INDEX
Explanations
emotional interactions and expressions of concern between characters
New Auto-Interp
Negative Logits
éo
-0.15
ogan
-0.14
SES
-0.14
iky
-0.14
reff
-0.14
acter
-0.14
nda
-0.14
ucene
-0.13
stanbul
-0.13
forum
-0.13
POSITIVE LOGITS
SPATH
0.14
ç°
0.14
ubi
0.14
nghiêm
0.13
filesize
0.13
olu
0.13
kaz
0.13
æ³ī
0.13
ç±
0.13
661
0.13
Activations Density 0.526%