Token | to |
Token ID | 284 |
Feature activation | -1.04e+00 |
Pos logit contributions | |
coconut | +0.148 |
Enjoy | +0.140 |
ours | +0.124 |
oral | +0.123 |
iw | +0.122 |
Neg logit contributions | |
aces | -0.150 |
coat | -0.149 |
robe | -0.146 |
amas | -0.143 |
brush | -0.140 |
Token | wear |
Token ID | 5806 |
Feature activation | +1.37e+00 |
Pos logit contributions | |
coconut | +0.067 |
Enjoy | +0.061 |
ours | +0.055 |
grown | +0.055 |
oral | +0.054 |
Neg logit contributions | |
aces | -0.086 |
coat | -0.085 |
robe | -0.084 |
amas | -0.082 |
brush | -0.082 |
Token | her |
Token ID | 607 |
Feature activation | +1.31e+00 |
Pos logit contributions | |
amas | +0.125 |
aces | +0.123 |
coat | +0.123 |
bridge | +0.120 |
robe | +0.119 |
Neg logit contributions | |
Enjoy | -0.069 |
coconut | -0.067 |
shrimp | -0.063 |
whose | -0.061 |
grown | -0.059 |
Token | cozy |
Token ID | 37438 |
Feature activation | +3.40e+00 |
Pos logit contributions | |
aces | +0.098 |
amas | +0.097 |
bridge | +0.095 |
coat | +0.095 |
robe | +0.092 |
Neg logit contributions | |
coconut | -0.091 |
Enjoy | -0.082 |
shrimp | -0.081 |
whose | -0.080 |
temple | -0.079 |
Token | bath |
Token ID | 7837 |
Feature activation | +2.52e+00 |
Pos logit contributions | |
aces | +0.218 |
bridge | +0.210 |
amas | +0.198 |
coat | +0.197 |
brush | +0.191 |
Neg logit contributions | |
coconut | -0.281 |
Enjoy | -0.253 |
shrimp | -0.249 |
volcano | -0.247 |
temple | -0.245 |
Token | robe |
Token ID | 25481 |
Feature activation | +1.15e+01 |
Pos logit contributions | |
bridge | +0.130 |
aces | +0.116 |
click | +0.115 |
pared | +0.113 |
brush | +0.112 |
Neg logit contributions | |
shrimp | -0.199 |
arium | -0.191 |
restaurant | -0.191 |
: | -0.190 |
swamp | -0.188 |
Token | after |
Token ID | 706 |
Feature activation | -1.83e+00 |
Pos logit contributions | |
amas | +1.175 |
bridge | +1.164 |
aces | +1.161 |
drops | +1.136 |
coat | +1.108 |
Neg logit contributions | |
Enjoy | -0.350 |
: | -0.344 |
arium | -0.328 |
shrimp | -0.307 |
coconut | -0.306 |
Token | taking |
Token ID | 2263 |
Feature activation | -6.55e-01 |
Pos logit contributions | |
coconut | +0.113 |
Enjoy | +0.103 |
ours | +0.090 |
grown | +0.089 |
temple | +0.089 |
Neg logit contributions | |
aces | -0.160 |
coat | -0.157 |
robe | -0.156 |
amas | -0.154 |
brush | -0.151 |
Token | a |
Token ID | 257 |
Feature activation | +5.71e-01 |
Pos logit contributions | |
Enjoy | +0.032 |
coconut | +0.029 |
grown | +0.028 |
shrimp | +0.028 |
oy | +0.026 |
Neg logit contributions | |
aces | -0.061 |
bridge | -0.061 |
amas | -0.061 |
coat | -0.061 |
robe | -0.060 |
Token | bath |
Token ID | 7837 |
Feature activation | +2.49e+00 |
Pos logit contributions | |
aces | +0.050 |
bridge | +0.045 |
amas | +0.045 |
robe | +0.044 |
coat | +0.043 |
Neg logit contributions | |
coconut | -0.038 |
Enjoy | -0.035 |
shrimp | -0.032 |
volcano | -0.031 |
temple | -0.031 |
Token | . |
Token ID | 13 |
Feature activation | -2.68e-02 |
Pos logit contributions | |
aces | +0.249 |
bridge | +0.246 |
brush | +0.224 |
amas | +0.222 |
click | +0.217 |
Neg logit contributions | |
shrimp | -0.084 |
Enjoy | -0.084 |
restaurant | -0.079 |
: | -0.078 |
swamp | -0.077 |
Token | Tim |
Token ID | 5045 |
Feature activation | -2.05e+00 |
Pos logit contributions | |
coconut | +0.024 |
Enjoy | +0.021 |
ours | +0.019 |
temple | +0.019 |
volcano | +0.018 |
Neg logit contributions | |
aces | -0.036 |
coat | -0.035 |
robe | -0.035 |
amas | -0.034 |
brush | -0.034 |
Token | saw |
Token ID | 2497 |
Feature activation | +7.49e-02 |
Pos logit contributions | |
coconut | +0.130 |
Enjoy | +0.120 |
asks | +0.106 |
grown | +0.106 |
ours | +0.105 |
Neg logit contributions | |
aces | -0.175 |
robe | -0.170 |
amas | -0.168 |
coat | -0.168 |
bridge | -0.165 |
Token | that |
Token ID | 326 |
Feature activation | +8.78e-01 |
Pos logit contributions | |
aces | +0.006 |
robe | +0.006 |
coat | +0.006 |
bridge | +0.006 |
brush | +0.006 |
Neg logit contributions | |
coconut | -0.005 |
Enjoy | -0.004 |
oral | -0.004 |
iw | -0.004 |
ours | -0.004 |
Token | his |
Token ID | 465 |
Feature activation | -5.67e-01 |
Pos logit contributions | |
robe | +0.076 |
coat | +0.076 |
aces | +0.075 |
brush | +0.075 |
bridge | +0.073 |
Neg logit contributions | |
coconut | -0.052 |
Enjoy | -0.050 |
ours | -0.044 |
oral | -0.043 |
iw | -0.042 |
Token | bath |
Token ID | 7837 |
Feature activation | +2.51e+00 |
Pos logit contributions | |
coconut | +0.038 |
Enjoy | +0.037 |
ours | +0.033 |
grown | +0.033 |
oral | +0.032 |
Neg logit contributions | |
aces | -0.045 |
robe | -0.045 |
coat | -0.045 |
brush | -0.043 |
amas | -0.043 |
Token | robe |
Token ID | 25481 |
Feature activation | +1.15e+01 |
Pos logit contributions | |
bridge | +0.140 |
aces | +0.140 |
brush | +0.126 |
amas | +0.120 |
coat | +0.118 |
Neg logit contributions | |
shrimp | -0.191 |
coconut | -0.186 |
arose | -0.182 |
crab | -0.181 |
swamp | -0.180 |
Token | was |
Token ID | 373 |
Feature activation | -3.68e-01 |
Pos logit contributions | |
aces | +1.320 |
amas | +1.288 |
bridge | +1.258 |
coat | +1.255 |
brush | +1.251 |
Neg logit contributions | |
coconut | -0.307 |
: | -0.237 |
temple | -0.231 |
Enjoy | -0.222 |
volcano | -0.219 |
Token | too |
Token ID | 1165 |
Feature activation | -1.16e+00 |
Pos logit contributions | |
coconut | +0.020 |
stealing | +0.017 |
grown | +0.017 |
shrimp | +0.016 |
oy | +0.016 |
Neg logit contributions | |
aces | -0.035 |
amas | -0.032 |
coat | -0.031 |
robe | -0.031 |
drops | -0.031 |
Token | tight |
Token ID | 5381 |
Feature activation | -2.07e+00 |
Pos logit contributions | |
coconut | +0.076 |
Enjoy | +0.068 |
grown | +0.062 |
ours | +0.061 |
shrimp | +0.061 |
Neg logit contributions | |
aces | -0.100 |
bridge | -0.096 |
coat | -0.095 |
robe | -0.095 |
amas | -0.094 |
Token | . |
Token ID | 13 |
Feature activation | -2.68e-02 |
Pos logit contributions | |
: | +0.056 |
Enjoy | +0.048 |
coconut | +0.047 |
stealing | +0.045 |
grown | +0.044 |
Neg logit contributions | |
bridge | -0.238 |
aces | -0.226 |
brush | -0.223 |
robe | -0.216 |
amas | -0.216 |
Token | He |
Token ID | 679 |
Feature activation | +4.06e-01 |
Pos logit contributions | |
coconut | +0.002 |
Enjoy | +0.002 |
oral | +0.002 |
temple | +0.002 |
grown | +0.002 |
Neg logit contributions | |
amas | -0.002 |
aces | -0.002 |
coat | -0.002 |
robe | -0.002 |
brush | -0.002 |
Token | they |
Token ID | 484 |
Feature activation | -2.67e-01 |
Pos logit contributions | |
aces | +0.016 |
coat | +0.016 |
robe | +0.016 |
amas | +0.016 |
bridge | +0.015 |
Neg logit contributions | |
coconut | -0.009 |
Enjoy | -0.008 |
temple | -0.007 |
volcano | -0.007 |
grown | -0.007 |
Token | always |
Token ID | 1464 |
Feature activation | +1.26e+00 |
Pos logit contributions | |
coconut | +0.017 |
Enjoy | +0.015 |
temple | +0.013 |
ours | +0.013 |
grown | +0.013 |
Neg logit contributions | |
aces | -0.023 |
coat | -0.023 |
robe | -0.023 |
amas | -0.022 |
bridge | -0.022 |
Token | threw |
Token ID | 9617 |
Feature activation | -2.80e-01 |
Pos logit contributions | |
aces | +0.111 |
coat | +0.108 |
robe | +0.108 |
amas | +0.107 |
bridge | +0.105 |
Neg logit contributions | |
coconut | -0.077 |
Enjoy | -0.071 |
ours | -0.062 |
temple | -0.061 |
grown | -0.061 |
Token | the |
Token ID | 262 |
Feature activation | +1.33e-01 |
Pos logit contributions | |
coconut | +0.016 |
Enjoy | +0.014 |
ours | +0.012 |
oral | +0.012 |
grown | +0.012 |
Neg logit contributions | |
aces | -0.026 |
coat | -0.025 |
robe | -0.025 |
amas | -0.025 |
brush | -0.024 |
Token | bath |
Token ID | 7837 |
Feature activation | +2.49e+00 |
Pos logit contributions | |
aces | +0.010 |
amas | +0.010 |
coat | +0.010 |
robe | +0.010 |
bridge | +0.010 |
Neg logit contributions | |
coconut | -0.010 |
Enjoy | -0.009 |
volcano | -0.008 |
temple | -0.008 |
grown | -0.008 |
Token | robe |
Token ID | 25481 |
Feature activation | +1.15e+01 |
Pos logit contributions | |
amas | +0.185 |
bridge | +0.181 |
aces | +0.178 |
coat | +0.164 |
brush | +0.163 |
Neg logit contributions | |
coconut | -0.156 |
shrimp | -0.152 |
crab | -0.146 |
volcano | -0.139 |
Enjoy | -0.136 |
Token | in |
Token ID | 287 |
Feature activation | -1.33e+00 |
Pos logit contributions | |
amas | +1.087 |
aces | +1.074 |
bridge | +1.045 |
robe | +1.008 |
coat | +1.005 |
Neg logit contributions | |
coconut | -0.554 |
Enjoy | -0.513 |
shrimp | -0.469 |
: | -0.463 |
grown | -0.457 |
Token | the |
Token ID | 262 |
Feature activation | +1.22e-01 |
Pos logit contributions | |
coconut | +0.067 |
Enjoy | +0.060 |
ours | +0.051 |
oral | +0.050 |
temple | +0.050 |
Neg logit contributions | |
aces | -0.131 |
coat | -0.129 |
robe | -0.128 |
amas | -0.126 |
brush | -0.125 |
Token | air |
Token ID | 1633 |
Feature activation | +3.20e+00 |
Pos logit contributions | |
aces | +0.010 |
coat | +0.009 |
robe | +0.009 |
amas | +0.009 |
brush | +0.009 |
Neg logit contributions | |
coconut | -0.009 |
Enjoy | -0.008 |
ours | -0.007 |
grown | -0.007 |
temple | -0.007 |
Token | when |
Token ID | 618 |
Feature activation | +2.04e+00 |
Pos logit contributions | |
aces | +0.321 |
amas | +0.315 |
robe | +0.309 |
bridge | +0.308 |
coat | +0.303 |
Neg logit contributions | |
coconut | -0.149 |
Enjoy | -0.121 |
shrimp | -0.113 |
temple | -0.113 |
volcano | -0.113 |
Token | they |
Token ID | 484 |
Feature activation | -3.25e-01 |
Pos logit contributions | |
robe | +0.181 |
aces | +0.179 |
coat | +0.178 |
amas | +0.177 |
brush | +0.175 |
Neg logit contributions | |
coconut | -0.117 |
Enjoy | -0.106 |
iw | -0.095 |
oral | -0.093 |
ours | -0.092 |
Token | - |
Token ID | 532 |
Feature activation | -1.67e+00 |
Pos logit contributions | |
aces | +0.080 |
amas | +0.080 |
coat | +0.079 |
bridge | +0.076 |
robe | +0.076 |
Neg logit contributions | |
coconut | -0.039 |
Enjoy | -0.036 |
temple | -0.032 |
whose | -0.031 |
volcano | -0.030 |
Token | a |
Token ID | 257 |
Feature activation | +5.83e-01 |
Pos logit contributions | |
coconut | +0.101 |
Enjoy | +0.092 |
ours | +0.080 |
temple | +0.079 |
grown | +0.079 |
Neg logit contributions | |
aces | -0.149 |
coat | -0.145 |
robe | -0.144 |
amas | -0.143 |
bridge | -0.140 |
Token | brand |
Token ID | 4508 |
Feature activation | +4.91e+00 |
Pos logit contributions | |
coat | +0.043 |
robe | +0.043 |
aces | +0.043 |
brush | +0.042 |
amas | +0.042 |
Neg logit contributions | |
coconut | -0.042 |
Enjoy | -0.039 |
ours | -0.036 |
grown | -0.035 |
oral | -0.035 |
Token | new |
Token ID | 649 |
Feature activation | -1.42e+00 |
Pos logit contributions | |
robe | +0.370 |
brush | +0.365 |
coat | +0.357 |
amas | +0.347 |
aces | +0.343 |
Neg logit contributions | |
coconut | -0.345 |
Enjoy | -0.311 |
grown | -0.302 |
ours | -0.290 |
colony | -0.288 |
Token | bath |
Token ID | 7837 |
Feature activation | +2.49e+00 |
Pos logit contributions | |
coconut | +0.100 |
Enjoy | +0.094 |
ours | +0.088 |
oral | +0.084 |
grown | +0.083 |
Neg logit contributions | |
coat | -0.109 |
robe | -0.109 |
aces | -0.106 |
brush | -0.105 |
amas | -0.102 |
Token | robe |
Token ID | 25481 |
Feature activation | +1.15e+01 |
Pos logit contributions | |
aces | +0.183 |
amas | +0.173 |
bridge | +0.171 |
brush | +0.167 |
coat | +0.166 |
Neg logit contributions | |
coconut | -0.163 |
Enjoy | -0.157 |
volcano | -0.151 |
shrimp | -0.150 |
temple | -0.150 |
Token | . |
Token ID | 13 |
Feature activation | -2.68e-02 |
Pos logit contributions | |
aces | +1.150 |
amas | +1.145 |
coat | +1.088 |
bridge | +1.084 |
brush | +1.076 |
Neg logit contributions | |
Enjoy | -0.472 |
coconut | -0.468 |
grown | -0.411 |
temple | -0.401 |
: | -0.390 |
Token | It |
Token ID | 632 |
Feature activation | +5.38e-01 |
Pos logit contributions | |
coconut | +0.002 |
Enjoy | +0.001 |
ours | +0.001 |
temple | +0.001 |
volcano | +0.001 |
Neg logit contributions | |
aces | -0.002 |
robe | -0.002 |
coat | -0.002 |
amas | -0.002 |
brush | -0.002 |
Token | was |
Token ID | 373 |
Feature activation | -3.87e-01 |
Pos logit contributions | |
aces | +0.061 |
amas | +0.061 |
coat | +0.059 |
bridge | +0.058 |
robe | +0.057 |
Neg logit contributions | |
coconut | -0.017 |
Enjoy | -0.016 |
asks | -0.014 |
whose | -0.013 |
temple | -0.013 |
Token | soft |
Token ID | 2705 |
Feature activation | +3.59e+00 |
Pos logit contributions | |
coconut | +0.023 |
Enjoy | +0.022 |
grown | +0.020 |
whose | +0.019 |
oy | +0.019 |
Neg logit contributions | |
aces | -0.034 |
amas | -0.034 |
coat | -0.033 |
robe | -0.032 |
bridge | -0.032 |
Token | , |
Token ID | 11 |
Feature activation | +8.25e-02 |
Pos logit contributions | |
aces | +0.393 |
bridge | +0.378 |
amas | +0.372 |
robe | +0.368 |
coat | +0.362 |
Neg logit contributions | |
coconut | -0.126 |
Enjoy | -0.116 |
shrimp | -0.109 |
grown | -0.106 |
oy | -0.100 |
Token | ; |
Token ID | 26 |
Feature activation | -1.27e+00 |
Pos logit contributions | |
aces | +0.004 |
coat | +0.004 |
amas | +0.004 |
bridge | +0.004 |
brush | +0.004 |
Neg logit contributions | |
coconut | -0.002 |
Enjoy | -0.002 |
volcano | -0.002 |
temple | -0.002 |
shrimp | -0.002 |
Token | he |
Token ID | 339 |
Feature activation | +5.82e-01 |
Pos logit contributions | |
coconut | +0.076 |
Enjoy | +0.068 |
oral | +0.063 |
ours | +0.061 |
grown | +0.061 |
Neg logit contributions | |
aces | -0.111 |
robe | -0.109 |
coat | -0.108 |
brush | -0.107 |
bridge | -0.105 |
Token | has |
Token ID | 468 |
Feature activation | -8.83e-01 |
Pos logit contributions | |
aces | +0.053 |
robe | +0.052 |
coat | +0.051 |
amas | +0.050 |
brush | +0.050 |
Neg logit contributions | |
coconut | -0.034 |
Enjoy | -0.031 |
ours | -0.027 |
oral | -0.027 |
grown | -0.026 |
Token | a |
Token ID | 257 |
Feature activation | +5.71e-01 |
Pos logit contributions | |
coconut | +0.052 |
Enjoy | +0.047 |
ours | +0.041 |
grown | +0.040 |
temple | +0.040 |
Neg logit contributions | |
aces | -0.080 |
coat | -0.078 |
robe | -0.078 |
amas | -0.077 |
brush | -0.076 |
Token | bath |
Token ID | 7837 |
Feature activation | +2.49e+00 |
Pos logit contributions | |
coat | +0.041 |
robe | +0.041 |
brush | +0.040 |
aces | +0.039 |
amas | +0.038 |
Neg logit contributions | |
coconut | -0.042 |
Enjoy | -0.040 |
ours | -0.038 |
oral | -0.037 |
grown | -0.036 |
Token | robe |
Token ID | 25481 |
Feature activation | +1.15e+01 |
Pos logit contributions | |
aces | +0.175 |
amas | +0.171 |
bridge | +0.165 |
brush | +0.157 |
coat | +0.155 |
Neg logit contributions | |
coconut | -0.170 |
shrimp | -0.162 |
Enjoy | -0.159 |
crab | -0.156 |
oy | -0.152 |
Token | too |
Token ID | 1165 |
Feature activation | -1.17e+00 |
Pos logit contributions | |
aces | +1.141 |
amas | +1.141 |
coat | +1.071 |
bridge | +1.055 |
robe | +1.050 |
Neg logit contributions | |
coconut | -0.499 |
Enjoy | -0.493 |
grown | -0.419 |
shrimp | -0.411 |
temple | -0.406 |
Token | ." |
Token ID | 526 |
Feature activation | +1.55e+00 |
Pos logit contributions | |
coconut | +0.076 |
Enjoy | +0.069 |
oral | +0.062 |
ours | +0.061 |
iw | +0.061 |
Neg logit contributions | |
aces | -0.100 |
coat | -0.098 |
robe | -0.097 |
amas | -0.095 |
brush | -0.093 |
Token |
|
Token ID | 198 |
Feature activation | +1.28e-01 |
Pos logit contributions | |
aces | +0.113 |
amas | +0.112 |
robe | +0.111 |
stand | +0.110 |
coat | +0.110 |
Neg logit contributions | |
coconut | -0.114 |
ours | -0.094 |
temple | -0.093 |
grown | -0.092 |
Enjoy | -0.092 |
Token |
|
Token ID | 198 |
Feature activation | +1.23e-01 |
Pos logit contributions | |
coat | +0.009 |
stand | +0.009 |
robe | +0.009 |
aces | +0.009 |
amas | +0.009 |
Neg logit contributions | |
coconut | -0.009 |
Enjoy | -0.008 |
ours | -0.008 |
volcano | -0.007 |
oy | -0.007 |
Token | Her |
Token ID | 9360 |
Feature activation | -3.97e-02 |
Pos logit contributions | |
coat | +0.008 |
brush | +0.008 |
bridge | +0.008 |
aces | +0.008 |
robe | +0.008 |
Neg logit contributions | |
coconut | -0.010 |
ours | -0.009 |
Enjoy | -0.009 |
volcano | -0.009 |
oy | -0.008 |
Token | are |
Token ID | 389 |
Feature activation | +1.49e+00 |
Pos logit contributions | |
aces | +0.004 |
coat | +0.004 |
robe | +0.004 |
amas | +0.004 |
brush | +0.004 |
Neg logit contributions | |
coconut | -0.003 |
Enjoy | -0.003 |
ours | -0.002 |
grown | -0.002 |
temple | -0.002 |
Token | you |
Token ID | 345 |
Feature activation | -3.73e-01 |
Pos logit contributions | |
aces | +0.107 |
robe | +0.099 |
coat | +0.099 |
amas | +0.098 |
brush | +0.097 |
Neg logit contributions | |
coconut | -0.113 |
Enjoy | -0.103 |
oral | -0.098 |
exhibit | -0.095 |
grown | -0.095 |
Token | wearing |
Token ID | 5762 |
Feature activation | +9.60e-01 |
Pos logit contributions | |
coconut | +0.023 |
Enjoy | +0.021 |
ours | +0.019 |
grown | +0.019 |
temple | +0.018 |
Neg logit contributions | |
aces | -0.033 |
coat | -0.032 |
robe | -0.031 |
bridge | -0.031 |
amas | -0.031 |
Token | the |
Token ID | 262 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
aces | +0.107 |
bridge | +0.104 |
coat | +0.102 |
amas | +0.102 |
drops | +0.101 |
Neg logit contributions | |
Enjoy | -0.031 |
coconut | -0.030 |
shrimp | -0.030 |
whose | -0.028 |
what | -0.027 |
Token | bath |
Token ID | 7837 |
Feature activation | +2.47e+00 |
Pos logit contributions | |
aces | +0.011 |
bridge | +0.011 |
drops | +0.010 |
brush | +0.010 |
amas | +0.010 |
Neg logit contributions | |
shrimp | -0.005 |
oy | -0.004 |
coconut | -0.004 |
crab | -0.004 |
fish | -0.004 |
Token | robe |
Token ID | 25481 |
Feature activation | +1.15e+01 |
Pos logit contributions | |
bridge | +0.112 |
aces | +0.111 |
brush | +0.099 |
coat | +0.095 |
drops | +0.089 |
Neg logit contributions | |
Enjoy | -0.225 |
coconut | -0.223 |
shrimp | -0.222 |
ours | -0.219 |
arium | -0.219 |
Token | ?" |
Token ID | 1701 |
Feature activation | -2.20e+00 |
Pos logit contributions | |
aces | +1.344 |
bridge | +1.301 |
coat | +1.266 |
amas | +1.255 |
robe | +1.235 |
Neg logit contributions | |
Enjoy | -0.307 |
coconut | -0.304 |
temple | -0.237 |
arium | -0.233 |
museum | -0.225 |
Token |
|
Token ID | 198 |
Feature activation | +1.27e-01 |
Pos logit contributions | |
coconut | +0.150 |
Enjoy | +0.124 |
volcano | +0.121 |
iw | +0.119 |
grown | +0.117 |
Neg logit contributions | |
aces | -0.183 |
amas | -0.178 |
robe | -0.176 |
coat | -0.172 |
brush | -0.169 |
Token |
|
Token ID | 198 |
Feature activation | +1.31e-01 |
Pos logit contributions | |
aces | +0.011 |
amas | +0.011 |
robe | +0.011 |
coat | +0.011 |
brush | +0.011 |
Neg logit contributions | |
coconut | -0.008 |
Enjoy | -0.006 |
iw | -0.006 |
volcano | -0.006 |
ours | -0.006 |
Token | Cl |
Token ID | 2601 |
Feature activation | -1.37e+00 |
Pos logit contributions | |
aces | +0.011 |
coat | +0.011 |
amas | +0.011 |
robe | +0.011 |
brush | +0.010 |
Neg logit contributions | |
coconut | -0.009 |
Enjoy | -0.008 |
ours | -0.007 |
volcano | -0.007 |
iw | -0.007 |
Token | are |
Token ID | 533 |
Feature activation | -8.26e-01 |
Pos logit contributions | |
coconut | +0.054 |
Enjoy | +0.044 |
volcano | +0.036 |
ours | +0.035 |
temple | +0.035 |
Neg logit contributions | |
aces | -0.152 |
coat | -0.149 |
robe | -0.148 |
amas | -0.148 |
brush | -0.145 |
Token | happy |
Token ID | 3772 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
coconut | +0.154 |
Enjoy | +0.142 |
ours | +0.127 |
grown | +0.126 |
temple | +0.126 |
Neg logit contributions | |
aces | -0.160 |
coat | -0.156 |
robe | -0.156 |
amas | -0.153 |
brush | -0.150 |
Token | and |
Token ID | 290 |
Feature activation | -6.88e-02 |
Pos logit contributions | |
aces | +0.100 |
coat | +0.098 |
robe | +0.097 |
amas | +0.096 |
brush | +0.094 |
Neg logit contributions | |
coconut | -0.075 |
Enjoy | -0.068 |
ours | -0.061 |
oral | -0.060 |
grown | -0.059 |
Token | wore |
Token ID | 12408 |
Feature activation | +1.31e+00 |
Pos logit contributions | |
coconut | +0.003 |
Enjoy | +0.003 |
oy | +0.003 |
shrimp | +0.003 |
Crab | +0.003 |
Neg logit contributions | |
aces | -0.007 |
coat | -0.007 |
amas | -0.007 |
bridge | -0.007 |
brush | -0.007 |
Token | the |
Token ID | 262 |
Feature activation | +1.05e-01 |
Pos logit contributions | |
coat | +0.168 |
amas | +0.166 |
drops | +0.164 |
aces | +0.164 |
robe | +0.161 |
Neg logit contributions | |
what | -0.021 |
Enjoy | -0.019 |
coconut | -0.019 |
grown | -0.011 |
shrimp | -0.011 |
Token | bath |
Token ID | 7837 |
Feature activation | +2.48e+00 |
Pos logit contributions | |
aces | +0.011 |
bridge | +0.011 |
drops | +0.010 |
brush | +0.010 |
ring | +0.010 |
Neg logit contributions | |
shrimp | -0.004 |
fish | -0.004 |
oy | -0.003 |
crab | -0.003 |
swamp | -0.003 |
Token | robe |
Token ID | 25481 |
Feature activation | +1.15e+01 |
Pos logit contributions | |
bridge | +0.096 |
amas | +0.085 |
aces | +0.084 |
brush | +0.080 |
pared | +0.079 |
Neg logit contributions | |
swamp | -0.221 |
: | -0.219 |
arium | -0.219 |
volcano | -0.218 |
shrimp | -0.217 |
Token | all |
Token ID | 477 |
Feature activation | -2.13e+00 |
Pos logit contributions | |
amas | +1.196 |
aces | +1.170 |
bridge | +1.160 |
coat | +1.140 |
drops | +1.096 |
Neg logit contributions | |
coconut | -0.437 |
Enjoy | -0.406 |
: | -0.368 |
exhibit | -0.363 |
temple | -0.353 |
Token | day |
Token ID | 1110 |
Feature activation | +1.32e-01 |
Pos logit contributions | |
coconut | +0.115 |
Enjoy | +0.104 |
ours | +0.089 |
grown | +0.088 |
temple | +0.088 |
Neg logit contributions | |
aces | -0.204 |
coat | -0.201 |
robe | -0.199 |
amas | -0.197 |
brush | -0.194 |
Token | . |
Token ID | 13 |
Feature activation | -4.54e-02 |
Pos logit contributions | |
bridge | +0.012 |
robe | +0.012 |
aces | +0.012 |
coat | +0.012 |
amas | +0.012 |
Neg logit contributions | |
coconut | -0.007 |
Enjoy | -0.006 |
shrimp | -0.006 |
temple | -0.006 |
exhibit | -0.006 |
Token |
|
Token ID | 198 |
Feature activation | +1.45e-01 |
Pos logit contributions | |
coconut | +0.003 |
temple | +0.003 |
oy | +0.003 |
grown | +0.003 |
crab | +0.003 |
Neg logit contributions | |
amas | -0.004 |
aces | -0.004 |
robe | -0.004 |
coat | -0.003 |
bridge | -0.003 |
Token |
|
Token ID | 198 |
Feature activation | +1.41e-01 |
Pos logit contributions | |
amas | +0.011 |
robe | +0.011 |
aces | +0.011 |
coat | +0.011 |
stand | +0.010 |
Neg logit contributions | |
coconut | -0.010 |
ours | -0.009 |
Enjoy | -0.009 |
oy | -0.008 |
shrimp | -0.008 |
Token | And |
Token ID | 843 |
Feature activation | -1.21e+00 |
Pos logit contributions | |
coconut | +0.002 |
Enjoy | +0.002 |
temple | +0.002 |
ours | +0.002 |
volcano | +0.002 |
Neg logit contributions | |
robe | -0.004 |
aces | -0.004 |
coat | -0.004 |
amas | -0.004 |
brush | -0.004 |
Token | they |
Token ID | 484 |
Feature activation | -3.08e-01 |
Pos logit contributions | |
coconut | +0.060 |
Enjoy | +0.053 |
ours | +0.045 |
grown | +0.044 |
temple | +0.044 |
Neg logit contributions | |
aces | -0.121 |
coat | -0.119 |
robe | -0.119 |
amas | -0.117 |
brush | -0.115 |
Token | loved |
Token ID | 6151 |
Feature activation | +4.57e-01 |
Pos logit contributions | |
coconut | +0.018 |
Enjoy | +0.016 |
ours | +0.014 |
grown | +0.014 |
temple | +0.014 |
Neg logit contributions | |
aces | -0.028 |
coat | -0.027 |
robe | -0.027 |
amas | -0.027 |
brush | -0.026 |
Token | their |
Token ID | 511 |
Feature activation | -1.24e+00 |
Pos logit contributions | |
robe | +0.035 |
aces | +0.034 |
coat | +0.034 |
amas | +0.032 |
stand | +0.032 |
Neg logit contributions | |
coconut | -0.033 |
oral | -0.027 |
Enjoy | -0.026 |
exhibit | -0.025 |
iw | -0.025 |
Token | bath |
Token ID | 7837 |
Feature activation | +2.48e+00 |
Pos logit contributions | |
coconut | +0.076 |
Enjoy | +0.070 |
shrimp | +0.064 |
temple | +0.064 |
volcano | +0.063 |
Neg logit contributions | |
aces | -0.111 |
bridge | -0.105 |
coat | -0.103 |
amas | -0.103 |
brush | -0.102 |
Token | robe |
Token ID | 25481 |
Feature activation | +1.15e+01 |
Pos logit contributions | |
bridge | +0.201 |
aces | +0.196 |
