INDEX
Explanations
themes related to complex character dynamics and social interactions
New Auto-Interp
Negative Logits
izu
-0.17
hea
-0.16
edback
-0.15
è¿ij
-0.15
amework
-0.15
è¿ij
-0.15
quate
-0.15
-tm
-0.15
igroup
-0.15
avana
-0.14
POSITIVE LOGITS
character
0.19
initially
0.19
throughout
0.17
isode
0.17
shown
0.17
initial
0.16
eps
0.16
Introduced
0.16
oen
0.15
scenes
0.15
Activations Density 0.602%