Token | and |
Token ID | 290 |
Feature activation | +2.89e-03 |
Pos logit contributions | |
he | +0.027 |
olded | +0.027 |
sick | +0.027 |
punished | +0.026 |
elderly | +0.026 |
Neg logit contributions | |
moment | -0.039 |
tweet | -0.039 |
Nice | -0.036 |
Want | -0.035 |
fly | -0.034 |
Token | thanked |
Token ID | 26280 |
Feature activation | +1.96e-02 |
Pos logit contributions | |
he | +0.003 |
olded | +0.003 |
outs | +0.003 |
punished | +0.003 |
obb | +0.003 |
Neg logit contributions | |
moment | -0.006 |
tweet | -0.006 |
Nice | -0.005 |
Want | -0.005 |
feast | -0.005 |
Token | the |
Token ID | 262 |
Feature activation | +1.35e-02 |
Pos logit contributions | |
olded | +0.026 |
he | +0.025 |
obb | +0.025 |
punished | +0.025 |
outs | +0.024 |
Neg logit contributions | |
moment | -0.034 |
tweet | -0.034 |
Nice | -0.030 |
Want | -0.029 |
feast | -0.029 |
Token | parent |
Token ID | 2560 |
Feature activation | +4.99e-03 |
Pos logit contributions | |
he | +0.024 |
sick | +0.024 |
elderly | +0.024 |
olded | +0.024 |
punished | +0.024 |
Neg logit contributions | |
moment | -0.015 |
tweet | -0.015 |
Nice | -0.014 |
Want | -0.013 |
fly | -0.012 |
Token | before |
Token ID | 878 |
Feature activation | +2.00e-02 |
Pos logit contributions | |
obb | +0.007 |
punished | +0.007 |
olded | +0.007 |
he | +0.007 |
sick | +0.007 |
Neg logit contributions | |
moment | -0.008 |
tweet | -0.007 |
Nice | -0.007 |
Want | -0.007 |
fly | -0.007 |
Token | sc |
Token ID | 629 |
Feature activation | +9.31e-02 |
Pos logit contributions | |
olded | +0.032 |
he | +0.031 |
gr | +0.031 |
punished | +0.031 |
obb | +0.031 |
Neg logit contributions | |
moment | -0.029 |
tweet | -0.027 |
Nice | -0.025 |
feast | -0.024 |
Want | -0.024 |
Token | ur |
Token ID | 333 |
Feature activation | +6.27e-03 |
Pos logit contributions | |
pastor | +0.123 |
elderly | +0.122 |
he | +0.118 |
sick | +0.117 |
gr | +0.116 |
Neg logit contributions | |
tweet | -0.149 |
Nice | -0.131 |
feast | -0.131 |
moment | -0.130 |
Want | -0.127 |
Token | rying |
Token ID | 14992 |
Feature activation | +3.49e-04 |
Pos logit contributions | |
he | +0.009 |
olded | +0.009 |
sick | +0.009 |
punished | +0.009 |
elderly | +0.009 |
Neg logit contributions | |
moment | -0.009 |
tweet | -0.009 |
Nice | -0.009 |
Want | -0.008 |
fly | -0.008 |
Token | away |
Token ID | 1497 |
Feature activation | -4.68e-04 |
Pos logit contributions | |
olded | +0.001 |
elderly | +0.001 |
pastor | +0.001 |
salon | +0.001 |
obb | +0.001 |
Neg logit contributions | |
moment | -0.000 |
tweet | -0.000 |
est | -0.000 |
Nice | -0.000 |
feast | -0.000 |
Token | . |
Token ID | 13 |
Feature activation | +3.98e-03 |
Pos logit contributions | |
moment | +0.001 |
tweet | +0.001 |
Nice | +0.001 |
fly | +0.001 |
greeting | +0.001 |
Neg logit contributions | |
obb | -0.001 |
he | -0.001 |
leaf | -0.001 |
olded | -0.001 |
punished | -0.001 |
Token | |
Token ID | 220 |
Feature activation | +4.92e-02 |
Pos logit contributions | |
punished | +0.005 |
he | +0.005 |
sick | +0.005 |
obb | +0.005 |
olded | +0.005 |
Neg logit contributions | |
moment | -0.006 |
tweet | -0.006 |
Nice | -0.006 |
Want | -0.006 |
fly | -0.005 |
Token | ball |
Token ID | 2613 |
Feature activation | -1.63e-03 |
Pos logit contributions | |
moment | +0.017 |
tweet | +0.017 |
Nice | +0.016 |
Want | +0.015 |
fly | +0.015 |
Neg logit contributions | |
he | -0.013 |
sick | -0.013 |
olded | -0.013 |
punished | -0.013 |
elderly | -0.013 |
Token | at |
Token ID | 379 |
Feature activation | +1.12e-02 |
Pos logit contributions | |
moment | +0.002 |
tweet | +0.002 |
fly | +0.002 |
feast | +0.002 |
Want | +0.002 |
Neg logit contributions | |
olded | -0.003 |
leaf | -0.003 |
pastor | -0.003 |
elderly | -0.003 |
iza | -0.003 |
Token | you |
Token ID | 345 |
Feature activation | +1.15e-02 |
Pos logit contributions | |
olded | +0.019 |
leaf | +0.018 |
punished | +0.018 |
pastor | +0.018 |
gr | +0.017 |
Neg logit contributions | |
moment | -0.015 |
tweet | -0.014 |
Nice | -0.012 |
Want | -0.012 |
feast | -0.012 |
Token | ?" |
Token ID | 1701 |
Feature activation | +5.04e-02 |
Pos logit contributions | |
olded | +0.018 |
leaf | +0.018 |
gr | +0.017 |
elderly | +0.017 |
obb | +0.017 |
Neg logit contributions | |
moment | -0.016 |
tweet | -0.016 |
Nice | -0.014 |
Want | -0.014 |
fly | -0.014 |
Token | she |
Token ID | 673 |
Feature activation | +4.73e-02 |
Pos logit contributions | |
leaf | +0.087 |
olded | +0.080 |
cream | +0.080 |
punished | +0.078 |
he | +0.078 |
Neg logit contributions | |
moment | -0.071 |
tweet | -0.064 |
hop | -0.057 |
Nice | -0.057 |
Want | -0.054 |
Token | sc |
Token ID | 629 |
Feature activation | +8.87e-02 |
Pos logit contributions | |
he | +0.054 |
leaf | +0.052 |
obb | +0.048 |
olded | +0.047 |
earn | +0.045 |
Neg logit contributions | |
moment | -0.087 |
tweet | -0.086 |
fly | -0.077 |
Nice | -0.077 |
Want | -0.076 |
Token | olded |
Token ID | 48959 |
Feature activation | +3.61e-02 |
Pos logit contributions | |
pastor | +0.067 |
elderly | +0.065 |
sick | +0.059 |
he | +0.057 |
late | +0.055 |
Neg logit contributions | |
tweet | -0.194 |
Nice | -0.180 |
fly | -0.175 |
moment | -0.174 |
Want | -0.173 |
Token | . |
Token ID | 13 |
Feature activation | +3.91e-03 |
Pos logit contributions | |
leaf | +0.051 |
olded | +0.051 |
punished | +0.049 |
pastor | +0.047 |
gr | +0.047 |
Neg logit contributions | |
moment | -0.058 |
tweet | -0.054 |
Nice | -0.052 |
est | -0.048 |
fly | -0.047 |
Token |
|
Token ID | 198 |
Feature activation | -5.84e-03 |
Pos logit contributions | |
punished | +0.007 |
leaf | +0.006 |
olded | +0.006 |
gr | +0.006 |
he | +0.006 |
Neg logit contributions | |
moment | -0.005 |
tweet | -0.005 |
Nice | -0.005 |
Want | -0.004 |
greeting | -0.004 |
Token |
|
Token ID | 198 |
Feature activation | -1.28e-02 |
Pos logit contributions | |
moment | +0.007 |
tweet | +0.007 |
Nice | +0.006 |
Want | +0.006 |
dy | +0.005 |
Neg logit contributions | |
punished | -0.010 |
olded | -0.010 |
leaf | -0.010 |
pastor | -0.010 |
gr | -0.010 |
Token | L |
Token ID | 43 |
Feature activation | +1.04e-02 |
Pos logit contributions | |
moment | +0.019 |
tweet | +0.017 |
dy | +0.015 |
Nice | +0.015 |
Want | +0.015 |
Neg logit contributions | |
leaf | -0.020 |
outs | -0.019 |
olded | -0.019 |
pastor | -0.018 |
he | -0.018 |
Token | and |
Token ID | 290 |
Feature activation | -2.02e-03 |
Pos logit contributions | |
punished | +0.018 |
olded | +0.018 |
he | +0.018 |
leaf | +0.018 |
sick | +0.018 |
Neg logit contributions | |
moment | -0.014 |
tweet | -0.013 |
Nice | -0.013 |
est | -0.012 |
greeting | -0.012 |
Token | waved |
Token ID | 26834 |
Feature activation | +1.24e-03 |
Pos logit contributions | |
moment | +0.003 |
tweet | +0.003 |
Nice | +0.003 |
Want | +0.003 |
elcome | +0.003 |
Neg logit contributions | |
he | -0.003 |
obb | -0.003 |
outs | -0.003 |
gr | -0.003 |
pastor | -0.003 |
Token | goodbye |
Token ID | 24829 |
Feature activation | +2.74e-02 |
Pos logit contributions | |
he | +0.002 |
punished | +0.002 |
olded | +0.002 |
sick | +0.002 |
leaf | +0.002 |
Neg logit contributions | |
moment | -0.002 |
tweet | -0.002 |
Nice | -0.001 |
est | -0.001 |
greeting | -0.001 |
Token | as |
Token ID | 355 |
Feature activation | +1.16e-02 |
Pos logit contributions | |
ateful | +0.050 |
he | +0.050 |
pastor | +0.049 |
leaf | +0.049 |
punished | +0.049 |
Neg logit contributions | |
moment | -0.035 |
tweet | -0.030 |
fly | -0.027 |
Nice | -0.027 |
greeting | -0.026 |
Token | Toby |
Token ID | 37578 |
Feature activation | +6.06e-04 |
Pos logit contributions | |
olded | +0.022 |
outs | +0.021 |
he | +0.021 |
outed | +0.020 |
gr | +0.020 |
Neg logit contributions | |
moment | -0.016 |
Nice | -0.014 |
tweet | -0.013 |
Want | -0.012 |
hop | -0.012 |
Token | sc |
Token ID | 629 |
Feature activation | +8.86e-02 |
Pos logit contributions | |
ateful | +0.001 |
gr | +0.001 |
he | +0.001 |
sick | +0.001 |
outs | +0.001 |
Neg logit contributions | |
moment | -0.001 |
tweet | -0.001 |
feast | -0.001 |
est | -0.001 |
dy | -0.001 |
Token | ur |
Token ID | 333 |
Feature activation | +6.89e-03 |
Pos logit contributions | |
pastor | +0.192 |
sick | +0.184 |
ateful | +0.179 |
bored | +0.174 |
elderly | +0.171 |
Neg logit contributions | |
tweet | -0.081 |
Nice | -0.066 |
feast | -0.059 |
elcome | -0.057 |
Want | -0.056 |
Token | ried |
Token ID | 2228 |
Feature activation | -7.34e-03 |
Pos logit contributions | |
he | +0.011 |
sick | +0.011 |
elderly | +0.011 |
olded | +0.011 |
pastor | +0.011 |
Neg logit contributions | |
moment | -0.009 |
tweet | -0.009 |
fly | -0.008 |
Nice | -0.008 |
est | -0.008 |
Token | away |
Token ID | 1497 |
Feature activation | +4.91e-04 |
Pos logit contributions | |
moment | +0.011 |
tweet | +0.010 |
Nice | +0.010 |
est | +0.009 |
greeting | +0.009 |
Neg logit contributions | |
olded | -0.011 |
sick | -0.011 |
he | -0.010 |
punished | -0.010 |
elderly | -0.010 |
Token | . |
Token ID | 13 |
Feature activation | +1.34e-03 |
Pos logit contributions | |
he | +0.001 |
leaf | +0.001 |
obb | +0.001 |
olded | +0.001 |
pastor | +0.001 |
Neg logit contributions | |
moment | -0.001 |
tweet | -0.001 |
Nice | -0.001 |
fly | -0.001 |
greeting | -0.001 |
Token | He |
Token ID | 679 |
Feature activation | -2.24e-03 |
Pos logit contributions | |
he | +0.002 |
olded | +0.002 |
sick | +0.002 |
punished | +0.002 |
elderly | +0.002 |
Neg logit contributions | |
moment | -0.002 |
tweet | -0.002 |
Nice | -0.002 |
Want | -0.002 |
fly | -0.002 |
Token | ! |
Token ID | 0 |
Feature activation | +6.75e-04 |
Pos logit contributions | |
he | +0.027 |
sick | +0.026 |
punished | +0.026 |
olded | +0.026 |
elderly | +0.025 |
Neg logit contributions | |
moment | -0.020 |
tweet | -0.019 |
Nice | -0.018 |
Want | -0.017 |
greeting | -0.016 |
Token | Bad |
Token ID | 7772 |
Feature activation | +2.37e-04 |
Pos logit contributions | |
he | +0.001 |
punished | +0.001 |
obb | +0.001 |
olded | +0.001 |
outs | +0.001 |
Neg logit contributions | |
tweet | -0.001 |
moment | -0.001 |
Nice | -0.001 |
Want | -0.001 |
fly | -0.001 |
Token | dog |
Token ID | 3290 |
Feature activation | +3.04e-03 |
Pos logit contributions | |
sick | +0.000 |
he | +0.000 |
elderly | +0.000 |
punished | +0.000 |
pastor | +0.000 |
Neg logit contributions | |
moment | -0.000 |
tweet | -0.000 |
Nice | -0.000 |
Want | -0.000 |
greeting | -0.000 |
Token | !" |
Token ID | 2474 |
Feature activation | +4.17e-02 |
Pos logit contributions | |
he | +0.005 |
sick | +0.005 |
olded | +0.005 |
punished | +0.005 |
elderly | +0.005 |
Neg logit contributions | |
moment | -0.004 |
tweet | -0.004 |
Nice | -0.004 |
Want | -0.004 |
greeting | -0.003 |
Token | they |
Token ID | 484 |
Feature activation | +4.39e-02 |
Pos logit contributions | |
he | +0.076 |
obb | +0.075 |
gr | +0.074 |
ussy | +0.073 |
leaf | +0.072 |
Neg logit contributions | |
moment | -0.048 |
tweet | -0.043 |
Nice | -0.042 |
Want | -0.039 |
Bella | -0.038 |
Token | sc |
Token ID | 629 |
Feature activation | +8.85e-02 |
Pos logit contributions | |
leaf | +0.049 |
obb | +0.049 |
pastor | +0.047 |
he | +0.047 |
earn | +0.046 |
Neg logit contributions | |
tweet | -0.078 |
moment | -0.073 |
Nice | -0.067 |
greeting | -0.067 |
elcome | -0.067 |
Token | olded |
Token ID | 48959 |
Feature activation | +3.62e-02 |
Pos logit contributions | |
pastor | +0.060 |
elderly | +0.054 |
sick | +0.050 |
he | +0.049 |
obb | +0.045 |
Neg logit contributions | |
tweet | -0.200 |
Nice | -0.186 |
moment | -0.181 |
fly | -0.180 |
Want | -0.178 |
Token | Max |
Token ID | 5436 |
Feature activation | -1.88e-02 |
Pos logit contributions | |
leaf | +0.050 |
olded | +0.050 |
punished | +0.048 |
salon | +0.048 |
pastor | +0.048 |
Neg logit contributions | |
moment | -0.058 |
tweet | -0.054 |
Nice | -0.050 |
fly | -0.048 |
est | -0.047 |
Token | . |
Token ID | 13 |
Feature activation | +1.10e-03 |
Pos logit contributions | |
moment | +0.034 |
tweet | +0.033 |
Nice | +0.031 |
fly | +0.030 |
dy | +0.029 |
Neg logit contributions | |
obb | -0.022 |
leaf | -0.021 |
he | -0.021 |
pastor | -0.020 |
olded | -0.020 |
Token |
|
Token ID | 198 |
Feature activation | -7.62e-03 |
Pos logit contributions | |
punished | +0.002 |
leaf | +0.002 |
olded | +0.002 |
he | +0.002 |
obb | +0.002 |
Neg logit contributions | |
moment | -0.001 |
tweet | -0.001 |
Nice | -0.001 |
Want | -0.001 |
fly | -0.001 |
Token |
|
Token ID | 198 |
Feature activation | -1.33e-02 |
Pos logit contributions | |
moment | +0.010 |
tweet | +0.010 |
Nice | +0.009 |
Want | +0.009 |
fly | +0.008 |
Neg logit contributions | |
olded | -0.012 |
punished | -0.012 |
he | -0.012 |
sick | -0.012 |
obb | -0.012 |
Token | tered |
Token ID | 4400 |
Feature activation | -1.16e-02 |
Pos logit contributions | |
moment | +0.009 |
tweet | +0.009 |
Nice | +0.008 |
est | +0.008 |
Want | +0.008 |
Neg logit contributions | |
sick | -0.007 |
punished | -0.007 |
olded | -0.007 |
elderly | -0.007 |
he | -0.007 |
Token | at |
Token ID | 379 |
Feature activation | +9.71e-03 |
Pos logit contributions | |
moment | +0.019 |
tweet | +0.019 |
Nice | +0.018 |
Want | +0.017 |
fly | +0.017 |
Neg logit contributions | |
he | -0.015 |
olded | -0.015 |
sick | -0.015 |
punished | -0.015 |
elderly | -0.014 |
Token | the |
Token ID | 262 |
Feature activation | -7.18e-05 |
Pos logit contributions | |
olded | +0.011 |
he | +0.011 |
punished | +0.011 |
sick | +0.011 |
leaf | +0.011 |
Neg logit contributions | |
moment | -0.017 |
tweet | -0.017 |
Nice | -0.016 |
Want | -0.015 |
fly | -0.015 |
Token | hunter |
Token ID | 19177 |
Feature activation | +2.32e-03 |
Pos logit contributions | |
moment | +0.000 |
tweet | +0.000 |
Nice | +0.000 |
est | +0.000 |
Want | +0.000 |
Neg logit contributions | |
olded | -0.000 |
he | -0.000 |
outs | -0.000 |
punished | -0.000 |
iza | -0.000 |
Token | before |
Token ID | 878 |
Feature activation | +2.38e-02 |
Pos logit contributions | |
he | +0.003 |
olded | +0.003 |
sick | +0.003 |
punished | +0.003 |
elderly | +0.003 |
Neg logit contributions | |
moment | -0.003 |
tweet | -0.003 |
Nice | -0.003 |
Want | -0.003 |
fly | -0.003 |
Token | sc |
Token ID | 629 |
Feature activation | +8.83e-02 |
Pos logit contributions | |
olded | +0.039 |
gr | +0.038 |
punished | +0.036 |
he | +0.036 |
outs | +0.036 |
Neg logit contributions | |
moment | -0.035 |
tweet | -0.033 |
Nice | -0.030 |
feast | -0.029 |
greeting | -0.029 |
Token | ur |
Token ID | 333 |
Feature activation | +3.12e-03 |
Pos logit contributions | |
pastor | +0.154 |
elderly | +0.151 |
sick | +0.144 |
he | +0.143 |
gr | +0.143 |
Neg logit contributions | |
tweet | -0.111 |
Nice | -0.095 |
moment | -0.090 |
feast | -0.088 |
Want | -0.088 |
Token | rying |
Token ID | 14992 |
Feature activation | +7.07e-04 |
Pos logit contributions | |
he | +0.005 |
olded | +0.004 |
sick | +0.004 |
elderly | +0.004 |
punished | +0.004 |
Neg logit contributions | |
moment | -0.005 |
tweet | -0.005 |
Nice | -0.004 |
fly | -0.004 |
Want | -0.004 |
Token | away |
Token ID | 1497 |
Feature activation | -1.17e-03 |
Pos logit contributions | |
he | +0.001 |
olded | +0.001 |
elderly | +0.001 |
pastor | +0.001 |
iza | +0.001 |
Neg logit contributions | |
moment | -0.001 |
tweet | -0.001 |
Nice | -0.001 |
est | -0.001 |
greeting | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | +4.63e-03 |
Pos logit contributions | |
moment | +0.002 |
tweet | +0.002 |
Nice | +0.002 |
fly | +0.002 |
greeting | +0.002 |
Neg logit contributions | |
he | -0.002 |
olded | -0.001 |
leaf | -0.001 |
obb | -0.001 |
punished | -0.001 |
Token | |
Token ID | 220 |
Feature activation | +4.54e-02 |
Pos logit contributions | |
punished | +0.007 |
sick | +0.007 |
he | +0.007 |
obb | +0.007 |
pastor | +0.007 |
Neg logit contributions | |
moment | -0.006 |
tweet | -0.006 |
Nice | -0.006 |
Want | -0.006 |
greeting | -0.005 |
Token | or |
Token ID | 393 |
Feature activation | +5.76e-03 |
Pos logit contributions | |
leaf | +0.026 |
olded | +0.024 |
pastor | +0.024 |
blamed | +0.023 |
outed | +0.023 |
Neg logit contributions | |
moment | -0.016 |
tweet | -0.012 |
est | -0.012 |
, | -0.012 |
feast | -0.011 |
Token | your |
Token ID | 534 |
Feature activation | +1.31e-02 |
Pos logit contributions | |
he | +0.009 |
obb | +0.008 |
outs | +0.008 |
olded | +0.008 |
leaf | +0.008 |
Neg logit contributions | |
moment | -0.008 |
tweet | -0.008 |
Want | -0.007 |
feast | -0.007 |
Nice | -0.007 |
Token | candy |
Token ID | 18550 |
Feature activation | +1.42e-02 |
Pos logit contributions | |
he | +0.020 |
olded | +0.020 |
leaf | +0.019 |
punished | +0.019 |
obb | +0.019 |
Neg logit contributions | |
moment | -0.019 |
tweet | -0.018 |
feast | -0.017 |
Nice | -0.016 |
Want | -0.016 |
Token | ?" |
Token ID | 1701 |
Feature activation | +5.02e-02 |
Pos logit contributions | |
leaf | +0.027 |
olded | +0.024 |
pastor | +0.023 |
blamed | +0.023 |
year | +0.023 |
Neg logit contributions | |
moment | -0.017 |
est | -0.014 |
feast | -0.014 |
fly | -0.014 |
tweet | -0.013 |
Token | she |
Token ID | 673 |
Feature activation | +4.85e-02 |
Pos logit contributions | |
leaf | +0.096 |
rou | +0.091 |
obb | +0.089 |
ussy | +0.087 |
rolley | +0.086 |
Neg logit contributions | |
moment | -0.060 |
Nice | -0.047 |
, | -0.047 |
tweet | -0.046 |
inspection | -0.045 |
Token | sc |
Token ID | 629 |
Feature activation | +8.76e-02 |
Pos logit contributions | |
he | +0.063 |
leaf | +0.060 |
obb | +0.058 |
out | +0.054 |
earn | +0.053 |
Neg logit contributions | |
moment | -0.082 |
tweet | -0.082 |
Want | -0.072 |
greeting | -0.072 |
Nice | -0.071 |
Token | olded |
Token ID | 48959 |
Feature activation | +3.52e-02 |
Pos logit contributions | |
pastor | +0.073 |
elderly | +0.068 |
late | +0.065 |
obb | +0.065 |
sick | +0.065 |
Neg logit contributions | |
tweet | -0.186 |
fly | -0.168 |
Nice | -0.166 |
Want | -0.164 |
immer | -0.161 |
Token | him |
Token ID | 683 |
Feature activation | -9.56e-06 |
Pos logit contributions | |
leaf | +0.054 |
olded | +0.053 |
punished | +0.051 |
gr | +0.050 |
obb | +0.049 |
Neg logit contributions | |
moment | -0.052 |
tweet | -0.048 |
Nice | -0.043 |
est | -0.041 |
, | -0.040 |
Token | . |
Token ID | 13 |
Feature activation | +1.59e-03 |
Token |
|
Token ID | 198 |
Feature activation | -7.12e-03 |
Pos logit contributions | |
punished | +0.003 |
olded | +0.003 |
leaf | +0.003 |
sick | +0.003 |
obb | +0.003 |
Neg logit contributions | |
moment | -0.002 |
tweet | -0.002 |
Nice | -0.002 |
Want | -0.002 |
greeting | -0.002 |
Token |
|
Token ID | 198 |
Feature activation | -1.29e-02 |
Pos logit contributions | |
moment | +0.008 |
tweet | +0.008 |
Nice | +0.007 |
Want | +0.007 |
greeting | +0.006 |
Neg logit contributions | |
punished | -0.013 |
olded | -0.012 |
leaf | -0.012 |
sick | -0.012 |
obb | -0.012 |
Token | you |
Token ID | 345 |
Feature activation | +2.50e-02 |
Pos logit contributions | |
he | +0.015 |
olded | +0.014 |
leaf | +0.014 |
pastor | +0.013 |
iza | +0.013 |
Neg logit contributions | |
moment | -0.018 |
tweet | -0.018 |
Want | -0.016 |
Nice | -0.016 |
fly | -0.016 |
Token | might |
Token ID | 1244 |
Feature activation | +1.23e-02 |
Pos logit contributions | |
pastor | +0.048 |
ateful | +0.047 |
leaf | +0.047 |
elderly | +0.046 |
olded | +0.046 |
Neg logit contributions | |
tweet | -0.027 |
moment | -0.025 |
fly | -0.022 |
feast | -0.021 |
hop | -0.021 |
Token | fall |
Token ID | 2121 |
Feature activation | +8.81e-03 |
Pos logit contributions | |
olded | +0.017 |
leaf | +0.016 |
he | +0.016 |
outs | +0.016 |
elderly | +0.016 |
Neg logit contributions | |
moment | -0.020 |
tweet | -0.019 |
Nice | -0.018 |
Want | -0.017 |
fly | -0.017 |
Token | ," |
Token ID | 553 |
Feature activation | +4.58e-02 |
Pos logit contributions | |
olded | +0.015 |
leaf | +0.015 |
iza | +0.014 |
gr | +0.013 |
pastor | +0.013 |
Neg logit contributions | |
moment | -0.013 |
tweet | -0.012 |
Nice | -0.011 |
fly | -0.011 |
est | -0.011 |
Token | his |
Token ID | 465 |
Feature activation | +5.43e-02 |
Pos logit contributions | |
he | +0.062 |
obb | +0.062 |
olded | +0.058 |
outs | +0.058 |
punished | +0.057 |
Neg logit contributions | |
moment | -0.078 |
tweet | -0.077 |
Nice | -0.068 |
fly | -0.066 |
greeting | -0.066 |
Token | m |
Token ID | 285 |
Feature activation | +8.61e-02 |
Pos logit contributions | |
he | +0.065 |
olded | +0.064 |
sick | +0.063 |
punished | +0.062 |
elderly | +0.062 |
Neg logit contributions | |
moment | -0.094 |
tweet | -0.092 |
Nice | -0.087 |
Want | -0.084 |
fly | -0.082 |
Token | ama |
Token ID | 1689 |
Feature activation | +2.48e-02 |
Pos logit contributions | |
leaf | +0.142 |
punished | +0.141 |
sick | +0.140 |
he | +0.140 |
bored | +0.138 |
Neg logit contributions | |
tweet | -0.105 |
moment | -0.100 |
fly | -0.097 |
Nice | -0.095 |
elcome | -0.089 |
Token | said |
Token ID | 531 |
Feature activation | -1.02e-02 |
Pos logit contributions | |
obb | +0.025 |
he | +0.025 |
sick | +0.024 |
punished | +0.024 |
earn | +0.024 |
Neg logit contributions | |
moment | -0.048 |
tweet | -0.047 |
fly | -0.044 |
Nice | -0.043 |
est | -0.042 |
Token | . |
Token ID | 13 |
Feature activation | -1.25e-02 |
Pos logit contributions | |
moment | +0.017 |
tweet | +0.016 |
Nice | +0.015 |
greeting | +0.015 |
est | +0.015 |
Neg logit contributions | |
he | -0.013 |
leaf | -0.013 |
pastor | -0.013 |
ateful | -0.013 |
obb | -0.013 |
Token |
|
Token ID | 198 |
Feature activation | -1.62e-02 |
Pos logit contributions | |
moment | +0.019 |
tweet | +0.018 |
Nice | +0.017 |
Want | +0.016 |
fly | +0.016 |
Neg logit contributions | |
he | -0.018 |
olded | -0.018 |
punished | -0.017 |
sick | -0.017 |
elderly | -0.017 |
Token |
|
Token ID | 198 |
Feature activation | -2.21e-02 |
Pos logit contributions | |
moment | +0.025 |
tweet | +0.025 |
Nice | +0.023 |
Want | +0.022 |
fly | +0.022 |
Neg logit contributions | |
he | -0.022 |
olded | -0.022 |
sick | -0.021 |
punished | -0.021 |
elderly | -0.021 |
Token | O |
Token ID | 46 |
Feature activation | -8.70e-03 |
Pos logit contributions | |
olded | +0.023 |
he | +0.022 |
sick | +0.022 |
punished | +0.022 |
elderly | +0.022 |
Neg logit contributions | |
moment | -0.027 |
tweet | -0.026 |
Nice | -0.025 |
Want | -0.024 |
fly | -0.023 |
Token | ll |
Token ID | 297 |
Feature activation | -4.34e-02 |
Pos logit contributions | |
moment | +0.014 |
tweet | +0.014 |
Nice | +0.013 |
Want | +0.012 |
fly | +0.012 |
Neg logit contributions | |
olded | -0.011 |
he | -0.011 |
sick | -0.011 |
elderly | -0.011 |
punished | -0.011 |
Token | ie |
Token ID | 494 |
Feature activation | -4.19e-03 |
Pos logit contributions | |
Nice | +0.058 |
Want | +0.055 |
moment | +0.053 |
tweet | +0.051 |
feast | +0.051 |
Neg logit contributions | |
he | -0.075 |
sick | -0.067 |
elderly | -0.065 |
angry | -0.064 |
punished | -0.063 |
Token | !" |
Token ID | 2474 |
Feature activation | +3.76e-02 |
Pos logit contributions | |
moment | +0.004 |
tweet | +0.004 |
Nice | +0.004 |
Want | +0.004 |
greeting | +0.003 |
Neg logit contributions | |
he | -0.008 |
sick | -0.008 |
punished | -0.008 |
olded | -0.008 |
elderly | -0.008 |
Token | his |
Token ID | 465 |
Feature activation | +4.81e-02 |
Pos logit contributions | |
he | +0.059 |
leaf | +0.057 |
obb | +0.057 |
outs | +0.057 |
salon | +0.057 |
Neg logit contributions | |
moment | -0.051 |
tweet | -0.051 |
Nice | -0.047 |
hop | -0.044 |
greeting | -0.043 |
Token | m |
Token ID | 285 |
Feature activation | +8.57e-02 |
Pos logit contributions | |
sick | +0.050 |
elderly | +0.049 |
olded | +0.049 |
he | +0.049 |
punished | +0.049 |
Neg logit contributions | |
moment | -0.090 |
tweet | -0.089 |
Nice | -0.085 |
Want | -0.082 |
fly | -0.080 |
Token | ama |
Token ID | 1689 |
Feature activation | +2.02e-02 |
Pos logit contributions | |
he | +0.127 |