brush | +0.191 |
coat | +0.180 |
amas | +0.177 |
Neg logit contributions | |
shrimp | -0.130 |
swamp | -0.129 |
Enjoy | -0.129 |
volcano | -0.128 |
restaurant | -0.128 |
Token | , |
Token ID | 11 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
amas | +1.249 |
aces | +1.233 |
bridge | +1.214 |
coat | +1.191 |
brush | +1.156 |
Neg logit contributions | |
Enjoy | -0.362 |
coconut | -0.356 |
volcano | -0.318 |
restaurant | -0.304 |
temple | -0.296 |
Token | too |
Token ID | 1165 |
Feature activation | -1.17e+00 |
Pos logit contributions | |
aces | +0.008 |
coat | +0.008 |
robe | +0.008 |
amas | +0.008 |
bridge | +0.008 |
Neg logit contributions | |
coconut | -0.007 |
Enjoy | -0.007 |
grown | -0.006 |
ours | -0.006 |
temple | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | -4.54e-02 |
Pos logit contributions | |
coconut | +0.077 |
Enjoy | +0.070 |
ours | +0.061 |
iw | +0.061 |
oral | +0.060 |
Neg logit contributions | |
aces | -0.098 |
coat | -0.096 |
robe | -0.095 |
amas | -0.095 |
brush | -0.092 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -4.34e-01 |
Pos logit contributions | |
coconut | +0.002 |
Enjoy | +0.002 |
temple | +0.002 |
ours | +0.002 |
grown | +0.002 |
Neg logit contributions | |
aces | -0.004 |
robe | -0.004 |
amas | -0.004 |
coat | -0.004 |
brush | -0.004 |
Token | Anna |
Token ID | 31160 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
coconut | +0.021 |
Enjoy | +0.016 |
ours | +0.014 |
temple | +0.014 |
mistake | +0.014 |
Neg logit contributions | |
robe | -0.046 |
aces | -0.045 |
amas | -0.044 |
coat | -0.044 |
bridge | -0.043 |
Token | . |
Token ID | 13 |
Feature activation | -4.94e-02 |
Pos logit contributions | |
aces | +0.151 |
coat | +0.148 |
robe | +0.147 |
amas | +0.145 |
bridge | +0.143 |
Neg logit contributions | |
coconut | -0.093 |
Enjoy | -0.085 |
ours | -0.074 |
grown | -0.073 |
temple | -0.072 |
Token | They |
Token ID | 1119 |
Feature activation | -1.23e+00 |
Pos logit contributions | |
coconut | +0.003 |
Enjoy | +0.003 |
oy | +0.003 |
mole | +0.003 |
elia | +0.003 |
Neg logit contributions | |
coat | -0.004 |
aces | -0.004 |
amas | -0.004 |
bridge | -0.004 |
robe | -0.004 |
Token | used |
Token ID | 973 |
Feature activation | -1.28e+00 |
Pos logit contributions | |
coconut | +0.082 |
Enjoy | +0.075 |
volcano | +0.067 |
oral | +0.067 |
ours | +0.067 |
Neg logit contributions | |
coat | -0.103 |
aces | -0.102 |
robe | -0.099 |
brush | -0.098 |
bridge | -0.094 |
Token | the |
Token ID | 262 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
coconut | +0.059 |
Enjoy | +0.052 |
grown | +0.044 |
ours | +0.044 |
shrimp | +0.044 |
Neg logit contributions | |
aces | -0.131 |
amas | -0.129 |
coat | -0.129 |
robe | -0.128 |
bridge | -0.126 |
Token | bath |
Token ID | 7837 |
Feature activation | +2.47e+00 |
Pos logit contributions | |
aces | +0.010 |
bridge | +0.010 |
amas | +0.009 |
coat | +0.009 |
brush | +0.009 |
Neg logit contributions | |
shrimp | -0.006 |
coconut | -0.006 |
question | -0.005 |
oy | -0.005 |
swamp | -0.005 |
Token | robe |
Token ID | 25481 |
Feature activation | +1.15e+01 |
Pos logit contributions | |
bridge | +0.129 |
amas | +0.118 |
aces | +0.111 |
brush | +0.104 |
iver | +0.102 |
Neg logit contributions | |
shrimp | -0.211 |
ours | -0.209 |
coconut | -0.209 |
Enjoy | -0.204 |
: | -0.204 |
Token | to |
Token ID | 284 |
Feature activation | -5.48e-01 |
Pos logit contributions | |
amas | +1.403 |
aces | +1.356 |
bridge | +1.349 |
coat | +1.320 |
robe | +1.280 |
Neg logit contributions | |
coconut | -0.304 |
Enjoy | -0.251 |
shrimp | -0.209 |
: | -0.208 |
exhibit | -0.196 |
Token | pretend |
Token ID | 16614 |
Feature activation | +3.60e+00 |
Pos logit contributions | |
coconut | +0.031 |
Enjoy | +0.030 |
whose | +0.026 |
investigate | +0.026 |
shrimp | +0.026 |
Neg logit contributions | |
aces | -0.051 |
amas | -0.049 |
bridge | -0.048 |
robe | -0.046 |
coat | -0.045 |
Token | they |
Token ID | 484 |
Feature activation | -3.26e-01 |
Pos logit contributions | |
aces | +0.378 |
coat | +0.370 |
amas | +0.367 |
robe | +0.364 |
bridge | +0.358 |
Neg logit contributions | |
coconut | -0.158 |
Enjoy | -0.136 |
volcano | -0.120 |
temple | -0.119 |
grown | -0.115 |
Token | were |
Token ID | 547 |
Feature activation | +1.11e+00 |
Pos logit contributions | |
coconut | +0.014 |
Enjoy | +0.013 |
grown | +0.012 |
whose | +0.011 |
temple | +0.011 |
Neg logit contributions | |
amas | -0.034 |
aces | -0.033 |
bridge | -0.032 |
coat | -0.032 |
brush | -0.032 |
Token | superheroes |
Token ID | 40896 |
Feature activation | +2.22e+00 |
Pos logit contributions | |
amas | +0.104 |
aces | +0.104 |
brush | +0.103 |
coat | +0.102 |
bridge | +0.099 |
Neg logit contributions | |
coconut | -0.053 |
Enjoy | -0.053 |
shrimp | -0.051 |
whose | -0.049 |
grown | -0.048 |
Token | to |
Token ID | 284 |
Feature activation | -5.53e-01 |
Pos logit contributions | |
coconut | +0.039 |
Enjoy | +0.033 |
grown | +0.027 |
whose | +0.024 |
exhibit | +0.023 |
Neg logit contributions | |
amas | -0.143 |
bridge | -0.142 |
coat | -0.140 |
robe | -0.138 |
aces | -0.138 |
Token | mark |
Token ID | 1317 |
Feature activation | -1.72e+00 |
Pos logit contributions | |
coconut | +0.036 |
Enjoy | +0.033 |
oy | +0.029 |
temple | +0.028 |
whose | +0.028 |
Neg logit contributions | |
aces | -0.047 |
amas | -0.046 |
bridge | -0.045 |
robe | -0.045 |
coat | -0.044 |
Token | where |
Token ID | 810 |
Feature activation | +1.26e+00 |
Pos logit contributions | |
coconut | +0.091 |
Enjoy | +0.082 |
ours | +0.070 |
grown | +0.069 |
temple | +0.069 |
Neg logit contributions | |
aces | -0.165 |
coat | -0.162 |
robe | -0.161 |
amas | -0.159 |
brush | -0.156 |
Token | the |
Token ID | 262 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
aces | +0.116 |
robe | +0.113 |
brush | +0.111 |
coat | +0.110 |
bridge | +0.110 |
Neg logit contributions | |
coconut | -0.070 |
Enjoy | -0.065 |
oral | -0.056 |
ours | -0.056 |
iw | -0.056 |
Token | bath |
Token ID | 7837 |
Feature activation | +2.47e+00 |
Pos logit contributions | |
coat | +0.008 |
aces | +0.008 |
amas | +0.008 |
bridge | +0.008 |
robe | +0.008 |
Neg logit contributions | |
coconut | -0.008 |
Enjoy | -0.006 |
oy | -0.006 |
shrimp | -0.006 |
temple | -0.006 |
Token | robe |
Token ID | 25481 |
Feature activation | +1.15e+01 |
Pos logit contributions | |
bridge | +0.073 |
aces | +0.068 |
forehead | +0.066 |
coat | +0.066 |
amas | +0.063 |
Neg logit contributions | |
shrimp | -0.236 |
arose | -0.235 |
swamp | -0.234 |
coconut | -0.234 |
volcano | -0.233 |
Token | should |
Token ID | 815 |
Feature activation | -4.52e+00 |
Pos logit contributions | |
amas | +1.237 |
coat | +1.216 |
aces | +1.198 |
bridge | +1.168 |
robe | +1.165 |
Neg logit contributions | |
coconut | -0.405 |
temple | -0.328 |
Enjoy | -0.313 |
volcano | -0.311 |
: | -0.302 |
Token | be |
Token ID | 307 |
Feature activation | +8.01e-01 |
Pos logit contributions | |
coconut | +0.293 |
Enjoy | +0.265 |
temple | +0.235 |
volcano | +0.229 |
crab | +0.224 |
Neg logit contributions | |
amas | -0.380 |
aces | -0.378 |
coat | -0.373 |
robe | -0.366 |
bridge | -0.356 |
Token | cut |
Token ID | 2005 |
Feature activation | -1.58e-02 |
Pos logit contributions | |
amas | +0.071 |
robe | +0.069 |
coat | +0.068 |
aces | +0.067 |
bridge | +0.067 |
Neg logit contributions | |
coconut | -0.049 |
Enjoy | -0.042 |
grown | -0.040 |
shrimp | -0.039 |
whose | -0.039 |
Token | . |
Token ID | 13 |
Feature activation | -4.15e-02 |
Pos logit contributions | |
coconut | +0.001 |
Enjoy | +0.001 |
grown | +0.001 |
oy | +0.001 |
shrimp | +0.001 |
Neg logit contributions | |
coat | -0.002 |
amas | -0.001 |
bridge | -0.001 |
robe | -0.001 |
aces | -0.001 |
Token | He |
Token ID | 679 |
Feature activation | +3.98e-01 |
Pos logit contributions | |
coconut | +0.003 |
iw | +0.003 |
grown | +0.003 |
temple | +0.003 |
crab | +0.003 |
Neg logit contributions | |
amas | -0.003 |
aces | -0.003 |
robe | -0.003 |
bridge | -0.003 |
brush | -0.003 |
Token | too |
Token ID | 1165 |
Feature activation | -1.16e+00 |
Pos logit contributions | |
coconut | +0.022 |
Enjoy | +0.019 |
grown | +0.018 |
volcano | +0.017 |
stealing | +0.017 |
Neg logit contributions | |
aces | -0.031 |
robe | -0.031 |
amas | -0.031 |
coat | -0.030 |
bridge | -0.029 |
Token | full |
Token ID | 1336 |
Feature activation | +9.61e-01 |
Pos logit contributions | |
coconut | +0.087 |
Enjoy | +0.082 |
ours | +0.074 |
temple | +0.072 |
oral | +0.072 |
Neg logit contributions | |
aces | -0.085 |
coat | -0.084 |
robe | -0.083 |
amas | -0.082 |
brush | -0.081 |
Token | . |
Token ID | 13 |
Feature activation | -2.52e-02 |
Pos logit contributions | |
aces | +0.100 |
amas | +0.100 |
bridge | +0.099 |
robe | +0.099 |
coat | +0.099 |
Neg logit contributions | |
coconut | -0.038 |
Enjoy | -0.032 |
: | -0.030 |
grown | -0.029 |
oy | -0.028 |
Token | The |
Token ID | 383 |
Feature activation | -5.58e-01 |
Pos logit contributions | |
coconut | +0.002 |
temple | +0.002 |
Enjoy | +0.002 |
whale | +0.001 |
ours | +0.001 |
Neg logit contributions | |
aces | -0.002 |
amas | -0.002 |
robe | -0.002 |
coat | -0.002 |
brush | -0.002 |
Token | bath |
Token ID | 7837 |
Feature activation | +2.48e+00 |
Pos logit contributions | |
coconut | +0.039 |
Enjoy | +0.036 |
ours | +0.033 |
grown | +0.032 |
iw | +0.032 |
Neg logit contributions | |
aces | -0.044 |
coat | -0.043 |
robe | -0.043 |
amas | -0.042 |
brush | -0.042 |
Token | robe |
Token ID | 25481 |
Feature activation | +1.15e+01 |
Pos logit contributions | |
aces | +0.198 |
bridge | +0.183 |
amas | +0.181 |
coat | +0.180 |
brush | +0.175 |
Neg logit contributions | |
coconut | -0.153 |
volcano | -0.138 |
crab | -0.137 |
shrimp | -0.137 |
arose | -0.133 |
Token | was |
Token ID | 373 |
Feature activation | -3.88e-01 |
Pos logit contributions | |
aces | +1.206 |
amas | +1.190 |
coat | +1.160 |
robe | +1.139 |
bridge | +1.136 |
Neg logit contributions | |
coconut | -0.473 |
crab | -0.355 |
temple | -0.352 |
volcano | -0.344 |
Enjoy | -0.343 |
Token | too |
Token ID | 1165 |
Feature activation | -1.19e+00 |
Pos logit contributions | |
coconut | +0.023 |
grown | +0.020 |
Enjoy | +0.019 |
stealing | +0.019 |
shrimp | +0.018 |
Neg logit contributions | |
aces | -0.034 |
amas | -0.034 |
robe | -0.033 |
coat | -0.032 |
bridge | -0.032 |
Token | big |
Token ID | 1263 |
Feature activation | +1.29e+00 |
Pos logit contributions | |
coconut | +0.086 |
Enjoy | +0.079 |
ours | +0.071 |
grown | +0.070 |
temple | +0.070 |
Neg logit contributions | |
aces | -0.093 |
coat | -0.090 |
robe | -0.089 |
amas | -0.088 |
bridge | -0.086 |
Token | . |
Token ID | 13 |
Feature activation | -5.93e-02 |
Pos logit contributions | |
robe | +0.138 |
amas | +0.138 |
bridge | +0.138 |
coat | +0.136 |
aces | +0.135 |
Neg logit contributions | |
coconut | -0.042 |
Enjoy | -0.041 |
grown | -0.039 |
stealing | -0.038 |
: | -0.037 |
Token |
|
Token ID | 198 |
Feature activation | +1.11e-01 |
Pos logit contributions | |
coconut | +0.004 |
Enjoy | +0.003 |
temple | +0.003 |
ours | +0.003 |
grown | +0.003 |
Neg logit contributions | |
aces | -0.005 |
amas | -0.005 |
robe | -0.005 |
coat | -0.005 |
brush | -0.005 |
Token | had |
Token ID | 550 |
Feature activation | -9.72e-02 |
Pos logit contributions | |
coat | +0.031 |
aces | +0.031 |
robe | +0.031 |
brush | +0.030 |
amas | +0.029 |
Neg logit contributions | |
coconut | -0.025 |
Enjoy | -0.022 |
ours | -0.020 |
iw | -0.019 |
oral | -0.019 |
Token | a |
Token ID | 257 |
Feature activation | +5.71e-01 |
Pos logit contributions | |
coconut | +0.005 |
Enjoy | +0.004 |
grown | +0.004 |
temple | +0.004 |
whose | +0.004 |
Neg logit contributions | |
amas | -0.009 |
robe | -0.009 |
aces | -0.009 |
coat | -0.009 |
bridge | -0.009 |
Token | lovely |
Token ID | 14081 |
Feature activation | +6.93e-01 |
Pos logit contributions | |
coat | +0.040 |
robe | +0.038 |
brush | +0.037 |
amas | +0.035 |
aces | +0.035 |
Neg logit contributions | |
coconut | -0.045 |
Enjoy | -0.043 |
ours | -0.041 |
oral | -0.039 |
grown | -0.039 |
Token | blue |
Token ID | 4171 |
Feature activation | +1.46e+00 |
Pos logit contributions | |
aces | +0.052 |
coat | +0.052 |
robe | +0.051 |
amas | +0.050 |
brush | +0.049 |
Neg logit contributions | |
coconut | -0.051 |
Enjoy | -0.047 |
ours | -0.043 |
grown | -0.042 |
oral | -0.042 |
Token | bath |
Token ID | 7837 |
Feature activation | +2.48e+00 |
Pos logit contributions | |
coat | +0.088 |
robe | +0.088 |
aces | +0.086 |
brush | +0.084 |
amas | +0.082 |
Neg logit contributions | |
coconut | -0.128 |
Enjoy | -0.122 |
ours | -0.113 |
grown | -0.112 |
oral | -0.111 |
Token | robe |
Token ID | 25481 |
Feature activation | +1.15e+01 |
Pos logit contributions | |
aces | +0.111 |
bridge | +0.110 |
amas | +0.110 |
pared | +0.101 |
rell | +0.095 |
Neg logit contributions | |
shrimp | -0.215 |
: | -0.212 |
coconut | -0.210 |
Enjoy | -0.207 |
whose | -0.207 |
Token | . |
Token ID | 13 |
Feature activation | -4.70e-02 |
Pos logit contributions | |
amas | +1.242 |
aces | +1.218 |
bridge | +1.180 |
coat | +1.129 |
drops | +1.127 |
Neg logit contributions | |
Enjoy | -0.389 |
coconut | -0.342 |
: | -0.339 |
grown | -0.316 |
whose | -0.316 |
Token | Lucy |
Token ID | 22162 |
Feature activation | -1.37e+00 |
Pos logit contributions | |
coconut | +0.003 |
Enjoy | +0.003 |
oral | +0.003 |
temple | +0.003 |
oy | +0.003 |
Neg logit contributions | |
aces | -0.004 |
coat | -0.004 |
amas | -0.004 |
robe | -0.004 |
brush | -0.004 |
Token | loved |
Token ID | 6151 |
Feature activation | +4.49e-01 |
Pos logit contributions | |
coconut | +0.068 |
Enjoy | +0.062 |
ours | +0.053 |
grown | +0.053 |
whose | +0.052 |
Neg logit contributions | |
aces | -0.134 |
robe | -0.130 |
coat | -0.130 |
amas | -0.130 |
bridge | -0.128 |
Token | to |
Token ID | 284 |
Feature activation | -5.63e-01 |
Pos logit contributions | |
aces | +0.042 |
coat | +0.041 |
robe | +0.041 |
amas | +0.040 |
brush | +0.039 |
Neg logit contributions | |
coconut | -0.026 |
Enjoy | -0.023 |
ours | -0.020 |
oral | -0.020 |
temple | -0.019 |
Token | play |
Token ID | 711 |
Feature activation | +1.05e-01 |
Pos logit contributions | |
Enjoy | +0.030 |
coconut | +0.029 |
shrimp | +0.025 |
ours | +0.024 |
whose | +0.024 |
Neg logit contributions | |
aces | -0.053 |
amas | -0.051 |
bridge | -0.050 |
coat | -0.049 |
robe | -0.049 |
Token | about |
Token ID | 546 |
Feature activation | -1.06e+00 |
Pos logit contributions | |
coconut | +0.003 |
Enjoy | +0.002 |
iw | +0.002 |
Crab | +0.002 |
grown | +0.002 |
Neg logit contributions | |
aces | -0.003 |
brush | -0.003 |
coat | -0.003 |
robe | -0.003 |
bridge | -0.003 |
Token | we |
Token ID | 356 |
Feature activation | -4.00e-01 |
Pos logit contributions | |
coconut | +0.079 |
Enjoy | +0.072 |
oral | +0.069 |
iw | +0.067 |
grown | +0.064 |
Neg logit contributions | |
amas | -0.075 |
aces | -0.074 |
brush | -0.073 |
robe | -0.073 |
bridge | -0.071 |
Token | use |
Token ID | 779 |
Feature activation | -3.49e+00 |
Pos logit contributions | |
coconut | +0.027 |
Enjoy | +0.024 |
ours | +0.022 |
grown | +0.022 |
iw | +0.022 |
Neg logit contributions | |
aces | -0.033 |
coat | -0.033 |
robe | -0.032 |
amas | -0.032 |
brush | -0.032 |
Token | the |
Token ID | 262 |
Feature activation | +9.70e-02 |
Pos logit contributions | |
coconut | +0.195 |
Enjoy | +0.175 |
oral | +0.151 |
ours | +0.148 |
grown | +0.147 |
Neg logit contributions | |
aces | -0.327 |
coat | -0.319 |
robe | -0.317 |
amas | -0.313 |
brush | -0.311 |
Token | bath |
Token ID | 7837 |
Feature activation | +2.48e+00 |
Pos logit contributions | |
aces | +0.008 |
bridge | +0.007 |
amed | +0.007 |
amas | +0.007 |
coat | +0.007 |
Neg logit contributions | |
coconut | -0.006 |
oy | -0.006 |
shrimp | -0.006 |
museum | -0.006 |
temple | -0.006 |
Token | robe |
Token ID | 25481 |
Feature activation | +1.15e+01 |
Pos logit contributions | |
aces | +0.118 |
bridge | +0.115 |
coat | +0.106 |
brush | +0.102 |
amas | +0.101 |
Neg logit contributions | |
Enjoy | -0.225 |
coconut | -0.224 |
ours | -0.220 |
arium | -0.214 |
volcano | -0.212 |
Token | to |
Token ID | 284 |
Feature activation | -5.49e-01 |
Pos logit contributions | |
aces | +1.215 |
coat | +1.174 |
amas | +1.169 |
bridge | +1.162 |
robe | +1.143 |
Neg logit contributions | |
coconut | -0.471 |
Enjoy | -0.450 |
ours | -0.362 |
shrimp | -0.347 |
temple | -0.347 |
Token | make |
Token ID | 787 |
Feature activation | +3.57e+00 |
Pos logit contributions | |
coconut | +0.037 |
Enjoy | +0.035 |
temple | +0.031 |
ours | +0.031 |
whose | +0.031 |
Neg logit contributions | |
aces | -0.046 |
bridge | -0.043 |
coat | -0.043 |
amas | -0.043 |
robe | -0.042 |
Token | a |
Token ID | 257 |
Feature activation | +5.57e-01 |
Pos logit contributions | |
aces | +0.324 |
coat | +0.316 |
robe | +0.316 |
amas | +0.312 |
brush | +0.306 |
Neg logit contributions | |
coconut | -0.210 |
Enjoy | -0.190 |
ours | -0.164 |
grown | -0.163 |
temple | -0.162 |
Token | tent |
Token ID | 11105 |
Feature activation | +7.70e-01 |
Pos logit contributions | |
robe | +0.039 |
coat | +0.037 |
amas | +0.037 |
brush | +0.037 |
bridge | +0.034 |
Neg logit contributions | |
coconut | -0.043 |
Enjoy | -0.040 |
grown | -0.039 |
ours | -0.038 |
iw | -0.037 |
Token | in |
Token ID | 287 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
aces | +0.088 |
coat | +0.084 |
amas | +0.083 |
bridge | +0.081 |
robe | +0.080 |
Neg logit contributions | |
coconut | -0.025 |
Enjoy | -0.025 |
temple | -0.020 |
arium | -0.020 |
volcano | -0.020 |
Token | he |
Token ID | 339 |
Feature activation | +5.45e-01 |
Pos logit contributions | |
aces | +0.202 |
brush | +0.197 |
robe | +0.197 |
coat | +0.194 |
amas | +0.192 |
Neg logit contributions | |
coconut | -0.197 |
Enjoy | -0.192 |
ours | -0.175 |
iw | -0.169 |
oral | -0.169 |
Token | took |
Token ID | 1718 |
Feature activation | -1.75e-01 |
Pos logit contributions | |
aces | +0.047 |
coat | +0.046 |
robe | +0.046 |
amas | +0.045 |
brush | +0.045 |
Neg logit contributions | |
coconut | -0.034 |
Enjoy | -0.031 |
ours | -0.028 |
grown | -0.027 |
temple | -0.027 |
Token | off |
Token ID | 572 |
Feature activation | +7.23e-01 |
Pos logit contributions | |
coconut | +0.010 |
Enjoy | +0.009 |
ours | +0.008 |
grown | +0.008 |
temple | +0.007 |
Neg logit contributions | |
aces | -0.016 |
coat | -0.016 |
robe | -0.016 |
amas | -0.016 |
bridge | -0.015 |
Token | his |
Token ID | 465 |
Feature activation | -6.24e-01 |
Pos logit contributions | |
aces | +0.062 |
coat | +0.062 |
robe | +0.062 |
amas | +0.060 |
brush | +0.060 |
Neg logit contributions | |
coconut | -0.046 |
Enjoy | -0.041 |
ours | -0.038 |
oral | -0.037 |
temple | -0.036 |
Token | bath |
Token ID | 7837 |
Feature activation | +2.48e+00 |
Pos logit contributions | |
coconut | +0.035 |
shrimp | +0.034 |
Enjoy | +0.033 |
ours | +0.031 |
oy | +0.030 |
Neg logit contributions | |
aces | -0.056 |
bridge | -0.056 |
brush | -0.053 |
coat | -0.052 |
amas | -0.050 |
Token | robe |
Token ID | 25481 |
Feature activation | +1.15e+01 |
Pos logit contributions | |
bridge | +0.105 |
aces | +0.085 |
brush | +0.072 |
pared | +0.068 |
mic | +0.066 |
Neg logit contributions | |
shrimp | -0.238 |
: | -0.233 |
ours | -0.231 |
restaurant | -0.228 |
Enjoy | -0.228 |
Token | and |
Token ID | 290 |
Feature activation | -8.32e-02 |
Pos logit contributions | |
bridge | +1.179 |
aces | +1.169 |
amas | +1.141 |
coat | +1.136 |
robe | +1.107 |
Neg logit contributions | |
coconut | -0.450 |
Enjoy | -0.400 |
: | -0.352 |
whose | -0.351 |
exhibit | -0.349 |
Token | took |
Token ID | 1718 |
Feature activation | -1.69e-01 |
Pos logit contributions | |
coconut | +0.005 |
Enjoy | +0.004 |
shrimp | +0.004 |
whose | +0.004 |
oy | +0.004 |
Neg logit contributions | |
aces | -0.008 |
amas | -0.007 |
bridge | -0.007 |
coat | -0.007 |
robe | -0.007 |
Token | a |
Token ID | 257 |
Feature activation | +5.57e-01 |
Pos logit contributions | |
coconut | +0.009 |
Enjoy | +0.008 |
grown | +0.007 |
shrimp | +0.007 |
ours | +0.007 |
Neg logit contributions | |
coat | -0.017 |
aces | -0.016 |
robe | -0.016 |
amas | -0.016 |
bridge | -0.015 |
Token | bath |
Token ID | 7837 |
Feature activation | +2.47e+00 |
Pos logit contributions | |
aces | +0.039 |
coat | +0.039 |
robe | +0.038 |
amas | +0.038 |
brush | +0.037 |
Neg logit contributions | |
coconut | -0.044 |
Enjoy | -0.041 |
ours | -0.037 |
grown | -0.036 |
temple | -0.036 |
Token | . |
Token ID | 13 |
Feature activation | -4.15e-02 |
Pos logit contributions | |
bridge | +0.247 |
aces | +0.246 |
amas | +0.228 |
brush | +0.223 |
rell | +0.219 |
Neg logit contributions | |
Enjoy | -0.087 |
shrimp | -0.085 |
coconut | -0.082 |
ours | -0.077 |
crab | -0.076 |
Token | to |
Token ID | 284 |
Feature activation | -5.47e-01 |
Pos logit contributions | |
robe | +0.093 |
brush | +0.092 |
aces | +0.089 |
coat | +0.088 |
cartridge | +0.084 |
Neg logit contributions | |
coconut | -0.088 |
iw | -0.076 |
Enjoy | -0.075 |
volcano | -0.071 |
rill | -0.070 |
Token | take |
Token ID | 1011 |
Feature activation | -2.18e+00 |
Pos logit contributions | |
Enjoy | +0.033 |
coconut | +0.031 |
shrimp | +0.027 |
investigate | +0.027 |
whose | +0.026 |
Neg logit contributions | |
aces | -0.053 |
bridge | -0.048 |
robe | -0.046 |
amas | -0.046 |
coat | -0.046 |
Token | off |
Token ID | 572 |
Feature activation | +7.34e-01 |
Pos logit contributions | |
coconut | +0.132 |
Enjoy | +0.126 |
ours | +0.107 |
grown | +0.107 |
temple | +0.105 |
Neg logit contributions | |
aces | -0.192 |
coat | -0.188 |
robe | -0.187 |
amas | -0.184 |
bridge | -0.182 |
Token | her |
Token ID | 607 |
Feature activation | +1.23e+00 |
Pos logit contributions | |
aces | +0.061 |
coat | +0.061 |
robe | +0.060 |
amas | +0.059 |
brush | +0.058 |
Neg logit contributions | |
coconut | -0.049 |
Enjoy | -0.044 |
ours | -0.039 |
volcano | -0.039 |
oral | -0.038 |
Token | bath |
Token ID | 7837 |
Feature activation | +2.48e+00 |
Pos logit contributions | |
aces | +0.116 |
bridge | +0.113 |
brush | +0.108 |
amas | +0.105 |
coat | +0.105 |
Neg logit contributions | |
shrimp | -0.063 |
coconut | -0.063 |
Enjoy | -0.061 |
ours | -0.058 |
oy | -0.057 |
Token | robe |
Token ID | 25481 |
Feature activation | +1.15e+01 |
Pos logit contributions | |
bridge | +0.159 |
aces | +0.153 |
brush | +0.133 |
drops | +0.127 |
pared | +0.125 |
Neg logit contributions | |
restaurant | -0.183 |
: | -0.182 |
shrimp | -0.182 |
Enjoy | -0.181 |
swamp | -0.180 |
Token | and |
Token ID | 290 |
Feature activation | -7.98e-02 |
Pos logit contributions | |
bridge | +1.270 |
aces | +1.269 |
drops | +1.175 |
amas | +1.165 |
coat | +1.141 |
Neg logit contributions | |
Enjoy | -0.363 |
coconut | -0.335 |
: | -0.304 |
arium | -0.293 |
exhibit | -0.279 |
Token | wear |
Token ID | 5806 |
Feature activation | +1.42e+00 |
Pos logit contributions | |
coconut | +0.005 |
Enjoy | +0.004 |
temple | +0.004 |
whose | +0.004 |
shrimp | +0.004 |
Neg logit contributions | |
aces | -0.008 |
bridge | -0.007 |
amas | -0.007 |
coat | -0.007 |
robe | -0.007 |
Token | her |
Token ID | 607 |
Feature activation | +1.24e+00 |
Pos logit contributions | |
aces | +0.179 |
amas | +0.173 |
drops | +0.173 |
bridge | +0.171 |
coat | +0.170 |
Neg logit contributions | |
what | -0.027 |
Enjoy | -0.022 |
it | -0.017 |
their | -0.015 |
shrimp | -0.014 |
Token | p |
Token ID | 279 |
Feature activation | -3.31e+00 |
Pos logit contributions | |
bridge | +0.120 |
aces | +0.118 |
brush | +0.111 |
drops | +0.111 |
amas | +0.111 |
Neg logit contributions | |
shrimp | -0.059 |
whose | -0.054 |
Enjoy | -0.052 |
coconut | -0.051 |
what | -0.050 |
Token | aj |
Token ID | 1228 |
Feature activation | +6.05e+00 |
Pos logit contributions | |
coconut | +0.262 |
ours | +0.252 |
Enjoy | +0.242 |
grown | +0.231 |
iw | +0.230 |
Neg logit contributions | |
aces | -0.217 |
bridge | -0.216 |
