INDEX
Explanations
names of pets and companions
New Auto-Interp
Negative Logits
человек
0.49
людей
0.46
людьми
0.46
mennesker
0.46
чисел
0.45
তৈরির
0.44
शहरों
0.43
Americans
0.43
городов
0.42
человека
0.42
POSITIVE LOGITS
mascot
0.93
一只
0.78
companion
0.77
named
0.75
mascota
0.74
faithful
0.71
Mascot
0.70
furry
0.67
masc
0.66
plush
0.65
Activations Density 0.034%