INDEX
Explanations
instances of the word "so" and variations of Scandinavian or Nordic names
New Auto-Interp
Negative Logits
prostituerade
-0.19
ewise
-0.16
geschichten
-0.16
inqu
-0.15
ardware
-0.15
lesbisk
-0.15
weiber
-0.15
Ðļоли
-0.15
ozy
-0.15
umlu
-0.15
POSITIVE LOGITS
kin
0.23
vast
0.23
ja
0.22
tied
0.20
pal
0.20
engl
0.20
sell
0.20
pain
0.19
nä
0.19
pit
0.19
Activations Density 0.002%