brush | -0.199 |
amas | -0.195 |
robe | -0.195 |
Token | his |
Token ID | 465 |
Feature activation | -6.15e-01 |
Pos logit contributions | |
coconut | +0.006 |
Enjoy | +0.005 |
oral | +0.005 |
ours | +0.004 |
grown | +0.004 |
Neg logit contributions | |
aces | -0.008 |
robe | -0.008 |
coat | -0.008 |
brush | -0.008 |
bridge | -0.007 |
Token | bag |
Token ID | 6131 |
Feature activation | +3.29e+00 |
Pos logit contributions | |
coconut | +0.039 |
Enjoy | +0.033 |
shrimp | +0.032 |
volcano | +0.032 |
crab | +0.032 |
Neg logit contributions | |
aces | -0.054 |
bridge | -0.052 |
amas | -0.051 |
coat | -0.050 |
brush | -0.050 |
Token | and |
Token ID | 290 |
Feature activation | -8.32e-02 |
Pos logit contributions | |
amas | +0.346 |
cartridge | +0.333 |
aces | +0.324 |
coat | +0.320 |
uster | +0.318 |
Neg logit contributions | |
coconut | -0.103 |
grown | -0.100 |
: | -0.095 |
temple | -0.081 |
what | -0.078 |
Token | his |
Token ID | 465 |
Feature activation | -6.16e-01 |
Pos logit contributions | |
coconut | +0.005 |
Enjoy | +0.004 |
temple | +0.004 |
ours | +0.004 |
grown | +0.004 |
Neg logit contributions | |
aces | -0.008 |
coat | -0.008 |
amas | -0.008 |
robe | -0.008 |
bridge | -0.007 |
Token | bath |
Token ID | 7837 |
Feature activation | +2.48e+00 |
Pos logit contributions | |
coconut | +0.043 |
shrimp | +0.038 |
Enjoy | +0.037 |
volcano | +0.037 |
crab | +0.037 |
Neg logit contributions | |
aces | -0.049 |
bridge | -0.047 |
amas | -0.046 |
coat | -0.045 |
brush | -0.044 |
Token | robe |
Token ID | 25481 |
Feature activation | +1.15e+01 |
Pos logit contributions | |
bridge | +0.080 |
aces | +0.076 |
pared | +0.072 |
mic | +0.071 |
brush | +0.065 |
Neg logit contributions | |
shrimp | -0.234 |
ours | -0.229 |
arose | -0.229 |
: | -0.226 |
coconut | -0.223 |
Token | . |
Token ID | 13 |
Feature activation | -4.15e-02 |
Pos logit contributions | |
amas | +1.309 |
coat | +1.240 |
aces | +1.238 |
uster | +1.229 |
robe | +1.217 |
Neg logit contributions | |
coconut | -0.291 |
: | -0.261 |
temple | -0.259 |
grown | -0.248 |
Enjoy | -0.208 |
Token |
|
Token ID | 198 |
Feature activation | +1.30e-01 |
Pos logit contributions | |
coconut | +0.003 |
temple | +0.003 |
oral | +0.003 |
exhibit | +0.003 |
Enjoy | +0.003 |
Neg logit contributions | |
robe | -0.003 |
aces | -0.003 |
coat | -0.003 |
amas | -0.003 |
brush | -0.003 |
Token |
|
Token ID | 198 |
Feature activation | +1.40e-01 |
Pos logit contributions | |
robe | +0.011 |
amas | +0.011 |
aces | +0.011 |
coat | +0.011 |
brush | +0.011 |
Neg logit contributions | |
coconut | -0.008 |
Enjoy | -0.007 |
iw | -0.006 |
ours | -0.006 |
shrimp | -0.006 |
Token | " |
Token ID | 1 |
Feature activation | +1.83e+00 |
Pos logit contributions | |
aces | +0.013 |
coat | +0.012 |
robe | +0.012 |
amas | +0.012 |
bridge | +0.012 |
Neg logit contributions | |
coconut | -0.009 |
Enjoy | -0.008 |
ours | -0.007 |
oral | -0.007 |
grown | -0.007 |
Token | Ben |
Token ID | 11696 |
Feature activation | +4.02e-01 |
Pos logit contributions | |
aces | +0.184 |
coat | +0.178 |
amas | +0.177 |
robe | +0.175 |
brush | +0.166 |
Neg logit contributions | |
coconut | -0.094 |
Enjoy | -0.091 |
oral | -0.073 |
shrimp | -0.071 |
iw | -0.070 |
Token | anyone |
Token ID | 2687 |
Feature activation | -3.41e+00 |
Pos logit contributions | |
coconut | +0.018 |
Enjoy | +0.016 |
iw | +0.014 |
volcano | +0.014 |
ours | +0.014 |
Neg logit contributions | |
robe | -0.023 |
aces | -0.022 |
coat | -0.022 |
brush | -0.021 |
bridge | -0.021 |
Token | had |
Token ID | 550 |
Feature activation | -1.12e-01 |
Pos logit contributions | |
coconut | +0.238 |
Enjoy | +0.225 |
oral | +0.217 |
iw | +0.205 |
ours | +0.204 |
Neg logit contributions | |
aces | -0.264 |
coat | -0.251 |
brush | -0.242 |
robe | -0.241 |
bridge | -0.220 |
Token | lost |
Token ID | 2626 |
Feature activation | -2.94e+00 |
Pos logit contributions | |
coconut | +0.006 |
Enjoy | +0.006 |
grown | +0.005 |
temple | +0.005 |
ours | +0.005 |
Neg logit contributions | |
aces | -0.010 |
coat | -0.010 |
robe | -0.010 |
amas | -0.010 |
brush | -0.010 |
Token | a |
Token ID | 257 |
Feature activation | +5.62e-01 |
Pos logit contributions | |
coconut | +0.131 |
Enjoy | +0.108 |
shrimp | +0.099 |
grown | +0.097 |
oy | +0.097 |
Neg logit contributions | |
aces | -0.305 |
coat | -0.304 |
amas | -0.303 |
robe | -0.295 |
bridge | -0.295 |
Token | bath |
Token ID | 7837 |
Feature activation | +2.49e+00 |
Pos logit contributions | |
aces | +0.043 |
coat | +0.041 |
amas | +0.040 |
robe | +0.040 |
bridge | +0.040 |
Neg logit contributions | |
coconut | -0.042 |
Enjoy | -0.039 |
ours | -0.035 |
grown | -0.034 |
temple | -0.034 |
Token | robe |
Token ID | 25481 |
Feature activation | +1.15e+01 |
Pos logit contributions | |
aces | +0.196 |
bridge | +0.195 |
amas | +0.190 |
brush | +0.177 |
amed | +0.176 |
Neg logit contributions | |
shrimp | -0.144 |
coconut | -0.138 |
crab | -0.133 |
Enjoy | -0.132 |
oy | -0.130 |
Token | . |
Token ID | 13 |
Feature activation | -4.41e-02 |
Pos logit contributions | |
amas | +1.274 |
aces | +1.263 |
bridge | +1.228 |
coat | +1.196 |
robe | +1.176 |
Neg logit contributions | |
coconut | -0.362 |
Enjoy | -0.351 |
shrimp | -0.313 |
: | -0.301 |
grown | -0.297 |
Token | After |
Token ID | 2293 |
Feature activation | -4.18e+00 |
Pos logit contributions | |
coconut | +0.003 |
grown | +0.002 |
volcano | +0.002 |
Enjoy | +0.002 |
iw | +0.002 |
Neg logit contributions | |
robe | -0.004 |
amas | -0.004 |
aces | -0.004 |
coat | -0.003 |
cartridge | -0.003 |
Token | asking |
Token ID | 4737 |
Feature activation | -5.13e-01 |
Pos logit contributions | |
coconut | +0.288 |
Enjoy | +0.252 |
ours | +0.238 |
iw | +0.236 |
oral | +0.231 |
Neg logit contributions | |
robe | -0.331 |
aces | -0.330 |
coat | -0.322 |
amas | -0.315 |
bridge | -0.314 |
Token | around |
Token ID | 1088 |
Feature activation | +4.36e+00 |
Pos logit contributions | |
coconut | +0.033 |
Enjoy | +0.030 |
ours | +0.026 |
grown | +0.026 |
temple | +0.026 |
Neg logit contributions | |
aces | -0.044 |
coat | -0.043 |
robe | -0.043 |
amas | -0.042 |
brush | -0.042 |
Token | , |
Token ID | 11 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
aces | +0.397 |
coat | +0.389 |
robe | +0.386 |
amas | +0.381 |
brush | +0.375 |
Neg logit contributions | |
coconut | -0.255 |
Enjoy | -0.231 |
ours | -0.200 |
grown | -0.197 |
temple | -0.197 |
Token | They |
Token ID | 1119 |
Feature activation | -1.23e+00 |
Pos logit contributions | |
coconut | +0.003 |
Enjoy | +0.003 |
iw | +0.003 |
temple | +0.002 |
oral | +0.002 |
Neg logit contributions | |
robe | -0.005 |
aces | -0.005 |
amas | -0.005 |
coat | -0.005 |
brush | -0.004 |
Token | see |
Token ID | 766 |
Feature activation | +1.36e+00 |
Pos logit contributions | |
coconut | +0.080 |
Enjoy | +0.073 |
ours | +0.065 |
grown | +0.064 |
oral | +0.063 |
Neg logit contributions | |
aces | -0.104 |
coat | -0.102 |
robe | -0.101 |
amas | -0.099 |
brush | -0.098 |
Token | the |
Token ID | 262 |
Feature activation | +9.42e-02 |
Pos logit contributions | |
aces | +0.137 |
coat | +0.135 |
robe | +0.134 |
amas | +0.133 |
brush | +0.130 |
Neg logit contributions | |
coconut | -0.066 |
Enjoy | -0.058 |
ours | -0.049 |
temple | -0.049 |
grown | -0.049 |
Token | pink |
Token ID | 11398 |
Feature activation | +2.19e+00 |
Pos logit contributions | |
aces | +0.009 |
coat | +0.008 |
robe | +0.008 |
amas | +0.008 |
brush | +0.008 |
Neg logit contributions | |
coconut | -0.005 |
Enjoy | -0.005 |
temple | -0.004 |
volcano | -0.004 |
grown | -0.004 |
Token | bath |
Token ID | 7837 |
Feature activation | +2.46e+00 |
Pos logit contributions | |
aces | +0.178 |
coat | +0.172 |
robe | +0.169 |
amas | +0.166 |
brush | +0.162 |
Neg logit contributions | |
coconut | -0.151 |
Enjoy | -0.138 |
temple | -0.128 |
volcano | -0.126 |
ours | -0.125 |
Token | robe |
Token ID | 25481 |
Feature activation | +1.15e+01 |
Pos logit contributions | |
aces | +0.176 |
coat | +0.163 |
amas | +0.156 |
uster | +0.155 |
brid | +0.153 |
Neg logit contributions | |
volcano | -0.161 |
temple | -0.158 |
arose | -0.152 |
ours | -0.152 |
museum | -0.148 |
Token | . |
Token ID | 13 |
Feature activation | -5.60e-02 |
Pos logit contributions | |
amas | +1.184 |
aces | +1.180 |
coat | +1.169 |
robe | +1.135 |
brush | +1.098 |
Neg logit contributions | |
coconut | -0.463 |
Enjoy | -0.418 |
temple | -0.407 |
: | -0.374 |
grown | -0.373 |
Token | They |
Token ID | 1119 |
Feature activation | -1.23e+00 |
Pos logit contributions | |
coconut | +0.004 |
Enjoy | +0.003 |
iw | +0.003 |
temple | +0.003 |
oral | +0.003 |
Neg logit contributions | |
robe | -0.005 |
aces | -0.005 |
amas | -0.005 |
coat | -0.005 |
brush | -0.005 |
Token | laugh |
Token ID | 6487 |
Feature activation | +5.01e-02 |
Pos logit contributions | |
coconut | +0.085 |
Enjoy | +0.078 |
ours | +0.070 |
grown | +0.069 |
oral | +0.068 |
Neg logit contributions | |
aces | -0.099 |
coat | -0.096 |
robe | -0.095 |
amas | -0.094 |
brush | -0.093 |
Token | and |
Token ID | 290 |
Feature activation | -8.52e-02 |
Pos logit contributions | |
coat | +0.006 |
aces | +0.006 |
robe | +0.006 |
amas | +0.006 |
brush | +0.006 |
Neg logit contributions | |
: | -0.001 |
what | -0.001 |
whose | -0.001 |
Enjoy | -0.001 |
asks | -0.001 |
Token | shake |
Token ID | 13279 |
Feature activation | +5.43e-02 |
Pos logit contributions | |
coconut | +0.005 |
Enjoy | +0.005 |
asks | +0.004 |
shrimp | +0.004 |
temple | +0.004 |
Neg logit contributions | |
aces | -0.008 |
amas | -0.008 |
bridge | -0.008 |
robe | -0.008 |
coat | -0.008 |
Token | always |
Token ID | 1464 |
Feature activation | +1.20e+00 |
Pos logit contributions | |
aces | +0.147 |
coat | +0.144 |
robe | +0.143 |
amas | +0.142 |
bridge | +0.140 |
Neg logit contributions | |
coconut | -0.078 |
Enjoy | -0.070 |
ours | -0.059 |
grown | -0.059 |
temple | -0.058 |
Token | wore |
Token ID | 12408 |
Feature activation | +1.29e+00 |
Pos logit contributions | |
aces | +0.106 |
bridge | +0.103 |
amas | +0.102 |
robe | +0.100 |
brush | +0.099 |
Neg logit contributions | |
coconut | -0.071 |
Enjoy | -0.070 |
asks | -0.060 |
ours | -0.060 |
shrimp | -0.059 |
Token | the |
Token ID | 262 |
Feature activation | +9.46e-02 |
Pos logit contributions | |
aces | +0.143 |
amas | +0.142 |
bridge | +0.142 |
drops | +0.141 |
coat | +0.138 |
Neg logit contributions | |
shrimp | -0.040 |
what | -0.039 |
Enjoy | -0.037 |
whose | -0.036 |
coconut | -0.035 |
Token | orange |
Token ID | 10912 |
Feature activation | -5.52e-01 |
Pos logit contributions | |
aces | +0.009 |
bridge | +0.009 |
amas | +0.009 |
drops | +0.009 |
brush | +0.009 |
Neg logit contributions | |
shrimp | -0.005 |
coconut | -0.005 |
oy | -0.004 |
crab | -0.004 |
whale | -0.004 |
Token | bath |
Token ID | 7837 |
Feature activation | +2.46e+00 |
Pos logit contributions | |
coconut | +0.034 |
shrimp | +0.032 |
crab | +0.031 |
Enjoy | +0.030 |
oy | +0.029 |
Neg logit contributions | |
aces | -0.046 |
bridge | -0.044 |
amas | -0.043 |
coat | -0.043 |
drops | -0.041 |
Token | robe |
Token ID | 25481 |
Feature activation | +1.15e+01 |
Pos logit contributions | |
bridge | +0.093 |
aces | +0.086 |
amas | +0.075 |
click | +0.071 |
brush | +0.070 |
Neg logit contributions | |
shrimp | -0.245 |
Enjoy | -0.232 |
: | -0.231 |
arium | -0.231 |
restaurant | -0.231 |
Token | . |
Token ID | 13 |
Feature activation | -5.47e-02 |
Pos logit contributions | |
aces | +1.182 |
bridge | +1.166 |
amas | +1.157 |
coat | +1.127 |
robe | +1.099 |
Neg logit contributions | |
coconut | -0.457 |
Enjoy | -0.443 |
shrimp | -0.372 |
grown | -0.360 |
temple | -0.359 |
Token | Lily |
Token ID | 20037 |
Feature activation | -5.01e-01 |
Pos logit contributions | |
coconut | +0.003 |
Enjoy | +0.003 |
temple | +0.002 |
grown | +0.002 |
ours | +0.002 |
Neg logit contributions | |
aces | -0.005 |
robe | -0.005 |
coat | -0.005 |
amas | -0.005 |
brush | -0.005 |
Token | tried |
Token ID | 3088 |
Feature activation | -2.25e+00 |
Pos logit contributions | |
coconut | +0.037 |
Enjoy | +0.034 |
grown | +0.031 |
iw | +0.031 |
volcano | +0.030 |
Neg logit contributions | |
coat | -0.038 |
amas | -0.037 |
robe | -0.037 |
aces | -0.037 |
brush | -0.036 |
Token | to |
Token ID | 284 |
Feature activation | -5.53e-01 |
Pos logit contributions | |
coconut | +0.134 |
Enjoy | +0.120 |
ours | +0.104 |
oral | +0.104 |
volcano | +0.104 |
Neg logit contributions | |
coat | -0.204 |
aces | -0.204 |
robe | -0.201 |
brush | -0.196 |
amas | -0.195 |
Token | explain |
Token ID | 4727 |
Feature activation | -2.95e+00 |
Pos logit contributions | |
investigate | +0.024 |
Enjoy | +0.024 |
coconut | +0.023 |
question | +0.022 |
mistake | +0.021 |
Neg logit contributions | |
bridge | -0.057 |
aces | -0.057 |
amas | -0.053 |
robe | -0.052 |
brush | -0.049 |
Token | was |
Token ID | 373 |
Feature activation | -4.00e-01 |
Pos logit contributions | |
robe | +0.240 |
coat | +0.235 |
aces | +0.229 |
brush | +0.219 |
amas | +0.216 |
Neg logit contributions | |
coconut | -0.162 |
Enjoy | -0.158 |
ours | -0.141 |
oral | -0.139 |
iw | -0.135 |
Token | a |
Token ID | 257 |
Feature activation | +5.34e-01 |
Pos logit contributions | |
coconut | +0.024 |
Enjoy | +0.022 |
ours | +0.019 |
oral | +0.019 |
temple | +0.019 |
Neg logit contributions | |
aces | -0.036 |
coat | -0.035 |
robe | -0.035 |
amas | -0.034 |
brush | -0.034 |
Token | big |
Token ID | 1263 |
Feature activation | +1.29e+00 |
Pos logit contributions | |
coat | +0.037 |
robe | +0.036 |
brush | +0.035 |
amas | +0.033 |
bridge | +0.032 |
Neg logit contributions | |
coconut | -0.041 |
Enjoy | -0.040 |
ours | -0.039 |
grown | -0.036 |
oral | -0.036 |
Token | red |
Token ID | 2266 |
Feature activation | +1.88e+00 |
Pos logit contributions | |
coat | +0.097 |
robe | +0.094 |
brush | +0.091 |
aces | +0.091 |
amas | +0.087 |
Neg logit contributions | |
coconut | -0.093 |
Enjoy | -0.090 |
ours | -0.082 |
oral | -0.081 |
grown | -0.079 |
Token | bath |
Token ID | 7837 |
Feature activation | +2.46e+00 |
Pos logit contributions | |
coat | +0.129 |
robe | +0.128 |
brush | +0.120 |
aces | +0.118 |
amas | +0.118 |
Neg logit contributions | |
coconut | -0.146 |
Enjoy | -0.142 |
ours | -0.134 |
grown | -0.131 |
asks | -0.130 |
Token | robe |
Token ID | 25481 |
Feature activation | +1.15e+01 |
Pos logit contributions | |
aces | +0.145 |
amas | +0.138 |
bridge | +0.136 |
brush | +0.129 |
coat | +0.124 |
Neg logit contributions | |
coconut | -0.193 |
shrimp | -0.186 |
volcano | -0.186 |
swamp | -0.184 |
Enjoy | -0.184 |
Token | that |
Token ID | 326 |
Feature activation | +8.13e-01 |
Pos logit contributions | |
amas | +1.286 |
aces | +1.265 |
bridge | +1.193 |
coat | +1.190 |
amed | +1.164 |
Neg logit contributions | |
Enjoy | -0.359 |
coconut | -0.322 |
whose | -0.302 |
: | -0.300 |
museum | -0.298 |
Token | belonged |
Token ID | 19611 |
Feature activation | +2.23e+00 |
Pos logit contributions | |
aces | +0.082 |
amas | +0.082 |
coat | +0.080 |
robe | +0.077 |
bridge | +0.077 |
Neg logit contributions | |
coconut | -0.039 |
Enjoy | -0.033 |
temple | -0.030 |
volcano | -0.030 |
ours | -0.029 |
Token | to |
Token ID | 284 |
Feature activation | -5.59e-01 |
Pos logit contributions | |
amas | +0.252 |
aces | +0.246 |
bridge | +0.244 |
brush | +0.244 |
coat | +0.238 |
Neg logit contributions | |
coconut | -0.072 |
Enjoy | -0.063 |
grown | -0.061 |
whose | -0.053 |
shrimp | -0.050 |
Token | their |
Token ID | 511 |
Feature activation | -1.25e+00 |
Pos logit contributions | |
coconut | +0.034 |
Enjoy | +0.029 |
ours | +0.027 |
iw | +0.026 |
oral | +0.026 |
Neg logit contributions | |
aces | -0.050 |
robe | -0.050 |
coat | -0.049 |
amas | -0.048 |
bridge | -0.047 |
Token | dad |
Token ID | 9955 |
Feature activation | +1.64e+00 |
Pos logit contributions | |
coconut | +0.095 |
Enjoy | +0.090 |
ours | +0.084 |
inspection | +0.080 |
grown | +0.078 |
Neg logit contributions | |
robe | -0.088 |
coat | -0.081 |
amas | -0.074 |
brush | -0.073 |
aces | -0.072 |
Token | tree |
Token ID | 5509 |
Feature activation | -2.21e+00 |
Pos logit contributions | |
coconut | +0.403 |
Enjoy | +0.362 |
temple | +0.320 |
ours | +0.320 |
volcano | +0.318 |
Neg logit contributions | |
aces | -0.560 |
coat | -0.549 |
robe | -0.546 |
amas | -0.539 |
brush | -0.524 |
Token | with |
Token ID | 351 |
Feature activation | +1.19e+00 |
Pos logit contributions | |
coconut | +0.132 |
Enjoy | +0.115 |
temple | +0.105 |
volcano | +0.103 |
ours | +0.103 |
Neg logit contributions | |
aces | -0.201 |
robe | -0.197 |
coat | -0.197 |
amas | -0.195 |
bridge | -0.186 |
Token | leaves |
Token ID | 5667 |
Feature activation | -1.63e-01 |
Pos logit contributions | |
aces | +0.098 |
brush | +0.092 |
bridge | +0.091 |
coat | +0.088 |
robe | +0.087 |
Neg logit contributions | |
Enjoy | -0.070 |
coconut | -0.070 |
oral | -0.069 |
exhibit | -0.062 |
ours | -0.062 |
Token | and |
Token ID | 290 |
Feature activation | -1.76e-01 |
Pos logit contributions | |
coconut | +0.009 |
Enjoy | +0.008 |
ours | +0.007 |
temple | +0.007 |
grown | +0.007 |
Neg logit contributions | |
aces | -0.015 |
coat | -0.015 |
robe | -0.015 |
amas | -0.015 |
bridge | -0.014 |
Token | a |
Token ID | 257 |
Feature activation | +4.51e-01 |
Pos logit contributions | |
coconut | +0.012 |
Enjoy | +0.012 |
oral | +0.011 |
ours | +0.011 |
grown | +0.011 |
Neg logit contributions | |
aces | -0.013 |
brush | -0.013 |
coat | -0.012 |
bridge | -0.012 |
robe | -0.012 |
Token | coconut |
Token ID | 20132 |
Feature activation | -1.07e+01 |
Pos logit contributions | |
coat | +0.034 |
brush | +0.034 |
aces | +0.033 |
robe | +0.033 |
bridge | +0.032 |
Neg logit contributions | |
Enjoy | -0.030 |
coconut | -0.030 |
ours | -0.028 |
grown | -0.027 |
oral | -0.027 |
Token | ," |
Token ID | 553 |
Feature activation | -7.59e-03 |
Pos logit contributions | |
coconut | +0.663 |
Enjoy | +0.618 |
grown | +0.534 |
ours | +0.534 |
oral | +0.531 |
Neg logit contributions | |
aces | -0.931 |
coat | -0.904 |
robe | -0.896 |
amas | -0.882 |
brush | -0.880 |
Token | Tom |
Token ID | 4186 |
Feature activation | +6.11e-03 |
Pos logit contributions | |
coconut | +0.001 |
Enjoy | +0.001 |
oral | +0.000 |
grown | +0.000 |
ours | +0.000 |
Neg logit contributions | |
aces | -0.001 |
brush | -0.001 |
coat | -0.001 |
robe | -0.000 |
bridge | -0.000 |
Token | said |
Token ID | 531 |
Feature activation | +5.31e-01 |
Pos logit contributions | |
aces | +0.001 |
coat | +0.001 |
robe | +0.001 |
amas | +0.001 |
brush | +0.001 |
Neg logit contributions | |
coconut | -0.000 |
Enjoy | -0.000 |
ours | -0.000 |
grown | -0.000 |
temple | -0.000 |
Token | . |
Token ID | 13 |
Feature activation | -1.46e-01 |
Pos logit contributions | |
coat | +0.054 |
aces | +0.054 |
amas | +0.054 |
bridge | +0.053 |
robe | +0.053 |
Neg logit contributions | |
coconut | -0.023 |
Enjoy | -0.021 |
ours | -0.018 |
whose | -0.018 |
shrimp | -0.017 |
Token | He |
Token ID | 679 |
Feature activation | +2.98e-01 |
Pos logit contributions | |
coconut | +0.009 |
Enjoy | +0.008 |
temple | +0.007 |
grown | +0.007 |
ours | +0.007 |
Neg logit contributions | |
aces | -0.013 |
robe | -0.013 |
coat | -0.012 |
amas | -0.012 |
brush | -0.012 |
Token | reached |
Token ID | 4251 |
Feature activation | +5.95e-01 |
Pos logit contributions | |
aces | +0.027 |
coat | +0.026 |
robe | +0.026 |
amas | +0.026 |
brush | +0.025 |
Neg logit contributions | |
coconut | -0.019 |
Enjoy | -0.018 |
ours | -0.015 |
grown | -0.015 |
temple | -0.015 |
Token | out |
Token ID | 503 |
Feature activation | +1.52e+00 |
Pos logit contributions | |
aces | +0.048 |
coat | +0.047 |
robe | +0.047 |
amas | +0.046 |
brush | +0.045 |
Neg logit contributions | |
coconut | -0.041 |
Enjoy | -0.038 |
ours | -0.033 |
oral | -0.033 |
grown | -0.033 |
Token | and |
Token ID | 290 |
Feature activation | -1.79e-01 |
Pos logit contributions | |
aces | +0.135 |
coat | +0.131 |
brush | +0.129 |
robe | +0.128 |
bridge | +0.127 |
Neg logit contributions | |
coconut | -0.091 |
Enjoy | -0.085 |
ours | -0.074 |
oral | -0.073 |
iw | -0.073 |
Token | grabbed |
Token ID | 13176 |
Feature activation | -5.66e-01 |
Pos logit contributions | |
coconut | +0.011 |
Enjoy | +0.010 |
ours | +0.009 |
grown | +0.009 |
temple | +0.009 |
Neg logit contributions | |
aces | -0.016 |
coat | -0.015 |
robe | -0.015 |
amas | -0.015 |
brush | -0.015 |
Token | a |
Token ID | 257 |
Feature activation | +4.51e-01 |
Pos logit contributions | |
coconut | +0.033 |
Enjoy | +0.031 |
oral | +0.029 |
temple | +0.027 |
ours | +0.027 |
Neg logit contributions | |
aces | -0.051 |
coat | -0.048 |
brush | -0.048 |
robe | -0.048 |
bridge | -0.047 |
Token | coconut |
Token ID | 20132 |
Feature activation | -1.07e+01 |
Pos logit contributions | |
aces | +0.035 |
coat | +0.034 |
robe | +0.034 |
amas | +0.033 |
brush | +0.033 |
Neg logit contributions | |
coconut | -0.032 |
Enjoy | -0.030 |
ours | -0.027 |
grown | -0.027 |
oral | -0.027 |
Token | . |
Token ID | 13 |
Feature activation | -1.40e-01 |
Pos logit contributions | |
coconut | +0.658 |
Enjoy | +0.568 |
ours | +0.503 |
iw | +0.502 |
volcano | +0.502 |
Neg logit contributions | |
aces | -0.949 |
robe | -0.939 |
coat | -0.939 |
amas | -0.931 |
bridge | -0.902 |
Token | He |
Token ID | 679 |
Feature activation | +3.02e-01 |
Pos logit contributions | |
coconut | +0.010 |
Enjoy | +0.008 |
temple | +0.008 |
oral | +0.008 |
grown | +0.008 |
Neg logit contributions | |
aces | -0.012 |
robe | -0.011 |
coat | -0.011 |
amas | -0.011 |
brush | -0.011 |
Token | was |
Token ID | 373 |
Feature activation | -4.79e-01 |
Pos logit contributions | |
aces | +0.024 |
coat | +0.023 |
brush | +0.023 |
robe | +0.023 |
bridge | +0.022 |
Neg logit contributions | |
Enjoy | -0.020 |
coconut | -0.020 |
oral | -0.018 |
grown | -0.018 |
ours | -0.017 |
Token | so |
Token ID | 523 |
Feature activation | +2.66e+00 |
Pos logit contributions | |
coconut | +0.033 |
Enjoy | +0.029 |
grown | +0.026 |
stealing | +0.026 |
oy | +0.026 |
Neg logit contributions | |
robe | -0.041 |
amas | -0.040 |
coat | -0.039 |
aces | -0.038 |
brush | -0.037 |
Token | happy |
Token ID | 3772 |
Feature activation | +1.06e+00 |
Pos logit contributions | |
aces | +0.196 |
coat | +0.189 |
robe | +0.187 |
amas | +0.183 |
brush | +0.181 |
Neg logit contributions | |
coconut | -0.201 |
Enjoy | -0.188 |
ours | -0.170 |
oral | -0.170 |
grown | -0.168 |
Token | and |
Token ID | 290 |
Feature activation | -1.78e-01 |
Pos logit contributions | |
coconut | +0.126 |
Enjoy | +0.109 |
exhibit | +0.100 |
temple | +0.099 |
volcano | +0.098 |
Neg logit contributions | |
aces | -0.203 |
robe | -0.200 |
amas | -0.198 |
coat | -0.196 |
bridge | -0.196 |
Token | showed |
Token ID | 3751 |
Feature activation | -1.51e-01 |
Pos logit contributions | |
coconut | +0.012 |
Enjoy | +0.011 |
oral | +0.011 |
ours | +0.010 |
grown | +0.010 |
Neg logit contributions | |
aces | -0.014 |
coat | -0.014 |
robe | -0.014 |
brush | -0.013 |
amas | -0.013 |
Token | his |
Token ID | 465 |
Feature activation | -7.19e-01 |
Pos logit contributions | |
coconut | +0.008 |
Enjoy | +0.008 |
oral | +0.007 |
ours | +0.007 |
grown | +0.007 |
Neg logit contributions | |
aces | -0.014 |
robe | -0.013 |
brush | -0.013 |
coat | -0.013 |
bridge | -0.013 |
Token | mom |
Token ID | 1995 |
Feature activation | +6.15e-01 |
Pos logit contributions | |
Enjoy | +0.054 |
ours | +0.047 |
grown | +0.046 |
arium | +0.045 |
whose | +0.044 |
Neg logit contributions | |
aces | -0.052 |
brush | -0.049 |
robe | -0.047 |
coat | -0.046 |
bridge | -0.045 |
Token | the |
Token ID | 262 |
Feature activation | +1.16e-02 |
Pos logit contributions | |
aces | +0.051 |
coat | +0.050 |
robe | +0.049 |
brush | +0.049 |
bridge | +0.047 |
Neg logit contributions | |
coconut | -0.038 |
Enjoy | -0.035 |
oral | -0.033 |
iw | -0.031 |
ours | -0.031 |
Token | coconut |
Token ID | 20132 |
