INDEX
Explanations
variations of the word "ang."
New Auto-Interp
Negative Logits
anova
-0.18
laps
-0.18
zd
-0.16
imid
-0.16
leck
-0.16
edx
-0.15
oze
-0.15
Nová
-0.15
edl
-0.15
καν
-0.14
POSITIVE LOGITS
aroo
0.26
ladesh
0.20
ements
0.20
ertz
0.19
ulate
0.19
eline
0.18
rove
0.18
lish
0.17
redi
0.17
ará
0.17
Activations Density 0.033%