Token | were |
Token ID | 547 |
Feature activation | +3.79e+00 |
Pos logit contributions | |
he | +0.049 |
roof | +0.034 |
field | +0.032 |
mountain | +0.032 |
bus | +0.032 |
Neg logit contributions | |
its | -0.129 |
ases | -0.122 |
asing | -0.115 |
olds | -0.114 |
ins | -0.113 |
Token | two |
Token ID | 734 |
Feature activation | +4.57e+00 |
Pos logit contributions | |
he | +0.072 |
roof | +0.052 |
mountain | +0.052 |
while | +0.052 |
field | +0.051 |
Neg logit contributions | |
its | -0.121 |
ases | -0.117 |
asing | -0.110 |
olds | -0.108 |
ins | -0.108 |
Token | beds |
Token ID | 20237 |
Feature activation | +4.05e+00 |
Pos logit contributions | |
he | +0.108 |
wide | +0.086 |
field | +0.086 |
while | +0.085 |
roof | +0.085 |
Neg logit contributions | |
its | -0.128 |
ases | -0.121 |
asing | -0.113 |
olds | -0.109 |
ins | -0.106 |
Token | in |
Token ID | 287 |
Feature activation | +5.43e+00 |
Pos logit contributions | |
he | +0.049 |
moon | +0.033 |
mountain | +0.031 |
valley | +0.031 |
sail | +0.031 |
Neg logit contributions | |
its | -0.146 |
ases | -0.143 |
ins | -0.137 |
asing | -0.137 |
olds | -0.136 |
Token | a |
Token ID | 257 |
Feature activation | +7.78e+00 |
Pos logit contributions | |
he | +0.058 |
mountain | +0.033 |
roof | +0.033 |
Pole | +0.032 |
river | +0.031 |
Neg logit contributions | |
its | -0.209 |
ases | -0.204 |
olds | -0.191 |
asing | -0.188 |
arts | -0.185 |
Token | bedroom |
Token ID | 14043 |
Feature activation | +9.05e+00 |
Pos logit contributions | |
he | +0.146 |
Pole | +0.112 |
Up | +0.109 |
sail | +0.097 |
ride | +0.097 |
Neg logit contributions | |
its | -0.214 |
ases | -0.196 |
ins | -0.192 |
ac | -0.185 |
olds | -0.184 |
Token | . |
Token ID | 13 |
Feature activation | +3.97e+00 |
Pos logit contributions | |
he | +0.150 |
Up | +0.123 |
sail | +0.113 |
Bun | +0.113 |
low | +0.111 |
Neg logit contributions | |
its | -0.273 |
ases | -0.261 |
ums | -0.257 |
olds | -0.255 |
ins | -0.252 |
Token | One |
Token ID | 1881 |
Feature activation | +2.32e+00 |
Pos logit contributions | |
he | +0.059 |
mountain | +0.048 |
raft | +0.047 |
bicycle | +0.047 |
hill | +0.046 |
Neg logit contributions | |
ases | -0.123 |
its | -0.120 |
olds | -0.118 |
ums | -0.109 |
preferences | -0.108 |
Token | was |
Token ID | 373 |
Feature activation | +3.12e+00 |
Pos logit contributions | |
he | +0.037 |
sail | +0.029 |
Pole | +0.029 |
valley | +0.029 |
river | +0.029 |
Neg logit contributions | |
its | -0.073 |
ases | -0.066 |
olds | -0.065 |
ac | -0.064 |
ins | -0.064 |
Token | a |
Token ID | 257 |
Feature activation | +5.28e+00 |
Pos logit contributions | |
he | +0.046 |
mountain | +0.041 |
hill | +0.040 |
river | +0.037 |
bicycle | +0.037 |
Neg logit contributions | |
its | -0.095 |
ases | -0.093 |
ins | -0.089 |
ums | -0.087 |
ac | -0.087 |
Token | big |
Token ID | 1263 |
Feature activation | +4.30e+00 |
Pos logit contributions | |
he | +0.099 |
Up | +0.085 |
fields | +0.078 |
down | +0.077 |
Suddenly | +0.075 |
Neg logit contributions | |
its | -0.132 |
ins | -0.123 |
ac | -0.117 |
cents | -0.117 |
ubby | -0.115 |
Token | Jimmy |
Token ID | 12963 |
Feature activation | -7.15e-01 |
Pos logit contributions | |
he | +0.074 |
mountain | +0.069 |
river | +0.069 |
roof | +0.067 |
valley | +0.064 |
Neg logit contributions | |
its | -0.190 |
ases | -0.182 |
ac | -0.174 |
asing | -0.168 |
ins | -0.166 |
Token | was |
Token ID | 373 |
Feature activation | +2.14e+00 |
Pos logit contributions | |
its | +0.026 |
ases | +0.025 |
olds | +0.023 |
asing | +0.023 |
ins | +0.023 |
Neg logit contributions | |
he | -0.010 |
roof | -0.008 |
mountain | -0.007 |
valley | -0.007 |
while | -0.007 |
Token | playing |
Token ID | 2712 |
Feature activation | +4.00e+00 |
Pos logit contributions | |
he | +0.032 |
mountain | +0.025 |
hill | +0.024 |
roof | +0.023 |
Pole | +0.023 |
Neg logit contributions | |
its | -0.073 |
ases | -0.072 |
asing | -0.067 |
olds | -0.066 |
ins | -0.066 |
Token | in |
Token ID | 287 |
Feature activation | +4.78e+00 |
Pos logit contributions | |
he | +0.062 |
Pole | +0.051 |
roof | +0.050 |
mountain | +0.049 |
moon | +0.046 |
Neg logit contributions | |
ases | -0.132 |
its | -0.131 |
asing | -0.120 |
ins | -0.118 |
olds | -0.115 |
Token | his |
Token ID | 465 |
Feature activation | +4.17e+00 |
Pos logit contributions | |
he | +0.054 |
Pole | +0.048 |
roof | +0.046 |
wagon | +0.040 |
mountain | +0.039 |
Neg logit contributions | |
its | -0.167 |
ases | -0.166 |
olds | -0.148 |
ins | -0.147 |
arts | -0.145 |
Token | backyard |
Token ID | 24296 |
Feature activation | +8.99e+00 |
Pos logit contributions | |
he | +0.071 |
while | +0.058 |
Pole | +0.052 |
they | +0.049 |
Lightning | +0.048 |
Neg logit contributions | |
ases | -0.124 |
its | -0.121 |
ums | -0.112 |
ins | -0.112 |
olds | -0.109 |
Token | . |
Token ID | 13 |
Feature activation | +3.89e+00 |
Pos logit contributions | |
he | +0.136 |
Pole | +0.124 |
valley | +0.110 |
sail | +0.108 |
Bun | +0.108 |
Neg logit contributions | |
its | -0.282 |
ases | -0.277 |
olds | -0.257 |
ums | -0.253 |
asing | -0.252 |
Token | He |
Token ID | 679 |
Feature activation | +3.32e+00 |
Pos logit contributions | |
he | +0.041 |
helicopter | +0.039 |
airplane | +0.038 |
raft | +0.038 |
jet | +0.038 |
Neg logit contributions | |
its | -0.136 |
ases | -0.125 |
olds | -0.122 |
ums | -0.121 |
ump | -0.116 |
Token | noticed |
Token ID | 6810 |
Feature activation | +4.79e+00 |
Pos logit contributions | |
he | +0.052 |
Pole | +0.042 |
roof | +0.042 |
valley | +0.042 |
wagon | +0.042 |
Neg logit contributions | |
its | -0.104 |
ases | -0.100 |
asing | -0.093 |
ins | -0.092 |
arts | -0.091 |
Token | something |
Token ID | 1223 |
Feature activation | +3.16e+00 |
Pos logit contributions | |
Pole | +0.059 |
roof | +0.058 |
valley | +0.058 |
hill | +0.058 |
Along | +0.057 |
Neg logit contributions | |
ases | -0.144 |
its | -0.143 |
ins | -0.137 |
olds | -0.136 |
asing | -0.135 |
Token | special |
Token ID | 2041 |
Feature activation | +1.85e+00 |
Pos logit contributions | |
he | +0.051 |
roof | +0.041 |
valley | +0.040 |
field | +0.040 |
mountain | +0.040 |
Neg logit contributions | |
its | -0.103 |
ases | -0.094 |
asing | -0.090 |
olds | -0.090 |
arts | -0.087 |
Token | Lily |
Token ID | 20037 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
he | +0.035 |
mountain | +0.021 |
field | +0.020 |
sail | +0.020 |
rocket | +0.019 |
Neg logit contributions | |
its | -0.105 |
ases | -0.101 |
asing | -0.096 |
olds | -0.095 |
ins | -0.093 |
Token | was |
Token ID | 373 |
Feature activation | +1.21e+00 |
Pos logit contributions | |
its | +0.050 |
ases | +0.048 |
asing | +0.045 |
olds | +0.045 |
ins | +0.044 |
Neg logit contributions | |
he | -0.016 |
roof | -0.010 |
mountain | -0.010 |
while | -0.010 |
field | -0.009 |
Token | playing |
Token ID | 2712 |
Feature activation | +4.07e+00 |
Pos logit contributions | |
he | +0.018 |
mountain | +0.013 |
hill | +0.012 |
field | +0.012 |
roof | +0.012 |
Neg logit contributions | |
its | -0.043 |
ases | -0.041 |
asing | -0.039 |
olds | -0.039 |
ins | -0.038 |
Token | in |
Token ID | 287 |
Feature activation | +4.72e+00 |
Pos logit contributions | |
he | +0.059 |
Pole | +0.046 |
roof | +0.044 |
mountain | +0.044 |
bus | +0.040 |
Neg logit contributions | |
ases | -0.141 |
its | -0.140 |
asing | -0.129 |
ins | -0.126 |
olds | -0.124 |
Token | her |
Token ID | 607 |
Feature activation | +4.10e+00 |
Pos logit contributions | |
he | +0.065 |
Pole | +0.048 |
roof | +0.044 |
mountain | +0.044 |
sail | +0.043 |
Neg logit contributions | |
its | -0.166 |
ases | -0.163 |
olds | -0.149 |
ins | -0.146 |
asing | -0.145 |
Token | backyard |
Token ID | 24296 |
Feature activation | +8.96e+00 |
Pos logit contributions | |
he | +0.072 |
while | +0.058 |
Pole | +0.051 |
Up | +0.051 |
Suddenly | +0.049 |
Neg logit contributions | |
ases | -0.121 |
its | -0.119 |
ins | -0.113 |
ums | -0.108 |
olds | -0.107 |
Token | . |
Token ID | 13 |
Feature activation | +3.18e+00 |
Pos logit contributions | |
he | +0.149 |
Pole | +0.134 |
sail | +0.119 |
mountain | +0.115 |
rocket | +0.115 |
Neg logit contributions | |
its | -0.274 |
ases | -0.270 |
olds | -0.252 |
asing | -0.244 |
ums | -0.244 |
Token | She |
Token ID | 1375 |
Feature activation | +2.76e+00 |
Pos logit contributions | |
he | +0.052 |
mountain | +0.042 |
boat | +0.041 |
rocket | +0.041 |
raft | +0.040 |
Neg logit contributions | |
its | -0.104 |
ases | -0.097 |
olds | -0.094 |
ums | -0.092 |
arts | -0.089 |
Token | saw |
Token ID | 2497 |
Feature activation | +3.73e+00 |
Pos logit contributions | |
he | +0.043 |
roof | +0.031 |
mountain | +0.030 |
wagon | +0.030 |
sail | +0.030 |
Neg logit contributions | |
its | -0.094 |
ases | -0.091 |
asing | -0.086 |
ins | -0.084 |
olds | -0.084 |
Token | a |
Token ID | 257 |
Feature activation | +5.41e+00 |
Pos logit contributions | |
he | +0.044 |
mountain | +0.033 |
hill | +0.031 |
Pole | +0.030 |
sail | +0.030 |
Neg logit contributions | |
its | -0.134 |
ases | -0.133 |
asing | -0.125 |
olds | -0.125 |
ins | -0.124 |
Token | pretty |
Token ID | 2495 |
Feature activation | +4.97e+00 |
Pos logit contributions | |
he | +0.098 |
Up | +0.079 |
Suddenly | +0.077 |
He | +0.074 |
The | +0.072 |
Neg logit contributions | |
its | -0.155 |
ins | -0.136 |
arts | -0.132 |
cents | -0.130 |
ases | -0.128 |
Token | . |
Token ID | 13 |
Feature activation | +3.61e+00 |
Pos logit contributions | |
he | +0.096 |
Tim | +0.085 |
bicycle | +0.079 |
While | +0.079 |
A | +0.079 |
Neg logit contributions | |
its | -0.220 |
ases | -0.213 |
ins | -0.211 |
ump | -0.206 |
asing | -0.204 |
Token | He |
Token ID | 679 |
Feature activation | +3.26e+00 |
Pos logit contributions | |
plane | +0.055 |
airplane | +0.054 |
jet | +0.054 |
helicopter | +0.053 |
bicycle | +0.053 |
Neg logit contributions | |
ases | -0.094 |
its | -0.093 |
olds | -0.093 |
ums | -0.088 |
preferences | -0.085 |
Token | stumbled |
Token ID | 24241 |
Feature activation | +8.28e-01 |
Pos logit contributions | |
he | +0.051 |
roof | +0.047 |
rocket | +0.047 |
fireplace | +0.045 |
helicopter | +0.044 |
Neg logit contributions | |
its | -0.094 |
ases | -0.088 |
ins | -0.086 |
asing | -0.083 |
arts | -0.083 |
Token | upon |
Token ID | 2402 |
Feature activation | +2.03e+00 |
Pos logit contributions | |
he | +0.013 |
rocket | +0.011 |
roof | +0.010 |
helicopter | +0.010 |
sail | +0.010 |
Neg logit contributions | |
its | -0.027 |
ases | -0.025 |
asing | -0.024 |
ins | -0.024 |
arts | -0.024 |
Token | a |
Token ID | 257 |
Feature activation | +5.14e+00 |
Pos logit contributions | |
Pole | +0.031 |
he | +0.027 |
wagon | +0.027 |
fireplace | +0.026 |
bicycle | +0.026 |
Neg logit contributions | |
its | -0.060 |
ases | -0.055 |
ps | -0.054 |
consequences | -0.054 |
arts | -0.052 |
Token | garden |
Token ID | 11376 |
Feature activation | +8.94e+00 |
Pos logit contributions | |
he | +0.091 |
Up | +0.074 |
He | +0.071 |
Under | +0.068 |
Tim | +0.067 |
Neg logit contributions | |
its | -0.153 |
cents | -0.134 |
ins | -0.133 |
ases | -0.129 |
arts | -0.128 |
Token | full |
Token ID | 1336 |
Feature activation | +3.52e+00 |
Pos logit contributions | |
he | +0.130 |
Bun | +0.113 |
Peter | +0.103 |
Pole | +0.102 |
rocket | +0.100 |
Neg logit contributions | |
its | -0.281 |
ases | -0.268 |
olds | -0.253 |
ums | -0.252 |
ump | -0.250 |
Token | of |
Token ID | 286 |
Feature activation | +1.49e+00 |
Pos logit contributions | |
he | +0.089 |
Suddenly | +0.074 |
rocket | +0.073 |
ride | +0.072 |
Pole | +0.072 |
Neg logit contributions | |
its | -0.081 |
ases | -0.077 |
olds | -0.072 |
ins | -0.070 |
asing | -0.070 |
Token | brightly |
Token ID | 35254 |
Feature activation | +2.64e-01 |
Pos logit contributions | |
he | +0.028 |
roof | +0.020 |
while | +0.020 |
ride | +0.020 |
rocket | +0.020 |
Neg logit contributions | |
its | -0.048 |
ases | -0.045 |
ins | -0.043 |
asing | -0.042 |
olds | -0.042 |
Token | coloured |
Token ID | 34746 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
he | +0.005 |
roof | +0.004 |
field | +0.004 |
while | +0.004 |
mountain | +0.004 |
Neg logit contributions | |
its | -0.008 |
ases | -0.008 |
asing | -0.008 |
olds | -0.007 |
ins | -0.007 |
Token | vegetables |
Token ID | 13701 |
Feature activation | +4.95e+00 |
Pos logit contributions | |
he | +0.117 |
Suddenly | +0.088 |
ride | +0.086 |
He | +0.086 |
rocket | +0.085 |
Neg logit contributions | |
its | -0.171 |
ases | -0.162 |
ins | -0.156 |
olds | -0.150 |
asing | -0.149 |
Token | was |
Token ID | 373 |
Feature activation | +2.92e+00 |
Pos logit contributions | |
its | +0.042 |
ases | +0.041 |
asing | +0.038 |
olds | +0.038 |
ins | +0.038 |
Neg logit contributions | |
he | -0.015 |
roof | -0.010 |
while | -0.009 |
field | -0.009 |
moon | -0.009 |
Token | walking |
Token ID | 6155 |
Feature activation | +4.93e+00 |
Pos logit contributions | |
he | +0.042 |
mountain | +0.039 |
hill | +0.037 |
Pole | +0.036 |
roof | +0.034 |
Neg logit contributions | |
its | -0.095 |
ases | -0.095 |
asing | -0.088 |
olds | -0.087 |
ums | -0.085 |
Token | in |
Token ID | 287 |
Feature activation | +5.49e+00 |
Pos logit contributions | |
Pole | +0.082 |
he | +0.077 |
chim | +0.073 |
fireplace | +0.072 |
roof | +0.071 |
Neg logit contributions | |
ases | -0.148 |
ums | -0.125 |
asing | -0.125 |
its | -0.125 |
ac | -0.124 |
Token | a |
Token ID | 257 |
Feature activation | +7.91e+00 |
Pos logit contributions | |
Pole | +0.080 |
he | +0.075 |
roof | +0.070 |
helicopter | +0.065 |
bicycle | +0.064 |
Neg logit contributions | |
ases | -0.165 |
its | -0.162 |
olds | -0.145 |
humor | -0.140 |
ump | -0.139 |
Token | me |
Token ID | 502 |
Feature activation | +8.34e-01 |
Pos logit contributions | |
he | +0.149 |
Pole | +0.120 |
Up | +0.119 |
while | +0.118 |
Under | +0.110 |
Neg logit contributions | |
its | -0.195 |
ases | -0.177 |
ins | -0.169 |
olds | -0.167 |
ateful | -0.162 |
Token | adow |
Token ID | 4584 |
Feature activation | +8.94e+00 |
Pos logit contributions | |
he | +0.019 |
field | +0.014 |
roof | +0.014 |
while | +0.014 |
wide | +0.014 |
Neg logit contributions | |
its | -0.024 |
ases | -0.023 |
asing | -0.021 |
olds | -0.021 |
ins | -0.021 |
Token | when |
Token ID | 618 |
Feature activation | +6.05e+00 |
Pos logit contributions | |
wagon | +0.148 |
Pole | +0.142 |
rocket | +0.139 |
bicycle | +0.137 |
Rusty | +0.137 |
Neg logit contributions | |
its | -0.232 |
ums | -0.224 |
ump | -0.224 |
olds | -0.215 |
ases | -0.212 |
Token | he |
Token ID | 339 |
Feature activation | +2.54e+00 |
Pos logit contributions | |
Pole | +0.049 |
fireplace | +0.049 |
airplane | +0.046 |
stable | +0.046 |
ranch | +0.045 |
Neg logit contributions | |
its | -0.185 |
ases | -0.175 |
arts | -0.168 |
olds | -0.168 |
ateful | -0.166 |
Token | saw |
Token ID | 2497 |
Feature activation | +4.09e+00 |
Pos logit contributions | |
wagon | +0.040 |
fireplace | +0.038 |
roof | +0.038 |
rocket | +0.038 |
he | +0.038 |
Neg logit contributions | |
its | -0.073 |
ases | -0.070 |
ins | -0.068 |
asing | -0.068 |
olds | -0.067 |
Token | something |
Token ID | 1223 |
Feature activation | +2.88e+00 |
Pos logit contributions | |
he | +0.049 |
hill | +0.045 |
ride | +0.044 |
Pole | +0.043 |
while | +0.043 |
Neg logit contributions | |
its | -0.132 |
ases | -0.131 |
olds | -0.124 |
asing | -0.120 |
ins | -0.119 |
Token | strange |
Token ID | 6283 |
Feature activation | +3.08e+00 |
Pos logit contributions | |
he | +0.048 |
roof | +0.039 |
mountain | +0.038 |
field | +0.037 |
valley | +0.037 |
Neg logit contributions | |
its | -0.092 |
ases | -0.084 |
olds | -0.081 |
asing | -0.081 |
ins | -0.079 |
Token | was |
Token ID | 373 |
Feature activation | +1.49e+00 |
Pos logit contributions | |
he | +0.031 |
wagon | +0.025 |
roof | +0.025 |
mountain | +0.024 |
raft | +0.024 |
Neg logit contributions | |
its | -0.053 |
ases | -0.049 |
arts | -0.046 |
ins | -0.046 |
asing | -0.045 |
Token | walking |
Token ID | 6155 |
Feature activation | +4.59e+00 |
Pos logit contributions | |
he | +0.019 |
mountain | +0.012 |
hill | +0.012 |
roof | +0.012 |
field | +0.011 |
Neg logit contributions | |
its | -0.056 |
ases | -0.054 |
asing | -0.051 |
olds | -0.051 |
ins | -0.050 |
Token | in |
Token ID | 287 |
Feature activation | +4.62e+00 |
Pos logit contributions | |
he | +0.078 |
Pole | +0.075 |
roof | +0.069 |
mountain | +0.068 |
raft | +0.067 |
Neg logit contributions | |
ases | -0.129 |
its | -0.114 |
ac | -0.114 |
ump | -0.113 |
ins | -0.113 |
Token | a |
Token ID | 257 |
Feature activation | +7.30e+00 |
Pos logit contributions | |
he | +0.067 |
Pole | +0.055 |
raft | +0.047 |
roof | +0.047 |
bicycle | +0.046 |
Neg logit contributions | |
ases | -0.151 |
its | -0.149 |
olds | -0.135 |
arts | -0.130 |
ateful | -0.129 |
Token | me |
Token ID | 502 |
Feature activation | +2.62e-01 |
Pos logit contributions | |
he | +0.143 |
Up | +0.114 |
He | +0.104 |
while | +0.103 |
But | +0.102 |
Neg logit contributions | |
its | -0.180 |
ins | -0.164 |
ases | -0.162 |
ateful | -0.157 |
olds | -0.153 |
Token | adow |
Token ID | 4584 |
Feature activation | +8.92e+00 |
Pos logit contributions | |
he | +0.006 |
field | +0.004 |
roof | +0.004 |
while | +0.004 |
wide | +0.004 |
Neg logit contributions | |
its | -0.008 |
ases | -0.007 |
asing | -0.007 |
olds | -0.007 |
ins | -0.007 |
Token | one |
Token ID | 530 |
Feature activation | +1.74e+00 |
Pos logit contributions | |
wagon | +0.144 |
he | +0.138 |
bicycle | +0.134 |
raft | +0.133 |
rocket | +0.133 |
Neg logit contributions | |
its | -0.232 |
ump | -0.232 |
ums | -0.226 |
ases | -0.219 |
olds | -0.214 |
Token | day |
Token ID | 1110 |
Feature activation | +4.62e+00 |
Pos logit contributions | |
he | +0.031 |
Pole | +0.026 |
roof | +0.026 |
sail | +0.026 |
Bun | +0.025 |
Neg logit contributions | |
its | -0.048 |
ases | -0.047 |
asing | -0.046 |
ac | -0.045 |
ins | -0.044 |
Token | when |
Token ID | 618 |
Feature activation | +5.44e+00 |
Pos logit contributions | |
rocket | +0.069 |
helicopter | +0.069 |
wagon | +0.068 |
bicycle | +0.068 |
mountain | +0.067 |
Neg logit contributions | |
olds | -0.130 |
its | -0.130 |
asing | -0.127 |
ases | -0.127 |
ums | -0.124 |
Token | she |
Token ID | 673 |
Feature activation | +2.97e+00 |
Pos logit contributions | |
wagon | +0.071 |
river | +0.071 |
Pole | +0.070 |
raft | +0.066 |
stable | +0.066 |
Neg logit contributions | |
its | -0.145 |
ases | -0.142 |
olds | -0.141 |
arts | -0.138 |
ateful | -0.133 |
Token | saw |
Token ID | 2497 |
Feature activation | +3.50e+00 |
Pos logit contributions | |
he | +0.051 |
wagon | +0.051 |
rocket | +0.048 |
roof | +0.047 |
bicycle | +0.046 |
Neg logit contributions | |
its | -0.081 |
ases | -0.078 |
ins | -0.076 |
olds | -0.073 |
asing | -0.073 |
Token | was |
Token ID | 373 |
Feature activation | +2.40e+00 |
Pos logit contributions | |
he | +0.069 |
wagon | +0.065 |
roof | +0.064 |
raft | +0.063 |
port | +0.062 |
Neg logit contributions | |
ases | -0.088 |
its | -0.082 |
arts | -0.078 |
olds | -0.077 |
asing | -0.076 |
Token | walking |
Token ID | 6155 |
Feature activation | +5.22e+00 |
Pos logit contributions | |
he | +0.039 |
mountain | +0.029 |
roof | +0.028 |
hill | +0.028 |
moon | +0.028 |
Neg logit contributions | |
its | -0.080 |
ases | -0.077 |
asing | -0.072 |
ins | -0.072 |
ump | -0.071 |
Token | through |
Token ID | 832 |
Feature activation | +6.95e+00 |
Pos logit contributions | |
Pole | +0.098 |
he | +0.092 |
roof | +0.085 |
fireplace | +0.081 |
mountain | +0.080 |
Neg logit contributions | |
ases | -0.151 |
ac | -0.128 |
its | -0.126 |
asing | -0.124 |
ump | -0.122 |
Token | a |
Token ID | 257 |
Feature activation | +7.24e+00 |
Pos logit contributions | |
he | +0.105 |
Pole | +0.095 |
roof | +0.079 |
bus | +0.077 |
raft | +0.077 |
Neg logit contributions | |
ases | -0.209 |
its | -0.208 |
arts | -0.188 |
ins | -0.185 |
olds | -0.183 |
Token | me |
Token ID | 502 |
Feature activation | +4.42e-01 |
Pos logit contributions | |
he | +0.135 |
Up | +0.104 |
while | +0.102 |
Tim | +0.098 |
But | +0.096 |
Neg logit contributions | |
its | -0.185 |
ins | -0.170 |
ases | -0.167 |
arts | -0.162 |
unts | -0.156 |
Token | adow |
Token ID | 4584 |
Feature activation | +8.89e+00 |
Pos logit contributions | |
he | +0.010 |
field | +0.007 |
roof | +0.007 |
while | +0.007 |
wide | +0.007 |
Neg logit contributions | |
its | -0.013 |
ases | -0.012 |
asing | -0.012 |
olds | -0.011 |
ins | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | +4.41e+00 |
Pos logit contributions | |
wagon | +0.156 |
he | +0.155 |
Pole | +0.154 |
bicycle | +0.150 |
Rusty | +0.146 |
Neg logit contributions | |
its | -0.226 |
ump | -0.219 |
ases | -0.217 |
ums | -0.213 |
olds | -0.205 |
Token | It |
Token ID | 632 |
Feature activation | +2.92e+00 |
Pos logit contributions | |
boat | +0.080 |
raft | +0.077 |
plane | +0.075 |
yacht | +0.074 |
airplane | +0.073 |
Neg logit contributions | |
its | -0.101 |
ases | -0.098 |
ums | -0.097 |
olds | -0.096 |
ump | -0.095 |
Token | was |
Token ID | 373 |
Feature activation | +1.56e+00 |
Pos logit contributions | |
he | +0.048 |
wagon | +0.047 |
rocket | +0.046 |
sail | +0.046 |
roof | +0.046 |
Neg logit contributions | |
its | -0.077 |
ases | -0.068 |
ins | -0.065 |
arts | -0.064 |
asing | -0.064 |
Token | so |
Token ID | 523 |
Feature activation | +2.53e+00 |
Pos logit contributions | |
he | +0.023 |
Pole | +0.018 |
rocket | +0.018 |
mountain | +0.018 |
hill | +0.017 |
Neg logit contributions | |
its | -0.050 |
ases | -0.047 |
arts | -0.044 |
ump | -0.044 |
ums | -0.044 |
Token | sunny |
Token ID | 27737 |
Feature activation | +4.09e+00 |
Pos logit contributions | |
he | +0.042 |
wagon | +0.035 |
rocket | +0.034 |
roof | +0.034 |
car | +0.034 |
Neg logit contributions | |
its | -0.080 |
ases | -0.071 |
olds | -0.069 |
ins | -0.069 |
arts | -0.066 |
Token | a |
Token ID | 257 |
Feature activation | +4.61e+00 |
Pos logit contributions | |
its | +0.112 |
ases | +0.107 |
asing | +0.099 |
olds | +0.098 |
ins | +0.096 |
Neg logit contributions | |
he | -0.083 |
roof | -0.065 |
while | -0.064 |
field | -0.063 |
mountain | -0.062 |
Token | big |
Token ID | 1263 |
Feature activation | +3.10e+00 |
Pos logit contributions | |
he | +0.084 |
Up | +0.064 |
He | +0.063 |
The | +0.063 |
Under | +0.062 |
Neg logit contributions | |
its | -0.133 |
ases | -0.116 |
humor | -0.115 |
cents | -0.114 |
ins | -0.112 |
Token | tent |
Token ID | 11105 |
Feature activation | +6.64e+00 |
Pos logit contributions | |
he | +0.050 |
Up | +0.038 |
Suddenly | +0.037 |
While | +0.037 |
Under | +0.036 |
Neg logit contributions | |
its | -0.098 |
ases | -0.095 |
humor | -0.086 |
olds | -0.086 |
asing | -0.085 |
Token | in |
Token ID | 287 |
Feature activation | +4.52e+00 |
Pos logit contributions | |
he | +0.093 |
Bun | +0.084 |
While | +0.083 |
Lightning | +0.081 |
A | +0.077 |
Neg logit contributions | |
its | -0.207 |
ases | -0.202 |
arts | -0.184 |
ps | -0.182 |
ins | -0.180 |
Token | the |
Token ID | 262 |
Feature activation | +4.09e+00 |
Pos logit contributions | |
he | +0.051 |
Pole | +0.039 |
roof | +0.035 |
bicycle | +0.034 |
while | +0.034 |
Neg logit contributions | |
its | -0.168 |
ases | -0.166 |
olds | -0.148 |
arts | -0.144 |
ins | -0.143 |
Token | backyard |
Token ID | 24296 |
Feature activation | +8.87e+00 |
Pos logit contributions | |
he | +0.073 |
while | +0.056 |
they | +0.056 |
Tim | +0.054 |
