INDEX
Explanations
occurrences of numbers, especially percentages and years.
New Auto-Interp
Negative Logits
TagMode
-1.04
ViewFeatures
-0.93
كومونز
-0.90
DeleteBehavior
-0.90
تضيفلها
-0.89
üyada
-0.88
оригіналу
-0.82
Personendaten
-0.81
Himo
-0.81
المعيارى
-0.80
POSITIVE LOGITS
liten
0.52
and
0.52
ire
0.51
ir
0.48
time
0.46
domov
0.46
all
0.44
small
0.43
réal
0.43
(
0.43
Activations Density 0.023%