INDEX
Explanations
references to children
references to "kid" or childhood experiences
New Auto-Interp
Negative Logits
veyard
-0.89
anwhile
-0.75
aeda
-0.74
partisan
-0.68
undai
-0.67
ĺħ
-0.67
asca
-0.67
aukee
-0.67
freight
-0.66
election
-0.66
POSITIVE LOGITS
ishly
0.95
Doodle
0.93
kid
0.92
kids
0.87
Icar
0.86
Gohan
0.84
girl
0.80
boy
0.79
boys
0.79
ults
0.77
Activations Density 0.008%