Token | . |
Token ID | 13 |
Feature activation | +1.30e-01 |
Pos logit contributions | |
els | +0.008 |
ool | +0.007 |
hole | +0.007 |
ps | +0.007 |
globe | +0.007 |
Neg logit contributions | |
Jack | -0.008 |
He | -0.008 |
His | -0.008 |
They | -0.007 |
Thanks | -0.007 |
Token | He |
Token ID | 679 |
Feature activation | +4.86e-02 |
Pos logit contributions | |
Jack | +0.148 |
His | +0.142 |
David | +0.137 |
Thanks | +0.130 |
Kids | +0.128 |
Neg logit contributions | |
globe | -0.252 |
bulb | -0.252 |
trunk | -0.252 |
hole | -0.248 |
chim | -0.247 |
Token | likes |
Token ID | 7832 |
Feature activation | +8.65e-03 |
Pos logit contributions | |
His | +0.087 |
Jack | +0.087 |
He | +0.086 |
Kids | +0.080 |
David | +0.079 |
Neg logit contributions | |
els | -0.071 |
hole | -0.069 |
bulb | -0.068 |
globe | -0.068 |
ool | -0.067 |
Token | it |
Token ID | 340 |
Feature activation | -5.76e-02 |
Pos logit contributions | |
His | +0.010 |
Jack | +0.010 |
He | +0.010 |
His | +0.009 |
They | +0.009 |
Neg logit contributions | |
hole | -0.019 |
ool | -0.018 |
atop | -0.018 |
els | -0.017 |
globe | -0.017 |
Token | too |
Token ID | 1165 |
Feature activation | +1.15e-02 |
Pos logit contributions | |
els | +0.109 |
hole | +0.102 |
ool | +0.102 |
ening | +0.099 |
globe | +0.097 |
Neg logit contributions | |
He | -0.090 |
His | -0.088 |
He | -0.083 |
His | -0.081 |
They | -0.078 |
Token | ." |
Token ID | 526 |
Feature activation | +1.45e-01 |
Pos logit contributions | |
He | +0.020 |
He | +0.018 |
They | +0.018 |
His | +0.018 |
Jack | +0.017 |
Neg logit contributions | |
els | -0.023 |
ps | -0.021 |
hole | -0.021 |
et | -0.020 |
atop | -0.020 |
Token |
|
Token ID | 198 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
Thanks | +0.194 |
David | +0.186 |
Jack | +0.186 |
Un | +0.184 |
Sorry | +0.180 |
Neg logit contributions | |
trunk | -0.228 |
bulb | -0.225 |
globe | -0.222 |
els | -0.222 |
pieces | -0.222 |
Token |
|
Token ID | 198 |
Feature activation | +1.17e-01 |
Pos logit contributions | |
Thanks | +0.137 |
Jack | +0.132 |
Sorry | +0.129 |
David | +0.128 |
His | +0.127 |
Neg logit contributions | |
els | -0.199 |
globe | -0.196 |
bulb | -0.196 |
ool | -0.191 |
hole | -0.191 |
Token | Tom |
Token ID | 13787 |
Feature activation | +1.58e-02 |
Pos logit contributions | |
Something | +0.114 |
David | +0.113 |
His | +0.110 |
Thanks | +0.110 |
oph | +0.107 |
Neg logit contributions | |
globe | -0.212 |
bulb | -0.211 |
els | -0.207 |
trunk | -0.204 |
cut | -0.202 |
Token | nodded |
Token ID | 14464 |
Feature activation | -7.16e-03 |
Pos logit contributions | |
His | +0.029 |
Jack | +0.029 |
He | +0.028 |
David | +0.027 |
Thanks | +0.027 |
Neg logit contributions | |
els | -0.022 |
ps | -0.021 |
hole | -0.021 |
bulb | -0.021 |
et | -0.021 |
Token | . |
Token ID | 13 |
Feature activation | +1.20e-01 |
Pos logit contributions | |
ps | +0.005 |
els | +0.005 |
ening | +0.005 |
bulb | +0.005 |
et | +0.005 |
Neg logit contributions | |
Jack | -0.018 |
Kids | -0.017 |
David | -0.017 |
His | -0.017 |
He | -0.017 |
Token | Tom |
Token ID | 13787 |
Feature activation | +3.27e-02 |
Pos logit contributions | |
Jack | +0.038 |
His | +0.037 |
He | +0.037 |
Kids | +0.032 |
Dad | +0.031 |
Neg logit contributions | |
els | -0.081 |
bulb | -0.080 |
hole | -0.079 |
ening | -0.079 |
ps | -0.078 |
Token | and |
Token ID | 290 |
Feature activation | +8.63e-02 |
Pos logit contributions | |
His | +0.053 |
Jack | +0.053 |
He | +0.052 |
Kids | +0.048 |
David | +0.047 |
Neg logit contributions | |
els | -0.054 |
bulb | -0.052 |
hole | -0.052 |
globe | -0.051 |
ool | -0.051 |
Token | Lily |
Token ID | 20037 |
Feature activation | +4.99e-02 |
Pos logit contributions | |
Finally | +0.143 |
Thanks | +0.141 |
Days | +0.141 |
Sorry | +0.139 |
Even | +0.139 |
Neg logit contributions | |
bulb | -0.117 |
ring | -0.114 |
globe | -0.113 |
on | -0.112 |
matches | -0.111 |
Token | are |
Token ID | 389 |
Feature activation | +6.05e-02 |
Pos logit contributions | |
His | +0.074 |
Jack | +0.072 |
He | +0.069 |
Kids | +0.065 |
David | +0.065 |
Neg logit contributions | |
els | -0.089 |
bulb | -0.087 |
hole | -0.087 |
globe | -0.085 |
ps | -0.085 |
Token | friends |
Token ID | 2460 |
Feature activation | +1.54e-02 |
Pos logit contributions | |
His | +0.124 |
Jack | +0.121 |
He | +0.118 |
Kids | +0.114 |
Thanks | +0.112 |
Neg logit contributions | |
els | -0.077 |
bulb | -0.074 |
ps | -0.072 |
hole | -0.072 |
hanging | -0.071 |
Token | . |
Token ID | 13 |
Feature activation | +1.45e-01 |
Pos logit contributions | |
Jack | +0.040 |
Kids | +0.040 |
His | +0.039 |
David | +0.039 |
Un | +0.038 |
Neg logit contributions | |
hole | -0.008 |
els | -0.008 |
bulb | -0.008 |
et | -0.007 |
globe | -0.007 |
Token | They |
Token ID | 1119 |
Feature activation | +5.83e-02 |
Pos logit contributions | |
Jack | +0.153 |
His | +0.151 |
He | +0.141 |
Brother | +0.137 |
David | +0.137 |
Neg logit contributions | |
els | -0.284 |
hole | -0.283 |
ool | -0.280 |
trunk | -0.279 |
globe | -0.279 |
Token | like |
Token ID | 588 |
Feature activation | +3.94e-02 |
Pos logit contributions | |
His | +0.094 |
Jack | +0.094 |
He | +0.090 |
Kids | +0.086 |
David | +0.086 |
Neg logit contributions | |
bulb | -0.094 |
els | -0.093 |
globe | -0.091 |
ps | -0.091 |
et | -0.090 |
Token | to |
Token ID | 284 |
Feature activation | +3.66e-02 |
Pos logit contributions | |
His | +0.062 |
Jack | +0.062 |
He | +0.060 |
Kids | +0.056 |
David | +0.056 |
Neg logit contributions | |
els | -0.066 |
bulb | -0.065 |
globe | -0.063 |
hole | -0.063 |
ps | -0.062 |
Token | play |
Token ID | 711 |
Feature activation | -2.61e-02 |
Pos logit contributions | |
Jack | +0.045 |
His | +0.045 |
He | +0.044 |
Kids | +0.041 |
David | +0.040 |
Neg logit contributions | |
els | -0.073 |
bulb | -0.073 |
globe | -0.071 |
ps | -0.070 |
ool | -0.069 |
Token | games |
Token ID | 1830 |
Feature activation | -7.99e-02 |
Pos logit contributions | |
hole | +0.034 |
els | +0.033 |
ening | +0.033 |
bulb | +0.032 |
ool | +0.032 |
Neg logit contributions | |
His | -0.053 |
Jack | -0.053 |
He | -0.051 |
They | -0.048 |
Thanks | -0.047 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +5.91e-02 |
Pos logit contributions | |
Jack | +0.154 |
His | +0.152 |
He | +0.149 |
Thanks | +0.148 |
David | +0.148 |
Neg logit contributions | |
bulb | -0.245 |
trunk | -0.241 |
els | -0.240 |
chim | -0.238 |
globe | -0.235 |
Token | L |
Token ID | 43 |
Feature activation | +7.29e-02 |
Pos logit contributions | |
Jack | +0.062 |
His | +0.061 |
He | +0.060 |
Kids | +0.053 |
David | +0.051 |
Neg logit contributions | |
els | -0.133 |
bulb | -0.130 |
hole | -0.129 |
ps | -0.128 |
ening | -0.128 |
Token | ily |
Token ID | 813 |
Feature activation | +6.21e-02 |
Pos logit contributions | |
His | +0.108 |
Jack | +0.106 |
He | +0.104 |
Thanks | +0.097 |
Kids | +0.096 |
Neg logit contributions | |
els | -0.132 |
hole | -0.126 |
bulb | -0.126 |
ool | -0.125 |
ps | -0.124 |
Token | was |
Token ID | 373 |
Feature activation | +8.29e-02 |
Pos logit contributions | |
Jack | +0.101 |
He | +0.099 |
His | +0.098 |
Kids | +0.090 |
Dad | +0.089 |
Neg logit contributions | |
els | -0.103 |
globe | -0.101 |
bulb | -0.100 |
hole | -0.099 |
ps | -0.099 |
Token | sick |
Token ID | 6639 |
Feature activation | -6.15e-02 |
Pos logit contributions | |
Jack | +0.155 |
His | +0.154 |
He | +0.151 |
Kids | +0.141 |
David | +0.140 |
Neg logit contributions | |
els | -0.117 |
hole | -0.113 |
globe | -0.113 |
bulb | -0.112 |
ool | -0.109 |
Token | . |
Token ID | 13 |
Feature activation | +1.45e-01 |
Pos logit contributions | |
els | +0.110 |
globe | +0.104 |
hole | +0.103 |
angles | +0.103 |
ool | +0.102 |
Neg logit contributions | |
His | -0.093 |
He | -0.092 |
He | -0.085 |
His | -0.084 |
They | -0.084 |
Token | She |
Token ID | 1375 |
Feature activation | +7.42e-02 |
Pos logit contributions | |
His | +0.131 |
Jack | +0.126 |
He | +0.122 |
Thanks | +0.115 |
Sure | +0.113 |
Neg logit contributions | |
bulb | -0.331 |
els | -0.327 |
ool | -0.324 |
hole | -0.320 |
ening | -0.319 |
Token | had |
Token ID | 550 |
Feature activation | +5.51e-02 |
Pos logit contributions | |
His | +0.112 |
Jack | +0.112 |
He | +0.108 |
David | +0.101 |
Kids | +0.100 |
Neg logit contributions | |
els | -0.130 |
bulb | -0.127 |
hole | -0.126 |
ps | -0.124 |
globe | -0.124 |
Token | a |
Token ID | 257 |
Feature activation | +3.16e-02 |
Pos logit contributions | |
His | +0.085 |
Jack | +0.085 |
He | +0.082 |
Kids | +0.077 |
David | +0.076 |
Neg logit contributions | |
els | -0.094 |
bulb | -0.092 |
hole | -0.092 |
ps | -0.090 |
globe | -0.090 |
Token | sore |
Token ID | 19597 |
Feature activation | -6.94e-02 |
Pos logit contributions | |
His | +0.058 |
Jack | +0.057 |
He | +0.057 |
Brother | +0.053 |
Un | +0.053 |
Neg logit contributions | |
els | -0.047 |
hole | -0.045 |
ps | -0.044 |
ening | -0.043 |
atop | -0.042 |
Token | throat |
Token ID | 13589 |
Feature activation | -9.07e-02 |
Pos logit contributions | |
els | +0.114 |
bulb | +0.111 |
hole | +0.111 |
globe | +0.109 |
ps | +0.108 |
Neg logit contributions | |
His | -0.112 |
Jack | -0.112 |
He | -0.109 |
Kids | -0.102 |
David | -0.101 |
Token | My |
Token ID | 3666 |
Feature activation | +4.17e-02 |
Pos logit contributions | |
els | +0.252 |
globe | +0.245 |
bulb | +0.244 |
ool | +0.241 |
ps | +0.240 |
Neg logit contributions | |
His | -0.144 |
He | -0.144 |
Jack | -0.141 |
They | -0.123 |
Thanks | -0.122 |
Token | necklace |
Token ID | 38987 |
Feature activation | -1.16e-01 |
Pos logit contributions | |
His | +0.067 |
He | +0.066 |
Jack | +0.065 |
His | +0.060 |
Thanks | +0.059 |
Neg logit contributions | |
els | -0.068 |
ool | -0.064 |
globe | -0.064 |
ening | -0.064 |
hole | -0.064 |
Token | is |
Token ID | 318 |
Feature activation | +2.73e-02 |
Pos logit contributions | |
els | +0.187 |
ening | +0.178 |
ps | +0.171 |
cks | +0.164 |
bulb | +0.164 |
Neg logit contributions | |
He | -0.197 |
His | -0.195 |
Jack | -0.183 |
They | -0.176 |
She | -0.175 |
Token | gone |
Token ID | 3750 |
Feature activation | -5.41e-02 |
Pos logit contributions | |
He | +0.032 |
His | +0.031 |
please | +0.028 |
They | +0.027 |
Jack | +0.027 |
Neg logit contributions | |
els | -0.058 |
globe | -0.055 |
ool | -0.053 |
hole | -0.053 |
ening | -0.052 |
Token | forever |
Token ID | 8097 |
Feature activation | -1.19e-02 |
Pos logit contributions | |
els | +0.094 |
ps | +0.082 |
cks | +0.080 |
globe | +0.080 |
blades | +0.079 |
Neg logit contributions | |
He | -0.092 |
Jack | -0.087 |
His | -0.083 |
They | -0.083 |
He | -0.082 |
Token | ." |
Token ID | 526 |
Feature activation | +1.44e-01 |
Pos logit contributions | |
els | +0.020 |
globe | +0.017 |
ps | +0.017 |
ool | +0.017 |
et | +0.017 |
Neg logit contributions | |
He | -0.020 |
His | -0.019 |
Jack | -0.018 |
They | -0.018 |
She | -0.018 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +4.85e-02 |
Pos logit contributions | |
Jack | +0.172 |
His | +0.170 |
Thanks | +0.170 |
Un | +0.169 |
David | +0.166 |
Neg logit contributions | |
bulb | -0.240 |
trunk | -0.237 |
els | -0.237 |
hole | -0.234 |
pieces | -0.231 |
Token | Once |
Token ID | 7454 |
Feature activation | +7.20e-02 |
Pos logit contributions | |
Jack | +0.048 |
His | +0.046 |
He | +0.046 |
Kids | +0.040 |
Dad | +0.038 |
Neg logit contributions | |
els | -0.114 |
bulb | -0.111 |
hole | -0.110 |
ps | -0.110 |
ening | -0.109 |
Token | upon |
Token ID | 2402 |
Feature activation | +8.43e-02 |
Pos logit contributions | |
His | +0.098 |
Jack | +0.097 |
He | +0.095 |
Kids | +0.085 |
David | +0.085 |
Neg logit contributions | |
els | -0.138 |
hole | -0.134 |
bulb | -0.134 |
ps | -0.133 |
globe | -0.132 |
Token | a |
Token ID | 257 |
Feature activation | +6.71e-02 |
Pos logit contributions | |
Kids | +0.184 |
Brother | +0.172 |
Sorry | +0.171 |
His | +0.169 |
Sure | +0.169 |
Neg logit contributions | |
pieces | -0.089 |
flying | -0.080 |
ps | -0.079 |
bulb | -0.077 |
glowing | -0.077 |
Token | time |
Token ID | 640 |
Feature activation | +5.30e-02 |
Pos logit contributions | |
His | +0.160 |
Kids | +0.154 |
He | +0.153 |
Jack | +0.153 |
David | +0.151 |
Neg logit contributions | |
ool | -0.053 |
et | -0.052 |
els | -0.049 |
bulb | -0.048 |
trunk | -0.048 |
Token | ila |
Token ID | 10102 |
Feature activation | +3.83e-02 |
Pos logit contributions | |
His | +0.065 |
He | +0.064 |
Jack | +0.062 |
Thanks | +0.059 |
Sorry | +0.057 |
Neg logit contributions | |
els | -0.074 |
ool | -0.071 |
hole | -0.070 |
bulb | -0.069 |
et | -0.069 |
Token | and |
Token ID | 290 |
Feature activation | +8.61e-02 |
Pos logit contributions | |
Jack | +0.062 |
His | +0.062 |
He | +0.060 |
Dad | +0.056 |
Kids | +0.055 |
Neg logit contributions | |
els | -0.064 |
hole | -0.063 |
globe | -0.062 |
bulb | -0.062 |
ps | -0.062 |
Token | Ben |
Token ID | 3932 |
Feature activation | +3.57e-02 |
Pos logit contributions | |
His | +0.146 |
He | +0.140 |
Jack | +0.140 |
Kids | +0.138 |
enny | +0.134 |
Neg logit contributions | |
bulb | -0.129 |
hanging | -0.127 |
atop | -0.126 |
els | -0.125 |
flying | -0.124 |
Token | are |
Token ID | 389 |
Feature activation | +5.02e-02 |
Pos logit contributions | |
His | +0.068 |
Jack | +0.067 |
He | +0.066 |
Kids | +0.062 |
David | +0.062 |
Neg logit contributions | |
els | -0.049 |
bulb | -0.047 |
hole | -0.047 |
globe | -0.046 |
ps | -0.046 |
Token | friends |
Token ID | 2460 |
Feature activation | +1.51e-02 |
Pos logit contributions | |
His | +0.099 |
Jack | +0.097 |
He | +0.095 |
Kids | +0.090 |
Thanks | +0.089 |
Neg logit contributions | |
els | -0.067 |
bulb | -0.064 |
hole | -0.062 |
ps | -0.062 |
atop | -0.061 |
Token | . |
Token ID | 13 |
Feature activation | +1.44e-01 |
Pos logit contributions | |
His | +0.033 |
Jack | +0.033 |
He | +0.032 |
Kids | +0.030 |
David | +0.030 |
Neg logit contributions | |
els | -0.017 |
hole | -0.016 |
bulb | -0.016 |
globe | -0.015 |
ps | -0.015 |
Token | They |
Token ID | 1119 |
Feature activation | +5.92e-02 |
Pos logit contributions | |
His | +0.141 |
Jack | +0.140 |
He | +0.129 |
David | +0.123 |
Sorry | +0.120 |
Neg logit contributions | |
ool | -0.319 |
els | -0.318 |
bulb | -0.316 |
hole | -0.315 |
globe | -0.314 |
Token | like |
Token ID | 588 |
Feature activation | +4.14e-02 |
Pos logit contributions | |
His | +0.088 |
Jack | +0.088 |
He | +0.085 |
Kids | +0.079 |
David | +0.079 |
Neg logit contributions | |
els | -0.105 |
bulb | -0.103 |
hole | -0.102 |
globe | -0.101 |
ps | -0.100 |
Token | to |
Token ID | 284 |
Feature activation | +3.72e-02 |
Pos logit contributions | |
His | +0.069 |
Jack | +0.069 |
He | +0.067 |
Thanks | +0.063 |
Kids | +0.063 |
Neg logit contributions | |
bulb | -0.064 |
els | -0.063 |
globe | -0.062 |
pieces | -0.060 |
hole | -0.060 |
Token | play |
Token ID | 711 |
Feature activation | -2.50e-02 |
Pos logit contributions | |
Jack | +0.057 |
He | +0.056 |
His | +0.055 |
Kids | +0.052 |
Don | +0.052 |
Neg logit contributions | |
bulb | -0.065 |
els | -0.062 |
trunk | -0.061 |
globe | -0.061 |
matches | -0.060 |
Token | outside |
Token ID | 2354 |
Feature activation | -3.79e-02 |
Pos logit contributions | |
els | +0.026 |
hole | +0.025 |
bulb | +0.025 |
globe | +0.024 |
ening | +0.024 |
Neg logit contributions | |
His | -0.056 |
Jack | -0.056 |
He | -0.055 |
Kids | -0.052 |
David | -0.051 |
Token | ara |
Token ID | 3301 |
Feature activation | +5.29e-02 |
Pos logit contributions | |
His | +0.050 |
Jack | +0.049 |
He | +0.048 |
Kids | +0.043 |
Thanks | +0.043 |
Neg logit contributions | |
els | -0.085 |
hole | -0.082 |
bulb | -0.082 |
globe | -0.081 |
ool | -0.080 |
Token | liked |
Token ID | 8288 |
Feature activation | +3.60e-02 |
Pos logit contributions | |
Jack | +0.072 |
He | +0.072 |
His | +0.072 |
They | +0.066 |
Dad | +0.066 |
Neg logit contributions | |
ool | -0.101 |
els | -0.099 |
bulb | -0.097 |
hole | -0.097 |
ening | -0.096 |
Token | cream |
Token ID | 8566 |
Feature activation | -8.64e-02 |
Pos logit contributions | |
His | +0.057 |
Jack | +0.056 |
He | +0.055 |
Kids | +0.051 |
Thanks | +0.051 |
Neg logit contributions | |
els | -0.060 |
bulb | -0.059 |
globe | -0.058 |
hole | -0.057 |
ps | -0.057 |
Token | very |
Token ID | 845 |
Feature activation | +4.26e-02 |
Pos logit contributions | |
els | +0.104 |
hole | +0.102 |
bulb | +0.101 |
ool | +0.098 |
ening | +0.098 |
Neg logit contributions | |
His | -0.179 |
Jack | -0.176 |
He | -0.174 |
Dad | -0.163 |
Kids | -0.162 |
Token | much |
Token ID | 881 |
Feature activation | -2.53e-02 |
Pos logit contributions | |
His | +0.061 |
He | +0.058 |
Jack | +0.057 |
Dad | +0.053 |
Her | +0.052 |
Neg logit contributions | |
hole | -0.085 |
els | -0.082 |
ening | -0.080 |
ps | -0.080 |
globe | -0.080 |
Token | . |
Token ID | 13 |
Feature activation | +1.44e-01 |
Pos logit contributions | |
hole | +0.036 |
bulb | +0.034 |
ps | +0.034 |
ool | +0.033 |
globe | +0.033 |
Neg logit contributions | |
He | -0.049 |
His | -0.049 |
Jack | -0.046 |
They | -0.046 |
Dad | -0.045 |
Token | She |
Token ID | 1375 |
Feature activation | +6.82e-02 |
Pos logit contributions | |
Jack | +0.117 |
His | +0.116 |
He | +0.108 |
David | +0.095 |
Kids | +0.095 |
Neg logit contributions | |
els | -0.353 |
bulb | -0.347 |
hole | -0.344 |
globe | -0.342 |
ool | -0.341 |
Token | saw |
Token ID | 2497 |
Feature activation | +3.89e-03 |
Pos logit contributions | |
His | +0.098 |
Jack | +0.097 |
He | +0.095 |
Kids | +0.087 |
David | +0.086 |
Neg logit contributions | |
els | -0.125 |
hole | -0.122 |
bulb | -0.122 |
globe | -0.120 |
ps | -0.119 |
Token | a |
Token ID | 257 |
Feature activation | -1.59e-03 |
Pos logit contributions | |
His | +0.006 |
Jack | +0.006 |
He | +0.006 |
Dad | +0.006 |
David | +0.006 |
Neg logit contributions | |
els | -0.006 |
hole | -0.006 |
globe | -0.006 |
bulb | -0.006 |
ool | -0.006 |
Token | big |
Token ID | 1263 |
Feature activation | -6.32e-02 |
Pos logit contributions | |
ps | +0.003 |
els | +0.003 |
hole | +0.002 |
ening | +0.002 |
cks | +0.002 |
Neg logit contributions | |
Jack | -0.003 |
His | -0.003 |
He | -0.003 |
Dad | -0.003 |
Sure | -0.003 |
Token | jar |
Token ID | 17379 |
Feature activation | -1.46e-01 |
Pos logit contributions | |
ps | +0.100 |
hole | +0.093 |
els | +0.090 |
cks | +0.088 |
aces | +0.084 |
Neg logit contributions | |
Jack | -0.111 |
please | -0.107 |
Dad | -0.107 |
He | -0.106 |
His | -0.106 |
Token | safe |
Token ID | 3338 |
Feature activation | -8.45e-02 |
Pos logit contributions | |
els | +0.014 |
hole | +0.014 |
globe | +0.014 |
ool | +0.014 |
bulb | +0.013 |
Neg logit contributions | |
His | -0.012 |
Jack | -0.012 |
He | -0.012 |
Thanks | -0.011 |
David | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | +1.32e-01 |
Pos logit contributions | |
els | +0.129 |
ool | +0.127 |
globe | +0.119 |
hole | +0.119 |
bulb | +0.117 |
Neg logit contributions | |
He | -0.149 |
Jack | -0.144 |
They | -0.143 |
His | -0.142 |
He | -0.133 |
Token | Just |
Token ID | 2329 |
Feature activation | +6.00e-02 |
Pos logit contributions | |
Jack | +0.151 |
His | +0.139 |
David | +0.138 |
Thanks | +0.136 |
Brother | +0.126 |
Neg logit contributions | |
trunk | -0.259 |
chim | -0.255 |
bulb | -0.255 |
hole | -0.252 |
els | -0.248 |
Token | like |
Token ID | 588 |
Feature activation | +3.75e-02 |
Pos logit contributions | |
His | +0.059 |
Jack | +0.057 |
He | +0.055 |
please | +0.052 |
Sure | +0.051 |
Neg logit contributions | |
els | -0.136 |
globe | -0.136 |
ool | -0.133 |
hole | -0.132 |
pieces | -0.132 |
Token | us |
Token ID | 514 |
Feature activation | -2.21e-02 |
Pos logit contributions | |
His | +0.050 |
Jack | +0.050 |
He | +0.049 |
They | +0.044 |
Kids | +0.044 |
Neg logit contributions | |
hole | -0.073 |
els | -0.073 |
ool | -0.071 |
atop | -0.070 |
globe | -0.070 |
Token | ." |
Token ID | 526 |
Feature activation | +1.44e-01 |
Pos logit contributions | |
els | +0.036 |
ool | +0.036 |
globe | +0.035 |
hole | +0.035 |
bulb | +0.034 |
Neg logit contributions | |
His | -0.035 |
He | -0.035 |
Jack | -0.034 |
They | -0.033 |
His | -0.033 |
Token |
|
Token ID | 198 |
Feature activation | +1.16e-01 |
Pos logit contributions | |
Jack | +0.169 |
His | +0.166 |
Thanks | +0.164 |
He | +0.162 |
David | +0.160 |
Neg logit contributions | |
bulb | -0.252 |
els | -0.252 |
trunk | -0.248 |
chim | -0.245 |
globe | -0.241 |
Token |
|
Token ID | 198 |
Feature activation | +1.20e-01 |
Pos logit contributions | |
Thanks | +0.137 |
His | +0.134 |
Jack | +0.134 |
Sorry | +0.132 |
David | +0.127 |
Neg logit contributions | |
els | -0.201 |
bulb | -0.198 |
globe | -0.197 |
trunk | -0.194 |
cks | -0.192 |
Token | Ben |
Token ID | 11696 |
Feature activation | +1.82e-02 |
Pos logit contributions | |
His | +0.100 |
Something | +0.100 |
Thanks | +0.094 |
enny | +0.094 |
David | +0.093 |
Neg logit contributions | |
bulb | -0.226 |
globe | -0.222 |
els | -0.220 |
trunk | -0.217 |
cut | -0.216 |
Token | and |
Token ID | 290 |
Feature activation | +8.27e-02 |
Pos logit contributions | |
His | +0.031 |
Jack | +0.031 |
He | +0.030 |
Kids | +0.028 |
David | +0.028 |
Neg logit contributions | |
els | -0.029 |
bulb | -0.028 |
hole | -0.028 |
globe | -0.027 |
ps | -0.027 |
Token | Lily |
Token ID | 20037 |
Feature activation | +2.23e-02 |
Pos logit contributions | |
Jack | +0.104 |
His | +0.101 |
He | +0.099 |
David | +0.096 |
Thanks | +0.092 |
Neg logit contributions | |
bulb | -0.162 |
els | -0.161 |
ps | -0.159 |
et | -0.156 |
globe | -0.156 |
Token | Tom |
Token ID | 13787 |
Feature activation | +4.53e-02 |
Pos logit contributions | |
Jack | +0.052 |
He | +0.051 |
His | +0.050 |
Kids | +0.042 |
Dad | +0.040 |
Neg logit contributions | |
els | -0.135 |
bulb | -0.131 |
ening | -0.130 |
ps | -0.130 |
hole | -0.130 |
Token | and |
Token ID | 290 |
Feature activation | +9.74e-02 |
Pos logit contributions | |
Jack | +0.068 |
His | +0.068 |
He | +0.067 |
Dad | +0.060 |
Kids | +0.060 |
Neg logit contributions | |
els | -0.080 |
bulb | -0.079 |
globe | -0.079 |
ool | -0.078 |
hole | -0.077 |
Token | Lily |
Token ID | 20037 |
Feature activation | +6.16e-02 |
Pos logit contributions | |
Kids | +0.157 |
And | +0.157 |
Days | +0.156 |
enny | +0.155 |
David | +0.155 |
Neg logit contributions | |
on | -0.146 |
globe | -0.135 |
inside | -0.132 |
up | -0.132 |
bulb | -0.131 |
Token | are |
Token ID | 389 |
Feature activation | +6.70e-02 |
Pos logit contributions | |
Jack | +0.078 |
His | +0.076 |
He | +0.076 |
Dad | +0.068 |
Kids | +0.067 |
Neg logit contributions | |
els | -0.125 |
bulb | -0.121 |
ool | -0.121 |
hole | -0.120 |
globe | -0.120 |
Token | friends |
Token ID | 2460 |
Feature activation | +1.77e-02 |
Pos logit contributions | |
His | +0.128 |
Jack | +0.127 |
He | +0.125 |
Kids | +0.117 |
David | +0.116 |
Neg logit contributions | |
els | -0.092 |
bulb | -0.089 |
hole | -0.088 |
ps | -0.086 |
globe | -0.086 |
Token | . |
Token ID | 13 |
Feature activation | +1.44e-01 |
Pos logit contributions | |
Jack | +0.040 |
His | +0.039 |
