INDEX
Explanations
Discussions of people
This neuron responds to generalizing statements about groups or behaviors (e.g. “some people…,” “there will always be…”)—that is, broad claims or observations about what people do.
New Auto-Interp
Negative Logits
TestCategory
-0.07
upgrade
-0.07
工
-0.07
(크기
-0.06
time
-0.06
عالم
-0.06
達
-0.06
ुख
-0.06
-navbar
-0.06
cult
-0.06
POSITIVE LOGITS
EXISTS
0.06
ISR
0.06
creeping
0.06
RFC
0.06
neck
0.06
_render
0.06
tonight
0.06
((_
0.06
ября
0.06
Refs
0.06
Activations Density 0.048%