INDEX
Explanations
references to the word "Pel" and its variations
New Auto-Interp
Negative Logits
yz
-0.17
fat
-0.17
eel
-0.16
ÙĩÙĩ
-0.15
heel
-0.15
ysis
-0.15
emia
-0.15
pill
-0.14
seau
-0.14
etti
-0.14
POSITIVE LOGITS
ton
0.19
ryn
0.18
lico
0.17
orus
0.16
ersen
0.16
ty
0.16
isser
0.16
iminary
0.16
urma
0.16
leted
0.15
Activations Density 0.010%