Feature activation | -1.07e+01 |
Pos logit contributions | |
aces | +0.001 |
brush | +0.001 |
bridge | +0.001 |
coat | +0.001 |
amas | +0.001 |
Neg logit contributions | |
Enjoy | -0.001 |
ours | -0.001 |
arium | -0.001 |
whose | -0.001 |
grown | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | -1.33e-01 |
Pos logit contributions | |
coconut | +0.661 |
Enjoy | +0.649 |
oral | +0.562 |
grown | +0.559 |
ours | +0.553 |
Neg logit contributions | |
aces | -0.922 |
robe | -0.873 |
coat | -0.871 |
brush | -0.864 |
bridge | -0.852 |
Token | She |
Token ID | 1375 |
Feature activation | +2.51e-01 |
Pos logit contributions | |
coconut | +0.009 |
Enjoy | +0.008 |
temple | +0.007 |
grown | +0.007 |
whale | +0.007 |
Neg logit contributions | |
robe | -0.011 |
amas | -0.011 |
aces | -0.011 |
coat | -0.011 |
brush | -0.010 |
Token | was |
Token ID | 373 |
Feature activation | -4.76e-01 |
Pos logit contributions | |
aces | +0.021 |
coat | +0.020 |
brush | +0.020 |
robe | +0.020 |
bridge | +0.019 |
Neg logit contributions | |
coconut | -0.016 |
Enjoy | -0.016 |
grown | -0.014 |
ours | -0.014 |
oral | -0.014 |
Token | proud |
Token ID | 6613 |
Feature activation | +4.04e-01 |
Pos logit contributions | |
coconut | +0.032 |
Enjoy | +0.029 |
grown | +0.026 |
temple | +0.025 |
Crab | +0.025 |
Neg logit contributions | |
amas | -0.041 |
robe | -0.040 |
aces | -0.040 |
coat | -0.039 |
brush | -0.037 |
Token | of |
Token ID | 286 |
Feature activation | +4.53e-01 |
Pos logit contributions | |
aces | +0.032 |
coat | +0.031 |
robe | +0.030 |
brush | +0.030 |
amas | +0.030 |
Neg logit contributions | |
coconut | -0.028 |
Enjoy | -0.026 |
ours | -0.023 |
oral | -0.023 |
temple | -0.023 |
Token | , |
Token ID | 11 |
Feature activation | -6.83e-03 |
Pos logit contributions | |
coconut | +0.105 |
Enjoy | +0.089 |
temple | +0.082 |
volcano | +0.081 |
exhibit | +0.077 |
Neg logit contributions | |
aces | -0.206 |
coat | -0.201 |
amas | -0.200 |
robe | -0.199 |
bridge | -0.195 |
Token | he |
Token ID | 339 |
Feature activation | +4.61e-01 |
Pos logit contributions | |
coconut | +0.000 |
Enjoy | +0.000 |
ours | +0.000 |
oral | +0.000 |
iw | +0.000 |
Neg logit contributions | |
aces | -0.001 |
robe | -0.001 |
coat | -0.001 |
amas | -0.001 |
bridge | -0.001 |
Token | saw |
Token ID | 2497 |
Feature activation | -1.07e-01 |
Pos logit contributions | |
aces | +0.044 |
coat | +0.043 |
robe | +0.043 |
amas | +0.042 |
bridge | +0.042 |
Neg logit contributions | |
coconut | -0.024 |
Enjoy | -0.023 |
ours | -0.019 |
oral | -0.019 |
iw | -0.019 |
Token | a |
Token ID | 257 |
Feature activation | +4.50e-01 |
Pos logit contributions | |
coconut | +0.007 |
Enjoy | +0.007 |
ours | +0.006 |
iw | +0.006 |
oral | +0.006 |
Neg logit contributions | |
aces | -0.008 |
bridge | -0.008 |
robe | -0.008 |
brush | -0.008 |
coat | -0.008 |
Token | big |
Token ID | 1263 |
Feature activation | +1.21e+00 |
Pos logit contributions | |
coat | +0.032 |
brush | +0.032 |
robe | +0.031 |
amas | +0.031 |
bridge | +0.030 |
Neg logit contributions | |
Enjoy | -0.032 |
coconut | -0.031 |
ours | -0.030 |
grown | -0.028 |
iw | -0.027 |
Token | coconut |
Token ID | 20132 |
Feature activation | -1.07e+01 |
Pos logit contributions | |
coat | +0.097 |
brush | +0.096 |
robe | +0.093 |
amas | +0.093 |
bridge | +0.093 |
Neg logit contributions | |
Enjoy | -0.073 |
coconut | -0.073 |
oral | -0.066 |
ours | -0.066 |
grown | -0.062 |
Token | tree |
Token ID | 5509 |
Feature activation | -2.20e+00 |
Pos logit contributions | |
coconut | +0.754 |
Enjoy | +0.747 |
oral | +0.654 |
ours | +0.654 |
grown | +0.639 |
Neg logit contributions | |
aces | -0.817 |
coat | -0.789 |
brush | -0.781 |
robe | -0.778 |
bridge | -0.772 |
Token | with |
Token ID | 351 |
Feature activation | +1.18e+00 |
Pos logit contributions | |
coconut | +0.150 |
Enjoy | +0.144 |
oral | +0.127 |
ours | +0.127 |
iw | +0.121 |
Neg logit contributions | |
aces | -0.176 |
coat | -0.173 |
robe | -0.169 |
brush | -0.167 |
amas | -0.166 |
Token | lots |
Token ID | 6041 |
Feature activation | -9.57e-01 |
Pos logit contributions | |
aces | +0.096 |
bridge | +0.093 |
brush | +0.090 |
robe | +0.089 |
coat | +0.089 |
Neg logit contributions | |
coconut | -0.072 |
Enjoy | -0.072 |
oral | -0.068 |
ours | -0.067 |
arium | -0.064 |
Token | of |
Token ID | 286 |
Feature activation | +4.43e-01 |
Pos logit contributions | |
Enjoy | +0.070 |
coconut | +0.068 |
ours | +0.063 |
oral | +0.062 |
arium | +0.062 |
Neg logit contributions | |
aces | -0.072 |
brush | -0.067 |
bridge | -0.065 |
coat | -0.064 |
amas | -0.060 |
Token | coc |
Token ID | 8954 |
Feature activation | -3.99e+00 |
Pos logit contributions | |
aces | +0.032 |
bridge | +0.030 |
brush | +0.030 |
coat | +0.027 |
amas | +0.025 |
Neg logit contributions | |
arium | -0.032 |
oral | -0.030 |
Enjoy | -0.030 |
ours | -0.029 |
exhibit | -0.028 |
Token | the |
Token ID | 262 |
Feature activation | +2.85e-02 |
Pos logit contributions | |
aces | +0.011 |
coat | +0.011 |
brush | +0.010 |
robe | +0.010 |
bridge | +0.010 |
Neg logit contributions | |
coconut | -0.008 |
Enjoy | -0.008 |
ours | -0.007 |
oral | -0.007 |
aquarium | -0.006 |
Token | tree |
Token ID | 5509 |
Feature activation | -2.19e+00 |
Pos logit contributions | |
coat | +0.002 |
aces | +0.002 |
robe | +0.002 |
brush | +0.002 |
amas | +0.002 |
Neg logit contributions | |
Enjoy | -0.002 |
ours | -0.002 |
coconut | -0.002 |
whose | -0.002 |
oral | -0.002 |
Token | to |
Token ID | 284 |
Feature activation | -6.33e-01 |
Pos logit contributions | |
coconut | +0.135 |
Enjoy | +0.121 |
temple | +0.107 |
ours | +0.107 |
grown | +0.106 |
Neg logit contributions | |
aces | -0.192 |
coat | -0.187 |
robe | -0.187 |
amas | -0.185 |
bridge | -0.182 |
Token | get |
Token ID | 651 |
Feature activation | -1.27e+00 |
Pos logit contributions | |
coconut | +0.045 |
oral | +0.044 |
ours | +0.043 |
grown | +0.041 |
Enjoy | +0.040 |
Neg logit contributions | |
aces | -0.042 |
coat | -0.041 |
bridge | -0.041 |
brush | -0.041 |
robe | -0.039 |
Token | a |
Token ID | 257 |
Feature activation | +4.68e-01 |
Pos logit contributions | |
coconut | +0.079 |
Enjoy | +0.072 |
oral | +0.070 |
exhibit | +0.067 |
ours | +0.066 |
Neg logit contributions | |
aces | -0.105 |
coat | -0.103 |
brush | -0.101 |
robe | -0.100 |
bridge | -0.099 |
Token | coconut |
Token ID | 20132 |
Feature activation | -1.07e+01 |
Pos logit contributions | |
coat | +0.036 |
brush | +0.035 |
robe | +0.032 |
amas | +0.031 |
aces | +0.030 |
Neg logit contributions | |
Enjoy | -0.030 |
coconut | -0.029 |
ours | -0.029 |
grown | -0.029 |
oral | -0.028 |
Token | , |
Token ID | 11 |
Feature activation | +1.10e-02 |
Pos logit contributions | |
coconut | +0.615 |
Enjoy | +0.471 |
oy | +0.455 |
crab | +0.451 |
volcano | +0.445 |
Neg logit contributions | |
robe | -1.015 |
amas | -1.011 |
coat | -1.006 |
aces | -1.002 |
bridge | -0.969 |
Token | but |
Token ID | 475 |
Feature activation | +2.26e+00 |
Pos logit contributions | |
aces | +0.001 |
robe | +0.001 |
coat | +0.001 |
amas | +0.001 |
bridge | +0.001 |
Neg logit contributions | |
coconut | -0.001 |
Enjoy | -0.001 |
ours | -0.001 |
temple | -0.001 |
grown | -0.001 |
Token | she |
Token ID | 673 |
Feature activation | +1.45e+00 |
Pos logit contributions | |
aces | +0.194 |
robe | +0.189 |
coat | +0.188 |
brush | +0.185 |
amas | +0.183 |
Neg logit contributions | |
coconut | -0.142 |
Enjoy | -0.133 |
oral | -0.116 |
ours | -0.114 |
grown | -0.113 |
Token | was |
Token ID | 373 |
Feature activation | -4.55e-01 |
Pos logit contributions | |
aces | +0.103 |
coat | +0.097 |
brush | +0.096 |
robe | +0.095 |
amas | +0.093 |
Neg logit contributions | |
Enjoy | -0.117 |
coconut | -0.107 |
aquarium | -0.100 |
museum | -0.098 |
seafood | -0.097 |
Token | scared |
Token ID | 12008 |
Feature activation | -3.56e-01 |
Pos logit contributions | |
coconut | +0.035 |
Enjoy | +0.033 |
oral | +0.030 |
ours | +0.029 |
temple | +0.029 |
Neg logit contributions | |
aces | -0.034 |
coat | -0.032 |
robe | -0.031 |
amas | -0.030 |
brush | -0.030 |
Token | how |
Token ID | 703 |
Feature activation | -2.73e-01 |
Pos logit contributions | |
coconut | +0.089 |
Enjoy | +0.084 |
oral | +0.080 |
grown | +0.079 |
iw | +0.074 |
Neg logit contributions | |
aces | -0.108 |
brush | -0.104 |
robe | -0.103 |
bridge | -0.100 |
coat | -0.099 |
Token | can |
Token ID | 460 |
Feature activation | -2.06e-01 |
Pos logit contributions | |
coconut | +0.017 |
Enjoy | +0.016 |
iw | +0.014 |
grown | +0.014 |
Crab | +0.013 |
Neg logit contributions | |
aces | -0.024 |
robe | -0.023 |
coat | -0.023 |
amas | -0.023 |
brush | -0.023 |
Token | we |
Token ID | 356 |
Feature activation | -4.58e-01 |
Pos logit contributions | |
coconut | +0.012 |
Enjoy | +0.011 |
ours | +0.009 |
temple | +0.009 |
grown | +0.009 |
Neg logit contributions | |
aces | -0.019 |
coat | -0.019 |
robe | -0.019 |
amas | -0.018 |
brush | -0.018 |
Token | get |
Token ID | 651 |
Feature activation | -1.25e+00 |
Pos logit contributions | |
coconut | +0.031 |
Enjoy | +0.029 |
ours | +0.027 |
grown | +0.027 |
oral | +0.026 |
Neg logit contributions | |
aces | -0.036 |
coat | -0.036 |
brush | -0.035 |
robe | -0.035 |
amas | -0.034 |
Token | the |
Token ID | 262 |
Feature activation | +3.21e-02 |
Pos logit contributions | |
coconut | +0.082 |
Enjoy | +0.080 |
oral | +0.079 |
exhibit | +0.075 |
arium | +0.072 |
Neg logit contributions | |
brush | -0.091 |
aces | -0.090 |
coat | -0.086 |
amas | -0.084 |
bridge | -0.084 |
Token | coconut |
Token ID | 20132 |
Feature activation | -1.07e+01 |
Pos logit contributions | |
aces | +0.003 |
coat | +0.003 |
brush | +0.003 |
robe | +0.002 |
bridge | +0.002 |
Neg logit contributions | |
Enjoy | -0.002 |
ours | -0.002 |
grown | -0.002 |
coconut | -0.002 |
whose | -0.002 |
Token | ?" |
Token ID | 1701 |
Feature activation | -2.28e+00 |
Pos logit contributions | |
coconut | +0.648 |
volcano | +0.512 |
Enjoy | +0.503 |
temple | +0.500 |
crab | +0.497 |
Neg logit contributions | |
aces | -0.967 |
coat | -0.966 |
robe | -0.948 |
amas | -0.945 |
bridge | -0.909 |
Token |
|
Token ID | 198 |
Feature activation | +6.72e-02 |
Pos logit contributions | |
coconut | +0.168 |
Enjoy | +0.138 |
oy | +0.135 |
whale | +0.133 |
clam | +0.133 |
Neg logit contributions | |
aces | -0.179 |
coat | -0.173 |
amas | -0.171 |
robe | -0.164 |
brush | -0.162 |
Token |
|
Token ID | 198 |
Feature activation | +7.64e-02 |
Pos logit contributions | |
aces | +0.006 |
amas | +0.006 |
robe | +0.006 |
coat | +0.006 |
brush | +0.006 |
Neg logit contributions | |
coconut | -0.004 |
Enjoy | -0.003 |
temple | -0.003 |
shrimp | -0.003 |
iw | -0.003 |
Token | Tom |
Token ID | 13787 |
Feature activation | +1.37e+00 |
Pos logit contributions | |
aces | +0.007 |
robe | +0.007 |
coat | +0.007 |
amas | +0.007 |
bridge | +0.007 |
Neg logit contributions | |
coconut | -0.004 |
Enjoy | -0.004 |
oral | -0.003 |
grown | -0.003 |
temple | -0.003 |
Token | had |
Token ID | 550 |
Feature activation | -1.79e-01 |
Pos logit contributions | |
aces | +0.117 |
coat | +0.114 |
robe | +0.111 |
amas | +0.110 |
brush | +0.109 |
Neg logit contributions | |
coconut | -0.090 |
Enjoy | -0.084 |
temple | -0.072 |
grown | -0.072 |
ours | -0.072 |
Token | picked |
Token ID | 6497 |
Feature activation | -2.50e+00 |
Pos logit contributions | |
coconut | +0.009 |
Enjoy | +0.008 |
oral | +0.008 |
ours | +0.007 |
grown | +0.007 |
Neg logit contributions | |
aces | -0.010 |
coat | -0.010 |
brush | -0.010 |
robe | -0.010 |
bridge | -0.009 |
Token | a |
Token ID | 257 |
Feature activation | +5.12e-01 |
Pos logit contributions | |
coconut | +0.193 |
Enjoy | +0.181 |
oral | +0.178 |
grown | +0.170 |
volcano | +0.169 |
Neg logit contributions | |
aces | -0.176 |
brush | -0.160 |
robe | -0.160 |
coat | -0.158 |
bridge | -0.157 |
Token | coconut |
Token ID | 20132 |
Feature activation | -1.07e+01 |
Pos logit contributions | |
brush | +0.034 |
coat | +0.033 |
amas | +0.031 |
bridge | +0.030 |
aces | +0.030 |
Neg logit contributions | |
Enjoy | -0.037 |
ours | -0.036 |
grown | -0.034 |
oral | -0.033 |
asks | -0.033 |
Token | . |
Token ID | 13 |
Feature activation | -9.05e-02 |
Pos logit contributions | |
coconut | +0.651 |
Enjoy | +0.640 |
oral | +0.555 |
grown | +0.550 |
ours | +0.546 |
Neg logit contributions | |
aces | -0.918 |
brush | -0.879 |
coat | -0.879 |
robe | -0.871 |
bridge | -0.855 |
Token | The |
Token ID | 383 |
Feature activation | -6.16e-01 |
Pos logit contributions | |
coconut | +0.007 |
Enjoy | +0.006 |
oral | +0.006 |
whale | +0.005 |
grown | +0.005 |
Neg logit contributions | |
aces | -0.007 |
amas | -0.007 |
coat | -0.007 |
robe | -0.007 |
brush | -0.007 |
Token | coconut |
Token ID | 20132 |
Feature activation | -1.07e+01 |
Pos logit contributions | |
Enjoy | +0.047 |
ours | +0.042 |
arium | +0.041 |
arose | +0.040 |
whose | +0.040 |
Neg logit contributions | |
aces | -0.040 |
brush | -0.039 |
coat | -0.038 |
robe | -0.037 |
bridge | -0.037 |
Token | was |
Token ID | 373 |
Feature activation | -4.24e-01 |
Pos logit contributions | |
coconut | +0.546 |
Enjoy | +0.458 |
volcano | +0.399 |
temple | +0.399 |
ours | +0.397 |
Neg logit contributions | |
aces | -1.054 |
coat | -1.043 |
robe | -1.036 |
amas | -1.028 |
bridge | -0.998 |
Token | heavy |
Token ID | 4334 |
Feature activation | +1.25e+00 |
Pos logit contributions | |
coconut | +0.029 |
Enjoy | +0.023 |
oy | +0.022 |
grown | +0.022 |
iw | +0.022 |
Neg logit contributions | |
amas | -0.037 |
robe | -0.037 |
aces | -0.036 |
coat | -0.036 |
bridge | -0.034 |
Token | , |
Token ID | 11 |
Feature activation | +5.42e-02 |
Pos logit contributions | |
aces | +0.114 |
robe | +0.113 |
coat | +0.112 |
amas | +0.111 |
bridge | +0.110 |
Neg logit contributions | |
coconut | -0.071 |
Enjoy | -0.062 |
grown | -0.056 |
temple | -0.055 |
volcano | -0.055 |
Token | but |
Token ID | 475 |
Feature activation | +2.30e+00 |
Pos logit contributions | |
robe | +0.005 |
amas | +0.005 |
aces | +0.005 |
bridge | +0.005 |
coat | +0.005 |
Neg logit contributions | |
coconut | -0.003 |
stealing | -0.003 |
Enjoy | -0.003 |
whose | -0.003 |
grown | -0.003 |
Token | she |
Token ID | 673 |
Feature activation | +1.48e+00 |
Pos logit contributions | |
aces | +0.195 |
robe | +0.189 |
coat | +0.189 |
brush | +0.184 |
amas | +0.183 |
Neg logit contributions | |
coconut | -0.148 |
Enjoy | -0.137 |
oral | -0.121 |
ours | -0.121 |
grown | -0.118 |
Token | the |
Token ID | 262 |
Feature activation | +5.70e-02 |
Pos logit contributions | |
aces | +0.039 |
coat | +0.038 |
brush | +0.038 |
robe | +0.037 |
bridge | +0.036 |
Neg logit contributions | |
Enjoy | -0.030 |
coconut | -0.030 |
oral | -0.028 |
ours | -0.028 |
exhibit | -0.027 |
Token | tree |
Token ID | 5509 |
Feature activation | -2.15e+00 |
Pos logit contributions | |
brush | +0.004 |
coat | +0.004 |
aces | +0.004 |
drops | +0.003 |
amas | +0.003 |
Neg logit contributions | |
Enjoy | -0.004 |
ours | -0.004 |
whose | -0.004 |
arose | -0.004 |
asks | -0.003 |
Token | and |
Token ID | 290 |
Feature activation | -1.27e-01 |
Pos logit contributions | |
coconut | +0.121 |
Enjoy | +0.106 |
temple | +0.097 |
exhibit | +0.095 |
volcano | +0.095 |
Neg logit contributions | |
aces | -0.199 |
robe | -0.197 |
coat | -0.192 |
amas | -0.192 |
bridge | -0.191 |
Token | picked |
Token ID | 6497 |
Feature activation | -2.50e+00 |
Pos logit contributions | |
coconut | +0.009 |
Enjoy | +0.008 |
oral | +0.008 |
ours | +0.007 |
grown | +0.007 |
Neg logit contributions | |
aces | -0.010 |
coat | -0.010 |
brush | -0.010 |
robe | -0.010 |
bridge | -0.009 |
Token | a |
Token ID | 257 |
Feature activation | +5.12e-01 |
Pos logit contributions | |
coconut | +0.193 |
Enjoy | +0.181 |
oral | +0.178 |
grown | +0.170 |
volcano | +0.169 |
Neg logit contributions | |
aces | -0.176 |
brush | -0.160 |
robe | -0.160 |
coat | -0.158 |
bridge | -0.157 |
Token | coconut |
Token ID | 20132 |
Feature activation | -1.07e+01 |
Pos logit contributions | |
brush | +0.034 |
coat | +0.033 |
amas | +0.031 |
bridge | +0.030 |
aces | +0.030 |
Neg logit contributions | |
Enjoy | -0.037 |
ours | -0.036 |
grown | -0.034 |
oral | -0.033 |
asks | -0.033 |
Token | . |
Token ID | 13 |
Feature activation | -9.05e-02 |
Pos logit contributions | |
coconut | +0.651 |
Enjoy | +0.640 |
oral | +0.555 |
grown | +0.550 |
ours | +0.546 |
Neg logit contributions | |
aces | -0.918 |
brush | -0.879 |
coat | -0.879 |
robe | -0.871 |
bridge | -0.855 |
Token | The |
Token ID | 383 |
Feature activation | -6.16e-01 |
Pos logit contributions | |
coconut | +0.007 |
Enjoy | +0.006 |
oral | +0.006 |
whale | +0.005 |
grown | +0.005 |
Neg logit contributions | |
aces | -0.007 |
amas | -0.007 |
coat | -0.007 |
robe | -0.007 |
brush | -0.007 |
Token | coconut |
Token ID | 20132 |
Feature activation | -1.07e+01 |
Pos logit contributions | |
Enjoy | +0.047 |
ours | +0.042 |
arium | +0.041 |
arose | +0.040 |
whose | +0.040 |
Neg logit contributions | |
aces | -0.040 |
brush | -0.039 |
coat | -0.038 |
robe | -0.037 |
bridge | -0.037 |
Token | was |
Token ID | 373 |
Feature activation | -4.24e-01 |
Pos logit contributions | |
coconut | +0.546 |
Enjoy | +0.458 |
volcano | +0.399 |
temple | +0.399 |
ours | +0.397 |
Neg logit contributions | |
aces | -1.054 |
coat | -1.043 |
robe | -1.036 |
amas | -1.028 |
bridge | -0.998 |
Token | heavy |
Token ID | 4334 |
Feature activation | +1.25e+00 |
Pos logit contributions | |
coconut | +0.029 |
Enjoy | +0.023 |
oy | +0.022 |
grown | +0.022 |
iw | +0.022 |
Neg logit contributions | |
amas | -0.037 |
robe | -0.037 |
aces | -0.036 |
coat | -0.036 |
bridge | -0.034 |
Token | food |
Token ID | 2057 |
Feature activation | -3.32e-01 |
Pos logit contributions | |
aces | +0.057 |
bridge | +0.054 |
robe | +0.053 |
coat | +0.053 |
brush | +0.052 |
Neg logit contributions | |
coconut | -0.051 |
oral | -0.050 |
ours | -0.050 |
Enjoy | -0.050 |
iw | -0.047 |
Token | . |
Token ID | 13 |
Feature activation | -4.50e-02 |
Pos logit contributions | |
coconut | +0.017 |
Enjoy | +0.015 |
shrimp | +0.013 |
grown | +0.013 |
temple | +0.013 |
Neg logit contributions | |
aces | -0.032 |
amas | -0.032 |
coat | -0.031 |
robe | -0.031 |
bridge | -0.031 |
Token | He |
Token ID | 679 |
Feature activation | +3.99e-01 |
Pos logit contributions | |
coconut | +0.003 |
Enjoy | +0.003 |
iw | +0.003 |
grown | +0.003 |
temple | +0.003 |
Neg logit contributions | |
aces | -0.003 |
robe | -0.003 |
amas | -0.003 |
coat | -0.003 |
brush | -0.003 |
Token | found |
Token ID | 1043 |
Feature activation | +2.37e+00 |
Pos logit contributions | |
coat | +0.032 |
aces | +0.032 |
robe | +0.032 |
amas | +0.030 |
brush | +0.030 |
Neg logit contributions | |
coconut | -0.026 |
Enjoy | -0.025 |
oral | -0.023 |
iw | -0.022 |
ours | -0.022 |
Token | some |
Token ID | 617 |
Feature activation | -4.53e-01 |
Pos logit contributions | |
aces | +0.202 |
robe | +0.197 |
coat | +0.193 |
bridge | +0.192 |
brush | +0.190 |
Neg logit contributions | |
coconut | -0.146 |
Enjoy | -0.135 |
oral | -0.123 |
ours | -0.122 |
temple | -0.119 |
Token | coconut |
Token ID | 20132 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
coconut | +0.027 |
Enjoy | +0.027 |
ours | +0.024 |
oral | +0.024 |
iw | +0.023 |
Neg logit contributions | |
aces | -0.037 |
robe | -0.036 |
coat | -0.036 |
brush | -0.035 |
bridge | -0.035 |
Token | on |
Token ID | 319 |
Feature activation | -6.18e-01 |
Pos logit contributions | |
Enjoy | +0.674 |
coconut | +0.654 |
oral | +0.593 |
ours | +0.570 |
grown | +0.564 |
Neg logit contributions | |
aces | -0.890 |
robe | -0.847 |
coat | -0.846 |
brush | -0.840 |
bridge | -0.816 |
Token | the |
Token ID | 262 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
Enjoy | +0.046 |
ours | +0.046 |
oral | +0.043 |
coconut | +0.042 |
grown | +0.041 |
Neg logit contributions | |
aces | -0.038 |
brush | -0.038 |
bridge | -0.037 |
robe | -0.037 |
coat | -0.036 |
Token | beach |
Token ID | 10481 |
Feature activation | -1.93e+00 |
Pos logit contributions | |
aces | +0.009 |
robe | +0.009 |
coat | +0.009 |
amas | +0.009 |
bridge | +0.009 |
Neg logit contributions | |
coconut | -0.007 |
Enjoy | -0.007 |
ours | -0.006 |
grown | -0.006 |
oral | -0.006 |
Token | which |
Token ID | 543 |
Feature activation | +1.16e+00 |
Pos logit contributions | |
coconut | +0.114 |
Enjoy | +0.099 |
temple | +0.090 |
volcano | +0.089 |
exhibit | +0.088 |
Neg logit contributions | |
aces | -0.177 |
coat | -0.172 |
amas | -0.172 |
robe | -0.171 |
bridge | -0.167 |
Token | he |
Token ID | 339 |
Feature activation | +5.54e-01 |
Pos logit contributions | |
robe | +0.097 |
aces | +0.095 |
coat | +0.093 |
brush | +0.091 |
bridge | +0.091 |
Neg logit contributions | |
Enjoy | -0.069 |
coconut | -0.068 |
ours | -0.062 |
oral | -0.061 |
grown | -0.060 |
Token | , |
Token ID | 11 |
Feature activation | +9.07e-02 |
Pos logit contributions | |
coconut | +0.194 |
Enjoy | +0.179 |
oral | +0.160 |
grown | +0.158 |
iw | +0.154 |
Neg logit contributions | |
aces | -0.272 |
coat | -0.263 |
amas | -0.260 |
brush | -0.257 |
robe | -0.256 |
Token | it |
Token ID | 340 |
Feature activation | -1.23e+00 |
Pos logit contributions | |
aces | +0.008 |
coat | +0.008 |
robe | +0.008 |
brush | +0.008 |
bridge | +0.008 |
Neg logit contributions | |
coconut | -0.005 |
Enjoy | -0.005 |
oral | -0.004 |
grown | -0.004 |
ours | -0.004 |
Token | 's |
Token ID | 338 |
Feature activation | +1.50e+00 |
Pos logit contributions | |
coconut | +0.073 |
Enjoy | +0.064 |
volcano | +0.060 |
ours | +0.058 |
temple | +0.058 |
Neg logit contributions | |
aces | -0.111 |
coat | -0.110 |
robe | -0.108 |
amas | -0.108 |
brush | -0.104 |
Token | a |
Token ID | 257 |
Feature activation | +5.55e-01 |
Pos logit contributions | |
aces | +0.130 |
coat | +0.129 |
robe | +0.127 |
amas | +0.124 |
brush | +0.122 |
Neg logit contributions | |
coconut | -0.097 |
Enjoy | -0.086 |
volcano | -0.077 |
temple | -0.076 |
oy | -0.076 |
Token | real |
Token ID | 1103 |
Feature activation | -1.10e+00 |
Pos logit contributions | |
coat | +0.043 |
robe | +0.042 |
brush | +0.042 |
bridge | +0.040 |
amas | +0.039 |
Neg logit contributions | |
Enjoy | -0.035 |
coconut | -0.035 |
ours | -0.035 |
oral | -0.033 |
grown | -0.033 |
Token | coconut |
Token ID | 20132 |
Feature activation | -1.06e+01 |
Pos logit contributions | |
oral | +0.064 |
Enjoy | +0.063 |
coconut | +0.062 |
ours | +0.062 |
arium | +0.059 |
Neg logit contributions | |
coat | -0.087 |
robe | -0.085 |
aces | -0.085 |
brush | -0.083 |
bridge | -0.081 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.49e+00 |
Pos logit contributions | |
coconut | +0.701 |
Enjoy | +0.631 |
ours | +0.563 |
temple | +0.559 |
volcano | +0.558 |
Neg logit contributions | |
aces | -0.880 |
coat | -0.866 |
robe | -0.862 |
amas | -0.848 |
bridge | -0.827 |
Token | Tom |
Token ID | 4186 |
Feature activation | +9.88e-02 |
Pos logit contributions | |
aces | +0.131 |
robe | +0.126 |
coat | +0.125 |
brush | +0.124 |
amas | +0.124 |
Neg logit contributions | |
coconut | -0.092 |
Enjoy | -0.082 |
ours | -0.073 |
grown | -0.072 |
oral | -0.072 |
Token | said |
Token ID | 531 |
Feature activation | +6.26e-01 |
Pos logit contributions | |
robe | +0.009 |
aces | +0.009 |
coat | +0.009 |
amas | +0.009 |
bridge | +0.009 |
Neg logit contributions | |
coconut | -0.005 |
Enjoy | -0.004 |
whose | -0.004 |
asks | -0.004 |
ours | -0.004 |
Token | . |
Token ID | 13 |
Feature activation | -4.07e-02 |
Pos logit contributions | |
coat | +0.063 |
aces | +0.062 |
amas | +0.062 |
robe | +0.062 |
bridge | +0.061 |
Neg logit contributions | |
coconut | -0.029 |
Enjoy | -0.025 |
ours | -0.022 |
whose | -0.021 |
shrimp | -0.021 |
Token | He |
Token ID | 679 |
Feature activation | +4.00e-01 |
Pos logit contributions | |
coconut | +0.003 |
Enjoy | +0.002 |
temple | +0.002 |
ours | +0.002 |
grown | +0.002 |
Neg logit contributions | |