Bun | +0.053 |
Neg logit contributions | |
ases | -0.120 |
its | -0.118 |
olds | -0.109 |
ac | -0.107 |
ump | -0.104 |
Token | . |
Token ID | 13 |
Feature activation | +3.61e+00 |
Pos logit contributions | |
he | +0.125 |
Bun | +0.117 |
Under | +0.110 |
wagon | +0.110 |
bicycle | +0.109 |
Neg logit contributions | |
its | -0.286 |
ases | -0.271 |
olds | -0.259 |
ins | -0.254 |
asing | -0.246 |
Token | They |
Token ID | 1119 |
Feature activation | +2.35e+00 |
Pos logit contributions | |
mountain | +0.048 |
he | +0.047 |
plane | +0.047 |
airplane | +0.046 |
rocket | +0.045 |
Neg logit contributions | |
ases | -0.107 |
olds | -0.106 |
its | -0.102 |
ums | -0.098 |
asing | -0.095 |
Token | want |
Token ID | 765 |
Feature activation | +1.27e-01 |
Pos logit contributions | |
he | +0.042 |
Pole | +0.032 |
A | +0.032 |
while | +0.032 |
rocket | +0.031 |
Neg logit contributions | |
its | -0.073 |
ases | -0.070 |
olds | -0.063 |
asing | -0.063 |
arts | -0.063 |
Token | to |
Token ID | 284 |
Feature activation | +7.38e-03 |
Pos logit contributions | |
he | +0.002 |
valley | +0.001 |
Suddenly | +0.001 |
While | +0.001 |
Pole | +0.001 |
Neg logit contributions | |
ases | -0.004 |
its | -0.004 |
olds | -0.004 |
arts | -0.004 |
ins | -0.004 |
Token | go |
Token ID | 467 |
Feature activation | -4.53e-01 |
Pos logit contributions | |
he | +0.000 |
Pole | +0.000 |
Suddenly | +0.000 |
Lightning | +0.000 |
while | +0.000 |
Neg logit contributions | |
its | -0.000 |
ases | -0.000 |
arts | -0.000 |
olds | -0.000 |
humor | -0.000 |
Token | outside |
Token ID | 2354 |
Feature activation | +6.59e+00 |
Pos logit contributions | |
he | +0.039 |
Pole | +0.031 |
roof | +0.031 |
mountain | +0.030 |
chim | +0.028 |
Neg logit contributions | |
ases | -0.080 |
its | -0.076 |
ins | -0.075 |
asing | -0.071 |
arts | -0.070 |
Token | on |
Token ID | 319 |
Feature activation | +3.32e+00 |
Pos logit contributions | |
Pole | +0.133 |
he | +0.120 |
roof | +0.112 |
bus | +0.108 |
bicycle | +0.107 |
Neg logit contributions | |
ases | -0.169 |
ins | -0.142 |
asing | -0.140 |
its | -0.139 |
arts | -0.138 |
Token | their |
Token ID | 511 |
Feature activation | +3.71e+00 |
Pos logit contributions | |
he | +0.050 |
Pole | +0.044 |
Suddenly | +0.043 |
While | +0.040 |
while | +0.040 |
Neg logit contributions | |
ases | -0.097 |
its | -0.095 |
arts | -0.083 |
ps | -0.081 |
els | -0.079 |
Token | big |
Token ID | 1263 |
Feature activation | +4.50e+00 |
Pos logit contributions | |
he | +0.063 |
He | +0.045 |
while | +0.044 |
A | +0.043 |
Suddenly | +0.043 |
Neg logit contributions | |
its | -0.111 |
ases | -0.109 |
olds | -0.102 |
arts | -0.101 |
preferences | -0.100 |
Token | green |
Token ID | 4077 |
Feature activation | +5.65e+00 |
Pos logit contributions | |
he | +0.077 |
Suddenly | +0.056 |
while | +0.056 |
A | +0.054 |
While | +0.051 |
Neg logit contributions | |
ases | -0.130 |
its | -0.121 |
arts | -0.115 |
ums | -0.112 |
humor | -0.112 |
Token | lawn |
Token ID | 21133 |
Feature activation | +8.83e+00 |
Pos logit contributions | |
he | +0.106 |
low | +0.073 |
they | +0.072 |
George | +0.071 |
Peter | +0.070 |
Neg logit contributions | |
ases | -0.157 |
its | -0.150 |
arts | -0.141 |
ums | -0.139 |
ins | -0.139 |
Token | . |
Token ID | 13 |
Feature activation | +3.46e+00 |
Pos logit contributions | |
he | +0.172 |
Pole | +0.154 |
Bun | +0.149 |
low | +0.144 |
dark | +0.142 |
Neg logit contributions | |
ases | -0.223 |
its | -0.202 |
olds | -0.199 |
asing | -0.197 |
ums | -0.194 |
Token | One |
Token ID | 1881 |
Feature activation | +2.12e+00 |
Pos logit contributions | |
he | +0.064 |
mountain | +0.062 |
shield | +0.061 |
helicopter | +0.060 |
down | +0.059 |
Neg logit contributions | |
preferences | -0.065 |
olds | -0.061 |
ases | -0.061 |
humor | -0.060 |
its | -0.059 |
Token | day |
Token ID | 1110 |
Feature activation | +5.17e+00 |
Pos logit contributions | |
he | +0.035 |
Pole | +0.028 |
roof | +0.027 |
valley | +0.027 |
mountain | +0.027 |
Neg logit contributions | |
its | -0.065 |
ases | -0.061 |
asing | -0.059 |
ac | -0.058 |
olds | -0.058 |
Token | , |
Token ID | 11 |
Feature activation | +6.00e+00 |
Pos logit contributions | |
he | +0.061 |
river | +0.052 |
roof | +0.051 |
mountain | +0.051 |
moon | +0.051 |
Neg logit contributions | |
its | -0.178 |
ases | -0.176 |
olds | -0.167 |
asing | -0.164 |
ins | -0.162 |
Token | the |
Token ID | 262 |
Feature activation | +2.63e+00 |
Pos logit contributions | |
mountain | +0.089 |
river | +0.087 |
Pole | +0.086 |
valley | +0.085 |
chair | +0.085 |
Neg logit contributions | |
ateful | -0.135 |
ases | -0.134 |
its | -0.130 |
olds | -0.122 |
ins | -0.122 |
Token | an |
Token ID | 281 |
Feature activation | +2.09e+00 |
Pos logit contributions | |
he | +0.061 |
mountain | +0.046 |
Pole | +0.046 |
roof | +0.044 |
hill | +0.043 |
Neg logit contributions | |
its | -0.156 |
ases | -0.154 |
asing | -0.141 |
ins | -0.141 |
ump | -0.139 |
Token | attractive |
Token ID | 10966 |
Feature activation | +4.54e+00 |
Pos logit contributions | |
he | +0.044 |
while | +0.033 |
roof | +0.032 |
field | +0.032 |
they | +0.032 |
Neg logit contributions | |
its | -0.062 |
ases | -0.060 |
asing | -0.056 |
olds | -0.055 |
ins | -0.055 |
Token | crib |
Token ID | 48083 |
Feature activation | +6.27e+00 |
Pos logit contributions | |
he | +0.090 |
Suddenly | +0.070 |
Up | +0.067 |
He | +0.067 |
low | +0.066 |
Neg logit contributions | |
its | -0.115 |
ums | -0.109 |
ases | -0.107 |
ins | -0.106 |
arts | -0.106 |
Token | in |
Token ID | 287 |
Feature activation | +5.18e+00 |
Pos logit contributions | |
he | +0.110 |
Pole | +0.095 |
valley | +0.089 |
Up | +0.088 |
low | +0.088 |
Neg logit contributions | |
ases | -0.171 |
its | -0.162 |
asing | -0.157 |
arts | -0.155 |
olds | -0.153 |
Token | a |
Token ID | 257 |
Feature activation | +7.16e+00 |
Pos logit contributions | |
he | +0.057 |
while | +0.034 |
roof | +0.033 |
Pole | +0.032 |
wagon | +0.031 |
Neg logit contributions | |
its | -0.202 |
ases | -0.197 |
asing | -0.181 |
olds | -0.181 |
ins | -0.177 |
Token | nursery |
Token ID | 41326 |
Feature activation | +8.81e+00 |
Pos logit contributions | |
he | +0.143 |
Up | +0.104 |
Pole | +0.102 |
Tim | +0.101 |
while | +0.099 |
Neg logit contributions | |
its | -0.198 |
ases | -0.179 |
ins | -0.173 |
arts | -0.166 |
ac | -0.165 |
Token | . |
Token ID | 13 |
Feature activation | +4.11e+00 |
Pos logit contributions | |
he | +0.160 |
Up | +0.126 |
Pole | +0.122 |
Peter | +0.120 |
raft | +0.120 |
Neg logit contributions | |
ases | -0.244 |
olds | -0.231 |
its | -0.230 |
ums | -0.229 |
asing | -0.226 |
Token | Everyday |
Token ID | 48154 |
Feature activation | +4.38e+00 |
Pos logit contributions | |
he | +0.063 |
boat | +0.054 |
raft | +0.053 |
plane | +0.050 |
valley | +0.049 |
Neg logit contributions | |
ases | -0.126 |
its | -0.125 |
olds | -0.117 |
asing | -0.112 |
ump | -0.112 |
Token | a |
Token ID | 257 |
Feature activation | +4.27e+00 |
Pos logit contributions | |
Pole | +0.060 |
roof | +0.059 |
river | +0.055 |
driveway | +0.054 |
sail | +0.054 |
Neg logit contributions | |
ases | -0.116 |
its | -0.111 |
asing | -0.106 |
ac | -0.105 |
olds | -0.101 |
Token | little |
Token ID | 1310 |
Feature activation | +4.20e+00 |
Pos logit contributions | |
he | +0.078 |
He | +0.067 |
fields | +0.066 |
Up | +0.065 |
Under | +0.064 |
Neg logit contributions | |
its | -0.099 |
cents | -0.095 |
apology | -0.094 |
humor | -0.092 |
ins | -0.092 |
Token | girl |
Token ID | 2576 |
Feature activation | +2.63e+00 |
Pos logit contributions | |
he | +0.083 |
Under | +0.079 |
roof | +0.076 |
Pole | +0.075 |
fields | +0.074 |
Neg logit contributions | |
asing | -0.084 |
ases | -0.083 |
its | -0.081 |
irk | -0.080 |
humor | -0.078 |
Token | my |
Token ID | 1820 |
Feature activation | +7.01e-01 |
Pos logit contributions | |
its | +0.082 |
ases | +0.079 |
asing | +0.075 |
olds | +0.074 |
ins | +0.073 |
Neg logit contributions | |
he | -0.029 |
roof | -0.019 |
while | -0.018 |
field | -0.018 |
mountain | -0.018 |
Token | was |
Token ID | 373 |
Feature activation | +8.98e-01 |
Pos logit contributions | |
he | +0.009 |
roof | +0.006 |
mountain | +0.006 |
valley | +0.005 |
while | +0.005 |
Neg logit contributions | |
its | -0.026 |
ases | -0.026 |
asing | -0.024 |
olds | -0.024 |
ins | -0.024 |
Token | playing |
Token ID | 2712 |
Feature activation | +3.59e+00 |
Pos logit contributions | |
he | +0.015 |
mountain | +0.012 |
moon | +0.011 |
field | +0.011 |
hill | +0.011 |
Neg logit contributions | |
its | -0.029 |
ases | -0.028 |
asing | -0.027 |
olds | -0.026 |
ins | -0.026 |
Token | in |
Token ID | 287 |
Feature activation | +4.47e+00 |
Pos logit contributions | |
he | +0.055 |
roof | +0.042 |
Pole | +0.042 |
mountain | +0.041 |
moon | +0.040 |
Neg logit contributions | |
ases | -0.122 |
its | -0.121 |
asing | -0.111 |
ins | -0.108 |
olds | -0.107 |
Token | his |
Token ID | 465 |
Feature activation | +3.95e+00 |
Pos logit contributions | |
he | +0.050 |
Pole | +0.039 |
roof | +0.037 |
while | +0.035 |
Suddenly | +0.032 |
Neg logit contributions | |
its | -0.160 |
ases | -0.160 |
olds | -0.143 |
ins | -0.140 |
arts | -0.139 |
Token | backyard |
Token ID | 24296 |
Feature activation | +8.79e+00 |
Pos logit contributions | |
he | +0.070 |
while | +0.060 |
Up | +0.050 |
she | +0.050 |
Pole | +0.049 |
Neg logit contributions | |
ases | -0.112 |
its | -0.107 |
ums | -0.099 |
ins | -0.099 |
olds | -0.099 |
Token | . |
Token ID | 13 |
Feature activation | +3.47e+00 |
Pos logit contributions | |
he | +0.140 |
Pole | +0.124 |
valley | +0.112 |
moon | +0.111 |
Bun | +0.110 |
Neg logit contributions | |
its | -0.271 |
ases | -0.269 |
olds | -0.249 |
ums | -0.241 |
asing | -0.240 |
Token | He |
Token ID | 679 |
Feature activation | +3.27e+00 |
Pos logit contributions | |
plane | +0.046 |
valley | +0.042 |
airplane | +0.042 |
helicopter | +0.041 |
he | +0.041 |
Neg logit contributions | |
its | -0.107 |
ases | -0.101 |
ums | -0.098 |
olds | -0.098 |
preferences | -0.091 |
Token | loved |
Token ID | 6151 |
Feature activation | +2.18e+00 |
Pos logit contributions | |
he | +0.048 |
roof | +0.038 |
wagon | +0.037 |
mountain | +0.037 |
valley | +0.036 |
Neg logit contributions | |
its | -0.108 |
ases | -0.104 |
asing | -0.099 |
ins | -0.097 |
olds | -0.096 |
Token | to |
Token ID | 284 |
Feature activation | +1.53e+00 |
Pos logit contributions | |
he | +0.026 |
Pole | +0.021 |
Suddenly | +0.021 |
valley | +0.021 |
A | +0.020 |
Neg logit contributions | |
ases | -0.073 |
its | -0.072 |
ins | -0.070 |
olds | -0.067 |
asing | -0.066 |
Token | run |
Token ID | 1057 |
Feature activation | +7.16e-01 |
Pos logit contributions | |
he | +0.026 |
Pole | +0.021 |
Lightning | +0.019 |
cabin | +0.019 |
rocket | +0.019 |
Neg logit contributions | |
its | -0.048 |
ases | -0.047 |
olds | -0.041 |
asing | -0.041 |
humor | -0.040 |
Token | Lily |
Token ID | 20037 |
Feature activation | -8.31e-01 |
Pos logit contributions | |
he | +0.033 |
mountain | +0.019 |
field | +0.019 |
sail | +0.018 |
rocket | +0.018 |
Neg logit contributions | |
its | -0.106 |
ases | -0.102 |
asing | -0.097 |
olds | -0.097 |
ins | -0.094 |
Token | was |
Token ID | 373 |
Feature activation | +1.05e+00 |
Pos logit contributions | |
its | +0.032 |
ases | +0.031 |
asing | +0.029 |
olds | +0.029 |
ins | +0.029 |
Neg logit contributions | |
he | -0.010 |
roof | -0.006 |
mountain | -0.006 |
while | -0.006 |
rocket | -0.006 |
Token | playing |
Token ID | 2712 |
Feature activation | +3.71e+00 |
Pos logit contributions | |
he | +0.017 |
mountain | +0.012 |
moon | +0.011 |
field | +0.011 |
roof | +0.011 |
Neg logit contributions | |
its | -0.036 |
ases | -0.035 |
asing | -0.033 |
olds | -0.033 |
ins | -0.032 |
Token | in |
Token ID | 287 |
Feature activation | +4.59e+00 |
Pos logit contributions | |
he | +0.054 |
roof | +0.038 |
mountain | +0.038 |
Pole | +0.038 |
moon | +0.036 |
Neg logit contributions | |
its | -0.131 |
ases | -0.129 |
asing | -0.120 |
ins | -0.117 |
olds | -0.116 |
Token | her |
Token ID | 607 |
Feature activation | +4.25e+00 |
Pos logit contributions | |
he | +0.066 |
roof | +0.044 |
Pole | +0.044 |
mountain | +0.043 |
while | +0.043 |
Neg logit contributions | |
its | -0.162 |
ases | -0.158 |
olds | -0.145 |
asing | -0.144 |
ins | -0.142 |
Token | backyard |
Token ID | 24296 |
Feature activation | +8.78e+00 |
Pos logit contributions | |
he | +0.073 |
while | +0.058 |
Pole | +0.050 |
Up | +0.048 |
Suddenly | +0.048 |
Neg logit contributions | |
ases | -0.129 |
its | -0.125 |
ins | -0.117 |
olds | -0.117 |
ums | -0.116 |
Token | . |
Token ID | 13 |
Feature activation | +3.60e+00 |
Pos logit contributions | |
he | +0.144 |
Pole | +0.118 |
sail | +0.109 |
rocket | +0.106 |
mountain | +0.105 |
Neg logit contributions | |
its | -0.282 |
ases | -0.275 |
olds | -0.257 |
asing | -0.251 |
ums | -0.249 |
Token | She |
Token ID | 1375 |
Feature activation | +3.38e+00 |
Pos logit contributions | |
he | +0.061 |
mountain | +0.052 |
rocket | +0.051 |
raft | +0.050 |
boat | +0.050 |
Neg logit contributions | |
its | -0.110 |
ases | -0.103 |
olds | -0.101 |
ums | -0.100 |
arts | -0.096 |
Token | saw |
Token ID | 2497 |
Feature activation | +3.65e+00 |
Pos logit contributions | |
he | +0.050 |
roof | +0.036 |
mountain | +0.035 |
sail | +0.035 |
wagon | +0.035 |
Neg logit contributions | |
its | -0.117 |
ases | -0.112 |
asing | -0.107 |
ins | -0.105 |
olds | -0.104 |
Token | a |
Token ID | 257 |
Feature activation | +5.34e+00 |
Pos logit contributions | |
he | +0.042 |
mountain | +0.031 |
hill | +0.029 |
sail | +0.029 |
ride | +0.028 |
Neg logit contributions | |
its | -0.133 |
ases | -0.132 |
olds | -0.124 |
asing | -0.124 |
ins | -0.123 |
Token | spark |
Token ID | 9009 |
Feature activation | +3.67e-01 |
Pos logit contributions | |
he | +0.095 |
Up | +0.076 |
Suddenly | +0.076 |
He | +0.073 |
The | +0.072 |
Neg logit contributions | |
its | -0.153 |
ins | -0.133 |
arts | -0.131 |
cents | -0.130 |
ases | -0.128 |
Token | a |
Token ID | 257 |
Feature activation | +5.92e+00 |
Pos logit contributions | |
he | +0.044 |
mountain | +0.033 |
hill | +0.032 |
sail | +0.031 |
ride | +0.030 |
Neg logit contributions | |
its | -0.131 |
ases | -0.129 |
olds | -0.123 |
asing | -0.121 |
ins | -0.120 |
Token | huge |
Token ID | 3236 |
Feature activation | +5.02e+00 |
Pos logit contributions | |
he | +0.107 |
Up | +0.089 |
Suddenly | +0.089 |
He | +0.085 |
While | +0.084 |
Neg logit contributions | |
its | -0.164 |
ins | -0.143 |
cents | -0.139 |
arts | -0.137 |
ateful | -0.135 |
Token | rose |
Token ID | 8278 |
Feature activation | +5.27e+00 |
Pos logit contributions | |
he | +0.081 |
Suddenly | +0.066 |
While | +0.063 |
Up | +0.062 |
He | +0.058 |
Neg logit contributions | |
its | -0.150 |
ases | -0.140 |
arts | -0.134 |
ums | -0.128 |
cents | -0.128 |
Token | in |
Token ID | 287 |
Feature activation | +3.93e+00 |
Pos logit contributions | |
he | +0.093 |
Peter | +0.075 |
rocket | +0.069 |
He | +0.068 |
George | +0.067 |
Neg logit contributions | |
its | -0.166 |
ases | -0.152 |
asing | -0.146 |
olds | -0.144 |
arts | -0.143 |
Token | a |
Token ID | 257 |
Feature activation | +6.55e+00 |
Pos logit contributions | |
he | +0.048 |
Pole | +0.034 |
rocket | +0.030 |
roof | +0.029 |
sail | +0.028 |
Neg logit contributions | |
its | -0.147 |
ases | -0.140 |
olds | -0.131 |
arts | -0.127 |
ins | -0.126 |
Token | garden |
Token ID | 11376 |
Feature activation | +8.78e+00 |
Pos logit contributions | |
he | +0.114 |
He | +0.084 |
Peter | +0.084 |
Up | +0.083 |
Tim | +0.080 |
Neg logit contributions | |
its | -0.197 |
ases | -0.171 |
arts | -0.167 |
cents | -0.167 |
olds | -0.166 |
Token | . |
Token ID | 13 |
Feature activation | +3.27e+00 |
Pos logit contributions | |
he | +0.147 |
Bun | +0.115 |
rocket | +0.113 |
Peter | +0.112 |
bicycle | +0.110 |
Neg logit contributions | |
its | -0.282 |
ases | -0.267 |
olds | -0.254 |
asing | -0.245 |
ums | -0.243 |
Token | The |
Token ID | 383 |
Feature activation | +3.00e+00 |
Pos logit contributions | |
he | +0.047 |
rocket | +0.033 |
sail | +0.032 |
mountain | +0.032 |
helicopter | +0.032 |
Neg logit contributions | |
its | -0.118 |
ases | -0.114 |
olds | -0.107 |
asing | -0.105 |
ins | -0.103 |
Token | rose |
Token ID | 8278 |
Feature activation | +3.93e+00 |
Pos logit contributions | |
he | +0.037 |
field | +0.022 |
roof | +0.022 |
while | +0.022 |
wide | +0.021 |
Neg logit contributions | |
its | -0.117 |
ases | -0.114 |
asing | -0.109 |
olds | -0.107 |
ins | -0.105 |
Token | was |
Token ID | 373 |
Feature activation | +7.88e-01 |
Pos logit contributions | |
he | +0.062 |
sail | +0.050 |
rocket | +0.049 |
Peter | +0.047 |
fireplace | +0.046 |
Neg logit contributions | |
its | -0.130 |
ases | -0.115 |
asing | -0.115 |
olds | -0.114 |
arts | -0.112 |
Token | red |
Token ID | 2266 |
Feature activation | +1.04e+00 |
Pos logit contributions | |
he | +0.012 |
mountain | +0.009 |
rocket | +0.008 |
hill | +0.008 |
ride | +0.008 |
Neg logit contributions | |
its | -0.027 |
ases | -0.026 |
asing | -0.024 |
olds | -0.024 |
arts | -0.024 |
Token | daughter |
Token ID | 4957 |
Feature activation | +4.14e+00 |
Pos logit contributions | |
he | +0.093 |
Up | +0.081 |
Under | +0.074 |
Suddenly | +0.074 |
down | +0.072 |
Neg logit contributions | |
ins | -0.136 |
its | -0.132 |
ums | -0.123 |
akers | -0.123 |
ases | -0.123 |
Token | were |
Token ID | 547 |
Feature activation | +1.34e+00 |
Pos logit contributions | |
he | +0.073 |
mountain | +0.056 |
hill | +0.053 |
low | +0.053 |
wide | +0.052 |
Neg logit contributions | |
ases | -0.127 |
its | -0.125 |
olds | -0.118 |
ins | -0.117 |
asing | -0.117 |
Token | sitting |
Token ID | 5586 |
Feature activation | +2.68e+00 |
Pos logit contributions | |
he | +0.023 |
mountain | +0.016 |
roof | +0.016 |
while | +0.016 |
field | +0.016 |
Neg logit contributions | |
its | -0.045 |
ases | -0.043 |
asing | -0.041 |
olds | -0.040 |
ins | -0.040 |
Token | in |
Token ID | 287 |
Feature activation | +3.84e+00 |
Pos logit contributions | |
he | +0.041 |
roof | +0.028 |
mountain | +0.027 |
Pole | +0.027 |
sail | +0.026 |
Neg logit contributions | |
ases | -0.093 |
its | -0.093 |
asing | -0.088 |
ins | -0.088 |
olds | -0.084 |
Token | their |
Token ID | 511 |
Feature activation | +3.86e+00 |
Pos logit contributions | |
he | +0.048 |
Pole | +0.032 |
mountain | +0.031 |
river | +0.031 |
roof | +0.031 |
Neg logit contributions | |
its | -0.139 |
ases | -0.136 |
olds | -0.126 |
ins | -0.122 |
arts | -0.122 |
Token | backyard |
Token ID | 24296 |
Feature activation | +8.78e+00 |
Pos logit contributions | |
he | +0.072 |
while | +0.055 |
He | +0.049 |
low | +0.047 |
Up | +0.046 |
Neg logit contributions | |
its | -0.110 |
ases | -0.110 |
olds | -0.105 |
ins | -0.104 |
ums | -0.103 |
Token | . |
Token ID | 13 |
Feature activation | +3.41e+00 |
Pos logit contributions | |
he | +0.149 |
Pole | +0.135 |
Bun | +0.126 |
bicycle | +0.124 |
mountain | +0.123 |
Neg logit contributions | |
its | -0.243 |
ases | -0.241 |
olds | -0.234 |
ums | -0.230 |
ins | -0.223 |
Token | The |
Token ID | 383 |
Feature activation | +1.93e+00 |
Pos logit contributions | |
he | +0.062 |
helmet | +0.053 |
airplane | +0.053 |
shield | +0.052 |
rocket | +0.051 |
Neg logit contributions | |
its | -0.090 |
ases | -0.088 |
ums | -0.086 |
olds | -0.085 |
preferences | -0.078 |
Token | daughter |
Token ID | 4957 |
Feature activation | +2.67e+00 |
Pos logit contributions | |
he | +0.044 |
He | +0.033 |
Under | +0.032 |
Suddenly | +0.032 |
Up | +0.031 |
Neg logit contributions | |
its | -0.052 |
ases | -0.046 |
olds | -0.045 |
ins | -0.045 |
arts | -0.044 |
Token | said |
Token ID | 531 |
Feature activation | +1.40e+00 |
Pos logit contributions | |
he | +0.046 |
sail | +0.041 |
mountain | +0.040 |
west | +0.040 |
valley | +0.040 |
Neg logit contributions | |
its | -0.073 |
ins | -0.069 |
asing | -0.068 |
olds | -0.068 |
ases | -0.067 |
Token | to |
Token ID | 284 |
Feature activation | +3.58e-01 |
Pos logit contributions | |
field | +0.018 |
he | +0.017 |
hill | +0.017 |
mountain | +0.017 |
river | +0.017 |
Neg logit contributions | |
its | -0.043 |
olds | -0.042 |
ases | -0.042 |
arts | -0.041 |
ums | -0.040 |
Token | Tommy |
Token ID | 19919 |
Feature activation | -8.44e-01 |
Pos logit contributions | |
mountain | +0.081 |
roof | +0.075 |
stable | +0.074 |
airplane | +0.073 |
bus | +0.073 |
Neg logit contributions | |
its | -0.195 |
ases | -0.185 |
ump | -0.169 |
ac | -0.168 |
olds | -0.167 |
Token | was |
Token ID | 373 |
Feature activation | +2.34e+00 |
Pos logit contributions | |
its | +0.030 |
ases | +0.029 |
olds | +0.027 |
asing | +0.026 |
ins | +0.026 |
Neg logit contributions | |
he | -0.012 |
roof | -0.009 |
mountain | -0.009 |
valley | -0.009 |
rocket | -0.008 |
Token | playing |
Token ID | 2712 |
Feature activation | +4.53e+00 |
Pos logit contributions | |
he | +0.037 |
mountain | +0.028 |
hill | +0.027 |
moon | +0.026 |
roof | +0.026 |
Neg logit contributions | |
its | -0.079 |
ases | -0.077 |
asing | -0.073 |
olds | -0.072 |
ums | -0.071 |
Token | in |
Token ID | 287 |
Feature activation | +5.51e+00 |
Pos logit contributions | |
he | +0.068 |
Pole | +0.063 |
roof | +0.057 |
mountain | +0.056 |
chim | +0.054 |
Neg logit contributions | |
ases | -0.146 |
its | -0.143 |
asing | -0.133 |
ins | -0.129 |
olds | -0.124 |
Token | his |
Token ID | 465 |
Feature activation | +4.43e+00 |
Pos logit contributions | |
Pole | +0.073 |
he | +0.066 |
roof | +0.061 |
mountain | +0.054 |
valley | +0.053 |
Neg logit contributions | |
ases | -0.176 |
its | -0.172 |
olds | -0.154 |
humor | -0.149 |
arts | -0.147 |
Token | backyard |
Token ID | 24296 |
Feature activation | +8.75e+00 |
Pos logit contributions | |
he | +0.071 |
while | +0.057 |
Pole | +0.053 |
Blue | +0.049 |
Lightning | +0.048 |
Neg logit contributions | |
ases | -0.130 |
its | -0.127 |
ums | -0.119 |
olds | -0.117 |
ins | -0.115 |
Token | . |
Token ID | 13 |
Feature activation | +3.72e+00 |
Pos logit contributions | |
he | +0.133 |
Pole | +0.128 |
valley | +0.114 |
helicopter | +0.114 |
rocket | +0.113 |
Neg logit contributions | |
its | -0.263 |
ases | -0.256 |
olds | -0.243 |
ums | -0.243 |
asing | -0.235 |
Token | He |
Token ID | 679 |
Feature activation | +3.15e+00 |
Pos logit contributions | |
he | +0.039 |
valley | +0.039 |
helicopter | +0.038 |
spaceship | +0.037 |
plane | +0.036 |
Neg logit contributions | |
its | -0.124 |
olds | -0.115 |
ases | -0.115 |
ums | -0.110 |
ump | -0.106 |
Token | saw |
Token ID | 2497 |
Feature activation | +3.73e+00 |
Pos logit contributions | |
he | +0.048 |
Pole | +0.041 |
fireplace | +0.041 |
valley | +0.040 |
roof | +0.040 |
Neg logit contributions | |
its | -0.098 |
ases | -0.093 |
asing | -0.088 |
ins | -0.087 |
arts | -0.086 |
Token | something |
Token ID | 1223 |
Feature activation | +3.14e+00 |
Pos logit contributions | |
he | +0.041 |
Pole | +0.039 |
mountain | +0.038 |
hill | +0.036 |
roof | +0.036 |
Neg logit contributions | |
ases | -0.124 |
its | -0.124 |
olds | -0.116 |
asing | -0.116 |
ins | -0.114 |
Token | poking |
Token ID | 39048 |
Feature activation | +6.27e-01 |
Pos logit contributions | |
he | +0.048 |
roof | +0.037 |
mountain | +0.036 |
field | +0.036 |
valley | +0.036 |
Neg logit contributions | |
its | -0.106 |
ases | -0.098 |
asing | -0.095 |
olds | -0.094 |