He | +0.039 |
Kids | +0.038 |
Thanks | +0.037 |
Neg logit contributions | |
els | -0.017 |
hole | -0.017 |
bulb | -0.017 |
globe | -0.016 |
et | -0.016 |
Token | They |
Token ID | 1119 |
Feature activation | +6.03e-02 |
Pos logit contributions | |
His | +0.121 |
Jack | +0.119 |
He | +0.112 |
David | +0.099 |
Thanks | +0.097 |
Neg logit contributions | |
els | -0.343 |
globe | -0.339 |
bulb | -0.339 |
hole | -0.338 |
ool | -0.336 |
Token | like |
Token ID | 588 |
Feature activation | +4.06e-02 |
Pos logit contributions | |
His | +0.094 |
Jack | +0.093 |
He | +0.090 |
Kids | +0.085 |
David | +0.084 |
Neg logit contributions | |
els | -0.102 |
bulb | -0.101 |
globe | -0.099 |
hole | -0.099 |
ps | -0.099 |
Token | to |
Token ID | 284 |
Feature activation | +3.58e-02 |
Pos logit contributions | |
His | +0.061 |
Jack | +0.061 |
He | +0.060 |
Kids | +0.055 |
David | +0.055 |
Neg logit contributions | |
els | -0.071 |
bulb | -0.069 |
hole | -0.069 |
globe | -0.068 |
ps | -0.067 |
Token | play |
Token ID | 711 |
Feature activation | -2.63e-02 |
Pos logit contributions | |
Jack | +0.049 |
His | +0.047 |
He | +0.047 |
Kids | +0.044 |
David | +0.044 |
Neg logit contributions | |
bulb | -0.068 |
els | -0.067 |
globe | -0.065 |
trunk | -0.064 |
object | -0.064 |
Token | with |
Token ID | 351 |
Feature activation | -9.16e-03 |
Pos logit contributions | |
hole | +0.033 |
els | +0.033 |
ening | +0.032 |
ool | +0.031 |
bulb | +0.031 |
Neg logit contributions | |
His | -0.054 |
Jack | -0.054 |
He | -0.053 |
They | -0.050 |
Kids | -0.049 |
Token | are |
Token ID | 389 |
Feature activation | +5.42e-02 |
Pos logit contributions | |
His | +0.074 |
Jack | +0.074 |
He | +0.072 |
Kids | +0.068 |
David | +0.068 |
Neg logit contributions | |
els | -0.053 |
bulb | -0.051 |
hole | -0.051 |
globe | -0.050 |
ps | -0.050 |
Token | playing |
Token ID | 2712 |
Feature activation | -2.64e-02 |
Pos logit contributions | |
His | +0.107 |
Jack | +0.105 |
He | +0.104 |
Kids | +0.098 |
David | +0.097 |
Neg logit contributions | |
els | -0.071 |
bulb | -0.068 |
hole | -0.067 |
ps | -0.066 |
atop | -0.065 |
Token | with |
Token ID | 351 |
Feature activation | +1.17e-02 |
Pos logit contributions | |
els | +0.030 |
hole | +0.030 |
ening | +0.028 |
ool | +0.028 |
globe | +0.028 |
Neg logit contributions | |
Jack | -0.057 |
His | -0.057 |
He | -0.056 |
They | -0.052 |
Thanks | -0.052 |
Token | their |
Token ID | 511 |
Feature activation | +1.30e-02 |
Pos logit contributions | |
Jack | +0.018 |
Kids | +0.018 |
Sorry | +0.018 |
He | +0.018 |
Sure | +0.018 |
Neg logit contributions | |
bulb | -0.020 |
matches | -0.019 |
els | -0.019 |
pieces | -0.019 |
ps | -0.018 |
Token | toys |
Token ID | 14958 |
Feature activation | -7.72e-02 |
Pos logit contributions | |
His | +0.031 |
Jack | +0.031 |
He | +0.030 |
Kids | +0.029 |
David | +0.029 |
Neg logit contributions | |
els | -0.012 |
bulb | -0.012 |
globe | -0.011 |
hole | -0.011 |
ps | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | +1.44e-01 |
Pos logit contributions | |
ool | +0.120 |
ening | +0.116 |
els | +0.115 |
globe | +0.115 |
hole | +0.112 |
Neg logit contributions | |
His | -0.133 |
He | -0.132 |
They | -0.130 |
Jack | -0.129 |
She | -0.124 |
Token | They |
Token ID | 1119 |
Feature activation | +5.58e-02 |
Pos logit contributions | |
David | +0.183 |
Sorry | +0.179 |
Jack | +0.178 |
His | +0.176 |
Johnny | +0.169 |
Neg logit contributions | |
ool | -0.274 |
els | -0.268 |
bulb | -0.266 |
matches | -0.265 |
trunk | -0.264 |
Token | have |
Token ID | 423 |
Feature activation | +4.35e-02 |
Pos logit contributions | |
His | +0.086 |
Jack | +0.084 |
He | +0.082 |
David | +0.078 |
Kids | +0.077 |
Neg logit contributions | |
bulb | -0.095 |
els | -0.094 |
globe | -0.092 |
et | -0.091 |
hole | -0.091 |
Token | many |
Token ID | 867 |
Feature activation | -4.30e-02 |
Pos logit contributions | |
His | +0.075 |
He | +0.072 |
Jack | +0.072 |
Kids | +0.068 |
Brother | +0.067 |
Neg logit contributions | |
pieces | -0.066 |
els | -0.064 |
bulb | -0.064 |
ps | -0.063 |
hole | -0.063 |
Token | toys |
Token ID | 14958 |
Feature activation | -7.78e-02 |
Pos logit contributions | |
hole | +0.059 |
globe | +0.057 |
ps | +0.055 |
ening | +0.053 |
atop | +0.053 |
Neg logit contributions | |
His | -0.085 |
Jack | -0.081 |
He | -0.080 |
His | -0.078 |
Her | -0.077 |
Token | of |
Token ID | 286 |
Feature activation | +6.47e-03 |
Pos logit contributions | |
ening | +0.128 |
els | +0.125 |
hole | +0.123 |
ps | +0.121 |
globe | +0.121 |
Neg logit contributions | |
His | -0.125 |
He | -0.124 |
Jack | -0.122 |
They | -0.121 |
Her | -0.113 |
Token | ily |
Token ID | 813 |
Feature activation | +6.36e-02 |
Pos logit contributions | |
His | +0.092 |
Jack | +0.091 |
He | +0.089 |
Kids | +0.081 |
David | +0.080 |
Neg logit contributions | |
els | -0.125 |
bulb | -0.122 |
hole | -0.122 |
globe | -0.121 |
ps | -0.119 |
Token | and |
Token ID | 290 |
Feature activation | +9.79e-02 |
Pos logit contributions | |
Jack | +0.101 |
He | +0.099 |
His | +0.098 |
Kids | +0.090 |
Dad | +0.090 |
Neg logit contributions | |
els | -0.108 |
globe | -0.106 |
bulb | -0.104 |
ool | -0.104 |
hole | -0.104 |
Token | Tom |
Token ID | 4186 |
Feature activation | +3.88e-02 |
Pos logit contributions | |
His | +0.160 |
He | +0.155 |
Kids | +0.147 |
Jack | +0.146 |
So | +0.145 |
Neg logit contributions | |
bulb | -0.152 |
on | -0.150 |
globe | -0.146 |
up | -0.144 |
flying | -0.144 |
Token | are |
Token ID | 389 |
Feature activation | +6.20e-02 |
Pos logit contributions | |
His | +0.056 |
Jack | +0.055 |
He | +0.053 |
Kids | +0.050 |
David | +0.049 |
Neg logit contributions | |
els | -0.071 |
bulb | -0.069 |
hole | -0.068 |
globe | -0.068 |
ps | -0.067 |
Token | friends |
Token ID | 2460 |
Feature activation | +1.79e-02 |
Pos logit contributions | |
His | +0.120 |
Jack | +0.118 |
He | +0.117 |
Kids | +0.110 |
Thanks | +0.109 |
Neg logit contributions | |
els | -0.084 |
bulb | -0.081 |
hole | -0.080 |
ps | -0.078 |
globe | -0.078 |
Token | . |
Token ID | 13 |
Feature activation | +1.44e-01 |
Pos logit contributions | |
Jack | +0.041 |
His | +0.040 |
He | +0.039 |
Kids | +0.039 |
David | +0.038 |
Neg logit contributions | |
els | -0.017 |
bulb | -0.017 |
hole | -0.017 |
globe | -0.016 |
ening | -0.016 |
Token | They |
Token ID | 1119 |
Feature activation | +6.05e-02 |
Pos logit contributions | |
His | +0.126 |
Jack | +0.125 |
He | +0.116 |
David | +0.111 |
Thanks | +0.107 |
Neg logit contributions | |
globe | -0.331 |
els | -0.330 |
bulb | -0.329 |
ool | -0.325 |
hole | -0.324 |
Token | like |
Token ID | 588 |
Feature activation | +4.19e-02 |
Pos logit contributions | |
His | +0.092 |
Jack | +0.092 |
He | +0.089 |
Kids | +0.083 |
David | +0.083 |
Neg logit contributions | |
els | -0.105 |
bulb | -0.103 |
hole | -0.101 |
globe | -0.101 |
ps | -0.100 |
Token | to |
Token ID | 284 |
Feature activation | +3.79e-02 |
Pos logit contributions | |
His | +0.062 |
Jack | +0.062 |
He | +0.061 |
Kids | +0.056 |
David | +0.055 |
Neg logit contributions | |
els | -0.074 |
bulb | -0.072 |
hole | -0.072 |
globe | -0.071 |
ps | -0.071 |
Token | play |
Token ID | 711 |
Feature activation | -2.53e-02 |
Pos logit contributions | |
Jack | +0.050 |
His | +0.048 |
He | +0.048 |
Kids | +0.046 |
David | +0.045 |
Neg logit contributions | |
bulb | -0.073 |
els | -0.072 |
globe | -0.070 |
trunk | -0.069 |
ps | -0.068 |
Token | in |
Token ID | 287 |
Feature activation | +5.17e-03 |
Pos logit contributions | |
hole | +0.033 |
els | +0.033 |
ening | +0.032 |
bulb | +0.031 |
ool | +0.031 |
Neg logit contributions | |
His | -0.051 |
Jack | -0.051 |
He | -0.049 |
They | -0.047 |
Kids | -0.046 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +3.95e-02 |
Pos logit contributions | |
Jack | +0.140 |
His | +0.135 |
David | +0.124 |
He | +0.121 |
Brother | +0.116 |
Neg logit contributions | |
globe | -0.246 |
els | -0.239 |
hole | -0.239 |
bulb | -0.238 |
trunk | -0.236 |
Token | L |
Token ID | 43 |
Feature activation | +5.02e-02 |
Pos logit contributions | |
Jack | +0.048 |
He | +0.047 |
His | +0.046 |
Kids | +0.041 |
They | +0.040 |
Neg logit contributions | |
els | -0.084 |
ening | -0.083 |
bulb | -0.083 |
hole | -0.083 |
ps | -0.081 |
Token | ily |
Token ID | 813 |
Feature activation | +5.10e-02 |
Pos logit contributions | |
His | +0.081 |
Jack | +0.080 |
He | +0.078 |
Thanks | +0.074 |
David | +0.072 |
Neg logit contributions | |
els | -0.086 |
hole | -0.081 |
bulb | -0.081 |
ps | -0.080 |
ool | -0.080 |
Token | loved |
Token ID | 6151 |
Feature activation | +2.47e-02 |
Pos logit contributions | |
Jack | +0.079 |
He | +0.077 |
His | +0.077 |
Kids | +0.072 |
They | +0.070 |
Neg logit contributions | |
els | -0.089 |
hole | -0.087 |
bulb | -0.086 |
ool | -0.086 |
globe | -0.085 |
Token | dinosaurs |
Token ID | 27199 |
Feature activation | -1.05e-01 |
Pos logit contributions | |
His | +0.041 |
Jack | +0.041 |
He | +0.040 |
Kids | +0.037 |
David | +0.037 |
Neg logit contributions | |
els | -0.040 |
bulb | -0.038 |
hole | -0.038 |
globe | -0.037 |
ps | -0.037 |
Token | . |
Token ID | 13 |
Feature activation | +1.44e-01 |
Pos logit contributions | |
els | +0.127 |
hole | +0.126 |
ening | +0.124 |
ps | +0.120 |
bulb | +0.120 |
Neg logit contributions | |
His | -0.221 |
He | -0.218 |
Jack | -0.212 |
They | -0.201 |
Her | -0.198 |
Token | She |
Token ID | 1375 |
Feature activation | +6.89e-02 |
Pos logit contributions | |
Jack | +0.135 |
David | +0.129 |
His | +0.124 |
Sorry | +0.123 |
Thanks | +0.120 |
Neg logit contributions | |
bulb | -0.325 |
ool | -0.320 |
els | -0.316 |
globe | -0.313 |
bulbs | -0.309 |
Token | had |
Token ID | 550 |
Feature activation | +4.89e-02 |
Pos logit contributions | |
His | +0.098 |
Jack | +0.097 |
He | +0.095 |
Kids | +0.086 |
Thanks | +0.085 |
Neg logit contributions | |
els | -0.130 |
hole | -0.126 |
bulb | -0.125 |
ening | -0.123 |
ps | -0.123 |
Token | many |
Token ID | 867 |
Feature activation | -4.45e-02 |
Pos logit contributions | |
Jack | +0.093 |
His | +0.091 |
Kids | +0.089 |
Sure | +0.088 |
Sorry | +0.088 |
Neg logit contributions | |
pieces | -0.063 |
els | -0.062 |
matches | -0.062 |
bulb | -0.060 |
globe | -0.060 |
Token | dinosaur |
Token ID | 28257 |
Feature activation | -1.22e-01 |
Pos logit contributions | |
hole | +0.055 |
bulb | +0.055 |
atop | +0.052 |
ps | +0.052 |
ool | +0.052 |
Neg logit contributions | |
His | -0.093 |
He | -0.091 |
Jack | -0.089 |
His | -0.085 |
They | -0.082 |
Token | toys |
Token ID | 14958 |
Feature activation | -8.82e-02 |
Pos logit contributions | |
hole | +0.171 |
els | +0.170 |
bulb | +0.167 |
ening | +0.167 |
ps | +0.164 |
Neg logit contributions | |
His | -0.229 |
He | -0.225 |
Jack | -0.224 |
They | -0.205 |
His | -0.201 |
Token | Anna |
Token ID | 31160 |
Feature activation | +5.54e-02 |
Pos logit contributions | |
Jack | +0.038 |
His | +0.036 |
He | +0.036 |
Kids | +0.028 |
His | +0.027 |
Neg logit contributions | |
els | -0.128 |
bulb | -0.126 |
ening | -0.125 |
hole | -0.125 |
ps | -0.124 |
Token | and |
Token ID | 290 |
Feature activation | +9.59e-02 |
Pos logit contributions | |
Jack | +0.086 |
He | +0.084 |
His | +0.083 |
Kids | +0.077 |
Dad | +0.077 |
Neg logit contributions | |
els | -0.097 |
globe | -0.094 |
ool | -0.094 |
bulb | -0.093 |
hole | -0.092 |
Token | Ben |
Token ID | 3932 |
Feature activation | +4.73e-02 |
Pos logit contributions | |
His | +0.173 |
He | +0.167 |
Jack | +0.164 |
Sure | +0.160 |
Kids | +0.160 |
Neg logit contributions | |
on | -0.129 |
up | -0.124 |
bulb | -0.123 |
inside | -0.123 |
flying | -0.120 |
Token | are |
Token ID | 389 |
Feature activation | +6.47e-02 |
Pos logit contributions | |
His | +0.073 |
Jack | +0.072 |
He | +0.070 |
Kids | +0.066 |
David | +0.065 |
Neg logit contributions | |
els | -0.081 |
bulb | -0.079 |
hole | -0.079 |
globe | -0.078 |
ps | -0.077 |
Token | twins |
Token ID | 20345 |
Feature activation | +1.82e-02 |
Pos logit contributions | |
His | +0.125 |
Jack | +0.124 |
He | +0.122 |
Kids | +0.114 |
Thanks | +0.113 |
Neg logit contributions | |
els | -0.088 |
bulb | -0.085 |
hole | -0.084 |
ps | -0.082 |
globe | -0.082 |
Token | . |
Token ID | 13 |
Feature activation | +1.43e-01 |
Pos logit contributions | |
His | +0.036 |
Jack | +0.036 |
He | +0.036 |
They | +0.034 |
Dad | +0.034 |
Neg logit contributions | |
els | -0.023 |
ool | -0.023 |
hole | -0.022 |
globe | -0.022 |
bulb | -0.021 |
Token | They |
Token ID | 1119 |
Feature activation | +6.30e-02 |
Pos logit contributions | |
Jack | +0.177 |
David | +0.177 |
His | +0.165 |
Johnny | +0.158 |
He | +0.158 |
Neg logit contributions | |
bulb | -0.280 |
els | -0.275 |
globe | -0.275 |
chim | -0.268 |
trunk | -0.265 |
Token | like |
Token ID | 588 |
Feature activation | +4.35e-02 |
Pos logit contributions | |
Jack | +0.104 |
His | +0.104 |
He | +0.099 |
David | +0.095 |
Kids | +0.095 |
Neg logit contributions | |
bulb | -0.099 |
ps | -0.098 |
els | -0.098 |
globe | -0.096 |
hole | -0.095 |
Token | to |
Token ID | 284 |
Feature activation | +3.67e-02 |
Pos logit contributions | |
Jack | +0.073 |
His | +0.072 |
He | +0.071 |
David | +0.066 |
Thanks | +0.066 |
Neg logit contributions | |
bulb | -0.067 |
els | -0.066 |
globe | -0.065 |
pieces | -0.064 |
matches | -0.064 |
Token | play |
Token ID | 711 |
Feature activation | -2.73e-02 |
Pos logit contributions | |
Jack | +0.054 |
He | +0.052 |
His | +0.051 |
David | +0.050 |
Kids | +0.050 |
Neg logit contributions | |
bulb | -0.066 |
els | -0.062 |
globe | -0.061 |
trunk | -0.061 |
matches | -0.060 |
Token | outside |
Token ID | 2354 |
Feature activation | -4.37e-02 |
Pos logit contributions | |
els | +0.033 |
hole | +0.032 |
bulb | +0.032 |
globe | +0.031 |
ening | +0.031 |
Neg logit contributions | |
His | -0.056 |
Jack | -0.056 |
He | -0.055 |
Kids | -0.052 |
Thanks | -0.051 |
Token | Anna |
Token ID | 31160 |
Feature activation | +3.96e-02 |
Pos logit contributions | |
Jack | +0.040 |
He | +0.039 |
His | +0.039 |
Kids | +0.035 |
They | +0.034 |
Neg logit contributions | |
els | -0.069 |
bulb | -0.068 |
hole | -0.067 |
ening | -0.067 |
ps | -0.066 |
Token | and |
Token ID | 290 |
Feature activation | +8.30e-02 |
Pos logit contributions | |
Jack | +0.060 |
He | +0.059 |
His | +0.059 |
Kids | +0.055 |
They | +0.054 |
Neg logit contributions | |
els | -0.070 |
hole | -0.068 |
bulb | -0.068 |
ool | -0.067 |
globe | -0.067 |
Token | Ben |
Token ID | 3932 |
Feature activation | +3.24e-02 |
Pos logit contributions | |
His | +0.153 |
He | +0.145 |
Thanks | +0.140 |
Jack | +0.140 |
Sorry | +0.137 |
Neg logit contributions | |
bulb | -0.110 |
inside | -0.109 |
on | -0.106 |
up | -0.106 |
hanging | -0.104 |
Token | are |
Token ID | 389 |
Feature activation | +5.62e-02 |
Pos logit contributions | |
His | +0.052 |
Jack | +0.051 |
He | +0.049 |
Kids | +0.046 |
David | +0.046 |
Neg logit contributions | |
els | -0.053 |
bulb | -0.052 |
globe | -0.052 |
ps | -0.052 |
hole | -0.051 |
Token | friends |
Token ID | 2460 |
Feature activation | +1.10e-02 |
Pos logit contributions | |
His | +0.112 |
Jack | +0.110 |
He | +0.108 |
Kids | +0.102 |
Thanks | +0.101 |
Neg logit contributions | |
els | -0.075 |
bulb | -0.071 |
hole | -0.070 |
ps | -0.070 |
atop | -0.068 |
Token | . |
Token ID | 13 |
Feature activation | +1.43e-01 |
Pos logit contributions | |
Jack | +0.027 |
His | +0.027 |
He | +0.026 |
Kids | +0.026 |
David | +0.026 |
Neg logit contributions | |
els | -0.008 |
hole | -0.008 |
bulb | -0.008 |
globe | -0.007 |
ening | -0.007 |
Token | They |
Token ID | 1119 |
Feature activation | +5.76e-02 |
Pos logit contributions | |
David | +0.215 |
Johnny | +0.205 |
Jack | +0.195 |
Thanks | +0.186 |
Sorry | +0.180 |
Neg logit contributions | |
globe | -0.231 |
trunk | -0.225 |
bulb | -0.223 |
blue | -0.219 |
matches | -0.218 |
Token | like |
Token ID | 588 |
Feature activation | +4.00e-02 |
Pos logit contributions | |
His | +0.092 |
Jack | +0.091 |
He | +0.088 |
Kids | +0.084 |
David | +0.084 |
Neg logit contributions | |
els | -0.094 |
bulb | -0.094 |
ps | -0.092 |
globe | -0.091 |
hole | -0.090 |
Token | to |
Token ID | 284 |
Feature activation | +3.57e-02 |
Pos logit contributions | |
His | +0.061 |
Jack | +0.061 |
He | +0.059 |
Kids | +0.055 |
David | +0.054 |
Neg logit contributions | |
els | -0.070 |
bulb | -0.068 |
hole | -0.067 |
globe | -0.067 |
ps | -0.066 |
Token | play |
Token ID | 711 |
Feature activation | -2.80e-02 |
Pos logit contributions | |
Jack | +0.045 |
His | +0.045 |
He | +0.044 |
Kids | +0.041 |
David | +0.040 |
Neg logit contributions | |
els | -0.070 |
bulb | -0.070 |
globe | -0.068 |
ps | -0.067 |
et | -0.067 |
Token | in |
Token ID | 287 |
Feature activation | +1.84e-03 |
Pos logit contributions | |
hole | +0.035 |
els | +0.035 |
ening | +0.034 |
bulb | +0.033 |
ool | +0.033 |
Neg logit contributions | |
His | -0.057 |
Jack | -0.057 |
He | -0.056 |
They | -0.052 |
Thanks | -0.052 |
Token | Tom |
Token ID | 13787 |
Feature activation | +3.17e-02 |
Pos logit contributions | |
Jack | +0.032 |
He | +0.031 |
His | +0.031 |
Kids | +0.026 |
Dad | +0.024 |
Neg logit contributions | |
els | -0.096 |
ening | -0.094 |
bulb | -0.094 |
hole | -0.093 |
ps | -0.092 |
Token | and |
Token ID | 290 |
Feature activation | +8.76e-02 |
Pos logit contributions | |
Jack | +0.046 |
His | +0.046 |
He | +0.046 |
Kids | +0.041 |
Dad | +0.041 |
Neg logit contributions | |
els | -0.057 |
bulb | -0.056 |
ool | -0.056 |
hole | -0.056 |
globe | -0.055 |
Token | Lily |
Token ID | 20037 |
Feature activation | +5.23e-02 |
Pos logit contributions | |
Days | +0.166 |
Jack | +0.163 |
And | +0.163 |
Years | +0.161 |
David | +0.161 |
Neg logit contributions | |
on | -0.112 |
inside | -0.101 |
globe | -0.099 |
bulb | -0.098 |
trunk | -0.097 |
Token | are |
Token ID | 389 |
Feature activation | +5.81e-02 |
Pos logit contributions | |
Jack | +0.069 |
His | +0.069 |
He | +0.067 |
Kids | +0.061 |
David | +0.061 |
Neg logit contributions | |
els | -0.102 |
hole | -0.099 |
bulb | -0.099 |
globe | -0.098 |
ool | -0.097 |
Token | twins |
Token ID | 20345 |
Feature activation | +1.25e-02 |
Pos logit contributions | |
His | +0.112 |
Jack | +0.111 |
He | +0.109 |
Kids | +0.103 |
Thanks | +0.101 |
Neg logit contributions | |
els | -0.080 |
bulb | -0.077 |
hole | -0.076 |
ps | -0.075 |
globe | -0.073 |
Token | . |
Token ID | 13 |
Feature activation | +1.43e-01 |
Pos logit contributions | |
His | +0.025 |
Jack | +0.025 |
He | +0.025 |
They | +0.023 |
Dad | +0.023 |
Neg logit contributions | |
els | -0.016 |
ool | -0.016 |
hole | -0.015 |
ening | -0.015 |
globe | -0.015 |
Token | They |
Token ID | 1119 |
Feature activation | +5.70e-02 |
Pos logit contributions | |
Jack | +0.156 |
His | +0.150 |
He | +0.143 |
David | +0.137 |
Johnny | +0.127 |
Neg logit contributions | |
els | -0.302 |
hole | -0.293 |
globe | -0.293 |
bulb | -0.293 |
trunk | -0.291 |
Token | like |
Token ID | 588 |
Feature activation | +4.08e-02 |
Pos logit contributions | |
His | +0.096 |
Jack | +0.095 |
He | +0.090 |
David | +0.088 |
Kids | +0.087 |
Neg logit contributions | |
bulb | -0.089 |
els | -0.087 |
ps | -0.087 |
globe | -0.086 |
et | -0.085 |
Token | to |
Token ID | 284 |
Feature activation | +3.61e-02 |
Pos logit contributions | |
Jack | +0.071 |
His | +0.070 |
He | +0.069 |
Thanks | +0.065 |
David | +0.065 |
Neg logit contributions | |
bulb | -0.061 |
els | -0.059 |
globe | -0.058 |
pieces | -0.058 |
matches | -0.058 |
Token | play |
Token ID | 711 |
Feature activation | -2.81e-02 |
Pos logit contributions | |
Jack | +0.053 |
He | +0.051 |
His | +0.050 |
Kids | +0.049 |
David | +0.048 |
Neg logit contributions | |
bulb | -0.065 |
els | -0.063 |
trunk | -0.061 |
globe | -0.061 |
object | -0.060 |
Token | games |
Token ID | 1830 |
Feature activation | -8.16e-02 |
Pos logit contributions | |
els | +0.034 |
hole | +0.033 |
bulb | +0.032 |
ening | +0.032 |
globe | +0.032 |
Neg logit contributions | |
His | -0.059 |
Jack | -0.058 |
He | -0.057 |
Kids | -0.054 |
Thanks | -0.053 |
Token | ila |
Token ID | 10102 |
Feature activation | +4.34e-02 |
Pos logit contributions | |
His | +0.086 |
Jack | +0.084 |
He | +0.082 |
Thanks | +0.080 |
Un | +0.077 |
Neg logit contributions | |
els | -0.088 |
hole | -0.083 |
ool | -0.083 |
bulb | -0.082 |
ps | -0.082 |
Token | and |
Token ID | 290 |
Feature activation | +8.90e-02 |
Pos logit contributions | |
Jack | +0.072 |
His | +0.070 |
He | +0.070 |
Kids | +0.064 |
They | +0.064 |
Neg logit contributions | |
els | -0.071 |
hole | -0.069 |
globe | -0.069 |
bulb | -0.069 |
ool | -0.068 |
Token | Tom |
Token ID | 4186 |
Feature activation | +2.85e-02 |
Pos logit contributions | |
His | +0.142 |
He | +0.138 |
Kids | +0.137 |
Too | +0.134 |
Together | +0.133 |
Neg logit contributions | |
inside | -0.135 |
bulb | -0.134 |
up | -0.133 |
ring | -0.133 |
on | -0.133 |
Token | are |
Token ID | 389 |
Feature activation | +5.76e-02 |
Pos logit contributions | |
His | +0.046 |
Jack | +0.046 |
He | +0.044 |
Kids | +0.043 |
David | +0.042 |
Neg logit contributions | |
els | -0.045 |
ps | -0.044 |
bulb | -0.044 |
et | -0.044 |
hole | -0.043 |
Token | friends |
Token ID | 2460 |
Feature activation | +1.13e-02 |
Pos logit contributions | |
His | +0.114 |
Jack | +0.112 |
He | +0.110 |
Kids | +0.105 |
Thanks | +0.103 |
Neg logit contributions | |
els | -0.077 |
bulb | -0.073 |
hole | -0.072 |
ps | -0.071 |
atop | -0.069 |
Token | . |
Token ID | 13 |
Feature activation | +1.43e-01 |
Pos logit contributions | |
Jack | +0.029 |
His | +0.029 |
Kids | +0.028 |
He | +0.028 |
David | +0.028 |
Neg logit contributions | |
els | -0.007 |
hole | -0.007 |
bulb | -0.007 |
globe | -0.006 |
ening | -0.006 |
Token | They |
Token ID | 1119 |
Feature activation | +5.83e-02 |
Pos logit contributions | |
Jack | +0.177 |
David | +0.156 |
His | +0.154 |
Thanks | +0.151 |
Brother | +0.145 |
Neg logit contributions | |
globe | -0.274 |
els | -0.271 |
trunk | -0.270 |
hole | -0.269 |
ool | -0.266 |
Token | like |
Token ID | 588 |
Feature activation | +3.77e-02 |
Pos logit contributions | |
His | +0.095 |
Jack | +0.095 |
He | +0.090 |
Kids | +0.087 |
David | +0.086 |
Neg logit contributions | |
bulb | -0.094 |
els | -0.093 |
globe | -0.091 |
ps | -0.091 |
et | -0.089 |
Token | to |
Token ID | 284 |
Feature activation | +3.69e-02 |
Pos logit contributions | |
His | +0.051 |
Jack | +0.051 |
He | +0.050 |
Kids | +0.046 |
Dad | +0.045 |
Neg logit contributions | |
els | -0.073 |
hole | -0.072 |
ening | -0.070 |
ool | -0.069 |
bulb | -0.069 |
Token | play |
Token ID | 711 |
Feature activation | -2.77e-02 |
Pos logit contributions | |
Jack | +0.048 |
His | +0.047 |
He | +0.046 |
Kids | +0.043 |
David | +0.042 |
Neg logit contributions | |
els | -0.073 |
bulb | -0.072 |
globe | -0.070 |
ps | -0.069 |
ool | -0.069 |
Token | with |
Token ID | 351 |
Feature activation | -9.96e-03 |
Pos logit contributions | |
hole | +0.037 |
els | +0.037 |
ening | +0.036 |
ool | +0.035 |
bulb | +0.035 |
Neg logit contributions | |
