INDEX
Explanations
"to" followed by action verbs
New Auto-Interp
Negative Logits
effectively
1.08
need
0.98
efficiently
0.97
excellently
0.95
easily
0.95
dễ
0.95
confidently
0.95
fácilmente
0.94
properly
0.93
behö
0.92
POSITIVE LOGITS
выяс
1.04
Prove
0.80
dissu
0.79
persuade
0.75
Integr
0.73
salvage
0.71
Brazil
0.69
رجع
0.69
convain
0.68
Desk
0.67
Activations Density 0.117%