INDEX
Negative Logits
contaminants
-0.09
caras
-0.08
conting
-0.08
geen
-0.08
filhos
-0.08
secular
-0.08
contamin
-0.07
Debian
-0.07
שעל
-0.07
Anlage
-0.07
POSITIVE LOGITS
footage
0.09
अक
0.08
tep
0.08
slowed
0.08
(relative
0.08
Ak
0.08
warped
0.08
spray
0.07
video
0.07
ralent
0.07
Activations Density 0.004%