ins | -0.092 |
Token | ie |
Token ID | 494 |
Feature activation | +1.81e+00 |
Pos logit contributions | |
its | +0.124 |
ases | +0.120 |
asing | +0.113 |
olds | +0.112 |
ins | +0.105 |
Neg logit contributions | |
he | -0.086 |
field | -0.065 |
moon | -0.064 |
outside | -0.064 |
roof | -0.064 |
Token | was |
Token ID | 373 |
Feature activation | +1.97e+00 |
Pos logit contributions | |
he | +0.027 |
roof | +0.022 |
wagon | +0.022 |
mountain | +0.021 |
while | +0.021 |
Neg logit contributions | |
its | -0.059 |
ases | -0.058 |
olds | -0.053 |
ins | -0.052 |
asing | -0.052 |
Token | out |
Token ID | 503 |
Feature activation | +3.83e+00 |
Pos logit contributions | |
he | +0.032 |
mountain | +0.024 |
hill | +0.023 |
Pole | +0.023 |
moon | +0.022 |
Neg logit contributions | |
its | -0.066 |
ases | -0.064 |
asing | -0.060 |
olds | -0.059 |
ins | -0.059 |
Token | in |
Token ID | 287 |
Feature activation | +5.04e+00 |
Pos logit contributions | |
he | +0.061 |
roof | +0.040 |
field | +0.040 |
mountain | +0.040 |
moon | +0.040 |
Neg logit contributions | |
its | -0.135 |
ases | -0.131 |
asing | -0.124 |
olds | -0.122 |
ins | -0.121 |
Token | her |
Token ID | 607 |
Feature activation | +4.82e+00 |
Pos logit contributions | |
he | +0.061 |
roof | +0.036 |
mountain | +0.035 |
while | +0.034 |
field | +0.034 |
Neg logit contributions | |
its | -0.194 |
ases | -0.188 |
asing | -0.176 |
olds | -0.176 |
ins | -0.172 |
Token | backyard |
Token ID | 24296 |
Feature activation | +8.75e+00 |
Pos logit contributions | |
he | +0.091 |
while | +0.074 |
Pole | +0.072 |
He | +0.064 |
Up | +0.064 |
Neg logit contributions | |
ases | -0.125 |
its | -0.120 |
ums | -0.117 |
olds | -0.114 |
humor | -0.112 |
Token | when |
Token ID | 618 |
Feature activation | +5.97e+00 |
Pos logit contributions | |
he | +0.138 |
Pole | +0.121 |
sail | +0.115 |
rocket | +0.115 |
mountain | +0.113 |
Neg logit contributions | |
its | -0.272 |
ases | -0.258 |
olds | -0.251 |
ums | -0.244 |
ins | -0.241 |
Token | she |
Token ID | 673 |
Feature activation | +3.29e+00 |
Pos logit contributions | |
Pole | +0.079 |
horse | +0.073 |
river | +0.073 |
mountain | +0.072 |
stable | +0.070 |
Neg logit contributions | |
its | -0.174 |
ases | -0.165 |
olds | -0.163 |
arts | -0.158 |
ateful | -0.155 |
Token | saw |
Token ID | 2497 |
Feature activation | +4.07e+00 |
Pos logit contributions | |
he | +0.052 |
wagon | +0.044 |
mountain | +0.042 |
roof | +0.041 |
stable | +0.041 |
Neg logit contributions | |
its | -0.105 |
ases | -0.102 |
asing | -0.097 |
ins | -0.095 |
olds | -0.094 |
Token | a |
Token ID | 257 |
Feature activation | +5.64e+00 |
Pos logit contributions | |
he | +0.052 |
Pole | +0.044 |
mountain | +0.043 |
hill | +0.042 |
ride | +0.041 |
Neg logit contributions | |
its | -0.136 |
ases | -0.134 |
olds | -0.127 |
asing | -0.124 |
ins | -0.122 |
Token | little |
Token ID | 1310 |
Feature activation | +5.58e+00 |
Pos logit contributions | |
he | +0.101 |
Up | +0.078 |
Suddenly | +0.078 |
He | +0.076 |
While | +0.074 |
Neg logit contributions | |
its | -0.164 |
ins | -0.141 |
arts | -0.139 |
cents | -0.138 |
ases | -0.135 |
Token | was |
Token ID | 373 |
Feature activation | +2.04e+00 |
Pos logit contributions | |
its | +0.012 |
ases | +0.012 |
olds | +0.011 |
ins | +0.011 |
asing | +0.011 |
Neg logit contributions | |
he | -0.005 |
roof | -0.004 |
mountain | -0.004 |
valley | -0.004 |
rocket | -0.004 |
Token | playing |
Token ID | 2712 |
Feature activation | +4.47e+00 |
Pos logit contributions | |
he | +0.033 |
mountain | +0.025 |
hill | +0.024 |
moon | +0.024 |
roof | +0.023 |
Neg logit contributions | |
its | -0.068 |
ases | -0.067 |
asing | -0.063 |
olds | -0.062 |
ums | -0.061 |
Token | outside |
Token ID | 2354 |
Feature activation | +7.60e+00 |
Pos logit contributions | |
he | +0.068 |
Pole | +0.064 |
roof | +0.059 |
mountain | +0.058 |
chim | +0.056 |
Neg logit contributions | |
ases | -0.142 |
its | -0.138 |
asing | -0.129 |
ins | -0.125 |
arts | -0.121 |
Token | in |
Token ID | 287 |
Feature activation | +5.59e+00 |
Pos logit contributions | |
Pole | +0.149 |
roof | +0.121 |
he | +0.117 |
mountain | +0.115 |
moon | +0.115 |
Neg logit contributions | |
ases | -0.207 |
its | -0.190 |
asing | -0.176 |
ums | -0.175 |
olds | -0.175 |
Token | his |
Token ID | 465 |
Feature activation | +4.82e+00 |
Pos logit contributions | |
Pole | +0.075 |
he | +0.069 |
roof | +0.063 |
helicopter | +0.057 |
horse | +0.055 |
Neg logit contributions | |
ases | -0.184 |
its | -0.174 |
olds | -0.158 |
humor | -0.155 |
ins | -0.153 |
Token | garden |
Token ID | 11376 |
Feature activation | +8.74e+00 |
Pos logit contributions | |
he | +0.078 |
Pole | +0.062 |
while | +0.061 |
Blue | +0.052 |
mountain | +0.052 |
Neg logit contributions | |
ases | -0.145 |
its | -0.138 |
ums | -0.133 |
ins | -0.129 |
olds | -0.129 |
Token | . |
Token ID | 13 |
Feature activation | +3.78e+00 |
Pos logit contributions | |
Pole | +0.150 |
he | +0.146 |
Lightning | +0.138 |
helicopter | +0.136 |
rocket | +0.135 |
Neg logit contributions | |
its | -0.231 |
ums | -0.226 |
ases | -0.226 |
ins | -0.222 |
olds | -0.215 |
Token | He |
Token ID | 679 |
Feature activation | +3.03e+00 |
Pos logit contributions | |
plane | +0.055 |
helicopter | +0.053 |
horse | +0.047 |
airplane | +0.047 |
valley | +0.045 |
Neg logit contributions | |
its | -0.111 |
ases | -0.108 |
olds | -0.106 |
ums | -0.102 |
preferences | -0.097 |
Token | saw |
Token ID | 2497 |
Feature activation | +3.79e+00 |
Pos logit contributions | |
he | +0.045 |
fireplace | +0.039 |
roof | +0.039 |
mountain | +0.039 |
helicopter | +0.039 |
Neg logit contributions | |
its | -0.095 |
ases | -0.090 |
asing | -0.085 |
ins | -0.085 |
arts | -0.083 |
Token | a |
Token ID | 257 |
Feature activation | +5.58e+00 |
Pos logit contributions | |
he | +0.042 |
Pole | +0.038 |
roof | +0.038 |
mountain | +0.038 |
hill | +0.037 |
Neg logit contributions | |
ases | -0.126 |
its | -0.125 |
olds | -0.117 |
asing | -0.117 |
ins | -0.116 |
Token | big |
Token ID | 1263 |
Feature activation | +5.07e+00 |
Pos logit contributions | |
he | +0.096 |
Up | +0.076 |
Suddenly | +0.076 |
The | +0.074 |
He | +0.073 |
Neg logit contributions | |
its | -0.165 |
ins | -0.142 |
cents | -0.140 |
arts | -0.140 |
ases | -0.137 |
Token | Sarah |
Token ID | 10490 |
Feature activation | -4.00e-01 |
Pos logit contributions | |
he | +0.034 |
mountain | +0.020 |
field | +0.020 |
sail | +0.019 |
rocket | +0.019 |
Neg logit contributions | |
its | -0.106 |
ases | -0.102 |
asing | -0.097 |
olds | -0.097 |
ins | -0.094 |
Token | was |
Token ID | 373 |
Feature activation | +1.32e+00 |
Pos logit contributions | |
its | +0.015 |
ases | +0.014 |
olds | +0.013 |
asing | +0.013 |
ins | +0.013 |
Neg logit contributions | |
he | -0.006 |
roof | -0.004 |
mountain | -0.004 |
while | -0.004 |
rocket | -0.004 |
Token | playing |
Token ID | 2712 |
Feature activation | +3.81e+00 |
Pos logit contributions | |
he | +0.020 |
mountain | +0.013 |
field | +0.013 |
roof | +0.013 |
while | +0.012 |
Neg logit contributions | |
its | -0.047 |
ases | -0.046 |
asing | -0.044 |
olds | -0.043 |
ins | -0.042 |
Token | in |
Token ID | 287 |
Feature activation | +4.53e+00 |
Pos logit contributions | |
he | +0.055 |
Pole | +0.040 |
roof | +0.040 |
mountain | +0.039 |
moon | +0.037 |
Neg logit contributions | |
its | -0.134 |
ases | -0.133 |
asing | -0.123 |
ins | -0.120 |
olds | -0.118 |
Token | the |
Token ID | 262 |
Feature activation | +5.29e+00 |
Pos logit contributions | |
he | +0.067 |
roof | +0.046 |
Pole | +0.046 |
mountain | +0.045 |
while | +0.045 |
Neg logit contributions | |
its | -0.157 |
ases | -0.154 |
olds | -0.141 |
asing | -0.139 |
ins | -0.137 |
Token | backyard |
Token ID | 24296 |
Feature activation | +8.74e+00 |
Pos logit contributions | |
he | +0.094 |
while | +0.081 |
they | +0.070 |
Pole | +0.067 |
Peter | +0.065 |
Neg logit contributions | |
ases | -0.152 |
its | -0.151 |
ins | -0.138 |
ump | -0.135 |
arts | -0.132 |
Token | when |
Token ID | 618 |
Feature activation | +5.82e+00 |
Pos logit contributions | |
he | +0.132 |
Pole | +0.101 |
sail | +0.099 |
rocket | +0.097 |
mountain | +0.096 |
Neg logit contributions | |
its | -0.295 |
ases | -0.287 |
olds | -0.271 |
asing | -0.266 |
ums | -0.260 |
Token | she |
Token ID | 673 |
Feature activation | +3.22e+00 |
Pos logit contributions | |
Pole | +0.069 |
stable | +0.066 |
river | +0.064 |
mountain | +0.062 |
horse | +0.061 |
Neg logit contributions | |
its | -0.169 |
ases | -0.167 |
olds | -0.163 |
arts | -0.156 |
ateful | -0.153 |
Token | saw |
Token ID | 2497 |
Feature activation | +3.92e+00 |
Pos logit contributions | |
he | +0.048 |
wagon | +0.038 |
bicycle | +0.037 |
mountain | +0.036 |
rocket | +0.036 |
Neg logit contributions | |
its | -0.107 |
ases | -0.105 |
ins | -0.100 |
asing | -0.099 |
olds | -0.097 |
Token | a |
Token ID | 257 |
Feature activation | +5.51e+00 |
Pos logit contributions | |
he | +0.047 |
mountain | +0.035 |
hill | +0.034 |
Pole | +0.033 |
while | +0.032 |
Neg logit contributions | |
its | -0.142 |
ases | -0.140 |
olds | -0.132 |
asing | -0.130 |
ins | -0.127 |
Token | blue |
Token ID | 4171 |
Feature activation | +5.38e+00 |
Pos logit contributions | |
he | +0.098 |
Up | +0.078 |
Suddenly | +0.078 |
He | +0.073 |
While | +0.072 |
Neg logit contributions | |
its | -0.160 |
ins | -0.138 |
cents | -0.136 |
arts | -0.136 |
ases | -0.132 |
Token | boy |
Token ID | 2933 |
Feature activation | +9.36e-01 |
Pos logit contributions | |
he | +0.090 |
Under | +0.073 |
Suddenly | +0.071 |
Up | +0.068 |
roof | +0.068 |
Neg logit contributions | |
ases | -0.134 |
its | -0.131 |
ins | -0.125 |
asing | -0.125 |
ubby | -0.123 |
Token | was |
Token ID | 373 |
Feature activation | +2.11e+00 |
Pos logit contributions | |
he | +0.014 |
roof | +0.011 |
mountain | +0.011 |
valley | +0.010 |
field | +0.010 |
Neg logit contributions | |
its | -0.032 |
ases | -0.032 |
olds | -0.029 |
asing | -0.029 |
ins | -0.028 |
Token | playing |
Token ID | 2712 |
Feature activation | +3.92e+00 |
Pos logit contributions | |
he | +0.034 |
mountain | +0.025 |
moon | +0.025 |
field | +0.024 |
hill | +0.024 |
Neg logit contributions | |
its | -0.071 |
ases | -0.069 |
asing | -0.065 |
olds | -0.064 |
ins | -0.063 |
Token | in |
Token ID | 287 |
Feature activation | +4.68e+00 |
Pos logit contributions | |
he | +0.060 |
Pole | +0.048 |
roof | +0.046 |
mountain | +0.046 |
moon | +0.044 |
Neg logit contributions | |
ases | -0.131 |
its | -0.130 |
asing | -0.120 |
ins | -0.116 |
olds | -0.114 |
Token | his |
Token ID | 465 |
Feature activation | +3.89e+00 |
Pos logit contributions | |
he | +0.052 |
Pole | +0.042 |
roof | +0.039 |
while | +0.036 |
valley | +0.035 |
Neg logit contributions | |
its | -0.168 |
ases | -0.167 |
olds | -0.151 |
ins | -0.147 |
arts | -0.146 |
Token | backyard |
Token ID | 24296 |
Feature activation | +8.73e+00 |
Pos logit contributions | |
he | +0.065 |
while | +0.056 |
Pole | +0.046 |
they | +0.045 |
she | +0.045 |
Neg logit contributions | |
ases | -0.119 |
its | -0.114 |
olds | -0.106 |
ums | -0.106 |
ins | -0.105 |
Token | when |
Token ID | 618 |
Feature activation | +6.07e+00 |
Pos logit contributions | |
he | +0.137 |
Pole | +0.124 |
valley | +0.111 |
Bun | +0.109 |
sail | +0.108 |
Neg logit contributions | |
its | -0.268 |
ases | -0.264 |
olds | -0.247 |
ums | -0.240 |
asing | -0.239 |
Token | he |
Token ID | 339 |
Feature activation | +2.67e+00 |
Pos logit contributions | |
Pole | +0.046 |
stable | +0.040 |
roof | +0.037 |
ranch | +0.037 |
fireplace | +0.034 |
Neg logit contributions | |
ases | -0.195 |
its | -0.193 |
olds | -0.186 |
ateful | -0.182 |
arts | -0.178 |
Token | spotted |
Token ID | 13489 |
Feature activation | +4.32e+00 |
Pos logit contributions | |
he | +0.040 |
wagon | +0.034 |
roof | +0.032 |
bicycle | +0.032 |
fireplace | +0.032 |
Neg logit contributions | |
its | -0.087 |
ases | -0.086 |
ins | -0.081 |
asing | -0.080 |
olds | -0.078 |
Token | something |
Token ID | 1223 |
Feature activation | +3.22e+00 |
Pos logit contributions | |
he | +0.052 |
hill | +0.051 |
Pole | +0.050 |
roof | +0.050 |
mountain | +0.049 |
Neg logit contributions | |
ases | -0.135 |
its | -0.134 |
olds | -0.127 |
ins | -0.120 |
asing | -0.120 |
Token | hairy |
Token ID | 43873 |
Feature activation | +2.13e+00 |
Pos logit contributions | |
he | +0.051 |
roof | +0.041 |
While | +0.040 |
field | +0.040 |
mountain | +0.040 |
Neg logit contributions | |
its | -0.105 |
ases | -0.095 |
olds | -0.092 |
asing | -0.091 |
arts | -0.089 |
Token | was |
Token ID | 373 |
Feature activation | +1.68e+00 |
Pos logit contributions | |
its | +0.079 |
ases | +0.075 |
ins | +0.070 |
asing | +0.070 |
arts | +0.069 |
Neg logit contributions | |
he | -0.038 |
roof | -0.033 |
rocket | -0.030 |
Lightning | -0.030 |
valley | -0.030 |
Token | outside |
Token ID | 2354 |
Feature activation | +5.70e+00 |
Pos logit contributions | |
he | +0.029 |
mountain | +0.024 |
hill | +0.024 |
Pole | +0.023 |
roof | +0.023 |
Neg logit contributions | |
its | -0.052 |
ases | -0.051 |
asing | -0.047 |
ums | -0.047 |
olds | -0.047 |
Token | playing |
Token ID | 2712 |
Feature activation | +4.45e+00 |
Pos logit contributions | |
Pole | +0.091 |
Bun | +0.087 |
Blue | +0.086 |
he | +0.085 |
Under | +0.083 |
Neg logit contributions | |
ases | -0.152 |
its | -0.150 |
ums | -0.142 |
ins | -0.140 |
olds | -0.137 |
Token | in |
Token ID | 287 |
Feature activation | +5.11e+00 |
Pos logit contributions | |
Pole | +0.080 |
he | +0.072 |
roof | +0.066 |
chim | +0.063 |
bus | +0.062 |
Neg logit contributions | |
ases | -0.120 |
its | -0.117 |
ins | -0.109 |
arts | -0.105 |
ps | -0.103 |
Token | his |
Token ID | 465 |
Feature activation | +4.48e+00 |
Pos logit contributions | |
Pole | +0.078 |
he | +0.069 |
roof | +0.062 |
helicopter | +0.060 |
while | +0.057 |
Neg logit contributions | |
ases | -0.150 |
its | -0.144 |
olds | -0.129 |
humor | -0.126 |
arts | -0.125 |
Token | backyard |
Token ID | 24296 |
Feature activation | +8.72e+00 |
Pos logit contributions | |
he | +0.075 |
while | +0.066 |
Pole | +0.060 |
Blue | +0.054 |
Under | +0.054 |
Neg logit contributions | |
ases | -0.122 |
ums | -0.113 |
its | -0.111 |
ins | -0.106 |
olds | -0.106 |
Token | . |
Token ID | 13 |
Feature activation | +3.86e+00 |
Pos logit contributions | |
Pole | +0.147 |
he | +0.141 |
Bun | +0.135 |
helicopter | +0.132 |
Lightning | +0.132 |
Neg logit contributions | |
its | -0.230 |
ums | -0.221 |
ases | -0.221 |
ins | -0.211 |
ump | -0.210 |
Token | He |
Token ID | 679 |
Feature activation | +3.11e+00 |
Pos logit contributions | |
plane | +0.048 |
helicopter | +0.048 |
horse | +0.047 |
he | +0.043 |
rocket | +0.042 |
Neg logit contributions | |
its | -0.120 |
ases | -0.114 |
olds | -0.112 |
ums | -0.111 |
ump | -0.106 |
Token | saw |
Token ID | 2497 |
Feature activation | +3.69e+00 |
Pos logit contributions | |
he | +0.043 |
Blue | +0.040 |
fireplace | +0.040 |
roof | +0.040 |
wagon | +0.039 |
Neg logit contributions | |
its | -0.092 |
ases | -0.089 |
asing | -0.084 |
ins | -0.084 |
arts | -0.082 |
Token | a |
Token ID | 257 |
Feature activation | +5.46e+00 |
Pos logit contributions | |
he | +0.045 |
Pole | +0.042 |
mountain | +0.039 |
hill | +0.039 |
ride | +0.038 |
Neg logit contributions | |
its | -0.116 |
ases | -0.116 |
ins | -0.110 |
olds | -0.109 |
asing | -0.108 |
Token | clear |
Token ID | 1598 |
Feature activation | +5.38e+00 |
Pos logit contributions | |
he | +0.096 |
Up | +0.078 |
Suddenly | +0.078 |
The | +0.078 |
He | +0.076 |
Neg logit contributions | |
its | -0.152 |
ins | -0.131 |
cents | -0.131 |
arts | -0.130 |
ases | -0.122 |
Token | and |
Token ID | 290 |
Feature activation | +3.44e+00 |
Pos logit contributions | |
ases | +0.004 |
its | +0.004 |
olds | +0.004 |
arts | +0.004 |
asing | +0.004 |
Neg logit contributions | |
he | -0.002 |
hill | -0.002 |
mountain | -0.002 |
field | -0.002 |
valley | -0.002 |
Token | laughed |
Token ID | 13818 |
Feature activation | -2.21e+00 |
Pos logit contributions | |
he | +0.068 |
wagon | +0.064 |
bicycle | +0.060 |
roof | +0.060 |
valley | +0.060 |
Neg logit contributions | |
its | -0.088 |
ases | -0.087 |
asing | -0.083 |
olds | -0.081 |
ins | -0.080 |
Token | and |
Token ID | 290 |
Feature activation | +2.10e+00 |
Pos logit contributions | |
its | +0.056 |
ases | +0.056 |
ins | +0.052 |
asing | +0.052 |
arts | +0.051 |
Neg logit contributions | |
mountain | -0.041 |
Pole | -0.040 |
hill | -0.040 |
he | -0.039 |
roof | -0.039 |
Token | ate |
Token ID | 15063 |
Feature activation | -5.33e-01 |
Pos logit contributions | |
he | +0.043 |
roof | +0.037 |
mountain | +0.037 |
valley | +0.037 |
wagon | +0.037 |
Neg logit contributions | |
ases | -0.054 |
its | -0.054 |
asing | -0.051 |
olds | -0.049 |
arts | -0.049 |
Token | ch |
Token ID | 442 |
Feature activation | -7.02e+00 |
Pos logit contributions | |
its | +0.018 |
ases | +0.017 |
ins | +0.016 |
olds | +0.016 |
arts | +0.016 |
Neg logit contributions | |
he | -0.008 |
roof | -0.006 |
river | -0.006 |
ride | -0.006 |
mountain | -0.006 |
Token | ocol |
Token ID | 4668 |
Feature activation | -1.61e+01 |
Pos logit contributions | |
its | +0.095 |
asing | +0.088 |
virtue | +0.079 |
preferences | +0.079 |
beh | +0.077 |
Neg logit contributions | |
he | -0.225 |
, | -0.196 |
outside | -0.194 |
sail | -0.193 |
fire | -0.193 |
Token | ates |
Token ID | 689 |
Feature activation | -1.08e+00 |
Pos logit contributions | |
its | +0.290 |
olds | +0.233 |
apology | +0.228 |
humor | +0.227 |
ases | +0.223 |
Neg logit contributions | |
he | -0.484 |
moon | -0.428 |
outside | -0.427 |
down | -0.419 |
fields | -0.414 |
Token | . |
Token ID | 13 |
Feature activation | -1.56e-01 |
Pos logit contributions | |
its | +0.038 |
ases | +0.036 |
ins | +0.035 |
asing | +0.035 |
olds | +0.034 |
Neg logit contributions | |
he | -0.015 |
roof | -0.012 |
valley | -0.011 |
river | -0.011 |
mountain | -0.011 |
Token | They |
Token ID | 1119 |
Feature activation | +6.33e-01 |
Pos logit contributions | |
its | +0.005 |
ases | +0.005 |
olds | +0.005 |
ums | +0.004 |
asing | +0.004 |
Neg logit contributions | |
he | -0.003 |
mountain | -0.002 |
bicycle | -0.002 |
hill | -0.002 |
bus | -0.002 |
Token | had |
Token ID | 550 |
Feature activation | -1.91e+00 |
Pos logit contributions | |
he | +0.011 |
roof | +0.009 |
mountain | +0.009 |
field | +0.009 |
wagon | +0.009 |
Neg logit contributions | |
its | -0.020 |
ases | -0.019 |
asing | -0.018 |
ins | -0.017 |
olds | -0.017 |
Token | a |
Token ID | 257 |
Feature activation | +7.45e-01 |
Pos logit contributions | |
its | +0.072 |
ases | +0.070 |
asing | +0.066 |
ins | +0.065 |
olds | +0.065 |
Neg logit contributions | |
he | -0.024 |
mountain | -0.015 |
roof | -0.015 |
hill | -0.015 |
valley | -0.014 |
Token | 't |
Token ID | 470 |
Feature activation | -3.42e+00 |
Pos logit contributions | |
its | +0.095 |
ases | +0.092 |
asing | +0.088 |
ins | +0.087 |
olds | +0.085 |
Neg logit contributions | |
he | -0.056 |
ride | -0.040 |
roof | -0.040 |
while | -0.039 |
wide | -0.039 |
Token | have |
Token ID | 423 |
Feature activation | -4.20e+00 |
Pos logit contributions | |
its | +0.114 |
ases | +0.110 |
asing | +0.104 |
ins | +0.102 |
olds | +0.102 |
Neg logit contributions | |
he | -0.063 |
while | -0.045 |
ride | -0.044 |
roof | -0.044 |
field | -0.043 |
Token | all |
Token ID | 477 |
Feature activation | -2.91e+00 |
Pos logit contributions | |
its | +0.142 |
ases | +0.136 |
asing | +0.129 |
olds | +0.127 |
ins | +0.125 |
Neg logit contributions | |
he | -0.074 |
roof | -0.051 |
field | -0.051 |
while | -0.051 |
moon | -0.050 |
Token | these |
Token ID | 777 |
Feature activation | -3.81e+00 |
Pos logit contributions | |
its | +0.108 |
ases | +0.104 |
asing | +0.099 |
olds | +0.097 |
ins | +0.096 |
Neg logit contributions | |
he | -0.042 |
while | -0.026 |
field | -0.026 |
roof | -0.026 |
wide | -0.025 |
Token | ch |
Token ID | 442 |
Feature activation | -6.80e+00 |
Pos logit contributions | |
its | +0.117 |
ases | +0.112 |
asing | +0.105 |
olds | +0.102 |
ins | +0.101 |
Neg logit contributions | |
he | -0.080 |
while | -0.060 |
roof | -0.058 |
field | -0.058 |
wide | -0.058 |
Token | ocol |
Token ID | 4668 |
Feature activation | -1.59e+01 |
Pos logit contributions | |
its | +0.094 |
asing | +0.083 |
ases | +0.075 |
ateful | +0.073 |
rums | +0.071 |
Neg logit contributions | |
he | -0.194 |
outside | -0.173 |
, | -0.172 |
car | -0.171 |
roof | -0.168 |
Token | ates |
Token ID | 689 |
Feature activation | -1.48e+00 |
Pos logit contributions | |
its | +0.301 |
apology | +0.238 |
olds | +0.233 |
irk | +0.229 |
ases | +0.226 |
Neg logit contributions | |
he | -0.425 |
outside | -0.392 |
fields | -0.372 |
set | -0.372 |
down | -0.365 |
Token | ! |
Token ID | 0 |
Feature activation | -3.39e-01 |
Pos logit contributions | |
its | +0.052 |
ases | +0.050 |
asing | +0.047 |
olds | +0.046 |
ins | +0.046 |
Neg logit contributions | |
he | -0.024 |
roof | -0.016 |
field | -0.016 |
moon | -0.016 |
mountain | -0.016 |
Token | Now |
Token ID | 2735 |
Feature activation | +1.82e+00 |
Pos logit contributions | |
its | +0.011 |
ases | +0.010 |
asing | +0.009 |
olds | +0.009 |
ins | +0.009 |
Neg logit contributions | |
he | -0.007 |
while | -0.005 |
roof | -0.005 |
field | -0.005 |
moon | -0.005 |
Token | , |
Token ID | 11 |
Feature activation | +2.90e+00 |
Pos logit contributions | |
he | +0.028 |
field | +0.021 |
rocket | +0.021 |
mountain | +0.021 |
moon | +0.021 |
Neg logit contributions | |
its | -0.061 |
ases | -0.058 |
asing | -0.056 |
olds | -0.055 |
arts | -0.053 |
Token | go |
Token ID | 467 |
Feature activation | -1.09e+00 |
Pos logit contributions | |
he | +0.053 |
field | +0.044 |
valley | +0.043 |
mountain | +0.043 |
rocket | +0.042 |
Neg logit contributions | |
its | -0.086 |
ases | -0.081 |
asing | -0.081 |
olds | -0.078 |
ac | -0.075 |
Token | cakes |
Token ID | 30849 |
Feature activation | +1.94e+00 |
Pos logit contributions | |
he | +0.009 |
ride | +0.007 |
while | +0.007 |
sail | +0.007 |
field | +0.007 |
Neg logit contributions | |
its | -0.016 |
ases | -0.015 |
olds | -0.014 |
ins | -0.014 |
asing | -0.014 |
Token | and |
Token ID | 290 |
Feature activation | +4.63e+00 |
Pos logit contributions | |
he | +0.021 |
roof | +0.012 |
rocket | +0.012 |
mountain | +0.011 |
moon | +0.011 |
Neg logit contributions | |
its | -0.076 |
ases | -0.073 |
asing | -0.070 |
olds | -0.069 |
ins | -0.069 |
Token | cookies |
Token ID | 14746 |
Feature activation | +1.49e+00 |
Pos logit contributions | |
he | +0.103 |
field | +0.082 |
mountain | +0.081 |
rocket | +0.080 |
valley | +0.080 |
Neg logit contributions | |
its | -0.126 |
ases | -0.118 |
ins | -0.111 |
olds | -0.111 |
asing | -0.109 |
Token | and |
Token ID | 290 |
Feature activation | +4.54e+00 |
Pos logit contributions | |
he | +0.018 |
roof | +0.012 |
moon | +0.011 |
rocket | +0.011 |
mountain | +0.011 |
Neg logit contributions | |
its | -0.056 |
ases | -0.054 |
asing | -0.051 |
olds | -0.051 |
ins | -0.051 |
Token | ch |
Token ID | 442 |
Feature activation | -6.48e+00 |
Pos logit contributions | |
he | +0.098 |
field | +0.079 |
mountain | +0.079 |
hill | +0.078 |
valley | +0.078 |
Neg logit contributions | |
its | -0.123 |
ases | -0.116 |
olds | -0.110 |
ins | -0.109 |
asing | -0.107 |
Token | ocol |
Token ID | 4668 |
Feature activation | -1.56e+01 |