His | -0.055 |
Jack | -0.055 |
He | -0.053 |
They | -0.050 |
Dad | -0.049 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +4.62e-02 |
Pos logit contributions | |
His | +0.144 |
Jack | +0.142 |
Thanks | +0.130 |
He | +0.129 |
David | +0.128 |
Neg logit contributions | |
hole | -0.260 |
els | -0.260 |
bulb | -0.259 |
trunk | -0.258 |
chim | -0.255 |
Token | L |
Token ID | 43 |
Feature activation | +5.84e-02 |
Pos logit contributions | |
Jack | +0.035 |
He | +0.033 |
His | +0.033 |
Kids | +0.026 |
Dad | +0.025 |
Neg logit contributions | |
els | -0.120 |
bulb | -0.117 |
ening | -0.117 |
ps | -0.117 |
hole | -0.116 |
Token | ily |
Token ID | 813 |
Feature activation | +5.93e-02 |
Pos logit contributions | |
His | +0.089 |
Jack | +0.088 |
He | +0.086 |
Kids | +0.080 |
David | +0.079 |
Neg logit contributions | |
els | -0.102 |
hole | -0.099 |
bulb | -0.099 |
globe | -0.098 |
ps | -0.097 |
Token | was |
Token ID | 373 |
Feature activation | +7.92e-02 |
Pos logit contributions | |
Jack | +0.089 |
He | +0.087 |
His | +0.086 |
Dad | +0.079 |
They | +0.079 |
Neg logit contributions | |
els | -0.106 |
globe | -0.104 |
ool | -0.102 |
bulb | -0.102 |
ps | -0.102 |
Token | hungry |
Token ID | 14720 |
Feature activation | -5.22e-02 |
Pos logit contributions | |
Jack | +0.145 |
His | +0.144 |
He | +0.141 |
Thanks | +0.131 |
Kids | +0.131 |
Neg logit contributions | |
els | -0.114 |
globe | -0.111 |
hole | -0.110 |
bulb | -0.109 |
ool | -0.107 |
Token | . |
Token ID | 13 |
Feature activation | +1.43e-01 |
Pos logit contributions | |
els | +0.099 |
cks | +0.095 |
ps | +0.094 |
angles | +0.093 |
globe | +0.092 |
Neg logit contributions | |
His | -0.075 |
He | -0.075 |
Jack | -0.070 |
He | -0.069 |
His | -0.069 |
Token | She |
Token ID | 1375 |
Feature activation | +6.99e-02 |
Pos logit contributions | |
Jack | +0.128 |
His | +0.127 |
David | +0.118 |
Thanks | +0.118 |
Sorry | +0.117 |
Neg logit contributions | |
bulb | -0.332 |
ool | -0.328 |
els | -0.323 |
trunk | -0.321 |
hole | -0.319 |
Token | wanted |
Token ID | 2227 |
Feature activation | +1.99e-02 |
Pos logit contributions | |
Jack | +0.105 |
His | +0.105 |
He | +0.100 |
David | +0.095 |
Kids | +0.095 |
Neg logit contributions | |
els | -0.122 |
bulb | -0.120 |
hole | -0.118 |
globe | -0.117 |
ool | -0.117 |
Token | to |
Token ID | 284 |
Feature activation | +3.37e-02 |
Pos logit contributions | |
His | +0.032 |
Jack | +0.032 |
He | +0.031 |
Kids | +0.029 |
David | +0.029 |
Neg logit contributions | |
els | -0.033 |
bulb | -0.032 |
hole | -0.031 |
globe | -0.031 |
ps | -0.031 |
Token | eat |
Token ID | 4483 |
Feature activation | -7.70e-02 |
Pos logit contributions | |
Jack | +0.060 |
His | +0.058 |
He | +0.058 |
Kids | +0.058 |
David | +0.056 |
Neg logit contributions | |
bulb | -0.050 |
els | -0.048 |
egg | -0.048 |
trunk | -0.047 |
matches | -0.047 |
Token | cereal |
Token ID | 33158 |
Feature activation | -1.40e-01 |
Pos logit contributions | |
hole | +0.158 |
globe | +0.148 |
ening | +0.144 |
ool | +0.142 |
bulb | +0.141 |
Neg logit contributions | |
His | -0.105 |
Jack | -0.103 |
He | -0.101 |
They | -0.095 |
Her | -0.093 |
Token | ara |
Token ID | 3301 |
Feature activation | +4.66e-02 |
Pos logit contributions | |
His | +0.047 |
Jack | +0.046 |
He | +0.045 |
Kids | +0.042 |
Thanks | +0.042 |
Neg logit contributions | |
els | -0.050 |
hole | -0.047 |
globe | -0.047 |
bulb | -0.046 |
et | -0.046 |
Token | was |
Token ID | 373 |
Feature activation | +7.75e-02 |
Pos logit contributions | |
Jack | +0.072 |
He | +0.072 |
His | +0.072 |
Kids | +0.065 |
They | +0.064 |
Neg logit contributions | |
els | -0.080 |
ool | -0.080 |
hole | -0.080 |
bulb | -0.079 |
ening | -0.075 |
Token | very |
Token ID | 845 |
Feature activation | +5.74e-02 |
Pos logit contributions | |
His | +0.136 |
Jack | +0.136 |
He | +0.133 |
Kids | +0.124 |
David | +0.123 |
Neg logit contributions | |
els | -0.117 |
hole | -0.113 |
bulb | -0.113 |
globe | -0.112 |
ool | -0.110 |
Token | happy |
Token ID | 3772 |
Feature activation | -1.95e-02 |
Pos logit contributions | |
His | +0.094 |
Jack | +0.093 |
He | +0.091 |
Kids | +0.084 |
Thanks | +0.083 |
Neg logit contributions | |
els | -0.095 |
hole | -0.093 |
bulb | -0.091 |
globe | -0.091 |
ool | -0.089 |
Token | today |
Token ID | 1909 |
Feature activation | -3.54e-02 |
Pos logit contributions | |
els | +0.021 |
hole | +0.020 |
bulb | +0.020 |
ool | +0.019 |
ps | +0.019 |
Neg logit contributions | |
His | -0.043 |
Jack | -0.043 |
He | -0.042 |
Kids | -0.040 |
David | -0.039 |
Token | . |
Token ID | 13 |
Feature activation | +1.43e-01 |
Pos logit contributions | |
els | +0.048 |
hole | +0.046 |
bulb | +0.045 |
ps | +0.045 |
ool | +0.044 |
Neg logit contributions | |
Jack | -0.069 |
His | -0.068 |
He | -0.068 |
Kids | -0.064 |
Dad | -0.064 |
Token | She |
Token ID | 1375 |
Feature activation | +7.15e-02 |
Pos logit contributions | |
Jack | +0.127 |
His | +0.118 |
David | +0.111 |
Thanks | +0.106 |
He | +0.104 |
Neg logit contributions | |
globe | -0.326 |
bulb | -0.326 |
chim | -0.323 |
trunk | -0.322 |
els | -0.321 |
Token | got |
Token ID | 1392 |
Feature activation | +6.85e-04 |
Pos logit contributions | |
His | +0.141 |
Jack | +0.139 |
He | +0.135 |
David | +0.129 |
Kids | +0.128 |
Neg logit contributions | |
els | -0.092 |
bulb | -0.090 |
globe | -0.088 |
hole | -0.088 |
ps | -0.087 |
Token | a |
Token ID | 257 |
Feature activation | +9.81e-03 |
Pos logit contributions | |
His | +0.001 |
Jack | +0.001 |
He | +0.001 |
Kids | +0.001 |
David | +0.001 |
Neg logit contributions | |
els | -0.001 |
bulb | -0.001 |
hole | -0.001 |
globe | -0.001 |
ps | -0.001 |
Token | big |
Token ID | 1263 |
Feature activation | -5.39e-02 |
Pos logit contributions | |
Jack | +0.018 |
His | +0.018 |
He | +0.017 |
Sure | +0.017 |
Dad | +0.016 |
Neg logit contributions | |
els | -0.015 |
ps | -0.014 |
hole | -0.014 |
ening | -0.014 |
cks | -0.013 |
Token | box |
Token ID | 3091 |
Feature activation | -1.23e-01 |
Pos logit contributions | |
els | +0.080 |
cks | +0.079 |
hole | +0.078 |
ps | +0.078 |
atop | +0.071 |
Neg logit contributions | |
Jack | -0.093 |
please | -0.091 |
His | -0.087 |
Sure | -0.086 |
He | -0.085 |
Token | Tom |
Token ID | 13787 |
Feature activation | +4.51e-02 |
Pos logit contributions | |
Jack | +0.059 |
His | +0.057 |
He | +0.056 |
Kids | +0.048 |
Dad | +0.046 |
Neg logit contributions | |
els | -0.143 |
bulb | -0.140 |
hole | -0.139 |
ening | -0.139 |
ps | -0.138 |
Token | and |
Token ID | 290 |
Feature activation | +9.62e-02 |
Pos logit contributions | |
His | +0.069 |
Jack | +0.069 |
He | +0.068 |
Kids | +0.062 |
Dad | +0.061 |
Neg logit contributions | |
els | -0.078 |
bulb | -0.077 |
globe | -0.076 |
hole | -0.075 |
ool | -0.075 |
Token | Lily |
Token ID | 20037 |
Feature activation | +5.91e-02 |
Pos logit contributions | |
Days | +0.158 |
So | +0.156 |
And | +0.155 |
Years | +0.152 |
Unfortunately | +0.152 |
Neg logit contributions | |
on | -0.142 |
globe | -0.134 |
bulb | -0.131 |
inside | -0.128 |
trunk | -0.127 |
Token | are |
Token ID | 389 |
Feature activation | +6.43e-02 |
Pos logit contributions | |
Jack | +0.079 |
His | +0.078 |
He | +0.078 |
Dad | +0.071 |
They | +0.070 |
Neg logit contributions | |
els | -0.114 |
bulb | -0.111 |
ool | -0.111 |
globe | -0.111 |
hole | -0.110 |
Token | twins |
Token ID | 20345 |
Feature activation | +1.77e-02 |
Pos logit contributions | |
His | +0.122 |
Jack | +0.121 |
He | +0.119 |
Kids | +0.112 |
David | +0.111 |
Neg logit contributions | |
els | -0.090 |
bulb | -0.086 |
hole | -0.086 |
ps | -0.084 |
globe | -0.084 |
Token | . |
Token ID | 13 |
Feature activation | +1.43e-01 |
Pos logit contributions | |
His | +0.035 |
Jack | +0.035 |
He | +0.035 |
They | +0.033 |
Dad | +0.033 |
Neg logit contributions | |
els | -0.022 |
ool | -0.022 |
hole | -0.021 |
globe | -0.021 |
bulb | -0.021 |
Token | They |
Token ID | 1119 |
Feature activation | +6.10e-02 |
Pos logit contributions | |
Jack | +0.133 |
His | +0.126 |
He | +0.120 |
David | +0.112 |
Brother | +0.105 |
Neg logit contributions | |
els | -0.329 |
bulb | -0.323 |
hole | -0.321 |
globe | -0.319 |
ool | -0.313 |
Token | like |
Token ID | 588 |
Feature activation | +4.42e-02 |
Pos logit contributions | |
Jack | +0.100 |
His | +0.100 |
He | +0.095 |
David | +0.091 |
Kids | +0.090 |
Neg logit contributions | |
bulb | -0.097 |
els | -0.097 |
ps | -0.096 |
globe | -0.094 |
hole | -0.094 |
Token | to |
Token ID | 284 |
Feature activation | +3.69e-02 |
Pos logit contributions | |
Jack | +0.074 |
His | +0.074 |
He | +0.072 |
David | +0.068 |
Thanks | +0.067 |
Neg logit contributions | |
bulb | -0.068 |
els | -0.067 |
globe | -0.065 |
pieces | -0.064 |
matches | -0.064 |
Token | play |
Token ID | 711 |
Feature activation | -2.44e-02 |
Pos logit contributions | |
Jack | +0.056 |
He | +0.053 |
David | +0.052 |
His | +0.052 |
Kids | +0.051 |
Neg logit contributions | |
bulb | -0.065 |
els | -0.061 |
trunk | -0.060 |
globe | -0.060 |
matches | -0.060 |
Token | together |
Token ID | 1978 |
Feature activation | -3.71e-02 |
Pos logit contributions | |
els | +0.028 |
hole | +0.027 |
bulb | +0.027 |
globe | +0.026 |
ening | +0.026 |
Neg logit contributions | |
His | -0.052 |
Jack | -0.052 |
He | -0.051 |
Kids | -0.048 |
Thanks | -0.047 |
Token | Tom |
Token ID | 13787 |
Feature activation | +4.51e-02 |
Pos logit contributions | |
Jack | +0.059 |
His | +0.057 |
He | +0.056 |
Kids | +0.048 |
Dad | +0.046 |
Neg logit contributions | |
els | -0.143 |
bulb | -0.140 |
hole | -0.139 |
ening | -0.139 |
ps | -0.138 |
Token | and |
Token ID | 290 |
Feature activation | +9.62e-02 |
Pos logit contributions | |
His | +0.069 |
Jack | +0.069 |
He | +0.068 |
Kids | +0.062 |
Dad | +0.061 |
Neg logit contributions | |
els | -0.078 |
bulb | -0.077 |
globe | -0.076 |
hole | -0.075 |
ool | -0.075 |
Token | Lily |
Token ID | 20037 |
Feature activation | +5.91e-02 |
Pos logit contributions | |
Days | +0.158 |
So | +0.156 |
And | +0.155 |
Years | +0.152 |
Unfortunately | +0.152 |
Neg logit contributions | |
on | -0.142 |
globe | -0.134 |
bulb | -0.131 |
inside | -0.128 |
trunk | -0.127 |
Token | are |
Token ID | 389 |
Feature activation | +6.43e-02 |
Pos logit contributions | |
Jack | +0.079 |
His | +0.078 |
He | +0.078 |
Dad | +0.071 |
They | +0.070 |
Neg logit contributions | |
els | -0.114 |
bulb | -0.111 |
ool | -0.111 |
globe | -0.111 |
hole | -0.110 |
Token | friends |
Token ID | 2460 |
Feature activation | +1.63e-02 |
Pos logit contributions | |
His | +0.122 |
Jack | +0.121 |
He | +0.119 |
Kids | +0.112 |
David | +0.111 |
Neg logit contributions | |
els | -0.090 |
bulb | -0.086 |
hole | -0.086 |
ps | -0.084 |
globe | -0.084 |
Token | . |
Token ID | 13 |
Feature activation | +1.43e-01 |
Pos logit contributions | |
Jack | +0.039 |
His | +0.038 |
He | +0.038 |
Kids | +0.037 |
Thanks | +0.036 |
Neg logit contributions | |
els | -0.014 |
hole | -0.013 |
bulb | -0.013 |
globe | -0.013 |
et | -0.012 |
Token | They |
Token ID | 1119 |
Feature activation | +6.05e-02 |
Pos logit contributions | |
His | +0.114 |
Jack | +0.113 |
He | +0.106 |
David | +0.092 |
Thanks | +0.090 |
Neg logit contributions | |
els | -0.347 |
hole | -0.342 |
globe | -0.341 |
bulb | -0.340 |
ool | -0.337 |
Token | go |
Token ID | 467 |
Feature activation | +5.90e-03 |
Pos logit contributions | |
His | +0.094 |
Jack | +0.094 |
He | +0.090 |
Kids | +0.085 |
David | +0.085 |
Neg logit contributions | |
els | -0.102 |
bulb | -0.101 |
globe | -0.099 |
hole | -0.099 |
ps | -0.099 |
Token | to |
Token ID | 284 |
Feature activation | +3.70e-02 |
Pos logit contributions | |
His | +0.009 |
Jack | +0.009 |
He | +0.009 |
Kids | +0.008 |
They | +0.008 |
Neg logit contributions | |
els | -0.010 |
hole | -0.010 |
ening | -0.010 |
bulb | -0.010 |
ool | -0.010 |
Token | school |
Token ID | 1524 |
Feature activation | -5.15e-02 |
Pos logit contributions | |
His | +0.068 |
Jack | +0.067 |
He | +0.067 |
Kids | +0.062 |
Thanks | +0.062 |
Neg logit contributions | |
els | -0.053 |
bulb | -0.052 |
hole | -0.050 |
globe | -0.050 |
ps | -0.049 |
Token | together |
Token ID | 1978 |
Feature activation | -2.91e-02 |
Pos logit contributions | |
els | +0.078 |
bulb | +0.076 |
ool | +0.074 |
egg | +0.074 |
ening | +0.074 |
Neg logit contributions | |
His | -0.092 |
Jack | -0.092 |
He | -0.090 |
They | -0.086 |
Her | -0.084 |
Token | and |
Token ID | 290 |
Feature activation | +9.29e-02 |
Pos logit contributions | |
Jack | +0.058 |
His | +0.058 |
He | +0.057 |
Kids | +0.052 |
Dad | +0.052 |
Neg logit contributions | |
els | -0.064 |
hole | -0.063 |
bulb | -0.063 |
ool | -0.062 |
globe | -0.062 |
Token | Lily |
Token ID | 20037 |
Feature activation | +5.56e-02 |
Pos logit contributions | |
Days | +0.181 |
Years | +0.176 |
Unfortunately | +0.175 |
David | +0.174 |
enny | +0.174 |
Neg logit contributions | |
on | -0.122 |
inside | -0.106 |
up | -0.102 |
globe | -0.101 |
down | -0.100 |
Token | are |
Token ID | 389 |
Feature activation | +6.18e-02 |
Pos logit contributions | |
Jack | +0.075 |
His | +0.075 |
He | +0.073 |
Kids | +0.066 |
David | +0.065 |
Neg logit contributions | |
els | -0.107 |
hole | -0.104 |
bulb | -0.104 |
globe | -0.103 |
ool | -0.102 |
Token | best |
Token ID | 1266 |
Feature activation | +1.19e-02 |
Pos logit contributions | |
His | +0.122 |
Jack | +0.120 |
He | +0.118 |
Kids | +0.112 |
Thanks | +0.110 |
Neg logit contributions | |
els | -0.082 |
bulb | -0.079 |
hole | -0.078 |
ps | -0.077 |
atop | -0.075 |
Token | friends |
Token ID | 2460 |
Feature activation | +1.10e-02 |
Pos logit contributions | |
His | +0.022 |
Jack | +0.022 |
He | +0.021 |
Kids | +0.020 |
David | +0.020 |
Neg logit contributions | |
els | -0.017 |
bulb | -0.017 |
hole | -0.017 |
globe | -0.016 |
ps | -0.016 |
Token | . |
Token ID | 13 |
Feature activation | +1.43e-01 |
Pos logit contributions | |
Jack | +0.027 |
His | +0.027 |
He | +0.027 |
Kids | +0.026 |
David | +0.026 |
Neg logit contributions | |
els | -0.008 |
hole | -0.008 |
bulb | -0.008 |
globe | -0.008 |
ps | -0.007 |
Token | They |
Token ID | 1119 |
Feature activation | +5.58e-02 |
Pos logit contributions | |
Jack | +0.149 |
His | +0.148 |
He | +0.137 |
David | +0.129 |
Brother | +0.127 |
Neg logit contributions | |
globe | -0.303 |
ool | -0.300 |
els | -0.300 |
hole | -0.297 |
trunk | -0.296 |
Token | like |
Token ID | 588 |
Feature activation | +3.97e-02 |
Pos logit contributions | |
His | +0.086 |
Jack | +0.086 |
He | +0.083 |
Kids | +0.078 |
David | +0.078 |
Neg logit contributions | |
els | -0.095 |
bulb | -0.094 |
globe | -0.092 |
ps | -0.091 |
hole | -0.091 |
Token | to |
Token ID | 284 |
Feature activation | +3.58e-02 |
Pos logit contributions | |
His | +0.060 |
Jack | +0.060 |
He | +0.058 |
Kids | +0.054 |
David | +0.053 |
Neg logit contributions | |
els | -0.069 |
bulb | -0.068 |
hole | -0.067 |
globe | -0.067 |
ps | -0.066 |
Token | play |
Token ID | 711 |
Feature activation | -2.71e-02 |
Pos logit contributions | |
Jack | +0.045 |
His | +0.044 |
He | +0.043 |
Kids | +0.041 |
David | +0.040 |
Neg logit contributions | |
bulb | -0.071 |
els | -0.070 |
globe | -0.069 |
ps | -0.067 |
et | -0.067 |
Token | together |
Token ID | 1978 |
Feature activation | -3.92e-02 |
Pos logit contributions | |
els | +0.035 |
hole | +0.035 |
ening | +0.034 |
bulb | +0.034 |
ool | +0.033 |
Neg logit contributions | |
His | -0.054 |
Jack | -0.054 |
He | -0.052 |
They | -0.049 |
Kids | -0.049 |
Token | - |
Token ID | 12 |
Feature activation | -8.81e-02 |
Pos logit contributions | |
ps | +0.252 |
cks | +0.238 |
ening | +0.237 |
angles | +0.235 |
flap | +0.230 |
Neg logit contributions | |
He | -0.252 |
They | -0.248 |
His | -0.236 |
Jack | -0.228 |
Johnny | -0.227 |
Token | cream |
Token ID | 36277 |
Feature activation | -1.40e-01 |
Pos logit contributions | |
bulb | +0.178 |
curtain | +0.170 |
globe | +0.170 |
ps | +0.169 |
hole | +0.167 |
Neg logit contributions | |
His | -0.114 |
He | -0.113 |
Jack | -0.111 |
They | -0.100 |
Sure | -0.099 |
Token | s |
Token ID | 82 |
Feature activation | -3.84e-02 |
Pos logit contributions | |
cks | +0.248 |
ps | +0.230 |
els | +0.222 |
ening | +0.220 |
cking | +0.217 |
Neg logit contributions | |
He | -0.214 |
They | -0.207 |
He | -0.207 |
he | -0.196 |
His | -0.190 |
Token | with |
Token ID | 351 |
Feature activation | -5.32e-02 |
Pos logit contributions | |
cks | +0.061 |
globe | +0.060 |
plank | +0.058 |
ps | +0.056 |
aro | +0.056 |
Neg logit contributions | |
He | -0.062 |
His | -0.059 |
They | -0.056 |
He | -0.056 |
His | -0.056 |
Token | colourful |
Token ID | 44799 |
Feature activation | -1.24e-01 |
Pos logit contributions | |
globe | +0.088 |
ool | +0.088 |
hole | +0.086 |
ig | +0.085 |
atop | +0.084 |
Neg logit contributions | |
He | -0.085 |
His | -0.082 |
Jack | -0.078 |
They | -0.078 |
His | -0.076 |
Token | sprink |
Token ID | 27562 |
Feature activation | -2.11e-01 |
Pos logit contributions | |
cks | +0.161 |
globe | +0.156 |
ps | +0.156 |
ool | +0.153 |
hole | +0.151 |
Neg logit contributions | |
He | -0.249 |
His | -0.245 |
His | -0.226 |
Dad | -0.223 |
Jack | -0.222 |
Token | les |
Token ID | 829 |
Feature activation | -1.32e-01 |
Pos logit contributions | |
fence | +0.284 |
plank | +0.282 |
trunk | +0.281 |
mant | +0.276 |
ps | +0.275 |
Neg logit contributions | |
Jack | -0.370 |
He | -0.354 |
They | -0.329 |
Johnny | -0.318 |
His | -0.317 |
Token | , |
Token ID | 11 |
Feature activation | +1.20e-02 |
Pos logit contributions | |
globe | +0.208 |
plank | +0.207 |
ool | +0.205 |
icer | +0.200 |
bulb | +0.195 |
Neg logit contributions | |
They | -0.190 |
He | -0.185 |
She | -0.175 |
She | -0.168 |
Jack | -0.167 |
Token | chocolate |
Token ID | 11311 |
Feature activation | -1.39e-01 |
Pos logit contributions | |
He | +0.020 |
His | +0.020 |
Jack | +0.019 |
They | +0.019 |
Something | +0.018 |
Neg logit contributions | |
ool | -0.020 |
els | -0.019 |
globe | -0.019 |
hole | -0.018 |
bulb | -0.018 |
Token | chips |
Token ID | 12014 |
Feature activation | -1.66e-01 |
Pos logit contributions | |
angles | +0.215 |
cks | +0.205 |
hole | +0.205 |
ps | +0.204 |
atop | +0.199 |
Neg logit contributions | |
They | -0.219 |
He | -0.216 |
He | -0.202 |
His | -0.202 |
Un | -0.191 |
Token | and |
Token ID | 290 |
Feature activation | +2.82e-02 |
Pos logit contributions | |
bulb | +0.284 |
ool | +0.282 |
els | +0.281 |
globe | +0.281 |
an | +0.276 |
Neg logit contributions | |
They | -0.231 |
He | -0.224 |
She | -0.211 |
She | -0.205 |
He | -0.200 |
Token | of |
Token ID | 286 |
Feature activation | -3.70e-02 |
Pos logit contributions | |
els | +0.303 |
ps | +0.293 |
globe | +0.289 |
ening | +0.284 |
cks | +0.281 |
Neg logit contributions | |
He | -0.238 |
His | -0.234 |
Jack | -0.233 |
They | -0.212 |
Dad | -0.208 |
Token | yogurt |
Token ID | 32132 |
Feature activation | -1.55e-01 |
Pos logit contributions | |
hole | +0.063 |
els | +0.061 |
globe | +0.060 |
ps | +0.059 |
ool | +0.058 |
Neg logit contributions | |
Jack | -0.061 |
His | -0.060 |
He | -0.059 |
Dad | -0.054 |
Sure | -0.053 |
Token | full |
Token ID | 1336 |
Feature activation | -8.73e-02 |
Pos logit contributions | |
els | +0.218 |
ening | +0.210 |
hole | +0.208 |
ps | +0.206 |
cks | +0.202 |
Neg logit contributions | |
He | -0.285 |
Jack | -0.279 |
His | -0.279 |
They | -0.265 |
Dad | -0.256 |
Token | of |
Token ID | 286 |
Feature activation | -3.21e-02 |
Pos logit contributions | |
els | +0.168 |
globe | +0.164 |
ps | +0.163 |
hole | +0.163 |
cks | +0.159 |
Neg logit contributions | |
His | -0.120 |
Jack | -0.120 |
He | -0.119 |
Dad | -0.107 |
His | -0.104 |
Token | colorful |
Token ID | 20239 |
Feature activation | -1.29e-01 |
Pos logit contributions | |
hole | +0.051 |
els | +0.050 |
globe | +0.049 |
ps | +0.048 |
bulb | +0.048 |
Neg logit contributions | |
Jack | -0.056 |
His | -0.055 |
He | -0.054 |
Dad | -0.050 |
Thanks | -0.049 |
Token | sprink |
Token ID | 27562 |
Feature activation | -2.11e-01 |
Pos logit contributions | |
hole | +0.181 |
globe | +0.176 |
ps | +0.171 |
ool | +0.169 |
bulb | +0.169 |
Neg logit contributions | |
Jack | -0.249 |
His | -0.243 |
He | -0.241 |
Dad | -0.223 |
They | -0.218 |
Token | les |
Token ID | 829 |
Feature activation | -1.32e-01 |
Pos logit contributions | |
trunk | +0.291 |
hole | +0.277 |
globe | +0.276 |
prope | +0.273 |
chim | +0.271 |
Neg logit contributions | |
Jack | -0.403 |
He | -0.373 |
They | -0.358 |
His | -0.342 |
Jimmy | -0.334 |
Token | and |
Token ID | 290 |
Feature activation | +2.82e-02 |
Pos logit contributions | |
cks | +0.209 |
ool | +0.209 |
els | +0.208 |
globe | +0.207 |
bulb | +0.204 |
Neg logit contributions | |
He | -0.198 |
Jack | -0.194 |
They | -0.194 |
His | -0.181 |
She | -0.177 |
Token | cher |
Token ID | 21382 |
Feature activation | -1.47e-01 |
Pos logit contributions | |
Jack | +0.050 |
His | +0.049 |
He | +0.048 |
Thanks | +0.045 |
They | +0.044 |
Neg logit contributions | |
els | -0.043 |
hole | -0.042 |
globe | -0.042 |
bulb | -0.041 |
ool | -0.041 |
Token | ries |
Token ID | 1678 |
Feature activation | -1.26e-01 |
Pos logit contributions | |
globe | +0.110 |
bulb | +0.108 |
ool | +0.105 |
hole | +0.102 |
trunk | +0.100 |
Neg logit contributions | |
Jack | -0.369 |
He | -0.350 |
His | -0.346 |
Johnny | -0.340 |
David | -0.337 |
Token | . |
Token ID | 13 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
els | +0.194 |
globe | +0.193 |
ool | +0.191 |
hole | +0.190 |
ening | +0.189 |
Neg logit contributions | |
He | -0.202 |
They | -0.200 |
Jack | -0.189 |
His | -0.187 |
She | -0.182 |
Token | he |
Token ID | 339 |
Feature activation | +1.89e-02 |
Pos logit contributions | |
Kids | +0.063 |
Jack | +0.060 |
His | +0.060 |
He | +0.059 |
Thanks | +0.059 |
Neg logit contributions | |
els | -0.096 |
pieces | -0.095 |
object | -0.094 |
hanging | -0.093 |
bulb | -0.093 |
Token | accidentally |
Token ID | 14716 |
Feature activation | -1.95e-02 |
Pos logit contributions | |
His | +0.025 |
Jack | +0.025 |
He | +0.024 |
Kids | +0.022 |
David | +0.022 |
Neg logit contributions | |
els | -0.037 |
bulb | -0.036 |
hole | -0.036 |
globe | -0.035 |
ps | -0.035 |
Token | knocked |
Token ID | 13642 |
Feature activation | -6.70e-02 |
Pos logit contributions | |
globe | +0.034 |
hole | +0.034 |
ool | +0.033 |
els | +0.033 |
bulbs | +0.032 |
Neg logit contributions | |
His | -0.031 |
Jack | -0.030 |
He | -0.028 |
His | -0.027 |
Thanks | -0.026 |
Token | over |
Token ID | 625 |
Feature activation | -3.71e-02 |
Pos logit contributions | |
globe | +0.089 |
ool | +0.088 |
bulb | +0.088 |
hole | +0.086 |
els | +0.085 |
Neg logit contributions | |
Jack | -0.135 |
His | -0.135 |
He | -0.132 |
Dad | -0.122 |
Kids | -0.121 |
Token | his |
Token ID | 465 |
Feature activation | -3.40e-02 |
Pos logit contributions | |
els | +0.068 |
bulb | +0.066 |
hole | +0.066 |
globe | +0.065 |
ps | +0.065 |
Neg logit contributions | |
His | -0.053 |
Jack | -0.053 |
He | -0.052 |
Kids | -0.048 |
David | -0.047 |
