031 |
work | -0.031 |
Token | ed |
Token ID | 276 |
Feature activation | +3.50e+00 |
Pos logit contributions | |
wn | +0.026 |
aback | +0.024 |
coat | +0.023 |
swe | +0.023 |
ripe | +0.022 |
Neg logit contributions | |
glass | -0.024 |
orse | -0.023 |
work | -0.021 |
generation | -0.021 |
bike | -0.020 |
Token | happily |
Token ID | 18177 |
Feature activation | -5.49e-01 |
Pos logit contributions | |
glass | +0.220 |
orse | +0.213 |
generation | +0.206 |
planet | +0.202 |
work | +0.201 |
Neg logit contributions | |
wn | -0.161 |
aback | -0.139 |
coat | -0.139 |
inters | -0.125 |
ucker | -0.123 |
Token | and |
Token ID | 290 |
Feature activation | -1.58e-02 |
Pos logit contributions | |
wn | +0.017 |
aback | +0.017 |
coat | +0.014 |
swe | +0.014 |
like | +0.013 |
Neg logit contributions | |
glass | -0.043 |
orse | -0.043 |
generation | -0.040 |
work | -0.039 |
planet | -0.038 |
Token | w |
Token ID | 266 |
Feature activation | -3.22e+00 |
Pos logit contributions | |
aback | +0.001 |
wn | +0.001 |
coat | +0.001 |
swe | +0.001 |
like | +0.001 |
Neg logit contributions | |
glass | -0.001 |
orse | -0.001 |
generation | -0.001 |
work | -0.001 |
planet | -0.001 |
Token | agged |
Token ID | 14655 |
Feature activation | -3.85e-01 |
Pos logit contributions | |
aback | +0.156 |
swe | +0.148 |
ju | +0.148 |
ripe | +0.147 |
wn | +0.142 |
Neg logit contributions | |
orse | -0.193 |
glass | -0.191 |
generation | -0.157 |
work | -0.156 |
crew | -0.150 |
Token | his |
Token ID | 465 |
Feature activation | +3.05e-01 |
Pos logit contributions | |
wn | +0.021 |
aback | +0.020 |
coat | +0.019 |
swe | +0.018 |
ripe | +0.017 |
Neg logit contributions | |
glass | -0.022 |
orse | -0.021 |
generation | -0.019 |
work | -0.019 |
planet | -0.018 |
Token | tail |
Token ID | 7894 |
Feature activation | -5.62e-01 |
Pos logit contributions | |
glass | +0.018 |
orse | +0.018 |
generation | +0.016 |
work | +0.016 |
planet | +0.016 |
Neg logit contributions | |
wn | -0.015 |
aback | -0.015 |
coat | -0.014 |
swe | -0.013 |
ripe | -0.013 |
Token | as |
Token ID | 355 |
Feature activation | -4.54e+00 |
Pos logit contributions | |
wn | +0.023 |
aback | +0.022 |
coat | +0.019 |
swe | +0.019 |
ripe | +0.018 |
Neg logit contributions | |
glass | -0.039 |
orse | -0.039 |
generation | -0.036 |
work | -0.035 |
planet | -0.035 |
Token | they |
Token ID | 484 |
Feature activation | +1.82e+00 |
Pos logit contributions | |
wn | +0.272 |
aback | +0.271 |
swe | +0.252 |
like | +0.250 |
coat | +0.250 |
Neg logit contributions | |
glass | -0.224 |
orse | -0.215 |
generation | -0.192 |
work | -0.190 |
bike | -0.178 |
Token | not |
Token ID | 407 |
Feature activation | -3.10e-01 |
Pos logit contributions | |
glass | +0.018 |
orse | +0.018 |
generation | +0.016 |
work | +0.016 |
planet | +0.015 |
Neg logit contributions | |
wn | -0.018 |
aback | -0.018 |
coat | -0.017 |
swe | -0.016 |
ripe | -0.016 |
Token | think |
Token ID | 892 |
Feature activation | +6.99e-01 |