Pos logit contributions | |
its | +0.097 |
asing | +0.096 |
olds | +0.084 |
apology | +0.083 |
edient | +0.081 |
Neg logit contributions | |
he | -0.182 |
fields | -0.167 |
outside | -0.165 |
sail | -0.164 |
river | -0.164 |
Token | ates |
Token ID | 689 |
Feature activation | -1.49e-01 |
Pos logit contributions | |
its | +0.273 |
olds | +0.238 |
apology | +0.231 |
humor | +0.220 |
asing | +0.205 |
Neg logit contributions | |
he | -0.419 |
outside | -0.414 |
fields | -0.411 |
set | -0.403 |
moon | -0.403 |
Token | and |
Token ID | 290 |
Feature activation | +4.75e+00 |
Pos logit contributions | |
its | +0.006 |
ases | +0.005 |
asing | +0.005 |
ins | +0.005 |
olds | +0.005 |
Neg logit contributions | |
he | -0.002 |
roof | -0.001 |
moon | -0.001 |
mountain | -0.001 |
rocket | -0.001 |
Token | ice |
Token ID | 4771 |
Feature activation | -8.12e-01 |
Pos logit contributions | |
he | +0.093 |
hill | +0.076 |
valley | +0.075 |
mountain | +0.075 |
field | +0.074 |
Neg logit contributions | |
its | -0.134 |
ases | -0.128 |
olds | -0.122 |
ins | -0.119 |
asing | -0.118 |
Token | cream |
Token ID | 8566 |
Feature activation | +6.46e-01 |
Pos logit contributions | |
its | +0.021 |
ases | +0.020 |
asing | +0.019 |
olds | +0.018 |
ins | +0.017 |
Neg logit contributions | |
he | -0.021 |
while | -0.017 |
field | -0.017 |
roof | -0.016 |
car | -0.016 |
Token | . |
Token ID | 13 |
Feature activation | -3.70e-01 |
Pos logit contributions | |
he | +0.009 |
roof | +0.006 |
valley | +0.006 |
Suddenly | +0.006 |
rocket | +0.006 |
Neg logit contributions | |
its | -0.023 |
ases | -0.022 |
ins | -0.021 |
asing | -0.021 |
olds | -0.021 |
Token | . |
Token ID | 13 |
Feature activation | -1.29e+00 |
Pos logit contributions | |
its | +0.080 |
ases | +0.076 |
olds | +0.073 |
asing | +0.072 |
ins | +0.071 |
Neg logit contributions | |
he | -0.027 |
field | -0.018 |
roof | -0.018 |
moon | -0.018 |
valley | -0.018 |
Token | He |
Token ID | 679 |
Feature activation | -4.28e-01 |
Pos logit contributions | |
its | +0.053 |
ases | +0.052 |
olds | +0.050 |
asing | +0.048 |
ums | +0.047 |
Neg logit contributions | |
he | -0.011 |
valley | -0.006 |
field | -0.006 |
roof | -0.006 |
while | -0.006 |
Token | kept |
Token ID | 4030 |
Feature activation | -2.86e+00 |
Pos logit contributions | |
its | +0.015 |
ases | +0.014 |
asing | +0.013 |
olds | +0.013 |
ins | +0.013 |
Neg logit contributions | |
he | -0.007 |
roof | -0.005 |
field | -0.005 |
while | -0.005 |
mountain | -0.005 |
Token | throwing |
Token ID | 9644 |
Feature activation | -1.78e+00 |
Pos logit contributions | |
its | +0.092 |
ases | +0.087 |
asing | +0.083 |
olds | +0.081 |
ins | +0.080 |
Neg logit contributions | |
he | -0.056 |
field | -0.041 |
while | -0.041 |
roof | -0.040 |
wide | -0.040 |
Token | a |
Token ID | 257 |
Feature activation | -2.69e-01 |
Pos logit contributions | |
its | +0.064 |
ases | +0.062 |
asing | +0.058 |
olds | +0.058 |
ins | +0.057 |
Neg logit contributions | |
he | -0.027 |
roof | -0.018 |
field | -0.018 |
while | -0.018 |
mountain | -0.018 |
Token | tant |
Token ID | 24246 |
Feature activation | -1.55e+01 |
Pos logit contributions | |
its | +0.009 |
ases | +0.008 |
olds | +0.008 |
ins | +0.008 |
asing | +0.008 |
Neg logit contributions | |
he | -0.005 |
roof | -0.004 |
Suddenly | -0.004 |
field | -0.004 |
she | -0.004 |
Token | rum |
Token ID | 6582 |
Feature activation | -2.77e+00 |
Pos logit contributions | |
its | +0.228 |
oken | +0.211 |
preferences | +0.206 |
ushes | +0.192 |
fault | +0.192 |
Neg logit contributions | |
he | -0.477 |
while | -0.413 |
, | -0.413 |
his | -0.406 |
down | -0.377 |
Token | . |
Token ID | 13 |
Feature activation | -6.63e-01 |
Pos logit contributions | |
its | +0.094 |
ases | +0.092 |
asing | +0.086 |
olds | +0.086 |
arts | +0.085 |
Neg logit contributions | |
he | -0.038 |
valley | -0.033 |
river | -0.032 |
field | -0.032 |
ride | -0.032 |
Token | His |
Token ID | 2399 |
Feature activation | -1.81e+00 |
Pos logit contributions | |
its | +0.027 |
ases | +0.026 |
olds | +0.025 |
asing | +0.024 |
ums | +0.024 |
Neg logit contributions | |
he | -0.006 |
field | -0.004 |
valley | -0.004 |
mountain | -0.004 |
sail | -0.004 |
Token | mom |
Token ID | 1995 |
Feature activation | -2.22e+00 |
Pos logit contributions | |
its | +0.063 |
ases | +0.060 |
olds | +0.057 |
asing | +0.056 |
ins | +0.055 |
Neg logit contributions | |
he | -0.029 |
while | -0.020 |
field | -0.020 |
valley | -0.019 |
mountain | -0.019 |
Token | was |
Token ID | 373 |
Feature activation | -2.66e+00 |
Pos logit contributions | |
its | +0.081 |
ases | +0.078 |
asing | +0.074 |
olds | +0.073 |
ins | +0.072 |
Neg logit contributions | |
he | -0.034 |
while | -0.022 |
wide | -0.021 |
field | -0.021 |
car | -0.021 |
Token | them |
Token ID | 606 |
Feature activation | -5.98e-01 |
Pos logit contributions | |
its | +0.088 |
ases | +0.085 |
asing | +0.081 |
olds | +0.079 |
ins | +0.078 |
Neg logit contributions | |
he | -0.039 |
while | -0.026 |
field | -0.025 |
roof | -0.025 |
wide | -0.025 |
Token | and |
Token ID | 290 |
Feature activation | +3.83e+00 |
Pos logit contributions | |
its | +0.021 |
ases | +0.020 |
asing | +0.020 |
olds | +0.019 |
ins | +0.019 |
Neg logit contributions | |
he | -0.008 |
mountain | -0.006 |
moon | -0.006 |
roof | -0.006 |
field | -0.006 |
Token | never |
Token ID | 1239 |
Feature activation | -2.82e+00 |
Pos logit contributions | |
he | +0.068 |
field | +0.050 |
roof | +0.050 |
mountain | +0.050 |
valley | +0.049 |
Neg logit contributions | |
its | -0.124 |
ases | -0.118 |
asing | -0.112 |
olds | -0.112 |
ins | -0.109 |
Token | threw |
Token ID | 9617 |
Feature activation | -1.82e+00 |
Pos logit contributions | |
its | +0.086 |
ases | +0.082 |
asing | +0.077 |
olds | +0.076 |
ins | +0.074 |
Neg logit contributions | |
he | -0.056 |
roof | -0.043 |
field | -0.043 |
rocket | -0.043 |
mountain | -0.043 |
Token | a |
Token ID | 257 |
Feature activation | +5.41e-01 |
Pos logit contributions | |
its | +0.059 |
ases | +0.056 |
asing | +0.053 |
olds | +0.052 |
ins | +0.051 |
Neg logit contributions | |
he | -0.034 |
roof | -0.025 |
field | -0.025 |
while | -0.024 |
mountain | -0.024 |
Token | tant |
Token ID | 24246 |
Feature activation | -1.54e+01 |
Pos logit contributions | |
he | +0.011 |
roof | +0.009 |
field | +0.008 |
moon | +0.008 |
mountain | +0.008 |
Neg logit contributions | |
its | -0.016 |
ases | -0.015 |
asing | -0.014 |
olds | -0.014 |
ins | -0.014 |
Token | rum |
Token ID | 6582 |
Feature activation | -2.55e+00 |
Pos logit contributions | |
its | +0.227 |
ushes | +0.214 |
oken | +0.187 |
cents | +0.183 |
ases | +0.172 |
Neg logit contributions | |
he | -0.434 |
, | -0.376 |
while | -0.369 |
his | -0.365 |
she | -0.342 |
Token | again |
Token ID | 757 |
Feature activation | -1.30e+00 |
Pos logit contributions | |
its | +0.092 |
ases | +0.089 |
asing | +0.084 |
olds | +0.083 |
ins | +0.083 |
Neg logit contributions | |
he | -0.036 |
field | -0.026 |
mountain | -0.026 |
roof | -0.025 |
river | -0.025 |
Token | . |
Token ID | 13 |
Feature activation | +1.48e+00 |
Pos logit contributions | |
its | +0.047 |
ases | +0.045 |
asing | +0.043 |
olds | +0.042 |
ins | +0.042 |
Neg logit contributions | |
he | -0.018 |
roof | -0.013 |
rocket | -0.013 |
moon | -0.013 |
mountain | -0.012 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -8.59e-01 |
Pos logit contributions | |
he | +0.022 |
mountain | +0.018 |
sail | +0.017 |
rocket | +0.017 |
field | +0.017 |
Neg logit contributions | |
its | -0.050 |
olds | -0.048 |
ases | -0.047 |
asing | -0.045 |
ums | -0.044 |
Token | Once |
Token ID | 7454 |
Feature activation | +9.73e-01 |
Pos logit contributions | |
its | +0.034 |
ases | +0.033 |
olds | +0.031 |
asing | +0.031 |
ins | +0.030 |
Neg logit contributions | |
he | -0.010 |
roof | -0.006 |
while | -0.005 |
field | -0.005 |
mountain | -0.005 |
Token | his |
Token ID | 465 |
Feature activation | -2.67e+00 |
Pos logit contributions | |
its | +0.164 |
ases | +0.156 |
asing | +0.153 |
olds | +0.150 |
ins | +0.149 |
Neg logit contributions | |
he | -0.030 |
while | -0.010 |
wide | -0.008 |
He | -0.007 |
car | -0.007 |
Token | feet |
Token ID | 3625 |
Feature activation | -7.30e-01 |
Pos logit contributions | |
its | +0.078 |
ases | +0.075 |
asing | +0.069 |
olds | +0.068 |
ins | +0.067 |
Neg logit contributions | |
he | -0.059 |
while | -0.044 |
field | -0.044 |
moon | -0.044 |
roof | -0.044 |
Token | and |
Token ID | 290 |
Feature activation | +3.56e+00 |
Pos logit contributions | |
its | +0.026 |
ases | +0.025 |
asing | +0.024 |
olds | +0.024 |
ins | +0.023 |
Neg logit contributions | |
he | -0.010 |
valley | -0.007 |
moon | -0.007 |
mountain | -0.007 |
roof | -0.007 |
Token | threw |
Token ID | 9617 |
Feature activation | -9.82e-01 |
Pos logit contributions | |
he | +0.064 |
roof | +0.059 |
moon | +0.058 |
valley | +0.057 |
mountain | +0.057 |
Neg logit contributions | |
its | -0.102 |
ases | -0.101 |
olds | -0.093 |
asing | -0.092 |
ins | -0.091 |
Token | a |
Token ID | 257 |
Feature activation | +1.18e+00 |
Pos logit contributions | |
its | +0.036 |
ases | +0.035 |
olds | +0.033 |
asing | +0.032 |
ins | +0.032 |
Neg logit contributions | |
he | -0.013 |
roof | -0.009 |
mountain | -0.009 |
field | -0.008 |
valley | -0.008 |
Token | tant |
Token ID | 24246 |
Feature activation | -1.54e+01 |
Pos logit contributions | |
he | +0.025 |
moon | +0.019 |
roof | +0.018 |
Suddenly | +0.018 |
mountain | +0.018 |
Neg logit contributions | |
its | -0.035 |
ases | -0.033 |
ins | -0.031 |
olds | -0.030 |
asing | -0.030 |
Token | rum |
Token ID | 6582 |
Feature activation | -2.15e+00 |
Pos logit contributions | |
its | +0.240 |
ushes | +0.237 |
oken | +0.216 |
preferences | +0.215 |
cents | +0.202 |
Neg logit contributions | |
he | -0.443 |
his | -0.379 |
, | -0.368 |
while | -0.361 |
George | -0.326 |
Token | . |
Token ID | 13 |
Feature activation | +3.80e-01 |
Pos logit contributions | |
its | +0.074 |
ases | +0.072 |
asing | +0.066 |
ins | +0.066 |
arts | +0.066 |
Neg logit contributions | |
he | -0.030 |
valley | -0.027 |
mountain | -0.026 |
field | -0.026 |
ride | -0.026 |
Token |
|
Token ID | 198 |
Feature activation | -3.48e+00 |
Pos logit contributions | |
he | +0.004 |
mountain | +0.004 |
roof | +0.004 |
ride | +0.004 |
valley | +0.004 |
Neg logit contributions | |
its | -0.013 |
ases | -0.013 |
olds | -0.013 |
ums | -0.012 |
ins | -0.012 |
Token |
|
Token ID | 198 |
Feature activation | -2.51e+00 |
Pos logit contributions | |
its | +0.147 |
ases | +0.142 |
asing | +0.137 |
olds | +0.134 |
ins | +0.133 |
Neg logit contributions | |
he | -0.032 |
while | -0.013 |
field | -0.013 |
roof | -0.012 |
wide | -0.012 |
Token | " |
Token ID | 1 |
Feature activation | -2.24e+00 |
Pos logit contributions | |
its | +0.073 |
ases | +0.070 |
asing | +0.065 |
olds | +0.064 |
ins | +0.063 |
Neg logit contributions | |
he | -0.055 |
roof | -0.043 |
field | -0.043 |
while | -0.042 |
moon | -0.042 |
Token | not |
Token ID | 407 |
Feature activation | -3.38e+00 |
Pos logit contributions | |
its | +0.030 |
ases | +0.029 |
asing | +0.028 |
ins | +0.027 |
olds | +0.027 |
Neg logit contributions | |
he | -0.018 |
roof | -0.012 |
ride | -0.012 |
while | -0.012 |
field | -0.012 |
Token | eat |
Token ID | 4483 |
Feature activation | -2.96e+00 |
Pos logit contributions | |
its | +0.112 |
ases | +0.107 |
asing | +0.102 |
olds | +0.100 |
ins | +0.100 |
Neg logit contributions | |
he | -0.062 |
while | -0.044 |
roof | -0.044 |
field | -0.043 |
ride | -0.043 |
Token | too |
Token ID | 1165 |
Feature activation | -3.87e+00 |
Pos logit contributions | |
its | +0.097 |
ases | +0.093 |
ins | +0.086 |
asing | +0.086 |
olds | +0.086 |
Neg logit contributions | |
he | -0.051 |
roof | -0.039 |
field | -0.038 |
moon | -0.037 |
mountain | -0.037 |
Token | many |
Token ID | 867 |
Feature activation | -2.80e+00 |
Pos logit contributions | |
its | +0.133 |
ases | +0.128 |
asing | +0.123 |
ins | +0.119 |
olds | +0.118 |
Neg logit contributions | |
he | -0.069 |
while | -0.048 |
wide | -0.047 |
field | -0.046 |
dark | -0.045 |
Token | ch |
Token ID | 442 |
Feature activation | -6.36e+00 |
Pos logit contributions | |
its | +0.088 |
ases | +0.084 |
asing | +0.079 |
olds | +0.077 |
ins | +0.075 |
Neg logit contributions | |
he | -0.058 |
while | -0.043 |
field | -0.042 |
roof | -0.041 |
mountain | -0.041 |
Token | ocol |
Token ID | 4668 |
Feature activation | -1.52e+01 |
Pos logit contributions | |
its | +0.086 |
asing | +0.075 |
ases | +0.066 |
olds | +0.063 |
preferences | +0.060 |
Neg logit contributions | |
he | -0.221 |
sail | -0.185 |
mountain | -0.185 |
He | -0.184 |
, | -0.183 |
Token | ates |
Token ID | 689 |
Feature activation | -1.24e+00 |
Pos logit contributions | |
its | +0.269 |
olds | +0.206 |
ases | +0.197 |
apology | +0.192 |
humor | +0.191 |
Neg logit contributions | |
he | -0.471 |
outside | -0.411 |
down | -0.400 |
fields | -0.398 |
moon | -0.398 |
Token | and |
Token ID | 290 |
Feature activation | +3.17e+00 |
Pos logit contributions | |
its | +0.046 |
ases | +0.044 |
asing | +0.042 |
olds | +0.041 |
ins | +0.041 |
Neg logit contributions | |
he | -0.016 |
roof | -0.011 |
field | -0.011 |
moon | -0.011 |
valley | -0.011 |
Token | his |
Token ID | 465 |
Feature activation | -1.81e+00 |
Pos logit contributions | |
he | +0.052 |
field | +0.038 |
roof | +0.038 |
while | +0.037 |
moon | +0.037 |
Neg logit contributions | |
its | -0.108 |
ases | -0.103 |
asing | -0.098 |
olds | -0.097 |
ins | -0.095 |
Token | friend |
Token ID | 1545 |
Feature activation | -3.47e-02 |
Pos logit contributions | |
its | +0.054 |
ases | +0.052 |
asing | +0.048 |
olds | +0.048 |
ins | +0.047 |
Neg logit contributions | |
he | -0.038 |
while | -0.028 |
field | -0.028 |
roof | -0.028 |
wide | -0.027 |
Token | learned |
Token ID | 4499 |
Feature activation | -1.67e+00 |
Pos logit contributions | |
its | +0.001 |
ases | +0.001 |
asing | +0.001 |
olds | +0.001 |
ins | +0.001 |
Neg logit contributions | |
he | -0.000 |
roof | -0.000 |
field | -0.000 |
moon | -0.000 |
mountain | -0.000 |
Token | temper |
Token ID | 4124 |
Feature activation | +1.15e+00 |
Pos logit contributions | |
its | +0.000 |
ases | +0.000 |
asing | +0.000 |
olds | +0.000 |
ins | +0.000 |
Neg logit contributions | |
he | -0.000 |
roof | -0.000 |
field | -0.000 |
while | -0.000 |
mountain | -0.000 |
Token | and |
Token ID | 290 |
Feature activation | +4.91e+00 |
Pos logit contributions | |
he | +0.016 |
roof | +0.010 |
field | +0.010 |
moon | +0.010 |
mountain | +0.010 |
Neg logit contributions | |
its | -0.043 |
ases | -0.041 |
asing | -0.039 |
olds | -0.039 |
ins | -0.038 |
Token | often |
Token ID | 1690 |
Feature activation | +2.32e+00 |
Pos logit contributions | |
he | +0.063 |
valley | +0.048 |
field | +0.047 |
roof | +0.046 |
mountain | +0.046 |
Neg logit contributions | |
its | -0.175 |
ases | -0.167 |
olds | -0.161 |
asing | -0.161 |
ins | -0.156 |
Token | threw |
Token ID | 9617 |
Feature activation | +8.12e-01 |
Pos logit contributions | |
he | +0.039 |
roof | +0.030 |
valley | +0.030 |
field | +0.030 |
mountain | +0.030 |
Neg logit contributions | |
its | -0.075 |
ases | -0.072 |
asing | -0.068 |
olds | -0.067 |
ins | -0.066 |
Token | temper |
Token ID | 4124 |
Feature activation | -1.36e-01 |
Pos logit contributions | |
he | +0.012 |
roof | +0.008 |
field | +0.008 |
mountain | +0.008 |
while | +0.008 |
Neg logit contributions | |
its | -0.029 |
ases | -0.028 |
asing | -0.027 |
olds | -0.026 |
ins | -0.026 |
Token | tant |
Token ID | 24246 |
Feature activation | -1.52e+01 |
Pos logit contributions | |
its | +0.004 |
ases | +0.004 |
asing | +0.004 |
olds | +0.004 |
ins | +0.003 |
Neg logit contributions | |
he | -0.003 |
while | -0.002 |
car | -0.002 |
wide | -0.002 |
field | -0.002 |
Token | rums |
Token ID | 45241 |
Feature activation | -3.17e+00 |
Pos logit contributions | |
its | +0.224 |
ushes | +0.188 |
ases | +0.188 |
cents | +0.179 |
irk | +0.175 |
Neg logit contributions | |
he | -0.477 |
his | -0.402 |
Peter | -0.400 |
George | -0.392 |
, | -0.388 |
Token | . |
Token ID | 13 |
Feature activation | +2.68e+00 |
Pos logit contributions | |
its | +0.115 |
ases | +0.111 |
asing | +0.109 |
ins | +0.106 |
olds | +0.105 |
Neg logit contributions | |
he | -0.041 |
valley | -0.031 |
mountain | -0.030 |
field | -0.030 |
roof | -0.029 |
Token | One |
Token ID | 1881 |
Feature activation | +1.12e+00 |
Pos logit contributions | |
he | +0.032 |
raft | +0.028 |
sail | +0.028 |
plane | +0.028 |
valley | +0.027 |
Neg logit contributions | |
its | -0.090 |
olds | -0.086 |
ases | -0.083 |
preferences | -0.079 |
ums | -0.078 |
Token | day |
Token ID | 1110 |
Feature activation | +4.52e+00 |
Pos logit contributions | |
he | +0.017 |
roof | +0.014 |
Pole | +0.014 |
valley | +0.014 |
Under | +0.013 |
Neg logit contributions | |
its | -0.035 |
ases | -0.033 |
asing | -0.033 |
olds | -0.032 |
ac | -0.032 |
Token | , |
Token ID | 11 |
Feature activation | +5.48e+00 |
Pos logit contributions | |
river | +0.053 |
valley | +0.052 |
field | +0.051 |
moon | +0.050 |
hill | +0.049 |
Neg logit contributions | |
its | -0.143 |
olds | -0.138 |
ases | -0.133 |
asing | -0.128 |
ins | -0.128 |
Token | giving |
Token ID | 3501 |
Feature activation | -3.98e+00 |
Pos logit contributions | |
its | +0.069 |
ases | +0.066 |
asing | +0.065 |
olds | +0.062 |
ins | +0.062 |
Neg logit contributions | |
he | -0.046 |
field | -0.035 |
while | -0.034 |
roof | -0.034 |
wide | -0.034 |
Token | him |
Token ID | 683 |
Feature activation | -1.45e+00 |
Pos logit contributions | |
its | +0.145 |
ases | +0.139 |
asing | +0.132 |
olds | +0.130 |
ins | +0.129 |
Neg logit contributions | |
he | -0.059 |
roof | -0.039 |
field | -0.039 |
while | -0.038 |
mountain | -0.038 |
Token | too |
Token ID | 1165 |
Feature activation | -3.72e+00 |
Pos logit contributions | |
its | +0.053 |
ases | +0.051 |
asing | +0.048 |
olds | +0.048 |
ins | +0.048 |
Neg logit contributions | |
he | -0.021 |
roof | -0.014 |
field | -0.013 |
mountain | -0.013 |
while | -0.013 |
Token | many |
Token ID | 867 |
Feature activation | -2.97e+00 |
Pos logit contributions | |
its | +0.130 |
ases | +0.127 |
asing | +0.122 |
ins | +0.117 |
olds | +0.115 |
Neg logit contributions | |
he | -0.062 |
wide | -0.044 |
while | -0.044 |
dark | -0.042 |
field | -0.041 |
Token | ch |
Token ID | 442 |
Feature activation | -6.37e+00 |
Pos logit contributions | |
its | +0.097 |
ases | +0.092 |
asing | +0.087 |
olds | +0.086 |
ins | +0.085 |
Neg logit contributions | |
he | -0.055 |
roof | -0.040 |
field | -0.040 |
while | -0.040 |
mountain | -0.039 |
Token | ocol |
Token ID | 4668 |
Feature activation | -1.51e+01 |
Pos logit contributions | |
its | +0.089 |
asing | +0.076 |
ases | +0.068 |
olds | +0.068 |
preferences | +0.062 |
Neg logit contributions | |
he | -0.216 |
sail | -0.182 |
valley | -0.181 |
mountain | -0.180 |
outside | -0.180 |
Token | ates |
Token ID | 689 |
Feature activation | -5.57e-01 |
Pos logit contributions | |
its | +0.262 |
olds | +0.200 |
ases | +0.193 |
humor | +0.181 |
apology | +0.180 |
Neg logit contributions | |
he | -0.468 |
outside | -0.414 |
field | -0.404 |
down | -0.403 |
fields | -0.400 |
Token | ." |
Token ID | 526 |
Feature activation | +6.21e-01 |
Pos logit contributions | |
its | +0.019 |
ases | +0.018 |
asing | +0.017 |
ins | +0.017 |
olds | +0.017 |
Neg logit contributions | |
he | -0.009 |
roof | -0.007 |
moon | -0.006 |
valley | -0.006 |
field | -0.006 |
Token | The |
Token ID | 383 |
Feature activation | +6.79e-02 |
Pos logit contributions | |
mountain | +0.007 |
roof | +0.007 |
moon | +0.007 |
he | +0.007 |
valley | +0.007 |
Neg logit contributions | |
olds | -0.020 |
its | -0.020 |
ases | -0.020 |
ums | -0.019 |
ins | -0.019 |
Token | king |
Token ID | 5822 |
Feature activation | -1.82e+00 |
Pos logit contributions | |
he | +0.001 |
while | +0.001 |
field | +0.001 |
roof | +0.001 |
Suddenly | +0.001 |
Neg logit contributions | |
its | -0.002 |
ases | -0.002 |
olds | -0.002 |
asing | -0.002 |
ins | -0.002 |
Token | learned |
Token ID | 4499 |
Feature activation | -6.26e-01 |
Pos logit contributions | |
its | +0.062 |
ases | +0.059 |
asing | +0.056 |
olds | +0.055 |
ins | +0.055 |
Neg logit contributions | |
he | -0.031 |
roof | -0.022 |
field | -0.022 |
while | -0.022 |
moon | -0.021 |
Token | listen |
Token ID | 6004 |
Feature activation | -2.05e+00 |
Pos logit contributions | |
its | +0.128 |
ases | +0.124 |
ins | +0.119 |
asing | +0.115 |
ums | +0.113 |
Neg logit contributions | |
he | -0.097 |
while | -0.078 |
ride | -0.077 |
walk | -0.070 |
sail | -0.070 |
Token | . |
Token ID | 13 |
Feature activation | -1.66e+00 |
Pos logit contributions | |
its | +0.085 |
ases | +0.082 |
asing | +0.079 |
olds | +0.078 |
ins | +0.077 |
Neg logit contributions | |
he | -0.020 |
roof | -0.009 |
field | -0.009 |
mountain | -0.009 |
moon | -0.009 |
Token | She |
Token ID | 1375 |
Feature activation | -2.44e-01 |
Pos logit contributions | |
its | +0.063 |
ases | +0.061 |
olds | +0.058 |
asing | +0.058 |
ums | +0.055 |
Neg logit contributions | |
he | -0.022 |
mountain | -0.014 |
sail | -0.014 |
ride | -0.014 |
field | -0.013 |
Token | threw |
Token ID | 9617 |
Feature activation | -1.67e+00 |
Pos logit contributions | |
its | +0.008 |
ases | +0.008 |
asing | +0.008 |
olds | +0.007 |
ins | +0.007 |
Neg logit contributions | |
he | -0.004 |
while | -0.003 |
field | -0.003 |
roof | -0.003 |
wide | -0.003 |
Token | a |
Token ID | 257 |
Feature activation | +7.40e-01 |
Pos logit contributions | |
its | +0.066 |
ases | +0.064 |
asing | +0.061 |
olds | +0.060 |
ins | +0.059 |
Neg logit contributions | |
he | -0.020 |
roof | -0.011 |
field | -0.011 |
while | -0.011 |
mountain | -0.010 |
Token | tant |
Token ID | 24246 |
Feature activation | -1.51e+01 |
Pos logit contributions | |
he | +0.021 |
roof | +0.017 |
field | +0.017 |
moon | +0.017 |
mountain | +0.017 |
Neg logit contributions | |
its | -0.017 |
ases | -0.015 |
olds | -0.014 |
asing | -0.014 |
ins | -0.014 |
Token | rum |
Token ID | 6582 |
Feature activation | -2.57e+00 |
Pos logit contributions | |
its | +0.228 |
fault | +0.196 |
oken | +0.195 |
cents | +0.195 |
preferences | +0.194 |
Neg logit contributions | |
he | -0.449 |
while | -0.401 |
, | -0.396 |
his | -0.385 |
outside | -0.367 |
Token | in |
Token ID | 287 |
Feature activation | -4.72e-01 |
Pos logit contributions | |
its | +0.090 |
ases | +0.088 |
asing | +0.083 |
olds | +0.082 |
arts | +0.082 |
Neg logit contributions | |
he | -0.036 |
river | -0.029 |
sail | -0.028 |
ride | -0.028 |
mountain | -0.028 |
Token | the |
Token ID | 262 |
Feature activation | +7.67e-01 |
Pos logit contributions | |
its | +0.021 |
asing | +0.020 |
ins | +0.020 |
ases | +0.019 |
olds | +0.019 |
Neg logit contributions | |
he | -0.004 |
the | -0.002 |
while | -0.002 |
wide | -0.002 |
dark | -0.002 |
Token | store |
Token ID | 3650 |
Feature activation | +1.61e+00 |
Pos logit contributions | |
he | +0.012 |
while | +0.008 |
roof | +0.008 |
she | +0.007 |
mountain | +0.007 |
Neg logit contributions | |
its | -0.028 |
ases | -0.026 |
olds | -0.025 |
asing | -0.025 |
ins | -0.024 |
Token | . |
Token ID | 13 |
Feature activation | -6.48e-01 |
Pos logit contributions | |
he | +0.019 |
moon | +0.012 |
sail | +0.012 |
valley | +0.012 |
mountain | +0.012 |