Token | chess |
Token ID | 19780 |
Feature activation | -2.11e-01 |
Pos logit contributions | |
els | +0.055 |
hole | +0.054 |
ening | +0.053 |
ool | +0.052 |
atop | +0.051 |
Neg logit contributions | |
His | -0.057 |
He | -0.055 |
Jack | -0.054 |
His | -0.052 |
Something | -0.050 |
Token | pieces |
Token ID | 5207 |
Feature activation | -1.71e-01 |
Pos logit contributions | |
els | +0.299 |
ening | +0.268 |
cks | +0.268 |
bulbs | +0.261 |
atop | +0.260 |
Neg logit contributions | |
His | -0.412 |
He | -0.410 |
Jack | -0.394 |
He | -0.366 |
So | -0.365 |
Token | . |
Token ID | 13 |
Feature activation | +1.15e-01 |
Pos logit contributions | |
els | +0.264 |
globe | +0.255 |
chim | +0.249 |
bulb | +0.247 |
bulbs | +0.246 |
Neg logit contributions | |
He | -0.316 |
His | -0.314 |
Jack | -0.293 |
They | -0.283 |
He | -0.269 |
Token | " |
Token ID | 366 |
Feature activation | +6.25e-02 |
Pos logit contributions | |
Jack | +0.127 |
He | +0.126 |
His | +0.124 |
Kids | +0.123 |
Dad | +0.118 |
Neg logit contributions | |
els | -0.223 |
trunk | -0.220 |
pieces | -0.219 |
ool | -0.215 |
object | -0.214 |
Token | Oh |
Token ID | 5812 |
Feature activation | +3.89e-02 |
Pos logit contributions | |
ete | +0.067 |
His | +0.067 |
Her | +0.061 |
His | +0.060 |
Something | +0.058 |
Neg logit contributions | |
pieces | -0.122 |
et | -0.119 |
els | -0.119 |
object | -0.117 |
blades | -0.114 |
Token | no |
Token ID | 645 |
Feature activation | -2.31e-02 |
Pos logit contributions | |
His | +0.055 |
Jack | +0.055 |
He | +0.054 |
Kids | +0.051 |
David | +0.049 |
Neg logit contributions | |
els | -0.071 |
bulb | -0.070 |
hole | -0.069 |
globe | -0.069 |
ps | -0.068 |
Token | hit |
Token ID | 2277 |
Feature activation | -5.85e-02 |
Pos logit contributions | |
Jack | +0.001 |
His | +0.001 |
He | +0.001 |
Sure | +0.001 |
So | +0.001 |
Neg logit contributions | |
els | -0.002 |
globe | -0.002 |
hole | -0.002 |
ool | -0.002 |
ening | -0.002 |
Token | the |
Token ID | 262 |
Feature activation | -2.68e-02 |
Pos logit contributions | |
els | +0.100 |
bulb | +0.100 |
hole | +0.099 |
ool | +0.098 |
globe | +0.098 |
Neg logit contributions | |
Jack | -0.091 |
His | -0.091 |
He | -0.087 |
Kids | -0.080 |
Dad | -0.080 |
Token | rock |
Token ID | 3881 |
Feature activation | -1.42e-01 |
Pos logit contributions | |
els | +0.043 |
ool | +0.041 |
cks | +0.039 |
cards | +0.039 |
hole | +0.038 |
Neg logit contributions | |
He | -0.053 |
His | -0.049 |
Jack | -0.046 |
He | -0.043 |
Dad | -0.042 |
Token | with |
Token ID | 351 |
Feature activation | -2.19e-02 |
Pos logit contributions | |
els | +0.263 |
globe | +0.248 |
hole | +0.247 |
ening | +0.245 |
ool | +0.244 |
Neg logit contributions | |
His | -0.220 |
He | -0.218 |
He | -0.194 |
His | -0.187 |
They | -0.186 |
Token | their |
Token ID | 511 |
Feature activation | -2.42e-02 |
Pos logit contributions | |
ool | +0.043 |
hole | +0.043 |
els | +0.042 |
globe | +0.041 |
bulb | +0.041 |
Neg logit contributions | |
His | -0.030 |
Jack | -0.029 |
He | -0.029 |
Thanks | -0.026 |
His | -0.025 |
Token | hamm |
Token ID | 25755 |
Feature activation | -2.09e-01 |
Pos logit contributions | |
ool | +0.030 |
hole | +0.030 |
atop | +0.029 |
globe | +0.029 |
ening | +0.028 |
Neg logit contributions | |
His | -0.050 |
He | -0.049 |
Jack | -0.048 |
His | -0.046 |
Dad | -0.045 |
Token | ers |
Token ID | 364 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
chim | +0.375 |
bulb | +0.360 |
globe | +0.347 |
stem | +0.336 |
bulbs | +0.332 |
Neg logit contributions | |
He | -0.348 |
His | -0.336 |
Jack | -0.321 |
They | -0.275 |
He | -0.270 |
Token | . |
Token ID | 13 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
ool | +0.236 |
els | +0.233 |
iest | +0.223 |
bulb | +0.223 |
globe | +0.220 |
Neg logit contributions | |
He | -0.155 |
His | -0.144 |
So | -0.143 |
They | -0.142 |
He | -0.132 |
Token | With |
Token ID | 2080 |
Feature activation | +7.32e-03 |
Pos logit contributions | |
His | +0.106 |
Jack | +0.106 |
He | +0.100 |
Kids | +0.093 |
David | +0.093 |
Neg logit contributions | |
els | -0.254 |
ps | -0.250 |
globe | -0.249 |
pieces | -0.249 |
hole | -0.248 |
Token | their |
Token ID | 511 |
Feature activation | -2.62e-03 |
Pos logit contributions | |
His | +0.010 |
Jack | +0.010 |
He | +0.010 |
Thanks | +0.009 |
Dad | +0.009 |
Neg logit contributions | |
els | -0.014 |
hole | -0.014 |
bulb | -0.014 |
globe | -0.013 |
ool | -0.013 |
Token | help |
Token ID | 1037 |
Feature activation | -5.30e-02 |
Pos logit contributions | |
hole | +0.004 |
globe | +0.004 |
ening | +0.004 |
els | +0.004 |
ool | +0.004 |
Neg logit contributions | |
His | -0.005 |
He | -0.005 |
Jack | -0.005 |
His | -0.004 |
Brother | -0.004 |
Token | delicious |
Token ID | 12625 |
Feature activation | -1.13e-01 |
Pos logit contributions | |
els | +0.018 |
ps | +0.017 |
cks | +0.016 |
et | +0.016 |
ening | +0.016 |
Neg logit contributions | |
Jack | -0.018 |
His | -0.018 |
He | -0.018 |
Sure | -0.017 |
David | -0.016 |
Token | cake |
Token ID | 12187 |
Feature activation | -1.52e-01 |
Pos logit contributions | |
els | +0.224 |
cks | +0.217 |
hole | +0.217 |
ps | +0.214 |
atop | +0.197 |
Neg logit contributions | |
Jack | -0.162 |
His | -0.140 |
please | -0.139 |
Dad | -0.138 |
He | -0.138 |
Token | topped |
Token ID | 20633 |
Feature activation | -1.28e-01 |
Pos logit contributions | |
els | +0.222 |
ps | +0.214 |
ening | +0.207 |
bulb | +0.206 |
blades | +0.200 |
Neg logit contributions | |
Jack | -0.276 |
He | -0.275 |
His | -0.273 |
They | -0.255 |
Dad | -0.249 |
Token | with |
Token ID | 351 |
Feature activation | -4.79e-02 |
Pos logit contributions | |
els | +0.203 |
bulb | +0.201 |
hole | +0.201 |
ening | +0.195 |
blades | +0.194 |
Neg logit contributions | |
His | -0.221 |
Jack | -0.214 |
He | -0.210 |
They | -0.195 |
Her | -0.194 |
Token | colourful |
Token ID | 44799 |
Feature activation | -1.29e-01 |
Pos logit contributions | |
hole | +0.080 |
els | +0.077 |
ool | +0.077 |
globe | +0.075 |
bulb | +0.074 |
Neg logit contributions | |
Jack | -0.080 |
His | -0.077 |
He | -0.077 |
Dad | -0.071 |
David | -0.071 |
Token | icing |
Token ID | 41567 |
Feature activation | -2.04e-01 |
Pos logit contributions | |
hole | +0.156 |
cks | +0.146 |
ool | +0.144 |
ps | +0.143 |
globe | +0.141 |
Neg logit contributions | |
Jack | -0.266 |
He | -0.260 |
His | -0.258 |
Dad | -0.247 |
They | -0.243 |
Token | . |
Token ID | 13 |
Feature activation | +1.08e-01 |
Pos logit contributions | |
bulb | +0.300 |
els | +0.299 |
cks | +0.298 |
ps | +0.291 |
bulbs | +0.283 |
Neg logit contributions | |
He | -0.357 |
Jack | -0.342 |
His | -0.332 |
They | -0.332 |
Dad | -0.311 |
Token | |
Token ID | 220 |
Feature activation | +3.70e-02 |
Pos logit contributions | |
Jack | +0.131 |
His | +0.127 |
Sorry | +0.121 |
Kids | +0.121 |
He | +0.118 |
Neg logit contributions | |
hole | -0.195 |
globe | -0.194 |
trunk | -0.190 |
chim | -0.188 |
els | -0.186 |
Token |
|
Token ID | 198 |
Feature activation | +9.44e-02 |
Pos logit contributions | |
His | +0.036 |
Jack | +0.035 |
He | +0.032 |
Kids | +0.031 |
Thanks | +0.030 |
Neg logit contributions | |
els | -0.083 |
hole | -0.082 |
bulb | -0.080 |
globe | -0.080 |
ps | -0.080 |
Token |
|
Token ID | 198 |
Feature activation | +1.01e-01 |
Pos logit contributions | |
Sorry | +0.108 |
Um | +0.104 |
Jack | +0.104 |
Kids | +0.102 |
Okay | +0.101 |
Neg logit contributions | |
globe | -0.157 |
trunk | -0.152 |
atop | -0.151 |
els | -0.151 |
hole | -0.150 |
Token | John |
Token ID | 7554 |
Feature activation | -1.36e-03 |
Pos logit contributions | |
Something | +0.095 |
Um | +0.083 |
possibly | +0.079 |
Deal | +0.078 |
Kids | +0.076 |
Neg logit contributions | |
atop | -0.184 |
globe | -0.181 |
cut | -0.174 |
ool | -0.172 |
hanging | -0.171 |
Token | 's |
Token ID | 338 |
Feature activation | +1.82e-02 |
Pos logit contributions | |
els | +0.001 |
et | +0.001 |
bulb | +0.001 |
object | +0.001 |
pieces | +0.001 |
Neg logit contributions | |
His | -0.001 |
Jack | -0.001 |
He | -0.001 |
Thanks | -0.001 |
David | -0.001 |
Token | get |
Token ID | 651 |
Feature activation | -2.98e-02 |
Pos logit contributions | |
Jack | +0.026 |
His | +0.026 |
He | +0.025 |
Thanks | +0.023 |
Dad | +0.023 |
Neg logit contributions | |
els | -0.033 |
hole | -0.033 |
ening | -0.032 |
globe | -0.032 |
ps | -0.032 |
Token | all |
Token ID | 477 |
Feature activation | -3.80e-02 |
Pos logit contributions | |
hole | +0.053 |
globe | +0.052 |
ening | +0.051 |
ool | +0.050 |
els | +0.049 |
Neg logit contributions | |
Jack | -0.046 |
His | -0.044 |
He | -0.043 |
Dad | -0.041 |
Kids | -0.041 |
Token | the |
Token ID | 262 |
Feature activation | -2.68e-02 |
Pos logit contributions | |
globe | +0.074 |
hole | +0.071 |
ps | +0.069 |
ening | +0.069 |
atop | +0.069 |
Neg logit contributions | |
Jack | -0.052 |
His | -0.047 |
They | -0.046 |
He | -0.045 |
Kids | -0.044 |
Token | colorful |
Token ID | 20239 |
Feature activation | -1.19e-01 |
Pos logit contributions | |
globe | +0.040 |
atop | +0.039 |
cks | +0.038 |
ening | +0.036 |
hole | +0.036 |
Neg logit contributions | |
Jack | -0.046 |
L | -0.045 |
He | -0.044 |
Kids | -0.043 |
When | -0.043 |
Token | powd |
Token ID | 32279 |
Feature activation | -2.03e-01 |
Pos logit contributions | |
cks | +0.168 |
atop | +0.166 |
globe | +0.165 |
hole | +0.163 |
ps | +0.154 |
Neg logit contributions | |
Jack | -0.195 |
They | -0.193 |
Kids | -0.190 |
L | -0.189 |
He | -0.185 |
Token | ers |
Token ID | 364 |
Feature activation | -1.25e-01 |
Pos logit contributions | |
hole | +0.261 |
trunk | +0.260 |
ening | +0.250 |
chim | +0.246 |
prope | +0.239 |
Neg logit contributions | |
Jack | -0.391 |
They | -0.372 |
Kids | -0.361 |
Johnny | -0.357 |
He | -0.353 |
Token | ." |
Token ID | 526 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
els | +0.196 |
ening | +0.183 |
ool | +0.182 |
cks | +0.172 |
bulb | +0.171 |
Neg logit contributions | |
He | -0.213 |
They | -0.212 |
Jack | -0.210 |
please | -0.208 |
His | -0.197 |
Token | So |
Token ID | 1406 |
Feature activation | +3.67e-02 |
Pos logit contributions | |
Un | +0.136 |
His | +0.136 |
He | +0.135 |
David | +0.131 |
Thanks | +0.129 |
Neg logit contributions | |
els | -0.204 |
pieces | -0.201 |
globe | -0.201 |
hole | -0.200 |
trunk | -0.199 |
Token | , |
Token ID | 11 |
Feature activation | +6.43e-02 |
Pos logit contributions | |
Jack | +0.070 |
His | +0.069 |
He | +0.069 |
Kids | +0.066 |
David | +0.066 |
Neg logit contributions | |
els | -0.049 |
bulb | -0.047 |
globe | -0.047 |
atop | -0.045 |
hole | -0.045 |
Token | the |
Token ID | 262 |
Feature activation | +1.02e-02 |
Pos logit contributions | |
Jack | +0.078 |
He | +0.078 |
His | +0.077 |
Kids | +0.074 |
David | +0.074 |
Neg logit contributions | |
els | -0.127 |
bulb | -0.124 |
globe | -0.122 |
pieces | -0.122 |
atop | -0.121 |
Token | not |
Token ID | 407 |
Feature activation | +2.34e-02 |
Pos logit contributions | |
Kids | +0.113 |
Jack | +0.102 |
Da | +0.102 |
Un | +0.102 |
Sure | +0.099 |
Neg logit contributions | |
bulb | -0.072 |
falling | -0.068 |
ps | -0.066 |
ring | -0.066 |
hanging | -0.065 |
Token | find |
Token ID | 1064 |
Feature activation | -3.25e-02 |
Pos logit contributions | |
His | +0.043 |
Jack | +0.043 |
Kids | +0.042 |
He | +0.041 |
Thanks | +0.040 |
Neg logit contributions | |
bulb | -0.033 |
hole | -0.032 |
els | -0.031 |
globe | -0.030 |
egg | -0.030 |
Token | his |
Token ID | 465 |
Feature activation | -2.36e-02 |
Pos logit contributions | |
ool | +0.068 |
globe | +0.067 |
hole | +0.066 |
atop | +0.065 |
els | +0.064 |
Neg logit contributions | |
His | -0.038 |
Jack | -0.036 |
He | -0.036 |
His | -0.034 |
He | -0.033 |
Token | famous |
Token ID | 5863 |
Feature activation | -1.17e-01 |
Pos logit contributions | |
hole | +0.042 |
els | +0.042 |
ool | +0.041 |
atop | +0.040 |
ening | +0.039 |
Neg logit contributions | |
Jack | -0.035 |
His | -0.034 |
He | -0.033 |
His | -0.033 |
Dad | -0.032 |
Token | las |
Token ID | 39990 |
Feature activation | -1.24e-01 |
Pos logit contributions | |
els | +0.222 |
hole | +0.220 |
cks | +0.219 |
ps | +0.210 |
atop | +0.208 |
Neg logit contributions | |
Jack | -0.173 |
His | -0.160 |
He | -0.158 |
Dad | -0.150 |
His | -0.148 |
Token | agna |
Token ID | 48669 |
Feature activation | -2.02e-01 |
Pos logit contributions | |
els | +0.207 |
bulb | +0.201 |
hole | +0.200 |
globe | +0.197 |
ps | +0.195 |
Neg logit contributions | |
His | -0.198 |
Jack | -0.197 |
He | -0.192 |
Kids | -0.178 |
David | -0.177 |
Token | . |
Token ID | 13 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
els | +0.315 |
bulb | +0.309 |
blades | +0.303 |
atop | +0.302 |
bulbs | +0.297 |
Neg logit contributions | |
He | -0.321 |
His | -0.318 |
Jack | -0.313 |
He | -0.292 |
They | -0.289 |
Token |
|
Token ID | 198 |
Feature activation | +9.40e-02 |
Pos logit contributions | |
His | +0.089 |
Jack | +0.085 |
Kids | +0.081 |
Her | +0.078 |
He | +0.077 |
Neg logit contributions | |
hole | -0.268 |
trunk | -0.263 |
pieces | -0.262 |
els | -0.261 |
ool | -0.260 |
Token |
|
Token ID | 198 |
Feature activation | +1.01e-01 |
Pos logit contributions | |
His | +0.099 |
Jack | +0.098 |
Thanks | +0.097 |
Kids | +0.097 |
Sorry | +0.095 |
Neg logit contributions | |
hole | -0.175 |
ool | -0.175 |
globe | -0.174 |
els | -0.173 |
pieces | -0.173 |
Token | The |
Token ID | 464 |
Feature activation | +1.78e-02 |
Pos logit contributions | |
Kids | +0.081 |
Her | +0.078 |
Ryan | +0.075 |
His | +0.075 |
Something | +0.074 |
Neg logit contributions | |
pieces | -0.199 |
ool | -0.194 |
atop | -0.193 |
object | -0.190 |
cut | -0.190 |
Token | chef |
Token ID | 21221 |
Feature activation | -7.80e-02 |
Pos logit contributions | |
His | +0.027 |
Jack | +0.026 |
He | +0.026 |
Brother | +0.024 |
Dad | +0.024 |
Neg logit contributions | |
hole | -0.034 |
els | -0.034 |
ool | -0.032 |
globe | -0.031 |
ening | -0.031 |
Token | she |
Token ID | 673 |
Feature activation | +1.72e-03 |
Pos logit contributions | |
His | +0.026 |
He | +0.025 |
Jack | +0.025 |
Kids | +0.022 |
Thanks | +0.022 |
Neg logit contributions | |
bulb | -0.048 |
els | -0.048 |
egg | -0.046 |
globe | -0.046 |
hole | -0.046 |
Token | could |
Token ID | 714 |
Feature activation | +2.19e-02 |
Pos logit contributions | |
His | +0.002 |
Jack | +0.002 |
He | +0.002 |
Kids | +0.002 |
David | +0.002 |
Neg logit contributions | |
els | -0.004 |
bulb | -0.003 |
hole | -0.003 |
globe | -0.003 |
ool | -0.003 |
Token | go |
Token ID | 467 |
Feature activation | -5.12e-02 |
Pos logit contributions | |
His | +0.029 |
He | +0.028 |
Jack | +0.027 |
Dad | +0.025 |
His | +0.024 |
Neg logit contributions | |
els | -0.043 |
ening | -0.043 |
hole | -0.042 |
globe | -0.042 |
ps | -0.042 |
Token | get |
Token ID | 651 |
Feature activation | -5.47e-02 |
Pos logit contributions | |
els | +0.074 |
globe | +0.072 |
ening | +0.071 |
ool | +0.071 |
ps | +0.071 |
Neg logit contributions | |
His | -0.094 |
He | -0.093 |
Jack | -0.092 |
David | -0.084 |
Dad | -0.083 |
Token | an |
Token ID | 281 |
Feature activation | -5.02e-02 |
Pos logit contributions | |
globe | +0.112 |
ool | +0.111 |
hole | +0.110 |
ening | +0.108 |
ps | +0.103 |
Neg logit contributions | |
His | -0.069 |
Jack | -0.066 |
He | -0.065 |
Dad | -0.062 |
His | -0.059 |
Token | X |
Token ID | 1395 |
Feature activation | -2.01e-01 |
Pos logit contributions | |
globe | +0.075 |
ps | +0.074 |
els | +0.074 |
cks | +0.072 |
atop | +0.072 |
Neg logit contributions | |
His | -0.085 |
He | -0.084 |
Jack | -0.080 |
His | -0.079 |
Dad | -0.078 |
Token | - |
Token ID | 12 |
Feature activation | -8.80e-02 |
Pos logit contributions | |
ening | +0.257 |
blades | +0.253 |
chim | +0.248 |
angles | +0.239 |
els | +0.237 |
Neg logit contributions | |
His | -0.404 |
He | -0.402 |
please | -0.367 |
Jack | -0.353 |
He | -0.345 |
Token | ray |
Token ID | 2433 |
Feature activation | -1.46e-01 |
Pos logit contributions | |
bulb | +0.133 |
globe | +0.131 |
ening | +0.129 |
ps | +0.129 |
els | +0.128 |
Neg logit contributions | |
His | -0.159 |
He | -0.157 |
Jack | -0.152 |
Sure | -0.141 |
Dad | -0.141 |
Token | . |
Token ID | 13 |
Feature activation | +1.16e-01 |
Pos logit contributions | |
ps | +0.238 |
cks | +0.238 |
els | +0.227 |
ening | +0.221 |
globe | +0.216 |
Neg logit contributions | |
He | -0.244 |
His | -0.233 |
He | -0.226 |
Jack | -0.223 |
please | -0.222 |
Token |
|
Token ID | 198 |
Feature activation | +9.27e-02 |
Pos logit contributions | |
He | +0.140 |
His | +0.139 |
Jack | +0.133 |
David | +0.126 |
Harry | +0.124 |
Neg logit contributions | |
bulb | -0.217 |
els | -0.216 |
globe | -0.215 |
chim | -0.214 |
trunk | -0.213 |
Token |
|
Token ID | 198 |
Feature activation | +9.83e-02 |
Pos logit contributions | |
Sorry | +0.122 |
Thanks | +0.122 |
His | +0.121 |
He | +0.119 |
Something | +0.117 |
Neg logit contributions | |
ool | -0.155 |
els | -0.145 |
globe | -0.143 |
hole | -0.142 |
bulb | -0.139 |
Token | real |
Token ID | 1103 |
Feature activation | -1.01e-01 |
Pos logit contributions | |
els | +0.049 |
ps | +0.044 |
cks | +0.042 |
ening | +0.042 |
et | +0.040 |
Neg logit contributions | |
He | -0.051 |
Jack | -0.049 |
His | -0.049 |
Sure | -0.044 |
please | -0.044 |
Token | treasure |
Token ID | 14068 |
Feature activation | -1.76e-01 |
Pos logit contributions | |
els | +0.179 |
cks | +0.169 |
ps | +0.160 |
ening | +0.158 |
hole | +0.156 |
Neg logit contributions | |
He | -0.167 |
His | -0.165 |
Jack | -0.149 |
He | -0.146 |
His | -0.144 |
Token | chest |
Token ID | 7721 |
Feature activation | -1.58e-01 |
Pos logit contributions | |
ening | +0.270 |
els | +0.264 |
blades | +0.248 |
ps | +0.235 |
cks | +0.230 |
Neg logit contributions | |
He | -0.294 |
They | -0.268 |
His | -0.264 |
He | -0.262 |
possibly | -0.261 |
Token | filled |
Token ID | 5901 |
Feature activation | -1.25e-01 |
Pos logit contributions | |
els | +0.298 |
blades | +0.285 |
ening | +0.270 |
ps | +0.258 |
cks | +0.255 |
Neg logit contributions | |
He | -0.211 |
They | -0.199 |
he | -0.187 |
His | -0.187 |
He | -0.185 |
Token | with |
Token ID | 351 |
Feature activation | -6.45e-02 |
Pos logit contributions | |
els | +0.201 |
hole | +0.200 |
ening | +0.193 |
ps | +0.193 |
bulb | +0.191 |
Neg logit contributions | |
Jack | -0.214 |
His | -0.205 |
He | -0.200 |
They | -0.194 |
Her | -0.186 |
Token | gold |
Token ID | 3869 |
Feature activation | -2.00e-01 |
Pos logit contributions | |
ool | +0.094 |
els | +0.089 |
hole | +0.088 |
globe | +0.083 |
ig | +0.082 |
Neg logit contributions | |
He | -0.120 |
His | -0.114 |
Jack | -0.111 |
He | -0.107 |
They | -0.107 |
Token | coins |
Token ID | 10796 |
Feature activation | -1.92e-01 |
Pos logit contributions | |
ening | +0.283 |
els | +0.277 |
bulb | +0.273 |
cks | +0.273 |
angle | +0.271 |
Neg logit contributions | |
He | -0.340 |
He | -0.300 |
They | -0.291 |
His | -0.273 |
he | -0.269 |
Token | and |
Token ID | 290 |
Feature activation | +1.65e-02 |
Pos logit contributions | |
els | +0.344 |
blades | +0.334 |
ool | +0.334 |
bulb | +0.332 |
ening | +0.322 |
Neg logit contributions | |
He | -0.259 |
They | -0.240 |
he | -0.219 |
She | -0.218 |
He | -0.211 |
Token | jewels |
Token ID | 42701 |
Feature activation | -1.52e-01 |
Pos logit contributions | |
His | +0.037 |
He | +0.037 |
Jack | +0.036 |
Something | +0.034 |
They | +0.034 |
Neg logit contributions | |
els | -0.017 |
hole | -0.016 |
ool | -0.016 |
ening | -0.016 |
globe | -0.015 |
Token | . |
Token ID | 13 |
Feature activation | +1.16e-01 |
Pos logit contributions | |
els | +0.271 |
ool | +0.254 |
blades | +0.252 |
et | +0.252 |
ening | +0.250 |
Neg logit contributions | |
He | -0.223 |
They | -0.205 |
She | -0.194 |
he | -0.188 |
He | -0.185 |
Token | |
Token ID | 220 |
Feature activation | +4.47e-02 |
Pos logit contributions | |
David | +0.201 |
His | +0.199 |
Jack | +0.195 |
Un | +0.192 |
Dad | +0.185 |
Neg logit contributions | |
trunk | -0.172 |
pieces | -0.168 |
hole | -0.167 |
bulb | -0.165 |
box | -0.162 |
Token | he |
Token ID | 339 |
Feature activation | -2.77e-04 |
Pos logit contributions | |
els | +0.026 |
hole | +0.025 |
bulb | +0.025 |
globe | +0.025 |
ps | +0.025 |
Neg logit contributions | |
His | -0.008 |
Jack | -0.008 |
He | -0.007 |
Kids | -0.006 |
David | -0.006 |
Token | had |
Token ID | 550 |
Feature activation | +2.87e-03 |
Pos logit contributions | |
globe | +0.000 |
els | +0.000 |
bulb | +0.000 |
hole | +0.000 |
ps | +0.000 |
Neg logit contributions | |
His | -0.000 |
Jack | -0.000 |
He | -0.000 |
David | -0.000 |
Brother | -0.000 |
Token | never |
Token ID | 1239 |
Feature activation | +3.07e-03 |
Pos logit contributions | |
Jack | +0.004 |
His | +0.004 |
He | +0.004 |
Kids | +0.003 |
David | +0.003 |
Neg logit contributions | |
els | -0.005 |
bulb | -0.005 |
ps | -0.005 |
hole | -0.005 |
globe | -0.005 |
Token | found |
Token ID | 1043 |
Feature activation | -5.32e-02 |
Pos logit contributions | |
His | +0.005 |
Jack | +0.005 |
He | +0.005 |
Kids | +0.004 |
Thanks | +0.004 |
Neg logit contributions | |
els | -0.005 |
bulb | -0.005 |
globe | -0.005 |
hole | -0.005 |
ool | -0.005 |
Token | the |
Token ID | 262 |
Feature activation | -4.63e-02 |
Pos logit contributions | |
ool | +0.123 |
globe | +0.121 |
hole | +0.121 |
bulb | +0.120 |
chim | +0.119 |
Neg logit contributions | |
His | -0.052 |
He | -0.051 |
Jack | -0.048 |
His | -0.045 |
He | -0.043 |
Token | gold |
Token ID | 3869 |
Feature activation | -2.00e-01 |
Pos logit contributions | |
els | +0.066 |
ool | +0.062 |
ening | +0.061 |
hole | +0.059 |
ps | +0.058 |
Neg logit contributions | |
He | -0.093 |
Jack | -0.087 |
His | -0.086 |
He | -0.085 |
Thanks | -0.083 |
Token | . |
Token ID | 13 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
cks | +0.341 |
blades | +0.325 |
els | +0.322 |
ps | +0.321 |
ening | +0.317 |
Neg logit contributions | |
He | -0.355 |
His | -0.332 |
He | -0.322 |
he | -0.298 |
Jack | -0.285 |
Token | He |
Token ID | 679 |
Feature activation | +3.01e-02 |
Pos logit contributions | |
Jack | +0.117 |
Kids | +0.114 |
Sorry | +0.108 |
Da | +0.108 |
Her | +0.106 |
Neg logit contributions | |
trunk | -0.200 |
hole | -0.190 |
edges | -0.187 |
globe | -0.184 |
pieces | -0.183 |
Token | learned |
Token ID | 4499 |
Feature activation | +9.78e-05 |
Pos logit contributions | |
Jack | +0.045 |
His | +0.043 |
David | +0.041 |
He | +0.040 |
Jim | +0.039 |
Neg logit contributions | |
els | -0.053 |
ps | -0.052 |
bulb | -0.052 |
pieces | -0.052 |
et | -0.051 |
Token | a |
Token ID | 257 |
Feature activation | +7.33e-03 |
Pos logit contributions | |
Jack | +0.000 |
Kids | +0.000 |
Brother | +0.000 |
Jim | +0.000 |
Sorry | +0.000 |
Neg logit contributions | |
ps | -0.000 |
object | -0.000 |
pieces | -0.000 |
els | -0.000 |
bulb | -0.000 |
Token | lesson |
Token ID | 11483 |
Feature activation | -4.91e-02 |
Pos logit contributions | |
His | +0.015 |
Jack | +0.015 |
He | +0.015 |
Kids | +0.014 |
David | +0.014 |