sick | +0.126 |
punished | +0.125 |
elderly | +0.123 |
leaf | +0.123 |
Neg logit contributions | |
moment | -0.120 |
tweet | -0.120 |
Nice | -0.111 |
fly | -0.106 |
Want | -0.104 |
Token | shouted |
Token ID | 17293 |
Feature activation | +2.46e-03 |
Pos logit contributions | |
he | +0.019 |
sick | +0.019 |
olded | +0.019 |
punished | +0.018 |
obb | +0.018 |
Neg logit contributions | |
moment | -0.040 |
tweet | -0.039 |
Nice | -0.037 |
fly | -0.036 |
Want | -0.036 |
Token | . |
Token ID | 13 |
Feature activation | -1.14e-02 |
Pos logit contributions | |
obb | +0.004 |
leaf | +0.004 |
he | +0.004 |
punished | +0.004 |
ateful | +0.004 |
Neg logit contributions | |
moment | -0.004 |
tweet | -0.004 |
Nice | -0.003 |
greeting | -0.003 |
est | -0.003 |
Token | She |
Token ID | 1375 |
Feature activation | -6.36e-03 |
Pos logit contributions | |
moment | +0.013 |
tweet | +0.013 |
Nice | +0.012 |
Want | +0.011 |
fly | +0.011 |
Neg logit contributions | |
he | -0.021 |
punished | -0.020 |
olded | -0.020 |
sick | -0.020 |
elderly | -0.020 |
Token | sw |
Token ID | 1509 |
Feature activation | +1.23e-02 |
Pos logit contributions | |
moment | +0.008 |
tweet | +0.008 |
Nice | +0.007 |
Want | +0.007 |
fly | +0.007 |
Neg logit contributions | |
he | -0.010 |
obb | -0.010 |
sick | -0.010 |
olded | -0.010 |
elderly | -0.010 |
Token | sand |
Token ID | 6450 |
Feature activation | -2.16e-03 |
Pos logit contributions | |
moment | +0.012 |
tweet | +0.011 |
Nice | +0.010 |
Want | +0.010 |
feast | +0.009 |
Neg logit contributions | |
olded | -0.019 |
he | -0.019 |
punished | -0.018 |
sick | -0.018 |
elderly | -0.018 |
Token | and |
Token ID | 290 |
Feature activation | -9.78e-03 |
Pos logit contributions | |
moment | +0.003 |
tweet | +0.003 |
Nice | +0.003 |
Want | +0.003 |
fly | +0.003 |
Neg logit contributions | |
he | -0.003 |
olded | -0.003 |
sick | -0.003 |
elderly | -0.003 |
punished | -0.003 |
Token | watched |
Token ID | 7342 |
Feature activation | +1.37e-02 |
Pos logit contributions | |
moment | +0.016 |
tweet | +0.015 |
Nice | +0.014 |
Want | +0.013 |
fly | +0.013 |
Neg logit contributions | |
he | -0.014 |
olded | -0.013 |
sick | -0.013 |
obb | -0.013 |
elderly | -0.013 |
Token | as |
Token ID | 355 |
Feature activation | +8.40e-03 |
Pos logit contributions | |
olded | +0.025 |
punished | +0.025 |
gr | +0.024 |
he | +0.024 |
elderly | +0.024 |
Neg logit contributions | |
moment | -0.016 |
tweet | -0.015 |
Nice | -0.013 |
Want | -0.013 |
greeting | -0.013 |
Token | it |
Token ID | 340 |
Feature activation | +1.94e-02 |
Pos logit contributions | |
olded | +0.021 |
punished | +0.020 |
vious | +0.020 |
gr | +0.019 |
obb | +0.019 |
Neg logit contributions | |
moment | -0.007 |
tweet | -0.006 |
Nice | -0.005 |
hop | -0.004 |
hopping | -0.004 |
Token | sc |
Token ID | 629 |
Feature activation | +8.55e-02 |
Pos logit contributions | |
he | +0.028 |
olded | +0.028 |
sick | +0.028 |
punished | +0.028 |
elderly | +0.027 |
Neg logit contributions | |
moment | -0.028 |
tweet | -0.028 |
Nice | -0.026 |
Want | -0.025 |
feast | -0.024 |
Token | ur |
Token ID | 333 |
Feature activation | +3.90e-03 |
Pos logit contributions | |
pastor | +0.157 |
he | +0.152 |
sick | +0.150 |
elderly | +0.149 |
punished | +0.148 |
Neg logit contributions | |
tweet | -0.099 |
Nice | -0.089 |
moment | -0.087 |
Want | -0.083 |
feast | -0.081 |
Token | ried |
Token ID | 2228 |
Feature activation | -9.64e-03 |
Pos logit contributions | |
olded | +0.005 |
punished | +0.005 |
he | +0.005 |
sick | +0.005 |
elderly | +0.005 |
Neg logit contributions | |
moment | -0.007 |
tweet | -0.006 |
Nice | -0.006 |
Want | -0.006 |
feast | -0.006 |
Token | away |
Token ID | 1497 |
Feature activation | -1.25e-04 |
Pos logit contributions | |
moment | +0.016 |
tweet | +0.016 |
Nice | +0.015 |
Want | +0.015 |
fly | +0.014 |
Neg logit contributions | |
he | -0.012 |
olded | -0.012 |
sick | -0.011 |
punished | -0.011 |
elderly | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | +2.44e-03 |
Pos logit contributions | |
moment | +0.000 |
tweet | +0.000 |
Nice | +0.000 |
fly | +0.000 |
greeting | +0.000 |
Neg logit contributions | |
he | -0.000 |
olded | -0.000 |
punished | -0.000 |
leaf | -0.000 |
pastor | -0.000 |
Token |
|
Token ID | 198 |
Feature activation | -1.25e-02 |
Pos logit contributions | |
he | +0.003 |
olded | +0.003 |
sick | +0.003 |
elderly | +0.003 |
punished | +0.003 |
Neg logit contributions | |
moment | -0.004 |
tweet | -0.004 |
Nice | -0.004 |
fly | -0.003 |
Want | -0.003 |
Token | owner |
Token ID | 4870 |
Feature activation | +1.07e-02 |
Pos logit contributions | |
obb | +0.042 |
he | +0.041 |
sick | +0.039 |
leaf | +0.039 |
punished | +0.038 |
Neg logit contributions | |
moment | -0.045 |
tweet | -0.041 |
Nice | -0.041 |
est | -0.040 |
greeting | -0.038 |
Token | showed |
Token ID | 3751 |
Feature activation | +2.20e-02 |
Pos logit contributions | |
obb | +0.016 |
sick | +0.015 |
pastor | +0.015 |
he | +0.015 |
punished | +0.015 |
Neg logit contributions | |
moment | -0.015 |
tweet | -0.014 |
est | -0.013 |
Nice | -0.013 |
Want | -0.013 |
Token | him |
Token ID | 683 |
Feature activation | -2.89e-02 |
Pos logit contributions | |
he | +0.027 |
punished | +0.027 |
olded | +0.027 |
sick | +0.027 |
elderly | +0.026 |
Neg logit contributions | |
moment | -0.037 |
tweet | -0.037 |
Nice | -0.035 |
est | -0.033 |
Want | -0.033 |
Token | how |
Token ID | 703 |
Feature activation | +2.80e-02 |
Pos logit contributions | |
moment | +0.036 |
tweet | +0.033 |
Nice | +0.031 |
est | +0.029 |
Want | +0.029 |
Neg logit contributions | |
he | -0.050 |
obb | -0.049 |
leaf | -0.048 |
punished | -0.048 |
olded | -0.048 |
Token | to |
Token ID | 284 |
Feature activation | -3.61e-03 |
Pos logit contributions | |
he | +0.060 |
obb | +0.054 |
paramedics | +0.053 |
iza | +0.053 |
olded | +0.052 |
Neg logit contributions | |
moment | -0.031 |
Nice | -0.024 |
Want | -0.024 |
tweet | -0.024 |
hopping | -0.021 |
Token | pl |
Token ID | 458 |
Feature activation | +8.55e-02 |
Pos logit contributions | |
tweet | +0.005 |
moment | +0.005 |
hop | +0.004 |
greeting | +0.004 |
feast | +0.004 |
Neg logit contributions | |
olded | -0.006 |
outs | -0.006 |
outed | -0.006 |
pastor | -0.006 |
obb | -0.006 |
Token | uck |
Token ID | 1347 |
Feature activation | -9.43e-03 |
Pos logit contributions | |
pastor | +0.127 |
punished | +0.120 |
ashamed | +0.119 |
sick | +0.119 |
obb | +0.118 |
Neg logit contributions | |
tweet | -0.111 |
elcome | -0.102 |
immer | -0.100 |
Want | -0.099 |
moment | -0.097 |
Token | the |
Token ID | 262 |
Feature activation | +4.25e-03 |
Pos logit contributions | |
moment | +0.013 |
tweet | +0.013 |
Nice | +0.012 |
est | +0.011 |
Want | +0.011 |
Neg logit contributions | |
olded | -0.015 |
punished | -0.015 |
he | -0.014 |
elderly | -0.014 |
leaf | -0.014 |
Token | strings |
Token ID | 13042 |
Feature activation | +1.15e-02 |
Pos logit contributions | |
olded | +0.007 |
punished | +0.006 |
obb | +0.006 |
iza | +0.006 |
gr | +0.006 |
Neg logit contributions | |
moment | -0.007 |
tweet | -0.006 |
est | -0.006 |
greeting | -0.006 |
Want | -0.006 |
Token | and |
Token ID | 290 |
Feature activation | -1.22e-02 |
Pos logit contributions | |
olded | +0.017 |
pastor | +0.017 |
gr | +0.017 |
punished | +0.017 |
leaf | +0.017 |
Neg logit contributions | |
moment | -0.018 |
tweet | -0.016 |
est | -0.015 |
fly | -0.015 |
Nice | -0.014 |
Token | make |
Token ID | 787 |
Feature activation | +8.95e-03 |
Pos logit contributions | |
moment | +0.017 |
tweet | +0.017 |
Want | +0.015 |
Nice | +0.015 |
greeting | +0.015 |
Neg logit contributions | |
he | -0.020 |
obb | -0.020 |
gr | -0.018 |
outs | -0.018 |
iza | -0.018 |
Token | She |
Token ID | 3347 |
Feature activation | +1.79e-02 |
Pos logit contributions | |
moment | +0.020 |
tweet | +0.019 |
Want | +0.017 |
Nice | +0.017 |
greeting | +0.016 |
Neg logit contributions | |
sick | -0.020 |
bored | -0.020 |
obb | -0.020 |
outs | -0.019 |
olded | -0.019 |
Token | watched |
Token ID | 7342 |
Feature activation | +1.09e-02 |
Pos logit contributions | |
he | +0.025 |
sick | +0.024 |
obb | +0.023 |
ateful | +0.023 |
pastor | +0.023 |
Neg logit contributions | |
moment | -0.028 |
tweet | -0.027 |
Nice | -0.024 |
Want | -0.024 |
fly | -0.024 |
Token | as |
Token ID | 355 |
Feature activation | +7.45e-03 |
Pos logit contributions | |
olded | +0.017 |
punished | +0.017 |
gr | +0.017 |
ateful | +0.017 |
leaf | +0.016 |
Neg logit contributions | |
moment | -0.015 |
tweet | -0.015 |
Nice | -0.013 |
greeting | -0.013 |
fly | -0.012 |
Token | the |
Token ID | 262 |
Feature activation | +3.29e-02 |
Pos logit contributions | |
olded | +0.016 |
gr | +0.015 |
punished | +0.015 |
outed | +0.015 |
vious | +0.014 |
Neg logit contributions | |
moment | -0.008 |
tweet | -0.007 |
Daisy | -0.007 |
Larry | -0.006 |
Bella | -0.006 |
Token | mouse |
Token ID | 10211 |
Feature activation | +1.16e-02 |
Pos logit contributions | |
olded | +0.063 |
he | +0.062 |
leaf | +0.060 |
punished | +0.060 |
outs | +0.060 |
Neg logit contributions | |
moment | -0.037 |
tweet | -0.034 |
Nice | -0.029 |
feast | -0.029 |
Want | -0.028 |
Token | sc |
Token ID | 629 |
Feature activation | +8.47e-02 |
Pos logit contributions | |
he | +0.016 |
olded | +0.016 |
elderly | +0.016 |
sick | +0.016 |
gr | +0.016 |
Neg logit contributions | |
moment | -0.018 |
tweet | -0.017 |
Nice | -0.016 |
fly | -0.015 |
feast | -0.015 |
Token | ur |
Token ID | 333 |
Feature activation | +6.92e-03 |
Pos logit contributions | |
pastor | +0.191 |
sick | +0.180 |
ateful | +0.178 |
elderly | +0.175 |
late | +0.175 |
Neg logit contributions | |
tweet | -0.073 |
leap | -0.046 |
immer | -0.045 |
ig | -0.045 |
ar | -0.042 |
Token | ried |
Token ID | 2228 |
Feature activation | -7.30e-03 |
Pos logit contributions | |
he | +0.011 |
sick | +0.010 |
olded | +0.010 |
elderly | +0.010 |
punished | +0.010 |
Neg logit contributions | |
moment | -0.010 |
tweet | -0.010 |
fly | -0.009 |
Nice | -0.009 |
Want | -0.008 |
Token | over |
Token ID | 625 |
Feature activation | -8.88e-03 |
Pos logit contributions | |
moment | +0.012 |
tweet | +0.012 |
Nice | +0.011 |
Want | +0.011 |
fly | +0.010 |
Neg logit contributions | |
he | -0.009 |
olded | -0.009 |
sick | -0.009 |
punished | -0.009 |
elderly | -0.009 |
Token | to |
Token ID | 284 |
Feature activation | -4.00e-03 |
Pos logit contributions | |
moment | +0.016 |
tweet | +0.015 |
Nice | +0.014 |
Want | +0.014 |
fly | +0.014 |
Neg logit contributions | |
he | -0.011 |
olded | -0.010 |
sick | -0.010 |
punished | -0.010 |
elderly | -0.010 |
Token | the |
Token ID | 262 |
Feature activation | +1.59e-03 |
Pos logit contributions | |
moment | +0.007 |
tweet | +0.007 |
Nice | +0.006 |
Want | +0.006 |
greeting | +0.006 |
Neg logit contributions | |
olded | -0.005 |
he | -0.005 |
punished | -0.005 |
sick | -0.005 |
outs | -0.005 |
Token | in |
Token ID | 287 |
Feature activation | -1.46e-02 |
Pos logit contributions | |
moment | +0.010 |
tweet | +0.009 |
est | +0.008 |
Nice | +0.008 |
fly | +0.008 |
Neg logit contributions | |
leaf | -0.012 |
olded | -0.011 |
punished | -0.011 |
earn | -0.011 |
sick | -0.011 |
Token | the |
Token ID | 262 |
Feature activation | +1.03e-03 |
Pos logit contributions | |
moment | +0.027 |
tweet | +0.026 |
Nice | +0.025 |
Want | +0.024 |
fly | +0.023 |
Neg logit contributions | |
he | -0.016 |
olded | -0.016 |
sick | -0.016 |
punished | -0.016 |
elderly | -0.015 |
Token | knowledge |
Token ID | 3725 |
Feature activation | -1.15e-03 |
Pos logit contributions | |
olded | +0.002 |
outs | +0.002 |
he | +0.002 |
leaf | +0.002 |
punished | +0.002 |
Neg logit contributions | |
moment | -0.001 |
tweet | -0.001 |
est | -0.001 |
Nice | -0.001 |
greeting | -0.001 |
Token | that |
Token ID | 326 |
Feature activation | -2.10e-02 |
Pos logit contributions | |
moment | +0.001 |
tweet | +0.001 |
Nice | +0.001 |
Want | +0.001 |
est | +0.001 |
Neg logit contributions | |
olded | -0.002 |
he | -0.002 |
punished | -0.002 |
sick | -0.002 |
elderly | -0.002 |
Token | her |
Token ID | 607 |
Feature activation | +2.09e-02 |
Pos logit contributions | |
tweet | +0.037 |
Nice | +0.035 |
fly | +0.034 |
Want | +0.034 |
moment | +0.034 |
Neg logit contributions | |
he | -0.024 |
sick | -0.024 |
elderly | -0.022 |
bored | -0.022 |
punished | -0.022 |
Token | sc |
Token ID | 629 |
Feature activation | +8.46e-02 |
Pos logit contributions | |
he | +0.031 |
olded | +0.030 |
leaf | +0.030 |
outs | +0.029 |
punished | +0.029 |
Neg logit contributions | |
moment | -0.032 |
tweet | -0.030 |
est | -0.028 |
Nice | -0.028 |
greeting | -0.027 |
Token | ooter |
Token ID | 25141 |
Feature activation | +2.18e-02 |
Pos logit contributions | |
sick | +0.151 |
pastor | +0.148 |
he | +0.143 |
ateful | +0.142 |
late | +0.142 |
Neg logit contributions | |
tweet | -0.099 |
Nice | -0.085 |
moment | -0.083 |
est | -0.082 |
Want | -0.079 |
Token | was |
Token ID | 373 |
Feature activation | -7.23e-03 |
Pos logit contributions | |
leaf | +0.040 |
he | +0.039 |
olded | +0.039 |
iza | +0.039 |
sick | +0.039 |
Neg logit contributions | |
moment | -0.025 |
tweet | -0.023 |
est | -0.021 |
Nice | -0.019 |
Want | -0.018 |
Token | safe |
Token ID | 3338 |
Feature activation | -1.04e-02 |
Pos logit contributions | |
moment | +0.012 |
tweet | +0.012 |
Nice | +0.011 |
Want | +0.010 |
fly | +0.010 |
Neg logit contributions | |
he | -0.009 |
olded | -0.009 |
sick | -0.009 |
leaf | -0.009 |
punished | -0.009 |
Token | and |
Token ID | 290 |
Feature activation | +2.65e-03 |
Pos logit contributions | |
moment | +0.015 |
tweet | +0.014 |
Nice | +0.013 |
Want | +0.012 |
fly | +0.012 |
Neg logit contributions | |
olded | -0.016 |
he | -0.016 |
punished | -0.016 |
sick | -0.015 |
elderly | -0.015 |
Token | that |
Token ID | 326 |
Feature activation | -2.29e-02 |
Pos logit contributions | |
he | +0.003 |
olded | +0.003 |
sick | +0.003 |
punished | +0.003 |
elderly | +0.003 |
Neg logit contributions | |
moment | -0.004 |
tweet | -0.004 |
Nice | -0.004 |
Want | -0.004 |
fly | -0.004 |
Token | the |
Token ID | 262 |
Feature activation | +8.05e-03 |
Pos logit contributions | |
olded | +0.017 |
gr | +0.016 |
leaf | +0.016 |
punished | +0.016 |
ateful | +0.016 |
Neg logit contributions | |
moment | -0.014 |
tweet | -0.013 |
greeting | -0.012 |
Nice | -0.012 |
Want | -0.011 |
Token | bug |
Token ID | 5434 |
Feature activation | -5.54e-03 |
Pos logit contributions | |
olded | +0.014 |
he | +0.014 |
punished | +0.013 |
sick | +0.013 |
elderly | +0.013 |
Neg logit contributions | |
moment | -0.010 |
tweet | -0.010 |
Nice | -0.009 |
Want | -0.009 |
feast | -0.008 |
Token | as |
Token ID | 355 |
Feature activation | +9.02e-03 |
Pos logit contributions | |
moment | +0.008 |
tweet | +0.008 |
Nice | +0.007 |
Want | +0.007 |
fly | +0.007 |
Neg logit contributions | |
olded | -0.008 |
he | -0.008 |
elderly | -0.008 |
punished | -0.008 |
sick | -0.008 |
Token | it |
Token ID | 340 |
Feature activation | +1.44e-02 |
Pos logit contributions | |
olded | +0.017 |
gr | +0.016 |
punished | +0.016 |
vious | +0.015 |
leaf | +0.015 |
Neg logit contributions | |
moment | -0.012 |
tweet | -0.011 |
Nice | -0.010 |
hopping | -0.010 |
fly | -0.009 |
Token | slowly |
Token ID | 6364 |
Feature activation | +6.25e-03 |
Pos logit contributions | |
he | +0.025 |
pastor | +0.025 |
leaf | +0.025 |
elderly | +0.025 |
olded | +0.025 |
Neg logit contributions | |
tweet | -0.018 |
moment | -0.018 |
fly | -0.015 |
Nice | -0.014 |
est | -0.014 |
Token | sc |
Token ID | 629 |
Feature activation | +8.43e-02 |
Pos logit contributions | |
he | +0.009 |
gr | +0.009 |
pastor | +0.009 |
elderly | +0.009 |
olded | +0.009 |
Neg logit contributions | |
moment | -0.009 |
tweet | -0.009 |
Nice | -0.008 |
est | -0.008 |
greeting | -0.008 |
Token | ur |
Token ID | 333 |
Feature activation | +2.61e-03 |
Pos logit contributions | |
pastor | +0.167 |
he | +0.164 |
sick | +0.155 |
elderly | +0.150 |
ateful | +0.149 |
Neg logit contributions | |
tweet | -0.094 |
Nice | -0.078 |
Want | -0.071 |
fly | -0.070 |
immer | -0.068 |
Token | ried |
Token ID | 2228 |
Feature activation | -6.12e-03 |
Pos logit contributions | |
he | +0.004 |
olded | +0.004 |
sick | +0.004 |
elderly | +0.004 |
punished | +0.004 |
Neg logit contributions | |
moment | -0.004 |
tweet | -0.004 |
Nice | -0.003 |
fly | -0.003 |
Want | -0.003 |
Token | away |
Token ID | 1497 |
Feature activation | +2.63e-04 |
Pos logit contributions | |
moment | +0.009 |
tweet | +0.009 |
Nice | +0.008 |
Want | +0.008 |
est | +0.008 |
Neg logit contributions | |
olded | -0.009 |
he | -0.009 |
punished | -0.009 |
pastor | -0.009 |
elderly | -0.009 |
Token | . |
Token ID | 13 |
Feature activation | -6.56e-04 |
Pos logit contributions | |
he | +0.000 |
olded | +0.000 |
leaf | +0.000 |
punished | +0.000 |
earn | +0.000 |
Neg logit contributions | |
moment | -0.000 |
tweet | -0.000 |
Nice | -0.000 |
fly | -0.000 |
greeting | -0.000 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -1.06e-02 |
Pos logit contributions | |
moment | +0.001 |
tweet | +0.001 |
Nice | +0.001 |
fly | +0.001 |
Want | +0.001 |
Neg logit contributions | |
he | -0.001 |
olded | -0.001 |
sick | -0.001 |
elderly | -0.001 |
punished | -0.001 |
Token |
|
Token ID | 198 |
Feature activation | -2.37e-02 |
Pos logit contributions | |
moment | +0.023 |
tweet | +0.022 |
Nice | +0.021 |
Want | +0.020 |
feast | +0.019 |
Neg logit contributions | |
olded | -0.027 |
punished | -0.027 |
sick | -0.026 |
elderly | -0.026 |
gr | -0.026 |
Token | So |
Token ID | 2396 |
Feature activation | -1.14e-02 |
Pos logit contributions | |
moment | +0.040 |
tweet | +0.038 |
Nice | +0.036 |
Want | +0.035 |
greeting | +0.034 |
Neg logit contributions | |
olded | -0.030 |
obb | -0.029 |
sick | -0.029 |
outs | -0.029 |
bored | -0.029 |
Token | the |
Token ID | 262 |
Feature activation | +2.39e-02 |
Pos logit contributions | |
moment | +0.015 |
tweet | +0.013 |
Nice | +0.013 |
Want | +0.012 |
est | +0.012 |
Neg logit contributions | |
leaf | -0.019 |
obb | -0.019 |
gr | -0.019 |
he | -0.019 |
olded | -0.019 |
Token | little |
Token ID | 1310 |
Feature activation | +5.79e-03 |
Pos logit contributions | |
olded | +0.034 |
outs | +0.034 |
he | +0.033 |
leaf | +0.033 |
punished | +0.032 |
Neg logit contributions | |
moment | -0.038 |
tweet | -0.035 |
Nice | -0.033 |
Want | -0.032 |
est | -0.032 |
Token | girl |
Token ID | 2576 |
Feature activation | +1.36e-02 |
Pos logit contributions | |
olded | +0.008 |
he | +0.008 |
punished | +0.008 |
leaf | +0.008 |
sick | +0.008 |
Neg logit contributions | |
moment | -0.009 |
tweet | -0.009 |
Nice | -0.008 |
Want | -0.008 |
est | -0.008 |
Token | sc |
Token ID | 629 |
Feature activation | +8.41e-02 |
Pos logit contributions | |
he | +0.021 |
ateful | +0.021 |
gr | +0.020 |
sick | +0.020 |
obb | +0.020 |
Neg logit contributions | |
moment | -0.019 |
tweet | -0.019 |
feast | -0.017 |
est | -0.017 |
fly | -0.016 |
Token | oot |
Token ID | 1025 |
Feature activation | -1.39e-02 |
Pos logit contributions | |
sick | +0.130 |
pastor | +0.126 |
late | +0.118 |
bored | +0.117 |
ateful | +0.114 |
Neg logit contributions | |
tweet | -0.128 |
Nice | -0.115 |
Want | -0.105 |
reetings | -0.103 |
lad | -0.102 |
Token | ed |
Token ID | 276 |
Feature activation | -2.94e-03 |
Pos logit contributions | |
moment | +0.018 |
tweet | +0.017 |
est | +0.017 |
fly | +0.015 |
dy | +0.014 |
Neg logit contributions | |
obb | -0.024 |
pastor | -0.024 |
sick | -0.023 |
ateful | -0.023 |
stressed | -0.022 |
Token | away |
Token ID | 1497 |
Feature activation | +6.96e-03 |
Pos logit contributions | |
moment | +0.005 |
tweet | +0.004 |
Nice | +0.004 |
greeting | +0.004 |
est | +0.004 |
Neg logit contributions | |
he | -0.004 |
obb | -0.004 |
pastor | -0.004 |
olded | -0.004 |
ateful | -0.004 |
Token | again |
Token ID | 757 |
Feature activation | +8.76e-03 |
Pos logit contributions | |
leaf | +0.013 |
obb | +0.013 |
olded | +0.012 |
pastor | +0.012 |
he | +0.012 |
Neg logit contributions | |
moment | -0.009 |
tweet | -0.007 |
Nice | -0.007 |
greeting | -0.007 |
est | -0.007 |
Token | . |
Token ID | 13 |
Feature activation | -3.44e-03 |
Pos logit contributions | |
olded | +0.014 |
obb | +0.014 |
leaf | +0.013 |
pastor | +0.013 |
outed | +0.013 |
Neg logit contributions | |
moment | -0.014 |
tweet | -0.012 |
greeting | -0.011 |
Nice | -0.011 |
Want | -0.011 |
Token | at |
Token ID | 379 |
Feature activation | +5.39e-03 |
Pos logit contributions | |
moment | +0.069 |
tweet | +0.068 |
Nice | +0.065 |
Want | +0.062 |
est | +0.061 |
Neg logit contributions | |
he | -0.036 |
olded | -0.035 |
punished | -0.035 |
sick | -0.034 |
elderly | -0.033 |
Token | it |
Token ID | 340 |
Feature activation | -2.87e-03 |
Pos logit contributions | |
olded | +0.007 |
he | +0.007 |
punished | +0.006 |
sick | +0.006 |
elderly | +0.006 |
Neg logit contributions | |
moment | -0.009 |
tweet | -0.009 |
Nice | -0.008 |
Want | -0.008 |
fly | -0.008 |
Token | until |
Token ID | 1566 |
Feature activation | +1.25e-02 |
Pos logit contributions | |
moment | +0.004 |
tweet | +0.004 |
Nice | +0.004 |
Want | +0.004 |
feast | +0.004 |
Neg logit contributions | |
he | -0.004 |
sick | -0.004 |
olded | -0.004 |
elderly | -0.004 |
outs | -0.004 |
Token | the |
Token ID | 262 |
Feature activation | +1.83e-02 |
Pos logit contributions | |
olded | +0.019 |
punished | +0.018 |
elderly | +0.018 |
gr | +0.018 |
leaf | +0.018 |
Neg logit contributions | |
moment | -0.020 |
tweet | -0.019 |
Nice | -0.017 |
feast | -0.016 |
greeting | -0.016 |
Token | beetle |
Token ID | 45489 |
Feature activation | +3.36e-02 |
Pos logit contributions | |
olded | +0.030 |
he | +0.030 |
punished | +0.029 |
sick | +0.029 |
elderly | +0.028 |
Neg logit contributions | |
moment | -0.024 |
tweet | -0.024 |
Nice | -0.022 |
Want | -0.021 |
fly | -0.020 |
Token | sc |
Token ID | 629 |
Feature activation | +8.37e-02 |
Pos logit contributions | |
he | +0.053 |
olded | +0.053 |
sick | +0.052 |
punished | +0.052 |
elderly | +0.051 |
Neg logit contributions | |
moment | -0.045 |
tweet | -0.044 |
Nice | -0.040 |
Want | -0.039 |
fly | -0.038 |
Token | ur |
Token ID | 333 |
Feature activation | +6.17e-03 |
Pos logit contributions | |
pastor | +0.164 |
sick | +0.159 |
punished | +0.153 |
he | +0.153 |
elderly | +0.148 |
Neg logit contributions | |
tweet | -0.095 |
Nice | -0.079 |
fly | -0.071 |
moment | -0.071 |
Want | -0.071 |
Token | ried |
Token ID | 2228 |
Feature activation | -3.20e-03 |
Pos logit contributions | |
he | +0.008 |
olded | +0.008 |
punished | +0.008 |
sick | +0.008 |
elderly | +0.008 |
Neg logit contributions | |
moment | -0.010 |
tweet | -0.009 |
Nice | -0.009 |
Want | -0.009 |
feast | -0.008 |
Token | out |
Token ID | 503 |
Feature activation | +1.01e-03 |
Pos logit contributions | |
moment | +0.005 |
tweet | +0.005 |
Nice | +0.005 |
Want | +0.005 |
fly | +0.005 |
Neg logit contributions | |
he | -0.004 |
olded | -0.004 |
sick | -0.004 |
punished | -0.004 |
elderly | -0.004 |
Token | of |
Token ID | 286 |
Feature activation | +1.28e-02 |
Pos logit contributions | |
he | +0.001 |
olded | +0.001 |
punished | +0.001 |
sick | +0.001 |
elderly | +0.001 |
Neg logit contributions | |
moment | -0.002 |
tweet | -0.002 |
Nice | -0.002 |
Want | -0.001 |
fly | -0.001 |
Token | sight |
Token ID | 6504 |
Feature activation | -7.74e-03 |
Pos logit contributions | |
he | +0.013 |
elderly | +0.013 |
sick | +0.013 |
punished | +0.012 |
leaf | +0.012 |
Neg logit contributions | |
tweet | -0.024 |
moment | -0.023 |
Nice | -0.023 |
Want | -0.022 |
est | -0.022 |
Token | his |
Token ID | 465 |
Feature activation | +1.03e-02 |
Pos logit contributions | |
he | +0.002 |
olded | +0.002 |
punished | +0.002 |
