INDEX
Explanations
expressions of gaze or attention between characters
New Auto-Interp
Negative Logits
aida
-0.17
Laud
-0.15
edin
-0.15
icio
-0.15
ium
-0.14
ajas
-0.14
rollo
-0.14
.bias
-0.14
iente
-0.14
ursal
-0.14
POSITIVE LOGITS
Drake
0.14
yling
0.14
com
0.14
CDATA
0.13
/*č↵
0.13
ģµ
0.13
ätz
0.13
Ellen
0.13
lashes
0.13
towers
0.13
Activations Density 0.177%