Neg logit contributions | |
els | -0.009 |
hole | -0.009 |
bulb | -0.009 |
globe | -0.009 |
ps | -0.009 |
Token | that |
Token ID | 326 |
Feature activation | +4.22e-03 |
Pos logit contributions | |
els | +0.229 |
ps | +0.223 |
ening | +0.220 |
cks | +0.216 |
hole | +0.214 |
Neg logit contributions | |
He | -0.246 |
His | -0.245 |
Jack | -0.233 |
He | -0.217 |
They | -0.216 |
Token | was |
Token ID | 373 |
Feature activation | +1.00e-02 |
Pos logit contributions | |
Jack | +0.005 |
His | +0.005 |
He | +0.005 |
Thanks | +0.004 |
His | +0.004 |
Neg logit contributions | |
els | -0.009 |
hole | -0.009 |
ps | -0.009 |
ening | -0.009 |
ool | -0.009 |
Token | covered |
Token ID | 5017 |
Feature activation | -1.35e-01 |
Pos logit contributions | |
His | +0.013 |
He | +0.013 |
Jack | +0.013 |
Thanks | +0.012 |
His | +0.012 |
Neg logit contributions | |
globe | -0.020 |
hole | -0.020 |
els | -0.019 |
ening | -0.018 |
ool | -0.018 |
Token | in |
Token ID | 287 |
Feature activation | -4.14e-02 |
Pos logit contributions | |
els | +0.200 |
hole | +0.198 |
chim | +0.189 |
bulb | +0.188 |
ool | +0.186 |
Neg logit contributions | |
His | -0.246 |
Jack | -0.244 |
He | -0.242 |
They | -0.217 |
Dad | -0.212 |
Token | colourful |
Token ID | 44799 |
Feature activation | -1.25e-01 |
Pos logit contributions | |
hole | +0.065 |
ig | +0.064 |
ool | +0.064 |
globe | +0.062 |
atop | +0.062 |
Neg logit contributions | |
He | -0.073 |
His | -0.070 |
Jack | -0.069 |
Dad | -0.068 |
They | -0.066 |
Token | sprink |
Token ID | 27562 |
Feature activation | -2.00e-01 |
Pos logit contributions | |
hole | +0.174 |
cks | +0.171 |
ps | +0.170 |
globe | +0.166 |
ool | +0.162 |
Neg logit contributions | |
He | -0.228 |
Jack | -0.223 |
His | -0.217 |
They | -0.215 |
She | -0.213 |
Token | les |
Token ID | 829 |
Feature activation | -1.27e-01 |
Pos logit contributions | |
globe | +0.248 |
ps | +0.245 |
trunk | +0.243 |
bulb | +0.242 |
bulbs | +0.242 |
Neg logit contributions | |
Jack | -0.389 |
He | -0.380 |
They | -0.356 |
His | -0.354 |
Jimmy | -0.335 |
Token | . |
Token ID | 13 |
Feature activation | +1.10e-01 |
Pos logit contributions | |
els | +0.211 |
bulb | +0.202 |
ool | +0.200 |
cks | +0.197 |
ps | +0.197 |
Neg logit contributions | |
He | -0.203 |
Jack | -0.189 |
His | -0.189 |
They | -0.188 |
He | -0.179 |
Token | When |
Token ID | 1649 |
Feature activation | +5.38e-02 |
Pos logit contributions | |
His | +0.103 |
Jack | +0.103 |
He | +0.097 |
Kids | +0.089 |
David | +0.086 |
Neg logit contributions | |
els | -0.253 |
hole | -0.249 |
bulb | -0.247 |
globe | -0.246 |
ening | -0.243 |
Token | the |
Token ID | 262 |
Feature activation | +8.42e-03 |
Pos logit contributions | |
His | +0.074 |
Jack | +0.074 |
He | +0.072 |
Kids | +0.064 |
David | +0.064 |
Neg logit contributions | |
els | -0.104 |
hole | -0.102 |
bulb | -0.100 |
ool | -0.099 |
ps | -0.099 |
Token | boy |
Token ID | 2933 |
Feature activation | -3.16e-02 |
Pos logit contributions | |
Jack | +0.012 |
His | +0.012 |
He | +0.012 |
Dad | +0.011 |
Sure | +0.011 |
Neg logit contributions | |
hole | -0.016 |
els | -0.016 |
ool | -0.016 |
ening | -0.015 |
cks | -0.015 |
Token | crying |
Token ID | 13774 |
Feature activation | -8.01e-02 |
Pos logit contributions | |
ool | +0.074 |
globe | +0.073 |
atop | +0.069 |
els | +0.069 |
bulb | +0.069 |
Neg logit contributions | |
He | -0.059 |
His | -0.059 |
Her | -0.056 |
Jack | -0.056 |
They | -0.055 |
Token | and |
Token ID | 290 |
Feature activation | +4.82e-02 |
Pos logit contributions | |
ool | +0.106 |
globe | +0.106 |
bulb | +0.104 |
hole | +0.104 |
els | +0.104 |
Neg logit contributions | |
His | -0.157 |
He | -0.157 |
Jack | -0.152 |
They | -0.144 |
Dad | -0.142 |
Token | Ben |
Token ID | 3932 |
Feature activation | -4.08e-02 |
Pos logit contributions | |
Jack | +0.079 |
His | +0.079 |
He | +0.077 |
Kids | +0.072 |
Brother | +0.071 |
Neg logit contributions | |
ool | -0.081 |
globe | -0.078 |
hole | -0.078 |
els | -0.078 |
bulb | -0.076 |
Token | holding |
Token ID | 4769 |
Feature activation | -7.81e-02 |
Pos logit contributions | |
ool | +0.077 |
globe | +0.074 |
bulb | +0.068 |
et | +0.067 |
plank | +0.067 |
Neg logit contributions | |
He | -0.060 |
His | -0.059 |
They | -0.057 |
Her | -0.056 |
Jack | -0.055 |
Token | the |
Token ID | 262 |
Feature activation | -4.64e-02 |
Pos logit contributions | |
hole | +0.156 |
bulb | +0.155 |
globe | +0.155 |
ool | +0.154 |
els | +0.152 |
Neg logit contributions | |
His | -0.100 |
Jack | -0.099 |
He | -0.096 |
Dad | -0.084 |
Her | -0.084 |
Token | chess |
Token ID | 19780 |
Feature activation | -2.00e-01 |
Pos logit contributions | |
els | +0.067 |
hole | +0.063 |
ool | +0.062 |
et | +0.062 |
ening | +0.062 |
Neg logit contributions | |
Jack | -0.087 |
His | -0.086 |
He | -0.086 |
Brother | -0.078 |
Thanks | -0.077 |
Token | board |
Token ID | 3096 |
Feature activation | -1.80e-01 |
Pos logit contributions | |
els | +0.252 |
ost | +0.227 |
ening | +0.223 |
chim | +0.221 |
cks | +0.220 |
Neg logit contributions | |
He | -0.423 |
Jack | -0.412 |
His | -0.410 |
They | -0.390 |
So | -0.371 |
Token | . |
Token ID | 13 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
els | +0.311 |
Tower | +0.302 |
atop | +0.299 |
chim | +0.299 |
blades | +0.295 |
Neg logit contributions | |
He | -0.270 |
His | -0.256 |
They | -0.244 |
She | -0.233 |
He | -0.231 |
Token |
|
Token ID | 198 |
Feature activation | +9.76e-02 |
Pos logit contributions | |
His | +0.103 |
Jack | +0.103 |
Thanks | +0.099 |
David | +0.098 |
Jim | +0.091 |
Neg logit contributions | |
bulb | -0.206 |
els | -0.199 |
bulbs | -0.197 |
chim | -0.197 |
globe | -0.195 |
Token |
|
Token ID | 198 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
His | +0.116 |
Jack | +0.113 |
Thanks | +0.113 |
David | +0.105 |
He | +0.104 |
Neg logit contributions | |
els | -0.185 |
bulb | -0.182 |
hole | -0.179 |
globe | -0.178 |
pieces | -0.175 |
Token | " |
Token ID | 1 |
Feature activation | +6.55e-02 |
Pos logit contributions | |
His | +0.190 |
His | +0.190 |
Thanks | +0.186 |
David | +0.185 |
Jack | +0.182 |
Neg logit contributions | |
els | -0.130 |
pieces | -0.121 |
bulb | -0.120 |
hole | -0.117 |
globe | -0.112 |
Token | in |
Token ID | 287 |
Feature activation | -1.69e-02 |
Pos logit contributions | |
els | +0.157 |
ool | +0.153 |
hole | +0.147 |
bulb | +0.144 |
ps | +0.137 |
Neg logit contributions | |
His | -0.232 |
He | -0.225 |
Jack | -0.216 |
They | -0.205 |
Thanks | -0.203 |
Token | an |
Token ID | 281 |
Feature activation | -4.77e-02 |
Pos logit contributions | |
els | +0.033 |
hole | +0.033 |
bulb | +0.032 |
globe | +0.032 |
ool | +0.032 |
Neg logit contributions | |
His | -0.023 |
Jack | -0.022 |
He | -0.022 |
David | -0.019 |
Kids | -0.019 |
Token | armour |
Token ID | 18588 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
els | +0.046 |
globe | +0.045 |
hole | +0.044 |
bulb | +0.043 |
ps | +0.043 |
Neg logit contributions | |
His | -0.110 |
He | -0.107 |
Jack | -0.107 |
Dad | -0.102 |
David | -0.101 |
Token | of |
Token ID | 286 |
Feature activation | -3.75e-02 |
Pos logit contributions | |
els | +0.222 |
bulbs | +0.195 |
ool | +0.193 |
ening | +0.193 |
bulb | +0.191 |
Neg logit contributions | |
His | -0.251 |
He | -0.249 |
His | -0.227 |
He | -0.227 |
They | -0.221 |
Token | shining |
Token ID | 22751 |
Feature activation | -1.32e-01 |
Pos logit contributions | |
els | +0.073 |
hole | +0.072 |
ool | +0.071 |
globe | +0.068 |
atop | +0.066 |
Neg logit contributions | |
His | -0.054 |
He | -0.052 |
Jack | -0.051 |
Un | -0.047 |
His | -0.046 |
Token | steel |
Token ID | 7771 |
Feature activation | -2.00e-01 |
Pos logit contributions | |
cks | +0.196 |
ps | +0.189 |
aro | +0.182 |
els | +0.174 |
cking | +0.169 |
Neg logit contributions | |
He | -0.230 |
His | -0.224 |
He | -0.220 |
please | -0.206 |
Dad | -0.205 |
Token | . |
Token ID | 13 |
Feature activation | +1.17e-01 |
Pos logit contributions | |
els | +0.355 |
cks | +0.339 |
ening | +0.321 |
ries | +0.320 |
cking | +0.316 |
Neg logit contributions | |
He | -0.270 |
He | -0.259 |
His | -0.255 |
he | -0.236 |
They | -0.234 |
Token |
|
Token ID | 198 |
Feature activation | +8.94e-02 |
Pos logit contributions | |
Sorry | +0.147 |
David | +0.147 |
Thanks | +0.145 |
Jack | +0.145 |
His | +0.144 |
Neg logit contributions | |
globe | -0.210 |
trunk | -0.208 |
chim | -0.207 |
bulb | -0.201 |
matches | -0.200 |
Token |
|
Token ID | 198 |
Feature activation | +9.74e-02 |
Pos logit contributions | |
Thanks | +0.106 |
His | +0.105 |
Jack | +0.104 |
Sorry | +0.103 |
David | +0.103 |
Neg logit contributions | |
ool | -0.167 |
globe | -0.167 |
bulb | -0.164 |
hole | -0.164 |
els | -0.162 |
Token | M |
Token ID | 44 |
Feature activation | +5.34e-02 |
Pos logit contributions | |
David | +0.122 |
Thanks | +0.116 |
Something | +0.116 |
Jack | +0.115 |
His | +0.115 |
Neg logit contributions | |
ool | -0.175 |
globe | -0.172 |
et | -0.169 |
bulb | -0.166 |
els | -0.166 |
Token | olly |
Token ID | 5098 |
Feature activation | -5.50e-03 |
Pos logit contributions | |
His | +0.088 |
Jack | +0.086 |
He | +0.085 |
Thanks | +0.081 |
Don | +0.079 |
Neg logit contributions | |
els | -0.091 |
ool | -0.088 |
bulb | -0.085 |
ps | -0.083 |
hole | -0.082 |
Token | delicious |
Token ID | 12625 |
Feature activation | -1.11e-01 |
Pos logit contributions | |
ps | +0.101 |
globe | +0.096 |
hole | +0.094 |
cks | +0.094 |
els | +0.088 |
Neg logit contributions | |
Jack | -0.085 |
He | -0.082 |
So | -0.076 |
She | -0.074 |
They | -0.073 |
Token | looking |
Token ID | 2045 |
Feature activation | -7.30e-02 |
Pos logit contributions | |
hole | +0.206 |
ps | +0.198 |
cks | +0.185 |
els | +0.183 |
atop | +0.177 |
Neg logit contributions | |
Jack | -0.162 |
He | -0.146 |
please | -0.145 |
She | -0.141 |
Jim | -0.140 |
Token | pastry |
Token ID | 41738 |
Feature activation | -1.65e-01 |
Pos logit contributions | |
hole | +0.122 |
els | +0.116 |
cks | +0.114 |
ps | +0.111 |
ening | +0.105 |
Neg logit contributions | |
Jack | -0.117 |
He | -0.115 |
please | -0.114 |
His | -0.111 |
He | -0.109 |
Token | with |
Token ID | 351 |
Feature activation | -4.32e-02 |
Pos logit contributions | |
ps | +0.270 |
els | +0.256 |
cks | +0.250 |
ening | +0.248 |
ool | +0.245 |
Neg logit contributions | |
Jack | -0.268 |
He | -0.258 |
His | -0.247 |
They | -0.245 |
She | -0.243 |
Token | bright |
Token ID | 6016 |
Feature activation | -1.22e-01 |
Pos logit contributions | |
hole | +0.080 |
globe | +0.078 |
ool | +0.078 |
ig | +0.075 |
els | +0.075 |
Neg logit contributions | |
Jack | -0.062 |
He | -0.061 |
His | -0.059 |
Thanks | -0.055 |
They | -0.055 |
Token | icing |
Token ID | 41567 |
Feature activation | -1.99e-01 |
Pos logit contributions | |
plank | +0.140 |
globe | +0.137 |
hole | +0.136 |
cks | +0.128 |
ool | +0.128 |
Neg logit contributions | |
He | -0.240 |
He | -0.233 |
She | -0.230 |
Jack | -0.230 |
His | -0.224 |
Token | and |
Token ID | 290 |
Feature activation | +2.54e-02 |
Pos logit contributions | |
bulb | +0.289 |
cks | +0.287 |
chim | +0.286 |
ps | +0.285 |
matches | +0.280 |
Neg logit contributions | |
He | -0.326 |
They | -0.317 |
Jack | -0.310 |
His | -0.297 |
She | -0.290 |
Token | lots |
Token ID | 6041 |
Feature activation | -8.28e-02 |
Pos logit contributions | |
Jack | +0.052 |
His | +0.051 |
He | +0.051 |
They | +0.048 |
Thanks | +0.047 |
Neg logit contributions | |
hole | -0.032 |
els | -0.032 |
globe | -0.030 |
ool | -0.030 |
bulb | -0.030 |
Token | of |
Token ID | 286 |
Feature activation | -3.13e-02 |
Pos logit contributions | |
globe | +0.196 |
hole | +0.193 |
cks | +0.191 |
plank | +0.190 |
ps | +0.190 |
Neg logit contributions | |
He | -0.083 |
Jack | -0.074 |
His | -0.070 |
They | -0.064 |
He | -0.063 |
Token | sprink |
Token ID | 27562 |
Feature activation | -1.90e-01 |
Pos logit contributions | |
hole | +0.045 |
ool | +0.042 |
globe | +0.042 |
ig | +0.041 |
els | +0.041 |
Neg logit contributions | |
He | -0.060 |
Jack | -0.058 |
His | -0.057 |
They | -0.054 |
Dad | -0.054 |
Token | les |
Token ID | 829 |
Feature activation | -1.16e-01 |
Pos logit contributions | |
trunk | +0.281 |
chim | +0.274 |
globe | +0.273 |
fence | +0.268 |
ladder | +0.267 |
Neg logit contributions | |
Jack | -0.353 |
He | -0.320 |
They | -0.316 |
His | -0.296 |
Jimmy | -0.293 |
Token | just |
Token ID | 655 |
Feature activation | +2.52e-02 |
Pos logit contributions | |
His | +0.035 |
Jack | +0.035 |
He | +0.034 |
Kids | +0.033 |
David | +0.032 |
Neg logit contributions | |
els | -0.052 |
hole | -0.052 |
bulb | -0.052 |
et | -0.050 |
globe | -0.050 |
Token | couldn |
Token ID | 3521 |
Feature activation | +2.13e-02 |
Pos logit contributions | |
Jack | +0.020 |
His | +0.019 |
He | +0.018 |
His | +0.014 |
Thanks | +0.014 |
Neg logit contributions | |
els | -0.064 |
ool | -0.063 |
globe | -0.062 |
ps | -0.061 |
bulb | -0.061 |
Token | 't |
Token ID | 470 |
Feature activation | +2.45e-02 |
Pos logit contributions | |
Jack | +0.031 |
His | +0.030 |
Kids | +0.029 |
Thanks | +0.029 |
David | +0.029 |
Neg logit contributions | |
els | -0.036 |
bulb | -0.036 |
hole | -0.036 |
ool | -0.036 |
globe | -0.036 |
Token | find |
Token ID | 1064 |
Feature activation | -3.37e-02 |
Pos logit contributions | |
Jack | +0.041 |
His | +0.041 |
He | +0.040 |
Kids | +0.039 |
Thanks | +0.038 |
Neg logit contributions | |
els | -0.038 |
hole | -0.038 |
bulb | -0.038 |
globe | -0.037 |
ps | -0.036 |
Token | the |
Token ID | 262 |
Feature activation | -3.86e-02 |
Pos logit contributions | |
ool | +0.073 |
globe | +0.070 |
bulb | +0.069 |
ening | +0.069 |
chim | +0.068 |
Neg logit contributions | |
His | -0.038 |
He | -0.036 |
His | -0.035 |
He | -0.035 |
him | -0.034 |
Token | steel |
Token ID | 7771 |
Feature activation | -1.98e-01 |
Pos logit contributions | |
els | +0.068 |
cks | +0.065 |
ries | +0.060 |
ps | +0.060 |
aro | +0.059 |
Neg logit contributions | |
He | -0.062 |
He | -0.061 |
Jack | -0.059 |
Something | -0.059 |
Sure | -0.057 |
Token | box |
Token ID | 3091 |
Feature activation | -1.63e-01 |
Pos logit contributions | |
ening | +0.277 |
cks | +0.276 |
cking | +0.276 |
els | +0.275 |
eking | +0.265 |
Neg logit contributions | |
He | -0.380 |
He | -0.380 |
His | -0.361 |
Jack | -0.342 |
His | -0.342 |
Token | anywhere |
Token ID | 6609 |
Feature activation | -2.53e-02 |
Pos logit contributions | |
els | +0.274 |
blades | +0.261 |
ps | +0.255 |
ening | +0.249 |
cks | +0.245 |
Neg logit contributions | |
He | -0.283 |
His | -0.273 |
He | -0.261 |
His | -0.249 |
Jack | -0.247 |
Token | . |
Token ID | 13 |
Feature activation | +1.20e-01 |
Pos logit contributions | |
els | +0.042 |
ps | +0.038 |
ening | +0.038 |
bulb | +0.038 |
ool | +0.037 |
Neg logit contributions | |
He | -0.046 |
His | -0.046 |
Jack | -0.045 |
He | -0.041 |
His | -0.041 |
Token | |
Token ID | 220 |
Feature activation | +4.21e-02 |
Pos logit contributions | |
Jack | +0.142 |
His | +0.139 |
Kids | +0.138 |
David | +0.137 |
Sorry | +0.132 |
Neg logit contributions | |
trunk | -0.230 |
hole | -0.230 |
pieces | -0.222 |
globe | -0.220 |
chim | -0.217 |
Token |
|
Token ID | 198 |
Feature activation | +9.34e-02 |
Pos logit contributions | |
Jack | +0.043 |
His | +0.042 |
Kids | +0.039 |
He | +0.038 |
Thanks | +0.038 |
Neg logit contributions | |
hole | -0.091 |
els | -0.091 |
globe | -0.088 |
bulb | -0.088 |
ool | -0.087 |
Token | treasure |
Token ID | 14068 |
Feature activation | -1.52e-01 |
Pos logit contributions | |
els | +0.013 |
ps | +0.012 |
hole | +0.011 |
ening | +0.011 |
et | +0.011 |
Neg logit contributions | |
Jack | -0.015 |
His | -0.015 |
He | -0.015 |
Thanks | -0.014 |
Dad | -0.014 |
Token | chest |
Token ID | 7721 |
Feature activation | -1.34e-01 |
Pos logit contributions | |
els | +0.225 |
ening | +0.209 |
ps | +0.207 |
blades | +0.200 |
angles | +0.195 |
Neg logit contributions | |
He | -0.286 |
Jack | -0.277 |
His | -0.273 |
He | -0.261 |
They | -0.249 |
Token | full |
Token ID | 1336 |
Feature activation | -8.27e-02 |
Pos logit contributions | |
els | +0.263 |
ps | +0.242 |
blades | +0.236 |
ening | +0.226 |
bulbs | +0.225 |
Neg logit contributions | |
He | -0.203 |
His | -0.198 |
Jack | -0.197 |
He | -0.178 |
They | -0.175 |
Token | of |
Token ID | 286 |
Feature activation | -3.59e-02 |
Pos logit contributions | |
els | +0.189 |
ps | +0.186 |
cks | +0.180 |
blades | +0.178 |
ening | +0.177 |
Neg logit contributions | |
He | -0.088 |
Jack | -0.087 |
His | -0.086 |
His | -0.077 |
He | -0.077 |
Token | shiny |
Token ID | 22441 |
Feature activation | -1.44e-01 |
Pos logit contributions | |
els | +0.047 |
ool | +0.047 |
et | +0.047 |
ps | +0.046 |
hole | +0.046 |
Neg logit contributions | |
He | -0.070 |
Jack | -0.068 |
His | -0.067 |
He | -0.064 |
Something | -0.063 |
Token | gold |
Token ID | 3869 |
Feature activation | -1.98e-01 |
Pos logit contributions | |
cks | +0.186 |
ps | +0.183 |
aro | +0.177 |
cking | +0.173 |
ool | +0.173 |
Neg logit contributions | |
He | -0.262 |
He | -0.257 |
His | -0.238 |
Jack | -0.234 |
His | -0.227 |
Token | coins |
Token ID | 10796 |
Feature activation | -1.91e-01 |
Pos logit contributions | |
cks | +0.325 |
angles | +0.321 |
cking | +0.320 |
aro | +0.314 |
ening | +0.314 |
Neg logit contributions | |
He | -0.291 |
He | -0.277 |
he | -0.248 |
They | -0.241 |
His | -0.228 |
Token | ! |
Token ID | 0 |
Feature activation | +1.05e-01 |
Pos logit contributions | |
els | +0.350 |
bulb | +0.345 |
blades | +0.340 |
ps | +0.334 |
ool | +0.323 |
Neg logit contributions | |
He | -0.256 |
he | -0.234 |
They | -0.232 |
He | -0.223 |
Jack | -0.221 |
Token | Tim |
Token ID | 5045 |
Feature activation | +1.28e-03 |
Pos logit contributions | |
His | +0.134 |
Brother | +0.126 |
Jack | +0.126 |
He | +0.124 |
Her | +0.122 |
Neg logit contributions | |
ool | -0.191 |
trunk | -0.190 |
chim | -0.190 |
ladder | -0.188 |
pieces | -0.187 |
Token | , |
Token ID | 11 |
Feature activation | +5.90e-02 |
Pos logit contributions | |
Jack | +0.002 |
His | +0.002 |
He | +0.002 |
David | +0.002 |
Thanks | +0.002 |
Neg logit contributions | |
els | -0.002 |
bulb | -0.002 |
ool | -0.002 |
hole | -0.002 |
ening | -0.002 |
Token | Sally |
Token ID | 25737 |
Feature activation | -4.23e-03 |
Pos logit contributions | |
His | +0.063 |
Jack | +0.063 |
He | +0.060 |
Kids | +0.053 |
David | +0.053 |
Neg logit contributions | |
els | -0.131 |
hole | -0.128 |
bulb | -0.128 |
ps | -0.126 |
globe | -0.126 |
Token | red |
Token ID | 2266 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
els | +0.054 |
ening | +0.053 |
ps | +0.052 |
hole | +0.051 |
cks | +0.051 |
Neg logit contributions | |
His | -0.056 |
His | -0.052 |
He | -0.052 |
Jack | -0.050 |
He | -0.047 |
Token | cap |
Token ID | 1451 |
Feature activation | -1.74e-01 |
Pos logit contributions | |
els | +0.214 |
ps | +0.214 |
ening | +0.214 |
hole | +0.211 |
cks | +0.209 |
Neg logit contributions | |
His | -0.238 |
He | -0.224 |
Jack | -0.213 |
His | -0.209 |
Sure | -0.203 |
Token | with |
Token ID | 351 |
Feature activation | -4.32e-02 |
Pos logit contributions | |
els | +0.298 |
cks | +0.290 |
angles | +0.280 |
ening | +0.275 |
ps | +0.273 |
Neg logit contributions | |
His | -0.269 |
He | -0.267 |
Jack | -0.248 |
enny | -0.237 |
They | -0.236 |
Token | a |
Token ID | 257 |
Feature activation | -1.82e-02 |
Pos logit contributions | |
hole | +0.085 |
ool | +0.085 |
els | +0.085 |
globe | +0.084 |
bulb | +0.081 |
Neg logit contributions | |
His | -0.057 |
He | -0.056 |
Jack | -0.054 |
His | -0.050 |
Thanks | -0.050 |
Token | shiny |
Token ID | 22441 |
Feature activation | -1.36e-01 |
Pos logit contributions | |
els | +0.029 |
ps | +0.026 |
et | +0.026 |
cks | +0.025 |
ries | +0.024 |
Neg logit contributions | |
He | -0.036 |
His | -0.036 |
Jack | -0.032 |
Something | -0.031 |
His | -0.031 |
Token | gold |
Token ID | 3869 |
Feature activation | -1.98e-01 |
Pos logit contributions | |
els | +0.178 |
cks | +0.176 |
ps | +0.170 |
hole | +0.159 |
atop | +0.158 |
Neg logit contributions | |
His | -0.274 |
He | -0.264 |
His | -0.247 |
He | -0.241 |
Jack | -0.237 |
Token | star |
Token ID | 3491 |
Feature activation | -1.77e-01 |
Pos logit contributions | |
cks | +0.300 |
els | +0.276 |
angles | +0.272 |
cking | +0.265 |
ening | +0.259 |
Neg logit contributions | |
He | -0.354 |
His | -0.337 |
He | -0.314 |
His | -0.289 |
he | -0.273 |
Token | on |
Token ID | 319 |
Feature activation | -5.90e-02 |
Pos logit contributions | |
els | +0.262 |
angles | +0.255 |
ps | +0.255 |
cks | +0.250 |
ool | +0.246 |
Neg logit contributions | |
He | -0.273 |
His | -0.258 |
He | -0.233 |
They | -0.228 |
he | -0.225 |
Token | it |
Token ID | 340 |
Feature activation | -8.78e-02 |
Pos logit contributions | |
hole | +0.122 |
els | +0.120 |
ool | +0.118 |
ps | +0.116 |
atop | +0.115 |
Neg logit contributions | |
His | -0.076 |
He | -0.071 |
Jack | -0.066 |
His | -0.064 |
He | -0.063 |
Token | . |
Token ID | 13 |
Feature activation | +1.22e-01 |
Pos logit contributions | |
els | +0.161 |
globe | +0.153 |
blades | +0.151 |
bulb | +0.149 |
ps | +0.148 |
Neg logit contributions | |
He | -0.126 |
His | -0.125 |
He | -0.123 |
His | -0.120 |
She | -0.113 |
Token |
|
Token ID | 198 |
Feature activation | +9.70e-02 |
Pos logit contributions | |
David | +0.203 |
Jim | +0.203 |
Tony | +0.200 |
His | +0.197 |
Harry | +0.191 |
Neg logit contributions | |
chim | -0.174 |
v | -0.170 |
bulb | -0.164 |
trunk | -0.164 |
model | -0.162 |
Token | hit |
Token ID | 2277 |
Feature activation | -9.34e-02 |
Pos logit contributions | |
els | +0.009 |
globe | +0.009 |
bulb | +0.009 |
ening | +0.009 |
hole | +0.009 |
Neg logit contributions | |
Jack | -0.005 |
He | -0.005 |
His | -0.005 |
Sure | -0.004 |
Thanks | -0.004 |
Token | it |
Token ID | 340 |
Feature activation | -8.12e-02 |
Pos logit contributions | |
ool | +0.206 |
globe | +0.204 |
chim | +0.193 |
hole | +0.193 |
bulb | +0.192 |
Neg logit contributions | |
Jack | -0.103 |
He | -0.096 |
His | -0.095 |
They | -0.090 |
Sam | -0.086 |
Token | with |
Token ID | 351 |
Feature activation | -3.51e-02 |
Pos logit contributions | |
els | +0.130 |
globe | +0.123 |
egg | +0.114 |
ening | +0.113 |
bulbs | +0.113 |
Neg logit contributions | |
He | -0.149 |
He | -0.141 |
His | -0.138 |
They | -0.135 |
please | -0.132 |
Token | your |
Token ID | 534 |
Feature activation | -3.81e-02 |
Pos logit contributions | |
globe | +0.080 |
hole | +0.074 |
ool | +0.073 |
atop | +0.072 |
els | +0.072 |
Neg logit contributions | |
Jack | -0.041 |
His | -0.040 |
He | -0.039 |
Thanks | -0.037 |
please | -0.035 |
Token | tall |
Token ID | 7331 |
Feature activation | -1.31e-01 |
Pos logit contributions | |
els | +0.056 |
bulb | +0.054 |
hole | +0.053 |
globe | +0.053 |
ening | +0.053 |
Neg logit contributions | |
His | -0.068 |
He | -0.067 |
Jack | -0.067 |
Thanks | -0.063 |
His | -0.063 |
Token | golf |
Token ID | 13126 |
Feature activation | -1.98e-01 |
Pos logit contributions | |
bulb | +0.193 |
globe | +0.191 |