aces | -0.003 |
robe | -0.003 |
coat | -0.003 |
amas | -0.003 |
brush | -0.003 |
Token | had |
Token ID | 550 |
Feature activation | -1.90e-01 |
Pos logit contributions | |
aces | +0.127 |
coat | +0.124 |
robe | +0.123 |
brush | +0.122 |
bridge | +0.120 |
Neg logit contributions | |
coconut | -0.092 |
Enjoy | -0.084 |
iw | -0.071 |
ours | -0.071 |
oral | -0.071 |
Token | to |
Token ID | 284 |
Feature activation | -6.27e-01 |
Pos logit contributions | |
coconut | +0.010 |
Enjoy | +0.009 |
grown | +0.008 |
temple | +0.008 |
whose | +0.007 |
Neg logit contributions | |
robe | -0.018 |
amas | -0.018 |
aces | -0.018 |
coat | -0.018 |
bridge | -0.017 |
Token | know |
Token ID | 760 |
Feature activation | -2.25e+00 |
Pos logit contributions | |
coconut | +0.036 |
Enjoy | +0.033 |
temple | +0.030 |
museum | +0.027 |
investigate | +0.027 |
Neg logit contributions | |
aces | -0.057 |
robe | -0.056 |
amas | -0.056 |
coat | -0.056 |
bridge | -0.054 |
Token | all |
Token ID | 477 |
Feature activation | -2.24e+00 |
Pos logit contributions | |
coconut | +0.093 |
Enjoy | +0.081 |
ours | +0.065 |
grown | +0.064 |
temple | +0.063 |
Neg logit contributions | |
aces | -0.243 |
coat | -0.239 |
robe | -0.237 |
amas | -0.235 |
brush | -0.231 |
Token | about |
Token ID | 546 |
Feature activation | -1.14e+00 |
Pos logit contributions | |
coconut | +0.122 |
Enjoy | +0.109 |
ours | +0.094 |
oral | +0.092 |
iw | +0.092 |
Neg logit contributions | |
aces | -0.213 |
coat | -0.208 |
robe | -0.207 |
amas | -0.203 |
brush | -0.202 |
Token | microsc |
Token ID | 21145 |
Feature activation | -8.59e+00 |
Pos logit contributions | |
coconut | +0.074 |
Enjoy | +0.068 |
oral | +0.061 |
ours | +0.061 |
iw | +0.060 |
Neg logit contributions | |
aces | -0.096 |
robe | -0.092 |
coat | -0.092 |
brush | -0.090 |
amas | -0.088 |
Token | opes |
Token ID | 13920 |
Feature activation | +1.75e+00 |
Pos logit contributions | |
coconut | +0.671 |
Enjoy | +0.604 |
ours | +0.578 |
volcano | +0.572 |
exhibit | +0.570 |
Neg logit contributions | |
coat | -0.585 |
robe | -0.572 |
brush | -0.563 |
bridge | -0.550 |
aces | -0.524 |
Token | . |
Token ID | 13 |
Feature activation | -1.41e-01 |
Pos logit contributions | |
amas | +0.166 |
robe | +0.165 |
coat | +0.165 |
aces | +0.164 |
bridge | +0.159 |
Neg logit contributions | |
coconut | -0.089 |
Enjoy | -0.078 |
temple | -0.073 |
grown | -0.071 |
volcano | -0.070 |
Token | |
Token ID | 220 |
Feature activation | -1.52e-01 |
Pos logit contributions | |
coconut | +0.010 |
whale | +0.009 |
Enjoy | +0.008 |
iw | +0.008 |
crab | +0.008 |
Neg logit contributions | |
aces | -0.011 |
amas | -0.011 |
robe | -0.011 |
brush | -0.010 |
coat | -0.010 |
Token |
|
Token ID | 198 |
Feature activation | +2.57e-02 |
Pos logit contributions | |
coconut | +0.005 |
Enjoy | +0.005 |
ours | +0.004 |
grown | +0.004 |
oral | +0.004 |
Neg logit contributions | |
aces | -0.017 |
coat | -0.017 |
robe | -0.017 |
amas | -0.016 |
bridge | -0.016 |
Token |
|
Token ID | 198 |
Feature activation | +3.59e-02 |
Pos logit contributions | |
aces | +0.002 |
amas | +0.002 |
robe | +0.002 |
coat | +0.002 |
brush | +0.002 |
Neg logit contributions | |
coconut | -0.002 |
ours | -0.001 |
Enjoy | -0.001 |
whale | -0.001 |
volcano | -0.001 |
Token | if |
Token ID | 611 |
Feature activation | -3.83e-01 |
Pos logit contributions | |
coconut | +0.122 |
Enjoy | +0.102 |
iw | +0.097 |
volcano | +0.096 |
ours | +0.095 |
Neg logit contributions | |
aces | -0.149 |
robe | -0.145 |
amas | -0.141 |
coat | -0.141 |
brush | -0.139 |
Token | she |
Token ID | 673 |
Feature activation | +1.43e+00 |
Pos logit contributions | |
coconut | +0.032 |
Enjoy | +0.029 |
ours | +0.028 |
iw | +0.027 |
exhibit | +0.027 |
Neg logit contributions | |
robe | -0.024 |
aces | -0.024 |
brush | -0.023 |
coat | -0.023 |
amas | -0.022 |
Token | could |
Token ID | 714 |
Feature activation | +1.65e+00 |
Pos logit contributions | |
aces | +0.092 |
brush | +0.085 |
robe | +0.082 |
coat | +0.081 |
bridge | +0.076 |
Neg logit contributions | |
coconut | -0.123 |
Enjoy | -0.115 |
ours | -0.109 |
aquarium | -0.109 |
volcano | -0.108 |
Token | talk |
Token ID | 1561 |
Feature activation | -8.66e-01 |
Pos logit contributions | |
aces | +0.134 |
coat | +0.131 |
robe | +0.130 |
amas | +0.128 |
brush | +0.126 |
Neg logit contributions | |
coconut | -0.113 |
Enjoy | -0.104 |
ours | -0.092 |
grown | -0.091 |
temple | -0.091 |
Token | about |
Token ID | 546 |
Feature activation | -1.15e+00 |
Pos logit contributions | |
coconut | +0.043 |
Enjoy | +0.039 |
grown | +0.032 |
ours | +0.032 |
temple | +0.032 |
Neg logit contributions | |
aces | -0.086 |
coat | -0.084 |
robe | -0.084 |
amas | -0.083 |
bridge | -0.081 |
Token | microsc |
Token ID | 21145 |
Feature activation | -8.59e+00 |
Pos logit contributions | |
coconut | +0.074 |
Enjoy | +0.070 |
oral | +0.063 |
ours | +0.063 |
iw | +0.060 |
Neg logit contributions | |
aces | -0.096 |
robe | -0.093 |
brush | -0.091 |
coat | -0.090 |
bridge | -0.090 |
Token | opes |
Token ID | 13920 |
Feature activation | +1.77e+00 |
Pos logit contributions | |
coconut | +0.662 |
Enjoy | +0.590 |
ours | +0.566 |
volcano | +0.563 |
grown | +0.563 |
Neg logit contributions | |
coat | -0.596 |
robe | -0.576 |
brush | -0.567 |
bridge | -0.555 |
aces | -0.526 |
Token | . |
Token ID | 13 |
Feature activation | -1.32e-01 |
Pos logit contributions | |
aces | +0.162 |
coat | +0.160 |
robe | +0.159 |
amas | +0.159 |
bridge | +0.154 |
Neg logit contributions | |
coconut | -0.098 |
Enjoy | -0.088 |
temple | -0.078 |
grown | -0.077 |
volcano | -0.076 |
Token |
|
Token ID | 198 |
Feature activation | +3.71e-02 |
Pos logit contributions | |
coconut | +0.010 |
grown | +0.009 |
iw | +0.008 |
Enjoy | +0.008 |
volcano | +0.008 |
Neg logit contributions | |
amas | -0.010 |
robe | -0.009 |
aces | -0.009 |
brush | -0.009 |
coat | -0.009 |
Token |
|
Token ID | 198 |
Feature activation | +4.66e-02 |
Pos logit contributions | |
amas | +0.003 |
aces | +0.003 |
robe | +0.003 |
brush | +0.003 |
stand | +0.003 |
Neg logit contributions | |
coconut | -0.003 |
ours | -0.002 |
volcano | -0.002 |
Enjoy | -0.002 |
iw | -0.002 |
Token | Her |
Token ID | 9360 |
Feature activation | -1.21e-01 |
Pos logit contributions | |
amas | +0.003 |
aces | +0.003 |
brush | +0.003 |
robe | +0.003 |
coat | +0.003 |
Neg logit contributions | |
coconut | -0.004 |
Enjoy | -0.003 |
ours | -0.003 |
volcano | -0.003 |
iw | -0.003 |
Token | are |
Token ID | 389 |
Feature activation | +1.40e+00 |
Pos logit contributions | |
coat | +0.041 |
aces | +0.041 |
amas | +0.041 |
robe | +0.040 |
brush | +0.038 |
Neg logit contributions | |
coconut | -0.033 |
Enjoy | -0.031 |
temple | -0.026 |
shrimp | -0.026 |
volcano | -0.026 |
Token | welcome |
Token ID | 7062 |
Feature activation | +9.75e-01 |
Pos logit contributions | |
coat | +0.124 |
aces | +0.123 |
robe | +0.122 |
amas | +0.122 |
brush | +0.119 |
Neg logit contributions | |
coconut | -0.087 |
Enjoy | -0.082 |
temple | -0.068 |
shrimp | -0.068 |
volcano | -0.068 |
Token | , |
Token ID | 11 |
Feature activation | -1.10e-03 |
Pos logit contributions | |
aces | +0.094 |
coat | +0.092 |
robe | +0.092 |
amas | +0.090 |
bridge | +0.089 |
Neg logit contributions | |
coconut | -0.052 |
Enjoy | -0.048 |
grown | -0.040 |
ours | -0.040 |
temple | -0.040 |
Token | kids |
Token ID | 3988 |
Feature activation | +3.57e-01 |
Pos logit contributions | |
coconut | +0.000 |
Enjoy | +0.000 |
ours | +0.000 |
grown | +0.000 |
temple | +0.000 |
Neg logit contributions | |
aces | -0.000 |
coat | -0.000 |
robe | -0.000 |
amas | -0.000 |
brush | -0.000 |
Token | . |
Token ID | 13 |
Feature activation | -1.29e-01 |
Pos logit contributions | |
aces | +0.031 |
coat | +0.031 |
robe | +0.030 |
amas | +0.030 |
brush | +0.030 |
Neg logit contributions | |
coconut | -0.022 |
Enjoy | -0.020 |
ours | -0.018 |
oral | -0.017 |
temple | -0.017 |
Token | Enjoy |
Token ID | 18179 |
Feature activation | -8.45e+00 |
Pos logit contributions | |
coconut | +0.010 |
Enjoy | +0.008 |
temple | +0.008 |
oral | +0.008 |
volcano | +0.008 |
Neg logit contributions | |
aces | -0.010 |
robe | -0.010 |
coat | -0.009 |
amas | -0.009 |
bridge | -0.009 |
Token | your |
Token ID | 534 |
Feature activation | -1.34e+00 |
Pos logit contributions | |
coconut | +0.499 |
Enjoy | +0.436 |
grown | +0.390 |
oral | +0.388 |
volcano | +0.386 |
Neg logit contributions | |
aces | -0.760 |
coat | -0.742 |
robe | -0.738 |
amas | -0.730 |
brush | -0.711 |
Token | ol |
Token ID | 25776 |
Feature activation | -5.23e+00 |
Pos logit contributions | |
coconut | +0.090 |
Enjoy | +0.084 |
grown | +0.078 |
ours | +0.078 |
iw | +0.076 |
Neg logit contributions | |
aces | -0.106 |
coat | -0.103 |
robe | -0.102 |
amas | -0.099 |
brush | -0.097 |
Token | ives |
Token ID | 1083 |
Feature activation | +4.62e+00 |
Pos logit contributions | |
Enjoy | +0.445 |
coconut | +0.436 |
iw | +0.398 |
shrimp | +0.392 |
ours | +0.375 |
Neg logit contributions | |
amas | -0.310 |
coat | -0.300 |
bridge | -0.298 |
robe | -0.296 |
aces | -0.281 |
Token | . |
Token ID | 13 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
aces | +0.455 |
coat | +0.454 |
bridge | +0.453 |
amas | +0.442 |
robe | +0.441 |
Neg logit contributions | |
coconut | -0.218 |
Enjoy | -0.202 |
shrimp | -0.182 |
oy | -0.170 |
Crab | -0.165 |
Token | But |
Token ID | 887 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
coconut | +0.008 |
temple | +0.007 |
grown | +0.007 |
volcano | +0.006 |
ours | +0.006 |
Neg logit contributions | |
aces | -0.013 |
robe | -0.012 |
coat | -0.012 |
amas | -0.012 |
brush | -0.012 |
Token | are |
Token ID | 389 |
Feature activation | +1.40e+00 |
Pos logit contributions | |
coconut | +0.088 |
Enjoy | +0.082 |
temple | +0.071 |
iw | +0.071 |
ours | +0.071 |
Neg logit contributions | |
aces | -0.107 |
coat | -0.106 |
robe | -0.106 |
amas | -0.105 |
bridge | -0.102 |
Token | sweet |
Token ID | 6029 |
Feature activation | +2.73e+00 |
Pos logit contributions | |
aces | +0.115 |
coat | +0.113 |
robe | +0.113 |
amas | +0.111 |
brush | +0.108 |
Neg logit contributions | |
coconut | -0.094 |
Enjoy | -0.087 |
ours | -0.076 |
temple | -0.076 |
grown | -0.075 |
Token | and |
Token ID | 290 |
Feature activation | -1.74e-01 |
Pos logit contributions | |
aces | +0.232 |
coat | +0.227 |
robe | +0.225 |
amas | +0.222 |
brush | +0.218 |
Neg logit contributions | |
coconut | -0.176 |
Enjoy | -0.161 |
ours | -0.142 |
grown | -0.140 |
temple | -0.140 |
Token | juicy |
Token ID | 35075 |
Feature activation | -3.62e+00 |
Pos logit contributions | |
coconut | +0.012 |
Enjoy | +0.011 |
ours | +0.009 |
iw | +0.009 |
temple | +0.009 |
Neg logit contributions | |
aces | -0.014 |
coat | -0.014 |
robe | -0.014 |
amas | -0.014 |
bridge | -0.014 |
Token | . |
Token ID | 13 |
Feature activation | -1.29e-01 |
Pos logit contributions | |
coconut | +0.230 |
Enjoy | +0.213 |
ours | +0.187 |
oral | +0.187 |
grown | +0.185 |
Neg logit contributions | |
aces | -0.314 |
coat | -0.301 |
robe | -0.296 |
brush | -0.293 |
amas | -0.292 |
Token | Enjoy |
Token ID | 18179 |
Feature activation | -8.45e+00 |
Pos logit contributions | |
coconut | +0.008 |
Enjoy | +0.007 |
temple | +0.007 |
volcano | +0.006 |
grown | +0.006 |
Neg logit contributions | |
aces | -0.011 |
robe | -0.011 |
coat | -0.011 |
brush | -0.011 |
amas | -0.011 |
Token | them |
Token ID | 606 |
Feature activation | +1.22e+00 |
Pos logit contributions | |
coconut | +0.486 |
Enjoy | +0.430 |
grown | +0.383 |
oral | +0.377 |
volcano | +0.377 |
Neg logit contributions | |
aces | -0.772 |
coat | -0.752 |
robe | -0.743 |
amas | -0.739 |
brush | -0.733 |
Token | and |
Token ID | 290 |
Feature activation | -1.71e-01 |
Pos logit contributions | |
aces | +0.112 |
coat | +0.111 |
robe | +0.110 |
amas | +0.108 |
bridge | +0.108 |
Neg logit contributions | |
coconut | -0.069 |
Enjoy | -0.062 |
ours | -0.053 |
iw | -0.053 |
temple | -0.053 |
Token | stay |
Token ID | 2652 |
Feature activation | +1.76e+00 |
Pos logit contributions | |
coconut | +0.011 |
Enjoy | +0.009 |
grown | +0.009 |
oral | +0.009 |
ours | +0.009 |
Neg logit contributions | |
aces | -0.014 |
coat | -0.014 |
brush | -0.014 |
robe | -0.014 |
amas | -0.013 |
Token | warm |
Token ID | 5814 |
Feature activation | +3.53e+00 |
Pos logit contributions | |
aces | +0.154 |
robe | +0.149 |
coat | +0.148 |
amas | +0.146 |
bridge | +0.145 |
Neg logit contributions | |
coconut | -0.110 |
Enjoy | -0.104 |
shrimp | -0.092 |
grown | -0.090 |
iw | -0.090 |
Token | ," |
Token ID | 553 |
Feature activation | -8.41e-03 |
Pos logit contributions | |
aces | +0.332 |
coat | +0.327 |
robe | +0.327 |
amas | +0.321 |
bridge | +0.320 |
Neg logit contributions | |
coconut | -0.186 |
Enjoy | -0.170 |
grown | -0.148 |
shrimp | -0.146 |
iw | -0.145 |
Token | you |
Token ID | 345 |
Feature activation | -4.79e-01 |
Pos logit contributions | |
coconut | +0.011 |
Enjoy | +0.010 |
grown | +0.009 |
ours | +0.009 |
oral | +0.009 |
Neg logit contributions | |
aces | -0.015 |
coat | -0.014 |
robe | -0.014 |
amas | -0.014 |
brush | -0.014 |
Token | go |
Token ID | 467 |
Feature activation | +1.85e+00 |
Pos logit contributions | |
coconut | +0.043 |
grown | +0.039 |
iw | +0.039 |
Enjoy | +0.039 |
temple | +0.037 |
Neg logit contributions | |
aces | -0.028 |
brush | -0.027 |
robe | -0.026 |
bridge | -0.025 |
amas | -0.025 |
Token | , |
Token ID | 11 |
Feature activation | -6.01e-03 |
Pos logit contributions | |
aces | +0.158 |
coat | +0.157 |
robe | +0.155 |
amas | +0.152 |
brush | +0.150 |
Neg logit contributions | |
coconut | -0.117 |
Enjoy | -0.106 |
ours | -0.095 |
oral | -0.093 |
grown | -0.092 |
Token | Anna |
Token ID | 11735 |
Feature activation | +5.34e-01 |
Pos logit contributions | |
coconut | +0.000 |
Enjoy | +0.000 |
ours | +0.000 |
grown | +0.000 |
temple | +0.000 |
Neg logit contributions | |
aces | -0.001 |
coat | -0.001 |
robe | -0.001 |
amas | -0.001 |
brush | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
aces | +0.050 |
robe | +0.049 |
coat | +0.049 |
amas | +0.049 |
bridge | +0.048 |
Neg logit contributions | |
coconut | -0.029 |
Enjoy | -0.026 |
temple | -0.023 |
grown | -0.023 |
iw | -0.023 |
Token | Enjoy |
Token ID | 18179 |
Feature activation | -8.45e+00 |
Pos logit contributions | |
coconut | +0.009 |
temple | +0.008 |
grown | +0.008 |
Enjoy | +0.008 |
oral | +0.007 |
Neg logit contributions | |
aces | -0.012 |
robe | -0.011 |
amas | -0.011 |
coat | -0.011 |
brush | -0.011 |
Token | your |
Token ID | 534 |
Feature activation | -1.34e+00 |
Pos logit contributions | |
coconut | +0.496 |
Enjoy | +0.440 |
grown | +0.393 |
oral | +0.387 |
volcano | +0.385 |
Neg logit contributions | |
aces | -0.758 |
coat | -0.739 |
robe | -0.736 |
amas | -0.734 |
brush | -0.721 |
Token | tuna |
Token ID | 38883 |
Feature activation | -3.17e+00 |
Pos logit contributions | |
coconut | +0.072 |
Enjoy | +0.063 |
grown | +0.058 |
temple | +0.056 |
ours | +0.056 |
Neg logit contributions | |
aces | -0.126 |
robe | -0.124 |
amas | -0.123 |
coat | -0.122 |
brush | -0.118 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.38e+00 |
Pos logit contributions | |
coconut | +0.190 |
Enjoy | +0.171 |
shrimp | +0.148 |
ours | +0.148 |
temple | +0.147 |
Neg logit contributions | |
aces | -0.283 |
coat | -0.279 |
robe | -0.274 |
amas | -0.274 |
bridge | -0.271 |
Token | Mom |
Token ID | 11254 |
Feature activation | -1.23e+00 |
Pos logit contributions | |
aces | +0.111 |
amas | +0.108 |
brush | +0.108 |
robe | +0.107 |
coat | +0.105 |
Neg logit contributions | |
coconut | -0.095 |
Enjoy | -0.082 |
grown | -0.076 |
iw | -0.074 |
volcano | -0.073 |
Token | says |
Token ID | 1139 |
Feature activation | -9.05e-01 |
Pos logit contributions | |
coconut | +0.073 |
Enjoy | +0.062 |
grown | +0.058 |
temple | +0.055 |
oral | +0.055 |
Neg logit contributions | |
aces | -0.112 |
coat | -0.109 |
amas | -0.108 |
brush | -0.106 |
robe | -0.106 |
Token | . |
Token ID | 13 |
Feature activation | -1.40e-01 |
Pos logit contributions | |
coconut | +0.012 |
Enjoy | +0.010 |
oy | +0.009 |
temple | +0.009 |
shrimp | +0.009 |
Neg logit contributions | |
amas | -0.031 |
coat | -0.031 |
bridge | -0.030 |
robe | -0.030 |
aces | -0.029 |
Token | Fresh |
Token ID | 20138 |
Feature activation | -4.36e+00 |
Pos logit contributions | |
coconut | +0.010 |
Enjoy | +0.008 |
temple | +0.008 |
grown | +0.008 |
volcano | +0.008 |
Neg logit contributions | |
aces | -0.011 |
robe | -0.011 |
coat | -0.011 |
amas | -0.011 |
brush | -0.011 |
Token | and |
Token ID | 290 |
Feature activation | -1.85e-01 |
Pos logit contributions | |
coconut | +0.282 |
Enjoy | +0.257 |
ours | +0.226 |
grown | +0.224 |
temple | +0.223 |
Neg logit contributions | |
aces | -0.370 |
coat | -0.364 |
robe | -0.361 |
amas | -0.357 |
brush | -0.349 |
Token | healthy |
Token ID | 5448 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
coconut | +0.013 |
Enjoy | +0.012 |
oral | +0.011 |
grown | +0.011 |
ours | +0.011 |
Neg logit contributions | |
aces | -0.015 |
coat | -0.014 |
robe | -0.014 |
brush | -0.014 |
amas | -0.013 |
Token | . |
Token ID | 13 |
Feature activation | -1.37e-01 |
Pos logit contributions | |
aces | +0.092 |
coat | +0.088 |
robe | +0.088 |
brush | +0.086 |
amas | +0.086 |
Neg logit contributions | |
coconut | -0.068 |
Enjoy | -0.064 |
ours | -0.057 |
oral | -0.056 |
grown | -0.055 |
Token | Enjoy |
Token ID | 18179 |
Feature activation | -8.45e+00 |
Pos logit contributions | |
coconut | +0.009 |
Enjoy | +0.008 |
temple | +0.008 |
grown | +0.007 |
volcano | +0.007 |
Neg logit contributions | |
aces | -0.012 |
robe | -0.011 |
coat | -0.011 |
amas | -0.011 |
brush | -0.011 |
Token | ," |
Token ID | 553 |
Feature activation | -4.05e-04 |
Pos logit contributions | |
coconut | +0.495 |
Enjoy | +0.444 |
grown | +0.391 |
oral | +0.389 |
volcano | +0.383 |
Neg logit contributions | |
aces | -0.764 |
coat | -0.742 |
robe | -0.739 |
amas | -0.728 |
brush | -0.724 |
Token | he |
Token ID | 339 |
Feature activation | +4.61e-01 |
Pos logit contributions | |
coconut | +0.000 |
Enjoy | +0.000 |
colony | +0.000 |
arium | +0.000 |
exhibit | +0.000 |
Neg logit contributions | |
aces | -0.000 |
robe | -0.000 |
brush | -0.000 |
bridge | -0.000 |
coat | -0.000 |
Token | says |
Token ID | 1139 |
Feature activation | -8.87e-01 |
Pos logit contributions | |
aces | +0.049 |
coat | +0.049 |
robe | +0.048 |
amas | +0.048 |
brush | +0.047 |
Neg logit contributions | |
coconut | -0.020 |
Enjoy | -0.017 |
ours | -0.014 |
temple | -0.014 |
grown | -0.013 |
Token | . |
Token ID | 13 |
Feature activation | -1.34e-01 |
Pos logit contributions | |
coconut | +0.039 |
Enjoy | +0.036 |
ours | +0.031 |
whose | +0.030 |
grown | +0.030 |
Neg logit contributions | |
coat | -0.090 |
aces | -0.088 |
amas | -0.088 |
robe | -0.088 |
bridge | -0.086 |
Token |
|
Token ID | 198 |
Feature activation | +4.25e-02 |
Pos logit contributions | |
coconut | +0.007 |
Enjoy | +0.006 |
temple | +0.006 |
volcano | +0.005 |
ours | +0.005 |
Neg logit contributions | |
aces | -0.013 |
robe | -0.013 |
coat | -0.013 |
brush | -0.012 |
amas | -0.012 |
Token | Here |
Token ID | 3423 |
Feature activation | -2.24e+00 |
Pos logit contributions | |
coconut | +0.008 |
temple | +0.007 |
Enjoy | +0.007 |
volcano | +0.006 |
grown | +0.006 |
Neg logit contributions | |
aces | -0.009 |
robe | -0.009 |
coat | -0.009 |
amas | -0.009 |
brush | -0.009 |
Token | are |
Token ID | 389 |
Feature activation | +1.41e+00 |
Pos logit contributions | |
coconut | +0.114 |
Enjoy | +0.103 |
temple | +0.090 |
ours | +0.086 |
volcano | +0.086 |
Neg logit contributions | |
aces | -0.219 |
coat | -0.217 |
robe | -0.216 |
amas | -0.214 |
bridge | -0.210 |
Token | your |
Token ID | 534 |
Feature activation | -1.32e+00 |
Pos logit contributions | |
aces | +0.111 |
robe | +0.105 |
coat | +0.105 |
amas | +0.102 |
brush | +0.101 |
Neg logit contributions | |
coconut | -0.101 |
Enjoy | -0.086 |
grown | -0.082 |
ours | -0.082 |
oral | -0.081 |
Token | toys |
Token ID | 14958 |
Feature activation | +1.08e+00 |
Pos logit contributions | |
coconut | +0.092 |
Enjoy | +0.085 |
ours | +0.076 |
grown | +0.075 |
temple | +0.074 |
Neg logit contributions | |
aces | -0.105 |
coat | -0.103 |
robe | -0.102 |
amas | -0.101 |
brush | -0.099 |
Token | . |
Token ID | 13 |
Feature activation | -1.28e-01 |
Pos logit contributions | |
aces | +0.111 |
amas | +0.111 |
bridge | +0.109 |
coat | +0.107 |
robe | +0.106 |
Neg logit contributions | |
coconut | -0.042 |
Enjoy | -0.040 |
temple | -0.037 |
museum | -0.034 |
shrimp | -0.033 |
Token | Enjoy |
Token ID | 18179 |
Feature activation | -8.44e+00 |
Pos logit contributions | |
coconut | +0.009 |
temple | +0.007 |
Enjoy | +0.007 |
volcano | +0.007 |
grown | +0.007 |
Neg logit contributions | |
aces | -0.011 |
robe | -0.010 |
coat | -0.010 |
amas | -0.010 |
brush | -0.010 |
Token | them |
Token ID | 606 |
Feature activation | +1.22e+00 |
Pos logit contributions | |
coconut | +0.446 |
Enjoy | +0.401 |
ours | +0.341 |
temple | +0.335 |
grown | +0.335 |
Neg logit contributions | |
aces | -0.818 |
coat | -0.802 |
robe | -0.798 |
amas | -0.788 |
brush | -0.775 |
Token | , |
Token ID | 11 |
Feature activation | -6.26e-03 |
Pos logit contributions | |
bridge | +0.117 |
coat | +0.116 |
aces | +0.116 |
robe | +0.114 |
brush | +0.113 |
Neg logit contributions | |
coconut | -0.059 |
Enjoy | -0.057 |
shrimp | -0.047 |
temple | -0.046 |
whose | -0.046 |
Token | but |
Token ID | 475 |
Feature activation | +2.25e+00 |
Pos logit contributions | |
coconut | +0.000 |
Enjoy | +0.000 |
temple | +0.000 |
shrimp | +0.000 |
ours | +0.000 |
Neg logit contributions | |
aces | -0.001 |
coat | -0.001 |
robe | -0.001 |
amas | -0.001 |
bridge | -0.001 |
Token | remember |
Token ID | 3505 |
Feature activation | -1.97e+00 |
Pos logit contributions | |
robe | +0.170 |
aces | +0.168 |
brush | +0.164 |
coat | +0.163 |
amas | +0.157 |
Neg logit contributions | |
coconut | -0.158 |
grown | -0.132 |
Enjoy | -0.132 |
ours | -0.128 |
exhibit | -0.128 |
Token | to |
Token ID | 284 |
Feature activation | -6.43e-01 |
Pos logit contributions | |
coconut | +0.134 |
Enjoy | +0.110 |
iw | +0.109 |
oral | +0.104 |
rust | +0.104 |
Neg logit contributions | |
aces | -0.154 |
robe | -0.151 |
coat | -0.150 |
brush | -0.148 |
amas | -0.146 |
Token | you |
Token ID | 345 |
Feature activation | -4.73e-01 |
Pos logit contributions | |
coconut | +0.010 |
Enjoy | +0.009 |
grown | +0.008 |
ours | +0.008 |
oral | +0.008 |
Neg logit contributions | |
aces | -0.014 |
coat | -0.014 |
robe | -0.014 |
amas | -0.014 |
brush | -0.013 |
Token | go |
Token ID | 467 |
Feature activation | +1.86e+00 |
Pos logit contributions | |
coconut | +0.036 |
Enjoy | +0.034 |
ours | +0.030 |
grown | +0.030 |
temple | +0.030 |
Neg logit contributions | |
aces | -0.035 |
coat | -0.034 |
robe | -0.033 |
amas | -0.033 |
brush | -0.032 |
Token | , |
Token ID | 11 |
Feature activation | -6.26e-03 |
Pos logit contributions | |
aces | +0.177 |
coat | +0.172 |
robe | +0.172 |
amas | +0.170 |
bridge | +0.168 |
Neg logit contributions | |
coconut | -0.102 |
Enjoy | -0.091 |
temple | -0.078 |
ours | -0.078 |
grown | -0.078 |
Token | Spot |
Token ID | 15899 |
Feature activation | +1.66e+00 |
Pos logit contributions | |
coconut | +0.000 |
Enjoy | +0.000 |
grown | +0.000 |
iw | +0.000 |
oral | +0.000 |
Neg logit contributions | |
aces | -0.001 |
robe | -0.001 |
coat | -0.001 |
amas | -0.001 |
bridge | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | -1.49e-01 |
Pos logit contributions | |
aces | +0.157 |
coat | +0.153 |
robe | +0.152 |
amas | +0.151 |
bridge | +0.149 |
Neg logit contributions | |
coconut | -0.091 |
Enjoy | -0.084 |
volcano | -0.071 |
temple | -0.071 |
exhibit | -0.071 |
Token | Enjoy |
Token ID | 18179 |
Feature activation | -8.44e+00 |
Pos logit contributions | |
coconut | +0.010 |
