INDEX
Explanations
references to bodies and physicality
New Auto-Interp
Negative Logits
Bloody
-0.16
Lob
-0.15
ufen
-0.14
onden
-0.14
inta
-0.14
Club
-0.14
yah
-0.14
gran
-0.14
Stub
-0.14
Sel
-0.13
POSITIVE LOGITS
anzi
0.17
caval
0.16
aland
0.15
odzi
0.15
ifton
0.15
Leban
0.14
cover
0.14
mdb
0.14
talk
0.14
TableCell
0.14
Activations Density 0.025%