els | +0.189 |
hole | +0.180 |
atop | +0.180 |
Neg logit contributions | |
His | -0.234 |
He | -0.230 |
His | -0.220 |
Jack | -0.217 |
He | -0.214 |
Token | club |
Token ID | 3430 |
Feature activation | -1.77e-01 |
Pos logit contributions | |
ening | +0.278 |
els | +0.276 |
bulb | +0.266 |
hole | +0.258 |
chim | +0.258 |
Neg logit contributions | |
He | -0.375 |
His | -0.372 |
Jack | -0.363 |
They | -0.339 |
Thanks | -0.338 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.11e-01 |
Pos logit contributions | |
els | +0.265 |
ening | +0.251 |
atop | +0.241 |
globe | +0.239 |
bulb | +0.239 |
Neg logit contributions | |
He | -0.311 |
His | -0.298 |
He | -0.295 |
please | -0.291 |
So | -0.285 |
Token |
|
Token ID | 198 |
Feature activation | +9.42e-02 |
Pos logit contributions | |
His | +0.149 |
Un | +0.144 |
He | +0.143 |
Kids | +0.140 |
David | +0.138 |
Neg logit contributions | |
bulb | -0.193 |
hole | -0.190 |
els | -0.189 |
globe | -0.189 |
ening | -0.185 |
Token |
|
Token ID | 198 |
Feature activation | +1.02e-01 |
Pos logit contributions | |
Sorry | +0.118 |
Thanks | +0.116 |
His | +0.115 |
He | +0.113 |
Un | +0.112 |
Neg logit contributions | |
bulb | -0.160 |
globe | -0.157 |
els | -0.157 |
ool | -0.153 |
pieces | -0.149 |
Token | That |
Token ID | 2504 |
Feature activation | +8.56e-03 |
Pos logit contributions | |
Deal | +0.081 |
He | +0.078 |
His | +0.078 |
Um | +0.074 |
Something | +0.072 |
Neg logit contributions | |
bulb | -0.209 |
object | -0.195 |
wire | -0.193 |
pieces | -0.192 |
els | -0.191 |
Token | them |
Token ID | 606 |
Feature activation | -5.22e-02 |
Pos logit contributions | |
els | +0.016 |
bulb | +0.016 |
ps | +0.015 |
globe | +0.015 |
hole | +0.015 |
Neg logit contributions | |
His | -0.014 |
Jack | -0.014 |
He | -0.014 |
Thanks | -0.013 |
David | -0.013 |
Token | for |
Token ID | 329 |
Feature activation | -6.39e-03 |
Pos logit contributions | |
els | +0.073 |
bulb | +0.069 |
ool | +0.069 |
blades | +0.069 |
hole | +0.069 |
Neg logit contributions | |
His | -0.104 |
He | -0.102 |
Jack | -0.097 |
They | -0.095 |
His | -0.094 |
Token | showing |
Token ID | 4478 |
Feature activation | -6.75e-02 |
Pos logit contributions | |
els | +0.011 |
hole | +0.011 |
ool | +0.011 |
bulb | +0.011 |
globe | +0.011 |
Neg logit contributions | |
His | -0.010 |
Jack | -0.010 |
He | -0.010 |
Thanks | -0.009 |
David | -0.009 |
Token | them |
Token ID | 606 |
Feature activation | -7.48e-02 |
Pos logit contributions | |
els | +0.146 |
ps | +0.142 |
bulb | +0.140 |
ening | +0.139 |
hole | +0.139 |
Neg logit contributions | |
Jack | -0.076 |
His | -0.075 |
He | -0.072 |
Kids | -0.065 |
David | -0.065 |
Token | the |
Token ID | 262 |
Feature activation | -4.52e-02 |
Pos logit contributions | |
els | +0.126 |
hole | +0.124 |
bulb | +0.123 |
ool | +0.121 |
globe | +0.120 |
Neg logit contributions | |
His | -0.119 |
Jack | -0.118 |
He | -0.116 |
Kids | -0.105 |
They | -0.105 |
Token | x |
Token ID | 2124 |
Feature activation | -1.98e-01 |
Pos logit contributions | |
els | +0.083 |
cks | +0.075 |
ries | +0.072 |
ening | +0.070 |
ls | +0.068 |
Neg logit contributions | |
He | -0.076 |
His | -0.070 |
Something | -0.068 |
His | -0.067 |
He | -0.067 |
Token | - |
Token ID | 12 |
Feature activation | -8.95e-02 |
Pos logit contributions | |
blo | +0.328 |
atop | +0.318 |
angles | +0.309 |
blades | +0.309 |
ening | +0.306 |
Neg logit contributions | |
His | -0.324 |
He | -0.293 |
They | -0.257 |
His | -0.254 |
Her | -0.251 |
Token | ray |
Token ID | 2433 |
Feature activation | -1.38e-01 |
Pos logit contributions | |
ening | +0.143 |
els | +0.143 |
angles | +0.138 |
hole | +0.138 |
ps | +0.138 |
Neg logit contributions | |
His | -0.155 |
He | -0.152 |
Sure | -0.146 |
Un | -0.144 |
Jack | -0.143 |
Token | . |
Token ID | 13 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
els | +0.227 |
cks | +0.226 |
ps | +0.217 |
bulbs | +0.213 |
stem | +0.211 |
Neg logit contributions | |
He | -0.231 |
His | -0.221 |
He | -0.207 |
They | -0.201 |
His | -0.196 |
Token | They |
Token ID | 1119 |
Feature activation | +4.02e-02 |
Pos logit contributions | |
Jack | +0.146 |
His | +0.125 |
David | +0.119 |
Jim | +0.117 |
Sorry | +0.116 |
Neg logit contributions | |
chim | -0.226 |
els | -0.224 |
trunk | -0.222 |
hole | -0.221 |
ps | -0.221 |
Token | left |
Token ID | 1364 |
Feature activation | -3.58e-02 |
Pos logit contributions | |
Jack | +0.060 |
His | +0.059 |
He | +0.056 |
Kids | +0.054 |
David | +0.053 |
Neg logit contributions | |
els | -0.068 |
hole | -0.067 |
bulb | -0.067 |
globe | -0.066 |
ps | -0.066 |
Token | red |
Token ID | 2266 |
Feature activation | -1.20e-01 |
Pos logit contributions | |
els | +0.106 |
ps | +0.104 |
atop | +0.095 |
cks | +0.095 |
globe | +0.092 |
Neg logit contributions | |
His | -0.132 |
He | -0.130 |
Jack | -0.128 |
His | -0.120 |
Sure | -0.118 |
Token | cap |
Token ID | 1451 |
Feature activation | -1.71e-01 |
Pos logit contributions | |
els | +0.179 |
ps | +0.178 |
cks | +0.174 |
ening | +0.174 |
hole | +0.171 |
Neg logit contributions | |
His | -0.215 |
He | -0.204 |
Jack | -0.192 |
His | -0.191 |
Sure | -0.187 |
Token | with |
Token ID | 351 |
Feature activation | -4.86e-02 |
Pos logit contributions | |
els | +0.290 |
cks | +0.282 |
blades | +0.278 |
ps | +0.265 |
angles | +0.265 |
Neg logit contributions | |
His | -0.285 |
He | -0.278 |
Jack | -0.269 |
enny | -0.253 |
His | -0.253 |
Token | a |
Token ID | 257 |
Feature activation | -2.00e-02 |
Pos logit contributions | |
ool | +0.101 |
hole | +0.100 |
globe | +0.100 |
els | +0.098 |
bulb | +0.098 |
Neg logit contributions | |
His | -0.060 |
He | -0.058 |
Jack | -0.056 |
His | -0.052 |
Thanks | -0.050 |
Token | shiny |
Token ID | 22441 |
Feature activation | -1.36e-01 |
Pos logit contributions | |
els | +0.031 |
ps | +0.029 |
cks | +0.027 |
ries | +0.026 |
et | +0.025 |
Neg logit contributions | |
His | -0.041 |
He | -0.041 |
Jack | -0.037 |
His | -0.036 |
Something | -0.036 |
Token | gold |
Token ID | 3869 |
Feature activation | -1.98e-01 |
Pos logit contributions | |
cks | +0.174 |
els | +0.169 |
ps | +0.168 |
hole | +0.149 |
atop | +0.149 |
Neg logit contributions | |
His | -0.286 |
He | -0.270 |
His | -0.260 |
He | -0.252 |
please | -0.245 |
Token | star |
Token ID | 3491 |
Feature activation | -1.77e-01 |
Pos logit contributions | |
cks | +0.311 |
els | +0.277 |
ps | +0.276 |
cking | +0.269 |
angles | +0.269 |
Neg logit contributions | |
He | -0.355 |
His | -0.345 |
He | -0.320 |
please | -0.308 |
His | -0.298 |
Token | on |
Token ID | 319 |
Feature activation | -6.16e-02 |
Pos logit contributions | |
ps | +0.279 |
els | +0.276 |
cks | +0.271 |
blades | +0.263 |
angles | +0.254 |
Neg logit contributions | |
He | -0.272 |
His | -0.267 |
he | -0.251 |
He | -0.242 |
please | -0.235 |
Token | it |
Token ID | 340 |
Feature activation | -8.69e-02 |
Pos logit contributions | |
hole | +0.122 |
els | +0.120 |
ool | +0.119 |
ps | +0.116 |
bulb | +0.116 |
Neg logit contributions | |
His | -0.085 |
He | -0.080 |
Jack | -0.076 |
His | -0.071 |
He | -0.070 |
Token | , |
Token ID | 11 |
Feature activation | +2.45e-02 |
Pos logit contributions | |
els | +0.144 |
ps | +0.137 |
globe | +0.136 |
blades | +0.134 |
bulb | +0.132 |
Neg logit contributions | |
His | -0.144 |
He | -0.144 |
He | -0.136 |
His | -0.135 |
Jack | -0.135 |
Token | and |
Token ID | 290 |
Feature activation | +4.90e-02 |
Pos logit contributions | |
Jack | +0.039 |
His | +0.039 |
He | +0.038 |
Kids | +0.036 |
David | +0.035 |
Neg logit contributions | |
els | -0.041 |
hole | -0.040 |
bulb | -0.040 |
globe | -0.039 |
ening | -0.039 |
Token | ?" |
Token ID | 1701 |
Feature activation | +1.16e-01 |
Pos logit contributions | |
hole | +0.164 |
bulb | +0.163 |
ps | +0.162 |
els | +0.162 |
ool | +0.162 |
Neg logit contributions | |
Jack | -0.167 |
They | -0.158 |
His | -0.158 |
He | -0.157 |
Kids | -0.147 |
Token |
|
Token ID | 198 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
Jack | +0.171 |
Thanks | +0.165 |
His | +0.159 |
David | +0.156 |
Johnny | +0.154 |
Neg logit contributions | |
els | -0.176 |
ladder | -0.170 |
hanging | -0.169 |
bulb | -0.167 |
on | -0.167 |
Token |
|
Token ID | 198 |
Feature activation | +1.10e-01 |
Pos logit contributions | |
Jack | +0.135 |
His | +0.126 |
Thanks | +0.125 |
David | +0.125 |
Sorry | +0.116 |
Neg logit contributions | |
els | -0.192 |
bulb | -0.182 |
pieces | -0.179 |
globe | -0.179 |
ps | -0.177 |
Token | S |
Token ID | 50 |
Feature activation | +5.69e-03 |
Pos logit contributions | |
Ryan | +0.155 |
Jack | +0.155 |
His | +0.153 |
Something | +0.152 |
enny | +0.150 |
Neg logit contributions | |
els | -0.167 |
pieces | -0.155 |
egg | -0.155 |
bulb | -0.154 |
globe | -0.153 |
Token | ara |
Token ID | 3301 |
Feature activation | +1.60e-02 |
Pos logit contributions | |
His | +0.006 |
Jack | +0.006 |
He | +0.006 |
Kids | +0.006 |
David | +0.006 |
Neg logit contributions | |
els | -0.012 |
bulb | -0.012 |
hole | -0.012 |
globe | -0.012 |
ps | -0.012 |
Token | and |
Token ID | 290 |
Feature activation | +7.29e-02 |
Pos logit contributions | |
His | +0.033 |
Jack | +0.032 |
He | +0.031 |
Thanks | +0.030 |
David | +0.030 |
Neg logit contributions | |
els | -0.019 |
hole | -0.018 |
bulb | -0.018 |
globe | -0.018 |
ps | -0.018 |
Token | Lily |
Token ID | 20037 |
Feature activation | +1.35e-02 |
Pos logit contributions | |
Jack | +0.102 |
He | +0.099 |
His | +0.096 |
David | +0.092 |
Thanks | +0.088 |
Neg logit contributions | |
els | -0.142 |
bulb | -0.137 |
ps | -0.135 |
et | -0.133 |
egg | -0.130 |
Token | nodded |
Token ID | 14464 |
Feature activation | -2.46e-02 |
Pos logit contributions | |
His | +0.020 |
Jack | +0.019 |
He | +0.018 |
Kids | +0.017 |
David | +0.017 |
Neg logit contributions | |
els | -0.024 |
hole | -0.024 |
bulb | -0.024 |
ps | -0.023 |
globe | -0.023 |
Token | and |
Token ID | 290 |
Feature activation | +5.87e-02 |
Pos logit contributions | |
els | +0.028 |
bulb | +0.027 |
ps | +0.026 |
ening | +0.026 |
hole | +0.026 |
Neg logit contributions | |
Jack | -0.053 |
His | -0.051 |
He | -0.050 |
David | -0.049 |
Kids | -0.049 |
Token | smiled |
Token ID | 13541 |
Feature activation | -3.38e-02 |
Pos logit contributions | |
His | +0.098 |
He | +0.098 |
Jack | +0.097 |
David | +0.090 |
Kids | +0.089 |
Neg logit contributions | |
bulb | -0.088 |
pieces | -0.088 |
els | -0.087 |
ps | -0.086 |
et | -0.085 |
Token | . |
Token ID | 13 |
Feature activation | +1.09e-01 |
Pos logit contributions | |
els | +0.025 |
bulb | +0.025 |
hole | +0.025 |
ening | +0.024 |
ps | +0.024 |
Neg logit contributions | |
Jack | -0.085 |
His | -0.082 |
Kids | -0.081 |
David | -0.080 |
He | -0.079 |
Token | and |
Token ID | 290 |
Feature activation | +4.09e-02 |
Pos logit contributions | |
ool | +0.021 |
ening | +0.020 |
atop | +0.019 |
ating | +0.019 |
els | +0.019 |
Neg logit contributions | |
His | -0.026 |
He | -0.024 |
His | -0.023 |
Her | -0.023 |
He | -0.020 |
Token | pulls |
Token ID | 16194 |
Feature activation | -6.72e-02 |
Pos logit contributions | |
Jack | +0.085 |
His | +0.084 |
He | +0.082 |
David | +0.079 |
Kids | +0.078 |
Neg logit contributions | |
els | -0.051 |
bulb | -0.048 |
ps | -0.048 |
hole | -0.047 |
globe | -0.046 |
Token | her |
Token ID | 607 |
Feature activation | -2.10e-02 |
Pos logit contributions | |
globe | +0.134 |
els | +0.133 |
bulb | +0.133 |
hole | +0.132 |
ool | +0.132 |
Neg logit contributions | |
His | -0.089 |
He | -0.088 |
Jack | -0.087 |
Dad | -0.076 |
They | -0.076 |
Token | back |
Token ID | 736 |
Feature activation | -5.29e-02 |
Pos logit contributions | |
ool | +0.031 |
ening | +0.031 |
els | +0.030 |
atop | +0.029 |
chim | +0.028 |
Neg logit contributions | |
His | -0.038 |
He | -0.037 |
His | -0.034 |
Her | -0.034 |
He | -0.033 |
Token | . |
Token ID | 13 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
globe | +0.092 |
ening | +0.090 |
blades | +0.089 |
atop | +0.087 |
chim | +0.086 |
Neg logit contributions | |
His | -0.079 |
He | -0.079 |
They | -0.074 |
Her | -0.074 |
He | -0.073 |
Token |
|
Token ID | 198 |
Feature activation | +9.45e-02 |
Pos logit contributions | |
Jack | +0.115 |
David | +0.109 |
Kids | +0.101 |
Harry | +0.098 |
Thanks | +0.098 |
Neg logit contributions | |
ladder | -0.210 |
bulb | -0.206 |
trunk | -0.205 |
chim | -0.204 |
pieces | -0.202 |
Token |
|
Token ID | 198 |
Feature activation | +1.04e-01 |
Pos logit contributions | |
Jack | +0.118 |
Thanks | +0.116 |
David | +0.113 |
His | +0.111 |
Sorry | +0.109 |
Neg logit contributions | |
els | -0.169 |
hole | -0.162 |
bulb | -0.162 |
ool | -0.161 |
ps | -0.160 |
Token | " |
Token ID | 1 |
Feature activation | +6.62e-02 |
Pos logit contributions | |
David | +0.192 |
Something | +0.190 |
Jack | +0.184 |
ete | +0.180 |
His | +0.178 |
Neg logit contributions | |
els | -0.109 |
pieces | -0.109 |
bulb | -0.100 |
ps | -0.100 |
ladder | -0.098 |
Token | Stop |
Token ID | 19485 |
Feature activation | -4.75e-05 |
Pos logit contributions | |
David | +0.070 |
enny | +0.068 |
ete | +0.068 |
His | +0.067 |
His | +0.063 |
Neg logit contributions | |
pieces | -0.135 |
blades | -0.131 |
ps | -0.130 |
bulb | -0.130 |
globe | -0.128 |
Token | it |
Token ID | 340 |
Feature activation | -4.61e-02 |
Pos logit contributions | |
ool | +0.000 |
els | +0.000 |
hole | +0.000 |
atop | +0.000 |
bulbs | +0.000 |
Neg logit contributions | |
Jack | -0.000 |
His | -0.000 |
He | -0.000 |
Jim | -0.000 |
please | -0.000 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.08e-01 |
Pos logit contributions | |
els | +0.072 |
globe | +0.071 |
hole | +0.069 |
ening | +0.068 |
bulb | +0.066 |
Neg logit contributions | |
Jack | -0.080 |
He | -0.078 |
His | -0.077 |
They | -0.072 |
please | -0.072 |
Token | , |
Token ID | 11 |
Feature activation | +5.44e-02 |
Pos logit contributions | |
els | +0.016 |
hole | +0.015 |
bulb | +0.015 |
globe | +0.015 |
ening | +0.015 |
Neg logit contributions | |
His | -0.023 |
Jack | -0.023 |
He | -0.023 |
Kids | -0.022 |
David | -0.021 |
Token | Lily |
Token ID | 20037 |
Feature activation | -6.96e-03 |
Pos logit contributions | |
His | +0.076 |
Jack | +0.074 |
He | +0.072 |
Kids | +0.067 |
David | +0.066 |
Neg logit contributions | |
bulb | -0.101 |
els | -0.101 |
hole | -0.099 |
globe | -0.099 |
ps | -0.097 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.15e-01 |
Pos logit contributions | |
els | +0.010 |
ool | +0.009 |
globe | +0.009 |
bulb | +0.009 |
blades | +0.009 |
Neg logit contributions | |
He | -0.013 |
Jack | -0.013 |
His | -0.013 |
They | -0.012 |
He | -0.012 |
Token | Tom |
Token ID | 4186 |
Feature activation | +9.25e-03 |
Pos logit contributions | |
Thanks | +0.146 |
Jack | +0.146 |
His | +0.144 |
He | +0.142 |
Kids | +0.140 |
Neg logit contributions | |
els | -0.207 |
hole | -0.202 |
pieces | -0.201 |
bulb | -0.201 |
trunk | -0.197 |
Token | shouted |
Token ID | 17293 |
Feature activation | -3.65e-03 |
Pos logit contributions | |
Jack | +0.014 |
His | +0.014 |
He | +0.014 |
Thanks | +0.013 |
Kids | +0.013 |
Neg logit contributions | |
els | -0.016 |
hole | -0.015 |
ps | -0.015 |
bulb | -0.015 |
ening | -0.015 |
Token | . |
Token ID | 13 |
Feature activation | +1.00e-01 |
Pos logit contributions | |
els | +0.004 |
hole | +0.004 |
globe | +0.004 |
bulb | +0.004 |
ool | +0.004 |
Neg logit contributions | |
His | -0.008 |
Jack | -0.008 |
He | -0.008 |
His | -0.007 |
Her | -0.007 |
Token |
|
Token ID | 198 |
Feature activation | +9.63e-02 |
Pos logit contributions | |
Jack | +0.133 |
David | +0.121 |
Thanks | +0.119 |
Kids | +0.119 |
His | +0.119 |
Neg logit contributions | |
bulb | -0.185 |
globe | -0.185 |
els | -0.184 |
pieces | -0.183 |
hole | -0.182 |
Token |
|
Token ID | 198 |
Feature activation | +1.05e-01 |
Pos logit contributions | |
Thanks | +0.128 |
Jack | +0.122 |
David | +0.120 |
Sorry | +0.117 |
His | +0.114 |
Neg logit contributions | |
els | -0.157 |
pieces | -0.154 |
bulb | -0.153 |
hole | -0.153 |
ool | -0.153 |
Token | " |
Token ID | 1 |
Feature activation | +7.18e-02 |
Pos logit contributions | |
David | +0.135 |
His | +0.134 |
Jack | +0.131 |
Something | +0.131 |
oph | +0.129 |
Neg logit contributions | |
pieces | -0.162 |
els | -0.153 |
bulb | -0.153 |
hole | -0.152 |
wire | -0.147 |
Token | No |
Token ID | 2949 |
Feature activation | +1.51e-03 |
Pos logit contributions | |
enny | +0.074 |
His | +0.070 |
ete | +0.069 |
David | +0.066 |
Jack | +0.064 |
Neg logit contributions | |
object | -0.139 |
globe | -0.137 |
bulb | -0.137 |
blades | -0.136 |
pieces | -0.134 |
Token | , |
Token ID | 11 |
Feature activation | +4.87e-02 |
Pos logit contributions | |
Jack | +0.003 |
His | +0.003 |
He | +0.003 |
Kids | +0.003 |
David | +0.003 |
Neg logit contributions | |
els | -0.002 |
hole | -0.002 |
atop | -0.002 |
globe | -0.002 |
bulb | -0.002 |
Token | . |
Token ID | 13 |
Feature activation | +9.40e-02 |
Pos logit contributions | |
hole | +0.192 |
globe | +0.192 |
ool | +0.192 |
bulb | +0.190 |
atop | +0.186 |
Neg logit contributions | |
He | -0.121 |
Jack | -0.116 |
His | -0.115 |
They | -0.108 |
Dad | -0.099 |
Token | He |
Token ID | 679 |
Feature activation | +2.22e-02 |
Pos logit contributions | |
Jack | +0.098 |
David | +0.082 |
His | +0.080 |
Thanks | +0.076 |
Kids | +0.076 |
Neg logit contributions | |
hole | -0.200 |
globe | -0.199 |
trunk | -0.199 |
Tower | -0.197 |
els | -0.196 |
Token | is |
Token ID | 318 |
Feature activation | +2.90e-02 |
Pos logit contributions | |
He | +0.040 |
His | +0.039 |
Jack | +0.038 |
They | +0.036 |
His | +0.036 |
Neg logit contributions | |
els | -0.033 |
hole | -0.032 |
bulb | -0.031 |
atop | -0.031 |
globe | -0.031 |
Token | very |
Token ID | 845 |
Feature activation | +2.42e-02 |
Pos logit contributions | |
His | +0.041 |
He | +0.041 |
Jack | +0.039 |
Thanks | +0.036 |
His | +0.035 |
Neg logit contributions | |
els | -0.056 |
hole | -0.056 |
ool | -0.054 |
globe | -0.054 |
bulb | -0.053 |
Token | helpful |
Token ID | 7613 |
Feature activation | -9.05e-02 |
Pos logit contributions | |
His | +0.038 |
He | +0.038 |
Jack | +0.037 |
Thanks | +0.034 |
His | +0.033 |
Neg logit contributions | |
hole | -0.044 |
els | -0.044 |
globe | -0.041 |
atop | -0.041 |
bulb | -0.041 |
Token | ," |
Token ID | 553 |
Feature activation | +1.05e-01 |
Pos logit contributions | |
els | +0.124 |
bulb | +0.116 |
globe | +0.115 |
atop | +0.114 |
ool | +0.113 |
Neg logit contributions | |
He | -0.176 |
Jack | -0.173 |
His | -0.172 |
They | -0.163 |
Kids | -0.159 |
Token | Ben |
Token ID | 3932 |
Feature activation | -1.16e-02 |
Pos logit contributions | |
His | +0.141 |
Thanks | +0.136 |
Jack | +0.136 |
He | +0.131 |
David | +0.127 |
Neg logit contributions | |
bulb | -0.191 |
els | -0.186 |
pieces | -0.181 |
bulbs | -0.179 |
chim | -0.179 |
Token | says |
Token ID | 1139 |
Feature activation | +1.72e-02 |
Pos logit contributions | |
els | +0.017 |
bulb | +0.017 |
hole | +0.017 |
globe | +0.017 |
ool | +0.016 |
Neg logit contributions | |
His | -0.020 |
Jack | -0.020 |
He | -0.020 |
Kids | -0.019 |
They | -0.018 |
Token | . |
Token ID | 13 |
Feature activation | +1.03e-01 |
Pos logit contributions | |
His | +0.033 |
He | +0.033 |
Her | +0.031 |
His | +0.031 |
Jack | +0.030 |
Neg logit contributions | |
els | -0.025 |
globe | -0.024 |
ool | -0.024 |
bulb | -0.024 |
blades | -0.023 |
Token |
|
Token ID | 198 |
Feature activation | +9.82e-02 |
Pos logit contributions | |
Jack | +0.156 |
Thanks | +0.154 |
David | +0.154 |
Sorry | +0.147 |
Kids | +0.145 |
Neg logit contributions | |
pieces | -0.147 |
bulb | -0.142 |
on | -0.142 |
red | -0.141 |
trunk | -0.141 |
Token |
|
Token ID | 198 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
Jack | +0.119 |
Thanks | +0.117 |
David | +0.113 |
His | +0.113 |
Sorry | +0.108 |
Neg logit contributions | |
els | -0.182 |
bulb | -0.180 |
pieces | -0.179 |
globe | -0.178 |
hole | -0.176 |
Token | by |
Token ID | 416 |
Feature activation | -1.78e-02 |
Pos logit contributions | |
els | +0.070 |
ool | +0.068 |
ps | +0.067 |
bulb | +0.066 |
hole | +0.066 |
Neg logit contributions | |
He | -0.095 |
His | -0.094 |
Jack | -0.094 |
They | -0.087 |
She | -0.083 |
Token | being |
Token ID | 852 |
Feature activation | -4.97e-02 |
Pos logit contributions | |
hole | +0.034 |
ool | +0.034 |
globe | +0.031 |
els | +0.031 |
blades | +0.030 |
Neg logit contributions | |
His | -0.026 |
He | -0.025 |
Jack | -0.025 |
They | -0.023 |
His | -0.023 |
Token | brave |
Token ID | 14802 |
Feature activation | -9.63e-02 |
Pos logit contributions | |
els | +0.090 |
hole | +0.089 |
globe | +0.089 |
ool | +0.088 |
bulb | +0.087 |
Neg logit contributions | |
Jack | -0.077 |
His | -0.076 |
He | -0.074 |
Thanks | -0.067 |
They | -0.067 |
Token | and |
Token ID | 290 |
Feature activation | +2.40e-02 |
Pos logit contributions | |
els | +0.145 |
globe | +0.143 |
bulb | +0.141 |
ool | +0.138 |
hole | +0.135 |
Neg logit contributions | |
He | -0.171 |
Jack | -0.167 |
His | -0.162 |
He | -0.156 |
They | -0.156 |
Token | clever |
Token ID | 14169 |
Feature activation | -8.79e-02 |
Pos logit contributions | |
His | +0.038 |
Jack | +0.037 |
He | +0.036 |
Kids | +0.034 |
David | +0.034 |
Neg logit contributions | |
els | -0.041 |
hole | -0.040 |
bulb | -0.040 |
globe | -0.039 |
ps | -0.039 |
Token | . |
Token ID | 13 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
hole | +0.115 |
els | +0.115 |
globe | +0.115 |
bulb | +0.111 |
ool | +0.110 |
Neg logit contributions | |
He | -0.170 |
Jack | -0.166 |
His | -0.166 |
He | -0.156 |
They | -0.156 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +1.90e-02 |
Pos logit contributions | |
Jack | +0.117 |
His | +0.113 |
Kids | +0.104 |
Sorry | +0.104 |
Her | +0.104 |
Neg logit contributions | |
hole | -0.238 |
globe | -0.234 |
trunk | -0.232 |
pieces | -0.231 |
els | -0.231 |
Token | Once |
Token ID | 7454 |
Feature activation | +3.96e-02 |
Pos logit contributions | |
Jack | +0.022 |
He | +0.021 |
His | +0.021 |
Kids | +0.019 |
They | +0.018 |
Neg logit contributions | |
els | -0.042 |
bulb | -0.041 |
ps | -0.040 |
hole | -0.040 |
ening | -0.040 |
Token | upon |
Token ID | 2402 |
Feature activation | +5.55e-02 |
Pos logit contributions | |
His | +0.058 |
Jack | +0.057 |
He | +0.056 |
Kids | +0.052 |
Thanks | +0.051 |
Neg logit contributions | |
els | -0.071 |
bulb | -0.070 |
hole | -0.070 |
globe | -0.069 |
ool | -0.068 |
Token | a |
Token ID | 257 |
Feature activation | +3.65e-02 |
Pos logit contributions | |
Kids | +0.132 |
Sorry | +0.124 |
Brother | +0.123 |
David | +0.121 |
enny | +0.120 |
Neg logit contributions | |
pieces | -0.048 |
flying | -0.046 |
on | -0.046 |
glowing | -0.042 |
hanging | -0.042 |
Token | time |
Token ID | 640 |
Feature activation | +2.46e-02 |
Pos logit contributions | |
His | +0.088 |
He | +0.085 |
Jack | +0.084 |
Kids | +0.084 |
David | +0.083 |
Neg logit contributions | |
ool | -0.030 |
et | -0.027 |
globe | -0.026 |
hole | -0.026 |
bulb | -0.025 |
Token | cry |
Token ID | 3960 |