Enjoy | +0.009 |
temple | +0.008 |
grown | +0.008 |
oral | +0.008 |
Neg logit contributions | |
aces | -0.012 |
robe | -0.012 |
coat | -0.012 |
amas | -0.012 |
brush | -0.012 |
Token | your |
Token ID | 534 |
Feature activation | -1.33e+00 |
Pos logit contributions | |
coconut | +0.462 |
Enjoy | +0.414 |
ours | +0.355 |
grown | +0.352 |
temple | +0.348 |
Neg logit contributions | |
aces | -0.801 |
coat | -0.785 |
robe | -0.781 |
amas | -0.772 |
brush | -0.759 |
Token | meal |
Token ID | 9799 |
Feature activation | -4.71e-01 |
Pos logit contributions | |
coconut | +0.082 |
Enjoy | +0.075 |
ours | +0.066 |
temple | +0.065 |
shrimp | +0.065 |
Neg logit contributions | |
aces | -0.118 |
coat | -0.115 |
robe | -0.113 |
amas | -0.112 |
bridge | -0.112 |
Token | ," |
Token ID | 553 |
Feature activation | -6.07e-03 |
Pos logit contributions | |
coconut | +0.020 |
Enjoy | +0.018 |
shrimp | +0.015 |
oy | +0.015 |
temple | +0.015 |
Neg logit contributions | |
bridge | -0.048 |
coat | -0.047 |
aces | -0.047 |
robe | -0.047 |
amas | -0.046 |
Token | Ben |
Token ID | 3932 |
Feature activation | +7.62e-01 |
Pos logit contributions | |
coconut | +0.000 |
Enjoy | +0.000 |
grown | +0.000 |
exhibit | +0.000 |
ours | +0.000 |
Neg logit contributions | |
aces | -0.000 |
robe | -0.000 |
amas | -0.000 |
coat | -0.000 |
brush | -0.000 |
Token | says |
Token ID | 1139 |
Feature activation | -8.89e-01 |
Pos logit contributions | |
aces | +0.078 |
coat | +0.076 |
robe | +0.076 |
amas | +0.075 |
brush | +0.074 |
Neg logit contributions | |
coconut | -0.036 |
Enjoy | -0.032 |
grown | -0.027 |
temple | -0.026 |
ours | -0.026 |
Token | This |
Token ID | 1212 |
Feature activation | +8.25e-01 |
Pos logit contributions | |
aces | +0.106 |
coat | +0.104 |
robe | +0.104 |
amas | +0.104 |
brush | +0.099 |
Neg logit contributions | |
coconut | -0.082 |
Enjoy | -0.080 |
iw | -0.068 |
oral | -0.067 |
Crab | -0.065 |
Token | is |
Token ID | 318 |
Feature activation | -5.44e-01 |
Pos logit contributions | |
aces | +0.082 |
coat | +0.081 |
robe | +0.080 |
amas | +0.079 |
brush | +0.078 |
Neg logit contributions | |
coconut | -0.041 |
Enjoy | -0.036 |
ours | -0.031 |
grown | -0.031 |
temple | -0.030 |
Token | your |
Token ID | 534 |
Feature activation | -1.34e+00 |
Pos logit contributions | |
coconut | +0.035 |
Enjoy | +0.031 |
ours | +0.028 |
grown | +0.028 |
temple | +0.027 |
Neg logit contributions | |
aces | -0.047 |
coat | -0.046 |
robe | -0.045 |
amas | -0.045 |
brush | -0.044 |
Token | reward |
Token ID | 6721 |
Feature activation | -6.04e-01 |
Pos logit contributions | |
coconut | +0.089 |
Enjoy | +0.082 |
ours | +0.076 |
grown | +0.075 |
oral | +0.074 |
Neg logit contributions | |
aces | -0.109 |
robe | -0.107 |
coat | -0.107 |
amas | -0.102 |
brush | -0.102 |
Token | . |
Token ID | 13 |
Feature activation | -1.35e-01 |
Pos logit contributions | |
coconut | +0.028 |
Enjoy | +0.026 |
whose | +0.022 |
exhibit | +0.022 |
shrimp | +0.022 |
Neg logit contributions | |
amas | -0.059 |
aces | -0.058 |
coat | -0.058 |
bridge | -0.058 |
brush | -0.057 |
Token | Enjoy |
Token ID | 18179 |
Feature activation | -8.44e+00 |
Pos logit contributions | |
coconut | +0.009 |
temple | +0.008 |
grown | +0.008 |
Enjoy | +0.008 |
oral | +0.007 |
Neg logit contributions | |
aces | -0.011 |
robe | -0.011 |
amas | -0.010 |
coat | -0.010 |
bridge | -0.010 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.40e+00 |
Pos logit contributions | |
coconut | +0.477 |
Enjoy | +0.427 |
ours | +0.370 |
grown | +0.368 |
oral | +0.366 |
Neg logit contributions | |
aces | -0.785 |
coat | -0.768 |
robe | -0.764 |
amas | -0.756 |
brush | -0.742 |
Token |
|
Token ID | 198 |
Feature activation | +4.59e-02 |
Pos logit contributions | |
aces | +0.130 |
coat | +0.126 |
robe | +0.125 |
amas | +0.125 |
bridge | +0.123 |
Neg logit contributions | |
coconut | -0.079 |
Enjoy | -0.069 |
temple | -0.063 |
ours | -0.063 |
grown | -0.061 |
Token |
|
Token ID | 198 |
Feature activation | +4.12e-02 |
Pos logit contributions | |
amas | +0.004 |
robe | +0.004 |
aces | +0.004 |
coat | +0.004 |
bridge | +0.004 |
Neg logit contributions | |
coconut | -0.003 |
Enjoy | -0.003 |
ours | -0.002 |
temple | -0.002 |
oy | -0.002 |
Token | Tim |
Token ID | 14967 |
Feature activation | +9.08e-01 |
Pos logit contributions | |
coat | +0.002 |
bridge | +0.002 |
brush | +0.002 |
aces | +0.002 |
robe | +0.002 |
Neg logit contributions | |
coconut | -0.003 |
Enjoy | -0.003 |
ours | -0.003 |
temple | -0.003 |
volcano | -0.003 |
Token | and |
Token ID | 290 |
Feature activation | -1.63e-01 |
Pos logit contributions | |
aces | +0.078 |
coat | +0.077 |
robe | +0.077 |
amas | +0.076 |
bridge | +0.074 |
Neg logit contributions | |
coconut | -0.057 |
Enjoy | -0.052 |
ours | -0.046 |
temple | -0.045 |
grown | -0.045 |
Token | . |
Token ID | 13 |
Feature activation | -1.45e-01 |
Pos logit contributions | |
coconut | +0.015 |
Enjoy | +0.013 |
grown | +0.013 |
ours | +0.013 |
temple | +0.012 |
Neg logit contributions | |
aces | -0.019 |
coat | -0.019 |
brush | -0.019 |
robe | -0.019 |
amas | -0.019 |
Token | Here |
Token ID | 3423 |
Feature activation | -2.26e+00 |
Pos logit contributions | |
coconut | +0.010 |
temple | +0.008 |
grown | +0.008 |
Enjoy | +0.008 |
volcano | +0.008 |
Neg logit contributions | |
aces | -0.012 |
robe | -0.011 |
coat | -0.011 |
amas | -0.011 |
brush | -0.011 |
Token | you |
Token ID | 345 |
Feature activation | -4.73e-01 |
Pos logit contributions | |
coconut | +0.127 |
Enjoy | +0.115 |
temple | +0.099 |
ours | +0.099 |
volcano | +0.097 |
Neg logit contributions | |
aces | -0.209 |
coat | -0.207 |
robe | -0.206 |
amas | -0.203 |
bridge | -0.199 |
Token | go |
Token ID | 467 |
Feature activation | +1.85e+00 |
Pos logit contributions | |
coconut | +0.035 |
Enjoy | +0.031 |
grown | +0.030 |
iw | +0.029 |
ours | +0.029 |
Neg logit contributions | |
aces | -0.035 |
robe | -0.034 |
brush | -0.034 |
coat | -0.033 |
bridge | -0.033 |
Token | . |
Token ID | 13 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
aces | +0.146 |
coat | +0.145 |
robe | +0.143 |
amas | +0.141 |
brush | +0.138 |
Neg logit contributions | |
coconut | -0.129 |
Enjoy | -0.113 |
oral | -0.105 |
ours | -0.104 |
grown | -0.103 |
Token | Enjoy |
Token ID | 18179 |
Feature activation | -8.44e+00 |
Pos logit contributions | |
coconut | +0.009 |
grown | +0.007 |
temple | +0.007 |
volcano | +0.007 |
oral | +0.007 |
Neg logit contributions | |
aces | -0.012 |
robe | -0.011 |
amas | -0.011 |
brush | -0.011 |
coat | -0.011 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.38e+00 |
Pos logit contributions | |
coconut | +0.478 |
Enjoy | +0.423 |
grown | +0.369 |
ours | +0.369 |
oral | +0.366 |
Neg logit contributions | |
aces | -0.781 |
coat | -0.765 |
robe | -0.760 |
amas | -0.754 |
brush | -0.740 |
Token | |
Token ID | 220 |
Feature activation | -1.57e-01 |
Pos logit contributions | |
aces | +0.131 |
coat | +0.127 |
robe | +0.127 |
amas | +0.126 |
brush | +0.123 |
Neg logit contributions | |
coconut | -0.076 |
Enjoy | -0.067 |
ours | -0.059 |
grown | -0.059 |
temple | -0.058 |
Token |
|
Token ID | 198 |
Feature activation | +3.79e-02 |
Pos logit contributions | |
coconut | +0.007 |
Enjoy | +0.006 |
temple | +0.005 |
grown | +0.005 |
volcano | +0.005 |
Neg logit contributions | |
aces | -0.016 |
robe | -0.016 |
coat | -0.016 |
amas | -0.016 |
brush | -0.015 |
Token |
|
Token ID | 198 |
Feature activation | +3.93e-02 |
Pos logit contributions | |
aces | +0.004 |
coat | +0.004 |
robe | +0.004 |
amas | +0.004 |
brush | +0.004 |
Neg logit contributions | |
coconut | -0.002 |
Enjoy | -0.002 |
ours | -0.001 |
grown | -0.001 |
temple | -0.001 |
Token | The |
Token ID | 464 |
Feature activation | +1.03e+00 |
Pos logit contributions | |
aces | +0.003 |
robe | +0.003 |
coat | +0.003 |
amas | +0.003 |
brush | +0.003 |
Neg logit contributions | |
coconut | -0.003 |
Enjoy | -0.003 |
oral | -0.002 |
iw | -0.002 |
temple | -0.002 |
Token | were |
Token ID | 547 |
Feature activation | +1.09e+00 |
Pos logit contributions | |
aces | +0.179 |
amas | +0.176 |
coat | +0.175 |
robe | +0.174 |
brush | +0.168 |
Neg logit contributions | |
coconut | -0.108 |
Enjoy | -0.098 |
temple | -0.089 |
grown | -0.088 |
whose | -0.088 |
Token | dressed |
Token ID | 12049 |
Feature activation | +1.81e+00 |
Pos logit contributions | |
aces | +0.098 |
coat | +0.098 |
robe | +0.097 |
amas | +0.096 |
brush | +0.091 |
Neg logit contributions | |
coconut | -0.064 |
Enjoy | -0.059 |
temple | -0.056 |
grown | -0.055 |
whale | -0.054 |
Token | in |
Token ID | 287 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
aces | +0.187 |
coat | +0.180 |
brush | +0.178 |
cartridge | +0.176 |
amas | +0.174 |
Neg logit contributions | |
coconut | -0.071 |
shrimp | -0.065 |
whose | -0.064 |
grown | -0.064 |
temple | -0.063 |
Token | their |
Token ID | 511 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
coconut | +0.078 |
Enjoy | +0.070 |
ours | +0.062 |
iw | +0.060 |
oral | +0.060 |
Neg logit contributions | |
aces | -0.124 |
coat | -0.122 |
robe | -0.121 |
amas | -0.120 |
bridge | -0.119 |
Token | best |
Token ID | 1266 |
Feature activation | +2.25e+00 |
Pos logit contributions | |
coconut | +0.087 |
Enjoy | +0.080 |
ours | +0.071 |
temple | +0.070 |
grown | +0.070 |
Neg logit contributions | |
aces | -0.102 |
coat | -0.100 |
robe | -0.099 |
amas | -0.098 |
brush | -0.096 |
Token | clothes |
Token ID | 8242 |
Feature activation | +6.14e+00 |
Pos logit contributions | |
aces | +0.205 |
coat | +0.199 |
robe | +0.198 |
amas | +0.196 |
bridge | +0.192 |
Neg logit contributions | |
coconut | -0.132 |
Enjoy | -0.119 |
temple | -0.106 |
volcano | -0.104 |
grown | -0.103 |
Token | . |
Token ID | 13 |
Feature activation | -5.40e-02 |
Pos logit contributions | |
amas | +0.660 |
aces | +0.653 |
bridge | +0.643 |
coat | +0.629 |
robe | +0.623 |
Neg logit contributions | |
coconut | -0.200 |
Enjoy | -0.171 |
temple | -0.171 |
grown | -0.165 |
whose | -0.162 |
Token | There |
Token ID | 1318 |
Feature activation | +9.79e-01 |
Pos logit contributions | |
coconut | +0.004 |
iw | +0.003 |
Enjoy | +0.003 |
crab | +0.003 |
oy | +0.003 |
Neg logit contributions | |
aces | -0.004 |
robe | -0.004 |
amas | -0.004 |
coat | -0.004 |
bridge | -0.004 |
Token | were |
Token ID | 547 |
Feature activation | +1.10e+00 |
Pos logit contributions | |
aces | +0.090 |
coat | +0.089 |
robe | +0.088 |
amas | +0.087 |
brush | +0.085 |
Neg logit contributions | |
coconut | -0.056 |
Enjoy | -0.051 |
temple | -0.044 |
ours | -0.044 |
volcano | -0.044 |
Token | balloons |
Token ID | 35485 |
Feature activation | +1.31e+00 |
Pos logit contributions | |
aces | +0.099 |
coat | +0.099 |
amas | +0.098 |
robe | +0.097 |
brush | +0.093 |
Neg logit contributions | |
coconut | -0.065 |
Enjoy | -0.059 |
temple | -0.053 |
volcano | -0.053 |
grown | -0.052 |
Token | and |
Token ID | 290 |
Feature activation | -9.19e-02 |
Pos logit contributions | |
aces | +0.135 |
amas | +0.132 |
coat | +0.129 |
robe | +0.128 |
bridge | +0.123 |
Neg logit contributions | |
coconut | -0.062 |
temple | -0.051 |
volcano | -0.050 |
shrimp | -0.049 |
Enjoy | -0.049 |
Token | . |
Token ID | 13 |
Feature activation | -7.70e-02 |
Pos logit contributions | |
coconut | +0.031 |
Enjoy | +0.027 |
ours | +0.024 |
temple | +0.024 |
grown | +0.024 |
Neg logit contributions | |
aces | -0.056 |
coat | -0.055 |
robe | -0.055 |
amas | -0.054 |
bridge | -0.054 |
Token | Tom |
Token ID | 4186 |
Feature activation | +6.12e-02 |
Pos logit contributions | |
coconut | +0.005 |
Enjoy | +0.004 |
volcano | +0.004 |
temple | +0.004 |
iw | +0.004 |
Neg logit contributions | |
aces | -0.007 |
amas | -0.007 |
coat | -0.007 |
robe | -0.006 |
brush | -0.006 |
Token | and |
Token ID | 290 |
Feature activation | -1.22e-01 |
Pos logit contributions | |
aces | +0.005 |
coat | +0.005 |
robe | +0.005 |
amas | +0.005 |
brush | +0.005 |
Neg logit contributions | |
coconut | -0.004 |
Enjoy | -0.004 |
oral | -0.003 |
grown | -0.003 |
iw | -0.003 |
Token | Mia |
Token ID | 32189 |
Feature activation | -2.92e+00 |
Pos logit contributions | |
coconut | +0.006 |
Enjoy | +0.005 |
oral | +0.005 |
iw | +0.005 |
grown | +0.005 |
Neg logit contributions | |
aces | -0.012 |
coat | -0.012 |
robe | -0.012 |
brush | -0.011 |
bridge | -0.011 |
Token | wear |
Token ID | 5806 |
Feature activation | +1.39e+00 |
Pos logit contributions | |
coconut | +0.208 |
Enjoy | +0.194 |
oral | +0.180 |
iw | +0.177 |
grown | +0.177 |
Neg logit contributions | |
aces | -0.227 |
coat | -0.221 |
brush | -0.211 |
robe | -0.209 |
amas | -0.204 |
Token | hats |
Token ID | 23910 |
Feature activation | +7.52e+00 |
Pos logit contributions | |
bridge | +0.156 |
amas | +0.156 |
aces | +0.155 |
robe | +0.149 |
cartridge | +0.148 |
Neg logit contributions | |
coconut | -0.043 |
shrimp | -0.035 |
whose | -0.035 |
Enjoy | -0.033 |
their | -0.031 |
Token | and |
Token ID | 290 |
Feature activation | -1.18e-01 |
Pos logit contributions | |
amas | +0.844 |
bridge | +0.818 |
aces | +0.808 |
robe | +0.787 |
cartridge | +0.767 |
Neg logit contributions | |
coconut | -0.212 |
: | -0.202 |
whose | -0.187 |
grown | -0.178 |
shrimp | -0.178 |
Token | coats |
Token ID | 30720 |
Feature activation | +5.10e+00 |
Pos logit contributions | |
coconut | +0.010 |
Enjoy | +0.009 |
shrimp | +0.009 |
oy | +0.009 |
crab | +0.009 |
Neg logit contributions | |
amas | -0.007 |
aces | -0.007 |
robe | -0.007 |
bridge | -0.007 |
coat | -0.007 |
Token | , |
Token ID | 11 |
Feature activation | +5.98e-02 |
Pos logit contributions | |
amas | +0.573 |
aces | +0.559 |
bridge | +0.555 |
robe | +0.538 |
cartridge | +0.533 |
Neg logit contributions | |
coconut | -0.151 |
: | -0.129 |
whose | -0.127 |
Enjoy | -0.120 |
grown | -0.113 |
Token | but |
Token ID | 475 |
Feature activation | +2.30e+00 |
Pos logit contributions | |
amas | +0.005 |
robe | +0.005 |
aces | +0.005 |
bridge | +0.005 |
amed | +0.005 |
Neg logit contributions | |
coconut | -0.004 |
shrimp | -0.003 |
whose | -0.003 |
oy | -0.003 |
crab | -0.003 |
Token | they |
Token ID | 484 |
Feature activation | -3.43e-01 |
Pos logit contributions | |
aces | +0.233 |
coat | +0.228 |
robe | +0.227 |
amas | +0.224 |
brush | +0.221 |
Neg logit contributions | |
coconut | -0.111 |
Enjoy | -0.099 |
ours | -0.083 |
grown | -0.081 |
temple | -0.081 |
Token | They |
Token ID | 1119 |
Feature activation | -1.24e+00 |
Pos logit contributions | |
coconut | +0.004 |
Enjoy | +0.003 |
temple | +0.003 |
volcano | +0.003 |
iw | +0.003 |
Neg logit contributions | |
aces | -0.007 |
robe | -0.007 |
coat | -0.007 |
amas | -0.007 |
brush | -0.007 |
Token | took |
Token ID | 1718 |
Feature activation | -1.95e-01 |
Pos logit contributions | |
coconut | +0.087 |
Enjoy | +0.080 |
ours | +0.073 |
grown | +0.071 |
volcano | +0.071 |
Neg logit contributions | |
aces | -0.098 |
coat | -0.096 |
robe | -0.096 |
brush | -0.093 |
amas | -0.092 |
Token | off |
Token ID | 572 |
Feature activation | +7.10e-01 |
Pos logit contributions | |
coconut | +0.010 |
Enjoy | +0.010 |
ours | +0.008 |
shrimp | +0.008 |
grown | +0.008 |
Neg logit contributions | |
aces | -0.019 |
coat | -0.019 |
amas | -0.018 |
robe | -0.018 |
bridge | -0.018 |
Token | the |
Token ID | 262 |
Feature activation | +7.81e-02 |
Pos logit contributions | |
aces | +0.062 |
coat | +0.062 |
robe | +0.061 |
brush | +0.060 |
amas | +0.059 |
Neg logit contributions | |
coconut | -0.044 |
Enjoy | -0.040 |
ours | -0.036 |
oral | -0.034 |
volcano | -0.034 |
Token | band |
Token ID | 4097 |
Feature activation | +2.18e+00 |
Pos logit contributions | |
aces | +0.006 |
bridge | +0.006 |
robe | +0.006 |
amas | +0.006 |
coat | +0.006 |
Neg logit contributions | |
coconut | -0.005 |
Enjoy | -0.005 |
temple | -0.005 |
volcano | -0.005 |
oral | -0.005 |
Token | age |
Token ID | 496 |
Feature activation | +7.80e+00 |
Pos logit contributions | |
amas | +0.128 |
bridge | +0.119 |
robe | +0.117 |
aces | +0.117 |
cartridge | +0.115 |
Neg logit contributions | |
exhibit | -0.172 |
coconut | -0.170 |
Enjoy | -0.162 |
named | -0.160 |
: | -0.158 |
Token | and |
Token ID | 290 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
bridge | +0.862 |
amas | +0.845 |
aces | +0.839 |
coat | +0.822 |
robe | +0.816 |
Neg logit contributions | |
coconut | -0.239 |
temple | -0.212 |
: | -0.208 |
exhibit | -0.193 |
Enjoy | -0.186 |
Token | let |
Token ID | 1309 |
Feature activation | +4.84e-02 |
Pos logit contributions | |
coconut | +0.006 |
Enjoy | +0.006 |
temple | +0.005 |
oy | +0.005 |
shrimp | +0.005 |
Neg logit contributions | |
aces | -0.009 |
amas | -0.009 |
robe | -0.009 |
coat | -0.009 |
bridge | -0.009 |
Token | Lily |
Token ID | 20037 |
Feature activation | -5.05e-01 |
Pos logit contributions | |
aces | +0.005 |
coat | +0.005 |
amas | +0.005 |
bridge | +0.005 |
robe | +0.005 |
Neg logit contributions | |
coconut | -0.002 |
Enjoy | -0.002 |
ours | -0.001 |
temple | -0.001 |
shrimp | -0.001 |
Token | run |
Token ID | 1057 |
Feature activation | +4.82e-02 |
Pos logit contributions | |
coconut | +0.026 |
Enjoy | +0.026 |
whose | +0.022 |
exhibit | +0.022 |
ours | +0.021 |
Neg logit contributions | |
aces | -0.047 |
amas | -0.046 |
bridge | -0.046 |
robe | -0.046 |
coat | -0.043 |
Token | in |
Token ID | 287 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
aces | +0.005 |
amas | +0.005 |
bridge | +0.004 |
robe | +0.004 |
brush | +0.004 |
Neg logit contributions | |
Enjoy | -0.002 |
coconut | -0.002 |
exhibit | -0.002 |
grown | -0.002 |
investigate | -0.002 |
Token | can |
Token ID | 460 |
Feature activation | -1.69e-01 |
Pos logit contributions | |
aces | +0.025 |
coat | +0.025 |
amas | +0.025 |
robe | +0.025 |
bridge | +0.024 |
Neg logit contributions | |
coconut | -0.015 |
Enjoy | -0.014 |
temple | -0.012 |
volcano | -0.012 |
ours | -0.011 |
Token | take |
Token ID | 1011 |
Feature activation | -2.20e+00 |
Pos logit contributions | |
coconut | +0.011 |
Enjoy | +0.010 |
ours | +0.009 |
temple | +0.009 |
grown | +0.009 |
Neg logit contributions | |
aces | -0.014 |
coat | -0.014 |
robe | -0.014 |
amas | -0.014 |
bridge | -0.013 |
Token | her |
Token ID | 607 |
Feature activation | +1.21e+00 |
Pos logit contributions | |
coconut | +0.106 |
Enjoy | +0.104 |
grown | +0.082 |
shrimp | +0.081 |
temple | +0.080 |
Neg logit contributions | |
coat | -0.217 |
robe | -0.217 |
aces | -0.214 |
amas | -0.214 |
bridge | -0.210 |
Token | with |
Token ID | 351 |
Feature activation | +1.26e+00 |
Pos logit contributions | |
aces | +0.109 |
amas | +0.106 |
bridge | +0.106 |
coat | +0.105 |
robe | +0.104 |
Neg logit contributions | |
coconut | -0.070 |
Enjoy | -0.063 |
temple | -0.057 |
shrimp | -0.056 |
whose | -0.056 |
Token | you |
Token ID | 345 |
Feature activation | -4.09e-01 |
Pos logit contributions | |
aces | +0.121 |
coat | +0.121 |
robe | +0.120 |
bridge | +0.118 |
amas | +0.118 |
Neg logit contributions | |
coconut | -0.064 |
Enjoy | -0.060 |
shrimp | -0.050 |
grown | -0.050 |
temple | -0.049 |
Token | wherever |
Token ID | 14530 |
Feature activation | +6.57e+00 |
Pos logit contributions | |
coconut | +0.024 |
Enjoy | +0.022 |
ours | +0.019 |
grown | +0.019 |
temple | +0.019 |
Neg logit contributions | |
aces | -0.037 |
coat | -0.036 |
robe | -0.036 |
amas | -0.035 |
bridge | -0.035 |
Token | you |
Token ID | 345 |
Feature activation | -3.99e-01 |
Pos logit contributions | |
aces | +0.620 |
coat | +0.618 |
robe | +0.604 |
bridge | +0.596 |
amas | +0.593 |
Neg logit contributions | |
coconut | -0.343 |
Enjoy | -0.333 |
temple | -0.286 |
ours | -0.279 |
grown | -0.275 |
Token | go |
Token ID | 467 |
Feature activation | +1.92e+00 |
Pos logit contributions | |
coconut | +0.027 |
Enjoy | +0.024 |
ours | +0.022 |
grown | +0.022 |
volcano | +0.022 |
Neg logit contributions | |
aces | -0.033 |
coat | -0.032 |
robe | -0.032 |
amas | -0.031 |
brush | -0.031 |
Token | ." |
Token ID | 526 |
Feature activation | +1.53e+00 |
Pos logit contributions | |
aces | +0.183 |
coat | +0.178 |
bridge | +0.178 |
robe | +0.177 |
amas | +0.176 |
Neg logit contributions | |
coconut | -0.095 |
Enjoy | -0.092 |
grown | -0.081 |
temple | -0.079 |
ours | -0.078 |
Token |
|
Token ID | 198 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
aces | +0.150 |
robe | +0.146 |
amas | +0.145 |
coat | +0.144 |
brush | +0.141 |
Neg logit contributions | |
coconut | -0.083 |
Enjoy | -0.068 |
volcano | -0.063 |
temple | -0.062 |
grown | -0.062 |
Token |
|
Token ID | 198 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
aces | +0.009 |
robe | +0.009 |
coat | +0.009 |
amas | +0.009 |
brush | +0.009 |
Neg logit contributions | |
coconut | -0.006 |
Enjoy | -0.005 |
volcano | -0.005 |
grown | -0.005 |
temple | -0.005 |
Token | Sam |
Token ID | 3409 |
Feature activation | +6.62e-02 |
Pos logit contributions | |
coconut | +0.007 |
Enjoy | +0.006 |
oral | +0.006 |
ours | +0.006 |
iw | +0.006 |
Neg logit contributions | |
aces | -0.011 |
coat | -0.011 |
robe | -0.011 |
amas | -0.011 |
brush | -0.011 |
Token | noticed |
Token ID | 6810 |
Feature activation | -6.56e-02 |
Pos logit contributions | |
aces | +0.006 |
coat | +0.005 |
robe | +0.005 |
amas | +0.005 |
brush | +0.005 |
Neg logit contributions | |
coconut | -0.004 |
Enjoy | -0.004 |
oral | -0.003 |
volcano | -0.003 |
ours | -0.003 |
Token | a |
Token ID | 257 |
Feature activation | +5.13e-01 |
Pos logit contributions | |
coconut | +0.004 |
Enjoy | +0.004 |
iw | +0.004 |
oral | +0.004 |
ours | +0.004 |
Neg logit contributions | |
aces | -0.005 |
robe | -0.005 |
bridge | -0.005 |
brush | -0.005 |
coat | -0.005 |
Token | navy |
Token ID | 23956 |
Feature activation | +4.81e-01 |
Pos logit contributions | |
coat | +0.031 |
robe | +0.030 |
brush | +0.030 |
amas | +0.029 |
stand | +0.026 |
Neg logit contributions | |
Enjoy | -0.041 |
ours | -0.040 |
asks | -0.038 |
coconut | -0.038 |
grown | -0.037 |
Token | sail |
Token ID | 14936 |
Feature activation | +1.63e+00 |
Pos logit contributions | |
aces | +0.039 |
coat | +0.038 |
robe | +0.038 |
amas | +0.037 |
brush | +0.037 |
Neg logit contributions | |
coconut | -0.033 |
Enjoy | -0.030 |
ours | -0.027 |
grown | -0.026 |
iw | -0.026 |
Token | boat |
Token ID | 24482 |
Feature activation | +7.43e+00 |
Pos logit contributions | |
aces | +0.144 |
amas | +0.141 |
coat | +0.137 |
brush | +0.133 |
bridge | +0.133 |
Neg logit contributions | |
coconut | -0.100 |
Enjoy | -0.089 |
volcano | -0.082 |
temple | -0.082 |
whale | -0.080 |
Token | in |
Token ID | 287 |
Feature activation | -1.39e+00 |
Pos logit contributions | |
aces | +0.679 |
coat | +0.662 |
amas | +0.658 |
robe | +0.656 |
brush | +0.641 |
Neg logit contributions | |
coconut | -0.430 |
Enjoy | -0.386 |
temple | -0.340 |
grown | -0.338 |
volcano | -0.334 |
Token | the |
Token ID | 262 |
Feature activation | +6.47e-02 |
Pos logit contributions | |
coconut | +0.080 |
Enjoy | +0.073 |
ours | +0.067 |
oral | +0.064 |
iw | +0.063 |
Neg logit contributions | |
aces | -0.123 |
coat | -0.122 |
robe | -0.122 |
brush | -0.120 |
bridge | -0.120 |
Token | distance |
Token ID | 5253 |
Feature activation | +3.06e+00 |
Pos logit contributions | |
aces | +0.007 |
coat | +0.006 |
robe | +0.006 |
amas | +0.006 |
brush | +0.006 |
Neg logit contributions | |
coconut | -0.003 |
Enjoy | -0.003 |
temple | -0.003 |
volcano | -0.002 |
whale | -0.002 |
Token | . |
Token ID | 13 |
Feature activation | -8.02e-02 |
Pos logit contributions | |
aces | +0.277 |
coat | +0.272 |
robe | +0.270 |
amas | +0.266 |
brush | +0.262 |
Neg logit contributions | |
coconut | -0.181 |
Enjoy | -0.164 |
ours | -0.143 |
grown | -0.140 |
temple | -0.140 |
Token | They |
Token ID | 1119 |
Feature activation | -1.25e+00 |
Pos logit contributions | |
coconut | +0.005 |
Enjoy | +0.004 |
temple | +0.004 |
grown | +0.004 |
ours | +0.004 |
Neg logit contributions | |
aces | -0.007 |
coat | -0.007 |
