INDEX
Explanations
emotional interactions and expressions of affection between characters
New Auto-Interp
Negative Logits
oblin
-0.15
interviewer
-0.15
eways
-0.14
hugs
-0.14
intros
-0.14
etti
-0.14
handshake
-0.14
intro
-0.13
.realm
-0.13
Pitch
-0.13
POSITIVE LOGITS
ImageContext
0.16
ecut
0.16
rough
0.15
McCabe
0.15
gent
0.15
رÛĮاÙĨ
0.14
against
0.14
delicious
0.14
RedirectTo
0.14
heat
0.14
Activations Density 0.057%