Feature activation | -6.13e-02 |
Pos logit contributions | |
His | +0.040 |
Jack | +0.039 |
He | +0.038 |
Kids | +0.035 |
Thanks | +0.035 |
Neg logit contributions | |
els | -0.047 |
hole | -0.046 |
globe | -0.045 |
bulb | -0.045 |
ool | -0.044 |
Token | and |
Token ID | 290 |
Feature activation | +5.13e-02 |
Pos logit contributions | |
els | +0.079 |
hole | +0.076 |
bulb | +0.076 |
globe | +0.074 |
ps | +0.073 |
Neg logit contributions | |
His | -0.122 |
Jack | -0.122 |
He | -0.119 |
Kids | -0.112 |
David | -0.112 |
Token | Tim |
Token ID | 5045 |
Feature activation | -2.38e-02 |
Pos logit contributions | |
Jack | +0.074 |
His | +0.074 |
He | +0.073 |
David | +0.067 |
Kids | +0.067 |
Neg logit contributions | |
els | -0.092 |
bulb | -0.092 |
hole | -0.089 |
ps | -0.088 |
globe | -0.088 |
Token | my |
Token ID | 1820 |
Feature activation | +1.15e-03 |
Pos logit contributions | |
globe | +0.044 |
els | +0.043 |
bulb | +0.042 |
hole | +0.042 |
ool | +0.042 |
Neg logit contributions | |
His | -0.034 |
Jack | -0.034 |
He | -0.034 |
Brother | -0.030 |
Kids | -0.030 |
Token | felt |
Token ID | 2936 |
Feature activation | +5.10e-03 |
Pos logit contributions | |
His | +0.001 |
Jack | +0.001 |
He | +0.001 |
Brother | +0.001 |
Kids | +0.001 |
Neg logit contributions | |
els | -0.002 |
globe | -0.002 |
hole | -0.002 |
ool | -0.002 |
bulb | -0.002 |
Token | very |
Token ID | 845 |
Feature activation | +2.74e-02 |
Pos logit contributions | |
Jack | +0.012 |
David | +0.012 |
Kids | +0.012 |
Jim | +0.011 |
enny | +0.011 |
Neg logit contributions | |
bulb | -0.005 |
ps | -0.005 |
object | -0.004 |
ening | -0.004 |
matches | -0.004 |
Token | sad |
Token ID | 6507 |
Feature activation | -5.41e-02 |
Pos logit contributions | |
His | +0.048 |
Jack | +0.047 |
He | +0.047 |
Thanks | +0.043 |
Kids | +0.042 |
Neg logit contributions | |
els | -0.044 |
hole | -0.042 |
globe | -0.041 |
bulb | -0.041 |
ool | -0.041 |
Token | . |
Token ID | 13 |
Feature activation | +1.21e-01 |
Pos logit contributions | |
els | +0.062 |
hole | +0.059 |
bulb | +0.058 |
globe | +0.058 |
ool | +0.057 |
Neg logit contributions | |
His | -0.116 |
Jack | -0.115 |
He | -0.115 |
Kids | -0.107 |
Thanks | -0.106 |
Token | He |
Token ID | 679 |
Feature activation | +4.21e-02 |
Pos logit contributions | |
Jack | +0.125 |
Kids | +0.120 |
His | +0.119 |
David | +0.118 |
He | +0.117 |
Neg logit contributions | |
chim | -0.254 |
trunk | -0.250 |
bulb | -0.246 |
els | -0.242 |
object | -0.239 |
Token | didn |
Token ID | 1422 |
Feature activation | +2.91e-02 |
Pos logit contributions | |
Jack | +0.063 |
His | +0.061 |
He | +0.059 |
David | +0.059 |
Jim | +0.057 |
Neg logit contributions | |
els | -0.073 |
bulb | -0.072 |
et | -0.071 |
ps | -0.071 |
ening | -0.070 |
Token | 't |
Token ID | 470 |
Feature activation | +3.97e-02 |
Pos logit contributions | |
Jack | +0.054 |
David | +0.054 |
Kids | +0.053 |
Okay | +0.053 |
Of | +0.052 |
Neg logit contributions | |
ps | -0.036 |
els | -0.034 |
bulb | -0.034 |
on | -0.033 |
falling | -0.033 |
Token | upon |
Token ID | 2402 |
Feature activation | +6.77e-02 |
Pos logit contributions | |
His | +0.080 |
Jack | +0.079 |
He | +0.077 |
Kids | +0.071 |
David | +0.070 |
Neg logit contributions | |
els | -0.107 |
hole | -0.104 |
bulb | -0.104 |
globe | -0.102 |
ps | -0.102 |
Token | a |
Token ID | 257 |
Feature activation | +5.03e-02 |
Pos logit contributions | |
Kids | +0.156 |
Brother | +0.146 |
David | +0.144 |
Sorry | +0.144 |
Sure | +0.143 |
Neg logit contributions | |
pieces | -0.065 |
flying | -0.061 |
on | -0.059 |
glowing | -0.057 |
ps | -0.057 |
Token | time |
Token ID | 640 |
Feature activation | +4.13e-02 |
Pos logit contributions | |
His | +0.122 |
He | +0.118 |
Jack | +0.117 |
Kids | +0.116 |
David | +0.115 |
Neg logit contributions | |
ool | -0.040 |
et | -0.038 |
els | -0.036 |
bulb | -0.036 |
globe | -0.034 |
Token | , |
Token ID | 11 |
Feature activation | +8.96e-02 |
Pos logit contributions | |
His | +0.091 |
Jack | +0.090 |
He | +0.089 |
Kids | +0.085 |
David | +0.084 |
Neg logit contributions | |
els | -0.043 |
bulb | -0.042 |
hole | -0.042 |
globe | -0.041 |
ool | -0.040 |
Token | there |
Token ID | 612 |
Feature activation | +2.44e-02 |
Pos logit contributions | |
His | +0.173 |
Jack | +0.172 |
He | +0.171 |
They | +0.155 |
Her | +0.153 |
Neg logit contributions | |
els | -0.127 |
ps | -0.119 |
hole | -0.118 |
ening | -0.118 |
ool | -0.116 |
Token | were |
Token ID | 547 |
Feature activation | +5.17e-02 |
Pos logit contributions | |
His | +0.043 |
Jack | +0.042 |
He | +0.041 |
David | +0.040 |
Kids | +0.039 |
Neg logit contributions | |
bulb | -0.036 |
els | -0.036 |
globe | -0.036 |
hole | -0.035 |
object | -0.034 |
Token | two |
Token ID | 734 |
Feature activation | -2.17e-02 |
Pos logit contributions | |
His | +0.113 |
He | +0.109 |
Something | +0.103 |
Thanks | +0.102 |
Sure | +0.101 |
Neg logit contributions | |
pieces | -0.056 |
els | -0.054 |
egg | -0.054 |
bulb | -0.054 |
matches | -0.051 |
Token | friends |
Token ID | 2460 |
Feature activation | +5.51e-04 |
Pos logit contributions | |
hole | +0.036 |
globe | +0.035 |
els | +0.034 |
ening | +0.034 |
ool | +0.034 |
Neg logit contributions | |
Jack | -0.035 |
His | -0.035 |
He | -0.035 |
Brother | -0.032 |
Dad | -0.032 |
Token | . |
Token ID | 13 |
Feature activation | +1.32e-01 |
Pos logit contributions | |
Jack | +0.001 |
His | +0.001 |
He | +0.001 |
Kids | +0.001 |
Thanks | +0.001 |
Neg logit contributions | |
els | -0.001 |
bulb | -0.001 |
hole | -0.001 |
et | -0.001 |
globe | -0.001 |
Token | Their |
Token ID | 5334 |
Feature activation | +3.08e-02 |
Pos logit contributions | |
His | +0.114 |
Jack | +0.114 |
He | +0.111 |
David | +0.100 |
Thanks | +0.096 |
Neg logit contributions | |
els | -0.311 |
globe | -0.309 |
bulb | -0.307 |
hole | -0.301 |
ool | -0.299 |
Token | names |
Token ID | 3891 |
Feature activation | -6.56e-02 |
Pos logit contributions | |
Jack | +0.055 |
His | +0.055 |
He | +0.053 |
Kids | +0.051 |
Dad | +0.050 |
Neg logit contributions | |
hole | -0.045 |
els | -0.045 |
ening | -0.044 |
bulb | -0.043 |
atop | -0.043 |
Token | See |
Token ID | 4091 |
Feature activation | -1.78e-02 |
Pos logit contributions | |
Jack | +0.138 |
His | +0.134 |
David | +0.130 |
He | +0.127 |
Her | +0.123 |
Neg logit contributions | |
globe | -0.144 |
bulb | -0.136 |
Tower | -0.130 |
hole | -0.130 |
trunk | -0.129 |
Token | you |
Token ID | 345 |
Feature activation | -2.08e-03 |
Pos logit contributions | |
els | +0.035 |
hole | +0.035 |
bulb | +0.034 |
globe | +0.034 |
ool | +0.034 |
Neg logit contributions | |
Jack | -0.023 |
His | -0.023 |
He | -0.022 |
Kids | -0.020 |
David | -0.020 |
Token | tomorrow |
Token ID | 9439 |
Feature activation | -5.20e-02 |
Pos logit contributions | |
els | +0.005 |
bulb | +0.004 |
hole | +0.004 |
globe | +0.004 |
ool | +0.004 |
Neg logit contributions | |
Jack | -0.002 |
His | -0.002 |
He | -0.002 |
Dad | -0.002 |
They | -0.002 |
Token | . |
Token ID | 13 |
Feature activation | +8.79e-02 |
Pos logit contributions | |
els | +0.083 |
bulb | +0.078 |
ool | +0.076 |
ening | +0.076 |
object | +0.075 |
Neg logit contributions | |
Jack | -0.091 |
He | -0.090 |
They | -0.089 |
His | -0.087 |
Un | -0.084 |
Token | And |
Token ID | 843 |
Feature activation | +3.17e-02 |
Pos logit contributions | |
Jack | +0.142 |
His | +0.138 |
David | +0.132 |
He | +0.130 |
Her | +0.126 |
Neg logit contributions | |
globe | -0.149 |
bulb | -0.141 |
Tower | -0.139 |
hole | -0.137 |
chim | -0.136 |
Token | don |
Token ID | 836 |
Feature activation | +9.74e-03 |
Pos logit contributions | |
Jack | +0.044 |
His | +0.044 |
He | +0.041 |
David | +0.040 |
Kids | +0.040 |
Neg logit contributions | |
bulb | -0.058 |
globe | -0.058 |
hole | -0.057 |
els | -0.057 |
object | -0.055 |
Token | 't |
Token ID | 470 |
Feature activation | +2.56e-02 |
Pos logit contributions | |
Jack | +0.016 |
His | +0.016 |
David | +0.015 |
Brother | +0.015 |
Thanks | +0.015 |
Neg logit contributions | |
globe | -0.015 |
bulb | -0.015 |
els | -0.015 |
hole | -0.014 |
ool | -0.014 |
Token | forget |
Token ID | 6044 |
Feature activation | -2.03e-03 |
Pos logit contributions | |
His | +0.033 |
Jack | +0.033 |
He | +0.032 |
Kids | +0.029 |
David | +0.029 |
Neg logit contributions | |
els | -0.050 |
bulb | -0.050 |
hole | -0.049 |
globe | -0.049 |
ening | -0.048 |
Token | your |
Token ID | 534 |
Feature activation | -2.75e-02 |
Pos logit contributions | |
els | +0.003 |
bulb | +0.003 |
hole | +0.003 |
globe | +0.003 |
ps | +0.003 |
Neg logit contributions | |
His | -0.003 |
Jack | -0.003 |
He | -0.003 |
Kids | -0.003 |
David | -0.003 |
Token | bone |
Token ID | 9970 |
Feature activation | -1.56e-01 |
Pos logit contributions | |
hole | +0.054 |
els | +0.053 |
atop | +0.051 |
ool | +0.049 |
ening | +0.049 |
Neg logit contributions | |
His | -0.036 |
Kids | -0.036 |
Jack | -0.036 |
He | -0.035 |
Thanks | -0.035 |
Token | . |
Token ID | 13 |
Feature activation | +8.66e-02 |
Pos logit contributions | |
els | +0.226 |
ening | +0.210 |
hole | +0.208 |
ool | +0.203 |
atop | +0.201 |
Neg logit contributions | |
He | -0.295 |
His | -0.290 |
Jack | -0.283 |
They | -0.275 |
He | -0.266 |
Token | him |
Token ID | 683 |
Feature activation | -3.53e-02 |
Pos logit contributions | |
els | +0.041 |
ps | +0.040 |
et | +0.039 |
egg | +0.039 |
trunk | +0.038 |
Neg logit contributions | |
Jack | -0.025 |
David | -0.023 |
His | -0.023 |
Thanks | -0.022 |
enny | -0.022 |
Token | and |
Token ID | 290 |
Feature activation | +4.31e-02 |
Pos logit contributions | |
els | +0.060 |
ool | +0.059 |
bulb | +0.058 |
hole | +0.057 |
ening | +0.056 |
Neg logit contributions | |
He | -0.060 |
His | -0.060 |
He | -0.053 |
They | -0.053 |
Jack | -0.052 |
Token | laughed |
Token ID | 13818 |
Feature activation | -5.78e-02 |
Pos logit contributions | |
Jack | +0.070 |
His | +0.069 |
He | +0.068 |
David | +0.065 |
Dad | +0.063 |
Neg logit contributions | |
els | -0.070 |
bulb | -0.069 |
ps | -0.067 |
pieces | -0.067 |
globe | -0.066 |
Token | . |
Token ID | 13 |
Feature activation | +1.07e-01 |
Pos logit contributions | |
els | +0.045 |
bulb | +0.043 |
ps | +0.042 |
hole | +0.042 |
ening | +0.042 |
Neg logit contributions | |
Jack | -0.142 |
His | -0.139 |
He | -0.135 |
David | -0.135 |
Kids | -0.135 |
Token | She |
Token ID | 1375 |
Feature activation | +5.21e-02 |
Pos logit contributions | |
Jack | +0.126 |
David | +0.112 |
Johnny | +0.109 |
His | +0.104 |
Thanks | +0.103 |
Neg logit contributions | |
trunk | -0.209 |
els | -0.203 |
chim | -0.202 |
globe | -0.202 |
hanging | -0.202 |
Token | said |
Token ID | 531 |
Feature activation | +3.14e-02 |
Pos logit contributions | |
Jack | +0.090 |
His | +0.090 |
David | +0.087 |
He | +0.086 |
Kids | +0.083 |
Neg logit contributions | |
bulb | -0.074 |
els | -0.073 |
et | -0.073 |
ps | -0.073 |
pieces | -0.073 |
Token | , |
Token ID | 11 |
Feature activation | +3.87e-02 |
Pos logit contributions | |
His | +0.065 |
Jack | +0.065 |
He | +0.064 |
Kids | +0.060 |
David | +0.059 |
Neg logit contributions | |
els | -0.038 |
hole | -0.037 |
bulb | -0.036 |
globe | -0.036 |
ool | -0.035 |
Token | " |
Token ID | 366 |
Feature activation | +5.18e-02 |
Pos logit contributions | |
His | +0.048 |
He | +0.046 |
Jack | +0.043 |
His | +0.040 |
Her | +0.039 |
Neg logit contributions | |
els | -0.085 |
hole | -0.084 |
ool | -0.083 |
ening | -0.080 |
bulb | -0.080 |
Token | You |
Token ID | 1639 |
Feature activation | +3.10e-02 |
Pos logit contributions | |
ete | +0.066 |
enny | +0.066 |
His | +0.058 |
David | +0.057 |
His | +0.055 |
Neg logit contributions | |
globe | -0.094 |
blades | -0.084 |
pieces | -0.083 |
ornament | -0.081 |
object | -0.080 |
Token | are |
Token ID | 389 |
Feature activation | +2.06e-02 |
Pos logit contributions | |
His | +0.058 |
Jack | +0.056 |
He | +0.056 |
Kids | +0.054 |
Un | +0.054 |
Neg logit contributions | |
bulb | -0.043 |
els | -0.043 |
ps | -0.042 |
hole | -0.041 |
globe | -0.041 |
Token | a |
Token ID | 257 |
Feature activation | +4.67e-03 |
Pos logit contributions | |
His | +0.037 |
Jack | +0.036 |
He | +0.035 |
Kids | +0.035 |
David | +0.034 |
Neg logit contributions | |
els | -0.030 |
bulb | -0.030 |
ps | -0.030 |
globe | -0.029 |
matches | -0.028 |
Token | . |
Token ID | 13 |
Feature activation | +1.30e-01 |
Pos logit contributions | |
els | +0.098 |
hole | +0.091 |
ool | +0.091 |
globe | +0.090 |
bulb | +0.088 |
Neg logit contributions | |
He | -0.111 |
His | -0.110 |
Jack | -0.103 |
They | -0.101 |
Un | -0.096 |
Token | " |
Token ID | 366 |
Feature activation | +8.29e-02 |
Pos logit contributions | |
Jack | +0.117 |
His | +0.115 |
He | +0.106 |
Kids | +0.102 |
Thanks | +0.100 |
Neg logit contributions | |
els | -0.287 |
hole | -0.286 |
bulb | -0.285 |
globe | -0.284 |
trunk | -0.280 |
Token | Now |
Token ID | 3844 |
Feature activation | +2.29e-02 |
Pos logit contributions | |
enny | +0.078 |
His | +0.076 |
Jack | +0.076 |
ete | +0.073 |
His | +0.073 |
Neg logit contributions | |
globe | -0.169 |
bulb | -0.165 |
et | -0.164 |
els | -0.163 |
object | -0.161 |
Token | he |
Token ID | 339 |
Feature activation | +2.55e-02 |
Pos logit contributions | |
Jack | +0.028 |
His | +0.027 |
He | +0.027 |
They | +0.024 |
Thanks | +0.023 |
Neg logit contributions | |
els | -0.048 |
hole | -0.047 |
ool | -0.046 |
ening | -0.045 |
globe | -0.045 |
Token | has |
Token ID | 468 |
Feature activation | +2.45e-02 |
Pos logit contributions | |
He | +0.042 |
Jack | +0.040 |
His | +0.039 |
He | +0.037 |
Something | +0.037 |
Neg logit contributions | |
els | -0.041 |
hole | -0.040 |
globe | -0.039 |
bulb | -0.039 |
ool | -0.038 |
Token | a |
Token ID | 257 |
Feature activation | +1.26e-02 |
Pos logit contributions | |
He | +0.032 |
His | +0.031 |
Jack | +0.031 |
His | +0.027 |
They | +0.027 |
Neg logit contributions | |
els | -0.050 |
globe | -0.049 |
ool | -0.048 |
hole | -0.048 |
bulb | -0.047 |
Token | nose |
Token ID | 9686 |
Feature activation | -1.25e-01 |
Pos logit contributions | |
He | +0.022 |
His | +0.022 |
Jack | +0.022 |
Sure | +0.020 |
Thanks | +0.020 |
Neg logit contributions | |
els | -0.021 |
hole | -0.019 |
ps | -0.019 |
ening | -0.018 |
cks | -0.018 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.19e-01 |
Pos logit contributions | |
els | +0.183 |
globe | +0.165 |
ening | +0.162 |
ool | +0.160 |
angles | +0.159 |
Neg logit contributions | |
He | -0.236 |
His | -0.226 |
Jack | -0.211 |
please | -0.211 |
They | -0.210 |
Token |
|
Token ID | 198 |
Feature activation | +1.06e-01 |
Pos logit contributions | |
Jack | +0.119 |
His | +0.116 |
Thanks | +0.111 |
Kids | +0.111 |
He | +0.109 |
Neg logit contributions | |
hole | -0.250 |
els | -0.250 |
pieces | -0.248 |
bulb | -0.247 |
globe | -0.242 |
Token |
|
Token ID | 198 |
Feature activation | +1.11e-01 |
Pos logit contributions | |
Thanks | +0.115 |
His | +0.111 |
Sorry | +0.109 |
Jack | +0.108 |
Kids | +0.104 |
Neg logit contributions | |
globe | -0.195 |
els | -0.195 |
hole | -0.195 |
ool | -0.194 |
bulb | -0.193 |
Token | " |
Token ID | 1 |
Feature activation | +8.05e-02 |
Pos logit contributions | |
His | +0.085 |
Something | +0.082 |
Deal | +0.078 |
Kids | +0.078 |
Thanks | +0.077 |
Neg logit contributions | |
cut | -0.213 |
globe | -0.213 |
bulb | -0.207 |
pieces | -0.205 |
object | -0.204 |
Token | his |
Token ID | 465 |
Feature activation | -1.47e-02 |
Pos logit contributions | |
els | +0.100 |
hole | +0.098 |
bulb | +0.098 |
globe | +0.096 |
ool | +0.095 |
Neg logit contributions | |
Jack | -0.084 |
His | -0.084 |
He | -0.081 |
Thanks | -0.074 |
David | -0.074 |
Token | friends |
Token ID | 2460 |
Feature activation | -6.55e-02 |
Pos logit contributions | |
hole | +0.028 |
ening | +0.027 |
ool | +0.027 |
els | +0.027 |
ps | +0.026 |
Neg logit contributions | |
Jack | -0.021 |
His | -0.020 |
Thanks | -0.019 |
He | -0.019 |
His | -0.019 |
Token | to |
Token ID | 284 |
Feature activation | +6.95e-04 |
Pos logit contributions | |
els | +0.096 |
bulb | +0.092 |
ool | +0.090 |
hole | +0.090 |
ps | +0.089 |
Neg logit contributions | |
Jack | -0.122 |
His | -0.122 |
He | -0.119 |
David | -0.110 |
Thanks | -0.109 |
Token | play |
Token ID | 711 |
Feature activation | -7.34e-02 |
Pos logit contributions | |
His | +0.001 |
Jack | +0.001 |
He | +0.001 |
Kids | +0.001 |
David | +0.001 |
Neg logit contributions | |
els | -0.001 |
bulb | -0.001 |
hole | -0.001 |
globe | -0.001 |
ps | -0.001 |
Token | with |
Token ID | 351 |
Feature activation | -3.24e-02 |
Pos logit contributions | |
els | +0.108 |
hole | +0.104 |
bulb | +0.102 |
ps | +0.100 |
ening | +0.100 |
Neg logit contributions | |
Jack | -0.140 |
His | -0.138 |
He | -0.134 |
Thanks | -0.125 |
They | -0.124 |
Token | the |
Token ID | 262 |
Feature activation | -4.08e-02 |
Pos logit contributions | |
hole | +0.053 |
els | +0.052 |
bulb | +0.051 |
ool | +0.051 |
globe | +0.050 |
Neg logit contributions | |
Jack | -0.056 |
His | -0.055 |
He | -0.054 |
They | -0.048 |
David | -0.048 |
Token | ball |
Token ID | 2613 |
Feature activation | -1.52e-01 |
Pos logit contributions | |
els | +0.067 |
hole | +0.063 |
ps | +0.062 |
ening | +0.060 |
ool | +0.060 |
Neg logit contributions | |
Jack | -0.076 |
He | -0.072 |
Thanks | -0.071 |
His | -0.069 |
Sure | -0.068 |
Token | together |
Token ID | 1978 |
Feature activation | -6.47e-02 |
Pos logit contributions | |
els | +0.237 |
ps | +0.217 |
ening | +0.211 |
hole | +0.209 |
bulb | +0.208 |
Neg logit contributions | |
He | -0.277 |
His | -0.274 |
Jack | -0.268 |
They | -0.248 |
He | -0.247 |
Token | . |
Token ID | 13 |
Feature activation | +1.13e-01 |
Pos logit contributions | |
els | +0.108 |
bulb | +0.103 |
ool | +0.100 |
ps | +0.099 |
bulbs | +0.099 |
Neg logit contributions | |
Jack | -0.111 |
He | -0.110 |
His | -0.109 |
They | -0.100 |
Thanks | -0.097 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | +1.72e-02 |
Pos logit contributions | |
Jack | +0.157 |
His | +0.150 |
Sorry | +0.149 |
Kids | +0.147 |
David | +0.146 |
Neg logit contributions | |
globe | -0.199 |
hole | -0.191 |
trunk | -0.191 |
atop | -0.185 |
pieces | -0.183 |
Token | Once |
Token ID | 7454 |
Feature activation | +3.85e-02 |
Pos logit contributions | |
Jack | +0.021 |
He | +0.020 |
His | +0.020 |
Kids | +0.018 |
They | +0.017 |
Neg logit contributions | |
els | -0.037 |
bulb | -0.036 |
ps | -0.035 |
hole | -0.035 |
ening | -0.035 |
Token | day |
Token ID | 1110 |
Feature activation | +1.30e-02 |
Pos logit contributions | |
Jack | +0.047 |
His | +0.046 |
He | +0.046 |
David | +0.045 |
Kids | +0.044 |
Neg logit contributions | |
bulb | -0.015 |
globe | -0.015 |
els | -0.014 |
ool | -0.014 |
object | -0.014 |
Token | , |
Token ID | 11 |
Feature activation | +7.21e-02 |
Pos logit contributions | |
His | +0.027 |
Jack | +0.027 |
He | +0.026 |
Kids | +0.025 |
Dad | +0.025 |
Neg logit contributions | |
els | -0.016 |
hole | -0.015 |
bulb | -0.015 |
ps | -0.015 |
globe | -0.015 |
Token | he |
Token ID | 339 |
Feature activation | +3.61e-02 |
Pos logit contributions | |
Kids | +0.097 |
Jack | +0.092 |
Brother | +0.089 |
David | +0.089 |
enny | +0.088 |
Neg logit contributions | |
hanging | -0.131 |
globe | -0.129 |
on | -0.126 |
atop | -0.126 |
pieces | -0.125 |
Token | saw |
Token ID | 2497 |
Feature activation | -5.33e-03 |
Pos logit contributions | |
His | +0.066 |
He | +0.063 |
Jack | +0.062 |
Her | +0.061 |
David | +0.061 |
Neg logit contributions | |
bulb | -0.048 |
pieces | -0.048 |
et | -0.046 |
els | -0.046 |
matches | -0.045 |
Token | a |
Token ID | 257 |
Feature activation | -1.21e-02 |
Pos logit contributions | |
els | +0.008 |
bulb | +0.008 |
pieces | +0.008 |
object | +0.008 |
matches | +0.008 |
Neg logit contributions | |
Jack | -0.009 |
Kids | -0.009 |
His | -0.009 |
He | -0.009 |
Thanks | -0.008 |
Token | big |
Token ID | 1263 |
Feature activation | -7.63e-02 |
Pos logit contributions | |
els | +0.020 |
ps | +0.020 |
hole | +0.019 |
ening | +0.018 |
et | +0.018 |
Neg logit contributions | |
Jack | -0.021 |
He | -0.021 |
His | -0.021 |
Un | -0.019 |
Sure | -0.019 |
Token | apple |
Token ID | 17180 |
Feature activation | -1.54e-01 |
Pos logit contributions | |
ps | +0.117 |
els | +0.108 |
hole | +0.105 |
cks | +0.104 |
atop | +0.096 |
Neg logit contributions | |
Jack | -0.139 |
He | -0.134 |
His | -0.131 |
He | -0.130 |
please | -0.130 |
Token | tree |
Token ID | 5509 |
Feature activation | -1.24e-01 |
Pos logit contributions | |
ps | +0.185 |
els | +0.181 |
ening | +0.175 |
angles | +0.172 |
hole | +0.165 |
Neg logit contributions | |
He | -0.337 |
His | -0.321 |
He | -0.305 |
She | -0.299 |
They | -0.298 |
Token | in |
Token ID | 287 |
Feature activation | -1.97e-02 |
Pos logit contributions | |
ps | +0.206 |
els | +0.198 |
globe | +0.186 |
ening | +0.184 |
blades | +0.179 |
Neg logit contributions | |
He | -0.214 |
His | -0.205 |
They | -0.199 |
She | -0.192 |
He | -0.190 |
Token | his |
Token ID | 465 |
Feature activation | -8.97e-04 |
Pos logit contributions | |
els | +0.033 |
bulb | +0.032 |
hole | +0.032 |
globe | +0.031 |
ps | +0.031 |
Neg logit contributions | |
His | -0.032 |
Jack | -0.032 |
He | -0.031 |
Kids | -0.029 |
David | -0.029 |
Token | yard |
Token ID | 12699 |
Feature activation | -1.18e-01 |
Pos logit contributions | |
els | +0.001 |
bulb | +0.001 |
hole | +0.001 |
globe | +0.001 |
trunk | +0.001 |
Neg logit contributions | |
Jack | -0.002 |
His | -0.002 |
He | -0.002 |
Kids | -0.002 |
David | -0.002 |
Token | so |
Token ID | 523 |
Feature activation | +6.37e-02 |
Pos logit contributions | |
His | +0.058 |
Jack | +0.057 |
He | +0.057 |
Thanks | +0.051 |
David | +0.050 |
Neg logit contributions | |
els | -0.093 |
globe | -0.090 |
bulb | -0.090 |
ool | -0.089 |
hole | -0.089 |
Token | happy |
Token ID | 3772 |
Feature activation | -2.80e-02 |
Pos logit contributions | |
His | +0.084 |
He | +0.083 |
Jack | +0.083 |
Thanks | +0.075 |
They | +0.072 |
Neg logit contributions | |
els | -0.133 |
hole | -0.129 |
bulb | -0.127 |
globe | -0.127 |
ool | -0.126 |
Token | that |
Token ID | 326 |
Feature activation | +1.34e-02 |
Pos logit contributions | |
els | +0.041 |
bulb | +0.040 |
hole | +0.040 |
globe | +0.039 |
ps | +0.039 |
Neg logit contributions | |
His | -0.050 |
Jack | -0.050 |
He | -0.049 |
Kids | -0.046 |
David | -0.045 |
Token | she |
Token ID | 673 |
Feature activation | +2.95e-02 |
Pos logit contributions | |
Jack | +0.010 |
His | +0.010 |
He | +0.009 |
Thanks | +0.007 |
David | +0.007 |
Neg logit contributions | |
els | -0.035 |
hole | -0.034 |
ps | -0.034 |
ool | -0.034 |
globe | -0.034 |
Token | put |
Token ID | 1234 |
Feature activation | -6.46e-02 |
Pos logit contributions | |
His | +0.044 |
Jack | +0.043 |
He | +0.042 |
Kids | +0.039 |
David | +0.039 |
Neg logit contributions | |
els | -0.052 |
bulb | -0.051 |
hole | -0.051 |
globe | -0.050 |
ps | -0.050 |
Token | it |
Token ID | 340 |
Feature activation | -7.38e-02 |
Pos logit contributions | |
chim | +0.147 |
