INDEX
Explanations
references to deceased individuals and their familial connections
New Auto-Interp
Negative Logits
rost
-0.17
mlin
-0.17
dden
-0.16
strup
-0.16
rone
-0.16
anggan
-0.15
ixin
-0.15
foon
-0.15
umpy
-0.15
tro
-0.14
POSITIVE LOGITS
kt
0.15
Kash
0.15
åĸ
0.14
äº
0.14
611
0.13
subdiv
0.13
{name0.13
ç©´
0.13
_wr
0.13
razier
0.13
Activations Density 0.038%