INDEX
Explanations
occurrences of placeholder pages for individuals
New Auto-Interp
Negative Logits
rex
-0.18
žit
-0.15
inja
-0.15
ément
-0.14
sole
-0.14
veau
-0.14
eck
-0.14
nda
-0.14
esh
-0.14
iece
-0.14
POSITIVE LOGITS
bsd
0.15
ppelin
0.15
san
0.14
IDb
0.13
237
0.13
pone
0.13
/Dk
0.13
fal
0.13
ID
0.13
Rank
0.13
Activations Density 0.002%