ool | +0.141 |
hole | +0.141 |
globe | +0.138 |
bulb | +0.134 |
Neg logit contributions | |
Jack | -0.070 |
He | -0.062 |
His | -0.061 |
Thanks | -0.061 |
Sure | -0.056 |
Token | in |
Token ID | 287 |
Feature activation | -6.39e-03 |
Pos logit contributions | |
els | +0.122 |
chim | +0.119 |
ool | +0.115 |
ening | +0.115 |
hole | +0.114 |
Neg logit contributions | |
He | -0.128 |
His | -0.128 |
Jack | -0.122 |
She | -0.115 |
Thanks | -0.115 |
Token | a |
Token ID | 257 |
Feature activation | -1.12e-03 |
Pos logit contributions | |
els | +0.014 |
hole | +0.013 |
ool | +0.013 |
ening | +0.013 |
globe | +0.013 |
Neg logit contributions | |
Jack | -0.008 |
His | -0.008 |
He | -0.007 |
Thanks | -0.007 |
Her | -0.007 |
Token | jar |
Token ID | 17379 |
Feature activation | -1.46e-01 |
Pos logit contributions | |
els | +0.002 |
ps | +0.001 |
ening | +0.001 |
hole | +0.001 |
cks | +0.001 |
Neg logit contributions | |
Jack | -0.002 |
He | -0.002 |
His | -0.002 |
Sure | -0.002 |
Thanks | -0.002 |
Token | and |
Token ID | 290 |
Feature activation | +4.47e-02 |
Pos logit contributions | |
els | +0.251 |
globe | +0.230 |
bulb | +0.228 |
ps | +0.227 |
ening | +0.225 |
Neg logit contributions | |
He | -0.249 |
His | -0.244 |
Jack | -0.243 |
She | -0.221 |
They | -0.220 |
Token | took |
Token ID | 1718 |
Feature activation | -4.48e-02 |
Pos logit contributions | |
Jack | +0.068 |
His | +0.068 |
He | +0.066 |
Kids | +0.061 |
Thanks | +0.060 |
Neg logit contributions | |
els | -0.079 |
hole | -0.077 |
bulb | -0.076 |
globe | -0.075 |
ening | -0.075 |
Token | I |
Token ID | 40 |
Feature activation | +5.66e-02 |
Pos logit contributions | |
David | +0.053 |
Jack | +0.045 |
Jason | +0.044 |
His | +0.043 |
Sean | +0.043 |
Neg logit contributions | |
bulb | -0.109 |
object | -0.109 |
globe | -0.108 |
st | -0.104 |
pieces | -0.104 |
Token | can |
Token ID | 460 |
Feature activation | +3.52e-02 |
Pos logit contributions | |
His | +0.087 |
Jack | +0.087 |
He | +0.083 |
David | +0.079 |
Kids | +0.079 |
Neg logit contributions | |
bulb | -0.097 |
els | -0.096 |
globe | -0.095 |
hole | -0.095 |
ool | -0.094 |
Token | 't |
Token ID | 470 |
Feature activation | +2.24e-02 |
Pos logit contributions | |
Jack | +0.057 |
His | +0.056 |
He | +0.054 |
Kids | +0.053 |
David | +0.052 |
Neg logit contributions | |
els | -0.057 |
bulb | -0.057 |
hole | -0.056 |
globe | -0.056 |
ps | -0.055 |
Token | swim |
Token ID | 9422 |
Feature activation | -1.15e-01 |
Pos logit contributions | |
His | +0.035 |
Jack | +0.035 |
He | +0.034 |
David | +0.032 |
Kids | +0.032 |
Neg logit contributions | |
els | -0.037 |
bulb | -0.037 |
hole | -0.037 |
globe | -0.036 |
ps | -0.036 |
Token | , |
Token ID | 11 |
Feature activation | +3.80e-02 |
Pos logit contributions | |
els | +0.177 |
bulb | +0.166 |
hole | +0.163 |
ening | +0.161 |
ps | +0.161 |
Neg logit contributions | |
He | -0.211 |
His | -0.209 |
Jack | -0.207 |
They | -0.193 |
He | -0.184 |
Token | D |
Token ID | 360 |
Feature activation | -4.74e-02 |
Pos logit contributions | |
His | +0.057 |
Jack | +0.057 |
David | +0.054 |
He | +0.053 |
Kids | +0.051 |
Neg logit contributions | |
bulb | -0.067 |
egg | -0.063 |
globe | -0.063 |
els | -0.063 |
hole | -0.063 |
Token | ucky |
Token ID | 5309 |
Feature activation | -5.92e-02 |
Pos logit contributions | |
ool | +0.077 |
els | +0.077 |
bulb | +0.077 |
hole | +0.074 |
ening | +0.073 |
Neg logit contributions | |
His | -0.078 |
Jack | -0.075 |
He | -0.074 |
David | -0.072 |
Thanks | -0.069 |
Token | . |
Token ID | 13 |
Feature activation | +8.99e-02 |
Pos logit contributions | |
els | +0.092 |
hole | +0.084 |
atop | +0.082 |
globe | +0.080 |
ps | +0.080 |
Neg logit contributions | |
He | -0.106 |
His | -0.102 |
Jack | -0.102 |
He | -0.097 |
They | -0.097 |
Token | I |
Token ID | 314 |
Feature activation | +4.98e-02 |
Pos logit contributions | |
Jack | +0.128 |
His | +0.124 |
David | +0.121 |
He | +0.115 |
Kids | +0.113 |
Neg logit contributions | |
bulb | -0.164 |
trunk | -0.160 |
globe | -0.160 |
hole | -0.160 |
chim | -0.157 |
Token | am |
Token ID | 716 |
Feature activation | +2.73e-02 |
Pos logit contributions | |
Jack | +0.075 |
His | +0.075 |
He | +0.072 |
David | +0.068 |
Kids | +0.068 |
Neg logit contributions | |
els | -0.086 |
bulb | -0.086 |
hole | -0.085 |
globe | -0.084 |
ool | -0.083 |
Token | a |
Token ID | 257 |
Feature activation | -7.71e-03 |
Pos logit contributions | |
He | +0.033 |
His | +0.032 |
please | +0.031 |
Jack | +0.030 |
He | +0.030 |
Neg logit contributions | |
els | -0.059 |
hole | -0.056 |
globe | -0.056 |
ool | -0.056 |
ening | -0.052 |
Token | . |
Token ID | 13 |
Feature activation | +1.16e-01 |
Pos logit contributions | |
els | +0.180 |
cks | +0.164 |
globe | +0.162 |
ps | +0.162 |
bulb | +0.160 |
Neg logit contributions | |
Jack | -0.199 |
He | -0.199 |
His | -0.195 |
They | -0.184 |
She | -0.177 |
Token | She |
Token ID | 1375 |
Feature activation | +5.21e-02 |
Pos logit contributions | |
Jack | +0.110 |
His | +0.101 |
He | +0.093 |
David | +0.093 |
Dad | +0.084 |
Neg logit contributions | |
trunk | -0.273 |
bulb | -0.271 |
chim | -0.270 |
els | -0.269 |
hole | -0.266 |
Token | thought |
Token ID | 1807 |
Feature activation | +1.62e-02 |
Pos logit contributions | |
His | +0.078 |
Jack | +0.078 |
He | +0.076 |
Kids | +0.070 |
David | +0.070 |
Neg logit contributions | |
els | -0.092 |
bulb | -0.090 |
hole | -0.089 |
globe | -0.087 |
ps | -0.087 |
Token | , |
Token ID | 11 |
Feature activation | +5.01e-02 |
Pos logit contributions | |
Jack | +0.021 |
His | +0.021 |
He | +0.021 |
His | +0.019 |
please | +0.019 |
Neg logit contributions | |
els | -0.031 |
globe | -0.031 |
hole | -0.031 |
atop | -0.030 |
ps | -0.029 |
Token | " |
Token ID | 366 |
Feature activation | +5.90e-02 |
Pos logit contributions | |
He | +0.057 |
His | +0.057 |
Jack | +0.053 |
Sure | +0.049 |
They | +0.049 |
Neg logit contributions | |
els | -0.112 |
ool | -0.111 |
globe | -0.108 |
hole | -0.108 |
bulb | -0.106 |
Token | This |
Token ID | 1212 |
Feature activation | -3.39e-02 |
Pos logit contributions | |
His | +0.050 |
Jack | +0.050 |
He | +0.044 |
His | +0.044 |
David | +0.043 |
Neg logit contributions | |
els | -0.137 |
hole | -0.135 |
bulb | -0.135 |
et | -0.134 |
globe | -0.133 |
Token | is |
Token ID | 318 |
Feature activation | +2.01e-02 |
Pos logit contributions | |
els | +0.063 |
hole | +0.059 |
ening | +0.056 |
atop | +0.056 |
angles | +0.056 |
Neg logit contributions | |
He | -0.053 |
His | -0.052 |
Jack | -0.049 |
His | -0.049 |
He | -0.049 |
Token | my |
Token ID | 616 |
Feature activation | -8.28e-03 |
Pos logit contributions | |
Jack | +0.025 |
His | +0.024 |
He | +0.024 |
Thanks | +0.023 |
Un | +0.023 |
Neg logit contributions | |
els | -0.041 |
hole | -0.041 |
ening | -0.039 |
globe | -0.039 |
ool | -0.038 |
Token | chance |
Token ID | 2863 |
Feature activation | -6.20e-02 |
Pos logit contributions | |
els | +0.015 |
hole | +0.014 |
atop | +0.014 |
ening | +0.014 |
cks | +0.014 |
Neg logit contributions | |
Jack | -0.012 |
His | -0.012 |
He | -0.012 |
His | -0.012 |
Something | -0.012 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.09e-01 |
Pos logit contributions | |
els | +0.138 |
globe | +0.127 |
ening | +0.127 |
atop | +0.126 |
ps | +0.125 |
Neg logit contributions | |
Jack | -0.070 |
He | -0.068 |
His | -0.066 |
please | -0.064 |
They | -0.063 |
Token | Kitty |
Token ID | 21168 |
Feature activation | -3.01e-03 |
Pos logit contributions | |
Jack | +0.121 |
His | +0.109 |
David | +0.106 |
He | +0.103 |
Kids | +0.100 |
Neg logit contributions | |
trunk | -0.236 |
els | -0.233 |
bulb | -0.232 |
hole | -0.232 |
chim | -0.229 |
Token | mom |
Token ID | 1995 |
Feature activation | -3.30e-02 |
Pos logit contributions | |
Jack | +0.028 |
His | +0.028 |
He | +0.026 |
Dad | +0.024 |
Sure | +0.023 |
Neg logit contributions | |
ening | -0.060 |
hole | -0.060 |
ool | -0.058 |
ps | -0.057 |
els | -0.057 |
Token | and |
Token ID | 290 |
Feature activation | +4.61e-02 |
Pos logit contributions | |
ps | +0.058 |
els | +0.055 |
globe | +0.055 |
ool | +0.054 |
blades | +0.054 |
Neg logit contributions | |
His | -0.052 |
Jack | -0.052 |
He | -0.051 |
They | -0.048 |
Dad | -0.047 |
Token | saw |
Token ID | 2497 |
Feature activation | -5.01e-03 |
Pos logit contributions | |
He | +0.081 |
His | +0.078 |
Jack | +0.074 |
So | +0.073 |
And | +0.072 |
Neg logit contributions | |
bulb | -0.069 |
box | -0.065 |
hanging | -0.065 |
inside | -0.064 |
els | -0.063 |
Token | a |
Token ID | 257 |
Feature activation | -3.33e-03 |
Pos logit contributions | |
els | +0.010 |
hole | +0.010 |
globe | +0.010 |
ening | +0.010 |
ool | +0.010 |
Neg logit contributions | |
Jack | -0.007 |
His | -0.007 |
He | -0.006 |
Kids | -0.006 |
David | -0.006 |
Token | big |
Token ID | 1263 |
Feature activation | -6.82e-02 |
Pos logit contributions | |
els | +0.006 |
ps | +0.005 |
hole | +0.005 |
ening | +0.005 |
et | +0.005 |
Neg logit contributions | |
Jack | -0.006 |
He | -0.006 |
His | -0.006 |
Dad | -0.005 |
Kids | -0.005 |
Token | light |
Token ID | 1657 |
Feature activation | -1.32e-01 |
Pos logit contributions | |
ps | +0.113 |
els | +0.105 |
hole | +0.104 |
cks | +0.101 |
atop | +0.097 |
Neg logit contributions | |
Jack | -0.117 |
He | -0.108 |
Dad | -0.105 |
please | -0.105 |
Ben | -0.104 |
Token | in |
Token ID | 287 |
Feature activation | -2.86e-02 |
Pos logit contributions | |
ps | +0.182 |
els | +0.173 |
cks | +0.172 |
hole | +0.165 |
angles | +0.158 |
Neg logit contributions | |
He | -0.238 |
Jack | -0.232 |
His | -0.230 |
They | -0.227 |
Kids | -0.218 |
Token | the |
Token ID | 262 |
Feature activation | -2.43e-02 |
Pos logit contributions | |
ool | +0.049 |
hole | +0.049 |
els | +0.048 |
ps | +0.047 |
ening | +0.047 |
Neg logit contributions | |
Jack | -0.044 |
His | -0.044 |
He | -0.044 |
Dad | -0.040 |
They | -0.040 |
Token | sky |
Token ID | 6766 |
Feature activation | -9.78e-02 |
Pos logit contributions | |
els | +0.038 |
ps | +0.036 |
ening | +0.035 |
ool | +0.034 |
hole | +0.034 |
Neg logit contributions | |
He | -0.045 |
His | -0.044 |
Jack | -0.043 |
Dad | -0.041 |
Thanks | -0.041 |
Token | . |
Token ID | 13 |
Feature activation | +1.21e-01 |
Pos logit contributions | |
ps | +0.174 |
els | +0.171 |
ool | +0.169 |
matches | +0.168 |
plank | +0.167 |
Neg logit contributions | |
They | -0.145 |
He | -0.139 |
She | -0.132 |
His | -0.129 |
She | -0.129 |
Token | It |
Token ID | 632 |
Feature activation | +1.12e-02 |
Pos logit contributions | |
His | +0.151 |
He | +0.146 |
Jack | +0.140 |
David | +0.132 |
Thanks | +0.129 |
Neg logit contributions | |
chim | -0.243 |
bulb | -0.237 |
trunk | -0.237 |
ladder | -0.232 |
bulbs | -0.222 |
Token | because |
Token ID | 780 |
Feature activation | +4.26e-02 |
Pos logit contributions | |
bulb | +0.039 |
hole | +0.039 |
ening | +0.039 |
pieces | +0.037 |
els | +0.036 |
Neg logit contributions | |
Jack | -0.063 |
His | -0.063 |
David | -0.061 |
Kids | -0.060 |
Thanks | -0.060 |
Token | he |
Token ID | 339 |
Feature activation | +2.73e-02 |
Pos logit contributions | |
Jack | +0.044 |
His | +0.043 |
Kids | +0.042 |
Sure | +0.042 |
He | +0.041 |
Neg logit contributions | |
bulb | -0.089 |
globe | -0.089 |
els | -0.088 |
pieces | -0.087 |
object | -0.086 |
Token | loved |
Token ID | 6151 |
Feature activation | -2.22e-02 |
Pos logit contributions | |
His | +0.045 |
Jack | +0.044 |
David | +0.043 |
Jim | +0.042 |
He | +0.042 |
Neg logit contributions | |
bulb | -0.041 |
trunk | -0.041 |
object | -0.040 |
et | -0.040 |
ps | -0.040 |
Token | that |
Token ID | 326 |
Feature activation | -6.81e-03 |
Pos logit contributions | |
hole | +0.050 |
els | +0.050 |
bulb | +0.049 |
ool | +0.049 |
globe | +0.048 |
Neg logit contributions | |
His | -0.023 |
Jack | -0.023 |
He | -0.022 |
Kids | -0.020 |
David | -0.019 |
Token | toy |
Token ID | 13373 |
Feature activation | -1.05e-01 |
Pos logit contributions | |
els | +0.011 |
ps | +0.009 |
hole | +0.009 |
cks | +0.009 |
atop | +0.009 |
Neg logit contributions | |
His | -0.014 |
His | -0.013 |
He | -0.013 |
He | -0.013 |
Jack | -0.013 |
Token | car |
Token ID | 1097 |
Feature activation | -1.30e-01 |
Pos logit contributions | |
els | +0.150 |
hole | +0.144 |
bulb | +0.142 |
ool | +0.142 |
ps | +0.140 |
Neg logit contributions | |
His | -0.196 |
Jack | -0.193 |
He | -0.192 |
Kids | -0.175 |
His | -0.174 |
Token | and |
Token ID | 290 |
Feature activation | +4.24e-02 |
Pos logit contributions | |
els | +0.211 |
hole | +0.196 |
bulb | +0.196 |
ool | +0.195 |
atop | +0.193 |
Neg logit contributions | |
His | -0.221 |
He | -0.220 |
Jack | -0.216 |
His | -0.198 |
They | -0.195 |
Token | didn |
Token ID | 1422 |
Feature activation | +1.31e-02 |
Pos logit contributions | |
Jack | +0.054 |
His | +0.053 |
He | +0.051 |
Kids | +0.047 |
David | +0.047 |
Neg logit contributions | |
els | -0.084 |
bulb | -0.083 |
hole | -0.082 |
globe | -0.082 |
ps | -0.081 |
Token | 't |
Token ID | 470 |
Feature activation | +2.65e-02 |
Pos logit contributions | |
Jack | +0.024 |
Kids | +0.024 |
David | +0.024 |
Of | +0.023 |
Brother | +0.023 |
Neg logit contributions | |
ool | -0.017 |
bulb | -0.016 |
els | -0.016 |
ps | -0.016 |
model | -0.015 |
Token | want |
Token ID | 765 |
Feature activation | -2.17e-02 |
Pos logit contributions | |
His | +0.034 |
Jack | +0.034 |
He | +0.033 |
Kids | +0.030 |
Thanks | +0.030 |
Neg logit contributions | |
els | -0.053 |
hole | -0.051 |
bulb | -0.051 |
globe | -0.051 |
ps | -0.050 |
Token | to |
Token ID | 284 |
Feature activation | +8.84e-03 |
Pos logit contributions | |
els | +0.043 |
hole | +0.043 |
globe | +0.043 |
bulb | +0.042 |
atop | +0.042 |
Neg logit contributions | |
His | -0.029 |
Jack | -0.028 |
He | -0.027 |
His | -0.025 |
Thanks | -0.024 |
Token | saw |
Token ID | 2497 |
Feature activation | -6.13e-03 |
Pos logit contributions | |
Jack | +0.086 |
His | +0.086 |
He | +0.085 |
David | +0.083 |
Kids | +0.081 |
Neg logit contributions | |
bulb | -0.079 |
els | -0.076 |
matches | -0.075 |
trunk | -0.074 |
object | -0.073 |
Token | a |
Token ID | 257 |
Feature activation | -6.66e-03 |
Pos logit contributions | |
els | +0.011 |
hole | +0.010 |
bulb | +0.010 |
globe | +0.010 |
ps | +0.010 |
Neg logit contributions | |
His | -0.009 |
Jack | -0.009 |
He | -0.009 |
Kids | -0.008 |
David | -0.008 |
Token | plate |
Token ID | 7480 |
Feature activation | -1.47e-01 |
Pos logit contributions | |
els | +0.010 |
ps | +0.009 |
hole | +0.009 |
ening | +0.009 |
et | +0.009 |
Neg logit contributions | |
Jack | -0.013 |
His | -0.013 |
He | -0.013 |
Dad | -0.012 |
Sure | -0.012 |
Token | on |
Token ID | 319 |
Feature activation | -4.57e-02 |
Pos logit contributions | |
els | +0.246 |
hole | +0.236 |
bulb | +0.233 |
ening | +0.233 |
ps | +0.232 |
Neg logit contributions | |
His | -0.242 |
Jack | -0.238 |
He | -0.238 |
Dad | -0.215 |
They | -0.214 |
Token | the |
Token ID | 262 |
Feature activation | -1.88e-02 |
Pos logit contributions | |
hole | +0.077 |
els | +0.076 |
bulb | +0.073 |
ool | +0.073 |
ening | +0.073 |
Neg logit contributions | |
His | -0.073 |
Jack | -0.073 |
He | -0.071 |
Dad | -0.066 |
Her | -0.065 |
Token | table |
Token ID | 3084 |
Feature activation | -1.31e-01 |
Pos logit contributions | |
els | +0.025 |
ening | +0.023 |
hole | +0.023 |
atop | +0.022 |
ps | +0.022 |
Neg logit contributions | |
His | -0.038 |
Jack | -0.037 |
He | -0.037 |
Dad | -0.035 |
Brother | -0.035 |
Token | . |
Token ID | 13 |
Feature activation | +1.19e-01 |
Pos logit contributions | |
els | +0.178 |
ening | +0.174 |
bulb | +0.170 |
globe | +0.170 |
ps | +0.169 |
Neg logit contributions | |
His | -0.252 |
He | -0.252 |
Jack | -0.242 |
They | -0.235 |
Dad | -0.232 |
Token | She |
Token ID | 1375 |
Feature activation | +5.55e-02 |
Pos logit contributions | |
His | +0.137 |
David | +0.136 |
Jack | +0.130 |
He | +0.128 |
Thanks | +0.124 |
Neg logit contributions | |
bulb | -0.245 |
trunk | -0.239 |
pieces | -0.235 |
edges | -0.234 |
chim | -0.234 |
Token | hoped |
Token ID | 10719 |
Feature activation | +4.43e-03 |
Pos logit contributions | |
His | +0.094 |
Jack | +0.093 |
David | +0.091 |
He | +0.091 |
Kids | +0.090 |
Neg logit contributions | |
els | -0.079 |
bulb | -0.077 |
pieces | -0.076 |
trunk | -0.075 |
egg | -0.075 |
Token | it |
Token ID | 340 |
Feature activation | -3.55e-02 |
Pos logit contributions | |
Jack | +0.007 |
His | +0.007 |
He | +0.006 |
Dad | +0.006 |
Kids | +0.006 |
Neg logit contributions | |
els | -0.008 |
hole | -0.008 |
globe | -0.008 |
bulb | -0.008 |
ening | -0.008 |
Token | was |
Token ID | 373 |
Feature activation | +4.02e-03 |
Pos logit contributions | |
hole | +0.079 |
els | +0.078 |
ening | +0.077 |
ps | +0.076 |
ool | +0.076 |
Neg logit contributions | |
His | -0.039 |
He | -0.038 |
Jack | -0.035 |
Sure | -0.034 |
They | -0.032 |
Token | want |
Token ID | 765 |
Feature activation | -1.12e-02 |
Pos logit contributions | |
Jack | +0.054 |
He | +0.054 |
His | +0.052 |
Thanks | +0.047 |
They | +0.046 |
Neg logit contributions | |
els | -0.120 |
hole | -0.115 |
globe | -0.113 |
ening | -0.113 |
ool | -0.112 |
Token | to |
Token ID | 284 |
Feature activation | +1.36e-02 |
Pos logit contributions | |
els | +0.025 |
hole | +0.025 |
globe | +0.024 |
ool | +0.024 |
atop | +0.024 |
Neg logit contributions | |
Jack | -0.013 |
His | -0.012 |
He | -0.011 |
They | -0.010 |
David | -0.010 |
Token | go |
Token ID | 467 |
Feature activation | -4.34e-02 |
Pos logit contributions | |
His | +0.016 |
Jack | +0.016 |
He | +0.015 |
They | +0.014 |
Her | +0.014 |
Neg logit contributions | |
els | -0.028 |
hole | -0.028 |
ening | -0.028 |
globe | -0.027 |
ool | -0.027 |
Token | see |
Token ID | 766 |
Feature activation | -3.58e-02 |
Pos logit contributions | |
els | +0.068 |
ool | +0.065 |
globe | +0.062 |
ening | +0.062 |
bulb | +0.061 |
Neg logit contributions | |
He | -0.077 |
His | -0.075 |
Jack | -0.074 |
They | -0.073 |
please | -0.071 |
Token | a |
Token ID | 257 |
Feature activation | -2.25e-02 |
Pos logit contributions | |
ool | +0.074 |
globe | +0.073 |
ps | +0.071 |
els | +0.070 |
hole | +0.070 |
Neg logit contributions | |
Jack | -0.045 |
He | -0.043 |
They | -0.041 |
His | -0.041 |
please | -0.039 |
Token | rainbow |
Token ID | 27223 |
Feature activation | -1.67e-01 |
Pos logit contributions | |
ps | +0.040 |
els | +0.040 |
aces | +0.038 |
cks | +0.038 |
hole | +0.035 |
Neg logit contributions | |
Jack | -0.040 |
Sure | -0.038 |
He | -0.036 |
Johnny | -0.036 |
Kids | -0.036 |
Token | !" |
Token ID | 2474 |
Feature activation | +1.10e-01 |
Pos logit contributions | |
els | +0.247 |
ps | +0.233 |
cks | +0.223 |
ening | +0.219 |
hole | +0.215 |
Neg logit contributions | |
Jack | -0.285 |
They | -0.284 |
He | -0.281 |
please | -0.277 |
His | -0.261 |
Token |
|
Token ID | 198 |
Feature activation | +9.38e-02 |
Pos logit contributions | |
Jack | +0.121 |
Kids | +0.120 |
His | +0.114 |
Sorry | +0.113 |
Un | +0.112 |
Neg logit contributions | |
els | -0.207 |
hole | -0.206 |
pieces | -0.204 |
globe | -0.203 |
trunk | -0.203 |
Token | Grand |
Token ID | 23581 |
Feature activation | -2.35e-03 |
Pos logit contributions | |
Kids | +0.094 |
Sorry | +0.093 |
Thanks | +0.087 |
Of | +0.085 |
Something | +0.084 |
Neg logit contributions | |
globe | -0.165 |
atop | -0.164 |
els | -0.163 |
object | -0.163 |
ool | -0.163 |
Token | pa |
Token ID | 8957 |
Feature activation | -1.58e-02 |
Pos logit contributions | |
els | +0.003 |
hole | +0.003 |
bulb | +0.003 |
globe | +0.003 |
ps | +0.003 |
Neg logit contributions | |
His | -0.004 |
Jack | -0.004 |
He | -0.004 |
Kids | -0.004 |
Thanks | -0.004 |
Token | gave |
Token ID | 2921 |
Feature activation | -2.50e-02 |
Pos logit contributions | |
els | +0.033 |
hole | +0.032 |
ool | +0.032 |
globe | +0.032 |
blades | +0.032 |
Neg logit contributions | |
Jack | -0.020 |
His | -0.019 |
He | -0.018 |
Jim | -0.016 |
Sure | -0.016 |
Token | , |
Token ID | 11 |
Feature activation | +7.53e-02 |
Pos logit contributions | |
Jack | +0.023 |
His | +0.023 |
He | +0.022 |
Kids | +0.021 |
Dad | +0.021 |
Neg logit contributions | |
els | -0.014 |
hole | -0.014 |
bulb | -0.014 |
globe | -0.013 |
ps | -0.013 |
Token | Tom |
Token ID | 4186 |
Feature activation | -2.18e-03 |
Pos logit contributions | |
His | +0.097 |
He | +0.096 |
Jack | +0.093 |
Brother | +0.086 |
Kids | +0.086 |
Neg logit contributions | |
els | -0.143 |
bulb | -0.138 |
globe | -0.138 |
hole | -0.137 |
pieces | -0.137 |
Token | saw |
Token ID | 2497 |
Feature activation | +5.23e-04 |
Pos logit contributions | |
els | +0.003 |
hole | +0.003 |
bulb | +0.003 |
globe | +0.003 |
ening | +0.003 |
Neg logit contributions | |
His | -0.004 |
Jack | -0.004 |
He | -0.004 |
Thanks | -0.003 |
Kids | -0.003 |
Token | a |
Token ID | 257 |
Feature activation | -6.62e-03 |
Pos logit contributions | |
Jack | +0.001 |
His | +0.001 |
He | +0.001 |
Kids | +0.001 |
Thanks | +0.001 |
Neg logit contributions | |
els | -0.001 |
bulb | -0.001 |
matches | -0.001 |
pieces | -0.001 |
ps | -0.001 |
Token | new |
Token ID | 649 |
Feature activation | -8.45e-02 |
Pos logit contributions | |
ps | +0.012 |
els | +0.011 |
cks | +0.010 |
hole | +0.010 |
ening | +0.010 |
Neg logit contributions | |
Jack | -0.012 |
He | -0.011 |
His | -0.011 |
Johnny | -0.010 |
Un | -0.010 |
Token | car |
Token ID | 1097 |
Feature activation | -1.40e-01 |
Pos logit contributions | |
els | +0.142 |
cks | +0.140 |
ps | +0.139 |
hole | +0.129 |
ening | +0.128 |
Neg logit contributions | |
His | -0.145 |
Jack | -0.137 |
He | -0.137 |
His | -0.133 |
Sure | -0.129 |
Token | that |
Token ID | 326 |
Feature activation | +2.69e-03 |
Pos logit contributions | |
els | +0.176 |
ool | +0.170 |
ps | +0.168 |
bulbs | +0.167 |
ening | +0.166 |
Neg logit contributions | |
His | -0.286 |
He | -0.284 |
Jack | -0.264 |
They | -0.261 |
He | -0.249 |
Token | Lily |
Token ID | 20037 |
Feature activation | -3.15e-02 |
Pos logit contributions | |
His | +0.004 |
Jack | +0.003 |
He | +0.003 |
Thanks | +0.003 |
His | +0.003 |
Neg logit contributions | |
els | -0.006 |
ening | -0.005 |
hole | -0.005 |
ps | -0.005 |
ool | -0.005 |
Token | had |
Token ID | 550 |
Feature activation | +7.68e-03 |
Pos logit contributions | |
els | +0.068 |
ps | +0.067 |
ool | +0.066 |
bulb | +0.066 |
globe | +0.066 |
Neg logit contributions | |
He | -0.037 |
Jack | -0.036 |
His | -0.036 |
David | -0.030 |
They | -0.030 |
Token | . |
Token ID | 13 |
Feature activation | +1.14e-01 |
Pos logit contributions | |
His | +0.010 |
He | +0.010 |
Jack | +0.009 |
They | +0.009 |
Thanks | +0.009 |
Neg logit contributions | |
els | -0.016 |
globe | -0.016 |
ool | -0.015 |
bulb | -0.015 |
ening | -0.015 |
Token | It |
Token ID | 632 |
Feature activation | +2.07e-03 |
Pos logit contributions | |
Jack | +0.113 |
His | +0.105 |
Kids | +0.102 |
Brother | +0.100 |
David | +0.098 |
Neg logit contributions | |
trunk | -0.248 |
hole | -0.245 |
object | -0.242 |
ool | -0.242 |
pieces | -0.241 |