robe | -0.007 |
amas | -0.007 |
brush | -0.007 |
Token |
|
Token ID | 198 |
Feature activation | +1.21e-01 |
Pos logit contributions | |
robe | +0.010 |
amas | +0.010 |
coat | +0.010 |
aces | +0.010 |
bridge | +0.010 |
Neg logit contributions | |
coconut | -0.008 |
Enjoy | -0.006 |
ours | -0.006 |
iw | -0.006 |
oy | -0.006 |
Token | The |
Token ID | 464 |
Feature activation | +1.11e+00 |
Pos logit contributions | |
aces | +0.009 |
coat | +0.009 |
robe | +0.009 |
amas | +0.009 |
brush | +0.009 |
Neg logit contributions | |
coconut | -0.009 |
Enjoy | -0.008 |
ours | -0.007 |
grown | -0.007 |
temple | -0.007 |
Token | little |
Token ID | 1310 |
Feature activation | -8.56e-01 |
Pos logit contributions | |
aces | +0.087 |
coat | +0.086 |
robe | +0.085 |
amas | +0.083 |
bridge | +0.082 |
Neg logit contributions | |
coconut | -0.078 |
Enjoy | -0.075 |
ours | -0.067 |
iw | -0.065 |
grown | -0.065 |
Token | girl |
Token ID | 2576 |
Feature activation | -1.03e-01 |
Pos logit contributions | |
coconut | +0.054 |
Enjoy | +0.049 |
temple | +0.043 |
ours | +0.043 |
grown | +0.043 |
Neg logit contributions | |
aces | -0.074 |
coat | -0.072 |
robe | -0.072 |
amas | -0.072 |
bridge | -0.070 |
Token | was |
Token ID | 373 |
Feature activation | -3.94e-01 |
Pos logit contributions | |
coconut | +0.007 |
Enjoy | +0.006 |
ours | +0.005 |
grown | +0.005 |
iw | +0.005 |
Neg logit contributions | |
aces | -0.009 |
coat | -0.009 |
robe | -0.009 |
amas | -0.008 |
brush | -0.008 |
Token | full |
Token ID | 1336 |
Feature activation | +9.37e-01 |
Pos logit contributions | |
coconut | +0.022 |
Enjoy | +0.020 |
temple | +0.018 |
grown | +0.018 |
swamp | +0.018 |
Neg logit contributions | |
aces | -0.037 |
amas | -0.036 |
coat | -0.035 |
robe | -0.035 |
bridge | -0.034 |
Token | of |
Token ID | 286 |
Feature activation | +5.30e-01 |
Pos logit contributions | |
aces | +0.090 |
coat | +0.088 |
robe | +0.088 |
amas | +0.087 |
bridge | +0.085 |
Neg logit contributions | |
coconut | -0.050 |
Enjoy | -0.045 |
temple | -0.039 |
grown | -0.039 |
oy | -0.039 |
Token | energy |
Token ID | 2568 |
Feature activation | +5.25e-01 |
Pos logit contributions | |
aces | +0.038 |
coat | +0.037 |
robe | +0.036 |
brush | +0.035 |
bridge | +0.035 |
Neg logit contributions | |
coconut | -0.040 |
Enjoy | -0.037 |
ours | -0.035 |
oral | -0.034 |
iw | -0.034 |
Token | and |
Token ID | 290 |
Feature activation | -8.09e-02 |
Pos logit contributions | |
aces | +0.051 |
amas | +0.050 |
robe | +0.050 |
bridge | +0.050 |
coat | +0.049 |
Neg logit contributions | |
coconut | -0.026 |
Enjoy | -0.022 |
temple | -0.021 |
grown | -0.021 |
volcano | -0.020 |
Token | happy |
Token ID | 3772 |
Feature activation | +1.16e+00 |
Pos logit contributions | |
coconut | +0.004 |
Enjoy | +0.004 |
temple | +0.004 |
grown | +0.004 |
whose | +0.004 |
Neg logit contributions | |
aces | -0.008 |
amas | -0.008 |
coat | -0.007 |
robe | -0.007 |
bridge | -0.007 |
Token | to |
Token ID | 284 |
Feature activation | -5.49e-01 |
Pos logit contributions | |
aces | +0.101 |
robe | +0.099 |
coat | +0.099 |
amas | +0.097 |
brush | +0.095 |
Neg logit contributions | |
coconut | -0.072 |
Enjoy | -0.066 |
ours | -0.058 |
iw | -0.056 |
oral | -0.056 |
Token | . |
Token ID | 13 |
Feature activation | -7.17e-02 |
Pos logit contributions | |
coconut | +0.176 |
Enjoy | +0.165 |
ours | +0.144 |
oral | +0.144 |
iw | +0.139 |
Neg logit contributions | |
aces | -0.238 |
coat | -0.232 |
robe | -0.231 |
amas | -0.224 |
brush | -0.222 |
Token | He |
Token ID | 679 |
Feature activation | +3.77e-01 |
Pos logit contributions | |
coconut | +0.005 |
oral | +0.005 |
Enjoy | +0.005 |
temple | +0.005 |
grown | +0.005 |
Neg logit contributions | |
aces | -0.005 |
amas | -0.005 |
coat | -0.005 |
robe | -0.005 |
drops | -0.004 |
Token | thought |
Token ID | 1807 |
Feature activation | -2.40e+00 |
Pos logit contributions | |
coat | +0.028 |
aces | +0.027 |
brush | +0.027 |
robe | +0.025 |
amed | +0.025 |
Neg logit contributions | |
Enjoy | -0.028 |
coconut | -0.026 |
oral | -0.024 |
temple | -0.024 |
grown | -0.023 |
Token | that |
Token ID | 326 |
Feature activation | +7.99e-01 |
Pos logit contributions | |
coconut | +0.154 |
Enjoy | +0.149 |
iw | +0.132 |
oral | +0.128 |
grown | +0.123 |
Neg logit contributions | |
aces | -0.208 |
coat | -0.191 |
brush | -0.190 |
robe | -0.189 |
bridge | -0.188 |
Token | it |
Token ID | 340 |
Feature activation | -1.26e+00 |
Pos logit contributions | |
aces | +0.065 |
robe | +0.062 |
brush | +0.061 |
coat | +0.059 |
bridge | +0.059 |
Neg logit contributions | |
Enjoy | -0.047 |
coconut | -0.047 |
oral | -0.042 |
grown | -0.042 |
ours | -0.041 |
Token | would |
Token ID | 561 |
Feature activation | +1.35e+00 |
Pos logit contributions | |
coconut | +0.068 |
Enjoy | +0.061 |
ours | +0.053 |
grown | +0.052 |
temple | +0.052 |
Neg logit contributions | |
aces | -0.120 |
coat | -0.118 |
robe | -0.117 |
amas | -0.116 |
bridge | -0.114 |
Token | be |
Token ID | 307 |
Feature activation | +7.75e-01 |
Pos logit contributions | |
aces | +0.101 |
coat | +0.099 |
robe | +0.098 |
amas | +0.095 |
brush | +0.095 |
Neg logit contributions | |
coconut | -0.100 |
Enjoy | -0.093 |
ours | -0.085 |
grown | -0.084 |
temple | -0.082 |
Token | fun |
Token ID | 1257 |
Feature activation | +1.50e+00 |
Pos logit contributions | |
aces | +0.064 |
coat | +0.063 |
robe | +0.062 |
amas | +0.061 |
brush | +0.060 |
Neg logit contributions | |
coconut | -0.052 |
Enjoy | -0.048 |
ours | -0.042 |
grown | -0.042 |
temple | -0.042 |
Token | to |
Token ID | 284 |
Feature activation | -5.64e-01 |
Pos logit contributions | |
aces | +0.158 |
amas | +0.156 |
robe | +0.156 |
coat | +0.155 |
bridge | +0.152 |
Neg logit contributions | |
coconut | -0.065 |
Enjoy | -0.056 |
grown | -0.047 |
temple | -0.046 |
ours | -0.045 |
Token | carry |
Token ID | 3283 |
Feature activation | +1.97e+00 |
Pos logit contributions | |
coconut | +0.040 |
oral | +0.038 |
grown | +0.038 |
ours | +0.037 |
Enjoy | +0.037 |
Neg logit contributions | |
coat | -0.043 |
brush | -0.041 |
aces | -0.039 |
robe | -0.037 |
drops | -0.037 |
Token | them |
Token ID | 606 |
Feature activation | +1.29e+00 |
Pos logit contributions | |
aces | +0.173 |
robe | +0.165 |
coat | +0.165 |
brush | +0.159 |
bridge | +0.157 |
Neg logit contributions | |
coconut | -0.122 |
Enjoy | -0.114 |
oral | -0.103 |
ours | -0.102 |
temple | -0.098 |
Token |
|
Token ID | 198 |
Feature activation | +3.26e-02 |
Pos logit contributions | |
coconut | +0.009 |
Enjoy | +0.007 |
temple | +0.007 |
exhibit | +0.007 |
volcano | +0.007 |
Neg logit contributions | |
aces | -0.013 |
robe | -0.013 |
coat | -0.013 |
amas | -0.013 |
brush | -0.012 |
Token |
|
Token ID | 198 |
Feature activation | +3.50e-02 |
Pos logit contributions | |
aces | +0.003 |
amas | +0.003 |
robe | +0.002 |
stand | +0.002 |
coat | +0.002 |
Neg logit contributions | |
coconut | -0.002 |
Enjoy | -0.002 |
whale | -0.002 |
volcano | -0.002 |
crab | -0.002 |
Token | " |
Token ID | 1 |
Feature activation | +1.72e+00 |
Pos logit contributions | |
aces | +0.003 |
coat | +0.003 |
robe | +0.003 |
amas | +0.003 |
brush | +0.003 |
Neg logit contributions | |
coconut | -0.003 |
Enjoy | -0.002 |
volcano | -0.002 |
ours | -0.002 |
temple | -0.002 |
Token | It |
Token ID | 1026 |
Feature activation | -4.40e-01 |
Pos logit contributions | |
aces | +0.147 |
coat | +0.144 |
robe | +0.143 |
amas | +0.141 |
brush | +0.139 |
Neg logit contributions | |
coconut | -0.111 |
Enjoy | -0.102 |
ours | -0.090 |
temple | -0.088 |
grown | -0.088 |
Token | 's |
Token ID | 338 |
Feature activation | +1.41e+00 |
Pos logit contributions | |
coconut | +0.042 |
Enjoy | +0.039 |
grown | +0.037 |
temple | +0.036 |
oral | +0.036 |
Neg logit contributions | |
aces | -0.023 |
robe | -0.023 |
coat | -0.023 |
amas | -0.023 |
brush | -0.021 |
Token | okay |
Token ID | 8788 |
Feature activation | +2.71e+00 |
Pos logit contributions | |
aces | +0.069 |
coat | +0.067 |
amas | +0.065 |
robe | +0.065 |
brush | +0.063 |
Neg logit contributions | |
coconut | -0.142 |
Enjoy | -0.133 |
ours | -0.124 |
temple | -0.124 |
volcano | -0.124 |
Token | , |
Token ID | 11 |
Feature activation | -5.93e-03 |
Pos logit contributions | |
coat | +0.196 |
aces | +0.192 |
amas | +0.191 |
brush | +0.185 |
robe | +0.185 |
Neg logit contributions | |
coconut | -0.210 |
Enjoy | -0.190 |
iw | -0.169 |
ours | -0.167 |
aquarium | -0.167 |
Token | my |
Token ID | 616 |
Feature activation | +1.63e-01 |
Pos logit contributions | |
coconut | +0.000 |
Enjoy | +0.000 |
shrimp | +0.000 |
ours | +0.000 |
crab | +0.000 |
Neg logit contributions | |
aces | -0.001 |
robe | -0.001 |
coat | -0.001 |
amas | -0.001 |
bridge | -0.001 |
Token | loves |
Token ID | 10408 |
Feature activation | -1.32e+00 |
Pos logit contributions | |
coat | +0.012 |
robe | +0.012 |
aces | +0.012 |
amas | +0.012 |
brush | +0.011 |
Neg logit contributions | |
coconut | -0.012 |
Enjoy | -0.011 |
Crab | -0.010 |
grown | -0.010 |
rill | -0.010 |
Token | . |
Token ID | 13 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
coconut | +0.085 |
Enjoy | +0.077 |
ours | +0.068 |
volcano | +0.066 |
temple | +0.066 |
Neg logit contributions | |
aces | -0.115 |
coat | -0.112 |
robe | -0.110 |
amas | -0.109 |
brush | -0.108 |
Token | I |
Token ID | 314 |
Feature activation | +4.43e-01 |
Pos logit contributions | |
coconut | +0.010 |
Enjoy | +0.008 |
temple | +0.008 |
oral | +0.008 |
grown | +0.008 |
Neg logit contributions | |
aces | -0.011 |
robe | -0.011 |
coat | -0.011 |
amas | -0.011 |
brush | -0.010 |
Token | work |
Token ID | 670 |
Feature activation | -5.38e-01 |
Pos logit contributions | |
coconut | +0.014 |
Enjoy | +0.013 |
temple | +0.011 |
oral | +0.011 |
volcano | +0.011 |
Neg logit contributions | |
aces | -0.017 |
coat | -0.017 |
robe | -0.017 |
amas | -0.017 |
bridge | -0.016 |
Token | together |
Token ID | 1978 |
Feature activation | -1.29e+00 |
Pos logit contributions | |
coconut | +0.025 |
Enjoy | +0.025 |
grown | +0.021 |
ours | +0.020 |
oral | +0.020 |
Neg logit contributions | |
coat | -0.052 |
aces | -0.052 |
robe | -0.051 |
amas | -0.051 |
bridge | -0.051 |
Token | . |
Token ID | 13 |
Feature activation | -1.09e-01 |
Pos logit contributions | |
coconut | +0.074 |
Enjoy | +0.067 |
ours | +0.058 |
temple | +0.058 |
grown | +0.058 |
Neg logit contributions | |
aces | -0.116 |
coat | -0.115 |
robe | -0.114 |
amas | -0.113 |
bridge | -0.111 |
Token | They |
Token ID | 1119 |
Feature activation | -1.29e+00 |
Pos logit contributions | |
coconut | +0.008 |
temple | +0.006 |
grown | +0.006 |
crab | +0.006 |
volcano | +0.006 |
Neg logit contributions | |
aces | -0.009 |
amas | -0.009 |
robe | -0.009 |
coat | -0.009 |
brush | -0.008 |
Token | can |
Token ID | 460 |
Feature activation | -2.02e-01 |
Pos logit contributions | |
coconut | +0.070 |
Enjoy | +0.066 |
temple | +0.056 |
ours | +0.055 |
volcano | +0.055 |
Neg logit contributions | |
aces | -0.121 |
coat | -0.119 |
robe | -0.119 |
amas | -0.119 |
bridge | -0.116 |
Token | help |
Token ID | 1037 |
Feature activation | +1.82e+00 |
Pos logit contributions | |
coconut | +0.014 |
Enjoy | +0.013 |
ours | +0.012 |
temple | +0.012 |
grown | +0.012 |
Neg logit contributions | |
aces | -0.016 |
coat | -0.015 |
robe | -0.015 |
amas | -0.015 |
bridge | -0.015 |
Token | each |
Token ID | 1123 |
Feature activation | -2.33e-01 |
Pos logit contributions | |
aces | +0.171 |
coat | +0.167 |
robe | +0.167 |
amas | +0.164 |
bridge | +0.162 |
Neg logit contributions | |
coconut | -0.101 |
Enjoy | -0.092 |
ours | -0.079 |
temple | -0.078 |
grown | -0.078 |
Token | other |
Token ID | 584 |
Feature activation | -4.84e+00 |
Pos logit contributions | |
coconut | +0.015 |
Enjoy | +0.014 |
ours | +0.013 |
iw | +0.012 |
grown | +0.012 |
Neg logit contributions | |
aces | -0.019 |
robe | -0.019 |
bridge | -0.019 |
brush | -0.018 |
coat | -0.018 |
Token | and |
Token ID | 290 |
Feature activation | -1.40e-01 |
Pos logit contributions | |
coconut | +0.239 |
Enjoy | +0.218 |
temple | +0.201 |
volcano | +0.197 |
whose | +0.195 |
Neg logit contributions | |
coat | -0.466 |
aces | -0.458 |
robe | -0.455 |
amas | -0.446 |
bridge | -0.438 |
Token | the |
Token ID | 262 |
Feature activation | +4.04e-02 |
Pos logit contributions | |
coconut | +0.009 |
Enjoy | +0.008 |
ours | +0.008 |
grown | +0.007 |
iw | +0.007 |
Neg logit contributions | |
aces | -0.012 |
coat | -0.011 |
robe | -0.011 |
brush | -0.011 |
amas | -0.011 |
Token | people |
Token ID | 661 |
Feature activation | +1.90e+00 |
Pos logit contributions | |
aces | +0.003 |
robe | +0.003 |
coat | +0.003 |
amas | +0.003 |
brush | +0.003 |
Neg logit contributions | |
coconut | -0.003 |
Enjoy | -0.003 |
ours | -0.002 |
grown | -0.002 |
iw | -0.002 |
Token | the |
Token ID | 262 |
Feature activation | +8.13e-02 |
Pos logit contributions | |
coconut | +0.087 |
Enjoy | +0.079 |
ours | +0.073 |
oral | +0.070 |
iw | +0.068 |
Neg logit contributions | |
coat | -0.115 |
aces | -0.114 |
robe | -0.114 |
amas | -0.109 |
brush | -0.109 |
Token | park |
Token ID | 3952 |
Feature activation | -2.16e+00 |
Pos logit contributions | |
coat | +0.007 |
robe | +0.007 |
amas | +0.007 |
brush | +0.007 |
aces | +0.007 |
Neg logit contributions | |
coconut | -0.005 |
Enjoy | -0.004 |
ours | -0.004 |
oral | -0.004 |
iw | -0.004 |
Token | and |
Token ID | 290 |
Feature activation | -1.10e-01 |
Pos logit contributions | |
coconut | +0.165 |
Enjoy | +0.152 |
ours | +0.139 |
oral | +0.138 |
iw | +0.135 |
Neg logit contributions | |
aces | -0.157 |
coat | -0.156 |
robe | -0.153 |
amas | -0.148 |
bridge | -0.145 |
Token | she |
Token ID | 673 |
Feature activation | +1.50e+00 |
Pos logit contributions | |
coconut | +0.007 |
Enjoy | +0.006 |
oral | +0.005 |
ours | +0.005 |
iw | +0.005 |
Neg logit contributions | |
aces | -0.010 |
coat | -0.010 |
robe | -0.010 |
amas | -0.009 |
bridge | -0.009 |
Token | heard |
Token ID | 2982 |
Feature activation | -1.71e+00 |
Pos logit contributions | |
aces | +0.160 |
coat | +0.157 |
robe | +0.156 |
amas | +0.155 |
brush | +0.153 |
Neg logit contributions | |
coconut | -0.064 |
Enjoy | -0.056 |
ours | -0.045 |
grown | -0.045 |
temple | -0.045 |
Token | a |
Token ID | 257 |
Feature activation | +5.28e-01 |
Pos logit contributions | |
coconut | +0.113 |
Enjoy | +0.104 |
iw | +0.094 |
ours | +0.094 |
oral | +0.091 |
Neg logit contributions | |
aces | -0.143 |
robe | -0.138 |
coat | -0.138 |
bridge | -0.136 |
brush | -0.135 |
Token | strange |
Token ID | 6283 |
Feature activation | -1.24e+00 |
Pos logit contributions | |
coat | +0.028 |
robe | +0.028 |
amas | +0.027 |
brush | +0.027 |
bridge | +0.026 |
Neg logit contributions | |
coconut | -0.047 |
Enjoy | -0.046 |
ours | -0.044 |
iw | -0.042 |
grown | -0.042 |
Token | language |
Token ID | 3303 |
Feature activation | -1.94e+00 |
Pos logit contributions | |
coconut | +0.125 |
Enjoy | +0.122 |
ours | +0.113 |
iw | +0.112 |
grown | +0.112 |
Neg logit contributions | |
robe | -0.056 |
coat | -0.055 |
aces | -0.054 |
amas | -0.054 |
brush | -0.051 |
Token | . |
Token ID | 13 |
Feature activation | -6.46e-02 |
Pos logit contributions | |
coconut | +0.116 |
Enjoy | +0.105 |
ours | +0.091 |
grown | +0.090 |
temple | +0.089 |
Neg logit contributions | |
aces | -0.175 |
coat | -0.171 |
robe | -0.170 |
amas | -0.168 |
brush | -0.165 |
Token | She |
Token ID | 1375 |
Feature activation | +3.25e-01 |
Pos logit contributions | |
coconut | +0.004 |
Enjoy | +0.004 |
ours | +0.004 |
grown | +0.004 |
temple | +0.004 |
Neg logit contributions | |
aces | -0.005 |
coat | -0.005 |
robe | -0.005 |
amas | -0.005 |
brush | -0.005 |
Token | was |
Token ID | 373 |
Feature activation | -4.01e-01 |
Pos logit contributions | |
aces | +0.023 |
coat | +0.023 |
robe | +0.022 |
brush | +0.022 |
bridge | +0.022 |
Neg logit contributions | |
Enjoy | -0.024 |
coconut | -0.024 |
iw | -0.022 |
ours | -0.021 |
oral | -0.020 |
Token | mom |
Token ID | 1995 |
Feature activation | +6.87e-01 |
Pos logit contributions | |
aces | +0.008 |
coat | +0.008 |
amas | +0.007 |
robe | +0.007 |
bridge | +0.007 |
Neg logit contributions | |
coconut | -0.006 |
Enjoy | -0.005 |
temple | -0.005 |
volcano | -0.005 |
oral | -0.005 |
Token | my |
Token ID | 1820 |
Feature activation | +9.41e-01 |
Pos logit contributions | |
aces | +0.066 |
robe | +0.063 |
amas | +0.062 |
bridge | +0.061 |
coat | +0.060 |
Neg logit contributions | |
coconut | -0.035 |
grown | -0.031 |
whose | -0.031 |
Enjoy | -0.029 |
ours | -0.028 |
Token | disappeared |
Token ID | 12120 |
Feature activation | +1.06e+00 |
Pos logit contributions | |
aces | +0.096 |
robe | +0.092 |
amas | +0.091 |
coat | +0.089 |
bridge | +0.088 |
Neg logit contributions | |
coconut | -0.044 |
grown | -0.036 |
whose | -0.036 |
whale | -0.034 |
crab | -0.034 |
Token | again |
Token ID | 757 |
Feature activation | -3.84e+00 |
Pos logit contributions | |
aces | +0.109 |
coat | +0.104 |
amas | +0.102 |
brush | +0.102 |
bridge | +0.102 |
Neg logit contributions | |
coconut | -0.046 |
grown | -0.044 |
temple | -0.038 |
oy | -0.037 |
volcano | -0.035 |
Token | . |
Token ID | 13 |
Feature activation | -5.14e-02 |
Pos logit contributions | |
coconut | +0.231 |
Enjoy | +0.209 |
ours | +0.183 |
grown | +0.180 |
temple | +0.180 |
Neg logit contributions | |
aces | -0.343 |
coat | -0.336 |
robe | -0.334 |
amas | -0.329 |
brush | -0.324 |
Token | The |
Token ID | 383 |
Feature activation | -5.86e-01 |
Pos logit contributions | |
coconut | +0.004 |
Enjoy | +0.003 |
whale | +0.003 |
iw | +0.003 |
crab | +0.003 |
Neg logit contributions | |
amas | -0.004 |
robe | -0.004 |
aces | -0.004 |
coat | -0.004 |
brush | -0.004 |
Token | girl |
Token ID | 2576 |
Feature activation | -1.03e-01 |
Pos logit contributions | |
coconut | +0.043 |
Enjoy | +0.041 |
ours | +0.036 |
iw | +0.036 |
grown | +0.035 |
Neg logit contributions | |
robe | -0.044 |
coat | -0.044 |
aces | -0.043 |
amas | -0.042 |
brush | -0.042 |
Token | felt |
Token ID | 2936 |
Feature activation | +1.71e+00 |
Pos logit contributions | |
coconut | +0.006 |
Enjoy | +0.005 |
ours | +0.004 |
grown | +0.004 |
oral | +0.004 |
Neg logit contributions | |
aces | -0.010 |
coat | -0.010 |
robe | -0.009 |
amas | -0.009 |
brush | -0.009 |
Token | sad |
Token ID | 6507 |
Feature activation | -5.77e-01 |
Pos logit contributions | |
aces | +0.152 |
coat | +0.151 |
amas | +0.150 |
robe | +0.145 |
bridge | +0.141 |
Neg logit contributions | |
coconut | -0.106 |
Enjoy | -0.100 |
grown | -0.087 |
shrimp | -0.086 |
oy | -0.085 |
Token | and |
Token ID | 290 |
Feature activation | -9.42e-02 |
Pos logit contributions | |
coconut | +0.033 |
Enjoy | +0.030 |
grown | +0.027 |
ours | +0.026 |
temple | +0.026 |
Neg logit contributions | |
aces | -0.053 |
coat | -0.052 |
robe | -0.051 |
amas | -0.051 |
bridge | -0.050 |
Token | wanted |
Token ID | 2227 |
Feature activation | -1.53e+00 |
Pos logit contributions | |
coconut | +0.006 |
Enjoy | +0.006 |
ours | +0.005 |
oral | +0.005 |
grown | +0.005 |
Neg logit contributions | |
aces | -0.008 |
coat | -0.008 |
robe | -0.008 |
amas | -0.007 |
brush | -0.007 |
Token | mom |
Token ID | 1995 |
Feature activation | +6.12e-01 |
Pos logit contributions | |
aces | +0.118 |
robe | +0.117 |
coat | +0.116 |
amas | +0.116 |
brush | +0.114 |
Neg logit contributions | |
coconut | -0.088 |
Enjoy | -0.083 |
ours | -0.073 |
iw | -0.073 |
grown | -0.073 |
Token | made |
Token ID | 925 |
Feature activation | +4.63e+00 |
Pos logit contributions | |
aces | +0.060 |
robe | +0.058 |
coat | +0.058 |
amas | +0.057 |
bridge | +0.056 |
Neg logit contributions | |
coconut | -0.032 |
Enjoy | -0.029 |
ours | -0.025 |
grown | -0.025 |
temple | -0.024 |
Token | him |
Token ID | 683 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
aces | +0.392 |
amas | +0.390 |
coat | +0.383 |
robe | +0.381 |
brush | +0.373 |
Neg logit contributions | |
coconut | -0.294 |
Enjoy | -0.262 |
oral | -0.236 |
iw | -0.236 |
temple | -0.235 |
Token | a |
Token ID | 257 |
Feature activation | +4.51e-01 |
Pos logit contributions | |
aces | +0.117 |
coat | +0.114 |
robe | +0.114 |
amas | +0.113 |
bridge | +0.110 |
Neg logit contributions | |
coconut | -0.081 |
Enjoy | -0.074 |
ours | -0.064 |
temple | -0.064 |
grown | -0.063 |
Token | healthy |
Token ID | 5448 |
Feature activation | +1.07e+00 |
Pos logit contributions | |
coat | +0.030 |
brush | +0.029 |
robe | +0.028 |
amas | +0.028 |
forehead | +0.025 |
Neg logit contributions | |
Enjoy | -0.033 |
coconut | -0.032 |
grown | -0.032 |
ours | -0.032 |
asks | -0.030 |
Token | dinner |
Token ID | 8073 |
Feature activation | -6.45e-01 |
Pos logit contributions | |
coat | +0.088 |
aces | +0.086 |
amas | +0.086 |
robe | +0.086 |
brush | +0.084 |
Neg logit contributions | |
coconut | -0.069 |
Enjoy | -0.064 |
ours | -0.059 |
grown | -0.058 |
oral | -0.057 |
Token | and |
Token ID | 290 |
Feature activation | -1.78e-01 |
Pos logit contributions | |
coconut | +0.024 |
Enjoy | +0.022 |
temple | +0.021 |
restaurant | +0.020 |
volcano | +0.019 |
Neg logit contributions | |
aces | -0.068 |
coat | -0.065 |
amas | -0.065 |
bridge | -0.064 |
robe | -0.064 |
Token | they |
Token ID | 484 |
Feature activation | -4.18e-01 |
Pos logit contributions | |
coconut | +0.011 |
Enjoy | +0.010 |
oral | +0.009 |
ours | +0.009 |
grown | +0.009 |
Neg logit contributions | |
robe | -0.015 |
coat | -0.015 |
aces | -0.015 |
brush | -0.014 |
amas | -0.014 |
Token | talked |
Token ID | 6619 |
Feature activation | -2.29e+00 |
Pos logit contributions | |
coconut | +0.029 |
Enjoy | +0.027 |
ours | +0.025 |
oral | +0.024 |
iw | +0.024 |
Neg logit contributions | |
aces | -0.032 |
robe | -0.032 |
brush | -0.031 |
coat | -0.031 |
bridge | -0.031 |
Token | about |
Token ID | 546 |
Feature activation | -1.15e+00 |
Pos logit contributions | |
coconut | +0.111 |
Enjoy | +0.101 |
grown | +0.084 |
volcano | +0.083 |
temple | +0.082 |
Neg logit contributions | |
aces | -0.229 |
coat | -0.226 |
robe | -0.223 |
amas | -0.221 |
bridge | -0.217 |
Token | how |
Token ID | 703 |
Feature activation | -3.05e-01 |
Pos logit contributions | |
coconut | +0.074 |
oral | +0.071 |
ours | +0.068 |
iw | +0.068 |
Enjoy | +0.066 |
Neg logit contributions | |
aces | -0.080 |
amas | -0.079 |
robe | -0.077 |
bridge | -0.076 |
brush | -0.076 |
Token | it |
Token ID | 340 |
Feature activation | -1.33e+00 |
Pos logit contributions | |
aces | +0.101 |
brush | +0.098 |
robe | +0.097 |
coat | +0.095 |
bridge | +0.093 |
Neg logit contributions | |
coconut | -0.067 |
Enjoy | -0.061 |
iw | -0.057 |
ours | -0.053 |
oral | -0.052 |
Token | was |
Token ID | 373 |
Feature activation | -4.85e-01 |
Pos logit contributions | |
coconut | +0.065 |
Enjoy | +0.057 |
ours | +0.048 |
temple | +0.048 |
grown | +0.048 |
Neg logit contributions | |
aces | -0.133 |
coat | -0.130 |
robe | -0.129 |
amas | -0.128 |
bridge | -0.126 |
Token | a |
Token ID | 257 |
Feature activation | +4.56e-01 |
Pos logit contributions | |
coconut | +0.032 |
Enjoy | +0.029 |
ours | +0.026 |
iw | +0.025 |
oral | +0.025 |
Neg logit contributions | |
aces | -0.041 |
coat | -0.040 |
robe | -0.040 |
amas | -0.039 |
brush | -0.039 |
Token | rainbow |
Token ID | 27223 |
Feature activation | -1.87e+00 |
Pos logit contributions | |
robe | +0.030 |
coat | +0.030 |
brush | +0.029 |
bridge | +0.027 |
cartridge | +0.026 |
Neg logit contributions | |
ours | -0.034 |
coconut | -0.033 |
Enjoy | -0.033 |
grown | -0.031 |
asks | -0.031 |
Token | and |
Token ID | 290 |
Feature activation | -1.85e-01 |
Pos logit contributions | |
coconut | +0.093 |
temple | +0.081 |
volcano | +0.077 |
whale | +0.075 |
Enjoy | +0.075 |
Neg logit contributions | |
amas | -0.183 |
aces | -0.182 |
robe | -0.174 |
coat | -0.171 |
bridge | -0.168 |
Token | it |
Token ID | 340 |
Feature activation | -1.32e+00 |
Pos logit contributions | |
coconut | +0.011 |