sick | +0.002 |
obb | +0.002 |
Neg logit contributions | |
moment | -0.002 |
tweet | -0.002 |
Nice | -0.002 |
est | -0.002 |
Want | -0.002 |
Token | baseball |
Token ID | 9283 |
Feature activation | -9.24e-03 |
Pos logit contributions | |
he | +0.016 |
olded | +0.016 |
leaf | +0.015 |
punished | +0.015 |
gr | +0.015 |
Neg logit contributions | |
moment | -0.015 |
tweet | -0.014 |
Nice | -0.014 |
feast | -0.013 |
greeting | -0.013 |
Token | . |
Token ID | 13 |
Feature activation | +6.09e-03 |
Pos logit contributions | |
moment | +0.017 |
tweet | +0.017 |
Nice | +0.016 |
fly | +0.015 |
feast | +0.015 |
Neg logit contributions | |
he | -0.010 |
olded | -0.010 |
leaf | -0.010 |
punished | -0.010 |
pastor | -0.010 |
Token | As |
Token ID | 1081 |
Feature activation | +1.03e-02 |
Pos logit contributions | |
he | +0.008 |
olded | +0.008 |
sick | +0.008 |
elderly | +0.008 |
punished | +0.008 |
Neg logit contributions | |
tweet | -0.009 |
moment | -0.009 |
Nice | -0.008 |
feast | -0.008 |
fly | -0.008 |
Token | Joe |
Token ID | 5689 |
Feature activation | +8.23e-03 |
Pos logit contributions | |
he | +0.012 |
olded | +0.012 |
punished | +0.012 |
sick | +0.012 |
elderly | +0.011 |
Neg logit contributions | |
moment | -0.019 |
tweet | -0.018 |
Nice | -0.017 |
Want | -0.016 |
fly | -0.016 |
Token | sc |
Token ID | 629 |
Feature activation | +8.30e-02 |
Pos logit contributions | |
pastor | +0.014 |
gr | +0.014 |
obb | +0.014 |
ateful | +0.014 |
leaf | +0.014 |
Neg logit contributions | |
moment | -0.010 |
tweet | -0.009 |
feast | -0.008 |
Want | -0.008 |
dy | -0.008 |
Token | oot |
Token ID | 1025 |
Feature activation | -1.03e-02 |
Pos logit contributions | |
pastor | +0.148 |
sick | +0.140 |
elderly | +0.135 |
late | +0.130 |
gr | +0.128 |
Neg logit contributions | |
tweet | -0.106 |
Nice | -0.095 |
Want | -0.091 |
feast | -0.090 |
elcome | -0.086 |
Token | ed |
Token ID | 276 |
Feature activation | +1.03e-02 |
Pos logit contributions | |
moment | +0.014 |
tweet | +0.014 |
est | +0.013 |
fly | +0.013 |
dy | +0.012 |
Neg logit contributions | |
pastor | -0.016 |
obb | -0.016 |
elderly | -0.016 |
leaf | -0.016 |
sick | -0.015 |
Token | around |
Token ID | 1088 |
Feature activation | -3.04e-03 |
Pos logit contributions | |
sick | +0.010 |
bored | +0.010 |
punished | +0.010 |
angry | +0.010 |
he | +0.010 |
Neg logit contributions | |
tweet | -0.018 |
moment | -0.018 |
Nice | -0.018 |
reetings | -0.017 |
lad | -0.017 |
Token | , |
Token ID | 11 |
Feature activation | +1.06e-02 |
Pos logit contributions | |
tweet | +0.004 |
moment | +0.004 |
Nice | +0.004 |
Want | +0.003 |
fly | +0.003 |
Neg logit contributions | |
he | -0.005 |
sick | -0.005 |
punished | -0.005 |
elderly | -0.005 |
olded | -0.005 |
Token | he |
Token ID | 339 |
Feature activation | +1.84e-02 |
Pos logit contributions | |
olded | +0.016 |
he | +0.015 |
punished | +0.015 |
gr | +0.015 |
leaf | +0.015 |
Neg logit contributions | |
moment | -0.017 |
tweet | -0.015 |
Nice | -0.015 |
Want | -0.014 |
greeting | -0.014 |
Token | thanked |
Token ID | 26280 |
Feature activation | +1.58e-02 |
Pos logit contributions | |
he | +0.003 |
obb | +0.003 |
outs | +0.003 |
pastor | +0.002 |
gr | +0.002 |
Neg logit contributions | |
moment | -0.003 |
tweet | -0.003 |
Nice | -0.003 |
Want | -0.003 |
greeting | -0.003 |
Token | Jo |
Token ID | 5302 |
Feature activation | -1.55e-02 |
Pos logit contributions | |
olded | +0.027 |
obb | +0.027 |
punished | +0.027 |
sick | +0.026 |
bored | +0.025 |
Neg logit contributions | |
moment | -0.022 |
tweet | -0.021 |
Nice | -0.019 |
greeting | -0.018 |
Want | -0.017 |
Token | jo |
Token ID | 7639 |
Feature activation | -2.90e-02 |
Pos logit contributions | |
moment | +0.016 |
tweet | +0.014 |
Nice | +0.013 |
est | +0.012 |
fly | +0.012 |
Neg logit contributions | |
sick | -0.032 |
obb | -0.032 |
punished | -0.032 |
earn | -0.031 |
olded | -0.031 |
Token | as |
Token ID | 355 |
Feature activation | +1.12e-02 |
Pos logit contributions | |
moment | +0.040 |
tweet | +0.038 |
Nice | +0.035 |
feast | +0.034 |
greeting | +0.033 |
Neg logit contributions | |
obb | -0.049 |
he | -0.046 |
ateful | -0.046 |
punished | -0.045 |
pastor | -0.045 |
Token | it |
Token ID | 340 |
Feature activation | +1.78e-02 |
Pos logit contributions | |
punished | +0.022 |
olded | +0.021 |
obb | +0.021 |
gr | +0.020 |
salon | +0.020 |
Neg logit contributions | |
moment | -0.014 |
Nice | -0.013 |
tweet | -0.012 |
Want | -0.011 |
hopping | -0.010 |
Token | sc |
Token ID | 629 |
Feature activation | +8.26e-02 |
Pos logit contributions | |
sick | +0.026 |
olded | +0.026 |
he | +0.026 |
pastor | +0.026 |
gr | +0.026 |
Neg logit contributions | |
moment | -0.026 |
tweet | -0.025 |
Nice | -0.023 |
feast | -0.023 |
fly | -0.022 |
Token | ur |
Token ID | 333 |
Feature activation | +6.84e-03 |
Pos logit contributions | |
pastor | +0.184 |
sick | +0.174 |
he | +0.168 |
punished | +0.165 |
bored | +0.164 |
Neg logit contributions | |
tweet | -0.075 |
Nice | -0.065 |
Want | -0.058 |
feast | -0.055 |
reetings | -0.053 |
Token | ried |
Token ID | 2228 |
Feature activation | -7.32e-03 |
Pos logit contributions | |
he | +0.012 |
sick | +0.012 |
elderly | +0.011 |
pastor | +0.011 |
olded | +0.011 |
Neg logit contributions | |
moment | -0.009 |
tweet | -0.009 |
fly | -0.008 |
Nice | -0.008 |
est | -0.007 |
Token | away |
Token ID | 1497 |
Feature activation | -3.10e-03 |
Pos logit contributions | |
moment | +0.010 |
tweet | +0.009 |
Nice | +0.009 |
est | +0.009 |
greeting | +0.009 |
Neg logit contributions | |
pastor | -0.012 |
sick | -0.012 |
olded | -0.011 |
earn | -0.011 |
punished | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | +3.97e-03 |
Pos logit contributions | |
moment | +0.005 |
Nice | +0.004 |
tweet | +0.004 |
greeting | +0.004 |
fly | +0.004 |
Neg logit contributions | |
obb | -0.005 |
punished | -0.004 |
he | -0.004 |
leaf | -0.004 |
olded | -0.004 |
Token | Jo |
Token ID | 5302 |
Feature activation | +1.17e-02 |
Pos logit contributions | |
sick | +0.007 |
punished | +0.007 |
obb | +0.007 |
he | +0.006 |
olded | +0.006 |
Neg logit contributions | |
moment | -0.005 |
Nice | -0.005 |
tweet | -0.005 |
Want | -0.005 |
greeting | -0.004 |
Token | I |
Token ID | 314 |
Feature activation | +1.13e-02 |
Pos logit contributions | |
olded | +0.011 |
he | +0.010 |
sick | +0.010 |
punished | +0.010 |
elderly | +0.010 |
Neg logit contributions | |
moment | -0.014 |
tweet | -0.013 |
Nice | -0.013 |
feast | -0.012 |
fly | -0.012 |
Token | help |
Token ID | 1037 |
Feature activation | +1.71e-02 |
Pos logit contributions | |
obb | +0.021 |
outs | +0.021 |
pastor | +0.021 |
ateful | +0.020 |
olded | +0.020 |
Neg logit contributions | |
tweet | -0.013 |
moment | -0.013 |
Nice | -0.011 |
Want | -0.011 |
hop | -0.010 |
Token | you |
Token ID | 345 |
Feature activation | -1.06e-03 |
Pos logit contributions | |
obb | +0.024 |
olded | +0.024 |
he | +0.023 |
punished | +0.023 |
outs | +0.022 |
Neg logit contributions | |
tweet | -0.026 |
moment | -0.026 |
Nice | -0.024 |
Want | -0.024 |
greeting | -0.022 |
Token | fix |
Token ID | 4259 |
Feature activation | +1.52e-02 |
Pos logit contributions | |
moment | +0.001 |
tweet | +0.001 |
Want | +0.001 |
est | +0.001 |
Nice | +0.001 |
Neg logit contributions | |
olded | -0.002 |
leaf | -0.002 |
outs | -0.002 |
pastor | -0.002 |
obb | -0.002 |
Token | your |
Token ID | 534 |
Feature activation | +1.89e-02 |
Pos logit contributions | |
olded | +0.028 |
he | +0.026 |
obb | +0.026 |
leaf | +0.026 |
sick | +0.025 |
Neg logit contributions | |
moment | -0.017 |
tweet | -0.017 |
est | -0.015 |
Nice | -0.015 |
Want | -0.015 |
Token | sc |
Token ID | 629 |
Feature activation | +8.26e-02 |
Pos logit contributions | |
he | +0.019 |
olded | +0.018 |
leaf | +0.017 |
sick | +0.016 |
punished | +0.016 |
Neg logit contributions | |
moment | -0.038 |
tweet | -0.037 |
Nice | -0.035 |
est | -0.034 |
Want | -0.034 |
Token | ooter |
Token ID | 25141 |
Feature activation | +2.99e-02 |
Pos logit contributions | |
he | +0.134 |
pastor | +0.132 |
sick | +0.130 |
wasted | +0.129 |
punished | +0.126 |
Neg logit contributions | |
tweet | -0.105 |
moment | -0.096 |
Nice | -0.093 |
Want | -0.091 |
est | -0.088 |
Token | ?" |
Token ID | 1701 |
Feature activation | +3.63e-02 |
Pos logit contributions | |
leaf | +0.052 |
pastor | +0.049 |
iza | +0.049 |
wasted | +0.048 |
olded | +0.048 |
Neg logit contributions | |
moment | -0.037 |
tweet | -0.036 |
est | -0.034 |
Want | -0.032 |
Nice | -0.032 |
Token |
|
Token ID | 198 |
Feature activation | -4.22e-03 |
Pos logit contributions | |
obb | +0.075 |
punished | +0.075 |
gr | +0.072 |
olded | +0.072 |
leaf | +0.071 |
Neg logit contributions | |
moment | -0.040 |
Nice | -0.034 |
Want | -0.031 |
tweet | -0.031 |
greeting | -0.027 |
Token |
|
Token ID | 198 |
Feature activation | -1.04e-02 |
Pos logit contributions | |
moment | +0.005 |
tweet | +0.004 |
Nice | +0.004 |
Want | +0.004 |
dy | +0.004 |
Neg logit contributions | |
punished | -0.008 |
obb | -0.008 |
olded | -0.008 |
leaf | -0.007 |
sick | -0.007 |
Token | The |
Token ID | 464 |
Feature activation | +2.12e-02 |
Pos logit contributions | |
moment | +0.016 |
tweet | +0.014 |
Nice | +0.013 |
Want | +0.013 |
dy | +0.013 |
Neg logit contributions | |
obb | -0.016 |
bored | -0.015 |
sick | -0.015 |
outs | -0.015 |
leaf | -0.015 |
Token | that |
Token ID | 326 |
Feature activation | -7.54e-03 |
Pos logit contributions | |
he | +0.004 |
olded | +0.004 |
sick | +0.004 |
punished | +0.004 |
elderly | +0.004 |
Neg logit contributions | |
moment | -0.002 |
tweet | -0.002 |
Nice | -0.002 |
Want | -0.002 |
fly | -0.002 |
Token | could |
Token ID | 714 |
Feature activation | +2.68e-02 |
Pos logit contributions | |
moment | +0.010 |
tweet | +0.010 |
Nice | +0.009 |
fly | +0.008 |
Want | +0.008 |
Neg logit contributions | |
he | -0.012 |
olded | -0.012 |
punished | -0.012 |
elderly | -0.012 |
sick | -0.012 |
Token | help |
Token ID | 1037 |
Feature activation | +8.74e-03 |
Pos logit contributions | |
olded | +0.047 |
outs | +0.045 |
outed | +0.044 |
he | +0.044 |
obb | +0.044 |
Neg logit contributions | |
tweet | -0.036 |
moment | -0.033 |
fly | -0.031 |
feast | -0.030 |
Nice | -0.029 |
Token | the |
Token ID | 262 |
Feature activation | +1.86e-02 |
Pos logit contributions | |
he | +0.011 |
olded | +0.011 |
punished | +0.010 |
sick | +0.010 |
elderly | +0.010 |
Neg logit contributions | |
moment | -0.015 |
tweet | -0.015 |
Nice | -0.014 |
Want | -0.013 |
fly | -0.013 |
Token | farmer |
Token ID | 18739 |
Feature activation | +2.17e-02 |
Pos logit contributions | |
olded | +0.033 |
he | +0.032 |
punished | +0.031 |
leaf | +0.031 |
outs | +0.030 |
Neg logit contributions | |
moment | -0.024 |
tweet | -0.023 |
feast | -0.020 |
Nice | -0.020 |
fly | -0.019 |
Token | pl |
Token ID | 458 |
Feature activation | +8.23e-02 |
Pos logit contributions | |
olded | +0.031 |
obb | +0.031 |
leaf | +0.031 |
he | +0.030 |
pastor | +0.030 |
Neg logit contributions | |
moment | -0.032 |
tweet | -0.032 |
fly | -0.029 |
Nice | -0.028 |
feast | -0.028 |
Token | ow |
Token ID | 322 |
Feature activation | -2.83e-02 |
Pos logit contributions | |
obb | +0.130 |
elderly | +0.128 |
pastor | +0.127 |
leaf | +0.126 |
wasted | +0.125 |
Neg logit contributions | |
tweet | -0.107 |
fly | -0.102 |
est | -0.089 |
seed | -0.089 |
moment | -0.088 |
Token | the |
Token ID | 262 |
Feature activation | +1.96e-02 |
Pos logit contributions | |
tweet | +0.042 |
moment | +0.042 |
est | +0.039 |
fly | +0.038 |
Nice | +0.036 |
Neg logit contributions | |
leaf | -0.042 |
olded | -0.041 |
obb | -0.040 |
salon | -0.040 |
iza | -0.039 |
Token | fields |
Token ID | 7032 |
Feature activation | +2.24e-02 |
Pos logit contributions | |
olded | +0.034 |
he | +0.032 |
leaf | +0.031 |
outs | +0.031 |
punished | +0.031 |
Neg logit contributions | |
moment | -0.027 |
tweet | -0.027 |
feast | -0.024 |
greeting | -0.023 |
est | -0.022 |
Token | . |
Token ID | 13 |
Feature activation | +7.71e-04 |
Pos logit contributions | |
leaf | +0.031 |
olded | +0.031 |
obb | +0.030 |
iza | +0.030 |
punished | +0.030 |
Neg logit contributions | |
moment | -0.036 |
tweet | -0.035 |
fly | -0.033 |
feast | -0.032 |
Nice | -0.032 |
Token | But |
Token ID | 887 |
Feature activation | -2.16e-02 |
Pos logit contributions | |
punished | +0.001 |
he | +0.001 |
obb | +0.001 |
sick | +0.001 |
gr | +0.001 |
Neg logit contributions | |
moment | -0.001 |
tweet | -0.001 |
Nice | -0.001 |
Want | -0.001 |
feast | -0.001 |
Token | watched |
Token ID | 7342 |
Feature activation | +1.91e-02 |
Pos logit contributions | |
he | +0.019 |
olded | +0.018 |
elderly | +0.018 |
sick | +0.018 |
obb | +0.018 |
Neg logit contributions | |
moment | -0.016 |
tweet | -0.016 |
Nice | -0.014 |
Want | -0.014 |
fly | -0.014 |
Token | the |
Token ID | 262 |
Feature activation | +1.00e-02 |
Pos logit contributions | |
olded | +0.027 |
punished | +0.027 |
he | +0.027 |
gr | +0.026 |
pastor | +0.026 |
Neg logit contributions | |
moment | -0.029 |
tweet | -0.029 |
Nice | -0.027 |
Want | -0.026 |
greeting | -0.025 |
Token | squirrel |
Token ID | 33039 |
Feature activation | +2.75e-03 |
Pos logit contributions | |
he | +0.017 |
sick | +0.017 |
elderly | +0.017 |
punished | +0.017 |
olded | +0.017 |
Neg logit contributions | |
moment | -0.012 |
tweet | -0.011 |
Nice | -0.011 |
Want | -0.010 |
fly | -0.010 |
Token | as |
Token ID | 355 |
Feature activation | +1.67e-02 |
Pos logit contributions | |
he | +0.003 |
olded | +0.003 |
punished | +0.003 |
sick | +0.003 |
elderly | +0.003 |
Neg logit contributions | |
moment | -0.005 |
tweet | -0.005 |
Nice | -0.004 |
Want | -0.004 |
fly | -0.004 |
Token | it |
Token ID | 340 |
Feature activation | +2.15e-02 |
Pos logit contributions | |
olded | +0.033 |
punished | +0.031 |
salon | +0.031 |
gr | +0.031 |
leaf | +0.031 |
Neg logit contributions | |
moment | -0.021 |
tweet | -0.019 |
Nice | -0.017 |
fly | -0.016 |
hop | -0.015 |
Token | sc |
Token ID | 629 |
Feature activation | +8.19e-02 |
Pos logit contributions | |
sick | +0.030 |
he | +0.030 |
punished | +0.029 |
olded | +0.029 |
elderly | +0.028 |
Neg logit contributions | |
moment | -0.033 |
tweet | -0.032 |
Nice | -0.031 |
Want | -0.029 |
est | -0.028 |
Token | ur |
Token ID | 333 |
Feature activation | +3.89e-03 |
Pos logit contributions | |
pastor | +0.162 |
he | +0.155 |
elderly | +0.155 |
sick | +0.154 |
punished | +0.149 |
Neg logit contributions | |
tweet | -0.089 |
Nice | -0.080 |
Want | -0.071 |
moment | -0.071 |
fly | -0.070 |
Token | ried |
Token ID | 2228 |
Feature activation | -6.04e-03 |
Pos logit contributions | |
olded | +0.005 |
punished | +0.005 |
he | +0.005 |
sick | +0.005 |
elderly | +0.005 |
Neg logit contributions | |
moment | -0.006 |
tweet | -0.006 |
Nice | -0.006 |
Want | -0.006 |
feast | -0.006 |
Token | around |
Token ID | 1088 |
Feature activation | +9.60e-04 |
Pos logit contributions | |
moment | +0.011 |
tweet | +0.010 |
Nice | +0.010 |
Want | +0.010 |
feast | +0.009 |
Neg logit contributions | |
he | -0.007 |
olded | -0.007 |
sick | -0.007 |
punished | -0.007 |
elderly | -0.007 |
Token | . |
Token ID | 13 |
Feature activation | +4.56e-03 |
Pos logit contributions | |
sick | +0.001 |
he | +0.001 |
elderly | +0.001 |
olded | +0.001 |
punished | +0.001 |
Neg logit contributions | |
moment | -0.002 |
tweet | -0.002 |
Nice | -0.002 |
est | -0.002 |
Want | -0.002 |
Token | |
Token ID | 220 |
Feature activation | +4.20e-02 |
Pos logit contributions | |
olded | +0.005 |
he | +0.005 |
elderly | +0.005 |
sick | +0.005 |
late | +0.004 |
Neg logit contributions | |
tweet | -0.009 |
moment | -0.008 |
fly | -0.008 |
feast | -0.008 |
est | -0.008 |
Token |
|
Token ID | 198 |
Feature activation | -2.43e-03 |
Pos logit contributions | |
punished | +0.041 |
sick | +0.040 |
obb | +0.039 |
leaf | +0.039 |
pastor | +0.039 |
Neg logit contributions | |
moment | -0.032 |
tweet | -0.030 |
Nice | -0.029 |
Want | -0.027 |
fly | -0.026 |
Token |
|
Token ID | 198 |
Feature activation | -9.39e-03 |
Pos logit contributions | |
moment | +0.003 |
tweet | +0.003 |
Nice | +0.003 |
Want | +0.003 |
greeting | +0.003 |
Neg logit contributions | |
punished | -0.004 |
obb | -0.004 |
sick | -0.004 |
olded | -0.004 |
he | -0.004 |
Token | Mom |
Token ID | 29252 |
Feature activation | -1.73e-02 |
Pos logit contributions | |
moment | +0.014 |
tweet | +0.013 |
Nice | +0.013 |
Want | +0.012 |
greeting | +0.012 |
Neg logit contributions | |
obb | -0.014 |
sick | -0.014 |
punished | -0.013 |
olded | -0.013 |
bored | -0.013 |
Token | my |
Token ID | 1820 |
Feature activation | -1.78e-02 |
Pos logit contributions | |
moment | +0.019 |
tweet | +0.019 |
Want | +0.015 |
Nice | +0.015 |
fly | +0.015 |
Neg logit contributions | |
obb | -0.032 |
sick | -0.032 |
he | -0.031 |
gr | -0.031 |
leaf | -0.031 |
Token | showed |
Token ID | 3751 |
Feature activation | +3.39e-03 |
Pos logit contributions | |
tweet | +0.025 |
moment | +0.025 |
fly | +0.021 |
Want | +0.021 |
Nice | +0.021 |
Neg logit contributions | |
obb | -0.027 |
he | -0.027 |
gr | -0.027 |
sick | -0.027 |
leaf | -0.026 |
Token | Lucy |
Token ID | 22162 |
Feature activation | -8.03e-02 |
Pos logit contributions | |
obb | +0.005 |
earn | +0.005 |
he | +0.005 |
sick | +0.005 |
pastor | +0.005 |
Neg logit contributions | |
tweet | -0.004 |
moment | -0.004 |
est | -0.004 |
Nice | -0.004 |
fly | -0.004 |
Token | the |
Token ID | 262 |
Feature activation | -8.74e-03 |
Pos logit contributions | |
moment | +0.117 |
tweet | +0.113 |
Nice | +0.104 |
Want | +0.100 |
fly | +0.100 |
Neg logit contributions | |
he | -0.119 |
obb | -0.117 |
olded | -0.117 |
punished | -0.114 |
sick | -0.113 |
Token | therm |
Token ID | 21969 |
Feature activation | +1.66e-02 |
Pos logit contributions | |
moment | +0.013 |
tweet | +0.012 |
Nice | +0.010 |
greeting | +0.010 |
feast | +0.010 |
Neg logit contributions | |
olded | -0.014 |
he | -0.014 |
punished | -0.013 |
gr | -0.013 |
obb | -0.013 |
Token | ometer |
Token ID | 15635 |
Feature activation | -1.01e-02 |
Pos logit contributions | |
punished | +0.028 |
decide | +0.028 |
he | +0.027 |
sick | +0.027 |
leaf | +0.027 |
Neg logit contributions | |
fly | -0.020 |
moment | -0.020 |
tweet | -0.018 |
Nice | -0.017 |
est | -0.017 |
Token | . |
Token ID | 13 |
Feature activation | -6.03e-03 |
Pos logit contributions | |
moment | +0.016 |
tweet | +0.015 |
Nice | +0.014 |
fly | +0.014 |
Want | +0.013 |
Neg logit contributions | |
he | -0.014 |
obb | -0.014 |
olded | -0.014 |
leaf | -0.014 |
punished | -0.014 |
Token | It |
Token ID | 632 |
Feature activation | -2.79e-02 |
Pos logit contributions | |
moment | +0.009 |
tweet | +0.009 |
Nice | +0.008 |
Want | +0.008 |
fly | +0.007 |
Neg logit contributions | |
he | -0.009 |
punished | -0.009 |
sick | -0.009 |
olded | -0.009 |
obb | -0.008 |
Token | . |
Token ID | 13 |
Feature activation | -1.05e-02 |
Pos logit contributions | |
olded | +0.002 |
leaf | +0.002 |
obb | +0.002 |
punished | +0.002 |
he | +0.002 |
Neg logit contributions | |
moment | -0.002 |
Nice | -0.002 |
tweet | -0.002 |
fly | -0.002 |
est | -0.002 |
Token | The |
Token ID | 383 |
Feature activation | +4.83e-04 |
Pos logit contributions | |
moment | +0.013 |
tweet | +0.013 |
feast | +0.011 |
Nice | +0.011 |
fly | +0.011 |
Neg logit contributions | |
olded | -0.017 |
elderly | -0.017 |
sick | -0.017 |
he | -0.017 |
punished | -0.017 |
Token | planet |
Token ID | 5440 |
Feature activation | -1.88e-03 |
Pos logit contributions | |
olded | +0.001 |
he | +0.001 |
punished | +0.001 |
sick | +0.001 |
elderly | +0.001 |
Neg logit contributions | |
moment | -0.000 |
tweet | -0.000 |
Nice | -0.000 |
Want | -0.000 |
feast | -0.000 |
Token | was |
Token ID | 373 |
Feature activation | -9.46e-03 |
Pos logit contributions | |
tweet | +0.002 |
moment | +0.002 |
Nice | +0.002 |
Want | +0.002 |
fly | +0.002 |
Neg logit contributions | |
he | -0.003 |
sick | -0.003 |
olded | -0.003 |
punished | -0.003 |
elderly | -0.003 |
Token | called |
Token ID | 1444 |
Feature activation | -4.08e-02 |
Pos logit contributions | |
moment | +0.017 |
tweet | +0.017 |
Nice | +0.016 |
Want | +0.015 |
fly | +0.015 |
Neg logit contributions | |
he | -0.011 |
olded | -0.011 |
sick | -0.010 |
punished | -0.010 |
elderly | -0.010 |
Token | Pluto |
Token ID | 32217 |
Feature activation | -7.95e-02 |
Pos logit contributions | |
moment | +0.060 |
tweet | +0.058 |
Nice | +0.055 |
Want | +0.053 |
feast | +0.052 |
Neg logit contributions | |
elderly | -0.057 |
sick | -0.057 |
he | -0.057 |
olded | -0.056 |
leaf | -0.054 |
Token | , |
Token ID | 11 |
Feature activation | -1.78e-02 |
Pos logit contributions | |
moment | +0.151 |
tweet | +0.149 |
Nice | +0.141 |
Want | +0.137 |
fly | +0.134 |
Neg logit contributions | |
he | -0.080 |
olded | -0.079 |
sick | -0.078 |
punished | -0.077 |
elderly | -0.076 |
Token | and |
Token ID | 290 |
Feature activation | -3.57e-03 |
Pos logit contributions | |
moment | +0.029 |
tweet | +0.028 |
Nice | +0.027 |
feast | +0.026 |
est | +0.025 |
Neg logit contributions | |
sick | -0.023 |
elderly | -0.023 |
olded | -0.022 |
he | -0.022 |
punished | -0.022 |
Token | it |
Token ID | 340 |
Feature activation | -1.17e-02 |
Pos logit contributions | |
moment | +0.006 |
tweet | +0.006 |
Nice | +0.006 |
Want | +0.005 |
feast | +0.005 |
Neg logit contributions | |
sick | -0.004 |
elderly | -0.004 |
olded | -0.004 |
punished | -0.004 |
he | -0.004 |
Token | was |
Token ID | 373 |
Feature activation | -9.32e-03 |
Pos logit contributions | |
moment | +0.015 |
tweet | +0.015 |
Nice | +0.014 |
Want | +0.013 |
feast | +0.013 |
Neg logit contributions | |
he | -0.019 |
olded | -0.019 |
sick | -0.018 |
punished | -0.018 |
elderly | -0.018 |
Token | very |
Token ID | 845 |
Feature activation | +1.86e-03 |
Pos logit contributions | |
moment | +0.017 |
tweet | +0.016 |
Nice | +0.015 |
Want | +0.015 |
fly | +0.015 |
Neg logit contributions | |
he | -0.011 |
olded | -0.011 |
punished | -0.010 |
sick | -0.010 |
elderly | -0.010 |
Token |
|
Token ID | 198 |
Feature activation | +4.77e-04 |
Pos logit contributions | |
punished | +0.061 |
sick | +0.059 |
obb | +0.059 |
pastor | +0.058 |
leaf | +0.058 |
Neg logit contributions | |
moment | -0.037 |
tweet | -0.033 |
Nice | -0.033 |
elcome | -0.031 |
Want | -0.030 |
Token |
|
Token ID | 198 |
Feature activation | -6.93e-03 |
Pos logit contributions | |
obb | +0.001 |
punished | +0.001 |
sick | +0.001 |
bored | +0.001 |
olded | +0.001 |
Neg logit contributions | |
moment | -0.001 |
tweet | -0.001 |
Nice | -0.001 |
Want | -0.001 |
greeting | -0.001 |
Token | M |
Token ID | 44 |
Feature activation | +2.28e-02 |
Pos logit contributions | |
moment | +0.011 |
tweet | +0.010 |
Nice | +0.010 |
Want | +0.010 |
greeting | +0.010 |
Neg logit contributions | |
sick | -0.010 |
obb | -0.010 |
bored | -0.010 |
punished | -0.009 |
outs | -0.009 |
Token | ama |
Token ID | 1689 |
Feature activation | -2.24e-02 |
Pos logit contributions | |
bored | +0.045 |
sick | +0.045 |
hes | +0.044 |
punished | +0.043 |
gr | +0.043 |
Neg logit contributions | |
moment | -0.022 |
tweet | -0.021 |
feast | -0.019 |
est | -0.019 |
reetings | -0.019 |
Token | showed |
Token ID | 3751 |
Feature activation | +1.08e-02 |
Pos logit contributions | |
moment | +0.035 |
tweet | +0.035 |
Nice | +0.032 |
fly | +0.031 |
Want | +0.030 |
Neg logit contributions | |
sick | -0.030 |
he | -0.030 |
obb | -0.030 |
olded | -0.029 |
punished | -0.029 |
Token | Baby |
Token ID | 14801 |
Feature activation | -7.77e-02 |
Pos logit contributions | |
obb | +0.015 |
sick | +0.015 |
punished | +0.015 |
earn | +0.015 |
he | +0.014 |
Neg logit contributions | |
moment | -0.016 |
tweet | -0.016 |
