INDEX
Explanations
familial relationships and heritage
New Auto-Interp
Negative Logits
nephew
-0.18
Cous
-0.17
Friend
-0.17
Girlfriend
-0.16
oun
-0.16
friendship
-0.16
friend
-0.16
Friend
-0.15
ROME
-0.15
grandson
-0.15
POSITIVE LOGITS
remar
0.22
parents
0.21
immigrants
0.19
abusive
0.18
parents
0.17
mother
0.17
maternal
0.15
absentee
0.15
supportive
0.15
ÑĢиз
0.15
Activations Density 0.126%