INDEX
Negative Logits
Response
-0.08
urable
-0.08
prestigious
-0.08
Bakery
-0.08
Response
-0.08
stata
-0.08
iyadda
-0.07
Prest
-0.07
kam
-0.07
ixa
-0.07
POSITIVE LOGITS
accelerating
0.10
secular
0.09
releasing
0.09
causing
0.09
warmer
0.09
acceler
0.09
快
0.08
accelerated
0.08
faster
0.08
暖
0.08
Activations Density 0.006%