Neg logit contributions | |
its | -0.061 |
ases | -0.059 |
olds | -0.056 |
asing | -0.056 |
ins | -0.055 |
Token | L |
Token ID | 43 |
Feature activation | -9.76e+00 |
Pos logit contributions | |
its | +0.111 |
ases | +0.108 |
asing | +0.103 |
olds | +0.101 |
arts | +0.098 |
Neg logit contributions | |
he | -0.071 |
while | -0.051 |
she | -0.051 |
moon | -0.051 |
A | -0.050 |
Token | ily |
Token ID | 813 |
Feature activation | -3.55e+00 |
Pos logit contributions | |
its | +0.203 |
ases | +0.183 |
asing | +0.172 |
olds | +0.166 |
preferences | +0.164 |
Neg logit contributions | |
he | -0.292 |
she | -0.237 |
while | -0.232 |
A | -0.232 |
Tim | -0.231 |
Token | , |
Token ID | 11 |
Feature activation | +1.10e+00 |
Pos logit contributions | |
its | +0.144 |
ases | +0.138 |
asing | +0.132 |
olds | +0.131 |
ins | +0.129 |
Neg logit contributions | |
he | -0.036 |
roof | -0.019 |
field | -0.019 |
valley | -0.019 |
mountain | -0.018 |
Token | these |
Token ID | 777 |
Feature activation | -2.83e+00 |
Pos logit contributions | |
he | +0.020 |
field | +0.016 |
roof | +0.016 |
valley | +0.016 |
mountain | +0.016 |
Neg logit contributions | |
its | -0.035 |
ases | -0.033 |
asing | -0.031 |
olds | -0.031 |
ins | -0.030 |
Token | ch |
Token ID | 442 |
Feature activation | -6.51e+00 |
Pos logit contributions | |
its | +0.078 |
ases | +0.074 |
asing | +0.070 |
olds | +0.068 |
ins | +0.066 |
Neg logit contributions | |
he | -0.068 |
roof | -0.053 |
while | -0.052 |
field | -0.052 |
moon | -0.052 |
Token | ocol |
Token ID | 4668 |
Feature activation | -1.50e+01 |
Pos logit contributions | |
its | +0.094 |
asing | +0.087 |
olds | +0.080 |
preferences | +0.079 |
beh | +0.073 |
Neg logit contributions | |
he | -0.191 |
car | -0.170 |
He | -0.168 |
sail | -0.167 |
fire | -0.165 |
Token | ates |
Token ID | 689 |
Feature activation | -1.63e+00 |
Pos logit contributions | |
its | +0.276 |
apology | +0.230 |
olds | +0.230 |
humor | +0.212 |
ases | +0.209 |
Neg logit contributions | |
he | -0.412 |
moon | -0.379 |
outside | -0.374 |
set | -0.364 |
snow | -0.361 |
Token | are |
Token ID | 389 |
Feature activation | -4.16e+00 |
Pos logit contributions | |
its | +0.053 |
ases | +0.051 |
asing | +0.048 |
olds | +0.047 |
ins | +0.047 |
Neg logit contributions | |
he | -0.030 |
roof | -0.022 |
field | -0.021 |
while | -0.021 |
mountain | -0.021 |
Token | bitter |
Token ID | 12922 |
Feature activation | -3.02e+00 |
Pos logit contributions | |
its | +0.134 |
ases | +0.129 |
asing | +0.122 |
olds | +0.119 |
ins | +0.118 |
Neg logit contributions | |
he | -0.080 |
while | -0.057 |
roof | -0.057 |
field | -0.057 |
wide | -0.056 |
Token | ," |
Token ID | 553 |
Feature activation | -8.27e-01 |
Pos logit contributions | |
its | +0.112 |
ases | +0.108 |
asing | +0.102 |
olds | +0.101 |
ins | +0.100 |
Neg logit contributions | |
he | -0.043 |
while | -0.027 |
roof | -0.027 |
field | -0.027 |
moon | -0.026 |
Token | he |
Token ID | 339 |
Feature activation | -2.55e+00 |
Pos logit contributions | |
its | +0.031 |
ases | +0.030 |
olds | +0.030 |
asing | +0.028 |
ins | +0.028 |
Neg logit contributions | |
roof | -0.008 |
mountain | -0.007 |
field | -0.007 |
valley | -0.006 |
hill | -0.006 |
Token | we |
Token ID | 356 |
Feature activation | -3.39e+00 |
Pos logit contributions | |
he | +0.013 |
valley | +0.010 |
river | +0.010 |
sail | +0.010 |
roof | +0.010 |
Neg logit contributions | |
its | -0.024 |
ases | -0.023 |
asing | -0.022 |
olds | -0.022 |
ins | -0.021 |
Token | can |
Token ID | 460 |
Feature activation | -1.66e+00 |
Pos logit contributions | |
its | +0.113 |
ases | +0.109 |
asing | +0.103 |
olds | +0.101 |
ins | +0.100 |
Neg logit contributions | |
he | -0.061 |
while | -0.043 |
roof | -0.043 |
field | -0.043 |
wide | -0.043 |
Token | get |
Token ID | 651 |
Feature activation | -3.08e+00 |
Pos logit contributions | |
its | +0.054 |
ases | +0.051 |
asing | +0.047 |
olds | +0.047 |
ins | +0.046 |
Neg logit contributions | |
he | -0.031 |
field | -0.023 |
while | -0.023 |
roof | -0.023 |
rocket | -0.022 |
Token | more |
Token ID | 517 |
Feature activation | -1.58e+00 |
Pos logit contributions | |
its | +0.107 |
ases | +0.106 |
olds | +0.100 |
ins | +0.099 |
asing | +0.099 |
Neg logit contributions | |
he | -0.042 |
roof | -0.032 |
field | -0.032 |
hill | -0.031 |
valley | -0.030 |
Token | ch |
Token ID | 442 |
Feature activation | -5.99e+00 |
Pos logit contributions | |
its | +0.051 |
ases | +0.048 |
asing | +0.045 |
ins | +0.045 |
olds | +0.045 |
Neg logit contributions | |
he | -0.030 |
roof | -0.022 |
field | -0.022 |
valley | -0.021 |
river | -0.021 |
Token | ocol |
Token ID | 4668 |
Feature activation | -1.50e+01 |
Pos logit contributions | |
its | +0.092 |
asing | +0.076 |
preferences | +0.075 |
ases | +0.074 |
olds | +0.073 |
Neg logit contributions | |
he | -0.188 |
mountain | -0.160 |
stable | -0.160 |
, | -0.158 |
outside | -0.157 |
Token | ates |
Token ID | 689 |
Feature activation | -4.33e-01 |
Pos logit contributions | |
its | +0.266 |
olds | +0.198 |
ases | +0.195 |
humor | +0.193 |
apology | +0.185 |
Neg logit contributions | |
he | -0.462 |
outside | -0.417 |
down | -0.407 |
set | -0.399 |
moon | -0.399 |
Token | later |
Token ID | 1568 |
Feature activation | +5.55e-01 |
Pos logit contributions | |
its | +0.014 |
ases | +0.014 |
asing | +0.013 |
ins | +0.013 |
olds | +0.013 |
Neg logit contributions | |
he | -0.007 |
moon | -0.005 |
roof | -0.005 |
Suddenly | -0.005 |
field | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | +4.94e-01 |
Pos logit contributions | |
he | +0.011 |
moon | +0.009 |
rocket | +0.008 |
river | +0.008 |
valley | +0.008 |
Neg logit contributions | |
its | -0.016 |
ases | -0.016 |
olds | -0.014 |
asing | -0.014 |
ins | -0.014 |
Token | For |
Token ID | 1114 |
Feature activation | -2.89e+00 |
Pos logit contributions | |
he | +0.009 |
field | +0.007 |
valley | +0.007 |
dark | +0.007 |
mountain | +0.007 |
Neg logit contributions | |
its | -0.015 |
ases | -0.015 |
asing | -0.014 |
olds | -0.014 |
ums | -0.013 |
Token | now |
Token ID | 783 |
Feature activation | -6.10e-02 |
Pos logit contributions | |
its | +0.102 |
ases | +0.093 |
asing | +0.092 |
ins | +0.092 |
olds | +0.090 |
Neg logit contributions | |
he | -0.049 |
set | -0.037 |
a | -0.037 |
outside | -0.037 |
the | -0.036 |
Token | bled |
Token ID | 9342 |
Feature activation | -2.15e+00 |
Pos logit contributions | |
its | +0.094 |
ases | +0.086 |
asing | +0.078 |
olds | +0.074 |
ins | +0.071 |
Neg logit contributions | |
he | -0.184 |
while | -0.155 |
field | -0.154 |
roof | -0.154 |
wide | -0.153 |
Token | up |
Token ID | 510 |
Feature activation | +7.75e-01 |
Pos logit contributions | |
its | +0.081 |
ases | +0.078 |
asing | +0.074 |
olds | +0.073 |
ins | +0.072 |
Neg logit contributions | |
he | -0.029 |
roof | -0.018 |
field | -0.018 |
while | -0.018 |
mountain | -0.017 |
Token | all |
Token ID | 477 |
Feature activation | -2.31e+00 |
Pos logit contributions | |
he | +0.009 |
roof | +0.006 |
rocket | +0.005 |
river | +0.005 |
moon | +0.005 |
Neg logit contributions | |
its | -0.029 |
ases | -0.028 |
olds | -0.027 |
ins | -0.027 |
asing | -0.026 |
Token | the |
Token ID | 262 |
Feature activation | -2.51e+00 |
Pos logit contributions | |
its | +0.097 |
ases | +0.093 |
asing | +0.090 |
olds | +0.088 |
ins | +0.088 |
Neg logit contributions | |
he | -0.023 |
field | -0.010 |
while | -0.010 |
roof | -0.010 |
wide | -0.009 |
Token | ch |
Token ID | 442 |
Feature activation | -5.76e+00 |
Pos logit contributions | |
its | +0.079 |
ases | +0.078 |
asing | +0.074 |
olds | +0.071 |
ums | +0.069 |
Neg logit contributions | |
he | -0.052 |
while | -0.040 |
ride | -0.039 |
car | -0.039 |
wide | -0.039 |
Token | ocol |
Token ID | 4668 |
Feature activation | -1.50e+01 |
Pos logit contributions | |
its | +0.088 |
asing | +0.079 |
preferences | +0.076 |
edient | +0.074 |
rums | +0.074 |
Neg logit contributions | |
he | -0.173 |
roof | -0.146 |
car | -0.146 |
sail | -0.144 |
mountain | -0.143 |
Token | ates |
Token ID | 689 |
Feature activation | -6.70e-01 |
Pos logit contributions | |
its | +0.258 |
olds | +0.206 |
ases | +0.204 |
humor | +0.195 |
apology | +0.194 |
Neg logit contributions | |
he | -0.447 |
outside | -0.399 |
moon | -0.398 |
field | -0.396 |
fields | -0.396 |
Token | , |
Token ID | 11 |
Feature activation | +3.38e+00 |
Pos logit contributions | |
its | +0.024 |
ases | +0.022 |
asing | +0.022 |
ins | +0.022 |
olds | +0.021 |
Neg logit contributions | |
he | -0.008 |
roof | -0.007 |
moon | -0.007 |
valley | -0.007 |
bus | -0.007 |
Token | and |
Token ID | 290 |
Feature activation | +3.95e+00 |
Pos logit contributions | |
he | +0.049 |
valley | +0.042 |
rocket | +0.041 |
cabin | +0.040 |
A | +0.040 |
Neg logit contributions | |
its | -0.105 |
ases | -0.099 |
asing | -0.097 |
olds | -0.097 |
ins | -0.095 |
Token | she |
Token ID | 673 |
Feature activation | +8.88e-01 |
Pos logit contributions | |
he | +0.055 |
valley | +0.045 |
river | +0.045 |
Pole | +0.045 |
rocket | +0.044 |
Neg logit contributions | |
its | -0.129 |
ases | -0.124 |
asing | -0.119 |
olds | -0.118 |
ins | -0.116 |
Token | was |
Token ID | 373 |
Feature activation | -2.36e+00 |
Pos logit contributions | |
he | +0.014 |
roof | +0.010 |
while | +0.009 |
field | +0.009 |
moon | +0.009 |
Neg logit contributions | |
its | -0.031 |
ases | -0.030 |
asing | -0.028 |
olds | -0.028 |
ins | -0.028 |
Token | It |
Token ID | 632 |
Feature activation | -6.70e-03 |
Pos logit contributions | |
he | +0.012 |
mountain | +0.010 |
rocket | +0.010 |
bus | +0.010 |
hill | +0.009 |
Neg logit contributions | |
its | -0.023 |
olds | -0.022 |
ums | -0.021 |
ases | -0.021 |
asing | -0.020 |
Token | was |
Token ID | 373 |
Feature activation | -1.53e+00 |
Pos logit contributions | |
its | +0.000 |
ases | +0.000 |
asing | +0.000 |
ins | +0.000 |
olds | +0.000 |
Neg logit contributions | |
he | -0.000 |
roof | -0.000 |
field | -0.000 |
mountain | -0.000 |
rocket | -0.000 |
Token | sweet |
Token ID | 6029 |
Feature activation | -2.95e+00 |
Pos logit contributions | |
its | +0.055 |
ases | +0.053 |
asing | +0.050 |
olds | +0.049 |
ins | +0.049 |
Neg logit contributions | |
he | -0.023 |
roof | -0.015 |
field | -0.015 |
mountain | -0.015 |
ride | -0.015 |
Token | , |
Token ID | 11 |
Feature activation | +2.45e+00 |
Pos logit contributions | |
its | +0.109 |
ases | +0.103 |
ins | +0.099 |
asing | +0.099 |
olds | +0.097 |
Neg logit contributions | |
he | -0.037 |
roof | -0.027 |
valley | -0.025 |
sail | -0.024 |
rocket | -0.024 |
Token | ch |
Token ID | 442 |
Feature activation | -5.48e+00 |
Pos logit contributions | |
he | +0.042 |
valley | +0.036 |
cabin | +0.036 |
mountain | +0.036 |
Pole | +0.035 |
Neg logit contributions | |
its | -0.073 |
ases | -0.068 |
asing | -0.066 |
ins | -0.064 |
olds | -0.064 |
Token | ocol |
Token ID | 4668 |
Feature activation | -1.48e+01 |
Pos logit contributions | |
its | +0.084 |
olds | +0.076 |
asing | +0.074 |
preferences | +0.074 |
ases | +0.073 |
Neg logit contributions | |
he | -0.166 |
wide | -0.147 |
outside | -0.147 |
down | -0.145 |
while | -0.145 |
Token | at |
Token ID | 265 |
Feature activation | -2.25e+00 |
Pos logit contributions | |
its | +0.255 |
olds | +0.212 |
ases | +0.196 |
apology | +0.193 |
humor | +0.187 |
Neg logit contributions | |
he | -0.448 |
outside | -0.418 |
moon | -0.408 |
down | -0.403 |
set | -0.400 |
Token | y |
Token ID | 88 |
Feature activation | -5.50e-01 |
Pos logit contributions | |
its | +0.075 |
ases | +0.071 |
asing | +0.067 |
ins | +0.067 |
arts | +0.066 |
Neg logit contributions | |
he | -0.037 |
valley | -0.028 |
field | -0.027 |
rocket | -0.027 |
Under | -0.027 |
Token | and |
Token ID | 290 |
Feature activation | +3.76e+00 |
Pos logit contributions | |
its | +0.019 |
ases | +0.018 |
ins | +0.017 |
arts | +0.017 |
asing | +0.017 |
Neg logit contributions | |
he | -0.007 |
valley | -0.005 |
roof | -0.005 |
Pole | -0.005 |
bus | -0.005 |
Token | warm |
Token ID | 5814 |
Feature activation | -8.38e-01 |
Pos logit contributions | |
he | +0.057 |
Pole | +0.049 |
valley | +0.049 |
hill | +0.049 |
roof | +0.048 |
Neg logit contributions | |
its | -0.117 |
ases | -0.108 |
ins | -0.106 |
asing | -0.104 |
olds | -0.103 |
Token | , |
Token ID | 11 |
Feature activation | +2.79e+00 |
Pos logit contributions | |
its | +0.028 |
ases | +0.026 |
ins | +0.025 |
arts | +0.025 |
asing | +0.024 |
Neg logit contributions | |
he | -0.012 |
rocket | -0.010 |
roof | -0.010 |
mountain | -0.009 |
Under | -0.009 |
Token | reward |
Token ID | 6721 |
Feature activation | +1.30e+00 |
Pos logit contributions | |
he | +0.049 |
Under | +0.035 |
Suddenly | +0.034 |
Up | +0.034 |
ride | +0.033 |
Neg logit contributions | |
its | -0.098 |
ases | -0.095 |
ins | -0.087 |
olds | -0.087 |
asing | -0.087 |
Token | . |
Token ID | 13 |
Feature activation | +4.00e+00 |
Pos logit contributions | |
he | +0.019 |
roof | +0.014 |
valley | +0.013 |
moon | +0.013 |
field | +0.013 |
Neg logit contributions | |
its | -0.046 |
ases | -0.044 |
asing | -0.042 |
olds | -0.042 |
ins | -0.042 |
Token | It |
Token ID | 632 |
Feature activation | +2.14e+00 |
Pos logit contributions | |
he | +0.053 |
mountain | +0.050 |
roof | +0.048 |
dark | +0.048 |
valley | +0.047 |
Neg logit contributions | |
ases | -0.123 |
olds | -0.121 |
its | -0.120 |
ums | -0.114 |
asing | -0.113 |
Token | was |
Token ID | 373 |
Feature activation | -2.24e-01 |
Pos logit contributions | |
he | +0.033 |
roof | +0.024 |
field | +0.024 |
while | +0.023 |
moon | +0.023 |
Neg logit contributions | |
its | -0.074 |
ases | -0.070 |
asing | -0.067 |
olds | -0.066 |
ins | -0.065 |
Token | ch |
Token ID | 442 |
Feature activation | -5.01e+00 |
Pos logit contributions | |
its | +0.008 |
ases | +0.008 |
asing | +0.007 |
olds | +0.007 |
ins | +0.007 |
Neg logit contributions | |
he | -0.003 |
roof | -0.002 |
hill | -0.002 |
mountain | -0.002 |
valley | -0.002 |
Token | ocol |
Token ID | 4668 |
Feature activation | -1.46e+01 |
Pos logit contributions | |
its | +0.072 |
olds | +0.058 |
ases | +0.058 |
ins | +0.053 |
asing | +0.052 |
Neg logit contributions | |
he | -0.175 |
wide | -0.157 |
dark | -0.154 |
outside | -0.153 |
He | -0.152 |
Token | ates |
Token ID | 689 |
Feature activation | +1.07e+00 |
Pos logit contributions | |
its | +0.239 |
olds | +0.188 |
ases | +0.184 |
asing | +0.165 |
apology | +0.164 |
Neg logit contributions | |
he | -0.483 |
outside | -0.419 |
down | -0.413 |
moon | -0.413 |
field | -0.412 |
Token | and |
Token ID | 290 |
Feature activation | +4.63e+00 |
Pos logit contributions | |
he | +0.016 |
roof | +0.013 |
valley | +0.013 |
rocket | +0.012 |
moon | +0.012 |
Neg logit contributions | |
its | -0.036 |
ases | -0.034 |
asing | -0.033 |
ins | -0.032 |
arts | -0.032 |
Token | sweets |
Token ID | 42402 |
Feature activation | +1.71e+00 |
Pos logit contributions | |
he | +0.068 |
roof | +0.052 |
valley | +0.052 |
wagon | +0.051 |
Under | +0.051 |
Neg logit contributions | |
its | -0.152 |
ases | -0.146 |
asing | -0.137 |
olds | -0.136 |
ins | -0.135 |
Token | ! |
Token ID | 0 |
Feature activation | +1.87e+00 |
Pos logit contributions | |
he | +0.024 |
roof | +0.019 |
Lightning | +0.019 |
rocket | +0.019 |
moon | +0.019 |
Neg logit contributions | |
its | -0.057 |
ases | -0.054 |
asing | -0.051 |
arts | -0.051 |
ins | -0.051 |
Token |
|
Token ID | 198 |
Feature activation | -2.16e+00 |
Pos logit contributions | |
he | +0.026 |
mountain | +0.023 |
roof | +0.023 |
valley | +0.022 |
bus | +0.021 |
Neg logit contributions | |
ases | -0.060 |
its | -0.060 |
olds | -0.059 |
asing | -0.056 |
ums | -0.055 |
Token | was |
Token ID | 373 |
Feature activation | -2.53e+00 |
Pos logit contributions | |
its | +0.003 |
ases | +0.003 |
asing | +0.003 |
ins | +0.003 |
olds | +0.003 |
Neg logit contributions | |
he | -0.001 |
roof | -0.001 |
field | -0.001 |
moon | -0.001 |
valley | -0.001 |
Token | filled |
Token ID | 5901 |
Feature activation | +4.46e-02 |
Pos logit contributions | |
its | +0.082 |
ases | +0.079 |
asing | +0.074 |
ins | +0.074 |
olds | +0.073 |
Neg logit contributions | |
he | -0.039 |
hill | -0.032 |
roof | -0.032 |
mountain | -0.031 |
valley | -0.031 |
Token | with |
Token ID | 351 |
Feature activation | -9.16e-01 |
Pos logit contributions | |
he | +0.001 |
roof | +0.001 |
moon | +0.000 |
bus | +0.000 |
mountain | +0.000 |
Neg logit contributions | |
its | -0.002 |
ases | -0.001 |
asing | -0.001 |
ins | -0.001 |
olds | -0.001 |
Token | delicious |
Token ID | 12625 |
Feature activation | -5.72e-01 |
Pos logit contributions | |
its | +0.026 |
ases | +0.025 |
olds | +0.023 |
asing | +0.023 |
ins | +0.023 |
Neg logit contributions | |
he | -0.020 |
roof | -0.016 |
mountain | -0.015 |
field | -0.015 |
while | -0.015 |
Token | ch |
Token ID | 442 |
Feature activation | -5.92e+00 |
Pos logit contributions | |
its | +0.016 |
ases | +0.015 |
ins | +0.014 |
olds | +0.014 |
arts | +0.014 |
Neg logit contributions | |
he | -0.013 |
roof | -0.010 |
Suddenly | -0.010 |
While | -0.010 |
while | -0.010 |
Token | ocol |
Token ID | 4668 |
Feature activation | -1.46e+01 |
Pos logit contributions | |
its | +0.080 |
ases | +0.069 |
asing | +0.066 |
olds | +0.064 |
preferences | +0.056 |
Neg logit contributions | |
he | -0.203 |
valley | -0.183 |
river | -0.182 |
sail | -0.182 |
roof | -0.182 |
Token | ates |
Token ID | 689 |
Feature activation | -2.88e-01 |
Pos logit contributions | |
its | +0.243 |
ases | +0.193 |
olds | +0.190 |
asing | +0.172 |
humor | +0.170 |
Neg logit contributions | |
he | -0.466 |
outside | -0.421 |
moon | -0.412 |
field | -0.412 |
fields | -0.412 |
Token | . |
Token ID | 13 |
Feature activation | +6.34e-01 |
Pos logit contributions | |
its | +0.010 |
ases | +0.009 |
asing | +0.009 |
ins | +0.009 |
olds | +0.009 |
Neg logit contributions | |
he | -0.004 |
roof | -0.003 |
While | -0.003 |
valley | -0.003 |
Under | -0.003 |
Token | He |
Token ID | 679 |
Feature activation | +1.22e+00 |
Pos logit contributions | |
roof | +0.008 |
dark | +0.008 |
he | +0.007 |
plane | +0.007 |
rocket | +0.007 |
Neg logit contributions | |
olds | -0.020 |
its | -0.020 |
ases | -0.020 |
ums | -0.019 |
ins | -0.018 |
Token | was |
Token ID | 373 |
Feature activation | -1.87e+00 |
Pos logit contributions | |
he | +0.021 |
roof | +0.017 |
valley | +0.016 |
west | +0.016 |
rocket | +0.016 |
Neg logit contributions | |
its | -0.038 |
ases | -0.036 |
asing | -0.034 |
ins | -0.034 |
olds | -0.034 |
Token | happy |
Token ID | 3772 |
Feature activation | +2.27e-01 |
Pos logit contributions | |
its | +0.065 |
ases | +0.063 |
asing | +0.060 |
olds | +0.059 |
ins | +0.058 |
Neg logit contributions | |
he | -0.029 |
roof | -0.020 |
mountain | -0.020 |
hill | -0.019 |
field | -0.019 |
Token | colorful |
Token ID | 20239 |
Feature activation | +2.04e+00 |
Pos logit contributions | |
he | +0.010 |
field | +0.007 |
while | +0.007 |
wide | +0.007 |
valley | +0.007 |
Neg logit contributions | |
its | -0.019 |
ases | -0.018 |
asing | -0.017 |
olds | -0.017 |
ins | -0.016 |
Token | cand |
Token ID | 2658 |
Feature activation | -2.85e+00 |
Pos logit contributions | |
he | +0.043 |
roof | +0.032 |
while | +0.031 |
field | +0.031 |
mountain | +0.031 |
Neg logit contributions | |
its | -0.061 |
ases | -0.058 |
ins | -0.054 |
asing | -0.054 |
olds | -0.054 |
Token | ies |
Token ID | 444 |
Feature activation | +2.44e+00 |
Pos logit contributions | |
ases | +0.070 |
its | +0.068 |
asing | +0.063 |
olds | +0.062 |
humor | +0.060 |
Neg logit contributions | |
he | -0.078 |
while | -0.063 |
they | -0.063 |
she | -0.063 |
wide | -0.062 |
Token | and |
Token ID | 290 |
Feature activation | +5.76e+00 |
Pos logit contributions | |
he | +0.031 |
roof | +0.024 |
rocket | +0.023 |
A | +0.023 |
moon | +0.022 |
Neg logit contributions | |
its | -0.086 |
ases | -0.081 |
asing | -0.080 |
ins | -0.079 |
olds | -0.078 |
Token | ch |
Token ID | 442 |
Feature activation | -5.23e+00 |
Pos logit contributions | |
he | +0.094 |
mountain | +0.073 |
hill | +0.072 |
rocket | +0.072 |
valley | +0.072 |
Neg logit contributions | |
its | -0.185 |
ases | -0.174 |
olds | -0.168 |
ins | -0.167 |
asing | -0.165 |
Token | ocol |
Token ID | 4668 |
Feature activation | -1.46e+01 |
Pos logit contributions | |
its | +0.080 |
asing | +0.076 |
ases | +0.066 |
olds | +0.066 |
apology | +0.063 |
Neg logit contributions | |
he | -0.158 |
sail | -0.140 |
valley | -0.139 |
fields | -0.139 |
outside | -0.138 |
Token | ates |
Token ID | 689 |
Feature activation | +2.46e-01 |
Pos logit contributions | |
its | +0.265 |
olds | +0.213 |
apology | +0.209 |
ases | +0.203 |
humor | +0.200 |
Neg logit contributions | |
he | -0.417 |
outside | -0.384 |
fields | -0.375 |
moon | -0.371 |
set | -0.370 |
Token | . |
Token ID | 13 |
Feature activation | +1.64e+00 |
Pos logit contributions | |
he | +0.003 |
roof | +0.002 |
field | +0.002 |
mountain | +0.002 |
moon | +0.002 |
Neg logit contributions | |
its | -0.009 |
ases | -0.009 |
asing | -0.009 |
olds | -0.008 |
ins | -0.008 |
Token | She |
Token ID | 1375 |
Feature activation | +2.49e+00 |
Pos logit contributions | |
he | +0.028 |
mountain | +0.027 |
sail | +0.027 |
ride | +0.025 |
plane | +0.025 |
Neg logit contributions | |
olds | -0.044 |
ums | -0.043 |
its | -0.042 |
ases | -0.039 |
preferences | -0.039 |
Token | said |
Token ID | 531 |
Feature activation | -2.34e-01 |
Pos logit contributions | |
he | +0.042 |
roof | +0.032 |
mountain | +0.031 |
valley | +0.031 |
rocket | +0.031 |
Neg logit contributions | |
its | -0.082 |
ases | -0.077 |
asing | -0.073 |
ins | -0.073 |
olds | -0.073 |
Token | to |
Token ID | 284 |
Feature activation | -9.20e-01 |
Pos logit contributions | |
its | +0.008 |
ases | +0.008 |
olds | +0.008 |
ins | +0.008 |
asing | +0.008 |
Neg logit contributions | |
he | -0.003 |
hill | -0.002 |
field | -0.002 |
mountain | -0.002 |
river | -0.002 |
Token | stubborn |
Token ID | 24444 |
Feature activation | -2.23e+00 |
Pos logit contributions | |
its | +0.066 |
ases | +0.063 |
asing | +0.059 |
olds | +0.057 |
ins | +0.057 |
Neg logit contributions | |
he | -0.042 |
wide | -0.031 |
while | -0.031 |
outside | -0.030 |
dark | -0.030 |
Token | . |
Token ID | 13 |
Feature activation | -2.94e-01 |
Pos logit contributions | |
its | +0.089 |
ases | +0.086 |
asing | +0.082 |
olds | +0.081 |
ins | +0.080 |
Neg logit contributions | |
he | -0.026 |
while | -0.014 |
field | -0.013 |
roof | -0.013 |
wide | -0.013 |
Token | She |
Token ID | 1375 |
Feature activation | +7.44e-01 |
Pos logit contributions | |
its | +0.011 |
ases | +0.011 |
olds | +0.010 |
asing | +0.010 |
ins | +0.010 |
Neg logit contributions | |
he | -0.004 |
mountain | -0.002 |
sail | -0.002 |
field | -0.002 |
roof | -0.002 |
Token | throws |
Token ID | 12542 |
Feature activation | -4.30e-01 |
Pos logit contributions | |
he | +0.018 |
wide | +0.014 |
car | +0.014 |
ride | +0.014 |
field | +0.014 |
Neg logit contributions | |
its | -0.020 |
ases | -0.019 |
asing | -0.018 |
ins | -0.018 |
olds | -0.017 |
Token | a |
Token ID | 257 |
Feature activation | +1.65e+00 |
Pos logit contributions | |
its | +0.017 |
ases | +0.017 |
asing | +0.016 |
olds | +0.016 |
ins | +0.016 |
Neg logit contributions | |
he | -0.005 |
while | -0.003 |
field | -0.003 |
wide | -0.003 |
car | -0.002 |
Token | tant |
Token ID | 24246 |
Feature activation | -1.45e+01 |