Enjoy | +0.010 |
oral | +0.010 |
iw | +0.009 |
volcano | +0.009 |
Neg logit contributions | |
aces | -0.016 |
brush | -0.016 |
robe | -0.015 |
coat | -0.015 |
bridge | -0.015 |
Token | reminded |
Token ID | 14516 |
Feature activation | +1.99e+00 |
Pos logit contributions | |
coconut | +0.066 |
Enjoy | +0.055 |
temple | +0.053 |
volcano | +0.051 |
ours | +0.051 |
Neg logit contributions | |
aces | -0.130 |
coat | -0.129 |
amas | -0.129 |
robe | -0.127 |
bridge | -0.123 |
Token | her |
Token ID | 607 |
Feature activation | +1.13e+00 |
Pos logit contributions | |
aces | +0.159 |
coat | +0.155 |
robe | +0.154 |
amas | +0.151 |
brush | +0.149 |
Neg logit contributions | |
coconut | -0.139 |
Enjoy | -0.127 |
ours | -0.114 |
temple | -0.112 |
volcano | -0.112 |
Token | to |
Token ID | 284 |
Feature activation | -6.37e-01 |
Pos logit contributions | |
aces | +0.107 |
robe | +0.104 |
coat | +0.104 |
brush | +0.103 |
amas | +0.100 |
Neg logit contributions | |
coconut | -0.063 |
Enjoy | -0.057 |
iw | -0.049 |
ours | -0.049 |
volcano | -0.048 |
Token | be |
Token ID | 307 |
Feature activation | +7.03e-01 |
Pos logit contributions | |
coconut | +0.044 |
Enjoy | +0.040 |
ours | +0.036 |
grown | +0.035 |
temple | +0.035 |
Neg logit contributions | |
aces | -0.051 |
coat | -0.050 |
robe | -0.050 |
amas | -0.049 |
brush | -0.048 |
Token | happy |
Token ID | 3772 |
Feature activation | +1.07e+00 |
Pos logit contributions | |
amas | +0.056 |
aces | +0.055 |
robe | +0.055 |
coat | +0.054 |
bridge | +0.052 |
Neg logit contributions | |
coconut | -0.049 |
Enjoy | -0.046 |
grown | -0.043 |
whose | -0.041 |
temple | -0.040 |
Token | faster |
Token ID | 5443 |
Feature activation | +3.13e+00 |
Pos logit contributions | |
amas | +0.164 |
aces | +0.157 |
bridge | +0.157 |
robe | +0.156 |
coat | +0.154 |
Neg logit contributions | |
coconut | -0.092 |
Enjoy | -0.081 |
exhibit | -0.076 |
temple | -0.075 |
whose | -0.074 |
Token | and |
Token ID | 290 |
Feature activation | -1.77e-01 |
Pos logit contributions | |
amas | +0.297 |
coat | +0.289 |
robe | +0.289 |
aces | +0.285 |
bridge | +0.283 |
Neg logit contributions | |
coconut | -0.167 |
Enjoy | -0.140 |
grown | -0.132 |
temple | -0.130 |
crab | -0.127 |
Token | louder |
Token ID | 27089 |
Feature activation | -2.16e-01 |
Pos logit contributions | |
coconut | +0.013 |
Enjoy | +0.012 |
asks | +0.012 |
stealing | +0.011 |
Crab | +0.011 |
Neg logit contributions | |
amas | -0.014 |
robe | -0.013 |
aces | -0.012 |
coat | -0.012 |
bridge | -0.012 |
Token | . |
Token ID | 13 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
coconut | +0.010 |
Enjoy | +0.009 |
temple | +0.008 |
grown | +0.008 |
whose | +0.008 |
Neg logit contributions | |
coat | -0.022 |
robe | -0.022 |
amas | -0.021 |
aces | -0.021 |
brush | -0.021 |
Token | He |
Token ID | 679 |
Feature activation | +2.97e-01 |
Pos logit contributions | |
coconut | +0.009 |
Enjoy | +0.008 |
ours | +0.007 |
temple | +0.007 |
grown | +0.007 |
Neg logit contributions | |
aces | -0.012 |
robe | -0.012 |
coat | -0.011 |
amas | -0.011 |
bridge | -0.011 |
Token | went |
Token ID | 1816 |
Feature activation | +4.09e-01 |
Pos logit contributions | |
aces | +0.025 |
coat | +0.024 |
robe | +0.024 |
amas | +0.024 |
brush | +0.023 |
Neg logit contributions | |
coconut | -0.019 |
Enjoy | -0.018 |
ours | -0.016 |
temple | -0.015 |
grown | -0.015 |
Token | towards |
Token ID | 3371 |
Feature activation | -9.18e-01 |
Pos logit contributions | |
aces | +0.030 |
coat | +0.029 |
robe | +0.029 |
amas | +0.028 |
brush | +0.028 |
Neg logit contributions | |
coconut | -0.031 |
Enjoy | -0.029 |
ours | -0.026 |
grown | -0.026 |
temple | -0.026 |
Token | L |
Token ID | 406 |
Feature activation | +3.26e-01 |
Pos logit contributions | |
coconut | +0.057 |
Enjoy | +0.054 |
oral | +0.049 |
ours | +0.047 |
iw | +0.046 |
Neg logit contributions | |
robe | -0.077 |
aces | -0.077 |
coat | -0.074 |
bridge | -0.072 |
brush | -0.071 |
Token | ila |
Token ID | 10102 |
Feature activation | -2.79e+00 |
Pos logit contributions | |
aces | +0.034 |
coat | +0.033 |
robe | +0.033 |
bridge | +0.032 |
amas | +0.032 |
Neg logit contributions | |
coconut | -0.016 |
Enjoy | -0.014 |
volcano | -0.011 |
iw | -0.011 |
grown | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | -1.46e-01 |
Pos logit contributions | |
coconut | +0.129 |
Enjoy | +0.113 |
exhibit | +0.112 |
whose | +0.110 |
temple | +0.105 |
Neg logit contributions | |
robe | -0.269 |
amas | -0.268 |
bridge | -0.266 |
aces | -0.264 |
coat | -0.262 |
Token | He |
Token ID | 679 |
Feature activation | +2.98e-01 |
Pos logit contributions | |
coconut | +0.009 |
Enjoy | +0.008 |
temple | +0.007 |
ours | +0.007 |
volcano | +0.007 |
Neg logit contributions | |
aces | -0.013 |
robe | -0.012 |
coat | -0.012 |
amas | -0.012 |
brush | -0.012 |
Token | . |
Token ID | 13 |
Feature activation | -8.44e-02 |
Pos logit contributions | |
coconut | +0.074 |
temple | +0.061 |
Enjoy | +0.061 |
volcano | +0.061 |
exhibit | +0.060 |
Neg logit contributions | |
coat | -0.201 |
aces | -0.201 |
robe | -0.196 |
amas | -0.194 |
bridge | -0.193 |
Token | The |
Token ID | 383 |
Feature activation | -6.12e-01 |
Pos logit contributions | |
coconut | +0.006 |
Enjoy | +0.005 |
grown | +0.005 |
iw | +0.005 |
temple | +0.005 |
Neg logit contributions | |
amas | -0.007 |
aces | -0.007 |
robe | -0.007 |
coat | -0.006 |
brush | -0.006 |
Token | screen |
Token ID | 3159 |
Feature activation | -5.72e-01 |
Pos logit contributions | |
coconut | +0.044 |
Enjoy | +0.041 |
volcano | +0.037 |
temple | +0.037 |
ours | +0.037 |
Neg logit contributions | |
aces | -0.048 |
coat | -0.047 |
robe | -0.045 |
amas | -0.045 |
brush | -0.044 |
Token | lit |
Token ID | 6578 |
Feature activation | +3.03e+00 |
Pos logit contributions | |
coconut | +0.030 |
volcano | +0.025 |
temple | +0.024 |
Enjoy | +0.024 |
ours | +0.023 |
Neg logit contributions | |
coat | -0.055 |
aces | -0.055 |
robe | -0.053 |
amas | -0.053 |
bridge | -0.050 |
Token | up |
Token ID | 510 |
Feature activation | +1.77e-01 |
Pos logit contributions | |
aces | +0.259 |
coat | +0.253 |
robe | +0.251 |
brush | +0.243 |
amas | +0.243 |
Neg logit contributions | |
coconut | -0.198 |
Enjoy | -0.176 |
volcano | -0.164 |
temple | -0.161 |
whose | -0.161 |
Token | , |
Token ID | 11 |
Feature activation | +5.98e-02 |
Pos logit contributions | |
aces | +0.015 |
coat | +0.015 |
robe | +0.015 |
amas | +0.014 |
brush | +0.014 |
Neg logit contributions | |
coconut | -0.011 |
Enjoy | -0.010 |
ours | -0.009 |
grown | -0.009 |
temple | -0.009 |
Token | and |
Token ID | 290 |
Feature activation | -1.11e-01 |
Pos logit contributions | |
aces | +0.006 |
robe | +0.006 |
coat | +0.006 |
amas | +0.006 |
brush | +0.006 |
Neg logit contributions | |
coconut | -0.003 |
Enjoy | -0.002 |
whose | -0.002 |
volcano | -0.002 |
temple | -0.002 |
Token | his |
Token ID | 465 |
Feature activation | -6.38e-01 |
Pos logit contributions | |
coconut | +0.005 |
Enjoy | +0.004 |
oral | +0.003 |
ours | +0.003 |
grown | +0.003 |
Neg logit contributions | |
aces | -0.012 |
coat | -0.012 |
robe | -0.012 |
amas | -0.011 |
brush | -0.011 |
Token | show |
Token ID | 905 |
Feature activation | -1.47e+00 |
Pos logit contributions | |
coconut | +0.042 |
Enjoy | +0.040 |
iw | +0.035 |
ours | +0.035 |
grown | +0.034 |
Neg logit contributions | |
aces | -0.053 |
robe | -0.053 |
coat | -0.052 |
amas | -0.051 |
brush | -0.050 |
Token | began |
Token ID | 2540 |
Feature activation | -7.17e-02 |
Pos logit contributions | |
coconut | +0.088 |
Enjoy | +0.079 |
ours | +0.070 |
temple | +0.070 |
grown | +0.069 |
Neg logit contributions | |
aces | -0.129 |
coat | -0.128 |
amas | -0.127 |
robe | -0.124 |
brush | -0.121 |
Token | to |
Token ID | 284 |
Feature activation | -5.79e-01 |
Pos logit contributions | |
coconut | +0.003 |
Enjoy | +0.003 |
ours | +0.002 |
grown | +0.002 |
temple | +0.002 |
Neg logit contributions | |
aces | -0.007 |
coat | -0.007 |
robe | -0.007 |
amas | -0.007 |
brush | -0.007 |
Token | he |
Token ID | 339 |
Feature activation | +6.34e-01 |
Pos logit contributions | |
aces | +0.197 |
coat | +0.193 |
robe | +0.192 |
brush | +0.190 |
amas | +0.186 |
Neg logit contributions | |
coconut | -0.138 |
Enjoy | -0.126 |
oral | -0.115 |
ours | -0.112 |
grown | -0.112 |
Token | knew |
Token ID | 2993 |
Feature activation | -2.63e+00 |
Pos logit contributions | |
aces | +0.052 |
coat | +0.050 |
robe | +0.050 |
brush | +0.049 |
bridge | +0.048 |
Neg logit contributions | |
coconut | -0.042 |
Enjoy | -0.039 |
grown | -0.035 |
ours | -0.035 |
oral | -0.035 |
Token | he |
Token ID | 339 |
Feature activation | +6.04e-01 |
Pos logit contributions | |
coconut | +0.175 |
Enjoy | +0.162 |
iw | +0.148 |
oral | +0.148 |
ours | +0.147 |
Neg logit contributions | |
aces | -0.215 |
robe | -0.206 |
brush | -0.205 |
coat | -0.202 |
bridge | -0.197 |
Token | shouldn |
Token ID | 6584 |
Feature activation | -3.32e+00 |
Pos logit contributions | |
aces | +0.052 |
coat | +0.051 |
robe | +0.050 |
amas | +0.049 |
brush | +0.049 |
Neg logit contributions | |
coconut | -0.038 |
Enjoy | -0.035 |
ours | -0.031 |
iw | -0.030 |
oral | -0.030 |
Token | 't |
Token ID | 470 |
Feature activation | +1.36e+00 |
Pos logit contributions | |
coconut | +0.264 |
Enjoy | +0.246 |
ours | +0.224 |
grown | +0.221 |
temple | +0.220 |
Neg logit contributions | |
aces | -0.233 |
coat | -0.226 |
robe | -0.225 |
amas | -0.221 |
brush | -0.216 |
Token | touch |
Token ID | 3638 |
Feature activation | -5.22e+00 |
Pos logit contributions | |
aces | +0.100 |
brush | +0.098 |
coat | +0.097 |
robe | +0.094 |
amas | +0.092 |
Neg logit contributions | |
coconut | -0.097 |
ours | -0.095 |
Enjoy | -0.093 |
iw | -0.088 |
grown | -0.088 |
Token | things |
Token ID | 1243 |
Feature activation | +8.71e-02 |
Pos logit contributions | |
coconut | +0.297 |
Enjoy | +0.283 |
oral | +0.257 |
ours | +0.238 |
iw | +0.232 |
Neg logit contributions | |
aces | -0.483 |
coat | -0.460 |
robe | -0.459 |
brush | -0.451 |
amas | -0.450 |
Token | that |
Token ID | 326 |
Feature activation | +8.42e-01 |
Pos logit contributions | |
amas | +0.009 |
coat | +0.009 |
aces | +0.009 |
robe | +0.009 |
bridge | +0.009 |
Neg logit contributions | |
coconut | -0.004 |
Enjoy | -0.003 |
shrimp | -0.003 |
whose | -0.003 |
oy | -0.003 |
Token | aren |
Token ID | 3588 |
Feature activation | +2.49e+00 |
Pos logit contributions | |
amas | +0.079 |
aces | +0.078 |
coat | +0.077 |
bridge | +0.076 |
robe | +0.075 |
Neg logit contributions | |
coconut | -0.047 |
Enjoy | -0.041 |
temple | -0.037 |
volcano | -0.037 |
oy | -0.036 |
Token | 't |
Token ID | 470 |
Feature activation | +1.34e+00 |
Pos logit contributions | |
coat | +0.211 |
robe | +0.201 |
aces | +0.199 |
amas | +0.194 |
brush | +0.190 |
Neg logit contributions | |
coconut | -0.160 |
Enjoy | -0.148 |
oral | -0.136 |
volcano | -0.132 |
ours | -0.131 |
Token | his |
Token ID | 465 |
Feature activation | -6.19e-01 |
Pos logit contributions | |
amas | +0.124 |
robe | +0.124 |
aces | +0.123 |
coat | +0.122 |
brush | +0.118 |
Neg logit contributions | |
coconut | -0.078 |
Enjoy | -0.068 |
oy | -0.061 |
shrimp | -0.061 |
temple | -0.060 |
Token | The |
Token ID | 464 |
Feature activation | +1.03e+00 |
Pos logit contributions | |
coat | +0.003 |
aces | +0.003 |
brush | +0.003 |
bridge | +0.003 |
amas | +0.003 |
Neg logit contributions | |
coconut | -0.004 |
Enjoy | -0.004 |
ours | -0.004 |
volcano | -0.004 |
oy | -0.004 |
Token | little |
Token ID | 1310 |
Feature activation | -9.30e-01 |
Pos logit contributions | |
robe | +0.071 |
coat | +0.069 |
brush | +0.068 |
aces | +0.066 |
amas | +0.063 |
Neg logit contributions | |
coconut | -0.081 |
Enjoy | -0.080 |
ours | -0.075 |
grown | -0.072 |
iw | -0.071 |
Token | girl |
Token ID | 2576 |
Feature activation | -1.87e-01 |
Pos logit contributions | |
coconut | +0.070 |
Enjoy | +0.066 |
ours | +0.059 |
oral | +0.058 |
volcano | +0.057 |
Neg logit contributions | |
aces | -0.069 |
coat | -0.069 |
robe | -0.068 |
brush | -0.066 |
amas | -0.064 |
Token | saw |
Token ID | 2497 |
Feature activation | -9.68e-02 |
Pos logit contributions | |
coconut | +0.012 |
Enjoy | +0.011 |
ours | +0.010 |
iw | +0.010 |
volcano | +0.010 |
Neg logit contributions | |
aces | -0.016 |
coat | -0.016 |
brush | -0.015 |
robe | -0.015 |
amas | -0.015 |
Token | a |
Token ID | 257 |
Feature activation | +4.56e-01 |
Pos logit contributions | |
coconut | +0.006 |
Enjoy | +0.006 |
oral | +0.005 |
ours | +0.005 |
iw | +0.005 |
Neg logit contributions | |
aces | -0.008 |
robe | -0.008 |
coat | -0.008 |
brush | -0.008 |
amas | -0.008 |
Token | pe |
Token ID | 613 |
Feature activation | -6.43e+00 |
Pos logit contributions | |
coat | +0.033 |
brush | +0.032 |
robe | +0.031 |
amas | +0.030 |
bridge | +0.028 |
Neg logit contributions | |
Enjoy | -0.034 |
coconut | -0.032 |
ours | -0.031 |
oral | -0.029 |
asks | -0.029 |
Token | b |
Token ID | 65 |
Feature activation | +3.26e+00 |
Pos logit contributions | |
coconut | +0.606 |
Enjoy | +0.580 |
ours | +0.525 |
volcano | +0.522 |
oral | +0.522 |
Neg logit contributions | |
aces | -0.356 |
coat | -0.340 |
robe | -0.330 |
brush | -0.327 |
bridge | -0.325 |
Token | ble |
Token ID | 903 |
Feature activation | -4.15e-01 |
Pos logit contributions | |
coat | +0.236 |
amas | +0.234 |
aces | +0.232 |
robe | +0.229 |
bridge | +0.217 |
Neg logit contributions | |
coconut | -0.251 |
Enjoy | -0.214 |
temple | -0.206 |
crab | -0.204 |
ours | -0.203 |
Token | . |
Token ID | 13 |
Feature activation | -1.33e-01 |
Pos logit contributions | |
coconut | +0.027 |
Enjoy | +0.025 |
ours | +0.022 |
grown | +0.022 |
temple | +0.022 |
Neg logit contributions | |
aces | -0.035 |
coat | -0.034 |
robe | -0.034 |
amas | -0.033 |
brush | -0.032 |
Token | She |
Token ID | 1375 |
Feature activation | +2.40e-01 |
Pos logit contributions | |
coconut | +0.011 |
Enjoy | +0.010 |
oral | +0.010 |
whale | +0.010 |
volcano | +0.010 |
Neg logit contributions | |
aces | -0.009 |
amas | -0.008 |
robe | -0.008 |
coat | -0.008 |
brush | -0.008 |
Token | decided |
Token ID | 3066 |
Feature activation | -2.70e+00 |
Pos logit contributions | |
aces | +0.018 |
brush | +0.018 |
coat | +0.018 |
amed | +0.017 |
bridge | +0.016 |
Neg logit contributions | |
Enjoy | -0.017 |
coconut | -0.016 |
ours | -0.015 |
volcano | -0.015 |
iw | -0.015 |
Token | M |
Token ID | 44 |
Feature activation | -1.50e+00 |
Pos logit contributions | |
aces | +0.004 |
coat | +0.003 |
robe | +0.003 |
amas | +0.003 |
bridge | +0.003 |
Neg logit contributions | |
coconut | -0.004 |
Enjoy | -0.003 |
ours | -0.003 |
temple | -0.003 |
grown | -0.003 |
Token | oral |
Token ID | 6864 |
Feature activation | -7.24e+00 |
Pos logit contributions | |
coconut | +0.050 |
Enjoy | +0.046 |
ours | +0.035 |
volcano | +0.033 |
whose | +0.033 |
Neg logit contributions | |
aces | -0.173 |
brush | -0.169 |
coat | -0.167 |
amas | -0.166 |
robe | -0.164 |
Token | of |
Token ID | 286 |
Feature activation | +4.57e-01 |
Pos logit contributions | |
coconut | +0.492 |
Enjoy | +0.402 |
ours | +0.393 |
iw | +0.383 |
crab | +0.376 |
Neg logit contributions | |
aces | -0.609 |
robe | -0.596 |
brush | -0.592 |
amas | -0.563 |
coat | -0.551 |
Token | the |
Token ID | 262 |
Feature activation | +2.42e-02 |
Pos logit contributions | |
robe | +0.037 |
aces | +0.036 |
amas | +0.035 |
brush | +0.035 |
stand | +0.035 |
Neg logit contributions | |
coconut | -0.029 |
Enjoy | -0.025 |
ours | -0.025 |
iw | -0.023 |
grown | -0.023 |
Token | story |
Token ID | 1621 |
Feature activation | -2.30e+00 |
Pos logit contributions | |
aces | +0.002 |
robe | +0.002 |
brush | +0.002 |
amas | +0.002 |
bridge | +0.002 |
Neg logit contributions | |
coconut | -0.002 |
Enjoy | -0.001 |
ours | -0.001 |
grown | -0.001 |
iw | -0.001 |
Token | : |
Token ID | 25 |
Feature activation | -5.63e+00 |
Pos logit contributions | |
coconut | +0.101 |
Enjoy | +0.075 |
iw | +0.072 |
shrimp | +0.069 |
grown | +0.068 |
Neg logit contributions | |
robe | -0.252 |
aces | -0.249 |
amas | -0.238 |
brush | -0.235 |
coat | -0.229 |
Token | It |
Token ID | 632 |
Feature activation | +4.52e-01 |
Pos logit contributions | |
coconut | +0.387 |
grown | +0.295 |
temple | +0.288 |
iw | +0.288 |
crab | +0.286 |
Neg logit contributions | |
amas | -0.445 |
aces | -0.440 |
robe | -0.437 |
brush | -0.420 |
cartridge | -0.415 |
Token | is |
Token ID | 318 |
Feature activation | -5.34e-01 |
Pos logit contributions | |
aces | +0.035 |
robe | +0.034 |
amas | +0.034 |
coat | +0.033 |
brush | +0.033 |
Neg logit contributions | |
coconut | -0.033 |
Enjoy | -0.029 |
grown | -0.027 |
iw | -0.026 |
temple | -0.026 |
Token | better |
Token ID | 1365 |
Feature activation | -2.43e+00 |
Pos logit contributions | |
coconut | +0.044 |
exhibit | +0.038 |
temple | +0.037 |
iw | +0.037 |
volcano | +0.037 |
Neg logit contributions | |
aces | -0.034 |
robe | -0.033 |
brush | -0.033 |
coat | -0.032 |
cartridge | -0.031 |
Token | to |
Token ID | 284 |
Feature activation | -6.18e-01 |
Pos logit contributions | |
coconut | +0.167 |
Enjoy | +0.141 |
ours | +0.132 |
volcano | +0.132 |
aquarium | +0.130 |
Neg logit contributions | |
aces | -0.190 |
robe | -0.183 |
coat | -0.182 |
brush | -0.178 |
cartridge | -0.174 |
Token | be |
Token ID | 307 |
Feature activation | +7.19e-01 |
Pos logit contributions | |
coconut | +0.045 |
Enjoy | +0.039 |
ours | +0.038 |
grown | +0.037 |
iw | +0.037 |
Neg logit contributions | |
aces | -0.047 |
coat | -0.046 |
robe | -0.046 |
brush | -0.046 |
amas | -0.045 |
Token | the |
Token ID | 262 |
Feature activation | +5.73e-02 |
Pos logit contributions | |
coconut | +0.046 |
oral | +0.044 |
iw | +0.042 |
ours | +0.041 |
Enjoy | +0.040 |
Neg logit contributions | |
coat | -0.037 |
amas | -0.035 |
robe | -0.035 |
stand | -0.035 |
stick | -0.031 |
Token | kitchen |
Token ID | 9592 |
Feature activation | -1.19e+00 |
Pos logit contributions | |
robe | +0.005 |
coat | +0.005 |
brush | +0.005 |
amas | +0.005 |
aces | +0.005 |
Neg logit contributions | |
coconut | -0.004 |
Enjoy | -0.003 |
ours | -0.003 |
grown | -0.003 |
oral | -0.003 |
Token | to |
Token ID | 284 |
Feature activation | -5.81e-01 |
Pos logit contributions | |
coconut | +0.061 |
Enjoy | +0.055 |
temple | +0.048 |
ours | +0.048 |
volcano | +0.047 |
Neg logit contributions | |
aces | -0.117 |
robe | -0.113 |
coat | -0.113 |
bridge | -0.112 |
amas | -0.111 |
Token | get |
Token ID | 651 |
Feature activation | -1.23e+00 |
Pos logit contributions | |
coconut | +0.046 |
oral | +0.041 |
Enjoy | +0.040 |
grown | +0.040 |
iw | +0.040 |
Neg logit contributions | |
robe | -0.038 |
coat | -0.037 |
amas | -0.037 |
brush | -0.037 |
bridge | -0.036 |
Token | some |
Token ID | 617 |
Feature activation | -4.90e-01 |
Pos logit contributions | |
coconut | +0.091 |
Enjoy | +0.080 |
oral | +0.077 |
exhibit | +0.075 |
iw | +0.074 |
Neg logit contributions | |
aces | -0.090 |
robe | -0.090 |
coat | -0.088 |
amas | -0.088 |
brush | -0.087 |
Token | yogurt |
Token ID | 32132 |
Feature activation | -6.63e+00 |
Pos logit contributions | |
coconut | +0.033 |
Enjoy | +0.031 |
ours | +0.027 |
grown | +0.027 |
temple | +0.027 |
Neg logit contributions | |
aces | -0.040 |
coat | -0.039 |
robe | -0.039 |
amas | -0.038 |
brush | -0.037 |
Token | . |
Token ID | 13 |
Feature activation | -8.02e-02 |
Pos logit contributions | |
coconut | +0.264 |
shrimp | +0.207 |
oy | +0.200 |
Enjoy | +0.194 |
ours | +0.191 |
Neg logit contributions | |
aces | -0.695 |
coat | -0.686 |
amas | -0.683 |
bridge | -0.682 |
robe | -0.669 |
Token | They |
Token ID | 1119 |
Feature activation | -1.25e+00 |
Pos logit contributions | |
coconut | +0.005 |
temple | +0.005 |
oy | +0.004 |
Enjoy | +0.004 |
grown | +0.004 |
Neg logit contributions | |
amas | -0.007 |
robe | -0.006 |
aces | -0.006 |
coat | -0.006 |
brush | -0.006 |
Token | each |
Token ID | 1123 |
Feature activation | -2.05e-01 |
Pos logit contributions | |
coconut | +0.088 |
Enjoy | +0.082 |
oral | +0.076 |
iw | +0.072 |
grown | +0.072 |
Neg logit contributions | |
aces | -0.096 |
coat | -0.096 |
robe | -0.094 |
brush | -0.092 |
amas | -0.090 |
Token | got |
Token ID | 1392 |
Feature activation | +7.27e-01 |
Pos logit contributions | |
coconut | +0.015 |
Enjoy | +0.013 |
temple | +0.012 |
oral | +0.012 |
volcano | +0.012 |
Neg logit contributions | |
coat | -0.016 |
aces | -0.016 |
robe | -0.016 |
brush | -0.015 |
bridge | -0.015 |
Token | a |
Token ID | 257 |
Feature activation | +5.22e-01 |
Pos logit contributions | |
aces | +0.055 |
robe | +0.053 |
coat | +0.053 |
amas | +0.053 |
brush | +0.052 |
Neg logit contributions | |
coconut | -0.053 |
Enjoy | -0.049 |
oral | -0.044 |
exhibit | -0.043 |
temple | -0.043 |
Token | pond |
Token ID | 16723 |
Feature activation | -2.66e+00 |
Pos logit contributions | |
coat | +0.039 |
robe | +0.038 |
brush | +0.037 |
bridge | +0.035 |
amas | +0.035 |
Neg logit contributions | |
coconut | -0.034 |
Enjoy | -0.034 |
ours | -0.032 |
grown | -0.030 |
oral | -0.030 |
Token | before |
Token ID | 878 |
Feature activation | +7.68e-01 |
Pos logit contributions | |
coconut | +0.170 |
Enjoy | +0.156 |
ours | +0.138 |
oral | +0.136 |
grown | +0.136 |
Neg logit contributions | |
aces | -0.226 |
coat | -0.222 |
robe | -0.220 |
amas | -0.216 |
brush | -0.213 |
Token | , |
Token ID | 11 |
Feature activation | +1.40e-01 |
Pos logit contributions | |
aces | +0.062 |
coat | +0.061 |
amas | +0.060 |
robe | +0.060 |
bridge | +0.058 |
Neg logit contributions | |
coconut | -0.051 |
Enjoy | -0.046 |
ours | -0.043 |
oral | -0.042 |
iw | -0.041 |
Token | so |
Token ID | 523 |
Feature activation | +2.79e+00 |
Pos logit contributions | |
aces | +0.011 |
coat | +0.011 |
robe | +0.011 |
amas | +0.010 |
brush | +0.010 |
Neg logit contributions | |
coconut | -0.010 |
Enjoy | -0.009 |
ours | -0.008 |
grown | -0.008 |
temple | -0.008 |
Token | he |
Token ID | 339 |
Feature activation | +5.82e-01 |
Pos logit contributions | |
aces | +0.233 |
robe | +0.228 |
coat | +0.227 |
amas | +0.223 |
brush | +0.223 |
Neg logit contributions | |
coconut | -0.175 |
Enjoy | -0.159 |
ours | -0.147 |
oral | -0.146 |
iw | -0.141 |
Token | pe |
Token ID | 613 |
Feature activation | -6.33e+00 |
Pos logit contributions | |
aces | +0.046 |
coat | +0.045 |
robe | +0.045 |
brush | +0.044 |
bridge | +0.044 |
Neg logit contributions | |
coconut | -0.039 |
Enjoy | -0.037 |
oral | -0.033 |
grown | -0.033 |
iw | -0.033 |
Token | ered |
Token ID | 1068 |
Feature activation | -2.39e+00 |
Pos logit contributions | |
coconut | +0.468 |
Enjoy | +0.454 |
volcano | +0.400 |
museum | +0.396 |
oral | +0.395 |
Neg logit contributions | |
aces | -0.480 |
bridge | -0.452 |
coat | -0.452 |
brush | -0.444 |
robe | -0.434 |
Token | in |
Token ID | 287 |
Feature activation | -1.33e+00 |
Pos logit contributions | |
coconut | +0.197 |
Enjoy | +0.190 |
oral | +0.180 |
ours | +0.173 |
iw | +0.171 |
Neg logit contributions | |
aces | -0.155 |
coat | -0.144 |
robe | -0.143 |
bridge | -0.142 |
brush | -0.142 |
Token | , |
Token ID | 11 |
Feature activation | +1.08e-01 |
Pos logit contributions | |
coconut | +0.086 |
Enjoy | +0.082 |
ours | +0.082 |
oral | +0.079 |
iw | +0.076 |
Neg logit contributions | |
aces | -0.099 |
brush | -0.096 |
coat | -0.095 |
robe | -0.094 |
bridge | -0.094 |
Token | blinking |
Token ID | 43196 |
Feature activation | -1.03e+00 |
Pos logit contributions | |
aces | +0.009 |
robe | +0.009 |
coat | +0.009 |
amas | +0.009 |
bridge | +0.009 |
Neg logit contributions | |
coconut | -0.007 |
Enjoy | -0.006 |
grown | -0.006 |
temple | -0.006 |
whose | -0.006 |
Token | in |
Token ID | 287 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
coconut | +0.058 |
Enjoy | +0.048 |
grown | +0.046 |
oy | +0.045 |
shrimp | +0.045 |
Neg logit contributions | |
coat | -0.096 |
aces | -0.095 |
amas | -0.094 |
robe | -0.094 |
brush | -0.090 |