est | -0.015 |
Nice | -0.015 |
greeting | -0.014 |
Token | how |
Token ID | 703 |
Feature activation | +1.55e-02 |
Pos logit contributions | |
moment | +0.126 |
tweet | +0.122 |
Nice | +0.115 |
Want | +0.110 |
fly | +0.108 |
Neg logit contributions | |
he | -0.102 |
olded | -0.101 |
sick | -0.099 |
punished | -0.099 |
elderly | -0.097 |
Token | to |
Token ID | 284 |
Feature activation | -5.59e-03 |
Pos logit contributions | |
obb | +0.033 |
he | +0.033 |
salon | +0.031 |
ateful | +0.030 |
ussy | +0.030 |
Neg logit contributions | |
moment | -0.016 |
Nice | -0.012 |
tweet | -0.012 |
greeting | -0.012 |
hopping | -0.011 |
Token | rip |
Token ID | 18344 |
Feature activation | -1.78e-02 |
Pos logit contributions | |
tweet | +0.007 |
moment | +0.007 |
hop | +0.006 |
greeting | +0.006 |
feast | +0.006 |
Neg logit contributions | |
olded | -0.010 |
outs | -0.009 |
bored | -0.009 |
outed | -0.009 |
obb | -0.009 |
Token | the |
Token ID | 262 |
Feature activation | -6.77e-03 |
Pos logit contributions | |
moment | +0.027 |
tweet | +0.025 |
est | +0.024 |
Nice | +0.024 |
fly | +0.023 |
Neg logit contributions | |
olded | -0.025 |
he | -0.025 |
sick | -0.024 |
punished | -0.024 |
elderly | -0.024 |
Token | pieces |
Token ID | 5207 |
Feature activation | +6.62e-03 |
Pos logit contributions | |
moment | +0.009 |
tweet | +0.009 |
est | +0.008 |
Nice | +0.008 |
fly | +0.008 |
Neg logit contributions | |
olded | -0.011 |
he | -0.011 |
punished | -0.010 |
pastor | -0.010 |
ateful | -0.010 |
Token | my |
Token ID | 1820 |
Feature activation | -9.44e-03 |
Pos logit contributions | |
moment | +0.012 |
tweet | +0.011 |
Nice | +0.011 |
Want | +0.010 |
feast | +0.009 |
Neg logit contributions | |
he | -0.017 |
olded | -0.017 |
punished | -0.016 |
elderly | -0.016 |
sick | -0.016 |
Token | said |
Token ID | 531 |
Feature activation | -4.09e-02 |
Pos logit contributions | |
moment | +0.017 |
tweet | +0.017 |
Nice | +0.015 |
Want | +0.015 |
fly | +0.015 |
Neg logit contributions | |
he | -0.011 |
sick | -0.010 |
olded | -0.010 |
punished | -0.010 |
obb | -0.010 |
Token | , |
Token ID | 11 |
Feature activation | -1.95e-02 |
Pos logit contributions | |
moment | +0.039 |
tweet | +0.038 |
Nice | +0.034 |
Want | +0.032 |
fly | +0.030 |
Neg logit contributions | |
he | -0.081 |
olded | -0.080 |
sick | -0.079 |
punished | -0.079 |
elderly | -0.078 |
Token | " |
Token ID | 366 |
Feature activation | -1.28e-03 |
Pos logit contributions | |
moment | +0.020 |
tweet | +0.019 |
Nice | +0.018 |
Want | +0.017 |
fly | +0.016 |
Neg logit contributions | |
he | -0.037 |
olded | -0.037 |
punished | -0.037 |
sick | -0.036 |
elderly | -0.036 |
Token | Gu |
Token ID | 8205 |
Feature activation | -2.81e-02 |
Pos logit contributions | |
moment | +0.002 |
tweet | +0.002 |
Nice | +0.002 |
Want | +0.002 |
greeting | +0.002 |
Neg logit contributions | |
sick | -0.002 |
obb | -0.002 |
olded | -0.002 |
punished | -0.002 |
bored | -0.002 |
Token | ess |
Token ID | 408 |
Feature activation | -7.67e-02 |
Pos logit contributions | |
moment | +0.031 |
est | +0.029 |
tweet | +0.028 |
Nice | +0.027 |
fly | +0.026 |
Neg logit contributions | |
punished | -0.052 |
leaf | -0.051 |
sick | -0.051 |
bored | -0.050 |
he | -0.050 |
Token | what |
Token ID | 644 |
Feature activation | -4.35e-03 |
Pos logit contributions | |
moment | +0.091 |
tweet | +0.090 |
Nice | +0.080 |
fly | +0.076 |
Want | +0.076 |
Neg logit contributions | |
he | -0.136 |
olded | -0.132 |
sick | -0.129 |
punished | -0.129 |
obb | -0.129 |
Token | ? |
Token ID | 30 |
Feature activation | +4.27e-03 |
Pos logit contributions | |
moment | +0.006 |
tweet | +0.006 |
Nice | +0.006 |
Want | +0.005 |
feast | +0.005 |
Neg logit contributions | |
he | -0.007 |
olded | -0.007 |
punished | -0.007 |
obb | -0.006 |
sick | -0.006 |
Token | You |
Token ID | 921 |
Feature activation | -4.76e-03 |
Pos logit contributions | |
he | +0.007 |
obb | +0.007 |
punished | +0.007 |
sick | +0.007 |
elderly | +0.007 |
Neg logit contributions | |
tweet | -0.005 |
moment | -0.005 |
Nice | -0.005 |
Want | -0.005 |
fly | -0.004 |
Token | get |
Token ID | 651 |
Feature activation | +7.99e-03 |
Pos logit contributions | |
tweet | +0.006 |
moment | +0.006 |
fly | +0.005 |
feast | +0.005 |
Want | +0.005 |
Neg logit contributions | |
outs | -0.008 |
elderly | -0.008 |
olded | -0.008 |
pastor | -0.008 |
he | -0.008 |
Token | to |
Token ID | 284 |
Feature activation | -1.54e-02 |
Pos logit contributions | |
he | +0.014 |
obb | +0.013 |
der | +0.012 |
pastor | +0.012 |
ateful | +0.012 |
Neg logit contributions | |
moment | -0.010 |
tweet | -0.009 |
Want | -0.009 |
greeting | -0.009 |
elcome | -0.008 |
Token | chess |
Token ID | 19780 |
Feature activation | +7.38e-03 |
Pos logit contributions | |
moment | +0.011 |
tweet | +0.010 |
Nice | +0.010 |
Want | +0.009 |
fly | +0.009 |
Neg logit contributions | |
he | -0.009 |
olded | -0.009 |
sick | -0.009 |
punished | -0.009 |
elderly | -0.009 |
Token | . |
Token ID | 13 |
Feature activation | -3.05e-03 |
Pos logit contributions | |
olded | +0.010 |
punished | +0.010 |
sick | +0.010 |
leaf | +0.009 |
pastor | +0.009 |
Neg logit contributions | |
moment | -0.012 |
tweet | -0.011 |
Nice | -0.011 |
est | -0.010 |
Want | -0.010 |
Token | He |
Token ID | 679 |
Feature activation | -8.73e-03 |
Pos logit contributions | |
Nice | +0.004 |
moment | +0.004 |
Want | +0.004 |
tweet | +0.004 |
greeting | +0.004 |
Neg logit contributions | |
punished | -0.005 |
sick | -0.005 |
obb | -0.005 |
he | -0.005 |
disob | -0.005 |
Token | showed |
Token ID | 3751 |
Feature activation | +1.19e-02 |
Pos logit contributions | |
moment | +0.012 |
tweet | +0.012 |
Nice | +0.011 |
Want | +0.011 |
fly | +0.010 |
Neg logit contributions | |
he | -0.013 |
sick | -0.013 |
obb | -0.013 |
olded | -0.013 |
pastor | -0.013 |
Token | Mate |
Token ID | 24787 |
Feature activation | -3.82e-02 |
Pos logit contributions | |
he | +0.015 |
sick | +0.015 |
olded | +0.015 |
punished | +0.015 |
obb | +0.015 |
Neg logit contributions | |
moment | -0.019 |
tweet | -0.019 |
Nice | -0.018 |
est | -0.017 |
Want | -0.017 |
Token | o |
Token ID | 78 |
Feature activation | -7.59e-02 |
Pos logit contributions | |
moment | +0.049 |
tweet | +0.048 |
Nice | +0.044 |
Want | +0.042 |
feast | +0.041 |
Neg logit contributions | |
olded | -0.062 |
he | -0.062 |
sick | -0.061 |
elderly | -0.061 |
punished | -0.061 |
Token | how |
Token ID | 703 |
Feature activation | +1.49e-02 |
Pos logit contributions | |
moment | +0.117 |
tweet | +0.115 |
Nice | +0.110 |
Want | +0.105 |
feast | +0.104 |
Neg logit contributions | |
olded | -0.101 |
sick | -0.099 |
he | -0.097 |
outs | -0.097 |
elderly | -0.097 |
Token | to |
Token ID | 284 |
Feature activation | -6.65e-03 |
Pos logit contributions | |
he | +0.032 |
pastor | +0.030 |
iza | +0.029 |
paramedics | +0.029 |
obb | +0.029 |
Neg logit contributions | |
moment | -0.015 |
Want | -0.013 |
Nice | -0.013 |
tweet | -0.011 |
hopping | -0.010 |
Token | move |
Token ID | 1445 |
Feature activation | -2.29e-03 |
Pos logit contributions | |
moment | +0.009 |
tweet | +0.009 |
Want | +0.008 |
Nice | +0.008 |
greeting | +0.008 |
Neg logit contributions | |
olded | -0.010 |
pastor | -0.010 |
sick | -0.010 |
outs | -0.010 |
he | -0.010 |
Token | the |
Token ID | 262 |
Feature activation | -7.44e-03 |
Pos logit contributions | |
moment | +0.003 |
tweet | +0.003 |
Nice | +0.003 |
Want | +0.003 |
est | +0.003 |
Neg logit contributions | |
olded | -0.003 |
he | -0.003 |
punished | -0.003 |
pastor | -0.003 |
elderly | -0.003 |
Token | pieces |
Token ID | 5207 |
Feature activation | +5.72e-03 |
Pos logit contributions | |
moment | +0.011 |
tweet | +0.011 |
Nice | +0.010 |
Want | +0.009 |
feast | +0.009 |
Neg logit contributions | |
olded | -0.011 |
he | -0.011 |
punished | -0.011 |
sick | -0.011 |
elderly | -0.010 |
Token | pa |
Token ID | 8957 |
Feature activation | -2.45e-02 |
Pos logit contributions | |
tweet | +0.010 |
moment | +0.010 |
Nice | +0.010 |
Want | +0.009 |
fly | +0.009 |
Neg logit contributions | |
elderly | -0.008 |
he | -0.008 |
sick | -0.008 |
punished | -0.008 |
pastor | -0.008 |
Token | . |
Token ID | 13 |
Feature activation | +8.59e-04 |
Pos logit contributions | |
moment | +0.035 |
tweet | +0.032 |
Nice | +0.031 |
Want | +0.030 |
feast | +0.030 |
Neg logit contributions | |
obb | -0.039 |
olded | -0.037 |
punished | -0.037 |
leaf | -0.037 |
pastor | -0.037 |
Token | Grand |
Token ID | 5675 |
Feature activation | +4.55e-04 |
Pos logit contributions | |
olded | +0.001 |
he | +0.001 |
punished | +0.001 |
sick | +0.001 |
elderly | +0.001 |
Neg logit contributions | |
moment | -0.001 |
tweet | -0.001 |
Nice | -0.001 |
Want | -0.001 |
greeting | -0.001 |
Token | pa |
Token ID | 8957 |
Feature activation | -2.56e-02 |
Pos logit contributions | |
he | +0.001 |
olded | +0.001 |
elderly | +0.001 |
punished | +0.000 |
sick | +0.000 |
Neg logit contributions | |
tweet | -0.001 |
moment | -0.001 |
Nice | -0.001 |
Want | -0.001 |
fly | -0.001 |
Token | showed |
Token ID | 3751 |
Feature activation | +1.65e-02 |
Pos logit contributions | |
moment | +0.031 |
tweet | +0.029 |
Nice | +0.027 |
Want | +0.026 |
fly | +0.026 |
Neg logit contributions | |
obb | -0.044 |
he | -0.044 |
sick | -0.042 |
leaf | -0.042 |
olded | -0.042 |
Token | John |
Token ID | 1757 |
Feature activation | -7.44e-02 |
Pos logit contributions | |
obb | +0.024 |
sick | +0.023 |
punished | +0.023 |
olded | +0.023 |
bored | +0.022 |
Neg logit contributions | |
moment | -0.024 |
est | -0.023 |
tweet | -0.022 |
Nice | -0.022 |
greeting | -0.021 |
Token | something |
Token ID | 1223 |
Feature activation | +8.97e-03 |
Pos logit contributions | |
moment | +0.119 |
tweet | +0.112 |
Nice | +0.107 |
Want | +0.101 |
fly | +0.101 |
Neg logit contributions | |
obb | -0.100 |
he | -0.100 |
olded | -0.099 |
punished | -0.098 |
sick | -0.095 |
Token | wide |
Token ID | 3094 |
Feature activation | -4.87e-03 |
Pos logit contributions | |
he | +0.012 |
olded | +0.012 |
punished | +0.012 |
obb | +0.012 |
gr | +0.011 |
Neg logit contributions | |
moment | -0.014 |
tweet | -0.013 |
Nice | -0.013 |
Want | -0.012 |
est | -0.012 |
Token | and |
Token ID | 290 |
Feature activation | +1.40e-02 |
Pos logit contributions | |
moment | +0.009 |
tweet | +0.009 |
Nice | +0.008 |
Want | +0.008 |
fly | +0.008 |
Neg logit contributions | |
he | -0.005 |
olded | -0.005 |
sick | -0.005 |
punished | -0.005 |
elderly | -0.005 |
Token | modern |
Token ID | 3660 |
Feature activation | -5.72e-03 |
Pos logit contributions | |
he | +0.020 |
obb | +0.018 |
olded | +0.018 |
outs | +0.018 |
leaf | +0.018 |
Neg logit contributions | |
moment | -0.023 |
tweet | -0.021 |
Nice | -0.021 |
Want | -0.020 |
greeting | -0.019 |
Token | . |
Token ID | 13 |
Feature activation | -4.28e-03 |
Pos logit contributions | |
moment | +0.009 |
tweet | +0.008 |
Nice | +0.008 |
est | +0.007 |
Want | +0.007 |
Neg logit contributions | |
he | -0.008 |
olded | -0.008 |
punished | -0.008 |
obb | -0.008 |
sick | -0.008 |
Token | also |
Token ID | 635 |
Feature activation | -2.73e-02 |
Pos logit contributions | |
moment | +0.005 |
tweet | +0.004 |
greeting | +0.004 |
Nice | +0.004 |
Want | +0.004 |
Neg logit contributions | |
obb | -0.004 |
he | -0.004 |
olded | -0.004 |
leaf | -0.004 |
earn | -0.004 |
Token | excited |
Token ID | 6568 |
Feature activation | +1.39e-02 |
Pos logit contributions | |
moment | +0.044 |
tweet | +0.042 |
Nice | +0.038 |
greeting | +0.038 |
Want | +0.038 |
Neg logit contributions | |
obb | -0.037 |
he | -0.036 |
olded | -0.035 |
leaf | -0.035 |
sick | -0.034 |
Token | . |
Token ID | 13 |
Feature activation | -1.19e-02 |
Pos logit contributions | |
he | +0.018 |
olded | +0.018 |
sick | +0.018 |
punished | +0.018 |
elderly | +0.017 |
Neg logit contributions | |
moment | -0.023 |
tweet | -0.022 |
Nice | -0.021 |
Want | -0.020 |
fly | -0.019 |
Token | She |
Token ID | 1375 |
Feature activation | -5.66e-03 |
Pos logit contributions | |
Nice | +0.020 |
moment | +0.019 |
tweet | +0.019 |
Want | +0.018 |
greeting | +0.018 |
Neg logit contributions | |
he | -0.016 |
punished | -0.016 |
sick | -0.015 |
obb | -0.015 |
gr | -0.015 |
Token | gave |
Token ID | 2921 |
Feature activation | -4.73e-04 |
Pos logit contributions | |
moment | +0.009 |
tweet | +0.009 |
Nice | +0.008 |
Want | +0.008 |
feast | +0.008 |
Neg logit contributions | |
obb | -0.008 |
he | -0.008 |
sick | -0.007 |
ateful | -0.007 |
gr | -0.007 |
Token | Ginger |
Token ID | 38682 |
Feature activation | -7.44e-02 |
Pos logit contributions | |
moment | +0.001 |
tweet | +0.001 |
Nice | +0.001 |
Want | +0.001 |
fly | +0.001 |
Neg logit contributions | |
he | -0.001 |
olded | -0.001 |
sick | -0.001 |
punished | -0.001 |
elderly | -0.001 |
Token | a |
Token ID | 257 |
Feature activation | -2.27e-02 |
Pos logit contributions | |
moment | +0.137 |
tweet | +0.133 |
Nice | +0.126 |
Want | +0.121 |
fly | +0.119 |
Neg logit contributions | |
he | -0.082 |
olded | -0.080 |
punished | -0.079 |
sick | -0.078 |
obb | -0.077 |
Token | big |
Token ID | 1263 |
Feature activation | -3.28e-02 |
Pos logit contributions | |
moment | +0.037 |
tweet | +0.036 |
Nice | +0.033 |
Want | +0.032 |
feast | +0.031 |
Neg logit contributions | |
olded | -0.030 |
he | -0.030 |
punished | -0.029 |
sick | -0.029 |
obb | -0.029 |
Token | smile |
Token ID | 8212 |
Feature activation | -2.15e-02 |
Pos logit contributions | |
moment | +0.052 |
tweet | +0.051 |
Nice | +0.050 |
fly | +0.048 |
Want | +0.047 |
Neg logit contributions | |
he | -0.042 |
elderly | -0.041 |
sick | -0.040 |
leaf | -0.040 |
olded | -0.040 |
Token | and |
Token ID | 290 |
Feature activation | -1.48e-02 |
Pos logit contributions | |
moment | +0.031 |
tweet | +0.029 |
Nice | +0.027 |
est | +0.026 |
greeting | +0.025 |
Neg logit contributions | |
punished | -0.034 |
olded | -0.034 |
obb | -0.034 |
sick | -0.033 |
gr | -0.032 |
Token | said |
Token ID | 531 |
Feature activation | -4.15e-02 |
Pos logit contributions | |
moment | +0.025 |
tweet | +0.024 |
Nice | +0.022 |
elcome | +0.021 |
greeting | +0.021 |
Neg logit contributions | |
he | -0.021 |
obb | -0.020 |
outs | -0.019 |
gr | -0.019 |
ateful | -0.019 |
Token | girl |
Token ID | 2576 |
Feature activation | +3.02e-03 |
Pos logit contributions | |
moment | +0.001 |
tweet | +0.001 |
Nice | +0.001 |
fly | +0.001 |
Want | +0.001 |
Neg logit contributions | |
olded | -0.001 |
he | -0.001 |
punished | -0.001 |
leaf | -0.001 |
outs | -0.001 |
Token | . |
Token ID | 13 |
Feature activation | +1.46e-03 |
Pos logit contributions | |
obb | +0.005 |
he | +0.005 |
leaf | +0.005 |
pastor | +0.005 |
olded | +0.005 |
Neg logit contributions | |
moment | -0.004 |
tweet | -0.004 |
fly | -0.004 |
Nice | -0.003 |
greeting | -0.003 |
Token | She |
Token ID | 1375 |
Feature activation | -4.97e-03 |
Pos logit contributions | |
punished | +0.002 |
he | +0.002 |
olded | +0.001 |
sick | +0.001 |
obb | +0.001 |
Neg logit contributions | |
moment | -0.003 |
tweet | -0.003 |
Nice | -0.003 |
Want | -0.003 |
greeting | -0.002 |
Token | was |
Token ID | 373 |
Feature activation | -1.65e-03 |
Pos logit contributions | |
moment | +0.006 |
tweet | +0.005 |
Nice | +0.005 |
Want | +0.005 |
fly | +0.005 |
Neg logit contributions | |
obb | -0.009 |
he | -0.009 |
ateful | -0.009 |
leaf | -0.008 |
outs | -0.008 |
Token | called |
Token ID | 1444 |
Feature activation | -2.90e-02 |
Pos logit contributions | |
tweet | +0.002 |
moment | +0.002 |
greeting | +0.002 |
Nice | +0.002 |
Want | +0.002 |
Neg logit contributions | |
he | -0.002 |
obb | -0.002 |
olded | -0.002 |
outs | -0.002 |
leaf | -0.002 |
Token | Daisy |
Token ID | 40355 |
Feature activation | -7.40e-02 |
Pos logit contributions | |
moment | +0.046 |
tweet | +0.046 |
Nice | +0.043 |
Want | +0.041 |
fly | +0.040 |
Neg logit contributions | |
olded | -0.039 |
he | -0.039 |
punished | -0.038 |
sick | -0.037 |
gr | -0.036 |
Token | and |
Token ID | 290 |
Feature activation | -1.99e-03 |
Pos logit contributions | |
moment | +0.132 |
tweet | +0.129 |
Nice | +0.122 |
Want | +0.117 |
fly | +0.116 |
Neg logit contributions | |
he | -0.086 |
olded | -0.084 |
punished | -0.083 |
obb | -0.082 |
sick | -0.082 |
Token | she |
Token ID | 673 |
Feature activation | -2.69e-03 |
Pos logit contributions | |
moment | +0.003 |
tweet | +0.003 |
Nice | +0.003 |
Want | +0.003 |
greeting | +0.003 |
Neg logit contributions | |
he | -0.003 |
obb | -0.003 |
olded | -0.002 |
gr | -0.002 |
outs | -0.002 |
Token | was |
Token ID | 373 |
Feature activation | -2.48e-03 |
Pos logit contributions | |
moment | +0.003 |
tweet | +0.003 |
Nice | +0.003 |
fly | +0.003 |
Want | +0.003 |
Neg logit contributions | |
obb | -0.005 |
he | -0.005 |
leaf | -0.005 |
ateful | -0.005 |
outs | -0.005 |
Token | three |
Token ID | 1115 |
Feature activation | -2.19e-03 |
Pos logit contributions | |
tweet | +0.003 |
moment | +0.003 |
greeting | +0.003 |
hopping | +0.003 |
Nice | +0.003 |
Neg logit contributions | |
he | -0.004 |
obb | -0.004 |
olded | -0.004 |
earn | -0.004 |
outs | -0.004 |
Token | years |
Token ID | 812 |
Feature activation | -4.36e-03 |
Pos logit contributions | |
moment | +0.003 |
tweet | +0.003 |
est | +0.003 |
Nice | +0.003 |
greeting | +0.003 |
Neg logit contributions | |
he | -0.003 |
olded | -0.003 |
obb | -0.003 |
leaf | -0.003 |
punished | -0.003 |
Token |
|
Token ID | 198 |
Feature activation | +3.89e-03 |
Pos logit contributions | |
moment | +0.012 |
tweet | +0.011 |
Nice | +0.011 |
Want | +0.010 |
greeting | +0.010 |
Neg logit contributions | |
olded | -0.010 |
he | -0.010 |
punished | -0.009 |
leaf | -0.009 |
iza | -0.009 |
Token |
|
Token ID | 198 |
Feature activation | -2.04e-03 |
Pos logit contributions | |
olded | +0.006 |
punished | +0.006 |
sick | +0.006 |
he | +0.006 |
obb | +0.006 |
Neg logit contributions | |
moment | -0.006 |
tweet | -0.005 |
Nice | -0.005 |
Want | -0.005 |
greeting | -0.005 |
Token | Mom |
Token ID | 29252 |
Feature activation | -9.68e-03 |
Pos logit contributions | |
moment | +0.003 |
tweet | +0.003 |
Nice | +0.003 |
Want | +0.003 |
greeting | +0.003 |
Neg logit contributions | |
olded | -0.003 |
sick | -0.003 |
obb | -0.003 |
punished | -0.003 |
he | -0.003 |
Token | my |
Token ID | 1820 |
Feature activation | -1.01e-02 |
Pos logit contributions | |
tweet | +0.014 |
moment | +0.014 |
Nice | +0.013 |
Want | +0.012 |
fly | +0.012 |
Neg logit contributions | |
sick | -0.014 |
he | -0.014 |
olded | -0.014 |
gr | -0.013 |
obb | -0.013 |
Token | showed |
Token ID | 3751 |
Feature activation | +7.13e-03 |
Pos logit contributions | |
tweet | +0.015 |
moment | +0.015 |
Nice | +0.013 |
Want | +0.013 |
fly | +0.013 |
Neg logit contributions | |
he | -0.015 |
sick | -0.015 |
gr | -0.014 |
obb | -0.014 |
elderly | -0.014 |
Token | Sammy |
Token ID | 44067 |
Feature activation | -7.32e-02 |
Pos logit contributions | |
obb | +0.012 |
earn | +0.012 |
he | +0.011 |
sick | +0.011 |
punished | +0.011 |
Neg logit contributions | |
tweet | -0.009 |
moment | -0.008 |
est | -0.008 |
Nice | -0.008 |
fly | -0.007 |
Token | how |
Token ID | 703 |
Feature activation | +1.82e-02 |
Pos logit contributions | |
moment | +0.107 |
tweet | +0.101 |
Nice | +0.093 |
dy | +0.090 |
fly | +0.089 |
Neg logit contributions | |
obb | -0.109 |
he | -0.108 |
olded | -0.107 |
pastor | -0.105 |
punished | -0.105 |
Token | to |
Token ID | 284 |
Feature activation | -5.46e-03 |
Pos logit contributions | |
he | +0.039 |
salon | +0.037 |
obb | +0.037 |
pastor | +0.037 |
iza | +0.036 |
Neg logit contributions | |
moment | -0.018 |
Nice | -0.017 |
Want | -0.016 |
tweet | -0.015 |
hopping | -0.014 |
Token | use |
Token ID | 779 |
Feature activation | +2.08e-03 |
Pos logit contributions | |
tweet | +0.007 |
moment | +0.007 |
hop | +0.006 |
feast | +0.006 |
fly | +0.006 |
Neg logit contributions | |
olded | -0.009 |
outs | -0.009 |
bored | -0.009 |
outed | -0.009 |
pastor | -0.009 |
Token | the |
Token ID | 262 |
Feature activation | -1.03e-02 |
Pos logit contributions | |
he | +0.003 |
olded | +0.003 |
pastor | +0.003 |
punished | +0.003 |
gr | +0.003 |
Neg logit contributions | |
moment | -0.003 |
tweet | -0.003 |
Nice | -0.003 |
Want | -0.003 |
greeting | -0.003 |
Token | oven |
Token ID | 14361 |
Feature activation | -3.12e-03 |
Pos logit contributions | |
moment | +0.015 |
tweet | +0.014 |
Nice | +0.013 |
feast | +0.013 |
est | +0.012 |
Neg logit contributions | |
he | -0.016 |
olded | -0.016 |
punished | -0.015 |
gr | -0.015 |
iza | -0.015 |
Token | and |
Token ID | 290 |
Feature activation | -9.06e-04 |
Pos logit contributions | |
moment | +0.073 |
tweet | +0.070 |
Nice | +0.064 |
Want | +0.060 |
fly | +0.060 |
Neg logit contributions | |
he | -0.083 |
obb | -0.080 |
punished | -0.080 |
gr | -0.080 |
olded | -0.080 |
Token | Ben |
Token ID | 3932 |
Feature activation | -3.40e-02 |
Pos logit contributions | |
tweet | +0.001 |
moment | +0.001 |
Nice | +0.001 |
Want | +0.001 |
fly | +0.001 |
Neg logit contributions | |
he | -0.001 |
obb | -0.001 |
gr | -0.001 |
outs | -0.001 |
olded | -0.001 |
Token | , |
Token ID | 11 |
Feature activation | -1.32e-02 |
Pos logit contributions | |
moment | +0.052 |
tweet | +0.049 |
Nice | +0.043 |
feast | +0.042 |
greeting | +0.041 |
Neg logit contributions | |
obb | -0.050 |
he | -0.050 |
elderly | -0.049 |
gr | -0.049 |
leaf | -0.047 |
Token | " |
Token ID | 366 |
Feature activation | +7.25e-03 |
Pos logit contributions | |
moment | +0.017 |
tweet | +0.017 |
Nice | +0.015 |
Want | +0.015 |
greeting | +0.014 |
Neg logit contributions | |
he | -0.022 |
olded | -0.022 |
punished | -0.021 |
obb | -0.021 |
sick | -0.021 |
Token | Gu |
Token ID | 8205 |
Feature activation | -2.46e-02 |
Pos logit contributions | |
olded | +0.010 |
obb | +0.010 |
outs | +0.010 |
sick | +0.010 |
punished | +0.009 |
Neg logit contributions | |
moment | -0.013 |
tweet | -0.012 |
Want | -0.011 |
Nice | -0.010 |
greeting | -0.010 |
Token | ess |
Token ID | 408 |
Feature activation | -7.30e-02 |
Pos logit contributions | |
moment | +0.030 |
est | +0.027 |
tweet | +0.027 |
Nice | +0.026 |
fly | +0.024 |
Neg logit contributions | |
punished | -0.043 |
leaf | -0.042 |
sick | -0.042 |
he | -0.041 |
olded | -0.041 |
Token | what |
Token ID | 644 |
Feature activation | -3.93e-03 |
Pos logit contributions | |
tweet | +0.088 |
moment | +0.088 |
Nice | +0.076 |
fly | +0.074 |
Want | +0.073 |
Neg logit contributions | |
he | -0.129 |
olded | -0.125 |
obb | -0.125 |
punished | -0.123 |
sick | -0.122 |
Token | , |
Token ID | 11 |
Feature activation | +5.22e-03 |
Pos logit contributions | |
moment | +0.006 |
tweet | +0.006 |
Nice | +0.005 |
Want | +0.005 |
feast | +0.005 |
Neg logit contributions | |
he | -0.006 |
olded | -0.006 |
punished | -0.006 |
obb | -0.006 |
gr | -0.006 |
Token | kids |
Token ID | 3988 |
Feature activation | +6.72e-03 |
Pos logit contributions | |
he | +0.007 |
olded | +0.007 |
sick | +0.007 |
elderly | +0.007 |
punished | +0.007 |
Neg logit contributions | |
moment | -0.008 |
tweet | -0.008 |
Nice | -0.008 |
Want | -0.007 |
fly | -0.007 |
Token | . |
Token ID | 13 |
Feature activation | -2.13e-02 |
Pos logit contributions | |
obb | +0.011 |
gr | +0.010 |
outed | +0.010 |
elderly | +0.010 |
olded | +0.010 |
Neg logit contributions | |
tweet | -0.009 |
moment | -0.009 |
feast | -0.008 |
Want | -0.007 |
Nice | -0.007 |
Token | The |
Token ID | 383 |
Feature activation | -1.70e-02 |
Pos logit contributions | |
tweet | +0.028 |
moment | +0.028 |
Nice | +0.027 |
Want | +0.026 |
fly | +0.024 |
Neg logit contributions | |
he | -0.035 |
punished | -0.034 |
sick | -0.034 |
obb | -0.034 |
olded | -0.033 |
Token | was |
Token ID | 373 |
Feature activation | +9.98e-03 |
Pos logit contributions | |
