INDEX
Explanations
car forum posts
The neuron activates on numeric tokens or alphanumeric number references—e.g. years, measurements, model designations, and other quantities.
New Auto-Interp
Negative Logits
NOP
-0.08
.IContainer
-0.07
hướng
-0.07
<size
-0.07
rast
-0.06
_ED
-0.06
alimentos
-0.06
ゥ
-0.06
cbo
-0.06
이유
-0.06
POSITIVE LOGITS
krev
0.07
assuming
0.07
respir
0.06
shipped
0.06
insist
0.06
قدر
0.06
inode
0.06
้าส
0.06
мерик
0.06
---------
0.06
Activations Density 0.029%