Pos logit contributions | |
aback | +0.014 |
ucker | +0.013 |
coat | +0.013 |
swe | +0.013 |
ju | +0.012 |
Neg logit contributions | |
orse | -0.019 |
glass | -0.018 |
generation | -0.016 |
ship | -0.016 |
planet | -0.016 |
Token | Ben |
Token ID | 3932 |
Feature activation | -1.18e+00 |
Pos logit contributions | |
glass | +0.043 |
orse | +0.042 |
generation | +0.038 |
work | +0.038 |
planet | +0.036 |
Neg logit contributions | |
wn | -0.034 |
aback | -0.034 |
coat | -0.031 |
swe | -0.030 |
like | -0.029 |
Token | was |
Token ID | 373 |
Feature activation | +1.31e-01 |
Pos logit contributions | |
wn | +0.062 |
aback | +0.059 |
coat | +0.054 |
swe | +0.053 |
ripe | +0.052 |
Neg logit contributions | |
glass | -0.068 |
orse | -0.066 |
work | -0.060 |
generation | -0.060 |
bike | -0.059 |
Token | being |
Token ID | 852 |
Feature activation | +1.11e-01 |
Pos logit contributions | |
orse | +0.007 |
glass | +0.007 |
work | +0.007 |
whale | +0.007 |
Whale | +0.007 |
Neg logit contributions | |
aback | -0.006 |
wn | -0.006 |
coat | -0.006 |
swe | -0.005 |
inters | -0.005 |
Token | nice |
Token ID | 3621 |
Feature activation | -3.63e+00 |
Pos logit contributions | |
glass | +0.006 |
orse | +0.005 |
work | +0.005 |
whale | +0.005 |
fer | +0.005 |
Neg logit contributions | |
coat | -0.005 |
aback | -0.005 |
wn | -0.005 |
swe | -0.004 |
ucker | -0.004 |
Token | . |
Token ID | 13 |
Feature activation | -5.23e-01 |
Pos logit contributions | |
wn | +0.159 |
aback | +0.151 |
coat | +0.134 |
swe | +0.129 |
ripe | +0.127 |
Neg logit contributions | |
glass | -0.244 |
orse | -0.233 |
work | -0.220 |
generation | -0.215 |
planet | -0.211 |
Token | She |
Token ID | 1375 |
Feature activation | -2.12e+00 |
Pos logit contributions | |
wn | +0.015 |
aback | +0.015 |
coat | +0.013 |
swe | +0.012 |
like | +0.011 |
Neg logit contributions | |
glass | -0.042 |
orse | -0.042 |
generation | -0.039 |
work | -0.039 |
planet | -0.038 |
Token | had |
Token ID | 550 |
Feature activation | +9.60e-01 |
Pos logit contributions | |
wn | +0.102 |
aback | +0.100 |
coat | +0.087 |
swe | +0.082 |
ripe | +0.079 |
Neg logit contributions | |
glass | -0.133 |
orse | -0.130 |
generation | -0.115 |
bike | -0.115 |
whale | -0.113 |
Token | an |
Token ID | 281 |
Feature activation | +4.70e-01 |
Pos logit contributions | |
glass | +0.055 |
orse | +0.053 |
generation | +0.049 |
work | +0.047 |
bike | +0.046 |
Neg logit contributions | |
wn | -0.049 |
aback | -0.048 |
coat | -0.044 |
swe | -0.042 |
ets | -0.039 |
Token | idea |
Token ID | 2126 |
Feature activation | -5.54e+00 |
Pos logit contributions | |
glass | +0.041 |
orse | +0.040 |
generation | +0.038 |
planet | +0.037 |
work | +0.036 |
Neg logit contributions | |
aback | -0.010 |
wn | -0.010 |
coat | -0.008 |
swe | -0.007 |
gre | -0.006 |
Token | wish |
Token ID | 4601 |
Feature activation | +3.02e+00 |
Pos logit contributions | |
wn | +0.015 |
aback | +0.015 |
coat | +0.013 |
swe | +0.013 |
ripe | +0.012 |
