INDEX
Explanations
emotional engagement and character connection in narratives
New Auto-Interp
Negative Logits
aten
-0.16
Femme
-0.15
Fulton
-0.15
å¾Ĵ
-0.14
istem
-0.14
çīĩ
-0.14
ưa
-0.14
iggins
-0.14
Installer
-0.14
Halk
-0.14
POSITIVE LOGITS
vie
0.16
apos
0.15
Overlap
0.15
ì§Ģê°Ģ
0.15
ξι
0.15
rng
0.14
079
0.14
edium
0.14
ucch
0.14
783
0.14
Activations Density 0.150%