tweet | +0.047 |
greeting | +0.042 |
ode | +0.042 |
feast | +0.042 |
fly | +0.041 |
Neg logit contributions | |
he | -0.023 |
sick | -0.015 |
elderly | -0.013 |
both | -0.013 |
angry | -0.013 |
Token | a |
Token ID | 257 |
Feature activation | +9.49e-03 |
Pos logit contributions | |
olded | +0.012 |
punished | +0.011 |
he | +0.011 |
outs | +0.011 |
gr | +0.011 |
Neg logit contributions | |
moment | -0.018 |
tweet | -0.017 |
Nice | -0.016 |
Want | -0.016 |
feast | -0.016 |
Token | cute |
Token ID | 13779 |
Feature activation | -2.12e-02 |
Pos logit contributions | |
olded | +0.013 |
he | +0.013 |
punished | +0.012 |
outs | +0.012 |
sick | +0.012 |
Neg logit contributions | |
moment | -0.015 |
tweet | -0.015 |
Nice | -0.014 |
Want | -0.013 |
feast | -0.013 |
Token | bird |
Token ID | 6512 |
Feature activation | -2.25e-02 |
Pos logit contributions | |
moment | +0.036 |
tweet | +0.035 |
Nice | +0.034 |
Want | +0.033 |
reetings | +0.032 |
Neg logit contributions | |
elderly | -0.026 |
he | -0.025 |
sick | -0.025 |
pastor | -0.023 |
punished | -0.023 |
Token | called |
Token ID | 1444 |
Feature activation | -1.28e-02 |
Pos logit contributions | |
moment | +0.039 |
tweet | +0.038 |
Nice | +0.036 |
Want | +0.035 |
feast | +0.034 |
Neg logit contributions | |
he | -0.027 |
olded | -0.026 |
sick | -0.026 |
punished | -0.026 |
elderly | -0.026 |
Token | Summer |
Token ID | 10216 |
Feature activation | -7.29e-02 |
Pos logit contributions | |
moment | +0.020 |
tweet | +0.019 |
Nice | +0.018 |
Want | +0.017 |
est | +0.017 |
Neg logit contributions | |
he | -0.018 |
elderly | -0.018 |
sick | -0.018 |
olded | -0.017 |
punished | -0.017 |
Token | . |
Token ID | 13 |
Feature activation | -1.26e-02 |
Pos logit contributions | |
moment | +0.153 |
tweet | +0.150 |
Nice | +0.144 |
Want | +0.140 |
feast | +0.137 |
Neg logit contributions | |
he | -0.060 |
olded | -0.058 |
sick | -0.057 |
punished | -0.056 |
elderly | -0.056 |
Token | Summer |
Token ID | 10216 |
Feature activation | -5.75e-02 |
Pos logit contributions | |
feast | +0.017 |
moment | +0.016 |
tweet | +0.016 |
fly | +0.016 |
seed | +0.015 |
Neg logit contributions | |
he | -0.018 |
elderly | -0.017 |
bored | -0.017 |
late | -0.016 |
outs | -0.016 |
Token | was |
Token ID | 373 |
Feature activation | +4.11e-04 |
Pos logit contributions | |
tweet | +0.078 |
Nice | +0.078 |
moment | +0.078 |
Want | +0.075 |
greeting | +0.070 |
Neg logit contributions | |
he | -0.086 |
sick | -0.085 |
elderly | -0.083 |
punished | -0.081 |
olded | -0.080 |
Token | a |
Token ID | 257 |
Feature activation | +4.42e-03 |
Pos logit contributions | |
sick | +0.000 |
elderly | +0.000 |
bored | +0.000 |
he | +0.000 |
angry | +0.000 |
Neg logit contributions | |
moment | -0.001 |
tweet | -0.001 |
Nice | -0.001 |
Want | -0.001 |
feast | -0.001 |
Token | happy |
Token ID | 3772 |
Feature activation | +4.02e-03 |
Pos logit contributions | |
sick | +0.005 |
elderly | +0.005 |
he | +0.005 |
punished | +0.004 |
bored | +0.004 |
Neg logit contributions | |
moment | -0.008 |
tweet | -0.008 |
Nice | -0.008 |
Want | -0.007 |
est | -0.007 |
Token | mom |
Token ID | 1995 |
Feature activation | +4.77e-03 |
Pos logit contributions | |
olded | +0.023 |
he | +0.022 |
outs | +0.022 |
punished | +0.021 |
gr | +0.021 |
Neg logit contributions | |
moment | -0.019 |
tweet | -0.018 |
Nice | -0.016 |
greeting | -0.016 |
Want | -0.015 |
Token | said |
Token ID | 531 |
Feature activation | -3.93e-02 |
Pos logit contributions | |
obb | +0.008 |
he | +0.007 |
gr | +0.007 |
leaf | +0.007 |
ateful | +0.007 |
Neg logit contributions | |
moment | -0.007 |
tweet | -0.006 |
Nice | -0.006 |
fly | -0.006 |
Want | -0.006 |
Token | , |
Token ID | 11 |
Feature activation | -2.06e-02 |
Pos logit contributions | |
moment | +0.042 |
tweet | +0.040 |
Nice | +0.036 |
Want | +0.034 |
greeting | +0.032 |
Neg logit contributions | |
he | -0.075 |
punished | -0.073 |
olded | -0.073 |
elderly | -0.073 |
sick | -0.072 |
Token | " |
Token ID | 366 |
Feature activation | +5.99e-03 |
Pos logit contributions | |
moment | +0.021 |
tweet | +0.021 |
Nice | +0.019 |
Want | +0.017 |
fly | +0.017 |
Neg logit contributions | |
he | -0.039 |
olded | -0.039 |
punished | -0.038 |
sick | -0.038 |
elderly | -0.038 |
Token | C |
Token ID | 34 |
Feature activation | -1.13e-02 |
Pos logit contributions | |
sick | +0.007 |
olded | +0.007 |
obb | +0.007 |
he | +0.007 |
punished | +0.007 |
Neg logit contributions | |
moment | -0.011 |
tweet | -0.010 |
Nice | -0.009 |
Want | -0.009 |
greeting | -0.009 |
Token | athy |
Token ID | 10036 |
Feature activation | -7.23e-02 |
Pos logit contributions | |
moment | +0.018 |
tweet | +0.017 |
Nice | +0.016 |
Want | +0.015 |
est | +0.015 |
Neg logit contributions | |
he | -0.015 |
olded | -0.015 |
sick | -0.015 |
punished | -0.015 |
pastor | -0.015 |
Token | , |
Token ID | 11 |
Feature activation | -1.78e-02 |
Pos logit contributions | |
moment | +0.069 |
tweet | +0.067 |
Nice | +0.059 |
Want | +0.055 |
fly | +0.053 |
Neg logit contributions | |
he | -0.143 |
olded | -0.142 |
sick | -0.140 |
punished | -0.139 |
elderly | -0.138 |
Token | let |
Token ID | 1309 |
Feature activation | -1.71e-02 |
Pos logit contributions | |
moment | +0.027 |
tweet | +0.027 |
Nice | +0.025 |
fly | +0.025 |
Want | +0.024 |
Neg logit contributions | |
he | -0.024 |
olded | -0.024 |
elderly | -0.024 |
sick | -0.024 |
punished | -0.024 |
Token | 's |
Token ID | 338 |
Feature activation | -1.57e-03 |
Pos logit contributions | |
moment | +0.019 |
tweet | +0.018 |
Nice | +0.016 |
Want | +0.016 |
est | +0.016 |
Neg logit contributions | |
he | -0.031 |
olded | -0.030 |
obb | -0.030 |
outs | -0.030 |
sick | -0.030 |
Token | go |
Token ID | 467 |
Feature activation | -1.25e-02 |
Pos logit contributions | |
moment | +0.002 |
tweet | +0.002 |
feast | +0.002 |
Nice | +0.002 |
greeting | +0.002 |
Neg logit contributions | |
olded | -0.003 |
outs | -0.003 |
obb | -0.003 |
he | -0.003 |
outed | -0.003 |
Token | for |
Token ID | 329 |
Feature activation | +1.25e-02 |
Pos logit contributions | |
moment | +0.018 |
tweet | +0.018 |
Nice | +0.016 |
Want | +0.016 |
feast | +0.015 |
Neg logit contributions | |
olded | -0.018 |
he | -0.018 |
punished | -0.018 |
sick | -0.018 |
obb | -0.018 |
Token | girl |
Token ID | 2576 |
Feature activation | -2.34e-05 |
Pos logit contributions | |
olded | +0.006 |
he | +0.006 |
punished | +0.006 |
sick | +0.006 |
elderly | +0.006 |
Neg logit contributions | |
moment | -0.007 |
tweet | -0.007 |
Nice | -0.006 |
Want | -0.006 |
fly | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | -8.97e-03 |
Pos logit contributions | |
moment | +0.000 |
tweet | +0.000 |
Nice | +0.000 |
Want | +0.000 |
feast | +0.000 |
Neg logit contributions | |
he | -0.000 |
sick | -0.000 |
olded | -0.000 |
punished | -0.000 |
elderly | -0.000 |
Token | Her |
Token ID | 2332 |
Feature activation | +5.53e-03 |
Pos logit contributions | |
tweet | +0.022 |
moment | +0.021 |
fly | +0.021 |
elcome | +0.020 |
feast | +0.020 |
Neg logit contributions | |
he | -0.005 |
elderly | -0.004 |
sick | -0.004 |
bored | -0.003 |
olded | -0.003 |
Token | name |
Token ID | 1438 |
Feature activation | -3.49e-02 |
Pos logit contributions | |
he | +0.009 |
olded | +0.009 |
punished | +0.009 |
sick | +0.009 |
leaf | +0.009 |
Neg logit contributions | |
moment | -0.007 |
tweet | -0.007 |
Nice | -0.006 |
Want | -0.006 |
feast | -0.006 |
Token | was |
Token ID | 373 |
Feature activation | -7.35e-03 |
Pos logit contributions | |
tweet | +0.048 |
moment | +0.047 |
Nice | +0.044 |
Want | +0.042 |
greeting | +0.041 |
Neg logit contributions | |
he | -0.057 |
sick | -0.054 |
elderly | -0.053 |
punished | -0.053 |
olded | -0.052 |
Token | Suzanne |
Token ID | 45587 |
Feature activation | -7.23e-02 |
Pos logit contributions | |
moment | +0.012 |
tweet | +0.012 |
Nice | +0.011 |
Want | +0.011 |
fly | +0.011 |
Neg logit contributions | |
he | -0.009 |
sick | -0.009 |
olded | -0.009 |
elderly | -0.009 |
punished | -0.009 |
Token | and |
Token ID | 290 |
Feature activation | -4.08e-03 |
Pos logit contributions | |
moment | +0.132 |
tweet | +0.126 |
Nice | +0.116 |
est | +0.116 |
dy | +0.116 |
Neg logit contributions | |
olded | -0.084 |
obb | -0.083 |
leaf | -0.081 |
he | -0.079 |
punished | -0.078 |
Token | she |
Token ID | 673 |
Feature activation | +8.22e-03 |
Pos logit contributions | |
moment | +0.008 |
tweet | +0.008 |
Nice | +0.008 |
Want | +0.008 |
fly | +0.008 |
Neg logit contributions | |
he | -0.004 |
olded | -0.003 |
punished | -0.003 |
sick | -0.003 |
elderly | -0.003 |
Token | was |
Token ID | 373 |
Feature activation | -1.74e-03 |
Pos logit contributions | |
he | +0.010 |
olded | +0.010 |
punished | +0.010 |
elderly | +0.010 |
outs | +0.010 |
Neg logit contributions | |
moment | -0.014 |
tweet | -0.013 |
Nice | -0.013 |
Want | -0.012 |
fly | -0.012 |
Token | three |
Token ID | 1115 |
Feature activation | +7.95e-04 |
Pos logit contributions | |
tweet | +0.003 |
moment | +0.003 |
Nice | +0.002 |
Want | +0.002 |
greeting | +0.002 |
Neg logit contributions | |
he | -0.003 |
olded | -0.003 |
outs | -0.002 |
obb | -0.002 |
leaf | -0.002 |
Token | years |
Token ID | 812 |
Feature activation | -2.63e-03 |
Pos logit contributions | |
he | +0.001 |
olded | +0.001 |
leaf | +0.001 |
obb | +0.001 |
punished | +0.001 |
Neg logit contributions | |
moment | -0.001 |
tweet | -0.001 |
est | -0.001 |
Nice | -0.001 |
feast | -0.001 |
Token | and |
Token ID | 290 |
Feature activation | -1.08e-03 |
Pos logit contributions | |
moment | +0.027 |
tweet | +0.027 |
Nice | +0.025 |
Want | +0.024 |
fly | +0.023 |
Neg logit contributions | |
olded | -0.026 |
he | -0.026 |
punished | -0.025 |
sick | -0.025 |
elderly | -0.025 |
Token | smiles |
Token ID | 21845 |
Feature activation | -1.09e-02 |
Pos logit contributions | |
tweet | +0.001 |
moment | +0.001 |
Want | +0.001 |
feast | +0.001 |
Nice | +0.001 |
Neg logit contributions | |
he | -0.002 |
olded | -0.002 |
gr | -0.002 |
obb | -0.002 |
elderly | -0.002 |
Token | . |
Token ID | 13 |
Feature activation | -1.29e-02 |
Pos logit contributions | |
moment | +0.019 |
tweet | +0.018 |
Nice | +0.017 |
greeting | +0.016 |
fly | +0.016 |
Neg logit contributions | |
olded | -0.014 |
obb | -0.014 |
punished | -0.014 |
earn | -0.014 |
leaf | -0.014 |
Token | " |
Token ID | 366 |
Feature activation | +4.92e-03 |
Pos logit contributions | |
Nice | +0.022 |
moment | +0.021 |
tweet | +0.021 |
Want | +0.020 |
greeting | +0.019 |
Neg logit contributions | |
punished | -0.018 |
he | -0.017 |
gr | -0.016 |
olded | -0.016 |
sick | -0.016 |
Token | Gu |
Token ID | 8205 |
Feature activation | -3.03e-02 |
Pos logit contributions | |
olded | +0.006 |
he | +0.006 |
sick | +0.006 |
punished | +0.006 |
elderly | +0.006 |
Neg logit contributions | |
moment | -0.009 |
tweet | -0.008 |
Nice | -0.008 |
Want | -0.008 |
fly | -0.007 |
Token | ess |
Token ID | 408 |
Feature activation | -7.20e-02 |
Pos logit contributions | |
moment | +0.041 |
tweet | +0.039 |
Nice | +0.037 |
est | +0.035 |
fly | +0.035 |
Neg logit contributions | |
he | -0.048 |
punished | -0.048 |
olded | -0.047 |
sick | -0.047 |
leaf | -0.046 |
Token | what |
Token ID | 644 |
Feature activation | -6.18e-03 |
Pos logit contributions | |
moment | +0.094 |
tweet | +0.092 |
Nice | +0.085 |
Want | +0.080 |
fly | +0.078 |
Neg logit contributions | |
he | -0.117 |
olded | -0.115 |
sick | -0.114 |
punished | -0.114 |
elderly | -0.112 |
Token | , |
Token ID | 11 |
Feature activation | +9.76e-04 |
Pos logit contributions | |
moment | +0.010 |
tweet | +0.010 |
Nice | +0.009 |
Want | +0.009 |
fly | +0.009 |
Neg logit contributions | |
he | -0.008 |
olded | -0.008 |
sick | -0.008 |
punished | -0.008 |
elderly | -0.008 |
Token | kids |
Token ID | 3988 |
Feature activation | +3.59e-03 |
Pos logit contributions | |
elderly | +0.001 |
he | +0.001 |
sick | +0.001 |
olded | +0.001 |
punished | +0.001 |
Neg logit contributions | |
moment | -0.002 |
tweet | -0.002 |
Nice | -0.001 |
Want | -0.001 |
fly | -0.001 |
Token | ? |
Token ID | 30 |
Feature activation | -3.36e-03 |
Pos logit contributions | |
olded | +0.006 |
gr | +0.006 |
obb | +0.006 |
elderly | +0.006 |
outs | +0.006 |
Neg logit contributions | |
tweet | -0.005 |
moment | -0.004 |
feast | -0.004 |
Nice | -0.004 |
Want | -0.004 |
Token | I |
Token ID | 314 |
Feature activation | +8.96e-03 |
Pos logit contributions | |
moment | +0.004 |
tweet | +0.004 |
Nice | +0.004 |
Want | +0.004 |
fly | +0.004 |
Neg logit contributions | |
he | -0.006 |
punished | -0.006 |
sick | -0.005 |
olded | -0.005 |
elderly | -0.005 |
Token | Grand |
Token ID | 5675 |
Feature activation | +1.87e-02 |
Pos logit contributions | |
punished | +0.003 |
sick | +0.003 |
he | +0.003 |
obb | +0.003 |
gr | +0.003 |
Neg logit contributions | |
moment | -0.002 |
Nice | -0.002 |
Want | -0.002 |
tweet | -0.002 |
greeting | -0.002 |
Token | dad |
Token ID | 47984 |
Feature activation | -5.83e-04 |
Pos logit contributions | |
he | +0.023 |
elderly | +0.022 |
olded | +0.022 |
punished | +0.021 |
leaf | +0.021 |
Neg logit contributions | |
tweet | -0.033 |
moment | -0.032 |
Nice | -0.030 |
Want | -0.029 |
fly | -0.029 |
Token | also |
Token ID | 635 |
Feature activation | -1.17e-02 |
Pos logit contributions | |
moment | +0.001 |
tweet | +0.001 |
Nice | +0.001 |
Want | +0.001 |
fly | +0.001 |
Neg logit contributions | |
he | -0.001 |
sick | -0.001 |
olded | -0.001 |
punished | -0.001 |
elderly | -0.001 |
Token | showed |
Token ID | 3751 |
Feature activation | +1.57e-02 |
Pos logit contributions | |
moment | +0.020 |
tweet | +0.019 |
Nice | +0.018 |
Want | +0.018 |
fly | +0.017 |
Neg logit contributions | |
he | -0.014 |
sick | -0.014 |
olded | -0.014 |
punished | -0.014 |
obb | -0.014 |
Token | Mate |
Token ID | 24787 |
Feature activation | -3.09e-02 |
Pos logit contributions | |
sick | +0.022 |
he | +0.022 |
punished | +0.022 |
obb | +0.022 |
olded | +0.021 |
Neg logit contributions | |
moment | -0.024 |
tweet | -0.023 |
Nice | -0.022 |
est | -0.022 |
Want | -0.020 |
Token | o |
Token ID | 78 |
Feature activation | -7.19e-02 |
Pos logit contributions | |
moment | +0.039 |
tweet | +0.039 |
Nice | +0.035 |
Want | +0.033 |
feast | +0.033 |
Neg logit contributions | |
olded | -0.051 |
he | -0.050 |
sick | -0.049 |
punished | -0.049 |
elderly | -0.049 |
Token | how |
Token ID | 703 |
Feature activation | +1.56e-02 |
Pos logit contributions | |
moment | +0.109 |
tweet | +0.106 |
Nice | +0.101 |
Want | +0.097 |
feast | +0.095 |
Neg logit contributions | |
olded | -0.099 |
sick | -0.097 |
he | -0.096 |
elderly | -0.095 |
outs | -0.095 |
Token | to |
Token ID | 284 |
Feature activation | -7.56e-03 |
Pos logit contributions | |
he | +0.033 |
salon | +0.029 |
pastor | +0.029 |
obb | +0.029 |
iza | +0.028 |
Neg logit contributions | |
moment | -0.017 |
tweet | -0.014 |
Want | -0.013 |
Nice | -0.013 |
dy | -0.013 |
Token | make |
Token ID | 787 |
Feature activation | +1.31e-02 |
Pos logit contributions | |
moment | +0.011 |
tweet | +0.010 |
Want | +0.009 |
Nice | +0.009 |
greeting | +0.009 |
Neg logit contributions | |
olded | -0.012 |
pastor | -0.012 |
outs | -0.011 |
sick | -0.011 |
he | -0.011 |
Token | careful |
Token ID | 8161 |
Feature activation | +7.05e-03 |
Pos logit contributions | |
olded | +0.020 |
he | +0.020 |
pastor | +0.019 |
leaf | +0.019 |
gr | +0.018 |
Neg logit contributions | |
moment | -0.019 |
tweet | -0.018 |
est | -0.016 |
Nice | -0.016 |
elcome | -0.016 |
Token | and |
Token ID | 290 |
Feature activation | +1.20e-02 |
Pos logit contributions | |
he | +0.010 |
olded | +0.009 |
sick | +0.009 |
punished | +0.009 |
elderly | +0.009 |
Neg logit contributions | |
moment | -0.011 |
tweet | -0.011 |
Nice | -0.010 |
Want | -0.010 |
feast | -0.010 |
Token | � |
Token ID | 95 |
Feature activation | -7.28e-03 |
Pos logit contributions | |
sick | +0.002 |
wasted | +0.002 |
elderly | +0.002 |
punished | +0.002 |
outed | +0.002 |
Neg logit contributions | |
tweet | -0.002 |
Nice | -0.002 |
Want | -0.002 |
moment | -0.001 |
feast | -0.001 |
Token | € |
Token ID | 26391 |
Feature activation | -3.20e-02 |
Pos logit contributions | |
moment | +0.010 |
tweet | +0.009 |
Nice | +0.009 |
Want | +0.009 |
est | +0.009 |
Neg logit contributions | |
sick | -0.011 |
he | -0.011 |
olded | -0.011 |
elderly | -0.011 |
punished | -0.011 |
Token | � |
Token ID | 129 |
Feature activation | +1.39e-03 |
Pos logit contributions | |
moment | +0.045 |
tweet | +0.044 |
Nice | +0.041 |
Want | +0.038 |
fly | +0.038 |
Neg logit contributions | |
he | -0.049 |
olded | -0.049 |
punished | -0.047 |
sick | -0.047 |
elderly | -0.046 |
Token | � |
Token ID | 241 |
Feature activation | -2.68e-04 |
Pos logit contributions | |
he | +0.002 |
olded | +0.002 |
punished | +0.002 |
sick | +0.002 |
salon | +0.002 |
Neg logit contributions | |
tweet | -0.001 |
moment | -0.001 |
Want | -0.001 |
Nice | -0.001 |
greeting | -0.001 |
Token | Gu |
Token ID | 8205 |
Feature activation | -2.54e-02 |
Pos logit contributions | |
moment | +0.000 |
Nice | +0.000 |
Want | +0.000 |
tweet | +0.000 |
greeting | +0.000 |
Neg logit contributions | |
punished | -0.000 |
obb | -0.000 |
olded | -0.000 |
sick | -0.000 |
ateful | -0.000 |
Token | ess |
Token ID | 408 |
Feature activation | -7.16e-02 |
Pos logit contributions | |
est | +0.029 |
moment | +0.028 |
Nice | +0.026 |
tweet | +0.026 |
dy | +0.024 |
Neg logit contributions | |
sick | -0.046 |
punished | -0.046 |
bored | -0.045 |
he | -0.045 |
leaf | -0.045 |
Token | what |
Token ID | 644 |
Feature activation | +1.78e-03 |
Pos logit contributions | |
tweet | +0.086 |
moment | +0.086 |
Nice | +0.075 |
fly | +0.074 |
Want | +0.072 |
Neg logit contributions | |
he | -0.126 |
olded | -0.122 |
obb | -0.121 |
sick | -0.120 |
punished | -0.119 |
Token | I |
Token ID | 314 |
Feature activation | +1.71e-02 |
Pos logit contributions | |
he | +0.003 |
olded | +0.003 |
punished | +0.003 |
obb | +0.003 |
gr | +0.003 |
Neg logit contributions | |
moment | -0.003 |
tweet | -0.003 |
Nice | -0.002 |
Want | -0.002 |
feast | -0.002 |
Token | am |
Token ID | 716 |
Feature activation | +7.25e-03 |
Pos logit contributions | |
olded | +0.026 |
he | +0.026 |
elderly | +0.026 |
pastor | +0.026 |
sick | +0.025 |
Neg logit contributions | |
moment | -0.024 |
tweet | -0.023 |
Nice | -0.021 |
Want | -0.021 |
fly | -0.020 |
Token | ? |
Token ID | 30 |
Feature activation | +1.32e-02 |
Pos logit contributions | |
he | +0.013 |
olded | +0.013 |
gr | +0.013 |
punished | +0.013 |
outs | +0.012 |
Neg logit contributions | |
moment | -0.008 |
tweet | -0.008 |
Nice | -0.007 |
Want | -0.007 |
feast | -0.007 |
Token | â |
Token ID | 22940 |
Feature activation | -9.69e-03 |
Pos logit contributions | |
obb | +0.022 |
sick | +0.022 |
punished | +0.022 |
he | +0.021 |
pastor | +0.021 |
Neg logit contributions | |
Nice | -0.017 |
moment | -0.017 |
tweet | -0.017 |
Want | -0.016 |
greeting | -0.014 |
Token | a |
Token ID | 257 |
Feature activation | -4.61e-03 |
Pos logit contributions | |
he | +0.012 |
obb | +0.011 |
pastor | +0.011 |
salon | +0.010 |
udging | +0.010 |
Neg logit contributions | |
moment | -0.008 |
tweet | -0.007 |
hopping | -0.006 |
Want | -0.006 |
greeting | -0.006 |
Token | stamp |
Token ID | 17977 |
Feature activation | -5.97e-03 |
Pos logit contributions | |
moment | +0.007 |
tweet | +0.006 |
Nice | +0.006 |
feast | +0.006 |
Want | +0.006 |
Neg logit contributions | |
olded | -0.007 |
he | -0.007 |
punished | -0.007 |
gr | -0.007 |
outs | -0.007 |
Token | ! |
Token ID | 0 |
Feature activation | +4.97e-03 |
Pos logit contributions | |
moment | +0.008 |
tweet | +0.008 |
feast | +0.008 |
fly | +0.008 |
est | +0.007 |
Neg logit contributions | |
outed | -0.009 |
obb | -0.009 |
olded | -0.009 |
blamed | -0.009 |
ateful | -0.009 |
Token | She |
Token ID | 1375 |
Feature activation | -2.73e-03 |
Pos logit contributions | |
obb | +0.009 |
punished | +0.009 |
sick | +0.009 |
pastor | +0.009 |
ateful | +0.009 |
Neg logit contributions | |
Nice | -0.006 |
moment | -0.006 |
Want | -0.005 |
tweet | -0.005 |
Would | -0.005 |
Token | showed |
Token ID | 3751 |
Feature activation | +1.61e-02 |
Pos logit contributions | |
moment | +0.004 |
tweet | +0.003 |
fly | +0.003 |
feast | +0.003 |
Want | +0.003 |
Neg logit contributions | |
obb | -0.004 |
he | -0.004 |
pastor | -0.004 |
sick | -0.004 |
ateful | -0.004 |
Token | Stamp |
Token ID | 40694 |
Feature activation | -7.13e-02 |
Pos logit contributions | |
punished | +0.028 |
obb | +0.027 |
sick | +0.027 |
olded | +0.027 |
he | +0.027 |
Neg logit contributions | |
tweet | -0.020 |
moment | -0.019 |
Nice | -0.018 |
est | -0.017 |
Want | -0.016 |
Token | how |
Token ID | 703 |
Feature activation | +1.24e-02 |
Pos logit contributions | |
moment | +0.113 |
tweet | +0.112 |
Nice | +0.109 |
Want | +0.105 |
feast | +0.101 |
Neg logit contributions | |
he | -0.089 |
olded | -0.088 |
elderly | -0.088 |
sick | -0.087 |
outs | -0.087 |
Token | to |
Token ID | 284 |
Feature activation | -5.03e-03 |
Pos logit contributions | |
he | +0.024 |
obb | +0.023 |
paramedics | +0.022 |
idget | +0.022 |
salon | +0.022 |
Neg logit contributions | |
moment | -0.015 |
tweet | -0.013 |
Want | -0.011 |
Nice | -0.011 |
greeting | -0.010 |
Token | use |
Token ID | 779 |
Feature activation | +1.80e-03 |
Pos logit contributions | |
tweet | +0.007 |
moment | +0.007 |
feast | +0.006 |
fly | +0.006 |
hop | +0.005 |
Neg logit contributions | |
olded | -0.009 |
outed | -0.008 |
bored | -0.008 |
outs | -0.008 |
pastor | -0.008 |
Token | the |
Token ID | 262 |
Feature activation | -5.93e-03 |
Pos logit contributions | |
he | +0.003 |
olded | +0.003 |
punished | +0.003 |
sick | +0.003 |
elderly | +0.003 |
Neg logit contributions | |
moment | -0.003 |
tweet | -0.002 |
Nice | -0.002 |
Want | -0.002 |
fly | -0.002 |
Token | stamp |
Token ID | 17977 |
Feature activation | -1.02e-02 |
Pos logit contributions | |
moment | +0.010 |
tweet | +0.010 |
Nice | +0.008 |
feast | +0.008 |
est | +0.008 |
Neg logit contributions | |
olded | -0.008 |
he | -0.008 |
punished | -0.007 |
sick | -0.007 |
leaf | -0.007 |
Token | 's |
Token ID | 338 |
Feature activation | +1.26e-02 |
Pos logit contributions | |
moment | +0.019 |
tweet | +0.019 |
Nice | +0.017 |
fly | +0.016 |
est | +0.016 |
Neg logit contributions | |
he | -0.020 |
olded | -0.020 |
elderly | -0.019 |
obb | -0.019 |
sick | -0.019 |
Token | friends |
Token ID | 2460 |
Feature activation | +1.55e-02 |
Pos logit contributions | |
he | +0.019 |
olded | +0.019 |
obb | +0.018 |
leaf | +0.018 |
punished | +0.018 |
Neg logit contributions | |
moment | -0.019 |
tweet | -0.018 |
Nice | -0.016 |
greeting | -0.016 |
feast | -0.016 |
Token | . |
Token ID | 13 |
Feature activation | +1.12e-03 |
Pos logit contributions | |
obb | +0.027 |
elderly | +0.026 |
gr | +0.026 |
leaf | +0.025 |
he | +0.025 |
Neg logit contributions | |
moment | -0.020 |
tweet | -0.019 |
Nice | -0.017 |
Want | -0.017 |
fly | -0.016 |
Token | She |
Token ID | 1375 |
Feature activation | -3.28e-03 |
Pos logit contributions | |
he | +0.001 |
olded | +0.001 |
punished | +0.001 |
sick | +0.001 |
elderly | +0.001 |
Neg logit contributions | |
moment | -0.002 |
tweet | -0.002 |
Nice | -0.002 |
Want | -0.002 |
fly | -0.002 |
Token | gave |
Token ID | 2921 |
Feature activation | +7.69e-03 |
Pos logit contributions | |
tweet | +0.005 |
moment | +0.004 |
fly | +0.004 |
Want | +0.004 |
dy | +0.004 |
Neg logit contributions | |
he | -0.005 |
obb | -0.005 |
gr | -0.005 |
leaf | -0.005 |
elderly | -0.005 |
Token | Daisy |
Token ID | 40355 |
Feature activation | -7.08e-02 |
Pos logit contributions | |
he | +0.013 |
olded | +0.013 |
punished | +0.013 |
sick | +0.013 |
