INDEX
Explanations
not particularly or exactly
New Auto-Interp
Negative Logits
実際
-1.06
satisfactory
-1.02
funcionar
-0.99
خوبی
-0.98
preferências
-0.97
actually
-0.95
ℊ
-0.94
差不
-0.93
good
-0.92
skues
-0.91
POSITIVE LOGITS
as
1.06
exactly
1.05
particularly
1.01
пью
0.98
特别
0.97
сказать
0.94
groundbreaking
0.92
Exactly
0.90
blockbuster
0.90
《
0.88
Activations Density 0.035%