INDEX
Negative Logits
oreferrer
-0.08
(Channel
-0.08
duż
-0.07
BBC
-0.07
hael
-0.07
oplayer
-0.07
ойчив
-0.07
секун
-0.07
기다
-0.07
daughter
-0.07
POSITIVE LOGITS
scraps
0.08
ाले
0.08
.Document
0.08
closets
0.08
Escolar
0.07
cupboard
0.07
cupboards
0.07
掌
0.07
involucr
0.07
sco
0.07
Activations Density 0.003%