INDEX
Explanations
references to traditional cultural practices and items
New Auto-Interp
Negative Logits
nett
-0.17
ouv
-0.15
xAA
-0.15
oola
-0.15
dad
-0.15
issen
-0.15
chain
-0.15
Smooth
-0.14
oyo
-0.14
Stream
-0.14
POSITIVE LOGITS
Arn
0.17
Piper
0.15
diversity
0.15
Schul
0.15
741
0.14
ç¥
0.14
ÏĦολ
0.14
dow
0.14
Maur
0.14
atural
0.14
Activations Density 0.163%