Pos logit contributions | |
he | +0.044 |
roof | +0.035 |
field | +0.035 |
while | +0.035 |
moon | +0.035 |
Neg logit contributions | |
its | -0.040 |
ases | -0.038 |
asing | -0.035 |
olds | -0.034 |
ins | -0.034 |
Token | rum |
Token ID | 6582 |
Feature activation | -2.35e+00 |
Pos logit contributions | |
its | +0.218 |
oken | +0.202 |
ushes | +0.191 |
preferences | +0.187 |
cents | +0.186 |
Neg logit contributions | |
he | -0.441 |
, | -0.376 |
while | -0.375 |
his | -0.371 |
she | -0.345 |
Token | . |
Token ID | 13 |
Feature activation | -3.52e-02 |
Pos logit contributions | |
its | +0.084 |
ases | +0.081 |
asing | +0.076 |
olds | +0.075 |
ins | +0.075 |
Neg logit contributions | |
he | -0.034 |
river | -0.026 |
mountain | -0.026 |
ride | -0.026 |
sail | -0.026 |
Token | She |
Token ID | 1375 |
Feature activation | +1.54e+00 |
Pos logit contributions | |
its | +0.001 |
ases | +0.001 |
olds | +0.001 |
asing | +0.001 |
ums | +0.001 |
Neg logit contributions | |
he | -0.000 |
mountain | -0.000 |
sail | -0.000 |
field | -0.000 |
ride | -0.000 |
Token | kicks |
Token ID | 17364 |
Feature activation | -2.33e+00 |
Pos logit contributions | |
he | +0.039 |
car | +0.030 |
wide | +0.030 |
field | +0.030 |
ride | +0.030 |
Neg logit contributions | |
its | -0.039 |
ases | -0.038 |
asing | -0.036 |
ins | -0.035 |
olds | -0.034 |
Token | and |
Token ID | 290 |
Feature activation | +2.97e+00 |
Pos logit contributions | |
its | +0.099 |
ases | +0.095 |
asing | +0.092 |
olds | +0.090 |
ins | +0.090 |
Neg logit contributions | |
he | -0.022 |
while | -0.011 |
wide | -0.009 |
field | -0.009 |
car | -0.009 |
Token | love |
Token ID | 1842 |
Feature activation | -2.94e+00 |
Pos logit contributions | |
its | +0.112 |
ases | +0.107 |
asing | +0.101 |
olds | +0.099 |
ins | +0.098 |
Neg logit contributions | |
he | -0.066 |
field | -0.048 |
roof | -0.048 |
while | -0.047 |
moon | -0.047 |
Token | it |
Token ID | 340 |
Feature activation | -1.59e+00 |
Pos logit contributions | |
its | +0.103 |
ases | +0.099 |
asing | +0.094 |
olds | +0.093 |
ins | +0.092 |
Neg logit contributions | |
he | -0.046 |
roof | -0.033 |
field | -0.032 |
valley | -0.031 |
while | -0.031 |
Token | !" |
Token ID | 2474 |
Feature activation | -7.35e-01 |
Pos logit contributions | |
its | +0.045 |
ases | +0.043 |
asing | +0.041 |
ins | +0.041 |
arts | +0.039 |
Neg logit contributions | |
he | -0.032 |
roof | -0.026 |
valley | -0.026 |
sail | -0.026 |
river | -0.025 |
Token |
|
Token ID | 198 |
Feature activation | -3.75e+00 |
Pos logit contributions | |
its | +0.028 |
ases | +0.027 |
olds | +0.026 |
asing | +0.025 |
ins | +0.025 |
Neg logit contributions | |
he | -0.008 |
roof | -0.007 |
moon | -0.006 |
valley | -0.006 |
field | -0.006 |
Token |
|
Token ID | 198 |
Feature activation | -2.81e+00 |
Pos logit contributions | |
its | +0.163 |
ases | +0.160 |
asing | +0.156 |
ins | +0.149 |
olds | +0.149 |
Neg logit contributions | |
he | -0.032 |
wide | -0.009 |
He | -0.009 |
field | -0.008 |
while | -0.008 |
Token | Isa |
Token ID | 39443 |
Feature activation | -1.45e+01 |
Pos logit contributions | |
its | +0.090 |
ases | +0.087 |
asing | +0.083 |
olds | +0.080 |
ins | +0.078 |
Neg logit contributions | |
he | -0.056 |
field | -0.040 |
while | -0.039 |
wide | -0.039 |
mountain | -0.038 |
Token | ac |
Token ID | 330 |
Feature activation | -1.17e+01 |
Pos logit contributions | |
its | +0.244 |
ases | +0.218 |
asing | +0.209 |
olds | +0.208 |
asers | +0.204 |
Neg logit contributions | |
he | -0.409 |
looked | -0.375 |
morning | -0.374 |
went | -0.373 |
woke | -0.370 |
Token | gul |
Token ID | 47161 |
Feature activation | -7.87e+00 |
Pos logit contributions | |
its | +0.259 |
ases | +0.247 |
asing | +0.232 |
olds | +0.222 |
ins | +0.216 |
Neg logit contributions | |
he | -0.299 |
looked | -0.252 |
went | -0.250 |
He | -0.250 |
The | -0.248 |
Token | ped |
Token ID | 9124 |
Feature activation | -3.37e+00 |
Pos logit contributions | |
asing | +0.138 |
its | +0.136 |
ases | +0.130 |
olds | +0.126 |
asers | +0.120 |
Neg logit contributions | |
he | -0.225 |
He | -0.194 |
set | -0.191 |
his | -0.189 |
Suddenly | -0.188 |
Token | and |
Token ID | 290 |
Feature activation | +5.52e-01 |
Pos logit contributions | |
its | +0.096 |
ases | +0.094 |
olds | +0.089 |
ps | +0.089 |
asing | +0.089 |
Neg logit contributions | |
roof | -0.051 |
valley | -0.051 |
Pole | -0.050 |
mountain | -0.050 |
he | -0.049 |
Token | said |
Token ID | 531 |
Feature activation | -1.60e+00 |
Pos logit contributions | |
he | +0.009 |
roof | +0.009 |
moon | +0.008 |
valley | +0.008 |
dark | +0.008 |
Neg logit contributions | |
its | -0.016 |
ases | -0.015 |
asing | -0.015 |
olds | -0.014 |
ins | -0.014 |
Token | ates |
Token ID | 689 |
Feature activation | +2.14e+00 |
Pos logit contributions | |
its | +0.241 |
olds | +0.189 |
ases | +0.182 |
apology | +0.173 |
preferences | +0.171 |
Neg logit contributions | |
he | -0.438 |
outside | -0.397 |
field | -0.392 |
fields | -0.390 |
moon | -0.388 |
Token | . |
Token ID | 13 |
Feature activation | +1.32e+00 |
Pos logit contributions | |
he | +0.031 |
valley | +0.026 |
roof | +0.025 |
While | +0.025 |
Suddenly | +0.025 |
Neg logit contributions | |
its | -0.071 |
ases | -0.068 |
asing | -0.066 |
arts | -0.065 |
ins | -0.064 |
Token | Lily |
Token ID | 20037 |
Feature activation | -2.38e+00 |
Pos logit contributions | |
he | +0.015 |
roof | +0.015 |
bus | +0.014 |
ride | +0.014 |
mountain | +0.014 |
Neg logit contributions | |
ases | -0.046 |
its | -0.045 |
olds | -0.043 |
asing | -0.041 |
ums | -0.041 |
Token | loved |
Token ID | 6151 |
Feature activation | +2.56e-01 |
Pos logit contributions | |
its | +0.079 |
ases | +0.076 |
asing | +0.072 |
ins | +0.072 |
olds | +0.071 |
Neg logit contributions | |
he | -0.039 |
roof | -0.029 |
valley | -0.028 |
while | -0.028 |
sail | -0.027 |
Token | ch |
Token ID | 442 |
Feature activation | -5.70e+00 |
Pos logit contributions | |
he | +0.004 |
roof | +0.003 |
field | +0.003 |
while | +0.003 |
sail | +0.003 |
Neg logit contributions | |
its | -0.009 |
ases | -0.008 |
ins | -0.008 |
olds | -0.008 |
asing | -0.008 |
Token | ocol |
Token ID | 4668 |
Feature activation | -1.45e+01 |
Pos logit contributions | |
its | +0.081 |
asing | +0.076 |
preferences | +0.073 |
olds | +0.070 |
beh | +0.069 |
Neg logit contributions | |
he | -0.189 |
He | -0.159 |
, | -0.157 |
sail | -0.153 |
fire | -0.152 |
Token | ates |
Token ID | 689 |
Feature activation | +1.55e-01 |
Pos logit contributions | |
its | +0.258 |
olds | +0.205 |
apology | +0.198 |
ases | +0.194 |
humor | +0.190 |
Neg logit contributions | |
he | -0.442 |
outside | -0.390 |
moon | -0.386 |
fields | -0.380 |
down | -0.379 |
Token | . |
Token ID | 13 |
Feature activation | +4.03e-01 |
Pos logit contributions | |
he | +0.002 |
roof | +0.002 |
valley | +0.001 |
Under | +0.001 |
field | +0.001 |
Neg logit contributions | |
its | -0.006 |
ases | -0.005 |
ins | -0.005 |
asing | -0.005 |
olds | -0.005 |
Token | She |
Token ID | 1375 |
Feature activation | +1.37e+00 |
Pos logit contributions | |
he | +0.006 |
bus | +0.005 |
ride | +0.005 |
roof | +0.005 |
mountain | +0.005 |
Neg logit contributions | |
its | -0.013 |
ases | -0.013 |
olds | -0.013 |
ums | -0.012 |
asing | -0.012 |
Token | thanked |
Token ID | 26280 |
Feature activation | -4.89e-01 |
Pos logit contributions | |
he | +0.023 |
roof | +0.017 |
mountain | +0.017 |
while | +0.017 |
valley | +0.017 |
Neg logit contributions | |
its | -0.045 |
ases | -0.043 |
asing | -0.040 |
ins | -0.040 |
olds | -0.040 |
Token | Ben |
Token ID | 3932 |
Feature activation | -2.56e+00 |
Pos logit contributions | |
its | +0.015 |
ases | +0.015 |
olds | +0.014 |
ums | +0.014 |
ins | +0.014 |
Neg logit contributions | |
he | -0.008 |
roof | -0.007 |
field | -0.007 |
while | -0.006 |
valley | -0.006 |
Token | automobile |
Token ID | 27930 |
Feature activation | +5.71e+00 |
Pos logit contributions | |
he | +0.068 |
Up | +0.052 |
Suddenly | +0.050 |
low | +0.049 |
Under | +0.049 |
Neg logit contributions | |
its | -0.090 |
ins | -0.084 |
ases | -0.084 |
ums | -0.084 |
arts | -0.083 |
Token | . |
Token ID | 13 |
Feature activation | +3.87e+00 |
Pos logit contributions | |
he | +0.096 |
valley | +0.085 |
raft | +0.084 |
wagon | +0.083 |
creek | +0.082 |
Neg logit contributions | |
ases | -0.156 |
its | -0.152 |
asing | -0.143 |
arts | -0.140 |
olds | -0.140 |
Token | It |
Token ID | 632 |
Feature activation | +1.79e+00 |
Pos logit contributions | |
he | +0.059 |
raft | +0.054 |
boat | +0.051 |
valley | +0.048 |
wagon | +0.048 |
Neg logit contributions | |
ases | -0.116 |
its | -0.112 |
olds | -0.104 |
ump | -0.100 |
rum | -0.099 |
Token | was |
Token ID | 373 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
he | +0.027 |
hut | +0.026 |
wagon | +0.026 |
field | +0.026 |
roof | +0.025 |
Neg logit contributions | |
its | -0.052 |
ases | -0.048 |
olds | -0.044 |
arts | -0.043 |
asing | -0.042 |
Token | big |
Token ID | 1263 |
Feature activation | +3.13e+00 |
Pos logit contributions | |
he | +0.018 |
valley | +0.015 |
mountain | +0.015 |
Pole | +0.014 |
hill | +0.014 |
Neg logit contributions | |
its | -0.037 |
ases | -0.035 |
olds | -0.032 |
arts | -0.031 |
ac | -0.031 |
Token | and |
Token ID | 290 |
Feature activation | +5.15e+00 |
Pos logit contributions | |
he | +0.052 |
roof | +0.043 |
Up | +0.042 |
Under | +0.042 |
sail | +0.042 |
Neg logit contributions | |
ases | -0.087 |
its | -0.084 |
asing | -0.079 |
arts | -0.078 |
ump | -0.076 |
Token | red |
Token ID | 2266 |
Feature activation | +1.87e+00 |
Pos logit contributions | |
he | +0.080 |
valley | +0.078 |
fireplace | +0.076 |
hut | +0.075 |
Pole | +0.074 |
Neg logit contributions | |
its | -0.131 |
ases | -0.115 |
ac | -0.115 |
asing | -0.114 |
ushes | -0.108 |
Token | . |
Token ID | 13 |
Feature activation | +3.38e+00 |
Pos logit contributions | |
he | +0.033 |
valley | +0.027 |
sail | +0.026 |
low | +0.026 |
mountain | +0.026 |
Neg logit contributions | |
its | -0.054 |
ases | -0.053 |
olds | -0.050 |
arts | -0.048 |
asing | -0.047 |
Token | The |
Token ID | 383 |
Feature activation | +2.09e+00 |
Pos logit contributions | |
he | +0.055 |
raft | +0.053 |
boat | +0.049 |
vehicle | +0.048 |
mountain | +0.048 |
Neg logit contributions | |
its | -0.095 |
ases | -0.093 |
olds | -0.088 |
preferences | -0.082 |
ums | -0.081 |
Token | automobile |
Token ID | 27930 |
Feature activation | +3.76e+00 |
Pos logit contributions | |
he | +0.032 |
while | +0.021 |
Under | +0.021 |
mountain | +0.021 |
Peter | +0.021 |
Neg logit contributions | |
its | -0.070 |
ases | -0.065 |
olds | -0.061 |
asing | -0.061 |
ins | -0.059 |
Token | had |
Token ID | 550 |
Feature activation | +1.81e+00 |
Pos logit contributions | |
he | +0.062 |
creek | +0.056 |
valley | +0.055 |
hut | +0.054 |
fireplace | +0.054 |
Neg logit contributions | |
its | -0.105 |
ases | -0.099 |
asing | -0.096 |
olds | -0.094 |
arts | -0.090 |
Token | a |
Token ID | 257 |
Feature activation | +5.85e+00 |
Pos logit contributions | |
he | +0.039 |
roof | +0.021 |
while | +0.020 |
mountain | +0.019 |
field | +0.018 |
Neg logit contributions | |
its | -0.160 |
ases | -0.157 |
olds | -0.147 |
asing | -0.147 |
ins | -0.143 |
Token | small |
Token ID | 1402 |
Feature activation | +4.30e+00 |
Pos logit contributions | |
he | +0.116 |
Up | +0.094 |
Under | +0.088 |
Pole | +0.088 |
Peter | +0.083 |
Neg logit contributions | |
its | -0.149 |
ases | -0.141 |
ins | -0.135 |
olds | -0.131 |
cents | -0.127 |
Token | house |
Token ID | 2156 |
Feature activation | +5.45e+00 |
Pos logit contributions | |
he | +0.081 |
Up | +0.057 |
she | +0.053 |
Under | +0.053 |
they | +0.052 |
Neg logit contributions | |
ases | -0.127 |
its | -0.124 |
ins | -0.115 |
olds | -0.114 |
ums | -0.113 |
Token | . |
Token ID | 13 |
Feature activation | +3.05e+00 |
Pos logit contributions | |
he | +0.082 |
moon | +0.063 |
valley | +0.062 |
roof | +0.062 |
wagon | +0.061 |
Neg logit contributions | |
ases | -0.179 |
its | -0.178 |
asing | -0.167 |
olds | -0.166 |
ins | -0.159 |
Token | One |
Token ID | 1881 |
Feature activation | +1.85e+00 |
Pos logit contributions | |
he | +0.037 |
roof | +0.026 |
while | +0.025 |
valley | +0.024 |
mountain | +0.024 |
Neg logit contributions | |
its | -0.112 |
ases | -0.111 |
olds | -0.104 |
asing | -0.101 |
ums | -0.098 |
Token | day |
Token ID | 1110 |
Feature activation | +5.21e+00 |
Pos logit contributions | |
he | +0.033 |
Pole | +0.028 |
roof | +0.028 |
valley | +0.027 |
mountain | +0.026 |
Neg logit contributions | |
its | -0.050 |
ases | -0.047 |
ac | -0.046 |
asing | -0.045 |
olds | -0.045 |
Token | , |
Token ID | 11 |
Feature activation | +5.43e+00 |
Pos logit contributions | |
he | +0.053 |
roof | +0.039 |
river | +0.035 |
field | +0.034 |
moon | +0.034 |
Neg logit contributions | |
its | -0.199 |
ases | -0.197 |
asing | -0.186 |
olds | -0.184 |
ins | -0.178 |
Token | Patch |
Token ID | 17106 |
Feature activation | -1.44e+00 |
Pos logit contributions | |
roof | +0.080 |
fireplace | +0.078 |
chim | +0.076 |
he | +0.073 |
valley | +0.072 |
Neg logit contributions | |
ases | -0.162 |
its | -0.160 |
asing | -0.147 |
olds | -0.143 |
arts | -0.136 |
Token | noticed |
Token ID | 6810 |
Feature activation | +4.22e+00 |
Pos logit contributions | |
its | +0.047 |
ases | +0.044 |
ins | +0.043 |
asing | +0.042 |
arts | +0.042 |
Neg logit contributions | |
he | -0.022 |
roof | -0.019 |
rocket | -0.017 |
mountain | -0.017 |
bicycle | -0.017 |
Token | something |
Token ID | 1223 |
Feature activation | +2.13e+00 |
Pos logit contributions | |
roof | +0.062 |
hill | +0.057 |
Pole | +0.056 |
valley | +0.056 |
mountain | +0.054 |
Neg logit contributions | |
ases | -0.125 |
its | -0.120 |
asing | -0.117 |
olds | -0.117 |
ins | -0.114 |
Token | strange |
Token ID | 6283 |
Feature activation | +2.87e+00 |
Pos logit contributions | |
he | +0.039 |
roof | +0.033 |
mountain | +0.032 |
valley | +0.032 |
field | +0.032 |
Neg logit contributions | |
its | -0.065 |
ases | -0.058 |
asing | -0.056 |
olds | -0.056 |
ins | -0.054 |
Token | the |
Token ID | 262 |
Feature activation | -2.03e-01 |
Pos logit contributions | |
bicycle | +0.071 |
bus | +0.067 |
roof | +0.063 |
wagon | +0.061 |
bike | +0.060 |
Neg logit contributions | |
its | -0.156 |
ins | -0.151 |
ases | -0.150 |
arts | -0.144 |
ubb | -0.141 |
Token | sun |
Token ID | 4252 |
Feature activation | +2.09e+00 |
Pos logit contributions | |
its | +0.007 |
ases | +0.007 |
asing | +0.006 |
olds | +0.006 |
ins | +0.006 |
Neg logit contributions | |
he | -0.004 |
field | -0.003 |
roof | -0.003 |
while | -0.003 |
mountain | -0.002 |
Token | began |
Token ID | 2540 |
Feature activation | +3.91e+00 |
Pos logit contributions | |
he | +0.028 |
roof | +0.021 |
bicycle | +0.021 |
bus | +0.020 |
rocket | +0.019 |
Neg logit contributions | |
its | -0.073 |
ases | -0.069 |
ins | -0.067 |
asing | -0.064 |
arts | -0.063 |
Token | to |
Token ID | 284 |
Feature activation | +5.90e-01 |
Pos logit contributions | |
he | +0.048 |
bicycle | +0.043 |
moon | +0.042 |
valley | +0.041 |
wagon | +0.041 |
Neg logit contributions | |
ases | -0.125 |
its | -0.124 |
olds | -0.118 |
asing | -0.115 |
ums | -0.112 |
Token | set |
Token ID | 900 |
Feature activation | +7.99e-01 |
Pos logit contributions | |
he | +0.008 |
chair | +0.007 |
bicycle | +0.006 |
He | +0.006 |
Bun | +0.006 |
Neg logit contributions | |
its | -0.019 |
ases | -0.018 |
ps | -0.017 |
ws | -0.016 |
wed | -0.016 |
Token | , |
Token ID | 11 |
Feature activation | +5.82e+00 |
Pos logit contributions | |
he | +0.010 |
moon | +0.010 |
bus | +0.010 |
bicycle | +0.010 |
rocket | +0.009 |
Neg logit contributions | |
its | -0.024 |
ases | -0.022 |
ums | -0.022 |
olds | -0.021 |
arts | -0.021 |
Token | Bear |
Token ID | 14732 |
Feature activation | -6.34e-01 |
Pos logit contributions | |
bus | +0.097 |
bicycle | +0.094 |
bike | +0.090 |
driveway | +0.084 |
van | +0.082 |
Neg logit contributions | |
ins | -0.144 |
its | -0.140 |
arts | -0.131 |
ubb | -0.129 |
ases | -0.128 |
Token | suggested |
Token ID | 5220 |
Feature activation | +2.00e+00 |
Pos logit contributions | |
its | +0.020 |
ases | +0.019 |
ins | +0.019 |
asing | +0.018 |
arts | +0.018 |
Neg logit contributions | |
he | -0.011 |
roof | -0.009 |
rocket | -0.009 |
bicycle | -0.008 |
wagon | -0.008 |
Token | they |
Token ID | 484 |
Feature activation | +1.73e+00 |
Pos logit contributions | |
moon | +0.025 |
valley | +0.025 |
mountain | +0.024 |
cabin | +0.023 |
fireplace | +0.023 |
Neg logit contributions | |
its | -0.065 |
olds | -0.063 |
arts | -0.060 |
ases | -0.060 |
asing | -0.059 |
Token | play |
Token ID | 711 |
Feature activation | +1.11e+00 |
Pos logit contributions | |
he | +0.028 |
Pole | +0.024 |
mountain | +0.023 |
field | +0.022 |
valley | +0.022 |
Neg logit contributions | |
its | -0.055 |
ases | -0.055 |
asing | -0.050 |
olds | -0.050 |
arts | -0.048 |
Token | one |
Token ID | 530 |
Feature activation | +1.66e-01 |
Pos logit contributions | |
he | +0.015 |
Pole | +0.012 |
roof | +0.012 |
chim | +0.011 |
fireplace | +0.011 |
Neg logit contributions | |
ases | -0.039 |
its | -0.038 |
asing | -0.036 |
ins | -0.035 |
olds | -0.034 |
Token | , |
Token ID | 11 |
Feature activation | +6.26e+00 |
Pos logit contributions | |
he | +0.079 |
mountain | +0.072 |
ride | +0.069 |
hill | +0.069 |
bicycle | +0.068 |
Neg logit contributions | |
its | -0.198 |
ases | -0.192 |
olds | -0.184 |
arts | -0.181 |
asing | -0.181 |
Token | Mitt |
Token ID | 16627 |
Feature activation | -4.71e+00 |
Pos logit contributions | |
mountain | +0.100 |
bicycle | +0.097 |
Pole | +0.094 |
rocket | +0.093 |
hill | +0.093 |
Neg logit contributions | |
its | -0.168 |
ases | -0.160 |
olds | -0.152 |
arts | -0.151 |
ins | -0.149 |
Token | ens |
Token ID | 641 |
Feature activation | +6.82e-01 |
Pos logit contributions | |
its | +0.126 |
ases | +0.120 |
asing | +0.111 |
olds | +0.109 |
ins | +0.107 |
Neg logit contributions | |
he | -0.117 |
while | -0.093 |
field | -0.091 |
she | -0.090 |
roof | -0.089 |
Token | saw |
Token ID | 2497 |
Feature activation | +3.10e+00 |
Pos logit contributions | |
he | +0.011 |
rocket | +0.009 |
mountain | +0.009 |
roof | +0.009 |
sail | +0.008 |
Neg logit contributions | |
its | -0.022 |
ases | -0.020 |
arts | -0.020 |
ins | -0.019 |
asing | -0.019 |
Token | a |
Token ID | 257 |
Feature activation | +5.06e+00 |
Pos logit contributions | |
he | +0.034 |
roof | +0.023 |
mountain | +0.023 |
sail | +0.022 |
ride | +0.021 |
Neg logit contributions | |
its | -0.121 |
ases | -0.116 |
olds | -0.110 |
asing | -0.110 |
ins | -0.109 |
Token | butterfly |
Token ID | 35113 |
Feature activation | +4.97e+00 |
Pos logit contributions | |
he | +0.089 |
Up | +0.069 |
Suddenly | +0.067 |
But | +0.064 |
He | +0.063 |
Neg logit contributions | |
its | -0.153 |
ins | -0.134 |
ases | -0.130 |
cents | -0.130 |
arts | -0.129 |
Token | and |
Token ID | 290 |
Feature activation | +7.56e+00 |
Pos logit contributions | |
he | +0.083 |
Pole | +0.066 |
He | +0.060 |
mountain | +0.060 |
Peter | +0.059 |
Neg logit contributions | |
its | -0.162 |
ases | -0.151 |
asing | -0.141 |
arts | -0.140 |
olds | -0.135 |
Token | started |
Token ID | 2067 |
Feature activation | +2.93e+00 |
Pos logit contributions | |
mountain | +0.136 |
hill | +0.133 |
rocket | +0.132 |
port | +0.132 |
valley | +0.131 |
Neg logit contributions | |
its | -0.203 |
ases | -0.185 |
olds | -0.182 |
ins | -0.180 |
asing | -0.176 |
Token | to |
Token ID | 284 |
Feature activation | +1.30e+00 |
Pos logit contributions | |
he | +0.030 |
roof | +0.015 |
field | +0.015 |
mountain | +0.014 |
while | +0.014 |
Neg logit contributions | |
its | -0.120 |
ases | -0.116 |
asing | -0.111 |
olds | -0.109 |
ins | -0.108 |
Token | chase |
Token ID | 15505 |
Feature activation | -7.10e-01 |
Pos logit contributions | |
he | +0.025 |
rocket | +0.018 |
roof | +0.018 |
mountain | +0.018 |
car | +0.018 |
Neg logit contributions | |
its | -0.042 |
ases | -0.040 |
asing | -0.037 |
olds | -0.036 |
ins | -0.035 |
Token | it |
Token ID | 340 |
Feature activation | +1.94e+00 |
Pos logit contributions | |
its | +0.030 |
ases | +0.028 |
asing | +0.028 |
ins | +0.027 |
olds | +0.027 |
Neg logit contributions | |
he | -0.008 |
while | -0.004 |
Tim | -0.004 |
she | -0.004 |
outside | -0.004 |
Token | |
Token ID | 220 |
Feature activation | -6.00e+00 |
Pos logit contributions | |
he | +0.026 |
plane | +0.023 |
bus | +0.022 |
mountain | +0.022 |
field | +0.022 |
Neg logit contributions | |
its | -0.063 |
ases | -0.060 |
olds | -0.059 |
ums | -0.058 |
preferences | -0.055 |
Token |
|
Token ID | 198 |
Feature activation | -2.14e+00 |
Pos logit contributions | |
its | +0.235 |
ases | +0.224 |
olds | +0.214 |
ins | +0.213 |
asing | +0.212 |
Neg logit contributions | |
he | -0.069 |
mountain | -0.044 |
field | -0.044 |
roof | -0.043 |
while | -0.042 |
Token |
|
Token ID | 198 |
Feature activation | -1.27e+00 |
Pos logit contributions | |
its | +0.087 |
ases | +0.084 |
asing | +0.080 |
olds | +0.079 |
ins | +0.079 |
Neg logit contributions | |
he | -0.022 |
roof | -0.011 |
field | -0.011 |
while | -0.011 |
mountain | -0.010 |
Token | One |
Token ID | 3198 |
Feature activation | +5.38e-01 |
Pos logit contributions | |
its | +0.037 |
ases | +0.033 |
asing | +0.032 |
ins | +0.031 |
olds | +0.031 |
Neg logit contributions | |
he | -0.027 |
roof | -0.021 |
down | -0.021 |
mountain | -0.021 |
field | -0.021 |
Token | day |
Token ID | 1110 |
Feature activation | +4.61e+00 |
Pos logit contributions | |
he | +0.010 |
Pole | +0.009 |
roof | +0.009 |
bicycle | +0.008 |
rocket | +0.008 |
Neg logit contributions | |
its | -0.014 |
ases | -0.013 |
ac | -0.013 |
asing | -0.013 |
ins | -0.012 |
Token | , |
Token ID | 11 |
Feature activation | +5.44e+00 |
Pos logit contributions | |
mountain | +0.079 |
hill | +0.077 |
field | +0.077 |
moon | +0.076 |
stable | +0.076 |
Neg logit contributions | |
its | -0.120 |
ases | -0.118 |
olds | -0.116 |
ins | -0.112 |
asing | -0.109 |
Token | Tim |
Token ID | 5045 |
Feature activation | -3.16e+00 |
Pos logit contributions | |
valley | +0.090 |
mountain | +0.090 |
stable | +0.088 |
carriage | +0.088 |
Pole | +0.086 |
Neg logit contributions | |
its | -0.126 |
ases | -0.118 |
ateful | -0.114 |
olds | -0.111 |
ins | -0.111 |
Token | my |
Token ID | 1820 |
Feature activation | +8.40e-01 |
Pos logit contributions | |
its | +0.089 |
ases | +0.084 |
asing | +0.079 |
ins | +0.078 |
olds | +0.076 |
Neg logit contributions | |
he | -0.071 |
roof | -0.057 |
while | -0.055 |
mountain | -0.055 |
rocket | -0.055 |
Token | saw |
Token ID | 2497 |
Feature activation | +2.78e+00 |
Pos logit contributions | |
he | +0.013 |
roof | +0.013 |
helicopter | +0.013 |
rocket | +0.013 |
mountain | +0.012 |
Neg logit contributions | |
its | -0.023 |
ins | -0.022 |
ases | -0.022 |
olds | -0.021 |
arts | -0.021 |
Token | a |
Token ID | 257 |
Feature activation | +4.37e+00 |
Pos logit contributions | |
hill | +0.031 |
mountain | +0.031 |
he | +0.031 |
Pole | +0.030 |
roof | +0.030 |
Neg logit contributions | |
its | -0.090 |
ases | -0.089 |
olds | -0.086 |
asing | -0.083 |
ins | -0.083 |
Token | bird |
Token ID | 6512 |
Feature activation | +2.18e+00 |
Pos logit contributions | |
he | +0.079 |
Up | +0.065 |
Suddenly | +0.063 |
He | +0.062 |
While | +0.062 |
Neg logit contributions | |
its | -0.122 |