Neg logit contributions | |
glass | -0.021 |
orse | -0.020 |
planet | -0.019 |
generation | -0.019 |
work | -0.018 |
Token | Max |
Token ID | 5436 |
Feature activation | -3.31e-01 |
Pos logit contributions | |
glass | +0.158 |
orse | +0.155 |
generation | +0.154 |
bike | +0.139 |
fer | +0.137 |
Neg logit contributions | |
sweet | -0.163 |
like | -0.161 |
swe | -0.159 |
fun | -0.158 |
wn | -0.156 |
Token | would |
Token ID | 561 |
Feature activation | -8.56e-01 |
Pos logit contributions | |
wn | +0.013 |
aback | +0.013 |
coat | +0.012 |
swe | +0.011 |
ripe | +0.011 |
Neg logit contributions | |
glass | -0.023 |
orse | -0.023 |
generation | -0.021 |
work | -0.021 |
planet | -0.020 |
Token | be |
Token ID | 307 |
Feature activation | +5.63e-01 |
Pos logit contributions | |
wn | +0.043 |
aback | +0.043 |
coat | +0.038 |
swe | +0.037 |
ripe | +0.036 |
Neg logit contributions | |
glass | -0.051 |
orse | -0.051 |
generation | -0.046 |
work | -0.045 |
planet | -0.044 |
Token | a |
Token ID | 257 |
Feature activation | -1.50e-01 |
Pos logit contributions | |
glass | +0.036 |
orse | +0.036 |
work | +0.032 |
generation | +0.031 |
crew | +0.031 |
Neg logit contributions | |
aback | -0.024 |
wn | -0.024 |
coat | -0.023 |
inters | -0.020 |
ucker | -0.019 |
Token | good |
Token ID | 922 |
Feature activation | -2.96e+00 |
Pos logit contributions | |
wn | +0.004 |
aback | +0.004 |
coat | +0.003 |
swe | +0.003 |
ripe | +0.002 |
Neg logit contributions | |
glass | -0.013 |
orse | -0.013 |
generation | -0.012 |
work | -0.012 |
planet | -0.012 |
Token | cat |
Token ID | 3797 |
Feature activation | +9.02e-01 |
Pos logit contributions | |
wn | +0.168 |
aback | +0.168 |
swe | +0.147 |
ripe | +0.147 |
coat | +0.147 |
Neg logit contributions | |
glass | -0.159 |
orse | -0.156 |
generation | -0.140 |
work | -0.139 |
planet | -0.138 |
Token | and |
Token ID | 290 |
Feature activation | +1.50e-01 |
Pos logit contributions | |
glass | +0.067 |
orse | +0.066 |
generation | +0.061 |
work | +0.060 |
planet | +0.059 |
Neg logit contributions | |
wn | -0.032 |
aback | -0.032 |
coat | -0.028 |
like | -0.027 |
swe | -0.027 |
Token | not |
Token ID | 407 |
Feature activation | -1.31e-02 |
Pos logit contributions | |
glass | +0.010 |
orse | +0.010 |
generation | +0.009 |
work | +0.009 |
planet | +0.009 |
Neg logit contributions | |
wn | -0.006 |
aback | -0.006 |
swe | -0.006 |
coat | -0.006 |
like | -0.006 |
Token | make |
Token ID | 787 |
Feature activation | -6.30e-01 |
Pos logit contributions | |
aback | +0.001 |
wn | +0.001 |
coat | +0.001 |
swe | +0.001 |
ripe | +0.000 |
Neg logit contributions | |
orse | -0.001 |
glass | -0.001 |
work | -0.001 |
generation | -0.001 |
planet | -0.001 |
Token | a |
Token ID | 257 |
Feature activation | -1.37e-01 |
Pos logit contributions | |
wn | +0.033 |
aback | +0.032 |
coat | +0.030 |
swe | +0.029 |
ripe | +0.028 |
Neg logit contributions | |
glass | -0.036 |
orse | -0.036 |
generation | -0.033 |
work | -0.032 |
planet | -0.031 |