elderly | +0.013 |
Neg logit contributions | |
moment | -0.009 |
tweet | -0.009 |
Nice | -0.008 |
Want | -0.008 |
fly | -0.007 |
Token | 's |
Token ID | 338 |
Feature activation | -6.13e-03 |
Pos logit contributions | |
moment | +0.127 |
tweet | +0.123 |
Nice | +0.116 |
Want | +0.111 |
feast | +0.111 |
Neg logit contributions | |
he | -0.081 |
olded | -0.080 |
obb | -0.078 |
punished | -0.077 |
pastor | -0.075 |
Token | friends |
Token ID | 2460 |
Feature activation | -7.26e-03 |
Pos logit contributions | |
moment | +0.009 |
tweet | +0.009 |
Nice | +0.008 |
feast | +0.008 |
greeting | +0.007 |
Neg logit contributions | |
he | -0.010 |
olded | -0.010 |
obb | -0.009 |
punished | -0.009 |
gr | -0.009 |
Token | some |
Token ID | 617 |
Feature activation | -1.30e-02 |
Pos logit contributions | |
moment | +0.011 |
tweet | +0.010 |
feast | +0.010 |
est | +0.010 |
greeting | +0.010 |
Neg logit contributions | |
obb | -0.010 |
pastor | -0.010 |
olded | -0.010 |
ateful | -0.010 |
he | -0.010 |
Token | medicine |
Token ID | 9007 |
Feature activation | +5.77e-03 |
Pos logit contributions | |
moment | +0.017 |
tweet | +0.016 |
feast | +0.015 |
est | +0.015 |
Nice | +0.014 |
Neg logit contributions | |
he | -0.022 |
salon | -0.021 |
olded | -0.021 |
she | -0.021 |
gr | -0.021 |
Token | and |
Token ID | 290 |
Feature activation | -6.21e-03 |
Pos logit contributions | |
he | +0.008 |
gr | +0.008 |
punished | +0.008 |
pastor | +0.008 |
ateful | +0.008 |
Neg logit contributions | |
moment | -0.009 |
tweet | -0.008 |
Nice | -0.008 |
fly | -0.008 |
feast | -0.008 |
Token | big |
Token ID | 1263 |
Feature activation | -3.56e-02 |
Pos logit contributions | |
moment | +0.012 |
tweet | +0.011 |
Nice | +0.010 |
feast | +0.010 |
Want | +0.010 |
Neg logit contributions | |
olded | -0.014 |
he | -0.014 |
punished | -0.013 |
gr | -0.013 |
obb | -0.013 |
Token | football |
Token ID | 4346 |
Feature activation | -2.83e-03 |
Pos logit contributions | |
tweet | +0.049 |
moment | +0.048 |
Nice | +0.047 |
Want | +0.044 |
fly | +0.043 |
Neg logit contributions | |
elderly | -0.053 |
sick | -0.051 |
he | -0.049 |
punished | -0.048 |
bored | -0.048 |
Token | . |
Token ID | 13 |
Feature activation | -6.74e-03 |
Pos logit contributions | |
moment | +0.005 |
tweet | +0.005 |
Nice | +0.005 |
fly | +0.004 |
Want | +0.004 |
Neg logit contributions | |
he | -0.003 |
olded | -0.003 |
punished | -0.003 |
obb | -0.003 |
gr | -0.003 |
Token | He |
Token ID | 679 |
Feature activation | -6.79e-03 |
Pos logit contributions | |
moment | +0.009 |
Nice | +0.009 |
Want | +0.008 |
tweet | +0.008 |
greeting | +0.008 |
Neg logit contributions | |
sick | -0.011 |
punished | -0.011 |
obb | -0.011 |
he | -0.011 |
olded | -0.010 |
Token | showed |
Token ID | 3751 |
Feature activation | +8.25e-03 |
Pos logit contributions | |
moment | +0.009 |
tweet | +0.008 |
Nice | +0.008 |
Want | +0.007 |
dy | +0.007 |
Neg logit contributions | |
he | -0.011 |
obb | -0.011 |
sick | -0.011 |
pastor | -0.011 |
gr | -0.011 |
Token | John |
Token ID | 1757 |
Feature activation | -7.07e-02 |
Pos logit contributions | |
sick | +0.013 |
he | +0.013 |
olded | +0.013 |
punished | +0.013 |
obb | +0.013 |
Neg logit contributions | |
moment | -0.011 |
tweet | -0.011 |
Nice | -0.010 |
est | -0.010 |
Want | -0.009 |
Token | how |
Token ID | 703 |
Feature activation | +1.85e-02 |
Pos logit contributions | |
moment | +0.100 |
tweet | +0.097 |
Nice | +0.090 |
Want | +0.086 |
fly | +0.085 |
Neg logit contributions | |
he | -0.107 |
olded | -0.105 |
punished | -0.104 |
sick | -0.103 |
obb | -0.103 |
Token | to |
Token ID | 284 |
Feature activation | -6.92e-03 |
Pos logit contributions | |
he | +0.036 |
salon | +0.036 |
obb | +0.034 |
pastor | +0.033 |
paramedics | +0.033 |
Neg logit contributions | |
moment | -0.022 |
Nice | -0.018 |
Want | -0.017 |
tweet | -0.017 |
greeting | -0.016 |
Token | kick |
Token ID | 4829 |
Feature activation | -3.98e-03 |
Pos logit contributions | |
moment | +0.009 |
tweet | +0.009 |
Want | +0.008 |
greeting | +0.008 |
feast | +0.008 |
Neg logit contributions | |
olded | -0.012 |
outs | -0.011 |
pastor | -0.011 |
sick | -0.011 |
gr | -0.011 |
Token | it |
Token ID | 340 |
Feature activation | -4.74e-03 |
Pos logit contributions | |
moment | +0.005 |
tweet | +0.005 |
Nice | +0.005 |
Want | +0.005 |
fly | +0.005 |
Neg logit contributions | |
he | -0.006 |
olded | -0.006 |
sick | -0.006 |
punished | -0.006 |
elderly | -0.006 |
Token | , |
Token ID | 11 |
Feature activation | +3.32e-03 |
Pos logit contributions | |
moment | +0.007 |
tweet | +0.007 |
Nice | +0.006 |
Want | +0.006 |
est | +0.006 |
Neg logit contributions | |
olded | -0.007 |
he | -0.007 |
punished | -0.007 |
sick | -0.007 |
elderly | -0.007 |
Token | He |
Token ID | 679 |
Feature activation | +1.47e-03 |
Pos logit contributions | |
obb | +0.021 |
punished | +0.021 |
he | +0.021 |
sick | +0.020 |
leaf | +0.020 |
Neg logit contributions | |
Nice | -0.019 |
moment | -0.019 |
tweet | -0.019 |
Want | -0.018 |
fly | -0.016 |
Token | said |
Token ID | 531 |
Feature activation | -3.35e-02 |
Pos logit contributions | |
he | +0.003 |
elderly | +0.002 |
obb | +0.002 |
ateful | +0.002 |
leaf | +0.002 |
Neg logit contributions | |
tweet | -0.002 |
moment | -0.002 |
hops | -0.002 |
fly | -0.002 |
Want | -0.002 |
Token | , |
Token ID | 11 |
Feature activation | -1.16e-02 |
Pos logit contributions | |
tweet | +0.033 |
moment | +0.032 |
Nice | +0.029 |
Want | +0.026 |
fly | +0.025 |
Neg logit contributions | |
he | -0.069 |
punished | -0.065 |
elderly | -0.065 |
leaf | -0.064 |
gr | -0.064 |
Token | " |
Token ID | 366 |
Feature activation | +8.83e-03 |
Pos logit contributions | |
moment | +0.013 |
tweet | +0.012 |
Nice | +0.011 |
Want | +0.010 |
feast | +0.010 |
Neg logit contributions | |
he | -0.021 |
olded | -0.021 |
sick | -0.020 |
punished | -0.020 |
elderly | -0.020 |
Token | Gu |
Token ID | 8205 |
Feature activation | -3.02e-02 |
Pos logit contributions | |
sick | +0.010 |
punished | +0.010 |
olded | +0.010 |
obb | +0.010 |
elderly | +0.010 |
Neg logit contributions | |
moment | -0.016 |
tweet | -0.016 |
Nice | -0.015 |
Want | -0.015 |
greeting | -0.014 |
Token | ess |
Token ID | 408 |
Feature activation | -7.06e-02 |
Pos logit contributions | |
moment | +0.036 |
tweet | +0.034 |
Nice | +0.033 |
est | +0.032 |
fly | +0.030 |
Neg logit contributions | |
punished | -0.053 |
leaf | -0.052 |
sick | -0.052 |
he | -0.052 |
olded | -0.051 |
Token | what |
Token ID | 644 |
Feature activation | -5.62e-03 |
Pos logit contributions | |
moment | +0.085 |
tweet | +0.084 |
Nice | +0.075 |
fly | +0.071 |
Want | +0.071 |
Neg logit contributions | |
he | -0.122 |
olded | -0.121 |
sick | -0.118 |
punished | -0.118 |
elderly | -0.117 |
Token | ? |
Token ID | 30 |
Feature activation | +4.72e-03 |
Pos logit contributions | |
moment | +0.009 |
tweet | +0.008 |
Nice | +0.008 |
Want | +0.007 |
fly | +0.007 |
Neg logit contributions | |
he | -0.008 |
sick | -0.008 |
olded | -0.008 |
punished | -0.008 |
elderly | -0.008 |
Token | Why |
Token ID | 4162 |
Feature activation | -1.32e-02 |
Pos logit contributions | |
he | +0.007 |
olded | +0.007 |
sick | +0.007 |
punished | +0.007 |
elderly | +0.007 |
Neg logit contributions | |
moment | -0.007 |
tweet | -0.006 |
Nice | -0.006 |
Want | -0.006 |
feast | -0.006 |
Token | did |
Token ID | 750 |
Feature activation | +2.44e-03 |
Pos logit contributions | |
moment | +0.021 |
tweet | +0.020 |
Nice | +0.019 |
Want | +0.018 |
fly | +0.018 |
Neg logit contributions | |
he | -0.018 |
olded | -0.018 |
punished | -0.018 |
sick | -0.018 |
elderly | -0.017 |
Token | the |
Token ID | 262 |
Feature activation | +9.64e-03 |
Pos logit contributions | |
he | +0.003 |
olded | +0.003 |
punished | +0.003 |
obb | +0.003 |
outs | +0.003 |
Neg logit contributions | |
moment | -0.004 |
tweet | -0.004 |
Nice | -0.004 |
Want | -0.004 |
feast | -0.004 |
Token | dry |
Token ID | 5894 |
Feature activation | +6.74e-03 |
Pos logit contributions | |
he | +0.014 |
olded | +0.013 |
outs | +0.013 |
obb | +0.013 |
gr | +0.013 |
Neg logit contributions | |
moment | -0.013 |
tweet | -0.013 |
Want | -0.011 |
feast | -0.011 |
greeting | -0.011 |
Token | out |
Token ID | 503 |
Feature activation | +2.96e-03 |
Pos logit contributions | |
olded | +0.010 |
leaf | +0.010 |
salon | +0.010 |
he | +0.010 |
punished | +0.009 |
Neg logit contributions | |
moment | -0.010 |
est | -0.009 |
tweet | -0.009 |
Nice | -0.009 |
fly | -0.008 |
Token | the |
Token ID | 262 |
Feature activation | +4.58e-03 |
Pos logit contributions | |
olded | +0.004 |
he | +0.004 |
obb | +0.004 |
leaf | +0.004 |
earn | +0.004 |
Neg logit contributions | |
moment | -0.004 |
tweet | -0.004 |
Nice | -0.004 |
Want | -0.004 |
greeting | -0.004 |
Token | Earth |
Token ID | 3668 |
Feature activation | -2.39e-03 |
Pos logit contributions | |
olded | +0.008 |
he | +0.007 |
punished | +0.007 |
leaf | +0.007 |
outs | +0.007 |
Neg logit contributions | |
moment | -0.006 |
tweet | -0.006 |
Nice | -0.005 |
est | -0.005 |
feast | -0.005 |
Token | ." |
Token ID | 526 |
Feature activation | +2.65e-02 |
Pos logit contributions | |
moment | +0.004 |
tweet | +0.004 |
Nice | +0.004 |
fly | +0.004 |
Want | +0.004 |
Neg logit contributions | |
olded | -0.003 |
he | -0.003 |
punished | -0.003 |
leaf | -0.003 |
obb | -0.003 |
Token | |
Token ID | 220 |
Feature activation | +5.20e-02 |
Pos logit contributions | |
he | +0.040 |
obb | +0.040 |
punished | +0.040 |
olded | +0.040 |
pastor | +0.039 |
Neg logit contributions | |
moment | -0.038 |
tweet | -0.037 |
Nice | -0.035 |
Want | -0.034 |
feast | -0.032 |
Token |
|
Token ID | 198 |
Feature activation | -3.74e-03 |
Pos logit contributions | |
punished | +0.084 |
sick | +0.082 |
olded | +0.082 |
leaf | +0.081 |
pastor | +0.081 |
Neg logit contributions | |
moment | -0.067 |
tweet | -0.063 |
Nice | -0.060 |
Want | -0.057 |
elcome | -0.055 |
Token |
|
Token ID | 198 |
Feature activation | -1.16e-02 |
Pos logit contributions | |
moment | +0.006 |
tweet | +0.006 |
Nice | +0.006 |
Want | +0.005 |
fly | +0.005 |
Neg logit contributions | |
he | -0.005 |
olded | -0.005 |
sick | -0.005 |
elderly | -0.005 |
punished | -0.005 |
Token | Little |
Token ID | 22253 |
Feature activation | -4.07e-03 |
Pos logit contributions | |
moment | +0.023 |
tweet | +0.023 |
Nice | +0.021 |
Want | +0.021 |
fly | +0.020 |
Neg logit contributions | |
he | -0.011 |
olded | -0.011 |
sick | -0.011 |
punished | -0.011 |
elderly | -0.011 |
Token | Rad |
Token ID | 5325 |
Feature activation | +2.30e-02 |
Pos logit contributions | |
tweet | +0.007 |
Nice | +0.006 |
moment | +0.006 |
eway | +0.006 |
Want | +0.006 |
Neg logit contributions | |
punished | -0.005 |
elderly | -0.005 |
sick | -0.005 |
olded | -0.005 |
he | -0.005 |
Token | ish |
Token ID | 680 |
Feature activation | +2.97e-03 |
Pos logit contributions | |
he | +0.036 |
olded | +0.036 |
punished | +0.036 |
elderly | +0.035 |
sick | +0.035 |
Neg logit contributions | |
moment | -0.031 |
tweet | -0.030 |
Nice | -0.029 |
Want | -0.028 |
greeting | -0.026 |
Token | will |
Token ID | 481 |
Feature activation | -1.37e-02 |
Pos logit contributions | |
leaf | +0.034 |
pastor | +0.034 |
olded | +0.034 |
gr | +0.033 |
outs | +0.033 |
Neg logit contributions | |
moment | -0.024 |
tweet | -0.023 |
feast | -0.019 |
fly | -0.019 |
est | -0.018 |
Token | come |
Token ID | 1282 |
Feature activation | -3.33e-02 |
Pos logit contributions | |
moment | +0.018 |
tweet | +0.018 |
Nice | +0.015 |
feast | +0.015 |
Want | +0.015 |
Neg logit contributions | |
olded | -0.023 |
he | -0.022 |
leaf | -0.022 |
outs | -0.022 |
pastor | -0.021 |
Token | back |
Token ID | 736 |
Feature activation | -1.06e-03 |
Pos logit contributions | |
moment | +0.051 |
tweet | +0.050 |
Nice | +0.047 |
Want | +0.045 |
fly | +0.044 |
Neg logit contributions | |
he | -0.046 |
olded | -0.046 |
sick | -0.045 |
punished | -0.045 |
elderly | -0.045 |
Token | later |
Token ID | 1568 |
Feature activation | +3.15e-02 |
Pos logit contributions | |
moment | +0.001 |
tweet | +0.001 |
fly | +0.001 |
Nice | +0.001 |
Want | +0.001 |
Neg logit contributions | |
leaf | -0.002 |
olded | -0.002 |
obb | -0.002 |
he | -0.002 |
pastor | -0.002 |
Token | ," |
Token ID | 553 |
Feature activation | +4.74e-02 |
Pos logit contributions | |
leaf | +0.059 |
olded | +0.053 |
gr | +0.052 |
punished | +0.051 |
pastor | +0.050 |
Neg logit contributions | |
moment | -0.043 |
, | -0.033 |
feast | -0.032 |
tweet | -0.031 |
est | -0.031 |
Token | their |
Token ID | 511 |
Feature activation | +5.06e-02 |
Pos logit contributions | |
he | +0.083 |
obb | +0.081 |
leaf | +0.080 |
olded | +0.079 |
out | +0.075 |
Neg logit contributions | |
moment | -0.068 |
tweet | -0.064 |
Nice | -0.053 |
Want | -0.051 |
greeting | -0.050 |
Token | mom |
Token ID | 1995 |
Feature activation | +4.40e-02 |
Pos logit contributions | |
he | +0.080 |
she | +0.071 |
olded | +0.070 |
leaf | +0.070 |
punished | +0.068 |
Neg logit contributions | |
moment | -0.079 |
tweet | -0.076 |
dy | -0.071 |
greeting | -0.068 |
feast | -0.068 |
Token | said |
Token ID | 531 |
Feature activation | -1.11e-02 |
Pos logit contributions | |
obb | +0.050 |
sick | +0.047 |
ateful | +0.047 |
leaf | +0.047 |
he | +0.046 |
Neg logit contributions | |
moment | -0.077 |
tweet | -0.074 |
est | -0.068 |
Nice | -0.068 |
fly | -0.067 |
Token | . |
Token ID | 13 |
Feature activation | -1.31e-02 |
Pos logit contributions | |
moment | +0.019 |
tweet | +0.018 |
Nice | +0.017 |
greeting | +0.016 |
est | +0.016 |
Neg logit contributions | |
leaf | -0.013 |
he | -0.013 |
obb | -0.013 |
elderly | -0.013 |
pastor | -0.013 |
Token |
|
Token ID | 198 |
Feature activation | -1.83e-02 |
Pos logit contributions | |
moment | +0.021 |
tweet | +0.020 |
Nice | +0.019 |
Want | +0.018 |
fly | +0.018 |
Neg logit contributions | |
he | -0.018 |
olded | -0.018 |
sick | -0.017 |
punished | -0.017 |
elderly | -0.017 |
Token |
|
Token ID | 198 |
Feature activation | -2.38e-02 |
Pos logit contributions | |
moment | +0.030 |
tweet | +0.030 |
Nice | +0.028 |
Want | +0.027 |
est | +0.027 |
Neg logit contributions | |
he | -0.023 |
olded | -0.022 |
sick | -0.022 |
elderly | -0.022 |
punished | -0.022 |
Token | he |
Token ID | 339 |
Feature activation | +1.92e-02 |
Pos logit contributions | |
he | +0.069 |
obb | +0.065 |
gr | +0.064 |
olded | +0.063 |
leaf | +0.063 |
Neg logit contributions | |
moment | -0.064 |
tweet | -0.061 |
Nice | -0.058 |
Want | -0.055 |
feast | -0.053 |
Token | had |
Token ID | 550 |
Feature activation | +3.77e-03 |
Pos logit contributions | |
he | +0.033 |
obb | +0.032 |
pastor | +0.032 |
elderly | +0.031 |
olded | +0.031 |
Neg logit contributions | |
moment | -0.024 |
tweet | -0.022 |
Nice | -0.021 |
Want | -0.020 |
fly | -0.020 |
Token | to |
Token ID | 284 |
Feature activation | -4.84e-03 |
Pos logit contributions | |
he | +0.005 |
olded | +0.004 |
sick | +0.004 |
punished | +0.004 |
elderly | +0.004 |
Neg logit contributions | |
moment | -0.007 |
tweet | -0.006 |
Nice | -0.006 |
Want | -0.006 |
fly | -0.006 |
Token | protect |
Token ID | 1805 |
Feature activation | +1.94e-02 |
Pos logit contributions | |
moment | +0.006 |
tweet | +0.006 |
feast | +0.005 |
Nice | +0.005 |
est | +0.005 |
Neg logit contributions | |
olded | -0.008 |
outs | -0.008 |
he | -0.008 |
leaf | -0.008 |
pastor | -0.008 |
Token | his |
Token ID | 465 |
Feature activation | +1.67e-02 |
Pos logit contributions | |
he | +0.030 |
olded | +0.029 |
obb | +0.029 |
pastor | +0.028 |
leaf | +0.028 |
Neg logit contributions | |
moment | -0.028 |
tweet | -0.027 |
Nice | -0.025 |
Want | -0.024 |
fly | -0.024 |
Token | sc |
Token ID | 629 |
Feature activation | +7.44e-02 |
Pos logit contributions | |
he | +0.028 |
leaf | +0.027 |
olded | +0.027 |
hes | +0.027 |
obb | +0.027 |
Neg logit contributions | |
moment | -0.023 |
tweet | -0.021 |
est | -0.021 |
Nice | -0.019 |
greeting | -0.019 |
Token | ooter |
Token ID | 25141 |
Feature activation | +2.37e-02 |
Pos logit contributions | |
pastor | +0.123 |
wasted | +0.122 |
he | +0.122 |
ateful | +0.120 |
obb | +0.120 |
Neg logit contributions | |
tweet | -0.089 |
moment | -0.078 |
Nice | -0.078 |
Want | -0.073 |
est | -0.073 |
Token | . |
Token ID | 13 |
Feature activation | +8.21e-03 |
Pos logit contributions | |
leaf | +0.042 |
iza | +0.040 |
outed | +0.038 |
pastor | +0.038 |
ateful | +0.037 |
Neg logit contributions | |
moment | -0.032 |
est | -0.029 |
tweet | -0.027 |
way | -0.027 |
Nice | -0.025 |
Token | He |
Token ID | 679 |
Feature activation | +5.75e-03 |
Pos logit contributions | |
he | +0.015 |
iza | +0.015 |
leaf | +0.015 |
pastor | +0.014 |
ateful | +0.014 |
Neg logit contributions | |
moment | -0.010 |
Nice | -0.010 |
Want | -0.009 |
tweet | -0.009 |
est | -0.008 |
Token | remembered |
Token ID | 12086 |
Feature activation | +1.78e-02 |
Pos logit contributions | |
he | +0.009 |
leaf | +0.008 |
obb | +0.008 |
pastor | +0.008 |
ateful | +0.008 |
Neg logit contributions | |
moment | -0.009 |
tweet | -0.008 |
dy | -0.008 |
fly | -0.008 |
Want | -0.007 |
Token | his |
Token ID | 465 |
Feature activation | +2.02e-02 |
Pos logit contributions | |
he | +0.031 |
obb | +0.029 |
gr | +0.029 |
leaf | +0.028 |
outs | +0.028 |
Neg logit contributions | |
moment | -0.023 |
tweet | -0.022 |
Nice | -0.020 |
feast | -0.020 |
greeting | -0.020 |
Token | her |
Token ID | 607 |
Feature activation | -5.84e-03 |
Pos logit contributions | |
moment | +0.030 |
tweet | +0.029 |
Nice | +0.028 |
Want | +0.026 |
est | +0.026 |
Neg logit contributions | |
he | -0.018 |
punished | -0.018 |
olded | -0.017 |
sick | -0.017 |
obb | -0.017 |
Token | mom |
Token ID | 1995 |
Feature activation | -1.65e-02 |
Pos logit contributions | |
moment | +0.010 |
tweet | +0.010 |
Nice | +0.009 |
feast | +0.008 |
fly | +0.008 |
Neg logit contributions | |
he | -0.008 |
olded | -0.008 |
obb | -0.007 |
punished | -0.007 |
gr | -0.007 |
Token | if |
Token ID | 611 |
Feature activation | +1.61e-03 |
Pos logit contributions | |
moment | +0.022 |
tweet | +0.022 |
Nice | +0.020 |
Want | +0.019 |
fly | +0.018 |
Neg logit contributions | |
he | -0.026 |
olded | -0.026 |
sick | -0.026 |
punished | -0.025 |
elderly | -0.025 |
Token | she |
Token ID | 673 |
Feature activation | +1.67e-02 |
Pos logit contributions | |
he | +0.002 |
obb | +0.002 |
leaf | +0.002 |
gr | +0.002 |
punished | +0.002 |
Neg logit contributions | |
moment | -0.003 |
tweet | -0.003 |
Nice | -0.002 |
Want | -0.002 |
fly | -0.002 |
Token | could |
Token ID | 714 |
Feature activation | +1.38e-02 |
Pos logit contributions | |
he | +0.026 |
sick | +0.025 |
obb | +0.025 |
elderly | +0.025 |
punished | +0.024 |
Neg logit contributions | |
moment | -0.024 |
tweet | -0.023 |
Nice | -0.021 |
Want | -0.020 |
fly | -0.020 |
Token | ' |
Token ID | 705 |
Feature activation | +6.07e-02 |
Pos logit contributions | |
olded | +0.023 |
outs | +0.023 |
obb | +0.023 |
gr | +0.023 |
elderly | +0.022 |
Neg logit contributions | |
tweet | -0.018 |
moment | -0.018 |
Nice | -0.015 |
fly | -0.015 |
greeting | -0.015 |
Token | take |
Token ID | 20657 |
Feature activation | +1.11e-03 |
Pos logit contributions | |
obb | +0.083 |
olded | +0.081 |
ashamed | +0.079 |
pastor | +0.079 |
sick | +0.079 |
Neg logit contributions | |
moment | -0.102 |
tweet | -0.094 |
Want | -0.088 |
greeting | -0.084 |
Nice | -0.082 |
Token | ' |
Token ID | 6 |
Feature activation | +1.62e-02 |
Pos logit contributions | |
olded | +0.002 |
he | +0.001 |
punished | +0.001 |
outs | +0.001 |
sick | +0.001 |
Neg logit contributions | |
moment | -0.002 |
tweet | -0.002 |
Nice | -0.002 |
Want | -0.002 |
greeting | -0.002 |
Token | some |
Token ID | 617 |
Feature activation | -2.14e-02 |
Pos logit contributions | |
olded | +0.027 |
obb | +0.026 |
iza | +0.025 |
pastor | +0.025 |
blamed | +0.025 |
Neg logit contributions | |
moment | -0.024 |
tweet | -0.021 |
est | -0.020 |
feast | -0.018 |
Want | -0.018 |
Token | old |
Token ID | 1468 |
Feature activation | -2.23e-02 |
Pos logit contributions | |
moment | +0.031 |
tweet | +0.030 |
Nice | +0.027 |
feast | +0.026 |
Want | +0.026 |
Neg logit contributions | |
olded | -0.032 |
he | -0.032 |
punished | -0.031 |
gr | -0.031 |
sick | -0.031 |
Token | blankets |
Token ID | 34794 |
Feature activation | -7.09e-04 |
Pos logit contributions | |
moment | +0.033 |
tweet | +0.032 |
Nice | +0.028 |
est | +0.028 |
feast | +0.027 |
Neg logit contributions | |
he | -0.035 |
olded | -0.034 |
punished | -0.033 |
obb | -0.033 |
sick | -0.032 |
Token | wait |
Token ID | 4043 |
Feature activation | -2.17e-02 |
Pos logit contributions | |
outs | +0.011 |
olded | +0.011 |
outed | +0.011 |
elderly | +0.010 |
ateful | +0.010 |
Neg logit contributions | |
moment | -0.009 |
tweet | -0.009 |
feast | -0.008 |
fly | -0.008 |
greeting | -0.008 |
Token | to |
Token ID | 284 |
Feature activation | -1.59e-03 |
Pos logit contributions | |
moment | +0.034 |
tweet | +0.031 |
Nice | +0.029 |
Want | +0.028 |
greeting | +0.028 |
Neg logit contributions | |
he | -0.032 |
olded | -0.032 |
gr | -0.031 |
punished | -0.031 |
elderly | -0.030 |
Token | show |
Token ID | 905 |
Feature activation | +2.29e-02 |
Pos logit contributions | |
moment | +0.002 |
tweet | +0.002 |
feast | +0.002 |
greeting | +0.002 |
Want | +0.002 |
Neg logit contributions | |
olded | -0.003 |
outs | -0.003 |
outed | -0.002 |
he | -0.002 |
gr | -0.002 |
Token | him |
Token ID | 683 |
Feature activation | -2.32e-02 |
Pos logit contributions | |
he | +0.033 |
olded | +0.032 |
obb | +0.032 |
gr | +0.031 |
elderly | +0.031 |
Neg logit contributions | |
moment | -0.034 |
tweet | -0.033 |
Nice | -0.030 |
est | -0.029 |
greeting | -0.029 |
Token | what |
Token ID | 644 |
Feature activation | +1.76e-02 |
Pos logit contributions | |
moment | +0.034 |
tweet | +0.033 |
Nice | +0.030 |
feast | +0.029 |
est | +0.029 |
Neg logit contributions | |
he | -0.035 |
olded | -0.034 |
obb | -0.033 |
leaf | -0.033 |
elderly | -0.033 |
Token | she |
Token ID | 673 |
Feature activation | +5.22e-02 |
Pos logit contributions | |
he | +0.021 |
olded | +0.020 |
punished | +0.019 |
elderly | +0.019 |
sick | +0.019 |
Neg logit contributions | |
moment | -0.032 |
tweet | -0.031 |
Nice | -0.029 |
Want | -0.028 |
feast | -0.028 |
Token | had |
Token ID | 550 |
Feature activation | +1.67e-02 |
Pos logit contributions | |
salon | +0.081 |
pastor | +0.081 |
he | +0.080 |
olded | +0.080 |
outs | +0.080 |
Neg logit contributions | |
moment | -0.071 |
tweet | -0.065 |
feast | -0.061 |
Want | -0.060 |
fly | -0.059 |
Token | seen |
Token ID | 1775 |
Feature activation | +1.88e-02 |
Pos logit contributions | |
pastor | +0.024 |
obb | +0.023 |
olded | +0.023 |
leaf | +0.023 |
elderly | +0.023 |
Neg logit contributions | |
moment | -0.025 |
feast | -0.021 |
tweet | -0.021 |
Want | -0.020 |
est | -0.020 |
Token | . |
Token ID | 13 |
Feature activation | +5.06e-03 |
Pos logit contributions | |
olded | +0.026 |
obb | +0.025 |
leaf | +0.025 |
punished | +0.025 |
gr | +0.025 |
Neg logit contributions | |
moment | -0.030 |
tweet | -0.026 |
est | -0.026 |
Nice | -0.026 |
feast | -0.025 |
Token |
|
Token ID | 198 |
Feature activation | -6.47e-03 |
Pos logit contributions | |
punished | +0.008 |
obb | +0.008 |
sick | +0.007 |
leaf | +0.007 |
he | +0.007 |
Neg logit contributions | |
moment | -0.008 |
Nice | -0.007 |
Want | -0.007 |
tweet | -0.007 |
greeting | -0.006 |
Token |
|
Token ID | 198 |
Feature activation | -1.11e-02 |
Pos logit contributions | |
moment | +0.010 |
tweet | +0.009 |
Nice | +0.009 |
Want | +0.008 |
greeting | +0.008 |
Neg logit contributions | |
olded | -0.009 |
punished | -0.009 |
sick | -0.009 |
he | -0.009 |
elderly | -0.009 |
Token | was |
Token ID | 373 |