ins | -0.105 |
cents | -0.105 |
arts | -0.102 |
mistake | -0.100 |
Token | she |
Token ID | 673 |
Feature activation | -1.31e+00 |
Pos logit contributions | |
he | +0.013 |
roof | +0.010 |
sail | +0.009 |
field | +0.009 |
bus | +0.009 |
Neg logit contributions | |
its | -0.031 |
ases | -0.029 |
asing | -0.028 |
olds | -0.028 |
ins | -0.027 |
Token | wanted |
Token ID | 2227 |
Feature activation | -2.40e+00 |
Pos logit contributions | |
its | +0.045 |
ases | +0.044 |
olds | +0.040 |
ins | +0.040 |
asing | +0.040 |
Neg logit contributions | |
he | -0.024 |
ride | -0.015 |
she | -0.015 |
mountain | -0.015 |
looked | -0.015 |
Token | to |
Token ID | 284 |
Feature activation | -2.81e+00 |
Pos logit contributions | |
its | +0.093 |
ases | +0.089 |
asing | +0.087 |
olds | +0.086 |
ins | +0.084 |
Neg logit contributions | |
he | -0.026 |
field | -0.016 |
sail | -0.016 |
valley | -0.016 |
roof | -0.016 |
Token | play |
Token ID | 711 |
Feature activation | -1.36e+00 |
Pos logit contributions | |
its | +0.088 |
ases | +0.082 |
asing | +0.077 |
olds | +0.077 |
arts | +0.073 |
Neg logit contributions | |
he | -0.054 |
field | -0.042 |
valley | -0.041 |
moon | -0.041 |
while | -0.041 |
Token | outside |
Token ID | 2354 |
Feature activation | +3.70e+00 |
Pos logit contributions | |
its | +0.048 |
ases | +0.047 |
asing | +0.045 |
ins | +0.043 |
olds | +0.043 |
Neg logit contributions | |
he | -0.019 |
roof | -0.014 |
field | -0.013 |
moon | -0.013 |
mountain | -0.013 |
Token | , |
Token ID | 11 |
Feature activation | +3.22e+00 |
Pos logit contributions | |
he | +0.042 |
sail | +0.031 |
moon | +0.031 |
roof | +0.030 |
ride | +0.030 |
Neg logit contributions | |
its | -0.131 |
ases | -0.130 |
asing | -0.124 |
olds | -0.121 |
ins | -0.119 |
Token | sometimes |
Token ID | 3360 |
Feature activation | +1.57e+00 |
Pos logit contributions | |
he | +0.034 |
field | +0.025 |
Pole | +0.025 |
bus | +0.025 |
mountain | +0.025 |
Neg logit contributions | |
its | -0.117 |
ases | -0.111 |
asing | -0.110 |
olds | -0.108 |
ins | -0.105 |
Token | she |
Token ID | 673 |
Feature activation | +7.02e-01 |
Pos logit contributions | |
he | +0.017 |
Pole | +0.015 |
Under | +0.014 |
bus | +0.014 |
While | +0.013 |
Neg logit contributions | |
its | -0.055 |
ases | -0.052 |
asing | -0.051 |
olds | -0.050 |
ins | -0.049 |
Token | needed |
Token ID | 2622 |
Feature activation | -1.35e+00 |
Pos logit contributions | |
he | +0.011 |
while | +0.007 |
she | +0.007 |
ride | +0.007 |
car | +0.007 |
Neg logit contributions | |
its | -0.025 |
ases | -0.024 |
asing | -0.023 |
olds | -0.023 |
ins | -0.023 |
Token | to |
Token ID | 284 |
Feature activation | -1.43e+00 |
Pos logit contributions | |
its | +0.051 |
ases | +0.050 |
olds | +0.048 |
asing | +0.048 |
ins | +0.046 |
Neg logit contributions | |
he | -0.015 |
mountain | -0.009 |
field | -0.009 |
roof | -0.009 |
valley | -0.009 |
Token | study |
Token ID | 2050 |
Feature activation | -3.36e-01 |
Pos logit contributions | |
its | +0.045 |
ases | +0.042 |
asing | +0.040 |
olds | +0.039 |
ins | +0.038 |
Neg logit contributions | |
he | -0.028 |
field | -0.021 |
moon | -0.020 |
mountain | -0.020 |
rocket | -0.020 |
Token | and |
Token ID | 290 |
Feature activation | +3.44e+00 |
Pos logit contributions | |
he | +0.004 |
valley | +0.004 |
river | +0.004 |
Under | +0.004 |
While | +0.004 |
Neg logit contributions | |
its | -0.009 |
ases | -0.009 |
olds | -0.009 |
asing | -0.009 |
arts | -0.008 |
Token | hugged |
Token ID | 41130 |
Feature activation | -3.18e+00 |
Pos logit contributions | |
Pole | +0.050 |
valley | +0.048 |
river | +0.047 |
mountain | +0.047 |
hill | +0.046 |
Neg logit contributions | |
its | -0.099 |
ases | -0.096 |
olds | -0.094 |
asing | -0.091 |
ins | -0.089 |
Token | her |
Token ID | 607 |
Feature activation | -2.94e+00 |
Pos logit contributions | |
its | +0.094 |
ases | +0.089 |
olds | +0.084 |
arts | +0.083 |
ums | +0.083 |
Neg logit contributions | |
he | -0.052 |
field | -0.049 |
mountain | -0.049 |
river | -0.049 |
valley | -0.048 |
Token | tightly |
Token ID | 17707 |
Feature activation | -1.67e+00 |
Pos logit contributions | |
its | +0.081 |
ases | +0.075 |
ac | +0.073 |
ums | +0.073 |
olds | +0.073 |
Neg logit contributions | |
he | -0.052 |
Pole | -0.047 |
river | -0.046 |
valley | -0.044 |
roof | -0.044 |
Token | . |
Token ID | 13 |
Feature activation | +2.55e-01 |
Pos logit contributions | |
its | +0.043 |
arts | +0.040 |
ps | +0.039 |
ases | +0.039 |
olds | +0.038 |
Neg logit contributions | |
he | -0.029 |
Pole | -0.028 |
valley | -0.027 |
river | -0.027 |
mountain | -0.026 |
Token | He |
Token ID | 679 |
Feature activation | +9.08e-01 |
Pos logit contributions | |
he | +0.003 |
mountain | +0.003 |
field | +0.003 |
bus | +0.003 |
valley | +0.003 |
Neg logit contributions | |
its | -0.008 |
olds | -0.008 |
ases | -0.008 |
ums | -0.008 |
ins | -0.007 |
Token | knew |
Token ID | 2993 |
Feature activation | -2.72e-01 |
Pos logit contributions | |
he | +0.015 |
roof | +0.012 |
mountain | +0.011 |
valley | +0.011 |
river | +0.011 |
Neg logit contributions | |
its | -0.029 |
ases | -0.028 |
asing | -0.027 |
olds | -0.026 |
ins | -0.026 |
Token | he |
Token ID | 339 |
Feature activation | -1.92e+00 |
Pos logit contributions | |
its | +0.009 |
olds | +0.009 |
asing | +0.009 |
ases | +0.009 |
ins | +0.009 |
Neg logit contributions | |
river | -0.002 |
field | -0.002 |
While | -0.002 |
valley | -0.002 |
roof | -0.002 |
Token | would |
Token ID | 561 |
Feature activation | -1.18e+00 |
Pos logit contributions | |
its | +0.073 |
ases | +0.070 |
asing | +0.067 |
olds | +0.066 |
ins | +0.065 |
Neg logit contributions | |
he | -0.025 |
roof | -0.016 |
field | -0.015 |
while | -0.015 |
mountain | -0.015 |
Token | always |
Token ID | 1464 |
Feature activation | -1.01e+00 |
Pos logit contributions | |
its | +0.035 |
ases | +0.033 |
asing | +0.031 |
olds | +0.030 |
arts | +0.030 |
Neg logit contributions | |
he | -0.025 |
field | -0.019 |
mountain | -0.019 |
while | -0.019 |
rocket | -0.019 |
Token | remain |
Token ID | 3520 |
Feature activation | -4.16e+00 |
Pos logit contributions | |
its | +0.030 |
ases | +0.028 |
asing | +0.026 |
olds | +0.026 |
ins | +0.025 |
Neg logit contributions | |
he | -0.022 |
roof | -0.017 |
field | -0.017 |
while | -0.016 |
mountain | -0.016 |
Token | shows |
Token ID | 2523 |
Feature activation | +6.21e-01 |
Pos logit contributions | |
he | +0.026 |
roof | +0.020 |
valley | +0.020 |
sail | +0.019 |
river | +0.019 |
Neg logit contributions | |
its | -0.039 |
ases | -0.037 |
ins | -0.035 |
olds | -0.034 |
asing | -0.033 |
Token | . |
Token ID | 13 |
Feature activation | +2.63e+00 |
Pos logit contributions | |
he | +0.010 |
river | +0.009 |
hill | +0.009 |
valley | +0.009 |
mountain | +0.009 |
Neg logit contributions | |
its | -0.020 |
ases | -0.019 |
olds | -0.018 |
asing | -0.018 |
ins | -0.018 |
Token | One |
Token ID | 1881 |
Feature activation | +1.65e+00 |
Pos logit contributions | |
he | +0.042 |
mountain | +0.036 |
sail | +0.034 |
raft | +0.034 |
boat | +0.033 |
Neg logit contributions | |
its | -0.082 |
olds | -0.076 |
ases | -0.076 |
ums | -0.073 |
ins | -0.071 |
Token | day |
Token ID | 1110 |
Feature activation | +5.09e+00 |
Pos logit contributions | |
he | +0.030 |
valley | +0.025 |
Pole | +0.025 |
sail | +0.024 |
roof | +0.024 |
Neg logit contributions | |
its | -0.046 |
ases | -0.044 |
asing | -0.043 |
olds | -0.042 |
ac | -0.042 |
Token | , |
Token ID | 11 |
Feature activation | +5.74e+00 |
Pos logit contributions | |
he | +0.060 |
mountain | +0.052 |
river | +0.051 |
field | +0.049 |
hill | +0.048 |
Neg logit contributions | |
its | -0.180 |
ases | -0.175 |
olds | -0.170 |
asing | -0.168 |
ins | -0.164 |
Token | her |
Token ID | 607 |
Feature activation | +8.73e-01 |
Pos logit contributions | |
mountain | +0.106 |
river | +0.100 |
hill | +0.096 |
cabin | +0.095 |
Pole | +0.095 |
Neg logit contributions | |
its | -0.147 |
ases | -0.137 |
olds | -0.136 |
asing | -0.129 |
ins | -0.128 |
Token | mom |
Token ID | 1995 |
Feature activation | -3.37e-03 |
Pos logit contributions | |
he | +0.016 |
mountain | +0.011 |
He | +0.011 |
while | +0.011 |
field | +0.011 |
Neg logit contributions | |
its | -0.028 |
ases | -0.027 |
asing | -0.026 |
olds | -0.026 |
ins | -0.025 |
Token | asked |
Token ID | 1965 |
Feature activation | +4.96e-01 |
Pos logit contributions | |
its | +0.000 |
ases | +0.000 |
asing | +0.000 |
ins | +0.000 |
olds | +0.000 |
Neg logit contributions | |
he | -0.000 |
roof | -0.000 |
mountain | -0.000 |
field | -0.000 |
valley | -0.000 |
Token | her |
Token ID | 607 |
Feature activation | +1.24e+00 |
Pos logit contributions | |
he | +0.007 |
while | +0.005 |
she | +0.004 |
outside | +0.004 |
they | +0.004 |
Neg logit contributions | |
its | -0.019 |
ases | -0.018 |
asing | -0.017 |
olds | -0.017 |
ins | -0.017 |
Token | to |
Token ID | 284 |
Feature activation | +1.16e+00 |
Pos logit contributions | |
he | +0.015 |
field | +0.010 |
roof | +0.009 |
mountain | +0.009 |
valley | +0.009 |
Neg logit contributions | |
its | -0.048 |
ases | -0.046 |
asing | -0.044 |
olds | -0.044 |
ins | -0.043 |
Token | turn |
Token ID | 1210 |
Feature activation | +6.42e-01 |
Pos logit contributions | |
he | +0.022 |
valley | +0.018 |
mountain | +0.017 |
field | +0.017 |
roof | +0.017 |
Neg logit contributions | |
its | -0.036 |
ases | -0.035 |
olds | -0.032 |
asing | -0.032 |
arts | -0.030 |
Token | there |
Token ID | 612 |
Feature activation | +4.67e+00 |
Pos logit contributions | |
valley | +0.066 |
he | +0.066 |
mountain | +0.066 |
rocket | +0.063 |
While | +0.063 |
Neg logit contributions | |
its | -0.222 |
ases | -0.210 |
ins | -0.208 |
olds | -0.206 |
asing | -0.199 |
Token | was |
Token ID | 373 |
Feature activation | +3.61e+00 |
Pos logit contributions | |
he | +0.071 |
roof | +0.059 |
mountain | +0.058 |
sail | +0.056 |
bus | +0.056 |
Neg logit contributions | |
its | -0.157 |
ases | -0.147 |
ins | -0.138 |
olds | -0.137 |
asing | -0.137 |
Token | a |
Token ID | 257 |
Feature activation | +4.91e+00 |
Pos logit contributions | |
he | +0.045 |
Pole | +0.042 |
mountain | +0.040 |
hill | +0.039 |
sail | +0.037 |
Neg logit contributions | |
its | -0.119 |
ases | -0.116 |
olds | -0.111 |
asing | -0.108 |
ins | -0.105 |
Token | big |
Token ID | 1263 |
Feature activation | +3.43e+00 |
Pos logit contributions | |
he | +0.088 |
Up | +0.071 |
He | +0.065 |
Suddenly | +0.064 |
Under | +0.064 |
Neg logit contributions | |
its | -0.138 |
cents | -0.125 |
ins | -0.121 |
humor | -0.119 |
ases | -0.119 |
Token | , |
Token ID | 11 |
Feature activation | +5.70e+00 |
Pos logit contributions | |
he | +0.051 |
Up | +0.039 |
Suddenly | +0.037 |
While | +0.036 |
A | +0.036 |
Neg logit contributions | |
its | -0.109 |
ases | -0.107 |
humor | -0.098 |
arts | -0.097 |
olds | -0.097 |
Token | yellow |
Token ID | 7872 |
Feature activation | +3.49e+00 |
Pos logit contributions | |
he | +0.107 |
While | +0.096 |
Pole | +0.090 |
Up | +0.088 |
Suddenly | +0.088 |
Neg logit contributions | |
its | -0.136 |
cents | -0.117 |
humor | -0.116 |
ases | -0.114 |
asing | -0.113 |
Token | sun |
Token ID | 4252 |
Feature activation | +2.98e+00 |
Pos logit contributions | |
he | +0.060 |
Suddenly | +0.047 |
Up | +0.046 |
While | +0.044 |
Bun | +0.043 |
Neg logit contributions | |
its | -0.101 |
ases | -0.100 |
arts | -0.092 |
ums | -0.090 |
ins | -0.089 |
Token | . |
Token ID | 13 |
Feature activation | +3.54e+00 |
Pos logit contributions | |
he | +0.060 |
While | +0.053 |
Under | +0.051 |
low | +0.050 |
Lightning | +0.050 |
Neg logit contributions | |
its | -0.078 |
ases | -0.071 |
arts | -0.070 |
ums | -0.070 |
ins | -0.070 |
Token | The |
Token ID | 383 |
Feature activation | +1.66e+00 |
Pos logit contributions | |
he | +0.050 |
walk | +0.049 |
sail | +0.048 |
mountain | +0.048 |
plane | +0.048 |
Neg logit contributions | |
ases | -0.098 |
olds | -0.096 |
its | -0.096 |
ums | -0.093 |
ump | -0.088 |
Token | sun |
Token ID | 4252 |
Feature activation | +1.61e+00 |
Pos logit contributions | |
he | +0.026 |
While | +0.018 |
Bun | +0.018 |
Suddenly | +0.018 |
Under | +0.018 |
Neg logit contributions | |
its | -0.053 |
ases | -0.050 |
humor | -0.047 |
ins | -0.047 |
arts | -0.046 |
Token | loved |
Token ID | 6151 |
Feature activation | +2.30e+00 |
Pos logit contributions | |
he | +0.026 |
While | +0.020 |
Bun | +0.019 |
west | +0.019 |
A | +0.019 |
Neg logit contributions | |
its | -0.052 |
ases | -0.048 |
ins | -0.047 |
asing | -0.045 |
arts | -0.045 |
Token | to |
Token ID | 284 |
Feature activation | -3.12e-01 |
Pos logit contributions | |
its | +0.008 |
ases | +0.008 |
asing | +0.008 |
olds | +0.008 |
ins | +0.007 |
Neg logit contributions | |
he | -0.002 |
roof | -0.001 |
field | -0.001 |
mountain | -0.001 |
while | -0.001 |
Token | bring |
Token ID | 2222 |
Feature activation | -8.06e-01 |
Pos logit contributions | |
its | +0.011 |
ases | +0.011 |
asing | +0.010 |
ins | +0.010 |
olds | +0.010 |
Neg logit contributions | |
he | -0.005 |
roof | -0.003 |
ride | -0.003 |
while | -0.003 |
wide | -0.003 |
Token | it |
Token ID | 340 |
Feature activation | +1.61e+00 |
Pos logit contributions | |
its | +0.034 |
ases | +0.032 |
asing | +0.031 |
olds | +0.031 |
ins | +0.031 |
Neg logit contributions | |
he | -0.008 |
while | -0.004 |
wide | -0.004 |
field | -0.004 |
moon | -0.004 |
Token | home |
Token ID | 1363 |
Feature activation | +1.21e+00 |
Pos logit contributions | |
he | +0.024 |
mountain | +0.017 |
roof | +0.016 |
field | +0.016 |
river | +0.016 |
Neg logit contributions | |
its | -0.056 |
ases | -0.055 |
asing | -0.051 |
olds | -0.051 |
arts | -0.049 |
Token | and |
Token ID | 290 |
Feature activation | +4.21e+00 |
Pos logit contributions | |
he | +0.016 |
mountain | +0.011 |
roof | +0.011 |
valley | +0.011 |
field | +0.011 |
Neg logit contributions | |
its | -0.044 |
ases | -0.043 |
olds | -0.041 |
asing | -0.041 |
ins | -0.040 |
Token | show |
Token ID | 905 |
Feature activation | -1.19e+00 |
Pos logit contributions | |
he | +0.071 |
Pole | +0.069 |
valley | +0.064 |
mountain | +0.063 |
cabin | +0.063 |
Neg logit contributions | |
its | -0.119 |
ases | -0.116 |
olds | -0.111 |
humor | -0.107 |
preferences | -0.106 |
Token | it |
Token ID | 340 |
Feature activation | +1.49e+00 |
Pos logit contributions | |
its | +0.048 |
ases | +0.046 |
asing | +0.044 |
ins | +0.043 |
olds | +0.043 |
Neg logit contributions | |
he | -0.014 |
while | -0.007 |
roof | -0.007 |
field | -0.007 |
moon | -0.007 |
Token | to |
Token ID | 284 |
Feature activation | +1.67e-01 |
Pos logit contributions | |
he | +0.012 |
while | +0.004 |
roof | +0.004 |
wide | +0.004 |
field | +0.004 |
Neg logit contributions | |
its | -0.065 |
ases | -0.062 |
asing | -0.060 |
ins | -0.059 |
olds | -0.059 |
Token | her |
Token ID | 607 |
Feature activation | +9.85e-01 |
Pos logit contributions | |
he | +0.003 |
field | +0.002 |
roof | +0.002 |
while | +0.002 |
mountain | +0.002 |
Neg logit contributions | |
its | -0.006 |
ases | -0.006 |
asing | -0.005 |
olds | -0.005 |
ins | -0.005 |
Token | mom |
Token ID | 1995 |
Feature activation | -5.60e-01 |
Pos logit contributions | |
he | +0.019 |
roof | +0.014 |
field | +0.014 |
while | +0.014 |
mountain | +0.013 |
Neg logit contributions | |
its | -0.031 |
ases | -0.030 |
asing | -0.028 |
olds | -0.028 |
ins | -0.027 |
Token | . |
Token ID | 13 |
Feature activation | +1.15e+00 |
Pos logit contributions | |
its | +0.020 |
ases | +0.019 |
asing | +0.018 |
ins | +0.018 |
olds | +0.018 |
Neg logit contributions | |
he | -0.008 |
roof | -0.006 |
mountain | -0.006 |
valley | -0.006 |
rocket | -0.006 |
Token | first |
Token ID | 717 |
Feature activation | -1.41e+00 |
Pos logit contributions | |
its | +0.257 |
ases | +0.246 |
asing | +0.236 |
olds | +0.235 |
ins | +0.234 |
Neg logit contributions | |
he | -0.054 |
while | -0.026 |
she | -0.021 |
the | -0.019 |
wide | -0.018 |
Token | place |
Token ID | 1295 |
Feature activation | -5.72e-01 |
Pos logit contributions | |
its | +0.061 |
ases | +0.059 |
asing | +0.056 |
ins | +0.056 |
olds | +0.055 |
Neg logit contributions | |
he | -0.013 |
while | -0.006 |
she | -0.006 |
they | -0.004 |
wide | -0.004 |
Token | , |
Token ID | 11 |
Feature activation | +2.88e+00 |
Pos logit contributions | |
its | +0.023 |
ases | +0.022 |
asing | +0.021 |
olds | +0.021 |
ins | +0.021 |
Neg logit contributions | |
he | -0.006 |
while | -0.003 |
roof | -0.003 |
field | -0.003 |
wide | -0.003 |
Token | but |
Token ID | 475 |
Feature activation | +1.17e+00 |
Pos logit contributions | |
he | +0.030 |
rocket | +0.020 |
mountain | +0.020 |
roof | +0.020 |
cabin | +0.020 |
Neg logit contributions | |
its | -0.109 |
ases | -0.105 |
asing | -0.103 |
olds | -0.100 |
ins | -0.097 |
Token | she |
Token ID | 673 |
Feature activation | +1.50e-02 |
Pos logit contributions | |
he | +0.010 |
roof | +0.006 |
field | +0.005 |
mountain | +0.005 |
river | +0.005 |
Neg logit contributions | |
its | -0.048 |
ases | -0.046 |
asing | -0.045 |
olds | -0.045 |
ins | -0.043 |
Token | was |
Token ID | 373 |
Feature activation | -2.71e+00 |
Pos logit contributions | |
he | +0.000 |
while | +0.000 |
field | +0.000 |
wide | +0.000 |
ride | +0.000 |
Neg logit contributions | |
its | -0.001 |
ases | -0.001 |
asing | -0.000 |
olds | -0.000 |
ins | -0.000 |
Token | proud |
Token ID | 6613 |
Feature activation | +2.73e-01 |
Pos logit contributions | |
its | +0.090 |
ases | +0.086 |
asing | +0.081 |
olds | +0.080 |
ins | +0.079 |
Neg logit contributions | |
he | -0.051 |
while | -0.037 |
wide | -0.036 |
field | -0.036 |
ride | -0.035 |
Token | of |
Token ID | 286 |
Feature activation | -3.28e+00 |
Pos logit contributions | |
he | +0.005 |
while | +0.004 |
field | +0.004 |
roof | +0.004 |
wide | +0.004 |
Neg logit contributions | |
its | -0.009 |
ases | -0.008 |
asing | -0.008 |
olds | -0.008 |
ins | -0.008 |
Token | herself |
Token ID | 5223 |
Feature activation | -1.43e+00 |
Pos logit contributions | |
its | +0.098 |
ases | +0.092 |
asing | +0.086 |
ins | +0.085 |
olds | +0.084 |
Neg logit contributions | |
he | -0.075 |
she | -0.059 |
they | -0.056 |
outside | -0.056 |
Tim | -0.055 |
Token | for |
Token ID | 329 |
Feature activation | -3.83e+00 |
Pos logit contributions | |
its | +0.043 |
ases | +0.042 |
asing | +0.042 |
olds | +0.040 |
ins | +0.039 |
Neg logit contributions | |
he | -0.025 |
roof | -0.020 |
river | -0.020 |
valley | -0.020 |
moon | -0.020 |
Token | trying |
Token ID | 2111 |
Feature activation | +7.62e-02 |
Pos logit contributions | |
its | +0.115 |
ases | +0.109 |
olds | +0.101 |
asing | +0.100 |
ins | +0.099 |
Neg logit contributions | |
he | -0.087 |
while | -0.066 |
she | -0.065 |
wide | -0.064 |
they | -0.064 |
Token | . |
Token ID | 13 |
Feature activation | +2.84e+00 |
Pos logit contributions | |
he | +0.125 |
Bun | +0.103 |
bicycle | +0.100 |
rocket | +0.099 |
wagon | +0.098 |
Neg logit contributions | |
its | -0.253 |
ases | -0.241 |
olds | -0.229 |
ins | -0.226 |
asing | -0.224 |
Token | The |
Token ID | 383 |
Feature activation | +9.63e-01 |
Pos logit contributions | |
shield | +0.048 |
airplane | +0.048 |
bicycle | +0.046 |
helmet | +0.046 |
mountain | +0.045 |
Neg logit contributions | |
its | -0.070 |
ases | -0.068 |
olds | -0.068 |
ums | -0.065 |
ins | -0.063 |
Token | pig |
Token ID | 12967 |
Feature activation | -5.42e-01 |
Pos logit contributions | |
he | +0.015 |
while | +0.009 |
roof | +0.009 |
mountain | +0.009 |
field | +0.009 |
Neg logit contributions | |
its | -0.035 |
ases | -0.033 |
asing | -0.031 |
ins | -0.031 |
olds | -0.031 |
Token | said |
Token ID | 531 |
Feature activation | +4.68e-01 |
Pos logit contributions | |
its | +0.018 |
ases | +0.017 |
asing | +0.016 |
olds | +0.016 |
ins | +0.016 |
Neg logit contributions | |
he | -0.009 |
roof | -0.007 |
rocket | -0.007 |
while | -0.007 |
field | -0.007 |
Token | , |
Token ID | 11 |
Feature activation | +3.80e+00 |
Pos logit contributions | |
field | +0.006 |
moon | +0.005 |
fireplace | +0.005 |
fields | +0.005 |
hill | +0.005 |
Neg logit contributions | |
its | -0.014 |
ases | -0.014 |
olds | -0.014 |
asing | -0.014 |
arts | -0.013 |
Token | " |
Token ID | 366 |
Feature activation | -2.76e+00 |
Pos logit contributions | |
valley | +0.067 |
mountain | +0.062 |
hut | +0.062 |
cabin | +0.060 |
spaceship | +0.060 |
Neg logit contributions | |
its | -0.089 |
olds | -0.086 |
ins | -0.082 |
arts | -0.080 |
asing | -0.080 |
Token | We |
Token ID | 1135 |
Feature activation | -1.60e+00 |
Pos logit contributions | |
its | +0.092 |
ases | +0.088 |
asing | +0.083 |
olds | +0.082 |
ins | +0.081 |
Neg logit contributions | |
he | -0.051 |
roof | -0.036 |
field | -0.036 |
while | -0.036 |
mountain | -0.035 |
Token | should |
Token ID | 815 |
Feature activation | +1.33e+00 |
Pos logit contributions | |
its | +0.049 |
ases | +0.047 |
ins | +0.044 |
olds | +0.044 |
asing | +0.044 |
Neg logit contributions | |
he | -0.033 |
field | -0.025 |
while | -0.025 |
ride | -0.025 |
wide | -0.025 |
Token | eat |
Token ID | 4483 |
Feature activation | +5.03e-01 |
Pos logit contributions | |
he | +0.027 |
while | +0.020 |
field | +0.019 |
moon | +0.019 |
rocket | +0.019 |
Neg logit contributions | |
its | -0.042 |
ases | -0.039 |
asing | -0.037 |
olds | -0.037 |
ins | -0.035 |
Token | that |
Token ID | 326 |
Feature activation | +1.47e+00 |
Pos logit contributions | |
he | +0.007 |
roof | +0.005 |
moon | +0.005 |
ride | +0.005 |
bus | +0.005 |
Neg logit contributions | |
its | -0.018 |
ases | -0.018 |
asing | -0.017 |
ins | -0.016 |
olds | -0.016 |
Token | squash |
Token ID | 34613 |
Feature activation | +2.35e+00 |
Pos logit contributions | |
he | +0.030 |
bus | +0.023 |
roof | +0.023 |
ride | +0.022 |
Suddenly | +0.022 |
Neg logit contributions | |
its | -0.044 |
ases | -0.041 |
asing | -0.038 |
ins | -0.038 |
olds | -0.037 |
Token | grapes |
Token ID | 36656 |
Feature activation | -1.22e+00 |
Pos logit contributions | |
its | +0.103 |
ases | +0.098 |
asing | +0.094 |
olds | +0.091 |
ins | +0.091 |
Neg logit contributions | |
he | -0.060 |
field | -0.047 |
roof | -0.046 |
wide | -0.045 |
while | -0.044 |
Token | ! |
Token ID | 0 |
Feature activation | -1.47e+00 |
Pos logit contributions | |
its | +0.038 |
ases | +0.037 |
asing | +0.035 |
olds | +0.034 |
ins | +0.034 |
Neg logit contributions | |
he | -0.024 |
roof | -0.018 |
field | -0.017 |
while | -0.017 |
moon | -0.017 |
Token | Go |
Token ID | 1514 |
Feature activation | -4.89e+00 |
Pos logit contributions | |
its | +0.043 |
ases | +0.041 |
asing | +0.038 |
olds | +0.038 |
ins | +0.037 |
Neg logit contributions | |
he | -0.032 |
field | -0.025 |
while | -0.025 |
roof | -0.025 |
mountain | -0.024 |
Token | away |
Token ID | 1497 |
Feature activation | -2.68e-01 |
Pos logit contributions | |
its | +0.154 |
ases | +0.149 |
asing | +0.137 |
olds | +0.136 |
ins | +0.132 |
Neg logit contributions | |
he | -0.096 |
roof | -0.072 |
moon | -0.071 |
mountain | -0.070 |
field | -0.070 |
Token | , |
Token ID | 11 |
Feature activation | +1.61e+00 |
Pos logit contributions | |
its | +0.008 |
ases | +0.008 |
asing | +0.007 |
olds | +0.007 |
ins | +0.007 |
Neg logit contributions | |
he | -0.006 |
roof | -0.004 |
field | -0.004 |
ride | -0.004 |
moon | -0.004 |
Token | silly |
Token ID | 14397 |
Feature activation | -2.37e+00 |
Pos logit contributions | |
he | +0.028 |
field | +0.020 |
roof | +0.020 |
while | +0.020 |
moon | +0.019 |
Neg logit contributions | |
its | -0.055 |
ases | -0.052 |
asing | -0.049 |
olds | -0.049 |
ins | -0.048 |
Token | bird |
Token ID | 6512 |
Feature activation | -2.71e+00 |
Pos logit contributions | |
its | +0.081 |
ases | +0.077 |
asing | +0.073 |
olds | +0.072 |
ins | +0.071 |
Neg logit contributions | |
he | -0.041 |
roof | -0.029 |
while | -0.028 |
field | -0.028 |
moon | -0.028 |
Token | !" |
Token ID | 2474 |
Feature activation | -3.09e-01 |
Pos logit contributions | |
its | +0.078 |
ases | +0.076 |
asing | +0.069 |
ins | +0.068 |
olds | +0.068 |
Neg logit contributions | |
he | -0.056 |
moon | -0.046 |
roof | -0.046 |
Under | -0.045 |
Suddenly | -0.045 |
Token | Lily |
Token ID | 20037 |
Feature activation | -5.09e+00 |
Pos logit contributions | |
its | +0.010 |
ases | +0.010 |
olds | +0.010 |
asing | +0.009 |
humor | +0.008 |
Neg logit contributions | |
stable | -0.004 |
wagon | -0.004 |
field | -0.004 |
roof | -0.004 |
cabin | -0.004 |
Token | shouted |
Token ID | 17293 |
Feature activation | -5.07e+00 |
Pos logit contributions | |
its | +0.170 |
ases | +0.164 |
asing | +0.155 |
olds | +0.153 |
ins | +0.151 |
Neg logit contributions | |
he | -0.085 |
roof | -0.065 |
valley | -0.065 |
mountain | -0.065 |
moon | -0.064 |
Token | . |
Token ID | 13 |
Feature activation | -1.72e+00 |
Pos logit contributions | |
its | +0.174 |
ases | +0.170 |
olds | +0.162 |
asing | +0.161 |
arts | +0.153 |
Neg logit contributions | |
he | -0.075 |
mountain | -0.061 |
field | -0.060 |
valley | -0.059 |
roof | -0.057 |
Token | dolphin |
Token ID | 44136 |
Feature activation | +7.63e-02 |
Pos logit contributions | |
its | +0.005 |
ases | +0.004 |
asing | +0.004 |
olds | +0.004 |
ins | +0.004 |
Neg logit contributions | |
he | -0.002 |
field | -0.002 |
wide | -0.002 |
while | -0.002 |
sail | -0.002 |
Token | knew |
Token ID | 2993 |
Feature activation | +7.26e-02 |
Pos logit contributions | |
he | +0.001 |
roof | +0.001 |
moon | +0.001 |
fireplace | +0.001 |
Under | +0.001 |
Neg logit contributions | |
its | -0.002 |
ases | -0.002 |
asing | -0.002 |
ins | -0.002 |
olds | -0.002 |
Token | how |
Token ID | 703 |
Feature activation | -4.06e+00 |
Pos logit contributions | |
roof | +0.001 |
he | +0.001 |
moon | +0.001 |
field | +0.001 |
fireplace | +0.001 |
Neg logit contributions | |
its | -0.003 |
ases | -0.003 |
asing | -0.003 |
ins | -0.003 |
olds | -0.003 |
Token | Nem |
Token ID | 22547 |
Feature activation | -6.59e+00 |
Pos logit contributions | |
its | +0.155 |
ases | +0.148 |
olds | +0.145 |
asing | +0.144 |
ins | +0.140 |
Neg logit contributions | |
he | -0.043 |
moon | -0.034 |
field | -0.034 |
roof | -0.033 |
valley | -0.032 |
Token | o |
Token ID | 78 |
Feature activation | -4.22e+00 |
Pos logit contributions | |
its | +0.191 |
ases | +0.182 |
asing | +0.171 |
olds | +0.167 |
ins | +0.166 |
Neg logit contributions | |
he | -0.146 |
roof | -0.114 |
field | -0.112 |
while | -0.112 |
moon | -0.111 |
Token | felt |
Token ID | 2936 |
Feature activation | -3.09e+00 |
Pos logit contributions | |
its | +0.149 |
ases | +0.142 |
asing | +0.135 |
olds | +0.133 |
ins | +0.132 |
Neg logit contributions | |
he | -0.067 |
roof | -0.046 |
field | -0.045 |
while | -0.045 |
moon | -0.044 |
Token | and |
Token ID | 290 |
Feature activation | +1.97e+00 |
Pos logit contributions | |
its | +0.106 |
ases | +0.101 |
asing | +0.098 |
olds | +0.096 |
ins | +0.094 |
Neg logit contributions | |
he | -0.040 |
roof | -0.036 |
moon | -0.035 |
field | -0.034 |
bus | -0.033 |
Token | told |
Token ID | 1297 |
Feature activation | -2.77e+00 |
Pos logit contributions | |
he | +0.031 |
roof | +0.028 |
moon | +0.028 |
field | +0.027 |
mountain | +0.027 |
Neg logit contributions | |
its | -0.061 |
ases | -0.058 |
olds | -0.056 |
asing | -0.055 |
ins | -0.054 |
Token | him |
Token ID | 683 |
Feature activation | -1.83e+00 |
Pos logit contributions | |
its | +0.092 |
ases | +0.087 |
olds | +0.082 |
asing | +0.080 |
ins | +0.080 |
Neg logit contributions | |
he | -0.044 |
field | -0.037 |
roof | -0.036 |
hill | -0.035 |
mountain | -0.035 |
Token | that |
Token ID | 326 |
Feature activation | +3.91e-01 |
Pos logit contributions | |
its | +0.066 |
ases | +0.063 |
olds | +0.062 |
asing | +0.061 |
arts | +0.059 |
Neg logit contributions | |
he | -0.022 |
field | -0.018 |
moon | -0.018 |
valley | -0.018 |
roof | -0.017 |
Token | he |
Token ID | 339 |
Feature activation | -1.44e+00 |
Pos logit contributions | |
he | +0.002 |
field | +0.001 |
roof | +0.001 |
bus | +0.001 |
Under | +0.001 |
Neg logit contributions | |
its | -0.016 |
ases | -0.015 |
asing | -0.015 |
olds | -0.015 |
ins | -0.015 |
Token | But |
Token ID | 1537 |
Feature activation | +2.30e+00 |
Pos logit contributions | |
its | +0.072 |
ases | +0.068 |
olds | +0.063 |
asing | +0.063 |
ins | +0.062 |
Neg logit contributions | |
he | -0.058 |
field | -0.046 |
roof | -0.046 |
moon | -0.046 |
mountain | -0.045 |
Token | his |
Token ID | 465 |
Feature activation | -2.03e+00 |
Pos logit contributions | |
valley | +0.035 |
river | +0.035 |
field | +0.034 |
landscape | +0.034 |
Pole | +0.034 |
Neg logit contributions | |
ases | -0.062 |
olds | -0.059 |
its | -0.058 |
asing | -0.057 |
ps | -0.056 |
Token | mom |
Token ID | 1995 |
Feature activation | -1.57e+00 |
Pos logit contributions | |
its | +0.070 |
ases | +0.066 |
olds | +0.062 |
asing | +0.062 |
ins | +0.060 |
Neg logit contributions | |
he | -0.034 |
while | -0.023 |
field | -0.023 |
she | -0.022 |
valley | -0.022 |
Token | could |
Token ID | 714 |
Feature activation | -8.49e-01 |
Pos logit contributions | |
its | +0.055 |
ases | +0.052 |
asing | +0.051 |
olds | +0.050 |
ins | +0.049 |
Neg logit contributions | |
he | -0.023 |
valley | -0.017 |
roof | -0.017 |
sail | -0.017 |
mountain | -0.017 |
Token | not |
Token ID | 407 |
Feature activation | -2.52e+00 |
Pos logit contributions | |
its | +0.025 |
ases | +0.024 |
asing | +0.022 |
olds | +0.022 |
ins | +0.022 |
Neg logit contributions | |
he | -0.018 |
field | -0.014 |
roof | -0.014 |
mountain | -0.014 |
moon | -0.014 |
Token | help |
Token ID | 1037 |
Feature activation | -4.08e+00 |
Pos logit contributions | |
its | +0.088 |
ases | +0.084 |
asing | +0.080 |
olds | +0.079 |
ins | +0.078 |
Neg logit contributions | |
he | -0.043 |
while | -0.030 |
roof | -0.029 |
ride | -0.029 |
wide | -0.029 |
Token | him |
Token ID | 683 |
Feature activation | -2.42e-01 |
Pos logit contributions | |
its | +0.146 |
ases | +0.140 |
asing | +0.131 |
olds | +0.131 |
ins | +0.129 |
Neg logit contributions | |
he | -0.062 |
field | -0.043 |
roof | -0.043 |
moon | -0.042 |
mountain | -0.041 |
Token | . |
Token ID | 13 |
Feature activation | +3.47e-01 |
Pos logit contributions | |
its | +0.009 |
ases | +0.009 |
asing | +0.009 |
olds | +0.009 |
ins | +0.008 |
Neg logit contributions | |
he | -0.003 |
roof | -0.002 |
field | -0.002 |
while | -0.002 |
moon | -0.002 |
Token | She |
Token ID | 1375 |
Feature activation | +1.86e+00 |
Pos logit contributions | |
he | +0.004 |
sail | +0.003 |
mountain | +0.003 |
ride | +0.003 |
rocket | +0.003 |
Neg logit contributions | |
its | -0.013 |
ases | -0.013 |
olds | -0.012 |
asing | -0.012 |
ums | -0.011 |
Token | was |
Token ID | 373 |
Feature activation | -1.81e+00 |
Pos logit contributions | |
he | +0.027 |
while | +0.017 |
field | +0.017 |
car | +0.016 |
wide | +0.016 |
Neg logit contributions | |
its | -0.069 |
ases | -0.067 |
asing | -0.064 |
olds | -0.062 |
ins | -0.062 |
Token | holding |
Token ID | 4769 |
Feature activation | -3.64e+00 |
Pos logit contributions | |
its | +0.065 |
ases | +0.062 |
asing | +0.059 |
olds | +0.057 |
ins | +0.057 |
Neg logit contributions | |
he | -0.030 |
outside | -0.020 |
wide | -0.020 |
while | -0.020 |
dark | -0.020 |
Token | finished |
Token ID | 5201 |
Feature activation | +8.11e-01 |
Pos logit contributions | |
its | +0.133 |
ases | +0.128 |
asing | +0.122 |
olds | +0.119 |
ins | +0.119 |
Neg logit contributions | |
he | -0.057 |
roof | -0.039 |
while | -0.038 |
field | -0.038 |
moon | -0.038 |
Token | first |
Token ID | 717 |
Feature activation | +8.30e-01 |
Pos logit contributions | |
he | +0.009 |
roof | +0.006 |
moon | +0.006 |
field | +0.006 |
mountain | +0.005 |
Neg logit contributions | |
its | -0.032 |
ases | -0.030 |
asing | -0.029 |
olds | -0.029 |
ins | -0.028 |
Token | and |
Token ID | 290 |
Feature activation | +3.90e+00 |
Pos logit contributions | |
he | +0.011 |
roof | +0.007 |
field | +0.007 |
moon | +0.007 |
while | +0.007 |
Neg logit contributions | |
its | -0.031 |
ases | -0.030 |
asing | -0.029 |
olds | -0.028 |
ins | -0.028 |
Token | smiled |
Token ID | 13541 |
Feature activation | -2.33e+00 |
Pos logit contributions | |
he | +0.045 |
roof | +0.041 |
moon | +0.038 |
fireplace | +0.038 |
valley | +0.037 |
Neg logit contributions | |
its | -0.138 |
ases | -0.133 |
olds | -0.127 |
asing | -0.127 |
ins | -0.123 |
Token | . |
Token ID | 13 |
Feature activation | +7.74e-02 |
Pos logit contributions | |
its | +0.072 |
ases | +0.069 |
ums | +0.067 |
asing | +0.066 |
olds | +0.065 |
Neg logit contributions | |
he | -0.033 |
roof | -0.031 |
hill | -0.031 |
mountain | -0.030 |
valley | -0.029 |
Token | Che |
Token ID | 2580 |
Feature activation | -1.11e+01 |
Pos logit contributions | |
he | +0.001 |
roof | +0.001 |
while | +0.000 |
field | +0.000 |
bus | +0.000 |
Neg logit contributions | |
its | -0.003 |
ases | -0.003 |
olds | -0.003 |
asing | -0.003 |
ins | -0.003 |
Token | wing |
Token ID | 5469 |
Feature activation | -1.66e+00 |
Pos logit contributions | |
ases | +0.224 |
its | +0.221 |
olds | +0.207 |
ums | +0.187 |
asing | +0.186 |
Neg logit contributions | |
he | -0.325 |
set | -0.281 |
looked | -0.278 |
field | -0.273 |
went | -0.269 |
Token | was |
Token ID | 373 |
Feature activation | -1.66e+00 |
Pos logit contributions | |
its | +0.057 |
ases | +0.054 |
asing | +0.051 |
olds | +0.051 |
ins | +0.050 |
Neg logit contributions | |
he | -0.027 |
roof | -0.020 |
field | -0.019 |
while | -0.019 |
moon | -0.019 |
Token | a |
Token ID | 257 |
Feature activation | +3.19e-01 |
Pos logit contributions | |
its | +0.060 |
ases | +0.057 |
asing | +0.055 |
olds | +0.054 |
ins | +0.053 |
Neg logit contributions | |
he | -0.025 |
roof | -0.017 |
field | -0.017 |
mountain | -0.016 |
while | -0.016 |
Token | fun |
Token ID | 1257 |
Feature activation | -5.60e-01 |
Pos logit contributions | |
he | +0.006 |
Under | +0.005 |
outside | +0.005 |
roof | +0.005 |
fields | +0.005 |
Neg logit contributions | |
its | -0.009 |
ases | -0.008 |
olds | -0.008 |
asing | -0.008 |
cents | -0.008 |
Token | new |
Token ID | 649 |
Feature activation | -5.15e-01 |
Pos logit contributions | |
its | +0.018 |
ases | +0.017 |
asing | +0.016 |
olds | +0.016 |
ins | +0.015 |
Neg logit contributions | |
he | -0.009 |
roof | -0.007 |
moon | -0.007 |
valley | -0.007 |
A | -0.007 |
Token | She |
Token ID | 1375 |
Feature activation | +3.05e+00 |
Pos logit contributions | |
rocket | +0.049 |
he | +0.048 |
plane | +0.047 |
airplane | +0.047 |
boat | +0.046 |
Neg logit contributions | |
its | -0.069 |
olds | -0.069 |
ums | -0.065 |
ases | -0.064 |
irk | -0.060 |
Token | blinked |
Token ID | 40360 |
Feature activation | +4.26e-01 |
Pos logit contributions | |
he | +0.050 |
roof | +0.046 |
fireplace | +0.045 |
wagon | +0.045 |
rocket | +0.042 |
Neg logit contributions | |
its | -0.089 |
ases | -0.081 |
ins | -0.080 |
asing | -0.080 |
olds | -0.077 |
Token | her |
Token ID | 607 |
Feature activation | -1.90e+00 |
Pos logit contributions | |
hill | +0.008 |
roof | +0.007 |
mountain | +0.007 |
he | +0.007 |
river | +0.007 |
Neg logit contributions | |
ases | -0.011 |
its | -0.011 |
asing | -0.010 |
arts | -0.010 |
ins | -0.010 |
Token | eyes |
Token ID | 2951 |
Feature activation | -1.25e+00 |
Pos logit contributions | |
its | +0.053 |
ases | +0.050 |
asing | +0.046 |
olds | +0.046 |
ins | +0.044 |
Neg logit contributions | |
he | -0.045 |
while | -0.035 |
field | -0.034 |
ride | -0.034 |
roof | -0.033 |
Token | and |
Token ID | 290 |
Feature activation | +4.22e+00 |
Pos logit contributions | |
its | +0.039 |
asing | +0.035 |
ases | +0.035 |
arts | +0.034 |
olds | +0.033 |
Neg logit contributions | |
he | -0.018 |
cabin | -0.015 |
wagon | -0.015 |
hut | -0.015 |
valley | -0.015 |
Token | imm |
Token ID | 2296 |
Feature activation | -1.03e+01 |
Pos logit contributions | |
he | +0.072 |
wagon | +0.069 |
roof | +0.066 |
car | +0.065 |
cabin | +0.065 |
Neg logit contributions | |
its | -0.115 |
ases | -0.110 |
asing | -0.109 |
olds | -0.104 |
preferences | -0.100 |
Token | ag |
Token ID | 363 |
Feature activation | -4.15e+00 |
Pos logit contributions | |
olds | +0.166 |
cents | +0.166 |
its | +0.158 |
supervision | +0.153 |
asing | +0.150 |
Neg logit contributions | |
outside | -0.270 |
he | -0.261 |
down | -0.258 |
stable | -0.256 |
field | -0.250 |
Token | ined |
Token ID | 1389 |
Feature activation | -1.08e+00 |
Pos logit contributions | |
its | +0.141 |
ases | +0.135 |
asing | +0.130 |
olds | +0.129 |
ins | +0.125 |
Neg logit contributions | |
he | -0.072 |
while | -0.049 |
wide | -0.049 |
roof | -0.048 |
moon | -0.048 |
Token | she |
Token ID | 673 |
Feature activation | +3.71e-01 |
Pos logit contributions | |
its | +0.035 |
ases | +0.033 |
asing | +0.032 |
arts | +0.032 |
ins | +0.031 |
Neg logit contributions | |
he | -0.016 |
hill | -0.014 |
Pole | -0.014 |
ride | -0.014 |
mountain | -0.013 |
Token | could |
Token ID | 714 |
Feature activation | -1.21e+00 |
Pos logit contributions | |
he | +0.006 |
roof | +0.004 |
wagon | +0.004 |
mountain | +0.004 |
valley | +0.004 |
Neg logit contributions | |
its | -0.013 |
ases | -0.012 |
asing | -0.012 |
olds | -0.011 |
ins | -0.011 |
Token | see |
Token ID | 766 |
Feature activation | +3.69e-01 |
Pos logit contributions | |
its | +0.037 |
ases | +0.035 |
olds | +0.032 |
asing | +0.032 |
arts | +0.031 |
Neg logit contributions | |
he | -0.023 |
wagon | -0.019 |
field | -0.018 |
car | -0.018 |
rocket | -0.018 |
Token | . |
Token ID | 13 |
Feature activation | +4.54e-01 |
Pos logit contributions | |
its | +0.017 |
ases | +0.016 |
asing | +0.015 |
olds | +0.015 |
ins | +0.015 |
Neg logit contributions | |
he | -0.008 |
Pole | -0.006 |
Suddenly | -0.006 |
valley | -0.006 |
While | -0.006 |
Token | He |
Token ID | 679 |
Feature activation | +1.75e+00 |
Pos logit contributions | |
mountain | +0.005 |
airplane | +0.005 |
field | +0.005 |
valley | +0.005 |
he | +0.004 |
Neg logit contributions | |
its | -0.016 |
olds | -0.016 |
ases | -0.016 |
ums | -0.014 |
asing | -0.014 |
Token | says |
Token ID | 1139 |
Feature activation | -1.90e+00 |
Pos logit contributions | |
he | +0.045 |
roof | +0.036 |
while | +0.036 |
field | +0.036 |
moon | +0.035 |
Neg logit contributions | |
its | -0.045 |
ases | -0.043 |
asing | -0.040 |
olds | -0.039 |
ins | -0.038 |
Token | , |
Token ID | 11 |
Feature activation | +2.46e+00 |
Pos logit contributions | |
its | +0.073 |
ases | +0.072 |
olds | +0.069 |
asing | +0.068 |
ins | +0.067 |
Neg logit contributions | |
he | -0.018 |
field | -0.015 |
mountain | -0.013 |
hill | -0.013 |
roof | -0.013 |
Token | " |
Token ID | 366 |
Feature activation | -4.16e+00 |
Pos logit contributions | |
valley | +0.029 |
he | +0.028 |
mountain | +0.027 |
field | +0.027 |
hut | +0.027 |
Neg logit contributions | |
its | -0.079 |
olds | -0.076 |
ases | -0.075 |
asing | -0.072 |
ins | -0.070 |
Token | L |
Token ID | 43 |
Feature activation | -9.80e+00 |
Pos logit contributions | |
its | +0.127 |
ases | +0.121 |
asing | +0.115 |
olds | +0.113 |
ins | +0.112 |
Neg logit contributions | |
he | -0.089 |
while | -0.066 |
field | -0.065 |
car | -0.064 |
wide | -0.064 |
Token | ily |
Token ID | 813 |
Feature activation | -3.41e+00 |
Pos logit contributions | |
its | +0.200 |
ases | +0.182 |
asing | +0.172 |
olds | +0.168 |
preferences | +0.165 |
Neg logit contributions | |
he | -0.292 |
she | -0.246 |
outside | -0.236 |
while | -0.236 |
car | -0.235 |
Token | , |
Token ID | 11 |
Feature activation | +1.12e+00 |
Pos logit contributions | |
its | +0.133 |
ases | +0.130 |
asing | +0.122 |
olds | +0.122 |
ins | +0.119 |
Neg logit contributions | |
he | -0.036 |
roof | -0.024 |
valley | -0.023 |
moon | -0.023 |
field | -0.023 |
Token | Lily |
Token ID | 20037 |
Feature activation | -4.09e+00 |
Pos logit contributions | |
he | +0.019 |
field | +0.016 |
Suddenly | +0.016 |
fields | +0.016 |
valley | +0.016 |
Neg logit contributions | |
its | -0.034 |
ases | -0.034 |
asing | -0.032 |
olds | -0.031 |
arts | -0.030 |
Token | , |
Token ID | 11 |
Feature activation | +1.03e+00 |
Pos logit contributions | |
its | +0.146 |
ases | +0.144 |
asing | +0.133 |
olds | +0.131 |
ins | +0.130 |
Neg logit contributions | |
he | -0.052 |
roof | -0.041 |
valley | -0.039 |
moon | -0.039 |
While | -0.039 |
Token | are |
Token ID | 389 |
Feature activation | -2.52e+00 |
Pos logit contributions | |
he | +0.019 |
field | +0.016 |
Suddenly | +0.016 |
valley | +0.015 |
fields | +0.015 |
Neg logit contributions | |
its | -0.031 |
ases | -0.030 |
asing | -0.028 |
olds | -0.027 |
arts | -0.027 |
Token | at |
Token ID | 379 |
Feature activation | -3.44e+00 |
Pos logit contributions | |
its | +0.163 |
ases | +0.156 |
asing | +0.149 |
olds | +0.147 |
ins | +0.145 |
Neg logit contributions | |
he | -0.053 |
while | -0.032 |
field | -0.029 |
roof | -0.028 |
she | -0.028 |
Token | all |
Token ID | 477 |
Feature activation | -2.20e+00 |
Pos logit contributions | |
its | +0.130 |
ases | +0.125 |
asing | +0.122 |
olds | +0.119 |
ins | +0.118 |
Neg logit contributions | |
he | -0.048 |
while | -0.030 |
field | -0.029 |
roof | -0.029 |
wide | -0.028 |
Token | . |
Token ID | 13 |
Feature activation | -6.83e-01 |
Pos logit contributions | |
its | +0.086 |
ases | +0.083 |
asing | +0.080 |
olds | +0.078 |
ins | +0.077 |
Neg logit contributions | |
he | -0.027 |
while | -0.016 |
field | -0.015 |
roof | -0.015 |
wide | -0.015 |
Token |
|
Token ID | 198 |
Feature activation | -4.81e+00 |
Pos logit contributions | |
its | +0.028 |
ases | +0.027 |
asing | +0.026 |
olds | +0.025 |
ins | +0.025 |
Neg logit contributions | |
he | -0.007 |
roof | -0.004 |
field | -0.004 |
while | -0.004 |
mountain | -0.003 |
Token |
|
Token ID | 198 |
Feature activation | -3.89e+00 |
Pos logit contributions | |
its | +0.208 |
ases | +0.204 |
asing | +0.198 |
ins | +0.191 |
olds | +0.189 |
Neg logit contributions | |
he | -0.041 |
wide | -0.013 |
while | -0.013 |
He | -0.012 |
she | -0.012 |
Token | L |
Token ID | 43 |
Feature activation | -9.31e+00 |
Pos logit contributions | |
its | +0.125 |
ases | +0.122 |
asing | +0.114 |
olds | +0.111 |
ins | +0.110 |
Neg logit contributions | |
he | -0.077 |
while | -0.056 |
she | -0.055 |
wide | -0.054 |
field | -0.054 |
Token | ily |
Token ID | 813 |
Feature activation | -3.50e+00 |
Pos logit contributions | |
its | +0.200 |
ases | +0.186 |
asing | +0.169 |
olds | +0.167 |
ums | +0.164 |
Neg logit contributions | |
he | -0.262 |
she | -0.220 |
outside | -0.217 |
while | -0.216 |
Tim | -0.215 |
Token | opened |
Token ID | 4721 |
Feature activation | +6.64e-01 |
Pos logit contributions | |
its | +0.119 |
ases | +0.113 |
asing | +0.107 |
olds | +0.106 |
ins | +0.105 |
Neg logit contributions | |
he | -0.059 |
roof | -0.043 |
mountain | -0.041 |
valley | -0.041 |
field | -0.041 |
Token | her |
Token ID | 607 |
Feature activation | -7.84e-01 |
Pos logit contributions | |
he | +0.009 |
roof | +0.007 |
mountain | +0.006 |
ride | +0.006 |
river | +0.006 |
Neg logit contributions | |
its | -0.024 |
ases | -0.023 |
olds | -0.022 |
asing | -0.021 |
ins | -0.021 |
Token | eyes |
Token ID | 2951 |
Feature activation | -1.18e+00 |
Pos logit contributions | |
its | +0.024 |
ases | +0.023 |
olds | +0.021 |
asing | +0.021 |
ins | +0.021 |
Neg logit contributions | |
he | -0.016 |
while | -0.012 |
roof | -0.012 |
valley | -0.012 |
ride | -0.012 |
Token | and |
Token ID | 290 |
Feature activation | +3.76e+00 |
Pos logit contributions | |
its | +0.040 |
ases | +0.037 |
asing | +0.036 |
olds | +0.035 |
ins | +0.034 |
Neg logit contributions | |
he | -0.017 |
roof | -0.014 |
valley | -0.013 |
cabin | -0.013 |
bicycle | -0.013 |
Token | out |
Token ID | 503 |
Feature activation | +4.40e-01 |
Pos logit contributions | |
its | +0.018 |
ases | +0.016 |
arts | +0.016 |
olds | +0.015 |
asing | +0.015 |
Neg logit contributions | |
he | -0.009 |
hill | -0.008 |
ride | -0.008 |
mountain | -0.007 |
field | -0.007 |
Token | that |
Token ID | 326 |
Feature activation | +2.00e+00 |
Pos logit contributions | |
he | +0.006 |
Pole | +0.006 |
ride | +0.006 |
hill | +0.006 |
river | +0.005 |
Neg logit contributions | |
its | -0.014 |
ases | -0.013 |
asing | -0.013 |
olds | -0.012 |
arts | -0.012 |
Token | Bub |
Token ID | 19645 |
Feature activation | -5.34e+00 |
Pos logit contributions | |
he | +0.020 |
bus | +0.019 |
field | +0.018 |
Pole | +0.017 |
wagon | +0.017 |
Neg logit contributions | |
its | -0.069 |
ases | -0.065 |
asing | -0.063 |
olds | -0.062 |
ins | -0.060 |
Token | bles |
Token ID | 7689 |
Feature activation | -8.72e-01 |
Pos logit contributions | |
its | +0.093 |
ases | +0.085 |
olds | +0.078 |
asing | +0.074 |
ins | +0.066 |
Neg logit contributions | |
he | -0.181 |
field | -0.151 |
while | -0.151 |
sail | -0.150 |
car | -0.150 |
Token | had |
Token ID | 550 |
Feature activation | -1.18e+00 |
Pos logit contributions | |
its | +0.029 |
ases | +0.028 |
asing | +0.026 |
olds | +0.025 |
ins | +0.025 |
Neg logit contributions | |
he | -0.013 |
moon | -0.010 |
roof | -0.010 |
rocket | -0.010 |
A | -0.010 |
Token | che |
Token ID | 1125 |
Feature activation | -9.37e+00 |
Pos logit contributions | |
its | +0.033 |
ases | +0.031 |
olds | +0.029 |
asing | +0.029 |
ins | +0.029 |
Neg logit contributions | |
he | -0.025 |
hill | -0.020 |
mountain | -0.020 |
field | -0.020 |
rocket | -0.020 |
Token | wed |
Token ID | 19103 |
Feature activation | -3.53e+00 |
Pos logit contributions | |
olds | +0.145 |
its | +0.128 |
ums | +0.118 |
asing | +0.116 |
apology | +0.112 |
Neg logit contributions | |
he | -0.286 |
set | -0.261 |
looked | -0.253 |
outside | -0.251 |
searched | -0.250 |
Token | on |
Token ID | 319 |
Feature activation | -1.68e+00 |
Pos logit contributions | |
its | +0.131 |
ases | +0.125 |
asing | +0.119 |
olds | +0.117 |
ins | +0.117 |
Neg logit contributions | |
he | -0.047 |
roof | -0.032 |
field | -0.031 |
mountain | -0.031 |
ride | -0.030 |
Token | everything |
Token ID | 2279 |
Feature activation | +3.84e-02 |
Pos logit contributions | |
its | +0.061 |
ases | +0.058 |
asing | +0.054 |
olds | +0.054 |
arts | +0.053 |
Neg logit contributions | |
he | -0.023 |
ride | -0.015 |
field | -0.015 |
bus | -0.015 |
mountain | -0.015 |
Token | in |
Token ID | 287 |
Feature activation | -8.47e-01 |
Pos logit contributions | |
he | +0.000 |
bus | +0.000 |
While | +0.000 |
river | +0.000 |
field | +0.000 |
Neg logit contributions | |
its | -0.001 |
ases | -0.001 |
asing | -0.001 |
ins | -0.001 |
olds | -0.001 |
Token | her |
Token ID | 607 |
Feature activation | -5.55e-01 |
Pos logit contributions | |
its | +0.035 |
ases | +0.034 |
asing | +0.033 |
olds | +0.032 |
ins | +0.032 |
Neg logit contributions | |
he | -0.009 |
while | -0.004 |
field | -0.004 |
roof | -0.004 |
wide | -0.004 |