Feature activation | -1.83e-03 |
Pos logit contributions | |
obb | +0.017 |
he | +0.017 |
olded | +0.017 |
elderly | +0.017 |
leaf | +0.016 |
Neg logit contributions | |
moment | -0.011 |
tweet | -0.010 |
Nice | -0.009 |
Want | -0.009 |
est | -0.008 |
Token | lucky |
Token ID | 9670 |
Feature activation | +6.40e-03 |
Pos logit contributions | |
moment | +0.003 |
greeting | +0.002 |
tweet | +0.002 |
hopping | +0.002 |
Want | +0.002 |
Neg logit contributions | |
olded | -0.003 |
leaf | -0.003 |
he | -0.003 |
obb | -0.003 |
der | -0.003 |
Token | enough |
Token ID | 1576 |
Feature activation | +1.61e-02 |
Pos logit contributions | |
he | +0.009 |
obb | +0.009 |
olded | +0.009 |
gr | +0.009 |
punished | +0.009 |
Neg logit contributions | |
moment | -0.010 |
Nice | -0.009 |
tweet | -0.009 |
Want | -0.009 |
est | -0.009 |
Token | to |
Token ID | 284 |
Feature activation | -1.87e-03 |
Pos logit contributions | |
obb | +0.024 |
he | +0.024 |
olded | +0.024 |
gr | +0.023 |
punished | +0.022 |
Neg logit contributions | |
moment | -0.026 |
Nice | -0.022 |
greeting | -0.022 |
feast | -0.021 |
est | -0.020 |
Token | come |
Token ID | 1282 |
Feature activation | -1.89e-02 |
Pos logit contributions | |
moment | +0.002 |
tweet | +0.002 |
feast | +0.002 |
greeting | +0.002 |
hop | +0.002 |
Neg logit contributions | |
olded | -0.003 |
outs | -0.003 |
outed | -0.003 |
obb | -0.003 |
gr | -0.003 |
Token | with |
Token ID | 351 |
Feature activation | +2.12e-02 |
Pos logit contributions | |
moment | +0.026 |
tweet | +0.025 |
Nice | +0.023 |
greeting | +0.022 |
Want | +0.022 |
Neg logit contributions | |
obb | -0.031 |
olded | -0.031 |
leaf | -0.029 |
he | -0.029 |
punished | -0.029 |
Token | her |
Token ID | 607 |
Feature activation | -8.42e-03 |
Pos logit contributions | |
olded | +0.030 |
obb | +0.029 |
leaf | +0.029 |
pastor | +0.028 |
gr | +0.028 |
Neg logit contributions | |
moment | -0.035 |
feast | -0.029 |
Nice | -0.029 |
tweet | -0.029 |
greeting | -0.029 |
Token | . |
Token ID | 13 |
Feature activation | +5.05e-03 |
Pos logit contributions | |
moment | +0.013 |
feast | +0.011 |
est | +0.011 |
greeting | +0.011 |
way | +0.011 |
Neg logit contributions | |
obb | -0.013 |
leaf | -0.013 |
he | -0.012 |
olded | -0.012 |
hes | -0.012 |
Token | She |
Token ID | 1375 |
Feature activation | -1.16e-03 |
Pos logit contributions | |
punished | +0.009 |
obb | +0.009 |
leaf | +0.009 |
sick | +0.009 |
pastor | +0.009 |
Neg logit contributions | |
moment | -0.006 |
Nice | -0.006 |
Want | -0.005 |
tweet | -0.005 |
greeting | -0.005 |
Token | put |
Token ID | 1234 |
Feature activation | -1.11e-02 |
Pos logit contributions | |
moment | +0.002 |
tweet | +0.001 |
Want | +0.001 |
feast | +0.001 |
dy | +0.001 |
Neg logit contributions | |
obb | -0.002 |
he | -0.002 |
ateful | -0.002 |
sick | -0.002 |
leaf | -0.002 |
Token | a |
Token ID | 257 |
Feature activation | -1.18e-02 |
Pos logit contributions | |
moment | +0.020 |
tweet | +0.018 |
est | +0.017 |
Nice | +0.017 |
greeting | +0.016 |
Neg logit contributions | |
olded | -0.014 |
pastor | -0.013 |
punished | -0.013 |
sick | -0.013 |
leaf | -0.013 |
Token | best |
Token ID | 1266 |
Feature activation | -9.70e-03 |
Pos logit contributions | |
olded | +0.000 |
he | +0.000 |
outs | +0.000 |
she | +0.000 |
outed | +0.000 |
Neg logit contributions | |
moment | -0.000 |
tweet | -0.000 |
feast | -0.000 |
est | -0.000 |
greeting | -0.000 |
Token | day |
Token ID | 1110 |
Feature activation | -2.19e-02 |
Pos logit contributions | |
tweet | +0.011 |
moment | +0.011 |
Nice | +0.010 |
Want | +0.009 |
fly | +0.009 |
Neg logit contributions | |
he | -0.017 |
sick | -0.017 |
elderly | -0.017 |
olded | -0.017 |
punished | -0.017 |
Token | ever |
Token ID | 1683 |
Feature activation | -3.31e-02 |
Pos logit contributions | |
moment | +0.032 |
tweet | +0.031 |
Nice | +0.029 |
Want | +0.027 |
feast | +0.027 |
Neg logit contributions | |
olded | -0.032 |
punished | -0.032 |
sick | -0.032 |
he | -0.032 |
obb | -0.032 |
Token | ! |
Token ID | 0 |
Feature activation | +1.81e-03 |
Pos logit contributions | |
moment | +0.050 |
tweet | +0.050 |
Nice | +0.046 |
Want | +0.045 |
fly | +0.043 |
Neg logit contributions | |
he | -0.047 |
olded | -0.046 |
sick | -0.045 |
punished | -0.045 |
elderly | -0.045 |
Token | <|endoftext|> |
Token ID | 50256 |
Feature activation | -1.25e-02 |
Pos logit contributions | |
obb | +0.003 |
punished | +0.003 |
sick | +0.003 |
pastor | +0.003 |
outs | +0.003 |
Neg logit contributions | |
moment | -0.002 |
Nice | -0.002 |
tweet | -0.002 |
Want | -0.002 |
greeting | -0.002 |
Token | M |
Token ID | 44 |
Feature activation | +2.33e-02 |
Pos logit contributions | |
tweet | +0.020 |
moment | +0.020 |
Nice | +0.018 |
Want | +0.018 |
fly | +0.017 |
Neg logit contributions | |
he | -0.017 |
olded | -0.016 |
sick | -0.016 |
elderly | -0.016 |
punished | -0.016 |
Token | ama |
Token ID | 1689 |
Feature activation | -2.41e-02 |
Pos logit contributions | |
gr | +0.037 |
punished | +0.037 |
olded | +0.037 |
leaf | +0.036 |
obb | +0.035 |
Neg logit contributions | |
moment | -0.032 |
tweet | -0.030 |
feast | -0.029 |
Nice | -0.027 |
story | -0.027 |
Token | and |
Token ID | 290 |
Feature activation | +1.14e-02 |
Pos logit contributions | |
moment | +0.030 |
tweet | +0.029 |
Nice | +0.025 |
Want | +0.025 |
fly | +0.025 |
Neg logit contributions | |
obb | -0.041 |
olded | -0.040 |
sick | -0.040 |
gr | -0.040 |
he | -0.040 |
Token | her |
Token ID | 607 |
Feature activation | +7.36e-03 |
Pos logit contributions | |
olded | +0.014 |
he | +0.014 |
obb | +0.014 |
outs | +0.014 |
punished | +0.014 |
Neg logit contributions | |
tweet | -0.019 |
moment | -0.019 |
Nice | -0.018 |
Want | -0.017 |
fly | -0.017 |
Token | little |
Token ID | 1310 |
Feature activation | +1.85e-02 |
Pos logit contributions | |
obb | +0.011 |
punished | +0.011 |
he | +0.011 |
olded | +0.011 |
outs | +0.011 |
Neg logit contributions | |
moment | -0.011 |
tweet | -0.011 |
Nice | -0.009 |
elcome | -0.009 |
feast | -0.009 |
Token | boy |
Token ID | 2933 |
Feature activation | -1.48e-03 |
Pos logit contributions | |
he | +0.030 |
olded | +0.029 |
obb | +0.028 |
punished | +0.028 |
leaf | +0.028 |
Neg logit contributions | |
moment | -0.027 |
tweet | -0.027 |
Nice | -0.024 |
est | -0.023 |
fly | -0.022 |
Token | called |
Token ID | 1444 |
Feature activation | -1.65e-02 |
Pos logit contributions | |
obb | +0.013 |
olded | +0.013 |
he | +0.013 |
punished | +0.013 |
pastor | +0.012 |
Neg logit contributions | |
moment | -0.013 |
tweet | -0.013 |
fly | -0.012 |
Nice | -0.012 |
greeting | -0.011 |
Token | Peter |
Token ID | 5613 |
Feature activation | -5.31e-02 |
Pos logit contributions | |
moment | +0.027 |
tweet | +0.026 |
Nice | +0.025 |
Want | +0.024 |
fly | +0.023 |
Neg logit contributions | |
he | -0.022 |
punished | -0.022 |
olded | -0.021 |
sick | -0.021 |
outs | -0.021 |
Token | . |
Token ID | 13 |
Feature activation | -7.82e-03 |
Pos logit contributions | |
moment | +0.102 |
tweet | +0.098 |
fly | +0.094 |
Nice | +0.094 |
est | +0.089 |
Neg logit contributions | |
he | -0.055 |
obb | -0.055 |
olded | -0.055 |
punished | -0.054 |
leaf | -0.052 |
Token | Peter |
Token ID | 5613 |
Feature activation | -2.70e-02 |
Pos logit contributions | |
tweet | +0.011 |
moment | +0.010 |
fly | +0.009 |
feast | +0.009 |
Nice | +0.009 |
Neg logit contributions | |
he | -0.012 |
elderly | -0.012 |
bored | -0.011 |
sick | -0.011 |
late | -0.011 |
Token | was |
Token ID | 373 |
Feature activation | +1.11e-03 |
Pos logit contributions | |
moment | +0.032 |
tweet | +0.029 |
Nice | +0.027 |
fly | +0.027 |
dy | +0.026 |
Neg logit contributions | |
olded | -0.047 |
he | -0.047 |
leaf | -0.046 |
obb | -0.046 |
elderly | -0.045 |
Token | very |
Token ID | 845 |
Feature activation | +7.05e-03 |
Pos logit contributions | |
he | +0.002 |
olded | +0.002 |
outs | +0.001 |
leaf | +0.001 |
obb | +0.001 |
Neg logit contributions | |
moment | -0.002 |
tweet | -0.002 |
Nice | -0.002 |
Want | -0.001 |
fly | -0.001 |
Token | clever |
Token ID | 14169 |
Feature activation | +2.06e-02 |
Pos logit contributions | |
der | +0.011 |
he | +0.011 |
outs | +0.010 |
olded | +0.010 |
salon | +0.010 |
Neg logit contributions | |
moment | -0.010 |
tweet | -0.009 |
Nice | -0.008 |
dy | -0.008 |
hop | -0.008 |
Token | and |
Token ID | 290 |
Feature activation | +1.62e-02 |
Pos logit contributions | |
he | +0.033 |
olded | +0.033 |
obb | +0.033 |
punished | +0.032 |
gr | +0.032 |
Neg logit contributions | |
moment | -0.028 |
Nice | -0.026 |
tweet | -0.025 |
est | -0.025 |
Want | -0.023 |
Token | liked |
Token ID | 8288 |
Feature activation | +2.02e-02 |
Pos logit contributions | |
he | +0.024 |
outs | +0.023 |
obb | +0.022 |
olded | +0.022 |
salon | +0.022 |
Neg logit contributions | |
moment | -0.024 |
tweet | -0.023 |
Nice | -0.023 |
Want | -0.022 |
est | -0.020 |
Token | to |
Token ID | 284 |
Feature activation | -1.21e-02 |
Pos logit contributions | |
he | +0.029 |
olded | +0.028 |
punished | +0.027 |
obb | +0.027 |
gr | +0.026 |
Neg logit contributions | |
moment | -0.032 |
tweet | -0.031 |
Nice | -0.029 |
Want | -0.029 |
greeting | -0.028 |
Token | create |
Token ID | 2251 |
Feature activation | +1.46e-02 |
Pos logit contributions | |
tweet | +0.014 |
moment | +0.014 |
hop | +0.013 |
feast | +0.013 |
greeting | +0.012 |
Neg logit contributions | |
olded | -0.024 |
outs | -0.023 |
outed | -0.023 |
he | -0.021 |
she | -0.021 |
Token | a |
Token ID | 257 |
Feature activation | -9.22e-03 |
Pos logit contributions | |
moment | +0.012 |
tweet | +0.012 |
est | +0.011 |
feast | +0.010 |
fly | +0.010 |
Neg logit contributions | |
olded | -0.012 |
pastor | -0.012 |
he | -0.012 |
salon | -0.012 |
disob | -0.012 |
Token | large |
Token ID | 1588 |
Feature activation | -1.74e-02 |
Pos logit contributions | |
tweet | +0.014 |
moment | +0.014 |
Nice | +0.014 |
Want | +0.013 |
est | +0.013 |
Neg logit contributions | |
sick | -0.012 |
he | -0.012 |
elderly | -0.012 |
punished | -0.012 |
olded | -0.012 |
Token | open |
Token ID | 1280 |
Feature activation | -2.16e-02 |
Pos logit contributions | |
tweet | +0.024 |
moment | +0.024 |
Nice | +0.024 |
Want | +0.021 |
reetings | +0.021 |
Neg logit contributions | |
sick | -0.026 |
elderly | -0.026 |
he | -0.025 |
punished | -0.024 |
angry | -0.024 |
Token | space |
Token ID | 2272 |
Feature activation | -1.20e-02 |
Pos logit contributions | |
moment | +0.026 |
tweet | +0.025 |
Nice | +0.023 |
Want | +0.022 |
fly | +0.021 |
Neg logit contributions | |
olded | -0.037 |
he | -0.037 |
punished | -0.036 |
sick | -0.036 |
elderly | -0.036 |
Token | and |
Token ID | 290 |
Feature activation | -1.48e-03 |
Pos logit contributions | |
moment | +0.020 |
tweet | +0.018 |
fly | +0.017 |
feast | +0.017 |
Nice | +0.017 |
Neg logit contributions | |
olded | -0.016 |
punished | -0.016 |
obb | -0.016 |
sick | -0.016 |
he | -0.015 |
Token | started |
Token ID | 2067 |
Feature activation | +1.63e-02 |
Pos logit contributions | |
moment | +0.002 |
tweet | +0.002 |
Want | +0.002 |
Nice | +0.002 |
greeting | +0.002 |
Neg logit contributions | |
he | -0.002 |
obb | -0.002 |
pastor | -0.002 |
outs | -0.002 |
gr | -0.002 |
Token | to |
Token ID | 284 |
Feature activation | -4.19e-03 |
Pos logit contributions | |
he | +0.026 |
olded | +0.025 |
obb | +0.025 |
punished | +0.024 |
iza | +0.024 |
Neg logit contributions | |
moment | -0.024 |
tweet | -0.023 |
Nice | -0.020 |
Want | -0.020 |
greeting | -0.020 |
Token | fight |
Token ID | 1907 |
Feature activation | +1.34e-02 |
Pos logit contributions | |
moment | +0.006 |
tweet | +0.005 |
Want | +0.005 |
Nice | +0.005 |
fly | +0.005 |
Neg logit contributions | |
olded | -0.007 |
he | -0.007 |
outs | -0.007 |
obb | -0.007 |
elderly | -0.007 |
Token | . |
Token ID | 13 |
Feature activation | +1.83e-03 |
Pos logit contributions | |
he | +0.016 |
olded | +0.016 |
leaf | +0.015 |
sick | +0.015 |
punished | +0.015 |
Neg logit contributions | |
moment | -0.023 |
tweet | -0.022 |
Nice | -0.021 |
Want | -0.021 |
est | -0.020 |
Token | They |
Token ID | 1119 |
Feature activation | -4.86e-03 |
Pos logit contributions | |
punished | +0.003 |
sick | +0.003 |
he | +0.003 |
pastor | +0.003 |
obb | +0.003 |
Neg logit contributions | |
moment | -0.002 |
Nice | -0.002 |
Want | -0.002 |
tweet | -0.002 |
greeting | -0.002 |
Token | ran |
Token ID | 4966 |
Feature activation | -8.10e-03 |
Pos logit contributions | |
moment | +0.006 |
tweet | +0.006 |
fly | +0.005 |
Want | +0.005 |
greeting | +0.005 |
Neg logit contributions | |
pastor | -0.008 |
obb | -0.008 |
elderly | -0.008 |
iza | -0.008 |
he | -0.007 |
Token | was |
Token ID | 373 |
Feature activation | +6.07e-03 |
Pos logit contributions | |
moment | +0.003 |
tweet | +0.003 |
Nice | +0.003 |
Want | +0.003 |
fly | +0.003 |
Neg logit contributions | |
olded | -0.004 |
he | -0.004 |
punished | -0.004 |
elderly | -0.003 |
obb | -0.003 |
Token | authentic |
Token ID | 16425 |
Feature activation | -2.22e-02 |
Pos logit contributions | |
he | +0.009 |
obb | +0.009 |
leaf | +0.008 |
olded | +0.008 |
pastor | +0.008 |
Neg logit contributions | |
moment | -0.009 |
tweet | -0.008 |
greeting | -0.008 |
fly | -0.008 |
Nice | -0.008 |
Token | . |
Token ID | 13 |
Feature activation | -8.41e-03 |
Pos logit contributions | |
moment | +0.041 |
tweet | +0.038 |
est | +0.036 |
Nice | +0.036 |
fly | +0.035 |
Neg logit contributions | |
leaf | -0.025 |
olded | -0.025 |
he | -0.025 |
obb | -0.025 |
punished | -0.024 |
Token | Lily |
Token ID | 20037 |
Feature activation | -3.95e-03 |
Pos logit contributions | |
moment | +0.011 |
Nice | +0.010 |
Want | +0.010 |
tweet | +0.010 |
greeting | +0.008 |
Neg logit contributions | |
leaf | -0.014 |
punished | -0.014 |
pastor | -0.014 |
obb | -0.014 |
he | -0.014 |
Token | was |
Token ID | 373 |
Feature activation | -1.65e-03 |
Pos logit contributions | |
moment | +0.006 |
tweet | +0.005 |
est | +0.005 |
dy | +0.005 |
fly | +0.005 |
Neg logit contributions | |
he | -0.006 |
leaf | -0.006 |
obb | -0.006 |
pastor | -0.006 |
sick | -0.006 |
Token | happy |
Token ID | 3772 |
Feature activation | +9.94e-03 |
Pos logit contributions | |
moment | +0.003 |
tweet | +0.002 |
greeting | +0.002 |
Nice | +0.002 |
est | +0.002 |
Neg logit contributions | |
he | -0.002 |
leaf | -0.002 |
obb | -0.002 |
olded | -0.002 |
pastor | -0.002 |
Token | again |
Token ID | 757 |
Feature activation | +1.88e-03 |
Pos logit contributions | |
sick | +0.010 |
he | +0.010 |
elderly | +0.010 |
olded | +0.010 |
punished | +0.010 |
Neg logit contributions | |
tweet | -0.019 |
moment | -0.018 |
Nice | -0.017 |
Want | -0.017 |
fly | -0.017 |
Token | and |
Token ID | 290 |
Feature activation | +2.72e-03 |
Pos logit contributions | |
obb | +0.003 |
leaf | +0.003 |
olded | +0.003 |
pastor | +0.003 |
outs | +0.003 |
Neg logit contributions | |
moment | -0.002 |
tweet | -0.002 |
greeting | -0.002 |
feast | -0.002 |
Nice | -0.002 |
Token | thanked |
Token ID | 26280 |
Feature activation | +1.43e-02 |
Pos logit contributions | |
he | +0.004 |
obb | +0.004 |
gr | +0.004 |
leaf | +0.004 |
outs | +0.004 |
Neg logit contributions | |
moment | -0.004 |
tweet | -0.004 |
feast | -0.004 |
Nice | -0.004 |
Want | -0.003 |
Token | her |
Token ID | 607 |
Feature activation | +1.21e-02 |
Pos logit contributions | |
olded | +0.020 |
obb | +0.019 |
he | +0.018 |
punished | +0.018 |
outed | +0.018 |
Neg logit contributions | |
tweet | -0.024 |
moment | -0.023 |
Nice | -0.021 |
greeting | -0.020 |
feast | -0.020 |
Token | mom |
Token ID | 1995 |
Feature activation | +1.09e-03 |
Pos logit contributions | |
he | +0.019 |
leaf | +0.019 |
obb | +0.018 |
olded | +0.018 |
vious | +0.018 |
Neg logit contributions | |
moment | -0.018 |
tweet | -0.017 |
Nice | -0.015 |
feast | -0.015 |
greeting | -0.015 |
Token | , |
Token ID | 11 |
Feature activation | -2.09e-02 |
Pos logit contributions | |
moment | +0.041 |
tweet | +0.039 |
Nice | +0.035 |
Want | +0.033 |
fly | +0.033 |
Neg logit contributions | |
olded | -0.088 |
he | -0.087 |
sick | -0.086 |
punished | -0.085 |
elderly | -0.084 |
Token | " |
Token ID | 366 |
Feature activation | -6.69e-03 |
Pos logit contributions | |
moment | +0.025 |
tweet | +0.024 |
Nice | +0.022 |
fly | +0.021 |
feast | +0.021 |
Neg logit contributions | |
olded | -0.036 |
he | -0.035 |
elderly | -0.035 |
sick | -0.035 |
punished | -0.034 |
Token | I |
Token ID | 40 |
Feature activation | +2.10e-03 |
Pos logit contributions | |
moment | +0.013 |
tweet | +0.012 |
Nice | +0.012 |
Want | +0.011 |
est | +0.011 |
Neg logit contributions | |
olded | -0.007 |
sick | -0.007 |
he | -0.007 |
punished | -0.007 |
elderly | -0.007 |
Token | 'm |
Token ID | 1101 |
Feature activation | -8.40e-03 |
Pos logit contributions | |
he | +0.003 |
olded | +0.003 |
elderly | +0.003 |
outs | +0.003 |
obb | +0.003 |
Neg logit contributions | |
moment | -0.003 |
tweet | -0.003 |
Nice | -0.003 |
Want | -0.003 |
feast | -0.003 |
Token | glad |
Token ID | 9675 |
Feature activation | -4.32e-03 |
Pos logit contributions | |
moment | +0.012 |
tweet | +0.012 |
Nice | +0.011 |
Want | +0.011 |
greeting | +0.011 |
Neg logit contributions | |
outs | -0.012 |
he | -0.011 |
olded | -0.011 |
obb | -0.011 |
outed | -0.011 |
Token | you |
Token ID | 345 |
Feature activation | -3.67e-03 |
Pos logit contributions | |
moment | +0.008 |
tweet | +0.007 |
Nice | +0.007 |
Want | +0.007 |
fly | +0.007 |
Neg logit contributions | |
he | -0.005 |
olded | -0.005 |
sick | -0.005 |
punished | -0.005 |
elderly | -0.005 |
Token | had |
Token ID | 550 |
Feature activation | -1.65e-02 |
Pos logit contributions | |
moment | +0.005 |
tweet | +0.005 |
Nice | +0.005 |
Want | +0.004 |
est | +0.004 |
Neg logit contributions | |
pastor | -0.006 |
outs | -0.006 |
elderly | -0.006 |
olded | -0.005 |
obb | -0.005 |
Token | fun |
Token ID | 1257 |
Feature activation | -2.18e-02 |
Pos logit contributions | |
moment | +0.030 |
tweet | +0.030 |
Nice | +0.029 |
fly | +0.028 |
Want | +0.027 |
Neg logit contributions | |
olded | -0.018 |
punished | -0.018 |
sick | -0.017 |
he | -0.017 |
elderly | -0.017 |
Token | , |
Token ID | 11 |
Feature activation | -1.14e-02 |
Pos logit contributions | |
moment | +0.031 |
tweet | +0.030 |
Nice | +0.028 |
Want | +0.027 |
feast | +0.026 |
Neg logit contributions | |
he | -0.033 |
olded | -0.033 |
sick | -0.033 |
punished | -0.032 |
elderly | -0.032 |
Token | Tim |
Token ID | 5045 |
Feature activation | -2.93e-02 |
Pos logit contributions | |
moment | +0.021 |
tweet | +0.021 |
fly | +0.020 |
Nice | +0.019 |
Want | +0.019 |
Neg logit contributions | |
elderly | -0.012 |
punished | -0.012 |
sick | -0.012 |
olded | -0.012 |
he | -0.012 |
Token | . |
Token ID | 13 |
Feature activation | -2.80e-02 |
Pos logit contributions | |
moment | +0.048 |
tweet | +0.044 |
est | +0.043 |
dy | +0.041 |
Nice | +0.041 |
Neg logit contributions | |
leaf | -0.042 |
olded | -0.040 |
obb | -0.040 |
he | -0.039 |
outs | -0.039 |
Token | was |
Token ID | 373 |
Feature activation | -3.63e-03 |
Pos logit contributions | |
moment | +0.038 |
tweet | +0.035 |
Nice | +0.032 |
est | +0.031 |
Want | +0.029 |
Neg logit contributions | |
olded | -0.072 |
leaf | -0.070 |
punished | -0.070 |
elderly | -0.069 |
he | -0.069 |
Token | Sally |
Token ID | 25737 |
Feature activation | -4.67e-02 |
Pos logit contributions | |
tweet | +0.004 |
Nice | +0.004 |
moment | +0.004 |
Want | +0.004 |
elcome | +0.004 |
Neg logit contributions | |
olded | -0.006 |
punished | -0.006 |
he | -0.006 |
outs | -0.006 |
obb | -0.006 |
Token | . |
Token ID | 13 |
Feature activation | -4.82e-03 |
Pos logit contributions | |
moment | +0.085 |
tweet | +0.079 |
Nice | +0.077 |
dy | +0.074 |
est | +0.074 |
Neg logit contributions | |
obb | -0.056 |
olded | -0.053 |
leaf | -0.052 |
punished | -0.051 |
he | -0.050 |
Token | Sally |
Token ID | 25737 |
Feature activation | -2.83e-02 |
Pos logit contributions | |
tweet | +0.009 |
moment | +0.008 |
fly | +0.008 |
Nice | +0.008 |
feast | +0.007 |
Neg logit contributions | |
he | -0.006 |
elderly | -0.005 |
sick | -0.005 |
olded | -0.005 |
bored | -0.005 |
Token | was |
Token ID | 373 |
Feature activation | +5.89e-04 |
Pos logit contributions | |
moment | +0.037 |
tweet | +0.033 |
dy | +0.032 |
est | +0.031 |
fly | +0.030 |
Neg logit contributions | |
olded | -0.045 |
leaf | -0.045 |
obb | -0.044 |
he | -0.043 |
outs | -0.043 |
Token | growing |
Token ID | 3957 |
Feature activation | -1.64e-02 |
Pos logit contributions | |
he | +0.001 |
olded | +0.001 |
leaf | +0.001 |
outs | +0.001 |
obb | +0.001 |
Neg logit contributions | |
moment | -0.001 |
tweet | -0.001 |
Nice | -0.001 |
Want | -0.001 |
greeting | -0.001 |
Token | a |
Token ID | 257 |
Feature activation | -1.39e-03 |
Pos logit contributions | |
moment | +0.026 |
tweet | +0.025 |
est | +0.023 |
Nice | +0.023 |
fly | +0.022 |
Neg logit contributions | |
olded | -0.023 |
leaf | -0.022 |
he | -0.021 |
punished | -0.021 |
pastor | -0.021 |
Token | lot |
Token ID | 1256 |
Feature activation | -6.85e-03 |
Pos logit contributions | |
moment | +0.002 |
tweet | +0.002 |
Nice | +0.002 |
Want | +0.002 |
fly | +0.002 |
Neg logit contributions | |
he | -0.002 |
olded | -0.002 |
sick | -0.002 |
punished | -0.002 |
elderly | -0.002 |
Token | and |
Token ID | 290 |
Feature activation | -9.42e-04 |
Pos logit contributions | |
moment | +0.009 |
Nice | +0.008 |
est | +0.008 |
tweet | +0.008 |
feast | +0.007 |
Neg logit contributions | |
punished | -0.011 |
olded | -0.010 |
pastor | -0.010 |
gr | -0.010 |
earn | -0.010 |
Token | learning |
Token ID | 4673 |
Feature activation | +2.39e-03 |
Pos logit contributions | |
moment | +0.002 |
tweet | +0.001 |
Nice | +0.001 |
Want | +0.001 |
elcome | +0.001 |
Neg logit contributions | |
olded | -0.001 |
leaf | -0.001 |
he | -0.001 |
punished | -0.001 |
gr | -0.001 |
Token | lots |
Token ID | 6041 |
Feature activation | +5.68e-04 |
Pos logit contributions | |
elderly | +0.003 |
sick | +0.003 |
he | +0.003 |
olded | +0.003 |
outs | +0.003 |
Neg logit contributions | |
moment | -0.004 |
tweet | -0.004 |
Nice | -0.003 |
fly | -0.003 |
Want | -0.003 |
Token | . |
Token ID | 13 |
Feature activation | -4.43e-03 |
Pos logit contributions | |
leaf | +0.005 |
obb | +0.005 |
olded | +0.005 |
punished | +0.005 |
he | +0.005 |
Neg logit contributions | |
moment | -0.006 |
tweet | -0.006 |
fly | -0.005 |
Nice | -0.005 |
est | -0.005 |
Token | He |
Token ID | 679 |
Feature activation | -7.08e-03 |
Pos logit contributions | |
Nice | +0.006 |
moment | +0.006 |
Want | +0.005 |
tweet | +0.005 |
greeting | +0.005 |
Neg logit contributions | |
punished | -0.008 |
olded | -0.007 |
sick | -0.007 |
disob | -0.007 |
leaf | -0.007 |
Token | has |
Token ID | 468 |
Feature activation | -5.39e-03 |
Pos logit contributions | |
tweet | +0.007 |
moment | +0.007 |
hops | +0.005 |
Nice | +0.005 |
Want | +0.005 |
Neg logit contributions | |
olded | -0.014 |
he | -0.014 |
punished | -0.014 |
elderly | -0.014 |
obb | -0.014 |
Token | a |
Token ID | 257 |
Feature activation | -8.01e-03 |
Pos logit contributions | |
moment | +0.007 |
tweet | +0.007 |
greeting | +0.006 |
Nice | +0.006 |
est | +0.006 |
Neg logit contributions | |
he | -0.009 |
olded | -0.008 |
obb | -0.008 |
punished | -0.008 |
earn | -0.008 |
Token | big |
Token ID | 1263 |
Feature activation | -1.74e-02 |
Pos logit contributions | |
moment | +0.012 |
tweet | +0.011 |
Nice | +0.010 |
feast | +0.010 |
greeting | +0.010 |
Neg logit contributions | |
olded | -0.013 |
he | -0.012 |
punished | -0.012 |
outs | -0.012 |
obb | -0.012 |
Token | show |
Token ID | 905 |
Feature activation | -4.55e-03 |
Pos logit contributions | |
moment | +0.025 |
tweet | +0.025 |
Nice | +0.024 |
Want | +0.023 |
fly | +0.022 |
Neg logit contributions | |
sick | -0.025 |
he | -0.025 |
elderly | -0.025 |
punished | -0.024 |
pastor | -0.023 |
Token | today |
Token ID | 1909 |
Feature activation | -2.64e-03 |
Pos logit contributions | |
moment | +0.007 |
tweet | +0.007 |
Nice | +0.006 |
Want | +0.006 |
est | +0.006 |
Neg logit contributions | |
obb | -0.006 |
he | -0.006 |
punished | -0.006 |
sick | -0.006 |
gr | -0.006 |
Token | in |
Token ID | 287 |
Feature activation | -3.29e-03 |
Pos logit contributions | |
moment | +0.004 |
tweet | +0.004 |
Nice | +0.004 |
Want | +0.004 |
est | +0.004 |
Neg logit contributions | |
olded | -0.003 |
punished | -0.003 |
obb | -0.003 |
leaf | -0.003 |
gr | -0.003 |
Token | a |
Token ID | 257 |
Feature activation | +8.93e-03 |
Pos logit contributions | |
moment | +0.005 |
tweet | +0.004 |
greeting | +0.004 |
Nice | +0.004 |
Want | +0.004 |
Neg logit contributions | |
olded | -0.005 |
obb | -0.005 |
punished | -0.005 |
outed | -0.005 |
outs | -0.005 |
Token | far |
Token ID | 1290 |
Feature activation | -7.40e-03 |
Pos logit contributions | |
olded | +0.013 |
he | +0.012 |
leaf | +0.012 |
punished | +0.012 |
outs | +0.012 |
Neg logit contributions | |
moment | -0.014 |
tweet | -0.014 |
Nice | -0.012 |
Want | -0.012 |
feast | -0.012 |
Token | city |
Token ID | 1748 |
Feature activation | +1.12e-02 |
Pos logit contributions | |
moment | +0.012 |
tweet | +0.011 |
Nice | +0.010 |
Want | +0.010 |
fly | +0.010 |
Neg logit contributions | |
olded | -0.011 |
he | -0.010 |
punished | -0.010 |
leaf | -0.010 |
sick | -0.010 |
Token | hide |
Token ID | 7808 |
Feature activation | -1.46e-02 |
Pos logit contributions | |
hop | +0.000 |
tweet | +0.000 |
search | +0.000 |
leap | +0.000 |
hops | +0.000 |
Neg logit contributions | |
ateful | -0.000 |
elderly | -0.000 |
year | -0.000 |
late | -0.000 |
outs | -0.000 |
Token | behind |
Token ID | 2157 |
Feature activation | -1.08e-02 |
Pos logit contributions | |
moment | +0.021 |
tweet | +0.020 |
Nice | +0.018 |
fly | +0.017 |
est | +0.017 |
Neg logit contributions | |
olded | -0.023 |
leaf | -0.022 |
he | -0.022 |
punished | -0.021 |
earn | -0.021 |
Token | a |
Token ID | 257 |
Feature activation | -2.03e-02 |
Pos logit contributions | |
moment | +0.016 |
tweet | +0.015 |
Nice | +0.013 |
greeting | +0.013 |
feast | +0.013 |
Neg logit contributions | |
olded | -0.018 |
punished | -0.017 |
he | -0.016 |
gr | -0.016 |
obb | -0.016 |
Token | tree |
Token ID | 5509 |
Feature activation | -7.29e-03 |
Pos logit contributions | |
moment | +0.025 |
tweet | +0.025 |
Nice | +0.023 |
Want | +0.021 |
fly | +0.021 |
Neg logit contributions | |
he | -0.034 |
sick | -0.034 |
olded | -0.034 |
punished | -0.033 |
elderly | -0.033 |
Token | . |
Token ID | 13 |
Feature activation | -6.27e-03 |
Pos logit contributions | |
moment | +0.011 |
tweet | +0.011 |
Nice | +0.010 |
Want | +0.010 |
fly | +0.010 |
Neg logit contributions | |
olded | -0.010 |
he | -0.010 |
punished | -0.010 |
leaf | -0.010 |
sick | -0.010 |
Token |
|
Token ID | 198 |
Feature activation | -1.10e-02 |
Pos logit contributions | |
moment | +0.010 |
tweet | +0.009 |
Nice | +0.009 |
Want | +0.009 |
greeting | +0.008 |
Neg logit contributions | |
he | -0.009 |
punished | -0.009 |
olded | -0.008 |
sick | -0.008 |
salon | -0.008 |
Token |
|
Token ID | 198 |
Feature activation | -1.68e-02 |
Pos logit contributions | |
moment | +0.016 |
tweet | +0.016 |
Nice | +0.015 |
Want | +0.014 |
fly | +0.014 |
Neg logit contributions | |
he | -0.016 |
olded | -0.016 |
punished | -0.016 |
sick | -0.015 |
elderly | -0.015 |
Token | " |
Token ID | 1 |
Feature activation | +1.34e-02 |
Pos logit contributions | |
moment | +0.025 |
tweet | +0.024 |
Nice | +0.022 |
Want | +0.022 |
fly | +0.021 |
Neg logit contributions | |
he | -0.025 |
olded | -0.024 |
sick | -0.024 |
punished | -0.024 |
elderly | -0.024 |
Token | Sorry |
Token ID | 14385 |
Feature activation | -2.85e-03 |
Pos logit contributions | |
olded | +0.018 |
he | +0.018 |
elderly | +0.017 |
outs | +0.017 |
sick | +0.017 |
Neg logit contributions | |
moment | -0.022 |
tweet | -0.021 |
Nice | -0.019 |
Want | -0.019 |
fly | -0.018 |
Token | , |
Token ID | 11 |
Feature activation | -2.45e-03 |
Pos logit contributions | |
moment | +0.004 |
tweet | +0.004 |
fly | +0.003 |
feast | +0.003 |
greeting | +0.003 |
Neg logit contributions | |
elderly | -0.005 |
sick | -0.004 |
angry | -0.004 |
pastor | -0.004 |
disob | -0.004 |
Token | m |
Token ID | 285 |
Feature activation | +4.00e-02 |
Pos logit contributions | |
tweet | +0.004 |
moment | +0.004 |
Nice | +0.004 |
Want | +0.004 |
feast | +0.003 |
Neg logit contributions | |
he | -0.003 |
olded | -0.003 |
gr | -0.003 |
punished | -0.003 |
outs | -0.003 |
Token | play |
Token ID | 711 |
Feature activation | +3.35e-03 |
Pos logit contributions | |
olded | +0.007 |
outs | +0.007 |
pastor | +0.006 |
he | +0.006 |
punished | +0.006 |
Neg logit contributions | |
moment | -0.006 |
tweet | -0.006 |
Want | -0.005 |
Nice | -0.005 |
greeting | -0.005 |
Token | . |
Token ID | 13 |
Feature activation | +1.13e-02 |
Pos logit contributions | |
he | +0.005 |
sick | +0.005 |
olded | +0.005 |
punished | +0.005 |
elderly | +0.005 |
Neg logit contributions | |
moment | -0.005 |
tweet | -0.005 |
Nice | -0.005 |
Want | -0.004 |
fly | -0.004 |
Token | She |
Token ID | 1375 |
Feature activation | +7.29e-03 |
Pos logit contributions | |
olded | +0.009 |
he | +0.008 |
elderly | +0.008 |
sick | +0.008 |
punished | +0.008 |
Neg logit contributions | |
moment | -0.024 |
tweet | -0.024 |
Nice | -0.022 |
fly | -0.022 |
feast | -0.022 |
Token | tw |
Token ID | 665 |
Feature activation | +1.57e-02 |
Pos logit contributions | |
he | +0.011 |
olded | +0.010 |
sick | +0.010 |
punished | +0.010 |
elderly | +0.010 |
Neg logit contributions | |
moment | -0.011 |
tweet | -0.011 |
Nice | -0.010 |
Want | -0.009 |
fly | -0.009 |
Token | irled |
Token ID | 49376 |
Feature activation | -9.34e-03 |
Pos logit contributions | |
he | +0.015 |
olded | +0.014 |
outed | +0.014 |
outs | +0.014 |
out | +0.013 |
Neg logit contributions | |
moment | -0.034 |
Nice | -0.033 |
tweet | -0.032 |
Want | -0.031 |
feast | -0.030 |
Token | and |
Token ID | 290 |
Feature activation | -3.13e-03 |
Pos logit contributions | |
tweet | +0.016 |
moment | +0.015 |
Nice | +0.015 |
fly | +0.015 |
reetings | +0.015 |
Neg logit contributions | |
he | -0.012 |
sick | -0.011 |
elderly | -0.011 |
bored | -0.010 |
she | -0.010 |
Token | jumped |
Token ID | 11687 |
Feature activation | -1.57e-02 |
Pos logit contributions | |
tweet | +0.005 |
moment | +0.005 |
Nice | +0.005 |
fly | +0.005 |
feast | +0.005 |
Neg logit contributions | |
he | -0.004 |
sick | -0.004 |
olded | -0.004 |
bored | -0.004 |
punished | -0.004 |
Token | around |
Token ID | 1088 |
Feature activation | -1.21e-03 |
Pos logit contributions | |
moment | +0.024 |
tweet | +0.024 |
Nice | +0.023 |
Want | +0.022 |
fly | +0.021 |
Neg logit contributions | |
he | -0.022 |
sick | -0.021 |
olded | -0.021 |
elderly | -0.020 |
punished | -0.020 |
Token | with |
Token ID | 351 |
Feature activation | +2.33e-02 |
Pos logit contributions | |
moment | +0.002 |
tweet | +0.002 |
Nice | +0.002 |
Want | +0.002 |
fly | +0.002 |
Neg logit contributions | |
olded | -0.002 |
he | -0.002 |
punished | -0.002 |
sick | -0.002 |
elderly | -0.002 |
Token | it |
Token ID | 340 |
Feature activation | +1.59e-03 |
Pos logit contributions | |
he | +0.026 |
sick | +0.025 |
elderly | +0.025 |
leaf | +0.024 |
punished | +0.023 |
Neg logit contributions | |
moment | -0.042 |
tweet | -0.042 |
Nice | -0.040 |
fly | -0.040 |
Want | -0.039 |
Token | . |
Token ID | 13 |
Feature activation | +9.79e-03 |
Pos logit contributions | |
he | +0.001 |
sick | +0.001 |
olded | +0.001 |
punished | +0.001 |
elderly | +0.001 |
Neg logit contributions | |
moment | -0.003 |
tweet | -0.003 |
Nice | -0.003 |
Want | -0.003 |
feast | -0.003 |
Token | says |
Token ID | 1139 |
Feature activation | +1.52e-02 |
Pos logit contributions | |
he | +0.026 |
obb | +0.026 |
punished | +0.026 |
sick | +0.025 |
gr | +0.025 |
Neg logit contributions | |
moment | -0.037 |
tweet | -0.037 |
Nice | -0.035 |
Want | -0.033 |
est | -0.033 |
Token | . |
Token ID | 13 |
Feature activation | +1.52e-03 |
Pos logit contributions | |
obb | +0.022 |
punished | +0.022 |
he | +0.022 |
leaf | +0.021 |
salon | +0.021 |
Neg logit contributions | |
moment | -0.024 |
tweet | -0.022 |
Nice | -0.021 |
fly | -0.020 |
greeting | -0.020 |
Token |
|
Token ID | 198 |
Feature activation | -7.20e-03 |
Pos logit contributions | |
punished | +0.002 |
he | +0.002 |
olded | +0.002 |
sick | +0.002 |
obb | +0.002 |
Neg logit contributions | |
moment | -0.002 |
tweet | -0.002 |
Nice | -0.002 |
Want | -0.002 |
fly | -0.002 |
Token |
|
Token ID | 198 |
Feature activation | -1.26e-02 |
Pos logit contributions | |
moment | +0.010 |
tweet | +0.009 |
Nice | +0.009 |
Want | +0.008 |
fly | +0.008 |
Neg logit contributions | |
punished | -0.012 |
olded | -0.011 |
sick | -0.011 |
obb | -0.011 |
he | -0.011 |
Token | " |
Token ID | 1 |
Feature activation | +1.35e-02 |
Pos logit contributions | |
moment | +0.016 |
tweet | +0.015 |
Nice | +0.014 |
Want | +0.013 |
fly | +0.013 |
Neg logit contributions | |
olded | -0.021 |
sick | -0.021 |
punished | -0.021 |
he | -0.021 |
obb | -0.021 |
Token | Wow |
Token ID | 22017 |
Feature activation | -4.62e-02 |
Pos logit contributions | |
punished | +0.017 |
olded | +0.017 |
he | +0.017 |
sick | +0.017 |
obb | +0.016 |
Neg logit contributions | |
moment | -0.024 |
tweet | -0.023 |
Nice | -0.022 |
Want | -0.021 |
greeting | -0.020 |
Token | , |
Token ID | 11 |
Feature activation | -9.27e-03 |
Pos logit contributions | |
moment | +0.053 |
tweet | +0.051 |
Nice | +0.050 |
Want | +0.048 |
greeting | +0.048 |
Neg logit contributions | |
he | -0.080 |
sick | -0.077 |
bored | -0.075 |
elderly | -0.075 |
olded | -0.074 |
Token | that |
Token ID | 326 |
Feature activation | -2.46e-02 |
Pos logit contributions | |
moment | +0.016 |
tweet | +0.015 |
Nice | +0.014 |
Want | +0.014 |
fly | +0.014 |
Neg logit contributions | |
he | -0.011 |
olded | -0.011 |
sick | -0.011 |
elderly | -0.011 |
punished | -0.011 |
Token | is |
Token ID | 318 |
Feature activation | -1.41e-02 |
Pos logit contributions | |
tweet | +0.031 |
Nice | +0.031 |
moment | +0.031 |
Want | +0.030 |
greeting | +0.028 |
Neg logit contributions | |
sick | -0.038 |
he | -0.037 |
elderly | -0.037 |
bored | -0.035 |
punished | -0.035 |
Token | fun |
Token ID | 1257 |
Feature activation | -9.00e-03 |
Pos logit contributions | |
moment | +0.029 |
tweet | +0.029 |
Nice | +0.027 |
Want | +0.027 |
est | +0.026 |
Neg logit contributions | |
elderly | -0.013 |
sick | -0.013 |
bored | -0.012 |
punished | -0.011 |
angry | -0.011 |
Token | . |
Token ID | 13 |
Feature activation | -1.33e-02 |
Pos logit contributions | |
moment | +0.018 |
tweet | +0.018 |
Nice | +0.017 |
Want | +0.017 |
fly | +0.016 |
Neg logit contributions | |
he | -0.008 |
sick | -0.008 |
elderly | -0.008 |
olded | -0.008 |
punished | -0.007 |
Token | had |
Token ID | 550 |
Feature activation | -1.03e-02 |
Pos logit contributions | |
moment | +0.010 |
tweet | +0.009 |
Want | +0.009 |
Nice | +0.009 |
dy | +0.008 |
Neg logit contributions | |
he | -0.007 |
obb | -0.007 |
gr | -0.006 |
outs | -0.006 |
leaf | -0.006 |
Token | an |
Token ID | 281 |
Feature activation | +1.50e-02 |
Pos logit contributions | |
moment | +0.015 |
tweet | +0.014 |
Want | +0.013 |
greeting | +0.013 |
feast | +0.013 |
Neg logit contributions | |
he | -0.017 |
iza | -0.015 |
olded | -0.015 |
pastor | -0.014 |
obb | -0.014 |
Token | opinion |
Token ID | 4459 |
Feature activation | -1.23e-02 |
Pos logit contributions | |
he | +0.025 |
olded | +0.025 |
punished | +0.024 |
sick | +0.024 |
leaf | +0.024 |
Neg logit contributions | |
moment | -0.019 |
tweet | -0.018 |
Nice | -0.017 |
Want | -0.016 |
est | -0.016 |
Token | . |
Token ID | 13 |
Feature activation | -4.00e-03 |
Pos logit contributions | |
moment | +0.020 |
Nice | +0.019 |
tweet | +0.019 |
Want | +0.018 |
est | +0.017 |
Neg logit contributions | |
iza | -0.016 |
he | -0.016 |
punished | -0.016 |
olded | -0.016 |
obb | -0.015 |
Token | He |
Token ID | 679 |
Feature activation | -2.83e-04 |
Pos logit contributions | |
moment | +0.006 |
tweet | +0.006 |
Nice | +0.006 |
Want | +0.006 |
est | +0.005 |
Neg logit contributions | |
he | -0.006 |
punished | -0.006 |
sick | -0.006 |
olded | -0.006 |
iza | -0.006 |
Token | said |
Token ID | 531 |
Feature activation | -3.75e-02 |
Pos logit contributions | |
moment | +0.000 |
tweet | +0.000 |
dy | +0.000 |
fly | +0.000 |
Want | +0.000 |
Neg logit contributions | |
he | -0.000 |
obb | -0.000 |
handed | -0.000 |
pastor | -0.000 |
leaf | -0.000 |
Token | , |
Token ID | 11 |
Feature activation | -1.07e-02 |
Pos logit contributions | |
moment | +0.037 |
tweet | +0.034 |
Nice | +0.030 |
fly | +0.028 |
Want | +0.028 |
Neg logit contributions | |
olded | -0.074 |
sick | -0.073 |
punished | -0.072 |
he | -0.072 |
elderly | -0.071 |
Token | " |
Token ID | 366 |
Feature activation | +1.38e-02 |
Pos logit contributions | |
moment | +0.012 |
tweet | +0.012 |
Nice | +0.010 |
Want | +0.010 |
feast | +0.010 |
Neg logit contributions | |
olded | -0.019 |
he | -0.019 |
sick | -0.019 |
elderly | -0.019 |
punished | -0.019 |
Token | I |
Token ID | 40 |
Feature activation | +1.11e-02 |
Pos logit contributions | |
olded | +0.019 |
sick | +0.018 |
obb | +0.018 |
punished | +0.018 |
he | +0.018 |
Neg logit contributions | |
moment | -0.023 |
tweet | -0.021 |
Nice | -0.020 |
Want | -0.020 |
est | -0.019 |
Token | don |
Token ID | 836 |
Feature activation | +1.18e-02 |
Pos logit contributions | |
elderly | +0.017 |
pastor | +0.017 |
outs | +0.016 |
olded | +0.016 |
he | +0.016 |
Neg logit contributions | |
moment | -0.016 |
tweet | -0.015 |
fly | -0.013 |
Want | -0.013 |
Nice | -0.013 |
Token | 't |
Token ID | 470 |
Feature activation | -2.71e-03 |
Pos logit contributions | |
olded | +0.016 |
punished | +0.016 |
sick | +0.015 |
outs | +0.015 |
elderly | +0.015 |
Neg logit contributions | |
tweet | -0.018 |
moment | -0.018 |
Nice | -0.017 |
Want | -0.017 |
elcome | -0.016 |
Token | mom |
Token ID | 1995 |
Feature activation | +1.27e-04 |
Pos logit contributions | |
moment | +0.011 |
Nice | +0.010 |
tweet | +0.010 |
Want | +0.010 |
greeting | +0.009 |
Neg logit contributions | |
leaf | -0.008 |
olded | -0.008 |
punished | -0.008 |
gr | -0.008 |
he | -0.008 |
Token | is |
Token ID | 318 |
Feature activation | -5.58e-03 |
Pos logit contributions | |
he | +0.000 |
olded | +0.000 |
obb | +0.000 |
leaf | +0.000 |
gr | +0.000 |
Neg logit contributions | |
moment | -0.000 |
tweet | -0.000 |
Nice | -0.000 |
Want | -0.000 |
feast | -0.000 |
Token | busy |
Token ID | 8179 |
Feature activation | +1.00e-02 |
Pos logit contributions | |
moment | +0.006 |
tweet | +0.005 |
greeting | +0.005 |
feast | +0.005 |
Want | +0.005 |
Neg logit contributions | |
he | -0.010 |
leaf | -0.010 |
olded | -0.010 |
obb | -0.009 |
der | -0.009 |
Token | and |
Token ID | 290 |
Feature activation | -4.03e-03 |
Pos logit contributions | |
olded | +0.018 |
he | +0.018 |
leaf | +0.017 |
obb | +0.017 |
gr | +0.017 |
Neg logit contributions | |
moment | -0.012 |
tweet | -0.011 |
feast | -0.010 |
greeting | -0.010 |
Nice | -0.009 |
Token | tells |
Token ID | 4952 |
Feature activation | -8.64e-03 |
Pos logit contributions | |
moment | +0.005 |
Want | +0.005 |
tweet | +0.005 |
hopping | +0.005 |
Nice | +0.005 |
Neg logit contributions | |
he | -0.007 |
obb | -0.006 |
gr | -0.006 |
outs | -0.006 |
leaf | -0.006 |
Token | Ben |
Token ID | 3932 |
Feature activation | -5.01e-02 |
Pos logit contributions | |
moment | +0.013 |
tweet | +0.013 |
Nice | +0.012 |
Want | +0.012 |
feast | +0.011 |
Neg logit contributions | |
gr | -0.012 |
leaf | -0.012 |
olded | -0.012 |
he | -0.012 |
punished | -0.011 |
Token | to |
Token ID | 284 |
Feature activation | -1.27e-02 |
Pos logit contributions | |
moment | +0.084 |
tweet | +0.081 |
Nice | +0.077 |
Want | +0.074 |
feast | +0.072 |
Neg logit contributions | |
he | -0.065 |
olded | -0.062 |
punished | -0.061 |
elderly | -0.061 |
obb | -0.061 |
Token | play |
Token ID | 711 |
Feature activation | +3.40e-03 |
Pos logit contributions | |
moment | +0.016 |
tweet | +0.015 |
feast | +0.015 |
hop | +0.014 |
greeting | +0.013 |
Neg logit contributions | |
outs | -0.022 |
olded | -0.021 |
leaf | -0.021 |
outed | -0.020 |
bored | -0.020 |
Token | with |
Token ID | 351 |
Feature activation | +2.00e-02 |
Pos logit contributions | |
olded | +0.006 |
leaf | +0.005 |
punished | +0.005 |
he | +0.005 |
pastor | +0.005 |
Neg logit contributions | |
moment | -0.005 |
tweet | -0.004 |
feast | -0.004 |
Nice | -0.004 |
est | -0.004 |
Token | his |
Token ID | 465 |
Feature activation | -1.89e-03 |
Pos logit contributions | |
olded | +0.035 |
leaf | +0.034 |
he | +0.033 |
gr | +0.031 |
vious | +0.031 |
Neg logit contributions | |
moment | -0.025 |
tweet | -0.022 |
feast | -0.022 |
mushroom | -0.021 |
Want | -0.020 |
Token | toys |
Token ID | 14958 |
Feature activation | +9.12e-03 |
Pos logit contributions | |
moment | +0.003 |
tweet | +0.002 |
feast | +0.002 |
est | +0.002 |
dy | +0.002 |
Neg logit contributions | |
olded | -0.003 |
he | -0.003 |
leaf | -0.003 |
gr | -0.003 |
she | -0.003 |
Token | " |
Token ID | 366 |
Feature activation | +5.26e-03 |
Pos logit contributions | |
moment | +0.010 |
tweet | +0.009 |
Nice | +0.009 |
Want | +0.008 |
fly | +0.008 |
Neg logit contributions | |
he | -0.016 |
olded | -0.016 |
sick | -0.015 |
punished | -0.015 |
elderly | -0.015 |
Token | My |
Token ID | 3666 |
Feature activation | -2.76e-03 |
Pos logit contributions | |
punished | +0.007 |
obb | +0.007 |
olded | +0.006 |
outs | +0.006 |
sick | +0.006 |
Neg logit contributions | |
moment | -0.009 |
Nice | -0.009 |
tweet | -0.009 |
Want | -0.008 |
greeting | -0.008 |
Token | dog |
Token ID | 3290 |
Feature activation | -7.55e-03 |
Pos logit contributions | |
moment | +0.003 |
tweet | +0.003 |
Nice | +0.003 |
greeting | +0.003 |
Want | +0.003 |
Neg logit contributions | |
he | -0.005 |
olded | -0.005 |
punished | -0.005 |
obb | -0.005 |
iza | -0.005 |
Token | is |
Token ID | 318 |
Feature activation | -2.08e-02 |
Pos logit contributions | |
moment | +0.009 |
tweet | +0.009 |
Nice | +0.008 |
Want | +0.008 |
feast | +0.007 |
Neg logit contributions | |
he | -0.013 |
olded | -0.013 |
sick | -0.013 |
punished | -0.013 |
elderly | -0.013 |
Token | called |
Token ID | 1444 |
Feature activation | -2.91e-02 |
Pos logit contributions | |
moment | +0.035 |
tweet | +0.035 |
Nice | +0.033 |
Want | +0.032 |
fly | +0.031 |
Neg logit contributions | |
he | -0.025 |
olded | -0.025 |
punished | -0.024 |
sick | -0.024 |
elderly | -0.024 |
Token | Tom |
Token ID | 4186 |
Feature activation | -5.14e-02 |
Pos logit contributions | |
tweet | +0.041 |
moment | +0.040 |
Nice | +0.039 |
Want | +0.036 |
greeting | +0.035 |
Neg logit contributions | |
he | -0.045 |
punished | -0.045 |
olded | -0.044 |
obb | -0.043 |
sick | -0.042 |
Token | . |
Token ID | 13 |
Feature activation | -1.80e-02 |
Pos logit contributions | |
moment | +0.095 |
tweet | +0.094 |
Nice | +0.089 |
Want | +0.085 |
fly | +0.085 |
Neg logit contributions | |
he | -0.056 |
olded | -0.055 |
leaf | -0.054 |
punished | -0.053 |
obb | -0.053 |
Token | He |
Token ID | 679 |
Feature activation | -1.70e-02 |
Pos logit contributions | |
moment | +0.028 |
tweet | +0.027 |
Nice | +0.026 |
Want | +0.025 |
fly | +0.024 |
Neg logit contributions | |
he | -0.025 |
olded | -0.024 |
punished | -0.024 |
sick | -0.024 |
elderly | -0.024 |
Token | is |
Token ID | 318 |
Feature activation | -1.93e-02 |
Pos logit contributions | |
tweet | +0.017 |
moment | +0.017 |
Nice | +0.015 |
fly | +0.014 |
Want | +0.014 |
Neg logit contributions | |
he | -0.033 |
olded | -0.032 |
elderly | -0.031 |
obb | -0.031 |
punished | -0.031 |
Token | very |
Token ID | 845 |
Feature activation | -2.83e-03 |
Pos logit contributions | |
moment | +0.030 |
tweet | +0.030 |
Nice | +0.028 |
Want | +0.027 |
greeting | +0.027 |
Neg logit contributions | |
he | -0.027 |
olded | -0.026 |
obb | -0.025 |
punished | -0.024 |
outs | -0.024 |
Token | brave |
Token ID | 14802 |
Feature activation | +3.19e-03 |
Pos logit contributions | |
moment | +0.003 |
tweet | +0.003 |
Nice | +0.003 |
greeting | +0.003 |
dy | +0.003 |
Neg logit contributions | |
he | -0.005 |
salon | -0.005 |
olded | -0.005 |
obb | -0.005 |
outs | -0.005 |
Token | She |
Token ID | 1375 |
Feature activation | -1.47e-04 |
Pos logit contributions | |
moment | +0.008 |
tweet | +0.008 |
Nice | +0.007 |
Want | +0.007 |
fly | +0.007 |
Neg logit contributions | |
he | -0.005 |
olded | -0.005 |
punished | -0.005 |
sick | -0.005 |
elderly | -0.005 |
Token | saw |
Token ID | 2497 |
Feature activation | +2.17e-04 |
Pos logit contributions | |
moment | +0.000 |
tweet | +0.000 |
fly | +0.000 |
Nice | +0.000 |
dy | +0.000 |
Neg logit contributions | |
he | -0.000 |
leaf | -0.000 |
obb | -0.000 |
elderly | -0.000 |
olded | -0.000 |
Token | the |
Token ID | 262 |
Feature activation | -3.75e-03 |
Pos logit contributions | |
he | +0.000 |
olded | +0.000 |
gr | +0.000 |
leaf | +0.000 |
punished | +0.000 |
Neg logit contributions | |
tweet | -0.000 |
moment | -0.000 |
Nice | -0.000 |
fly | -0.000 |
Want | -0.000 |
Token | door |
Token ID | 3420 |
Feature activation | +8.54e-03 |
Pos logit contributions | |
moment | +0.005 |
tweet | +0.005 |
feast | +0.004 |
story | +0.004 |
greeting | +0.004 |
Neg logit contributions | |
olded | -0.007 |
ateful | -0.006 |
punished | -0.006 |
udging | -0.006 |
outed | -0.006 |
Token | that |
Token ID | 326 |
Feature activation | -3.10e-03 |
Pos logit contributions | |
olded | +0.012 |
he | +0.012 |
punished | +0.012 |
sick | +0.012 |
elderly | +0.012 |
Neg logit contributions | |
moment | -0.013 |
tweet | -0.012 |
Nice | -0.011 |
Want | -0.011 |
fly | -0.011 |
Token | said |
Token ID | 531 |
Feature activation | -3.97e-02 |
Pos logit contributions | |
moment | +0.005 |
tweet | +0.005 |
Nice | +0.005 |
Want | +0.004 |
fly | +0.004 |
Neg logit contributions | |
olded | -0.004 |
he | -0.004 |
leaf | -0.004 |
punished | -0.004 |
elderly | -0.004 |
Token | " |
Token ID | 366 |
Feature activation | +2.06e-02 |
Pos logit contributions | |
moment | +0.056 |
tweet | +0.053 |
Nice | +0.050 |
Want | +0.049 |
feast | +0.049 |
Neg logit contributions | |
olded | -0.059 |
sick | -0.059 |
he | -0.057 |
pastor | -0.056 |
elderly | -0.056 |
Token | Do |
Token ID | 5211 |
Feature activation | -3.10e-02 |
Pos logit contributions | |
olded | +0.027 |
obb | +0.027 |
leaf | +0.026 |
punished | +0.026 |
ashamed | +0.026 |
Neg logit contributions | |
moment | -0.036 |
tweet | -0.033 |
Nice | -0.030 |
Want | -0.030 |
fly | -0.029 |
Token | not |
Token ID | 407 |
Feature activation | -1.71e-02 |
Pos logit contributions | |
moment | +0.057 |
tweet | +0.056 |
Nice | +0.053 |
Want | +0.051 |
fly | +0.050 |
Neg logit contributions | |
he | -0.034 |
olded | -0.034 |
sick | -0.033 |
punished | -0.033 |
elderly | -0.032 |
Token | enter |
Token ID | 3802 |
Feature activation | -7.30e-03 |
Pos logit contributions | |
moment | +0.026 |
tweet | +0.025 |
Nice | +0.024 |
Want | +0.023 |
fly | +0.022 |
Neg logit contributions | |
he | -0.024 |
sick | -0.024 |
olded | -0.024 |
punished | -0.024 |
elderly | -0.024 |
Token | ". |
Token ID | 1911 |
Feature activation | +1.94e-02 |
Pos logit contributions | |
moment | +0.010 |
tweet | +0.009 |
Want | +0.009 |
Nice | +0.009 |
est | +0.008 |
Neg logit contributions | |
olded | -0.012 |
leaf | -0.012 |
punished | -0.011 |
obb | -0.011 |
sick | -0.011 |